摘要:

Russian Chemical Reviews 68 (11) 889 ± 907 (1999) Fullerenes: functionalisation and prospects for the use of derivatives E N Karaulova, E I Bagrii Contents I. Introduction II. General aspects, synthesis and separation III. Complexes and inclusion compounds IV. Addition reactions V. Fullerenes in polymerisation reactions VI. Fullerenes in catalysis VII. Prospects for the practical use of fullerenes VIII. Conclusion Abstract. Published data on the functionalisation of fullerenes (mainly, C60) including those using addition reactions (cyclo addition, hydrogenation, oxidation, etc.) are surveyed. The participation of fullerenes in catalysis and fullerene-containing polymers are considered and the prospects for the practical use of fullerene derivatives are discussed.The bibliography includes 191 references. I. Introduction Synthesis of polyhedral carbon clusters, fullerenes C60, C70 and so on,1 from hot carbon plasma has initiated and triggered vigorous development of a new line of research in the chemistry of cage compounds, namely, study of the reactivity of fullerenes. In 1996, Kroto, Curl and Smalley were awarded the Nobel Prize in chemistry for the discovery of fullerenes, a new allotropic modification of carbon. In this respect, it is pertinent to consider the essence of this discovery, the history of which is presented in detail in the Nobel lectures of the authors.2±4 A necessary condition of the success was the original appara- tus designed by Smalley for the preparation and investigation of clusters formed on heating of high-melting compounds.This was a supersonic source of cluster beams with laser vaporisation. In this setup, a slowly rotating graphite disc was exposed to the radiation of a pulsed laser. The carbon vapour was treated with a helium flow in the supersonic flash regime and cooled to a temperature of several Kelvins by adiabatic expansion. The cluster beams were directed into a time-of-flight mass spectrometer with photoionisa- tion of the molecular beam. By varying the conditions of clustering and the time of cooling, the researchers attained the predominance of the stable C60 cluster in the mixture. The researchers 2±4 proposed and substantiated the structure of C60 fullerene as a spherical hollow molecule shaped like a truncated icosahedron (Fig.1). To verify the presence of a cavity inside theC60 spheroid, the graphite disc of E N Karaulova, E I Bagrii A V Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Leninsky prosp. 29, 117912 Moscow, Russian Federation. Fax (7-095) 230 22 24. Tel. (7-095) 955 41 25. E-mail: tips@tips.aha.ru Received 24 February 1999 Uspekhi Khimii 68 (11) 979 ± 998 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 0.83.72 : 541.42 : 541.64 : 541.427 : 541.572.53 : 546.26 : 547.7 : 547.591 889 890 890 891 902 904 904 904 the setup was impregnated with an aqueous solution of lanthanum chloride; in this case, laser vaporisation gave a stable complex with the metal located inside the fullerene cage.Figure 1. C60 fullerene (see Ref. 3). Calculations for the C60 molecule, which gave rise to the suggestion that this molecule should be stable, had been per- formed 5, 6 fifteen years ahead of the experiments of Smalley et al.1 (according to Kroto's assertion,3 they were not acquainted with these data). Vigorous development of the physics and chemistry of full- erene became possible only when KraÈ tschmer, Lamb, Fostiro- poulos and Huffman 7 had developed a method for the preparation of macroscopic quantities of fullerenes. The fact that fullerenes are formed during vaporisation of graphite electro- des in an electric arc in a helium atmosphere, discovered by these investigators, served as the basis for this method.The researchers 7 prepared and studied crystals of the C60 cluster. The development of the preparative-scale method for the synthesis of fullerenes brought about an avalanche-like increase in the number of publications devoted to this topic. Osawa, who had predicted the stability of icosahedral fullerenes in 1970,5 composed a computer database (fullerene literature data base, FLDB), based on reprinting from Current Contents.8 The journal Fullerene Science and Technology was founded in 1993, the to-be Nobel Prize winners being among its editors. Reviews on fullerenes are published routinely (see, for example Refs 9 ± 17). A widely cited monograph by Hirsch 18 on fullerene chemistry appeared in Germany in 1994 (a year later it was re-edited in the USA).In the present review, we cover the most important recent achievements in the functionalisation of fullerenes, mainly C60890 fullerene, and new prospects for the use of fullerenes including participation in polymerisation and in catalysis. II. General aspects, synthesis and separation Fullerenes constitute a subclass of polyhedranes, closed cage structures consisting of tricoordinated carbon atoms and having 12 pentagonal and n/2710 hexagonal faces (n520), each pentagon adjoining only hexagons. The nomenclature and atom numbering in fullerene systems were proposed in 1993 by Taylor.19 In 1995 Diederich et al.20 introduced a system of numbering for chiral fullerene derivatives.The chemical and electronic structures of fullerene have been considered comprehensively in the report by Sokolov and Stanke- vich.9 An algorithm for the bond designation based on the spiral pattern of numbering generally accepted for fullerene derivatives 9 has been proposed;21 it permits unambiguous assignment of the [6-6] and [5-6] bonds in addition products (compatible with the rules proposed in another publication 20). In 1997, the Commis- sion on Nomenclature of the IUPAC, having Taylor as a member, accepted a terminology and a tentative nomenclature of fullerenes (see Refs 22 and 23 and the Atlas of Fullerenes 24). A system of numbering for fullerenes based on the Schlegel diagrams 9 (Fig. 2) is described in detail in a special appendix to a study by Wudl et al.22 In this review, we mainly adhere to the nomenclature used in the publications cited.Carbon clusters, i.e. chains of carbon atoms, were discovered by astrophysicists in the spectra of some extraterrestrial objects (carbon stars, comet tails and interstellar dust),25 but the hypoth- esis that these include C60, C70 or Cá60 clusters has not been confirmed.2 Corannulene C20H10, a possible hydrocarbon pre- cursor of C60, which could be formed upon condensation of corannulene with carbon vapour, has been detected in meteor- ites.26 Natural fullerenes have been found in some minerals. The C70 :C60 ratio in them varies from 0.21 to 0.36. 25 Fullerenes are assumed to favour migration and accumulation of some elements, in particular, they presumably influence the formation of dia- mond deposits.25 Kinetic measurements and mathematical simu- lation of the fullerene synthesis in an arc reactor led researchers to the conclusion 27, 28 that the C60 :C70 ratio remains invariant over a broad range of the arc process parameters and amounts to *85 : 15.One of the preparative methods for the synthesis of higher fullerenes (up to C108) is combustion of benzene in a mixture of oxygen and argon.29 To increase the yield of higher fullerenes in 57 44 43 45 56 58 42 26 25 27 24 41 46 28 23 9 10 8 29 47 22 1 6 11 1213 30 40 23 7 21 20 39 4 5 14 38 31 15 1918 16 48 17 37 32 54 36 49 35 33 34 53 55 59 50 52 51 60 Figure 2.Numbering of carbon atoms in C60 fullerene (Schlegel dia- gram).19 E N Karaulova, E I Bagrii the arc synthesis, carbon electrodes are doped with hafnium, zirconium or titanium.30 The lower limit of the temperature of fullerene formation is 1200 ± 1500 8C. Recently, a pyrogenetic process for the production of fullerene soot has been developed.31 According to this method, a hot mixture of H2 and a carbon- containing compound is fed through a nozzle into a water-cooled spherical reactor; the best results are attained with C2Cl4 as the carbon precursor . At the outlet of the nozzle, plasma is formed. The yield is 0.2 g of fullerenes per kV, which corresponds to the conversion of 2.1% of the carbon from the C2Cl4 introduced.The C60 :C70 ratio depends on the cooling conditions. Several mechanisms have been proposed for the formation of fullerenes from the precursor nuclei. Thus it has been assumed 32 that the first stage of this process is a gas ± liquid transition in the expanding flow of supersaturated carbon vapour and this is followed by the formation of small clusters with a dendrite 32 or corannulene (curled sheet) structure.33 It has been found 34 that on heating, pure C60 crystals are transformed into amorphous carbon. Based on this finding, the researchers conclude that the C60 liquid phase, described previously, does not actually exist. Mixtures of C60 with 5%± 15% of C70 are often used for the functionalisation of C60 because C70 usually does not enter into reactions.However, in some cases, especially in the measurements of physical characteristics of processes, it is required that C60 should have a purity of 599%. Nowadays, C60 and C70 are normally separated by HPLC, which sharply increases the cost of the product; therefore, the interest in the optimisation of methods for fullerene isolation does not abate. The sorbents currently used for this purpose consist of polyaromatic compounds, p-acids, porphyrins and phthalocyanines immobilised on silica gel.35 The commercially available Buckyclutcher, which is 3,3,3- tris(dinitrobenzene)propyl silica gel, is commonly used. The use of tetraphenylporphyrin immobilised on silica gel ensures the preparation ofC60 with 98.5% purity andC70 with a>99%purity for a 5-mg batch.36 Preparative quantities of C60 and C70 have been successfully separated using the chromotropic acid tetramer with formaldehyde bound to g-aminopropyl silica gel as the sorbent. 35 A sorbent based on Al(X) tetraphenylporphyrin fixed on hydroxydimethylsilyl silica gel has been proposed for the separation of large amounts of fullerenes.37 Recently it has been found 38 that the solubility of C60 exhibits an abnormal temperature dependence: the maximum of the fullerene solubility in toluene and carbon disulfide falls down to 0 8C, and that in o-xylene is observed at 30 8C. In the case of C70, the temperature variation of the solubility follows a standard pattern. This fact might prove useful for the separation of mixtures of C60 and C70 .The main distinctive features of the C60 fullerene molecule permitting its functionalisation are the `pseudo-aromaticity',9 which governs the reactions with nucleophiles, carbenes and free-radical species to give various adducts, and the presence of the spherical cavity within the cage.2 III. Complexes and inclusion compounds A unique chemical feature of fullerenes is the formation of metallofullerenes, endohedral complexes, in which one or several metal atoms reside inside the carbon cage.{ A metal atom occurring as a cation cannot be extracted from the fullerene polyhedron without complete destruction of the cage, which can take place, for example, in the presence of oxygen. Fullerenes are capable of forming charge transfer complexes (CTC).39 Strictly speaking, the formation of CTC and endome- tallofullerenes cannot be regarded as functionalisation of fullerenes; nevertheless, we present here some novel data concern- { Endohedral complexes of fullerenes are designated using the character @. The metal atoms occurring inside the cage are written to the left of this character, e.g., La@C82.Fullerenes: functionalisation and prospects for the use of derivatives ing this type of reaction in order to characterise the properties of fullerenes more fully and because these compounds are highly promising. Even a weak nucleophile such as iodine forms a CTC with fullerene of the composition C60 .I2. This complex is unstable, its stability constant is <0.1 litre mol71 (see Ref.40). In an earlier study,9 it was reported that this complex was not formed. A new method for the synthesis of CTC of fullerenes, in particular the complex with iodine, has been proposed;41 the method involves doping with iodine of other CTC prepared beforehand (for example, complexes with substituted tetrathia- fulvalenes). In the syntheses of potential precursors of new superconduc- tors, special attention is paid to CTC of C60 fullerene having layered structures. For example, a complex of this type with tetramethyltetraselenofulvalene has been synthesised.42 Doping of the octamethylenetetrathiafulvalene .C60 . benzene complex with potassium resulted in the detection of superconductivity with Tc=17 ± 18.8 K, while doping with rubidium produced superconductivity with Tc= 23 ± 26 K.43 A new type of superconductor has been obtained by the inclusion of potassium into non-superconducting Ba3C60.44 The compound KBa3C60 , which possesses block superconductivity at 5.6 K, is synthesised by heating a mixture of K and Ba3C60 powders in a sealed glass tube for three days in a high vacuum at 260 8C. Endohedral fullerene complexes M@Cn are usually prepared in an arc discharge with a metal ± graphite composite as the positive electrode. These complexes were isolated in a pure state only in 1994,45 and since 1996 they have been studied successfully using scanning tunnelling microscopy.45 The C82 fullerene forms complexes with Sc, La, Y, Ce, Pr and Gd; they are most stable in air.The data obtained by electron paramagnetic resonance showed the occurrence of `intrafullerene' transfer of electrons from the endohedral atom to the carbon cage. According to X-ray diffraction data,45 the metal atom does not reside in the centre of the cage but, perhaps, wanders inside the spheroid; strong metal ± cage interaction takes place in this system. Endohedral complexes are mixtures of isomers with different arrangements of five- and six-membered rings in the cage with respect to the metal.45 The cages of some higher fullerenes can accommodate two or three metal atoms.46 For instance, the complexes Sc2@C84, Sc3@C82 and La2@C80 have been isolated. It was established for the latter complex that the La atoms execute a circular motion inside the spheroid.45 The data from high-energy spectroscopy provided the basis for the conclusion that the lanthanide ion in Tm@C82 is in the divalent state.47 The synthesis, isolation, and, what is the most important, purification of the compound Dy@C82 have been described in detail.48 Endometallofullerenes with calcium were synthesised for the first time by evaporation of a graphite anode filled with calcium carbide particles in an arc discharge.49 Extraction with carbon disulfide gave a reaction product in which the majority of C58±C100 fullerenes contained one calcium atom but some held two calcium atoms inside the cage. Endohedral complexes of fullerenes find use in studies of cells in molecular biology.45 The variation of the EPR spectra of solutions of the endohedral metallofullerenes La@C82 and Y@C82 (their preparative synthesis has been reported 50) as functions of the oxygen concentration can be employed for the monitoring of oxygen in biological, physical and chemical processes.51, 52 IV.Addition reactions The most important line in functionalisation of C60 is exohedral addition to the fullerene cage. The exohedral covalent attachment 891 to the fullerene cage is possible for many classes of compounds. From 1990 to 1995, cycloaddition, radical and nucleophilic addition, hydrogenation, oxidation and halogenation as well as complex formation with transition metals had been studied. The features of the fullerene reactivity have been considered in detail in reviews by Diederich and Thilgen 12, 17 and Hirsch,13 which cover publications up to the end of 1995.The C60 molecule is constructed of 12 five-membered rings, each being surrounded by six-membered rings (20 altogether). It has thirty bonds between six-membered rings ([6-6] bonds) and sixty bonds between six- and five-membered rings ([6-5] bonds). The fact that [6-6] bonds are shorter (1.38A) than the [6-5] bonds (1.45A) influences the chemical behaviour of C60. The conjugated p-system in C60 is not completely delocalised, the reactivity of this cluster being similar to that of an electron- deficient polyolefin. The electrophilic double [6-6] bonds readily add nucleophiles and radicals.The addition of 34 methyl radicals to C60 has been detected by mass spectrometry.10 Therefore, C60 fullerene is referred to as a `radical sponge'. In the adducts, some of the carbon atoms in the functionalised fullerene cage have changed the hybridisation from sp2 to sp3. Nucleophilic and carbene attacks, cycloaddition and hydrogena- tion usually occur as 1,2-addition, while radical reactions, for example, halogenation, follow the 1,4-addition pattern. However, the latter type of addition is also possible for nucleophilic reactions if they involve bulky fragments. Thus adducts A and B have been described 12 (Fig. 3). Ph H But B A Figure 3. Addition to the C60 molecule. A is 1,2-adduct, B is 1,4-adduct. When porphyrins or phthalocyanines contain fullerene frag- ments, the photoinitiated electrical conductivity of their solutions increases.In this case, photoinitiated electron transfer with participation of C60 and C70 takes place to give C60 ¡./C¡70. (see Ref. 53). The unpaired electron in the RC60 . radicals is mainly delocalised over the atoms of the two six-membered rings adjacent to the C7CR bond. 10 The molecule of C70, unlike C60, contains five types of nonequivalent atoms; therefore, free-radical addition to C70 can yield five isomeric RC70 . radicals.10 1. Cycloaddition reactions Cycloaddition to fullerenes gives mainly products of monoaddi- tion, commonly, 1,2-addition to a [6-6] bond, which is energeti- cally more favourable than addition to a [6-5] bond.13 The reaction products are used as the initial compounds for further functionalisation.The [m+2] cycloaddition reactions with C60 (m=1 ± 4) have found wide use, because they are facilitated by the low LUMO energy; [4+2] cycloaddition and, in particular, Diels ± Alder reactions, proceed especially easily; [6-6] bond is always dienophilic. The addition of various alkenes has been considered in a review.13 In recent years, quinodimethanes are successfully used as dienes in reactions with fullerenes. These compounds are either synthesised beforehand or formed in situ upon thermally induced elimination of SO2 from heterocyclic sulfones or sultines892 C60, D C60 SO2 n n51. This reaction underlies the so-called `reactive extraction' method, which permits recovery of higher fullerenes from fullerene soot.54 The soot is first extracted with toluene and 1,2,4-trichlorobenzene and the residue is heated with quinodimethane precursor 1. Me(CH2)14CONH SO2 1 This gives rise to a product soluble in conventional solvents and containing C60±C400 clusters including multisubstituted ones. The ordinary extraction from commercial soot recovers 8.4% of fullerenes, whereas reactive extraction gives additionally 11.8%.Derivatives 2a,b and 3a,b result from reactions of C60 with sulfones 4, 5, which act as precursors of pyrimidine- and pyrimidonequinodimethanes.55 X X C60 , 1,2,4-Cl3C6H3 , N2 , 214 8C N N C60 SO2 7SO2 Ph N Ph N 4 2a,b X=OMe (a, 77%), N NH (b, 95%).O O NR RN C60, D C60 SO2 N N 5 3a,b R = H (a, 82%), Me (b, 52%). The biological activities of these derivatives can be assayed, if they have been transformed into water-soluble products, for example, into acetyl derivatives 6.55 Ac N Ac2O, Py 2b N N C60 Ph N 6 Fullerene derivatives 7 ± 9 have been prepared by [4+2] cycloaddition to C60 of six-membered heterocyclic analogues of o-quinodimethane generated in situ from the corresponding thiophenes, furans and thiazoles.56 R N N N O C60 Ni N N N N N R 12 R=CH2OC5H11 . E N Karaulova, E I Bagrii R1 N C60 C60 C60 R2 O S S 9 8 7 R1=R2=H; R1=H, R2=CO2Me; R17R2=(CH=CH)2 . When the reactants are exposed to microwave radiation, the reaction time becomes shorter and the product yields increase.Thus microwave irradiation was used to synthesise compound 10 by the reaction of C60 with quinodimethanes prepared in situ from the corresponding o-bis(bromomethylated) quinoxaline and pyr- idine derivatives on treatment with NaI in the presence of a crown ether.56 The use of microwave radiation for other Diels ± Alder reactions and for 1,3-dipolar cycloaddition to C60 has also been proposed .57 The reaction time decreases to several minutes, which makes it possible to avoid decomposition to the initial compo- nents. HN R1 C60 R2 N 10 CH ; R1=R2=CN. R17R2=(CH=CH)2, CH The redox potentials for all newly synthesised organofuller- enes were determined by voltammetry in solution.56 The first reduction potentials for these compounds were found to undergo slight cathodic shifts with respect to that for C60 .In another study, 58 sultines (prepared from commercially available rongalite) were used as the precursors of quinodime- thanes. On heating in toluene in the presence of C60, sultines eliminate SO2; the resulting adducts are converted into quinones 11 on further modification. OR2 OR2 R1 R1 O PBr3 C60, PhMe, D C60 S 7SO2 R1 R1 O OR2 OR2 O OH R1 R1 DDQ C60 C60 R1 R1 11 OH O R1=H, R2=C6H13 , Me; R1=Br, R2=Me; R17R1=(CH=CH)2 , R2=Me; DDQ is 2,3-dichloro-5,6-dicyano-1,4-benzoquinone. Quinodimethanes have also been used to prepare adducts of C60 with phthalocyanines and hemiporphyrazine; electrochemical characteristics of these compounds have been determined.59 In particular, triad 12, containing two fullerene fragments was synthesised for the first time in the 1-chloronaphthalene medium (Scheme 1).Scheme 1 RO C60 RFullerenes: functionalisation and prospects for the use of derivatives Yet another example of functionalisation of C60 via [4+2] cycloaddition is its reaction with indene, which isomerises into isoindene under the reaction conditions. The yield of the reaction product 13 was 30%, while about 35% of the C60 recovered unchanged. The electron-withdrawing capacity of the adduct was somewhat lower than that of C60 (see Ref. 60). 1,2-Cl2C6H4 C60 C60 >180 8C, 10 ± 12 h 13 The researchers cited 60 believe that the indene adducts might be useful for the development of new materials for optics and electronics. The new method for the synthesis of 1,3-diazines fused to fullerene, described by Torres-Garcia et al.,61 can be extended to a large number of diamines.It should be noted that the last stage is carried out in an acid medium in order to prevent nucleophilic addition to fullerenes. OSiMe3 OSiMe3 a b, c C60 OSiMe3 OSiMe3 R1 N O d C60 C60 N R2 O 14a ± d (a) C60, 1,2-Cl2C6H4 , 180 8C, >50%; (b) Br2 ,777 8C; (c) Et3N. 3HF; R1 (d ) NH2, 1,2-Cl2C6H4 , AcOH; NH2 R2 R1=R2=H(a, 37%); R1=H; R2=OMe (b, 29%); (d, 10%). R17R2=(CH=CH)2 (c, 32%), 1,3-Bis(tert-butyldimethylsilyloxy)-2-azabuta-1,3-diene (15) adds to C60 fullerene at room temperature.62 Hydrolysis, replace- ment of the silyloxy group by an alkoxy group (in order to avoid the retro-reaction) and dealkoxylation yield a fullerene derivative of d-valerolactam 16, which is a very interesting starting com- pound for the synthesis of amino acids and polyamides containing a fullerene fragment. These derivatives can be prepared after opening the six-membered ring.OR1 H2C O OR1 b a c N C60 C60 NH N HC OR1 OR1 OR1 15 (R1=SiMe2But) O O d C60 C60 NH NH 16 OR2 (a) C60 , PhH, 20 8C, Ar, 4 h; (b) 10% HCl (77%); (c) conc. HCl, R2OH, CHCl3 , D, 12 h: R2=Et (100%), (CH2)3OH (62%); (d) Et3SiH, CF3CO2H, CHCl3 , D, 12 h (88%). On prolonged heating, C60 adds 7-substituted 4-hydroxytro- pone.63 The tropone moiety exists as two tautomeric forms, A and B.Since the reactivity of the sterically more hindered form B increases at elevated pressures, in most cases, the yields of reaction products 17a ± c increase when the reaction is carried out under a pressure of 300 MPa. 893 X X OH O C60, 100 8C HO O B A C60 C60 HO O X X 17a ± c O O X = H (a), OMe (b), Cl (c). g-(9-Anthryl)-7-oxaoctanoic acid 18 adds to C60 in a similar way.64 Adduct 19 forms true Langmuir films on an aqueous phase; they can be transferred as monolayer and multilayer films onto substrates, quartz, Si[111] or calcium fluoride plates.64 X C60 X C60 PhH, 60 8C, 7 days 19 18 X=CH2O(CH2)5CO2H. At 160 ± 180 8C, fullerene C60 reacts with norbornadiene and its 7-spirocyclopropane derivatives (a 200- to 300-fold excess of the diene is used in the reaction) to give, depending on the substituent in the cyclopropane ring, [1+3]-, [1+4]- or [2+2]- type adducts, including polyaddition products.65 The fullerene analogue of thiochroman 20 was obtained by the hetero-Diels ± Alder reaction with thioquinonemethide (21),66 formed in situ from benzothietane (22).C60 180 8C, Ar S C60 5 min S S 21 22 20 The bulky fullerene fragment does not prevent oxidation of compound 20 at the sulfur atom bym-chloroperbenzoic acid.This reaction at 20 8C gave a sulfoxide (yield 89%) and sulfone (yield 45%). At room temperature, C60 readily reacts with 2,5-dime- thylthiophene S-oxide and with 2,5-dimethylthiophene S,S-diox- ide; the reaction with the sulfoxide affords monoadduct 23, and the reaction with the sulfone gives tetraadduct 24.67 SO SO2 Me Me C60 C60 Me Me 4 24 23 The goals of the synthesis of the adducts 23 and 24 were firstly, to carry out further functionalisation at the SO and SO2 groups and, secondly, to study their biological activities, because cyclo- adducts with thiophene oxide fragments are used as biologically important reactive metabolites.Llacay et al. 68 were the first to synthesise the fullerene adducts with organosulfur compounds containing a sulfonyl group capa- ble of being eliminated, for example, with tetrathiofulvalene 25.894 CO2Me S S C60, PhMe, PhCN, D O2S 7SO2 S S CO2Me 25 CO2Me S S C60 S S CO2Me 26 S S C60, PhCl, PhCN, D C60 S S O2S 7SO2 S S Thione 26 was isolated in 41% yield as a highly stable clathrate including 0.5 molecules of CS2 .Fulleropyrrolidines, which are fairly promising compounds for the preparation of stationary phases for liquid chromatogra- phy of biological objects, have been synthesised by stereospecific thermal 1,3-dipolar cycloaddition of C60 to cis- and trans-N- benzyl-1,2-diphenylaziridines.69, 70 Thermally induced cleavage of the aziridine ring in N-[3-(triethoxysilyl)propyl]-2-methoxycar- bonylaziridine (27) in the presence of C60 gives rise to fulleropyr- rolidine 28, which can be attached by a covalent bond to silica gel used for HPLC. MeO2C C60 N(CH2)3Si(OEt)3 PhCl, D, 30 h 27CO2Me SiO2 C60 N(CH2)3Si(OEt)3 PhMe, D, 6 h 28 CO2Me OEt O C60 N(CH2)3Si O This new adsorbent with the immobilised fullerene-containing group displays an exceptionally high separating capacity with respect to cyclic oligomeric compounds, calixarenes or cyclo- dextrins in both organic and mixed aqueous-organic media.70 Optically pure fulleropyrrolidines have also been synthesised.71 Stationary phases modified with chiral fullerenes can be used for separation of enantiomers.When a mixture of a-amino acid, a carbonyl compound and C60 fullerene is heated in chlorobenzene, fullerene undergoes a [2+3]-type cycloaddition to the intermediate azomethine ylide 29, giving rise to oligoadducts containing one to three pyrrolidine fragments, depending on the reaction conditions.72 ± 77 PhCl, D Me2C(NH2)CO2H+ 7CO2 O Me Me + 7 C60 C60 NH NH Me2C N2 29 n n=1±3.The reaction with methyl nonyl ketone follows a similar route.72 Heating of sarcosine, terephthalic aldehyde and C60 in toluene (in this case, the intermediate ylide is stabilised by the phenyl group) gave compound 30, which was converted into the amphiphilic 1-methyl-2-(4-dimethoxymethylphenyl)fullereno- [C60][c]pyrrolidine (31).74 E N Karaulova, E I Bagrii CHO C60 MeNHCH2CO2H+ CHO C60 C60 NMe NMe H2NCH2CO2Me. HCl PhMe, MeOH, D, 4 h 31 30 CHO CH(OMe)2 Compound 31 forms Langmuir ± Blodgett films and mono- layers on an air/water interface; these films are also formed when 22-tricosenoic acid is added.They possess good transparency and thermal stability.73, 74 Fluorescence of aryl-substituted fullerenopyrrolidines has an interesting feature, namely, its intensity is much higher than that of C60, which is due to the decrease in the symmetry of the molecule. At the same time, the electron-withdrawing ability of the C60 spheroid in these adducts is markedly lower than that of fullerene.77 Silylamines can be used in this type of synthesis instead of the amino acid; the compound 32 obtained in this way has been oxidised to a quinone.78 CHO OH C60 PhNHSiMe3+ 1,2-Cl2C6H4, D HO O HO O OH DDQ PhH, AcOEt C60 C60 NH NH 32 (50%) Ph Ph By varying the amino acid and the aldehyde used, one can prepare biologically active compounds.For example, a fullerene analogue of natural nicotine 33 has been prepared in order to estimate the influence of the fullerene fragment on the toxicity of nicotine.79 However, the solubility of compound 33 in solvents miscible with water was found to be too low. N CHO C60 MeNHCH2CO2H+ PhMe, D N C60 NMe 33 The amino acid was modified by an efficient solubilisation agent, triethylene glycol; this gave compound 34, which reacted with aldehydes and C60 fullerene to give fulleropyrrolidones 35 (including those containing a nicotine fragment) soluble in an H2O±DMSO mixture.79 C60, D MeO(CH2)2O(CH2)2O(CH2)2NHCH2CO2H+RCHO 34 C60 N(CH2)2O(CH2)2O(CH2)2OMe R 35a ± c R=3-Py (a), 4-Py (b), CH2O(CH2)2O(CH2)OMe (c).Fullerenes: functionalisation and prospects for the use of derivatives Compound 35c exhibits an antimicrobial activity including that with respect to a large number of pathogenic microorgan- isms.79 On oxidation with m-chloroperbenzoic acid, fulleropyrroli- dines are converted into stable nitroxides possessing ferromag- netic properties.75, 76 Me Me Me Me 3-ClC6H4CO3H, N2 , 40 8C C60 C60 NH NO.Me (CH2)8Me Me (CH2)8Me Paramagnetic nitroxide 36 with a spirocyclohexane fragment has been synthesised from C70 fullerene.76 + 7 C70 NH Me2C(NH2)CO2H+ Me2C N2, PhCl, D 29 O 3-ClC6H4CO3H, PhH C70 C70 NO. NH D 36 Me Me Me Me Nitrile oxides generated from oximes or in another way add to C60 giving rise to fullerenoisoxazolines.77 ± 84 Thus fulminic acid (37) is converted into fullerenoisoxazoline 38, while its esters 39a ± c afford derivatives 40a ± c with retention of the alkoxycar- bonyl group.80 O C60, PhMe, 80 8C C60 N HON C(Cl)CO2H 7CO2,7HCl 38 37 O Na2CO3 C60 C60 [ONC CO2R] HON C(Cl)CO2R N PhMe, H2O 39a ± c 40a ± c CO2R OMe MeO R=Bn (a), But (b), (c).OMe MeO CO2Me The readily available fullerenoisoxazoline 38 is a convenient starting compound for the synthesis of diverse functional deriva- tives of fullerene.81 OCOBun OH a b 38 C60 C60 CN CN NMe2 , (a) Et3N, N2, PhMe, 70 8C, 6 h,410%; (b) BunCO2H, N N C N , PhMe, MeCN, 25 8C (100%). Fullerenoisoxazolines decompose under relatively mild con- ditions with quantitative regeneration of C60 .This process occurs on refluxing in chlorobenzene in the presence of Mo(CO)6 or on treatment with excess diisobutylaluminium hydride in toluene at 20 8C. 82 The decomposition of the isoxazoline fragment can be used either to control fullerene functionalisation or to perform further functionalisation. Thus treatment of substituted fullere- noisoxazoline 41 with formaldehyde yields tricyclic derivative 42. When the latter compound is made to react with Mo(CO)6, N-substituted fullerenopyrrolidine 43 is produced.82 N O + HO2CCH2NH2(CH2)4 F3CCO¡2 41 (CH2)4 N N O42 (a) CH2O, PhMe, D, 67%; (b) Mo(CO)6 , PhCl, D. The method for the synthesis of cyclopentenes based on cycloaddition of 2,3-dienoates or 2-ynoates to electron-deficient alkenes in the presence of phosphine has been employed success- fully for the synthesis of fullerenocyclopentenes.The driving force of this reaction is nucleophilic attack by a tertiary phosphine on a b-allene or acetylene carbon atom. A series of successive proto- tropic shifts is followed by b-elimination of phosphine giving rise to fullerenocyclopentene 44.85, 86 PBu3 C CHCO2Et CH2 PhMe, 20 8C, N2, 17 h + 7 C60 PBu3 CO2Et C6044 CO2Et The first example of cycloaddition of ylide 45, prepared from N-benzyl-1-(4-nitrophenyl)chloromethylideneamine (46), to C60 fullerene was reported. 87 CCl NCH2Ph O2N 46 + 7 4-NO2C6H4 C N CHPh 45 However, based on the 13C and 1H NMR data, the authors suggest that, in addition to the normal [6-6]-adduct 47, the reaction yields an approximately equal amount of a [5-6]-adduct with an open annulene structure as a mixture of two diaster- eoisomers.It is known that photoexcited C60 fullerene in the triplet state has stronger electron-withdrawing capacity than that in the ground state. Thus it is able to react with moderately electron- releasing reagents according to the [2+2]-cycloaddition pat- tern.88 ± 90 For example, anthracene adds to C60 on exposure to a laser radiation withl5500 nm giving rise to adduct 48.91 895 a (CH2)4CN N b 43 +PBu3 C60 CH2 7 CHCO2Et + C60 PBu3 7 7PBu3 CO2Et Et3N PhBr Ph C60 C60 N PhBr, 130 8C 47 C6H4NO2-4896 C60, hn PhH 4-Propenylanisole does not react with C60 fullerene without light.On exposure to the light of a xenon lamp (l>500 nm) at 20 8C, both cis- and trans-isomers are converted into monoad- ducts (yield 40%, the trans-compound being the major product), which, however, decompose to the initial components under conditions of mass spectroscopy.88 Studies of the isotope effects and the 1H NMR spectra of the adducts formed in the reaction of C60 fullerene with cis-2-deuterio-1-(4-methoxyphenyl)propene made it possible to propose a two-step reaction mechanism involving a bipolar or biradical intermediate.89 Jensen et al.90 have studied the behaviour of C60 fullerene in the De Mayo reaction, i.e. [2+2] photocycloaddition of enolys- able 1,3-diketones and 1,2-diketones to alkenes followed by opening of the cyclobutane ring giving rise to cyclooctadienones.The reactions of C60 with compounds 49 resulted in the synthesis of fused fullerenofurans, both chiral (50a,b) and achiral (51a,b). O C60, hn R2 PhH R2 R1O 49 O H O H 50a,b R1=H,Me3Si; R2=H(a), Me (b). Presumably, the reaction involves either intermolecular oxi- dation by singlet oxygen of the cyclobutane intermediates formed initially or intramolecular oxidation of these intermediates by a triplet fullerene fragment. Synthesis of methanofullerenes using cycloaddition reactions is especially significant. Methanofullerenes can serve as the initial compounds for the preparation of functional derivatives of fullerene by introducing substituents into the side chains of the adducts.Routes to methanofullerenes are diverse. They include thermal addition of diazo compounds followed by thermolysis or photolysis, addition of carbenes, and reactions with ylides, which follow an addition ± elimination mechanism. Cyclopropanation by diethyl bromomalonate (Bingel reaction 92) is among the best and most widely used cyclopropanation methods. Recently this reaction has been modified 93 in such a way that bromomalonates (which are difficult to prepare) are formed in situ during the cyclopropanation of fullerene. Fullerene is made to react with malonates in the presence of CBr4 and diazabicyclo[5.4.1]undec-7- ene (DBU) at 20 8C in toluene; the reagents are taken in a 10-fold excess (DBU in a 20-fold excess) with respect to fullerene.CBr4, DBU C60+CH2(CO2R)2 R=3,5-(3-BnOC6H4CH2O)2C6H3CH2 (29%), Et (57%), C18H37 (65%), S (55%). Me Me C60 48 O R2 R2 R2 R2 O + 51a,b CO2R C60 6 h CO2R E N Karaulova, E I Bagrii The reaction of monoesters of malonic acid, I2 and DBU with C60 resulted unexpectedly in iodinated derivatives 52a ± c.94 CO2R I2, DBU C60 C60+RO2CCH2CO2H PhMe, 20 8C, 24 h I 52a ± c R=Bn (a), 3,5-(C12H25O)2C6H3CH2 (b), (CH2)2O(CH2)2OEt (c). Presumably, the carbanion arising at the first stage is not sufficiently nucleophilic to be converted into diiodomalonate; instead, it is decarboxylated to give RO2CCH2I, which adds to fullerene. The reaction with diester 53 yielded fullerenocyclopro- panedicarboxylate 54 and monoester 55. CO2But CF3CO2H C60, I2, DBU C60 ButO2CCH2CO2R CH2Cl2 PhMe, 20 8C CO2R 53 54 N NMe2 , H CO2H CH2Cl2 C60 C60 7CO2 CO2R CO2R 55 R=3,5-(C12H25O)2C6H3CH2 .Amphiphilic C60 derivatives, 56 and 57, containing an unde- canoic acid fragment, have been synthesised (in 45% total yield); the purpose of the synthesis was to study the properties of thin films prepared from these compounds. The monoadduct 56 was resolved into isomers A±C using column elution chromatogra- phy.95 C60 H2NCON(NO)(CH2)10CO2Et KOH, PhMe, 20 8C, 12 h H H + C60 C60 (CH2)9CO2Et (CH2)9CO2Et 2 56 57 H H X X H X C B A X=(CH2)9CO2Et. The side chain in methanofullerenes is often modified in order to obtain biologically active compounds.Mono- and di-carbox- ylic acids, especially 1,2-methanofullerene[60]-61-carboxylic (fulleromalonic) acid, are used most frequently as the starting compounds. Thus N-phenyl(1,2-methanofullerene[60])-61-for- mohydroxamic acid has been prepared from this acid.96 When 1,2-methanofullerene[60]-61,61-dicarboxylic acid (58) reacts with N-hydroxysuccinimide (NHS) and chloro-N,N-di- cyclohexylcarbodiimide (DCC), it eliminates water and CO2 being thus converted into compound 59, which reacts with primary and secondary amines, 1-aza-12-crown-4 and 1,6-aza- homo[60]fullerene.97 H CO2H NHS, DCCl C60 C60 CO2 CO2H OC NOC 59 58Fullerenes: functionalisation and prospects for the use of derivatives Syntheses of diesters 60 (from porphyrin, open crown ether and methanofullerenecarboxylic acid) and 61 (from azafullerene with an `open' [6-5] bond and porphyrin) have been per- formed 98, 99 for biological purposes, in particular, for photo- dynamic therapy, and with the aim of elucidating the possibility of generation of singlet oxygen in the presence of fullerene derivatives in the triplet state.C60 CO2H C60C(O)O R C60 C(O)O 60 C60 NCH2CH2O 61 Ph N NH Ph . R= HN N Ph The reaction of C60 with tert-butyl diazoacetate in benzene or toluene in the presence of dirhodium tetraacetate affords a new type of cycloadducts, tricyclic compounds 62a,b, instead of the expected methanofullerenes.It is assumed that initially, cyclo- propanation of the solvent gives norcaradiene 63, and it is this product that reacts with fullerene.100 Hydrolysis of the ester group yields acids 64a,b, which are then converted into derivatives 65a,b at the amino group of leucine. R +N2CHCO2But C60 R62a,b HC60 R64a,b HC60 R 65a,b R = H (a), Me (b); DCC is dicyclohexylcarbodiimide. However, when 1-methylnaphthalene is used as the solvent, fullerene smoothly reacts with ethyl diazoacetate in the presence HO O O 2 C(O)O O O 2 O O 2 O O 2 R O(O)C O 2 H C60 Rh2(OAc)4 CO2But 7N2 R 63 CF3CO2H H CH2Cl2, D CO2But PriCH2CH(NH2)CO2But N H DCC, N, Et3N N CO2H OH H C(O)NHCHCH2Pri CO2But 897 of a stoichiometric amount of dirhodium tetraacetate to give normal methanofullerenes, mainly, at a [6-6] bond.101 A mecha- nism including the formation of metallacycle 66 was proposed for the reaction.C60+N2CHCO2Et+Rh2(OAc)4 + 7 Rh H C60 C60 7Rh2(OAc)4 CO2Et CO2Et 66 The addition to C60 of sulfonium ylides stabilised by cyano, carboxy or carbonyl groups followed by elimination of dimethyl sulfide affords fullerene derivatives which are highly promising as the starting compounds for subsequent functionalisation. The reaction is carried out either under phase transfer catalysis conditions or by generating the ylide in situ upon disproportiona- tion of a sulfonium salt on treatment with K2CO3.102 This gives rise to a mixture of mono-, bis- and tris-adducts 67.Unlike methanofullerenes prepared from diazo compounds, products 67 have a methylene bridge at a [6-6] bond; no isomers containing an `open' [6-5]-ring system (type A, see above) were detected. H + K2CO3 C60 Bun4 NBr, PhMe, H2O C60+Me2S CH2R Br7 R n 67 R=CN, CO2Et, COPh, COCMe2CH2CO2Et, CO2But; n=1±3. A new methanofullerene containing a fragment of a stable phosphonium ylide has been prepared by the reaction of C60 with a mixture of equal amounts of triphenylphosphine and dimethyl acetylenedicarboxylate.103 The reaction involves the formation of carbene ylide intermediate 68a or zwitter-ion 68b. MeO2CC CCO2Me+Ph3P + CCO2Me O7 Ph3P C60 C C C Ph3P C PhMe, D, 120 h OMe MeO2C CO2Me 68b 68a C60 CO2MeOMe + O7 Ph3P Me C60 Spiro derivative of C60 fullerene, adduct 69, containing an `open' [5-6] bond has formed unexpectedly in the reaction of C60 with 4,4,5,5-tetramethyl-1,3-diazolidine-2-thione (70) and DL- valine.104 Me Me Me +Me2CHCH(NH2)CO2H NH HN N2, PhCl, D, 20 h S 70 HN MeMe C60 Me Me NH 69 The mechanism assumed for this reaction includes the reaction of thione 70 with amino acid to give a carbene, which adds to fullerene.Shen et al.105 were the first to demonstrate that the reaction of C60 fullerene with chiral 1,4-di-tert-alkoxy-2,3-bisazidobutanes gives rise to chiral bisazafulleroids. Enantiomeric pairs with the898 (2R,3R) and (2S,3S ) symmetry were obtained in *51% yields.The enantiomers exhibit mirror image circular dichroism curves. Me2(R)CO OC(R)Me2 N N 71 (R=Me, Ph) Apparently, removal of the tert-alkyl groups from compounds 71 would provide a route to water-soluble chiral fullerene derivatives, which is important for the investigation of their biological activities. A bisazafulleroid fragment has been inserted in a crown-ether chain.106 2 O O O N3 According toNMRdata, the bisadduct 72 has a structure with `open' [5-6] bonds. The fullereno-crowns are excellent agents for binding metals. Perhaps, new superconducting substances would be found among these complexes.106 The reaction of fullerene with tetrabromides 73a ± d has resulted in selective multifunctionalisation of fullerene, the goal of which was to prepare disubstituted bisphenols with substituents located at particular rings in the cage.107 (CH2)n O O CH2Br BrCH2 CH2Br CH2Br 73a ± d (CH2)n O O 75a ± d n=2 (a), 3 (b), 4 (c), 5 (d). O 2 O O N N C60 , PhCl N3 72 (28%) C60 PhMe, D, KI, 18-C-6 OH OH BBr3 PhH 74a ± d (100%) E N Karaulova, E I Bagrii The reaction is carried out under conditions of high dilution (1074 mol litre71) and the guiding (CH2)n chain is subsequently split off.107 When n=2 ± 5, the addition occurs selectively within one hemisphere of the C60 cage. When n=2, 3, only cis-2- and cis- 3-isomers are formed and for n=5, bisaddition yields the e-adduct. When n=4, the reaction gives a complex mixture of bisadducts. For both the cyclic precursors and acyclic bisadducts, types of symmetry and the relationship between the symmetry and chirality were established by analysing the data of mass, NMR, UV and visible spectra (all adducts with C1 and C2 symmetry are chiral and those with Cs symmetry are achiral).The chiral cis-3- and e-adducts were resolved into enantiomers using HPLC with a chiral phase [cellulose tris(3,5-dimethylphenyl carbamate)-coated silica gel (CHIRAL CEL OD), hexane ± isopropyl alcohol]. For example, for n=2, the bisphenol 74a was obtained. Both the precursor 75a and the final diol 74a are cis-2 isomers (this was observed for the first time for fullerene bisadducts) and are achiral; the compound 75a corresponds to Cs symmetry. The cyclic precursor 75d and the bisphenol 74d for n=5 are e-isomers with C1 symmetry; they are chiral and optically active.The total yield of the isomers at the final stage is 100%. The [a]D values for the ethers 75a,d (5.76102 and75.26102) and the diols 74a,d (4.4610 and77.4610) were determined. The structures of optical isomers of the bisphenol 74d are presented below.OH HO OH HO 74d The compounds 74a ± d belong to the few examples of chiral derivatives with achiral substituents known in the chemistry of cage compounds, in particular, in the fullerene chemistry. The reaction of C60 fullerene with diazo compound 76 has resulted in the synthesis of spirane methanofullerenes 77 with quinone type addends.108 O N2 R R C60, hn 1,2-Cl2C6H4, 0±10 8C R R C60 77 O76 R=H, Me, But.Spiro[10-anthrone-9,610]methanofullerene (78) and spiro[cyclohexanone-4,610]methanofullerene (the latter as a mix- ture of [6-6] and [5-6] isomers) were prepared in a similar way. The conjugated dienone 78 exhibits stronger electron-withdrawing capacity than the initial C60 fullerene. The anthrone 78 was converted into compounds 79 and 80.108 CN NC O NCN a b C60 C60 C60 79 78 80 (a) Me3SiN=C=NSiMe3 , TiCl4 , CHCl3 , D, 5 days (60%); (b) CH2(CN)2 , TiCl4 , Py, CHCl3 (26%).Fullerenes: functionalisation and prospects for the use of derivatives Chemical modification of more than two double bonds in the C60 molecule gives, as a rule, a mixture of regio- and stereo- isomers. Therefore, study of the possibility of synthesising a single isomer by multifunctionalisation of C60 is of interest.These molecules could serve as useful tools in the studies of functional catalysis and in the materials science; they could find application in bioorganic chemistry (for example, for reactions with DNA, enzyme inhibition, design of compounds for drug transport, and for other purposes 109). Hirsch has studied the regiochemistry and stereochemistry of multiple addition to the double bonds in C60 and has developed a theoretical substantiation of these reactions.13, 21 Double (and multiple) cyclopropanation with dialkyl bromo- malonates according to the Bingel reaction pathway is used most often to prepare polysubstituted fullerenes.92 Thus C60 has been made to react 93 with the malonate and CBr4 in the presence of DBUon the addition of 10 equiv.of 9,10-dimethylanthracene and at longer reaction times; under these conditions, symmetrical hexakis(adducts) 81 were obtained in good yields. R C60 R 6 81 R=CO2CH2Me, CO2(CH2)16Me. When pentakis(adducts) 82 with C2v symmetry are again treated with diazomethane, heptakis- and octakis-adducts with Cs symmetry, pyrazoline derivatives, are formed with high regioselectivity. These products are converted into methanofuller- enes by thermally induced abstraction of nitrogen. The direction of nitrogen abstraction was assumed to be controlled by orbital symmetry.110 O O O O X X 82 X=C(CO2Et)2 , C(CO2CH2CO2Et)2 .The tetrakis(adduct) 83 contains a fulvene p-subsystem located on the spheroid surface. This compound is prepared from a mixture of C60 , fluorenylpotassium and fluorene in THF. The mixture is kept under an argon atmosphere in the presence of traces of oxygen for 72 h and the adduct is then isolated by preparative chromatography.111 The tetrakis(adduct) 83 adds nucleophiles, which is typical of a fulvene p-system. Thus it reacts with 1-octynyllithium in THF or with NaCN in a DMF± chlorobenzene mixture to give penta- kis(adducts) 84a,b in 63% and 68% yields, respectively. H H 1) RM 2) H+ H H 83 899 H H H H HR 84a,b RM=C6H13C CLi (a, 63%), NaCN (b, 68%). Long-term heating of C60 with compounds 85 in anhydrous 1,2-dichlorobenzene in the presence of 4Amolecular sieves results in double cycloaddition giving rise to tricyclic compounds 86.109 (CH2)n (CH2)n H+ C60 C60 R R 85 R R R R R R 86 (CH2)n C60 O O 87 R7R=OCMe2O; n=3, 4, 6.Acid hydrolysis affords quantitatively the corresponding diketones 87. A [2]catenane containing the C60 bishomofullerene fragment was synthesised for the first time using a sequence of two modified Bingel reactions (yield 18%).112 O O O O O O O OEt O N+ N+ O + + O N N OEt O 4PF¡6O O O O O 2. Nucleophilic, electrophilic and radical addition Nucleophilic addition to C60 fullerene gives, as a rule, mono- adducts; for example, C60 adds one amino acid molecule.113 Dipeptides react in a similar way.114 H C60 C60+H2NR NHR R is an amino acid or dipeptide residue.Fullerene dipeptide derivatives are soluble in water owing to association with amino acid residues.114 These adducts penetrate900 inside the liposomes, they exhibit adjuvant { properties and antiviral activity.113, 115, 116 The reaction of C60 in 1,2-dichlorobenzene with NaCN in DMF affords not only the [C60(CN)]7 anion but also the [C60(CN)]27 and [C60(CN)]37 anions, although the monoion predominates (as shown by negative-ion mass spectrometry).117 Mechanochemical activation (mixing of solid reactants in a vibratory mill 118) can be used for functionalisation of fullerenes. The reaction of C60 with NaCN carried out in this way gave the dimer C120 88, in which two spheroids are linked via a cyclobutane ring. The fact that the reaction pathway differs from that observed in solutions is due to the unique ability of the cyanide ion to serve as both a nucleophile and a leaving group.7 C60 C60 C60 C60 C60 7CN7 CN 88 CN7 The reactivity of the dimer 88 does not differ much from that of C60 . For instance, it enters into the Bingel reaction with diethyl bromomalonate and DBU (in equimolar amounts) to give a mixture of isomeric monoadducts (at least, three compounds) in 44% total yield. 118 Tanaka et al.119 made use of the fact that the fullerene spheroid readily forms anions and synthesised an unusual class of compounds, hydrocarbon salts. But But RClO4 C60 C60 7KClO4 7 R+ 7 K+ R+ is tris[1-(5-isopropyl-3,8-dimethylazulenyl)]cyclopropenylium. Vinyl ethers add to the stable 2-(1-octynyl)-1,2-dhydrofulle- ren[60]-1-ide ion (89) in the presence of HCl; this affords [6-6] bisadducts 90, which are promising monomers for cationic polymerisation.120 C CC6H13 HCl, C6H14 CHOR C60 +CH2 7 0 8C, THF, 2 h 89 C CC6H13 C60 CH(Me)OR 90 R=CH2Pri, (CH2)2OAc.The fullerene C60 adds Grignard reagents ClMgCH2SiR3 (R=Alk, Ar, AlkO) yielding mainly compounds C60(H)CH2SiR3 in THF and only products of the formula C60(CH2SiR3)2 in toluene. 121 Fullerenylmethylsilanes such as C60(H)CH2SiMe2Cl, C60(H)CH2SiMen(OR)37n and C60[CH2SiMen(OR)37n]2 , where R=Alk, n=0, 1, are used to modify silica gel. The products thus formed are used as sorbents for HPLC, as heterogeneous photo- sensitisers and in catalysis.121 Irradiation of C60 and allylstannanes in a benzene or toluene solution results in the regioselective formation of fullerene mono- allylation products.The process is inhibited by oxygen or benzonitrile.122 Fullerene derivatives 91 have been synthesised in this way. This reaction occurs by a mechanism involving triplet C60 and single-electron transfer.122 { Adjuvants are compounds stimulating an increase in the number of antibodies. E N Karaulova, E I Bagrii R3 R1 R3 R1 H 3 + (C60) + 7SnBu3 C60 R2 R2 SnBu3 91 R1, R2, R3=H, Me. Mixed adducts 92 ± 94 were directly synthesised for the first time from C60 and 2,2-dibenzyl-1,3-diazidopropane 123 by reflux- ing the reactants in chlorobenzene.N N N N Bn Bn Bn Bn N N N N 93 (25%) 92 (18%) N Bn Bn N 94 (11%) Bistriazolines are precursors of bisdiazolines and can be converted into the latter by thermolysis in chlorobenzene. Ther- molysis of the bistriazoline 92 afforded the compound 94. The formation of the adducts 93 and 94 was interpreted 123 in terms of a radical mechanism, similar to the mechanism of triazoline thermolysis.N N2 C60 C60 N NR NR NR NR C60 C60 RN C60 NR C60 3. The addition of hydrogen and the properties of hydrides All attempts to prepare completely hydrogenated fullereneC60H60 failed. This compound appears to be highly unstable due to the enormous strain which arises upon the formation of twenty planar cyclohexane rings and numerous shielding H7H interactions.13 By 1997, hydrides with the compositions C60Hx (x=2, 4, 18, 32, 36, 36 ± 50, 42 ± 44) and C70Hx (x=2, 4, 8, 10, 34 ± 36, 36 ± 38, 36 ± 44) were known.124, 125 The methods used to prepare fullerene hydrides have all been considered in detail; 11, 125 they include Birch reduction, hydro- boration, hydrozirconation, solid-phase and liquid-phase hydro- genation with hydrogen, electrochemical reduction, reduction with diimide or with chromium acetate, photoinitiated electron transfer and reduction with metal hydrides.Despite the fact that fullerenes could be promising catalysts for activation of the C7H bonds in hydrocarbons via the formation of fullerene hydrides, as was convincingly demon- strated in a comprehensive review,11 this aspect has not yet received adequate attention.Studies are mainly concerned with the structure 124, 125 and spectral and photophysical properties of fullerene hydrides.124 ± 128 It is assumed that the solid hydride C60H36 corresponds to T symmetry and contains four isolatedFullerenes: functionalisation and prospects for the use of derivatives benzenoid rings, located in tetrahedral positions on the surface of the closed shell of the molecule. A preparative synthesis of fullerene hydrides was carried out by treatment of fullerenes with metals and water serving as a source of protons;125 although the number of possible isomers is enormous, only a few of them are formed.For instance, heating of a solution of C60 fullerene in toluene with a Zn(Cu) mixture (freshly prepared Zn dust is employed) gives rise to a mixture of hydrides, which are isolated by preparative liquid chromatogra- phy. Depending on the temperature, reaction time and the ratio of the metals, the compounds C60H2, C60H4 and C60H6 (two isomers) can be obtained in 66%, 45% and 38% yields, respec- tively. Reduction of C70 fullerene afforded C70H8 and C70H10 (three isomers) in a total yield of 62%. The hydrides C70Hn are all unstable in air and during storage. Fullerene hydrides and deuterides containing up to 24 hydro- gen atoms per fullerene molecule have been obtained upon hydrogenation (with either hydrogen or deuterium) at 1 ± 2.5 MPa and 573 ± 673K of a mixture of fullerite (83% of C60 , 15% of C70 and 2% of higher fullerenes) and powdered intermetallic compounds LaNi5 , LaNi14.65 and CeO3.129 At 800 K, fullerene hydrides (deuterides) decompose with evolution of gaseous hydrogen (or deuterium); at 1000 K, the reaction of fullerite with intermetallic compounds is accompanied by evolu- tion of the corresponding 3d metal.129 The hydrides prepared in this way can serve as sources of high-purity hydrogen. When fullerite (78% of C60 and 22% of C70) is deuterated in the presence of Pd clusters (Pd powder and Pd/C), the maximum content of deuterium corresponds to [C60+C70]D26 and is attained after ten repetitions of the heating ± cooling cycle (673 ± 293 K) at 2.5 MPa.130 Fullerene dihydride C60H2 has been synthesised by ultrasonic treatment of C60 fullerene in decalin for 2 ± 4 h in a sealed vessel under argon.131 The hydrogen needed to hydrogenate fullerene is formed upon sonolysis of the solvent.The dehydrating capacity of C60 fullerene may be useful in the studies of biological objects. Thus C60 dehydrates the Hantsch ester (the reaction simulates the action of dihydronicotinamide- adeninedinucleotide) at 100 8Cunder anaerobic conditions to give diethyl 2,6-dimethylpyridine-3,5-dicarboxylate; in this reaction, fullerene is converted into mixed hydrides C60Hn (n=2, 4, 6).132 It is of interest that this dehydrogenation can be carried out by leaving the reaction mixture in the light at 20 8C; in this case, C60 acts as a catalyst because it is recovered unchanged.132 If the oxidation is performed in an aqueous medium, C60 forms a colloidal solution containing non-crystalline and quite 60. is generated from C60 .The radical cation reacts with homogeneous particles with a dimension of *10 nm (as estab- On irradiation in the presence of sensitisers such as 9,10- dicyanoanthracene (and diphenyl as a co-sensitiser), the radical cation Cá hydrogen donors, for example, with tert-butyl methyl ether, propionaldehyde and alcohols giving rise to 1-substituted 1,2- dihydro[60]fullerenes.133 The hydride 1,2-C60H2 has been prepared in 70% yield by the reaction of photoexcited C60 fullerene with 10-methyl-9,10-dihy- droacridine.134 The proton at the 4-position of stable 1-tert-butyl-1,4-dihy- dro[60]fullerene undergoes base-catalysed migration to give a thermodynamically more favoured isomer, 1-tert-butyl-1,2-dihy- dro[60]fullerene.135 4.Oxidation The electronegative fullerene molecules can be easily reduced but cannot be readily oxidised; however, the reaction of fullerene with oxygen on exposure toUV light (in hexane) or heating of fullerene in the presence of oxygen results in vigorous oxidation accom- panied by fragmentation, i.e. rupture of the C=C bonds in the cage.13 Upon irradiation, C60 passes initially into the triplet state; the transfer of energy from this species affords singlet oxygen, which oxidises the fullerene.9 Oxygen derivatives of fullerenes receive little attention yet.The monoxide, namely, 1,2-epoxy[60]fullerene (95), is prepared by reactions with various oxidants including systems with cyto- 901 chrome P450;136 however, the yields are relatively low. A prepa- rative pathway to 1,2-epoxy[60]fullerene (95) has been described 137 in which fullerene is treated with the system `methyltrioxorhenium ± hydrogen peroxide (as a complex with urea)'. O Me Re MeReO3+H2O2 O+H2O2 O O Me O O C60 Re C60 O PhH O O O 95 (*35%) Interestingly, either component of this oxidative system taken alone give virtually no target product. The epoxide C70O, which is fairly difficult to obtain, has been isolated in *0.5% yield upon the use of an arc-discharge reactor with a rotating cathode and an anode made of carbon plates (the reactor operation time was 24 h).138 Study by 13C NMR spectro- scopy showed the presence of two isomers, 1,2-epoxy-C70 and 5,6- epoxy-C70 in 47 : 57 ratio.The predominance of the [5-6] isomer was interpreted 138 as being due to the lower strain in its molecule. The compound C120O (its synthesis has been reported 139) disproportionates at 400 8C to give 60% of C60 , 15% of C120O2 , and a mixture containing oligomeric (C60)nOm oxides (n=3, 5; m=3 ± 9).140 The dioxide C120O2 is stable at room temperature both in a pure state and in solutions. After thermolysis of C120O at 550 ± 560 8C, product C119 containing two C58 cages was isolated from the reaction mixture in 2% yield. The cages are connected by a bridge comprising three sp3-hybridised carbons, two of which are equivalent.141 Oxidation of C¡ 60 60 permits the preparation of an aqueous colloidal solution of C60 fullerene .142 The C¡60 anion is formed when a solution of C60 in THF is treated successively with an Ni7Al alloy and an aqueous solution of sodium hydroxide; C2¡ can be prepared in a similar way except that DMSO is used in place of THF.The anions are inert with respect to water but highly sensitive to oxygen. Oxidation occurs via consecutive steps C¡ C60 , 60 C260¡ the overall reaction being written as 4C¡60+O2+2H2O 4C60 (soln)+4OH7. lished by electron microscopy); in an anhydrous organic solvent, C60 precipitates as a black powder. Electrochemical reduction of C60O affords an electrochemi- cally active film; a different active film is formed upon the reduction of C60 to the dianion in the presence of O2 .143 Films of conducting polymers, which are deposited on the electrodes upon electrochemical reduction of C60O, are promising for the design of important materials for engineering (e.g., for microelectronics).144 An interesting feature of pentaphenyl-substituted fullerenes has been discovered.145 The compound C60Ph5Cl was kept for two weeks in air and in the light, and C60Ph5H was stored for a year.Both substances were spontaneously oxidised yielding complex mixtures of products. In both cases, a compound with the composition C60Ph4C6H4O2 was isolated from the mixture in 60% yield. Relying on NMR and mass spectroscopy, it was established that elimination of hydrogen or chlorine atoms linked to the cage and an ortho-hydrogen atom of the neighbouring phenyl group followed by oxidation gives rise to fullereno- benzo[b]furan 96.The mechanism of this reaction is not yet entirely clear.902 O Ph [O] Ph O C60Ph5X Ph Ph 96 (X=H, Cl) The radicals generated in the reaction of fullerene with tert- butyl peroxide on irradiation are highly stable under a nitrogen atmosphere and react only slowly with oxygen, water and with radical traps such as nitrobenzene.146 hn 2ButO (ButO)2 n ButO +C60 (ButO)nC60 n51. Fullerenols eliminate active oxygen radicals including super- acidic radical anions formed in biological systems,147 which opens up the way for using them in biochemistry and pharmacology.No compounds containing the Cá60 cation have been obtained. However, the salt 97, containing the Cá76 cation, which has a much lower electrochemical oxidation potential, has been synthes- ised.148 CH2Cl2 R3N+Ag(CB11H6Br6) + 12 Br2 7AgBr C76, 1,2-Cl2C6H4 [R3N]+[CB11H6Br6]7 7R3N [C76]+[CB11H6Br6]7 97 R=2,4-Br2C6H3 . The salt 97 can be stored for several weeks under a nitrogen atmosphere. Apparently, this is the first compound with a carbocation consisting only of carbon. 5. Halogenation It was shown back in 1991 ± 1992 that C60 fullerene reacts with gaseous chlorine and fluorine and with liquid bromine giving complex mixtures of addition products.13 Subsequently, these studies have not progressed greatly due to the difficulties in the isolation of individual compounds and in controlling halogena- tion reactions.It had been considered 9 that iodine does not react with fullerene; however, later,40, 41 CTC involving iodine and C60 were prepared. The C60 fullerene readily reacts with chlorine under free radical reaction conditions to give C60Cl40 (see Ref. 149). This compound can be used as an intermediate in the further fullerene functionalisation, namely, synthesis of polyfullerol and poly- fluorofullerene. The reaction of C60 with excess iodine chloride (toluene, 20 8C) gives rise to an individual compound, C60Cl6 ; the reaction with liquid bromine affords C60Br2 , while the reaction in CS2 solvent results in C60Br5 (see Ref.9). Fluorination of fullerenes is receiving somewhat more atten- tion because fluorinated derivatives might be of interest for materials science.150 Thus fluorofullerenes containing up to 60 fluorine atoms per molecule, including C60F48 , 151 can be prepared by fluorination with fluorine gas, while hyperfluorofullerenes C60F76 can be synthesised on treatment with fluorine with exposure to UV radiation; this reaction is accompanied by cleavage of the spheroidal cage.149 When C60 is treated with MnF3 at 330 8C (24 h), C60F36 is formed.152 Judging by the mass spectrometric data, the mixture of fluorinated derivatives resulting from the reaction of C60 with KrF2 in anhydrous HF contains mainly E N Karaulova, E I Bagrii C60F44 and C60F46 (see Ref.150). Treatment of C60/C70 with vapours of the fluoridesWF6 , TaF5 , NbF5 and TiF4 results in the preparation of fluorides MFn .C60 (the MFn :C60 ratio is <1: 2).153 A preparative route to fullerene derivatives with organo- fluorine substituents (for instance, derivative 98) involves the reaction of C60 with halofluoroalkanes and tributyltin hydride under free-radical reaction conditions.154 Bu3Sn Bu3SnH Bu3Sn C60 CF2CO2Et BrCF2CO2Et 7Bu3SnBr PhH, N2, D, 30 h Bu3SnH EtCO2CF2C60 7Bu3Sn C60[6-6](H)CF2CO2Et 98 When n-C6F13I, BrCF2Br, n-C12F25I and I(CF2)6I are used, initiation of the reaction requires the addition of azoisobutyroni- trile. The hydrogen atom in the C60 ring in the resulting compounds is highly acidic and can be replaced by other func- tional groups in a weakly alkaline medium.154 V.Fullerenes in polymerisation reactions When C60 fullerene pellets are exposed to a pressure of 1.2 GPa at 600K for 5 h, [2+2] cycloaddition occurs to give polymers including linear polymers in which the C60 fragments are linked by cyclobutane rings.155 The C70 fullerene is more difficult to polymerise,156 which has been explained by steric hindrance to the [2+2] cycloaddition involving the double bonds in the equatorial region of the ellipsoidal cage. Judging by spectroscopic data,156 heating solid C70 fullerene at 750 8C under a pressure of at least 7.5 GPa affords dimers. Shvartsburg et al.157 have designed a setup permitting gen- eration of theCá=¡ 120 , Cá130 =¡ and C140 á=¡ ions, which are formed upon laser-induced desorption of crude C60 films , and measured their mobility and the trajectories.The trajectories of these ions were in satisfactory agreement with those calculated for the dimers produced upon [2+2] cycloaddition. This study is the first experimental proof of the structure of fullerene dimers.157 Simultaneous laser-induced vaporisation of C60 and graphite at a low helium pressure and a temperature of 400 ± 500 8C or 1900 ± 2400 8C gives rise to polymers of the composition C60(CC60)n71 (n=2 ± 7), the dimer C121 being the predominant species.158 When the laser radiation intensity is low, polymers dissociate to give the monomeric components, whereas high radiation intensity results in the elimination of C2 .During polymerisation of ordinary monomers, C60 can be introduced both in the backbone and in the side chain of the polymer formed. The chain growth reaction during polymerisation of polystyr- ene in the presence of C60 yields stable free radical (polystyr- ene)C60 . (see Ref. 159). The C60 and C70 fullerenes copolymerise with styrene by a standard free radical mechanism both in block and in aromatic solvents;160 this affords dark-brown polymers soluble in the same solvents as polystyrene and fluorescing more intensely than polystyrene (fluorescence is shifted to the blue region). High-pressure (0.1 to 1.5 GPa) polymerisation of styrene in the presence of C60 gives rise to a mixture containing products both soluble and insoluble in chloroform.161 An insoluble poly- styrene with the chains cross-linked by fullerene molecules was isolated from this mixture.Modification of polymers by C60 fullerene fragments changes markedly their physicochemical properties and, in some cases, substantially increases the solubility. A new method for the synthesis of soluble fullerenated polymers by the reaction of carbanionic polymeric intermediates with C60 has been devel- oped.162Fullerenes: functionalisation and prospects for the use of derivatives 7 CH2CH CH2C C60 NaH, THF N2, 20 8C, 48 h PhH, 7 days Br Br n n C CH CH2 CH2 17x H C60 x Br n 99 Poly-4-bromostyrene 99 with pendant fullerenes is readily soluble in most common organic solvents (CCl4 , CHCl3 , PhH, THF).Poly(vinylbenzyl chloride) modified with C60 fullerene and soluble in organic solvents has been prepared in a similar way.163 It was found164, 165 that C60 fullerene substantially inhibits free- radical polymerisation of vinyl monomers in the presence of AIBN by reacting with the initiator and/or by terminating the chain growth.164, 165 This is true for copolymerisation with methyl acrylate, methyl methacrylate, acrylonitrile, vinyl acetate and N-vinylpyrrolidine.164 Only with highly reactive monomers such as cyanoethyl acrylate and cyanovinyl acetate, does C60 give copolymers (yields 53% and 65%, respectively). Free-radical copolymerisation of 4-vinylbenzoic acid and 2- or 4-vinylpyr- idine with C60 results in polyelectrolytes.166 Apparently, in addition to non-linear copolymers, the process yields homopol- ymers and microparticles, which can be separated from copoly- mers only using microporous filters.When styrene or methyl methacrylate is subjected to radical polymerisation in the presence of C60 (1 mass % in the reaction mixture), fullerene reacts at early steps of the process. This gives rise to branched structures; the proportion of the linear polymer increases with time. Physicochemical properties of these polymers have been characterised in detail. 167 As noted above, the synthesis of water-soluble fullerene derivatives, in particular water-soluble polymers, is important regarding the investigation of the biological properties of full- erenes.Derivatives of this type have been prepared by covalent attachment of hydrophilic linear polymers to the fullerene frag- ment.168 Thus the poly(propionylethyleneimine ± ethyleneimine) copolymer is added to methano[60]fullerenedicarboxylic acid in the presence of 1-ethyl-3-(dimethylaminopropionyl)carbodi- imide. The reaction is carried out in chloroform at 0 8C. This gives water-soluble polymer 100 with pendant C60 groups (molecular mass 49 000) in about 15% yield. CH2CH2N CH2CH2N CH2CH2NH n CO EtO2C HO2C z C60 y 100 The high solubility of this polymer is due to the presence of the N,N-diethylenepropioamide fragment, which is structurally O R a C60 C60 C N N HO2C CO2H H H H H H 101 CO2H HO2C (a) , H2NRNH2, (PhO)3P, LiCl, , 110 8C, 3 h; R= O NMe 903 related to N,N-dimethylformamide and N,N-dimethylacetamide, miscible with water.New copolyamides containing dimethanofullerene fragments in the backbone have been synthesised 169 using [60]fullerene- bismethanocarboxylic acid (101) (Scheme 2). The polymer forms a brown brittle film capable of coating glass. Cationic polymerisation involving C60 fullerene has been studied. 170 Multicharged cations derived from C60 (they are generated 171 using a special ion source) are highly reactive in the gas phase. Dications and trications initiate cyclochain polymer- isation of allene or propyne under these conditions.170 More than sixteen allene molecules add to the C2á 60 dication, odd-numbered adducts reacting approximately 10 times faster than even-num- bered ones.It is assumed that multiaddition occurs during the chain growth; the alternation of the reaction rates is attributed to the difference between charge distribution and stability due to the formation of either acyclic or cyclic cation, for example, 102 or 103. C C + C C60 102 C60 C C C C C 103 C + Studies dealing with the development of conducting and semiconducting polymeric systems involving C60 are in progress. One method for the synthesis of conducting polymeric composites is polymerisation on a supporting template.172 The template is a cross-linked fullerenol ± polyurethane elastomer. It is impreg- nated with aniline; this gives a film, which is brought in contact with an oxidising aqueous solution containing LiCl, (NH4)2S2O8 and HCl at 715 8C and is thus converted into a new elastomer with inserted polyaniline.This product is characterised by much higher conductance than the initial `prepolymer' without losing the elasticity and tensile properties. The presence of the template hampers migration of the conducting species, which is a signifi- cant reason for deterioration of the physical properties of the material. The bisthiophene derivative 104, functionalised with fullerene, polymerises at the electrode during voltammetry in CH2Cl2 in the presence of Bun4 NPF6.173 S S C60 N(CH2)7 C60 104 S S 105 An electrically conducting polymer of the `necklace with a pendant' type was prepared for the first time by anodic oxidation of the bisthiophene derivative containing a fullerene fragment 105.174 The change in the photophysical and optical properties of conducting polymers upon the formation of composites with C60 fullerene is due to complex formation (resulting in the change in Scheme 2 O O O R C C N N C H H H n m O .904 the polymer conformation 175) and, what is more important, to the transfer of electrons from the polymer to C60 (see Ref.176). The spectrum of the film prepared from a mixture of C60 with poly(3- hexylthiophene) exhibits a substantial shift in the UV and visible region.175 Doping of poly(3-octadecylthiophene) with C60 causes quenching of the polymer photoluminescence.176 The poly-2- methoxy-5-(20-ethylhexyloxy)-1,4-phenylvinylene film doped with C60 fullerene is promising for the manufacture of holo- graphic recording materials.177 VI.Fullerenes in catalysis Fullerene C60 participates in catalysed processes as either a part of the catalyst and/or a substrate which modifies the action of the catalyst. The complex Z2-C60Pd(PPh3)2 has been used as the catalyst in the homogeneous (in toluene) and heterogeneous (supported on Sibunit) hydrogenation of 3,7-dimethyloct-6-en-1-yn-3-ol (dehy- drolinalool).178 The process occurs almost selectively (99.5%) as hydrogenation of the triple bond to a double bond; the catalyst activity in the heterogeneous process is maximal and is an order of magnitude higher than that of the traditional catalyst, namely, Pd supported on Sibunit.178 The tetralin extract of carbon was subjected to hydrocracking in petroleum ether with the addition of a dispersed C60/C70 mixture (90 : 10) and without this mixture.Fullerenes exerted no noticeable effect on the course of hydrocracking.179 The complex C60[Pd(OAc)2(PPh3)]3 was reduced with hydro- gen on heating to give a Pd ±C60 complex . Phenylacetylene, diphenylacetylene, cyclohexene and hex-1-ene were hydrogen- ated in the presence of 1 mol.% of this catalyst.180 Alkenes were completely reduced. The yields were comparable to those attained in the hydrogenation over Pd/C but the reaction occurred faster. The mechanism of the catalytic influence of the C60 substrate is unknown; however, it is assumed that metal fragments form active sites on the C60 spheroid.180 Photochemical oxidation of alkenes in the presence of silica gel modified with C60(H)CH2SiMe2(OPri) has been carried out.121 In order to prepare new film-forming materials including those promising as semiconducting films (for example, for the manufacture of diodes), polyaddition of benzene to fullerene in the presence of Kovacik catalyst (CuCl2+AlCl3) has been carried out. 181 This gave the compound C60Ph19H19 , which might have resulted from the reaction of protonated C60 with benzene.At 400 8C, this compound eliminates hydrogen to giveC60Ph19 . Both compounds form films from toluene solutions and C60Ph19 exhibits clear-cut luminescence in the blue-green region.The structure and catalytic properties of ruthenium on the C60 substrate (as well as on crude and extracted fullerene black and on the fullerene cathodic deposit) have been studied.182, 183 Highly dispersed amorphous Ru was prepared from Ru3(CO)2 by the impregnation/activation method. This system serves as an active catalyst for hydrogenation of CO and 2-cyclohexanone under atmospheric pressure. The molecular sieve MCM-41 is used to prepare catalysts with high surface areas for reactions with bulky molecules. The addition of C60 changes substantially the behaviour of the sieves during their dehydroxylation on heating, due to the abstraction of the hydroxy groups from the silanol fragments of the sieve to give C7H and C7OH groups followed by condensation of the C7OH groups or combination of C7OH with C7H.184 Precursors of fullerene-containing chelates of the type NMe2 R C60 Ni X NMe2 R=H, Me; X=H, Br, E N Karaulova, E I Bagrii have been synthesised;185 these compounds may find application as supported catalysts and in homogeneous catalysis. VII.Prospects for the practical use of fullerenes Extensive practical use of fullerenes would be largely determined by their availability. According to the data for 1995,13 fullerenes were supplied by the Hoechst AG (Germany), SES Research (USA) andMER(USA) companies. The price ofC60 fullerene was about $500 per kg in 1997.186 Back in 1995 the MER company developed a contact arc process which provided a daily output equal to 30 kg of carbon soot with a high fullerene content.Methanofullerene[60]-61-carboxylic acid is also commercially available (Fluka, Switzerland). According to the data of MER,186 attention is attracted first of all by the possibility of using fullerenes for the production of rechargeable galvanic cells, with electrodes made of C60/C70 films applied onto various substrates and for the storage of hydrogen, in particular, in the fuel cells of electric cars, and the storage of adsorbed oxygen (which may be useful in medicine and in military engineering). Fullerenes have been patented as precursors of diamonds and diamond coatings, production of which from fullerenes is energetically more favourable than that from graph- ite.The use of fullerenes for strengthening metal surfaces upon insertion of carbide particles into a metal matrix is also promising. Films of compounds containing fullerene fragments could be used in nonlinear optics, optical computerisation and in sensor tech- nology as solid sensors. Molecular donor ± acceptor charge transfer complexes can present considerable interest as the initial compounds for the manufacture of new materials,187 for example, superconduc- tors 43, 153, 188 (superconductivity transitions up to 40K have been observed).188. Nitroxides with a fullerene fragment and various fullereno- pyrrolidines can find use in the synthesis of organic ferromag- nets.75, 76 Films possessing valuable properties have been prepared based on fullerenopyrrolidines.73, 74 Methano- and pyrrolidino- fullerenes are efficient photosensitisers for the generation of singlet oxygen.102 The formation of C60 and C70 fullerene adducts with porphyrins of phthalocyanines increases the photoinitiated electrical conductivity of the latter (see Ref.53 and references therein). Nitrogen-containing fullerene derivatives immobilised on silica gel make it possible to develop new adsorbents.70 The potential use of fullerenes as biologically active com- pounds has stimulated vigorous development of the chemistry of functional derivatives of fullerene, especially after 1993, when it was shown 189 that water-soluble fullerene derivatives with meth- ane bridges inhibit HIV enzymes, namely, protease and reverse transcriptase; data on the biological properties of fullerene derivatives have been published.190, 191 For biological assays, it is necessary to prepare functional derivatives soluble in water or in polar media with a stabilising fragment linked to the fullerene by a covalent bond.Nakamura et al.191 have classified the biologically active fullerene derivatives into two types: the detergent type (in which the side chain is modified by, for example, attachment of carboxylic acid residues via organic binding groups) and the spherical type, in which there are polar groups distributed over the fullerene sphere (this class includes fullerenols and amino adducts). VIII. Conclusion In recent years, exohedral addition and, especially, cycloaddition has remained the main line of functionalisation of fullerenes. These reactions have been employed to prepare a large number of new oxygen- and nitrogen-containing derivatives such as qui- nones, pyridones, nitrile oxides and pyrrolidine derivatives.The photophysical and photochemical properties of these compounds (fluorescence, phosphorescence) and the redox potentials of these derivatives depend on the degree of functionalisation and posi-Fullerenes: functionalisation and prospects for the use of derivatives tions of functional groups in fullerene fragments. Fullerene derivatives containing heteroatoms can be used to model the photosensitisation and photosynthesis processes, including those involving singlet oxygen, which is important for the research into biological systems.An important aspect in the studies is develop- ment of the methods for selective multifunctionalisation, which would direct the addend to a particular site of the carbon cage. Among the fundamental studies devoted to the properties of fullerenes, the syntheses of a catenane with a fullerene fragment, the preparation of hydrocarbon salts and the generation of the carbocation consisting of only carbon atoms deserve attention. While Kroto, one of the fullerene discoverers, was amazed by the ability of pure carbon to be dissolved in organic solvents, the preparation of an aqueous colloidal solution of C60 seems even more remarkable. The potential practical application of fullerenes and their functional derivatives is being studied along several paths.In the case of C60 fullerene, this is mainly production of films deposited on various substrates including diamond-like ones, sensitisers and adsorbents for gases. The synthesis of new complexes and incarceranes (fullerenes containing an element incorporated in the cage) is aimed at the preparation of new superconductors, dielectrics and ferromagnets. Numerous publications have been devoted to the synthesis of C60 adducts in order to use them as the starting compounds in the synthesis of biologically active prod- ucts.The preparation of fullerene-based catalysts and catalytic systems and the synthesis of fullerene-containing polymers have been less developed.The progress along these lines might be expected in the future. References 1. H W Kroto, J R Heath, S C O'Brien, R F Curl, R E Smalley Nature (London) 318 162 (1985) 2. R F Curl Angew. Chem., Int. Ed. Engl. 36 1567 (1997) 3. H W Kroto Angew. Chem., Int. Ed. Engl. 36 1579 (1997) 4. R E Smalley Angew. Chem., Int. Ed. Engl. 36 1595 (1997) 5. E Osawa Kagaku (Kyoto) 25 854 (1970); Chem. Abstr. 74 75 698 (1971) 6. D A Bochvar, E G Gal'pern Dokl. Akad. Nauk SSSR 209 610 (1973) a 7. W KraÈ tschmer, L D Lamb, K Fostiropoulos, D R Huffman Nature (London) 347 354 (1990) 8. T Braun, E Osawa Full. Sci. Techn. 5 R3 (1997) 9. V I Sokolov, I V Stankevich Usp. Khim. 62 455 (1993) [Russ. Chem. Rev. 62 419 (1993)] 10. B L Tumanskii Izv. Akad.Nauk, Ser. Khim. 2396 (1996) b 11. N F Gol'dshleger, A P Moravskii Usp. Khim. 66 353 (1997) [Russ. Chem. Rev. 66 323 (1997)] 12. F Diederich, C Thilgen Science 271 317 (1996) 13. A Hirsch Synthesis 895 (1995) 14. J Mattay, G Torresgarsia, J Averdung, C Wolff, I Schlachter, H Luftmann, C Siedschlag, P Luger,M Ramm J. Phys. Chem. Solids 58 1929 (1997) 15. C Bellavialund, J C Hummelen, M Keshavarzk, R Gonzalez, F Wudl J. Phys. Chem. Solids 58 1983 (1997) 16. Y Rubin Chimia 52 118 (1998) 17. F Diederich, L Isaacs, D Philp Chem. Soc. Rev. 244 (1994) 18. A Hirsch Chemistry of Fullerenes (Stuttgart: Thieme Verlag, 1994) 19. R J Taylor J. Chem. Soc., Perkin Trans. 2 813 (1993) 20. A Hermann,M Ruttimann, C Thilgen, F Diederich Helv. Chim. Acta 78 1673 (1995) 21.A Hirsch, J Lamparth, G Schick Liebigs Ann. Chem. 1725 (1996) 22. F Wudl, R E Smalley, A B Smith, R Taylor, E Wasserman, E W Godly Pure Appl. Chem. 69 1412 (1997) 23. E W Godly, R Taylor Full. Sci. Techn. 5 1667 (1997) 24. P W Fowler, D E Manolopoulos Atlas of Fullerenes (ISBN 019 855 787) (Oxford: Clarendon Press, 1995) 25. S F Vinokurov, I N Novikov, A V Usatov Geokhimiya 937 (1997) c 26. D Heymann Astrophys. J. 489 L 111 (1997) 905 27. A G Ryabenko, A A Ryabenko, A P Moravskii, P V Fursikov Dokl. Akad. Nauk 351 215 (1996) a 28. A V Krestinin, A P Moravski Chem. Phys. Lett. 286 479 (1998) 29. H Richter, A J Labrocca, W J Grieco, K Taghizadeh, A L Lafleur, J B Howard J. Phys. Chem. B 101 1556 (1997) 30. B Pietzak, C Wolf, H J Mockel, A Weidinger Chem.Phys. Lett. 265 200 (1997) 31. T Alexakis, P G Tsantrizos,Y S Tsantrizos Appl. Phys. Lett. 70 2102 (1997) 32. T Yu Astakhova, G A Vinogradov,MMEl'yashevich, Sh A Shaginyan Khim. Fiz. 15 (10) 39 (1996) d 33. Yu E Lozovik, A M Popov Usp. Fiz. Nauk 167 751 (1997) e 34. M R Stetzer, P A Heiney, J E Fischer, A R McGhie Phys. Rev. B, Condens. Matter 55 127 (1997) 35. X-H Zhou, J-B Liu, Z-X Jin, Z-N Gu, Y-Q Wu, Y-L Sun Full. Sci. Techn. 5 285 (1997) 36. L L Gumanov, B L Korsunskii, V P Bubnov, E B Yagubskii, L S Pomogaeva Izv. Akad. Nauk, Ser. Khim. 1544 (1997) b 37. M Kele, R N Compton, G Guiochon J. Chromat. A 786 31 (1997) 38. L L Gumanov, B L Korsounskii Mendeleev Commun. 158 (1997) 39. D V Konarev, R N Lyubovskaya Usp.Khim. 68 23 (1999) [Russ. Chem. Rev. 68 19 (1999)] 40. M T Bek, G Mandi, Sh Keki Izv. Akad. Nauk, Ser. Khim. 2129 (1996) b 41. D V Konarev, R N Lyubovskaya, V N Semkin, A Graja Pol. J. Chem. 71 96 (1997) 42. S V Konovalikhin, O A D'yachenko, G V Shilov, N G Spitsyna, K V Van, E B Yagubskii Izv. Akad. Nauk, Ser. Khim. 1480 (1997) b 43. A Otsuka, G Saito, A A Zakhidov, K Yakushi Synth. Met. 85 1459 (1997) 44. Y Iwasa, H Hayashi, T Furudate, M Kawaguchi, T Mitani Synth. Met. 86 2309 (1997) 45. H Shinohara Sci. Rep. Res. Inst. Tohoku University A44 47 (1997) 46. T Akasaka, S Nagase, K Kobayashi,M Walchei, K Yamamoto, H Funasaka, M Kako, T Hoshino, T Erata Angew. Chem., Int. Ed. Engl. 36 1643 (1997) 47. T Pichler, M S Golden, M Knupfer, J Fink, U Kirbach, P Kuran, L Dunsch Phys. Rev.Lett. 79 3026 (1997) 48. Y Takabayashi, Y Kubozono, K Hiraoka, T Inoue, K Mimura, H Maeda, S Kashino Chem. Lett. 1019 (1997) 49. B-P Cao, Z-Y Shi, X-H Zhou, Z-N Gu, H-Z Xiao, J-Z Wang Chem. Lett. 937 (1997) 50. V P Bubnov, V K Koltover, E E Laukhina, Ya I Estrin, E B Yagubskii, in International Workshop in Russia `Fullerene and Atomic Clusters' IWFAC'97 (Abstracts of Reports), St. Petersburg, 1997 p. 130 51. V K Koltover, V P Bubnov, Ya I Estrin, L T Kasumova, E E Laukhina, E B Yagubskii, in International Workshop in Russia `Fullerene and Atomic Clusters' IWFAC'97 (Abstracts of Reports), St. Petersburg, 1997 p. 129 52. V K Kol'tover, E E Laukhina, Ya I Estrin, V P Bubnov, E B Yagubskii Dokl.Akad. Nauk 353 57 (1997) a 53. T Nojiri, M M Alam, H Konami, A Watanabe, O Ito J. Phys. Chem. A 101 7943 (1997) 54. F Beer, A Gugel, K Martin, J Rader, K Mullen J. Mater. Chem. 1327 (1997) 55. A C Tome, R F Enes, J A S Cavaliro, J Elguero Tetrahedron Lett. 38 2557 (1997) 56. U M Fernandez-Paniagua, B Illescas, N Martin, C Seoane, P de la Cruz, A de la Hoz, F Langa J. Org. Chem. 62 3705 (1997) 57. P de la Cruz,Ade la Hoz, F Langa, B Illeskas,N Martin Tetrahedron 53 2599 (1997) 58. B M Illescas, N Martin, C Seoane, E Orti, P M Viruela, R Viruela, A de la Hoz J. Org. Chem. 62 7585 (1997) 59. K DuÈ rr, S Fiedler, T Linssen, A Hirsch, M Hanack Chem. Ber. 130 1375 (1997) 60. A Puplovskis, J Kacens, O Neilands Tetrahedron Lett.38 285 (1997) 61. G Torres-Garcia, H Luftmann, C Wolff, J Mattay J. Org. Chem. 62 2752 (1997) 62. M Ohno, S Kojima, Y Shirakawa, S Eguchi Tetrahedron Lett. 37 9211 (1996) 63. A Mori, Y Takamori, H Takeshita Chem. Lett. 395 (1997) 64. T Kawai, S Scheib, M P Cava, R M Metzger Langmuir 13 5627 (1997)906 65. R I Khusnutdinov, Kh Kh Gainullin, Z S Muslimov, U M Dzhemilev, O M Nefedov, in 16-i Mendeleevskii S'ezd po Obshchei i Prikladnoi Khimii (Tez. Dokl.), Moskva, 1998 [The 16th Mendeleev Congress on General and Applied Chemistry (Abstracts of Reports), Moscow, 1998] Vol. 1, p. 324 66. M Ohno, S Kojima, Y Shirakawa, S Eguchi Tetrahedron Lett. 36 6899 (1995) 67. H-P Zeng, S Eguchi Synlett 175 (1997) 68. J Llacay, M Mas, E Molins, J Veciana, D Povell, C Rovira J.Chem. Soc., Chem. Commun. 659 (1997) 69. K Matsumoto,M Ciobanu, H Katsura, R Ohta, T Uchida Heterocycl. Commun. 2 545 (1996) 70. A Bianco, F Gasparrini,M Maggini,D Misiti, A Polese,M Prato, G Scorrano, C Toniolo, C Villani J. Am. Chem. Soc. 119 7550 (1997) 71. F Novello, M Prato, T Da Ros,M De Amici, A Bianco, C Toniolo,M Maggini J. Chem. Soc., Chem. Commun. 903 (1996) 72. Y Li, D Zheng, J Xu, Z Mao, Y K Yang, F L Bai, D B Zhu Chin. Sci. Bull. 42 1180 (1997) 73. D Zhou, L Gan, C Luo, C Huang, Y Wu Solid State Commun. 102 891 (1997) 74. D Zhou,G J Ashwell, R Rajan, L Gan, C Luo, C Huang J. Chem. Soc., Faraday Trans. 93 2077 (1997) 75. Y-L Li, Z Mao, J-H Xu, Y-K Yang, Z-X Guo, D-B Zhu, J-W Li, B Yin Chem. Phys.Lett. 361 (1997) 76. Y-L Li, J-H Xu, D-G Zheng, J-K Yang, C-Y Pan, D-B Zhu Solid State Commun. 101 123 (1997) 77. D-J Zhou, L-B Gan, C-P Luo, X-H Huang, C-Q Yao Chines J. Chem. 15 35 (1997); Chem. Abstr. 126 269 885 (1997) 78. M Iyoda, F Sultana, A Kato,M Yoshida, Y Kuwatani, M Komatsu, S Nagase Chem. Lett. 63 (1997) 79. T Da Ros,M Prato, F Novello,M Maggiki, E Banfi J. Org. Chem. 61 9070 (1996) 80. H Irngartinger, A Weber, T Escher Liebigs Ann. Chem. 1845 (1996) 81. H Irngartinger, A Weber Tetrahedron Lett. 38 2075 (1997) 82. T Da Ros,M Prato, F Novello, M Maggini, M De Amici, C De Micheli J. Chem. Soc., Chem. Commun. 59 (1997) 83. V N Drozd, V N Knyazev, F M Stoyanovich, F M Dolgushin, A I Yanovskii Izv. Akad.Nauk Ser. Khim. 118 (1997) b 84. V N Knyazev, N M Przheval'skii, in 16-i Mendeleevskii S'ezd po Obshchei i Prikladnoi Khimii (Tez. Dokl.), Moskva, 1998 [The 16th Mendeleev Congress on General and Applied Chemistry (Abstracts of Reports), Moscow, 1998] Vol. 1, p. 141 85. L-H Shu, W-Q Sun, D-W Zhang, S-H Wu, H-M Wu, J-F Xu, X-F Lao J. Chem. Soc., Chem. Commun. 79 (1997) 86. B F O'Donovan, P B Hitchcock,M F Meidine, H W Kroto, R Taylor,D R M Walton J. Chem. Soc., Chem. Commun. 81 (1997) 87. A A Ovcharenko, V A Chertkov, A V Karchava, M A Yurovskaya Tetrahedron Lett. 38 6933 (1997) 88. G Wassilikogiannakis,M Orfanopoulos Tetrahedron Lett. 38 4323 (1997) 89. G Wassilikogiannakis,M Orfanopoulos J. Am. Chem. Soc. 119 7394 (1997) 90. A W Jensen, A Khong,M Saunders, S R Wilson, D I Schuster J.Am. Chem. Soc. 119 7303 (1997) 91. N F Gol'dshleger, N N Denisov, A S Lobach, V A Nadtochenko, Yu M Shul'ga, V N Vasilets Dokl. Akad. Nauk 340 630 (1995) a 92. C Bingel Chem. Ber. 126 1957 (1993) 93. X Camps, A Hirsch J. Chem. Soc., Perkin Trans. 1 1595 (1997) 94. J E Nierengarten, J F Nicoud Tetrahedron Lett. 38 7737 (1997) 95. C-C Zhu, Y Xu, Y-Q Liu, D-B Zhu J. Org. Chem. 62 1996 (1997) 96. Y K Agrawal Full. Sci. Techn. 5 275 (1997) 97. J Lamparth, G Schik, A Hirsch Liebigs Ann. Chem. 253 (1997) 98. I G Safonov, P S Baran, D I Schuster Tetrahedron Lett. 38 8133 (1997) 99. P S Baran, R R Monaco, A U Khan, D I Schuster, S R Wilson J. Am. Chem. Soc. 119 8363 (1997) 100. H Duczek, W Radeck, H J Niclas,M Ramm, B Costisella Tetrahedron Lett.38 6651 (1997) 101. R Pellicciari, D Annibali,G Costantino,M Marinozzi, B Natalani Synlett 1196 (1997) 102. B Ma, C E Bunker, R Guduru, X F Zhang, Y P Sun J. Phys. Chem. A 101 5626 (1997) E N Karaulova, E I Bagrii 104. J Xu, Y Li, D-G Zheng, J-K Yang, Z Mao, D-B Zhu Tetrahedron 103. H Yamaguchi, S Murata, T Akasaka, T Suzuki Tetrahedron Lett. 38 3529 (1997) Lett. 38 6613 (1997) 105. C-K F Shen, K-M Chien, C-G Juo, G-R Her, T-Y Luh J. Org. Chem. 61 9242 (1996) 109. H Isobe, H Tokuyama, M Sawamura, E Nakamura J. Org. Chem. 106. A Ikeda, C Fukuhara, S Shinkai Chem. Lett. 407 (1997) 107. M Taki, S Sugita, Y Nakamura, E Kasashima, E Yashima, Y Okamoto, J Nishimura J. Am. Chem.Soc. 119 926 (1997) 108. B Knight, N Martin, T Ohno, E Orti, C Roviro, J Veciana, J Vidal-Gancedo, P Virnela, R Viruela, F Wudl J. Am. Chem. Soc. 119 9871 (1997) 62 5034 (1997) 110. R F Haldimann, F G Klirner, F Diederich J. Chem. Soc., Chem. Commun. 237 (1997) 111. Y Murata,M Shiro,K Komatsu J. Am. Chem. Soc. 119 8117 (1997) 112. P R Ashton, F Diederich, M Gdmez-Ldpez, J F Nierengarten, J A Preece, F M Raymo, J F Stoddart Angew. Chem., Int. Ed. Engl. 36 1448 (1997) 113. M E Vol'pin, Z N Parnes, V S Romanova Izv. Akad. Nauk, Ser. Khim. 1050 (1998) b 114. E M Belavtseva, E V Kichenko, V S Romanova, Z N Parnes, M E Vol'pin Izv. Akad. Nauk, Ser. Khim. 876(1996) b 115. R A Kotelnikova, A I Kotelnikov, G N Bogdanov, V S Romanova, E F Kuleshova, Z N Parnes, M E Vol'pin FEBS Lett.389 111 (1996); Chem. Abstr. 125 135 709 (1996) 116. G N Bogdanov, R A Kotelnikova, V S Romanova, Z N Parnes, in International Workshop in Russia `Fullerene and Atomic Clusters' IWFAC'97 (Abstracts of Reports), St. Petersburg, 1997 p. 47 117. G Khairallah, J B Peel J. Chem. Soc., Chem. Commun. 253 (1997) 118. G W Wang, K Komatsu, Y Murata, M Shiro Nature (London) 387 583 (1997) 119. T Tanaka, T Kitagawa, K Komatsu, K Takeuchi J. Am. Chem. Soc. 119 9313 (1997) 120. H Okamura, Y Murata, M Mineda, K Komatsu, T Miyamoto, T S Wan J. Org. Chem. 61 8500 (1996) 121. H Nagashima, K Jinno, K Itoh Nippon Kagaku Kaishi (2) 91 (1997); Chem. Abstr. 126 225 334 (1997) 122. K Mikami, S Matsumoto, T Tonoi, T Suenobi, A Ishida, S Fukuzumi Synlett 85 (1997) 123.C-K F Shen, H-H Yu, C-G Juo, K-M Chien, G-R Her, T-Y Luh Chem. A-Eur. J. 3 744 (1997) 124. A S Lobach, A A Perov, A I Rebrov, O S Roshchupkina, V A Tkacheva, A N Stepanov Izv. Akad. Nauk, Ser. Khim. 671 127. L B Alemany, A Gonzalez, W E Billups, M R Willeott, E Ezell, (1997) b 125. R G Bergosh, M S Meier, J A L Cooke, H P Spielmann, B R Weedon J. Org. Chem. 62 7667 (1997) 126. R V Bensasson, T J Hill, E J Land, S Leach, D J Mcgarvey, T G Truscott, J Ebenhoch, M Gerst, C Ruchardt J. Chem. Phys. 215 111 (1997) E Gozansky J. Org. Chem. 62 5771 (1997) 128. P W Fowler, J P B Sandall, R Taylor J. Chem. Soc., Perkin Trans. 2 419 (1997) 129. B P Tarasov, V N Fokin, A P Moravskii, Yu M Shul'ga Zh.Neorg. Khim. 42 920 (1997) f 130. B P Tarasov, V N Fokin, A P Moravskii, Yu M Shul'ga Izv. Akad. Nauk, Ser. Khim. 1867 (1996) b 131. D Mandrus, M Kele, R L Hettlich, G Guiochon, B C Sales, L A Boatner J. Phys. Chem. B 123 (1997) 132. N F Gol'dshleger, A S Astakhova, A S Lobach, G N Boiko, N N Denisov, V A Nadtochenko, O S Roshchupkina, 133. C Siedeschlag, H Luftmann, C Wolf, J Mattay Tetrahedron 53 135. F Banim, D J Cardin, P Heath J. Chem. Soc., Chem. Commun. 25 Yu M Shul'ga Izv. Akad. Nauk, Ser. Khim. 2531 (1996) b 3587 (1997) 134. S Fukuzumi, T Suenobu,M Patz, T Hirasaka, S Itoh, M Fujitsuka, O Ito J. Am. Chem. Soc. 120 8060 (1998) (1997) 136. T Hamano, T Mashino,M Hirobe J. Chem. Soc., Chem. Commun. 1537 (1995) 137. R W Murray, K Iyanar Tetrahedron Lett.38 335 (1997)Fullerenes: functionalisation and prospects for the use of derivatives 138. V N Bezmelnitsin, A V Eletskii, N G Schepetov, A G Avent, R Taylor J. Chem. Soc., Perkin Trans. 2 683 (1997) 139. S Lebedkin, S Ballenweg, J Gross, R Taylor, W Kritschmer Tetrahedron Lett. 36 4971 (1995) 140. A Gromov, S Lebedkin, S Ballenweg, A G Avent, R Taylor, W KraÈ tschmer J. Chem. Soc., Chem. Commun. 209 (1997) 141. A Gromov, S Ballenweg, S Giesa, S Lebedkin, W E Hull, W Kritschmer Chem. Phys. Lett. 267 460 (1997) 142. X Wei,M Wu, L Qi, Z Xu J. Chem. Soc., Perkin Trans. 2 1389 (1997) 143. K Winkler, D A Costa,W R Fawcett, A L Balch Adv. Matter 9 153 (1997); Chem. Abstr. 126 244 035 (1997) 144. M Fedurco, D A Costa, A L Balch, W R Fawcett Angew.Chem., Int. Ed. Engl. 34 194 (1995) 145. A G Avent, P R Birkett, A D Darwish, H W Kroto, R Taylor, D R M Walton J. Chem. Soc., Chem. Commun. 1579 (1997) 146. Y K Zhang, E G Janzen, Y Kotake J. Chem. Soc., Perkin Trans. 2 1191 (1996) 147. D Sun, Y Zhu, Z Liu, G-Z Liu, X-H Guo, R-Y Zhan, X-Y Liu Chin. Sci. Bull. 42 748 (1997) 148. R D Bolskar, R S Mathur, C A Reed J. Am. Chem. Soc. 118 13093 (1996) 149. F Cataldo Full. Sci. Techn. 5 257 (1997) 150. O V Boltalina, A K Abdul-Sada, R Taylor J. Chem. Soc., Perkin Trans. 2 981 (1995) 151. V F Bagryantsev, A S Zapol'skii, O V Boltalina, N A Galeva, L N Sidorov Dokl. Akad. Nauk 357 487 (1997) a 152. N Liu, Y Morio, F Okino, H Touhara, O V Boltalina, V V Pavlovich Synth. Met. 86 2289 (1997) 153. A Hamwi, C Latouche, B Burteau, J Dupuis Full. Sci. Techn. 4 1213 (1996) 154. M Yoshida, D Suzuki, M Iyoda Chem. Lett. 1097 (1996) 155. B Renker, H Schober, R Heid, P von Stein Solid State Commun. 104 527 (1997) 156. M Premila, C S Sundar, P C Sahu, A Bharathi, Y Hariharan, D V S Muthu, A K Sood Solid State Commun. 104 237 (1997) 157. A A Shvartsburg, R R Hudgins, P Dugourd,M F Jarrold J. Phys. Chem. A 101 1684 (1997) 158. F Tast, N Malinowski, I M L Billas, M Heinebrodt, W Branz, T P Martin Theor. J. Chem. Phys. 107 6980 (1997) 159. D Stewart, C T Imrie J. Chem. Soc., Chem. Commun. 1383 (1996) 160. T Cao, S E Webber Macromolecules 28 3741 (1995) 161. Y Kojima, T Matsuoka,H Takahasi, T Kurauchi J. Mat. Sci. Lett. 16 999 (1997) 162. Y Chen, Z Huang, R Cai, S-Q Kong, S Chen, Q Shao, X Yan, F Zhao, D Fu J. Polym. Sci. Part A, Polym. Chem. 34 3297 (1996) 163. Y Chen, Z-E Huang, R-F Cai, B-C Yu,W-W MaEur. Polym. J. 33 291 (1997); Chem. Abstr. 126 34 971 (1997) 164. S Mehrotra, A Nigam, R Malhotra J. Chem. Soc., Chem. Commun. 463 (1997) 165. K Kirkwood, D Stewart, C T Imrie J. Polym. Sci. Part A, Polym. Chem. 35 3323 (1997) 166. P L Nayak, S Alva, K Yang, P K Dhal, J Kumar, S K Tripathy Macromolecules 30 7351 (1997) 167. W T Ford, T D Graham, T H Mourey Macromolecules 30 6422 (1997) 168. Y-P Sun, B Liu, D K Moton J. Chem. Soc., Chem. Commun. 2699 (1996) 169. J Li, T Yoshizawa,M Ikuta,M Ozawa, K Nakahara, T Hasegawa, K Kitazawa, M Hayashi, K Kinbara, M Nohara, 170. V Baranov, J Wang, G Javahery, S Petrie, A C Hopkinson, 171. A B Raksit,D K Bohme Int. J. Mass Spectrom. Ion Processes 55 69 172. L-Y Chiang, L-Y Wang, C-S Kuo, J-G Lin, C-Y Huang Synth. 173. A Yassar,M Hmiene, D C Loveday, J P Ferraris Synth. Met. 84 174. T Benincori, E Brenna, F Sannicolo, L Trimarco, G Zotti, 175. H L Wang,M Grigorova, E S Maniloff, D W McBranch, K Saigo Chem. Lett. 1037 (1997) D K Bohme J. Am. Chem. Soc. 119 2040 (1997) (1983) Met. 84 721 (1997) 231 (1997) P Sozzani Angew. Chem., Int. Ed. Engl. 35 648 (1996) B R Mattes Synth. Met. 84 781 (1997) 907 176. S B Lee, A A Zakhidov, V Yu Sokolov, P K Khabibullaev, K Tada, T Kawai, K Yoshino Synth. Met. 84 567 (1997) 177. E Maniloff, D Vacar, D McBranch, H L Wang, B Mattes, A J Heeger Synth. Met. 84 547 (1997) 178. E M Sul'man, V G Matveeva, V V Bashilov, V I Sokolov Kinet. Katal. 38 274 (1997) g 179. S-F Zhang, J R Shearman,M Domin, M-J Lazaro, A A Herod, R Kandiyoti Fuel 76 207 (1997); Chem. Abstr. 126 240 564 (1997) 180. R Yu, Q Liu, K Tan, G-Q Xu, S-C Ng, H S O Chan, T S A Hor J. Chem. Soc., Faraday Trans. 93 2207 (1997) 181. Ya L Kogan, M Yu Belov, L S Fokeeva, V R Fogel', G I Davydova, Yu A Volkov Izv. Akad. Nauk, Ser. Khim. 1219 (1997) b 182. T Braun,M Wohlers, T Belz, G Nowitzke, G Wortmann, Y Uchida, N PfaÈ nder, R SchloÈ gl Catal. Lett. 43 167 (1997) 183. T Braun,M Wohlers, T Belz, R SchloÈ gl Catal. Lett. 43 175 (1997) 184. J Chen, Q Li, H Ding, W Pang, R Xu Langmuir 13 2050 (1997) 185. M D Meijer,M Rump, R A Gossage, J H T B Jastrzebski, G van Koten Tetrahedron Lett. 39 6773 (1998) 186. J C Withers, R O Loutfy, T P Lowe Full. Sci. Techn. 5 1 (1997) 187. E Sobczak, A Travers, R Nietubyc, Y Swilem, P Byszewski, D Zymierska Acta Phys. Pol. A 91 877 (1997) 188. W K Fullagar, P A Reynolds, J W White Solid State Commun. 104 23 (1997) 189. S H Friedman, D L DeCamp, R P Sijbesma, G Srdanov, F Wudl, G L Kenyon J. Am. Chem. Soc. 115 6506 (1993) 190. A W Jensen, S R Wilson, D I Schuster Bioorg. Med. Chem. 4 767 (1996) 191. E Nakamura, H Tokuyama, S Yamago, T Schiraki, Y Sugiura Bull. Chem. Soc. Jpn. 69 2143 (1996) a�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Transl.) c�Geochem. Int. (Engl. Transl.) d�Russ. J. Chem. Phys. (Engl. Transl.) e�Physics-Uspekhi (Engl. Transl.) f�Russ. J. Inorg. Chem. (Engl. Transl.) g�Kinet. Catal. (

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (11) 909 ± 923 (1999) Stable tellurols and their metal derivatives I D Sadekov, A V Zakharov Contents I. Introduction II. Synthesis, reactions and structures of tellurols III. Synthesis, reactions and structures of metal tellurolates IV. Conclusion Abstract. Data on the synthesis, reactions and structures of stable aromatic tellurols and sterically hindered tellurols containing E7TeH bonds (E=C, Si or Ge) are systematised and general- ised. Methods for the preparation of various metal derivatives of tellurols as well as applications of these derivatives in the low- temperature synthesis of metal tellurides are considered. The bibliography includes 90 references. I. Introduction Tellurium analogues of alcohols and phenols, viz., tellurols (RTeH), belong to a class of organotellurium compounds which are rather poorly studied.First representatives of these com- pounds, viz., alkanetellurols, were synthesised back in 1926,1 while rather stable aromatic tellurols, viz., 2,4,6-trialkylbenzene- tellurols and tris(trimethylsilyl)silanetellurols, were prepared only in 1991.2 Based on these compounds and their analogues and derivatives, a wide range of tellurolates of various metals have been synthesised in recent years. Tellurolates of sterically hindered organoelement tellurols (Me3Si)3ETeH (E=C, Si or Ge) are of particular interest because these compounds can be used in the pyrolytic synthesis of metal tellurides. Since these compounds contain E7Te7H bonds (E=Si or Ge), they do not formally belong to tellurols.However, it is reasonable to consider here methods of their preparation, reactions and spectral character- istics, because these derivatives possess unusual properties, in particular, enhanced stability in the crystalline state and in solutions compared to true tellurols. In addition, tellurolates based on the above-mentioned compounds allow one to prepare metal tellurides at relatively low temperatures.3±7 II. Synthesis, reactions and structures of tellurols 1. Synthesis and reactions of tellurols The first alkanetellurol, viz., ethanetellurol 1a, was prepared by high-temperature alcoholysis of Al2Te3 at 250 8C.1 Later on, this method was modified by Baroni.8 Oxidation of the resulting alkanetellurols can be prevented if the process is performed I D Sadekov, A V Zakharov Research Institute of Physical and Organic Chemistry of the Rostov State University, prosp.Stachki 194/2, 344090 Rostov-on-Don, Russian Federation. Fax (7-863) 228 56 67. Tel. (7-863) 228 08 94. E-mail: sadek@ipoc.rnd.runnet.ru Received 17 June 1999 Uspekhi Khimii 68 (11) 999 ± 1014 (1999); translated by T N Safonova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.269.9+547.286.4 909 909 911 922 under an atmosphere of hydrogen. Apparently, the overall reaction is described by the following equation: D Al2Te3+3ROH 3RTeH +Al2O3 1a ± d R=Et (a), Me (b), Pr (c), Bu (d). Alkanetellurols 1c,d were also prepared by the reactions of the corresponding alkyl halides with a solution of sodium ethoxide saturated with hydrogen telluride at low temperature.8 However, attempts at synthesising methane- and ethanetellurols 1a,b according to this procedure have failed.RBr EtOH +NaHTe EtONa+H2Te 7NaBr RTeH 1c,d R=Pr (c), Bu (d). The method which is most generally used for the preparation of different types of tellurols involves the reactions of their lithium and sodium salts (methods for their synthesis are discussed in Section III.1) with protic acids. Thus treatment of sodium methanetellurolate, which has been prepared from dimethyl ditelluride upon cleavage of the Te7Te bond with sodium in liquid ammonia, afforded methanetellurol 1b.9 ±11 Trideuteriome- thanetellurol 1e was prepared analogously.11 HX RTeNa R2Te2+Na 7NaX RTeH 1b,e R=Me (b), CD3 (e); HX=H2SO4, H3PO4.The acidH3PO4 is a protonating agent of choice 11 because the use of even diluteH2SO4 leads to partial oxidation of tellurol 1b to dimethyl ditelluride. Alkanetellurols 1 are liquids with an unpleasant odour. The yellowish colour of these compounds is caused by the presence of dialkyl ditellurides, which were formed either upon thermal decomposition of tellurols or upon their oxidation with atmos- pheric oxygen.9± 11 Tellurols can be stored at the temperature of liquid nitrogen with exclusion of air because they readily decom- pose at room temperature.11 Decomposition of methanetellurol under UV irradiation occurs in two stages.First, dimethyl ditelluride and hydrogen are formed and then more profound decomposition occurs to form methane, higher hydrocarbons and tellurium.11 Attempts to synthesise benzene- and thiophene-2-tellurols by protonation of their lithium salts with trifluoroacetic acid in THF or of their sodium salts in an ethanol ± ethyl acetate mixture have been unsuccessful.12 Attempts to prepare benzenetellurol upon910 methanolysis of (phenyltelluro)trimethylsilane have also failed.13 ± 15 However, benzenetellurol12, 14, 15 and thiophene-2- tellurol,12 which are prepared according to the above-described methods in situ, are efficient reducing agents for organic com- pounds.12, 15 In the presence of ZnI2, they are used for converting carbonyl compounds into asymmetrical ethers.14 Arenetellurols were synthesised by protonolysis of suspen- sions of sterically hindered lithium arenetellurolates 2 (prepared by reducing the corresponding diaryl ditellurides with lithium triethylborohydride) with HBF4 etherate in light petroleum.Compounds 3 were obtained in nearly quantitative yields.2 R R R TeLi R TeH HBF4 . Et2O 778 8C 3a ± c R 2a ± c R R=Me (a), Pri(b), But (c). Arenetellurols 3 are colourless crystalline compounds, which can be recrystallised from light petroleum at7110 8C. However, these compounds are light-sensitive and begin to decompose at 740 to730 8C to form metallic tellurium.2 The thermal stability of arenetellurols 3 is independent of the bulkiness of the sub- stituents.In particular, tellurol 3a is more stable than 3c.2 The most stable tellurols are organoelement tellurols (Me3Si)3ETeH (E=C, Si or Ge). Due to the presence of bulky substituents, metal tellurolates prepared from these tellurols are readily soluble in nonpolar organic solvents and exist as mono- mers or weakly associated aggregates. These properties together with the volatility and high thermal stability allow one to use salts of organoelement tellurols in the pyrolytic synthesis of metal tellurides (see Section III). The first tellurol containing the Si7TeH bond to be synthes- ised was triethylsilanetellurol 4a.16 This compound was prepared by treatment of bis(triethylsilyl) telluride with an equimolar amount of trifluoroacetic acid at 20 8C and was isolated by vacuum distillation. CF3CO2H CF3CO2H (Et3Si)2Te 7CF3CO2SiEt3 7CF3CO2SiEt3 Et3SiTeH 4a (>70%) O2 Te H2Te 7H2O Under the action of an excess of CF3CO2H, the Si7Te bond in the tellurol 4a was cleaved and triethylsilyl trifluoroacetate and elemental tellurium were formed as the final reaction products.16 Apparently, the latter was formed upon oxidation of intermediate H2Te.Sterically hindered tris(trimethylsilyl)silanetellurol (4b) was synthesised by protonolysis of lithium salts 5a 17 ± 19 or 5b 20 with trifluoroacetic acid in hexane 17 ± 19 or with HCl in toluene.20 The salts 5a 21 and 5b 22 were prepared according to a standard procedure for the synthesis of these salts, which involves the insertion of the Te atom at the Si7Li bond in the complexes of tris(trimethylsilyl)silyllithium with THF (6a )21 or with dime- thoxyethane (DME) (6b).22 HX 7LiX (Me3Si)3SiTeH 4b (90%) LnLiSi(SiMe3)3+Te 6a,b LmLiTeSi(SiMe3)3 5a,b L=THF: n=3 (6a), m=2 (5a); L=DME: n=1.5 (6b), m=1 (5b); HX=HCl, CF3CO2H.Colourless crystalline tellurol 4b is stable under an argon atmosphere in the dark for at least several days. In sunlight, solid tellurol 4b turns dark but the spectral characteristics remain unchanged. In solutions, the tellurol 4b rapidly decomposes under exposure to light to form a tellurium mirror.17, 18 I D Sadekov, A V Zakharov Protonolysis of lithium tris(trimethylsilyl)methanetellurolate (5c),23 tris(trimethylsilyl)germanetellurolate (5d) 19 and tris(triphenylsilyl)silanetellurolate (5e),24 which have been pre- liminarily isolated and purified, with trifluoromethanesulfonic acid in ether 17 ± 19 or with trifluoroacetic acid in hexane 24 afforded tellurols 4c ± e in 57%± 81% yields.Protonolysis of lithium salts 5e in diethyl ether, which is more polar than hexane, was accompanied by partial decomposition of the tellurol 4e resulting also in bis(triphenylsilyl) telluride and H2Te.24 HX HTeM(SiR3)3 4c ± e LiTeM(SiR3)3 . (THF)n 5c ± e 7LiX, 7THF M=C, R=Me, n=2 (5c); M=Ge, R=Me, n=3 (5d);M=Si, R=Ph, n=3 (5e). Tellurols 4c ± e are crystalline compounds the colour of which varies from yellow to orange. These compounds melt at temper- atures above 120 8C.Compounds 4c,d, like tellurol 4b, were isolated and purified by sublimation under high vacuum, while tellurol 4e was purified by recrystallisation from hexane at low temperature. The stability of tellurols 4c,d is similar to that of tellurol 4b,19 whereas tellurol 4e is more sensitive to atmospheric oxygen.24 The reactions of tellurols governed by the presence of the hydrotelluride function are poorly studied. Oxidation of the tellurol 4b with oxygen afforded ditelluride 7. Compound 7 can be more easily synthesised by the oxidation of the tellurolate 5a with oxygen or CuCl 17, 18 or by the oxidation of the tellurolate 5b with iodine, because these procedures exclude the stages of isolation and purification of the tellurol.20 O2 7H2O [(Me3Si)3Si]2Te2 7 (Me3Si)3SiTeH 4b The pKa values of tellurols 4b ± d and of a number of related thiols and selenols were determined (photometrically, for solu- tions in isobutyl methyl ketone).19 The following values were obtained: 7.3 for HTeSi(SiMe3)3 (4b), 9.3 for HTeC(SiMe3)3 (4c), 7.4 for HTeGe(SiMe3)3 (4d), 8.3 for HSe(SiMe3)3, 10.7 for HSSi(SiMe3)3 (the synthesis was described in Ref.25) and 10.8 for HSeC(SiMe3)3 (the synthesis was described in Ref. 26). The acid strengths of silane- and germanetellurols 4b,d are approximately 100 times higher than that of their carbon analogue 4c. This is also true for the corresponding selenium derivatives. This tendency for the change in the pKa values was attributed 19 to the lower electronegativity of Si (1.9) and Ge (2.0) compared to C (2.5).For compounds HM(SiMe3)3 (M=S, Se or Te), the pKa values increase in the series Te>Se>S.19 Thus the tellurol 4b is a stronger acid than the corresponding selenol and thiol by a factor of 10 and 1000, respectively. The same tendency for the change in the acidity is typical of inorganic hydrides of the corresponding elements.27 2. X-Ray diffraction analysis and 1Hand 125TeNMR spectra of tellurols The molecular and crystal structure of tris(triphenylsilyl)silane- tellurol 4e was established by X-ray diffraction analysis.24 The Te7Si bond length is 2.511 Aand the Te7Si7C angles are in the range of 106.0 ± 109.6 8. Unlike Ph3SiOH, which exists in the crystal as an H-bound tetramer,28 the tellurol 4e exists as a monomer (like triphenylsilanethiol Ph3SiSH 29).In the crystal, there are no shortened or specific intermolecular contacts smaller than the sum of the van der Walls radii of the elements. Thus the Te . . . Te distances are 4.64 A(the sum of the van der Waals radii is 4.40 A).30 Evidently, hydrogen bonds in the crystal of the tellurol 4e are absent due to the low electronegativity of the tellurium atom. The 1H and 125Te NMR spectroscopic data are given in Table 1.Stable tellurols and their metal derivatives Table 1. Data of 1H and 125Te NMR spectroscopy of tellurols 2 (CDCl3, 720 8C), tellurols 4 (PhH, 20 8C) and their analogues. 1 Compound H NMR 125Te NMR JTe7H/ Ref. d d Hz (seea) (seeb) 63 48.6 51.9 2,4,6-Me3C6H2TeH (2a) 2,4,6-Pri3C6H2TeH (2b) 2,4,6-But3C6H2TeH (2c) (Me3Si)3SiTeH (4b) 791.4 7134.8 153.8 7955 7955.1 763 7940 7684 77 72.68 72.95 71.25 78.82 78.83 75.1 79.02 76.45 77.46 77.44 74 74.6 101 95 27.4 58 52 (Me3Si)3CTeH (4c) (Me3Si)3GeTeH Ph3SiTeH H3SiTeH c H3GeTeH c 222 17 ± 19 20 19 19 24 31 31 a The chemical shifts of the protons bound to the Te atoms are given; b the chemical shifts relative toMe2Te; c the spectra were recorded in CDCl3 at 20 8C.31 The signals for the protons bound to the Te atoms in the tellurols 4b ± e containing the trimethylsilyl group are observed at substantially higher field (at d from 76.45 to 78.82) than the analogous signals in the 1HNMRspectra of the aromatic tellurols 2a ± c (at d from 71.25 to72.95).A comparison of the chemical shifts of the protons bound to the chalcogen atoms in the 1HNMR spectra of tris(trimethylsilyl)silanechalcogenols (Me3Si)3EH (E=S, Se or Te) are indicative of successive downfield shifts of these signals in the series Te<Se<S.19 The resonance signals for the 125Te nuclei are observed as doublets in a very broad region (more than 1100 ppm), viz., from d 153.8 for the aromatic tellurol 2c (see Ref. 2) to d 7955 for tris(trimethylsilyl)silanetellurol 4b.17 ± 20 A substantial upfield shift for the compound 4b was related to the lower electro- negativity of the silyl substituents compared to that of aromatic substituents.18 III. Synthesis, reactions and structures of metal tellurolates Until recently, alkane- or arenetellurolates of alkali metals, viz., of lithium, sodium, and, to a lesser degree, of potassium, were the best studied representatives of metal tellurolates.These com- pounds were prepared either by reduction of the corresponding ditellurides or by insertion of the tellurium atom at the C7Li bond of organolithium derivatives.32 ± 36 However, they are unstable and are readily oxidised; therefore, none of these tellurolates was isolated in the individual form and structurally characterised. These compounds prepared in situ find wide use in the synthesis of various classes of organotellurium compounds and primarily of ditellurides and tellurides.Mercury(II) telluro- lates Hg(TePh)2 , 37, 38 Hg(TeC6H4Me-4)2 , 38 Hg(TeC6H4 ± OEt-4)2 , 39 Hg(TeBut)2 , 40 Hg(TeBun)2 , 41 Hg(TeC2F5)2 (see Ref. 42) and Hg(TeCF3)2 (see Ref. 43) were isolated in the individual state. However, these compounds in the solid state exist as polymers, which precluded their studies by X-ray diffraction analysis. Only the structure of a complex mercury tellurolate 8 was established by this method 45 (the synthesis was described in Ref. 44). M[Hg(TePh)3] PhTeM+Hg(TePh)2 Ph4PCl 7MCl (Ph4P)[Hg(TePh)3] 8 M=Na, K. Mercury tellurolates are soluble in coordinating solvents such as 1,2-bis(diethylphosphino)ethane (DEPE) or 1,2-bis(dimethyl- phosphino)methane (DMPE).41 When the tellurolate and the 911 solvent were taken in a ratio of 1 : 2 or 1 : 1, coordination polymers or dimeric compounds of the [Hg(TePh)2(DEPE)]2 type, respectively, were formed.41 The structure of the latter compound was established by X-ray diffraction analysis.41 When heated in the solid state, the complexes of this type decompose to form mercury telluride, phosphine and the corresponding dio- rganyl ditelluride.41 The dimeric complexes decompose upon heating in solution.Rather stable tellurols containing bulky organic radicals have been synthesised recently, which made it possible to carry out systematic studies of their reactivities and the properties of the respective metal tellurolates (these data were most completely surveyed in a review 4).In this section, we consider methods for the synthesis, reactions and crystal structures of metal tellurolates containing bulky organic radicals. We restrict ourselves to consideration of tellurolates in which metal atoms are bound only to tellurium atoms. Tellurolates of other types, for example, tellurolates of metal cyclopentadienyl derivatives are not considered in detail (the data on these compounds were reported in the review 4). 1. General methods for the synthesis of metal tellurolates Presently, four major procedures for the preparation of metal tellurolates are known. These tellurolates were isolated in the individual state and in some cases were characterised by X-ray diffraction analysis. 1. Tellurolysis of the M7N bond in metal N-trimethyl- silylamides with stable free tellurols (most widely used).(RTe)nM RTeH +M[N(SiMe3)2]n 7(Me3Si)2NH This method was used for the preparation of tellurolates of Li,18, 19 Na,18 K,18 Ca,46 ± 48 Sr,46, 47 Ba,46, 47 Zn,17, 49, 50 Cd,2, 49 Hg,49, 51 Mn(II),52, 53 Sn(II),54 Pb(II),54 Fe(II),52 divalent lantha- nides (Eu,55 Sm55 and Yb55) and trivalent lanthanides (La 56 and Ce 56) as well of as of Al(III),57 Sb(III) 57 {Sb(NMe2)3 was used in the synthesis of Sb[TeSi(SiMe3)3]3} and Bi(III).57 Tellurolates of Mg,46, 47 Zn,17 In(III),57 Tl(I),57 Zr(IV) 58 and Hf(IV) 58 were prepared by tellurolysis of the M7C bonds in metal alkyls or metal cyclopentadienyl derivatives with the use of tris(trimethyl- silyl)silanetellurol (4b).M[TeSi(SiMe3)3]n RnM+HTeSi(SiMe3)3 7RH 2. Exchange reactions of metal tellurolates (most often, of lithium derivatives) with metal halides. These reactions are of a rather general character as well. (RTe)nM RTeM +M0Xn 7M0X M=Li, Na; X=Cl, Br. This method was used for the preparation of tellurolates of Cu(I),59 Au(I),60 Cd,61 Hg,49, 51 Mn(II),52 Fe(II),52 Co(I),52 Cr(II),52 Ga(III) 57 and In(I).57 Sodium tellurolates were used in the synthesis of Zn and Cd tellurolates.62 Tellurolates of divalent lanthanides (Yb, Sm and Eu,63, 64) were synthesised with the use of potassium derivatives of sterically hindered tellurols because it is impossible to separate lanthanide tellurolates from lithium or sodium halides formed in the exchange reactions with the use of lithium or sodium salts.3. Reduction of the corresponding ditellurides with lithium triethylborohydride,61, 65 sodium amalgam,18, 61, 66 potassium amalgam 18 or potassium tris(sec-butyl)borohydride 61, 66 allows one to synthesise alkali metal tellurolates with sterically hindered ligands. The insertion of some metals (Hg,37 ± 39, 41 ± 43, 51 Cd 43 or Yb 67) at the Te7Te bond of tellurides also belongs to reduction reactions. This method was used primarily for the preparation of metal tellurolates 9a ± f, i, j containing not-very-bulky substituents at the tellurium atom. However, tellurolates 9g, h, j were also synthesised by this method.51912 TeR+M RTe (RTe)2M 9a ± j R M Ref. Compound 9 Ph 4-MeC6H4 4-EtOC6H4 Bu C2F5 CF3 2,4,6-Me3C6H2 2,4,6-Pri3C6H2 37, 38 38 39 41 42 43 51 51 43 67 Hg Hg Hg Hg Hg Hg Hg Hg Cd Yb CF3 Ph abcdefghij It should be noted that dimesityl ditelluride readily reacts with metallic mercury in toluene at room temperature without a solvent to form mercury mesitylenetellurolate 9g,51 while the reaction of bis(2,4,6-triisopropylphenyl) ditelluride is performed in a polar solvent, such as methanol, at room temperature.The latter reaction is reversible and the tellurolate 9h decomposes to the corresponding ditelluride and metallic mercury upon heating in methanol or upon recrystallisation from nonpolar solvents. Attempts to prepare mercury bis(2,4,6-tri-tert-butyl- phenyl)tellurolate by the reaction of the corresponding ditellur- ide with metallic mercury have been unsuccessful.51 Ytterbium di(benzenetellurolate) was synthesised in liquid ammonia.After evaporation of the solvent and extraction with pyridine, the complex with composition (PhTe)2Yb(Py)5 was isolated.67 Attempts to prepare Yb(III) tellurolate by the reaction of Yb(III) chloride with sodium benzenetellurolate have failed because Yb(III) was reduced to Yb(II) and the same complex of divalent Yb, viz., (PhTe)2Yb(Py)5, has been obtained as the only product.67 The only copper(I) tellurolate prepared by insertion of the copper atom at the Te7Te bond was Cu(I) trifluoromethane- tellurolate.43 CF3TeCu CF3Te TeCF3+Cu Boiling of diaryl ditellurides with a copper powder in dioxane resulted in elimination of one tellurium atom to form diaryl tellurides in virtually quantitative yields.35, 36 4.Insertion of the tellurium atom at the C7Li bond of the corresponding lithium derivatives was used for the preparation of lithium derivatives of some sterically hindered tellur- ols.17 ± 20, 23, 61, 68 Thus the reactions of organic derivatives of Al(I) 69 and In(I) 70 with elemental tellurium in toluene or hexane gave compounds 10a, b in yields of >80%. The structures of both compounds contain distorted heterocubane molecular cores M4Te4 (M=Al or In) (data from X-ray diffraction analysis).69, 70 M4R4+Te [RMTe]4 10a,b M=Al, R=C5Me5 (a); M=In, R=(Me3Si)3C (b). The insertion of the Te atom at the M7C bond of tris(tert- butyl)gallium 71, 72 or -aluminium 71 occurs on stirring of a mixture of the reagents in toluene at room temperature.In this case, tellurolates 11a, b were obtained in *60% yields. In boiling toluene, compounds 10c, d were formed. These compounds were also formed upon heating of dimers 11a, b.72 The structure of aluminium tellurolate 11b was established by X-ray diffraction analysis.71 I D Sadekov, A V Zakharov But Te But But M M But3M+Te [ButMTe]4 10c,d But But Te But 11a,b M=Ga (10c, 11a), Al (10d, 11b). The two latter methods are less common because they allow one to prepare tellurolates of only a limited number of metals. 2. Alkali metal tellurolates Alkali metal tellurolates were synthesised according to methods 1, 3 and 4 (primarily, according to methods 3 and 4).Tellurolysis of lithium and sodium bis(trimethylsilyl)amides with tris(trimethyl- silyl)silanetellurol 4b afforded lithium tellurolate 5f 18 and sodium tellurolate 5g.18 The reactions performed in hexane yielded alkali metal tellurolates, which are not coordinated by solvent molecules or donor ligands. C6H14 7(Me3Si)2NH MTeSi(SiMe3)3 5f,g (Me3Si)3SiTeH +MN(SiMe3)2 4b M=Li (f, 54%), Na (g, 61%). Potassium tellurolate 5h was prepared by the reaction of the tellurol 4b with ButOK.18 ButOK 4b (Me3Si)3SiTeK 5h the containing metal Alkali bulky tellurolates tris(trimethylsilyl)silyl or aryl substituents were prepared by reducing the corresponding ditellurides (method 3). The reac- tions are generally carried out in THF at room temperature.The resulting tellurolates are isolated as complexes with THF or with other donor ligands. Thus reduction of the ditelluride 7 with potassium or sodium amalgam gave tellurolates 5i, j as complexes with THF or tetramethylethylenediamine (TMEDA).18 Na/Hg, THF (Me3Si)3SiTeNa . (THF)0.5 5i (70%) [(Me3Si)3Si]2Te2 7 1. K/Hg, THF 2. TMEDA (Me3Si)3SiTeK .(TMEDA) 5j (49%) The reactions of ditellurides 12 (R=Me or Pri) with lithium triethylborohydride in THF,61 with sodium amalgam in the presence of TMEDA61, 66 or DME61 as well as with potassium tris(sec-butyl)borohydride either in THF61 or followed by treat- ment with crown ether (18-C-6) 66 afforded complexes of tellur- olates 13a ± f.(2,4,6-R3C6H2)2TeMLn 13a ± f (2,4,6-R3C6H2)2Te2 12 n L M R Yield (%) Compound Reducing 13 agent Me LiBEt3H Na/Hg Na/Hg THF THF TMEDA DME THF 18-C-6 Li LiBEt3H PriLi Na Na KK Me Pri Pri Pri abcdef 80 48 80 64 40 47 1.5 2.5 211.33 1 KBus3BH KBus3BH Complexes of lithium tellurolates were also prepared by the reactions of elementary tellurium with organolithium compounds. Complexes 5b,k were synthesised by the reaction of powdered tellurium with organolithium compound 14.18Stable tellurols and their metal derivatives L Te, THF (THF)2LiTeSi(SiMe3)3 5a (THF)3LiSi(SiMe3)3 14 LiTeSi(SiMe3)3 .L 5b,k L=DME (5b, 82%), 12-C-4 (k, 79%). The reaction of the lithium derivative 14 with Te inDMEgave the complex 5b in 72% yield.20 Complexes of tellurolates 5c,d,19 15 61, 68 and 16 61 were synthesised analogously. Te, THF (THF)2LiTeM(SiMe3)3 5c,d (THF)nLiM(SiMe3)3 14 M=C(c, 56%), Ge (d, 48%).THF 2,4,6-But3C6H2TeLi(THF)3 2,4,6-But3C6H2Li+Te 15 (56%) 1. Te, THF 2. DME 2-Me2NCH2C6H4Li 2-Me2NCH2C6H4TeLi .DME 16 (91%) Complexes of alkali metal tellurolates containing bulky substituents are thermally stable at room temperature and can be stored without decomposition under an inert atmosphere for a long period of time. However, these compounds are readily oxidised in the solid state and in solution with atmospheric oxygen to give the corresponding ditellurides.As can be seen from the above-considered reactions, THF in the complexes is readily replaced by stronger donor ligands such as DME, TMEDA, 18-C-6 or 12-C-4. Some reactions which were studied using alkali metal tellur- olates 5a 17 ± 19 and 5b20 as examples are shown in Scheme 1. Scheme 1 + [(Me3Si)3Si]2Te2 (95%) [Et4N][TeSi(SiMe3)3] (76%) (Me3Si)3SiTeSiMe3 (82%) Cl7 + Et4NCl Me3SiCl O2(CuCl) or I2 LnLiTeSi(SiMe3)3 5a,b Me3Te+I7 7Te E, THF, 755 8C 7LiI, 7Me2Te (Me3Si)3SiTeMe (THF)2LiESi(SiMe3)3 E=Se (93%), S (90%). It is known 36 that lithium and sodium alkane- and arenetel- lurolates are oxidised to the corresponding ditellurides. In this respect, the behaviour of lithium tris(triphenylsilyl)silane- tellurolate (5e) is unusual.24 Its oxidation with atmospheric oxygen or the reaction with CuCl afforded only (Ph3Si)2Te.24 The exchange reactions of tellurium atoms in diaryl ditellur- ides 36 and in some tellurium-containing heterocycles 73 for sulfur or selenium atoms have long been known.These reactions proceed at high temperature. The tellurolate 5a reacts with S or Se at 755 8C. Based on this fact, Bonasia et al.19 considered the reactions of the tellurolate 5a with selenium and sulfur (meta- thesis) yielding lithium tris(trimethylsilyl)silaneselenolate and -thiolate to be a new type of reactions in the chemistry of chalcogens. 913 X (Me3Si)3SiXLi(THF)2+Te (Me3Si)3SiTeLi(THF)2 5a X=Se, S.Reactions of the tellurolate 5a with chalcogens occur as the replacement of tellurium by a lighter chalcogen. It is known that in the reactions of lithium alkane- and arenechalcogenolates, lighter chalcogens are inserted into the bond between heavier chalcogen and lithium atoms to form lithium chalcogenylchalcogenolates 17.74RE1Li+E2 RE1E2Li 17 E1=Te; E2=Se, S; E1=Se, E2=S. Unlike with lithium tris(trimethylsilyl)silanetellurolate 5a, the reaction of the carbon analogue of this compound, viz., lithium tris(trimethylsilyl)methanetellurolate 5c, with selenium at 755 8C resulted in a dark-red solution containing, according to the 125Te NMR spectrum, lithium tellurenylselenolate 18. Attempts at isolating this compound in the solid state have led to its decomposition into a mixture of nonidentified compounds.19 Se, THF, 755 8C (THF)2LiTeC(SiMe3)3 5c (THF)2LiSeTeC(SiMe3)3 18 The exchange reactions of tellurolates 5a and 5e with cyclo- pentadienyl complexes of Ti, Zr and Hf made it possible to synthesise new tellurium-containing complexes 19a ± j.24, 75 The structures of the compounds 19d, g 75 and 19f 24 were established by X-ray diffraction analysis.Cp2M[TeSi(SiR3)3]2 19a ± j Cp2MCl2+(THF)nLiTeSi(SiR3)3 5a,e Yield (%) Cp Compound 19 R M 75 55 86 70 71 74 75 67 33 72 Ph Ph Me Me Ph Ph Me Ph Ph Me C5H5 C5Me5 C5H5 C5Me5 C5H5 C5H4But C5H5 C5H5 C5H4But C5H5 Ti Ti Ti Ti Zr Zr Zr Hf Hf Hf abcdefghij Compounds of the type 19 were also formed in the reactions of lithium tellurolate 5c with metal complexes.75 However, the resulting products rapidly decomposed to give eventually Te, HC(SiMe3)3 and a mixture of nonidentified metal-containing compounds.In the reactions of Ti, Zr and Hf cyclopentadienyl complexes with the tellurolate 5a, the nature of the solvent is of considerable importance. Thus the reactions in hexane afforded tellurolates 19, while the reaction of the compound 5a with titanium complex in THF yielded the dimeric complex 20 and the ditelluride 7.75 THF, 25 8C Cp2TiCl2+(THF)2LiTeSi(SiMe3)3 [Cp2TiCl]2+[(Me3Si)3SiTe]2 7 20 The reactions of tellurolates 13a, b with cis-bis(triphenyl- phosphine)platinum dichloride resulted in displacement of one chloride ligand to form tellurium-containing platinum complexes 21a, b.The structure of the complex 21b was established by X-ray diffraction analysis.61914 THF 2,4,6-R3C6H2TeLi(THF)n+(PPh3)2PtCl2 7LiCl 13a,b (PPh3)2Pt[Te(2,4,6-R3C6H2)]Cl cis-21a,b R=Me (a), Pri (b). The structures of many complexes of alkali metal tellurolates with Lewis bases were established by X-ray diffraction analysis. Below are considered the characteristic structural features of these compounds. In the crystal, lithium tellurolate (THF)2LiTe ± Si(SiMe3)3 (5a) exists as dimers and contains the Li2Te2 core. The Te7Li bond lengths in the ring (2.82 and 2.88 A) are close to the values predicted based on the ionic radii (Li+, 0.73 A; Te27, 2.07 A).76 The Te7Si and Li7O bond lengths are 2.48 and 1.90 A, respectively.The Li7Te7Li and Te7Li7Te bond angles are 90.4 and 89.6 8, respectively.18 In lithium tellurolate 16, the Li atom is tetrahedrally coordi- nated by two oxygen atoms of DME, the nitrogen atom and the tellurium atom.61 The Te7Li distance in this compound (2.717 A) is smaller than that in the tellurolate 5a. The Te7C bond length (2.142 A) has the standard value.77 The Te7Li7O bond angles are 125.8 and 118.2 8 and the Te7Li7N angle is 99.6 8. However, it remains unclear whether this molecule retains N?Te coordination, which has been observed in analogous compounds with tellurium-containing substituents in the ortho position with respect to the CH(R)NMe2 group (R=H or Me) in the aromatic rings [the N?Te bond lengths are 2.366 78 and 2.406 A79 in 2-iodotellurenylbenzyldimethylamine and 2-(1-dime- thylaminoethyl)benzenetellurium trichloride, respectively, which are substantially smaller than the sum of the van der Waals radii of Te and N (3.70 A) 30].In the complex 15 61 containing a bulky aryl substituent, the C7Te7Li angle is 99.1 8, whereas the corresponding angles in the structurally similar sodium tellurolate 13c 66 and potassium tellurolate 13f 66 are 102.2 8 and 77.7 8, respectively. The aryl ligand in the tellurolate 15 is planar. The Te atom deviates from the plane by 0.56 A. The Te7Li bond length (2.82 A) is close to the corresponding bond length in the tellurolate 5a.The Te7C and Li7O bond lengths are 2.184 and 1.95 A, respectively. The Te7Li7O angles are in the range of 108.8 ± 115.2 8. In the crystal, potassium tellurolate 13f exists as monomers. The Te7K distance (3.499 A) is indicative of a weak interaction between the alkali metal atom and the tellurolate anion.66 The Te7C distance is 2.15 A and the C7Te7K angle is 77.68 8. The average K7O bond length is 2.849 A. 3. Alkaline-earth metal tellurolates The major method for the preparation of alkaline-earth metal 7(Me3Si)2NH tellurolates 22 involves tellurolysis of di[bis(trimethylsilyl)amides] of the corresponding metals with tris(trimethylsilyl)silanetellurol 4b in hexane at room temperature (the amide : tellurol ratio is 1 : 2).In most cases, the resulting tellurolates were isolated as adducts with different Lewis bases.46, 47 M[N(SiMe3)2]2+(Me3Si)3SiTeH 4b L M[TeSi(SiMe3)3] [(Me3Si)3SiTe]2M. Ln 22a ± i n Yield (%) M L Compound 22 68 73 50 46 66 25 75 52 73 THF Py 12-C-4 THF Py THF Py THF Py Mg Mg Mg Ca Ca Sr Sr Ba Ba abcdefghi 222444445 I D Sadekov, A V Zakharov The reaction performed in toluene afforded compound 22d in 85% yield.48 Magnesium tellurolate 22j, which does not contain coordinat- ing solvents, was synthesised by tellurolysis of the Mg7C bond of dibutylmagnesium.46, 47 C6H14 MgBu2+HTeSi(SiMe3)3 7BuH [(Me3Si)3SiTe]2Mg 22j (77%) A complex of magnesium 22a with tetrahydrofuran was prepared 47 by the reaction of mercury bis[tris(trimethyl- silyl)silanetellurolate] 49 with an excess of powdered Mg in THF (the yield was 44%).Tellurolates 22a ± j are yellow crystalline compounds, which are readily soluble in hydrocarbons and ether. These compounds are sensitive to atmospheric oxygen and moisture but can be stored under an inert atmosphere at room temperature for a long period of time. The reactions of tellurolates 22 with oxygen afforded ditelluride 7 and the reactions with water yielded the tellurol 4b. The tetrahydrofuran complexes slowly lose coordinated solvent molecules at room temperature and are readily converted into the pyridine complexes upon recrystallisation from pyri- dine.46, 47 Treatment of magnesium tellurolates 22a, b with 12- crown-4 resulted in replacement of both the coordinated solvent molecules (pyridine or THF) and the tellurolate ligands to form the ionic complex 22c.46, 47 72L [Mg(12-C-4)2][TeSi(SiMe3)3]2 22c [(Me3Si)3SiTe]2Mg. L2+12-C-4 22a,b The 1H NMR spectra of magnesium tellurolate 22j in toluene-d8 are indicative of the stereochemical flexibility of the latter in solutions.47 At 20 8C, the 1HNMRspectrum has a singlet signal, which is broadened as the temperature is decreased and is split into two peaks at790 8C.47 Apparently, this is due to a slow (within the NMR time scale) intramolecular exchange of the bridging and terminal ligands in the dimeric structure of the tellurolate at low temperature.47 R2 Te R2 Te MgTeR1 MgTeR1 R1TeMg R1TeMg Te R2 Te R2 R1 Te MgTeR1 R2TeMg Te R2 In the 125 Te NMR spectra of pyridine adducts 22b,e,g,i, the signals for the 125Te nuclei are observed in a rather narrow region from d71578 to71405 46, 47 (Table 2).The molecular and crystal structures of the tellurolates 22a,46, 47 22d 46 ± 48 and 22i 47 were established by X-ray diffrac- tion analysis. In magnesium tellurolate 22a, the Mg atom is tetrahedrally coordinated by two oxygen atoms of two THF molecules and two tellurium atoms.46, 47 The Mg7Te bond lengths have similar values (2.720 and 2.714 A), the Mg7O bond lengths are in the range of 2.035 ± 2.038 A and the Te7Mg7Te and O7Mg7O angles are 135.48 and 94.8 8, respectively.The Te7Mg7O angles are nonequivalent, one angle at each tellurium atom being large (111.77 and 144.37 8) and one angle being small (96.37 and 94.79 8). This may be due to steric repulsions from the lone electron pairs of the tellurium atoms although a crystal packing effect is not ruled out.46, 47 The results of two independent crystallographic studies 46, 48 of calcium tellurolate 22d agree with each other. The coordination polyhedron about the Ca atom is a distorted octahedron. The Ca7Te distance is 3.19 A and the Ca7O distances are 2.36 andStable tellurols and their metal derivatives Table 2. Data of 125Te NMR spectroscopy of metal tellurolates.a Ref. Compound Solvent d C7D8 71578 18, 19 19 20 (Me3Si)3SiTeLi(THF)2 (5a) C6D6 71622 (Me3Si)3SiTeLi(THF) (Me3Si)3SiTeLi(DME) (5b) C6D6 71641 (Ph3Si)2TeLi(THF)3 (5e) C6D6 71337 24 (Me3Si)3CTeLi(THF)2 (5c) C7D8 7287 (Me3Si)3GeTeLi(THF)2 (5d) C6D6 71515 2,4,6-Me3C6H2TeLi(THF)1.5 (13a) 2,4,6-Pri3C6H2TeLi(THF)2.5 (13b) 2,4,6-But3C6H2TeLi(THF)3 (15) THF THF THF THF-d8 2,4,6-Me3C6H2TeK(THF)1.3 19 19222 63 46, 47 46, 47 THF-d8 72204 48 46, 47 46, 47 7383.7 7478.5 9.2 7296 [(Me3Si)3SiTe]2Mg(Py)2 (22b) C6D6 71578 [(Me3Si)3SiTe]2Ca(Py)4 (22e) C6D6 71458 [(Me3Si)3SiTe]2Ca(THF)4 (22d) [(Me3Si)3SiTe]2Sr(Py)4 (22g) C6D6 71482 [(Me3Si)3SiTe]2Ba(Py)5 (22i) C6D6 71405 [(Me3Si)3SiTe]3Sb (23d) C6D6 7584 57 [(Me3Si)3SiTe]3Bi (23c) C6D6 7627 57 49 49 CDCl3 71469 {[(Me3Si)3SiTe]2Zn}2 (22k) C7D8 7783, 71215 {[(Me3Si)3SiTe]2Cd}2 (22l) C7D8 7933,b 71338 c [(Me3Si)3SiTe]2Hg (22m) C7D8 7850 [(Me3Si)3SiTe]2Zn(Py)2 (31a) [(Me3Si)3SiTe]2Cd(DMPE) (31d) CD2 Cl2 71565 d [(Me3Si)3SiTe]2Sn(DMPE)2 (33b) C7D8 71175 e [(Me3Si)3SiTe]2Sn(PMe3) (33a) C7D8 71165 f [(Me3Si)3SiTe]2Fe(DMPE)2 (34c) C6D6 71189 g [(Me3Si)3SiTe]2Fe(Cl)(DMPE) (36) C6D6 71350 h 459 193 203 see i see i see i 49 49 49 54 54 52 52444 see i 4 714 [(Me3Si)3SiTe]4Zr (46a) [(Me3Si)3SiTe]4Hf (46b) [(Me3Si)3SiTe]4Zr(NCC6H3Me2-2,6)2 (48a) [(Me3Si)3SiTe]4Hf(NCC6H3Me2-2,6)2 (48b) [(Me3Si)3SiTe]4Zr(DMPE) (49a) [(Me3Si)3SiTe]4Hf(DMPE)2 (49b) [(Me3Si)3SiTe]2Te=Zr(DMPE)2 (50a) see i see i see i 4 see i [(Me3Si)3SiTe]2Te=Hf(DMPE)2 (50b) 151,7116 4 128,7349 4 7706, 71173,j 71197 7701, 71212,k 71250 4 [(Me3Si)3SiTe]Au (41b) C6D6 71112 60 [(Me3Si)3GeTe]Au (41c) C6D6 71093 60 [(Me3Si)3CTe]4Au4 (41a) C6D6 7302 [(Me3Si)3CTe]Au(PPh3) (42a) C6D6 7148 [(Me3Si)3SiTe]2Yb(TMEDA)2 (55a) C6D6 610 [(Me3Si)3SiTe]2Sm(TMEDA)2 (55b) C6D6 3243 THF-d8 THF-d8 THF-d8 7280 7270 7260 60 60 55 55 64 63 64 (2,4,6-Me3C6H2Te)2Yb(THF)2 (54a) (2,4,6-Me3C6H2Te)2Yb(THF)3 (54a) (2,4,6-Me3C6H2Te)2Yb(diglym)2 (54d) [(Me3Si)3SiTe]3La (58a) C7D8 1018 56 [(Me3Si)3SiTe]3La(DMPE)2 (59a) C7D8 7894, 71074 C6D6 (Ph3Si)2Te 7851 56 [(Me3Si)3SiTe]2 (7) C6D6 7678 18 24 (Me3Si)3SiTeSiMe3 (44) C6D6 71081 18 18 C6D6 (Me3Si)3SiTeMe 7637.7 JTe7P=137 Hz; c JTe7Cd=1821 Hz; a The chemical shifts are given relative to Me2Te; b JTe7Cd=1059 Hz; d JTe7Cd=744 Hz, e JTe7P=64 Hz; f JTe7P=79 Hz; g JTe7P=80 Hz; h JTe7P=78 Hz; i the solvent is not mentioned; j t, JTe7P=166 Hz; k t, JTe7P=190 Hz.915 2.41 A. The tellurium ligands are in trans positions. The Ca7Te7Si angle is 128.5 8. In barium tellurolate 22i, the five pyridine ligands are located in the equatorial plane and the two tellurium ligands are in axial positions. The Ba7Te distances are 3.38 A, the Ba7N distance (average value) is 2.888 A and the Ba7Te7Ba angle is 171.9 8.47 4. Aluminium, gallium, indium, antimony and bismuth tellurolates In this section, we discuss methods for the synthesis of aluminium, gallium and indium tellurolates containing bulky substituents at the tellurium atom.Tellurolates of Group IIIA elements were synthesised by tellurolysis of the N7M and C7M bonds with the tellurol 4b (method 1).57 This procedure was used for the preparation of tellurolates of Al 23a, In(III) 23b and Bi(III) 23c. 4b, C6H14, 20 8C Al[N(SiMe3)2]3 7(Me3Si)2NH Al[TeSi(SiMe3)3]3 23a (44%) 4b, C6H14, 20 8C Cp3In 7CpH In[TeSi(SiMe3)3]3 23b (60%) 4b, C6H14,778 to 0 8C Bi[N(SiMe3)2]3 7(Me3Si)2NH Bi[TeSi(SiMe3)3]3 23c (68%) Tellurolysis of the C7M bonds afforded tellurolates of In(I) 24a and Tl(I) 24b. Compound 24b was isolated as a solvate with composition Tl[TeSi(SiMe3)3](C6H14)0.125. C6H14, 20 8C CpM+HTeSi(SiMe3)3 7CpH M[TeSi(SiMe3)3] 24a,b M=In (a, 43%), Tl (b, 80%).The synthesis of Sb(III) tellurolate 23d was performed by the tellurolysis of tris(dimethylamino)antimony (25) instead of the antimony trimethylsilylamide derivative.57 4b, C6H14,778 to 20 8C 7Me2NH Sb[TeSi(SiMe3)3]3 23d (75%) Sb(NMe2)3 25 It should be noted that tris(dimethylamino) derivatives of Sb and Bi are convenient starting compounds in the low-temperature synthesis of tellurides of the corresponding elements. Thus polycrystalline Sb2Te3 and Bi2Te3 were prepared in high yields by the reactions of (Me2N)3M(M=Sb and Bi, respectively) with bis(trimethylsilyl) telluride.80C6H14,730 8C M2Te3 M(NMe2)3+(Me3Si)2Te 7Me3SiNMe2 M=Sb, Bi. Method 2 consisting in the exchange reaction of metal tellurolates with metal halides is of lesser importance due to lower yields and lesser purity of the compounds formed.Thus tellurolates of Ga(III) 23e and In(I) 24b were synthesised by the reactions of the corresponding chlorides with lithium tris(trimethylsilyl)silanetellurolate 5a.57 5a, C6H14, 20 8C GaCl3 7LiCl Ga[TeSi(SiMe3)3]3 23e (67%) 5a, C6H14, 20 8C InCl 7LiCl In[TeSi(SiMe3)3] 24b (65%) Attempts to prepare Tl(III) tellurolate by the exchange reaction have been unsuccessful because the tellurolate 5a was oxidised to the ditelluride 7 in the course of the reaction.57916 The tellurolates 23a,b,e are coloured (from yellow to orange) air- and light-sensitive crystalline compounds, which are readily soluble in hydrocarbons.Tellurolates of Sb(III) and Bi(III) are particularly light-sensitive and thermally unstable. Because of this, these compounds are stored at low temperature in the dark.57 The In(I) and Tl(I) derivatives 24a,b form solvates with the solvents (including hexane) from which they were recrystallised. The solvates are stable and lose solvent molecules with difficulty upon heating in vacuo.57 Reactions of tellurolates 23 remain virtually unknown. It was noted 57 that In(III) tellurolate 23b reacted with different Lewis bases. However, in all cases, the compound 23b underwent substantial decomposition to form [(Me3Si)3Si]2Te and [(Me3Si)3Si]2Te2 (7). The reaction of Al tellurolate 23a with Cp2TiCl2 afforded the above-described titanocene tellurolate 19c (the yield was 44%).57 The molecular and crystal structure of one of the tellurolates discussed in this section, viz., of Ga(III) tellurolate 23e, was established by X-ray diffraction analysis.57 5.Zinc, cadmium and mercury tellurolates Zinc, cadmium and mercury tellurolates containing both ordinary alkyl and aryl substituents and bulky aryl and trimethylsilyl substituents at the tellurium atom were synthesised. The most general procedure for the preparation of zinc, cadmium and mercury tellurolates involves tellurolysis of metal di[bis(trimethylsilyl)amides] with sterically hindered tellurols 3a 2, 50, 51 and 4b.17, 49 778 to 20 8C M[N(SiMe3)2]2+2,4,6-Me3C6H2TeH light petroleum 3a [(2,4,6-Me3C6H2Te)2M]n 26a ± c Yield (%) Ref.M Compound 22 abc 502 51 52 50 65 Zn Cd Hg C6H14, 20 8C [(Me3Si)3SiTe]2M 22k ±m M[N(SiMe3)2]2+HTeSi(SiMe3)3 4b Ref. Compound 22 Yield (%) M klm 17, 49 49 49 75 69 82 Zn Cd Hg Tellurolysis of theC7Te bond was used only for preparing Zn tellurolate 22k (in quantitative yield).17 C6H14, 20 8C Et2Zn+HTeSi(SiMe3)3 7C2H6 [(Me3Si)3SiTe]2Zn 22k Dealkystannylation (dealkylsilylation) reactions are not so common. For example, these reactions were used for the prepara- tion of sterically hindered cadmium tellurolate 26b 2 and its benzene analogue 27.81 2,4,6-Me3C6H2TeSnBu3+CdCl2 7Bu3SnCl [(2,4,6-Me3C6H2Te)2Cd]n 26b PhTeSiMe3+CdMe2 7Me4Si (PhTe)2Cd 27 Stuczynski et al.81 stated that aryl stannyl(silyl) tellurides offer advantages over tellurols in the synthesis of different types of metal arenetellurolates.In fact, compounds containing the Te7E7C bonds (E=Si or Sn) are substantially more stable than tellurols. These compounds are obtained in high yields from readily available initial compounds and can be stored without decomposition over a long period of time. I D Sadekov, A V Zakharov So far, the exchange reactions have not been used for the preparation of Zn, Cd and Hg tellurolates with sterically hindered aryl substituents. However, Cd and Zn bis(2-amino- phenyl)tellurolates 28a,b 62 and Cd tellurolate 2961 were synthes- ised by these reactions. These compounds were obtained in rather high yields by the reactions of the corresponding lithium arene- tellurolates 61 or sodium arenetellurolates 62 with Cd(II) and Zn(II) salts and are the first examples of intracomplex compounds with an MN2Te2 coordination unit.2-H2NC6H4TeNa +M(OAc)2 7AcONa (2-H2NC6H4Te)2M 28a,b M=Cd (a), Zn (b). 2-Me2NCH2C6H4TeLi CdCl2 7LiCl (2-Me2NCH2C6H4Te)2Cd 29 The exchange reactions are not commonly used for the preparation of Zn, Cd and Hg tellurolates containing the TeSi(SiMe3)3 group. The reaction of HgCl2 with the tellurolate 5a afforded Hg tellurolate 22m in lower yield than in the tellurolysis.49 The reactions of ZnCl2 and CdBr2 with the tellur- olate 5a gave the respective Zn and Cd complexes with the solvent.49 Tellurolate complexes 26c,d were prepared by insertion of the Hg atom at the Te7Te bond of diaryl ditellurides 12a, b in methanol or toluene.51 [(2,4,6-RC6H2Te)2Hg]n 26c,d (2,4,6-RC6H2)2Te2+Hg 12a,b R=Me (26c), Pri (26d, 90%).Mercury tellurolate 26c was synthesised by the reaction of the ditelluride 12a with Hg[N(SiMe3)2]2 in light petroleum at 20 8C (the yield of the product was not reported).51 The mechanism of this sluggish reaction remains unclear. The assumption that N2(SiMe3)4 was formed as a by-product was not confirmed. Zinc and cadmium silylamides do not enter into analogous reactions.51 The tellurolates 26a ± d are virtually insoluble in polar solvents because they have the polymeric structures in the solid state.However, these compounds are readily soluble in strongly coordinating solvents such as DMSO, DMF and pyridine.50, 51 The reactions of zinc tellurolate 26a with strong Lewis bases afforded four-coordinate monomeric adducts 30a ± e, which were isolated in the individual state in 49%± 62% yields.50 Adducts 30a ± e are more soluble than the tellurolate 26a. These com- pounds rather readily lose ligand molecules upon heating. [Zn(TeMes)2]n+L 26a (Mes2Te)2Zn . Ln 30a ± eNMe , n=2 (30c); L=PMe3: n=1 (30a), 2 (30b); L= N L=DMPE, n=1 (30d); L=Py, n=2 (30e). Unlike zinc tellurolate 26a, neither cadmium tellurolate 26b nor mercury tellurolate 26c form stable adducts with pyridine due to the lower Lewis acidity of cadmium and mercury compared to that of zinc.50 Zinc tellurolate 26a and the complexes 30a ± e decompose upon heating in the solid state50 and cadmium tellurolate 26b decomposes upon boiling in mesitylene2 to form the correspond- ing tellurides, viz., microcrystalline ZnTe and polycrystalline CdTe, respectively.Cadmium telluride was also prepared by heating cadmium tellurolate 29 at 180 8C in vacuo.61 D Mes2Te+MTe (2-Me2NCH2C6H4)2Te 7CdTe M(TeMes)2 26a,b (2-Me2NCH2C6H4Te)2Cd 29Stable tellurols and their metal derivatives Mercury tellurolate 26c decomposed upon heating to form metallic mercury and bis(2,4,6-trimethylphenyl) ditelluride.51 Tellurolates 26 are of limited utility as precursors of metal tellurides due to their polymeric structures, low volatilities and stability.51, 82 Zinc, cadmium and mercury tris(trimethylsilyl)silanetellurol- ates 22k ±m are yellow-green crystalline compounds, which are stable under a nitrogen atmosphere at room temperature for a long period of time.17, 49 Their ability to undergo oxidation is determined by the Lewis acidity of the metal atoms.Thus, zinc tellurolate 22k was almost immediately oxidised to ditelluride 7 when exposed to air, while mercury derivative 22m remained unchanged on exposure to air for at least 45 min.49 The tellur- olates 22k ±m are thermally stable compounds, sublime without decomposition under reduced pressure,49 are readily soluble in different organic solvents and are stable in solutions in the absence of atmospheric oxygen.According to the data of mass spectrometry, the compounds 22k ±m exist in the gas phase as monomers. In solutions, zinc and cadmium tellurolates (22k and 22l, respectively) exist as dimers with the bridging and terminal RTe ligands (the data of multi- nuclear NMR spectroscopy 49) analogous to those suggested for magnesium tellurolate 22j.47 The exchange between the bridging and terminal RTe ligands is rapid within the NMR time scale at 20 8Cand is slow at low temperatures. The 1HNMRspectra of the tellurolates 22k, l in toluene-d8 at 20 8C have singlet signals of the Si(SiMe3)3 groups at d 0.43 ± 0.45 (see Table 2). These signals are gradually broadened as the temperature decreases and are split into two signals with equal intensities at750 8C and734 8C for zinc tellurolate 22k and cadmium tellurolate 22l, respectively.The temperatures of coalescence indicate that the rate of exchange in the tellurolates 22k, l is substantially smaller than that in Mg tellurolate 22j (the coalescence temperature is 790 8C). The data of 125Te NMR spectroscopy (see Table 2) and 113Cd NMR spectroscopy support this conclusion.49 Thus in the 125Te NMR spectrum of zinc tellurolate 22k, the signal which is observed as a singlet at room temperature is split at 760 8C into signals at d 71215 [the terminal (Me3Si)3SiTe group] and7783 [the bridging (Me3Si)3SiTe group].49 In the temperature range from +80 to 780 8C, the 1H and 125Te NMR spectra of mercury tellurolate 22m have singlet signals, which does not allow one to choose between the mono- meric structure and the dimeric structure with rapid (within the NMR time scale) exchange between the terminal and bridging ligands.49 The reactions of the tellurolates 22k, l with different Lewis bases afforded four-coordinate complexes 31a ± d in high yields.49 [(Me3Si)3SiTe]2M.Ln 31a ± d [(Me3Si)3SiTe]2M+L 22k,l M=Zn: L=Py, n=2 (31a, 90%); L=bipy, n=1 (31b, 71%); M=Cd: L=bipy, n=1 (31c, 76%); L=DMPE, n=1 (31d, 77%). The complexes 31a ± d are substantially more resistant to air and are less soluble in hydrocarbons compared to the initial tellurolates. Tris(trimethylsilyl)silanetellurolates have considerable advan- tages over metal arenetellurolates in the pyrolytic synthesis of metal tellurides owing to their high thermal stability, solubility in inert solvents and ability to sublimation leading to monomeric molecules.Thus Zn tellurolate 22k gave cubic ZnTe at 250 ± 300 8C and Cd tellurolate 22l gave hexagonal CdTe at 280 ± 350 8C. The only organic reaction product was bis[tris(trimethylsilyl)silyl] telluride 32.83 D [(Me3Si)3SiTe]2M MTe+[(Me3Si)3Si]2Te 32 M=Zn, Cd. 917 The molecular and crystal structures of the tellurolates 26b 2 and 22k 49 and some complexes, in particular, 30e,50 31a49 and 31d,49 were established by X-ray diffraction analysis. In the crystal, the compound 26b exists as a polymer with the tetrahedrally coordinated cadmium atom and the bridging mesi- tylenetellurolate ligands.2 The average Cd7Te and C7Te bond lengths are 2.837 and 2.164 A, respectively.The Cd7Te7Cd bond angles are in the range of 84.0 ± 116.7 8 and the C7Te7Cd angle is 107 8. In the crystal, zinc bis(tristrimethylsilyl)silanetellurolate 22k exists as dimers.49 The Zn2Te2 fragment is nonplanar; the Zn7Te7Zn and Te7Zn7Te angles are 79.29 8 and 95.62 8, respectively. The bridging tellurium atoms have a pyramidal configuration. The coordination environment about each zinc atom is planar-trigonal. The bond lengths in the Zn2Te2 core (2.628 ± 2.650 A) are substantially larger than the Zn7Tetermin bond lengths (2.50 A). The average Te7Si bond lengths in the bridging and terminal ligands have similar values (2.552 and 2.516 A, respectively). The pyridine complex of zinc tellurolate 30e has a distorted tetrahedral geometry.50 The Te7Zn7Te angle (126.9 8) is larger than the standard value due to steric repulsions between the bulky substituents, while the N7Zn7N angle is 89.2 8.The Zn7Te bond length (2.573 A) is equal to that in the pyridine complex of zinc silanetellurolate 31a (2.574 A49). The Te7C and Zn7N bond lengths are 2.140 and 2.110 A, respectively. 6. Tin, lead, manganese, chromium, iron and cobalt tellurolates Most of tin, lead, manganese, chromium, iron and cobalt tellurolates that have been synthesised contain the TeSi(SiMe3)3 fragment. These compounds were prepared according to two procedures, viz., by tellurolysis of the N7metal bonds in metal silylamides (method a) and by the exchange reactions of metal halides with lithium tellurolate 5a (method b).Method a was used for the preparation of Sn(II) tellurolate 22n and Pb(II) tellurolate 22o.54 C6H14, 20 8C 7(Me3Si)2NH [(Me3Si)3SiTe]2M] 22n,o M[N(SiMe3)2]2+HTeSi(SiMe3)3 4b M=Sn (22n, 81%), Pb (22o, 82%). Method b is less suitable for the preparation of tellurolates 22n,o. Thus the exchange reaction of SnCl2 with tellurolate 5a afforded the tellurolate 22n in substantially lower yield than that obtained by tellurolysis. The reaction of PbCl2 with lithium tellurolate 5a was accompanied by reduction of PbCl2 to metallic lead and oxidation of the compound 5a to the ditelluride 7.54 The tellurolates 22n, o are orange or red crystalline com- pounds, which are stable under nitrogen for a long period of time and are rapidly oxidised in air to form the ditelluride 7.54 The structures of the tellurolates 22n,o in solution cannot be unambiguously described on the basis of the 1H NMR spectral data.Seligson and Arnold 54 assumed that the presence of non- equivalent signals for the TeSi(SiMe3)3 groups in the low-temper- ature 1H NMR spectra may be ascribed either to a monomer ± oligomer equilibrium or to cis ± trans isomerisation in dimeric structures A and B. Te Te Te Te Te Te Te M M M M Te B A Like other Sn(II) complexes, the tellurolate 22n forms adducts with phosphines (33a, b) in high yields.54 [(Me3Si)3SiTe]2Sn+L 22n {[(Me3Si)3SiTe]2Sn}Ln 33a,b n=1, L=PMe3 (33a, 80%); n=2, L=DMPE (33b, 75%).918 The tellurolate 22n forms analogous adducts with a number of other monodentate phosphorus-containing ligands, viz., with Me2PPh, Et3P and (cyclo-C6H11)3P.However, these adducts were not isolated in the individual state. Their formation was judged from the presence of high-field signals in the 31P NMR spectra of solutions. Weak Lewis bases, such as Ph3P and (MeO)3P, do not form complexes with 22n.54 Pyrolysis of the tellurolates 22n,o under a stream of nitrogen at 250 8C afforded cubic tellurides SnTe and PbTe. However, these compounds contained admixtures of small amounts of carbon and hydrogen.54 Thermolysis of the tellurolates 22n, o can be per- formed by heating them in aromatic hydrocarbons. Thus PbTe was prepared by heating the tellurolate 22o in toluene at 100 8C for 22 h.The addition of Lewis bases to solutions of the tellurolates 22n, o promotes their decomposition to form metal tellurides. Thus the tellurolate 22o was quantitatively converted into PbTe upon addition of an excess of MeCN to an ethereal solution. Like silylamides of other metals, Fe(II) and Mn(II) di[bis(trimethylsilyl)amides] reacted with the tellurol 4b in hexane at room temperature. However, attempts to purify the resulting tellurolates failed because of their very high solubility.52 Tellur- olysis of Fe(II) and Mn(II) silylamides in the presence of Lewis bases afforded stable crystalline adducts 34a ± c.52 C6H14, 20 8C M[N(SiMe3)2]2+HTeSi(SiMe3)3+L 7(Me3Si)2NH 4b [(Me3Si)3SiTe]2M. Ln 34a ± c M=Mn: n=1, L=DMPE (34a, 79%); N (34b, 81%); n=2, L=But M=Fe, n=2, L=DMPE (34c, 71%).The reaction of manganese silylamide with the sterically hindered arenetellurol 3a was performed in the usual fashion and gave Mn(II) arenetellurolate 26e in 95% yield.53 light petroleum, 20 8C Mn[N(SiMe3)2]2+2,4,6-Me3C6H2Te2H 7(Me3Si)2NH 3a [(2,4,6-Me3C6H2Te)2Mn]n 26e Fe(II), Mn(II) and Cr(II) tellurolates 34a, c, d were also prepared by the exchange reactions.52 Since the reactions of dihalides of these elements with lithium tellurolate 5a afford the ditelluride 7 and insoluble pyrophoric products with unidentified structures, the exchange reactions are carried out either in the presence of Lewis bases or starting from adducts of metal halides with DMPE.52 MX2(DMPE)2+(THF)2LiTeSi(SiMe3)3 7LiX,7THF 5a [(Me3Si)3SiTe]2M.(DMPE)n 34a,c,d X=Cl, I; M=Mn, n=1 (34a, 80%);M=Fe, n=2 (34c, 79%); M=Cr, n=2 (34d, 74%). The reaction of lithium tellurolate 5a with CoBr2 yielded Co(I) tellurolate 35 as the final product due to strong reducing proper- ties of the compound 5a.52 PhMe, 20 8C CoBr2(PMe3)3+(THF)2LiTeSi(SiMe3)3 7LiBr,7THF 5a [(Me3Si)3SiTe]Co . (PMe3)3+7 35 (64%) The reaction of the adduct FeCl2(DMPE)2 with the tellurolate 5a (taken in an equimolar ratio) afforded tellurolate 36 with mixed I D Sadekov, A V Zakharov ligands (the yield was 88%).52 This compound was also prepared in 82% yield by ligand exchange between the tellurolate 34c and FeCl2(DMPE)2.FeCl2(DMPE)2+(THF)2LiTeSi(SiMe3)3 5a [(Me3Si)3SiTe]2Fe(DMPE)2+FeCl2(DMPE)2 34c [(Me3Si)3SiTe]Fe(DMPE)Cl 36 The Fe(II) tellurolate complex 34c differs from other tellur- olates and their complexes in the lability of the tellurolate ligands. This is evidenced by the above-considered ligand exchange reaction as well as by the data on the conductivity of a solution of the compound 34c in MeCN.52 The molar conductivity of the solution (182 O71 cm2 mol71) corresponds to that of an 1 : 1 electrolyte. Recrystallisation of the complex 34c from MeCN afforded a salt-like complex 37.52 [(Me3Si)3SiTe]2Fe(DMPE)2+MeCN 34c {[(Me3Si)3SiTe]Fe(DMPE)2(MeCN)}+[TeSi(SiMe3)3]7 37 (67%) The complexes 34a ± c are coloured crystalline compounds, which are readily soluble in hydrocarbons.These compounds are air- and moisture-sensitive but are stable under nitrogen at room temperature. Among them, Mn(II) complexes 34a, b are the most air-sensitive (they are oxidised in a few seconds). The complexes 34a ± c cannot be sublimed because they decompose on heating (100 ± 200 8C, 1073 Torr).52 The complexes 34a,b,d are paramagnetic. The magnetic moments are 2.7 ± 5.9 mB.52 In the 1H NMR spectra, the signals for the Si(SiMe3)3 groups are broad and are shifted downfield by 2 ± 4 ppm compared to the signals for the analogous groups in diamagnetic compounds. Unlike silane complexes 34a, b, Mn(II) arenetellurolate complex 26e is soluble only in coordinating solvents like other polymeric complexes.The complex 26e is extremely air-sensitive and does not sublime.53 The reaction of the tellurolate 26e with bipyridine gave the complex [(2,4,6- Me3C6H2Te)2Mn]2(bipy)3, which is substantially more soluble than the initial tellurolate. The molecular and crystal structures of Sn(II) tellurolate 22n 54 and of tellurolate complexes of Sn 33a,54 Mn 34a,52 Fe (34c and 36b) 52 and Co 35 52 were established by X-ray diffraction analysis. In the crystal, the tellurolate 22n has the dimeric structure.54 The Sn2Te2 core is nonplanar. The Sn7Te7Sn and Te7Sn7Te angles are 91.65 and 78.71 8, respectively. The ligands are in cis positions. Thus, the fact that the metal atom bears a lone electron pair is the reason for the formation of a structure which differs from that of {Zn[TeSi(SiMe3)3]2}2 (22k).49 In the compound 22k, the tellurolate ligands are in trans positions.The bridging Sn7Te bonds (2.946 and 2.956 A) are substantially longer than the Sn7Te termin bonds (2.800 A). The Te7Si bond length is 2.531 A. The average Te bridg7 Sn7Tetermin and Sn7Te bridg7Si angles are 89.27 and 107.22 8, respectively. In the complex 33a, the geometry of the bonds about the tin atom is trigonal-bipyramidal.54 The Te7Sn7Te angle (98.86 8) is somewhat larger than that in the initial tellurolate. The Te7Sn bonds (2.834 and 2.849 A) are somewhat longer than the Sn7Te termin bonds in the compound 22n. 7. Copper, silver and gold tellurolates Of copper, silver and gold tellurolates, only Cu(I) tellurolates which do not contain bulky organic radicals at the Te atom have been described previously.These compounds were prepared by the exchange reactions of sodium tellurolates with CuCl 84, 85 or by the reactions of CuCl with triphenylstannyl tellurides 38a ± e.85, 86Stable tellurols and their metal derivatives The latter method is a procedure of choice because it allows one to obtain analytically pure tellurolates 39a ± e in higher yields. RTeNa CuCl RTeCu 39a ± e Ph3SnTeR 38a ± e R=Et (a), Bu (b), C5H11 (c), Ph (d), 4-EtOC6H4 (e). The synthesis of Cu(I) benzenetellurolate 39d was performed with the use of the corresponding germanium and lead derivatives, viz., Ph3GeTePh and Ph3PbTePh, instead of Ph3SnTePh.87 The tellurolates 39a ± d are brown insoluble powders, which have apparently the polymeric structures.85, 86 Attempts to prepare Cu(I) and Ag(I) tellurolates with bulky organic radicals by the exchange reactions of metal chlorides with lithium tellurolate 5a have been unsuccessful.59 Since tris(trimethylsilyl)silanetellurolate 5a is a stronger reducing agent than the usual lithium and sodium tellurolates, it is readily oxidised by Cu(I) and Ag(I) chlorides to the ditelluride 7.[(Me3Si)3Si]2Te2+M 7 MCl+(THF)2LiTeSi(SiMe3)3 5a 7LiCl, 7THF M=Cu, Ag. However, the reactions with the use of a weaker reducing agent, viz., lithium [tris(trimethylsilyl)]methanetellurolate (5c), instead of the tellurolate 5a afforded Cu and Ag tellurolates 40a, b with the bulky organic substituent at the tellurium atom.59 The tellurolates 40a, b were synthesised by the reactions of the tellurolate 5c with tricyclohexylphosphine complexes of metal halides in toluene at room temperature and were isolated as the complexes with P(cyclo-C6H11)3.59 MX.P(cyclo-C6H11)3+(THF)2LiTeC(SiMe3)3 7LiX 5c [(Me3Si)3CTe]M . P(cyclo-C6H11)3 40a,b M=Cu, X=BF4 (a, 51%);M=Ag, X=Br (b). The synthesis of Au(I) tellurolates with sterically hindered radicals was carried out only by the exchange reactions.60 Thus free Au tellurolates 41a ± c were prepared by the reactions of AuCl.THF with lithium tellurolates 5a, c,d. The conditions for the preparation of Au(I) tellurolates are determined by the stability of the resulting compounds.Alkane-type tellurolate 41a was synthesised in benzene at room temperature, while thermally unstable analogues, viz., silane analogue 41b and germane analogue 41c, were prepared in ether at 778 8C. The latter two tellurolates were contaminated by LiCl. Attempts to purify these compounds led to their decomposition.7LiCl, 7THF AuCl .THF+(THF)nLiTeE(SiMe3)3 5a,c,d [(Me3Si)3ETe]Au 41a ± c E=C (41a, 42%), Si (41b), Ge (41c). Complexes of tellurolates 41a ± c with Ph3P were synthesised by the exchange reactions between tellurolates 5a,c,d and the AuCl triphenylphosphine complex.60 The reactions were carried out under conditions identical with those used in the reactions with free tellurolates. The difference was that alkanetellurolate complex 42a was also prepared in ether.The complex 42a was synthesised by the reaction of the tellurolate 41a with Ph3P in benzene.60 Et2O [(Me3Si)3ETe]Au . PPh3 42a ± c AuCl . PPh3+(THF)nLiTeE(SiMe3)3 5a,c,b 919 PhH 42a [(Me3Si)3CTe]4Au4+PPh3 41a E=C (42a, 65%), Si (42b), Ge (42c). The tellurolates 41a ± c are crystalline compounds, which differ substantially in thermal stabilities. Of these compounds, Au(I) alkanetellurolate 41a is most stable, it decomposes only on melting and is resistant to water and air. Tellurolates 41b, c decompose in solution even at 740 8C, which precludes their recrystallisation and purification. In the solid state, these com- pounds are resistant to water and air for at least 1 ± 3 h.60 The nature of the element bound to three trimethylsilyl groups affects the composition of thermal decomposition products of Au(I) tellurolates and their triphenylphosphine complexes.Only alkanetellurolate 41a decomposes analogously to tellurolates of other metals containing the bulky (Me3Si)3SiTe group [Zn, Cd,83 Sn, Pb,54 lanthanides(II),55 lanthanides(III) 56 and Zr 58] to form bis[tris(trimethylsilyl)methyl] telluride 43 and, apparently, gold telluride. D Au[TeC(SiMe3)3] 41a [(Me3Si)3C]2Te+AuTe 43 Decomposition of silane- and germanetellurolates proceeded in a different manner. Thus the tellurolate 41b gave compound 44 as the major product. The analogous mixed germyl silyl telluride 45 was obtained from the tellurolate 41c.D Me3SiTeSi(SiMe3)3+(Me3Si)2Te+ 44 Au[TeSi(SiMe3)3] 41b +[(Me3Si)3Si]2Te+Au 32 D Au[TeGe(SiMe3)3] 41c Me3SiTeGe(SiMe3)3+Au 45 In the 125Te NMR spectra of the tellurolates 41a ± c (see Table 2), the signals are substantially shifted upfield on going from alkanetellurolate 41a (d 7302) to silanetellurolate 41b (d 71112) and germanetellurolate 41c (d 71093). Bonasia et al. 60 attributed this fact to stronger shielding of the tellurium atom in the compounds 41b, c. The molecular and crystal structures of the tellurolate 41a 60 and its complex with triphenylphosphine 42a 60 were established by X-ray diffraction analysis. The structure of the tellurolate 41a comprises a nonplanar-square Au4Te4 core. The tellurium atoms occupy the vertices, each Te atom being bound to two Au atoms.The Au4Te4 core adopts a butterfly conformation with both wings of the butterfly being nearly coplanar. One wing comprising Te(1), Au(1), Te(2), Au(4) and Te(4) is coplanar to within 0.04A, while the second half consisting of Te(3), Au(2), Te(2), Au(3) and Te(4) is coplanar to within 0.09 A.60 The dihedral angle between these two mean planes is 144.98(6) 8. All Au7Te bond lengths are 2.56 A (average values). The Te7Au7Te fragments are almost linear; the Te7Au7Te angle is 175 8 (the average value). The Au7Te7Au angles vary in the range of 87 ± 90 8. The Te7C bond length in the tellurolate 41a (2.23 A) is virtually equal to that in the triphenylphosphine complex 42a (2.224 A).60 In the complex 42a, the Te7Au7P fragment is almost linear.The Te7Au7P and Au7Te7C angles are 174.3 and 104.1 8, respectively. The Au7Te bond length (2.566 A) is equal to that in the tellurolate 41a. The intermolecular Au7Te contacts (3.74 A) virtually coincide with the sum of the van der Waals radii of the elements (3.72 A). 8. Zirconium and hafnium tellurolates Zirconium tellurolate 46a and hafnium tellurolate 46b were prepared in 70%± 85% yields by tellurolysis of the M7C bonds of metal tetraalkyl derivatives 47 with the tellurol 4b.58920 7RH [(Me3Si)3SiTe]4M 46a,b MR4+(Me3Si)3SiTeH 4b 47 R=Bn, ButCH2; M=Zr (a), Hf (b). The tellurolates 46a,b are high-melting crystalline compounds (46a is green and 46b is red), which can be recrystallised from hexane.These compounds readily react with water and oxygen to give the tellurol 4b and the ditelluride 7.58 Heating of zirconium tellurolate 46a in vacuo from 200 to 375 8C afforded zirconium telluride ZrTe3 contaminated by admixtures of carbon and hydrogen. In this case, volatile organic thermolysis products, viz., bis[tris(trimethylsilyl)silyl] telluride (32) and tris(trimethylsilyl)silyl trimethylsilyl telluride (44), as well as nonidentified compounds containing the SiMe3 group were obtained.58 200 to 375 8C ZrTe3+[(Me3Si)3Si]2Te+ 32 [(Me3Si)3SiTe]4Zr 46a +Me3SiTeSi(SiMe3)3 44 The reactions with Lewis bases are typical of the compounds 46a,b, as of other tellurolates. Thus the reactions with 2,6- dimethylphenyl isocyanide afforded six-coordinate complexes 48a,b of 1 : 2 composition.58 The reactions with the bidentate DMPE ligand gave complexes 49a,b of 1 : 1 composition.58 RN2 C TeR1 R1Te R2N C M R1Te TeR1 (R1Te)4M 46a,b C 48a,b RN2 Te R1Te R1TeP P DMPE DMPE 46a,b M M 7R12 Te R1Te P P R1Te P TeR1 50a,b P TeR1 49a,b R1=(Me3Si)3Si, R2= 2,6-Me2C6H3; M=Zr (a), Hf (b).Treatment of the complexes 49a, b with 1 equiv. of DMPE afforded seven-coordinate complexes 50a,b in 50%± 80% yields as well as the telluride 32.58 An analogous tungsten complex 51 containing the terminal W=Te bonds was synthesised.87 H Te PMe2 PMe3 Me3P Te W (Me3P)4W 7Me3P Me3P CH2 PMe3 Te 51 Later, complexes containing the Nb=Te,88 Ta=Te 89 and Ge=Te 90 bonds were synthesised.The complexes 50a, b are red-brown crystalline air-sensitive compounds.58 In solutions, these compounds are stereochemically fluxional. The 1H and 31P NMR spectra recorded in toluene at 785 8C have sharp singlets. The signals for the SiMe3 groups are broadened as the temperature is increased to 20 8C and coalesce at a temperature above 40 8C. The 125Te NMR spectrum of the compound 50a (see Table 2) has three signals at d 7706, 71173 and 71197. The high-field signals were assigned to two non- equivalent TeSi(SiMe3)3 groups and the low-field signal was assigned to the tellurium atom of the Zr=Te bond. The structures of the tellurolate 46a and the telluride 50a 58 were established by X-ray diffraction analysis.The zirconium I D Sadekov, A V Zakharov atom in the molecule of the compound 46a is bound to four tellurium atoms and has a pseudotetrahedral coordination envi- ronment. The Te7Zr bond lengths are in the range of 2.724 ± 2.751 A and the Te7Zr7Te angles vary in a rather wide range from 101.6 to 115.9 8, which is apparently due to repulsions between the bulky (Me3Si)3SiTe groups.58 In the complex 50a, the coordination polyhedron about the zirconium atom is a distorted pentagonal bipyramid. Two DMPE ligands and one [tris(trimethylsilyl)]silanetellurolate ligand lie in the equatorial plane. The second (Me3Si)3SiTe ligand and the tellurium atom of the Zr=Te bond occupy axial positions. The Zr7TeSi(SiMe3)3 bond lengths (2.939 and 3.028 A for the axial and equatorial ligands, respectively) are substantially larger than those in the tellurolate 46a, while the Zr=Te bond is substantially shorter (2.650 A).58 9.Lanthanide tellurolates Recently, considerable attention has been focused on the synthesis of lanthanide tellurolates because these compounds can serve as precursors of lanthanide tellurides, which are of interest from the viewpoint of magnetic and luminescent properties.64 The synthesis of lanthanide tellurides from elements is performed at high temperatures and is accompanied by formation of substantial amounts of admixtures due to the high oxophilicity of rare-earth elements. In this connection, the low-temperature synthesis of lanthanide tellurides from the corresponding tellurolates, which can be readily purified according to different procedures, is of particular interest.One of the simplest tellurolates of this type, viz., Yb(II) tellurolate 9j, was synthesised by the reaction of diphenyl ditelluride with Yb in liquid ammonia.67 Since the tellurolate 9j is insoluble in usual organic solvents, it was isolated as a complex 52 after evaporation of ammonia and extraction of the residue with pyridine. The complex 52 is unstable and eliminates pyridine yielding the complex 53.67 Py NH3 (liq.) PhTeTePh +Yb (PhTe)2Yb 9j 74Py (PhTe)2Yb(Py) 53 (PhTe)2Yb(Py)5 52 An attempt to prepare Yb(III) tellurolate by the exchange reaction of YbCl3 with sodium benzenetellurolate led only to Yb(II) derivative 53.1. PhTeNa 2. Py YbCl3 Yb(TePh)2(Py)+Ph2Te2 53 The structure of the complex 52 was established by X-ray diffraction analysis.67 The coordination polyhedron about the Yb atom is a pentagonal bipyramid. The equatorial positions are occupied by five pyridine ligands and the axial positions are occupied by two benzenetellurolate ligands. The Te7Yb7Te and Yb7Te7C angles are 175.5 and 103.6 8, respectively. The average Yb7Te, Te7C and Yb7Nbond lengths are 3.282, 2.192 and 2.618 A, respectively. Lanthanide(II) tellurolates 54a ± c containing the bulky mesi- tylenetellurolate residue were synthesised by the exchange reac- tions of lanthanide(II) halides 55 with potassium 2,4,6- trimethylbenzenetellurolate.63, 64 When the reactions are carried out with the use of sodium or lithium tellurolates, lithium and sodium halides formed in the reactions cannot be completely separated from the major reaction products. Potassium bromide and iodide are readily precipitated from the reaction mixtures because they are less soluble in THF than the corresponding lithium and sodium halides.64Stable tellurols and their metal derivatives THF 7KX Ln(TeMes)2(THF)n 54a ± c LnX2(THF)2+MesTeK 55 Ln=Yb, n=2±3 (54a, 85% ± 93%); Ln=Sm, n=2 (54b, 67%); Ln=Eu, n=2, (54c, 71%), X=Br, I.The complexes of Yb(II) tellurolate with diglyme and DME (54d,e, respectively) were prepared analogously.64 YbI2L2+MesTeK 7KI Yb(TeMes)2L2 54d,e L=diglyme (d), DME (e).An alternative approach to the synthesis of lanthanide tellurolates, which makes it possible to prevent the formation of alkali metal halides, involves the reactions of lanthanide alkoxides (aryloxides) with mesityl trimethylsilyl telluride. The driving force for these reactions is the formation of Si7O bonds. For example, this procedure was used for the preparation of the tellurolate 54a.64 Yb(OMes)2(THF)3+MesTeSiMe3 7ArOSiMe3 Yb(TeMes)2(THF)2 54a The preparative usefulness of this procedure is impaired because the resulting trimethylsilyl ethers cocrystallise with, or coordinate, tellurolates 54 and are removed with difficulty. The tellurolates 54a ± e are coloured crystalline compounds (54a, b, d are orange and 54c is dark-green).These compounds are very air- and moisture-sensitive; Yb tellurolate 54a is diamagnetic and the tellurolates 54c, d are paramagnetic. The effective mag- netic moments are 3.06 and 6.95 mB, respectively.64 Pyrolysis of the complex 54a in boiling toluene at 250 ± 300 8C in vacuo afforded dimesityl telluride and a black insoluble precipitate. Annealing of the precipitate at 900 8C for 24 h gave cubic YbTe.63 YbTe +Mes2Te (MesTe)2Yb 54a Representatives of lanthanide(II) tellurolates of yet another type, viz., bis[tris(trimethylsilyl)silane]tellurolates 55a ± c, were synthesised by a slightly modified procedure of tellurolysis of silylamides of the corresponding metals 56a ± c. The reaction in diethyl ether in the presence of an excess of TMEDA afforded six- coordinate complexes 55a ± c in 40%± 63% yields.55 Et2O, 20 8C 7(Me3Si)2NH Ln[N(Me3Si)2]2+HTeSi(SiMe3)3+TMEDA 4b 56a ± c [(Me3Si)3SiTe]2Ln(TMEDA)2 55a ± c Ln=Yb (a), Sm (b), Eu (c).The reaction performed with DMPE as a Lewis base yielded a seven-coordinate dimer 57 55 containing the bridging DMPE ligand. The formation of the Eu(II) complex with the coordina- tion number as large as seven is a consequence of the longer Eu7P bond compared to the Eu7N bond.55 Et2O, 20 8C 7(Me3Si)2NH Eu[N(Me3Si)2]2+HTeSi(SiMe3)3+DMPE 4b {[(Me3Si)3SiTe]2Eu(DMPE)2}2(m-DMPE) 57 (50%) In the reaction with Eu(III) silylamide, the tellurol 4b, which is a strong reducing agent, converted Eu(III) into Eu(II).The tellurol 4b also reduced Yb(III) and Sm(III) silylamides.55 The tellurolates 55a ± c are deeply coloured air- and moisture- sensitive crystalline compounds. After storage under nitrogen at room temperature for 12 h, these compounds begin to decom- pose. At 740 8C, these compounds remain unchanged within 921 several weeks.55 The large magnetic moments of the complexes 55c and 57 (*8.0 mB) preclude the use of NMR spectroscopy. Like most of trimethylsilanetellurolates of other metals, the tellurolate 55a decomposes upon heating (150 ± 200 8C, 1072 Torr) to form the telluride 32 and YbTe contaminated by a carbon admixture.55 D 7TMEDA YbTe+[(Me3Si)3Si]2Te 32 [(Me3Si)3SiTe]2Yb(TMEDA)2 55a The molecular and crystal structure of the tellurolate 57 was established by X-ray diffraction analysis.55 Unlike ytterbium, samarium and europium tris[di(trimethyl- silyl)amides], which gave lanthanide(II) derivatives upon tellurol- ysis,55 La(III) and Ce(III) silylamides were converted into lantha- nide(III) tellurolates upon treatment with 3 equiv.of the tellurol 4b 56 as evidenced by the data of 1Hand 125TeNMRspectroscopy of the reaction mixtures. However, free La(III) and Ce(III) tellurolates 58a ,b are thermally unstable. Because of this, these compounds are isolated as complexes with DMPE 59a ,b, which are formed upon addition of an excess of the ligand to the reaction mixture. The yields of the complexes 59a, b were no higher than 75%.56 Ln[N(Me3Si)2]3+(Me3Si)3SiTeH 7(Me3Si)2NH 4b DMPE [(Me3Si)3SiTe]3Ln(DMPE)2 59a,b [(Me3Si)3SiTe]3Ln 58a,b Ln=La (a),Ce (b).Crystalline complexes 59 are coloured, viz., 59a is yellow and 59b is orange. The complex 59a is diamagnetic and the complex 59b is paramagnetic (meff=2.28 mB). The data of dynamic multi- nuclear spectroscopy (1H, 13C, 31P and 125 Te) are indicative of the stereochemical fluxionality of the compounds 59a, b in solutions. Thus the 1H NMR spectrum of the complex 59a at 23 8C has a broad singlet of three equivalent tellurolate ligands and broad peaks of the methyl and methylene groups of the phosphorus- containing ligands. When the temperature is decreased to 784 8C, the tellurolate ligands give two peaks with the intensity ratio of 2 : 1 and the methyl groups of DMPE give two broad peaks with equal intensities.The 125Te NMR spectra at 770 8C have two signals with the intensity ratio of 2 : 1. An analogous temperature dependence is also typical of the 13C and 31P NMR spectra.56 As mentioned above, lanthanide(III) tellurolates in the absence of donor ligands decomposed to form bis[tris(tri- methyl)silyl] telluride 32 and a complex mixture of products from which compounds 60a, b were isolated in low yields. The metal ± tellurium clusters 60a, b were considered 56 as intermedi- ates in the pyrolytic conversions of the tellurolates 59a,b into lanthanide tellurides. Pyrolysis of the compounds 59a, b and 60a, b in vacuo at 600 8C afforded lanthanide tellurides.56 600 8C 20 8C Ln5Te3[TeSi(Me3Si)3]9 60a,b Ln[TeSi(Me3Si)3]3(DMPE)2 59a,b Ln2Te3 The structures of the compounds 59a and 60b were established by X-ray diffraction analysis.56 The La atom in the complex 59a has a pseuodopentagonal-bipyramidal environment.Both phos- phine ligands and one tellurolate ligand lie in the equatorial plane and two other tellurolate ligands are located in the axial plane. The La7Te eq bond length is 3.141 A and the La7Te ax bond lengths are 3.168 and 3.170 A, respectively. The Te ax7La7Teax angle is 145.7 8. Steric interactions between the equatorial tellurolate and phosphine ligands result in the difference in the La7P bond lengths (3.127, 3.143 A and 3.208, 3.216 A).922 [Te7Ce ] 3 The molecule of the compound 60b contains a six-membered ring with the m3-bridging Ce(TeR)3 fragments located above and below this ring.The Ce7Te termin bond lengths are smaller (3.026 A) than the Ce7Te bridg bond lengths (3.183 and 3.235 A). The Ce3Te3 bond lengths vary from 3.124 to 3.278 A. IV. Conclusion As is evident from the present review, the synthesis of new metal tellurolates and their pyrolysis to the corresponding tellurides are of prime interest. It is in this field that the progress should be expected. Yet another line of investigation involves the synthesis of new stable tellurols. The stability of aromatic tellurols is determined by steric factors. Two-coordinate tellurium deriva- tives can be stabilised also by introducing substituents, such as COR, CH=N or N=N, which can form intramolecular coordi- nation bonds with the tellurium atom, at the ortho position with respect to the tellurium-containing function.This method has not been applied in the chemistry of tellurols so far. Apparently, the synthesis of stable tellurols and intracomplex metal compounds based on these tellurols will be developed and will be the subject of further studies. References 1. G Natta Giorn. Chim. Ind. Applicata 8 367 (1926); Chem. Abstr. 20 3273 (1926) 2. MBochmann, A P Coleman, K J Webb, MB Hursthouse, MMazid Angew. Chem. 103 975 (1991) 3. U Siemeling Angew. Chem., Int. Ed. Engl. 32 67 (1993) 4. J Arnold Progr. Inorg. Chem. 43 353 (1995) 5. T Chivers J. Chem. Soc., Dalton Trans.1185 (1996) 6. H B Singh, N Sudha Polyhedron 15 745 (1996) 7. S V Larionov, S M Zemskova Ross. Khim. Zh. 40 171 (1996) a 8. A Baroni Atti Accad. Naz. Lincey, Rend., Cl. Sci., Fis. Mat. Nat. 27 238 (1938); Chem. Abstr. 33 163 (1939) 9. C W Sink, A B Harvey J. Chem. Soc., D 1023 (1969) 10. K Hamada, H Morishita Synth. React. Inorg. Met.-Org. Chem. 7 355 (1977) 11. C W Sink, A B Harvey J. Chem. Phys. 57 4434 (1972) 12. M Akiba,M P Cava Synth. Commun. 14 1119 (1984) 13. J E Drake, R T Hemmings Inorg. Chem. 19 1879 (1980) 14. K Nagakawa, M Osuka, K Sasaki Chem. Lett. 1331 (1987) 15. N Ohira, Y Aso, T Otsubo, F Ogura Chem. Lett. 853 (1984) 16. M N Bochkarev, L P Sanina, N S Vyazankin Zh. Obshch. Khim. 39 135 (1969) b 17. B O Dabbousi, P J Bonasia, J Arnold J.Am. Chem. Soc. 113 3186 (1991) 18. P J Bonasia, D E Gindelberger, B O Dabbousi, J Arnold J. Am. Chem. Soc. 114 5209 (1992) 19. P J Bonasia,V Christou, J Arnold J. Am. Chem. Soc. 115 6777 (1993) 20. G Becker, K W Klinkhammer, S Lartiges, P Bottcher, W Poll Z. Anorg. Allg. Chem. 613 7 (1992) 21. G Gutekunst, A G Brook J. Organomet. Chem. 225 1 (1982) 22. G Becker, H M Hartmann, A Munch, H Riffel Z. Anorg. Allg. Chem. 530 29 (1985) 23. F Sladky, B Bildstein, C Rieker, A Gieren, H Betz, T HuÈ bner J. Chem. Soc., Chem. Commun. 1800 (1985) 24. D E Gindelberger, J Arnold Organometallics 13 4462 (1994) 25. M Ballestri, C Chatgilialoglu, G Seconi J. Organomet. Chem. 408 C1 (1991) 26. B Bildstein, K Giselbrecht, F Sladky Chem.Ber. 122 2279 (1989) 27. N N Greenwood, A Earnshaw, in Chemistry of the Elements (Oxford: Pergamon Press, 1984) p. 900 28. H Puff, K Braun, H Reuter J. Organomet. Chem. 409 119 (1991) 29. R Minkwitz, A Kornath, H Preut Z. Anorg. Allg. Chem. 619 877 (1993) 30. L Pauling The Nature of the Chemical Bond (Ithaka, NY: Cornell University Press, 1960) 31. C Glidewell, D W H Rankin, G M Sheldrick J. Chem. Soc., Trans. Farad. Soc. 1409 (1969) I D Sadekov, A V Zakharov 32. W R McWhinnie, in The Chemistry of Organic Selenium and Tellurium Compounds (Ed. S Patai) (New York: Wiley, 1987) Vol. 2, Part 13 33. K J Irgolic The Organic Chemistry of Tellurium (New York: Gordon and Breach, 1974) 34. I D Sadekov, A A Maksimenko, V I Minkin Khimiya Tellurorga- nicheskikh Soedinenii (The Chemistry of Organic Tellurium Com- pounds) (Rostov-on-Don: Rostov-on-Don State University, 1983) 35.I D Sadekov, B B Rivkin, A A Maksimenko, E I Sadekova Sulfur Rep. 17 1 (1995) 36. I D Sadekov, V I Minkin Sulfur Rep. 19 285 (1997) 37. Y Okamoto, T Yano J. Organomet. Chem. 29 99 (1971) 38. M L Steigerwald, C R Sprinkle J. Am. Chem. Soc. 109 7200 (1987) 39. N S Dance, C H W Jones J. Organomet. Chem. 152 175 (1978) 40. D C Harris, R A Nissan, K T Higa Inorg. Chem. 26 765 (1987) 41. J G Brennan, T Siegrist, P J Carroll, S M Stuczynski, P Reynders, L E Brus,M L Steigerwald Chem. Mater. 2 403 (1990) 42. H L Paige, J Passmore Inorg. Nucl. Chem. Lett. 9 277 (1973) 43. J Kischkewitz, D Naumann Z.Anorg. Allg. Chem. 547 167 (1987) 44. U Behrens, K Hoffmann, G Klar Chem. Ber. 110 3672 (1977) 45. J Liesk, G Klar Z. Anorg. Allg. Chem. 435 103 (1977) 46. D E Gindelberger, J Arnold J. Am. Chem. Soc. 114 6242 (1992) 47. D E Gindelberger, J Arnold Inorg. Chem. 33 6293 (1994) 48. G Becker, KW Klinkhammer, WSchwarz,MWesterhausen, T Hildebrand Z. Naturforsch., B Chem. Sci. 47 1225 (1992) 49. P J Bonasia, J Arnold Inorg. Chem. 31 2508 (1992) 50. M Bochmann, G C Bwembya, A K Powell, X Song Polyhedron 14 3495 (1995) 51. M Bochmann, K J Webb J. Chem. Soc., Dalton Trans. 2325 (1991) 52. D E Gindelberger, J Arnold Inorg. Chem. 32 5813 (1993) 53. M Bochmann, A K Powell, X Song J. Chem. Soc., Dalton Trans. 1645 (1995) 54. A L Seligson, J Arnold J. Am. Chem. Soc. 115 8214 (1993) 55. D R Cary, J Arnold Inorg. Chem. 33 1791 (1994) 56. D R Cary, J Arnold J. Am. Chem. Soc. 115 2520 (1993) 57. S P Wuller, A L Seligson, G P Mitchell, J Arnold Inorg. Chem. 34 4854 (1995) 58. V Christou, J Arnold J. Am. Chem. Soc. 114 6240 (1992) 59. P J Bonasia, G P Mitchell, F J Hollander, J Arnold Inorg. Chem. 33 1797 (1994) 60. P J Bonasia,DE Gindelberger, J Arnold Inorg. Chem., 32 5126 (1993) 61. P J Bonasia, J Arnold J. Organomet. Chem. 449 147 (1993) 62. I D Sadekov,G M Abakarov,S G Kuren',A D Garnovskii,V I M- inkin Zh. Obshch. Khim. 56 2168 (1986) b 63. A R Strzelecki, P A Timinski, B A Helsel, P A Bianconi J. Am. Chem. Soc. 114 3159 (1992) 64. A R Strzelecki, C L Likar, B A Helsel, T Utz,M C Lin, P A Bian- coni Inorg. Chem. 33 5188 (1994) 65. W W DuMont, R Hensel, S Kubiniok, L Lange, T Severengiz Phosphorus Sulfur Silicon Relat. Elem. 38 85 (1988) 66. P J Bonasia, J Arnold J. Chem. Soc., Chem. Commun. 1299 (1990) 67. M Brewer, D Khasnis, M Buretea, M Berardini, T J Emge, J G Brennan Inorg. Chem. 33 2743 (1994) 68. L Lange, W-W Du Mont J. Organomet. Chem. 286 C1 (1985) 69. S Schulz, H W Roesky, H J Koch, G M Sheldrick, D Stalke, A Kuhn Angew. Chem. 105 1828 (1993) 70. W Uhl, R Graupner, M Pohlmann, S Pohl,W Saak Chem. Ber. 129 143 (1996) 71. A H Cowley, R A Jones, P R Harris, D A Atwood, L Contreras, C J Burek Angew. Chem., Int. Ed. Engl. 30 1143 (1991) 72. M B Power, J W Ziller, A N Tyler, A R Barron Organometallics 11 1055 (1992) 73. I D Sadekov, V I Minkin Adv. Heterocycl. Chem. 58 48 (1993) 74. C Kollemann, D Obendorf, F Sladky Phosphorus Sulfur Silicon Relat. Elem. 38 69 (1988) 75. V Christou, S P Wuller, J Arnold J. Am. Chem. Soc. 115 10545 (1993) 76. R D Shannon Acta Crystallogr., Sect. A 32 751 (1976) 77. F H Allen, O Kennard, D G Watson, L Brammer, A G Orpen, R Taylor J. Chem. Soc., Perkin Trans. 2 S1 (1987) 78. R Kaur, H B Singh, R J Butcher Organometallics 14 4755 (1995) 79. H B Singh, N Sudha, R J Butcher Inorg. Chem. 31 1431 (1992) 80. T J Groshens, R W Gedridge, C K Lowe-Ma Chem. Mater. 6 727 (1994)923 Stable tellurols and their metal derivatives 81. S M Stuczynski, J G Brennan, M L Steigerwald Inorg. Chem. 28 4431 (1989) 82. M Bochmann, K J Webb, J E Hails, D Wolverson Eur. J. Solid State Inorg. Chem. 29 155 (1992) 83. J Arnold, J M Walker, K M Yu, P J Bonasia, A L Seligson, E D Bourret J. Cryst. Growth 124 647 (1992) 84. J Davies, W R McWhinnie Inorg. Nucl. Chem. Lett. 12 763 (1976) 85. J Davies, W R McWhinnie, N S Dance, C H W Jones Inorg. Chim. Acta 29 L217 (1978) 86. S A Gardner, P J Trotter, H J Gysling J. Organomet. Chem. 212 35 (1981) 87. D Rabinovich, G Parkin J. Am. Chem. Soc. 113 9421 (1991) 88. U Siemeling, V C Gibson J. Chem. Soc., Chem. Commun. 1670 (1992) 89. V Christou, J Arnold Angew. Chem. 105 1550 (1993) 90. M C Kuchta, G Parkin J. Chem. Soc., Chem. Commun. 1351 (1994) a�Mendeleev Chem. J. (Engl. Transl.) b�Russ. J. Gen. Chem. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (11) 925 ± 936 (1999) Current problems of the electrochemistry of titanium and boron. Synthesis of titanium diboride and titanium intermetallic compounds in ionic melts V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova Contents I. Introduction II. Electrochemical behaviour of titanium III. Electrochemical behaviour of boron IV. Methods for the synthesis of titanium diboride. High-temperature electrochemical synthesis V. Electrochemical properties and stability of halide systems containing titanium in various oxidation states and boron VI. Electrochemical synthesis of titanium diboride and titanium intermetallic compounds in a chloride ± fluoride melt Abstract. The current state of research into the titanium and boron electrochemistry and high-temperature electrochemical synthesis of titanium diboride and titanium intermetallic com- pounds with the iron triad metals is analysed.The prospects for the development of these studies are discussed. Primary attention is paid to the electrochemical properties and stability of halide systems containing titanium in various oxidation states together with boron. The bibliography includes 81 references. I. Introduction Studies on the electrochemical properties of titanium and boron and the processes of joint electroreduction of two or more components from ionic melts present interest for both theoretical and practical purposes. On the one hand, the results of these studies help one to understand characteristic features of multi- electron processes and to follow the development of high-temper- ature electrochemical synthesis; on the other hand, they underlie the development of new technologies for the production of nontraditional materials with unique properties.1±3 Despite the numerous publications devoted to the electrochemistry of tita- nium and boron as well as titanium diboride, data on the forms of their existence, the mechanisms of joint and separate electro- reduction in ionic melts (especially chloride and fluoride melts) are quite scarce and contradictory.These processes serve as the basis for the technology of the electrochemical production of titanium, boron, titanium diboride and titanium intermetallic compounds; therefore, studies in this field are topical for the development of modern high-temperature electrochemistry. II.Electrochemical behaviour of titanium Electrolysis of titanium melts is a promising method for the production of metallic titanium and the synthesis of high-melting compounds (carbides, borides, silicides, intermetallic compounds) based on it. However, the complex pattern of the behaviour of V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova V I Vernadskii Institute of General and Inorganic Chemistry, National Academy of Sciences of Ukraine, prosp. Palladina 32/34, 252680 Kiev, Ukraine. Fax (38-044) 444 30 70. Tel. (38-044) 444 01 11 (V I Shapoval). E-mail: postmaster@ionc.kiev.ua Received 16 February 1999 Uspekhi Khimii 68 (11) 1015 ± 1028 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 541.135:621.357:546.271 925 925 928 929 930 933 titanium-containing melts accounts for the stringent requirements in the process engineering and selection of the salt systems.Thus it is necessary to study systematically and comprehensively the electrochemical and chemical properties of titanium in various oxidation states in molten electrolytes. 1. Electrolysis of titanium-containing melts Titanium metal is mainly produced either by electrolysis of the melts of titanium oxygen compounds, fluorides or chlorides or by electrolytic refinement of crude titanium and its compounds. Electrolysis of oxygen compounds of titanium has several advantages over electrolysis of other compounds, first of all, ready availability and low cost of the raw materials, namely, titanium dioxide and alkali metal titanates.These advantages also include simplicity of the equipment and the environmental safety of the process. Numerous publications have been devoted to the elec- trolysis of TiO2 in melts of alkali metal chlorides or fluorides or potassium (sodium) hexafluorotitanate. The low solubility of titanium dioxide in various melts (0.02% ± 0.07%) precludes its use for the production of titanium. In addition, during electrolysis of the NaCl7K2TiF6 melt containing titanium dioxide, the dioxide is reduced at the cathode to Ti2O3. This is accompanied by the formation of a non-decomposing precipitate near the cathode.Electrolytic preparation of pure titanium in this system is impossible because of the formation of titanium monoxide at the cathode. Study of a similar system based on calcium chloride showed that in this case, titanium metal is formed in a secondary process, namely, reduction of TiO2 at the cathode by calcium produced in electrochemical processes. Normal operation of the electrolytic cell requires a low concentration of TiO2 in the electrolyte, its continuous feed to the cell and replacement of the electrolyte after it has been saturated with CaO. Electrolysis of TiO2 dissolved in molten alkali metal borates or phosphates has been considered. 4, 5 The researchers point to the necessity of additional purification of the resulting metal from oxygen.Thus, the use of oxygen compounds for the electrolytic preparation of titanium and the synthesis of its derivatives is hardly promising, because lower titanium oxides, which contam- inate the product, are formed at the cathode. Calculations and experimental results show that electrolysis in molten salts of oxygen-containing acids, alkalis, oxides or many other melts does not ensure the preparation of pure titanium. The range of molten media is mainly limited to metal halides. However, these compounds, too, are not all suitable for electro- lysis. Indeed, metals of Group IV and of subsequent Groups of the926 Periodic Table are usually much more electropositive than titanium; hence, halides of these metals are, most often, thermally unstable or possess low electrical conductivity.Alkali and alkaline earth metal chlorides and fluorides are best suited for this purpose. Numerous studies have been devoted to the development of a method for the production of titanium by electrolysis of fluoride melts. Potassium and sodium fluorotitanates are recognised as the most promising titanium raw materials.5±9 It is not expedient to subject pure hexafluorotitanate to electrolysis. During electroly- sis, it is reduced to K2TiF5 and, after that, the electrodes become passivated. The system subjected to electrolysis is usually K2TiF6 dis- solved in molten salts. An electrolyte consisting of potassium, lithium and sodium fluorides has been proposed as the medium for the electrolytic production of titanium.6 However, a process carried out in a purely fluoride melt can result in passivation of the anode by a gas film (anodic effect).In addition, since this melt is highly corrosive, the equipment soon goes out of service. There- fore, a mixture of alkali metal chlorides and potassium fluoroti- tanate has been proposed as the electrolyte for the electrolysis of titanium fluoride derivatives. 5±9 When titanium metal reacts with a chloride ± fluoride melt, a black crust insoluble in a mixture of alkali metal chlorides and fluorides is formed gradually on its surface. The existence of a fluoride complex of divalent titanium, M2[TiF4], has been reported. 8 Insoluble compounds formed at the cathode contaminate the cathodic product, which casts doubt on the expediency of using chloride ± fluoride ± fluorotitanate melts for producing pure titanium.Preparation of titanium from titanium chlorides presents substantial interest. Titanium, having a variable valence, forms several chlorides, namely, TiCl4, TiCl3 and TiCl2. The most readily available compound is TiCl4; the production and purifi- cation of this compound on an industrial scale is now a routine practice. In the electrolysis of alkali or alkaline earth metal chlorides, TiCl4 is either supplied onto the electrolyte surface or bubbled through the electrolyte, either neat or with a flow of an inert gas (see, for example, Refs 10, 11). Titanium tetrachloride is poorly soluble in molten alkali metal chlorides (the solubility is up to 2%at 1073 K); therefore, this technology does not provide high output.If the current density at the cathode is high, the alkali metal will be evolved together with titanium. It has been proposed to make use of this phenomenon, i.e. to carry out electrolysis at current densities sufficient for the evolution of the alkali metal. In this case, titanium tetrachloride is partially reduced by the reducing metal. The solubility of titanium chlorides in which the metal oxidation number is less than four is substantially higher than the solubility of titanium tetrachloride. This fact has been taken into account in the development of a method for the preparation of titanium by electrolysis of its lower chlorides dissolved in chloride melts.10 Salts used in this process should be relatively pure as regards oxygen and elements that are more electropositive than titanium.Lower titanium chlorides can easily be prepared from titanium tetrachloride; they are readily soluble in molten salts and fairly stable at temperatures of electrolysis. These compounds possess high electrical conductivities and relatively low volatilities; however, they are chemically reactive. Under conditions of electrolysis, lower titanium chlorides react with almost any standard refractory lining material; evidently, the rate of such a reaction would depend substantially on the presence of oxygen in the cell. A method has been proposed 10 for continuous production of titanium by electrolysis of TiCl3 and TiCl2, which are formed in a separate reactor upon the reaction of titanium tetrachloride with hydrogen.To simplify this method, the synthesis of lower titanium chlorides and electrolysis can be carried out in one vessel. In this case, electrolysis occurs in two stages, namely, reduction of TiCl4 to TiCl2 and TiCl3 at an auxiliary cathode and reduction of lower titanium chlorides to the metal on the main cathode.11 V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova It has been considered until recently that the cathodic process is a secondary reaction, consisting in the reduction of titanium by the alkali metal, which was evolved at the cathode. However, recent studies demonstrated that, under certain conditions, direct electrochemical discharge of the titanium ions occurs at the cathode. For this discharge to be possible, it is required, first of all, that the concentration of titanium ions in low oxidation states in the near-cathode layer be sufficiently high.Electrolysis of titanium chlorides is carried out under the atmosphere of an inert gas (argon, helium). For rough purification of the electrolyte from moisture and occluded gases, the electro- lytic cell is usually evacuated at 400 ± 650 8C to a residual pressure of 10 ± 20 mm Hg. Preliminary electrolysis at an auxiliary cathode with a voltage on the cell terminals equal to 1.8 ± 2.0 V is also employed for this purpose. Other methods for the purification of the electrolyte also exist.To remove moisture, dry hydrogen hydrochloride is passed through the melt for a short period of time. Oxygen can be removed by passing tetrachloromethane through the melt. The cathode materials used in these processes are mainly tungsten, molybdenum, silver, steel or nickel. The best adhesion of the titanium deposits has been noted for silver cathodes; however, this metal is expensive. Liquid cathodes made of zinc or lead can hardly find wide application because they form intermetallic compounds with titanium, which are difficult to destroy. During electrolysis of titanium chlorides, chlorine formed at the anode oxidises lower titanium chlorides present in the anode chamber to give TiCl4, which is distilled off together with chlorine.In addition, electrolytic oxidation of TiCl2 and TiCl3 at the anode, leading to the same consequences, is possible. Meanwhile, titanium tetrachloride entering the cathode chamber would react with titanium metal, which would result in the loss of the pure metal. Therefore, the cathode and anode chambers should be separated by a diaphragm. The problem of selection of the diaphragm materials has not been solved yet, due to the high chemical reactivity of lower titanium chlorides; research along this line is now in progress. For instance, it has been proposed to manufacture the diaphragm from alumina (these diaphragms are not virtually metallised under certain conditions) or graphite (in this case, half-sintered metal is formed at the cathode at 1305 8C).4, 5 It should also be noted that working with titanium chloride requires a high voltage at the electrolytic cell terminals (usually 9 ± 12 V); therefore, power expenditure per unit product is high.Whereas the expediency of one or another method for electro- lytic production of titanium is still debated and electrolysis has not been adopted on an industrial scale, the efficiency of electrolytic refining of titanium is beyond doubt. The soluble anode can be manufactured from any material possessing electronic conduc- tion, capable of reacting with the electrolyte and insoluble in it. Titanium forms lots of systems and compounds with such proper- ties.Powders with various chemical and grain-size compositions are currently produced from titanium by electrolysis with a soluble anode in the melts of alkali and/or alkaline earth metal chlorides.Electrolytic refinement of titanium is carried out using various electrolytes, namely, NaCl, NaCl7KCl, NaCl7SrCl2, NaCl7BaCl2, KCl7LiCl and KCl7NaCl7MgCl2. Electrolytes based on sodium chloride, an equimolar mixture of sodium and potassium chlorides and the NaCl7KCl7MgCl2 ternary system are the most widespread. The electrolytic refinement of titanium consists in the anodic dissolution of the metal to be refined and cathodic deposition of the pure metal. Electrochemical investiga- tions show that in the temperature range 750 ± 8508C, which is the best for the refinement, titanium passes from the anode to the melt as Ti2+.An increase in the temperature or in the anode current density increases the yield of the Ti3+ ions. Upon anodic dissolution, titanium passes into the melt mostly in an ionic form, the average oxidation state being 2.02 ± 2.13.Current problems of the electrochemistry of titanium and boron. The total content of impurities, which are to be eliminated during titanium refinement (oxygen, nitrogen, hydrogen, carbon, iron) is normally relatively low, 1 mass %¡¾ 2 mass %. In the studies dealing with the electrochemical refinement of titanium, special role belongs to the purification from oxygen. When the process is carried out under precisely pure conditions, the only way by which oxygen can pass from the anode to the cathode is the formation of the titanyl ion TiO2+.Thermodynamic calculations indicate that the probability of formation of this ion is fairly low even at high concentrations of oxygen in the anode material. A significant role in the production of a high-quality powder at the cathode is played by the ratio of the titanium concentration in the electrolyte to the current density at the cathode. The best powders are obtained if this ratio ranges from 2 to 3. The anode current density during electric refinement should be 4 ¡¾ 8 times lower than the cathode current density. In this case, divalent titanium disproportionates in the electrolyte (especially in the near-anode layer), resulting in the formation of highly dispersed metallic titanium in the bulk of the melt.The use of the refining tank for the synthesis of titanium borides is not expedient, due to the formation of an uncontrollable amount of dispersed titanium, which disrupts the stoichiometry of the product. Various opinions concerning the role of fluoride ions have been expressed in studies dealing with the electrolytic refinement of titanium in mixed chloride ¡¾ fluoride electrolytes (see, for example, Ref. 12). Some researchers believe that the addition of fluorides has a negative effect on the electrolysis characteristics (the anode material disintegrates, the current yield decreases). Conversely, some other investigators believe that fluoride addi- tives are favourable for the performance characteristics of the process. It should be noted that anodic dissolution of titanium in a mixed chloride and fluoride electrolyte affords poorly soluble compounds of divalent titanium, which contaminate the cathodic product.The foregoing demonstrates that halide melts are the most suitable systems for the high-temperature electrochemical syn- thesis of titanium compounds and production of the pure metal. However, when using chloride melts, one should create conditions for increasing the solubility of TiCl4 and suppressing dispropor- tionation, and when mixed chloride ¡¾ fluoride electrolytes are used, the formation of poorly soluble divalent titanium fluoride should be prevented and the electroreduction of titanium complex ions should be performed in one stage. 2. State of titanium in halide melts a.Thermodynamics of titanium-containing halide systems Titanium halides react with titanium metal in molten alkali metal halides. The following heterogeneous equilibrium is established with time: (1) 3 TiCl2 . 2 TiCl3+Ti As a rule, one oxidation state of titanium predominates in the melt; the ratio of various oxidation states depends on the temper- ature and the nature of the salt acting as the solvent. Heteroge- neous equilibrium (1) can be characterised by several parameters such as equilibrium constant, apparent standard potentials E*(Ti2+/Ti), E*(Ti3+/Ti) and the formal redox potential E*(Ti3+/Ti2+).13, 14 The equilibrium potential of the titanium electrode (E) would be simultaneously the potential of the Ti/ TiCl2, Ti/TiCl3 and TiCl2/TiCl3 systems.The apparent standard potentials and the formal redox potential are related (in accord- ance with the Luter rule) by the equation 13 (2) 3E*(Ti3+/Ti)=E*(Ti3+/Ti2+) + 2E*(Ti2+/Ti) . The task of finding these values reduces to the determination of the number of ionic forms of titanium in the potential measure- ment. The equilibrium potentials of titanium in an equimolar mixture of potassium and sodium chlorides in the temperature 927 range 973 ¡¾ 1148 K have been reported. 14 The temperature variations of E*(Ti2+/Ti) and E*(Ti3+/Ti) are described by the following empirical equations: (3) E*(Ti2+/Ti)=72.38 + 4.861074 T, (4) E*(Ti3+/Ti)=72.16 + 3.261074 T. The equilibrium potentials of titanium measured at 843 K vs.chlorine reference electrode amount to E*(Ti2+/Ti)=71.97 V and E*(Ti3+/Ti)=71.90 V. In order to identify the full picture of equilibrium between metallic titanium and a chloride melt containing minor amounts of Ti4+, in addition to Ti2+ and Ti3+, the E*(Ti4+/Ti3+) redox potential, and, hence, the standard E*(Ti4+/Ti) potential should be determined. The temperature variation of this potential in the KCl7NaCl melt is described by the equation (5) E*(Ti4+/Ti)=71.91 + 3.8461074 T. Various values for the arbitrary formal redox potential E*(Ti3+/Ti2+) in a molten equimolar mixture of sodium and potassium chlorides, determined experimentally by potentiomet- ric titration, have been reported in the literature: 71.807 (T=973 K), 71.762 (T=943 K), 71.726 V (T=984 ¡¾ 1130 K).The E*(Ti3+/Ti2+) value calculated in accordance with the Luter rule from Eqn (2) using the known experimental E*(Ti3+/Ti) and E*(Ti2+/Ti) values is described by the equation (6) E*(Ti3+/Ti2+)=71.71070.1761074 T. It can be seen that the theoretical value of the potential E*(Ti3+/Ti2+) depends slightly on the temperature and at 1000 K it is equal to71.727 V vs. chlorine reference electrode. The known standard potentials for the Ti3+/Ti and Ti4+/Ti systems were used to calculate the Ti4+/Ti3+ redox potential and to elucidate its temperature dependence (7) E*(Ti4+/Ti3+)=(71.473 + 5.8661074 T)0.007 . The variation of the equilibrium potential of titanium as a function of the concentration of titanium ions in the melt have shown that in alkali metal chloride melts, titanium is equilibrated with the Ti2+ and Ti3+ ions.Knowing the standard electrode potentials of the Ti2+/Ti and Ti3+/Ti systems, it can be written 15 (8) logK a log aTi2aa3 aTi3aa2 a 61:98410T4 [E*(Ti3+/Ti)7E*(Ti2+/Ti)] Thus the equilibrium constants for the reaction (9) 3Ti2+. 2Ti3+ +Ti can be calculated. For lithium chloride, (10) logK=74.234 + 6351 , T for potassium chloride, (11) logK=76.956 + 8165 , T for cesium chloride, (12) logK=76.956 + 7440 , T and for the KCl7NaCl melt, (13) logK=75.038 + 6774 . T As the temperature increases, the equilibrium constant decreases, which points to disproportionation of Ti2+. If the temperature remains constant, the equilibrium constant decreases with an increase in the radius (r) of the alkali metal cation of the molten salt, i.e.the equilibrium shifts towards disproportionation of Ti2+. This dependence has the form (14) logK a 8468 1512 T ¢§ Tr ¢§ 9:073 ¢§ 3:175 r .928 A substantial number of studies have been devoted to the process of disproportionation (see, for example, Refs 16 ± 18). The oxidation states of titanium can be described by four dispropor- tionation reactions (15) 3Ti2+, 2Ti3+ + Ti (16) Ti2+ + Ti4+, 2Ti3+ (17) Ti + 3 Ti4+, 4Ti3+ (18) Ti+Ti4+. 2Ti2+ These reactions occur simultaneously and are heterogeneous. The outcome depends on the experimental conditions, namely, temperature, composition of the atmosphere, surface area, and so on.It has also been noted that Ti3+ rapidly disproportionates in the lithium chloride melt. When lithium chloride is replaced by cesium chloride, the stability of Ti3+ increases. In halide melts, titanium metal reacts with the components of the melt. When the radius of the alkali metal cation in the melt is greater than the radius of potassium, Ti2+ ions appear; they disproportionate to give highly dispersed titanium metal; therefore, a high concen- tration of Ti2+ cannot be attained. In the case where the LiCl melt is used, the solution becomes rich in Ti2+. The main reaction which proceeds at a substantial rate in molten potassium chloride is (19) 3 TiCl2 (liq). 2 TiCl3 (liq) + Ti(solid) Thus, vast experimental data have been accumulated and a number of important practical conclusions have been drawn concerning the thermodynamics of titanium-containing chloride melts, namely, the equilibrium and standard potentials of tita- nium in various oxidation states were determined; the equilibrium constants for chemical reactions in the melts were calculated; and disproportionation of titanium ions in low oxidation states was detected.However, in some cases, the thermodynamic functions are rough. For example, the measurement of the equilibrium potentials of the Ti4+/Ti, Ti3+/Ti and Ti2+/Ti systems does not appear quite correct because actually, the near-electrode layer would contain titanium ions in various oxidation states formed in chemical reactions.In addition, thermodynamic characteristics do not provide exhaustive information on the mechanism of electro- reduction of titanium, which needs to be known for performing successful electrolysis of titanium-containing melts. Therefore, in addition to the thermodynamic characteristics, the mechanism of the electroreduction of titanium has also been studied. b. Mechanism of the electroreduction of titanium It is not always possible to interpret unambiguously experimental data characterising the mechanism and kinetics of electroreduc- tion and electrooxidation of titanium complexes. The complexity of interpretation of these data is due to the fact that side processes, not related directly to electrode reactions, proceed in the melt.In some studies, it is suggested that titanium metal is evolved in the reaction of titanium-containing complex ions with an alkali metal, resulting from reduction at the cathode. In other studies, it is suggested that two processes of titanium deposition occur simul- taneously; titanium is formed in the primary electrochemical process at the cathode Ti0 Ti2+ + 2e and in a secondary chemical process, thermal reduction by sodium in the bulk of the electrolyte. Some researchers believe that in the case of chloride melts, an alkali metal reduces Ti2+ ions (see, for example, Refs 19, 20) and in the chloride ± fluoride melts, the species being reduced is Ti3+. Titanium in higher oxidation states is preliminarily reduced to lower oxidation states by either electrochemical or chemical reactions.However, voltammo- grams characterising the cathodic evolution of titanium display a wave corresponding to the discharge of titanium complexes.18 In several studies (see, for example, Ref. 21), the mechanism of direct V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova reduction of TiCl3 to the metal without the intermediate forma- tion of Ti2+ has been assumed. A study of the electroreduction of TiCl3 in the LiCl, CsCl and KCl7LiCl systems at 973 Kby linear sweep voltammetry demonstrated that the reduction of Ti3+ in all three electrolytes involves the stage of recharge to Ti2+ (20) Ti2+, Ti3+ + e (21) Ti0. Ti2+ + 2e The stepwise character of the process was pointed out in several other publications.18, 22 ± 26 Electroreduction of titanium tetrachloride in the BaCl27CaCl27NaCl,22 LiCl7KCl, KCl7LiCl7NaCl 23 and BaCl27KCl7LiCl 24 melts occurs as three consecutive stages:(22) Ti3+, Ti4+ + e (23) Ti2+, Ti3+ + e (24) Ti0. Ti2+ + 2e The voltammograms for the electroreduction of Ti2+ in the KCl ± NaCl ± TiCl3 ± NaF ± Ti(solid) melt exhibit one clear-cut wave at potentials of 7(2.53 ± 2.6) V vs. chlorine reference elec- trode, which corresponds to the electroreduction of titanium.20 The diffusion coefficients for titanium in chloride melts have been calculated;21, 26, 27 in KCl7NaCl at 973 K, they are equal to 3.7261075 and 1.9461075 cm2 s71 for Ti3+ and Ti4+, respec- tively. The diffusion coefficient for divalent titanium in the LiCl7KCl melt at 773K amounts to (1.54 0.17)6 1075 cm2 s71 (see Ref.28). Thus, certain progress in the study of chloride and chloride ± fluoride titanium-containing melts has been achieved by now; however, the overall picture of the electrochemical and chemical behaviour of titanium in various oxidation states does not appear clear and unambiguous. There is no consensus of opinion on the mechanism of the cathodic process. Several aspects of the stability of systems containing titanium in various oxidation states are poorly documented, namely, which of the lower titanium chlor- ides is more stable and what is the limiting concentration up to which this stability is retained; what are the concentration limits for the stability of other titanium-containing species; what are the ratios of Ti2+ to Ti3+ in the melt at which the system remains stable, i.e.no disproportionation reactions occur, and what are the ways of suppressing these reactions. Answers to these questions present considerable interest from the practical view- point, because they would make it possible to create conditions for long continuous operation of an electrolysis cell and to estimate the influence of unstable titanium compounds on the technolog- ical parameters of the production of titanium and of the electro- chemical synthesis of titanium compounds. III. Electrochemical behaviour of boron The electrochemical method permits preparation of high-purity boron in large quantities. In the first studies, carried out 70 years ago, elementary boron was produced by the electrolysis of borides.The product contained not more than 60%± 70% of the target material, 20%± 30% of oxygen and 5%± 6% of the metal that served as the cathode material. When magnesium boride was used, elementary boron was formed at the cathode together with low-stability magnesium borides; the latter can be partially removed by treatment of the cathodic product with an acid. Currently, oxygen-containing compounds of boron and boron halides are mainly used for the electrolytic production of boron. Numerous studies have been devoted to the electrolysis of oxygen compounds of boron. In all systems boron oxide serves as the source of boron.29 When oxygen-containing melts are used, the process does not give high-purity boron because the product is contaminated by substantial amounts of oxygen.This is the major drawback of this method.Current problems of the electrochemistry of titanium and boron. The use of purely fluoride electrolytes is limited because fluoride melts are corrosive and toxic.30 The use of mixed chloride ± fluoride electrolytes presents considerable interest; in this case, high-purity boron free from oxygen impurities can be obtained at relatively low temperatures, the corrosiveness of the melts being relatively low.30 ± 36 Several studies have been devoted to the mechanism of electroreduction of boron. 37 ± 46 The cathodic processes occur- ring during the electrodeposition of boron from a purely fluoride (LiF7KF7KBF4) and fluoride ± oxide (LiF7KF7B2O3) elec- trolytes have been studied.36, 37 It was shown that in a purely fluoride melt, boron is reduced directly from the tetrafluoroborate complex anion in one step involving three electrons. In the LiF7KF7B2O3 system, boron oxide reacts with fluorine present in the electrolyte to give the [BF4]7 anion and oxyfluoride complexes, mainly those with a molar ratio of oxygen to boron equal to 1.66. The ratio of the concentrations of [BF4]7 and the oxyfluoride complexes is 0.1. The potentials of deposition of boron from these two electrochemically active species differ by 0.47 V, and the diffusion coefficient for [BF4]7 calculated from experimental data is 4.461079 m2 s71 at 973 K.In a study of the electrochemical properties of boron in chloride and chloride ± fluoride melts, the solubility of BCl3 in the CsCl melt was found to be 1.7461071 mol.% at 973 K; thermodynamic functions of BCl3 in CsCl were calculated, the equilibrium potentials of boron were determined and equations for them were reported.32 The equilibrium potentials of boron shift to the negative region in chloride ± fluoride electrolytes, indicating that the reactivity of boron ions in fluoride complexes is lower than that in chloride complexes.31, 32, 34 The calculation of the average number of fluoride ions (three or four) in the complex ions by the procedure proposed by Ivanovsky et al.31 attests indirectly that in the temperature range studied, most of the boron ions in the chloride ± fluoride melts occur in the [BF4]7 and [BF3Cl]7 groups.Based on the results of voltammetric studies, Chemezov 32 arrived at the conclusion that electroreduc- tion of boron in a chloride ± fluoride electrolyte occurs by a one- stage mechanism; conversely, Tsiklauri et al.41 have discovered a recharge stage, B3+ B+, followed by the discharge of B+ at the cathode to give elementary boron. For low current densities, charge transfer is the rate-determin- ing step in the electrodeposition of boron, whereas for high current densities, the rate of the process is limited by the diffusion of boron from the bulk of the melt towards the surface. This is typical of both chloride and chloride ± fluoride melts.Asomewhat different opinion on the kinetics of electroreduction of boron is held by the authors of another study,39 who claim that the rate of this process is limited by mass transfer over a broad range of polarisation rates and that the proper electrochemical stage is irreversible.Examination of the switching-on curves (potential ± time) for various polarisation current densities demonstrated 32 that neither a homogeneous chemical reaction (e.g., dissociation of the [BF4]7 or [BCl4]7 complexes), nor chemisorption of electrically active species on the electrode surface retards the reduction of boron in the melts studied. The logD± I/T plots (D is the diffusion coefficient of boron calculated from the Sand equation using the results of chronopotentiometric measurements, I is the current density and T is the temperature) in the temperature range 932 ± 1123K are straight lines; the calculation of the activation energy gives 35.52.3 kJ mol71 (see Refs.31, 32). However, in another study,41 it has been stated that electrochemically active species (mainly molecular boron fluoride) are produced in a chloride ± fluoride melt upon dissociation of boron complexes coupled with homogeneous exchange reactions between the initial complexes and the electrolyte. Study of the solubility and the kinetics of electroreduction of BF3 in the KCl7NaCl melt at 973Kin the range of BF3 pressures of (1.01 ± 5.05)6105 Pa showed the presence of two sorts of electrochemically active species.44 This accounts for the two 929 waves displayed on the voltammograms; the heights of the waves increase proportionally to the pressure. In the opinion of the researcher cited, the first wave corresponds to a two-electron transition involving BF3; at the potential of the second wave, the product formed in the first stage is reduced, the mixed chloride ± fluoride complex [BClxF3]x7 being involved in the cathodic process.Thus, despite the large number of publications dealing with the electrochemical properties of boron, no clear views on the mechanism and kinetics of its electroreduction from halide electrolytes have been developed so far. IV. Methods for the synthesis of titanium diboride. High-temperature electrochemical synthesis The principal methods used to synthesise high-melting com- pounds have been classified 23 into the following groups: (1) direct synthesis from elements, (2) synthesis from solutions in melts, (3) thermal reduction of metal oxides by metals, (4) reduction of oxides and other compounds by nonmetals and their derivatives, (5) gas-phase synthesis, and (6) electrolysis of salt melts.This classification is also applicable to the methods used to prepare titanium diboride. Direct synthesis from elements is based on the reaction of a metal (titanium) with a nonmetal (boron), which can be carried out (at various physical states of the elements) by fusing, sintering or hot pressing. In accordance with this, some researchers suggest conducting this synthesis by keeping samples of pure titanium for 5 h in amorphous boron at a temperature of 1273Kand a residual pressure of 0.133 Pa; according to other researchers, mixtures of titanium and boron powders should be heated by an electric discharge.47 The advantages of this method include the possibility of preparing large amounts of the product, the possibility of hot moulding the product in order to manufacture the required articles, simplicity of the equipment and the relatively short time needed for the synthesis.The facts that it is difficult to prepare borides with a precise composition and that the temperature of the process is too high are drawbacks of the direct synthesis. The synthesis from solutions in melts involves a chemical reaction between transition metal atoms and nonmetal atoms (or molecules of a nonmetal compound) occurring in molten salts or metals.An example is provided by the synthesis of titanium borides in an iron melt at temperatures of 1673 ± 2273 K. This route can afford both pure, doped and complex (mixed) com- pounds. In addition, it ensures the recovery of metal serving as the solvent and is relatively simple. The drawbacks of this method include low yields and high cost of the product, caused by the large amounts of salts and metallic solvents consumed, the possible contamination of the product by the solvent, the necessity of vigorous stirring of solutions in melts in order to obtain homoge- neous compounds and high temperatures of the process. The thermal reduction of oxides consists in the reduction of titanium and boron oxides by a reducing metal (in the presence of oxygen-evolving compounds) to give borides at temperatures of 2273 ± 3273K over periods of 2 ± 3 min.This is the simplest method among those listed above. However, it is not in general use because of the low quality and inhomogeneous composition of the products obtained in this way and also because it is difficult to isolate the product from the slag. The fourth method is based on the reduction of oxides and other compounds by nonmetals and their compounds. For instance TiB2 can be prepared by the reaction of BCl3 with highly dispersed titanium at 873 ± 1373 K. The advantages of this method include the possibility of preparing large quantities of the product and the fact that oxides and other readily available compounds of titanium and boron are suitable for the synthesis.The main drawback of this method is that the products of the synthesis are contaminated by metal and nonmetal oxides. Gas-phase synthesis includes decomposition of chemical compounds followed by reactions of the fragments (both ions or930 radicals and the atoms or molecules produced upon their reduction) occurring in the gas (vapour) phase to give high- melting compounds. A method for the synthesis of titanium diboride by the reduction of a vapour ± gas mixture consisting of TiCl3 and BCl3 by hydrogen at temperatures of 1273 ± 3773K and pressures of 1 ± 3 atm has been described.47 The major advantage of this method is the possibility of preparing single crystals and coatings on various materials.The inhomogeneity of the phase composition and the relatively low yields of the products as well as complexity of the equipment and the processes involved are drawbacks of this technique. High-temperature electrochemical synthesis from molten salts can be regarded as being one of the most promising methods for the preparation of high-melting compounds. The advantages of this method are relative simplicity of the equipment, the ready accessibility of the initial compounds, the possibility of preparing simple or complex composite coatings of high-melting com- pounds, the relatively low temperature of the process, and the possibility of controlling the morphology and the composition of the cathodic deposit by varying the electrolysis parameters.The theoretical grounds of this method have been described in detail, 48 and examples of its practical implementation can be found in reviews. 49, 50 Depending on the nature of the electrolyte used, two variants of electrochemical synthesis of titanium diboride can be distin- guished, namely, syntheses from oxygen-containing systems and from halide systems.51 The former variant commonly employs systems based on alkali and alkaline earth metal borates, which differ in the manner of saturation of the melt by the synthesis components. Boron can be introduced in the melt as B2O3, and titanium can be added as TiO2; the concentration of titanium in the electrolyte can be maintained by anodic dissolution of titanium metal.51 However, though seemingly advantageous, the synthesis based on oxygen-containing systems suffers from serious drawbacks.Specifically, the reaction temperatures are too high (>1023 K), and a pure product (free from oxygen) cannot be prepared in this way. The second variant can be implemented using either purely fluoride electrolytes, for example F/Li, Na, K, LiF7KF, Li, Rb, Cs/F and cryolite and mixed chloride ± fluoride electro- lytes. The sources of titanium and boron in this case are either titanium and boron fluorides or soluble anodes made of boron, titanium and boron carbides or titanium diboride. The use of mixed chloride ± fluoride electrolytes with the lowest possible content of fluorine is the most interesting.This decreases the corrosion activity and the cost of the melt; in the case where a soluble anode is used, the formation of fluorocarbons can be avoided and the stage of hydrotreating of the cathodic deposits can be simplified. The mechanism of the cathodic process in the electrochemical synthesis of titanium diboride in the LiF7KF melt has been studied.52, 53 Potassium hexafluorotitanate and tetrafluoroborate were used as the sources of titanium and boron. It was found that reduction of Ti 4+ in the LiF7KF7K2TiF67KBF4 system occurs in two steps Ti . Ti3+ Ti 4+ This is followed by the reduction of boron complexes to elementary boron, which then reacts with the reduced titanium to yield titanium diboride.This is accompanied by electrode depolarisation, which is due, in the opinion of the researchers, to the great negative value of the Gibbs energy for the formation of TiB2. Thus, analysis of the studies devoted to the methods of electrochemical synthesis of titanium diboride shows that mainly applied studies can be found in the literature. The information available from the literature is inadequate to interpret the mechanism of the synthesis; therefore, this process needs to be studied more comprehensively. V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova V. Electrochemical properties and stability of halide systems containing titanium in various oxidation states and boron 1.Study of electrochemical properties and stability of titanium-containing systems Experiments on the stability of halide systems containing titanium in various oxidation states and boron and on the electrochemical synthesis of TiB2 and titanium intermetallic compounds have been performed under an argon atmosphere, inert with respect to the melts under study. The voltammograms characterising the electroreduction of Ti3+ at polarisation rates of >0.5 V s71 (Fig. 1) exhibit two persistent waves with clear-cut peak currents. The waves are due to the recharge and discharge of titanium ions (25) [TiCl4]27 + 2Cl7, [TiCl6]37 + e (26) Ti + 4Cl7. [TiCl4]27 + 2e i 3 21 0.05 A cm72 1.5 0.5 7E/ V Figure 1. Voltammograms for the electroreduction of Ti3+ ions in the KCl7NaCl7TiCl3 melt.T=1000 K, [TiCl3]=8.361075 mol cm73, chlorine reference elec- trode; polarisation rate /V s71: (1) 1.0; (2) 0.1; (3) 0.05. To confirm the assumption that the potentials of the second wave correspond to the reduction of titanium complexes to the metal, potentiostatic electrolysis was carried out at these poten- tials. X-Ray diffraction identification of the electrolysis products showed the presence of titanium metal on the cathode. When the potential sweep rate is <0.1 V s71, the peak current of process (25) becomes feebly defined and at polarisation rates of <0.05 V s71, the waves of processes (25) and (26) coalesce. As a consequence, the overall process appears as occurring in one stage (see Fig.1, curve 3). In all probability, this is the reason for the disputes concerning the number of stages in the reduction of Ti3+ (see Refs 47, 51), because most of the studies have been carried out at polarisation rates close to the steady-state values (not higher than 10 mVs71). Analysis of voltammograms of the type presented in Fig. 2 showed that at high polarisation rates, processes (25) and (26) are not complicated by any additional reactions and, in the range of TiCl3 concentrations studied (up to 10 mass %), they completely obey the Faraday law. This is confirmed by the fact that the peak currents follow a linear dependence on the concentration of TiCl3. The dependence of the peak potential on the polarisation rate and the magnitude of the transport coefficient an, equal to 0.45, imply that at high polarisation rates, the charge transfer step in the recharge process (25) is retarded.The diffusion coefficient is equal to (3.5 0.3)61075 cm2 s71. The abnormal increase in the ip/V1/2 ratio at V<0.5 V s71 (see Fig. 2) is evidently due to the fact that the recharge is complicated by a catalytic reaction, i.e. electroreduction is accompanied by regeneration of electrochemically active species.Current problems of the electrochemistry of titanium and boron. ip/V1/2/ A cm72 V71/2 s1/2 0.1 21 V1/2/ V1/2 s71/2 1.0 0 Figure 2. Variation of ip/V1/2 vs. V1/2 (ip is the peak current, V is the polarisation rate) for processes (25) (1) and (26) (2) in the KCl7NaCl7TiCl3 melt.T=1000 K, [TiCl3]=5.561075 mol cm73. This might be disproportionation of the divalent titanium formed. For a redox process not accompanied by catalytic reactions, the ratio of the peak currents of the waves obtained in the forward (cathodic) and reverse (anodic) recording of voltammograms should be equal to unity. If this ratio is greater than unity, the product of electroreduction enters into a subsequent chemical reaction; when this is disproportionation, it increases the current and changes correspondingly the ip/V1/2 value. Study of cyclic voltammograms for process (25) has confirmed 54 that dispropor- tionation of divalent titanium occurs in the near-electrode layer (27) 3 [TiCl4]27=2 [TiCl6]37 + Ti . The minimum lifetime of the divalent titanium species is 0.08 ± 0.8 s for the TiCl3 concentration ranging from 561074 to 561075 mol cm73; the rate constant for disproportionation is (2.2 0.2)6104 cm2 mol s71.Methods for controlling the course of electroreduction by varying the acid ± base properties of the medium have been considered 53 ± 57 in relation to oxygen-containing complex ions. According to the idea of these studies, the electrochemical activity of oxygen-containing anions is due to the mechanism of formation of electrochemically active species peculiar to molten salts, namely, the acid ± base interactions in the melt. The formation and interactions of electrochemically active species in the near- electrode layer occur at a limited rate because these processes require rearrangement of the structure of reacting species.Thus, by changing the acid ± base properties of the medium, one can influence the mechanism and the rate of the process in a specified manner. Elimination of disproportionation reactions by changing the mechanism of the process can be attained in two ways, namely, by addition of either cations possessing greater polarising ability than potassium and sodium or anions having stronger basic properties than the chloride anion. In the former case, the acidity of the melt diminishes, while in the latter case, the basic properties of the electrolyte are enhanced (for instance, when sodium fluoride is introduced). The mechanism of the cathodic process in the electroreduction of titanium trichloride in chloride melts with different cationic compositions has been studied.58 It was found that reduction of titanium in the cesium chloride melt occurs via recharge stage (25), whereas in an eutectic LiCl7KCl melt, when the temperature diminishes to 673 K, trivalent titanium is reduced to the metal in one stage Ti .Ti3+ + 3e However, the LiCl7KCl system cannot be used for the synthesis of titanium diboride because the minimum temperature for the onset of the electrochemical synthesis markedly exceeds 673 K and amounts to 843 K. 931 It has been found 59 that fluoride ions added to the melt replace the chlorine in the titanium complex species and, as a consequence, the electroreduction process switches to a one-stage mechanism. In order to elucidate the potential for the synthesis of titanium diboride from chloride and fluoride electrolytes, it is significant to study the influence of fluoride ions, in particular, the [Ti3+] : [F7] molar ratio, on the mechanism of titanium electrodeposition and the type of titanium species in the melt.It is also necessary to determine the optimum electrolyte composition,60 which would preclude both disproportionation reactions and the formation of lower titanium fluorides, insoluble in the melt. As shown above, electroreduction of trivalent titanium in the chloride melt includes the recharge to the divalent state. Upon introduction into the chloride melt of fluoride ions in the ratio [Ti3+] : [F7]>1 : 2, the peak current of the recharge wave decreases proportionally to the content of fluoride ions in the electrolyte. In addition, as the current of the recharge wave decreases, the peak current of the Ti2+ Ti discharge wave proportionally increases.When the molar ratio [Ti3+] : [F7]=1 : 2 is attained, the recharge wave is no longer displayed on the voltammogram and the discharge of the complex ions of titanium becomes a one-stage process (Fig. 3). The switching to a different mechanism is accompanied by the simultaneous shift of the wave potentials towards more negative values. Experimental results imply that the change in the anionic composition of the melt upon the introduction of fluoride ions results in the formation of mixed chloride- and fluoride-contain- ing complexes. The replacement of chlorine in the [TiCl6]37 complex (this is the species in which TiCl3 occurs in the KCl7NaCl melt) by the smaller fluoride ion makes this complex stronger and changes the mechanism of its reduction.The fact that the mechanism changes at [Ti3+] : [F7]=1 : 2 suggests that the composition of the resulting titanium complex is [TiF2Cl4]37. Titanium chloride ± fluoride complexes occur in equilibrium with components of the melt; generally, the equilibrium is described by the equation (28) [TiFx]37y + y Cl7. [TiFxCly]37 Electroreduction involves dissociated species [TiFx]37y, the concentration of which is lower than the total concentration of titanium in the electrolyte. The dependence of the potential of the reduction wave peak on the polarisation rate in the broad range of potential sweep rates and also the value an =1.0 0.05, found from the Mat- suda ± Ayabe equation, imply that the charge transfer step is slow.The diffusion coefficient calculated from the Delahay equa- tion using peak currents corresponding to the total concentration i 1 2 0.025 A cm72 3 4 2.0 1.0 7E/ V Figure 3. Voltammograms for the electroreduction of titanium ions in the KCl7NaCl7NaF7TiCl3 melt at a [Ti3+] : [F7] molar ratio, equal to 1 : 0 (1), 1 : 1 (2), 1 : 2 (3), <1: 2 (4). T=1000 K, polarisation rate 1.0 Vs71, [Ti3+]=1.5561074 mol cm73, chlorine electrode as the reference electrode.932 of titanium in the melt is equal to (3.00.4)61075 cm2 s71.The underestimated value of the diffusion coefficient, (0.70.4)6 1075 cm2 s71, found using the currents corresponding to high polarisation rates confirms the assumption that the concentration of elecrochemically active species under these conditions does not coincide with the total titanium concentration in the melt. The number of ligands attached to the initial [TiF2Cl4]37 species during titration of the melt by fluoride ions, determined on the basis of the slope of the straight line plotted in the `shift in the wave peak ± depolariser concentration' coordinates (DE ± lncF7) (Fig. 4) is equal to two. With allowance for the composition of the initial species, the composition of the titanium complex in the bulk of a melt containing an excess of fluoride ions should be [TiF4Cl2]37. Then the formation and discharge of electrochemi- cally active titanium complexes can be described by the following scheme: (29) [TiF2Cl4]37+2Cl7, [TiCl6]37 +2F7 (30) [TiF4Cl2]37+2Cl7, [TiF2Cl4]37+2F7 (31) [TiF2Cl4]37 TiFá2 +4Cl7, (32) Ti+2F7.TiFá2 +3e DE/ V 0.3 0.2 a 0.1 cF7 72 0 ln Figure 4. Calculation of the coordination number of titanium in the complexes occurring in the KCl7NaCl7NaF7TiCl3 melt. Thus, the electroreduction of trivalent titanium in the near- electrode layer of a chloride melt involves disproportionation of divalent titanium. In a chloride melt, the Ti3+ Ti2+ recharge is a catalysed process and the charge transfer step is retarded over a broad range of polarisation rates.Stabilisation of the KCl7NaCl7TiCl3 system can be attained by changing the acid ± base properties of the melt by introducing fluoride ions into the chloride electrolyte, which results in the formation of mixed chloride and fluoride complexes such as [TiF2Cl4]37. The reduction of these complexes at the molar ratio [Ti3+] : [F7]<1 : 2 occurs in one stage involving the transfer of three electrons. 2. Study of electrochemical properties and stability of boron- containing systems In order to arrange a high-temperature electrochemical synthesis of titanium diboride, it is necessary that the potentials of the evolution of products used in the synthesis be close. Therefore, investigations into the electrochemical properties and stability of boron-containing chloride ± fluoride melts are topical.Analysis of the published data indicates that, despite the fact that numerous studies have been devoted to electroreduction of boron com- plexes, there are still no clear views on the electroreduction mechanism or the composition of the boron-containing species existing in chloride ± fluoride melts. The voltammograms of the electroreduction of boron exhibit one wave in the potential range from 72.35 to 72.7 V with a clear-cut peak current (Fig. 5). Further polarisation results in an increase in the current, caused by the discharge of alkali metal ions. The potentiostatic electrolysis at the potential corresponding to the reduction wave affords elementary boron.The dependence of the peak potential of the wave of electroreduction of boron V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova i 1 0.2 A cm72 23 2.0 1.5 7E/ V Figure 5. Cyclic voltammograms for the B3+ B process occurring in the KCl7NaCl7NaF7NaBF4 melt (vs. chlorine reference electrode). [B3+] : [F7] : (1) 1 :0, (2) 1 :7, (3) 1 : 14. complexes on the polarisation rate, the substantial difference between the wave peak potentials found in the forward and reverse recording of voltammograms, and the values an=0.96 ± 1.08 for the three-electron process calculated from the Matsuda ± Ayabe equation lead to the following conclusion. In the range of polarisation rates from 0.1 to 20.0 Vs71, the discharge of boron complexes is irreversible, i.e.it is controlled by the rate of charge transfer.61 The [BF4]7 ions added to a chloride melt enter into an exchange reaction with chloride anions to give mixed chloride ± fluoride complexes (33) [BF [BF47xClx]7+x F7. 4]7+x Cl7 The unstable boron complexes undergo thermal dissociation, which can be generally represented by the following scheme: (34) [BF47x]x71+x Cl7. [BF47xClx]7 Electroreduction involves species resulting from dissociation of the initial complexes; the concentration of these species in the melt bulk is small compared to the total concentration of boron ions in the electrolyte. Thus, the melt contains simultaneously two sorts of species, namely, the initial [BF47xClx]7 complexes and electrochemically active species [BF47x]x71.Upon the discharge of boron complexes, the near-electrode layer becomes depleted in the [BF47xClx]7 complexes, which displaces the equilibrium towards the formation of electrochemically active species. Anal- ysis of the ip/V1/27V1/2 dependence shows 62 that a decrease in the potential sweep rate diminishes the influence of dissociation on the electroreduction of boron complexes. The influence of ip/V1/2 onV1/2 can be extrapolated to the ip/V1/2 axis if the peak current of the wave corresponds to the total concentration of boron in the melt. The linear type of dependence of the peak currents on the boron concentration in the melt implies that the preceding chemical reaction exerts no retarding influence under these conditions. The diffusion coefficient calculated from the Delahay equation using the peak current values corresponding to the total concentration of boron in the melt is equal to (3.30.4)61075 cm2 s71.Thermal dissociation of alkali metal fluoroborates added to a chloride melt yields volatile boron fluorides, poorly soluble in the melt and passing into the gas phase. This results in a lower concentration of electrochemically active species and, hence, in a lower rate of boron electroreduction.63 The non-controllable change in the concentration of one component hampers the selection of operating conditions suitable for the electrochemical synthesis of titanium diboride. Therefore, it is necessary to find a method for stabilisation of the chloride melt containing an alkali metal fluoroborate.It has been shown 61, 62 that the introduction of NaF into the KCl7NaCl7NaBF4 melt induces a shift of theCurrent problems of the electrochemistry of titanium and boron. peak potential of the wave corresponding to the discharge of mixed chloride ± fluoride boron complexes, due to the replace- ment of chloride ions by fluoride ions. When [B3+] : [F7]51 : 2, the melt is stabilised upon the formation of mixed chloride ± fluoride complexes of boron and the rate of the reduction of boron no longer varies with time. It is of interest to determine the composition of boron-containing complexes in the initial chlor- ide melt and the complexes formed upon titration of the melt by fluoride ions.The coordination number of boron in the chloride ± fluoride complex was calculated using a procedure taking account of the dependence of the shift of the wave peak potential for the reduction of boron complexes on the concentration of the free ligand in the melt. The concentration of the free ligand was assumed, as the first approximation, to be equal to the total concentration of fluorine added to the melt. The coordination number determined from the slope of the straight line plotted in the DE ± lncF¡ coordinates (Fig. 6) is equal to unity, i.e. during titration of the KCl7NaCl7NaBF4 melt by fluoride ions, one fluoride ion adds to the initial species. Hence, the mixed chloride ± fluoride complex of boron in the NaCl7KCl melt has the composition [BF3Cl]7 and the species occurring in the melt with excess fluoride ions is BF¡4 .The application of a correction for the concentration of the bound ligand, determined by the molar ratio [B3+] : [F7]=1 : 1, did not alter the results. DE/ V 0.2 0.1 a cF7 71 ln 73 Figure 6. Calculation of the coordination number of boron in the complexes occurring in the KCl7NaCl7NaF7NaBF4 melt. Based on the foregoing, it can be concluded that the formation of mixed chloride ± fluoride complexes of boron can be described by the following equations. In the absence of excess fluoride ions in the melt (35) [BF3Cl]7+F7, BF¡4 +Cl7 (36) BF3+Cl7; [BF3Cl]7 and in the presence of excess fluoride ions in the melt (37) BF¡ BF3+F7.4 VI. Electrochemical synthesis of titanium diboride and titanium intermetallic compounds in a chloride ± fluoride melt 1. Joint electroreduction of titanium and boron in the chloride ± fluoride melt and electrodeposition of titanium diboride powders It has been shown 64 that the addition to the KCl7NaCl7NaF7KBF4 melt of titanium trichloride in a concentration close to half that of potassium tetrafluoroborate results in a sharp increase in the current at potentials 0.25 V more positive than the potential of reduction of titanium complexes. This brings about a wave characterised by a steeper slope than the waves for the discharge of the titanium and boron complex ions. This wave is displayed on voltammograms over broad ranges of polarisation rates and TiCl3 and MBF4 concentrations (Fig.7). If the ratio [Ti3+] : [B3+]=1 : 2 does not hold, then electroreduction 933 i 2.561072 A cm72 3 2 1 2.5 2.0 7E/ V Figure 7. Voltammograms for various processes (T=1000 K, polar- isation rate 1.0 Vs71, chlorine reference electrode); (1) electroreduction of titanium from the NaCl ± KCl ± NaF melt containing 4.56 1075 mol cm73 of TiCl3; (2) electroreduction of boron from the NaCl7KCl7NaF melt containing 861075 mol cm73 of NaBF4; (3) electrochemical synthesis of titanium diboride from the NaCl ± KCl ± NaF ± TiCl3 ± NaBF4 melt ([Ti]+[B]=1.261074 mol cm73). of TiB2 is followed (before the discharge of the alkali metal) by the discharge of the excess (with respect to this ratio) of a component (titanium or boron).When this ratio is lower than 1 : 2, this wave is exhibited at less negative potentials than in the case where the ratio is greater than 1 : 2. The cyclic voltammograms (reverse recording) for [Ti3+] : [B3+]=1 : 2 were found to contain one peak, while those for [Ti3+] : [B3+]=1 : 2 exhibit at least two peaks. Identification of the products of potentiostatic electrolysis demonstrated that the electrolysis product obtained at the potential of the first cathodic wave is titanium diboride, while that deposited at the second cathodic wave potential is titanium diboride together with the component (either titanium or boron) in excess with respect to the ratio [Ti3+] : [B3+]=1:2. The formation of titanium diboride is accompanied by evolution of a substantial amount of energy; therefore, its syn- thesis proceeds at more positive potentials than the separate reduction of either component of the synthesis.Depolarisation is equal to 0.15 ± 0.3 V (with respect to the initial components), depending on the electrolyte composition. Comparison of the dependences of ip/V1/2 on V1/2 for the joint and individual electroreduction of titanium and boron shows that at high polar- isation rates, the peak current of the synthesis wave is much lower than the sum of the partial currents corresponding to the discharge of the titanium and boron complexes. This fact implies that the process is retarded due to the preceding chemical reaction; the higher the polarisation rate, the more pronounced the retardation (Fig.8). This is due to the fact that the titanium and boron complex ions present in the melt form mixed complexes such as [B2TiFxCly].65 The absence of a horizontal section on the ip/V1/2 vs. V1/2 plot (as well as on the plots for the partial currents for titanium or boron), which would be due to the flow of electrochemically active species moving from the bulk melt, provides grounds for considering, according to the diagnostic electrochemical criteria, that the lifetime of the arising species is short with respect to the lifietime of species formed by titanium and boron in the corresponding systems. A decrease in the polarisation rate results in a lower contribution of the formation of joint complexes.Extrapolation of this dependence for the limiting current to the rate axis gives the current corresponding to the total concentration of titanium and boron in the melt. The introduction of boron as BF¡4 into the chloride melt containing [TiCl6]37 ions gives rise to exchange reaction (35). The fluoride ion thus liberated adds to the titanium complex [TiF2Cl4]37+2Cl7. [TiCl6]37+2F7 When there is no excess fluoride ions in the melt, the composition of the joint complex is [B2TiF8Cl4]37. The wave of934 ip/V1/2 A cm72 V71/2 s1/2 0.5 3 0.25 21 3 1 V1/2/ V1/2 s71/2 0 Figure 8. Variation of ip/V1/2 vs. V1/2 for a number of processes; (1) electroreduction of titanium complexes, [Ti]=661075 mol cm73; (2) electroreduction of boron complexes, [B]=1.261074 mol cm73; (3) electrochemical synthesis of titanium diboride, [Ti]+[B]= 1.861074 mol cm73, [Ti3+] : [B3+]=1:2.the joint electroreduction (synthesis) is extended along the potential axis; the peak width reaches 0.5 V and the potential difference between the peaks of the waves displayed during the forward and reverse recording of voltammograms is 1.0 V. The introduction of fluoride ions into the melt decreases the width of the wave and shifts the wave to the negative region; the difference between the potentials of the peaks of the cathodic and anodic waves decreases simultaneously. For the molar ratio ([Ti3+]+[B3+]) : [F7]51 : 15, the wave width and the differ- ence between the peak potentials barely change.The shift of the peak potentials of the synthesis wave towards negative values is due to the formation of stronger complexes upon the attachment of the fluoride ions in the melt to the initial complex species (38) [B2TiF8Cl4]37+n F7 [B2TiF8+nCl47n]37+n Cl7. The fact that the peak and half-peak potentials of the synthesis waves do not depend on the polarisation rate as well as the S-shape of the voltammograms imply a reversible step of charge transfer at polarisation rates of <2.0 Vs71. The number of electrons involved in the elementary electrochemical step, calculated using the Matsuda ± Ayabe equation, is equal to 9. This means that the discharge of the arising joint complexes at a polarisation rate of <2.0 Vs71 involves simultaneous transfer of nine electrons without kinetic limitations.Study of the electrochemical behav- iour of a titanium diboride electrode in a chloride ± fluoride melt showed that there is no equilibrium between the electrode and the melt and that the electrochemical step proper is reversible. During dissolution of titanium diboride, limitations appear associated with the chemical interaction at the electrode ± melt interface. As noted above, the charge transfer steps in the separate discharge of titanium and boron complexes are retarded and the electrochemical stage of the synthesis of titanium diboride is reversible. This confirms the assumption that the complexes react chemically in the bulk of the melt to give electrochemically active species, the discharge of which is reversible. The calculation of the coordination number for a mixed titanium and boron complexes shows that during titration of the melt with fluoride ions, four fluoride ions are attached to the initial species, [TiB2F8Cl4]37, to give [TiB2F12]37. It has been established relying on chronovoltammetric meas- urements 72, 73 that, for the discharge of joint titanium and boron complexes to persist, the molar ratio ([Ti3+]+[B3+]) : [F7]= 1 : 4 (for [Ti3+] : [B3+]=1 : 2) should be maintained in the melt.The results obtained were taken as the basis for practical implementation of the high-temperature electrochemical synthesis of titanium diboride. The influence of temperature, current density, and the concentrations of the synthesis components on the composition and properties of the product were studied.A V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova crucial condition for this process is that the molar ratio [Ti3+] : [B3+]=1 : 2 is maintained in the melt. In this case, a homogeneous phase of titanium diboride is deposited at the cathode over a broad range of current densities. If this condition is not fulfilled, titanium diboride is deposited at the cathode at relatively low current densities (up to 0.5 Acm72), while at higher current densities, it is deposited together with the component present in excess with respect to this ratio. The total concentration of titanium and boron in the electrolyte in the 0.6 mol.% ± 30 mol.% range does not influence significantly the product composition or economical characteristics of the process.A decrease in the total concentration to <0.6 mol.% is not expedient for technological reasons (frequent correction of the cell, low current density and low output), whereas an increase in the concentration to >30 mol.% brings about a greater loss of boron due to the thermal dissociation of potassium tetrafluor- oborate, enhanced corrosiveness of the melt, and difficulty in washing out the cathodic deposit. The temperature threshold for the onset of the electrochemical synthesis of titanium diboride in the system studied is 873 ± 893 K.66 The optimum temperature range is 963 ± 1123 K. An increase in the temperature above 1123K is not expedient as it diminishes the thermal stability of the cell and deteriorates the conditions of the formation of the metal salt precipitate.In the temperature range of 963 ± 1123 K, highly dispersed titanium diboride powders with a specific surface area of up to 10 m2 g71 are produced. Dispersed powders of titanium diboride were obtained using the KCl7NaCl ± NaF (30 mol.%) ± TiCl3 (5 mol.%) ±KBF4 (10 mol.% ) system as the electrolyte, current densities of 0.5 to 1.6 Acm72, and a temperature of 1023 K. The degree of disper- sion of the powders increased with an increase in the current density; when the current density was <0.5 Acm72, dendrites were formed, while at a current density of>1.6 Acm72, titanium diboride was deposited at the cathode together with an alkali metal.2. Joint electroreduction of titanium, nickel, cobalt and iron and electrodeposition of titanium intermetallic compounds In the electrolysis of melts containing simultaneously titanium and iron (nickel, cobalt) chloride, powdered intermetallic com- pounds (IMC) of titanium, TiFe, TiFe2, Ti2Ni, TiNi3 and TiCo, are formed at the cathode.67 ± 69 The ways of controlling the processes of formation of IMC have been studied allowing for the mechanisms proposed for high-temperature electrochemical synthesis. Electrochemical behaviour of titanium and iron chlorides occurring in the melt has been studied by linear chronovoltam- metry with cyclic potential sweep according to the three-electrode scheme.67 The dependence of the limiting current on the concen- tration was non-linear (unlike this dependence for simple rever- sible or irreversible processes).The fact that the height of the wave remains constant when the potential sweep rate increases above 0.5 Vs71 implies a retarding influence of the preceding chemical reaction. The illegibility of the peak wave also confirms that the chemical and electrochemical stages are coupled.70 It was noted 67 that potential of the cathode current shifts to the negative region and that the difference between the peak and half-peak potentials of the cathode current increases with an increase in the potential sweep rate. In view of the foregoing, with the assumption of first- order chemical reactions, two stages of high-temperature electro- chemical synthesis can be distinguished.1. Chemical reaction to give the equilibrium (39) [TixFey]z+ + 6x Cl7 , x [TiCl6]37+y Fe2+ where z=3x+2y, which shifts towards the formation of the heteronuclear complex [TixFey]z+; the heat of formation of the corresponding intermetallic compound increases. 2. Electrochemical reduction of the heteronuclear complex cation after the discharge potential has been attainedCurrent problems of the electrochemistry of titanium and boron. (40) [TixFey]z++ze=TixFey . Analysis of the diffusion coefficients shows that charge numbers of 3 and 4 are preferred for complex IMC (the particular number depends on the particular compound). Therefore, the intermetallic chloride complexes discharged at the cathode are [TiFe2]Cl3 and [TiFe]Cl3.A series of experiments dealing with the influence of the concentration of titanium on the peak potential of nickel dis- charge has been carried out. 68 Titanium is formally considered as a `quasi-ligand', although, in accordance with the views of the coordination chemistry of melts,71 it is incorporated in the nucleus of the heteronuclear complex [TixNiy]n+. The variation of the difference between the potentials of nickel reduction DEpNi in the presence and in the absence of TiCl3 as a function of the Napierian logarithm of the ratio of TiCl3 and NiCl2 concentrations has three linear sections characterised by different slopes and a horizontal section for a large excess of titanium (Fig.9). The formal coordination numbers of nickel with respect to titanium (the number of titanium atoms per nickel atom in the complex), determined from the slopes of the linear sections, are 0.4, 0.95 and 1.78. They are close to theoretical values of 0.33, 1.0 and 2.0 respectively, which correspond to the heteronuclear complexes [TiNi3]n+, [TiNi]n+, [Ti2Ni]n+. The assumed formation of these complexes has been confirmed 68 by X-ray diffraction analysis of the products of electrolysis at the potential of the intermediate wave of the synthesis; in addition to nickel, the powder contained intermetallic compounds TiNi3, TiNi and Ti2Ni. The reaction of NiCl2 and TiCl3 can be represented in the following way. According to spectroscopic measurements, 72 Ni2+ and Ti3+ cations form octahedral complexes [NiCl6]47 and [TiCl6]37 with chloride anions. Evidently, when the titanium concentrations are low (561075 mol cm73) and the ratio [Ti] : [Ni]=1 : 1, only these nickel and titanium complexes exist. The fact that upon an increase in the titanium concentration, a wave for the discharge of the intermetallic compound [TixNiy]n+ appears between the discharge waves for titanium and nickel implies 68 a reaction between the two species to give heteronuclear complexes according to the following equations (41) [TiCl6]37 .[NiCl6]4¡ [TiCl6]37+3 [NiCl6]47 3 , (42) [TiCl6]37. [NiCl6]47, [TiCl6]37+[NiCl6]47 (43) [TiCl6]3¡ . [NiCl6]47. 2 [TiCl6]37+[NiCl6]47 2 When titanium is present in a 25-fold (or higher) excess with respect to nickel, the potential of nickel discharge does not shift to negative values.Apparently, this situation corresponds to that in systems with excess complex-forming ions. In these systems, the structures of complexes are distorted; they are destroyed or turn into chain structures or into structures of individual molten salts.71 A study of the electrochemical behaviour of titanium and cobalt present simultaneously in chloride melts 69 has been the D EpNi/ mV 1.78 100 0.95 50 0.40 0 1 2 ln(cTiCl3/cNiCl2) Figure 9. DEpNi vs. ln(cTiCl/cNiCl) plot for the KCl ±NaCl ± NiCl2 ± TiCl3 melt; T=973 K. The numbers on the curve are explained in the text. 935 generalising stage in the investigations into the possibility of producing intermetallic compounds of titanium with iron group elements by high-temperature electrochemical synthesis.The appearance of the synthesis wave and the character of its variation for Ti7Co and Ti7Ni systems are similar.68 The negative slope ratios for the variation of the product of the limiting cathode current by the polarisation rate vs. the limiting cathode current of the synthesis wave makes it possible to propose a process of electrochemical reduction of the heteronuclear complex [TiCo]n+. The electrolysis yields a black finely dispersed metallic powder, TiCo. The powder is based on aggregates of dendrites with an average size of 20 mm. The specific surface area of the powder is 30 m2g71. The works that are worth mentioning among the most recent studies devoted to the electrochemistry of titanium and its alloys include investigation of the influence of the cationic and anionic compositions of the melts on the corrosion of titanium in alkali and alkaline earth metal halides,73 deposition of silicon,74 silver 75, 76 and copper 77,78 protective coatings onto titanium articles and preparation of titanium and its alloys for deposi- tion.79 Electrochemical behaviour of titanium and boron and electrodeposition of titanium intermetallic compounds have been studied in detail. 80, 81 References 1.V I Shapoval, V V Malyshev, I A Novoselova, Kh B Kushkhov Ukr. Khim. Zh. 60 37 (1994) 2. V I Shapoval, V V Malyshev, I A Novoselova, Kh B Kushkhov Zh.Prikl. Khim. 67 928 (1994) a 3. V I Shapoval, V V Malyshev, I A Novoselova, Kh B Kushkhov Usp. Khim. 64 133 (1995) [Russ. Chem. Rev. 64 125 (1995)] 4. V S Balikhin, V A Reznichenko, in Protsessy Proizvodstva Titana i ego Dvuokisi (Processes of Production of Titanium and Its Dioxide) (Moscow: Nauka, 1973) p. 182 5. L Piszczek Zesz. Nauk. Politech. Slask., Chem. 68 3 (1974); Chem. Abstr. 83 123 157 (1975) 6. F R Clayton, G A Mamantov J. Electrochem. Soc. 120 1193 (1973) 7. V I Shapoval, V I Taranenko, V V Nerubashchenko, L L Kaidanovich, in III Ukr. Konf. po Elektrokhimii (Tez. Dokl.), Kiev, 1980 [The Third Ukrainian Conference on Electrochemistry (Abstracts of Reports), Kiev, 1980] p. 157 8. A A Kazain, Yu A Andreev, V S Drozet Trudy Giredmeta (Proceedings of State Research Institute of the Rare Metals Industry) (Moscow: Nauka, 1972) Vol.43, p. 39 9. A Avaliani, O Tsiklauri, V I Shapoval, N Dvali, in Katodnye Protsessy v Ionnykh Rasplavakh (Cathodic Processes in Ionic Melts) (Tbilisi: Mesnereba, 1978) p. 1 10. A L Beskin Tekhnologiya i Oborudovanie Proizvodstva Titana i Magniya za Rubezhom (Technology and Equipment for Titanium and Magnesium Production Abroad) (Moscow: Mintsvetmet, 1978) 11. US P. 4 113 581; Chem. Abstr. 88 56 493 (1978) 12. V V Nerubashchenko, Yu G Olesov, V I Shapoval, G P Dovgaya Tsvet. Met. 10 83 (1984) 13. B F Markov Termodinamika Rasplavlennykh Solei (Thermo- dynamics of Molten Salts) (Kiev: Naukova Dumka, 1974) p. 160 14. MV Smirnov Elektrodnye Potentsialy v Rasplavlennykh Khloridakh (Electrode Potentials in Molten Chlorides) (Moscow: Nauka, 1973) p.247 15. A N Baraboshkin Elektrokristallizatsiya Metallov iz Rasplavov Solei (Electrocrystallisation of Metals from Molten Salts) (Moscow: Nauka, 1976) 16. V I Ermolenko, L T Savchenko, EMMatorin, Ya G Goroshchenko, V A Drozdenko, V Ya Shub Ukr. Khim. Zh. 41 1252 (1975) 17. D Inman, S White, in International Symposium on Molten Salts Elec- trolysis Metals Production (Abstracts of Reports), Grenoble, 1977 p. 50 18. E Chassaing, F Basile, in Titanium Science Technology Processes (Abstracts of Reports of the IVth International Conference), New York, 1980 Vol. 3, p. 1963 19. I V Zarutskii, V V Malyshev, V I Shapoval Zh. Prikl. Khim.70 1475 (1997) a936 20. V I Taranenko, V A Bidenko, I V Zarutskii, S V Devyatkin, in Boridy (Borides) (Kiev: Institute of Science of Metals Problems) 1990 21. L Piszczek Zesz. Nauk. Politech. Slask., Chem. 65 237 (1973); Chem. Abstr. 81 71 840 (1974) 22. M Dinu Metalurgia 36 261 (1984) 23. F Quemper,D Deroo,M Rigaud J. Electrochem. Soc. 119 1353 (1972) 24. E Chassaing, F Basile,GLorthioir J. Less-Common Met. 68 153 (1979) 25. A I Ivanov, S V Aleksandrovskii Elektrokhimiya 7 838 (1971) b 26. V S Balikhin, S B Makarov Izv. Akad. Nauk SSSR, Ser. Met. 49 (1983) 27. S B Makarov, A V Balikhin Ukr. Khim. Zh. 50 498 (1984) 28. I A Filipova, A N Baraboshkin Tr. Inst. Elektrokhim., Ural NTs Akad. Nauk SSSR 27 35 (1978) 29. H Kaheenberg Trans. J.Electrochem. Soc. 47 23 (1978) 30. G P Dovgaya, Candidate Thesis in Chemical Sciences, Institute of Science of Metals Problems, Kiev, 1983 31. L E Ivanovsky, O V Chemezov, V N Nekrasov, in The 37th Meeting of the International Electrochemical Society (Abstracts of Reports), Vilnus, 1986 p. 16 32. O V Chemezov, Candidate Thesis in Chemical Sciences, Institute of Electrochemistry, Sverdlovsk 1987 33. O V Chemezov, V P Batukhin, in IV Ural'sk. Konf. po Vysoko- temperaturnoi Fizicheskoi Khimii i Elektrokhimii (Tez. Dokl.), Sverdlovsk, 1985 [The Fourth Ural Conference on High-Temperature Physical Chemistry and Electrochemistry (Abstracts of Reports), Sverdlovsk, 1985] Vol. 1, p. 134 34. O V Chemezov, V P Batukhin, L E Ivanovskii, in IV Ural'sk. Konf.po Vysokotemperaturnoi Fizicheskoi Khimii i Elektrokhimii (Tez. Dokl.), Sverdlovsk, 1985 [The Fourth Ural Conference on High- Temperature Physical Chemistry and Electrochemistry (Abstracts of Reports), Sverdlovsk, 1985] Vol. 1, p. 132 35. O V Chemezov, L E Ivanovskii, V P Batukhin, in IV Ural'sk. Konf. po Vysokotemperaturnoi Fizicheskoi Khimii i Elektrokhimii (Tez. Dokl.), Sverdlovsk, 1985 [The Fourth Ural Conference on High- Temperature Physical Chemistry and Electrochemistry (Abstracts of Reports), Sverdlovsk, 1985] Vol. 1, p. 143 36. N N Kaiguzhskii, Yu G Olesov, I V Zarutskii, in Voprosy Khimii i Khimicheskoi Tekhnologii (Problems of Chemistry and Chemical Technology) (Kiev: Vysshaya Shchkola, 1974) 37. M Makyta, K Matiasovsky, P Fellner Electrochim.Acta 29 1653 (1984) 38. P Fellner, M Makita, K Matiashovski, A Silny, in V Konf. Sots. Stran po Khimii Rasplavlennykh Solei (Tez. Dokl.) [The Fifth Conference of Socialist States on the Chemistry of Molten Salts (Abstracts of Reports)] (Kiev: Naukova Dumka, 1984) p. 43 39. O G Tsiklauri, G A Gelovani Rasplavy (5) 35 (1989) 40. O G Tsiklauri, I A Bairamashvili,M N Kobalava Soobshch. Akad. Nauk Gr. SSR 134 35 (1989) 41. O G Tsiklauri, A Sh Avaliani, in The 37th Meeting of the International Electrochemical Society (Abstracts of Reports), Vilnus, 1986 p. 68 42. O G Tsiklauri, V I Shapoval, A Sh Avaliani, N V Dvali, in Ionnye Rasplavy i Tverdye Elektrolity (Ionic Melts and Solid Electrolytes) (Kiev: Naukova Dumka, 1986) Vol.1, p. 47 43. J Mindin, O G Tsiklauri, in The 9th International Symposium on Boron, Borides and Related Compounds (Abstracts of Reports), Duisburg, 1987 p. 54 44. G A Gelovani, Candidate Thesis in Chemical Sciences, Research Institute of Inorganic Chemistry and Electrochemistry, Academy of Sciences , Gr. SSR, Tbilisi, 1990 45. O G Tsiklauri, G A Gelovani, in IX Vsesoyuz. Konf. po Fizicheskoi Khimii i Elektrokhimii Ionnykh Rasplavov i Tverdykh Elektrolitov (Tez. Dokl.), Sverdlovsk, 1987 [The IXth All-Union Conference on Physical Chemistry and Electrochemistry of Ionic Melts and Solid Electrolytes (Abstracts of Reports), Sverdlovsk, 1987] p. 249 46. V N Gurin Usp. Khim. 41 616 (1972) [Russ. Chem. Rev. 41 323 (1972)] 47. V A Reznichenko, V S Ustinov, I A Karyakin, A N Petrun'ko Elektrometallurgiya i Khimiya Titana (Electrometallurgy and Chemistry of Titanium) (Moscow: Nauka, 1982) 48.I A Novoselova, V V Malyshev, V I Shapoval, Kh B Kushkhov Teor. Osnovy Khim. Tekhnol. 31 286 (1997) 49. V V Malyshev,N N Uskova, V I Shapoval Poroshk. Metall. (5/6) 66 (1997) 50. V V Malyshev, A D Pisanenko, V I Shapoval Rasplavy (5) 76 (1997) V I Shapoval, I V Zarutskii, V V Malyshev, N N Uskova 51. V A Garmata (Ed.) Titan, Svoistva, Syr'evaya Baza, Fiziko-Khimi- cheskie Osnovy i Sposoby Polucheniya (Titanium, Properties, Resources, Physicochemical Foundations and Methods of Produc- tion) (Moscow: Metallurgiya, 1983) 52. M Makyta, K Matiasovsky, V Taranenko Electrochim. Acta 34 861 (1989) 53. V V Malyshev, I A Novoselova, V I Shapoval Zh. Prikl. Khim. 69 1233 (1996) a 54. V I Shapoval, V I Taranenko, I V Zarutskii Ukr. Khim. Zh. 53 370 (1985) 55. V I Shapoval, Yu K Delimarskii, V F Grishchenko, in Ionnye Ras- plavy (Ionic Melts) (Kiev: Naukova Dumka, 1974) Vol. 1, p. 212 56. Yu K Delimarskii, V I Shapoval, O G Tsiklauri, V A Vasilenko Ukr. Khim. Zh. 40 8 (1974) 57. Yu K Delimarskii, V I Shapoval Teor. Eksp. Khim. 8 459 (1972) c 58. E Chassaing, F Basile,G Lorthioir J. Appl. Electrochem. 11 193 (1981) 59. L I Zarubitskaya, V I Taranenko, V I Shapoval Ukr. Khim. Zh. 51 1243 (1985) 60. L I Zarubitskaya, Yu V Korobka, V I Shapoval Zh. Prikl. Khim. 71 412 (1998) a 61. V V Malyshev, N N Uskova, V I Shapoval Zh. Neorg. Khim. 41 1774 (1996) d 62. V I Taranenko, I V Zarutskii, in IX Vsesoyuz. Konf. po Fizicheskoi Khimii i Elektrokhimii Ionnykh Rasplavov i Tverdykh Elektrolitov (Tez. Dokl.), Sverdlovsk, 1987 [The IXth All-Union Conference on Physical Chemistry and Electrochemistry of Ionic Melts and Solid Electrolytes (Abstracts of Reports), Sverdlovsk, 1987] p. 174 63. L P Polyakova, G A Bukatova, E G Polyakov, E Kristensen, I Barner, N Ya B'errum Elektrokhimiya 33 674 (1997) b 64. V I Shapoval, V I Taranenko, I V Zarutskii, in VI Konf. po Khimii Rasplavlennykh Solei, (Tez. Dokl.), Smolenitse, 1985 [The Sixth 66. M Makyta,K Matiasovsky,V I Taranenko Electrochim. Acta 34 961 Conference on Chemistry of Molten Salts (Abstracts of Reports), Smolenitse, 1985]p. 105 65. V I Taranenko, I V Zarutskii, V I Shapoval,M Makyta, K Matiasovsky Electrochim. Acta 37 263 (1992) (1989) 67. V I Shapoval, Yu V Korobka, L I Zarubitskaya Ukr. Khim. Zh. 59 378 (1993) 68. L I Zarubitskaya, Yu V Korobka, V I Shapoval Ukr. Khim. Zh. 61 42 (1995) 69. Yu B Korobka, L I Zarubitskaya, V I Shapoval Ukr. Khim. Zh. 60 521 (1994) 70. Z Galus Theoretyczne Podstawy Electroanalizy Chemiczney (Warsaw: Panstwawe Wydawnictwo Wankowe, 1971) 71. S V Volkov, V F Grishchenko, Yu K Delimarskii Koordinatsion- naya Khimiya Solevykh Rasplavov (Coordination Chemistry of Molten Salts) (Kiev: Naukova Dumka, 1977) p. 330 72. S V Volkov, K B Yatsimirskii Spektroskopiya Rasplavlennykh Solei (Spectroscopy of Molten Salts) (Kiev: Naukova Dumka, 1977) p. 224 73. V Tkhai, Yu G Dikunov, O V Kovalik, S P P'yankova Rasplavy (6) 56 (1997) 74. V Tkhai, Yu G Dikunov, O V Kovalik, S P P'yankova Rasplavy (6) 60 (1997) 75. G N Shardakova, V Ya Kudyakov, N D Shamanova, V G Zyryanov Rasplavy (4) 84 (1997) 76. V Ya Kudyakov, G N Shardakova, N T Shardakov Rasplavy (4) 78 (1997) 77. T I Manukhina, I N Ozernaya, N D Shamanova, V Ya Kudyakov Rasplavy (3) 71 (1997) 78. T I Manukhina, I N Ozernaya, N D Shamanova Rasplavy (3) 63 (1997) 79. E A Fedorova, T N Kuznetsova, V N Flerov Zh. Prikl. Khim. 71 1311 (1998) a 80. V I Shapoval, I V Zarutskii, V V Malyshev Elektrokhimiya 34 1107 (1998) b 81. V V Malyshev, I A Novoselova, V I Shapoval Molten Salts Bull. Sels Fonds 63 2 (1997) a�Russ. J. Appl. Chem. (Engl. Transl.) b�Russ. J. Electrochem. (Engl. Transl.) c�Theor. Exp. Chem. (Engl. Transl.) d�Russ. J. Inorg. Chem. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (11) 937 ± 956 (1999) Reactivity of lignin and problems of its oxidative destruction with peroxy reagents V A Demin, V V Shereshovets, J B Monakov Contents I. Introduction II. Lignin structure III. Oxidation of model compounds IV. Combined action of oxidants on the sulfate pulp lignin V. Lignin activation towards oxidative action of oxygen and hydrogen peroxide VI. Ozonolysis of sulfate pulps VII. A schematic approach to the selective destruction of residual lignin with peroxy reagents VIII. Kinetic methods for assessment of residual lignin reactivity IX. Conclusion Abstract. Published data on reactivity and oxidation of lignin and model compounds with hydrogen peroxide, ozone and chlorine dioxide as well as on oxidative destruction of the sulfate pulp lignin with various reagents during bleaching are systematised and generalised.Concepts of lignin activation towards its selective oxidation and kinetic features of sulfate pulp oxidative delignifi- cation are considered. The bibliography includes 157 references. I. Introduction Wood represents a complex polymeric composition in which lignin and cellulose are the main components.1± 7 During its processing, lignin and cellulose undergo a variety of parallel redox transformations and behave as a single complex (technical pulp). Selective oxidative destruction of lignin in pulp, especially in sulfate pulp where lignin has been modified during the sulfate pulping process of the native polymer, is a difficult problem, mostly due to the complexity of lignin, which is built up of a multitude of non-uniform structural elements, and the extremely high variability of properties of this polyfunctional polymer.The problem of oxidative destruction of residual lignin { with peroxy reagents [O3, RC(O)OOH, H2O2] is of considerable practical importance since current manufacture of pulps (pre- dominantly of sulfate pulps) consumes a million tons of chlorine for bleaching and results in toxic chlorine-containing pollutants, including dioxins. The known multi-step schemes for complete oxidative destruction of the residual lignin with various sets of reagents V A Demin Institute of Chemistry, Komi Scientific Centre of the Russian Academy of Sciences, Pervomaiskaya ul., 167000 Syktyvkar, Russian Federation.Fax (7-821) 243 66 77. Tel. (7-821) 243 24 27. E-mail: kav@chemi.komi.ru V V Shereshovets, J B Monakov Institute of Organic Chemistry of the Ufa Scientific Centre of the Russian Academy of Sciences, prosp. Oktyabrya 71, 450054 Ufa, Russian Federation. Fax (7-347) 235 60 66. Tel. (7-347) 235 61 66 (J B Monakov). E-mail: chemorg@academy.bashnet.ru Received 26 October 1998 Uspekhi Khimii 68 (11) 1029 ± 1050 (1999); translated by S V Chapyshev #1999 Russian Academy of Sciences and Turpion Ltd have been developed empirically. Such schemes usually have no solid physicochemical background and, as a result, do not always take into account the impact of the preceding technological operations on the physicochemical properties of the residual lignin and the effect of acid-base catalysis. Consideration of these factors takes on especial significance on replacement of the chlorine and chlorine-containing compounds (used so far in the pulp industry) by hydrogen peroxide, oxygen and ozone, which are ecologically friendly but possess relatively low reactivities towards lignin.The reactivities of the native and residual lignins in pulps and the mechanisms of lignin reactions with various reagents during pulp preparation from a wood raw material have been considered in a series of reviews.4, 8 ± 10 The last two of them,9, 10 present a survey of the mechanisms of elementary reactions and character- istics of the majority of reagents used in wood chemistry and depict reaction schemes for the low-molecular-weight compounds that model lignin fragments.However, all these data allowed one only to explain rather than to predict the results of the action of any particular oxidant on the residual lignin. This is associated with the fact that the oxidative destruction of the residual lignin during pulp delignification is usually performed in several stages, because lignin cannot be selectively and completely oxidised in one step with any particular reagent. It should be noted that lignin (more correctly, lignins) represent chemically labile polyfunctional polymers, which can- not be isolated as individual compounds in native states. Consec- utive stages of oxidative destruction on each occasion deal with a `different' lignin of a lignino-cellulose composition (technical pulp). Any chemical action on lignin not only alters its functional and polymeric composition but it also changes the number, ratio and strength of electrophilic and nucleophilic centres and the contents of `weak' bonds in a lignin macromolecule. The necessity to account for, and systematise, all the above- mentioned factors has led to the development of a new approach, in which the multi-step process of oxidative destruction of the residual lignin is considered as a united combination of redox transformations, wherein each particular stage `prepares' and activates lignin for the subsequent technological stage through creation in lignin of new centres reactive towards nucleophilic or electrophilic reagents.{ Lignin that remained in the sulfate pulp after cooking. UDC 547.992.3 : 634.0.813 937 938 940 949 950 951 952 953 954938 HO 4 MeO R=H, OMe. This review includes a brief description of the lignin structure, a consideration of the reactions of some model compounds which are hard to oxidise with oxygen, hydroperoxide anions, ozone and chlorine dioxide as well as delignification of the sulfate pulp with hydrogen peroxide and ozone and activation of lignin towards its oxidative destruction with various reagents. It also introduces a concept of selective lignin oxidation by consecutive action of peroxy reagents, which has led to the development of the most efficient schemes for pulp bleaching.In addition, the kinetic features of oxidative lignin destruction in the processes of the sulfate pulp delignification are discussed. II. Lignin structure The structural units are linked to each other with various ether and carbon ± carbon bonds.18 ± 21 Because of structural irregularity and great variety of the chemical bonds in lignin, the description of its macromolecule as a simple combination of a few (or a multitude of) monomers (substituted phenol ± propane units) seems to be impossible.22 ± 26 The structure of lignin is still the subject of model studies.27 ± 30 Several lignin structures, e.g., that shown in Scheme 1,28 have been proposed based on the results of studies of lignin ethanolysis, thioacetolysis, oxidation 31 ± 33 and hydrogenolysis 34 products, data of elemental analysis, functional group determinations, comparative analysis of properties of native lignin and synthetic dehydropolymers (DHP) built of substituted phenol ± propane units, spectroscopic data (1H and 13C NMR, IR, UV),20, 22, 35, 36 the results of gel and gas chromatography and mass spectrometry.A segment of lignin macromolecule represented in the Freuden- berg scheme contains 18 phenol ± propane units and reflects the ratio of the main types of structural units, bonds and functional groups. This scheme can be supplemented by alternative struc- tural elements, e.g., 6b, 13b, 14b, 13c and 14c. Later, Adler 37 suggested a structural scheme for a segment of the fir-wood lignin Lignins represent macromolecular polyfunctional aromatic com- pounds; a structural model of softwood lignin according to Freudenberg is described in Scheme 1.Being the components of wood tissues, lignins are responsible for vital mechanical, chem- ical and biological functions,1± 4 viz., they impart durability and protection against oxidation and microbial action. Lignin is biosynthesised in close proximity to macromolecules of cellulose and non-cellulosic polysaccharides, which results in a lig- nin ± carbohydrate polymeric matrix. Chemical, physicochemical and mechanical interactions of lignin with cellulose ensure the formation of a complex polymeric composition, viz., the wood tissue, and predetermine the reactivities of its main components in chemicotechnological processes. Wood lignins are built up of substituted phenol ± propane units.Guaiacyl- and syringylpropa- nol units are the most common structural elements.11 ± 17 H2COH HC HCOH H2COH 16 CO MeO H2COH CH2 O CH H2COH HC HC O 2 1/2 HC 1/2 HCO H2COH OMe 1 OH HC HC MeO H2COH 3 CH O HC CO (OMe)m O HC OH 10 MeO 4 H2COH OMe 9 OMe H2COH O O CHMeO O CH HC H2C HC HCO(C6H10O5)nH CH HC 11 HC 5 O CH2 MeO H2COH OH 8 CH O OMe H2COH HCOH O HC O HC 6a MeO 7 O OMe 6b OH MeO O V A Demin, V V Shereshovets, J B Monakov R 5 6 1 OH C C Cg b a 2 3 H2COH CO CH2 H2COH 18 HC H2COH OMe O HC CH HC 17 OMe 15 OH OMeO H2COH OC CH2 CH HC O HC HCOH HCOH HC 13a 14a 12 OMe O O OHH2COH CH CH 14b H2COH OMe O HC HCOH 13b OMe OH Scheme 1 OMe OMe H2COH CH H2COH CH CH 14c C OMe O 13c OMe O939 Reactivity of lignin and problems of its oxidative destruction with peroxy reagents Table 1.Types and number of linkages between the structural elements in different lignin models.18 Linkage type Number of linkages per 100 phenol ± propane units a III II I55 65 49 ± 51 7 7 6 ±8 16 6 9 ± 15 9 15 2 9 9.5 3 3.5 2 2 7 However, not all combinations in the coupling of monomers are of the equal probability. It has been shown by quantum- chemical calculations that the electron density distribution favours the formation of the ether C(b)7O7C(4) linkage result- ing from the coupling of radicals 1a and 1b.Indeed, this linkage type predominates in lignins (Table 1). The functional group distribution with regards to the linkage types for the C(a), C(b) and C(g) atoms in the side chain is described by the Ridholm diagram 38 (Table 2). Thus, the literature data available to date do not provide an unambiguous definition of lignin as an individual chemical compound and this substantially prevents assessment of its reactivity by the methods of classical organic chemistry, i.e., by studying reactions of simple monomers and dimers mimicking different substituted phenol ± propane structural units of lignin. Such studies may only provide helpful information regarding the elementary reactions of the lignin functional groups with oxidants rather than predict unambiguously their reactions with lignin itself. C(b)7O7C(4) C(a)7O7C(4) C(b)7O7C(5) C(b)7C(1) C(5)7C(5) 2.3 C(4)7O7C(5) 1.5 C(b)7C(b) 5.5 C(a)7O7C(g) 10 7 7 C(a)7C(b) 11 7 2.5 C(b)7C(6), C(6)7C(5) 4.5 ± 5.0 b 2 a I, the Glasser model,39 II, the Nimz model,41 III, the Erickson model.40 b Including the C(1)7O7C(4) and C(1)7C(5) bonds. The residual lignin in sulfate pulp from Pinus radiata has been studied by means of mass spectrometry and 13C NMR spectro- scopy.36 Residual lignin was isolated by treating the sulfate pulp with cellulase.As the duration of sulfate cooking is increased, the number of the C7C bonds increases and the portion of the b-aryl- ether bonds in the residual lignin is decreased, thus indicating the higher degree of lignin condensation.Despite the reduction of a Table 2. Functional group distribution in lignin according to bond types based on 13C NMR spectroscopic data (the Ridholm diagram). Structural unit a Group Pro- portion C(g)7OH C(g)7OH C(g)7OH C(g)7OH C(g)=O 0.04 0.03 0.02 0.60 0.03 0.18 coniferyl alcohol type dihydroconiferyl alcohol type guaiacylglycerol type other primary alcohols coniferyl aldehyde type C(g)7O7C(a) pinoresinol, phenylisochroman, open-type units C(g)7O7C(4) ether group methyl group g-lactone (ring 14a) the same (ring 13a) macromolecule comprising 16 phenol ± propane units.It is obvious that in such small segments it is rather difficult to take precisely into consideration the ratio of certain lignin structural elements and bonds. A recently proposed 38 model of fir-wood lignin contained 28 phenol ± propane units and included some alternative structural elements that were previously absent in the Adler scheme. In addition to the models with a relatively small number of phenol ± propane units built up by random combina- tion of functional groups, bonds and structural elements, several more complete computer structural models for fir-wood lignin have been suggested. Thus the Glasser model 39 for lignin with molecular weight >17 000 contains 94 phenol ± propane units and takes into account most of the known analytical data.The results of simulations of the lignin structure (models suggested by Glasser,39 Erickson 40 and Nimz 41) are presented in Table 1. The structural heterogeneity of lignins is associated with specific features of lignin biogenesis in wood tissues. The first stage of lignin biosynthesis is the enzymic oxidative deprotonation of p-hydroxycinnamyl alcohols (coumaryl, coniferyl and sinapyl alcohols) leading to the formation of a system containing the radicals 1a ± e (Scheme 2).38 C(g)H3 C(g)H2 C(g)=O 0.02 0.04 0.02 0.02 Scheme 2 total 1.00 g CH2OH CH2OH CH2OH b CH CH CH a CH CH CH1 6 23 7e,7H+ 5 OMe OMe OMe 4 guaiacylglycerol type C(b)7O7C(4) general type C(b)7C(5) C(b)7C(6) C(b)7C(b) C(b)7C(1) C(b)7C(a) C(b)H2 C(b)7OH 0.44 the same 0.14 " 0.04 " 0.14 " 0.10 coniferyl alcohol and coniferyl aldehyde type 0.07 coniferyl alcohol type 0.03 0.04 1b 1a O O OH total 1.00 CH2OH CH2OH CH2OH 0.25 C(a)7OH CH CH CH CH C CH C(a)=O C(a)=O etc.0.06 0.11 0.20 OMe OMe OMe 1d O 1c OH 1e O coupling of rings 12 ± 15; C(b)7C(1) bonds; guaiacylglycerol type keto group aldehyde group C(a)7O7C(g) pinoresinol, phenylisochroman, open-type units C(a)7O7C(4) coupling of rings 17 ± 18 C(a)7C(b) C(a)H2 bonds with carbohydrates and others 0.11 coniferyl alcohol and coniferyl aldehyde type 0.07 dihydroconiferyl alcohol type and others 0.06 C(a)7O7C(4) general type, e.g., in coupling of rings 3 ± 4 0.10 0.04 C(a)7O7 total 1.00 a For structural units and ring numbers, see Scheme 1.Recombination of two radicals affords one of 36 possible dimers. Upon the action of peroxidases, the latter undergo further oxidation producing dimeric radicals which, in turn, can form several hundred combinations of tetramers. The random charac- ter of the electron localisation predetermines the sporadic charac- ter and irregularity of the lignin structure.3, 5940 Table 3. Functional group content in LMP and sulfate lignin.18 Functional groups Number of groups per Method of analysis 100 phenol ± propane units fir-wood pine-wood LMPa sulfate lignin b OH (total amount) guaiacol type catechol type 120 60 12 120 30 7 aliphatic COOH CO (total) 48 16 15 905 203 7 acetylation oxidation with KIO4 reaction with Fe2+, calorimetry see footnote c methylation treatment with (NH2OH) .HCl, volumetry, KBH4 reduction, Dei-method d the same see footnote c 5 10 7 10 coniferyl aldehyde type a-CO b-CO and others a Molecular formula for a structural unit of fir-wood LMP is C9H8.8O2.4(OCH3)0.96. b Molecular formula for a structural unit of pine- wood sulfate lignin is C9H7.9O2.1S0.1(OCH3)0.82. c Calculated by differ- ence. d Method of differential UVspectroscopy based on the measurement of a difference in absorption of compounds with ionised and non-ionised phenolic and hydroxy groups in solutions of different pH.12 percentage of the b-aryl-ether bonds in lignin from 70% to 50%, the portion of these bonds in the residual lignin remained to be considerable even in the cases of low lignin content in pulp upon profound delignification.The sulfate lignin differs from slightly modified lignin of mechanical pulps (LMP) in the content of functional groups (Table 3). The sulfate pine-wood lignin contains fewer carbonyl and aliphatic hydroxy groups and more carboxy and phenolic OH groups.1 ±3 It is free from the coniferyl ± aldehyde groups, which are lost during the sulfate cooking,42 and virtually free from the benzyl alcohol OH or ether groups. During the sulfate cooking partial loss of the g-hydroxymethyl and methoxy groups is also observed. In addition, contrarily to LMP, sulfate lignins contain the thiol group and elementary sulfur.8, 42 Softwood lignins are mostly built up of guaiacol derivatives, while guaiacol and syringol derivatives are the main constituents of hardwood lignins. Thus non-fractionated sulfate birch-wood lignin contains approximately equal amounts of both types of derivatives.36 The fractionation of lignin has shown that the ratio of syringol and guaiacol derivatives is increased from 0.8 to 1.5 with diminishing of the lignin molecular weight from 3500 to 650.29 Examination of DHP has shown that these compounds have different structures and properties depending on the manner of addition of initial monomers (substituted phenol ± propane units) to the reaction mixture.When the monomer is added in one portion, the couplings of dimeric radicals with a monomer occur more rarely than in the gradual addition, whereas the couplings of a `monomer ± monomer' type become more abundant.In this case, the C(b)7C(5) bonds predominate.15 The probability of the radical coupling of a `polymer ± polymer' type is higher on gradual addition of the monomer due to the low concentration of the latter. In this case, the C(b)7O7C(4) bonds predominate. It should be noted that recent hydrodynamic investigations of DHP and slightly modified softwood lignins do not quite agree with this hypothesis of the existence of two types of lignins.2 The native lignin is synthesised in plants from monomeric molecules that usually have from two to four functional groups.15, 28 The formation of both branched chains and network structures (at gel-point) is theoretically possible.24 As has already V A Demin, V V Shereshovets, J B Monakov been pointed out, the interaction of the growing lignin polymeric chains with surrounding macromolecules of non-cellulosic poly- saccharides in plant tissues results in the formation of an amorphous lignocellulosic matrix.According to the Erin'sh hypothesis,7, 26 the three-dimensional structure of this matrix represents a superposition of three networks, viz., a network of hydrogen bonds between lignin and carbohydrates (H-network), a network of the lignin ± carbohydrate bonds (LC-network) and a network of lignin itself. The most important arguments in favour of this hypothesis are insolubility of native lignin in all neutral solvents without preliminary chemical or mechanical treatment and high polydispersity of lignins isolated from wood tissues.Based on the theory for branching processes, the average molec- ular masses for sol fractions of model lignin compounds have been calculated 30 to be 1300 ± 13 800. According to these values, the lignin extracted with ethanol in the absence of catalysts can be considered as a sol fraction of the native lignin. A new promising direction in the study of biosynthesis and the structure of lignin is a computer-assisted simulation of the growth of fractal clusters with fractional dimentionalities.24 Thus the currently available results of direct and indirect chemical, physicochemical and physical investigations taken altogether do not allow one to define unambiguously the formula and all levels of structural organisation for lignin.The properties and reactivities of industrial lignins depend to a considerable extent on the preparation methods.1, 17 The residual lignin is the most condensed and high-molecular-weight part of the lignin matrix involved into a lignin ± cellulose composition, and its selective removal from cellulose fibres can only be achieved with the use of oxidants.43 The presence of different types of functional groups and bonds in sulfate lignin and in the residual one predetermines a wide range of reagents used for lignin oxidative destruction in the preparation of cellulose from raw wood materials. The main method for the study of lignin transformations is based on the analysis of reaction products resulting from the reaction of model compounds with oxidants.The most efficient reagent used for oxidative destruction and solubilisation of residual lignin in the pulp bleaching are molecular chlorine,44 ± 46 sodium hypochlorite,47 ± 49 hypochlorous acid,48, 50 sodium chlor- ite,50 chlorine dioxide,51, 52 organic 9, 10, 53 and mineral 54 peroxy acids, hydrogen peroxide,55 ± 57 oxygen 58, 59 and ozone.60 ± 62 One of promising approaches to oxidative destruction of residual lignin in industrial pulps during bleaching is the biotechnological method.63, 64 Molecular chlorine is widely employed in the pulp industry for oxidation of lignin during the pulp bleaching, however, its application is accompanied by concomitant formation of a number of toxic aromatic chlorine-containing products,65, 66 including dioxins.67 The use of sodium hypochlorite for the pulp bleaching is accompanied by the formation of chloroform.68 Chlorinated hydrocarbons are dangerous for both living organ- isms and the Earth ozone layer. Therefore, the oxidation processes based on the use of environmentally safer reagents, such as hydrogen peroxide, chlorine dioxide and ozone, are of great importance.The reactions of model lignin compounds with these reagents are discussed below. III. Oxidation of model compounds 1. Oxidation with hydrogen peroxide and oxygen In an alkaline medium, hydrogen peroxide can undergo various transformations.When the concentration of a base is high, the decomposition of H2O2 follows predominantly a radical mecha- nism with the formation of molecular oxygen:69, 70 (1) O2¡ H2O2+2OH7 2 +2H2O, (2) O2¡ O2¡.+.OH+OH7, 2 +H2O2 (3) HO2.+H2O, .OH+H2O2Reactivity of lignin and problems of its oxidative destruction with peroxy reagents (4) O2¡.+H2O HO2 .+OH7, (5) H2O2+O2. 2HO2. accompanied by generation of the radical anions O¡ radicals and singlet oxygen (1O2). The schemes of generation of the reactive oxygen forms and their mutual transformations are described by the following equations (6) .O2. +e7 O¡2 ., (7) The oxidation of lignin with oxygen in an alkaline medium is 2 ., hydroxyl 12.25, the rate of the catechol oxidation is sharply increased, which is rationalised 70 as the result of the second-step dissociation of catechol (pKa*12.5, for the first step pKa=9.45).Kinetic studies provide evidence on an induction period prior to the appearance of H2O2 in the system, which catalyses the process described by the following equations R HO2. , O2¡.+H+ (8) HO¡2 +1O2, O¡2 .+HO2. (9) H2O2+1O2, HO2.+HO2. (10) HO¡2 +H2O2 O2¡.+HO.+H2O, (11) 2 (1O2) 2O2 +hn. Comparison of these equations with the equations (1) ± (5) shows that both systems contain all the above-mentioned species independently of the oxidant used (O2 or H2O2).70, 71 The equations (8), (9) and (11) account for the chemiluminescence observed in the oxidation of lignin with oxygen.71 ± 73 The decom- position ofH2O2 is catalysed by transition metal ions (Fe, Mn, Cu, Zn, etc.) that are present in plant tissues and lignin.59, 74, 75 The rate and the depth of the lignin oxidation with H2O2 and O2 depends on the reaction conditions.The reaction (6) leads to the single-electron oxidation of a phenolic fragment in lignin. R R +O¡ (12) +O2 2 . OMe OMe O7 O and The superoxide radical anion O¡2 . takes part in the single- phenyl)propane-1,3-diol (2b). It was preliminarily shown that compound 2b forms 3-(3-ethoxy-4-hydroxyphenyl)-2-(4- electron transfer as a reducing agent and in nucleophilic deproto- nation [equation (7)] and substitution reactions.In aqueous and non-aqueous media, this radical anion can oxidise hydroqui- nones 53 and aromatic dianions.69, 70 Vanillin, acetovanillone and guaiacol are resistant to the action of the radical anion in aprotic (pyridine, DMSO) solvents, while a-methylvanillyl alcohol is rapidly oxidised by this oxidant to acetovanillone and hydro- quinone is oxidised to acids.76 Despite the oxidation of some model compounds, the sulfate lignin itself does not undergo any noticeable changes except the partial loss of the phenolic hydroxy groups (according to the data of liquid chromatography, 1H NMR, IR and UV spectroscopy).77 ± 79 The hydroxyl radical HO. acts as an electrophile to form s-complexes with aromatic structural units of lignin.The latter undergo subsequent transformations involving cleavage of the side-chain C7C bonds of lignin, cleavage of the aryl ± aryl bonds and demethylation.80, 81 The radicals HO. also induce lignin hydroxylation, cleavage of biphenyl bonds and degradation of aromatic rings.43 In this respect, these radicals are more reactive than O2¡.. On the contrary, in reactions with carbonhydrates,71 ± 73 the radical HO. is less reactive: its addition to the aromatic fragments of lignin occurs fivefold faster than hydrogen abstrac- tion from carbohydrates. Unlike HO., the superoxide radical anions react with lignin with difficulty but easily destroy cellu- lose.70 This apparently is a result of a different ratio of the nucleophilic and electrophilic reactive centres in lignin and carbohydrates. Analogous stilbene derivatives were earlier 80, 81 obtained from diols of the type 2a,b.However, the a,b-elimination 2 .. The latter are detected by the products are formed in higher yields only in the case of structures It was found 70 that lignin and catechol are readily oxidised at pH 14 with hydrogen peroxide even at room temperature produc- ing the radical anions O¡ appearance of an intense violet colour in the presence of Tetrazolium Violet 1 ± 2 min after the beginning of the reaction. Such model compounds as guaiacol, isoeugenol, vanillic and 941 ferulic acids, b-guaiacyl ether of a-guaiacylethanone and b-guaiacyl ether of a-guaiacylethanol did not react with H2O2 even at 60 8C for 4 h.An important practical outcome of this study was the finding that catechol and the Pepper lignin do not virtually undergo oxidation at pH411.0. On increasing pH to R +O¡2 ., +O2 O7 O7 O7 OR R +2HO¡2 , +O2¡. H2O O7 O7 O O7 H2O HO¡2 H2O2+OH7, HO¡2 +H2O2 O2¡. +HO.+H2O. Thus, at high concentration of a base (NaOH), low content of oxygen and moderate temperatures the most probable reaction centres in lignin are the catechol structural units, and hydrogen peroxide acts on lignin in the same manner as does oxygen. The oxidation of dimeric model lignin compounds of the 1,2- diarylpropane-1,3-diol type with hydrogen peroxide and oxygen has been studied 79 in the example of 1-(4-hydroxy-3-methoxy- phenyl)-2-(4-hydroxy-3-methoxyphenyl)propane-1,3-diol (2a) 1-(3-ethoxy-4-hydroxyphenyl)-2-(4-hydroxy-3-methoxy- hydroxy-3-methoxyphenyl)prop-2-en-1-ol (4b) and 3-ethoxy- 4,40-dihydroxy-30-methoxystilbene (5b) (side reaction) dianions at pH 11 ± 13 in the absence of oxygen (in vacuum), the equili- brium mixture containing ca.95% of the initial dimer (Scheme 3). Scheme 3 HOCH2 HOCH2 HC O7 OH HC HC HCOH 7H+,7H2O OMe OMe H+, H2O 7H+ OR OR O 2a,b OH 3a,b HOCH2 CH C O7 7O OMe RO 4a,b 7O O7 CH CH 7CH2O OMe RO 5a,b R=Me (a), Et (b). containing hydroxymethylphenyl fragments.82 Vanillin and ethylvanillin instead of stilbene dianions 4b and 5b were obtained in experiments with preliminary passage of air or942 nitrogen through the system, thus indicating the oxidative character of the process. This result can be explained by rapid autooxidation of the initially formed stilbenes with a trace amount of oxygen dissolved in the alkaline solution. The oxidation of diol 2b with a solution of H2O2 stabilised at pH 10.5 results in the formation of ethanol, methanol ethylvanillin and small amounts of other compounds (the conversion of the initial diol is 30%) (Table 4). After 4 h of the treatment, 85% of hydrogen peroxide still remained unconsumed. Some amount of H2O2 is lost due to decomposing, and about 1%± 2% of the oxygen thus generated is consumed in the oxidation of diols as follows from the analysis of stilbene and the initial diol autoox- idation products. 2 Diols of the type 2 are less reactive towards H2O2 than the products of their initial oxidation.79 According to the composition of the oxidation products, the main direction of the oxidation of these diols with the stabilised hydrogen peroxide (i.e., in an oxygen-free system) involves the nucleophilic attack of the HO¡ anions on quinonemethide anion 3 and the oxidation of stilbene dianion 4.The hydroperoxide thus formed undergoes consecutive rearrangement and fragmentation yielding the hydroquinone dianion (6) and a-(4-hydroxy-3-methoxyphenyl)-b-hydroxypro- panol anion (11). The latter is converted into the guaiacylglycol anion (8) (Scheme 4). The oxidation of phenylpropan-2-one with hydrogen peroxide follows the same mechanism, this reaction may be regarded as an alkaline analogue of the Baeyer ± Villiger reaction.82 It should be noted that dihydroxyphenyl derivatives, like guaiacylglycol, have also been obtained by the action of basidiomycetes Phanerochaete chrysosporium on diols of the type 2.83 The same mechanism for conversion of a-hydroperoxide into guaiacylglycol and ethyl- vanillin has been proposed in another study.84 The mechanism of oxidation of stilbenes 4b and 5b with hydrogen peroxide in alkaline media in the absence (or virtual Table 4.Oxidation products of the diol 2b in a stabilised solution of hydrogen peroxide (A), with hydrogen peroxide in the absence of a stabiliser (B) and with atmospheric oxygen (C).70 Component of the reaction mixture Yield (with respect to the theoretical one) (mol.%)Cc Bb Aa 14 ± 18 31 70 1-(3-Ethoxy-4-hydroxyphenyl)-2- (4-hydroxy-3-methoxyphenyl)- propane-1,3-diol (2b) Ethoxyhydroquinone (6) Ethylvanillic acid d traces *1 0 " 4 0 Ethylvanillin (7) 6±7 *14 60 Guaiacylglycol (8) o-Hydroxyacetoguaiacone (9) Methoxyhydroquinone 1,2,4-Trihydroxybenzene (10) Methoxy- and ethoxybenzoquinones 3-(3-Ethoxy-4-hydroxyphenyl)-2- 2 4 7.5 0 2 traces 7 7 " 7 3.5 0 7 traces 1 ± 2 0 7 7 1 7 7 Acid (4-hydroxy-3-methoxyphenyl)prop- 2-en-1-ol dianion (4b) 3-Ethoxy-4,40-dihydroxy-30-meth- oxystilbene dianion (5b) 29 17 5 15 785 Methanol Ethanol Residual H2O2 36 4 ± 5 7 a Reaction conditions: 45 8C, 4 h, pH 10.5, [dimer]=0.0761073, [DTPA]=0.461073, [MgSO4]=0.1661073 mol litre71, [H2 O2]/ [dimer]=3.b Same reaction conditions but without addition of DTPA and MgSO4. c Reaction conditions: 45 8C, 4 h, pH 10.5. d Hydroxyethyl- vanillin in the case of the method B. V A Demin, V V Shereshovets, J B Monakov Scheme 4 HOCH2 HC O7 HO¡2 OH7 HCOH 2b OMe 7OH7 7H2O OEt O7 HOCH2 HOCH2 HC HC O7 O7 HC HCOOH HO7 O OMe OMe 7OH7 OEt OEt O O7 O7 HOCH2 O7 HC + , OEt CHO OMe O7 6 11 HOCH2 O7 HC HO¡211 HO O C O7 OMe H HOCH2 HOCH2 HO7 O7 + HCOO7. HC O7 HCOH O OMe OMe 8 CHO 12 absence) of oxygen is rather difficult to describe in detail since the hydroperoxide anion, as a strong nucleophilic reagent, cannot directly attack the electrophilic alkene structures.The oxidation occurs as the result of preliminary rearrangement of the stilbene dianion 4b into quinonemethide and subsequent attack by the hydroperoxide anion. o-Hydroxyacetoguaiacone, ethoxyhydro- quinone and acetoguaiacone were expected to be the products of these reactions. However, only a trace amount of ethoxyhydro- quinone has been detected (Table 4). Thus, the reaction of non- dissociated hydrogen peroxide with the stilbene 4b seems to be more probable.79 Comparison of the results of oxidation of the diol 2 with a non-stabilised alkaline solution of hydrogen peroxide at constant pH 10.5 with those obtained in its oxidation with stabilised H2O2 in the absence of oxygen (Table 4) allows one to reveal a number of differences, viz., the decomposition of non-stabilised hydrogen peroxide is almost complete, the degree of conversion for the initial diol is increased from 30% to 69%, the yields of the oxidation products (especially of ethanol, methanol, ethylvanillin and alkoxyhydroquinone) are higher, and the products contain the following organic acids:70 Yield (%) Succinic Oxalic Hydroxymaleic Malonic Glycolic 41 41 1 ± 2 51 ± 2 These differences are explained by the presence in the reaction mixtures of some additional oxidants: oxygen, hydroxyl radicals and superoxide radical anions along with non-stabilised hydrogen peroxide, which are formed in the reactions (1) ± (4).In the otherReactivity of lignin and problems of its oxidative destruction with peroxy reagents words, the oxidation process induces the generation of new oxidants.R1 R1 HO¡2 /H2O2 .OH or O2 OR2 OR2 OH O7R1 R1 ArH O. OH O7 O7 R1=OH, OAc; R2=Me, Et. The action of oxygen generated upon decomposition of hydrogen peroxide results in the formation of quinoid products. R R O2, H+ 7 OMe OMe O7 O H2COH 7 C OO CH OR O 13 H+ H2COH HOO C HO¡2 , OH7 CH EtOH + RCOOH OR OOH7 H2COH O HO C H O CH OR O7 16 943 R R R1 OH 7MeOH O OMe OH O O 7R2OH OR2 O7 Quinonemethides of the type 3 and o-quinones which appear in the reaction mixture are oxidised by hydroxyperoxide anions to carboxylic acids.CHR CHR CHR H H HO¡2 HO¡2 O OH O 7OH7 OMe OMe OMe O O O7 CHR CHR O OH R O7 HOO OMe OMe OOH O O OMe O When the diol 2b is allowed to react with H2O2 in the absence of a stabiliser and without maintenance of pH at 12.5, the pH value is decreased to 10 due to the accumulation in the reaction H2COH Scheme 5 O7 C HC OMe OR O2 O2 O7 4b OMe H2COH H2COH O7 O C O C O7 CH O CH OO 7 OMe OMe OR OR O7 O7 15 14 H+ OMe CH2OH H2COH C O CHO O7 O C HO¡2 , OH7 HO O CH OMe OMe MeOH + RCOOH + ArCOH + OR 7 O7 O7 9 OR O7OH7 H2COH H O C O7 O7 HO O CH OMe OMe OR O7 17944 mixture of carboxylic acids. Simultaneously, there is observed an increase in the portion of nonconsumed diol and an increase in the total yield of ethylvanillin and ethoxyhydroquinone (from 15% to 29%) as a consequence of the increased concentrations of hydro- peroxide anion and non-dissociated hydrogen peroxide in the solution.The particular contribution of oxygen generated in situ in the total oxidation process was estimated in the oxidation of the diol 2b with molecular oxygen, which allowed the authors 79 to propose a more detailed mechanism for this reaction (Scheme 5), which takes into account the published data.80, 81 As has already been shown, the diols 2a,b are partially converted in alkaline media into the stilbene dianions 4 and 5, which undergo intense oxidative destruction with oxygen. According to the much higher yield of methanol compared with that for ethanol, the reaction predominantly proceeds by a pathway involving the intermediate formation of hydroperoxide 15.The data considered above show that model structures like 2a,b have low reactivity towards hydroperoxide anion but can be relatively easily oxidised with oxygen to low-molecular-weight products. In the presence of hydrogen peroxide, stabilisers, and at pH410.5, the main oxidising reagent is the hydroperoxide anion, whereas molecular oxygen and its other reactive forms, i.e., O2¡., 2 and .OH, become the major oxidants upon increase in pH O¡ to 12. Alkaline hydrogen peroxide readily oxidises a,b-unsaturated ketones of the coniferyl aldehyde type and a,b-diketones but not a-ketones with an etherified phenolic hydroxy group, nor (virtually) propioveratrone and a-hydroxypropioveratrone.79 a-Ketones are oxidised through addition of the hydroperoxide anion to the electrophilic carbon atom of the carbonyl group with the intermediate formation of a quinoid compounds.Following recovery of the aromatic system, an ester is formed, which is then hydrolysed. The products of such reactions are vanillyl alcohol and a carboxylic acid with the same aliphatic radical as in the initial ketone 82 (Scheme 6). Scheme 6 H O R R O O R C O OH C C HOOH OH7 OMe OMe OMe O7 OH O7 O H R C O O OH R C O OH7 H2O +RCOOH OMe OMe OMe O7 O O7 In the case of model compounds of the general formula ArCH2C(O)Me, the intermediate formation of enolic tautomers is observed.10 An investigation of the oxidation of a series of ketones bearing different subsituents in the positions 3 and 4 of the benzene ring has shown 82 that these reactions mainly result in the rapid cleavage of the initial phenylpropan-2-one into benzalde- hyde and acetic acid (Scheme 7).The mechanism of these reactions apparently involves an attack by the hydroperoxide anion of an enolic tautomer on its C(a) atom and an intermediate formation of b-hydroxyhydroper- oxide anion, fast protonation of this anion by water and subsequent attack by the hydroperoxide anion on the C(b) atom of the initial carbonyl group, leading eventually to synchronous cleavage of the O7O bond in the intermediate organic peroxide V A Demin, V V Shereshovets, J B Monakov Scheme 7 OH O OH 7 HO¡2 H2O PhCHC PhCH C PhCH2C Me Me Me OOH HOO7 H HO¡2PhCH CMe PhCH CHMe 7OH7 7OH7 OH HO OH HO O O PhCHO+MeCOOH+H2O and the C(a)7C(b) bond.The oxidation of a,b-unsaturated ketones of the coniferyl aldehyde type to benzaldehyde occurs via the corresponding hydroperoxides and epoxides (Scheme 8). Scheme 8 O O H H O7 H C C C CH CH CH O CH CH HOOCH HOO7 HOO7 7OH7 OMe OMe OMe O[O7] O[O7] O[O7] CHO HCOO7 HOO7 low-molecular-weight products OMe O[O7] CHO CHO 7 CH CH CHO CH O OH HC CHO HOOH, OH7 7CH + OH OMe OMe OMe The reactions of the hydroperoxide anions with quinoid structures result either in the ring opening or in the splitting off of the side chains (Scheme 9).8, 85 In the oxidation of methoxy- substituted quinones, the demethoxylation and formation of methanol were also observed.The oxidation of the quinone rings represents apparently a longer reaction pathway in lignin oxidative destruction, even if decomposition of non-terminal units is considered. This requires a larger amount of an oxidising reagent for the scission of the carbon chains than for the side-chain oxidation. An important feature of the reactions shown in Schemes 6 ± 9 consists in participation of the hydroperoxide anion as a reactive nucleophilic species, which reacts with electron-withdrawing (electrophilic) sites of lignin. In an acidic medium, the mechanism of reaction of hydrogen peroxide with organic compounds is analogous to that of peroxy acids.In this case, either a protonated hydrogen peroxideH3Oá2 or a hydroxonium ion are those reactive electrophilic species which attack the aromatic structures.10 The oxidation of model lignin compounds with hydrogen peroxide and peroxy acids mainly results in the hydroxylation of the aromatic nuclei, oxidative demethoxylation, oxidative ben- zene ring opening, splitting off of the side-chains, cleavage of the C(b)7O ether bonds and formation of epoxides (Scheme 10).Reactivity of lignin and problems of its oxidative destruction with peroxy reagents R R O7 HOO7 7H+ O OO7 O O R R HOO7 O O O R O HOO7 O 7OH7 O7 O O7 O7 O HOO7 7H+ OMe O O O HOO7 O 7OH7 O7 O O7 CHR HRCOOH HOO7 OMe O O7 O HOO7 7OH7 OMe O Salts of peracetic, performic and mono- and diperoxysulfuric acids in alkaline media are employed as the oxidants for oxidative delignification of wood tissues and mechanical pulps.8 ± 10, 86 ± 88 Under conditions favouring stabilisation of the peroxy anion, this acts as a nucleophilic reagent.The oxidation of structures of the benzyl alcohol type with hydrogen peroxide and peroxy acids to ketones and aldehydes may follow the radical mechanism involving hydroxyl radi- cals.8 ± 10, 85 It should be noted that studies on the oxidation of compounds that mimic the lignin structural fragments are only useful in drawing a probable scheme (and in some cases, in elucidation of the mechanism) of reactions of the residual pulp lignin.In fact, the processes of oxidative destruction of the residual lignin under the action of a majority of peroxides do not go to completion (except for ozonolysis) and kinetic parameters determined for the oxida- tion of model compounds substantially differ from those observed during the lignin destruction in a polymeric lignocellulosic matrix. Scheme 9 R R O O7 O7 O 7 O O O O7 R OOH O HOO7 7OH7 O O7 O low-molecular-weight products O7 H2O OMe OMe 7MeOH OO7 O 7O O low-molecular-weight products HRC O HOO7 7RCHO, 7OH7 7OH7 OMe OMe O low-molecular-weight products R OH+ OMe OR OH+ OMe OH R OH+ OMe O [OH] R O O [OH] HRCOH OH+ OMe O [OH] MeO HC O CHOHOMe O [OH] MeO HO C O CH+ OMe O [O+H]OH OH+ C C C C 2.Ozonolysis of model compounds Reactions of ozone with organic compounds are widely used in organic synthesis, qualitative and quantitative analyses and chemical technology. The mechanisms of these reactions have been considered in detail.89, 90 In application to lignin, ozonolysis has been employed to prove its aromatic nature, to determine the nature of substituents in the aromatic rings 91 and to study the three-dimensional structure of side chains in the phenol ± propane structural units.92 945 Scheme 10 R R H HO HO ; 7H+ OMe O+Me O O R R OH ; O OMe 7H+, 7MeOH O O+H R OMe 7H+ OH O+[O+H] R OH+ OMe OMe ; 7H+ O O O [OH] OH HRCOHOH 7RCHO, OMe 7H+ + ; OMe O [OH] O [O+H] MeO R C O R CH OH+ 7H2O OMe O [OH] R CO CHOH H2O ; R, 7HO OMe MeO O [OH] 7H+ O + C C .7H+946 Scheme 11 R1 R1 R1 O3 + ; H MeO OH 7O2 O O O7 MeO MeO OR2 OR2 OR2 R1 R1 R1 O3 ; + OMe 7O2,7MeOH O O O7 O OMe OH OH O R1 R1 R1 H2O O3 OMe OMe 7H2O2 O OMe R2O R2O O OR2 O O O R1=Alk, R2=H, Alk. The first stage of the reaction of ozone with the aromatic fragments of lignin involves its electrophilic addition, and o-quinones or muconic acid derivatives (following ring opening) are the common final products of ozonolysis (Scheme 11).The ozonolysis of aromatic compounds is accompanied by the formation of strong oxidants, viz., oxygen, hydrogen peroxide, ozonides and trioxides, which react in an acidic medium with lignin as electrophiles.89, 90 On comparison with monocyclic aromatic compounds, poly- cyclic derivatives react with ozone more readily, however, after incorporation of one or two molecules of ozone the rate of the process sharply decreases. Thus the reaction with naphthalene is stopped after addition of two molecules of ozone, and the reaction rate with phenanthrene is reduced by two orders of magnitude after the addition of the first molecule of O3.90 In the preparation of ozone from oxygen, the concentration of the latter in the system is high, and all intermediate radicals are predominantly converted into the peroxide ones (according to EPR spectroscopy).These radicals are responsible for the chem- iluminescence and chain oxidation of hydrocarbons. The oxidation of alkenes, alcohols and ethers in the presence of ozone leads to the formation of carbonyl compounds; alde- hydes are oxidised to carboxylic acids.93 The cleavage of the C7H bonds occurs upon oxidation of oxygen-containing compounds.90 Reactions with phenols predominantly proceed through the abstraction of the hydrogen atoms from the hydroxy groups, leading to the formation of phenoxyl radicals and subsequent aromatic ring opening. Based on the analysis of the final products, one can judge about the structure of a particular lignin elementary unit.92 ± 94 Thus the formation of two molecules of oxalic acid per guaiacyl fragment points either to the absence of aliphatic Scheme 12 R1 1 6 2 COOH O3 2 3 5 COOH ; OMe 4O R1 R1 R2 COOH COOH O3 O3 2 .; COOH COOH OMe OMe R2 O O R1=R2=Alk, Ar. V A Demin, V V Shereshovets, J B Monakov substituents in the benzene ring or to its presence in position 6, while the formation of one molecule of the acid suggests the presence of a substituent in position 5 (Scheme 12). It has been found 91 that preliminary reduction of lignin with sodium borohydride favours a higher yield of water-soluble acids upon subsequent ozonolysis. Obviously, the reduction of the constituent aldehydes, ketones and quinones simplifies the lignin structure and accelerates the oxidation process, which mostly affect the aromatic structures.However, sodium borohydride is rather expensive to be used for the industrial pulp bleaching.{ The introduction of electron-withdrawing substituents, e.g., NO2, Cl, etc., into a molecule of an unsaturated compound substantially reduces the rate of its oxidation with ozone due to a decrease in the electron density on the carbon atoms of the p- bonds.90 This is quite important from the practical point of view, since pulp bleaching involves sequential treatment with various oxidants, including nitrogen oxides.96 Oxalic acid and its monomethyl ester, carbon dioxide and esters are the products of the oxidative destruction of the lignin C(b)7O7C(4) fragment 93 (Scheme 13).Scheme 13 g CH2OH O 4 CH2OH HC b O R2 HC O R2 a CHOH CHOH MeO O MeO O3 O3 O OMe OMe O OR1 OR1 g O CH2OH HC b O O HO R3 aCHOH + + R2+ R1O R4 O OH O COOMe COOH + + CO2 + COOH COOH Thus ozone can react with virtually all types of the lignin structural units, including those condensed and linked to each other with carbon ± carbon bonds, e.g., C(b)7C(1).92 An important point in the bleaching of sulfate pulp is that the depth, selectivity and rate of lignin oxidative destruction can be controlled by varying the functional composition of lignin using reduction, oxidation, etc. Ozonolysis of lignin-containing technical pulp aimed at its bleaching results in the oxidative destruction of the residual lignin, which is accompanied by oxidation and destruction of polysac- charides and cellulose (the latter process is usually undesirable).97 The accumulation of the carbonyl groups in the cellulose macro- molecule reduces its chemical resistance, results in yellowing and enhances pulp destruction upon subsequent processing.It is known that a decrease in the degree of polymerisation below a certain limit affects the durability of the paper products. The oxidative destruction of cellulose with ozone is a serious problem, which for a long time has been (and still is) an obstacle for the wide application of ozone in the pulp industry, together with the high cost of ozone and complexity of technological equipment for such processes. However, targeted destruction of polysaccharides, e.g., cellu- lose, chitin and chitosans,98 is of particular interest for their conversion into the corresponding oligomers.In this case, the ozonolysis compares favourably in efficiency and ecological safety to other methods of oxidative destruction. { The cost of 1 kg of NaBH4 is approximately equal to that of all other chemicals consumed in the bleaching of 1 ton of a pulp.95Reactivity of lignin and problems of its oxidative destruction with peroxy reagents 3. Oxidation of lignin with chlorine dioxide Chlorine dioxide oxidises lignin in lignin ± cellulose matrices without affecting cellulose. Lignin phenolic structural units are oxidised with ClO2 upon abstraction of the hydrogen atoms from the phenolic hydroxy groups.Muconic acids and quinoid com- pounds are the final products of such reactions, which are formed without loss of the methoxy groups 99 ± 101 (Scheme 14). R OMe O ClO2, H2O 7HClO R OMe O HO O In the case of aromatic ethers, e.g., 4-methyl-2,30,40-trime- thoxydiphenyl ether (18), three different directions for the reaction with chlorine dioxide are possible.99, 100 These involve the formation of quinoid structures, opening of one of the aromatic rings and cleavage of the diaryl ether bond leading to the formation of hydroquinone and phenol derivatives (Scheme 15). Veratryl b-guaiacyl ether (19) undergoes homolytic fragmen- tation under the action of chlorine dioxide, one of the reaction products being veratraldehyde. In parallel, the oxidative dehy- drogenation of the a-hydroxy group occurs (Scheme 16).The oxidation of 4,40-dihydroxy-3,30-dimethylstilbene (20), 4-hydroxy-3,30,40-trimethoxystilbene (21) and 3,4,30,40-tetra- methoxystilbene (22) (the most common model compounds of native and residual lignins) with chlorine dioxide has been studied.99 ± 101 The main transformations of stilbenes 20 ± 22 are shown in Scheme 17. Yet another model compound of the residual lignin is 3,4- dimethoxy-b-(2-methoxyphenoxy)styrene (23). It reacts with chlorine dioxide by a similar mechanism (Scheme 18). The oxidation involves the formation of a complex of the oxidant with the p-electronic system of the alkene fragment.101 In an acidic medium, this complex is protonated to form a resonance-stabi- lised radical cation and HClO2.The addition of ClO2 to the b-radical generated leads first to the chlorous ester (A), which is H2COH HCOAr C O 7H+ OMe OMe Ar=C6H4OMe-o. Scheme 14 R R OMe OMe O O ClO2 ClO2 7R,7ClO7 7HClO2 O R OMe OMe O OH H2COH H2COH HCOAr HCOAr HC O HCOH ClO2 ClO2, H+ 7HClO2 7HClO2 OMe OMe OMe 19 OMe H2COH 7 HCOAr CHO OMe OMe 947 Scheme 15 OAr OAr OClO ClO2 H + 7H+ OMe OMe OMe OMe 18 OAr OAr O OClO H2O 7HClO, 7MeOH O OMe OMe OMe OAr OAr OAr H2O ClO2 18 OH O + 7HClO OMe 7H+ OMe OClO OMe OClO MeO MeO MeO O ArO OClO OH ArO H2O ClO2 18 7H+, 7HClO2 OMe OMe OMe MeO OH + O ArOH + OMe O Me.Ar=MeO further converted into the ester B, from which aldehydes, ketones and esters are formed as the final products. A possibility of the formation of 1,1-bis(3,4-dimethoxy- phenyl)ethylene (24) from the corresponding stilbene as a result of aryl radical migration has been shown.102 MeO CH2 OMe MeO C OMe 24 Scheme 16 H2COH H2COH H2COH HCOAr HCOAr HCOAr CO HCOH HCOH +. 7H, 7H. OMe OMe OMe OMe O+Me OMeH2COH 7 HCOAr HC O H CHO + 7H+ OMe OMe OMe OMe948 MeO R1O CH CH 20 ± 22 MeO R1OMeO R1O MeO R1O R1=R2=H(20); R1=H, R2=Me (21); R1=R2=Me (22). OMe CH CH O 23 OMe OH CH O CH OClO B OMe OH CH OH + OHC OHC OMe CO OMe Unlike stilbenes 20 ± 23, compound 24 does not undergo fragmentation under the action of chlorine dioxide.Among the ClO2: OR2 OMe +CH CH CH CH H CH C O+R2 OClO OMe ClO2: OMe OMe OMe HOH OMe 7H+ OMe 7HClO OMe OMe OHCCH2 OMe V A Demin, V V Shereshovets, J B Monakov MeO ClO2 CH R1O CH OR2 OMe OR2 MeO OMe +. CH R1O CH OR2 O+R2 OMe OMe HOH aryl migration CH OClOCH 7HClO2,7H+ MeO OR1 OR2 OMe hydride shift OClOC+ 7HClO2,7H+ CH2 MeO OR1 OMe ClO2 CH CH OMe O OMe O CH CH O+Me MeO OClO OMe A MeO OH OMe CH O CO OMe Cl2 OHCCHCl OMe OMe OMe reaction products, only mono- and dichlorinated derivatives have been found. Oxidative demethoxylation also takes place.Scheme 17 H+ 7HClO2 ClO2: O+R2 OMe OR2 OMe OHCCH MeO OR1 OR2 OMe HOH CO CH2 MeO OR1 Scheme 18 MeO MeO O CH CHOReactivity of lignin and problems of its oxidative destruction with peroxy reagents In general, the early stages of the reactions of model lignin compounds with chlorine dioxide result in the formation of conjugated carbonyl groups, i.e., in new chromophoric groups and systems which, as has been shown earlier, are readily oxidised with hydrogen peroxide according to the ionic mechanism involv- ing the nucleophilic addition of the hydroperoxide anion. IV. Combined action of oxidants on the sulfate pulp lignin As has already been mentioned, lignin remaining in the sulfate pulp after cooking represents the most condensed, and strongly bound to polysaccharides, part of modified natural lignin (residual lignin).Despite the fact that the composition (various combinations of bonds, functional groups and structural units) and three-dimensional (network 5, 7, 26 and fractal 22, 24, 25) poly- meric structure of lignin do not allow prediction of the degree and the rate of its oxidative destruction, analysis of the results of the studies on an oxidation of model lignin compounds with hydrogen peroxide, oxygen, ozone, chlorine dioxide and other reagents (vide supra) makes it possible to elucidate general features of this process.8 ¡À10 The reactivity of lignin towards oxidants depends on the nature of the reactant and the presence of nucleophilic and electrophilic sites in the lignin macromolecule. Under conditions of acid and base catalysis, each particular reagent can produce differently reacting species depending on the medium acidity. A classification of reagents 10 used for lignin oxidation according to their electron-donor and electron-acceptor properties (or nucleo- philicity and electrophilicity, respectively) is given in Table 5.The mechanisms of their interaction with lignin and the reactivity of residual lignin itself substantially affect the rate and the depth of the sulfate pulp delignification and a number of the operation steps in pulp bleaching. The oxidation of nucleophilic alkene fragments in lignin occurs as the initial addition of an electrophile followed by the attack of a nucleophile (Scheme 19).10 The addition of electro- philes to lignin results in the appearance of new electrophilic sites in the lignin macromolecule and their activation towards the action of nucleophiles.If one form of the reagent predominates and no cooperation exists between nucleophilic and electrophilic species, the rate of dissolution of the electrophilic reaction products of residual lignin is substantially decreased, which apparently leads to deceleration of the sulfate pulp delignification and to non-rational consump- tion of bleaching reagents in the chemicotechnological processes. The efficiency of the sulfate pulp delignification is especially low when such environmentally safe reagents as hydrogen peroxide and ozone are used compared to chlorine-containing reagents (Table 6).103 Normally, complete dissolution of the reaction products of lignin with electrophilic oxidants requires treatment with alkali, i.e., the second stage in the pulp bleaching process is the treatment of pulps with a nucleophilic reagent (alkali).As a rule, Table 5. Reactive forms of the reagents, involved in pulp bleaching processes. Substrate Reactive form Reagent type Process (medium) electrophile cation radical aromatic and unsaturated structures Oxidation (acidic) nucleophile anion radical non-conjugated and conjugated carbonyl compounds Reduction (alkaline) Note. The list of reacting species includes both commercial bleaching reagents and the species that are not commercial bleaching reagents but can take part in various pulp bleaching processes.9 949 Scheme 19 C X C+ X C C C X2 d+ d7 Y7 X X 7X2 7X7 C Y C C + C X C X, Y=Cl, OH.Table 6. The depth of oxidative destruction of lignin in sulfate softwood pulps with equivalent amounts of different reagents.95 Reagent (conditions) Oxidant Depth of consumption residual (% of the lignin pulp mass) a oxidation b 0.19 0.30 0.42 0.59 1.26 0.59 Ozone Hydrogen peroxide Ozone+extraction with a NaOH solution Sodium hypochlorite (pH 10) Sodium hypochlorite+hypochlorous 2.63 c 2.63 c 0.53 0.64 0.67 1.00 acid (pH 7.5) Chlorine dioxide a With respect to the reagent mass.b Corresponds to the degree of lignin conversion; the initial lignin content in sulfate softwood pulp is expressed as 76 permanganate units (1 p. u. is referred to the content of lignin in pulps according to Bj��rkman, State standard 6845-54). c Per active chlorine unit (active chlorine is defined as an overall amount of chlorine- containing oxidants determined by iodometry). Cl2+NaOH and O3+NaOH pairs are used for this purpose. In the second stage, some part of chlorine in chlorolignin is replaced by hydroxy groups, thus enhancing the hydrophilicity of the chlorinated lignin. Delignification of the sulfate pulp can be enhanced by applying `oxidative alkalification', whereon oxidants such as sodium hypochlorite or hydrogen peroxide are added to an alkaline pulp solution.104 The hypochlorite anion (ClO7) acts on lignin as a nucleophilic reagent, while H2O2 in the absence of stabilisers and at high alkalinity of the medium generates various forms of oxidants: 1 O7.+.OH O2.HO2. H2O2 O22 ¡¦ The electrophilic mechanism predominates in the oxidation of lignin;10, 15, 18, 105 this has earlier been shown for some lignin model compounds.75 ¡À 80 Reacting species Number of electrons involved in the reaction 2 H+, OH+, O3, Cl+, NO+, NO�¢ 2 H., HO., HO2., .O2., Cl., ClO., ClO2., 1 NO., NO2. 4 H7, HO7, HO¡¦2 , ClO7, ClO¡¦2 , S2O2¡¦ 21 H., eaq, SO2¡¦.950 The most important characteristic of a delignifying reagent is the minimum content of lignin in pulp rather than the portion of lignin removed by this reagent from the pulp (bearing in mind that the initial content of lignin in pulps may vary) at which delignifi- cation with this reagent, taken in small excess, is terminated. According to this criterion, the reagents used for delignification of the sulfate pulp (after cooking) can be arranged in the following order: oxygen<hydrogen peroxide<chlorine (with subsequent extraction with a NaOH solution)<sodium hypochlor- ite<chlorine dioxide.103 Lignin in industrial pulps (including sulfate pulp) cannot be selectively destroyed to give water-soluble products and com- pletely removed from pulps by either of the oxidants.To achieve complete delignification (and concomitant bleaching) of sulfate pulp, it is necessary to combine oxidative treatmth different reagents in several (from two 106 to five ¡À seven 107) stages.Since the 1950s, a large number of publications has appeared that are devoted to the problem of more efficient lignin oxidation with oxygen and hydrogen peroxide aimed at the development of environmentally safe technologies. It should be noted that hydro- gen peroxide was used first only for the `oxygen-induced' type of lignin transformations (quite often together with molecular oxy- gen 107) because most of researchers referred to the hydroperoxide anion exclusively as the class of `lignin-preserving reagents',10, 75 probably judging from the results of industrial application of these reagents in pulp bleaching. V.Lignin activation towards oxidative action of oxygen and hydrogen peroxide Not only chlorine but also nitrogen oxides NO and NO2 can be used in the bleaching of sulfate pulps.108 The pulp treated with these oxides is then better bleached with sodium hypochlorite or hydrogen peroxide rather than undergo alkaline extraction as a version of processing. The use of nitrogen oxides or a system HNO37NaNO2 to increase the efficiency of subsequent sulfate pulp delignification with oxygen has been described.109 ¡À 112 The treatment with nitro- gen oxides was applied as an `activating stage'. According to the published data,111 the effect of activation is based on lignin fragmentation as the result of electrophilic substitution and elimination of side chains in phenol ¡À propane structural units, together with partial destruction of the residual lignin.The electrophilic reagent in this process is NO�¢2 generated in the reaction mixture. 113 fast + NO¡¦ HO NO2+HNO3 3 +HO NO2, H slow + + +NO2, H3O++NO¡¦3 HO NO2+HNO3 HCH(OH)R CH(OH)R O2N +NO�¢2 OMe OMe O2N +OH OH NO2 H . +H++R O OMe O2N OH Preliminary activation with nitrogen oxides makes it possible to reduce twofold the content of lignin in pulps (hardness) following treatment with oxygen (or H2O2 under conditions of the `oxygen' 109 ¡À 112 mechanism): from 10 to 5 Kappa units } for } The Kappa unit characterises the lignin content in industrial pulps accepted as a standard unit in Russian State Standard 1 007 074 and in a number of foreign standards.V A Demin, V V Shereshovets, J B Monakov different sorts of pulps. The minimum value of pulp hardness given in some studies 109 ¡À 112 equals 3.5 Kappa units, while that in another study was 114*7.0 ¡À 7.5 Kappa units. In the latter work, the pulp pretreatment with other oxidants (chlorine, chlorine dioxide, ozone, etc.) was proved to be efficient. It has also been found 111 that the selectivity of the sulfate pulp delignification with oxygen and hydrogen peroxide is higher in the case of preliminary pulp treatment with chlorine and ozone than with nitrogen oxides. The minimum hardness of pulps was 6 ¡À 9 Kappa units after treatment with chlorine (chlorine usage rate was 1%¡À 3% with respect to the mass of absolutely dry pulp) and 6.1 Kappa units after ozonolysis (O3 usage rate D 1%) at the degree of polymer- isation in the range of 1300 ¡À 1400. This degree of polymerisation without pulp pretreatment was observed only for pulps with hardness of about 15 Kappa units (after pulp delignification with oxygen under pressure of 0.5 MPa in the presence of 0.5% MgSO4 at 110 and 120 8C or with hydrogen peroxide under the same conditions in the presence of 4% NaOH).It should be noted that a mixture of nitrogen oxides and oxygen was initially proposed to be used in the pulp pretreatment step for lignin activation.112 However, oxygen only caused stronger destruction of cellulose. Methods involving lignin acti- vation with nitric acid permit removal of up to 75% of residual lignin from pulps.} One of extremely toxic side products of the lignin nitration is hydrogen cyanide, which is formed at high concentrations of nitrogen oxides.Sodium nitrite, nitrous acid, peroxynitric acid as well as a redox-composition NaNO27H2SO4 and some others have been proposed as safer reagents.116, 117 After oxidation in these systems and delignification with oxygen, the content of lignin in pulps is reduced to 8 ¡À 9 Kappa units. The common principle of the methods of lignin oxidation considered above (oxidation during bleaching of the sulfate pulp) based on lignin activation by the action of nitrogen oxides, chlorine and chlorine dioxide is as follows: (E++H3O+)7(E++OH7), i.e., an electrophile in an acidic medium ¡À an electrophile in an alkaline medium (see also Refs 118 ¡À 121). It has been shown 122 ¡À 126 that hydroperoxide anions also delignify pulps even somewhat better than oxygen and concom- itant radicals generated upon hydrogen peroxide decomposition (or directly upon the action of O2).In the absence of a hydrogen peroxide stabiliser and at pH512, the realisation of the mecha- nism H2O2?O2 leads to 45% delignification of the sulfate softwood pulp, while in the presence of a stabiliser (sodium silicate, etc.) and at pH about 10 solubilisation of the residual lignin reaches 50% at the initial content of lignin of 76 permanga- nate units (p.u.).127 However, even this depth of lignin conversion is achieved only in the case of pretreatment of unbleached pulp with a weakly acidic solution for removal of cations of transition metals. Preliminary washing of sulfate pulps with acids or solutions of chelating reagents reduces the loss of hydrogen peroxide due to its decomposition and to some extent decreases pulp destruction with superoxide radical anions. Methods of acid-catalysed and acid-electrophilic activation of the residual lignin based on acid treatment of the residual lignin in sulfate pulps and aimed at partial hydrolysis of the C(b)7O7C(4) ether bonds at temperatures above 70 8C have been proposed.123 ¡À 126, 128 ¡À 131 Such pretreatment not only changes the structure of the residual lignin, but also activates it to the action of the hydroperoxide anions owing to formation of the carbonyl groups near the site of the ether bond cleavage.Following acid-catalysed activation, the degree of lignin conver- sion in sulfate hardwood pulps is increased from 0.5 to 0.8 upon the oxidative destruction with hydroperoxide anions. The addi- tion of an oxidant (hydrogen peroxide, monoperoxysulfuric acid) } Quantification of lignin in pulps expressed in Kappa units gives overestimated values for the depth of delignification compared to that determined by photometry.115Reactivity of lignin and problems of its oxidative destruction with peroxy reagents at the stage of activation leads to additional fragmentation of lignin and its more rapid solubilisation in the next technological step including oxidation with the hydroperoxide anions. The acid-catalysed activation makes somewhat more efficient the lignin destruction under the action of oxygen in highly basic media.However, this effect is nearly the same as that achieved on treatment of lignin with acids at room temperature (removal of cations and destruction of sulfides 130) or with chelating reagents 132, 133 where the maximum degree of lignin conversion is about 0.5 and cellulose undergoes higher hydrolytic destruction. In general, all schemes of oxidative delignification of the type `reagent+catalysis' can be arranged according to the degree of lignin conversion in sulfate pulps in the following order:131 oxygen in alkaline media (O2+OH7); hydroperoxide anions in alkaline media (HOO7+OH7); acid-catalysed activation+oxygen in alkaline media [(H3O+)7(O2+OH7)]; acid-electrophilic acti- vation+oxygen in alkaline media [(NOá2 +H3O+)7(O2+ OH7)]; acid-catalysed activation+hydroperoxide anions in alkaline media [(H3O+)7(HOO7+OH7)]; acid-electrophilic activation+hydroperoxide anions in alkaline media [(OH++ H3O+)7(HOO7+OH7)]. The selectivity of action of hydroperoxide anions on the residual lignin in sulfate pulps is described by a plot of the degree of cellulose polymerisation as a function of the ee of lignin conversion (Fig.1). PV 1400 1 2 1200 1000 800 [L] 0.6 0.8 6001.0 0.4 Figure 1. Dependence of the degree of polymerisation (P V, measured by viscosimetry) on the relative lignin content {[L]} of sulfate softwood cellulose upon delignification under the action of HOO7: (1) without preliminary activation; (2) after preliminary acid-catalysed activation.In the studies on the mechanism of the acid-catalysed activation it was found that the reduction of activated lignin in sulfate pulps with sodium borohydride at pH 10 leads to partial deactivation of lignin 129, 130 towards the hydroperoxide anions. In this particular case, successive treatment of lignin with two different nucleophilic reagents did not give the expected result, although the former (NaBH4) acts as a reducing agent, and the latter (HOO7) acts as an oxidant. VI. Ozonolysis of sulfate pulps Ozonolysis is one of the methods of delignification of sulfate pulps in the course of bleaching.The selectivity of lignin oxidation in the ozonolysis depends on the conditions and, first of all, on the pH of a medium. In neutral media (pure water), destruction of cellulose predominates.97 To perform the sulfate pulp delignification, ozonolysis is carried out in acid media (pH 2 ± 3) in which, as well as in alkaline media, the destruction of cellulose is substan- tially lower.97 The selectivity of delignification is also influenced by the character of the preozonolysis treatment and by the lignin content in pulps subject to ozonolysis.131 The products of lignin ozonolysis formed upon treatment of unbleached sulfate softwood pulps with ozone are not dissolved instantly (Figs 2 and 3, curves 1).Using photometric assay of the lignin content, it was found that the changes in selectivity of the 951 [L] /p. u. 100 80 60 1 40 2 20 2.0 1.0 0.5 0 1.5 [O3 (consumed)] expressed in percent of the pulp mass Figure 2. Dynamics of changes in lignin content {[L]} in sulfate softwood pulp during its delignification with ozone: (1) ozonolysis; (2) ozonolysis and subsequent extraction with a solution of NaOH; p. u. are permanga- nate units.Pv 1400 1200 2 1 1000 800 0.8 0 0.6 0.4 0.2 y Figure 3. Selectivity of delignification of sulfate softwood pulp with ozone (y is the degree of lignin conversion equal to the portion of solubilised lignin): (1) after ozonolysis; (2) after ozonolysis and extraction with a solution of NaOH.oxidative processes during the ozonolysis of sulfate softwood pulps have a rather unusual character. At high lignin contents in pulps, the destruction of cellulose is predominant. Then, in parallel with a decrease of the lignin content, the direction of the process is gradually changing and, after dissolution of a half of the initial lignin, the selective pulp delignification becomes the major process.{ A part of ozonolysed lignin is dissolved in an alkaline medium (Figs 2 and 3, curves 2). Special tests have shown that destruction of cellulose upon ozonolysis in the presence of lignin is higher indeed. The ozonolysis of pulps partially delignified by other reagents (i.e., semibleached) is also accompanied by destruc- tion of cellulose.However, this destruction is always less than in the ozonolysis of unbleached pulps with high initial lignin content. The highest degree of the residual lignin oxidation and complete solubilisation of the ozonolysis products in acid media are observed in the case of lignin pretreatment with hydroperoxide anions.136 A similar effect is achieved on treating lignin with sodium borohydride. Obviously, a decrease in the number of carbonyl and aldehyde groups in the side chains of the lignin phenol ± propane units due to their oxidation or reduction (upon the action of HOO7 or BH¡4 , respectively) favours more selective oxidation of lignin aromatic rings with ozone. The best results are achieved by successive treatment of lignin, at first, with hydrogen peroxide in an acid medium (with { Assay of lignin content {[L]} by permanganate method 127 (especially in Kappa units 107, 127, 134, 135) in pulps with a low lignin content typical of semibleached pulps can be incorrect because of deviation of the relation- ship `Kappa number ± lignin rate' from linearity at [L]41%± 2%.952 Table 7.Effect of sulfate softwood pulp pretreatment on solubilisation of the residual lignin in the ozonolysis (reagent usage rate constituted 0.225% from the mass of absolutely dry pulp).131 Lignin content after ozonolysis a Pretreatment scheme Lignin content prior to ozonolysis a O3 ±NaOH O3 Without pretreatment OH+±HOO7 68 11 23 33 41 91 33 OH+± (H2O2 ? O2)/NaOH 39 51 58 87 11 30 38 46 OH+±NaOH H3O+ a Expressed in permanganate units.b At pH&2. peroxyacids), then with a solution of hydrogen peroxide stabilised at pH 10 and finally with ozone in an acid medium (OH+?HOO7?O3, Table 7).131 Thus, according to both absolute and relative parameters, the maximum solubilisation of the residual lignin in the ozonolysis of semibleached sulfate pulps is achieved after pretreatment with hydroperoxide anions. On the whole, with consideration of sulfate pulp cooking and preliminary acid-catalysed electrophilic activa- tion, the optimum sequence of the operations is as follows: sulfate cooking (SH7+OH7)?acid-catalysed electrophilic activation (OH++H3O+)?delignification with hydroperoxide anions (HOO7+OH7)?ozonolysis (O3+H3O+), i.e., an alterna- tion of treatments in acid and alkaline media.VII. A schematic approach to the selective destruction of residual lignin with peroxy reagents The oxidative destruction of lignin as a heterochain polymer containing ether bonds predominantly of the C(b)7O7C(4) type occurs under conditions of acid or base catalysis.12, 13, 137 The alternation of the action of acids (H3O+) and bases (OH7) allows one to achieve partial sulfate pulp delignification.124 It should be noted, however, that the treatment of pulps according to the HCl7NaOH scheme requires rather drastic conditions and the selectivity of delignification is low: along with twofold reduction of the lignin content, the degree of polymerisation of cellulose is also diminished almost twice.122 The analysis of the published data on delignification of industrial pulps with chlorine- containing reagents and oxygen in alkaline media as well as the experimental studies of various compositions of oxidising reagents and media have shown that the oxidation schemes without alternation of treatments in acid and alkaline media are less efficient (Table 8).131 Table 8.Schematic presentation and operational schemes for oxidative destruction of residual sulfate lignin. Operational scheme Schematic presentation (N7)7E+7N7 (Na2S)7H2SO57H2O2 (Na2S)7HCl7H2O2 (Na2S)7H2SO47H2O2 (Na2S)7Cl27NaOH (Na2S)7NO27O2 /NaOH (N7)7E+7E+ (Na2S)7NO27H2O2 /NaOH (N7)7E+ (Na2S) ±O2 /NaOH (Na2S) ±H2O2 /NaOH (Na2S) ±H2O2 (N7)7N7 Note.Here and in Table 9, the reagents and species participating in the process of sulfate cooking are presented in parentheses; E+, electrophile, N7, nucleophile. V A Demin, V V Shereshovets, J B Monakov Degree of delignification in one and two stages (%) Portion of solubilised lignin (%) b O3 O37NaOH 16 100 56 71 72 25 67 41 35 29 4 67 23 25 21 The properties and reactivities of lignins depend on the methods of their preparation. Therefore, the schemes in Table 8 include the stage of sulfate cooking. In order to achieve selective oxidative lignin destruction under the action of peroxy reagents with minimal changes in cellulose, it is necessary to apply a complex of different chemical operations which combine a pulp delignification stage and a stage of activation of the residual lignin towards the action of other reagents. The residual lignin is more labile than cellulose, and its reactivity depends on the method of its preparation or pretreat- ments.Therefore, enhancement of reactivity of lignin as a result of activation is also a method for an increase in the selectivity of the sulfate pulp delignification. An electrophilic attack of lignin by peroxy reagents (hydrogen peroxide, peroxy acids, ozone) is realised in acid media, whereas the presence of stabilisers and weakly alkaline media are necessary for treatment of lignin with nucleophiles.8±10 On the basis of the aforesaid, the most important factors affecting the selectivity and depth of the sulfate pulp (and other pulps) delignification can be deduced.These are: (1) the reactive form of a reagent; (2) the nature and the number of reactive sites in the residual lignin; (3) acid ± base catalysis; (4) cooking method and type of pulp pretreatment. As the result of an analysis of the selectivity and depth of oxidative destruction of lignin with different chlorine- and oxy- gen-containing reagents, a general statement concerning the selectivity of oxidative lignin destruction with peroxy reagents has been formulated:131 the oxidative destruction of the residual lignin is a maximum provided the stages based on the action of electrophilic and nucleophilic reagents alternate, and that the Ref. Catalysis Degree of lignin conversion (%) 80 ± 83 70 ± 80 70 ± 80 70 ± 80 50 ± 70 ±(OH)7H3O+7OH7 (OH)7H3O+7OH7 ±(OH7)7H3O+7OH7 50 ± 70 (OH7)7H3O+7OH7 450 450 <50 104, 116 112, 114 116 39, 124, 125 98 ± 102, 105, 107 ± 111 98 ± 102, 105, 107 ± 111 37, 96, 98, 105 37, 96, 98, 105 104, 112, 114 (OH7)±OH7 (OH7)±OH7 (OH7)±OH7Reactivity of lignin and problems of its oxidative destruction with peroxy reagents most complete destruction of the chromophore-containing struc- tures is achieved by the action of electrophilic reagents in acid media and of nucleophilic reagents in alkaline media as described by the general formula [(E++H3O+)7(N7+OH7)]n. At first glance (from the position of organic chemistry), this statement seems to be obvious.However, its formulation became possible only after development of the methods of pulp delignifi- cation with hydroperoxide anions, acid-catalysed and acid-elec- trophilic activation of the residual lignin as well as after studies on ozonolysis of unbleached and semibleached sulfate pulps 114, 124 ± 126, 128 ± 130 and generalisation of studies on acti- vated pulp delignification with oxygen in alkaline media.109 ± 112, 118 ± 121 All this made it possible to establish a close relationship between the general concepts of organic chemistry and new experimental results obtained in studies on the sulfate pulp delignification with peroxy reagents. Deviations from the order of operations described above (including the cooking stage) lead to substantial decrease in the lignin reactivity towards oxidants (even to such a reactive one as chlorine dioxide) and, as a consequence, to a decrease in the depth of oxidative delignification and to loss of the pulp brightness (Table 9).Table 9. The effect of the sequence of oxidative treatments on the depth of delignification of sulfate softwood pulps. Reagents Run Schematic presentation Pulp bright- ness (%) Degree of lignin con- version (%) 79 62 80 65 64 57 91 85 82 87 52 40 (N7)7E+7N7 ClO2±H2O2 (N7)7N77E+ H2O2 ± ClO2 (N7)7E+7N7 ClO2±H2O2 (N7)7N77E+ H2O2 ± ClO2 (N7)7E+7N7 H2O2±H2O2 (N7)7N77E+ H2O2±H2O2 123456Note.Usage rates for reagent are the same in pairs of runs: (1) and (2), (3) and (4), (5) and (6). Usage rate: ClO2 in runs (1), (2), 0.8%, in runs (3), (4), 1.3% from the pulp mass; H2O2 in runs (1) ± (4), 2.0%, in runs (5), (6), 1.0% [run (5) at pH 2 and 10, run (6) at pH 10 and 2 in the first and second stages, respectively]. On the basis of the statement mentioned above, a number of methods for sulfate pulp bleaching and ways of activation of the residual lignin with hydrogen peroxide and ozone have been proposed 124, 125, 129, 130 which are not inferior in the selectivity of delignification to the bleaching processes based on the use of chlorine-containing compounds. VIII. Kinetic methods for assessment of residual lignin reactivity Depending on the mechanism of lignin destruction, different kinetic models for assessment of its reactivity are employed.Most often, these are correlations for homogeneous reactions with an assumption that these reactions are homogeneous and heterophasic 138 and that the volume of the phase where the reaction takes place is constant.139, 140 In addition to the kinetic methods for examination of homogeneous reactions,141 kinetic models of topochemical reac- tions are widely applied for description of chemical transforma- tions of lignin, first of all, the Avraami ± Kolmogorov ± Erofeev equation.142 The validity of this equation is substantiated only formally, while no physical correlation between topochemical models and the processes of delignification has been found.It was shown 143 that a diffusion kinetics model is more appropriate to the description of the kinetics of softwood pulp delignification in organic solvents in which the rate-determining step is diffusion 953 of lignin and hemicellulose destruction products into the cooking solution. Since lignin is a polymer which is chemically and physically bound in pulp to polysaccharides the processes of the lignin chemical transformations should be considered as reactions in a polymeric matrix. The kinetics of chemical reactions in polymers is featured by the influence of various levels of the local matrix organisation and temperature-dependent matrix reorganisations on the conformity to the Arrhenius law, what is manifested in the so-called polychronous kinetics.144 ± 147 A model of polychronous kinetics has been used for the fist time by Bel'kova et al.148, 149 in the study of pulp delignification with nitric acid.Later, this model was applied to the study of softwood lignin destruction in alkaline media 150 ± 153 and to oxidative lignin destruction in pulp bleaching proc- esses.130, 154 ± 156 The application of the polychronous kinetics theory to the description of processes of chemical destruction and solubilisation of lignin (i.e., wood delignification) under the action of various reagents is based on three postulates. The first one is that reactions occur either in a solid phase or in an interface and are characterised by kinetic non-equivalence of chemically homogeneous species.The second one is that under isothermic conditions the processes do not proceed to completion, and at a certain depth of the conversion, the sharp deceleration of the reactions (the so-called kinetic stop) takes place. The third postulate says that the depth of the conversion depends on the temperature. Thus the major features of chemical processes that follow the polychronous kinetics are their kinetic stop and non-observance of the principle of the temperature ± time superposition. Although at higher temperature the reactions proceed deeper, they also stop at a certain point. According to the `classical' approach,144 f ÖlnkÜdlnk, f ÖlnkÜGÖk; tÜdlnk à n0 nÖt; TÜ à n0 Ö lnk Ö lnkmax lnk lnk min min where n(t, T) is the number of species which has entered into the reaction after time t at temperature T; f(ln k) is a distribution function; G(k, t) is a function describing the kinetics of the process in an elementary group with fixed values of the activation energy (E); k(E)=k0exp(7E/RT) is the reaction rate constant; kmin and kmax are the minimum and maximum rate constants, respectively; G for the simplest case is expressed as G1=exp(7k1t).The kinetic curves are straightened in the coordinates C/C07ln t (straight lines of the type y=ax+b). From these equations the values for kmin, kmax and the parameter of the kinetic non-equivalence [S=ln (kmax/kmin)] are found using the relation- ships 144 17y=a7b ln t, where a=17ln 1.76kmax/ln (kmax/kmin), b=(lnkmax7ln kmin)71.In general case, ln t=lnk. Recent publications 146, 147 consider the kinetic peculiarities of reactions in a polymeric matrix not as a violation of the classical Arrhenius law but, in contrast, as a particular case of a general model for polychronous kinetics (the so-called integral Arrhenius law). As a rule, in studies of the delignification processes (including those for oxidative destruction of the residual lignin), the necessity of imposing limitations in the framework of the formal kinetic analysis often arises. Such a lignocellulosic composite as unbleached sulfate pulp is so complex in structure and functional composition that the formulation of a direct kinetic model for the description of its oxidative destruction under the action of radicals and ions generated from peroxy reagents seems to be impossible. By comparison with other kinetic approaches, the model of polychronous kinetics allows one to characterise more completely the processes of oxidative destruction of the residual lignin by the effects observed in delignification of a lignocellulosic material (see Refs 154 ± 157 on polychronous kinetics for the oxidative deligni-954 ln K0eff 70 50 30 10 150 100 50 200 Eeff /kJ mol71 Figure 4.Kinetic compensative effect. eff fication of sulfate pulps with hydroperoxide anions and oxygen). The main kinetic peculiarities of oxidative lignin destruction (on bleaching of sulfate softwood pulps) with peroxy reagents include the existence of the kinetic stop and kinetic compensative effect (KCE, a linear relationship between preexponential factors K0 for separate ensembles of macromolecules and effective activation energy Eeff, Fig.4), the dependence of the highest degree of lignin conversion (from 0.07 to 0.55 157) on the parameter of kinetic non- equivalence S and temperature (Fig. 5) and the non-linear character of the relationship between ln t and 1/T in the temper- ature range from 293 to 373 K (Fig. 6). The values of the rate constants (k) for oxidative delignifica- tion in the temperature interval from 80 to 100 8C range approximately from 1017 exp(7143 kJ mol71/RT) s71 to 16 exp(719 kJ mol71/RT) s71. The larger values correspond to low-temperature processes with a degree of lignin conversion y=0.35, while the smaller values of the rate constant correspond to high-temperature processes with a degree of lignin conversion y=0.10.At a low degree of lignin conversion (y40.1) under the action of hydroperoxide anions, the rate of delignification is apparently determined by diffusion of the soluble products of lignin destruc- tion (their molecular weight can reach 15 000 ± 25 000) from the lignocellulosic matrix.148 Preliminary treatment of the residual lignin does not virtually affect the rate of the pulp delignification, since this treatment does not change the structure of cellulose fibres. With an increase in the degree of lignin conversion, the rate of the process sharply (exponentially) falls down, while the effective energy of activation rises.These effects are probably caused by slow molecular relaxation in a polymeric matrix so that the time for realisation of the most energetically favourable activated complex during an elementary reaction act (a kinetic constituent) is insufficient.144 On the one hand, the acid-catalysed and acid-electrophilic activations of the residual lignin lead to its destruction and better solubilisation (due to the smaller size) in alkaline media. On the other hand, they activate the lignin electrophilic sites. In the range of moderate temperatures (313 ± 333 K), the preactivation results S 80 400 340 360 T /K 300 320 Figure 5. Effect of temperature on the parameter of lignin kinetic non- equivalence (S) in the oxidative pulp delignification with hydrogen peroxide.V A Demin, V V Shereshovets, J B Monakov ln t 65 50 4 40 3 30 2 20 1 100 2.6 3.0 2.8 3.2 103 T71 /K71 Figure 6. Plot of ln t as a function of 1/T for various ensembles of lignin macromolecules in the sulfate pulp delignification in a stabilised solution of hydrogen peroxide at y=0.1 (1); 0.2 (2); 0.3 (3); 0.4 (4); 0.5 (5); 0.55 (6). in diminishing of the kinetic non-equivalence for lignin reactive sites (at 333 K the activated lignin has S=21 ± 23, while non- activated one has S*30). Despite the fact that the values of the minimum rate constant differ for different methods of lignin preactivation and that this constant bears an obvious physical meaning [7ln kmin=lnt (y=1.00), where ln t (y=1.00) is a loga- rithm of the time of complete lignin conversion], unambiguous interpretations based on analysis of this constant are hardly possible due to the very low values of this constant.The main manifestation of the lignin activation is an increase in the maximum degree of lignin conversion in its oxidative destruction with hydroperoxide anions and certain decrease in the parameter of kinetic non-equivalence associated probably with destruction- induced changes in the morphological structure of cellulose fibres. IX. Conclusion Lignin substantially differs in its behaviour in chemical reactions from low-molecular-weight model compounds, first of all, in a diversity of transformations and their depths.The reactivity of the residual lignin and the depth of its oxidative destruction with peroxy reagents depend to a considerable extent on the character of chemical pretreatments, which functionalise lignin and affect various structural changes in the polymeric lignocellulosic matrix. The depth of lignin destruction in the processes of pulp bleaching with peroxy reagents can be enhanced owing to alternation of the action of electrophilic and nucleophilic reagents under conditions of acid and base catalysis, respectively. However, the quantitative assessment of the lignin reactivity is rather difficult to perform because of the chemical and structural (topological and supramolecular) heterogeneity of industrial lignin, structural heterogeneity of the polymeric lignocellulosic matrix and some kinetic peculiarities of reactions in polymers.Aforementioned features of lignin, its polyfunctionality as well as diversity of oxidants, generated in reactions of oxygen, hydrogen peroxide and ozone, and the differences in mechanisms of the reactions of these reagents with lignin in acid and alkaline media provide a reason to consider the oxidative destruction of the residual lignin as a complex fundamental problem that demands further investigations. Despite some achievements in the studies of lignin reactions, an increase in the reactivity of the residual lignin remains a very important practical problem due to the permanently rising requirements for the environmental safety of technological processes in the pulp industry.Examination of the behaviour of lignin in chemical reactions and technological processes involving hydrogen peroxide, oxygen and ozone is not unambiguous, and further investigations and data systematisation are necessary in this field. Exploration of selective oxidative lignin destruction is aReactivity of lignin and problems of its oxidative destruction with peroxy reagents very important direction in wood chemistry, organic chemistry, chemistry of high-molecular-weight compounds and related fields of the sciences. References 1. B D Bogomolov Khimiya Drevesiny i Osnovy Khimii Vysokomole- kulyarnykh Soedinenii (The Chemistry of Wood and Foundations of the Chemistry of High-Molecular Compounds) (Moscow: Lesnaya Promst, 1973) 2.V Yu Belyaev, Candidate Thesis in Chemical Sciences, Institute of Chemistry, Komi Scientific Centre of Ural Branch, Russian Academy of Sciences, Syktyvkar, 1998 3. J B Monakov, V Yu Belyaev, A P Karmanov, T V Moskvicheva Dokl. Akad. Nauk 333 200 (1993) a 4. J S Gratzl Das Papier 41 120 (1987) 5. Ya A Gravitis, Doctoral Thesis in Chemical Sciences, Institute of Chemistry of Wood, Riga, 1989 6. A P Karmanov, Doctoral Thesis in Chemical Sciences, Institute of Chemistry, Komi Scientific Centre of Ural Branch, Russian Academy of Sciences, Syktyvkar, 1995 7. P P Erin'sh, I F Kul'kevitsa Khim. Drev. (6) 61 (1990) 8. J Gierer Holzforschung 36 55 (1982) 9. J Gierer Holzforschung 44 387 (1990) 10.J Gierer Holzforschung 44 395 (1990) 11. J P Simon, K-E L Eriksson Holzforschung 49 429 (1995) 12. K Sarkanen, K Lyudvig (Ed.) Ligniny (Struktura, Ssvoistva i Reaktsii) [Lignins (Structure, Properties and Reactions)] (Moscow: Lesnaya Promst, 1975) 13. V M Nikitin, A V Obolenskaya, V P Shchegolev Khimiya Drevesiny i Tsellyulozy (The Chemistry of Wood and Cellulose) (Moscow: Lesnaya Promst, 1978) 14. J F D Dean, K-E L Eriksson Holzforschung 46 135 (1992) 15. D Fengel, G Wegener Wood: Chemistry, Ultrastructure, Reactions (Berlin: Walter de Gruyter, 1983) 16. H H Nimz, A Berg, C Granzow, R Casten Zolz-Zentralblatt 95/96 1360 (1988) 17. D H Page J. Pulp Paper Sci. 9 15 (1983) 18. S M Shevchenko, Doctoral Thesis in Chemical Sciences, Wood Technology Academy, St.-Petersburg, 1992 19.K Poppius, Hortling Bo, I Sundquist, in International Symposium on Wood and Pulping Chemistry (Abstracts of Reports), Atlanta, 1989 p. 145 20. H Nimz, I Mogharab, H D Luedemann Makromol. Chem. 175 2563 (1974) 21. Y Tsutsumi, R Kondo, K Sakai, H Imamura Holzforschung 49 423 (1995) 22. E I Evstigneev, A V Kurzin, Yu A Platonov, E D Maiorova Zh. Prikl. Khim. 69 148 (1996) b 23. G Gellerstedt, E-L Lindfors, in Proceedings of the International Pulp Bleaching Conference, Stockholm, 1991 Vol. 2, p. 73 24. A P Karmanov, J B Monakov Vysokomol. Soedin., Ser. B 38 1631 (1996) c 25. A P Karmanov, J B Monakov Vysokomol. Soedin., Ser. B 37 328 (1995) c 26. P P Erin'sh Khim.Drev. (1) 8 (1977) 27. D C Smith, W G Glasser, H R Glasser, T C Ward Cellul. Chem. Technol. 22 171 (1988) 28. K Freudenberg, A S Neish Constitution and Biosynthesis of Lignin (Berlin: Springer, 1968) 29. M Matsukura, A Sakakibara J. Jpn. Wood Res. Soc. 1 35 (1969) 30. Ya A Gravitis, V G Ozol'-Kalnin Khim. Drev. (3) 24 (1977) 31. Y-Z Lai, K V Sarkanen Tappi J. 51 449 (1968) 32. Y-Z Lai, K V Sarkanen Cellul. Chem. Technol 9 239 (1975) 33. Y-Z Lai, X-P Guo,W Situ J. Wood Chem. Technol. 10 365 (1990) 34. T Yasuda, A Sakakibara Holzforschung 35 1831 (1981) 35. L V Kanitskaya, I P Deineko, D F Kushnarev, A V Klemper, G A Kalabin Khim. Drev. (6) 17 (1989) 36. G J Leary, J A Lloyd, K R Morgan Holzforschung 42 199 (1988) 37. E Adler Wood Sci.Technol. 11 169 (1977) 38. A Sakakibara Wood Sci. Technol. 14 89 (1980) 39. W G Glasser, H R Glasser Pap. Puu 63 71 (1981) 955 40. M Erickson, S Larsson, G E Miksche Acta Chem. Scand., Ser. C 27 903 (1973) 41. H Nimz Angew. Chem., Int. Ed. Engl. 13 313 (1974) 42. J Gierer Sven. Papperstidn. 73 571 (1970) 43. J Gierer, N-O Nilvebrant Holzforschung (Suppl.) 40 107 (1986) 44. L A Milovidova, G V Komarova, G A Ivanova, L A Smirnova Izv. Vyssh. Uchebn. Zaved., Lesn. Zh. (5 ± 6) 122 (1993) 45. V V Sokolov, D A Sukhov, A V Fedorov Tsellyul. Bum. Karton (9 ± 10) 22 (1996) 46. V Martin, K Burneson, H Lee, M Hewitt Holzforschung 49 453 (1995) 47. Yu A Malkov, N I Egorova, V A Zhalina, N M Tarannikov Tsellyul. Bum. Karton (11 ± 12) 10 (1997) 48.V A Demin, V D Davydov, B D Bogomolov Elektrokhimicheskaya Otbelka Sul'fatnoi Tsellyulozy (Electrochemical Bleaching of Sulfate Cellulose) (Leningrad: Nauka, 1982) 49. E B Althouse, J H Bortwick, D K Jain Tappi J. 70 113 (1987) 50. T A Tumanova Fiziko-Khimicheskie Osnovy Otbelki Tsellyulozy (Physicochemical Foundations of Whitening of Cellulose) (Moscow: Lesnaya Promst, 1984) p. 215 51. A Singh J. Pulp Paper Sci. 2 48 (1990) 52. T Elder Holzforschung 52 371 (1998) 53. J H Conkey Tappi J. 64 101 (1981) 54. J L Minor, E L Springer, in International Symposium on Wood and Pulping Chemistry (Abstracts of Reports), Atlanta, 1989 p.133 55. E N Medvedeva, V V Vershal', V A Babkin Khim. Interes. Ustoichiv. Razvitiya 4 343 (1996) d 56.T I Makovskaya, E N Medvedeva, A N Zakazov, F M Gizetdinov, V A Babkin Izv. Vyssh. Uchebn. Zaved., Lesn. Zh. (2 ± 3) 96 (1993) 57. D Lachenal, C De Choudens Cellul. Chem. Technol. 20 553 (1986) 58. J Gierer, N-O Nilvebrant Holzforschung (Suppl.) 48 51 (1994) 59. D Lachenal, C De Choudens, L Bourson Tappi J. 69 90 (1986) 60. J Colom, B Roncero, in The Fifth European Workshop on Lignocellulosics and Pulp. ELWP`98, University of Aveiro, Aveiro, 1998 p. 509 61. C Chirat, D Lachenal Holzforschung (Suppl.) 48 133 (1994) 62. C-A Lindholm Pap. Puu 72 254 (1990) 63. V A Solov'ev, G A Pazukhina PAP-FOR 93 (Pulp and Paper Forum), St.-Petersburg, 1993 p. 173 64. S A Medvedeva I V Volchatova, V A Babkin Khim. Interes. Ustoichiv. Razvitiya 4 321 (1996) d 65.V J J Martin, B K Burnison, H Lee, L M Hewitt Holzforschung 49 453 (1995) 66. T J Smith, R H Wearne, A F A Wallis Holzforschung 49 279 (1995) 67. T Norris, I Luke Pulp Paper J. 41 16 (1988) 68. R J Crawford,M N Stryker, S W Jett,W L Carpenter, R P Fisher, A K Jain Tappi J. 70 123 (1987) 69. V M Reznikov Khim. Drev. (2) 3 (1991) 70. E I Evstigneev, E I Chupka Khim. Drev. (5) 27 (1990) 71. J Gierer, E Yang, T Reitberger Holzforschung 48 405 (1994) 72. A D Sergeev, E I Chupka Khim. Drev. (1) 90 (1983) 73. A D Sergeev, E I Chupka Khim. Drev. (2) 73 (1985) 74. A Ya Sychev, V G Isak Usp. Khim. 64 1183 (1995) [Russ. Chem. Rev. 64 1105 (1995)] 75. D Lachenal, L Bourson Tappi J. 63 119 (1980) 76. C W Dence, S Omori Tappi J. 69 120 (1986) 77.E I Chupka, S I Zhirnova, A D Sergeev Khim. Drev. (1) 83 (1988) 78. E I Chupka, M E Kondakova, I M Luzhanskaya, S V Egorova, S M Trofimova, V V Fomina, N A Ivanov Khim. Drev. (2) 29 (1981) 79. A I Nonni, C W Dence Holzforschung 42 37 (1988) 80. G Gellerstedt, B Pattersson Sven. Papperstidn. 83 314 (1980) 81. G Gellerstedt, E-L Lindfors Tappi J. 70 120 (1987) 82. D D Jones, D C Johnson J. Org. Chem. 32 1402 (1967) 83. A Enoki,M N Gold Arch. Microbiol. 132 123 (1982) 84. R T Morrison, R N Boyd Organic Chemistry (Boston: Allyn and Bacon, 1971) 85. J Gierer, F Imsgard, in International Symposium on Chemistry of Delignification with Oxygen, Ozone and Peroxides (Abstracts of Reports), Tokyo, 1980 p. 133 86. M A Zil'bergleit, T V Korneichuk, V M Reznikov Khim.Drev. (6) 21 (1987) 87. M A Zil'bergleit, B S Simkhovich, V M Reznikov Khim. Drev. (6) 28 (1987)956 88. M A Zil'bergleit, T V Korneichuk, V M Reznikov Khim. Drev. (2) 56 (1988) 89. S D Razumovskii Kislorod�Elementarnye Formy i Svoistva (Oxygen�Elementary Forms and Properties) (Moscow: Khimiya, 1979) 90. S D Razumovskii, G E Zaikov Usp. Khim. 49 2344 (1980) [Russ. Chem. Rev. 49 1163 (1980)] 91. H S Kaneko, S Hosoya K Iiyama, J Nakano J. Wood Chem. Techol. 3 399 (1983) 92. Y Matsumoto, A Ishizu, J Nakano Holzforschung (Suppl.) 40 81 (1986) 93. S C Puri, S M Anand Cellul. Chem. Technol. 20 535 (1986) 94. H Taneda, N Habu, J Nakano Holzforschung 43 187 (1989) 95. D Bjorn, P Walter Pap. Puu 74 720 (1992) 96.P W-C Ku, J S Hsieh, D M Jaywant, L L Houle Tappi J. 75 146 (1992) 97. V A Demin, N N Kabal'nova, G Ya Osipova, V V Shereshovets Zh. Prikl. Khim. 66 2562 (1993) b 98. G G Galiaskarova, Candidate Thesis in Chemical Sciences, Institute of Organic Chemistry, Ural Scientific Centre, Russian Academy of Sciences, Ufa, 1997 99. C Brage, T Eriksson, J Gierer Holzforschung 45 23 (1991) 100. C Brage, T Eriksson, J Gierer Holzforschung 45 147 (1991) 101. C Brage, T Eriksson, J Gierer Holzforschung 49 127 (1995) 102. J Gierer J. Wood Chem. Technol. 12. 367 (1992) 103. V A Demin, A G Dontsov, E V German, T P Shcherbakova, E A Fel'de, in Lesokhimiya i Organicheskii Sintez (Wood Chemistry and Organic Synthesis) (Syktyvkar: Komi Scientific Centre, Ural Branch, Russian Academy of Sciences, 1996) p.71 104. N Liebergott, B Van Lierop Pulp Paper Can. 87 58 (1986) 105. S M Shevchenko, A G Apushkinskii Usp. Khim. 61 195 (1992) [Russ. Chem. Rev. 61 105 (1992)] 106. V A Demin, E I Fedorova, L A Nikulina, E V German, N F Pestova, T P Shcherbakova Otbelka Sul'fatnoi Tsellyulozy bez Molekulyarnogo Khlora (Bleaching of Sulfate Cellulose without Molecular Chlorine) (Syktyvkar: Komi Scientific Centre, Ural Branch, Russian Academy of Sciences, 1995) 107. H Sixta, L Lotzingen, A Schrittwieser, P Hendel Papier 45 610 (1991) 108. D Rasmusson, O Samuelsen Holzforschung (Suppl.) 40 147 (1986) 109. L Andersson, O Samuelson, P Mansson Sven. Papperstidn. 87 59 (1984) 110. L Andersson, O Samuelson Sven.Papperstidn. 88 102 (1984) 111. D Lachenal, L Bourson,M Muguet, A Chauvet Cellul. Chem. Technol. 24 593 (1990) 112. U Carlson, O Samuelson Holzforschung 39 41 (1985) 113. N N Shorygina, V M Reznikov, V V Elkin Reaktsionnaya Sposobnost' Lignina (Reactivity of Lignin) (Moscow: Nauka, 1976) 114. V A Demin . Drev. (3) 29 (1994) 115. G Fossum, A Marklund Tappi J. 71 79 (1988) 116. A F Gogotov, Yu V Panasenkov, I L Panchukov, E V Shilkin, V A Babkin Khim. Interes. Ustoichiv. Razvitiya 4 245 (1996) d 117. A F Gogotov, Doctoral Thesis in Chemical Sciences, Siberian State Technological University, Krasnoyarsk, 1998 118. P Larsson, O Samuelson, U Ojteg Holzforschung 43 121 (1989) 119. R Larsson, O Samuelson J. Wood Chem. Technol. 10 311 (1990) 120. D Rasmusson, O Samuelson Cellul. Chem. Technol. 23 285 (1988) 121. O Samuelson, U Ojteg Tappi J. 71 175 (1988) 122. Yu A Bobrov, M I Cherezova, V A Demin, in Khimiya i Tekhno- logiya Proizvodstva Tsellyulozy (The Chemistry and Technology of Production of Cellulose) (Leningrad: Leningrad Technological Academy, 1987) p. 48 123. V A Demin, A P Karmanov, Yu A Bobrov Khimicheskaya Pere- rabotka Drevesiny i Drevesnykh Otkhodov (Chemical Processing of Wood and Wood Waste) (Leningrad: Leningrad Technological Academy, 1988) p. 91 124. Russ. P. 1 193 192; Byull. Izobret. (43) 112 (1985) 125. Russ. P. 2 061 134; Byull. Izobret. (15) 244 (1996) 126. Russ. P. 2 019 612; Byull. Izobret. (17) 98 (1994) 127. A V Obolenskaya, V P Shchegolev, G L Akim, E L Akim, N L Kossovich, I Z Emel'yanova Prakticheskie Raboty po Khimii Drevesiny i Tsellyulozy (Practical Works on Chemistry of Wood and Cellulose) (Moscow: Lesnaya Promst, 1965) 128. Russ. P. 2 019 613; Byull. Izobret. (17) 98 (1994) V A Demin, V V Shereshovets, J B Monakov 129. Russ. P. 2 075 566; Byull. Izobret. (8) 190 (1997) 130. V A Demin, E V German Khim. Drev. (3) 38 (1994) 131. V A Demin, Doctoral Thesis in Chemical Sciences, Institute of Organic Chemistry, Ural Scientific Centre, Russian Academy of Sciences, Ufa, 1997 132. L A Milovidova, G V Komarova Tsellyul. Bum. Karton (11 ± 12) 14 (1997) 133. A D Sergeev, V A Petrenev, O P At'man Tsellyul. Bum. Karton (5 ± 6) 5 (1998) 134. R A D Mbachu, R St J Manley J. Polym. Sci., Polym. Chem. Ed. 19 2053 (1981) 135. H-J Mielisch, J Odermatt, O Kordachia, R Patt Holzforschung 49 445 (1995) 136. A G Dontsov, V A Demin, in Lesokhimiya i Organicheskii Sintez (Wood Chemistry and Organic Synthesis) (Syktyvkar: Komi Sci- entific Centre, Ural Branch, Russian Academy of Sciences, 1996) p. 75 137. Y-Z Lai, S-P Mun Holzforschung 48 203 (1994) 138. M A Zil'bergleit, V M Reznikov Khim. Drev. (6) 62 (1988) 139. M A Kushner, L G Matusevich, I V Sen'ko, L V Chirich Khim. Drev. (3) 50 (1990) 140. E M Ben'ko, V V Kovaleva, A N Mitrofanova, V A Voblikova, A N Pryakhin, V V Lunin Zh. Fiz. Khim. 68 1580 (1994) e 141. E T Denisov Kinetika Gomogennykh Khimicheskikh Reaktsii (Kinetics of Homogeneous Chemical Reactions) (Moscow: Vysshaya Shkola, 1988) 142. P Barret Kine'tique He'te'roge'ne' (Paris: Jauthier-Villars, 1973) 143. G B Tibermane, L T Purina, M Ya Ioelovich, A P Treimanis Khim. Drev. (4 ± 5) 36 (1992) 144. N M Emanuel', A L Buchachenko Khimicheskaya Fizika Molekulyarnogo Razrusheniya i Stabilizatsii Polimerov (Chemical Physics of Molecular Destruction and Stabilisation of Polymers) (Moscow: Nauka, 1988) 145. O N Karpukhin Usp. Khim. 47 1119 (1978) [Russ. Chem. Rev. 47 587 (1978)] 146. V A Kutyrkin Khim. Fiz. 14 (5) 71 (1995) f 147. V A Kutyrkin Khim. Fiz. 14 (5) 90 (1995) f 148. L P Bel'kova, V S Gromov, A I Mikhailov Khim. Drev. (6) 50 (1980) 149. L P Bel'kova, V S Gromov, A I Mikhailov Khim. Drev. (6) 59 (1980) 150. V V Vershal', E I Chupka, O V Ushakovskii, A I Mikhailov Khim. Drev. (6) 47 (1988) 151. R Z Pen, I L Shapiro, V R Pen Khim. Drev. (3) 33 (1990) 152. R Z Pen, I L Shapiro, V R Pen Khim. Drev. (5) 106 (1990) 153. R Z Pen, V R Pen, I L Shapiro, M Yu Katrukhina Khim. Drev. (4 ± 5) 31 (1992) 154. A P Karmanov, V A Demin Khim. Interes. Ustoichiv. Razvitiya 4 289 (1996) d 155. A I Mikhailov, V A Demin, L D Kaplun, G A Kolpakov, I A Shilova, in The Fifth European Workshop on Lignocellulosics and Pulp. ELWP`98, University of Aveiro, Aveiro, 1998 p. 79 156. V A Demin, A P Karmanov, L S Kocheva, in The Fifth European Workshop on Lignocellulosics and Pulp. ELWP`98, University of Aveiro, Aveiro, 1998 p. 141 157. V A Demin, E V German-Udoratina, T P Sherbakova, in The Fifth European Workshop on Lignocellulosics and Pulp. ELWP`98, University of Aveiro, Aveiro, 1998 p. 137 a�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) b�Russ. J. Appl. Chem. (Engl. Transl.) c�Polym. Sci. (Engl. Transl.) d�Chem. Sustain. Devel. (Engl. Transl.) e�Russ. J. Phys. Chem. (Engl. Transl.) f�Russ. J. Chem. Phys. (En

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (11) 957 ± 966 (1999) Alginic acids and alginates: analytical methods used for their estimation and characterisation of composition and primary structure A I Usov Contents I. Introduction II. Chemical methods III. Enzymic and immunological methods IV. Nondestructive physicochemical methods V. Conclusion Abstract. The history and state-of-the-art in the detection, quantitative determination and characterisation of the primary structure of alginic acids and their salts (alginates) are reviewed. A brief survey of the structure and properties of these polysacchar- ides is given. Numerous analytical methods including chemical, physicochemical and enzymic procedures for the structural analysis of alginates which can also be used for the investigation of other uronic acid-containing polysaccharides are discussed.The bibliography includes 211 references. I. Introduction Alginic acids were first isolated from several brown algae by Stanford more than 100 years ago.1 Later, these polysaccharides were detected in all brown algal species without exception as cell wall and intercellular matrix components. Their content in the biomass may amount to 40% of the dry weight; this strongly depends on the species and growth conditions.2± 4 Some bacteria (mostly, the representatives of the genera Azotobacter 5 and Pseudomonas 6) also produce alginic acids. In 1982, alginates were unexpectedly detected in one of the species of calcareous red algae belonging to the family Corallinaceae.7 Our recent studies 8, 9 have shown that the presence of alginates is virtually a characteristic feature of the representatives of this family. In other plants, these polysaccharides have not been found.Alginic acids are polyuronides, i.e., polysaccharides the molecules of which are built up of uronic acid residues. At first, D-mannuronic (1) 10, 11 and then L-guluronic (2) 12 acids were identified as the components of alginic acids. COOHO O OH OH COOHHO HO OH OH HO HO 1 2 Partial hydrolysis and isolation of the corresponding disaccharide revealed that the residues of both uronic acids are the components A I Usov N D Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 117913 Moscow, Leninsky prosp.47, Russian Federation. Fax (7-095) 135 53 28. Tel. (7-095) 137 67 91. E-mail: usov@ioc.ac.ru Received 18 May 1999 Uspekhi Khimii 68 (11) 1051 ± 1061 (1999); translated by R L Birnova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.458.7 957 958 961 962 963 of the same polymeric molecule.13 According to present-day concepts based on numerous studies including methylation,14, 15 periodate oxidation 16, 17 and several physicochemical methods of structural analysis (mostly, NMR spectroscopy 18 ± 21), molecules of alginic acids are linear and made up of b-D-mannuronic and a-L-guluronic acid residues (in the pyranose form) linked by 1?4 bonds. Bacterial alginic acids differ from the algal ones in that some of their hydroxy groups are usually acetylated.22, 23 Acetyl groups occupy positions 2 or 3 (sometimes, both simultaneously) in D-mannuronic acid residues.24 The degree of polymerisation of native polymers is rather high and the number of monosaccharide residues may amount to 1000 ± 10 000.Alginic acids isolated from various sources can differ in the ratios of mannuronic and guluronic acids (M/G) 25 and in the distribution of the monomers along the polymeric chain. Fractio- nation of oligomeric products formed upon partial hydrolysis has shown that some regions in the molecules of alginic acids are built of only one uronic acid species (M-blocks and G-blocks). These blocks are separated by fragments which contain approximately equal amounts of both monosaccharides arranged into more or less strictly alternating sequences (MG-blocks).26, 27 Such a peculiar structure is a result of stepwise biosynthesis of polysac- charide molecules where the formation of linear mannuronan precursors is followed by epimerisation of some b-D-mannuronic acid residues at the C(5) atom under the action of a specific enzyme, mannuronan-C(5)-epimerase, resulting in the appear- ance of a-L-guluronic acid residues within the polymer struc- ture.28 The size and mutual distribution of individual blocks strongly affect the properties of alginic acids and their salts (alginates).As a matter of fact, b-D-mannuronic and a-L-guluronic acid residues, being structurally very close [they differ only in the configuration at one of the asymmetric centres, viz., at the C(5) atom], have opposite conformations of the pyranose rings (4C1 and 1C4, respectively), which results in completely different shapes of the polymeric molecules in the poly-M and poly-G regions (Fig.1). This creates spatial conditions for the tight binding of bivalent metal cations with G-blocks. Moreover, the coordination with such cations favours the cooperative binding of different mole- cules of the polymer with one another,29 eventually resulting in the formation of ionotropic gels (Fig. 2). Therefore, the total content of a-L-guluronic acids (in the first approximation) or, more precisely, the relative length of G-blocks are important criteria of the most valuable property of alginates, viz., their ability to form gels.30 ± 32 The block structure is responsible for many other characteristics of alginates including their biological activity.33 ± 35958 HOOC HOOC OHO HO O O OHO OH O O O HOOC OH M M OH G Figure 1.The conformations of monosaccharide units in the molecule of alginic acid. By virtue of their unique physicochemical properties, alginic acids and their salts have found wide application in the production of foods and in some other branches of industry 36, 37 including biotechnology and medicine.38 This explains the fact that the majority of reviews devoted to alginic acids provide information about their structure, biosynthetic mechanisms and practical application, but pay insufficient attention to problems concerned with the analytical chemistry of alginates.Thus a special issue of the journal Carbohydrates in Europe, which is all devoted to alginates and consists of four reviews describing the chemistry and physical properties,39 genetics and biosynthesis of alginates,40 as well as biotechnological and clinical applications of alginate gels 41 and commercial applications of alginates,42 is laid out on this principle. Because of considerable practical utility of alginates, their isolation from brown sea weeds has become the main goal of their processing. It is obvious that estimation of alginic acid content in the algal biomass and standardisation of alginic acid preparations and alginate-containing products demand elaboration of conven- ient analytical procedures, while structural characterisation of these polysaccharides is necessary for determining the most rational approaches to the utilisation of particular preparations containing alginic acids.Proper allowance must be made for the fact that the analytical chemistry of alginates has specific features as compared with that of other polysaccharides. Alginic acids form water-soluble salts with monovalent cations but are precipitated upon acidification. Alginates of many bivalent cations, particularly, of Ca2+, Sr2+ and Ba2+, are insoluble in water. This property is utilised in the isolation of alginic acids from algae.43 At first, bivalent cations and concom- itant water-soluble polysaccharides are extracted from the bio- aCa2+ O OH 7O OH O O O O OH OO 7OH O b Ca Ca Ca Ca Ca Ca Ca Ca Ca Figure 2.The coordination of calcium ions by a-L-guluronic acid residues in the molecule of an alginate (a) and the role of these ions in intermolecular interactions during gel formation (b) (the `egg-box' model).29 OH O OH COOH G A I Usov mass with a dilute mineral acid; then alginic acids are extracted with an alkaline solution with subsequent precipitation from the extract by treatment with ethanol,44 acidification or conversion into Ca2+ salt.45 Preparative yields of alginic acids may be used for determination of their content in the algal biomass; it should be remembered, however, that the results obtained with the use of different isolation procedures can give substantially different results for the same algal specimen.This depends on a number of factors, such as reagents used, the number of precipitations and the degree of achieved purity. Thus special experiments aimed at comparing the efficiency of three most popular laboratory procedures 46 ± 48 revealed that the relative scatter in the values of preparative yields of alginates may amount to 20%± 40%.49 Additional problems may arise when this approach is used for the analysis of food compositions where the behaviour of pectin components is similar to that of alginates as regards their extraction and precipitation. Some procedures for the estimation of alginic acids are based on specific reactions of uronic acids.These reactions include decarboxylation and degradation under the action of concen- trated sulfuric acid with subsequent formation of coloured products in the presence of carbazole, indole, 3-hydroxybi- phenyl, etc. These methods have found wide application, although it is evident that the presence of other polysaccharides containing uronic acids in the specimen under study will inevitably give erroneous results of alginate assays. The use of acid hydrolysis of alginates for the preparation of monomeric uronic acids is a rather complex problem. Glycosidic bonds of uronic acid residues are highly resistant to acid hydrolysis. Low solubility of alginic acids in acid media creates an additional problem. Total acid hydrolysis of alginic acids is accompanied by the destruction of some monomers. It is of note that L-guluronic acid residues are destroyed faster than D-mannuronic acid residues and specific losses of each monomer depend on the pattern of their block distribution within the polymeric molecule.Therefore, it is very difficult to select versatile conditions for hydrolysis and to find correction factors for the destruction of monomers that would be equally suitable for the determination of the composition of alginates differing considerably in the ratios and the distribution of uronic acid residues in their molecules. Nondestructive physicochemical methods, in the first place, 1H and 13C NMR spectroscopy, have assumed great importance in structural characterisation of alginic acids.These methods can be used both for the determination of the monomer ratio in alginic acid molecules and their characterisation in terms of their arrangement into individual blocks. In the subsequent sections, the above-mentioned analytical approaches will be considered in more detail and illustrated, mostly by the studies of sea weeds, although these methods are equally suitable for the character- isation of bacterial alginates. Additional information can be retrieved from several reviews devoted mostly to alginates of bacterial origin.50 ± 55 II. Chemical methods 1. Assays based on cation exchange If a starting material is used in sufficiently large amounts (i.e., the alginic acid content in a sample under study is no less than 100 mg), alginic acids can be determined by the gravimetric method.Polymeric alginic acids are precipitated upon acidifica- tion of weakly alkaline solutions, while calcium alginate is precipitated from neutral sodium alginate solutions upon addi- tion of calcium chloride. Drying of the precipitates is problematic, since these compounds tightly retain water and are readily decomposed on heating. Volumetric analysis is more preferable in experiments with smaller amounts of the material. In this case, the polysaccharide is also precipitated as calcium salt, this is washed with a dilute acid to remove metal cations, and the insoluble alginic acid is washed with water and titrated withAlginic acids and alginates: analytical methods used for their estimation and characterisation of composition and primary structure sodium hydroxide in the presence of phenolphthalein or using conductometric 56 or potentiometric 57 control.The formation of viscous sodium alginate solutions in the course of titration can be avoided if an excess of calcium acetate is added to alginic acid, and the liberated acetic acid is titrated.58 A procedure is described which combines titration with colorimetry and is designed for the determination of the content of alginic acid in algae.59 An alternative, rather lengthy procedure for quantitative analysis of alginates is based on the ability of alginic acid- containing algal biomass to bind calcium ions.60 To this end, samples of a ground plant material are dialysed against dilute hydrochloric acid to remove metal cations; the mineral acid is removed by dialysis against water.Then, alginic acid is converted into calcium salt by dialysis against a calcium acetate solution; the excess of the reagent is removed by repeated dialysis against water. Upon this treatment, all the residues of uronic acids in the alginate are converted into calcium salts where one cation corresponds to two uronic acid residues. Bound calcium ions are then quantita- tively extracted by dialysis against dilute hydrochloric acid. The calcium concentration in the diffusate is determined by atomic adsorption spectrophotometry, and the alginate content in the original biomass is calculated.The main drawback of this method is that the presence of other polyanions (e.g., sulfated polysac- charides) in the biomass can overestimate the experimental results. As polyanions, alginic acid molecules form salts with coloured inorganic (bivalent copper cations 61 or ruthenium red 62) and organic cations (cationic dyes of the Methylene Blue type 63 ± 65). In this case, quantitative assays are based on the measurement of the decrease in the optical density of the original coloured solution upon formation of poorly soluble salts or on the changes in the spectral characteristics of the dye (metachromasia) due to its binding to the polymer. This method, which is characterised by high sensitivity and precision, has been recently proposed for quantitative determination of alginates in microcapsules used in cell transplantation.66 The experimental procedure is based on the analysis of metachromatic changes in the spectrum of 1,9- dimethyl-Methylene Blue as a result of its interaction with the alginate. This type of reactions also includes precipitation of polyanions (as salts) with cationic detergents (Cetavlon precip- itation).67 Polyhexamethylenebiguanidinium chloride used for the precipitation of alginic acids is the most studied reagent of this class.68 Precipitation of alginic acids with this reagent is virtually independent of the presence of salts (< 2.5%), pH (5.5 ± 7.6), temperature (10 ± 60 8C) and the M/G ratio.The analytical procedure is based on the measurement of the decrease in the reagent absorption at 235 nm.Although this method was recom- mended for the analysis of industrial extracts of brown algae,69, 70 other acid polysaccharides can be precipitated by cationic deter- gents together with alginic acids. It was proposed 71 to use coloured hydrazides formed by the reaction of 2-nitrophenylhy- drazine in the presence of water-soluble carbodiimide in colori- metric assays of alginates and their fragments instead of coloured salts;71 however, this method has not found wide application. 2. Assays based on degradation of uronic acids Heating of alginates results in decarboxylation and evolution of CO2.72 The samples under study are usually heated with 19% HCl, and the CO2 evolved is absorbed and weighed as in the micro- analytical determination of carbon in organic compounds by combustion.73, 74 Although this method was described in a popular practical guide to biochemical assays of algae,75 it requires special skill and equipment (cf.Refs 76 ± 78). This is also true for a modified decarboxylation procedure in which polysac- charide samples are heated with 57% HI and the CO2 evolved is assayed by gas-liquid chromatography (GLC)79 or direct non- aqueous titration.80 It is obvious that other uronic acid-containing polysaccharides can interfere with the alginate assays. Colorimetric assays of polyuronides based on acid degrada- tion of their constituent monosaccharides have found much wider 959 application. The main reaction product formed upon heating of uronic acid derivatives with concentrated sulfuric acid is 5-formylfuran-2-carboxylic acid, which is rather stable under these conditions.81 The formation of degradation products of neutral sugars (5-hydroxymethylfurfural and furfural) from hexoses and pentoses, respectively, prevents its direct spectropho- tometric analysis.It is therefore more expedient to use the ability of 5-formylfuran-2-carboxylic acid to form coloured products with various organic reagents, particularly with some aromatic nitrogen heterocycles and phenols, e.g., carbazole,82, 83 indole,84 3-hydroxybiphenyl 85 and 3,5-dimethylphenol.86 Along with indisputable advantages (in the first place, the possibility to experiment with microquantities of material), colorimetric assays have obvious drawbacks, such as interference due to neutral sugars and different reactivities of individual uronic acids.This is the reason why the analytical procedure was repeatedly modified, although the reaction with carbazole had long been used for the determination of polyuronides including alginic acids.87 Thus it was proposed to supplement the reaction mixture with boric acid in order to intensify the colour formation and to equalise the reactivities of different uronic acids.88, 89 The addition of sulfamate 90, 91 or different temperature regimes 92 were used in order to minimise the effects of neutral sugars. It was demonstrated that the composition of a specimen containing three uronic acids (glucuronic, galacturonic and mannuronic) can be calculated by conducting the reaction with carbazole in four different regimes.93 The effect of ballast neutral sugars was estimated during the elaboration of a procedure for pectin assays.94, 95 It was found that 3,5-dimethylphenol 86 and 3-hydroxybiphenyl 96 have salient advantages over carbazole in pectin and hemicellulose assays.In a more recent study,97 Filisetti-Cozzi and Carpita proposed a method for the colorimetric assay of polyuronides including alginic acids, which took account of recommendations from previous investigations. This method utilised 3-hydroxybiphenyl in combination with the borate anion and sulfamate. Our recent studies have shown that 3,5-dimethylphenol is a more preferable specific reagent for alginic acids.8 We modified the procedure proposed by Scott 86 for the analysis of pectins and hemicelluloses to fit with the specific composition of alginates using greater amounts of borate and more prolonged heating of reaction mixtures at 70 8C.The results of these experiments weakly depended on changes in the ratios of D-mannuronic and L-guluronic acids in different alginate specimens. The problem of selectivity of the colour reaction and the necessity to eliminate interference due to the presence of neutral sugars in the specimen under study is of prime importance in the analysis of heteropolysaccharides, such as plant hemicelluloses and complex pectins, animal mucopolysaccharides and many bacterial and fungal polysaccharides.In the case of alginic acids, the effect of neutral sugars can be eliminated by preliminary removal of ballast substances,98 although this procedure compli- cates the experimental protocol. For example, ion-exchange chromatography on Dowex 162 was used for purification of an alginate isolated from the extracellular material of mucoid strains of Pseudomonas aeruginosa,99 although DEAE-cellulose or DEAE-Sephadex are evidently more suitable.100, 101 In alginate assays in food products, it is recommended to conduct enzymic deproteination with subsequent precipitation of calcium alginate. The latter is carefully washed off from soluble carbohydrates and dissolved in cold 80% sulfuric acid or in hot sodium hexameta- phosphate solution; alginates are assayed in the resulting solutions using naphthoresorcinol,102 phenol 103, 104 or trivalent iron sul- fate 105 in concentrated sulfuric acid.The selectivity of these reagents is lower than that of 3-hydroxybiphenyl, 3,5-dimethyl- phenol and carbazole. The reaction with Fe2(SO4)3 was success- fully used for the study of seasonal changes and anatomic distribution of alginates in some species of brown algae.106 In a more complicated procedure described by Toyoda et al.,107 naphthoresorcinol-HCl was added to an alginate after its960 precipitation as Cu2+ salt; the coloured reaction product formed thereupon was extracted with butyl acetate and quantitatively assayed using HPLC. In the majority of the above-described reactions, concentrated sulfuric acid was used to destroy uronic acids; however, similar results can be obtained in polyuronide assays utilising the Bial reagent (orcinol in concentrated hydrochloric acid, usually supplemented with FeCl3 108). In the case of alginates, the reaction with orcinol has certain advantages over the reaction with carbazole,109 although analytical results also depend on the composition of the original polymer.110 Obviously, this depend- ence is the main limitation of the majority of alginic acid assays, since chemical properties of uronic acid residues strongly depend on whether it forms a part of a homo- or a heteropolymeric block.This circumstance is the main obstacle on the way to the design of versatile calibration curves for the determination of alginic acids with any ratio and any distribution of b-D-mannuronic and a-L- guluronic acids along the polymeric chain.3. Assays based on acid hydrolysis of glycosidic bonds As mentioned above, alginic acids, like other polyuronides, are hydrolysed with difficulty and under drastic conditions. This leads to partial degradation of the monosaccharides liberated, in the first place, of guluronic acid. In the hydrolysis products, the monomeric uronic acids are in an equilibrium with lactones. This all makes the analysis of alginic acid hydrolysates a rather complicated procedure. The method for determining the M/G ratio by polarimetry of the total hydrolysate of alginates in the form of brucine salts 111 has not found wide application, since it requires large quantities of the starting material and its high purity, for ballast carbohydrates significantly distort the results of polarimetric analysis.Therefore, in the overwhelming majority of cases the hydrolysates are analysed following preliminary separation of the hydrolysis products. Some time ago, it was proposed to use paper chromatography for separation of the hydrolysis products; this procedure allowed for the first time the proof of the presence of guluronic acid residues in the polymers.12 Later, conditions were elaborated for the separation of guluronic, mannuronic, galacturonic and glucuronic acids by paper electrophoresis.112 However, these methods are suitable for qualitative rather than for quantitative analysis of hydrolysates.In more recent studies, thin-layer chromatography (TLC) on silica gel was used for the identifica- tion of depolymerisation products of alginic acids. This method is especially convenient for detecting small amounts of alginates in beer 113 ± 115 and other foodstuff; acid methanolysis is more preferable than hydrolysis. Thus TLC of the methanolysis products allows unambiguous identification of alginates even in the presence of other uronic acid-containing food polysacchar- ides.116, 117 High-performance TLC was used to study alginate biosynthesis in Pseudomonas aeruginosa.118 A procedure based on ion-exchange chromatography has been developed for quantitative determination of uronic acids in hydrolysates and, correspondingly, for the calculation of the M/G ratio.110, 119 The conditions of hydrolysis proposed by Haug and Larsen 110, 119 were used (almost without modification) in a great number of more recent studies.After treatment of alginate with 80% sulfuric acid at 0 8C, the mixture was incubated for 18 h at 20 8C, then water was added up to the acid concentration of 1 mol litre71 and the resulting mixture was heated for 5 h at 100 8C. After cooling, the hydrolysate was neutralised with calcium carbonate, alkalified to pH 8.0, incubated for 30 min for lactone opening and chromatographed on a Dowex-1 column (acetate form) in a linear gradient of acetic acid (from 0.5 to 2 mol litre71). Under these conditions, guluronic acid was eluted from the column before mannuronic acid.The uronic acid content in the fractions was determined using the reaction with orcinol; the experimental value of M/G was multiplied by 0.66 in order to correct for guluronic acid losses due to its predominant degrada- A I Usov tion during hydrolysis. This method was used to characterise the alginates isolated from a great number of algal species.25 Some more recent modifications were aimed at the improve- ment of hydrolysis conditions 120 and refinement of the correction coefficients.121 In particular, it was proposed to use 88% formic acid instead of sulfuric acid for diminishing the degradation of monomeric uronic acids.22 However, the main progress in this field is connected with elaboration of new technologies for separation and identification of the hydrolysis products.Thus high-performance liquid chromatography was widely used for ion-exchange separation of uronic acids;122 ± 128 quantification of these compounds in the eluate can efficiently be performed by measuring intrinsic absorption at 210 nm,127 by using refracto- metric 124 or pulse amperometric detectors 125 or in a reaction with cyanoacetamide 126 with subsequent spectrophotometry or spectrofluorimetry.129 The new separation technique for uronic acids markedly accelerated and simplified the experimental procedure, but did not eliminate the distortions introduced by hydrolysis of alginates. A method for the analysis of real specimens of alginic acids isolated from different brown algal species 127 which utilises ion- exchange chromatography in an isocratic regime and UV-detec- tion for monitoring elution, has gained especially wide popularity.Yet another popular procedure for the quantification of hydrolysis of products of alginic acids is GLC, which requires additional chemical conversion of uronic acids into their volatile derivatives. Hydrolysis products normally represent a mixture of uronic acids with lactones. Substitution of methanolysis for hydrolysis yields a multicomponent mixture of methyl esters of methyl glycosides. Analysis of such mixtures (e.g., in the form of trimethylsilyl derivatives) is a rather complex problem,130 ± 132 and the results obtained are of limited value for quantitative assays.In order to simplify the composition of the mixtures to be separated, it was proposed to convert uronic acids into aldonic acid lactones (by reduction of the aldehyde groups with subse- quent lactonisation 133, 134) or alditols (by more profound reduc- tion) with subsequent chromatographic separation as acetates.134 ± 136 In the latter case, several approaches are used to distinguish uronic acid derivatives from the reduction products of neutral monosaccharides. For example, uronic acids are separated by treatment of hydrolysates with ion-exchange resins prior to reduction,136 otherwise sodium borodeuteride is used for reduc- tion followed by chromato-mass spectrometry for the identifica- tion of reaction products.137 It was also recommended to separate hydrolysates into two portions, in one of which neutral sugars are converted into aldononitrile acetates (the procedure employed results in degradation of uronic acids), and the other portion is used for complete reduction.Comparison of analytical data allows one to calculate the amounts of both neutral monosac- charides and uronic acids.138 Neutral sugars and uronic acids can also be assayed simultaneously by converting the former into alditol acetates and the latter into N-propylaldonamide acetates via aldonolactones.139 Attempts have been made to circumvent difficulties in alginic acid hydrolysis by reducing the carboxy groups in the polymeric molecule. This can be done either by preliminary esterification of the carboxy groups 140, 141 or by their activation with water- soluble carbodiimide with subsequent treatment with sodium borohydride at controlled pH.142 The reduction procedure for the determination of the composition of alginic acids was proposed in 1981;143 however, further studies of the preparative reduction of alginic acids with water-soluble carbodiimide revealed that this reaction cannot be brought to comple- tion.144, 145 Thus treatment with an excess of carbodiimide caused a significant destruction of the original polysaccharide 144 con- comitantly with the preferential reduction of the L-guluronic acid carboxy groups;145 in the deficiency of this reagent, preferential reduction of the D-mannuronic acid carboxy groups took place.144Alginic acids and alginates: analytical methods used for their estimation and characterisation of composition and primary structure A more recent study was specially devoted to comparison of different methods of reduction of the carboxy groups in polymers containing uronic acids.146 The method based on activation of the carboxy groups with 1-cyclohexyl-3-(2-morpholinoethyl)carbo- diimide metho-p-toluenesulfonate in 8 M urea 147 was found to be the most efficient procedure; however, in this case subsequent treatment with sodium borohydride caused only partial reduction of the carboxy groups and the results obtained for polysaccharides yielding viscous solutions were poorly reproducible.We have simplified the reduction procedure 8 by using prolonged treatment of alginates with the so-called magic methanol 140 (a metha- nol ± chloroform ± concentrated HCl mixture, 10 : 1 : 1) for ester- ification and by reducing the formed ester with a large excess of sodium borohydride in a concentrated imidazole buffer (pH 7.0, 0 8C), by analogy with a method proposed for the identification of esterified galacturonic acid residues in pectins.148 The hydrolysis products of reduced polysaccharides were converted into aldono- nitrile acetates and analysed by GLC.The chromatograms revealed two peaks corresponding to gulose derivatives (1,6- anhydro-b-L-gulose and L-gulononitrile acetates) and a peak of acetylated D-mannononitrile. Despite incomplete esterification and, correspondingly, incomplete reduction, the M/G values calculated from the peak areas of the corresponding mannose and gulose derivatives coincided with the results of analysis of the 13C NMR spectra.A procedure for the determination of the M/G ratio which consists in the reduction of the carboxy groups with subsequent methylation, reductive splitting of the polymer with triethylsilane in the presence of trimethylsilyl methanesulfonate and boron trifluoride etherate and identification of the anhydroalditols (in the form of the corresponding derivatives) using HPLC or GLC, proposed by Zeller and Gray,149 seems to be too complicated for routine assays of alginic acids. 4. Partial hydrolysis for determining the block composition As stated in the introduction, physicochemical properties of alginates are largely determined by the mode of distribution of the two uronic acid residues along the polymeric chain.This distribution pattern can be characterised chemically using partial acid hydrolysis.26, 27 After short-term heating of alginic acid suspensions in dilute mineral acids (20 min, 100 8C, 0.3 M HCl), an oligomeric fraction with an average degree of polymerisation of *20 and containing approximately equal amounts of both uronic acid residues (MG-blocks) passes into the solution. The insoluble residue additionally heated for 20 h with acid is separated into two fractions, viz., a fraction soluble at pH 2.85 and insoluble one. These fractions also represent oligomers with average degrees of polymerisation of *20, but the former contains exclusively M-blocks (the mannuronic acid content exceeds 90%), while the latter consists of G-blocks (the mannuronic acid content is less than 10%).From the yields of all the three fractions, one can establish the distribution pattern of uronic acid residues in the linear chain of alginates. Using partial hydrolysis, it was demon- strated for the first time that alginates of different origin differ in fact in their block composition.150 In subsequent studies, a similar approach was repeatedly used for characterisation of alginates (see, e.g., Ref. 151). A simplified procedure for the determination of the block composition of alginates was described by Penman and Sander- son.18 These authors proposed to use a one-step treatment of alginates with an acid (5 h, 100 8C, 0.3 MHCl) instead of two-step hydrolysis and to determine the yields of soluble and insoluble fractions colorimetrically, by reaction with phenol and concen- trated sulfuric acid. The insoluble fraction obtained was analysed by 1H NMR spectroscopy as a solution in D2O at 80 8C rather than separated into blocks of different composition. The low-field region of the spectrum has a singlet for H(1) of the residue M at t=5.34 (peak B), a doublet for H(1) of the residue G at t=4.96 (peak A) and a singlet for H(5) of the residue G at t=5.56 (peak C).The ratios of integral intensities B/A or B/C (these 961 values can be averaged) and the yield of the soluble fraction permit one to calculate the block composition of the original polysac- charide.Yet another modification of the partial hydrolysis procedure is the use of ion-exchange and gel-permeation chromatography for the separation of oligomeric products.152, 153 Using several commercial alginate preparations as examples, it was shown that the size of M-blocks varies in the range of 3 ± 10, that of G-blocks varies in the range of 4 ± 15 monosaccharide residues, while the distribution of monomers within MG-blocks is random.152 Electrophoresis in polyacrylamide gel 155 was recommended for separation and quantitative determination of oligomannur- onic and oligoguluronic acids with a degree of polymerisation varying from 2 to 9 obtained by partial hydrolysis of alginates.154 In order to increase the electrophoretic mobility of the oligosac- charides and sensitivity of their assays, these were fluorescently labelled by reductive amination with 8-aminonaphthalene-1,3,6- trisulfonic acid.III. Enzymic and immunological methods All currently known enzymes able to depolymerise alginate molecules are lyases, i.e., they catalyse b-elimination of the substituent from position 4 of the uronic acid residue rather than hydrolysis of glycosidic bonds.156 ± 158 This cleavage results in the formation of a 4,5-unsaturated uronic acid residue (Scheme 1) at the newly formed end of one of the fragments. Since elimination results in disappearance of asymmetric centres at C(4) and C(5), the same unsaturated uronic acid is formed from both b-D- mannuronic and a-L-guluronic acids.This unsaturated residue has an intense absorption band at 230 ± 240 nm, and the increase in optical density at 230 nm can be used as a convenient method for determining the enzymic activity of alginate lyases.159 A mixture of oligomers with low degrees of polymerisation is a final product of enzymic treatment of polymeric alginic acids with alginate lyases.160 Scheme 1 HOOC HOOC OHO OHO HO O O HO O HO O HOOC OH OH HOOC OHO HOOC OHO O O OH O O H O HO COOH HOOC OHO HOOC OHO HOOC OHO OH + O HO O HO HO Alginate lyases have been detected in many organisms, e.g., sea molluscs, echinoderms, bacteria and fungi.161 Molluscs (Gastropoda) 162 and bacteria 163 ± 165 are the most convenient sources for the isolation of these enzymes.Alginate lyases of different origin differ essentially in their stabilities,166 specific- ities 167 ± 169 and abilities to hydrolyse acetylated bacterial algi- nates.170, 171 Thus M-lyases (polymannuronate lyases) normally isolated from molluscs are endoenzymes, which easily hydrolyse M- and MG-blocks without any effect on G-blocks.172, 173 In contrast, G-lyases (polyguluronate lyases) hydrolyse G- and MG- blocks but are inert with respect to M-blocks.162 The substrate specificities of alginate lyases can be examined using a recently proposed procedure for depolymerisation of alginates on agar plates in the presence of calcium chloride 174 or using homopoly- meric substrates containing a terminal fluorescent label.175 A convenient and versatile procedure which takes into account the specificity of alginate lyases and is suitable for detection and analysis of alginates in solutions at concentrations962 of 0.01 to 1 mg ml71 (see Ref.176) was developed for quantitative determination of various alginic acids. The material under study was treated with a mixture of two enzymes of different origin, viz., a mixture of poly-M-lyase isolated from the mollusc Haliotis tuberculata and poly-G-lyase isolated from the bacterium Kleb- siella pneumoniae; the absorption of reaction products was measured at 230 nm. With the use of an excess of enzymes it was shown that the final values of optical density were proportional to the initial concentration of the substrate.A combination of two enzymes made it possible to eliminate the effect of the variable composition of different alginates. It is noteworthy that the separate application of these enzymes can provide information about the block structure of various alginates. This recently developed procedure holds much promise, especially for future studies provided alginate lyases become commercially available. Alginate lyases of known specificity are also used for obtain- ing information about the distribution of uronic acid residues and the sizes of individual blocks in alginates. It is bacterial enzymes that made it possible to establish, for the first time, the deviation of the uronic acid sequence in poly-MG from a strictly alternating pattern.160, 177 Some alginates were found to contain poly-M- blocks that are homogenous in size and have an average degree of polymerisation of 24;160 however, more recent studies with other subjects established more variable distribution patterns of poly- G- and poly-M-blocks according to their size.178, 179 The specificity of action of the enzyme isolated from the sea mollusc Haliotis tuberculata was studied in most detail.High- performance liquid chromatography was specially adapted for the separation of saturated and unsaturated oligoguluronates and oligomannuronates formed upon enzymic hydrolysis of algi- nates 180 and different procedures of NMR spectroscopy 181 were specially developed for the elucidation of the structure of these oligosaccharides. It was found that alginate lyase effectively cleaves M7M and G7M bonds (but not G7G or M7G bonds) and that the pentamer of mannuronic acid is best bound to the enzyme active centre.In an analogous study of recombinant bacterial alginate lyase,182 it was established that the enzyme cleaved exclusively M7M bonds. The enzymic action is blocked by the presence of the acetyl groups in mannuronic acid residues, which is character- istic of bacterial alginates. This enzyme was recommended for the preparative isolation of poly-M-blocks from acetylated polyman- nuronate of Pseudomonas aeruginosa and of poly-MG-blocks from another bacterial alginate containing residues of both uronic acids.183 Alginates possess antigenic activity.Several attempts were undertaken to raise antibodies against definite structural elements of alginate molecules and to use these antibodies for the analysis of alginates by standard immunological procedures.184 Specific antibodies (both polyclonal 185, 186 and monoclonal 187) against alginates were used to localise the polysaccharide in various algal tissues; however, this assay was complicated by gel formation. Considering that gel formation decreases the ability of the corresponding sites in the alginate molecule to bind to antibodies and that the reaction conditions (pH, temperature, accessibility of Ca2+ ions and the nature of other buffer components) can strongly influence the results of immunological assays,188 it may be concluded that these studies are still far from being completed and that the immunological behaviour of alginates and the specificity of the corresponding antibodies need additional inves- tigation.IV. Nondestructive physicochemical methods 1. IR spectroscopy The total amount of polymers consisting of uronic acids can be determined from the intensity of the absorption band of the ionised carboxyl at 1607 cm71 in the IR spectrum of a sample solution in a D2O-phosphate buffer.189 It was recommended to determine semi-quantitatively the M/G value in alginates by A I Usov measuring the ratio of absorption band intensities at 808 (M) and 787 cm71 (G) in the IR spectra of specially prepared films.190 Comparison of band intensities at 1320 (M) and 1290 cm71 (G) can also be used for this purpose.191 However, because of low precision and interference due to different factors these two methods have not found wide application. On the other hand, IR spectra provide valuable information about the presence of acetyl groups in bacterial alginates 22, 99, 101 judging from intense absorption bands at 1250 and 1730 cm71.Recently, it was proposed to use spectroscopy in the near IR region (400 ± 2500 nm) for determination of the alginate content in brown algae biomass.192 The reflectance spectra recorded for powdered biomass samples are computer-processed at 9 parame- ters. The experimentally determined calibration curves allow the determination of alginate in the starting material with high precision.This method was successfully used for characterisation of biodegradation of the alga Laminaria hyperborea in which the alginate content varies from 2.2% to 40.8%. Horn et al.,192 who have developed this procedure, gave it a high estimate, although this analysis requires the use of a specially designed spectropho- tometer. It was noted also that variable content of moisture and polyphenols in the specimens under study interfere with the analysis of alginates. 2. Circular dichroism The three types of blocks present in alginate molecules differ essentially in their circular dichroism (CD) spectra; the poly-G spectrum lies completely in the region of negative molar ellipticity values, [y], the poly-M spectrum has an intense positive band, while the poly-MG spectrum is characterised by intermediate values of [y].193, 194 The CD spectra of natural alginates are characterised by a peak at 200 nm and a trough at 215 nm.It was found that the M/G ratio may be calculated from the observed ratio of peak height to trough depth using simple equations.195 Therefore, this important parameter can be deter- mined from the CD spectra even for milligram quantities of the alginate without any destruction of the polymer. Since the CD spectra of mixed blocks (poly-MG) are not identical with the spectra of equimolar poly-M and poly-G mixtures, the determination of the block composition of alginates from their CD spectra is possible.A relatively simple computer- assisted calculation was proposed which mimics the real spectrum owing to a linear combination of three types of spectra character- istic of isolated blocks.195 Although the shape of poly-MG spectra depends on the distribution of uronic acid residues within the blocks, which is not taken into consideration in the calculation procedure, the results obtained with known alginate specimens showed good coincidence with those earlier established using partial hydrolysis or NMR spectra. The CD spectra were subsequently used to characterise the block composition of alginates isolated from many species of brown algae or from different types of algal tissues.47, 196, 197 3. 13C NMR Spectroscopy A preliminary report concerning the interpretation of 13C NMR spectra of alginates was published in 1977 by Grasdalen et al.20 who showed that the alginate spectra recorded at 25 MHz contained four signals in the anomeric resonance region rather than two signals, as might be expected for a polysaccharide made up of two uronic acid residues.This implies that the position of C(1) signals of both monosaccharides (M and G) depends on the nature of the neighbouring residue. The intensities of these signals could be used for estimating both M/G and the relative content (F ) of pairs of monomers (MM, MG, GM and GG). Moreover, the signal corresponding to the C(5) atom of the residue M was represented in the spectrum by three lines, which suggests the dependence of the signal position on the nature of both neigh- bouring residues.This phenomenon could be used for quantifica- tion of the number of triplets with a central M residue within the polymer. It was found that the central line of the signal corre-Alginic acids and alginates: analytical methods used for their estimation and characterisation of composition and primary structure and sponds to the central residuesMof theMMMandGMGtriplets, whereas the low-field and high-field lines represent the corre- sponding signals for GMM and MMG, respectively. Thus, the relative content of the latter two triplets in the alginate molecule can be measured directly, from the intensity of the corresponding signals, and the probability of the presence of triplets MMM and GMG in the polymeric chain can be calculated from the ratios: FMMM+FGMM=FMM and FGMG+FMMG=FMG.The data calculated for several alginates showed good correlation with the results of chemical analysis. However, no possibility to determine the number of triplets with a central (G) residue was found in this pioneering study. Later, in detailed studies Grasdalen et al.21 used a higher- resolution NMR spectrometer (50 MHz) and carefully selected the conditions for recording NMR spectra. The signals for the anomeric carbon atoms gave eight lines corresponding to the central units of eight theoretically possible triplets of the primary structure. The intensities of these lines were used to calculate the content of each type of triplets in several specimens of natural alginates.The validity of these calculations was partially con- firmed by an analysis of signals corresponding to the carboxyl carbon atoms and to C(4) and C(5) atoms of the residue M, which also have multiplet structures, and by comparison of these data with the results of chemical assays. The average lengths of homopolymeric blocks can be calculated from the triplet con- tent: NM>1 = (FM7 NG>1=(FG7FMGM)/FGGM FGMG)/FMMG. It is of note that the most probable distribution of residues along the polymeric chain in synthetic binary copolymers formed by polymerisation of two different monomers can be calculated, taking into account the initial concentrations and relative reac- tivities of the monomers.However, the block composition of alginates usually differs from the calculated sequence, which is at variance with the mechanism of formation of these polysacchar- ides by copolymerisation of two monosaccharide precursors and is a convincing argument in favour of an alternative route of their biosynthesis involving mannuronan-C(5)-epimerase.198, 199 13C NMR spectroscopy is an extremely convenient procedure for structural characterisation of alginates isolated from different natural sources (see, e.g., Refs 8, 9, 200) or modified by enzymic treatment.201 This was also used to monitor alginate biosynthesis in bacteria.202 However, recent experiments have shown 49, 203 that the reliability of determination of the block composition depends on M/G of the specimen under study.More adequate results were obtained for alginates with a high content of one of the monomers; if M/G is close to unity, the calculated values of individual block lengths can differ substantially from the results obtained upon chemical or enzymic hydrolysis. 13C NMR spectra were obtained not only for solutions but also for solid alginate samples.204 In the latter case, the resolution of peaks was much lower, especially for the anomeric carbon signals. Nevertheless, a complex spectral pattern could be repre- sented as a sum of individual symmetric signals corresponding to the C(2)7C(5) atoms of both uronic acid residues, which made it possible to estimate the M/G ratio with a sufficiently high accuracy.204 4. 1H NMR spectroscopy The 1H NMR spectra of alginates were studied in parallel with 13C NMR spectra in the hope to elaborate a more sensitive and accurate method for the determination of the content of mono- mers and their block distribution.It was shown 19 that informative spectra of alginates can be obtained using a 100 MHz spectrom- eter after short-term mild acid treatment of samples, which reduces the viscosity of solutions. Although not all signals could be resolved in these spectra, the low-field region contained three well-defined peaks, viz., A [H(1) of the residue G], B [a sum of H(1) of the residue M and H(5) of the residue MG] and C [H(5) of the residue GG], which made it possible to calculate the following values: FG=IA/(IB+IC) and FGG=IC(IB+IC).Taking into 963 FG+FM=1, FGG+FGM=FG account and that FMM+FMG=FM (for lengthy chains, FMG=FGM), the M/G ratio and the content of the four possible monomeric pairs in alginate molecules can be calculated. The parameters established for well-known specimens coincided with the results obtained by chemical methods and 13C NMR spectroscopy. In a continuation of these studies, 1H NMR spectra of alginates and their fractions were recorded at 400 MHz.205 As in the above-cited paper,19 the M/G values and contributions of the dimeric sequences (MM, MG=GM and GG) were calculated from the analysis of 1H NMR spectra. In addition, higher resolution made it possible to isolate four signals corresponding to the H(5) atom of the central residue of a-L-guluronic acid in the four possible triads (GGG, MGG, GGM and MGM). The content of these trimers showed a good correlation with the values obtained by other analytical methods.The 1H NMR spectra of poly-M and poly-G, the components of a bacterial alginate produced by Azotobacter vinelandii, were further analysed using both one-dimensional and modern two- dimensional NMR techniques.206, 207 The unambiguous assign- ment of signals was performed on the basis of chemical shift theory, analysis of spin ± spin coupling constants and measure- ments of the nuclear Overhauser effects. These data were used to study the interaction of poly-M- and poly-G-blocks with Ca2+ ions in aqueous media.A model for binding of calcium ions was proposed, which is based on the measurement of the relative intensity of cross-peaks in the two-dimensional spectra (NOESY) of the free polymer and its calcium salt. The experimental results are important for interpreting the gel-forming ability of alginates in the presence of calcium ions and elucidation of the role of calcium as a cofactor essential for the functioning of mannuronan- C(5)-epimerase. High-resolution 1H NMR spectroscopy was efficiently used for characterisation of many natural alginates, of both algal 196, 208 and bacterial 209, 210 origin, particularly for establishing the degree of acetylation and localisation of acetyl groups in bacterial alginates.24 It was noted, however,204 that partial hydrolysis traditionally used for decreasing the viscosity of solutions of high-molecular-weight alginates prior to recording their proton NMRspectra can induce the precipitation of a portion of alginate and thus significantly distort the experimental results (M/G ratio).V. Conclusion As may be inferred from the survey, the analytical chemistry of alginic acids is a well-grounded and well-studied field. Numerous approaches have been proposed for the solution of analytical problems of different degrees of complexity, from quantitative determination of alginates in mixtures with other natural com- pounds to the analysis of uronic acid sequences in the linear molecules of these polysaccharides. Alginic acids differ from other natural polyuronides in that they are made up of unique monosaccharides and are character- ised by a peculiar distribution of monomers along the polymeric chain.Special approaches were elaborated for investigation of these structural features among which partial acid hydrolysis, enzymic degradation and some nondestructive physicochemical methods (in the first place,NMRspectroscopy) have proved to be especially efficient. The experimental findings have made it possible to formulate a concept of the primary block structure of alginic acids and to gain a deeper insight into the pathways of their biosynthesis leading to this structure. These results not only have a great theoretical significance, but are of great practical value, since they permit direct alteration of physicochemical properties of alginates by enzymic treatment.201, 211 However, it should be remembered that specific analytical approaches were elaborated for the solution of concrete practical problems and have certain limitations which have been mentioned in the review.Further in- depth studies into the structure and properties of alginates will964 probably demand further improvement of existing methods of structural analysis of these polysaccharides. The significance of analytical chemistry of alginates goes far beyond the confines of this class of biopolymers. As uronic acid- containing polysaccharides are widely spread in nature, many chemical methods designed for the detection and quantitative determination of alginates, particularly, the approaches based on specific degradation of uronic acids or chromatography of acid hydrolysates, are equally efficient in the analysis of plant muci- lages, pectins, hemicelluloses, bacterial and fungal polysacchar- ides and animal glycuronoglycans.References 1. E C C Stanford J. Chem. Soc. 44 943 (1883) 2. E Percival, R H McDowell,in Chemistry and Enzymology of Marine Algal Polysaccharides (London: Academic Press, 1967) p. 99 3. E Percival, R H McDowell, in Encyclopedia of Plant Physiology Vol. 13B (EdsWTanner, FALoewus) (Berlin: Springer, 1981) p. 277 4. T J Painter, in The Polysaccharides Vol. 2 (Ed. G O Aspinall) (New York: Academic Press, 1983) p. 195 5. P A J Gorin, J F T Spencer Can. J.Chem. 44 993 (1966) 6. A Linker, R S Jones Nature (London) 204 187 (1964) 7. M Okazaki, K Furuya, K Tsukayama, K Nisizawa Bot. Mar. 25 123 (1982) 8. A I Usov,M I Bilan, N G Klochkova Bot. Mar. 38 43 (1995) 9. A I Usov,M I Bilan Bioorg. Khim. 22 126 (1996) a 10. W L Nelson, L H Cretcher J. Am. Chem. Soc. 51 1914 (1929) 11. W L Nelson, L H Cretcher J. Am. Chem. Soc. 52 2130 (1930) 12. F G Fischer, H Dorfel Z. Physiol. Chem. 302 186 (1955) 13. E L Hirst, E Percival, J K Wold J. Chem. Soc. 1493 (1964) 14. E L Hirst, D A Rees J. Chem. Soc. 1182 (1965) 15. D A Rees, J W B Samuel J. Chem. Soc. C 2295 (1967) 16. T J Painter, B Larsen Acta Chem. Scand. 24 813 (1970) 17. T J Painter, B Larsen Acta Chem. Scand. 27 1957 (1973) 18. A Penman, G R Sanderson Carbohydr.Res. 25 273 (1972) 19. H Grasdalen, B Larsen, O Smidsrod Carbohydr. Res. 68 23 (1979) 20. H Grasdalen, B Larsen, O Smidsrod Carbohydr. Res. 56 C11 (1977) 21. H Grasdalen, B Larsen, O Smidsrod Carbohydr. Res. 89 179 (1981) 22. A Linker, R S Jones J. Biol. Chem. 241 3845 (1966) 23. L R Evans, A Linker J. Bacteriol. 116 915 (1973) 24. G Skjak-Braek, H Grasdalen, B Larsen Carbohydr. Res. 154 239 (1986) 25. A Haug, Composition and Properties of Alginates (Trondheim: Norwegian Institute of Seaweed Research, 1964) Rep. 30, p. 1 26. A Haug, B Larsen, O Smidsrod Acta Chem. Scand. 20 183 (1966) 27. A Haug, B Larsen, O Smidsrod Acta Chem. Scand. 21 691 (1967) 28. G Skjak-Braek Biochem. Soc. Trans. 20 27 (1992) 29. D A Rees, E J Welsh Angew.Chem., Int. Ed. Engl. 16 214 (1977) 30. B T Stokke, O Smidsrod, P Bruheim, G Skjak-Braek Macromolecules 24 4637 (1991) 31. B T Stokke, O Smidsrod, F Zanetti, W Strand, G Skjak-Braek Carbohydr. Polym. 21 39 (1993) 32. K I Draget, G Skjak-Braek, O Smidsrod Carbohydr. Polym. 25 31 (1994) 33. M Fujihara, T Nagumo Carbohydr. Res. 224 343 (1992) 34. M Fujihara, T Nagumo Carbohydr. Res. 243 211 (1992) 35. T Suzuki, K Nakai, Y Yoshie, T Shirai, T Hirano Nippon Suisan Gakkaishi 59 545 (1993); Chem. Abstr. 119 94 363 (1993) 36. W H McNeely, D J Pettitt, in Industrial Gums, Polysaccharides and Their Derivatives (Ed. R L Whistler) (New York: Academic Press, 1973) p. 49 37. K Clare, in Industrial Gums, Polysaccharides and Their Derivatives (Eds R L Whistler, J N BeMiller) (New York, San Diego: Academic Press, 1993) p.105 38. G Skjak-Braek, A Martinsen, in Seaweed Resources in Europe, Uses and Potential (EdsMD Guiry, G Blunden) (New York: Wiley, 1991) p. 219 39. O Smidsrod, K I Draget Carbohydr. Eur. (14) 6 (1996) 40. S Valla, H Ertesvag, G Skjak-Braek Carbohydr. Eur. (14) 14 (1996) 41. G Skjak-Braek, T Espevik Carbohydr. Eur. (14) 19 (1996) 42. E Onsoyen Carbohydr. Eur. (14) 26 (1996) A I Usov 43. V J Chapman, D J Chapman Seaweeds and Their Uses (London, New York: Chapman and Hall, 1980) p. 194 44. A Haug Methods Carbohydr. Chem. 5 69 (1965) 45. D W Drummond, E L Hirst, E Percival J. Chem. Soc. 1208 (1962) 46. K G Rosell, L M Srivastava Can. J. Bot. 62 2229 (1984) 47.J S Craigie, E R Morris, D A Rees, D Thom Carbohydr. Polym. 4 237 (1984) 48. D J Wedlock, B A Fasihuddin Food Hydrocolloids 4 41 (1990) 49. R Panikkar, D J Brasch Carbohydr. Res. 293 119 (1996) 50. P Gacesa Carbohydr. Polym. 8 161 (1988) 51. N J Russell, P Gacesa Mol. Aspects Med. 10 1 (1988) 52. T B May, A M Chakrabarty Methods Enzymol. 235 295 (1994) 53. B H A Rehm, S Valla Appl. Microbiol. Biotechnol. 48 281 (1997) 54. F Clementi Crit. Rev. Biotechnol. 17 327 (1997) 55. P Gacesa Microbiology (UK) 144 1133 (1998) 56. T Akahane, S Takeuchi, A Minakata Polym. Bull. 24 437 (1990) 57. M Uddin Pak. J. Sci. Ind. Res. 30 343 (1987) 58. M C Cameron, A G Ross, E G V Percival J. Soc. Chem. Ind. 67 161 (1948) 59. G O Kim, C G Pu,M H Li, G O Lyu Punsok Hwahak (3) 13 (1982); Chem.Abstr. 98 140 058 (1983) 60. A Jensen, M Indergaard, T J Holt Bot. Mar. 28 375 (1985) 61. A Haug, O Smidsrod Acta Chem. Scand. 24 843 (1970) 62. E Murano, S Paoletti, A Cesaro, R Rizzo Anal. Biochem. 187 120 (1990) 63. G R Seely, R R Knotts Carbohydr. Polym. 3 109 (1983) 64. M K Pal, N Mandal Biopolymers 29 1541 (1990) 65. D F Day, M L Marceau-Day Ann. New York Acad. Sci. 613 751 (1990) 66. J P Halle, D Landry, A Fournier, M Beaudry, F A Leblond Cell Transplantation 2 429 (1993) 67. J E Scott Methods Carbohydr. Chem. 5 38 (1965) 68. J F Kennedy, I J Bradshaw Br. Polym. J. 16 95 (1984) 69. J F Kennedy, I J Bradshaw Carbohydr. Polym. 7 35 (1987) 70. J F Kennedy, I J Bradshaw Carbohydr. Polym.7 409 (1987) 71. Y Murata, S Kawashima, E Miyamoto, N Masauji, A Honda Carbohydr. Res. 208 289 (1990) 72. A S Perlin Can. J. Chem. 30 278 (1952) 73. R M McCready,H A Swenson,W D Maclay Ind. Eng. Chem., Anal. Ed. 18 290 (1946) 74. A Jensen, I Sunde Proc. Int. Seaweed Symp. 2 125 (1956) 75. B Larsen, in Handbook of Physiological Methods. Physiological and Biochemical Methods (Eds J A Hellebust, J S Craigie) (Cambridge: Cambridge University Press, 1978) p. 143 76. D M W Anderson Talanta 2 73 (1959) 77. J N C Whyte, J R Englar Anal. Biochem. 59 426 (1974) 78. A E Castagne, I R Siddiqui Carbohydr. Res. 42 382 (1975) 79. L P Kosheleva, G I Il'chenko, L I Glebko Anal. Biochem. 73 115 (1976) 80. L I Glebko, G Ya Il'chenko, L P Kosheleva Khim.Prir. Soedin. 10 (1976) 81. R W Scott, W E Moore,M J Effland, M A Millett Anal. Biochem. 21 68 (1967) 82. Z Dische J. Biol. Chem. 167 189 (1947) 83. Z Dische J. Biol. Chem. 183 489 (1950) 84. B E F de Arcuri,M E F de Recondo, E F Recondo Carbohydr. Res. 85 165 (1980) 85. N Blumenkrantz, G Asboe-Hansen Anal. Biochem. 54 484 (1973) 86. R W Scott Anal. Chem. 51 936 (1979) 87. E G V Percival, A G Ross J. Soc. Chem. Ind. 67 420 (1948) 88. J D Gregory Arch. Biochem. Biophys. 89 157 (1960) 89. T Bitter, H M Muir Anal. Biochem. 4 330 (1962) 90. J T Galambos, J R McCain Anal. Biochem. 19 119 (1967) 91. J T Galambos, J R McCain Anal. Biochem. 19 133 (1967) 92. C A Knutson, A Jeanes Anal. Biochem. 24 470 (1968) 93. C A Knutson, A Jeanes Anal. Biochem.24 482 (1968) 94. M P Filippov, T V Vlas'eva Prikl. Biokhim. Mikrobiol. 9 134 (1973) 95. M P Filippov, T V Vlas'eva, V I Kuz'minov, N P Dzyuba, V N Chushenko, V V Sayanova Izv. Akad. Nauk Mold. SSR, Ser. Biol. Khim. Nauk (1) 75 (1976) 96. P K Kintner, J P Van Buren J. Food Sci. 47 756 (1982) 97. T M C C Filisetti-Cozzi, N C Carpita Anal. Biochem. 197 157 (1991)Alginic acids and alginates: analytical methods used for their estimation and characterisation of composition and primary structure 98. D A T Southgate, in Gums and Stabilisers for the Food Industry Vol. 2 (Eds G O Phillips, D J Wedlock, P A Williams) (Oxford: Pergamon Press, 1984) p. 3 99. V Sherbrock-Cox, N J Russell, P Gacesa Carbohydr. Res. 135 147 (1984) 100. G B Pier, J M Saunders, P Ames, M S Edwards, H Auerbach, J Goldfarb, D P Speert, S Hurwitch New Engl.J. Med. 317 793 (1987) 101. S S Pedersen, F Espersen,N Hoiby,G H Shand J. Clin. Microbiol. 27 691 (1989) 102. K Kawana, T Nakaoka, S Watanabe Kanagawa-ken Eisei Kenkyusho Kenkyu Hokoku 21 25 (1991); Chem. Abstr. 117 46 824 (1992) 103. M Dubois, K A Gilles, J K Hamilton, P A Rebers, F Smith Anal. Chem. 28 350 (1956) 104. H D Graham J. Dairy Sci. 52 443 (1969) 105. H D Graham J. Food Sci. 35 494 (1970) 106. N E Aponte de Otaola,M Diaz-Piferrer, H D Graham J. Agric. Univ. Puerto Rico 47 464 (1983); Chem. Abstr. 100 65 018 (1984) 107. M Toyoda, C Yomota, Y Ito,M Harada Shokuhin Eiseigaku Zassi 26 189 (1985); Chem. Abstr. 103 195 032 (1985) 108.A H Brown Arch. Biochem. 11 269 (1946) 109. E G Brown, T J Hayes Analyst 77 445 (1952) 110. A Haug, B Larsen Acta Chem. Scand. 16 1908 (1962) 111. I R Siddiqui Carbohydr. Res. 67 289 (1978) 112. A Haug, B Larsen Acta Chem. Scand. 15 1395 (1961) 113. R Fey, F J Mack Lebensm. Gerichtl. Chem. 35 50 (1981) 114. F Schur, P Anderegg, H Pfenninger Brau-Rundsch. 92 25 (1981) 115. H B Pfenninger, P Anderegg, E Vykoukal J. Am. Soc. Brew. Chem. 41 40 (1983) 116. H Scherz Z. Lebensm.-Unters. Forsch. 179 17 (1984) 117. H Scherz Lebensm. Gerichtl. Chem. 39 32 (1985) 118. J Wingender, V Sherbrock-Cox, P Gacesa, N J Russell Biochem. Soc. Trans. 13 1148 (1985) 119. B Larsen, A Haug Acta Chem. Scand. 15 1397 (1961) 120. H Anzai, N Uchida, E Nishide Nippon Suisan Gakkaishi 56 73 (1990); Chem.Abstr. 113 37 203 (1990) 121. M-H Ji, W D Cao, L J Han Oceanol. Limnol. Sin. 12 240 (1981); Chem. Abstr. 95 183 279 (1981) 122. A G J Voragen, H A Schols, J A De Vries, W Pilnik J. Chromatogr. 244 327 (1982) 123. A G J Voragen, H A Schols,W Pilnik Progr. Food Nutr. Sci. 6 379 (1982) 124. G Annison, N W H Cheetham, I Couperwhite J. Chromatogr. 264 137 (1983) 125. L H Krull, G L Cote Carbohydr. Polym. 17 205 (1992) 126. S Honda, S Suzuki,M Takahashi, K Kakehi, S Ganno Anal. Biochem. 134 34 (1983) 127. P Gacesa, A Squire, P J Winterburn Carbohydr. Res. 118 1 (1983) 128. C Bouffar-Roupe, A Heyraud Food Hydrocolloids 1 559 (1987) 129. S Honda, K Sudo, K Kakehi, H Yuki, K Takiura Anal. Chim. Acta 77 199 (1975) 130.J R Clamp, J E Scott Chem. Ind. 652 (1969) 131. J F Kennedy, S M Robertson, M Stacey Carbohydr. Res. 49 243 (1976) 132. J F Kennedy, S M Robertson, M Stacey Carbohydr. Res. 57 205 (1977) 133. M B Perry, R K Hulyalkar Can. J. Biochem. 43 573 (1965) 134. E Sjlstrlm, P Haglund, J Janson Acta Chem. Scand. 20 1718 (1966) 135. J D Blake, G N Richards Carbohydr. Res. 8 275 (1968) 136. T M Jones, P Albersheim Plant Physiol. 49 926 (1972) 137. E Sjostrom, K Pfister, E Seppala Carbohydr. Res. 38 293 (1974) 138. J Lehrfeld Anal. Biochem. 115 410 (1981) 139. J Lehrfeld J. Chromatogr. 408 245 (1987) 140. S A Fazio, D J Uhlinger, J H Parker, D C White Appl. Environ. Microbiol. 43 1151 (1982) 141. G L Cote, L H Krull Carbohydr.Res. 181 143 (1988) 142. R L Taylor, H E Conrad Biochemistry 11 1383 (1972) 143. L Vadas, H S Prihar, B K Pugashetti, D S Feingold Anal. Biochem. 114 294 (1981) 144. K Larsson, O Larm, K Granath Acta Chem. Scand., Ser. B 38 15 (1984) 145. M Fujihara, T Nagumo Nippon Kagaku Kaishi (4) 576 (1986); Chem. Abstr. 106 5362 (1987) 965 146. E Quintero, K Ishida, G Gordon, G Geesey J. Microbiol. Methods 9 309 (1989) 147. W S York, A G Darvill,M McNeil, T T Stevenson, P Albersheim Methods Enzymol. 118 3 (1986) 148. N O Maness, J D Ryan, A J Mort Anal. Biochem. 185 346 (1990) 149. S G Zeller, G R Gray Carbohydr. Res. 226 313 (1992) 150. A Haug, B Larsen, O Smidsrod Carbohydr. Res. 32 217 (1974) 151. E Nishide, H Anzai, N Uchida Hydrobiologia 151/152 551 (1987) 152.A Heyraud, C Leonard Food Hydrocolloids 4 59 (1990) 153. A Heyraud, C Leonard Carbohydr. Res. 215 105 (1991) 154. T Shimokawa, S Yoshida, T Takeuchi, K Murata, T Ishii, I Kusakabe Biosci. Biotech. Biochem. 60 1532 (1996) 155. T Shimokawa, S Yoshida, I Kusakabe, T Takeuchi, K Murata Carbohydr. Res. 299 95 (1997) 156. R J Linhardt, P M Galliher, C L Cooney Appl. Biochem. Biotechnol. 12 135 (1986) 157. F Haugen, F Kortner, B Larsen Carbohydr. Res. 198 101 (1990) 158. P Gacesa Int. J. Biochem. 24 545 (1992) 159. H I Nakada, P C Sweeney J. Biol. Chem. 242 845 (1967) 160. J Boyd, J R Turvey Carbohydr. Res. 66 187 (1978) 161. D M Butler, L V Evans, B Kloareg, in Introduction to Applied Phycology (Ed. I Akatsuka) (The Hague: SPB Academic Publishing, 1990) p.647 162. C Boyen, B Kloareg, M Polne-Fuller, A Gibor Phycologia 29 173 (1990) 163. C Boyen, Y Bertheau, T Barbeyron, B Kloareg Enzyme Microbiol. Technol. 12 885 (1990) 164. I M Aasen, K Folkvord, D Levine Appl. Microbiol. Biotechnol. 37 55 (1992) 165. K Ostgaard, S H Knutsen, N Dyrset, I M Aasen Enzyme Microbiol. Technol. 15 756 (1993) 166. S Takeshita, N Sato,M Igarashi, T Muramatsu Biosci. Biotech. Biochem. 57 1125 (1993) 167. B J Brown, J F Preston Carbohydr. Res. 211 91 (1991) 168. F Eftekhar, N L Schiller Current Microbiol. 29 37 (1994) 169. B Larsen, K Hooen, K Ostgaard Hydrobiologia 261 557 (1993) 170. T Hisano, M Nishimura, Y Yonemoto, S Abe, T Yamashita, K Sakaguchi, A Kimura, K Murata J.Ferment. Bioeng. 75 332 (1993) 171. T Hisano, H Yamaguchi, Y Yonemoto, K Sakaguchi, T Yamashita, S Abe, A Kimura, K Murata J. Ferment. Bioeng. 75 220 (1993) 172. K Ostgaard, B Larsen Carbohydr. Res. 246 229 (1993) 173. K Ostgaard, B T Stokke, B Larsen Carbohydr. Res. 260 83 (1994) 174. T Hisano, M Nishimura, T Yamashita, T Imanaka, T Muramatsu, A Kimura, K Murata J. Ferment. Bioeng. 78 182 (1994) 175. F S Wusteman, P Gacesa Carbohydr. Res. 241 237 (1993) 176. K Ostgaard Carbohydr. Polym. 19 51 (1992) 177. K H Min, S F Sasaki, Y Kashiwabara, M Umekawa, K Nisizawa J. Biochem. 81 555 (1977) 178. F Haugen, B Larsen Polym. Prepr. 30 363 (1989) 179. K Ostgaard Hydrobiologia 260/261 513 (1993) 180. A Heyraud, P Colin-Morel, S Girond, C Richard, B Kloareg Carbohydr. Res. 291 115 (1996) 181. A Heyraud, C Gey, C Leonard, C Rochas, S Girond, B Kloareg Carbohydr. Res. 289 11 (1996) 182. F Chavagnat, A Heyraud, P Colin-Morel,M Guinand, J Wallach Carbohydr. Res. 308 409 (1998) 183. A Heyraud, P Colin-Morel, C Gey, F Chavagnat, M Guinand, J Wallach Carbohydr. Res. 308 417 (1998) 184. V Vreeland, E Zablackis, B Doboszewski, W M Laetsch Hydrobiologia 151/152 155 (1987) 185. V Vreeland, D J Chapman J. Immunol. Methods 23 227 (1978) 186. V Vreeland,W M Laetsch Proc. Int. Seaweed Symp. 10 531 (1981) 187. V Vreeland,M Slomich, W M Laetsch Planta 162 506 (1984) 188. B Larsen, V Vreeland, W M Laetsch Carbohydr. Res. 143 221 (1985) 189. S M Bociek, D Welti Carbohydr. Res. 42 217 (1975) 190. W Mackie Carbohydr. Res. 20 413 (1971) 191. M P Filippov, R Kohn Chem. Zvesti 28 817 (1974) 192. S J Horn, E Moen, K Ostgaard, in The XVIth International Seaweed Symposium (Abstracts of Reports), Cebu, 1998 p. 69A I Usov 966 193. E R Morris, D A Rees, D Thom J. Chem. Soc., Chem. Commun. 245 (1973) 194. E R Morris, D A Rees, G R Sanderson, D Thom J. Chem. Soc., Perkin Trans. 2 1418 (1975) 195. E R Morris, D A Rees, D Thom Carbohydr. Res. 81 305 (1980) 196. B Stockton, L V Evans, E R Morris, D A Powell, D A Rees Bot. Mar. 23 563 (1980) 197. B Stockton, L V Evans, E R Morris, D A Rees Int. J. Biol. Macromol. 2 176 (1980) 198. B Larsen Proc. Int. Seaweed Symp. 10 8 (1981) 199. B Larsen, G Skjak-Braek, T Painter Carbohydr. Res. 146 342 (1986) 200. A I Usov, E A Kosheleva, A P Yakovlev Bioorg. Khim. 11 830 (1985) a 201. G Skjak-Braek,O Smidsrod, B Larsen Int. J. Biol. Macromol. 8 330 (1986) 202. A Narbad, M J E Hewlins, P Gacesa, N J Russell Biochem. J. 267 579 (1990) 203. R Panikkar, D J Brasch Carbohydr. Res. 300 229 (1997) 204. F Llanes, F Sauriol, F G Morin, A S Perlin Can. J. Chem. 75 585 (1997) 205. H Grasdalen Carbohydr. Res. 118 255 (1983) 206. C A Steginsky, J M Beale, H G Floss, R M Mayer Carbohydr. Res. 225 11 (1992) 207. J M Beale, C A Steginsky, B M Mamiya Environ. Sci. Res. 44 209 (1992) 208. M Indergaard, G Skjak-Braek Hydrobiologia 151/152 541 (1987) 209. J R W Govan, P Sarasola, D Taylor, P J Tatnell, N J Russell, P Gacesa J. Clin. Microbiol. 30 595 (1992) 210. P J Tatnell, N J Russell, P Gacesa Biochem. Soc. Trans. 22 310 (1994) 211. D F Day, R D Ashby, J W Lee ACS Symp. Ser. 684 175 (1998) a�Russ. J. Bioorg. Chem. (Engl. Transl.) b�Chem. Nat. Compd. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (11) 967 ± 982 (1999) Site-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents MI Dobrikov Contents I. Introduction II. Free-radical cycloaddition and electrophilic addition photoreactions of nucleic acids III. Sensitised photomodification of DNA with binary systems IV. Two-quantum photosensitised modification of DNA V. The prospects for the use of photoreagents for targeted modification of nucleic acids in vivo Abstract. Photosensitised oxygen-independent chemical modifi- cation of nucleic acids including free-radical cycloaddition and electrophilic addition (type III photodynamic effect) are consid- ered. The main attention is given to site-specific photomodifica- tion. New approaches to highly specific affinity modification of nucleic acids, viz., sensitised photomodification with binary photoreagent ± sensitiser systems and two-photon photosensi- tised reactions, are described.The prospects for application of electrophilic photoreagents for in vivo modification of nucleic acids are discussed. The bibliography includes 119 references. I. Introduction The development of methods for targeted chemical modification of nucleic acids (NA) at definite nucleotide sequences is of great importance for the study of cellular processes connected with gene expression,1, 2 design of gene-addressed biologically active sub- stances 3 and the development of new technologies based on the use of oligonucleotide chips and microelements for new gener- ations of computers.4 The targeted modification of NA is carried out with affinity reagents which can bind specifically to definite nucleotide sequen- ces and contain reactive groups bringing about NA modification upon binding.Fragments of antibiotics,5, 6 oligonucleotides 3, 7, 8 and their derivatives 2, 9 are used for the recognition of definite nucleotide sequences. Compounds generating free radicals,3, 10 alkylating reagents 3, 8 and photoreactive compounds 2, 3, 7, 11 are used as the affinity reagents. Chemically reactive oligonucleotide derivatives can react with different cell components, which results in undesirable toxicity and considerable expenditure of reagents in side reactions. Photo- reactive derivatives possess considerable advantages, viz., they are nontoxic in the dark and their photoactivation is easily controlled.Considerable experience has been gained in the last few years in the area of synthesis, studies of physicochemical properties, biological application of novel reagents containing photoreactive groups and the determination of the structure of NA photomodification products under the action of these reagents. MI Dobrikov Novosibirsk Institute of Bioorganic Chemistry, Siberian Branch of the Russian Academy of Sciences, prosp. Akad. Lavrent'eva 8, 630090 Novosibirsk, Russian Federation. Fax (7-383) 233 36 76. Tel. (7-383) 233 37 62. E-mail: Dobmi@niboch.nsc.ru Received 13 April 1999 Uspekhi Khimii 68 (11) 1062 ± 1079 (1999); translated by R L Birnova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.963.3 : 577.113.6 967 967 975 978 980 The design of DNA-cleaving reagents which are chemically stable and can be photoactivated is a topical problem. These reagents are helpful tools in the study ofDNAstructure when used as photonucleases, photofootprinting reagents and potential photoreagents in the phototherapy of cancer and various viral infections.11 ± 13 Photosensitised NA modification reactions are commonly referred to as photoinduced reactions occurring at wavelengths greater than the intrinsic absorption of NA.14 These reactions involve three main types,15, 16 viz., photooxidation with radical species or by photoinduced electron transfer (type I photody- namic effect), photooxidation with singlet oxygen (type II photo- dynamic effect) and oxygen-independent free-radical and electrophilic addition (type III photodynamic effect).The classification of photosensitised reactions of NA is problematic and any description of them is ambiguous, for photochemical reactions are numerous, they occur in parallel and in low quantum yields. Their study is complicated by the fact that the same photomodification end products are formed by different mechanisms.7 This review is devoted to the analysis of the most important photosensitised, oxygen-independent free-radical and electro- philic addition reactions. One-photon 17, 18 and two-photon 19, 20 photochemical reactions of NA induced by far UV light (190 ± 300 nm), photooxidation of DNA with singlet oxy- gen 3, 7, 21 and radical species 22 ± 24 as well as oxidation caused by photoinduced electron transfer 11, 25, 26 have been studied in sufficiently great detail and will not be considered in this review.II. Free-radical cycloaddition and electrophilic addition photoreactions of nucleic acids Several types of oxygen-independent photosensitised chemical modifications of NA are known,14,16 e.g., free-radical photo- cycloaddition to the double bonds of bases,27 ± 45 photoinduced alkylation,46 ± 58 arylation,59 ± 63 acylation 27, 64 and other electro- philic addition reactions. Modification of NA with highly electro- philic species formed during photochemical reactions is especially attractive, since the electrophilic species manifest pronounced selectivity with respect to nucleophilic bases, such as guanine and adenine.In turn, photocycloaddition reactions occur with high selectivity at the double bonds of the pyrimidine bases.27 1. Free-radical photocycloaddition Excitation of most ketones and thioketones gives rise to biradical species which initiate radical cycloaddition reactions involving the double bonds of the heterocyclic bases of NA. Psoralen and968 thiopyrimidine derivatives are the most popular and efficient photoreagents of this type. a. Photomodification of nucleic acids with psoralen derivatives Psoralens (7H-furo[3,2-g][1]benzopyran-7-ones) are tricyclic aro- matic compounds, which have a planar structure and can intercalate between base pairs of double-stranded DNA (ds- DNA).They contain two photoreactive double bonds, viz., C(3)=C(4) in the pyrone ring and C(40)=C(50) in the furan ring. R24 4 0 3 5 0 R1 O O O R3 R1, R2, R3=H, Alk. Irradiation of DNA duplexes at 310 ± 420 nm in the presence of psoralens results in covalent binding of the latter to pyrimidine bases of DNA (preferentially, to thymine) at the C(5)=C(6) bond to give cyclobutane adducts (Scheme 1). Photoaddition reaction of psoralens to cytidine occurs 15-fold more slowly than that to thymidine.1 Psoralens are bifunctional photoreagents; they can cross-link both DNA strands, especially when intercalating in the 50-TpA-30 sequence.1, 28 ± 38 50-TpA/30-ApT fragments are present in signal sequences of many genes as potential targets for psoralen-containing oligonucleotides. O Me HN O N O + R1O O R3L 2 R2O O Me O HN N O R1O R3L O 3 R2O O Me O HN N O R1O R3L O R2O 1 L is CH2NH(CH2)2NH-p, R17R5 are fragments of oligonucleotide chains.Using molecular modelling methods, it was shown 1 that psoralen incorporates between the nucleotides of a target in two different ways, viz., between nucleosides Nn+1 and Nn or between nucleosides Nn and Nn71, where Nn is the nucleoside complemen- tary to the 50-terminal nucleoside of the psoralen conjugate. Spatial structure of the reaction centre which is the most favourable for photoaddition, is realised in the former case.The incorporation of psoralen is especially efficient where thymidine residues occupy positions n+1 and n71, and adenosine residues occupy position n of the target. In this case, the reaction occurs Scheme 1 O O hn (419 nm) O R4 T R5 O hn (310 ± 400 nm) O O Me O N HN O O OR5 R4O MI Dobrikov simultaneously in both directions. The yield of the cross-linked product of the type 1 amounts to 95%. In recent years, psoralen-containing oligonucleotide deriva- tives have been used mostly for photomodification of ds-DNA. Studies of Gunther et al.29 have shown that the 16-membered conjugate Pso-CH2NH(CH2)2NH-p-50-TTTTCTTTTCCCCC- CT 2 forms a triplex with DNA regions of HIV-1 containing homopurine ± homopyrimidine sequences.Psoralen intercalates into ds-DNA thus increasing the melting temperature of the triplex from 20 to 29 8C. Irradiation with UV light at wavelengths >310 nm yields mono- and bis-adducts 3 and 1, which is confirmed by their electrophoretic mobility. The addition of the furan ring of psoralen to the C(5)=C(6) bond of thymine next to the 50-end of the homopurine sequence of the DNA-target gives the cyclobutane adduct 3. This is followed by photochemical cycloaddition of the pyrone ring of the adduct 3 to the C(5)=C(6) bond of 30-terminal thymine in the homopyrimidine chain, resulting in the bisadduct 1. The DNA region of HIV-1 contains two sequences which can serve as targets for the conjugate 2. Both sequences are involved in cross-linkage with high efficiency.Besides, at higher exposures and higher concentrations of the oligonucleotide reagent 2, cross- linking occurs in the regions where the 50-terminal 8-membered fragment (T4CT3) of the conjugate 2 binds to ds-DNA. The reason for such modification is that the sequence (pC)6 of the conjugate 2 cannot be fully protonated under physiological conditions, and the overall specificity of the 16-membered conjugate 2 is deter- mined by the fragment (T4CT4). Substitution of (pG)6 for (pC)6 results in photomodification at only two specific sequences.29 The dependence of the length of the oligonucleotide address on the degree of modification of target ds-DNA in vitro was studied.29 It was shown that 10- and 16-membered psoralen- containing oligonucleotide conjugates were effective in sequence- specific photocross-linking of mouse ds-DNA, while an 8-mem- bered oligonucleotide reagent had little effect.Binding of the 10-membered conjugate to ds-DNA occurred rapidly (in less than 10 min). A conjugate of 5-alkoxypsoralen and 50-pTTTTC*C*TC*T- C*C*C*TC*T, where C* is 5-methylcytosine residue introduced for increasing the triplex stability, was used to inhibit the expression of the interleukin 2 a-chain receptor gene.30 The conjugate bound to the promoter homopurine ± homopyrimidine sequence to yield a triplex. The mutant form of the promoter containing a transversion of three nucleotides in the centre of the homopurine ± homopyrimidine sequence was used as a control.The conjugate did not bind to the mutant promoter. Irradiation with UV light (365 nm) was accompanied by photocycloaddition of the psoralen 3,4- and 40,50-double bonds to the C(5)=C(6) bonds of the sequence 50-TpA-30 at the duplex ± triplex junction, resulting in the bisadduct 1. The for- mation of a covalently bound triplex inhibited transcription in vitro, while transcription of the mutant form of the promoter was not inhibited. After in vitro photomodification of DNA, the photocross-linked triplex was transfected into human HeLa cells. The irradiated mutant promoter was transcribed with the same efficiency as the nonirradiated one; the activity of the irradiated promoter was inhibited by 80% (according to the photocross-linkage yield).The formation of photoadducts in a triplex of a bacteriophage l ds-DNA with a conjugate 40-hydroxymethyl-4,50,80-trimethyl- Pso-pAGGAAGGGGG was most extensively studied by Gas- parro et al.31 The quantum yield (j) of the monoadduct 3 was 0.0065, while that of the bisadduct 1 was 5 times as great. The formation of the bisadduct 1 implies that the C(40)=C(50) bond of the furan ring of psoralen reacted first and then the C(3)=C(4) bond of the pyrone ring. Otherwise, the bisadduct is not formed. It was therefore proposed to carry out the irradiation in a stepwise manner, viz., first, with visible light (419 ± 447 nm) and then with UV light (> 310 nm). The efficiency of photomodification upon irradiation with visible light was 400 times lower than with UVSite-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents light, but in this case the monoadduct 3 is formed at the T166 residue (yield 40%).Subsequent UV irradiation gives the bisad- duct 1 (yield 62%) and monoadducts (yields at pyrimidine and purine residues were 9% and 22%, respectively). In this two-step irradiation, the total degree of photomodification of ds-DNA reached 95%. Irradiation with visible light (419 nm, 30 min) of ds-DNA within a complex with a triplex-forming psoralen conjugate 2 yielded the monoadduct 3 in 87% yield. Further irradiation with UV light for 5 min resulted in the bisadduct 1 (in up to 92% yield).31 A conjugate 40-hydroxymethyl-Pso-4,50,8-trimethyl- pAGGAAGGGGG was used for site-specific injury of the supF- gene incorporated into vectors of E.coli phage l32 and virus SV40.33 It was shown that the formation and photocross-linking of the triplex can also occur in vivo.34 Inhibition of gene expression with a conjugate 5-hydroxy-Pso- pTC*TTC*TTC*C*TC*T-C*TT, where C* is 5-methylcytosine, in transfected HeLa cells is reported.38 After UV irradiation (>310 nm) of the triplex in vitro, the amplification of a DNA fragment was inhibited by 95%.Plasmid pCH110 was photo- modified and incorporated into HeLa cells. The activity of the gene of the incorporated b-galactosidase 5 h after transfection was 5%. Penetration of psoralen derivatives through hydrophobic cell membranes and their resistance to nuclease hydrolysis could be enhanced if psoralen was coupled to distamycin-like imidazole (4a) and pyrrole (4b and 5) derivatives.35 ± 37 O O O O HN O NMe 4a,b [X=N (a), CH (b)] O HN N HO OO S O N O P O O OH OH S HN N HO OO O O N O P O O OH OH X HN X O HN N NMe2 O Me O HN N HO OO NH hn O O O P O OH OH 6 S HN N HO OO hn NH O O O P O OH 9 Me O O O MeO Me NH HN O 5 The imidazole analogues of distamycin 4a can deliver psoralen to GC-rich DNAregions, while the pyrrole analogues 4b and 5, to AT-rich DNA regions.The imidazole analogue 4a photomodifies DNA 10 times more efficiently than the original psoralen.The efficiency of photomodification of ds-DNA with the pyrrole analogue 4b is 1000 times greater than that with psoralen and 100 times greater than that with the imidazole analogue.1 These data suggest that psoralen conjugates are promising reagents for photomodification of NA in vivo. High phototoxicity of psoralens as regards modification of ds-DNA is determined by the formation of bisadducts 1. b. Photomodification of nucleic acids with 4-thiopyrimidine derivatives Along with psoralens, thiopyrimidine nucleotides, e.g., 4-thiour- idine (s4U) and 4-thiothymidine (s4T), are used as photoreagents able to undergo free-radical cycloaddition. The presence of a sulfur atom instead of an oxygen atom at position 4 causes a shift of the absorption spectrum towards the long-wave region (l= 330 nm) and the appearance of new photochemical properties.39, 40 Irradiation of thiopyrimidines with UV light (330 nm) leads to the population of the excited state S2 which is rapidly (within 5 ps) and in high quantum yield (j=0.9) converted into state T1.Photolysis of d(s4U) in water in the presence of oxygen yields dU.40 Thiopyrimidine nucleotides can form covalent adducts with all natural nucleic bases, but the quantum yield of the photo- reaction with pyrimidine bases is higher than that with purines. Irradiation of the dinucleotide dTp(s4U) (6) gives two stable products 7 and 8 (quantum yield 0.02) (Scheme 2). It is assumed HN HO S OO NH O N O O P O O OH 7 HN HO OO O + NH O N O O P O O OH OH 10 969 NMe2 HN N O NH Me N O Me Scheme 2 O SH N N N O O OH 8 SN H O NH N O O OH 11970 U G G A AGUX GAAA A30 Figure 1.The structure of a ribozyme complex (bold-type) containing an RNA fragment. that the [2+2]-cycloaddition of thiopyrimidine involving the excited C=S bond to the C(5)=C(6) bond of a pyrimidine results in the adduct 7 and subsequent chemical reactions yield the final product 8.40 Irradiation of the dinucleotide d(s4U)pT (9) afforded the adducts 10 and 11 (Scheme 2). In all probability, they are formed due to hydrogen abstraction from the methyl group of thymine by photoexcited thiouridine followed by recombination of radical species.40 A significant advantage of thiopyrimidines as photoreagents is that the van der Waals radius of the sulfur atom is by only 0.45 A larger than that of the oxygen atom.This does not affect the structure of oligo- and polynucleotides with incorporated thiopyrimidine residues.39 The photochemical properties of thio- pyrimidines were most completely employed in the study of the tertiary structure and mechanism of action of ribozymes (Fig. 1).39 ± 45 Thiouridine residues were introduced into the RNA substrate at positions X. Irradiation of the ribozyme complex with RNA was accompanied by the formation of cross- links (indicated by arrows in Fig. 1). The photoaffinity modifica- tion data suggest that several alternative conformations appear in the ribozyme complex with RNA in different stages of a catalytic cycle of RNA hydrolysis.40 ± 42 2.Photoinduced alkylation of DNA Numerous photochemical methods for the generation of electro- philic species that alkylate DNA have been described.27, 46 ± 58 These methods involve the preparation of azacycloheptatetraene via an intermediate singlet nitrene by photodecomposition of aromatic azides,46 ± 52 generation of quinolylmethylcarbocations by heterolytic photodecomposition of quinolylmethylisothiouro- nium salts,53 preparation of quinonemethide by photoenolisation of 5- and 2-methylnaphthoquinone derivatives 54 ± 57 and gener- ation of oxonium cations from monothioacetals by photoinduced electron transfer.58 30 X X A 50 X AC C A G U a.Photomodification of nucleic acids by aromatic azides Oligonucleotide derivatives of aromatic azides are widely used in photomodification of DNA46, 47, 50 ± 52 and RNA48, 49 owing to their high photoreactivity and accessibility. Conformational changes of a p-azidophenacyl 50-thiophos- phate derivative of ribozyme L-21 Sca I were studied in different stages of binding and catalytic cleavage of an RNA-substrate.48, 49 Irradiation of the photoreactive ribozyme derivative with UV light (312 nm) in the absence of the RNA-substrate gives covalent adducts at the A88\A89\A90 residues (yield 12%). The photo- modification affected the A114\A115 residues of ribozyme (yield 34%) within the complex with the fragment GGCCCU- CUAAAAA where ribozyme hydrolysed the phosphodiester bond between the uridine and adenine residues; in the absence of Mg2+ cations, the C102\A103\A104 residues were photomodified in 38% yield.48 The data obtained made it possible to establish the position of the ribozyme internal guide sequence (IGS) containing the p-azidophenacyl residue in different stages of the catalytic cycle of RNA hydrolysis.48, 49 X50 Ethidium derivatives 12 and 13 proved to be the most efficient photoreagents for DNA modification.50, 51 In all probability, this is associated with the ability of these residues to intercalate into DNA and with the photochemical properties of aromatic azides.The photolysis products of the azides 12 and 13 in water have been studied.51 This reaction yields intermediate electrophilic species, viz., the azacycloheptatetraene 14 and the aziridine 15; which add water to give the adducts 16 and 17.In both cases, the triplet nitrenes 18 and 19 were formed, which underwent nonproductive decomposition as regards NA photomodification, viz., reduction to the amines 20 and 21, respectively. nucleophiles should give arylation products of the type 17; however, these products have not been obtained and this azide is considered for convenience of comparison. LN3 Formally, photolysis of the azide 13 in the presence of N3 hn j =1 +NEt 12 Ph N H2O +NEt L Ph 14 N H2O +NEt L 18 PhL hn j=0.5 +NEt 13 Ph MI Dobrikov ONH +NEt L Ph 16 NH2 +NEt L Ph 20Site-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents L L H2O + + NEt NEt HO N Ph 15 Ph 17 NH2 L L H2O + + NEt NEt N H2N 19 Ph Ph 21 L=NHCO(CH2)2NH2 .Heptanucleotide pCCAAACA conjugates with the azides 12 and 13 linked through 50-phosphate were used for site-specific photomodification of the 32P-labelled complementary octanu- cleotide 50-*pTGTTTGGC.51 The introduction of azidoethidium residues strongly stabilises the complementary complexes by increasing their melting temperature by 20 ± 30 8C. 50-*pTGTTTGGC . . . . . . . ACAAACCp-R Irradiation of these duplexes with UV light (320 nm) results in covalent adducts in 40% and 60% yields, respectively.After treatment with piperidine which results in the cleavage of the DNA chain at the modified bases, the total degree of the octanucleotide photomodification increased up to 70% and 80%, respectively. The results of piperidine-catalysed cleavage led the authors to conclude that modification predominantly affected the C8 residue and, to a lesser degree, the G7 residue (indicated by arrows). The formation of up to 10 different adducts (many of which are unstable) was established by HPLC, but the attempts to isolate these adducts and to determine their structure were unsuccessful. The oligonucleotides p(TC)6Tp containing the azides 12 and 13 at their terminal phosphates were used for photomodification of ds-DNA in the triplex 22.50 12 18 10 20 22 8 4 2 16 14 6 30-TGCTTCTCTCTCTCTCTGTGGA .. . . . . . . . . . . . . . . . . . . . . 50-ACGAAGAGAGAGAGAGACACCT . . . . . . . . . . . . . 50-R-pTC TCTCTCTCT CT(p-R) 22 Irradiation of the triplex 22 (320 ± 420 nm) containing the conjugate 50-(12)p(TC)6T as a photoreagent resulted in sponta- neous cleavage of the purine-rich chain at the residue G3 (the degree of photomodification was 15%) as well as in the formation of covalent adducts with this residue (yield 30%). The pyrimidine- rich chain was photomodified to a far lesser degree; this was accompanied by the formation of covalent adducts with the residue T4 (yield 10%). With the 30-(12)p(TC)6T derivative as a photoreagent, the total degree of photomodification of both chains did not exceed 10%.Irradiation of the triplex 22 with the 50-(13)p(TC)6T reagent resulted in highly selective photomodification of the pyrimidine- rich chain. No direct cleavage was observed in this case; the covalent adducts (yield 35%) and the product of modification of the residue T4 were detected using treatment with piperidine. 971 b. Photomodification of DNA with benzylisothiouronium salts A study of photolysis of a series of benzyl compounds of the type ArCH2X capable of generating carbocations revealed that the quinolyl-2(4)-methylisothiouronium salts 23 and 24 manifested the highest photoreactivity. Their irradiation with UV light (350 nm) in water gave the carbocations ArCHá2 (yield 90%, quantum yield 0.09).53 Cl7 CH2SC(NH2) NHá2 MeO Ph N 23 N Cl¡ CH2SC(NH2) NHá2 24 hn (ArCH2X)* ArCH2X [ArCHá2 X¡] ArCH2(dG+) dG 7X¡ ArCHá2 +X7 e¡ ArCH2 dG is deoxyguanosine.Photomodification of ds-DNA with the salts 23 and 24 resulted in insignificant cleavage of the target independent of its sequence. Most probably, the reaction occurs as hydrogen abstraction from deoxyribose by a benzyl radical (ArCH2. ) formed simultaneously with subsequent oxygen-dependent cleav- age of the DNA-target. After treatment with piperidine, the degree of DNA modification increased 10-fold due to G-specific cleavage. It is assumed that quinoline residues intercalate into DNA in such a way that the positively charged isothiouronium residues of the reagents 23 and 24 are located on the duplex surface.Alkylation of guanine residues with a photoinduced quinolylmethylcarbocation at position N(7) is considered to be the most probable mechanism of G-specific modification. The quantum yield of photoalkylation product formation is rather low [j=(5 ±6)61076]; the yield of adducts with respect to the photoreagent does not exceed 1%.53 Because of low efficiency of photoalkylation, these photoreagents have not found wide appli- cation. c. Photoalkylation of nucleic acids with methylnaphthoquinones Modification of single-stranded DNA (ss-DNA) with an oligonu- cleotide derivative of 5-methyl-1,4-naphthoquinone upon irradi- ation with UV light (>345 nm) is the best studied example of photoinduced alkylation of NA.54 ± 56 Unlike photomodification with the isothiouronium salts 23 and 24, the products of oxidative cleavage of the DNA-target were not found and the alkylation product was exclusively formed in *20% yield. Neither piper- idine (DNA cleavage at the sites of bases alkylated at the nitrogen atoms of a heterocycle), nor sodium borohydride [cleavage at the sites of pyrimidine bases with a saturated C(5)7C(6) bond] effected cleavage of the DNA-target.Pre-irradiation of the photoreagent resulted in its complete inactivation due to the photodegradation of the substituted naphthoquinone.54 It was shown that photogenerated quinonemethides alkylate G and C residues at the exocyclic amino groups.Such adducts are more resistant to alkaline hydrolysis of DNA than the ordinary alkylation products.56 The efficiency of photoalkylation of DNA with oligonucleo- tide derivatives of 3-(1,4-naphthoquinonyl)thiopropionic acid with substituents at positions 2, 5 and 6 has been studied.55, 56 5-Methyl-substituted naphthoquinones proved to be the most photoreactive agents, the degree of alkylation of ss-DNA regions reached 27%. The yields of products of DNA alkylation with 2- and 6-methylnaphthoquinones did not exceed 6%± 8%.972 The mechanism of photoalkylation of DNA with 5-methyl- naphthoquinone (25) in aqueous solutions has not been studied until very recently.57 Upon photoexcitation, the naphthoquinone 25 is converted into 4-hydroxy-5-methylidenenaphthalen-1(5H)- one (26a), which can be deprotonated to give the naphthoqui- none-5-carbanion (26b) or protonated to give the corresponding carbocation 26c.The reaction of the carbocation 26c with nucleophiles (HX) is the limiting step of photoinduced alkylation. CH3 O CH2 OH hn [25]* 26a 25 O O 7CH2 O 7H+ H+ 26b O + OH OH XCH2 CH2 H+ HX 7H+ OH 26c OH Photolysis under physiological conditions results predomi- nantly in the carbanion 26b (pKa=6.5); the corresponding value for the carbocation 26c is 1.1. This fact is the main reason for the low efficiency of DNA photoalkylation with methylnaphthoqui- nones. d. Photomodification of nucleic acids with monothioacetals Irradiation of 1,8-naphthalimidoacetaldehyde monothioacetals with UV light (> 330 nm) generates S-centred radical cations due to intramolecular electron transfer to the photoexcited naphthalimide.The radical cation dissociates into a carbocation and an alkylthio radical, which is rapidly reduced by the naphthalimide radical anion.58 R R7 R7 +SMe SMe hn + +SMe OMe OMe OMe RCH2 MeO R NH2 dA H+ N + ... +S7Me N OMe N N ON R= , dA is deoxyadenosine. O The carbocation reacts with nucleophiles (e.g., the reaction with MeOH in the absence of oxygen occurs with a 98% yield and a quantum yield of 0.038) to give alkylation products. 1,8-Naphthalimidoacetaldehyde monothioacetal was irradi- ated with UV light (366 nm) in the presence of a triplet sensitiser (acetophenone); in this case, 99% of the light was absorbed by acetophenone. It was thus demonstrated that the electron transfer occurs via the triplet state of naphthalimide.Irradiation of supercoiled DNA of pBR 322 with UV light (365 nm) in the presence of this monothioacetal resulted in its conversion into cyclic DNA after heating. Adenine alkylated at position N(7) was the main DNA photomodification product.58 Thus, several types of electrophilic photoreagents are cur- rently known which can be used for photoalkylation of NA upon MI Dobrikov irradiation with proximal UV light (320 ± 400 nm). However, low efficiency of these reagents and the lack of photoreactivity in the visible region of the spectrum restricts their application for the in vivo photomodification of NA.3. Photoarylation of DNA Ruthenium(II) polypyridyl complexes photosensitise the forma- tion of single-strand breaks in ds-DNA in low quantum yields [j=(1 ±7)61076].59, 60 The mechanism of cleavage was inves- tigated with short 50-labelled oligonucleotides, which were irradi- ated with visible light (436 nm) together with a Ru(TAP)3 complex, where TAP is 1,4,5,8-tetraazaphenanthrene. It was found that the main photosensitised reaction is not the cleavage of oligonucleotide chains, but rather the formation of covalent adducts of the Ru(II) complex with the oligonucleotide.59, 61 The yield of adducts in the Ru(bipy)nL37n series, where L=TAP or HAT, bipy is 2,20-bipyridyl and HAT is 1,4,5,8,9,12-hexaazatri- phenylene was rather high for n=0,1, while no covalent adducts were formed for n=2,3.With an increase in n, the photochemical properties of Ru(II) and Os(II) complexes change, which is manifested in a bath- ochromic shift of absorption and fluorescence maxima and a decrease in the reducing potential (Ered) of the excited state (Table 1). Covalent adducts with the DNA-target are formed only in the case of Ru(II) and Os(II) complexes the E red values of which exceed the redox potential of guanosine monophosphate (Eox=0.92 V). Table 1. Photochemical properties of Ru(II) and Os(II) complexes: the absorption (lmax) and fluorescence maxima (lfl); the values of the reduction potential (Ered) of photoexcited triplet states and the possibility of formation of covalent adducts with DNA.61 Adducts Complexes lmax /nm lfl /nm Ered /V 596 661 742 608 649 714 800 820 440 472 484 437 465 484 572 564 ++7++7+++ 1.46 1.12 0.83 1.32 1.06 0.86 1.03 1.02 7 7 1.11 Ru(HAT)3 Ru(bipy)(HAT)2 Ru(bipy)2(HAT) Ru(TAP)3 Ru(bipy)(TAP)2 Ru(bipy)2(TAP) [Ru(bipy)2]2HAT [Ru(phen)2]2HAT Os(TAP)3 bimetallic The and [Ru(bipy)2]2HAT complexes [Ru(phen)2]2HAT (where phen is o-phenanthroline) attract sub- stantial interest as electrophilic photoreagents, since they also form covalent adducts with DNA; their absorption maxima show bathochromic shifts by about 100 nm relative to monometallic complexes.The covalent adduct formed by photomodification of ds-DNA with the Ru(TAP)3 complex was isolated and charac- terised. The exocyclic 2-amino group of guanine was covalently bound to the C(2) position of TAP.Photomodification of DNA with Ru(bpz)3 (bpz is 2,20-bipyridazinyl), Ru(TAP)2bipy,61 [Ru(phen)2]2HAT60 and Os(TAP)3 59 complexes gave similar photoadducts. N NH NH HN N N N N (TAP)2Ru2+ N O Although the Ru(II) polypyridyl complexes intercalate into ds-DNA through the major groove, the formation of covalent adducts takes place in the minor groove. The mechanism of photosensitised arylation of DNA with Ru(II) complexes has not been finally elucidated; however, it isSite-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents believed that the photoinduced electron transfer from the guanine residue to the excited ligand (TAP*) is followed by proton transfer from the radical cation G+.to the radical anion TAP7. as well as by recombination of radicals and oxidative rearomatisation of the TAPH2 residue.59, 61 Oligonucleotide conjugates of Ru(II) polypyridyl complexes were used to study photoinduced electron transfer,62, 63 but experimental support for complementarily addressed photomodi- fication of DNA by these derivatives is lacking. It should be noted that metal polypyridyl complexes open up new opportunities for the design of photoreagents that can be employed for the in vivo modification of NA. This is due to the fact that the properties of such complexes can be varied within a very wide range by changing the central metal atom and the ligands. 4.Photoacylation of DNA It is known that photoinduced dissociation of dibenzoyldiazo- methane (27) results in dibenzoylcarbene (28), which is converted into the electrophilic ketene 29 capable of reacting with nucleo- philes following the Wolff rearrangement.27, 64 Thus its photolysis (366 nm) in the presence of 30,50-di-O-(tert-butyldiphenylsilyl)-20- deoxyguanosine gives a guanosine derivative 30 with an acylated exocyclic amino group.27 O O O O hn Ph Ph Ph Ph 27 N2 28 O O N O dG O NH O C Ph N Ph N Ph NH 29 dr Ph 30 dr is deoxyribose. Irradiation of dibenzoyldiazomethane (27) together with ss-DNA in the presence of the singlet oxygen trap, NaN3, gave covalent adducts.Treatment of the latter with piperidine is accompanied by G-specific cleavage owing to the reaction with the ketene 29 and insignificant nonspecific cleavage due to hydrogen abstraction from deoxyribose by the carbene 28.27 Site-specific photomodification of a 25-membered fragment of ss-DNA with 8- and 15-membered complementary oligonu- cleotide conjugates having a substituent (R) at the 50-terminal phosphate in the presence of NaN3 results exclusively in covalent adducts in 32% and 47% yields, respectively. 2 20 22 4 8 10 12 14 16 6 50-AGTGCCACCTGACGTCTGCTCTCTC . . . . . . . . 30-GGAC TGCAp-R 10 20 12 6 4 2 22 8 50-AGTGCCACCTGACGTCTGCTCTCTC .. . . . . . . . . . . . . . 30-TCACGGTGGACTGCAp-R R=4-(PhC(O)C(=N2)CO)C6H4CH2OC(O)NH(CH2)6NH. In both cases, treatment of photomodification products with piperidine caused cleavage of the DNA-target at the residue G18 located in the single-stranded fragment of the duplexes; in the case of the 15-membered reagent, additional cleavage occurred at the residue G14 (indicated by arrows). The greater part of the adducts is not cleaved with piperidine. The total degree of modification after treatment with piperidine increased up to 33% and 56%, respectively. The formation of piperidine-labile adducts is due to the acylation at the N(7) position of guanine, whereas that of piperidine-stable adducts, by the acylation at the exocyclic amino group.27 973 Thus, substituted diazo ketones are promising reagents for G-specific photoacylation of NA.5. Other electrophilic addition reactions A series of 8-alkoxy-5-azidopsoralens were synthesised with the aim of integrating the photochemical properties of aromatic azides and psoralens and investigating their ability in the photo- modification of NA. Irradiation of azidopsoralens in acetonitrile with UV light (351 nm) affords a triplet nitrene and an azacyclo- tetraene, whereas irradiation in water yields exclusively a triplet nitrene. The reactivity of these species is insufficient for efficient photomodification of DNA. Being a triplet sensitiser, psoralen favours the formation of a triplet nitrene. Such a detrimental effect of the psoralen residue on the photochemistry of the azido group is the reason for the inability of pBR322 DNA to undergo cleavage with 8-alkoxy-5-azidopsoralens upon UV irradiation.65 Oligonucleotide reagents containing aromatic azido groups were also fairly inefficient in the photomodification of NA due to numerous side reactions.In the case of aromatic azides, the reaction does not occur via the highly reactive singlet nitrene 31 but rather via the much less reactive azacycloheptatetraene 32 formed from the nitrene 31 by rapid intramolecular ring expan- sion. Besides, nitro-, cyano-, formyl-, acyl- and, particularly, amino-substituents in aromatic azides accelerate intersystem crossing into a more stable but less reactive triplet nitrene 33 upon irradiation.52 The nitrene 33 is usually consumed in radical dimerisation reactions to give azo derivatives or is reduced to the corresponding aniline without formation of NA modification products (Scheme 3).Scheme 3 + Py N N3 R N N 7 hn 7N N + Py N R R R R 32 31 N NH2 R0H R N N R R R33 Perfluoroaromatic azides represent a more promising class of photoreagents, since their photodissociation predominantly results in the singlet nitrene 31.52, 66 ± 68 The presence of fluorine atoms in o-positions relative to the azido group decreases the rate of azacycloheptatetraene formation 1700-fold.66 In addition, the acceptor fluorine atoms increase the electrophilicity of the singlet nitrene 10-fold 67 and reactivities of the triplet nitrene and the azacycloheptatetraene, 1000 and 10 000-fold, respectively,68 in comparison with nonfluorinated derivatives.Dobrikov et al.46, 47 studied the effect of the aryl azide structure on the efficiency of site-specific photomodification of DNA by comparing the reactivities of oligonucleotide derivatives of 4-azido-, 5-azido-2-nitro- and 4-azidotetrafluorobenzamides. Upon irradiation with UV light, these compounds generate different species, viz., azacycloheptatetraene 32, the triplet (33) and the singlet (31) nitrenes, respectively; the yields of covalent adducts in site-specific photomodification were 5%, 15% and 70%, respectively.67 The reactivity of the species 32 and 33 is insufficient for modification of NA.The specificity of photo- modification of DNA-targets containing T, C, A or G residues opposite to the photoreagent was studied with an example of an974 oligonucleotide reagent containing a perfluoroaryl azide residue 34 coupled through 50-phosphate. Treatment of modification products with piperidine showed that the degree of modification of these residues was *50%, the largest degree of modification being 40%, 45%, 45% and 65%, respectively.46, 47, 69 The reac- tions of the singlet nitrene 31 with the nitrogen atoms of the pyridine type of the heterocycles result in N-amination (Scheme 3). Photomodification of ss-DNA is most probably accompanied by electrophilic addition at position N(7) of gua- nine. The use of two oligonucleotide reagents complementary to the neighbouring regions of the DNA-target and containing two perfluoroaryl azide residues 34 at the contact site made it possible to increase the yield of covalent photomodification adducts to 80%.70 A highly efficient site-specific photomodification of 27-mem- bered fragments of ss- and ds-DNA with a 4-azidotetrafluoro- benzamide derivative of p(T)16 has been carried out.71, 72 CONH(CH2)nNH2 N3 F 30-(T)16p-34 .50-*pGCGCACG(A)16GTCG 34 (n=2, 3) . 50-34-p(T)16 35 30-CGCGTGC(T)16CAGCp* . . . . . . . . . . . . 50-*pGCGCACG(A)16GTCG . 50-34-p(T)16 36 Photomodification of ss-DNA within the triplex 35 occurred mainly at the bases G7 and G24 (bold-typed); the yields of covalent adducts amounted to 77%.Residues G7 of A-rich and G-22 of T-rich strands (the yields of covalent adducts were 37% and 16%, respectively) were the main sites of photomodification of ds-DNA within the triplet 36. After piperidine treatment, the total degree of modification of the triplex 36 increased up to 50% in A-rich and up to 42% in T-rich strands due to noncovalent modification at pyrimidine bases.71, 72 The effects of p-substituents in perfluoroaromatic azides on the efficiency of site-specific photomodification of ss-DNA was also studied. The increase in the conjugation chain in substituted perfluoroaryl azides accelerates the intersystem crossing of the singlet nitrene into the triplet one and decreases the efficiency of DNA photomodification. Studies with pyridines as the chemical trap for the singlet nitrene 31 revealed that the yield of covalent adducts with DNA correlates with the yield of pyridine ylides in the photolysis of perfluoroaryl azides in 6 M pyridine.It was assumed that the electrophilic addition of the singlet nitrene to the nitrogen atoms of the pyridine type in NA bases is the most feasible mechanism of covalent adduct formation.73 The efficiency of site-specific photomodification (irradiation with UV light, l=303 ± 370 nm, irradiation power 0.5 mW cm72, 5 min, 4 8C, pH 7.0) of ss-RNA and ss-DNA fragments (5 mmol litre71) with oligonucleotide conjugates of perfluoroaryl azide 34 (50 mmol litre71) was compared by Levina et al.74 Yield of covalent adducts for targets (%) Reagent 50-UGUUUGGC 50-d(TGTTTGGC) 50-34-pd(AAACA) 50-34-pd(CAAACA) 50-34-pd(CCAAACA) *5 30 55 *5 65 75 The efficiency of photomodification increases with an increase in the length of the oligonucleotide address.Photomodification of RNA occurs less efficiently than that of DNA, which most probably depends on the structures and stabilities of duplexes. Nevertheless, 6 ± 7-membered ribonucleotide conjugates carrying a perfluoroaryl azide residue 34 at the 50-terminal phosphate were MI Dobrikov used to study the decoding centre of human 80S ribosomes.75 Irradiation of a ribosome ±tRNA ± photoreactive mRNA ana- logue complex with UV light (270 ± 380 nm) resulted in binding of the analogue to the 40S subunit (45% ± 55%) and 18S sRNA (8% ± 22%).Only one base, viz., the residue G1207, underwent photomodification within sRNA. Kazantsev et al.76 studied the kinetics of site-specific photo- modification of 26-membered ss-DNA with oligonucleotide con- jugates of perfluoroaryl azide 34. Irradiation with UV light (>300 nm) resulted in the formation of covalent adducts; upon treatment with piperidine, cleavage (predominantly, at residue G19 and, to a lesser degree, at residues G17, A18 and G20) (indicated by arrows) occurred. 25 15 5 10 50-pTTGCCTTGAATGGGAAGAGGGTCATT 30-CTTACCCTTCTp-34 30-ACCC T TCTp-34 30-ACCC T Tp-34 The kinetic data shown in Table 2 suggest that the efficiency of photomodification with 50-p-azidotetrafluorobenzamide con- jugates is not very high.The maximum degree of photomodifica- tion with a 11-membered conjugate was higher than that with 6- and 8-membered ones due to the higher value of the constant Kx. Irradiation of the conjugates in the absence of a DNA target resulted in photodeactivation and an approximately 200-fold decrease in affinity for the target.76 Table 2. The association constants (Kx) of the conjugates with the target, the efficiency of photomodification within the duplexes (gef) and max- imum degrees of photomodifications (PZ) at 37 8C. Photoreagent PZ (%) Kx /mol litre71 gef 0.2 0.33 0.27 70 55 40 8.46106 1.16105 16105 CTTACCCTTCTp-34 ACCCTTCTp-34 ACCCTTp-34 Note. The concentrations of a DNA-target and the conjugates are 10 nmol litre71 and 60 mmol litre71, respectively (see Ref.76). Site-specific photomodification of a ss-DNA fragment with an oligonucleotide conjugate containing the azide 34 was carried out in order to elucidate the possible reasons for such photodeactiva- tion.77 50-*pGTGTGA 30-ACACACp-34 Irradiation of a labelled reagent in the absence of the target also resulted in a partial loss of its affinity for the target; treatment with piperidine resulted in cleavage at all the bases of the address, the total degree of cleavage was 44%. Irradiation with an excess of a target and subsequent hydrolysis revealed that the modification of the address occurred at the residue A4 (yield 29%), i.e., the formation of the duplex did not prevent the modification of the address.With a 5000-fold excess of the conjugate, photomodifi- cation affected the residue G3 of the target localised in the double- stranded fragment. The maximum yield of photomodification reaches 80%; geff is close to unity.77 Site-specific photomodification of a 37-membered DNA fragment with an oligonucleotide reagent carrying a perfluoro- aryl azide residue at 30-phosphate in complexes with human immunodeficiency virus reverse transcriptase (HIV RT)78, 79 or with DNA-polymerase a has been carried out.80Site-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents 35 5 30 15 10 25 50-(d)*pGGTTAAATAAAATAGTAAGAATGTATAGCCCCTACCA . . . . . . . . . . . . . . . . . . . 30-34-CTTACATATCGGGGATGGT Irradiation of these complexes with UV light (>300 nm) has led to cross-linking of the primer to matrices with 60%78, 79 and 70%80 efficiency. Treatment with piperidine resulted in cleavage of the target at the residue G19 which is located opposite the photoreagent incorporated into the primer. Thus, by virtue of high reactivity of the species formed upon irradiation of perfluorinated azides, these have found wide application in photomodification of NA.The main disadvantage of these photoreagents is the lack of noticeable photoreactivity in the visible region of the spectrum, which restricts their application in complex biological systems. III. Sensitised photomodification of DNA with binary systems 1. Photomodification of DNA with two-component (sensitiser ± photoreagent) systems Irradiation of DNA containing 5-bromo- or 5-iododeoxyuridine residues with UV light (302 ± 313 nm) induces numerous photo- chemical reactions, viz., modification of bases, cross-linking, single- and double-strand breaks, etc.(Scheme 4).23, 81 ± 88 The mechanism of formation of single-strand breaks initiated by photolysis of halogenated pyrimidines was studied in detail for oligonucleotides 37 containing 50-ABrU sequences.23, 81, 82 Irradi- ation with UV light (302 nm) resulted in photoinduced electron transfer from adenine to the excited triplet state of bromouridine 3(BrU)* with the formation of a radical cation of adenine Ade+., and a radical anion of uridine (compound 38).This is followed by elimination of Br7 from the radical anion; the 5-uracil radical formed abstracts hydrogen from the adjacent C(10) atom to give the carbocation 39.81 Addition of water to the C(10) atom of the carbocation 39 results in the 10-hydroxy derivative 40. Apurinisa- tion of compound 40 and subsequent reactions bring about cleavage of the DNA chain. The structure of stable DNA cleavage products has been considered in a review of Pratviel et al.81 Similar reactions take place in self-complementary duplexes of the type 50-d(GCABrUGC)2 containing 50-ABrU sequences.83 Irradiation (313 nm) of DNA containing 5-dIU fragments results in alkali-dependent single-strand cleavage with the effi- ciency 0.5; its mechanism is somewhat different.84 The photo- reaction occurs via the homolytic cleavage of the C(5)7I bond of uridine.The 5-uracyl radical formed competitively abstracts hydrogen atoms from the C(10) or C(20) atoms of deoxyribose on its 50-side.81 The efficiency of analogous cleavage induced by irradiation with UV light (l=365 nm) in the presence of Hoechst Ade R1O R1O O O Br HN hn O N O P O O P O OO OH OR2 37 Ade R1O O O + HN H2O O N 7H+ O P O OO OH 39 OR2 975 33258 bisbenzoimidazole 41 reaches 0.9. Under these conditions, the initial rate of sensitised photomodification is 4 ± 5 times greater than the rate of photomodification in the absence of bisbenzoimidazole. Binding of compound 41 to DNA occurs in the minor groove, where it strongly absorbs UV light (365 nm, 23 000 mol71 litre cm71) and undergoes excitation.The electron transfer from bisbenzoimidazole 41 to 5-dIU induces elimination of I7 and the formation of an uracil radical; reactions of the latter bring about photomodification of DNA.84 Similar results were obtained in the photo-induced cleavage of DNA containing 5-dBrUfragments sensitised to long-wave UV light with bisbenzo- imidazole 41.85 OH MeN N N N NH NH 41 It is believed that bisbenzoimidazole dyes bind preferentially to (AT)4- and (AT3)A(5-BrU) sequences of DNA86 and thus ensure high efficiency of electron transfer to the adjacent residues of halogenated pyrimidines. In cells cultured on 5-dBrU-containing media, one 5-dBrU residue per *200 nucleotides is incorporated into DNA.Irradi- ation of cells with UV light (365 nm) in the presence of bisbenzoi- midazole 41 induces the formation of protein ± nucleic acid cross- links as well as single- and double-strand breaks of DNA and modification of bases,87, 88 resulting in multiple aberrations of chromosomes and cell death. The death of cells irradiated in the presence of the photoreagent 5-dBrU and the sensitiser 41 occurs 30 000 times more efficiently than in the case of UV-irradiated native cells and 20 times more efficiently than in the case of 5-dBrU-containing cells irradiated in the absence of the sensi- tiser.87 The use of two-component sensitiser ± photoreagent systems has made it possible to combine high reactivity of the photore- agent with spectral-luminescent properties of the sensitiser.2. Sensitised photomodification of DNA with binary systems of oligonucleotide reagents For the majority of photoreactive derivatives, light absorption and dark reactions within a complementary complex with a DNA- target and in solution occur at the same rate. At best, this results in nonproductive decomposition of the photoreagent in solution and triggers undesirable side reactions. However, the specificity of oligonucleotide derivatives is insufficient for modification of nucleic acids in vivo, since under physiological conditions (37 8C) stable complexes are formed at Scheme 4 Ade+ O7 O Br HN O 7Br7 N OO OH OR2 38 Ade R1O O O HN OH O N O P O OO OH 40 OR2976 the oligonucleotide lengths of 12 ± 13 bases.Oligonucleotides having the lengths of 18 ± 20 bases, which is necessary for highly specific recognition of unique nucleotide sequences within the DNA structure, can form numerous imperfect complexes with incompletely complementary sequences.89 To overcome these drawbacks, the separation of light absorp- tion and photomodification was suggested. For this purpose, binary systems of the oligonucleotide conjugates 42 complemen- tary to the adjacent sequences of a DNA-target and containing chemically inert groups, viz., a sensitiser (S) and a photoreagent (R), have been developed. The groups S and R being brought together within the structure of the complementary complex form a photoreactive centre which can be activated by irradiation with long-waveUVor visible light due to a non-radiative transfer of the electron excitation energy from the sensitiser to the photoreagent.The thus excited photoreagent induces photomodification of only that DNA-target which is an integral part of the complementary complex.90 ± 100 The lengths of both nucleotide addresses (both were decanucleotides) were chosen so that the melting temper- ature of the duplexes was *37 8C; such short oligonucleotides cannot form imperfect duplexes with incomplete complementary sequences under physiological conditions.89 . . . . . . . . S R 42Me F CH2C(O)NH(CH2)2NH2 44 CH2NH2 46 F CH NNHC(CH2)2NH2 N3 O 48 Pyren-1-yl-methylamines (43a ± c),89, 94 ± 96 N-(5-fluoro-12- methylbenzo[a]anthracen-7-ylacetyl)ethylenediamine (44),90 ± 93 anthracen-9-ylmethylamine (45) 98 and perylen-1-ylmethylamine (46) 97, 99, 101, 102 derivatives were used as sensitisers, whereas perfluoroaromatic azides (p-azidotetrafluorobenzamides 34,94, 95 O-(4-aminobutyl)-p-azidotetrafluorobenzaldoxime (47) 96 and N-(4-azidotetrafluorobenzylidene)-N0-(3-aminopropionyl)hydr- azine (48) 89 ± 93, 101, 102 were used as photoreagents.Several sensitisation mechanisms, e.g., singlet ± singlet energy transfer from a photoexcited sensitiser to a photoreagent,89 ± 94, 98 a two-quantum triplet ± triplet energy transfer 95, 96 and photo- induced electron transfer from a photoreagent to an excited sensitiser,97, 99, 101, 102 have been suggested.a. Singlet ± singlet energy transfer Singlet ± singlet energy transfer occurs provided the energy of the 0,0-transition between the lowest vibrational levels of the ground and the first electronically excited states in the sensitiser molecule is greater than the energy of the corresponding transition in the photoreagentRmolecule and the emission spectrum of the excited sensitiser (S*) overlaps with the absorption spectrum of the photoreagent R.103 43a: X =H 43b: X=(CH2)3NH2 43c: X =(CH2)5NH2 CH2NHX CH2NH2 45F CH NO(CH2)4NH2 N3 47 MI Dobrikov The energy of the 0,0-transition in aromatic azides is *70 kcal mol71 and their weak np*-absorption is observed at 450 nm.104 Derivatives of polyaromatic hydrocarbons, viz., pyr- ene 43, benzoanthracene 44 and anthracene 45, have high quantum yields of fluorescence and low Stokes shifts (i.e., the minimum energy losses in photoexcitation and subsequent fluo- rescence).The energies of 0,0-transitions in unsubstituted pyrene, benzoanthracene and anthracene are equal to 76,104 72 105 and 73 kcal mol71 (see Ref. 104) and exceed those of 0,0-transitions in the azides 34, 47 and 48. Thus, the fluorescence spectra of compounds 43 ± 45 overlap with the absorption spectra of these azides.91 The efficiency of the energy transfer strongly depends on the distance between the sensitiser and the photoreagent and at commonly used micromolar concentrations of oligonucleotide derivatives the energy transfer from the sensitiser to the photore- agent does not occur in solution.UV irradiation (365 ± 390 nm) of ss-DNA complexes with an oligonucleotide conjugate containing a photoreagent in the absence (complex 49) and in the presence (complex 50) of a sensitiser results in covalent adducts. The azide 34 was used as a photoreagent (R), while the pyrene derivatives 43a ± c were used as sensitisers (S). Treatment of reaction products with piperidine showed that direct (within the complex 49) and sensitised (within the complex 50) photomodification affects the residue G11.89 10 5 20 15 50-*pATCTTTAACTGATGAACTTC T . . . . . . . . . . . . . . . . . . . . AGAAATTGACTACTTGAAGA R 49 15 10 5 20 50-*pATCTTTAACTGATGAACTTCT .. . . . . . . . . . . . . . . . . . . AGAAATTGACTACTTGAAGA R S50 The initial rate and efficiency of direct (within the complex 49) and sensitised (within the complex 50) photomodification of the target depended on the localisation of the photoreagent and the sensitiser on the 30- or 50-phosphates as well as on the length of the linker between the sensitiser and the oligonucleotide address. Oligonucleotide derivatives containing a photoreagent at 30-phosphate appeared to be more efficient, viz., the yield of covalent adducts amounted to 70%. The rate of photomodifica- tion sensitised to UV light (365 ± 390 nm) was 100 ± 1500 times higher than that of direct complementarily addressed modifica- tion and decreased with an increase in the linker length in the series of the pyrenyl derivatives 43a ± c.94 Photomodification of ss-DNA in the complexes 49 and 50 with the oligonucleotide azide derivative 48 has been carried out.At equimolar ratios of conjugate and target, the largest yields of covalent adducts of direct photomodification within the complex 49 (Fig. 2, curve 1) and of photomodificataion sensitised with the oligonucleotide benzoanthracene derivative 44 within the com- plex 50 (Fig. 2, curve 2) are identical (*30%), but the initial rate of the sensitisation reaction is 120 times greater than that of direct photomodification. The observed increase in the rate of photo- modification is due to efficient energy transfer from the photo- excited sensitiser S* to the photoreagent R. With a 50-fold excess of oligonucleotide conjugates (curve 3), the initial rate of sensi- tised photomodification was not practically changed in compar- ison with the rate of sensitised modification represented by the curve 2, but the yield of covalent adducts increases up to 72%.The yield of direct photomodification products increases insignifi- cantly under these conditions.90, 93 To account for these results, the dependence of the yields of covalent adducts was studied with a deficit of the sensitiser and an excess of the photoreagent on the concentration ratio of theSite-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents A (%) 3 60 40 1 2 200 2 3 4 1 log t (s) Figure 2.The yield (A) of covalent adducts of an oligonucleotide conjugate containing an azide residue 48 to a complementary DNA- target upon irradiation with UV light (365 ± 390 nm) of the complexes 49 (1) and 50 (2, 3); concentrations/ mmol litre71; oligonucleotide conjugates, 50 (1 ±3), DNA-target, 50 (1, 2), 1 (3). sensitiser to the DNA-target (Fig. 3). In the absence of the catalytic effect of the sensitiser it should be linear as in the case of irradiation under such conditions (10 min, 4 8C) where the components of the complex 50 are not exchanged for the solution components (curve 1). With an increase in the irradiation time and in temperature, each sensitiser molecule initiates the modification of several target molecules (curves 2 and 3).Thus with a 100-fold deficit of the sensitiser (irradiation at 42 8C), each sensitiser molecule induces photomodification of up to 24 target mole- cules.93 The specificity of direct and sensitised photomodification of a natural DNA-target was studied.91, 92 A mixture of 18 fragments of ss-DNA of M13BMR4 was obtained by enzymatic hydrolysis. One of these fragments having the length of *150 bases contained a sequence-target complementary to oligonucleotide addresses (similar to the complexes 49 and 50); while other fragments did not contain this sequence. The results of direct and sensitised photomodification of a mixture of DNA restricts with an oligonucleotide conjugate carrying an azide residue 48 at 32P-labelled 50-phosphate are shown in Fig.4. A (%) 3 80 2 60 40 1 200 0.6 0.4 0.2 0.8 cS/cDNA Figure 3. The dependence of the relative yields (A) of the covalent adducts of sensitised photomodification within the complex 50 on the concentration ratio of the sensitiser (cS) and DNA-target (cDNA) upon irradiation with UV light (365 ± 390 nm). The yield of adducts at an equimolar cs/cDNA ratio is taken as 100%. The concentration of the DNA- target and the oligonucleotide photoreagent is 50 mmol litre71. The irradiation was performed for 10 min, 4 8C (1); 90 min, 42 8C (2) and 300 min, 42 8C (3). 977 A (%) 6 80 5 60 40 4 20 1 ± 3 0 4 1 2 3 log t (s) Figure 4. The temperature dependence of the yields of the covalent adducts (A) upon photomodification of a mixture of M13BMR4 ss- DNA restricts within the complexes 49 (1 ±3) and 50 (4 ±6) upon irradiation with visible light (400 ± 480 nm)/ 8C: (1, 4) 4; (2, 5) 20; (3, 6) 37.An excess of noncomplementary DNA fragments has practi- cally no effect on geff of sensitised modification (curve 6) in comparison with modification within the model complex 50. The efficiency of direct photomodification of a mixture of restricts is noticeably decreased (curves 1 ± 3) in comparison with geff for the complex 49, apparently due to the competition with partially complementary sequences of different fragments of M13 DNA. The efficiency and rate of direct photomodification weakly depend on temperature (curves 1 ±3). In the case of sensitised photomodification, its efficiency increases sharply with an increase in temperature from 4 to 37 8C (curves 4 ± 6).In this case, the rate of the reaction increases first (curves 4, 5) and then decreases 2-fold as the melting temperature of the complex has been reached (curve 6). Such a temperature dependence is due to the ability of the single-stranded fragment to form hairpins near the modification site; at low temperatures, the target sequence becomes inaccessible for oligonucleotide conjugates. The accessibility of the target sequence and the efficiency of sensitised photomodification increase with an increase in temperature, which causes the unwinding of the structured sites of ss-DNA. At 37 8C, the initial rate of sensitised photomodification is four orders of magnitude higher (curve 6), and the degree of modification is 60 times greater than in the case of a direct reaction (curve 1).At the same temperature, the level of nonspecific modification of fragments of M13 DNA containing no comple- mentary insert is *1.5 times lower for the photosensitised reaction than for the direct one.92 It was found that the selectivity of sensitised photomodification determined as a ratio, geff, of specific to nonspecific modification is 90 times greater than that of the direct reaction.91, 92 Modifications predominantly affect the G residues, the PZ value being dependent on their number in the (pG)n-sequence of DNA-targets at the modification site. The yields of covalent adducts of direct modification at n=1, 2 and 4 are 33%± 41%, 68%± 72% and 82%± 85%, respectively, whereas those of photo- sensitised 62%± 68%,89, 98 modification products are 75%± 82%94, 98 and 98%± 99%,98 respectively.The photomodification of a DNA-target (50-*pATGGATT- CTTGAAGGGGACCGCATT) with a binary system of conju- gates complementary to the underlined regions and containing anthracene 45 and azide 48 residues has been carried out. The results of treatment of the reaction products with piperidine show that photomodification predominantly affected the residues G15 and G16 and, to a lesser degree, the residues G14 and G17.98 This finding made it possible to establish the structure of the binary978 system of oligonucleotide conjugates with the DNA-target (data from two-dimensional 1H NMR).100 The probabilities of the contacts of perfluoroaryl azide residues with bases G14 ± G17 of the DNA-target were as follows: G14 : G15 : G16 : G17= 12 : 40 : 40 : 8.98 b.Photoinduced electron transfer In the foregoing section, it was demonstrated that the use of binary systems of oligonucleotide conjugates increases the initial rate of photosensitised photomodification of ss-DNA (100 ± 10 000-fold in comparison with the rate of direct photo- modification) and the efficiency of the specific reaction (by nearly two orders of magnitude). However, sensitisation due to sin- glet ± singlet energy transfer does not fully prevent nonspecific modification, as is the case with long-chain (57500 bases) sequences.The occurrence of nonspecific reactions is due to the following. A conjugate containing a photoreagent binds to those complementary sequences of a DNA-target which have no adjacent sequences complementary to the conjugate containing a sensitiser.90 In order to prevent nonspecific modification in the formation of imperfect complexes, each of the groups in the photoreactive centre should be inactive under irradiation conditions. With this in mind, Dobrikov et al.97, 99 developed a binary system of oligonucleotide derivatives of the sensitiser 46 and the photoreagent 48 complementary to the neighbouring regions of a DNA-target within the complex 50; this system can be activated with visible light (440 ± 580 nm).The excitation energy (E0,0) of the perylene derivative 46, which was chosen as a sensitiser, is less (64 kcal mol71) 104 than the energy of the 0,0-transition in aromatic azides, but perylene is capable of sensitising their photodecomposition due to photo- induced electron transfer from the azido group to the photo- excited sensitiser.104 Figure 5 shows the results of direct photomodification of a DNA-target within the complex 49 upon irradiation with visible light (420 ± 580 nm) (curve 1) and photomodification within the complex 50 sensitised with oligonucleotide derivatives of ben- zoanthracene 44 (curve 2) and perylene 46 (curve 3). The initial rate of perylene-sensitised photomodification is nearly 300 000 times higher and that of benzoanthracene-sensitised reactions is nearly 10 times higher in comparison to direct photomodification. Irradiation of the complex 49 with UV light (365 ± 390 nm) results A (%) 3 60 40 2 20 1 0 log H (J cm72) 4 2 Figure 5.The dependence of the yields of the covalent adducts of DNA photomodification in the complexes 49 (1) and 50 (2, 3) upon irradiation with visible light on exposure (H). The concentration of the DNA target is 1 mmol litre71, that of oligonucleotide conjugates is 50 mmol litre71. The irradiation was performed at the wavelengths/ nm: 420 ± 580 (1), 450 ± 580 (2, 3). The oligonucleotide conjugate containing a benzoanthracene 44 (2) or perylene 46 (3) residue was used as a sensitiser. MI Dobrikov in adducts (yield *30%) which are cleaved by piperidine at the guanosine residue G14 of the DNA-target (marked by an arrow). The visible light-induced photomodification within the complex 50 with a conjugate of perylene 46 gives covalent adducts (yield *70%) cleaved at the same guanosine residue, which is most likely due to electron transfer from the lone electron pair of the nitrogen atom in the azido group of the photoreagent to the photoexcited perylene.97, 104 The perylene residue generates a radical anion which is rapidly oxidised by atmospheric oxygen to the corresponding 1,12-perylenequinone.97 This is accompanied by a simultaneous formation of a radical cation from perfluoro- aryl azide, which after a series of dark conversions induces DNA photomodification similar to that triggered by the singlet nitrene.10 5 20 25 50-*pATGGATTCTTGAAGGGGACCGCATT . . . . . . . . . . . . . . 30-AACTTCCCCTGGCG 48 46 51 Sensitised photomodification (450 ± 580 nm, 15 min) of a DNA sequence from the promoter region of the multiple drug resistance gene (MDR-1) containing four G residues in the region of the photoreactive centre with a binary system of oligonucleo- tide conjugates (complex 51) occurs with nearly quantitative yield.101, 102 Treatment of the reaction products with piperidine revealed that modification affected predominantly residues G15 and G16 and, to a lesser degree, residues G14 and G17. Photo- modification did not take place in the absence of the sensitiser. Thus, binary systems of oligonucleotide reagents containing perylene 46 and perfluoroaryl azide 48 residues can be used for quantitative photomodification of a DNA-target upon irradiation with visible light in the blue-green region of the spectrum (450 ± 580 nm) where both components of the system are singly practically inactive.IV. Two-quantum photosensitised modification of DNA The .OH radical and singlet 1O2 are the main reagents for nucleic acid photooxidation and possess high toxicity. It was therefore proposed to use photoreagents, which do not form active diffusing species. The chemical activity of photoreagents can be triggered by two-quantum excitation of higher electronic states.106, 107 The progress in picosecond laser technologies has made it possible to conduct multiquantum reactions. Two-quantum excitation can be mediated by two different mechanisms, viz., resonant two-quan- tum excitation and nonresonant two-photon excitation.The former consists in a consecutive absorption of two photons via a real intermediate state of the absorbing species. This is populated in the absorption of the first quantum and is a starting point for the absorption of the second quantum. The real excited state has a definite lifetime (normally, 1074 ± 1079 s). Photochemical modification of DNA can be performed both from higher singlet and triplet excited states.15, 106, 108 The second mechanism, viz., nonresonant two-quantum excitation, does not involve any real intermediate state. In this case, the second photon will be absorbed simultaneously with the first one, i.e., within *10715 s.Two-quantum (especially, non- resonant) excitation requires high light intensities (I5 361014 W m72).108 Summation of light quantum energy needed for a photo- chemical reaction to take place is an important feature of two- quantum excitation mechanisms. Yet another peculiarity of two- quantum reactions is their correspondence to spectroscopic selection rules. If a photosensitiser molecule has a centre of symmetry, the absorption of each light quantum changes the parity of the wave function. Hence, two-quantum excitation can generate reactive states which are forbidden in single-quantum excitation.108, 109Site-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents 1.Two-quantum nonresonant excitation Two-quantum excitation is a promising field for practical appli- cation of known photoreagents, since it allows one to trigger photochemical reactions by irradiation with red and infrared light instead of ultraviolet and violet light. IR irradiation (750 ± 1000 nm) can penetrate deeply into tissues, is not absorbed by cell components and does not induce cell injuries. The specificity of its action can thus be significantly increased.108 Psoralens belong to the most promising candidates for photo- reagents, but their spectral photosensitivity was limited by the range of 320 ± 436 nm.108, 110 Considerable progress was achieved in the past few years in overcoming this obstacle.Thus irradiation of 8-methoxy-, 40-amino-4,5,8-trimethyl- 109 and 40-hydroxy- methyl-4,50,8-trimethylpsoralens 111 with a titanium-sapphire laser (730 nm) was accompanied by nonresonant two-quantum excitation and anti-Stokes fluorescence at 452 nm the intensity of which showed a quadratic dependence on the peak power of the laser.108 The same photochemical reactions were triggered as in the case of single-quantum activation of psoralens with UV light (365 nm). Photobiological properties of psoralen in two-quantum exci- tation were studied in vivo in inactivation of Salmonella typhimu- rium His-operon. Short-pulse (t*0.2 ns), high-flux laser irradiation (I=361014 W m72) in the absence of psoralen was characterised by low mutagenicity; in the presence of psoralen, laser irradiation (730 nm) fully suppressed the expression of the His-operon due to the formation of covalent psoralen bisadducts with ds-DNA.108 2.Two-quantum resonant excitation Two-quantum resonant excitation consists in sequential absorp- tion of two quanta of light, resulting in both higher excited singlet states (Sn): hn hn S1 Sn , S0 and higher excited triplet states (Tn): hn hn Tn . T1 S1 S0 a. Two-quantum resonant singlet ± singlet excitation Photoexcitation generates primarily singlet states, Sn, character- ised by higher energy but several orders of magnitude shorter lifetimes than the corresponding triplet states. Thus the lifetime of state Sn is less than 10711 s, while that of state S1 is usually 1078 s.The appropriate conditions for bimolecular reactions from states Sn are attained in the case where the sensitiser forms a donor ± acceptor complex with the molecule-target in the ground state or an exciplex in an excited state.111 For realisation of this mechanism, the photoreagents should have a low yield of intersystem crossing (ISC) into the triplet state and the corresponding S1?Sn conversions. Some polymethine dyes (jISC40.03), tetraazaporphyrins (jISC40.15), Mg2+- and Zn2+-phthalocyanines (MPC2+, M=Mg, Zn) (jISC40.06), etc., meet these requirements.106, 112 Nonlinear absorption of MgPC2+ (e681=2.26105, jfl=0.6, tfl=5 ns) was observed in pulse (t=1.6 ns) laser excitation (680 nm) with a pulse density of >1024 Einstein cm72 s71.The yield of two-quantum excited molecules reached *100% at the pulse density of 1027 Einstein cm72 s71. The absorption spec- trum of *MgPC2+ from state S1 has a long-wave maximum at l=710 nm. Simultaneous irradiation of MgPC2+ with light of two different wavelengths (l=680 and 710 nm) results in efficient population of state Sn (t*15 ps), which is insufficient for the normal course of bimolecular reactions but sufficient for a photoinduced electron transfer which is completed within 0.3 ps.106 Laser-induced two-quantum photomodification of DNA sensitised by different intercalating dyes among which ethidium 979 bromide proved to be the most efficient one has been studied.113 ± 115 Site-specific two-quantum photomodification of ss- and ds- DNA with oligonucleotide conjugates of ethidium derivatives 20 and 21 has been carried out.Irradiation with a nitrogen laser (337 nm, I=150 W m72) results in stepwise two-quantum exci- tation of the chromophore intercalated into DNA with subse- quent modification of the target. The energy of two quanta was sufficient for initiating different photomodification reactions including the cleavage of phosphodiester bonds.113, 114 The focussed laser irradiation of the duplex 50-pAGAGTATTGACTTA . . . . . . . . TCTCATAAp-R resulted in a spontaneous cleavage of the DNA-target at the residue G9 (yield *10%) and the formation of covalent adducts (yields 20%± 40%). The efficiency of spontaneous cleavage of the DNA-target showed a quadratic dependence on the irradiation power density.After piperidine-induced cleavage, the total degree of modifica- tion of the DNA-target increased due to cryptic modification and reached 90%. Unfocussed laser irradiation (2 W m72) of the duplex gave exclusively covalent adducts (yield 40%).115 The mechanism of two-quantum cleavage is not finally elucidated yet. However, it was shown that an addition of chemical traps of radicals and 1O2 did not decrease substantially the efficiency of spontaneous cleavage. The majority of breaks are at a distance of not more than one base from the ethidium residue. Consequently, the migration radius of two-quantum excitation along the DNA chain does not exceed 4 ± 5 A.113, 114 The attachment of ethidium 20 to the heptanucleotide 50-ACAAACCp caused an increase in the melting temperature of the duplex with the complementary octanucleotide 50-pTGTTTGGC from 23 to 50 8C.Irradiation of the duplex with a nitrogen laser (337 nm, t=8 ns, I=1200 W m72, 4 h) up to complete photodegradation of the ethidium residue in the case of the 50-ACAAACCp-20 conjugate gives rise to covalent adducts (yield 40%). Their treatment with piperidine induced the cleavage at the residues G6 and G7 and increased the degree of injury of the target to 65% due to cryptic modification.50 Irradiation of the octanucleotide 50-pTGTTTGGC duplex with the 50-ACAA- ACCp-21 conjugate yields 55% of covalent adducts and 5% of direct cleavage products.Treatment with piperidine induces cleavage at the same residues (G6 and G7) (total yield 70%). Irradiation of a ds-DNA triplex with the 50-(TC)6T derivative of ethidium 20 results in a direct cleavage (yield 10%) of the polypurine chain at residues G(71), A(0), A(1) and G(2) relative to the position of ethidium within the triplex. This was not accompanied by any noticeable photomodification of the poly- pyrimidine chain.50 The type of injuries and noticeable specificity of photomodi- fication with respect to the electron-donor purine bases is consistent with the following mechanism. In the case of nano- second duration of the laser pulse, the triplet state cannot be formed and a two-quantum excitation passes into state Sn and is accompanied by ionisation and formation of an ethidium radical cation and a free electron.The former causes the electron abstraction from the purine bases of DNA or hydrogen atoms from deoxyribose. The quantum yield of ethidium bromide- sensitised photodegradation of DNA is very low (j=561078) due to a reverse electron transfer; however, in the presence of the chemical trap of electrons, viz., methyl viologen, it increases by an order of magnitude.116 b. Two-quantum resonant triplet ± triplet excitation Binary systems of oligonucleotide reagents open up broad opportunities for the application of different mechanisms of two-quantum sensitisation in site-specific photomodification of980 E T2 S1 S1 hn2 T1 hn1 T1 0 S0 S0 Photoreagent Sensitiser Figure 6.The distribution pattern of the ground and excited states of pyrene 43a and perfluoroarylazide 34. DNA. Sensitised photomodification was carried out by irradia- tion with UV and visible light (365 ± 580 nm) simultaneously of a complex of a DNA-target with a binary system of oligonucleotide conjugates 50 containing pyrene 43a and azide 34 residues. This was accompanied by a sixfold increase (in comparison with irradiation at 365 ± 390 nm) in modification rate and a change of its directionality from the guanine residue G11 to the thymine residue T13.95 It was assumed that these conditions favour two- quantum triplet ± triplet sensitisation of the azide 34 with the pyrene 43a (its mechanism is shown in Fig.6). Pyrene has rather high quantum yield of intersystem crossing into the triplet state (jISC=0.37),117 but the energy of the triplet level of pyrene [E(T1)=48.7 kcal mol71]118 is much lower than that of aromatic azides [E(T1)&68 kcal mol71].104 Conse- quently, one-quantum triplet sensitisation of the azide 34 with pyrene is impossible in this case. However, pyrene is characterised by intense triplet ± triplet absorption in the visible region. The lifetime of the triplet state (T1) is rather high (t&0.5 s) and pyrene, being in this state, can absorb the second quantum [in this case, of visible light (hn2)] in order to pass to higher triplet levels (Fig. 6). The energy of the excited triplet level of pyrene [E(T2)=77.8 kcal mol71]118 mark- edly exceeds the energy of the triplet state of the azide 34.As a result, the two-quantum triplet ± triplet transfer of energy from the sensitiser to the T1-level of the photoreagent is thermodynamically allowed and should occur at a diffusionally controlled rate. The azide 34 excited by a triplet ± triplet energy transfer further dissociates into nitrogen and the triplet nitrene 33. The latter causes photomodification of the DNA-target with the formation of products characteristic of this biradical species. Figure 7 presents the logarithmic dependences of the initial rates of accumulation of covalent adducts in the complexes 49 and 50 on the light flux intensity upon irradiation at various wave- lengths. In these coordinates, the slope of the straight line log k (min71) y=1.6x70.4 3 III y=0.95x70.6 II �1 �2 1 I y=0.97x72.2 71 2 log I (W m72) 7371 0 1 Figure 7.The dependence of the initial rate constant (k) of direct (I) and sensitised (II, III) photomodification ofDNAon the light flux intensity (I) within the irradiation range/ nm: 365 ± 440 (1), 365 ± 390 (2). MI Dobrikov corresponds to the reaction order with respect to light. It can be seen that under any irradiation conditions, the slope of the photomodification straight line is close to unity (straight line I). The slope of the straight line for sensitised photomodification upon irradiation with UV light (365 ± 390 nm) is also close to unity (straight line II), whereas upon simultaneous irradiation with UV and visible light (365 ± 580 nm) it is equal to 1.6 (curve III ).This suggests that such irradiation conditions favour two- quantum DNA photomodification which occurs in parallel with one-quantum one. The critical distance for a two-quantum triplet ± triplet energy transfer is equal to 5 ± 6 A, which is 2 ± 3 times less than that for one-quantum energy transfer.119 The occurrence of two-quantum sensitisation within the complementary complex 50 suggests that the sensitiser and photoreagent residues can approach each other to a distance of up to 6 A, at least at the moment of light quantum absorption. This was confirmed in 2D NMR studies of the structure of a complex of a binary system with a DNA-target. 6 4 2 8 10 12 50-*pGTATCA.GTTTCT-30 .. . . . . . . . . . 30-CATAGC CAAAGA 24 13 p-34 43a-p The pyrene residue 43a stabilises the complex by increasing the melting temperature by 7 8C due to hydrophobic interactions with heterocyclic bases. It is localised on the side of the small groove and reacts with the residues T4 ±C5 ± A6 ±G7 of the DNA-target. The photoreagent occupies a position on the side of the major groove near the residue G7. The distance between the centres of pyrene and azide residues is equal to 8 ± 9 A. This convergence of the sensitiser and the photoreagent ensures highly efficient trans- fer.100 Irradiation of the duplex results in the target photomodi- fication at the residue G7. Thus, the results presented above demonstrate high prospects of two-quantum methods for NA photomodification, since they allow the use of light of longer wavelengths and ensure a control over the reactivity of photoreagents under variable irradiation conditions.V. The prospects for the use of photoreagents for targeted modification of nucleic acids in vivo The mechanisms of photosensitised reactions (modification and cleavage of nucleic acids by the majority of photoreagents) considered in this review have not been finally elucidated. A more detailed knowledge of these mechanisms can be extremely useful for elaboration of novel, more efficient reagents capable of direct cleavage of phosphodiester bonds. High specificity of photomodification is the most obvious advantage of oligonucleotide conjugates of psoralens and per- fluoroaryl azides as photoreagents.Apart from the presence of an address, this is due to the absence of diffusing reactive species and the selectivity toward definite heterocyclic bases (residues T andU in the case of psoralens and residues G in the case of perfluoro azides). Sensitised photomodification of NA with binary systems of oligonucleotide conjugates combines favourably the advantages of physical (viz., the catalytic effect of the sensitiser) and chemical sensitisation (viz., the lack of freely diffusing reactive species). Yet another merit of this approach is that the photoreactive centre is formed from two inert elements only after independent recogni- tion of an extended sequence of the DNA target by two oligonucleotide addresses.Two-quantum sensitisation has a number of advantages as regards the development of efficient approaches to targeted modification of nucleic acids in vivo. Excitation of reagents with two quanta of long-wave light allows one to increase considerably the specificity of sensitised photomodification of a DNA-target by inhibiting one-quantum side reactions, while a change in theSite-specific photosensitised modification of nucleic acids with biradical and electrophilic reagents directionality of photomodification in the two-quantum sensitisa- tion can be more useful in the study of biological complexes thath one-quantum sensitisation. The use of nonresonant two-quantum excitation makes it possible to initiate photochemical reactions by irradiation with red and IR light but will hardly increase the specificity of photomodification.At such high peak intensities, all cell compo- nents which absorb in the UV region at a double frequency will be excited in a two-quantum mode. In this respect, two-quantum resonant excitation at two different wavelengths seems to be a more promising approach. In this case, only compounds the absorption of which is close to that of photosensitisers in both ground and excited states will be excited together with them, but the probability of this reaction is rather low. The results described in the review suggest that a combination of binary systems of oligonucleotide conjugates with two-quan- tum excitation methods is the most helpful tool for selective photomodification of nucleic acids in vivo.This review has been partially supported by the project `Leading Research Schools' (Grant No. 96-15-97732). References 1. S I Antsypovich, T S Oretskaya Usp. Khim. 67 274 (1998) [Russ. Chem. Rev. 67 245 (1998)] 2. C Helene, J-J Toulme Biochim. Biophys. Acta 1049 99 (1990) 3. D G Knorre, V V Vlassov, V F Zarytova, A V Lebedev, O S Fedorova Design and Targeted Reactions of Oligonucleotide Derivatives (Boca Raton, FL: CRC Press, 1994) 4. E Braun, Y Eichen, U Sivan, G Ben-Yoseph Nature (London) 391 775 (1998) 5. H P Spielmann, P A Fagan, T M Bregant, R D Little, D E Wemmer Nucl. Acids Res. 23 1576 (1995) 6. R Kuroda, H Satoh,M Shinomiya, T Watanabe, M Otsuka Nucl.Acids Res. 23 1524 (1995) 7. A S Boutorine, D Brault,M Takasugi, O Delgado, C Helene J. Am. Chem. Soc. 118 9469 (1996) 8. V V Vlassov The Lock and Key Principle (Ed. J-P Behr) (New York: Wiley, 1994) p. 89 9. B Armitage, T H Koch, H Frydenlund, H Orum, H-G Batz, G B Schuster Nucl. Acids Res. 25 4674 (1997) 10. D S Sigman, A Mazunder, D M Perrin Chem. Rev. 93 2295 (1993) 11. I Siato,M Takayama, H Sugiyama, T Nakamura NATO ASI Ser. 479 163 (1996) 12. D G Knorre Bioorg. Khim. 23 3 (1997) a 13. D S Sigman Biochemistry 29 9098 (1990) 14. G Laustriat Biochimie 68 771 (1986) 15. R P Wayne Principles and Applications of Photochemistry (Oxford, New York, Tokyo: Oxford University Press, 1988) 16. J Cadet,M Berger, C Decarroz, J R Wagner, J E Van Lier, Y M Ginot, P Vigny Biochimie 68 813 (1986) 17.R V Bensasson, E J Land, E J Truscott Flash Photolysis and Pulse Radiolysis. Contributions to the Chemistry of Biology and Medicine (Oxford, New York, Toronto: Pergamon Press, 1983) 18. N K Kochetkov, E I Budovskii, E D Sverdlov, N A Simukova, M F Turchinskii, V N Shibaev Organicheskaya Khimiya Nukleino- vykh Kislot (Organic Chemistry of Nucleic Acids) (Moscow: Khimiya, 1970) p. 615 19. D N Nikogosyan Laser Picosecond Spectroscopy and Photochemistry of Biomolecules (Ed. V S Letokhov) (Bristol: Hilger, 1987) p. 212 20. H Gorner, M Wala, D Schulte-Frohlinde Photochem. Photobiol. 55 173 (1992) 21. D Magda, R A Miller, M Wright, J L Sessler, B L Iverson, P I Sanson NATO ASI Ser.479 337 (1996) 22. D G Knorre, O S Fedorova, E I Frolova Usp. Khim. 62 70 (1993) [Russ. Chem. Rev. 62 65 (1993)] 23. J C Quada Jr, M J Levy, S M Hecht J. Am. Chem. Soc. 115 12171 (1993) 24. I H Holdberg Acc. Chem. Res. 24 191 (1991) 25. D A Dunn, V H Lin, I E Kochevar Biochemistry 31 16620 (1992) 26. D B Hall, J K Barton J. Am. Chem. Soc. 119 5045 (1997) 981 27. K Nakatani, J Shirai, S Sando, I Saito J. Am. Chem. Soc. 119 7626 (1997) 28. B L Lee, K R Blake, P S Miller Nucl. Acids Res. 16 10681 (1988) 29. E J Gunter, P A Havre, F P Gasparro, P M Glazer Photochem. Photobiol. 63 207 (1996) 30. C Giovannangeli, N T Thuong, C Helene Nucl. Acids Res. 20 4275 (1992) 31. F P Gasparro, P A Havre, G A Olack, E J Gunther, P M Glazer Nucl.Acids Res. 22 2845 (1994) 32. P A Havre, E J Gunther, F P Gasparro, P M Glazer Proc. Natl. Acad. Sci. USA 90 7879 (1993) 33. P A Havre, P M Glazer J. Virol. 67 7324 (1993) 34. M Grigoriev, D Praseuth, A L Guieysse, P Robin, N T Thuong, C Helene, A Harel-Bellan Proc. Natl. Acad. Sci. USA 90 3501 (1993) 35. J A Hartley, S R McAdam, S Das,M C Roldan, M K Haskell, M Lee Anti-Cancer Drug Des. 9 181 (1994) 36. K E Rao, G Gosselen, D Mrani, C Perigaut, J L Imbach, C Bailly, J P Henichart, P Colson, C Houssier, J W Lown Anti-Cancer Drug Des. 9 221 (1994) 37. C Bailly, J B Chaires Bioconjugate Chem. 9 513 (1998) 38. G Delons, J-P Clarenc, B Lebleu, J-P Leonetti J. Biol. Chem. 269 16 933 (1994) 39. K M Meisenheimer, T H Koch Critical Rev.Biochem. Mol. Biol. 32 101 (1997) 40. A Favre, C Saintome, J-L Fourrey, P Clivio, P Laugaa J. Photochem. Photobiol., B Biol. 42 109 (1998) 41. P Laugaa, A Woisard, J-L Fourray, A Favre C. R. Hebd. Seances Acad. Sci., Life Sci. 318 307 (1995) 42. H W Pley, K M Flaherty, D B McKay Nature (London) 372 68 (1994) 43. A Woisard, J-L Fourrey, A Favre J. Mol. Biol. 239 366 (1994) 44. W G Scoott, J B Murray, J Arnold, B L Stoddard, A Klug Science 274 2065 (1996) 45. C Bravo, F Lescure, P Laugaa, J-L Fourrey, A Favre Nucl. Acids Res. 24 1351 (1996) 46. M I Dobrikov, V F Zarytova, N I Komarova, A S Levina, S A Lokhov, T A Prikhod'ko, G V Shishkin, D R Tabatadze, M M Zaalishvili Bioorg. Khim. 18 540 (1992) a 47. A S Levina, D R Tabatadze, M I Dobrikov, G V Shishkin, L M Khalimskaya, V F Zarytova Antisense Nucl. Acid Drug Develop.6 119 (1996) 48. J-F Wang,W D Downs, T R Cech Science 260 504 (1993) 49. J-F Wang, T R Cech J. Am. Chem. Soc. 116 4178 (1994) 50. A A Koshkin, K Yu Kropachev, S V Mamaev, N V Bulychev, S G Lokhov, V V Vlassov, A V Lebedev J. Mol. Recogn. 7 177 (1994) 51. N P Gritsan, A A Koshkin, A Yu Denisov, Yu Ya Markushin, E V Cherepanova, A V Lebedev J. Photochem. Photobiol., B Biol. 37 40 (1997) 52. J Brunner Annu. Rev. Biochem. 62 483 (1993) 53. U Henriksen, C Larsen, G Karup, C Jeppesen, P E Nielsen, O Buchardt Photochem. Photobiol. 53 299 (1991) 54. M Chatterjee, S E Rokita J. Am. Chem. Soc. 112 6397 (1990) 55. M Chatterjee, S E Rokita J.Am. Chem. Soc. 113 5116 (1991) 56. M Chatterjee, S E Rokita J. Am. Chem. Soc. 116 1690 (1994) 57. Y Chiang, A J Kresge, B Hellrung, P Schunemann, J Wirz Helv. Chim. Acta 80 1106 (1997) 58. I Saito,M Takayama, T Sakurai J. Am. Chem. Soc. 116 2653 (1994) 59. J M Kelly,M M Feeney, L Jacquet, A K-D Mesmaeker, J-P Lecomte Pure Appl. Chem. 69 767 (1997) 60. A K-D Mesmaeker, I Ortmans, O V Gijte,W Bannwarth, in The 30th International Conference on Coordination Chemistry (Abstracts of Reports), Kyoto, 1994 p. 68 61. C Mousheron, A K-D Mesmaeker, J M Kelly J. Photochem. Photobiol., B Biol. 40 91 (1997) 62. C J Murphy,M R Arkin, Y Jenkins, N D Ghatlia, H S Bossmann, N J Turro, J K Barton Science 262 1025 (1993) 63. D B Hall,R E Holmlin, J K Barton Nature (London) 382 731 (1996) 64.K Nakatani, J Shirai, R Tamaki, I Saito Tetrahedron Lett. 36 5363 (1995) 65. T Chen, J Michalak,M S Platz Photochem. Photobiol. 61 600 (1995) 66. R Poe, K Schnapp, M J T Young, J Grayzar,M S Platz J. Am. Chem. Soc. 114 5054 (1992) 67. G B Schuster,M S Platz Adv. Photochem. 17 69 (1992)982 68. K A Schnapp, R Poe, E Leyva, N Soundararajan,M S Platz Bioconjugate Chem. 4 172 (1993) 69. M I Dobrikov, V V Gorn, V F Zarytova, A S Levina, T A Prikhod'ko, G V Shishkin, D R Tabatadze,M M Zaalishvili Bioorg. Khim. 18 1190 (1992) a 70. A S Levina, D R Tabatadze, V F Zarytova, M I Dobrikov, V V Gorn, L M Khalimskaya Bioorg. Khim. 20 21 (1994) a 71. A S Levina, D R Tabatadze, V F Zarytova, M I Dobrikov, V V Gorn, S A Lokhov Bioorg.Khim. 20 30 (1994) a 72. A S Levina, D R Tabatadze, M I Dobrikov, G V Shishkin, V F Zarytova Antisense Nucl. Acid Drug Develop. 6 127 (1996) 73. M I Dobrikov, R Yu Dudko, A S Levina, L M Khalimskaya, G V Shishkin Bioorg. Khim. 22 200 (1996) a 74. A S Levina,M V Berezovsky, A G Veniaminova,M I Dobrikov, M N Repkova, V F Zarytova Biochimie 75 25 (1993) 75. K N Bulygin, D M Graifer, M N Repkova, I A Smolenskaya, A G Veniaminova, G G Karpova RNA 3 1480 (1997) 76. A V Kazantsev, G A Maksakova, O S Fedorova Bioorg. Khim. 21 767 (1995) a 77. V V Koval', G A Maksakova, O S Fedorova Bioorg. Khim. 23 266 (1997) a 78. S V Doronin, M I Dobrikov, M Buckle, P Roux, H Buc, O I Lavrik FEBS Lett. 354 200 (1994) 79. M I Dobrikov, S V Doronin, I V Safronov, G V Shishkin Khimiya v Interesakh Ustoichivogo Razvitiya 2 605 (1994) b 80.M I Dobrikov, S V Doronin, O I Lavrik FEBS Lett. 313 31 (1992) 81. G Pratviel, J Bernadou, B Meunier Angew. Chem., Int. Ed. Engl. 34 746 (1995) 82. H Sugiyama, Y Tsutsumi, K Fujimoto, I Saito J. Am. Chem. Soc. 115 4443 (1993) 83. H Sugiyama, Y Tsutsumi, I Saito J. Am. Chem. Soc. 112 6720 (1990) 84. R O Rahn, H G Sellin Photochem. Photobiol. 55 309 (1992) 85. B S Rosenstein, R B Setlow, F E Ahmed Photochem. Photobiol. 31 215 (1980) 86. T Hard, P Fan, D R Kearns Photochem. Photobiol. 51 77 (1990) 87. C L Limoli, J P Day, J F Ward, W F Morgan Photochem. Photobiol. 67 233 (1998) 88. C L Limoli, J F Ward Radiat. Res. 138 312 (1994) 89. M I Dobrikov, S A Gaidamakov, A A Koshkin, N P Luk'yanchuk, G V Shishkin, V V Vlassov Dokl. Akad. Nauk 344 122 (1995) c 90. V V Vlassov,MI Dobrikov, S A Gaidamakov, E K Gaidamakova, T I Gainutdinov, A A Koshkin NATO ASI Ser. 479 195 (1996) 91. M I Dobrikov, S A Gaidamakov, A A Koshkin, G V Shishkin, V V Vlassov Dokl. Akad. Nauk 351 547 (1996) c 92. M I Dobrikov, S A Gaidamakov, A A Koshkin, V V Vlassov Dokl. Akad. Nauk 351 687 (1996) c 93. MI Dobrikov, S A Gaidamakov, E K Gaidamakova, T I Gainutdinov, A A Koshkin, V V Vlassov Antisense Nucl. Acid Drug Develop. 7 309 (1997) 94. MI Dobrikov, S A Gaidamakov, A A Koshkin, T I Gainutdinov, N P Luk'yanchuk, G V Shishkin, V V Vlassov Bioorg. Khim. 23 191 (1997) a 95. MI Dobrikov, S A Gaidamakov, A A Koshkin, T I Gainutdinov, N P Luk'yanchuk, G V Shishkin, V V Vlassov Bioorg. Khim. 23 553 (1997) a 96. M I Dobrikov, S A Gaidamakov, G V Shishkin, V V Vlassov Bioorg. Khim. 24 831 (1998) a 97. M I Dobrikov, S A Gaidamakov, T I Gainutdinov, E D Tenetova, G V Shishkin, V V Vlassov Bioorg. Khim. 25 31 (1999) a 98. M I Dobrikov, S A Gaidamakov, T I Gainutdinov, T M Ivanova, V V Vlassov Bioorg. Khim. 25 137 (1999) a 99. M I Dobrikov, S A Gaidamakov, T I Gainutdinov, E D Tenetova, G V Shishkin, V V Vlassov Dokl. Akad. Nauk 358 403 (1998) c 100. E V Bichenkova, D S Marks, S G Lokhov,M I Dobrikov, V V Vlassov, K T Douglas J. Biomol. Struct. Dyn. 15 307 (1997) 101. M I Dobrikov, T I Gainutdinov, V V Vlassov, in The XIIIth International Round Table `Nucleosides, Nucleotides and their Bio- logical Applications' (Abstracts of Reports), Montpellier, 1998 p. 332 MI Dobrikov 102. M I Dobrikov, T I Gainutdinov, V V Vlassov Nucleosides Nucleotides 18 1517 (1999) 103. M Ya Mel'nikov, V A Smirnov Fotokhimiya Organicheskikh Radikalov (Photochemistry of Organic Radicals) (Moscow: Moscow State University, 1994) p. 90 104. L J Leyshon, A Reiser J. Chem. Soc., Faraday Trans. 2 11 1918 (1972) 105. H A Sharifian, C-H Puin, F-B Jiang, S-M Perk J. Photochem. 30 229 (1985) 106. K Teuchner, A Pfarrherr, H Stiel,W Freyer, D Leupold Photochem. Photobiol. 57 465 (1993) 107. D Leupold,W Freyer J. Photochem. Photobiol., B Biol. 12 311 (1992) 108. W G Fisher,W P Partridge Jr, C Dees, E A Wachter Photochem. Photobiol. 66 141 (1997) 109. C R Lambert, H Stiel, D Leupold,M C Lynch, I E Kochevar Photochem. Photobiol. 63 154 (1996) 110. D H Oh, R J Stanley,M Lin, W K Hoeffler, S G Boxer, M W Berns, E A Bauer Photochem. Photobiol. 65 91 (1997) 111. G O Bekker, A V El'tsov Vvedenie v Fotokhimiyu Organicheskikh Soedinenii (Introduction to Photochemistry of Organic Compounds) (Leningrad: Khimiya, 1976) 112. D Leopold, I E Kochevar Photochem. Photobiol. 66 562 (1997) 113. N V Bulychev, Doctoral Thesis in Chemical Sciences, Novosibirsk Institute of Bioorganic Chemistry, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, 1991 114. L Z Benimetskaya, N V Bulychev, A L Kozionov, A A Koshkin, A A Lebedev, S Yu Novozhilov, M I Shtokman Bioorg. Khim. 14 48 (1988) a 115. L Z Benimetskaya, N V Bulychev, A L Kozionov, A V Lebedev, S Yu Novozhilov, M I Stockmann Biopolymers 28 1129 (1989) 116. Z Shaquiri, E Keskinova, A Spassky, D Angelov Photochem. Photobiol. 65 517 (1997) 117. F Wilkinson Organic Molecular Photophysics Vol. 2 (Eds J B Birks, J London) (New York: Willey, 1975) p. 95 118. H Labhart,W Heinzermann Organic Molecular Photophysics Vol. 1 (Eds J B Birks, J London) (New York: Willey, 1973) p. 297 119. V L Ermolaev, E N Bodunov, E B Sveshnikova, T A Shakhverdov, in Bezyzluchatel'nyi Perenos Energii Elektronnogo Vozbuzhdeniya (Nonradiative Transfer of Energy of Electron Excitation) (Leningrad: Nauka, 1977) p. 94 a�Russ. J. Bioorg. Chem. (Engl. Transl.) b�Chem. Sustain. Devel. (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC