摘要:

Russian Chemical Reviews 69 (8) 639 ± 660 (2000) Synthesis and properties of functionalised dendrimers I P Beletskaya, A V Chuchurjukin Contents I. Introduction II. Hydrocarbon dendrimers containing conjugated p-systems III. Conjugated dendrimers containing transition metal atoms IV. Dendrimers containing pincer-type ligands and their metal complexes V. Dendrimers containing transition metal coordination complexes VI. Dendrimers containing complexes with metal ± carbon s-bonds VII. Dendrimers containing terminal ferrocenyl groups VIII. Dendrimers containing transition metal p-complexes IX. Dendrimers containing central metal-containing fragments Abstract. proper- structure, synthesis, the on data published The The published data on the synthesis, structure, proper- ties containing dendrimers of applications and ties and applications of dendrimers containing conjugated conjugated p-sys- tems, transition metal complexes and organometallic compounds tems, transition metal complexes and organometallic compounds are surveyed and systematised.The bibliography includes 83 are surveyed and systematised. The bibliography includes 83 references. I. Introduction The chemistry of dendritic molecules, which is a fundamentally new branch of chemistry, began to develop explosively in the past few years. The beauty of these flower- or star-shaped structures is so amazing that it may spark the imagination of artists. However, apart from the great attractiveness of these experimental subjects the interest of investigators in this field of chemistry is mainly due to the potential applications of dendrimers, for example, in industry as novel catalysts and materials for molecular electronic devices and light energy conversion, in medicine and in other fields.In contrast with polymers formed by spontaneous polymer- isation, dendrimers, which are oligomeric compounds (in princi- ple, they can be polymeric), are prepared by multistep synthesis. There are two different methodologies, viz., the divergent and convergent approaches. In the divergent approach, dendrimers are constructed by the attachment of various fragments to a dendritic backbone (core). In this case, the molecule is built up layer-by-layer in the direction from the centre of the dendritic molecule to its periphery.The convergent methodology involves initial synthesis of dendrons, i.e., the branches of a dendrimer, which are then coupled to the dendrimer core. This process resembles the assembly of individual sections on a common base. More complex dendritic molecules built out of individual fragments are constructed using methods of organic, organo- metallic or coordination chemistry employed for the preparation of supramolecular structures. I P Beletskaya, A V Chuchurjukin Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 932 88 46. Tel. (7-095) 939 36 18. E-mail: beletska@org.chem.msu.su (I P Beletskaya), Tel. (7-095) 939 11 39 (A V Chuchurjukin) Received 23 February 2000 Uspekhi Khimii 69 (8) 699 ± 720 (2000); translated by R L Birnova #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n08ABEH000577 639 639 641 644 646 652 653 654 655 Dendrimers have been described in numerous reviews;1 ±12 however, the number of new types of dendritic molecules contin- ues to increase.In writing this review, we were guided by our own interests and thus confined ourselves to the consideration of those dendrimers which incorporate ethynyl and aromatic fragments synthesised in reactions catalysed by transition metal complexes and dendrimers containing transition metal complexes as their cores or at the periphery where the metal is linked to the ligand by a coordination bond or by a carbon ± metal bond.In many cases, we followed the path of the authors of the original publications who have discussed zero-generation molecules as matrices for future dendritic systems of the first, second and higher genera- tions. It should be noted that our subdivision into sections is arbitrary, since many dendrimers comprise both arylethynyl fragments (which is the object of our particular interest) and transition metal atoms, leaving silicon-containing and many other dendrimers beyond the scope of consideration. In our attempt at a maximally complete coverage of the available material, we kept on adding novel data, but soon realised that we had to stop, otherwise this review would have never been written, for the number of publications is increasing every day.II. Hydrocarbon dendrimers containing conjugated p-systems Cross-coupling reactions catalysed by palladium complexes are widely employed in the synthesis of dendrimers containing ethynyl or p-phenylene fragments. This type of dendrimer presents sub- stantial practical interest because it ensures structural rigidity and conjugation between the fragments, which are essential for directed energy and electron transfer and for the preparation of photoreactive materials.13, 14 1,3,5-Triethynylbenzene, which does not ensure conjugation between the substituents 15 on the periph- ery, can nevertheless transfer photoexcitation energy 13, 14 and represents a convenient core for the construction of such den- drimers.The photoreactive dendrimer 1, the dendritic branches in which are linked together through ethynyl and p-phenylene bridges with a common focal point { of the dendrimer,16 viz., the { By the term `focal point', the terminal point of the photoexcitation energy transfer in the molecule is understood in this case.640 But But But But But But But But perylene fragment, may serve as an example. This dendrimer was synthesised by cross-coupling of a terminal acetylene with tribro- mobenzene according to the Sonogashira reaction.17 Studies of the photophysical properties of this dendrimer revealed that on irradiation with light with the wavelengths corresponding to the absorption maxima of peripheral triethynylbenzene groups, the radiation intensity of the perylene fragment in the dendrimer exceeds that of free ethynylperylene by more than two orders of magnitude under identical conditions.In this case, the efficiency But N But But N But But But But But But But But But 1 But But N H N N N N But N But 2 But But But But But But But of the energy transfer from the periphery to the focal point of the dendrimer reaches 98%.16 Further investigations showed that the dendrimer of the next generation absorbs more light, but the efficiency of energy transfer to the focal point decreases.18 The dendrimers 2 containing 9-phenylcarbazole fragments linked through ethynyl bridges have been synthesised.19 Such dendrimers are designed for the preparation of organic photo- conductors (i.e., compounds with non-linear optical properties), photo- and electroluminescent materials.But But N N N N But N But I P Beletskaya, A V Chuchurjukin But But But But But But But But But But N But But N ButSynthesis and properties of functionalised dendrimers A dendrimer containing 1,3,5-triethynylbenzene fragments was used as the monomer unit in the synthesis of a polyacetylenic polymer (Scheme 1).20 H Me3Si SiMe3 Me3Si SiMe3 A series of zero-generation dendrimers containing thiophene groups were prepared by trimerisation of acetylthiophene deriv- atives in the presence of SiCl4.21 R S R S S R R=Ph, .S These compounds can be used in the synthesis of dendrimers, organic polymers of higher generations, semiconductors and liquid-crystalline materials. III. Conjugated dendrimers containing transition metal atoms Conjugated systems, which contain several polydentate ligands coordinated to metal atoms within a single molecule, present substantial interest. Examples are provided by bi- (3, 4) and trinuclear (5) complexes of ruthenium with a phenanthroline ligand synthesised by cross-coupling of a Ru ± 3-bromo-1,10- phenanthroline complex with 1,4-diethynylbenzene, 4,40-dieth- ynyl-1,10-biphenyl and 1,3,5-triethynylbenzene, respectively.15 N N Ru (bipy)2 3 Scheme 1 SiMe3 [Rh(C7H8)Cl2]2 SiMe3 n SiMe3 SiMe3 S S S N N Ru (bipy)2 Measurements of the UV-spectra of the complexes thus prepared revealed that the main absorption bands of the bipyr- idine ligands do not change.Conjugation of the phenanthroline ligands in the binuclear complexes based on 1,4-diethynylbenzene (complex 3) and 4,40-diethynylbiphenyl (complex 4) produced an additional low-energy absorption band. The absorption spectrum of the trinuclear complex 5 based on 1,3,5-triethynylbenzene more closely resembles that of the mononuclear complex, which points to the lack of conjugation between the substituents at positions 1, 3 and 5 of the central benzene ring. N N Ru (bipy)2 4 N (bipy)2Ru N N N 5 (bipy)2Ru Ru(bipy)2 Luminescence studies of the heteronuclear complexes 6 ± 8 containing both Ru and Ni, Ru and Cu or Ru and Os cations, which have been prepared by cross-coupling of Ru± phenanthro- line complexes and the bipyridine complexes of Ni, Cu and Os with 1,3,5-triethynylbenzene, established that such complexes can transfer photoexcitation energy.13 But N N N Ru N N N XN N (6), X= N N H2O Ni H2O OH2 OH2 But Thus the fragment 7 was used to build a more complex molecule 9 14 in which the Ni ± bipyridine complex is the central part of the molecular architecture. It is of note that the syntheses of complex molecules on the basis of fragments 6 ± 8 have been accomplished with 67% to 78%yields.1,3,5-Triethynylbenzene was used as a basis for the synthesis of the tridentate phosphine ligand 1,3,5-(Ph2PC:C)3C6H3 (10).To this end, 1,3,5-triethynylbenzene was metallated with 3 equiv. of n-butyllithium and the lithium derivative formed was intro- duced into the reaction with chlorodiphenylphosphine.22 The 641 4+ N N Ru (bipy)2 6+ N N 6+ But N N N Ru N N N 6 ± 8 N N N N N Ru M (8) (7), N N N N N M=Cu, Os.642 But But N N N Ru N N N N N N N Ru N N But But phosphine ligand 10 was first introduced into the reaction with the molybdenum or tungsten carbonyls M(CO)6 (M=Mo, W) and (CO)5M 10 M(CO)6 Ph2P M(CO)5 (CO)5M (OC)3Co (OC)3Co Ph2P Co(CO)3 (OC)3Co M(CO)5 N N N Ru N N N N NRu N N N N N NRu N N N N N N N N RuN N N NRu N N N N N N N N RuN N N NN Ru N N N PPh2 Co2(CO)8 PPh2 11 M(CO)5 PPh2 Co(CO)3 Co(CO)3 PPh2 12 M(CO)5 I P Beletskaya, A V Chuchurjukin 38+ But But N N N Ru NN N N N But N Ru N N N N N N N Ru But N N N N N N RuN N N N N N N Ru N Ni N N N N N N N N N Ru N N N N N But N N RuN N N N But N N RuN N N N N Ru N N N But But 9the complexes 11 formed were treated with cobalt carbonyl Co2(CO)8. These reactions afforded compound 12.The osmium complex 13 was obtained in a similar way. In the latter case, the phosphine ligand 10 was coordinated to three Os3(CO)11 clusters. Os3(CO)11 PPh2 PPh2 Ph2P 13 Os3(CO)11 Os3(CO)11 True organometallic compounds constitute an additional type of dendritic conjugated systems.The platinum complex 14 with triethynylmesitylene ligands prepared by cross-coupling of triio- domesitylene with trimethylsilylacetylene with subsequent desily- lation is an example.23 Triethynylmesitylene was introduced into the reaction with Pt(Bu3P)2Cl2 in the presence of CuCl and further converted into the complex 15 by a reaction with phenylacetylene. Substitution ofSynthesis and properties of functionalised dendrimers PhBu3P Ph PBu3 Bu3P Pt Bu3P Pt Bu3P Bu3P PhPh Bu3P triethynylmesitylene for phenylacetylene yields first- and second- generation dendrimers.23 The organoruthenium dendrimer 16 was synthesised follow- ing the convergent approach.24 It was shown that the introduction of ruthenium atoms into a conjugated p-system significantly enhances both the non-linear optical properties of the original Pt(PBu3)2Cl2 CuCl Cl(PBu3)2Pt Ph Pt PBu3 Bu3P Bu3P Pt Bu3P PBu3 Ph 15 Ph Ph Ph Bu3P PBu3 Bu3P PBu3 Pt Pt Pt Pt PBu3 Bu3P PBu3 Pt PBu3 Bu3P Pt PBu3 Bu3P Ph Bu3P PBu3 PBu3 Bu3P Pt Pt Pt Pt Pt PBu3 Bu3P PBu3 Bu3P PBu3 Pt Pt PBu3 Bu3P PBu3 Bu3P Ph Ph PBu3 Bu3P Bu3P PBu3 Pt Pt Pt Pt PBu3 Bu3P PBu3 Pt Pt Bu3P Bu3P PBu3 Bu3P Ph Ph 14 compound and the two-photon absorption.These characteristics are also enhanced with the growth of the dendrimer. X Pt(PBu3)2Cl PhC CH Pt(PBu3)2Cl X 16 Ph Ph2P PPh2 Ru Ph2P Ph2P PPh2 Ph2P PBu3 Ru X= Pt PPh2 Ph2P Ph PPh2 Ph2P Ru PPh2 Ph2P Ph 643 X644 this case the pincer group is incorporated into the dendrimer structure as the connecting link.IV. Dendrimers containing pincer-type ligands and their metal complexes PhS Cl Pd PhS O O The presence of pincer-type ligands at the periphery of zero- generation dendrimers facilitates their metallation with the for- mation of metallacycles. Thus the complexes 17 ± 19 containing pincer-type SCS ligands were used as building blocks in the synthesis of more complex dendrimers.25, 26 For example, the pyridine group in compound 18 or the cyano group in compound Cl SPh PhS Pd NH O 18 N O CN O O SPh PhS O O PhS Pd Pd Cl Cl Pd 17 SPh PhS Cl 19 PhS The presence of SCS pincer-type ligands strongly restricts the application of non-palladium transition metal complexes, since only cyclopalladation is possible.Replacement of SCS by PCP 19 is coordinated to palladium atoms to yield cationic complexes. The dendrimer 20 synthesised on the basis of the complexes 17 and 18 represents an organometallic compound the dendritic structure of which is formed by coordination bonds. It is noteworthy that in Cl Cl Pd SPh PhS SPh PhS Pd O O HN O NH O PhS O O SPh PhS Pd Pd Cl N N Pd Pd SPh PhS SPh PhS O O O NH Cl SPh PhS NPd Pd SPh PhS O O NH O O O PhS SPh PhS SPh PhS O Pd Pd Pd N N N N Cl Pd Pd NH SPh PhS O PhS SPh PhS O O NH O O OPd SPh PhS SPh PhS Pd N N NH O NH O O O O O SPh PhS PhS SPh Pd Pd Pd Pd Cl Cl Cl Cl SPh PhS PhS SPh 20 I P Beletskaya, A V Chuchurjukin SPh Cl Pd SPh SPh Pd Cl SPh 9+ SPh Cl Cl Pd SPh PhS O O HN SPh O Pd Cl SPhSynthesis and properties of functionalised dendrimers ligands made it possible to conduct cyclometallation with both Pd complexes and a variety of other transition metal compounds.We made an attempt to synthesise zero-generation dendrimers on the basis of triethynyl- (21) and hexaethynylbenzene (22).27, 28 Ph2P Ph2PPh2P Ph2P Cyclopalladation of triethynylbenzene 21 was successful. The corresponding trinuclear palladium complex 23 was obtained in 90% yield. However, cyclopalladation of the ligand 22 designed for the synthesis of a hexanuclear complex failed, since this reaction gave a mixture of compounds.We attribute this failure to the high degree of conjugation within the molecule, which prevents the introduction of six electron-withdrawing substitu- ents. Yet another trinuclear palladium complex 24 prepared by cyclopalladation of the ligand 21 is an efficient catalyst for the Heck reaction.29 Ph2P Cl Pd Ph2PPh2P MeCN Ph2P The applications of other transition metal complexes, e.g., the Ni and Pt complexes 25 and 26 containing PCP pincer-type ligands, have been described.30 PPh2PPh2 21 PPh2 Ph2P PPh2 Ph2P PPh2 Ph2P PPh2 PPh2 PPh2 Ph2P PPh 22 2 Ph2P PPh2 Cl Pd PPh2 Pd Ph2P PPh2 Cl 23 3+ NCMe Pd PPh2 Ph2P PPh2 Pd Pd NCMe Ph2P 24 Cl M PPh2 Ph2P OO O Ph2PM Cl Ph2P 25 CNO O Ph2PM Cl Ph2P 26 M=Ni, Pt.A method for the introduction of the pincer-type NCN and CN ligands, e.g., 1-[3,5-(Me2NCH2)2C6H3] and 1-(4-Me2N± CH2C6H4) units, at the periphery of organosilicon dendrimers of zero (G0) and first (G1) generations has been proposed.31 To this end, the corresponding bromo-substituted derivatives, viz., 1-Br- 3,5-(CH2NMe2)2C6H3 or 1-Br-4-Me2NCH2C6H4, were lithiated and introduced into the reaction with an organosilicon dendrimer containing the chlorosilane groups Si(CH2CH2CH2Cl)3 at the periphery. This approach was used in the synthesis of the organo- silicon dendrimers G07SiMe27 C6H3(CH2NMe2)2-1,3, G07SiMe27C6H4(CH2NMe2)-4, G17SiMe27C6H3(CH2NMe2)2-1,3 and G17SiMe27 C6H4(CH2NMe2)-4.The dendrimers synthesised and model com- pounds were metallated with ButLi and further deuterated. Compound G07SiMe27C6H2(CH2NMe2)2-1,3-Li-2 (27, X=Li) was converted into a platinum derivative under the action of 12 equiv. of PtCl2(SEt2)2. X NMe2 Me2N Me2N Si X Si Me2N Si Me2N Si X Me2N Si Si Si Me2NX Si Si Me2NMe2N X Me2N X Me2N Me2N X=H, Li, D. Catalytic activities of nickel complexes prepared on the basis of organosilicon dendrimers containing peripheral NCN pincer- type ligands were studied in the Kharash-type addition of CCl4 to alkenes.32 The kinetic data suggest that the catalytic activities of 645 PPh2 M Cl PPh2 PPh2 M Cl PPh2 NMe2 NMe2 X X NMe2 NMe2 Si NMe2 X Si Si NMe2 Si NMe2 X Si NMe2 Si NMe2 Si Si X NMe2 X NMe2 Me2N 27646 zero- and first-generation dendrimers are lower than that of the monomeric catalyst (by 20% and 30%, respectively).However, complete separation of the catalyst from the reaction product by filtration through a membrane is an obvious advantage of dendrimers.32 The dendrimer 28 containing cycloplatinated fragments with pincer groups was used as a sensor in the determination of SO2.33, 34 According to calculations, the molecule 28 is planar. In contrast with organosilicon dendrimers, this compound is poorly soluble in common organic solvents, which can be attributed to its structural peculiarities. Earlier studies have shown 33 that plati- num and nickel complexes withNCNpincer ligands can reversibly bind SO2; the resulting compound has the structure of a tetrahe- dral pyramid.The solubility of the complex formed is much higher than that of the starting compound. The dendrimer 28 (both in solution and in a solid state) reacts instantaneously with gaseous SO2 to yield the corresponding addition products, which is accompanied by the appearance of an orange colour. Me2N Cl Pt Me2N O O O O O O Me2N Cl Pt O Me2N O O O O O O Me2N Pt NMe2 Me2N 28 Cl Me2N Cl Pt O2S Me2N O O O O O Me2N Cl Pt O2S Me2N O O O O O O Me2N Pt NMe2 Me2NCl SO2 29 The formation of the adducts 29 was confirmed by 1H NMR and UV spectroscopic data.The addition of SO2 is reversible. The desorption of SO2 occurs on heating of the adduct to 40 8C or in vacuo. NMe2 Cl PtNMe2 NMe2 Cl Pt O O NMe2 O SO2 O OPt NMe2 Cl NMe2 Cl Pt SO2 NMe2 O NMe2 Cl Pt O SO2 NMe2 O O O O OPt NMe2 Cl SO2 I P Beletskaya, A V Chuchurjukin V. Dendrimers containing transition metal coordination complexes 1. Dendrimers containing coordination complexes with phosphine ligands A search for novel catalytic systems stimulated the design of new ligands including those based on dendritic molecules with differ- ent electronic and spatial characteristics. It was anticipated that the use of dendritic ligands in combination with heterogenisation of a catalyst and biphasic systems would provide a clue to the problem of the catalyst's regeneration.Moreover, grafting of a catalyst on the surface of the dendrimer might enhance the regioselectivity by virtue of fixation of its conformation due to steric hindrances created by the dendrimer.35 However, the experimental data available so far evidence that dendritic com- plexes manifest lower activities than their mononuclear ana- logues.35 The reason for this may be related to steric hindrances and other problems due to larger sizes of dendrimers. Besides, the activity of a catalyst may be lost owing to the stabilisation of the complex by various functional groups present in the dendrimer.35 Tertiary phosphines can form complexes with virtually all transition metals; many of those efficiently catalyse various reactions. It is not surprising therefore that the overwhelming majority of publications are devoted to the dendrimers containing phosphine complexes of transition metals.Thus the synthesis of ruthenium complexes with phosphine-containing dendritic ligands was described and their catalytic activity was investi- gated.35 However, reactivities of these dendritic complexes appeared to be lower than that of the original mononuclear complexes RuH2(PPh3)4. Unlike the latter, dendrimers did not react with chloroform and formed insoluble compounds with alkenes which could not be characterised; these compounds slowly reacted with CO to give the corresponding carbonyls.Preliminary studies showed that dendritic ruthenium complexes displayed high catalytic activities in the hydrogenation of ketones.35 Synthesis of dendrimers containing bidentate phosphine ligands [7CH=NN(CH2PPh2)2] at the periphery and the prepa- ration of platinum, palladium and rhodium complexes based on them have been described.36 Dendritic platinum and palladium complexes were methylated in the reaction with MeMgBr and carbonylated by treatment with CO. Dendritic polyamines containing terminal diphenylalkylphos- phine groups, which are ideal ligands for the complexation with transition metals in lower oxidation states, have been synthes- ised.37, 38 The structure of one such compound (its right half is shown), viz., a multinuclear gold(I) complex with a fourth- generation dendrimer containing 32 terminal diphenylphosphine groups (complex 30), is given below.37 X-Ray diffraction analysis of analogous gold(I) mono- and binuclear complexes revealed the presence of intermolecular NH.. .O hydrogen bonds in their crystal structures. It was suggested that similar intramolecular hydrogen bonds are also present in the Au complex with dendritic ligands. Such gold(I) multinuclear complexes represent a new class of metal-containing polymers with definite, probably spheric structures with dendritic molecules as the supporting matrices. These complexes can be used as catalysts and in medicine for biochemical diagnostics and as anti-inflammatory and anti-tumour drugs.37 Dendritic molecules containing chiral ferrocenyl phosphine ligands present special interest.Zero- (compound 31) and first- generation dendrimers containing up to eight chiral diphosphine ligands were synthesised on the basis of benzene-1,3,5-tricarbonyl chloride and adamantane-1,3,5,7-tetracarbonyl chloride 39 and a silylated ferrocenyldiphosphine derivative [C5H3(PPh2)(CHMe. .PCy2)]Fe(C5H4SiMe2CH2CH2CH2Cl), where Cy is cyclohexyl.Synthesis and properties of functionalised dendrimers O CN H NN N N N N N NNN N C N H H C O Cy2P Cy2P PPh2 Fe Si Cy2P PPh2 Fe Si Catalytic properties of rhodium complexes prepared by the reaction of [Rh(COD)2]BF4 (COD is cycloocta-1,5-diene) with these dendritic ligands were studied in asymmetric hydrogenation of dimethyl itaconate.An enantiomer of dimethyl 2-methylbu- tane-1,4-diolate was obtained with ee 98%, which is only 1% lower than that achieved with a mononuclear complex.{ This { Mononuclear catalysts (ferrocene derivatives) are used in the commer- cial production of pesticides.40 Ph2 P AuCl Ph2 P AuCl N C H O N C P AuCl Ph2Ph2 H O P AuCl N C N N C H O P AuCl Ph2 O H C N N P AuCl Ph2 H O AuCl N N N C PPh2 H O AuCl N C PPh2 Ph2 P AuCl H N N Ph2 P AuCl H O N CON C H O N C N Ph2 P AuCl H H O O H O C N AuCl N N C Ph2 P AuCl PPh2 H O N C AuCl PPh2 AuCl O PPh2 AuCl PPh2 30 PPh2 PCy2 Fe Fe PPh2 Si PCy2 Si O NH Fe PPh2 O HN Si O O O HN NH PCy2 O O O O O O HN Fe PPh2 O O HN Si O O HN HN O Si Si Fe Ph2P Fe Ph2P PCy2 PCy2 31 647 result is remarkable in that in this case the catalyst can be fully separated from the reaction products using membrane technol- ogy.Dendrimers of the type 32 with peripheral diphenylphosphi- nomethyl ligands were synthesised from the organosilicon den- drimers Gx7CH2CH2SiMeCl2 (x=0, 1) by reacting them with Ph2PCH2Li .TMEDA (TMEDA is N,N,N0,N0-tetramethyl- ethylenediamine).41 PPh2 PPh2 PPh2 PPh2 PPh2 Si Si PPh2 Ph2P Si PPh2 Ph2P Si Si Si PPh2 Si Ph2P Si PPh2 Si Ph2P Si PPh2 Si Ph2P Si Si Si PPh2 Ph2P Si PPh2 Ph2P Si Si Ph2P Ph2P PPh2 PPh2 Ph2P 32 The reaction of [(Z3-C3H5)PdCl]2 with the dendrimer 32 afforded a palladium complex which was further used as a catalyst in the allylic alkylation of diethyl methylmalonate with sodium allyltrifluoroacetate in both a batch process and in continuous process using a membrane reactor.In the former case, the catalyst manifested an exceedingly high activity. The yield of the reaction product exceeded 80% after 30 min at a substrate : catalyst ratio of 2000 : 1 (see }). However, in a continuous membrane reactor at the same ratio of the reactants, the initially high catalytic activity was gradually almost extinguished following passage of 15 reactor volumes of the reaction mixture. The reaction mixture at the outlet from the reactor did not contain any catalyst.This loss of the catalytic activity was attributed to the decomposition of the palladium complex. Synthesis of silica gel-grafted polyamidoamine dendrimers containing peripheral diphenylphosphinomethyl groups has been carried out.42, 43 These dendrimers were used to prepare the corresponding rhodium complexes 33 which present interest as recyclisable catalysts in the reaction of alkene hydroformylation and can be used, in particular, in the production of drugs of the ibuprofen and naproxen type.42 Studies of the catalytic activities of such dendritic catalysts revealed that these activities increased with the increase in the length of the carbon chain between the amide and amine nitrogen atoms from two to six carbon atoms and did not change further.Such an increase in the catalytic activity seems to be due to the reduction of steric hindrances on the surface of the dendrimer. Hydroformylation of styrene and vinyl acetate involving such catalysts occurred with virtually quantitative yields and high regioselectivity. The catalysts retained high catalytic activities over at least four cycles. However, slow wash-out } of rhodium from the dendrimer immobilised on silica gel did occur. } With respect to 1 mol of the catalytically active groups. } It is noteworthy that rhodium is not washed out from the catalyst with a pure solvent, but is washed out in the presence of CO or a CO7H2 mixture, apparently, as a result of exchange of dendritic phosphine ligands in the rhodium complexes for CO.648 O Si OO 2.Dendrimers based on metal complexes with polypyridine ligands In the foregoing section we considered the methodology of synthesis of metal-containing dendrimers based on the formation of new carbon ± carbon or carbon-heteroatom bonds. Coordina- tion of polydentate organic ligands around metal atoms is an alternative route to the construction of metal-containing den- drimers.44 This procedure was used, in particular, in the synthesis of the dendrimer 34 containing ruthenium cations coordinated by 2,3-bis(2-pyridyl)pyrazine.44 In such dendrimers, transition metal coordination complexes function both as branching and linking blocks. These dendrimers are prepared using the `complex-as-a-metal ± complex-as-a- ligand' synthetic strategy in which a complex with free chelating groups (`complex-as-a-ligand') chelates a complex with labile ligands (`complex-as-a-metal').This approach allows stepwise monitoring over the dendrimer growth and involvement of build- ing blocks containing various metals or ligands at each stage of the synthesis. Dendrimers comprising complexes of different chemical nature can possess specific chemical properties, e.g., provide gradients for electron transfer. Because of their remarkable photophysical and electrochemical properties, ruthenium(II) and osmium(II) complexes with polypyridine ligands represent ideal components for the synthesis of luminescent materials and other materials which can transform the light energy.44 Numerous dendritic compounds were synthesised based on 2,3- and 2,5-bis(2-pyridyl)pyrazine (2,3- and 2,5-dpp) as the connecting ligands and 2,20-bipyridine (bipy) as the terminal Cl(CO)2RhPPh2 O NH N NH O NO Ph2 P N Cl(CO)2Rh PPh2 Ph2P Ph2 P N N HN NH O O N O N O NHHN O O NO O HN NH NH N NH O N O O NH HN PPh2 N Rh(CO)2Cl Ph2P 33 ligand.44 These complexes contain several chromophore groups and a large number of redox centres.NN I P Beletskaya, A V Chuchurjukin Rh(CO)2Cl PPh2 Rh(CO)2Cl Ph2P PPh2 N NH NH N PPh2 Ph2P Rh(CO)2Cl Rh(CO)2Cl Ph2P PPh2 NN PPh2 Ph2P Rh(CO)2Cl N N N N RuN N N N N N N N N N Ru Ru N N N N N N N N N N N N Ru Ru N N N N N N N N Ru N N N N N N N Ru N N N N N N Ru N N 34 26+ N N N RuN N N N N N N N N N Ru Ru N N N N N N N Ru N N N N N Ru N NSynthesis and properties of functionalised dendrimers 2,20 : 60,200-Terpyridine (terpy) and its derivatives were used as ligands in the synthesis of metal-containing fragments of den- drimers.The synthesis of 1,3,5-tris(2,20 : 60,200-terpyridin-40- yl)benzene, N N N N N N N N N trinuclear 45 and hexanuclear (35) 46 ruthenium complexes pre- pared on its basis has been described.45 The dendritic complex 36 containing 12 ruthenium atoms was obtained by a reaction of 12 equiv. of (terpy)RuCl3 with a polyamide-based dendritic ligand.47 The reaction of 1 equiv.of 1,3,5-tribromomethyl-2,4,6-trime- thylbenzene with 3 equiv. of the ruthenium complex 37 was used to prepare the dendritic complex 38 containing six ruthenium atoms (Scheme 2).48 A similar reaction of 1 equiv. of hexabromomethylbenzene with 6 equiv. of compound 39 gave the dendritic complex 40 containing 18 ruthenium atoms (Scheme 3).49 Studies of electro- RO N N N N N N Ru N N O N N O N N RO Ru N N N HH N N N Ru O N N RO N N O N N Ru N N N N N N RO R=(CH2)3C[(CH2)3OBn]. OR N N Ru N NO O OO OON N Ru N NOR N N N Ru N N N Cl chemical properties of compound 40 revealed that all ruthenium centres are independent and do not apparently interact with one another.OR N N N Ru N N NO O NH OO HN O ON N N Ru N N NOR 36 12+ Cl N N N Ru N N NON N N Ru N N N N N N N N N Ru Ru N N N N N N O 35 24+ OR N N N N Ru N N O N N N O N Ru N N N NRu O N N N N O N N Ru N N N N OR 649 O N N N Ru N N N Cl OR OR650 Br N N Ru + N N Br Br HO Br Br N Br+ Br N Br HO Br 4+ O N N N N Ru N N N N 37O N N N N N Ru Ru N N N N N O 39 N N N Ru N N NO N N Ru N NNN N N N Ru N N NO N N N N Ru N N N Ru N N O N O ON N N Ru N N NO N N N Ru N N N 38 6+ N N N Ru 6 PF¡6N N N N N N Ru N N N O N N N Ru N N N N N N O Ru N N N N N N Ru O N N N N N N Ru N N N N N Ru O O N N O O N N N N O N O Ru N N N N N Ru N O N N N N N N Ru N N N N O N N N N N N Ru Ru N N N N N N O N N Ru N N N 40 I P Beletskaya, A V Chuchurjukin Scheme 2 12+ O N N N N Ru N N N N Scheme 3 36+ O N N N Ru N N N O N N N N Ru N N N N ON N N Ru N N N ON N Ru N NOSynthesis and properties of functionalised dendrimers The dendrimers 41 containing ruthenium atoms at the periph- ery and iron or cobalt atoms in the core were prepared on the basis of polypyridine ligands.50 N NN N RuN N O O N N N N O N M N N O N N RuN N O O N N N N RuN N 41 M=Fe, Co.The central fragment of the dendrimer 41 was synthesised by complexation of a substituted bipyridine ligand to iron(II) or cobalt(II) ions.50 The complex formed was introduced into reac- tion with a terpyridine ruthenium complex. Studies of electro- chemical properties of the dendrimers revealed that a reversible redox process Ru(II)>Ru(III) can occur at +0.81 V; however, ButO ButO O O ButOO ButOO HN ButO O ButO O OBut O ButOO ButO O HN ButO O ButO O ButO 14+ N N N N RuN N O O N N N N O Ru O N N N O O N N N N RuN N ButO O OBut O O OBut NH O NH O NH O O N N HN O Ru N N N NH O N N HN N Ru N N N HNO NH HN O O O O HN O O OBut O O ButO OBut no apparent redox reaction involving iron or cobalt ions in the dendrimer core could be detected.50 The incorporation of four bulky carborane groups at the periphery of the pentaerythritol-based ruthenium-containing den- drimer (complex 42) 51 resulted in restrictions of the conforma- tional mobility, which was confirmed by magnetic non- equivalence of four methylene groups in the core and of two methyl groups at the periphery of the dendrimer.This had no effect on the electrochemical characteristics of the polypyridine ruthenium complexes of the type 42.ButMe2Si ButMe2Si The dendrimer 43 containing four ruthenium complexes with bipyridine ligands has been described.52 The NMR spectra of this dendrimer are extremely complex; therefore, its structure was established on the basis of UV spectroscopy, MALDI-TOF mass spectrometry and cyclic voltammetry. N N O NH N O O O O O HN O N N N Ru N N N 43 651 8+ N N N N N N SiMe2But Ru Ru N N N N N N O OO O N N N N N N Ru Ru SiMe2But N N N N N N 42 8+ ButO OBut O ButO O O OBut O OBut O O OBut NH O NH O O O O OBut O NH HN OBut O O NH OBut N N Ru N N O HN O OBut O O O OBut HN O HN O OBut O NH O HN OBut O ButO O O OBut OBut O OBut O ButO652 O O O HO O K2CO3 + 18-C-6 OH Br O HO O O 45 O 44a,b (G27X) 44a: X=OH CBr4/PPh3 44b: X = Br VI. Dendrimers containing complexes with metal ± carbon s-bonds The convergent approach was used 53 in the synthesis of the dendrimers 44 with 3,5-dibenzyloxybenzyl groups at the periphery (Scheme 4).3,5-Dihydroxybenzyl alcohol (45) has been used as the monomer. It was noted 54 that the advantage of the convergent strategy is that it allows accurate monitoring of the number and localisation of functional groups in the final product. The inter- mediate products were isolated, purified and characterised after each step. Since all the intermediate and final products are easily soluble, the course of the reaction was monitored by NMR spectroscopy.Thus, any potential defects which might occur in the synthesis of dendrons could be detected prior to their attach- ment to the central fragment of the dendrimer. The organoruthenium compounds of the type 46 were syn- thesised in a similar fashion.54 The resulting metal-containing dendrimers were rather stable, which made it possible to synthe- sise larger dendritic structures containing Ru7C p- and s-bonds. CO CO CO Ru Ru OC O O CO CO CO Ru OC Ru O O O O Br 46 This approach was used in the synthesis of third- and fourth- generation dendrimers containing up to 48 ruthenium atoms.54 All the organoruthenium compounds thus prepared are readily soluble in common organic solvents. A first-generation organosilicon dendrimer containing 12 peripheral organopalladium units was synthesised by oxidative coupling of the [Pd2(dba)3 .dba/TMEDA] complex (dba is diben- zylidenacetone) to a dendrimer containing iodoarene groups. Further reaction of the metal-containing dendrimer with MeLi and bipyridine 55 resulted in the dendrimer 47. The oxidative addition of benzyl bromide to the Pd-contain- ing dendrimer 47 was studied. The stability of the newly formed OO 45 OO X K2CO3, 18-C-6 O O dendrimer containing organic derivatives of Pd(IV) and the main pathways of its decomposition were determined. N N Pd OO Pd N N O Pd N N N Pd N The synthesis of organoplatinum dendrimers of the type 48 has been described.56 This involved the attachment of benzyl halides (RX) to [PtMe2(NN)] (where NN is the diimine ligand, e.g., 2,20-bipyridine).The reaction occurs with quantitative yields; however, the synthesis of larger dendrimers is complicated, since the solubility of the reaction products decreases as the size of the molecule increases. This reaction was followed by a colour change from orange-red (PtII complexes) to pale yellow (PtIV complexes). The increase in the chain length was repeated until NMR spectro- scopy showed that the products contained structural defects. Larger organoplatinum dendrimers were obtained if the ligands were modified by introduction of tert-butyl groups, which increased the solubility of the dendrimers.57 This synthesis employed a convergent strategy.The [PtIV 2 Me4(m-SMe2)2] com- plex and 4,40-bis(bromomethyl)-2,20-bipyridine were used as the monomer units. The dendrimer 48 containing 14 platinum atoms was the last member of the series that has been synthesised. An attempt at preparing a dendrimer of the next generation contain- ing 30 platinum atoms failed. I P Beletskaya, A V Chuchurjukin Scheme 4 O O O O 45 etc. X O K2CO3, 18-C-6 O O 44 (G37X) O X= OH CBr4/PPh3 X=Br N N Pd N N N Pd Pd N O O O N Si N Pd Si Si Si O Si Si Si Si Si N Si Si Si O Pd N Si Si Si Si O Si Pd N O O N O Pd N N Pd N N 47Synthesis and properties of functionalised dendrimers But But But But But But But But N N N N N N N N Br Pt Br Pt Pt Br Br Pt N N N N N N Br Pt Pt Br N N Pt Pt N Br N Br Br Pt Br Pt N N N N Br Pt Pt Br Br Pt Br Pt N N N N N N N N But But But But But But But But 48 More recent studies 58 were aimed at investigating the possi- bility of synthesising multinuclear organopalladium compounds and bimetallic palladium ± platinum dendrimers by the conver- gent approach.These studies showed that in this case the use of organopalladium compounds is limited because of the tendency of palladium(IV) complexes to undergo a reverse reaction, viz., to the reductive elimination of alkyl halide. VII. Dendrimers containing terminal ferrocenyl groups The majority of iron-containing dendrimers reported so far contain iron in the form of ferrocenyl groups.This can be explained by the fact that the chemistry of ferrocenes is well- studied and its derivatives are rather stable compounds. The synthesis of such systems aimed at the construction of dendrimers which could undergo redox reactions involving electron interac- tions between the metal-containing fragments.59 The electronic interaction between the transition metal atoms within one of such dendrimers (49) has been studied; its structure is shown below. Fe Fe Fe Si Si Fe Si Si Fe Si Fe Si Si Fe Si Fe Si Fe Si Fe Si Si Si Fe Fe Si Si Fe Si Fe Si Fe Fe 49 Polymers containing ferrocene derivatives are currently used for modification of electrodes and redox catalysts and biosen- sors.60 The cation [FeCp(arene)]+ can be used as a starting material in the synthesis of dendrimers by the polyalkylation reaction.[FeCp(C6Me6)]PF6 reacts with an excess of a base and ferrocenylalkyl halide I(CH2)4Fc (Fc is C5H5FeC5H4) resulting in selective hexasubstitution to give {FeCp[C6(CH2)5Fc]6}PF6. According to cyclic voltammetry data, all the six ferrocenyl groups are electrochemically equivalent, being oxidised at the same potential. The positive charge in the central fragment of the molecule has no effect on the oxidation process.60 The dendrimer containing peripheral ferrocenyl groups can form inclusion compounds with cyclodextrins (see, e.g., com- pound 50) 61 localised at its periphery where they form a supra- molecular structure of an even greater size.Fe Fe O O NH NH Fe N HN HN N O N N O N N NH HN Fe HN HN O O Fe Fe 50 �b-cyclodextrin. The number of ferrocenyl groups incorporated into cyclo- dextrin molecules is limited by steric hindrances on the surface of the dendrimer. In the case of the dendrimer 50, the reaction between the ferrocenyl groups and cyclodextrin is complete. The synthesis of an unusual dendritic molecule 51, which simultaneously incorporates Fe and Pt atoms (the platinum frag- ments in the centre and the ferrocenyl groups at the periphery), has been described.62 Fe Fe Fe Fe N N Fe N Pt N Pt N Pt N Fe N Pt N Pt N Pt N N N Fe Fe Fe 51 653 Fe OO FeFe Fe Fe654 In this case, the dendritic molecule was assembled from fragments containing one platinum atom and two ferrocenyl units, which were attached to the central fragment, viz., hexa- kis(bromomethyl)benzene.This dendrimer was characterised by electrochemical and spectroscopic methods. Dendrimers of the type 52 containing peripheral ferrocenyl groups linked to the central fragment through the amide groups have been used as sensors in the determination of inorganic anions (Cl7, NO¡3 and HSO¡4 , see Ref. 63). Cyclic voltammograms of such dendrimers have the forms of simple reversible waves, which suggests that all of the ferrocenyl groups are equivalent in the oxidation process of Fe(II)/Fe(III).63 In this case, the `sensitivity' of the dendrimers to the nature of the anion in the redox process increases on going from zero- (G0) to the first-generation den- drimers (G1), reaching a maximum for a second-generation dendrimer (G2) preceding the third-generation dendrimer (G3) where steric saturation of the surface with the ferrocenyl groups takes place.With an increase in the generation number, the ferrocenyl groups on the surface of the dendrimer become more closely spaced; this results in a decrease in the sizes of the cavities where the anions could penetrate. Fe Fe CO HN Fe HN N CO Fe COHN N HN CO Fe CO NH Fe An optimal dendrimer for smaller anions is that belonging to a generation in which the number of ferrocenyl groups precedes steric saturation of the surface, which ensures the most suitable bulk of the cavities.63 VIII.Dendrimers containing transition metal p-complexes One of the ways of modification of dendrimers containing arene groups at the periphery consists in the formation of p-arene complexes of transition metals. Thus, the synthesis of first- and second-generation organosilicon dendrimers (e.g., the dendrimer 53) containing aromatic rings with the Cr(CO)3 groups linked has been described.64 Fe Fe Fe OC CO NH HN HN CO OC NH N N O O O O O O O O O N N N NH HNCO OC HN NH CO CO Fe Fe Fe Fe 52 Fe OCNH Fe HN N CO CO NH Fe N NHCO Fe HN COFe I P Beletskaya, A V Chuchurjukin (CO)3Crr(CO)3 Si Si Si Si Si Si Si Si Si Si Si (CO)3Cr Si Si Cr(CO)3 53 The starting organosilicon dendrimer was obtained by stepwise hydrosilylation and allylation.To this end, tetravinylsi- lane was introduced into the hydrosilylation reaction with methyl- dichlorosilane; the product formed was allylated with allylmagnesium bromide and repeatedly hydrosilylated with phe- nyldimethylsilane. The complexation involved only four out of eight benzene rings. The attempts to prepare a dendrimer with eight chromium atoms failed, since this required more drastic reaction conditions which resulted in the decomposition of the dendrimer. Some first-generation organosilicon dendrimers containing peripheral alkyne groups further converted into the cluster p-complexes 54a and 54b by a reaction with Co2(CO)8 have recently been synthesised.65 Co(CO)3 (OC)3Co O Si O Co(CO)3 Co(CO)3 Si O Si O O Si O O Co(CO)3 Co(CO)3 Si O Co(CO)3 (OC)3Co 54a SiMe3 Co(CO)3 (OC)3Co Si Co(CO)3 O Si O (CO)3Co Si SiMe3 Co(CO)3 Si O O Me3Si(CO)3Co Si Co(CO)3 (OC)3Co 54b SiMe3Synthesis and properties of functionalised dendrimers The organometallic dendrimer 55 containing six peripheral rhenium atoms was synthesised from the 3,5-dihydroxybenzyl alcohol using the convergent approach.66 (CO)3Re O Me O O O (CO)3Re O O O Re(CO)3 (CO)3Re 55 However, an attempt at using this compound as a starting material for the synthesis of a dendritic oxo-rhenium complex by conversion of the carbonyl groups, which could be used as a catalyst in oxidation reactions, was unsuccessful. IX.Dendrimers containing central metal- containing fragments The syntheses of dendrimers containing chiral titanium complexes in the centre of the molecule have recently been reported.67,68 In these studies, (R,R)-a,a,a0,a0-tetraaryl-1,3-dioxolane-4,5-dime- thanol (TADDOL) has been used as the core to which four dendrons prepared on the basis of 3,5-dihydroxybenzyl and/or octyl fragments were attached by the benzylation reaction.67 The catalytic activity of titanium complexes of the type 56 was studied in addition of diethylzinc to benzaldehyde. A slight O O O O O O O O O O O O O O O O O O O O O Ti(OPri)2 O O O O O O O OO O O O O O O O O O O O O 56 decrease in the catalytic activity and stereoselectivity with the increase in the number of dendrimer generations from zero to third was observed.However, on going to the fourth-generation dendrimers the decrease in the catalytic activity was more pro- nounced, being nearly two-fold. These results leave hope that second- and third-generation dendrimers can have possible appli- cations as catalysts in membrane reactors. Re(CO)3 OO Re(CO)3 O O O O O O OO O O OO O O OO O O O O O O Analogously, dendrimers based on 1,3,5-triethynylbenzene with a central optically active 1,10-bis(2-naphthol) fragment were prepared. 68 But But But But But But But But But But But But But ButBut But But But But But But These dendrimers present special interest as chiral ligands and are used to recognise enantiomers. Catalytic activity of the titanium complex 57 was studied in the addition of diethylzinc to aldehydes, the enantiomeric excess attained being 89% ± 90%. Yet another possible application of these dendrimers is the construc- tion of a fluorescent probe for optical determinations of enantio- meric compositions of chiral compounds.A dendrimer containing phthalocyanine as the central frag- ment and its cobalt complex 58 have been described.69 Studies of the electrochemical properties of the dendrimer 58 showed that the dendritic shell prevents intermolecular interactions of the phthalocyanine fragments and thus impedes the electron transfer to the centre of the molecule.In the oxidation of 2-mercaptoetha- nol with oxygen, the catalytic activity of the dendrimer was 20% R O RO NH O O O R N N MeO Co N R NH O N O N O O R MeOHN O R R 58 R=7NHC[CH2CH2CONHC(CH2CH2COOBut)3]3. 655 But But But But But But Ti(OPri)2 OO But But But But But 57R O OMe R O O N O O R HN N OMe N R O O O OR O656 RO O O RO ROO RO O O O RO OO RO O RO O O lower than that of the analogous non-dendritic phthalocyanine catalyst but its stability was higher in comparison with the latter.A great number of papers are devoted to the synthesis of dendrimers containing metal-porphyrin complexes as the central fragments. Such compounds present interest as selective oxidation catalysts mimicking enzymes.70 The dendritic shell enhances the regioselectivity of the catalyst thus increasing its similarity to enzymes. In addition, such compounds were used in model studies of electron transfer in biological objects 71 and as fluorescent probes.72 Yet another possible application of dendrimers is the use of their shells as protectors for the cores. The ruthenium(II) complex 59 coordinated to three dendritic bipyridyl ligands provides an example.73 It was shown that the lifetime of this complex in the excited state exceeds that of its non-dendritic analogue, since the dendritic shell protects the central part of the dendritic molecule from contacts with a solvent and dissolved oxygen and thus prevents luminescence quenching.The organoerbium dendrimers of the first (60) and second (61) generations 74 were synthesised in the reaction of Er[N(SiMe3)2]3 with phenylethynyl dendrons. Er 60 OR RO RO OO RO RO O RO O O O O O RO O O O O RO NH NH O O O O RO O O O O O O O NH O O O RO OR HN O N O O NH O O O Ru2+ O HN O NH O N O O O N NH O O O NH O OR O O O RO HN O O O OO HN RO O NH O O O ORO O O O O RO O OR RO O O O OO RO OR OR RO 59 I P Beletskaya, A V Chuchurjukin RO RO RO O OR O OO O O OR O O O O HN ORO NH OO OO O OR O O O O O NH O O O ORRO OR HN OO O N O OR O OR OO HNO OO O O OR O HN O HN O O N O OR O O O NH N O OR O O O O O HN OR O OR O O O OR HN O O O O O O OR HN O HN O O O OR O O O O OR O O O O OO RO RO OR RO Er 61 The starting phenylethynyl dendrons of the first and second generations were synthesised by using the convergent approach developed earlier for phenylethynyl dendrons with peripheral 3,5- di-tert-butylphenyl groups.75 According to the computer simula- tion data, phenylethynyl dendrons represent rigid planar mole- cules which apparently do not provide sufficient shielding of the central erbium atom despite their considerable sizes.For example, the dendrimers 60 and 61 synthesised are as sensitive to atmos- pheric oxygen and moisture as other compounds of the organo- lanthanide series.74 Lanthanide derivatives are widely employed as lumino- phores.76 However, the quantum yields of luminescence of lan- thanide cations decrease as a result of cluster formation. In order to prevent luminescence quenching, lanthanide cations are encap- sulated into a dendritic shell.76, 77 To this end, lanthanide triace-Synthesis and properties of functionalised dendrimers O O O O O O O O O O O OO O tates are introduced into a reaction with dendrons prepared on the basis of 3,5-dihydroxybenzyl fragments containing carboxy groups.The quantum yields of luminescence of the lanthanide complexes thus prepared (62) increased with the increase in the generation number for both solid samples and strongly diluted solutions of dendrimers in toluene. This effect can be attributed to the shielding of the lanthanide cations by dendritic ligands and the `antenna effect', i.e., the transfer of photoexcitation energy from the dendritic ligands to the lanthanide cations in the centre of the + R N N O2 R N R N 2MeCN Cu N CH2Cl2, 7788C N R 63 OMe O OMe R= , ,OMe , OMe O G0 OMe G1 O O O O O O O O O O O O O O O O Ln O O O O O O O O OO O O O O O O O O 62 2+ R R N O Cu Cu R N O N R R 64 MeO OMe O OMe O O OMe O O OMe O OMe G2 MeO O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O dendrimers despite the fact that dendrons based on 3,5-dihydr- oxybenzyl halides are not conjugated molecules.Synthesis of a series of dendrimers of the first, second and third generations the nuclei of which represent monovalent copper complexes with 1,4,7-triazacyclononane have been described.78 The rate constants for the oxidation of the dendritic complexes 63 synthesised with gaseous oxygen at 778 8C leading to the dendritic complexes 64 have been measured (Scheme 5). The oxidation rate constant for the second-generation dendrimer 63 OMeOMe MeO O O O O OO O O O O MeO OMe OMe 657 OO O O O O OO O O O O O O O O O Scheme 5 OMe OMe OMe O OMe O OMe .OMe O OMe O OMe OMe OMe G3658 OO OO OO (R=G2), k2=1.361072 litre mol71 s71, appeared to be two orders of magnitude lower than that of the first-generation dendrimer (R=G1), k2=1.39 litre mol71 s71. The third-gener- ation dendrimer (R=G3) was resistant to oxidation under identical conditions. With an increase in temperature to710 8C, the complexes 64 are decomposed as a result of intramolecular oxidation with the cleavage of the C7N bond. The half-life of the non-dendritic complex 64 (R=G0) is 7 s, whereas those of the first- (R=G1) and second-generation (R=G2) dendrimers are 24 and 3075 s, respectively, which seems to be due to steric hindrances in the transition state of the decomposition reaction created by the dendritic substituents.This opens a way to the synthesis of thermally stable bimetallic complexes mimicking the effects of enzymes. A method for encapsulation of noble metal microparticles in the core of polyamidoamine dendrimers (PAMAM) has been proposed.79 ± 81 To obtain such materials, Pd(II) or Pt(II) cations were adsorbed by a polyamidoamine dendrimer containing hydroxy groups in the centre of the molecule with subsequent reduction with NaBH4.79 Such dendrimers contain microparticles of noble metals in their central fragment and are readily soluble in water, which makes it possible to conduct organic reactions (e.g., hydrogenation of alkenes) involving these species in the aqueous phase.79 The solubility of such catalysts in non-polar solvents can be increased through the use of surfactants, e.g., fatty acids, which form salts with the surface amine groups of the dendrimer.80 The use of perfluorinated carboxylic acids in biphasic systems con- taining perfluorinated solvents presents special interest, since this allows easy separation of the catalyst from the reaction product after the termination of the reaction.81 Metallodendrimers of the zero, first, second, third and fourth generations which contain Fe4S4(S2¡ 4 ) as the dendrimer core and O O O O O O O O O O O MeN O O S Fe S S Fe S O O MeN O O O O O O O O O 65 (G3) I P Beletskaya, A V Chuchurjukin 27 O O O O O O O O O O O O NMe O O O S S Fe Fe S O S O O NMe O O O O O O O O O O O O can exist in two stable oxidation states, which can be used for storage of information, have been synthesised.82 The dendritic shell in compound 65 prevents possible electron transfer between closely arranged molecules, which would result in the loss of stored information.82 Voltammetric studies demonstrated that the electron transfer to the core of the molecule becomes more difficult with an increase in the number of the dendrimer gen- erations.82 The iridium complex with dendritic phosphine ligands 66 was used for reversible binding of fullerene C60.83 The complex 66 was O O O O O O PPh2 + CO Ir Cl PPh2 C60 O O O O O O 66Synthesis and properties of functionalised dendrimers O O O O O O PPh2CO Ir Cl PPh2 O O O O O O prepared by the reduction of Na3IrCl6 with carbon monoxide with subsequent attachment of dendritic phosphine ligands. On addi- tion of fullerene to the bright yellow solution of the dendrimer 66, the solution of the dendrimer ± fullerene complex changes its colour to dark green.References 1. A M Muzafarov, E A Rebrov, V S Papkov Usp. Khim. 60 1596 (1991) [Russ. Chem. Rev. 60 807 (1991)] 2. M N Bochkarev,M A Katkova Usp. Khim. 64 1106 (1995) [Russ. Chem.Rev. 64 1035 (1995)] 3. J-P Majoral, A-M Caminade Chem. Rev. 99 845 (1999) 4. M A Hearshaw, J R Moss Chem. Commun. 1 (1999) 5. A Archut, F VoÈ gtle Chem. Soc. Rev. 27 233 (1998) 6. H-F Chow, T K-K Mong, M F Nongrum, C-W Wan Tetrahedron 54 8543 (1998) 7. C Gorman Adv. Mater. 10 295 (1998) 8. H W I Peerlings, E W Meijer Chem. Eur. J. 3 1563 (1997) 9. G van Koten, J T B H Jastrzebski J. Mol. Catal. A. 146 317 (1999) 10. F J Stoddart, T Welton Polyhedron 18 3575 (1999) 11. M Fischer, F VoÈ gtle Angew. Chem., Int. Ed. Engl. 38 884 (1999) 12. I Cuadrado,M Mora n, C M Casado, B Alonso, J Losada Coord. Chem. Rev. 193 ± 195 395 (1999) 13. M Osawa, H Sonoki,M Hoshino, Y Wakatsuki Chem. Lett. 1081 (1998) 14. M Osawa, M Hoshino, S Horiuchi, Y Wakatsuki Organometallics 18 112 (1999) 15.D Tzalis, Y Tor J. Chem. Soc., Chem. Commun. 1043 (1996) 16. Z Xu, J S Moore Acta Polym. 45 83 (1994) 17. S Takahashi, Y Kuroyama, K Sonogashira, N Hagihara Synthesis 627 (1980) 18. C Devadoss, P Bharathi, J S Moore J. Am. Chem. Soc. 118 9635 (1996) 19. Z Zhu, J S Moore J. Org. Chem. 65 116 (2000) 20. T Kaneko, T Horie,M Asano, T Aoki, E Oikawa Macromolecules 30 3118 (1997) 21. S Kotha, K Chakraborty, E Brahmachary Synlett 1621 (1999) 22. E C Constable, C E Housecroft, B Krattinger, MNeuburger, M Zehnder Organometallics 18 2565 (1999) 23. N Ohshiro, F Takei, K Onitsuka, S Takahashi J. Organomet. Chem. 569 195 (1998) 24. A M McDonagh,M G Humphrey,M Samoc, B Luther-Davies Organometallics 18 5195 (1999) 25.W T S Huck, F C J Mvan Veggel, D N Reinhoudt Angew. Chem., Int. Ed. Engl. 35 1213 (1996) 26. W T S Huck, L J Prins, R H Fokkens, N M M Nibbering, F C J M van Veggel, D N Reinhoudt J. Am. Chem. Soc. 120 6240 (1998) 27. I P Beletskaya, A V Chuchurjukin, H P Dijkstra, G P M van Klink, G van Koten Tetrahedron Lett. 41 1075 (2000) 28. I P Beletskaya, A V Chuchurjukin, H P Dijkstra, G P M van Klink, G van Koten Tetrahedron Lett. 41 1081 (2000) 659 29. I P Beletskaya, A V Chuchurjukin, I L Eremenko, H P Dijkstra, G van Klink, G van Koten J. Organomet. Chem. (2000) (in the press) 30. W T S Huck, B Snellink-Ruel, F C J Mvan Veggel, D N Reinhoudt Organometallics 16 4287 (1997) 31. A W Kleij, H Kleijn, J T B H Jastrzebski,W J J Smeets, A L Spek, G van Koten Organometallics 18 268 (1999) 32.JW J Knapen, A Wvan der Made, J C de Wilde, PWN Mvan Leeuwen, P Wijkens, D M Grove, G van Koten Nature (London) 372 659 (1994) 33. M Albrecht, R A Gossage, A L Spek,G van Koten Chem. Commun. 1003 (1998) 34. M Albrecht, R A Gossage,M Lutz, A L Spek, G van Koten Chem. Eur. J. 6 1431 (2000) 35. M Bardaji, A -M Caminade, J-P Majoral, B Chaudret Organometallics 16 3489 (1997) 36. M Bardaji, M Kustos, A-M Caminade, J-P Majoral, B Chaudret Organometallics 16 403 (1997) 37. P Lange, A Schier, H Schmidbaur Inorg. Chem. 35 637 (1996) 38. P Lange, A Schier, H Schmidbaur Inorg. Chim. Acta 235 263 (1995) 39. K Kollner, B Pugin, A Togni J. Am. Chem. Soc. 120 10274 (1998) 40.A Togni Angew. Chem., Int. Ed. Engl. 35 1475 (1996) 41. D de Groot, E B Eggeling, J C de Wilde, H Kooijman, R J van Haaren, A W van der Made, A L Spek, D Vogt, J N H Reek, P C J Kamer, P W N M van Leeuwen Chem. Commun. 1623 (1999) 42. S C Bourque, F Maltais, W-J Xiao, O Tardif, H Alper, P Arya, L E Manzer J. Am. Chem. Soc. 121 3035 (1999) 43. S C Bourque, H Alper J. Am. Chem. Soc. 122 956 (2000) 44. V Balzani, S Campagna, G Denti, A Juris, S Serroni, M Venturi Acc. Chem. Res. 31 26 (1998) 45. E C Constable, A M W Cargill Thompson J. Chem. Soc., Dalton Trans. 3467 (1992) 46. E C Constable, A M W Cargil Thompson, P Harveson, L Macko, M Zehnder Chem. Eur. J. 1 360 (1995) 47. G R Newkome, F Cardullo, E C Constable, C N Moorefield, A M W Cargill Thompson J. Chem. Soc., Chem. Commun. 925 (1993) 48. E C Constable, P Harverson J. Chem. Soc., Chem. Commun. 33 (1996) 49. E C Constable, P Harverson Inorg. Chim. Acta 252 9 (1996) 50. E C Constable, P Harverson,M Oberholzer J. Chem. Soc., Chem. Commun. 1821 (1996) 51. D Armspach, M Cattalini, E C Constable, C E Housecroft, D Philips J. Chem. Soc., Chem. Commun. 1823 (1996) 52. G R Newkome,A K Patri, L A Godinez Chem. Eur. J. 5 1445 (1999) 53. C J Hawker, J M J Frechet J. Am. Chem. Soc. 112 7638 (1990) 54. Y-H Liao, R Moss Organometallics 15 4307 (1996) 55. N J Hovestad, J L Hoare, J T B H Jastrzebski, A J Canty, W J J Smeets, A L Spek, G van Koten Organometallics 18 2970 (1999) 56. S Achar, J J Vittal, R J Puddephatt Organometallics 15 43 (1996) 57. S Achar, R J Puddephatt Angew. Chem., Int. Ed. Engl. 33 847 (1994) 58. G-X Liu, R J Puddephatt Organometallics 15 5257 (1996) 59. I Cuadrado, C Casado, B Alonso,M Moran, J Losada, V Belsky J. Am. Chem. Soc. 119 7613 (1997) 60. J-L Fillaut, J Linares, D Astruc Angew. Chem., Int. Ed. Engl. 33 2460 (1994) 61. R Castro, I Cuadrado, B Alonso, C M Casado, M Moran, A E Kaifer J. Am. Chem. Soc. 119 5760 (1997) 62. S Achar, C E Immoos, M G Hill, V J Catalano Inorg. Chem. 36 2314 (1997) 63. C Valerio, J L Fillaut, J Ruiz, J Guittard, J-C Blais, D Astruc J. Am. Chem. Soc. 119 2588 (1997) 64. F Lobete, I Cuadrado,C M Casado, B Alonso,M Moran, J Losada J. Organomet. Chem. 509 109 (1996) 65. K BruÈ ning, H Lang J. Organomet. Chem. 592 147 (1999) 66. I J Mavunkal, J R Moss, J Bacsa J. Organomet. Chem. 593 ± 594 361 (2000) 67. P B Rheiner, D Seebach Chem. Eur. J. 5 3221 (1999) 68. Q -S Hu, V Pugh,M Sabat, L Pu J. Org. Chem. 64 7528 (1999) 69. M Kimura, Y Sugihara, T Muto, K Hanabusa, H Shirai, N Kobayashi Chem. Eur. J. 5 3495 (1999) 70. P Bhyrappa, J K Young, J S Moore, K S Suslick J. Am. Chem. Soc. 118 5708 (1996) 71. P J Dandliker, F Diederich, J-P Gisselbrecht, A Louati, M Gross Angew. Chem., Int. Ed. Engl. 34 2725 (1995) 72. R Sadamoto,N Tomioka, T Aida J. Am. Chem. Soc. 118 3978 (1996)I P Beletskaya, A V Chuchurjukin 660 73. J Issberner, F VoÈ gtle, L De Cola, V Balzani Chem. Eur. J. 3 706 (1997) 74. M N Bochkarev,M A Katkova, S Ya Khorshev, N P Makarenko Izv. Akad. Nauk, Ser. Khim. 356 (1998) a 75. Z Xu,M Kahr, K L Walker, C L Wilkins, J S Moore J. Am. Chem. Soc. 116 4537 (1994) 76. M Kawa, J M J Fre chet Thin Solid Films 331 259 (1998) 77. M Kawa, J M J Fre chet J. Chem. Mater. 10 286 (1998) 78. M Enomoto, T Aida J. Am. Chem. Soc. 121 874 (1999) 79. M Zhao, R M Crooks Angew. Chem., Int. Ed. Engl. 38 364 (1999) 80. V Chechik,M Zhao,R M Crooks J. Am. Chem. Soc. 121 4910 (1999) 81. V Chechik, R M Crooks J. Am. Chem. Soc. 122 1243 (2000) 82. C B Gorman, B L Parkhurst,W Y Su, K-Y Chen J. Am. Chem. Soc. 119 1141 (1997) 83. V J Catalano, N Parodi Inorg. Chem. 36 537 (1997) a�Russ. Chem. Bull. (Engl. Transl

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (8) 661 ± 682 (2000) Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium K N Gavrilov, A I Polosukhin Contents I. Introduction II. P,N-Bidentate phosphorus-substituted ferrocenes III. P,N-Bidentate phosphorus-containing 4,5-dihydrooxazoles IV. Other groups of chiral P,N-bidentate ligands V. Conclusion Abstract. palla- and rhodium of synthesis the on data Published Published data on the synthesis of rhodium and palla- dium ligands P,N-bidentate active optically with complexes dium complexes with optically active P,N-bidentate ligands and and their catalysis asymmetric homogeneous in applications their applications in homogeneous asymmetric catalysis are are summarised the of nature the of effect The discussed.and summarised and discussed. The effect of the nature of the P,N- P,N- bidentate complexes metal the of structure the on compounds bidentate compounds on the structure of the metal complexes and and on Allylic examined. is catalysis in enantioselectivity on enantioselectivity in catalysis is examined. Allylic substitution, substitution, cross-coupling, catalysed hydrosilylation and hydroboration cross-coupling, hydroboration and hydrosilylation catalysed by by Rh P,N-bidentate active optically with complexes Pd or Rh or Pd complexes with optically active P,N-bidentate ligands ligands are field this of development the for prospects The considered. are considered. The prospects for the development of this field of of chemistry 186 includes bibliography The demonstrated.are chemistry are demonstrated. The bibliography includes 186 refer- refer- ences. I. Introduction A fundamental problem studied at the border of chemical disciplines is the coordination behaviour of bi- and polyfunctional ligands. This problem includes competition of ligands for metal atoms, donor ± acceptor interactions, isomerism of complexes, stereo- and regioselective synthesis of coordination compounds and specific complex formation in biologically important objects.1 Of particular interest are phosphorus-containing compounds, which play an extremely important role in the development of views on coordination bonds, the geometry and the structure of complexes and on the mutual influence of inner-sphere ligands as well as in metal complex catalysis.2, 3 Systems containing an electron-donating centre of a different nature, apart from the phosphorus centre, are being vigorously studied.P,N-Bidentate ligands are highly important. They simultaneously exhibit proper- ties of both soft and hard bases and promote redistribution of functions in a catalytic cycle, formation of bimetallic structures and a specific arrangement of the metal coordination sphere. Thus P,N-bidentate ligands can provide cis- or trans-chelation, bridging binding in either the `head-to-tail' or `head-to-head' fashions (in particular, in structures with metal7metal bonds) and P-mono- dentate coordination. The particular type of coordination K N Gavrilov S A Esenin Ryazan State Pedagogical University, ul.Svobody 46, 390000 Ryazan, Russian Federation. Fax: (7-091) 244 43 90. Tel. (7-091) 277 01 62. E-mail: chem@ttc.ryazan.ru AI PolosukhinANNesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, ul. Vavilova 28, 117813 Moscow, Russian Federation. Fax (7-095) 135 64 71. Tel. (7-095) 135 25 48. E-mail: davank@ineos.ac.ru Received 29 February 2000 Uspekhi Khimii 69 (8) 721 ± 743 (2000); translated by Z P Bobkova #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n08ABEH000559 661 662 665 669 678 depends, first of all, on the length of the bridge linking the donor atoms. Systems with two linking units are typical chelating agents; the bridging type of coordination is barely known for them, whereas three-unit species are capable of bridging coordination.Coordination of a P,N-system to a central atom gives rise to two moieties differing in kind, containing electron-donating centres with different steric and electronic characteristics. Being structur- ally asymmetric, P,N-bidentate ligands also possess clear-cut electronic asymmetry. In addition, both steric and electronic parameters of the electron-donating centres and the nature of the bridge linking them can be varied over wide limits. The features listed above are important in asymmetric catalysis in which very fine structural and electronic changes in the ligand nature can be significant. Therefore, it is quite natural that the greatest progress in the enantioselective catalysis of cross-coupling, allylation, hydroboration, hydrosilylation and other processes has been achieved by the use of metal complexes with P,N-bidentate chiral ligands.Data on the role of optically active P,N-bidentate ligands in the coordination chemistry and catalysis are reported in several monographs and reviews.4±10 However, the information pre- sented there concerns either a particular narrow group of com- pounds (for example, ferrocenyl systems 8) or a particular catalytic reaction. The present review is the first attempt to summarise and describe systematically the role of chiral P,N-bidentate ligands in coordination chemistry and asymmetric catalysis. Attention is concentrated on the studies published over the last 15 years, most of all, in the 1990s, which were the years of vigorous development of the field of chemistry in question.This review deals only with P,N-bidentate systems with no additional electron-donating centres (i.e. devoid of O, S, Se and As or additional N or P atoms), only rhodium and palladium complexes being considered. Although examples of the use of W(0),11 Au(I),12, 13 Ir(I) 14 and Pt(II) 15 complexes in enantioselective catalysis do exist, the leading position of the two above-mentioned metals is beyond doubt. We have classified the published data on the basis of the ligand structure. This approach appears expedient because it allows one to cover the achievements of both coordination chemistry and catalysis.Classification in terms of the types of organic reactions is more reasonable in considering the problems of enantioselective catalysis, whereas that accepted in coordination chemistry is based on the nature of the metal and the main initial metal complexes used. When these two aspects are discussed simulta- neously, the nature of the P,N-bidentate ligand is a unifying point. Phosphorus-substituted ferrocenes and dihydrooxazoles are the subjects of separate Sections. This is due to the outstanding role of662 these derivatives in asymmetric catalysis. Ligands of this type ensure high yields and high enantioselectivity of processes and provide the possibility of investigating some aspects of the catalysis mechanisms, for example, the role of the planar chirality of the ferrocenyl fragment.Phosphorus-substituted dihydrooxa- zoles represent a specific group of ligands developed for highly efficient allylation. The change in the steric and electronic param- eters of the sp2-hybridised donor nitrogen atom incorporated in the five-membered ring influences the properties of the ligand and the catalytic activity of its metal complexes. The rather patchy group of other ligands is classified according to the nature of the electron-donating nitrogen centre; compounds with the phosphite donor centre are distinguished as a separate group. In addition, in arranging the information, we attempted, whenever possible, to group studies of the most productive scientific schools chronolog- ically and to consider comprehensively the most important catalytic processes.II. P,N-Bidentate phosphorus-substituted ferrocenes 1. Ligands with sp3-hybridised electron-donating nitrogen atoms. Aminoalkylferrocenylphosphines In 1974, in the very beginning of the development ofmetal complex asymmetric catalysis, a publication appeared 16 dealing with the synthesis of (S,R)-1-(1-dimethylaminoethyl)-2-diphenylphosphino- ferrocene [(S,R)-1, (S,R)-PPFA] { and (R,S)-1-(1-dimethylami- noethyl)-2-dimethylphosphinoferrocene [(R,S)-2, (R,S)-MPFA], which were the first representatives of chiral P,N-bidentate phos- phorus-containing ligands. PMe2 NMe2 NMe2 Me Me Fe Fe PPh2 (R,S)-2 (S,R)-1 The novel ligands were tested 16 in enantioselective hydro- silylation of a series of prochiral ketones.The best result (yield 89%, ee 49%) was obtained in the reaction of acetophenone with diphenylsilane in the presence of [Rh(C6H10)Cl]2 . 4L, where L=(R,S)-MPFA [(R,S)-2]. 1) Cat 2) H3O+ PhCH(OH)Me PhCOMe+Ph2SiH2 Cat=[Rh(C6H10)Cl]2 . 4 (R,S)-MPFA. The compound (R,S)-2 was found to be markedly more efficient than the best mono- and bidentate phosphorus-contain- ing ligands known at that time, namely, (R)-PhCH2(Ph)(Me)P* and 2,3-O-isopropylidene-2,3-dihydroxy-1,4-bis(diphenylphos- phino)butane [(7)-DIOP].16 It was also found that replacement of the dimethylamino group in the side chain of the ligand (R,S)-2 by a hydrogen atom results in a sharp decrease in the stereo- selectivity.These studies opened up a new line in the search for bidentate chiral ligands for metal complex catalysis. The research- ers used simultaneously two fruitful concepts, namely, combina- tion of two different electron-donating centres (phosphorus and nitrogen) and two types of chirality (central and planar) in the molecule of an optically active ligand. PPFA [(S,R)-1] has been modified, for example, by replacing the methyl substituents at the nitrogen atom by Et, Pri, Bui, etc.17 P,N-Bidentate ligands have been assayed in other asymmetric catalytic processes. Cross- coupling of (1-phenylethyl)magnesium chloride with vinyl bro- { Hereinafter, the first character denotes the configuration of the ferro- cenyl core.K N Gavrilov, A I Polosukhin mide catalysed by the complex PdCl2[(S,R)-PPFA] gave (S)-3- phenylbut-1-ene in 82% yield (ee 61%).17, 18 PdCl2[(S,R)-PPFA] PhCHMgCl +BrCH CH2 PhCHCH CH2 Me Me The enantiomer of the ligand (S,R)-1, namely, (R,S)-PPFA, also as a complex with PdCl2, ensures an even higher enantiose- lectivity in this type of reaction.19 PdCl2[(R,S)-PPFA] Me3Si R1 R1 Me3SiCHMgCl+ Br Ph Ph R2 H R2 R1=Me, R2 = H (ee 85%); R1=Ph, R2=H(ee 95%); R1=R2=H(ee 96%). The bidentate ligand (S,R)-1 is able to induce not only central but also planar chirality.20 Cl Cl [Pd(All)Cl]2 . 2.4[(S,R)-PPFA] +RM Cl R Cr(CO)3 Cr(CO)3 (ee<44%) R=CH=CH2, C(Me)=CH2, CH=CHBu; M=SnBu3, MgCl, ZnCl, B(OH)2; All=CH2=CHCH2. For (S,R)-1 and for its close analogue having two tert-butyl substituents at the phosphorus atom, a number of rhodium(I) cationic complexes, including [(NBD)Rh(L)]+X7 (NBD is nor- bornadiene, X7=PF¡6 , ClO¡4 ), [(COD)Rh(L)]+Y7 (COD is cyclooctadiene, Y7=BF¡4 , BPh¡4 ), have been prepared.The complexes have been thoroughly studied by 1H NMR, 31P NMR and IR spectroscopy and conductometry. The structure of the complex {(NBD)Rh[(R,S)-PPFA]}+PF¡6 was established by X-ray diffraction analysis.21 The formation of a metal-containing chelate ring was demonstrated in all cases. The use of these complexes as catalysts of hydrogenation of N-acetyl-a-aminocinnamic acid ensures ee of up to 84%.21, 22 The further investigation into chiral phosphorus-containing ferrocenes has been mostly carried out by two research groups, namely, Weissensteiner's 23 ± 28 and Togni's 29 ± 39 groups.Thus Weissensteiner et al.23, 24 synthesised epimeric ferrocenyl P,N- bidentate ligands, (R,R)-PTFA 3a and (R,S)-PTFA 3b. H Me2N H Me2N Ph2P Ph2P Fe Fe 3b 3a In the model cross-coupling of (1-phenylethyl)magnesium chloride with vinyl bromide, the complex PdCl2[(R,R)-PTFA] provides an ee of 79% for a nearly quantitative chemical yield of 3-phenylbutene; in the presence of PdCl2[(R,S)-PTFA], the ee is only 4%. Thus, in this case, the process enantioselectivity is governed by the central chirality of the ligands 3a,b. This outcome contrasts sharply with the results obtained in the catalysis by nickel complexes, which gave close values of the optical yield in the presence of (R,S)- and (R,R)-PPFA.17 The structure of the catalyst PdCl2[(R,R)-PTFA] was studied by 1H, 13Cand 31PNMR spectroscopy.According to X-ray diffraction data, the Pd7P bond length [2.207(2)A] is close to the average values observed in cis-palladium(II) dichloride complexes with aminophosphine ligands.24 Subsequently, the palladium(II) complex {Pd(Z3- C4H7)[(R,R)-PTFA]}+Tf7 and the palladium(0) complex {Pd[(R,R)-PTFA]}L, where Tf=CF3SO3, L is maleic anhydride or dimethyl fumarate, have been studied in detail 25 using multi- nuclearNMRdata (including the INDOR and EXSY techniques)Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium and X-ray diffraction analysis.It was concluded that PTFA, which is conformationally more rigid than PPFA, possesses also stronger chelating properties. An attempt 26 to use (S,S)-PTFA as a ligand in asymmetric hydroformylation of styrene proved unsuccessful: for rhodium(I)-based catalyst, ee was 0.5%, while in the case of platinum(II)-based catalyst, it is 7%¡À 21%. This was explained 26 by the fact that the chelate angle formed by PTFA is rather small, especially in the case of complexation with platin- um(II). Enantiomeric ferrocenyl bidentate ligands 4a,b, related in structure to PPFA and PTFA, have been prepared.27, 28 Com- plexes of the ligand 4a with PdCl2 and the ligand 4b with PdClMe, PdMe2 and PdPh2 were synthesised.NMe2 Me2N H H PPh2 Ph2P Fe Fe (R,R)-4b (S,S)-4a The structures of the ligands 4a,b and complexes derived from them, which are promising in asymmetric catalysis, were studied by X-ray diffraction analysis and multinuclear NMR spectro- scopy including low-temperature procedures. Broadening of the resonance signals of the diastereotopic methyl groups in the palladium complexes of 4a,b suggests 28 that the energy barrier to the rupture of the Pd7N bond is low. 2. Ligands with sp2-hybridised electron-donating nitrogen atoms a. Ferrocenylphosphinopyrazoles Researches into ferrocene-based P,N-bidentate ligands have gained a new incentive due to the studies of Togni et al.29 ¡À 39 Having started with minor structural changes, i.e.introduction of one or five methyl groups into the unsubstituted ring of (S,R)- or (R,S)-PPFA,29 they used subsequently a fairly advantageous expedient, namely, they replaced the dimethylamino group by a pyrazole ring. This gave rise to a series of 1-diarylphosphino-2- [1-(pyrazol-1-yl)ethyl]ferrocenes 2 3 ¡À 5a,b 29 and 6a ¡À g.30 30 ¡À 32 R2 R R1 R3 R PPh2 N N PAr2 N N Me Me Fe Fe (R,S)-6a ¡À g (R,S)-5a,b R = H (a), Me (b) Ar R3 R2 Compound 6 R1 Me H Me MeH H PhPh H CF3 Ph Ph H Me Ph H PriPh Ph Ph Ph cyclo-C6H11 4-FC6H4 4-CF3C6H4 4-MeOC6H4 HMe CF3 Me CF3 Pri Me Ph Me Ph Me Me Me Me Me Me Me Me Me Me Br HNO2 HHHH abcdefghijklm 663 Ar R3 R2 Compound 6 R1 Me Me Me Me Me CF3 1-naphthyl 1-adamantyl cyclo-C6H11 Ph2-naphthyl 4-pyridyl 3-NO2C6H4 2,4-(MeO)2C6H3 9-anthryl 9-tripticyl H H Me 3,5-Me2C6H3 H Me 4-MeC6H4 H Me 4-PhC6H4 H Me 3,5-(CF3)2C6H3 H Me 4-NMe2C6H4 H CF3 4-MeOC6H4 H Me Ph H H Ph H Me Ph H H Ph H Me Ph H Me Ph H H Ph H Me Ph H Me Ph H Ph nopqrstuvwxyzabg The ligands 6a ¡À t were tested 30, 31 in the hydroboration ¡À oxidation of styrene with catecholborane.O 1) Cat BH PhCH CH2+ 2) H2O2, NaOH O PhCH(OH)Me+Ph(CH2)2OH Cat={[Rh(COD)2]+BF¡¦4 } ¡À (6a ¡À t). In some cases, the reaction gives a substantial amount of an achiral primary alcohol but, in general, the enantioselectivity is high.The optical yield was found to be clearly dependent on the electronic properties of substituents at the donor centres D electron-withdrawing substituents in the pyrazole nucleus decrease the optical yield, whereas substituents in the phenyl groups at the phosphorus atom increase the optical yield [the ee values for the formation of (R)-1-phenylethanol in the presence of the ligands 6c, e, a, l, q are 33%, 44%, 65%, 98% and 98.5%, respectively]. Rhodium(I) chloro carbonyl complexes 7a ¡À t were synthesised by reactions of 6a ¡À t with [Rh(CO)2Cl]2. They were characterised by the data of 1H, 31P (1JP7Rh=161 ¡À 176 Hz) and 13C NMR and IR spectroscopy (nCO=1980 ¡À 2008 cm71) and mass spectrometry.31 X-ray diffraction analysis of the compounds 7n, o was performed.31 Cl OC R1 Rh Ar2P N N R2 R3 Me FeCp 7a ¡À t 4 An acceptable correlation has been found between the nCO and ee values for the complexes 7a ¡À t; the frequency of the carbonyl group vibrations is affected by changes in the s-donor properties of the nitrogen centre and in the p-acceptor properties of the phosphorus centre.The cationic complexes [Rh(COD)(L)]+BF¡¦ have been prepared (L=6b, c) and also studied by multinuclear magnetic resonance and X-ray diffraction analysis. As in the complexes 7a ¡À t, in this case, too, a seven-membered chelate metallacycle is formed. The ligands 6t ¡À g have found application in a reaction significant from the synthetic standpoint, namely, allylic amina- tion 32 (the enantioselectivity reaches 99%).NHBn OCO2Et Cat +BnNH2 Ph Ph Ph Ph Cat=[Pd2(dba)3] . 3L, dba is dibenzylideneacetone; L=6t ¡À g.664 The reaction mechanism, including allylic intermediates [Pd(Z3-PhC3H3Ph)(L)]+PF¡6 , was studied in detail using X-ray diffraction analysis and 2D NMR. It was found that the attack of a nitrogen-containing nucleophile is directed at the carbon atom in the allylic ligand located in the trans-position relative to the phosphorus centre of the coordinated P,N-bidentate ligand. The activation barrier to the formation of the C7N bond in the trans- position relative to the phosphorus atom is 3 kcal mol71 lower than that for the trans-position to nitrogen.32 Calculation meth- ods 33 and conformational analysis 34 [using 31P, 13C and 1H(NOESY) data] of model allylic complexes have also been invoked to study the reaction mechanism and stereochemical features of the process.4 The ligand 6u, containing an adamantyl substituent in the pyrazole ring, proved to be the most efficient in allylic amination (ee 99%).32, 35 By the addition of coordinating anions, F7, BH¡ or OH7, in co-catalytic amounts, a virtually quantitative optical yield can be attained (>99.5%).36 Similar derivatives of ruthenocene 8a,b 37 and bimatallic systems 9a,b, in which the pyrazole ring contains a second ferro- or rutheno-cene fragment, have been synthesised.38 M PPh2 X N N Me Me Ru Fe PPh2(S,R)-9a,b (R,S)-8a,b M=Fe (a), Ru (b) Me Me (b) X=NMe2 (a), N N The efficiencies of the ligands 6u and 9a,b in the catalysed allylic amination are comparable (ee >99.5%).38 In catalytic hydroboration ± oxidation, the ligand 8b is somewhat inferior to the ferrocenyl analogue 5b (the optical yields for 100% conversion of styrene are 87% and 94%, respectively).37 Recently, 39 ligands 10a ± d, close analogues of the compounds 6a ± g, have been proposed.In the presence of these ligands, asymmetric hydrosilylation ± oxidation of norbornene has been carried out. R1 R2 N N Me Fe PAr2 (S,R)-10a ± d 1) Cat OH +HSiCl3 2) [O] Cat=PdCl2 . L, L=10a ± d. Ar R2 Optical yield of hydroxy- norbornane (%) Com- R1 pound 10 398291 3,5-(CF3)2C6H3 99 Ph 2,4,6-(MeO)3C6H2 2,4,6-Me3C6H2 2,4,6-Me3C6H2 abcd Me Ph H Ph H Ph H K N Gavrilov, A I Polosukhin In the case of the ligands 10, the optical yield increases on passing from 10a to 10d.It has been concluded 39 that optimisa- tion of the ligand regarding steric factors should be done with allowance for its electronic asymmetry because the s-basicity of the pyrazole fragment and the p-acidity of the phosphine fragment are important in hydroxylation ± oxidation, as well as in hydro- boration ± oxidation. b. Phosphinoferrocenyl-4,5-dihydrooxazoles Phosphinodihydrooxazole ligands will be discussed in Section III. This Section is devoted to structures related to the ligands 6a ± g but having a second chiral centre in the heterocyclic fragment.In particular, epimeric ligands of type 11a ± c have been synthes- ised.40 ± 43 They are designated by the common abbreviation DPOF. PPh2 O O R R N N Fe Fe PPh2 (R,S)-11a ± c (S,S)-11a ± c R=Pri (a), But (b), Me (c). The reaction of the compounds (S,S)- and (R,S)-11b with PdCl2(MeCN)2 gave complexes {PdCl2[(S,S)-11b]} and {PdCl2[(R,S)-11b]}; the latter was characterised by X-ray diffrac- tion analysis.40 These complexes were used as cross-coupling catalysts. Cat PhCH(Me)CH CHR PhCH(Me)MgCl+BrCH CHR R=H, Ph; Cat=[Pd(All)Cl2]L, L=(S,S)-11a, (S,S)-11c, (S,R)-11a. The complex {PdCl2[(S,S)-11a]} provides a higher enantiose- lectivity of this process than its epimer (ee 45% and 8%, respectively).The chemical yield is also substantially greater in the former case (74% and 23%, respectively). Thus, unlike the ligands (R,R)-PTFA 3a or (R,S)-PTFA 3b, in this case, the enantioselectivity is mainly due to planar chirality. In the classical alkylation of malonates with allyl esters catalysed by [Pd(All)Cl]2, the ligand (S,S)-11b ensures ee 90%; in the case of the ligand (S,S)-11c, the enantioselectivity is more than 99% (the chemical yield is nearly quantitative in both cases). In the presence of (R,S)-11b, ee is only 73%.42, 43 In this reaction, the crucial role is also played by the stereochemistry of the ferrocenyl core of the ligand. OAc CH(CO2Me)2 Cat Ph Ph Ph Ph+CH2(CO2Me)2 12 Cat=[Pd(All)Cl]2L, L=(S,S)-11b, (S,S)-11c, (R,S)-11b.The enantioselectivity of the reaction depends also on the steric requirements of the substrate. On passing from 1,3-diphe- nylprop-2-enyl acetate 12 to the methyl analogue 13, the optical yield sharply drops (it is not more than 36%). OAc CH(CO2Me)2 Cat Me Me Me Me +CH2(CO2Me)2 13 Cat=[Pd(All)Cl]2L, L=(S,S)-11b. Tetrasubstituted ferrocenyl ligands 14a,b, constructed as combinations of two fragments of the ligands 11a or 11b, have also been synthesised. 44Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium R Cl Cl N Pd N O PPh2 O PPh2 Fe Fe PPh2 O PPh2 O N N PdCl Cl 14a,b R 15 R=Pri(a), But (b). The compound 14a was converted into the palladium complex 15; unfortunately, this complex was only scantily characterised by spectroscopic data (data on 1H NMR and mass spectra were reported).Interestingly, the ligand prepared by replacement of the dihydrooxazole rings in the structure 14a by the CH(Me)NMe2 groups behaves as a P,P- rather than a P,N-bidentate ligand with respect to Pd(II) because the amino-group nitrogen atom is bound to palladium more weakly than the dihydrooxazole nitrogen.44 In the Pd(II)-catalysed alkylation of the allyl acetate 12, the ligands 14a,b provide ee 96%¡À 99%, the chemical yield being more than 90%; the epimer of compound 14a with the opposite planar configuration displays the same enantioselectivity. Thus, in this case, unlike the ligands 11, the planar chirality does not have a governing influence on the product configuration; this might be due to different reaction mechanisms. Ligand (S,S,S)-DIPOF 16 has been studied 14, 45 in the asym- metric reduction of a series of prochiral aromatic and aliphatic ketones with diphenylsilane {with [Rh(COD)Cl]2 as the pre- catalyst}.Ph O Ph N Fe PPh2 16 The chemical yields in these reactions were 45%¡À 100%, while the optical yields were 57%¡À 91%. The best results were attained in the reduction of acetophenone to (R)-1-phenylethanol. In the presence of (R,R,R)-DIPOF, the reaction affords (S)-1-phenyl- ethanol with ee 91%. The use of [Ir(COD)Cl]2 as the pre-catalyst increases the optical yield to 96%. In the hydrosilylation of acetophenone with diphenylsilane, the use of related dihydrooxazole systems, namely, arylideneami- noalkylferrocenylphosphines 17a ¡À d leads to commensurable results (ee 87%¡À 90% for chemical yields of 86%¡À 94%) 46 {with [Rh(NBD)Cl]2 as the pre-catalyst}.Ar N NH2 Me Me Fe Fe PPh2 PPh2 (S,R)-18 (S,R)-17a ¡À d Ar=Ph (a), 3-CF3C6H4 (b), 4-CF3C6H4 (c), C6F5 (d) The presence of the imino group in the ligand structure is of prime importance. Indeed, in the presence of the amine (S,R)-18, ee is 39%, while in the presence of (S,R)-PPFA [(S,R)-1], it is only 16%. However, the nature of the arylidene group barely influences the optical yield. 665 In conclusion, we would like to mention several other P,N- bidentate phosphorus-substituted ferrocenes which hold pros- pects as asymmetric catalysts. The first one is ligand 19, for which cationic complex 20 has been obtained.47 The P,N-biden- tate manner of binding to palladium(II) in the complex 20 was established by 1H, 31P and 13C NMR spectroscopy and X-ray diffraction analysis.+ N N [Pd(All)Cl]2 PF¡¦6Fe Pd P PPh2 CH2OMe NH4PF6, MeOH (S,S)-19 Ph Ph 20 Phosphaferrocenes 21a ¡À c, representing a new class of ligands with planar chirality,48 and analogous compounds 22,49 contain- ing peripheral aza heterocycles have been synthesised. X P Fe Fe R P 22a,b R=2-pyridyl, 8-quinolyl 21a ¡À c X=NMe2 (a), CH2NH2 (b), CH2NMe2 (c) A recent publication 50 describes the synthesis and resolution of diastereomers (S,S)-23 and (S,R)-23 with a chiral centre on the phosphorus atom. NMe2 NMe2 P P Fe Fe Me Me Ph Ph (S,R)-23 (S,S)-23 3 A complex of the compound (S,R)-23 with PdI2D[(1-MePhP- 2-Me2HN+CH2C5H3)FeC5H5)]PdI¡¦ D has been prepared.50 X-Ray diffraction analysis showed a P-monodentate binding to palladium of the ligand protonated at the nitrogen atom.The researchers suggested that the complex was formed upon opening of the chelate ring on treatment with the hydrogen iodide, resulting from decomposition of PdI2, followed by the reaction of I2 with traces of water. Hence, it was concluded that the six-membered metallacycle involving the ligand (S,R)-23 is relatively unstable. However, this statement seems arguable because lots of metal complexes formed by related systems, for example, PPFA 1, PTFA 3 or the ligands 4a,b are known.The wrong choice of the initial palladium(II) compound appears to be a more probable reason. Thus, the use of nitrogen-and-phosphorus-containing ferro- cenes in asymmetric metal complex catalysis makes it possible to attain more than 90% optical yields in some important reactions such as cross-coupling, hydroboration ¡À oxidation and hydro- silylation ¡À oxidation of alkenes, hydrosilylation of ketones, and allylic substitution. III. P,N-Bidentate phosphorus-containing 4,5-dihydrooxazoles This group of ligands includes compounds containing an sp2- hybridised donor nitrogen atom. In 1993, three research groups working independently, namely, Helmchen's group 51 from the Institute of Organic Chemistry of Heidelberg University, Williams' group 52 from the Technological University (Lawboro), and Pfaltz's 53 group from the Institute of Organic Chemistry of Basel University, reported almost simultaneously the synthesis and the use in enantiomeric catalysis of the first representatives of a new class of P,N-bidentate666 ligands D phosphorus-containing 4,5-dihydrooxazoles.There is an opinion,54 however, that Williams et al.55 came out with the main concepts a year earlier. Subsequently, the effort of the three groups mentioned above resulted in the synthesis of a large series of phosphorus-substituted dihydrooxazoles such as 24, 25 and 26a ¡À t.56 ¡À 76O PPh2 N Pri 24 O R P NAr1 26a ¡À t Ar2 Ar1 Compound 26 Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph 1-naphthyl 1-naphthyl 2-biphenylyl 2-biphenylyl 3,5-(CF3)2C6H3 cyclo-C6H11 abcdefghijklmnopqrst When ligands of the type 26 are employed in the model alkylation of dimethyl malonate with the allyl acetate 12 {with [Pd(All)Cl]2 as the pre-catalyst},51 ¡À 53, 56 optical and chemical yields reach 99%.The highest enantioselectivity is exhibited by the chiral substituted dihydrooxazoles 26b, c, e. Other nucleo- philic substitution reactions at the allylic carbon atom in the allyl acetates 27 were also performed with high enantioselectivity in the presence of these compounds. The results are summarised in Table 1.OAc R2 [Pd(All)Cl2]2 .L +Nu7 R3 R1 27 It should be noted that, in the presence of the epimeric compounds 26m,n, the corresponding derivatives are formed in the same optical yield (77%) and with the same (S) absolute configuration, i.e. the asymmetric induction is accomplished by the dihydrooxazole fragment.56 The distance between the elec- tron-donating centres is also significant D the ligand 24, which forms a five-membered metal-containing chelate ring, provides a moderate selectivity (ee 73%).51 Several palladium-catalysed intramolecular alkylation reac- tions have also been successfully performed;67 the best result was attained with the ligand 26e (ee 80% ¡À 87%). Enatioselective allylic substitution is utilised to prepare natural products includ- ing optically active starting compounds for the synthesis of a- and Ph O CH2OSiButMe2 N PPh2 25 R Ar2 Me Pri But Bui Ph CH2Ph 1-naphthyl CH2But CH2SBui CMe2SCH2Pri (CH2)2SMe (3-indolyl)methyl Pri Pri Pri Pri Ph Pri Pri Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph 1-naphthyl Ph 1-naphthyl Ph Ph 3,5-(CF3)2C6H3 cyclo-C6H11 Pri Nu R2 R3 R1 K N Gavrilov, A I Polosukhin b-amino acids and their derivatives.58, 66 A palladium(II) cationic complex was prepared from the compound 26b and characterised by 1H NMR spectroscopy (NOESY, ROESY) and X-ray diffrac- tion analysis.+ O SbF¡¦6N PPh2 Pd PriPh Ph Resorting to low-temperature 2DNMRspectroscopy made it possible to propose a catalytic cycle for the alkylation of dimethyl malonate with the acetate 12 in the presence of the ligand 26b.68 Ph PhCH(OAc) Pri P Ph 12 C H Pd N 26b Ph Nu (S) Pri Pri Ph Ph P P H Pd Pd N N Nu CH(OAc)Ph CPhPri P Ph Ph Pd Nu7 N exo-28 Nu=CH(CO2Me)2. The predominant (S)-enantiomer of the reaction product is formed via an attack by the nucleophile on the C(3) atom of the allylic fragment, located in the trans-position relative to the phosphorus centre, in the exo-28. Detailed investigation of the reaction mechanism using model reactions permitted ligands of the type 26 to be employed in other asymmetric allylic substitution processes.Thus ee values of more than 99.9% are attained in some cases in the alkylation of nitromethane with the allyl carbonates RCH=CHCH(R)O. .CO2Me (R=Me, Et, Ph).69 Among the ligands 26b,c,e,i,m,n,o tested {with [Pd2(dba)3] .CHCl3 as the pre-catalyst}, the best results were obtained for 26b,c,i. In the case of cyclic substrates 29, the best results were obtained for the ligands 26p,q.70 The epimeric compounds 26m,n provide close enantioselectivities. 7CH(CO2R)2 (CH2)n (CH2)n PdLm CO2R X 29a ¡À d RO2C n = 1 (a), 2 (b), 3 (c), 4 (d); X =OAc, OCO2Me, Br; R =Me, Et; m=1.5 ¡À 3; L=26b,m,n,p,q. Phosphorus-containing cymantrenyldihydrooxazoles 30a ¡À i and 31a,b 71 ensure better asymmetric induction in some cases {with [Pd(All)Cl]2 as the pre-catalysts}.In the reaction of the cyclic allyl acetates 29a,b,c (X=OAc) with the anion derived from dimethyl malonate in the presence of the ligand 30d, the optical yields are 96%, 93% and >99%, respectively. The configuration of the phosphorus centre is of fundamental importance in this case: in the presence of the epimeric ligand 30c, the reaction with the compound 29b results in ee of only 31%. Compounds achiral at phosphorus (includingChiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium Table 1. Enantioselectivity in allylic substitution in the compounds 27 in the presence of the ligands 26b,c,e. Compound 27 abcdefghia Pre-catalyst is [Pd2(dba)3] .CHCl3; b the corresponding phosphate was used; c Boc is ButOCO.30a,b) are relatively ineffective, the ee values being not more than 44%.71 ON R Ar1 30a ± i Compound 30 abcdefghi Ph2P ON R R=Pri(a), But (b) The ligand 30d was also found to be highly effective in related reactions. Thus, when 3-chlorocyclopentene reacts with the sodium salt 32, the reaction product 33 is formed in >94% yield and ee >98.5%.72 Cl R=Me (a), Et (b); L=30d. R3 R2 R1 Ph H Ph Ph Ph HH Et Me Pri Ph Ph HPh Ph P(O)(OEt)2 (see b) Me Ph Ph Ph Ph Mes H Ph Ph PriMn(CO)3 P Ar2 R Ar2 Ar1 Ph Ph 2-biphenylyl Ph 2-biphenylyl 3,5-Me2C6H3 2-biphenylyl 3,5-(CF3)2C6H3 2-biphenylyl Ph Ph Ph 2-biphenylyl 3,5-Me2C6H3 2-biphenylyl 3,5-(CF3)2C6H3 2-biphenylyl 2-biphenylyl Pri But But But But But But But But Mn(CO)3 31a,b CO2R [Pd(All)Cl]2 .2L +NaC OAc 33 32 CO2R L HNu 26b 26b CH2(SO2Ph)2 H2NSO2C6H4-4 CO 26b NH CO (Ph)2C=NCH2P(O)(OEt)2 (Ph)2C=NCH2CO2But CH2(CO2Me)2 H2NNHBz CH(NHAc)(CO2Et)2 HO2SBut HO2SBut HO2SBut CH2(CO2Me)2 CH2(CO2Me)2 H2NBn HN(Boc)2 c CH2(CO2Me)2 CH2(CO2Me)2 26b 26c 26e 26c 26e 26b 26c 26e 26c 26c 26c 26c 26b 26b 26b CH2(CO2Me)2 26b 26c CH2(CO2Me)2 CH2 (CO2Me)2A similar reaction with the N-substituted tetrahydropyridine acetate 34 results in a product yield of 84% and ee 98%.73 CO2Me OAc 32a, [Pd(All)Cl]2 .1.5L NBoc 34 L=30d. Compound 35 is a valuable optically active initial compound for the preparation of aza sugars and (R)-nicotinic acid. The use of phosphorus-containing dihydrooxazoles in the formation of the C7C bond according to the Heck reaction is equally effective. Thus the ligand 26c ensures a fairly good enantioselectivity in the reactions with triflates.58, 74, 75 Cat +PhOTf 36 O + TfO O O Cat O +36 O Cat +36 NCO2Me CO2R OAc CO2R Cat=[Pd(dba)2] . 2L, L=26c. 667 Ref. ee (%) Yield (%) 57 58 93 90 78 84 58 96 65 96 97 97 97 97 91 90 93 79 71 89 97 98 89 98 95 98 69 70 71 98 96 87 29 59 60 53 61 53 62 a 62 62 63 57, 63 61 61 64 ± 66 64 ± 66 64 ± 66 64 ± 66 53 96 88 OAc CO2Me NBoc 35 Ph O [87% (ee 97%)] Cat O [95% (ee 88%)] Ph O [70% (ee 92%)] Ph N [88% (ee 85%)] CO2Me668 In this reaction, phosphorus-substituted dihydrooxazoles are preferred over diphosphines including the known BINAP ligand [2,20-bis(diphenylphosphino)-1,10-dinaphthalene] because they do not induce undesirable side processes of double bond migration.75 Complexes 37a ¡À d, prepared from the corresponding ligands and PdCl2(MeCN)2 [followed by treatment with Ag(OTf) and water], have been employed for copolymerisation of styrene with CO.76 This yielded poly(1-oxo-2-phenylpropane-1,3-diyl) with a molecular mass of*20 000.In the presence of the achiral complex 37a, an atactic polymer is formed, whereas catalysis by the optically active complexes 37b ¡À d results in a stereoregular isotactic polymer.2+ O OTf ¡¦2Ph2P N Pd R H2O OH2 37a ¡À d R = H (a), Bui (b), Ph (c), Bn (d). Ligand 38 has been employed for copolymerisation of styrene and ethylene with CO in the presence of the complex [Pd(OAc)LS]+TfO7, where L=38 and S is a solvent.77 O N P Bn 38 Other researchers working independently of the three leading groups mentioned above used successfully the ligands 26b,e in allylic sulfonylation,78 while the ligand 26c was used in the intramolecular Heck reaction.79 The allylation mechanism was analysed theoretically by the extended Hu�� ckel method.80 New phosphorus-containing oxazolines 39 and 40 have been synthesised.The ligand 39 81 ensures ee of 95%¡À 99% in the allylic amination of the acetates 27d,h,i with benzylamine or phthali- mide. ORO RO RO N O O N Ph2P Ph2P Me Me 40 (R=ButCO) 39 NHBn Ph a Me R OAc Ph [Pd(All)Cl]2 . 2.4L X Ph Me R b 27d,h,i R Me (a) BnNH2; (b) HX, X= R=Ph (d, 43%, ee 99%), 4-ClC6H4 (h), 4-BrC6H4 (i); L=39; CON . CO The compound 40, containing a carbohydrate fragment fused with a dihydrooxazole fragment, has been used as a ligand in the alkylation of dimethyl malonate with the allyl acetates 12 (ee 98%) and 13 (ee 69%) in the presence of [Pd(All)Cl]2.82 This resulted in the synthesis of phosphorus-containing dihydro-1,3-oxazine iso- K N Gavrilov, A I Polosukhin mers 41.In the reaction with the allyl acetate 12 {[Pd(All)Cl]2 as the pre-catalyst}, ee depends on the ligand: it is 95%in the presence of cis-41 and only 64%in the presence of trans-41.83 O O N N PPh2 PPh2 trans-41 cis-41 On the basis of the compound 42, neutral and cationic palladium(II) complexes have been prepared.84 To the best of our knowledge, 42 is the only P,N-bidentate ligand of the dihydrooxazole type containing a phosphinite group. It would be of interest to study this compound in the asymmetric catalysis and to compare it with analogous phosphine systems. OPPh2 N Et O 42 R2 Recently, a series of phosphorus-containing dihydrooxazoles 43a ¡À n have been synthesised. Unlike the ligands 26 ¡À 42, in this case, the diphenylphosphine fragment is not attached to the aromatic nucleus.85, 86 N Ph2P O 43a ¡À n R1 R1=H:R2=Me (a), Bn (b), Pri (c), Ph (d); R1 = (S)-Me, R2 =Pri (e); R1 = (R)-Me, R2 =Pri (f); R1 = (S)-Ph: R2 =Bn (g), Pri (h), Ph (i), But (j); R1 =(R)-Ph: R2=Bn (k), Pri (l), Ph (m), But (n).The alkylation with the allyl acetate 12 {[Pd(All)Cl]2 as the pre-catalyst} in the presence of the ligands 43a ¡À d provides ee of 11%¡À 90%, while this reaction in the presence of 43e ¡À n results in ee of 22%¡À 98%. The best results were attained in the presence of the ligands 43c and 43n. The optical yield depends on the solvent; for example, with the ligand 43c inCH2Cl2, the ee is 90%, whereas that in MeCN is only 11%.Acetonitrile is the optimal solvent for ligands with two chiral centres such as 43e ¡À n. The research- ers 85, 86 did not explain this fact. Apparently, the chiral centre of the dihydrooxazole ring in the compounds 43e ¡À n is responsible for the asymmetric induction because the diastereomers 43h and 43l provide ee of 90% and 93%, respectively, all other factors being the same. A number of phosphorus-substituted dihydrooxazoles 44a ¡Àm, differing in the nature of the substituent R in the 2-position, have been synthesised.87 O R N44a ¡Àm PPh2 R=Me (a), But (b), CHPh2 (c), CPh3 (d), 1-adamantyl (e), 3,5-But2C6H3 (f), ferrocenyl (g), CF3 (h), Ph (i), 4-MeOC6H4 (j), 4-MeC6H4 (k), 4-NO2C6H4 (l), C6F5 (m). These ligands ensure ee of up to 94% in the alkylation of dimethyl malonate with the allyl acetate 12 in the presence of [Pd(All)Cl]2, the compound 44e being the most effective.Electron- donating substituents in the dihydrooxazole ring were found to increase the optical yield. In general, the enantioselectivity attained with the ligands 44a ¡Àm is somewhat lower than that observed with the ligands 26a ¡À t. This was attributed 87 to the fact that the relative strain and the tendency for opening of the five- membered chelate metallacycles formed by the ligands 44 areChiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium greater than those in the case of 26. Indeed, according to 31PNMR spectroscopy, the reaction of [Pd(All)Cl]2 with the compound 44k taken in a twofold excess relative to the palladium complex results in a complex containing two organophosphorus ligands in the palladium coordination sphere, while the reaction of the com- pound 26e with the same Pd complex yields a metallacycle with only one phosphorus-containing ligand.In view of these data together with the results obtained using the ligand 24, one can formulate a general rule: the phosphorus-substituted dihydroox- azoles forming five-membered chelate metallacycles are less effective than those forming six-membered rings. In this connec- tion, compounds 45a ± f synthesised recently 88 are worthy of note. R2 Ph2P N R1 O NBoc45a ± f R1=H:R2=Pri(a), Ph (b), But (c); R1=Pri: R2 =H (d); R1=Ph, R2 = H (e); R1=R2=Me (f).For example, in the alkylation of dimethyl malonate with cyclopent-2-enyl acetate (46), ee reaches 96%, the yield of the final product being 79%± 99%. CO2Me [Pd(All)Cl]2 . 3L OAc+CH2(CO2Me)2 CO2Me 46 L=45a ± f. The optical yield and the absolute configuration of the reaction product are controlled by the pyrrolidine fragment because in the presence of the ligand 45f, ee is only 80%. The ligands 45a ± f should form complexes containing seven-mem- bered metallacycles, which are moreover rather strained; never- theless, they ensure a high level of asymmetric induction. Thus, the enantioselectivity in this reaction does not depend on the size of the metal ring. In addition to the two, to some extent, relatrocesses, i.e., allylic substitution and Heck reaction, phosphorus-containing dihydrooxazole P,N-bidentate ligands have found application in the asymmetric version of catalytic hydrosilylation.Thus the ligands 26a ± c,f,i,k,o,r ± t have been studied in the reduction of a series of prochiral aromatic and aliphatic ketones with diphenyl- silane {with [Rh(COD)Cl]2 as the pre-catalyst}.89, 90 In the hydrosilylation of acetophenone, the degree of conversion is 80%± 99% and the optical yield is up to 86%. The highest enantioselectivity was found for the ligands 26b (ee 82%) and 26r (ee 86%); the latter ligand, which contains electron-with- drawing substituents at the phosphorus centre ensures also the highest catalytic activity.90 It should be noted that in the presence of the ligand 26c, which displayed very good properties in allylic alkylation, ee is only 40%.Too bulky substituents at the phos- phorus centre do not enhance the enantioselectivity. Indeed, with the ligand 26o, ee is only 17%. It is noteworthy that the replace- ment of the PAr1Ar2 fragment in the compounds 26 by SPh or SePh almost eliminates the asymmetric induction. The use of the ligand 39 for the reduction of acetophenone with diphenylslane {the catalytic system [Rh(COD)Cl]2 . 4L} 91 proved especially successful (ee 94%). In the general case, the selectivity is inversely proportional to the size of the substituent at the 4-position of the dihydrooxazole fragment of the ligand.However, the asymmetric induction provided by the ligand 39 was found to be abnormally high, despite the fact that the two methyl substituents in the compound 39 create greater steric hindrance than one tert-butyl substituent in the compound 26c. In the opinion of the investigators cited,91 interpretation of these effects requires further research into the mechanism of the rhodium-catalysed hydrosilylation and the structure of intermedi- ates. 669 IV. Other groups of chiral P,N-bidentate ligands The two classes of optically active P,N-bidentate ligands consid- ered above mainly include particular series of compounds prepared specially to be investigated in catalytic reactions. The other compounds used as P,N-bidentate ligands are difficult to classify because they belong to different classes and types. We classified these compounds in terms of the structures of the nitrogen and phosphorus electron-donating centres; three groups of ligands were thus distinguished.Within each group, the data are mainly arranged according to the type of the catalytic reaction. 1. Ligands with sp3-hybridised electron-donating nitrogen atoms Chiral P,N-bidentate ligands with sp3 hybridised electron-donat- ing nitrogen atoms became known earlier than other compounds (except for aminoalkylferrocenylphosphines). Back in 1977, syn- theses of compound 47a with two asymmetric centres, carbon and phosphorus, and the complex of the phosphine 47a with PdCl2 were reported.92 The formation of the metal chelate ring and the absolute configuration of the chiral centres in the complex were proved by X-ray diffraction analysis.Subsequently,93 several more aminophosphines with benzene or naphthalene nuclei have been reported, for example, (S)-AMPHOS 48, 1,2-DPNEA 49, 1,8-DPNEA 50, 2,1-DPNEA 51 and 2,3-DPNEA 52. H Me Me Me NMe2 H H PPh2 NMe2 NMe2 PPh2 PPhBu 49 (S)-48 (SC, RP)-47a, (SC, SP)-47b H NMe2 Ph2P Me Ph2P H H Me Me NMe2 NMe2 PPh2 52 51 50 Neutral or cationic Rh(I) complexes with ligands of the type 47b and 48 ± 51 were used as catalysts of hydrogenation of prochiral unsaturated acids. Thus hydrogenation of itaconic acid in the presence of 1,2-DPNEA 49 proceeds with an optical yield of up to 43%.93 Other ligands, for example, the phosphorus-containing ephe- drine derivative 53, have also found application in the hydro- genation of prochiral unsaturated acids.The reduction of N-acetyl-a-aminocinnamic acid in the presence of [Rh(C8H14)Cl]2 and the ligand 53 gives the product in an optical yield reaching 77%.94 PhCHCH(Me)NHMe PPh2 53 Ligands 54, 55a,b, 56, 57a,b, 58 and 59a,b 95, 96 have been used in the hydrogenation of N-acetyl-a-aminocinnamic and N-benzoyl-a- aminocinnamic acids in the presence of [Rh(C6H8)Cl]2. The max- imum optical yield (68%) was attained in the case of (S)-1-[2- (methyl-tert-butylphosphino)phenyl]ethylamine (58).96 NMe2 Me2N H NMe2 Me Ph Me P PPriMe PPhMe (SC)-55a, (RC)-55b 56 Me 54670 Ph2P NMe2 NMe2 Me2N H Me Me P(Me)C6H11-cyclo PButMe 58 (SC)-57a,(RC)-57b (S)-59a, (R)-59b The cationic rhodium(I) complexes 60a,b have been synthes- ised and characterised in detail by 1H, 31P and 13C NMR and IR spectroscopy and conductometry.97 The ligands present in these complexes contain rarely encountered carbamide donor centres.However, the reduction of Z-(a)-ethylacetaminocinnamic acid in the presence of the complex 60b resulted in only a moderate enantioselectivity (ee 34%). + N PPh2 PF¡6 Rh(COD) O NHR 60a,b R=But (a), (CH2)3Si(OEt)3 (b). Yet another traditional field of application of chiral amino- phosphines with an sp3-hybridised nitrogen atom is cross-cou- pling. Both acyclic (61a ± g, 62) and heterocyclic (63, 64) aminophosphines have been employed in these reactions.98, 99 Me2N R61a ± g Ph2P Abbreviation R Compound 61 Me Pri Bui But Ph Bn cyclo-C6H11 abcdefg (S)-ALAPHOS (S)-VALPHOS (S)-LEUPHOS (R)-t-LEUPHOS (R)-PhGLYPHOS (S)-PHEPHOS (R)-ChGLYPHOS H Bn Me2N H CH2PPh2 PPh2 N H NMe Ph2P Me 64 62 63 (S)-PROPHOS Cross-coupling of 1-phenylethylmagnesium chloride with vinyl bromide in the presence of complexes of PdCl2 with the ligands 61a ± g, 62, 63 and 64 proceeds with a moderate enantio- selectivity [ee up to 55% for (S)-PHEPHOS 61f].The catalytic system NiCl2 . L was found to be more efficient; indeed, in the case of the ligand 61d, the optical yield reaches 94%. It should be noted that movement of the chiral centre towards the phosphorus atom results in a decrease in the asymmetrising activity (ee 38% for 61a and ee 25% for 62).99 For example, in the cross-coupling of 1-phenylethylmagnesium chloride with b-bromostyrene, the Ni complex (S)-AMPHOS 48 exhibits no catalytic effect, while PdCl2 .[(S)-AMPHOS] ensures a 95% chemical yield and a 40% optical yield.100 Palladium(II) cis-dichloro compounds 65a,b have been synthesised and characterised by X-ray diffraction data.101 These complexes and nickel complexes 66a,b having a similar structure were tested in the cross-coupling of 1-phenylethylmag- nesium chloride with vinyl bromide. In the presence of the palladium complexes 65a,b, the optical yield was moderate (ee up to 45%). The use of the nickel complexes 66a,b resulted in a higher enantioselectivity (for 66b, ee 88%).K N Gavrilov, A I Polosukhin R M R Compound PPh2 Me2N M Cl Cl 65a 65b 66a 66b Pd Pd Ni Ni (S)-(CH2)2SMe (R)-(CH2)2SMe (S)-(CH2)2SMe (R)-(CH2)2SMe 65a,b, 66a,b The synthesis of P,N-bidentate ligands 67a ± k, containing a pyrrolidine ring, has been described.102, 103 NR2 Ph(R1)P 67a ± k R2 R1 Compound 67 abcdefghijk Ph H Ph Bn Ph 2-MeOC6H4CH2 Ph Et H H 2-MeOC6H4CH2 H 2,4,6-(MeO)3C6H2CH2 H 2-MeOC6H4CH2 Bn 2,4,6-(MeO)3C6H2CH2 Bn H Bn H Ph The compounds 67b,c,g,h,i were used to prepare the cis-diiodo metal chelates PdI2 . L, having a heteronorbornane skeleton; diastereomeric palladium complexes with the ligands 67a ± g were separated by column chromatography.The structure of the complex compounds was convincingly proved by IR spectroscopy and 1H, 13C and 31P NMR spectroscopy (including the use of solid-state techniques), FAB mass spectrometry and X-ray dif- fraction analysis. However, these products were not selective in cross-coupling, the ee values being not more than 10%. Mean- while, in the case of NiCl2 . 67b, the chemical yield is 98% and the optical yield is 75%.103 P,N-Bidentate ligands 68 and 69, containing a phosphinite group, have been described.104 Ph Me Ph Ph2PO Me2NCH2 Bn OPPh2 Me2N Me 68 69 Palladium complexes formed by the ligands 68 and 69 proved to be inefficient in the cross-coupling of 1-phenylethylmagnesium bromide with vinyl bromide; the optical yield was only 6%± 9% for a degree of conversion of 60%± 74%.104 This outcome was explained by insufficient rigidity of six- and seven-membered palladacycles and by lower basicity of the phosphinite group in comparison with the phosphine group.However, this interpreta- tion is quite debatable. The palladium complexes were character- ised only by 31P and 1H NMR spectra. No data of long- wavelength IR spectroscopy which would provide information on the arrangement of the chloro ligands at the Pd atom based on the position of the nPd7Cl band are available. No data on the molecular masses of the complexes can be found either. Therefore, it cannot be ruled out that the reaction of the ligand 69 with PdCl2 gives a bi- or poly-nuclear compound [PdCl2 .L]n, in which the P,N-bidentate ligands function as bridges linked according to the `head-to-tail' pattern, rather than a chelate. As regards the basicity of the phosphorus centre in cross-coupling catalysts, the statistical data available to date do not permit an unambiguous conclusion about its role in the enantioselectivity. It is quite probable that in the case of aminophosphinites 68 and 69, an appropriate catalytic system has not been chosen.Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium Several chiral aminophosphines have been tested in asymmet- ric hydrosilylation. For instance, the reduction of acetophenone with diphenylsilane with participation of (S)-AMPHOS 48 {[Rh(C2H4)2Cl]2 as the pre-catalyst} gave the corresponding alcohol with ee 33%, and in the case of reduction of tert-butyl methyl ketone, ee was 72%.105 Subsequently,106 this result was somewhat improved: catalysis by [Rh(COD)Cl]2 .10L, where L=(S)-AMPHOS 48, gave a-phenylethyl alcohol in an optical yield of 51% and a chemical yield of 97%. Other pre-catalysts proved to be less efficient: for Pd(COD)Cl2, ee reaches 50.5% but the degree of conversion is only 13%, while for Pt(COD)Cl2 and [Ir(COD)Cl]2, ee does not exceed 27%. It should be noted that close analogues of (S)-AMPHOS 48 containing a di(tert-butyl)phosphine group in the benzene ring, instead of the diphenylphosphine group, are known,107 namely, (S)- (70a) and (R)-DIBUTPHOS 70b.According to X-ray dif- fraction analysis,107 in the rhodium complex 71, the ligand (R)- DIBUTPHOS 70b forms a metal chelate. + Me ClO¡4But NMe2 P RhNMe2 But PBut2 Me 71 70a,b However, the use of the ligands 70a,b in the hydrosilylation of acetophenone with phenyldimethylsilane {with [Rh(C2H4)2Cl]2 as the pre-catalyst} gave a quite unexpected result: for a chemical yield of up to 92%, no enantioselectivity was observed at all. The ligand (S)-PHEPHOS 61f ensures an optical yield of 54% in the hydrosilylation of acetophenone with diphenylsilane {the [Rh(COD)Cl]2 .4L system}.108 When cyclic imine 72 was subjected to hydrosilylation with diphenylsilane in the presence of the [Rh(COD)Cl]2 .2.8L system, where L=(S)-PHEPHOS 61f, com- pound 73, the enantiomer of the antidepressant pyrazidol, was obtained in 69% yield and ee 73%.109 Me Me 1) Ph2SiH2, [Rh(COD)Cl]2, 2.8 equiv. 67f N 2) H+ N NH N 73 72 The behaviour of (S)-PHEPHOS 67f and (S)-AMPHOS 48 in the reaction with [Rh(COD)Cl]2 and [Rh(CO)2Cl]2 has been studied.110 The reaction products, binuclear Rh(I) complexes of the [Rh(L)Cl]2 type and mononuclear complexes of the Rh(CO)Cl(L) type, were isolated and characterised by spectro- scopy (IR, X-ray electron, 1HNMR, 31PNMRand mass spectra). To summarise the foregoing, it should be noted that rhodium and palladium derivatives of chiral aminophosphines with an sp3- hybridised peripheral nitrogen atom provide moderate enantiose- lectivity as catalysts of hydrogenation, cross-coupling and hydro- silylation.However, they are more effective in allylic substitution. Specifically, amino-substituted phosphinous amide 74, in which the asymmetric phosphorus atom is incorporated in a heterocycle, ensured a chemical yield of 99% and an optical yield of 62% in the alkylation of dimethyl malonate with the acetate 12 {with [Pd(All)Cl]2 as the pre-catalyst}.111 Me H N(CH2)2N P Ph 74 671 The same model reaction was used to test a series of amino phosphinous amides 75a ± j, in which the phosphorus atom is achiral {[Pd(All)Cl]2 as the pre-catalyst}.112 In the presence of these ligands, ee is 76% for a chemical yield of 41%± 82%. The best optical yield was attained with the ligand 75h, having no substituents at the amide nitrogen atom.R2 R1 Compound 75 NHR1 PPh2 NR2 75a ± j abcdefghij H H Me Me Bn Bn Pri Pri CH2CH2Pri CH2CH2Pri 2-MeC6H4CH2 2-MeC6H4CH2 3-MeC6H4CH2 3-MeC6H4CH2 Bn H 2-MeC6H4CH2 H ButCH2 H Ligands 76a,b, 77, 78a,b and 79 with axial or central chirality of the peripheral nitrogen atom have been prepared.113 Me Ph2P(CH2)nN Ph2P(CH2)2NMe 77 76a,b n = 2 (a), 3 (b) R Ph Ph2P(CH2)2N Ph2P(CH2)2N Ph R 79 78a,b R=Et (a), Bn (b) In the model allylic alkylation using [Pd(All)Cl]2 as the pre- catalyst, the use of the ligands 76a,b results in ee values of 93% and 96%, respectively.The chemical yields are 96%. An increase in the optical yield on passing from the compound 76a to 76b, was attributed 113 to an increase in the size of the chelate metallacycle it forms. The ligands 77, 78a,b and 79 are much less enantioselective (ee up to 39%). In the presence of the ligands 80a ± c, in which the phosphorus center is further removed from the nitrogen centre, the asymmetric induction decreases with an increase in the distance between the donor centres.113 n m Compound 80 (CH2)mPPh2 N (CH2)n abc 1 0 2 0 1 1 80a ± c In the alkylation of dimethyl malonate with the acetate 12, the ligands 80a ± c [with Pd(OAc)2 as the pre-catalyst] ensure ee of 96%, 79% and 18%, respectively.114 The low optical yield observed in the last-mentioned case was explained by assuming monodentate coordination of the compound 80c for the forma- tion of the intermediate.The ligands 80a ± c were also employed in the cross-coupling of (1-phenylethyl)magnesium chloride with vinyl bromide. The best result (ee 46%) was observed for the ligand 80b; in other cases, ee did not exceed 9%. Compounds 81a,b, having a similar structure and containing additional chiral centres in the dihydroazepine ring, offer no advantages over the ligands 80a ± c.115672 RN PPh2 R 81a,b [R =Me (a), Et (b)] In the same model allylic alkylation reaction {with [Pd(All)Cl]2 as the pre-catalyst}, ee is 59% for the ligand 81a and 37% for the ligand 81b. A higher enantioselectivity (ee 68%) was achieved in the alkylation of dimethyl malonate with the sterically less shielded acetate 13 {the [Pd(All)Cl]2 . 81a catalytic system}.When dinaphthalene derivatives 82a ± h, which also have a chiral axis, are employed in allylic alkylation [the acetate 12, pre- catalyst Pd(dba)2], the chemical yields are 77%± 95% and the optical yields are 53%± 73% (the ligand 82f is the most effective). R2 R1 Compound 82 R2 N Me R1 i PPh2 82a ± h cyclo-C6H11 Pri cyclo-C6H11 H HMe H Et H Pr HMe Me cyclo-C6H11 cyclo-C6H11 abcdefgh The reaction of aminophosphine 82b with PdCl2(PhCN)2 has been studied.116 According to the 1H, 13C and 31P NMR spectra, the reaction gives an equilibrium mixture of three complexes, 83, 84 and 85, in a ratio of 10 : 85 : 5. Cl Cl Ph Ph Ph Ph Ph Pd 7 P P P Ph PdCl2 PdCl2 N N 83 85 NMe2 +84 It was demonstrated by X-ray diffraction analysis that the complex 84, which predominates in the mixture, is a chelate formed due to the P7Cs binding. It was noted that this type of binding can play an essential role in the formation of intermedi- ates in allylation processes.Chiral N-sp3-hybridised aminophosphines, which have not yet found application in catalysis but, in our opinion, hold consid- erable promise, deserve special consideration. Most of these compounds contain asymmetric phosphorus and/or nitrogen atoms. In particular, the reaction of compound 86 with PdCl2(MeCN)2 results in two epimers of the palladium chelate complexes, PdCl2 .[(SN, SC, SP)-86] and PdCl2 . [(SN, SC, RP)-86], which were separated by fractional crystallisation.117 Bun PPh 86 NH Palladium and platinum cis-dichloro complexes 87a,b with two chiral centres have also been described.118 Whereas the platinum complex 87a exists as a single dia- stereomer (the asymmetric nitrogen atom has only one config- K N Gavrilov, A I Polosukhin Me Pri Me N Cl N PPh2 Pd Me Ph H M P Cl Cl Me Cl Ph Ph 88 87a,b [M=Pt (a), Pd (b)] uration), the palladium analogue 87b undergoes epimerisation with a half-transformation period of 30 min (according to 1H NMR spectra and polarimetry). Unlike the complex 87b, com- pound 88 contains a configurationally stable chiral nitrogen centre due to the stereochemical assistance of the neighbouring endocyclic carbon centre.119 The chiral nitrogen centre in the compound 89 also possesses configurational stability, as shown 120 by the data of 1H (2D NOESY), 13C, 15Nand 31P NMR, IR and mass spectra and X-ray diffraction analysis.Pri NH2 C6H4OMe-4 N (4-MeC6H4)2P Me P Pd H Cl Cl 89 R 90a,b [R =Ph (a), 2-ClC6H4 (b)] The complex exists as a single diastereomer both in solution and in the solid state but the palladacycle retains configurational mobility. The conformers differ little in energy, which detracts from the potential use of the compound 89 as a ligand in stereo- selective catalysis. Some aminophosphines with chiral centres at the phosphorus atom also appear interesting, for example, compounds 90a,b, for which both neutral and cationic palladium(II) chelates have been obtained.121, 122 The original bicyclic compound 91 contains five asymmetric carbon atoms in addition to the chiral phosphorus atom.This compound can bind to a metal as both a P-monodentate or a P,N- bidentate ligand (according to 31P, 13C and 1H NMR spectro- scopy).123 H H a N P Me Ph Rh OMe Cl CO OBut H H ButO H H N P N P b Ph Me Me Ph Rh 91 COD Cl (a) 0.5 equiv. [Rh(CO)2Cl]2; (b) 0.5 equiv. [Rh(COD)2Cl]2. The complex PdBr(R)(L), where L=(R)-AMPHOS [(R)-48], R=Me, C:CSiMe3, has been synthesised.124 Several ligands 92a ± d based on (R)-AMPHOS [(R)-48] have been prepared.125 Me HNMe2 PPh2 Cr(CO)2L 92a ± d L=CO (a), PPh3 (b), P(OMe)3 (c), P(OPh)3 (d).The complexes 92a ± d have found application so far only in the amination of 4-bromotoluene with piperidine catalysed byChiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium palladium complexes; however, they are also of interest for asymmetric synthesis, especially in view of the possibility of changing the electronic properties of the phosphorus centre by changing the nature of the ligand L.125 The dimethylhydrazone of the phosphorus-substituted (R)- camphor 93 reacts with PdCl2(PhCN)2 to give initially six- membered chelate metallacycle 94 (according to the IR spectra, 31P and 1H NMR spectra and X-ray diffraction analysis).126 On treatment of a solution of the complex 94 in CH2Cl2 with dry hydrogen chloride, the metal chelate rearranges giving rise to complex 95 with a coordinated sp2-hybridised nitrogen atom.Ph Ph P Cl PPh2 HCl PdCl2(PhCN)2 Pd Cl CH2Cl2 N N Me 94 93 Me NNMe2 Ph Ph P Cl Pd Cl N 95 NMe2 2. Ligands with sp2-hybridised electron-donating nitrogen atoms 6 )2 The interest in ligands with sp2-hybridised donor nitrogen atoms arose somewhat later than that in compounds with sp3-hybridised nitrogen centres. In 1982, phosphines 96a,b and their neutral palladium derivatives PdCl2(L) were synthesised.127 Later,128 the cationic complexes [Pd(L)2]2+(PF¡ and [PdCl(L)2]+X7 (X=Cl, PF6) were prepared. N P(Me)Ph (RP)-96a, (SP)-96b Several other fairly interesting ligands and complexes based on them, for example, 97 ± 99, have been reported.129 ± 131 X-ray diffraction analysis 129 showed that the Pd7Cl(1) bond length in the complex 97 [2.394(2)A)] is markedly greater than the typical value, due to the strong trans-effect of the phosphorus centre with an unusual geometry.The ligand 98 has a nine-atom bridge between phosphorus and nitrogen. It is thus able to form 12-membered macrometalla- cycles.130 Nevertheless, the enantioselectivity (ee 86%) in the model alkylation of dimethyl malonate with the allyl acetate 12 {with [Pd(All)Cl]2 as the pre-catalyst} is quite comparable with the enantioselectivity attained when using catalysts with a standard size of the chelate ring.N Cl1 O O Pd NH NH Cl2 P Ph N Me 98 Me PPh2 (SP)-97 4 4 Cationic complexes of the phosphinite 99, [Rh(CO). .(PPh3)(L)]+ClO¡ and [Rh(COD)(L)]+ClO¡ (the latter was studied by X-ray diffraction analysis), have found use as catalysts of hydroformylation of a number of unsaturated substrates such as vinyl acetate, styrene, vinylnaphthalene and methyl acrylate.131 The optical yields were 12%, 6%, 78% and 92%, respectively for rather high degrees of conversion and regioselectivities. However, later it was shown 132 that the last-mentioned value was erroneous. 673 Me N OPPh2 Me Me 99 The formation of complexes by phosphine 100 (PYDIPHOS) was thoroughly studied by spectral methods (IR spectroscopy, 31P, 1H and 13C NMR spectroscopy) and mass spectrome- try.133, 134 Me Me O O 0.5 equiv.[Rh(CO)2Cl]2 Ph Me Me P N Ph Rh O O 101 Cl OCMe Me O O N PPh2 100 0.5 equiv. PdCl2(PhCN)2 Ph P N Ph Pd 102 Cl OC Both rhodium (101) and palladium (102) metal complexes were described as chelates (the structure of the complex 102 was confirmed by X-ray diffraction analysis); however, the data obtained for the complex 101 do not rule out unambiguously an alternative highly symmetrical binuclear structure with bridging coordination of the P,N-bidentate ligands of the `head-to-tail' type.134 The enantioselectivity observed in catalytic reactions involving PYDIPHOS (100) is moderate. For hydroformylation of styrene in the presence of the complex 101, the ee is 28%; ethoxycarbonylation in the presence of the complex 102 results in ee of 20%; in the model allylic alkylation {the [Pd(All)Cl]2 .4L system}, ee is 9%.135 Studies by Brown et al.136 ± 141 deserve special attention. These researchers proposed effective ligands QUINAP 103 and PHENAP 104a,b. N N PPh2 PPh2 (R)-104a, (S)-104b 103 Cationic rhodium complexes 105a,b with these ligands have found successful use as catalysts of hydroboration of vinylarenes with catecholborane.137, 138 The optical yields were 64% ± 95% and the chemical yields were 21%± 71%. For example, in the hydroboration of styrene in the presence of the complexes 105a,b, the ee values were 91% and 67%, respectively.138 + N CF3SO¡3Rh P Ph Ph 105a,b Ph N =103 (a), 104a (b).P Ph674 The complex 105a has also been used as the catalyst for hydroboration ± amination of vinylarenes;139 the corresponding primary amines were obtained in optical yields of 77% ± 98% and in chemical yields of 50% ± 64%. Allylic palladium complexes 106a ± d with the P,N-bidentate ligand 103 have been synthesised. Their structure was established based on the spectral data (IR spectroscopy and 31P, 13C and 1H NMR spectroscopy and mass spectrometry).140 +BF¡4R2 Pd N P R1 Ph Ph R3 106a ± d R3 R2 R1 Compound 106 H H H Ph Ph Ph HPh Ph HHPh abcd In the classical alkylation of dimethyl malonate with the allyl acetate 12 catalysed by the complex 106a, the optical yield fluctuates from 67% to 98%, depending on the reaction condi- tions.For other allylic substrates, the optical yield varies from 47% to 67%. The replacement of dimethyl malonate by diethyl malonate decreases somewhat the enantioselectivity (ee 75%± 78%). The use of the neutral complex PdCl2 . 103 instead of the cationic complex 106a increases the optical yield (ee 79%).140 A similar complex in which [Pd(All)(104a)]+BF¡4 is used instead of QUINAP 103 ensures a commensurable level of asymmetric induction in this reaction (ee 95%).141 Ligands 107a ± d 142 used in hydroboration and compound 108 143 containing an indole fragment have been prepared recently. N N PAr2 PPh2 NMe 108 107a ± d Configuration Ar Compound 107 abcd SSRR 3-MeC6H4 3,5-Me2C6H3 2-biphenylyl 2-furyl Unfortunately, both the ligand 108 itself and its complex with PdCl2 109 are readily racemised.The corresponding cationic allylpalladium complex 110 is the only diastereomer. + Ph N N Cl BF¡ Pd Pd 4 Cl P P Ph Ph Ph Ph Ph 109 110 =108 Ph2P N K N Gavrilov, A I Polosukhin The structures of 109 and 110 were studied 143 using spectral methods (IR and 31P, 1H and 13C NMR spectroscopy and mass spectrometry). The complex 110 was studied by X-ray diffraction analysis. The metal atoms in both complexes were found to be coordinated to the isoquinoline nitrogen atom. The configura- tional stability of the allylic cationic complexes of Pd(II) of the type 110 gives hope for their successful use in asymmetric allylic substitution.143 In addition to ligands with a pyridine or isoquinoline frag- ment, a large group of ligands with a peripheral imino group, iminophosphines, have been described.They were studied by Brunner et al.144, 145, 148 ± 150 In particular, (R)-IMINPHOS 111 and (R)-AMINPHOS 112 (prepared by the reduction of 111 with NaBH4) were compared in the hydrogenation of N-acetyl-a- aminocinnamic acid {[Rh(C2H4)2Cl]2 as the pre-catalyst} and hydrosilylation of acetophenone with diphenylsilane {[Rh(COD)Cl]2 as the pre-catalyst}.144 Ph2P N Ph2P HMe HMe NH Ph Ph (R)-112 (R)-111 The ligand (R)-111 was more enantioselective in hydrogena- tion than (R)-112 (ee 17% and 14%, respectively).The ee value attained in hydrosilylation is 29% in the presence of (R)- IMINPHOS 111 and 53% for (R)-AMINPHOS 112.144 Note that the ferrocene arylideneaminophosphines 17a ± d and the aminophosphines (S,R)-18 exhibit a diametrically opposite pat- tern of dependence (see Section II.2.b). Subsequently,145 other imino- and amino-phosphines such as 113a,b ± 115a,b and 116 have been studied in the hydrosilylation of acetophenone {[Rh(COD)Cl]2 as the pre-catalyst}; however, they all proved to be middling asymmetric inductors because they ensured ee of not more than 20%. The best results were obtained with the ligands 113a and 115a. R R N N N Me H Ph2P Ph2P Me H NH Ph Ph 114a,b 113a,b R N NH N N Me H Me H Ph2P Ph2P Ph 116 Ph 115a,b R = H (a), Me (b). A later publication 146 describes the synthesis of the (S)- enantiomer of IMINOPHOS [(S)-111] and a large series of its palladium and platinum complexes such as PdCl(Me)(L), [Pd(Z3-All)(L)]+X7 (Z3-All=2-RC3H4, 2-MeC3Me4, 2-RC3H2Me2; X=Cl, O3SCF3), PdCl2(L).149 The chelating P,N-bidentate coordination of the ligand in these complexes was proved by 31P, 1H and 13C NMR spectroscopy; neutral Pd(II) complexes were studied by X-ray diffraction analysis. Close analogues of IMINPHOS 111, compounds 117a,b, are also known.147 The enantioselectivity exhibited by these com- pounds in the alkylation of dimethyl malonate with the allyl acetate 12 catalysed by Pd complexes is also low (ee does not exceed 17%).Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium N Ph2P HMe R 117a,b R=cyclo-C6H11 (a), 2-naphthyl (b).Quite recently, a new large series of iminophosphines 118a ± g have been proposed;148 ± 150 these products were used to prepare biologically active compound 119, which belongs to the group of barbiturates.149 In the reaction catalysed by palladium acetylacet- onate in the presence of (S)-118e, the yield of the compound 119 was 89% (ee*34%). PPh2 NR1 118a ± g Ph R1= ( C Me a), HC(Me)CH2OH (b), HC(Et)CH2OH (c), H HC(Pri)CH2OH (d), HC(But)CH2OH (e), HC(Ph)CH2OH (f), Pri Me CH H2C (g). Me CH2OH Me Me O O O O Pd(acac)2, . 4L OAc+ NMe HN NMe HN O 119 O L=(S)-118e.Although the optical yield attained is relatively low, this reaction opens up the way for accumulation of a valuable medicinal preparation, namely, the anesthetic metohexital 120, which was prepared in the presence of the ligand (S)-118e in an optical yield of 80%.150 H H Me Me Et Et O O O O Pd(acac)2 . 4L OAc+ NMe HN NMe HN O 120 O L=(S)-118e. Chiral aminoiminophosphines 121 and 122a,b were found to be fairly effective in the alkylation of dimethyl malonate with the allyl acetate 12.151 ± 153 N N Pri CH2OMe N PPh2 Me2N PPh2 (S)-122a, (R)-122b 121 When the reaction was carried out with the ligand 121 {[Pd(All)Cl]2 as the pre-catalyst}, the chemical yield was 85% and the optical yield was 92%.In the allylic palladium intermedi- ate, the ligand was found 151 to be coordinated through the phosphorus atom and the imine nitrogen atom. The iminophos- phines 122a,b are also effective in this reaction {[Pd(All)Cl]2 .3L as the catalyst}: ee is 90%± 92% for a nearly quantitative chemical yield.152 The ligand 121 was used in processes in which ketene silyl 675 acetals are used as nucleophiles.153 The yields in these reactions reach 95% and ee is 93%. R1R2CCO2R3 R1 OSiR4 OCOBut 3 [Pd(All)Cl]2 . 4L + Ph Ph Ph Ph OR3 R2 R1=R2=R3=R4=Me: R1±R2=(CH2)5, R3=R4=Me; R1=R2=H, R3=Me, R43 =ButMe2; R1=CO2Me, R2=H, R3=R4=Me; L=121. Recently, a new series of phosphorus-substituted imines 123a ± f, containing substituents of different nature in the arylme- thylidene fragment, have been prepared.154 In the alkylation of dimethyl malonate with 3-pivaloyloxy-1,3-diphenylprop-1-ene {catalysed by [Pd(All)Cl]2 .4L}, the catalytic activity was found to be directly correlated with the enantioselectivity, on the one hand, and the electron-donating capacity of the substituent X on the other hand. 4-XC6H4CH NCHCH2PPh2 Pri 123a ± f ee (%) Yield (%) X Compound 123 19 38 42 46 CO2Me CF3H 57 52 Me OMe NMe2 74 85 92 76 88 99 abcdef In the author's opinion,154 this is due to the increase in the stability of the Pd(II) cationic complexes on passing from the ligand 123a to 123f. It should be noted that ligands with sp2-hybridised donor nitrogen atoms have made a greater contribution to asymmetric catalysis than the ligands with sp3-hybridised nitrogen because, in addition to allylic alkylation, they are effective in hydroboration ± oxidation and hydroboration ± amination processes.3. P,N-Bidentate derivatives of phosphorous acid About 250 chiral P,N-bidentate ligands are currently used in the synthesis of coordination compounds and in asymmetric catalysis. The natures of the nitrogen centres in these compounds are quite diverse, while the phosphorus centres are usually acyclic phos- phines (a PAr2 fragment is encountered most often). (Phosphinites such as 42, 68, 69 and 99 and phosphinous amides 74 and 113 ± 116 are the scanty exceptions.) The introduction of electron-donating or electron-withdrawing substituents into the corresponding posi- tions of the benzene nucleus is almost the only method of controlling the inductive effect of phosphorus.However, there exists another, more effective method different in kind, namely, replacement of the carbon atoms in the first coordination sphere of phosphorus by heteroatoms (oxygen and/or nitrogen). This method allows fine adjustment of the chemical stability, the donor ± acceptor properties and steric requirements of the ligand. A phosphorus centre of this type is present in P,N-bidentate ligands derived from phosphorous acid. It is noteworthy that, starting from the early 1990s, the number of publications devoted to the use of chiral phosphites and phosphorous amides, especially P,P-bidentate ones, in the coordination chemistry and catalysis, has substantially increased.155, 156 The obvious advantages of these ligands have been outlined.157 (i) They are readily available because most of them, including functionally substituted ones, can be prepared fairly easily (in one or two steps) and in a high yield from widely used optically active organic precursors, for example, amino alcohols or diols.(ii) No P7C bonds, which would considerably hamper the oxidative destruction of the ligand, are present in the molecule.676 (iii) Owing to the clear-cut p-acceptor properties inherent in phosphites,158 they are more electrophilic ligands; coordinated phosphites are able to stabilise low oxidation states of complex- forming metals. It should be noted that chiral phosphinophosphites are the best ligands for asymmetric hydroformylation.132, 156 The intro- duction of an N-donor centre into the molecule of an ester or an amidoester of phosphorous acid extends the scope of application of chiral phosphites.The first publication reporting the synthesis and complex formation of optically active aminoalkyl phosphites 124 and 125a,b appeared in 1993.159 The reaction of the compounds 124 and 125 with [Rh(CO)2Cl]2 afforded mono- and bi-nuclear Rh(I) chloro carbonyl complexes.159, 160 The ligand 124 forms metal chelate 126, while the ligand 125b gives dimer 127b. The reaction of [Rh(CO)2Cl]2 with the ligand 125a gives rise to a mixture of monomer 128 and dimer 127a; during storage of the solution, the latter product gradually rearranges to complex 128.It should be noted that the complexes 127a,b and 128 have been studied only in solution. Bn O O O [Rh(CO)2Cl]2 Bn P P O NMe2 O O Rh NMe2 124 126 OC CO PriO2C O [Rh(CO)2Cl]2 PO(CH2)nN (CH2)m O PriO2C 125a,b (CH2)m O N(CH2)nP P (CH2)2 CO OC Rh Rh + Rh N Cl Cl P(CH2)nN OC (CH2)2 (CH2)m Cl 128 127a,b n=m = 2 (a); n=3, m = 1 (b). However, the use of the chiral P,N-bidentate derivatives of phosphorous acid 124 and 125a,b in the hydrosilylation of acetophenone with diphenylsilane {[Rh(CO)2Cl]2 and [Rh(COD)Cl]2 as pre-catalysts} was not very effective (ee did not exceed 8%).159 Ligand 129 was prepared 161 by successive phosphorylation of diisopropyl tartrate and 4-aminomorpholine.It is coordinated as a P-monodentate ligand in the complex (acac)Rh(CO)L. O PriO2C O N P NH O 129 PriO2C The complexes of phosphites 130 ± 132 and phosphorous amide 133 with Pd(COD)Cl2were studied by 1H, 31P and 13C NMR, X-ray electron and IR spectroscopy and plasma desorp- tion mass spectrometry. OR RO Pri (a), (b), OR=O P NMe2 O O130a ± c (c). O K N Gavrilov, A I Polosukhin O NPri MeN O P OPh O OP O H 132 131 O NMe2 P O Me N 133 It was shown that the compounds 130a ± c 162 ± 164 and 133 165 form mononuclear chelates of the same type. According to X-ray diffraction data, the Pd7P distance in the palladium complex of the ligand 130a (2.189A) is shorter than this bond in palladium dichloro complexes with P,N-bidentate ligands forming a six- membered metallacycle (2.205 ± 2.241A).162 The bond shortening is due to the p-acceptor properties of the phosphite centre.The resonance signals of phosphorus in the 31P NMR spectra of Pd(II) cis-dichloro complexes with amino phosphites and amino phosphorous amides occur at a higher field than the signals of free ligands due to shielding of the phosphorus nucleus. However, in the case of similar aminophosphine complexes, the signals of the 31P nuclei are located at a lower field than those for the free ligands.166 Dioxaazaphosphocane 131 forms a polynuclear compound of Pd(II), namely, cis-[PdCl2(L)]20, in which the P,N-bidentate ligands function as bridges arranged in the `head-to-tail' man- ner.167 From the 1,3,2-dioxaphospholane derivative 132, the dimeric complex 134 with bridging chloro ligands and an intra- active palladium7nitrogen bond was synthesised.168 This unusual pattern of complexation observed for the ligand 132 is due to the presence of the bulky substituent in the peripheral amino group of its molecule.As a consequence, the amino group becomes a very weak nucleophile, an analogue of HuÈ nig bases. Pri MePhCH Cl N O O Cl O P Pd P Pd OO Cl O NPri Cl CHPhMe 134 Recently,169 aminoalkyl phosphite 135 based on a unique amino alcohol, quincoridin, has been synthesised. O P N O O 135 To the best of our knowledge, the compound 135 is the only known example of a P,N-bidentate ligand with a configurationally stable chiral nitrogen.This ligand was used to synthesise rho- dium(I) chloro carbonyl chelate, [Rh(CO)Cl(L)]. It should be noted that the increased (even in comparison with the most electron-withdrawing amino phosphites 31, 160) p-acidity of the aminoalkyl phosphites 124, 125a,b and 135 influences the spectral characteristics of chloro carbonyl complexes of rhodium with these ligands. Thus the 1JP7Ph spin ± spin coupling constants observed for these complexes are, on the average, *100 Hz greater, while n(CO) values are greater by*20 cm71. Hydrophosphorane compounds are seldom used for the preparation of metal complexes with P,N-bidentate derivatives of phosphorous acid.Complex formation is accompanied by rupture of the P7H and P7N bonds and simultaneous for- mation of M7P and M7N bonds. The main studies dealing with the complexation of chiral hydrophosphoranes have beenChiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium surveyed in several publications.155, 170, 171 Therefore, here we shall consider only the tricyclic hydrophosphorane 136, prepared not long ago.172 The reaction of this compound with Pd(COD)Cl2 gave the first stable palladated phosphorane 137 (in an almost quantitative yield).173 Et HN Et P N O Cl Cl O N Et Pd Pd Et Pd(COD)Cl2 O Cl Cl P O N O N P H Et O NH Et 137 136 After the proton has migrated from phosphorus to nitrogen, the compound 136 acts as a P-monodentate phosphoranide ligand.When 136 is made to react with Pt(COD)Cl2, the phos- phorane ring is opened at the apical P7N bond. In the complex [PtCl2(L)], the ligand occurs in the `open' form and is bound to the metal in the P,N-bidentate fashion.174 The synthesis and coordination to metals of phosphite and phosphorous amide derivatives of alkaloids, quinine and codeine, have been the subject of a detailed review;175 therefore, we do not discuss them here. In 1997, diastereospecific synthesis of three amino-substituted phosphorous diamide ligands 138a ± c was reported.176 XO CH2 P N N N N Ph N (c). (b), (a), X= H 138a ± c In the classical allylic alkylation with the allyl acetate 12 {with [Pd(All)Cl]2 as the pre-catalyst}, the ligands 138a,b ensured a degree of conversion of 100% and ee of 85%± 87%, and the ligand 138c provided 85% and 76%, respectively.Nearly identical results were attained in the related alkylation of dimethyl malo- nate with 3-methoxycarbonyloxy-1,3-diphenylprop-1-ene. Subse- quently,177 ligand 139 has been synthesised. The compounds 138a ± c and 139 (yields 70%± 100%, ee 73% ± 94%) have found application as catalysts of allylic amination. Ph Ph Ph Ph [Pd(All)Cl]2 . 8L +R2R3NH N OR1 R2 R3 HN R1=Ac, CO2Me; R2R3N=BnNH, OMe, N O; OMe N L= N P H(139). N O When diphenylallyl acetate is used as the substrate and benzyl- amine is employed as the nucleophile, the ligands 138a and 138c are the most effective (degree of conversion 95% and ee 93%).In the allylic amination of acetate 140 in the presence of the ligand 138a, the yield is 93% and the optical yield is 89%.178 CH2OAc Pd(dba)2 . 5L O +HN 140 CH2OAc 677 CH2OAc O + CH2N O CH2N CH2OAc (1R, 4S) (1S, 4R) L=138a. It is worth noting that known bisphosphines, (+)-DIOP and (7)-BINAP, ensure ee < 5%in this reaction.178 4,5-Dihydrooxazoles 141 ± 143, containing cyclic phosphite groups as substituents, are effective ligands for catalytic allylla- tion.179, 180 Thus alkylation of dimethyl malonate with the allyl acetate 27 in the presence of the [Pd(All)Cl]2 . 2.5L catalytic system results in ee of 77%± 92%.Me Me O O P O O N But 141 R1 Me Me O O P O O N R2 R1 142a ± d R1=H: R2=But (a), Pri (b), Ph (c); R1=Me, R2=But (d). Me Me O O P O O N But 143 Ph OAc+CH2(CO2Me)2 27 CO2Me MeO2C CO2Me MeO2C + Ph Ph 144b 144a The best result was achieved in the presence of the ligand 142d, the total yield of branched (144a) and linear (144b) reaction products being equal to 75% ±92% Asymmetric induction is mainly accomplished due to the dihydrooxazole fragment. Thus for the ligand 141, the ee is 83% and for the ligand 143, it is 79%. In addition, the enantio- and regioselectivity of the process increase with an increase in the size and, especially, electronegativity of substituents at the phospho- rus atom.The results of alkylation of dimethyl malonate with the acetates 27a,b are listed in Table 2. It can be seen that both an increase in the size of substituents in the ligand and an increase in Table 2. Enantio- and regioselectivity of the alkylation of dimethyl malonate with the acetates 27 {[Pd(All)Cl]2 as the pre-catalyst}. The ratio 144a : 144b ee (%) Ligand Compound 27 a 23c 145 142a 142a 4 : 96 47 : 53 76 : 24 95 : 5 78 84 90 96 b678 the steric requirements in the compound 27 results in higher enantioselectivity and regioselectivity of the reaction. It should be noted that the selectivity increases also on passing from the ligand 23c to its analogue 145, having perfluorophenyl groups attached to phosphorus.O (C6F5)2P N 145 But Ar Ar H Me Me O Me O Me O OP O O N R2 R1 HAr Ar 146a ± e R2 R1 Ar Compound 146 HPh HPri Ph Ph Ph Ph 2-naphthyl Pri abcde Ph HPri HH Ligands 146a ± e, prepared using widespread chiral inductors, a,a,a 0,a 0-tetraaryl-1,2-dioxolano-4,5-dimethanols (TADDOL), were employed 181 to prepare Rh(I) cationic complexes 147b,c (the complex 147b was characterised by X-ray diffraction analy- sis) and to carry out hydrosilylation of a series of aliphatic and aromatic ketones, including reduction of acetophenone with diphenylsilane {with [Rh(COD)Cl]2 as the pre-catalyst}. + N [Rh(COD)Cl]2 BF¡ 146b,c 4 Rh NH4BF4, CH2Cl2 P 147b,c N =146b,c. P The best results were obtained with the ligand 146d: the ee value in the reduction of acetophenone was 88% (chemical yield 91%) and in the reduction of tert-butyl methyl ketone, the ee reached 95%.V. Conclusion The data considered here demonstrate that the research into chiral P,N-bidentate ligands is developing rather vigorously. More than 270 compounds of this type are being used in asymmetric catalysis for allylic alkylation, cross-coupling, hydrosilylation of the C=O and C=C bonds, hydroboration ± oxidation, hydroboration ± amination and for the Heck reaction. The use of rhodium and palladium complexes with optically active P,N-bidentate ligands in asymmetric catalysis is briefly summarised in Table 3. The table presents the results of the most thoroughly developed processes, namely, hydrosilylation, hydro- boration ± oxidation and allylic substitution.The enantioselective allylation was considered taking the model allylic alkylation of dimethyl malonate with the acetate 12 as an example. It follows from the data of Table 3 that phosphorus-substituted ferrocenes and dihydrooxazoles are the most effective in the hydrosilylation of the C=O bond; ferrocenylphosphinopyrazoles and QUINAP 103 are the best compounds for hydroboration ± oxidation; and all the groups of chiral P,N-bidentate compounds proved efficient for K N Gavrilov, A I Polosukhin allylic substitution. It should also be noted that ferrocenylami- noalkylphosphines are effective stereoselective ligands for some cross-coupling reactions.However, several topical problems in the chemistry of P,N- bidentate ligands are still to be solved. (i) Among the problems of design and synthesis of these ligands, preparation of compounds with a chiral carbon atom at the phosphorus centre deserves attention in the first place because at present, predominant structures are those having asymmetric carbon atoms at the nitrogen centre and/or in the bridge linking phosphorus and nitrogen. It is also important to synthesise ligands with asymmetric donor atoms, especially, with configurationally stable chiral nitrogen atoms. Combination in a chiral ligand molecule of asymmetric atoms with different absolute configura- tions and chemical natures as well as combination of different types of chirality (for example, central and planar chirality, central and axial chirality, etc.) can provide effective transfer of chiral information from the metal complex to the substrate.For exam- ple, recently, phosphinodihydrooxazoles 148a,b with a chirality axis were synthesised;182, 183 allylic amination of the acetate 27 with substituted hydrazines 183 in the presence of these compounds results in ee of up to 99%, while the allylic alkylation of the allyl acetate 12 182 gives ee of up to 91%. It was shown 182 that the binaphthalene cage is responsible for the increase in the stereo- selectivity. O O N N Ph2P Ph2P Pri Pri 148a 148b As noted above, an increase in the p-acidity of the phosphorus centre favours an increase in the chemical and optical yields in allylic substitution, hydroboration ± oxidation and hydrosilyla- tion ± oxidation of alkenes and hydrosilylation of ketones.The inductive effect of the phosphorus centre increases on passing from aminophosphines to amino phosphites or amino phospho- rous amides. In these compounds, the inductive effect can be readily optimised. In the development of new chiral P,N-bidentate ligands, optically active natural products can be used such as steroids, amino acids and alkaloids. No P,N-bidentate derivatives of phosphorous acid having a plane of chirality or a peripheral imino group are known to date. However, the synthesis of these compounds is quite practicable. It should also be noted that the synthesis of phosphites and phosphorous amides involves phosphorylation of the correspond- ing proton-donating functional groups, which proceeds, as a rule, with high yields; therefore, one can expect that a new generation of P,N-bidentate ligands would contain a phosphite (or amino phosphite) centre, possess planar or axial chirality of the carbon framework, and the nitrogen-containing fragment would be represented by an oxazoline, oxazine, imine or pyrazoline group.(ii) Regarding asymmetric catalysis, it is necessary to master those reactions in which optically active P,N-bidentate ligands have received little use, if any. This refers first of all to the addition of electrophilic reagents to alkenes (for example, hydrocyanation) and to hydroformylation and copolymerisation of alkenes or dienes in the presence of CO.Non-symmetrical phosphinophos- phites,156 i.e. systems with donor centres with different chemical natures, showed good results in these reactions. An example of using P,N-bidentate ligands in non-standard reactions is the aza Claisen rearrangement catalysed by the complexes PdCl2L [L=(S,S)-11b, 26b,c,f].184Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium Table 3. Comparative efficiency of some P,N-bidentate ligands in model catalytic reactions. Reagent Substrate PhCOMe Ph2SiH2 Ph2SiH2 Ph2SiH2 Ph2SiH2 Ph2SiH2 O PhCH=CH2 BHa OOBHa OOBHa O CH2(CO2Me)2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 PhCH=CHCH(OAc)Ph [Pd(All)Cl]2 a Followed by treatment with H2O2 and NaOH.Ph Ar PdCl2L, AgBF4 O N R The use of phosphinodihydrooxazoles as ligands in cross- coupling, hydroboration and hydrosilylation of alkenes appears quite promising. It is necessary to study already known amino phosphites, amino phosphorous amides and phosphitodihy- drooxazoles not only in allylic substitution reactions but also in all other catalytic processes in which ligands with electron-with- drawing P-donor centres can participate.These studies would substantially extend the scope of application of chiral P,N- bidentate ligands in enantioselective catalysis. Further perfection of well-developed processes is needed, first of all, in order to achieve quantitative enantioselectivity (for example,185 it is known that certification of pharmaceutical preparations requires ee 599%). The enantioselectivity can be increased by introducing non-chiral additives.185 (iii) Investigation into the mechanisms of catalytic reactions can provide the basis for recommendations on the reaction conditions and for targeted synthesis of optimal ligands. Despite the impressive progress in modern asymmetric catalysis, it largely relies on the rules obtained empirically.This is partly due to the necessity of taking into account too many factors influencing the outcome of a catalytic reaction such as the nature of the ligand and the metal, the ligand : metal and catalyst : substrate ratios, the temperature, the concentration, the nature of the solvent, the additives present, etc. and, to some extent, this is due to the absence of a detailed investigation of the mechanisms of catalytic reactions. Ligand Pre-catalyst (S,S,S)-DIPOF (16) PhCH(OH)Me PhCH(OH)Me PhCH(OH)Me PhCH(OH)Me PhCH(OH)Me [Rh(COD)Cl]2 [Rh(NBD)Cl]2 [Rh(COD)Cl]2 [Rh(COD)Cl]2 [Rh(COD)Cl]2 (S,R)-176 26r 39 146d 6g [Rh(COD)2]+BF¡4 8b [Rh(COD)2]+BF¡4 QUINAP (103) [Rh(COD)2]+SO3CF¡3 (S,S)-11b 26b,c,e 40 41 43n 75h 76b 98 QUINAP (103) 121 138b 148b O Ar Ph N R (iv) In the field of coordination chemistry, it appears topical to continue the studies on the regularities of complexation of the main groups of ligands.Unfortunately, isolation and reliable characterisation of complexes with transition metals is far from being performed for all the P,N-bidentate ligands effective in catalysis . Thus, despite the considerable contribution of iminophos- phine derivatives, such as the ligands 118a ± g or 123a ± f, to asymmetric catalysis, the coordination behaviour of these ligands has scarcely been studied. This also concerns several other effective stereoselective ligands, 138, 139 and 141 ± 145.The complex formation of related compounds 124, 125, 129 and 130 ± 135 has been extensively studied; however, no data on the use of these compounds in catalytic processes can be found. Evidently, catalytic experiment and the study of the structures of pre-catalysts and related metal complexes for each group of ligands should be combined with investigation of the coordination and catalysis mechanism. It appears useful to synthesise a series of related complexes with different ligands (as has been done, for example, for chloro carbonyl rhodium derivatives of phosphinoferrocenylpyrazoles 6a ± t 31). Comparison of the characteristic features of coordina- tion of these compounds to metals and their spectroscopic proper- ties and estimation of the electronic effects of the electron- donating centres would permit more efficient prediction of catalytic experiments and elucidation of the corresponding corre- lations.This integrated approach has been realised to some extent for ferrocenylaminophosphines (including PPFA 1 and PTFA 3), phosphinoferrocenylpyrazoles and phosphinodihydrooxazoles. However, even for these compounds, further studies are needed because P,N-bidentate coordination and chelation cannot be postulated a priori even for well-known ligands (for example, in the ligand 80b, P,Cs- rather than P,N-binding predominates). Reaction product PhCH(OH)Me PhCH(OH)Me PhCH(OH)Me PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 PhCH=CHCH(Ph)CH(CO2Me)2 679 Yield (%) Ref.[ee (%)] 45 46 90 91 181 100 (91) 90 (90) 99 (86) 84 (94) 91 (88) 68 (98.5) 31 87 100 (87) 138 71 (91) 43 51 ± 53, 56 82 83 86 112 113 113 140 151 176 182 98 (99) 99 (99) 94 (98) 99 (95) 78 (98) 70 (76) 96 (96) 82 (86) 94 (98) 85 (92) 100 (87) 99 (91)680 It is clear that the current state of asymmetric catalysis and the prospects for its development are determined by not only P,N- bidentate ligands. Considerable contribution is made by P,P-, P,O- and N,N-bidentate systems (see, for example, several stud- ies 4 ± 10, 186).It can be stated with confidence that the use of P,N-bidentate compounds opens a new stage in the development of coordination chemistry and asymmetric catalysis. References 1. A D Garnovskii, D A Garnovskii, I S Vasil'chenko, A S Burlov, A P Sadimenko, I D Sadekov Usp. Khim. 66 434 (1997) [Russ. Chem. Rev. 66 389 (1997)] 2. Yu N Kukushkin Reaktsionnaya Sposobnost' Koordinatsionnykh Soedinenii (Reactivity of Coordination Compounds) (Leningrad: Khimiya, 1987) 3. H M Colquhoun, J Holton, D J Thompson, MV Twigg New Path- way of Organic Synthesis. Practical Application of Transition Metals (New York: Plenum, 1984) 4. M Nogradi Stereoselective Synthesis (Weinheim: VCH, 1987) 5.H Brunner, W Zettlmeier Handbook of Enantioselective Catalysis with Transition Metal Compounds (Weinheim: VCH, 1993) 6. R Noyori Asymmetric Catalysis in Organic Synthesis (New York: Wiley, 1994) 7. I Ojima (Ed.) Catalytic Asymmetric Synthesis (Weinheim: VCH, 1995) 8. A Togni, T Hayashi Ferrocenes. Homogeneous Catalysis, Organic Synthesis, Material Science (Weinheim: VCH, 1995) 9. V V Dunina, I P Beletskaya Zh. Org. Khim. 28 1929 (1992) a 10. V V Dunina, I P Beletskaya Zh. Org. Khim. 28 2369 (1992) a 11. G C Lloyd-Jones, A Pfaltz Z. Naturforsch., B Chem. Sci. 50 361 (1995) 12. Y Ito,M Sawamura, T Hayashi Tetrahedron Lett. 28 6215 (1987) 13. A Togni, S D Pastor J. Org. Chem. 55 1649 (1990) 14.Y Nishibayashi, K Segawa, H Takada, K Ohe, S Uemura J. Chem. Soc., Chem. Commun. 847 (1996) 15. W R Cullen, S V Evans, N F Han, J Trotter Inorg. Chem. 26 514 (1987) 16. T Hayashi,K Yamamoto,M Kumada Tetrahedron Lett. 4405 (1974) 17. T Hayashi, M Konishi, M Fukushima, T Mise,M Kagotani, M Tajika,M Kumada J. Am. Chem. Soc. 104 180 (1982) 18. N Kumada Pure Appl. Chem. 52 669 (1980) 19. T Hayashi, M Konishi, H Ito,M Kumada J. Am. Chem. Soc. 104 4962 (1982) 20. M Uemura, H Nishimura, T Hayashi Tetrahedron Lett. 34 107 (1993) 21. W R Cullen, F W B Einstein, C-H Huang, A C Willis, E-S Yeh J. Am. Chem. Soc. 102 988 (1980) 22. T D Appleton, W R Cullen, S V Evans, T-J Kim, J Trotter J. Organomet. Chem. 279 5 (1985) 23. B Jedlicka, C Kratky,W Weissensteiner,M Widhalm J.Chem. Soc., Chem. Commun. 1329 (1993) 24. B Jedlicka, R E RuÈ lke,W Weissensteiner, R Ferna ndez-Gala n, F A Jalun, B R Manzano, J Rodrõ guez-de la Fuente, N Veldman, H Kooijman, A L Spek J. Organomet. Chem. 516 97 (1996) 25. R Ferna ndez-Gala n, F A Jalon, B R Manzano, J Rodrõ guez-de la Fuente,M Vrahami, B Jedlicka, W Weissensteiner, G Jogl Organometallics 16 3758 (1997) 26. B Jedlicka,W Weissensteiner, T Ke gl, L Kolle r J. Organomet. Chem. 563 37 (1998) 27. A Mernyi, C Kratky, W Weissensteiner, M Widhalm J. Organomet. Chem. 508 209 (1996) 28. F Go mez-de la Torre, F A Jalo n, A Lo pez-Agenjo, B R Manzano, A Rodrõ guez, T Sturm,W Weissensteiner, M Martõ nez-Ripoll Organometallics 17 4634 (1998).29. H C L Abbenhuis, U Burckhardt, V Gramlich, A Togni, A Albinati, B MuÈ ller Organometallics 13 4481 (1994) 30. A Schnyder, L Hintermann, A Togni Angew. Chem., Int. Ed. Engl. 34 931 (1995). 31. A Schnyder, A Togni, U Wiesli Organometallics 16 255 (1997) K N Gavrilov, A I Polosukhin 32. A Togni, U Burckhardt, V Gramlich, P S Pregosin, R Salzmann J. Am. Chem. Soc. 118 1031 (1996) 33. P E BloÈ chl, A Togni Organometallics 15 4125 (1996) 34. U Burckhardt, V Gramlich, P Hofmann, R Nesper, P S Pregosin, R Salzmann, A Togni Organometallics 15 3496 (1996) 35. A Togni Angew. Chem., Int. Ed. Engl. 35 1475 (1996) 36. U Burckhardt, M Baumann, A Togni Tetrahedron Asymmetry 8 155 (1997) 37. H C L Abbenhuis, U Burckhardt, V Gramlich, A Martelletti, J Spencer, I Steiner, A Togni Organometallics 15 1614 (1996) 38.U Burckhardt, M Baumann, G Trabesinger, V Cramlich, A Togni Organometallics 16 5252 (1997) 39. A Togni, R Dorta, C KoÈ llner, G Pioda Pure Appl. Chem. 70 1477 (1998) 40. C J Richards, D E Hibbs, M B Hursthouse Tetrahedron Lett. 36 3745 (1995) 41. C J Richards, A W Mulvaney Tetrahedron Asymmetry 7 1419 (1996) 42. K H Ahn, C-W Cho, J Park, S Lee Tetrahedron Asymmetry 8 1179 (1997). 43. K H Ahn, C-W Cho, J Park, S Lee Bull. Korean Chem. Soc. 18 789 (1997) 44. W Zhang, T Hirao, I Ikeda Tetrahedron Lett. 37 4545 (1996) 45. Y Nishibayashi, K Segawa, K Ohe, S Uemura Organometallics 14 5486 (1995) 46. T Hayashi, C Hayashi, Y Uozumi Tetrahedron Asymmetry 6 2503 (1995) 47.C Ganter, T Wagner Chem. Ber. 128 1157 (1995) 48. C Ganter, L Brassat, C GlinsboÈ ckel, B Ganter Organometallics 16 2862 (1997) 49. B Deschamps, L Ricard, F Mathey J. Organomet. Chem. 548 17 (1997) 50. L L Troitskaya, Z A Starikova, T V Demeshchik, V I Sokolov Izv. Akad. Nauk, Ser. Khim. 1760 (1999) b 51. J Sprinz, G Helmchen Tetrahedron Lett. 34 1769 (1993) 52. G J Dawson, C G Frost, J M J Williams, S J Coote Tetrahedron Lett. 34 3149 (1993) 53. P von Matt, A Pfaltz Angew. Chem., Int. Ed. Engl. 32 566 (1993) 54. SÏ VyskocÆ il, M Smrcina, V Hanu, M Pola sÏ ek, P KocÆ ovsky' J. Org. Chem. 63 7738 (1998) 55. C G Frost, J Howarth, J M J Williams Tetrahedron Asymmetry 3 1089 (1992) 56. J Sprinz, M Kiefer, G Helmchen, M Reggelin, G Huttner, O Walter, L Zsolnai Tetrahedron Lett.35 1523 (1994) 57. J V Allen, S J Coote, G J Dawson, C G Frost, C J Martin, J M J Williams J. Chem. Soc., Perkin Trans. 1 2065 (1994) 58. R Jumnah, A C Williams, J M J Williams Synlett 821 (1995) 59. I C Baldwin, J M J Williams, R P Beckett Tetrahedron Asymmetry 6 679 (1995) 60. I C Baldwin, J M J Williams, R P Beckett Tetrahedron Asymmetry 6 1515 (1995) 61. P von Matt, O Loiseleur, G Koch, A Pfaltz, C Lefeber, T Feucht, G Helmchen Tetrahedron Asymmetry 5 573 (1994) 62. H-J Gais, H Eichelmann, N Spalthoff, F Gerhards, M Frank, G Raabe Tetrahedron Asymmetry 9 235 (1998) 63. A Pfaltz Acta Chem. Scand. 50 189 (1996) 64. G J Dawson, J M J Williams, S J Coote Tetrahedron Lett.36 461 (1995) 65. G J Dawson, J M J Williams, S J Coote Tetrahedron Asymmetry 6 2535 (1995) 66. J F Bower, J M J Williams Synlett 685 (1996) 67. G Koch, A Pfaltz Tetrahedron Asymmetry 7 2213 (1996) 68. H Steinhagen, M Reggelin, G Helmchen Angew. Chem., Int. Ed. Engl. 36 2108 (1997) 69. H Rieck, G Helmchen Angew. Chem., Int. Ed. Engl. 34 2687 (1995) 70. P Sennhenn, B Gabler, G Helmchen Tetrahedron Lett. 35 8595 (1994) 71. S Kudis, G Helmchen Angew. Chem., Int. Ed. Engl. 37 3047 (1998) 72. S Kudis, G Helmchen Tetrahedron 54 10449 (1998) 73. S Schleich, G Helmchen Eur. J. Org. Chem. 2515 (1999) 74. O Loiseleur, P Meier, A Pfaltz Angew. Chem., Int. Ed. Engl. 35 200 (1996) 75. O Loiseleur,M Hayashi, N Schmees, A Pfaltz Synthesis 1338 (1997) 76.M Sperrle, A Aeby, G Consiglio, A Pfaltz Helv. Chim. Acta 79 1387 (1996)Chiral P,N-bidentate ligands in coordination chemistry and organic catalysis involving rhodium and palladium 77. A Aeby, G Consiglio J. Chem. Soc., Dalton Trans. 655 (1999) 78. H Eichelmann, H-J Gais Tetrahedron Asymmetry 6 643 (1995) 79. L Ripa, A Hallberg J. Org. Chem. 62 595 (1997) 80. T R Ward Organometallics 15 2836 (1996) 81. A Sudo, K Saigo J. Org. Chem. 62 5508 (1997) 82. B GlaÈ ser, H Kunz Synlett 53 (1998) 83. P A Evans, T A Brandt Tetrahedron Lett. 37 9143 (1996) 84. M Gomez-Simon, S Jansat,G Miller,D Panyella,M Font-Bardia, X Salans J. Chem. Soc., Dalton Trans. 3755 (1997) 85. S R Gilbertson, C-W T Chang J. Chem. Soc., Chem. Commun. 975 (1997) 86.S R Gilbertson, C-W T Chang J. Org. Chem. 63 8424 (1998) 87. A M Porte, J Reibenspies, K Burgess J. Am. Chem. Soc. 120 9180 (1998) 88. S R Gilbertson, D Xie Angew. Chem., Int. Ed. Engl. 38 2750 (1999) 89. L M Newman, J M J Williams, R McCague, G A Potter Tetrahedron Asymmetry 7 1597 (1996) 90. T Langer, J Janssen, G Helmchen Tetrahedron Asymmetry 7 1599 (1996) 91. A Sudo, H Yoshida, K Saigo Tetrahedron Asymmetry 8 3205 (1997) 92. A Takenaka, Y Sasada, K Yamamoto, J Tsuji Bull. Chem. Soc. Jpn. 50 3177 (1977) 93. K Yamamoto, A Tomita, J Tsuji Chem. Lett. 3 (1978) 94. U Nagel,H Menzel, P W Lednor,W Beck, A Guyot,M Bartholin Z. Naturforsch., B Chem. Sci. 36 578 (1981) 95. L Horner, G Simons Z. Naturforsch., B Chem. Sci. 39 504 (1984) 96.L Horner, G Simons Z. Naturforsch., B Chem. Sci. 39 512 (1984) 97. A Carmona, A Corma,M Iglesias, A San Jose , F Sa nchez J. Organomet. Chem. 492 11 (1995) 98. T Hayashi, M Fukushima, M Konishi, M Kumada Tetrahedron Lett. 21 79 (1980) 99. T Hayashi, M Konishi, M Fukushima, K Kanehira, T Hioki, M Kumada J. Org. Chem. 48 2195 (1983) 100. H-J Kreuzfeld, C DoÈ bler, H-P Abicht J. Organomet. Chem. 336 287 (1987) 101. G Cross, B K Vriesema, G Boven, R M Kellogg, F van Bolhuis J. Organomet. Chem. 370 357 (1989) 102. U Nagel, H G Nedden Chem. Ber. 130 385 (1997) 103. U Nagel, H G Nedden Chem. Ber. 130 989 (1997) 104. G FrancioÁ , C G Arena,M Panzalorto,G Bruno, F Faraone Inorg. Chim. Acta 277 119 (1998) 105. N C Payne, D W Stephan Inorg.Chem. 21 182 (1982) 106. A Kinting, H-J Kreuzfeld, H-P Abicht J. Organomet. Chem. 370 343 (1989) 107. I D McKay, N C Payne Can. J. Chem. 64 1930 (1986) 108. E Yu Zhorov, V A Pavlov, Yu D Koreshkov, E I Klabunovskii Izv. Akad. Nauk SSSR, Ser. Khim. 1951 (1990) b 109. E Yu Zhorov, V A Pavlov, O A Fedotova, V I Shvedov, E A Mistryukov, D N Platonov, L S Gorshkova, E I Klabunovskii Izv. Akad. Nauk SSSR, Ser. Khim. 865 (1991) b 110. E Yu Zhorov, K N Gavrilov, V A Pavlov, L S Gorshkova, E I Klabunovskii Izv. Akad. Nauk SSSR, Ser. Khim. 1464 (1991) b 111. G Brenchley, E Merifield,M Wills,M Fedouloff Tetrahedron Lett. 35 2791 (1994) 112. I C F Vasconcelos, N P Rath, C D Spilling Tetrahedron Asymmetry 9 937 (1998) 113. H Kubota, K Koga Tetrahedron Lett.35 6689 (1994) 114. P Wimmer, M Widhalm Tetrahedron Asymmetry 6 657 (1995) 115. M Bourghida,M Widhalm Tetrahedron Asymmetry 9 1073 (1998) 116. P KocÆ ovsky', SÏ VyskocÆ il, I CisarÆ ova , J Sejbal, I TisÆ lerova, M SmrcÆ ina, G C Lloyd-Jones, S C Stephen, C P Butts, M Murray, V Langer J. Am. Chem. Soc. 121 7714 (1999) 117. I Kinoshita, K Kashiwabara, J Fujita Chem. Lett. 831 (1977) 118. A de Renzi, G Morelli,M Scalone Inorg. Chim. Acta 65 L119 (1982) 119. V V Dunina, E B Golovan', N S Gulyukina, Yu K Grishin, I P Beletskaya Izv. Akad. Nauk, Ser. Khim. 1385 (1997) b 120. A Albinati, F Lianza, H Berger, P S Prigosin, H RuÈ egger, R W Kunz Inorg. Chem. 32 478 (1993) 121. C E Barclay, G Deeble, R J Doyle, S A Elix, G Salem, T L Jones, S B Wild, A C Willis J.Chem. Soc., Dalton Trans. 57 (1995) 122. R J Doyle, G Salem, A C Willis J. Chem. Soc., Dalton Trans. 2713 (1997) 681 123. C Bianchini, S Cicchi,M Peruzzini, K M Pietrusiewicz, A Brandi J. Chem. Soc., Chem. Commun. 833 (1995) 124. W de Graaf, S Harder, J Boersma, G van Koten, J A Kanters J. Organomet. Chem. 358 545 (1988) 125. K Kamikawa, S Sugimoto,M Uemura J. Org. Chem. 63 8407 (1998) 126. S D Perera, B L Shaw,M Thornton-Pett J. Chem. Soc., Dalton Trans. 999 (1992) 127. D J Allen, G M McLaughlin, G B Robertson,W L Steffen, G Salem, S B Wild Inorg. Chem. 21 1007 (1982) 128. G Salem, S B Wild Inorg. Chem. 31 581 (1992) 129. G He, S-K Loh, J J Vittal, K F Mok, P-H Leung Organometallics 17 3931 (1998) 130. B M Trost, J D Oslob J.Am. Chem. Soc. 121 3057 (1999) 131. C G Arena, F NicoloÁ , D Drommi, G Bruno, F Faraone J. Chem. Soc., Chem. Commun. 2251 (1994) 132. F Agbossou, J-F Carpentier, A Mortreux Chem. Rev. 95 2485 (1995) 133. G Chelucci, M A Cabras, C Botteghi,M Marchetti Tetrahedron Asymmetry 5 299 (1994) 134. S Stoccoro, G Chelucci, A Zucca, M A Cinellu, G Minghetti, M Manassero J. Chem. Soc., Dalton Trans. 1295 (1996) 135. G Chelucci, M A Cabras, C Botteghi, C Basoli, M Marchetti Tetrahedron Asymmetry 7 885 (1996) 136. N W Alcock, J M Brown, D I Hulmes Tetrahedron Asymmetry 4 743 (1993) 137. J M Brown, D I Hulmes, T P Layzell J. Chem. Soc., Chem. Commun. 1673 (1993) 138. J-M Valk, G A Whitlock, T P Layzell, J M Brown Tetrahedron Asymmetry 6 2593 (1995) 139. E Fernandez,M W Hooper, F I Knight, J M Brown Chem. Commun. 173 (1997) 140. J M Brown, D I Hulmes, P J Guiry Tetrahedron 50 4493 (1994) 141. J-M Valk, T D W Claridge, J M Brown, D Hibbs, MB Hursthouse Tetrahedron Asymmetry 6 2597 (1995) 142. H Doucet, J M Brown Tetrahedron Asymmetry 8 3775 (1997) 143. T D W Claridge, J M Long, J M Brown, D Hibbs, M B Hursthouse Tetrahedron 53 4035 (1997) 144. H Brunner, A F M M Rahman Chem. Ber. 117 710 (1984) 145. H Brunner, H Weber Chem. Ber. 118 3380 (1985) 146. H A Ankersmit, B H Lùken, H Kooijman, A L Spek, K Vrieze, G van Koten Inorg. Chim. Acta 252 141 (1996) 147. P Barbaro, P S Pregosin, R Salzmann, A Albinati, R W Kunz Organometallics 14 5160 (1995) 148. H Brunner, P Faustmann, A Diett, B Nuber J. Organomet. Chem. 542 255 (1997) 149. H Brunner, I Deml,W Dirnberger, B Nuber, W Reiber Eur. J. Inorg. Chem. 43 (1998) 150. H Brunner, I Deml,W Dirnberger, K-P Ittner, W Reiber, M Zimmermann Eur. J. Inorg. Chem. 51 (1999) 151. A Saitoh, T Morimoto, K Achiwa Tetrahedron Asymmetry 8 3567 (1997) 152. T Mino, W Imiya, M Yamashita Synlett 583 (1997) 153. A Saitoh, K Achiwa, T Morimoto Tetrahedron Asymmetry 9 741 (1998) 154. A Saitoh,M Misawa, T Morimoto Synlett 483 (1999) 155. K N Gavrilov, I S Mikhel' Koord. Khim. 24 883 (1998) c 156. K N Gavrilov, I S Mikhel' Koord. Khim. 25 83 (1999) c 157. D J Wink, T J Kwok, A Yee Inorg. Chem. 29 5006 (1990) 158. C A Tolman Chem. Rev. 77 313 (1977) 159. K N Gavrilov, E Yu Zhorov Zh. Neorg. Khim. 38 1544 (1993) d 160. K N Gavrilov Zh. Neorg. Khim. 42 433 (1997) d 161. K N Gavrilov, I S Mikhel' Zh. Neorg. Khim. 40 1858 (1995) d 162. K N Gavrilov, I S Mikhel', K A Lysenko, M Yu Antipin, G I Timofeeva, A I Polosukhin, A V Korostylev Zh. Neorg. Khim. 42 593 (1997) d 163. K N Gavrilov, A V Korostylev, G I Timofeeva, A I Polosukhin, O G Bondarev, P V Petrovsky, S E Lyubimov, in The VIIth Inter- national Conference `The Problems of Solvation and Complex Formation in Solution' (Abstracts of Reports), Ivanovo, 1998 p. 218 164. K N Gavrilov, A V Korostylev, G I Timofeeva, A I Polosukhin, O G Bondarev, P V Petrovskii Koord. Khim. 24 610 (1998) cK N Gavrilov, A I Polosukhin 682 165. I S Mikhel', K N Gavrilov, A I Rebrov Zh. Neorg. Khim. 44 1634 (1999) d 166. K N Gavrilov, Doctoral Thesis in Chemical Sciences, Ryasan State Pedagogical University, Ryazan, 1998 167. K N Gavrilov, A I Polosukhin, L A Lopanova, G I Timofeeva Koord. Khim. 21 218 (1995) c 168. K N Gavrilov, I S Mikhel' Koord. Khim. 23 560 (1997) c 169. S E Lyubimov, O G Bondarev, A I Polosukhin, A V Korostylev, K A Lysenko,M Yu Antipin, K N Gavrilov, in XIX Vserosiiskoe Chugaevskoe Soveshchanie po Khimii Kompleksnykh Soedinenii (Tez. Dokl.), Ivanovo, 1999 [The XIXth Chugaev Meeting on the Chemistry of Complex Compounds (Abstracts of Reports), Ivanovo, 1999] p. 189. 170. K N Gavrilov, I S Mikhel' Usp. Khim. 65 242 (1996) [Russ. Chem. Rev. 65 225 (1996)] 171. K N Gavrilov Koord. Khim. 24 136 (1998) c 172. K N Gavrilov, D V Lechkin Koord. Khim. 21 432 (1995) c 173. K N Gavrilov, I S Mikhel', D V Lechkin, A I Rebrov, G V Cherkaev, G I Timofeeva, M G Vinogradov Zh. Neorg. Khim. 42 1676 (1997)d 174. I S Mikhel', K N Gavrilov, D V Lechkin, A I Rebrov Izv. Akad. Nauk, Ser. Khim. 1416 (1997) b 175. K N Gavrilov, D V Lechkin Koord. Khim. 24 3 (1998) c 176. J-M Brunel, T Constantieux, A Labande, F Lubatti, G Buono Tetrahedron Lett. 38 5971 (1997) 177. T Constantieux, J-M Brunel, A Labande, G Buono Synlett 49 (1998) 178. G Muchow, J M Brunel,M Maffei, O Pardigon, G Buono Tetrahedron 54 10435 (1998) 179. R Pre toà t, A Pfaltz Angew. Chem., Int. Ed. Engl. 37 323 (1998) 180. R Pre toà t, G C Lloyd-Jones, A Pfaltz Pure Appl. Chem. 70 1035 (1998) 181. D K Heldmann, D Seebach Helv. Chim. Acta 82 1096 (1999) 182. M Ogasawara, K Yoshida, H Kamei, K Kato, Y Uozumi, T Hayashi Tetrahedron Asymmetry 9 1779 (1998) 183. K Selvakumar, M Valentini, M WoÈ rle, P S Pregosin, A Albinati Organometallics 18 1207 (1999) 184. Y Uozumi, K Kato, T Hayashi Tetrahedron Asymmetry 9 1065 (1998) 185. E M Vogl, H GroÈ ger,M Shibasaki Angew. Chem., Int. Ed. Engl. 38 1571 (1999) 186. B M Trost, D L van Vranken Chem. Rev. 96 395 (1996) a�Russ. J. Org. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Transl.) a�Russ. J. Coord. Chem. (Engl. Transl.) a�Russ. J. Inorg. Chem. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (8) 683 ± 704 (2000) Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands Yu F Oprunenko Contents I. Introduction II. Classification and major regularities of inter-ring haptotropic rearrangements III. Studies of mechanisms and kinetics of rearrangements in tricarbonylchromium complexes of polycyclic arenes IV. Conclusion Abstract. rear- haptotropic inter-ring of studies of results The The results of studies of inter-ring haptotropic rear- rangements polycyclic with metals transition of rangements in in p-complexes -complexes of transition metals with polycyclic aromatic of structures and Syntheses discussed. are ligands aromatic ligands are discussed. Syntheses and structures of model model compounds dynamic of classification the and surveyed are compounds are surveyed and the classification of dynamic proc- proc- esses major The proposed.is compounds these in esses in these compounds is proposed. The major regularities, regularities, mechanisms and kinetics of inter-ring haptotropic rearrange- mechanisms and kinetics of inter-ring haptotropic rearrange- ments in these systems are analysed. The bibliography includes ments in these systems are analysed. The bibliography includes 143 references 143 references. I. Introduction Diversified dynamic processes occur in molecules of different heteroorganic compounds (HOC) and, in particular, organo- metallic compounds (OMC), due to changes in the mutual arrangement of various groups in HOC molecules (rotation, small-amplitude oscillation, formation and cleavage of agostic bonds, inversion, etc.).Generally, these processes are degenerate and occur rather rapidly and hence, they are studied primarily by dynamic NMR spectroscopy.1 One of the dynamic processes most commonly observed in HOC and OMC involves rapid migration of an element (metal) together with the ligand environment from one position in the organic fragment to another. For this type of tautomerism, Academician A N Nesmeyanov introduced the term `metallo- tropic tautomerism' (metallotropy). He was the first 2 to find metallotropic rearrangements in nitrosophenyl and quinoxime compounds of mercury. OHgEt ONOHgEt NO Hoffmann 3 extended the term `sigmatropic rearrangements' to p-complexes of transition metals and introduced the term `haptotropic' rearrangements for these processes.Metallotropy is among the most important problems of chemistry because the Yu F Oprunenko Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 932 00 67. Tel. (7-095) 939 53 78. E-mail: yopr@nmr.chem.msu.su Received 4 April 2000 Uspekhi Khimii 69 (8) 744 ± 766 (2000); translated by T N Safonova #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n08ABEH000589 683 685 691 702 reactivities and catalytic activities of OMC are to a large extent determined by the tautomer ratio.4 Metallotropic tautomerism has been studied for more than 40 years.A rich variety of metallotropic rearrangements have been thoroughly studied and a number of basic concepts have been offered, for example, the concepts of dual reactivity and s,p- conjugation,5 stereochemically nonrigid organometallic com- pounds 6 and catalytic and intramolecular activation of CH bonds,7 which made a noticeable impact on the development of the general theory of chemical bonds, catalytic activity and reactivity of OMC. Degenerate s,s- and p,p-rearrangements in HOC and OMC were among the first to be discovered.8, 9 For example, s,s- rearrangements in allylic complexes of magnesium,10 boron 11 and silicon 12 ([1,3]-sigmatropic rearrangements according to Woodward ± Hoffmann) proceed through an intramolecular mechanism.13 CH2=CHCH2MRn RnMCH2CH=CH2 n=1,M=Mg; n=2,M=B; n=3,M=Si. Rearrangements of C- and O-substituted derivatives of car- bonyl compounds containingMR13 groups (M=Ge or Sn) 14, 15 as well as the nitrosophenol ± quinoxime rearrangement of mercury compounds described above (the first metallotropic rearrange- ment observed experimentally) 2 may also be assigned to this type of rearrangements.R13 MCH2CR2 R13 MOC CH2 O R2 M=Ge, Sn. Migration of the organometallic group in the compounds described above is accompanied by rearrangement of the system of multiple bonds. Cross experiments demonstrated that these processes can occur both according to intramolecular and inter- molecular mechanisms, but an intramolecular mechanism pre- vails.Formally, s,s-rearrangements are analogues of prototropic rearrangements in organic compounds, whereas s,p- and p,p- rearrangements have no analogues because migrations of organo- metallic groups in the latter cases lead to a change in the character of a delocalised multicentre bond. For example, an interesting diotropic s,p-rearrangement was observed in a binuclear thio- phene complex.16 Cr(CO)5 Mn(CO)5 S S (CO)3Mn (CO)3Cr684 The [1,5]-sigmatropic rearrangements in cyclopentadienyl and indenyl compounds of many main-group and transition metals have received the most study.17 4 3 MRn 3 4 3 4 MRn 5 2 5 ... ... 2 5 2H 1 H 1 H 1 MRn M=Si, Sn, Ge. These rearrangements were also observed in cyclohepta- trienyl 18 and cyclononatetraenyl 19 compounds.RnM H n n H MRn n=1, 2. Generally, the activation barriers to these migrations are rather low and the migrations are degenerate in the sense that the molecule either undergoes internal conversion or is converted into a mirror image. Hence, these migrations are studied by dynamic NMR spectroscopy.20 Molecules of this type can be described as `stereochemically nonrigid' in terms proposed by Muetterties.21 Cotton 22 introduced the term `fluxional molecule' for their description. Haptotropic rearrangements in monocyclic p-complexes were considered in numerous reviews and mono- graphs.23 ± 25 For p-complexes of open polyenes, rearrangements of Z2-allene 26 and Z4-triene 27 derivatives have been studied most extensively.Me Me Me Me C C C C C C Me Me Me Me MRn RnM + MRn=Fe(CO)4, Fe(CO)2(Z5-C5H5). R2 R2 R1 R1 Fe(CO)3 Fe(CO)3 p,p-Rearrangements are often observed in p-complexes of open or cyclic conjugated polyenes with transition metals. These rearrangements are very diverse.28 The Z4,Z4-rearrangement of the tricarbonyliron complex of cyclooctatetraene provides a typical example. In the course of isomerisation, the Fe(CO)3 group migrates along the system of the conjugated double bonds of the ligand.29 8 7 1 8 1 1 2 2 3 3 2 4 7 8 6 ... ... 3 4 5 6 5 5 4 7 6 Fe(CO)3 Fe(CO)3 Fe(CO)3 It should be noted that haptotropic rearrangements occur in p-complexes in which the p-ligand can be coordinated to the metal atom through several coordination sites and not all these sites are used, i.e., some double bonds are not involved in coordination.In (Z4-cyclooctatetraene)(Z6-cyclooctatetraene)iron, migra- tion of the metal atom is accompanied by a change in the mode of bonding in each ligand (i.e., the Z4,Z6-rearrangement).30 Fe Fe Yu F Oprunenko In early studies, metallotropy in OMC, including metallo- tropic rearrangements with the participation of the hydrogen atom and alkyl groups, was investigated primarily in open or monocyclic compounds.30 ± 34 Much later, inter-ring rearrange- ments, rapid within theNMRtime scale, were found, for example, in nickel complexes of naphthalene 35 and anthracene,36 phena- lenyl palladium complex 37 and dibenzenepentalenyl manganese complex.38 Formally, p-complexes of transition metals (in particular, carbonyl chromium complexes, which are the most readily avail- able and well-studied compounds) with polycyclic aromatic neutral ligands (naphthalene, acenaphthene, acenaphthylene, biphenylene, biphenyl, fluoranthene, etc.) are promising models for searching for and revealing new types of haptotropic rear- rangements. In these systems, inter-ring haptotropic rearrange- ments (IHR) involving migration of the transition metal atom together with the ligand environment between the rings of the polyaromatic ligand can proceed.39 MLn MLn Studies of the general regularities of IHR are of considerable importance because they can allow one to obtain further insight into the nature of the transition metal ± ligand bond and to elucidate the mechanisms of catalytic processes occurring in solution under the action of metal-complex catalysts because IHR are involved as particular elementary stages in these proc- esses. In 1971, Nicholas et al.40 found for the first time nondegener- ate IHR.Thus deprotonation of a solution of tricarbonyl(Z6- fluorene)chromium (1) in DMSO with potassium tert-butoxide afforded tricarbonyl(Z5-fluorenyl)chromium (2). The authors suggested that tricarbonyl(Z6-fluorenyl)chromium (3) formed in the initial step of the reaction underwent a Z6,Z5-rearrangement. The anion 3 was not detected by spectral methods and the kinetics and mechanism of its conversion into the isomeric Z5-anion 2 were not studied in detail. 2 7Cr(CO)3 7H+ 1 Cr(CO)3 7 3 Cr(CO)3 Later, inter-ring haptotropic rearrangement in the Z6-anion 3 was studied in detail by Russian 41 ± 43 and Italian 44 researchers using IR and NMR spectroscopy and the kinetic and thermody- namic parameters of this rearrangement were determined.It was found that deprotonation of the complex 1 with a base gave rise to the Z6-anion 3, which was isomerised to form the Z5-anion 2. This rearrangement is reversible and entropy-controlled. The kinetic and thermodynamic parameters (Ea=16.29 kcal mol71 and DS==719.62 e.u.) are to a large extent determined by the equilibrium between different ion pairs of the corresponding fluorenyl salts [contact ion pairs (CIP), separated ion pairs (SIP) and ionic associates]. More recently, analogous Z6,Z5-rearrange- ments were found in complexes of a series of carbo- and hetero- cyclic analogues of fluorene.45 Nondegenerate inter-ring migrations in fluorenyl complexes are fairly well studied by stationary kinetic methods.Inter-ring migrations in complexes of naphthalene and related unsubstituted polyaromatic systems are degenerate and cannot be studied by conventional procedures. The researchers hoped to find rapidInter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands migrations of the Cr(CO)3 group between two equivalent rings in the simplest model complexes, viz., tricarbonyl(Z6-naphthalene)- chromium (4) 40, 46 and related (Z6-acenaphthylene)tricarbonyl- chromium (5),47 byNMRspectroscopy.Specimens were heated to rather high temperatures; however, no dynamic processes were observed in the spectra. Until the mid-1980s, procedures for the investigation of these degenerate processes were lacking. (CO)3Cr 5 4 Cr(CO)3 As early as 1966, Deubzer, a coworker of the Nobelist Fischer, found 48 that isomeric tricarbonylchromium p-complexes of 2,3- dimethylnaphthalene 6 and 7 underwent reversible Z6,Z6-IHR (Ea&30 kcal mol71) at 90 ¡À 110 8C. Both complexes involved in the rearrangement have been isolated from the reaction mixture in pure form (by fractional recrystallisation) and identified for the first time.Me Me Me 7 6 Cr(CO)3 Me Cr(CO)3 However, these pioneering results were not published and remained unknown over 17 years until they were cited in Ref. 47. In succeeding years, inter-ring haptotropic rearrangements were studied by several research groups. Teams headed by P M Treichel (USA), E P Ku�� ndig (Switzerland), A Ceccon (Italy), Ch Elschenbroich (Germany), T A Albright (USA), M J McGlinchey (Canada), Yu A Ustynyuk and N A Ustynyuk (both from Russia) have made the major contribution to this line of investigation. II. Classification and major regularities of inter- ring haptotropic rearrangements The classification of IHR is as yet inadequate. I shall follow the classification of IHR depending on the type of bonding of the metal atom with the ligand proposed by N A Ustynyuk,43 which is, in my opinion, the most reasonable and convenient one.According to this classification, the 3>2 and 6>7 rearrange- ments are assigned to Z6,Z5- and Z6,Z6-IHR, respectively. 1. Z2,Z2-Rearrangements in nickel complexes of naphthalene and anthracene Recently, this type of rapid reversible degenerate rearrangement has been observed in phosphine complexes of nickel(0) with naphthalene 35 and anthracene.36 The reactions of [Pri2P(CH2)nPPri2]2Ni (n=2 or 3) with naphthalene 49, 50 in THF afforded complexes, which have Z2-structures [according to the solid-state 13C NMR spectra measured with the use of cross- polarisation and the magic-angle-spinning technique (CP MAS NMR)].Solid-state and solution-phase two-dimensional ex- change 13C NMR spectra measured with magnetisation transfer demonstrated that the organonickel group migrates between double bonds both within one six-membered ring (Ea are 6 and (CH2)n PPri2 Pri2P Ni Ni Ni PPri PPri 2 Pri2P Pri2P 2 (CH2)n (CH2)n n=1, 2. 23 kcal mol71 in solution and in the solid state, respectively) and between the rings. The migration between the rings is observed only in solution (Ea=15 kcal mol71). The reactions of (Bu3P)3Ni with 5-alkyl- and 5,10-dialkylan- thracenes 49 in benzene gave rise to Z2-complexes whose structures were established by CP MAS 13C NMR spectroscopy. R1 (Bu3P)2Ni R2 R1=Me, Et; R2=H, Me, Et. Based on the data of 1H and 13C NMR spectroscopy, these complexes in solutions in benzene-d6 underwent rapid Z2,Z2- rearrangements, the organometallic group `running around' the perimeter of the polycyclic ligand. The dynamic behaviour of these systems was studied using the spin saturation transfer technique and total line-shape analysis.In anthracene complexes, as in naphthalene complexes, the organonickel group was dem- onstrated to migrate within one ring and between the terminal rings both through intramolecular (DH==8 kcal mol71) and intermolecular (DH==20 kcal mol71) mechanisms. 2. Z3,Z3-Rearrangements in phenalenyl complexes of palladium and rhodium Z3,Z3-Rearrangements were studied in most detail for Z3-phe- nalenyl complexes. Thus the reactions of di-m-chlorobis(Z3-phe- nalenyl)dipalladium complexes dissolved in acetone with N,N,N0,N0-tetramethylethylenediamine and KPF6 afforded cati- onic palladium complexes of the allylic type 8a ¡À c.37, 51 At 160 8C, coalescence of the signals for the aromatic protons was observed in the 1H NMRspectrum of the complex 8a, which is indicative of a dynamic process, viz., a haptotropic shift of the organometallic group between the six-membered rings.37 Me + Me N R1 Pd N PF¡¦6Me Me R3 R2 8a ¡À c R1 Me Pd Me R3 R2 R1=R2=R3=H(a); R1=R2=R3=Me (b); R1=Me, R2=R3=H(c).In the 13C NMRspectrum of the compound 8b, the signals for the allylic carbon atoms also coalesce at 102 8C. In this complex, the rate of the haptotropic rearrangement is higher and coales- cence starts at lower temperature due apparently to the electronic effect of the methyl groups.The following activation parameters (at 90 8C) were determined from total line-shape analysis of the signals for the methyl groups in the spectra of the compounds 8b,c. DS= /e.u. DG= /kcal mol71 Compo- DH= und 8 /kcal mol71 bc 71.30.3 4.11.9 21.40.1 21.60.1 20.90.4 23.10.7 685 R1 Ni(PBu3)2 R2 + Me N PF¡¦6 N Me Ea /kcal mol71 21.60.4 23.80.7686 The reaction of phenalene with Li2PdCl4 gave rise to the Z3-phenalenyl palladium complex in 90% yield. Subsequent treat- ment of the latter with a solution of acetylacetone in benzene afforded Z3-phenalenyl complex 9 in 50% yield.52, 53 The possi- bility of Z3,Z3-IHR occurring in the complex 9 was examined.Pd(acac) Pd(acac) Pd(acac) 11 10 9 Calculations by the extended Hu�� ckel method demonstrated that structure 10 is less stable than structure 11 as the possible transition state. The difference between their energies is 1.21 kcal mol71. The energy of the structure 9 is 15.77 kcal mol71 lower than that of the structure 11. Based on the results of calculations, it was concluded that the inter-ring haptotropic shift should occur through the transition state 10. However, no dynamic processes were observed in the complex 9 by NMR spectroscopy, although Z3,Z3-IHR would be expected to occur by analogy with the complexes 8. Z3-Phenalenyl platinum complexes 12 prepared as yellow crystals by the reaction of 1-ethoxyphenalenium tetraflu- oroborate with bis(triphenylphosphine)(Z3-ethylene)platinum (the 12a : 12b ratio was 10 : 1).54 The structures of the complexes 12a,b were determined by 31P NMR spectroscopy and X-ray diffraction analysis, but their dynamic behaviour in solution was not studied and the possibility of interconversion of the complexes 12a and 12b through inter-ring haptotropic rearrangements was not proved.+ + OEt EtO BF¡¦ BF¡¦ 4 4 Pt(PPh3)2 Pt(PPh3)2 12b 12a Calculations by the extended Hu�� ckel method demonstrated that the inter-ring haptotropic rearrangement in the Z3-phena- lenyl platinum complex 13a should occur according to an intra- molecular mechanism with a relatively low activation energy (20.1 kcal mol71) through transition state 13b.55 The difference between the energies of the complexes 13a and 13b was found to be 3.83 kcal mol71.+Pt(PPh3)2 +Pt(PPh3)2 +Pt(PPh3)2 13b 13a It was postulated that the Z3-phenalenyl rhodium complex was formed in the reaction of the catalytic conversion of cyclo- prop[a]acenaphthylene into phenalene in the presence of the complex [Rh(CO)2Cl]2.56 The deuterium label introduced in the starting compound, for example, at position 7, is statistically distributed over positions 1, 3, 4, 6, 7 and 9 in the final product, viz., in phenalene. This is attributable to the fact that the intermediate Z3-phenalenyl rhodium complex undergoes Z3,Z3- IHR involving migration of the metallohydride group between the six-membered rings with the `ricochet' elimination of hydrogen.Yu F Oprunenko H [Rh] 7 [Rh(CO)2Cl]2 D ... 8 80 D H [Rh]H D 1 3 D ... 4 9 6 7 D D 3. Z5,Z5-Rearrangements in dibenzopentalenyl complexes of manganese Only one example of Z5,Z5-IHR is available in the literature. Refluxing of 9-methylindeno[1,2-a]indene or indeno[1,2-a]indene with Mn2(CO)10 in dodecane for 1.5 h afforded isomeric dibenzo- pentalenyl complexes 14 ¡À 16, which differ in the type of fusion of the indene fragments.57 Me H Mn2(CO)10 H Me Me H H + 15 Mn(CO)3 (CO)3Mn 14 ButOK Me Me 7 7 >740 8C 17 Mn(CO)3 Mn(CO)3 H H Mn2(CO)10 ButOK H H (CO)3Mn H H 16 7 18 (CO)3Mn Deprotonation of the complexes 14 and 16 with potassium tert-butoxide gave rise to complexes 17 and 18, respectively.In the compound 18, no migration of the Mn(CO)3 group, rapid within the NMR time scale, was observed, whereas the complex 17 and the related benzyl complex 19 underwent rearrangements (in the case of the benzyl derivative, the rearrangement was reversible) to form the corresponding isomeric Z5-complexes at temperatures above740 8C. Bn Bn Me Me 7 7 >740 8C Mn(CO)3 19 Mn(CO)3Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands The difference in the behaviour of the complexes 17 ± 19 was explained based on analysis of the node properties and energies of the highest occupied p-MO of the dibenzopentalenyl complexes 17 and 18 calculated by the INDO method.43 4.Z6,Z5-Rearrangements in fluorenyl complexes of chromium Z6,Z5-Inter-ring haptotropic rearrangements have been studied comprehensively for indenyl and fluorenyl complexes of transi- tion metals of the chromium and iron subgroups. For example, isomerisation in the anionic complexes of such ligands as indene, fluorene, their analogues containing different substituents in five- and six-membered rings and heterocyclic analogues of fluorene was considered in detail in reviews.41 ± 43 In the present review, these data are briefly discussed and supplemented with new results. Thus tricarbonyl(Z6-fluorene)chromium (1) is readily deprotonated with different bases, such as phosphoranes (for example, Ph3P=CH2), proton sponges, alkali metals in hexame- thylphosphoramide (HMPA), BuLi and ButOM(M=Na, K, Rb or Cs),58 in THF, ether, HMPA, dioxane and some other polar solvents.In these cases, the anion 3 is initially formed at rather low temperatures. At temperatures above 720 8C, the Z6-form 3 undergoes rapid rearrangement into the Z5-form 2. This accounts for the fact that Nicholas et al.40 failed to detect the anion 3 at room temperature after deprotonation of the neutral complex 1. k1 B 7 k2 7BH Cr(CO)3 Cr(CO)3 3 1 7 2 Cr(CO)3K/Na 2 Hg Cr(CO)3 2 20 B is base. The isomeric Z5-anion 2 was also generated in the pure form by an independent synthesis involving reduction of symmetric mercury derivative 20 with a potassium ± sodium alloy in ethereal solutions.According to the data of IR spectroscopy, salts of the isomeric anions 2 and 3 exist in solution as equilibrium mixtures of contact ion pairs (CIP) and separated ion pairs (SIP) (Tables 1 and 2).59 ± 62 In the IR spectra of these salts, the absorption bands in the region of nCO stretching vibrations of the Cr(CO)3 fragment are very characteristic, which allows one to obtain data on the type of ion pairs and to infer the position of the cation. In particular, no splitting of the E band of antisymmetric vibrations was observed in the IR spectrum of the contact ion pair formed by the anion 3. Therefore, it can be concluded that the cation in this ion pair is located above the plane of the five-membered ring of the anion in the trans position with respect to the chromium atom.M+ 7 Cr CO CO CO In the contact ion pair formed by the anion 3, the negative charge in the ring is substantially decreased due to the strong Coulomb cation ± anion interaction. As a result, the symmetry C3u of the Cr(CO)3 fragment is retained and the E band is degenerate (see Table 1). On the contrary, splitting of the E band is Table 1. IR spectra of salts of the Z6-anion 3 (nCO). Solvent Cation Solvating additive Li+ Na+ K+ Rb+ Cs+ Bun4 N+ THF HMPA THF HMPA THF THF HMPA THF THF HMPA THF THF HMPA THF THF Ph3P+Me THF HMPA 7 1927 1820, 1839 SIP 7 1928 1820, 1844 SIP 7 1925 1812, 1835 SIP 7 1928 1820, 1844 SIP DCH-18-C-6 1925 1821, 1840 SIP 7 1927 1817, 1841 SIP 7 1928 1820, 1844 SIP [2.2.2]-cryptand 1905 1822, 1842 SIP 7 1930 1840 CIP 7[2.2.2]-cryptand 1927 1820, 1844 SIP 7 1929 1840 CIP 7[2.2.2]-cryptand 1928 1821, 1843 SIP 7 1924 1829 CIP 7 1927 1832 CIP DCH-18-C-6 Note.DCH-18-K-6 is dicyclohexyl-18-crown-6. Table 2. IR spectra of salts of the Z5-anion 2 (nCO). Solvent Cation Solvating additive Li+ Na+ K+ Rb+ Cs+ Bun4 N+ THF HMPA THF HMPA THF THF HMPA THF THF HMPA THF THF HMPA THF THF Ph3P+Me THF 7777DCH-18-C-6 77[2.2.2]-cryptand 1905 1800 77[2.2.2]-cryptand 1905 1800 77[2.2.2]-cryptand 1905 1800 77 pronounced in the IR spectra of the contact ion pair formed by the anion 2, the degree of splitting depending on the nature of the cation (see Table 2).Such splitting indicates that the counterion is coordinated to the oxygen atom of one of the CO groups. 7 Cr CO CO M+ CO The anions 2 and 3 and bulky cations characterised by high charge delocalisation, such as Rb+, Cs+, Bun4 N+ and Ph3P+Me, form predominantly contact ion pairs. The Z6,Z5-inter-ring haptotropic rearrangement of the anion 3 into the anion 2 and the reverse rearrangement were studied by 1H NMR44, 60, 63 and IR spectroscopy.59, 64 For these rearrange- ments, the equilibrium constants (Keq) and the rate constants of the forward and back reactions (k1 and k2, respectively) were determined (Tables 3 and 4).The Keq values (see Table 3) indicate that the metallotropic Z6,Z5-equilibrium studied for a wide range 687 nE nA /cm71 /cm71 Type of ion pair 1928 1819, 1843 SIP 1927 1821, 1843 SIP 1927 1820, 1842 SIP nE nA Type of ion pair /cm71 /cm71 1904 1799 1904 1798 1903 1800 1904 1798 1904 1798 1904 1789 1915 1733, 1820 CIP SIP 1907 1762, 1811 CIP SIP SIP 1905 1766, 1807 CIP SIP SIP 1904 1766, 1804 CIP SIP SIP 1904 1778, 1798 CIP SIP SIP 1903 1778, 1798 CIP CIP 1905 1802688 Table 3. Equilibrium constants for Z6,Z5-IHR of salts of anions 2 and 3. Solvent Cation Keq=2k1/k2 Type of 3>2 equilibrium Li+ Na+ K+ Rb+ Cs++ 5.40 see b 5.00 13.7 4.92 14.3 4.40 5.0 4.60 1.40 3.40 SIP > SIP SIP > SIP SIP > SIP SIP > SIP SIP > SIP SIP > SIP SIP > SIP CIP > CIP SIP > SIP CIP > CIP CIP > CIP HMPA THF HMPA THF HMPA THF THFa THF THFa THF Ph3PMe aHMPA was used as the solvating additive.b The Keq value was not determined because the lithium salt of the anion 3 in a solution inTHF was unstable. Table 4. Rate constants and equilibrium constants of the 3>2 rearrange- ment.59 T /K Keq=2k1/k2 (see a) 1074k2 /s71 1074k1 /s71 THF 11.1 11.62 12.7 13.9 14.3 0.2 0.21 0.63 1.28 2.4 1.11 1.22 4.0 8.9 17.1 273 280 288.5 298 301 THF with a DCH-18-C-6 additive 3.1 4.91 5.4 0.62 2.75 18.5 273 285.5 298 0.4 1.12 6.86 a The possibility of the 3>2 rearrangement occurring in both directions was ignored in the study 59 and, hence, only the refined Keq values for the doubly degenerate process are given.of systems is very sensitive to the structure of the ion pairs. The thermodynamic data given below testify that these Z6,Z5-IHR are entropy-controlled processes. It should be noted that many reactions involving ion pairs are entropy-controlled, which is attributed to the fact that several solvent molecules are coordi- nated about the cation (solvate coat) in separated ion pairs resulting in a substantial decrease in the entropy of this state.62 The activation parameters of the 3>2 rearrangement in THF in the presence and in the absence of the solvating additive (DCH- 18-C-6) are given below: Solvent DS= /e.u.Ea /kcal mol71 DH= /kcal mol71 719.62.55 0.20.02 15.72.04 21.32.10 THF 16.32.18 THF+DCH-18-C-6 21.92.13 Cross experiments demonstrated that at moderate temper- atures Z6,Z5-IHR in the anions 2 and 3 proceed through an intramolecular mechanism. In these processes, the tricarbonyl- chromium group is not eliminated and does not go into solution to form a complex with the solvent (for example, with THF).41 5. Rearrangements in anionic tricarbonylchromium Z6- and Z5-complexes of carbo-and heterocyclic analogues of fluorene By going from the unsubstituted tricarbonylchromium complex of fluorene to its carbo- and heterocyclic analogues, one can sub- Yu F Oprunenko stantially extend the range of systems in which Z6,Z5-IHR occur and study the dependence of these processes on the electronic and steric effects of the substituents and heteroatoms.65 The introduc- tion of a heteroatom or a substituent into one of the rings of the p-ligand leads to a substantial change in the electron density in the latter.The isomer in which the organometallic group is bound to the ring possessing a higher electron density is thermodynamically more stable resulting in an increase in the content of this isomer in the equilibrium mixture. In the case of carbocyclic analogues of 9-substituted fluorenes (9-methyl-, 9-tert-butyl- and 9-phenyl- fluorenes), the ratio between the Z5 and Z6-isomers depends on the electronic and steric effects of the substituent.The donor effect of the methyl group leads to an increase in the electron density in the five-membered ring of the fluorenyl anion and, correspond- ingly, to an increase in the proportion of the Z5-isomer (Keq=[Z5]/[Z6]=10). On the contrary, the electron-withdrawing effect of the phenyl substituent leads to a decrease in the electron density in the five-membered ring and, correspondingly, to a decrease in the portion of the Z5-isomer in the equilibrium mixture (Keq=[Z5]/[Z6]=0.35). Tricarbonyl(Z6-fluoradenyl)chromium (21) formed upon deprotonation of fluoradene complex 22 does not undergo rearrangement to form the Z5-anion 23.66 This is attributable to the fact that the negative charge is delocalised over two five-membered rings and the six-membered ring fused with these rings.H 7 ButOK (CO)3Cr (CO)3Cr 21 22 7 (CO)3Cr 23 On the contrary, delocalisation of the negative charge in the five-membered ring in (Z6-indenyl)tricarbonylchromium (24) is smaller than that in the fluorenyl analogue and hence, the anion of 24 is completely rearranged into the Z5-isomer 25 in THF even at temperatures above 740 8C. The rate of the Z6,Z5-inter-ring haptotropic 24?25 rearrangement is noticeably higher than that of the 3?2 rearrangement. 7 7 25 24 Cr(CO)3 Cr(CO)3 The phenyl substituent in the five-membered ring of the complex 24 (like that in the 9-phenylfluorenyl complex) decreases the electron density in the ring and hence, the equilibrium is shifted toward the Z6-anion.43 The OÈ fele reaction of tetraphenylcyclopentadiene with Py3Cr(CO)3 afforded tricarbonyl[(Z6-1-phenyl)-2,3,4-triphenyl- cyclopentadiene]chromium (26),67 which is the nonannelated open analogue of (Z6-indenyl)tricarbonylchromium.Deprotona- tion of the latter giving rise to the Z6-anion 27 and the ability of the tricarbonylchromium group to migrate from the six-membered ring to the five-membered ring to form the Z5-anion 28 were examined. Ph Ph ButOK Ph 770 8C, THF 26 (CO)3CrInter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands Ph Ph Ph Ph 7 >40 8C Ph Ph 7Cr(CO)3 28 27 (CO)3Cr Ceccon et al.studied irreversible Z6,Z5-IHR in the phenyl- cyclopentadienyl (29), 1-phenylindenyl (30) and 9-phenylfluor- enyl (31) complexes of chromium 68 and the rearrangement in the very interesting open phenylpentadienyl system 32.69 7 7 7 30 29 31 Cr(CO)3 Cr(CO)3 32 Cr(CO)3 Cr(CO)3 The activation parameters and the rate constants (k) of Z6,Z5-IHR in the anions 29 ± 31 are given below. T /K 29: DS==77.11.1 e.u., DH==22.70.4 kcal mol71, 342 6.210.02 337 3.850.02 332 2.320.01 325 104 k /s71 1.030.03 T /K 30: DS==77.60.6 e.u., DH==22.30.4 kcal mol71, 346 12.50.2 341 7.860.03 325 4.320.02 324 104 k /s71 1.30.02 T /K 31: DS==2.60.7 e.u., DH==24.30.02 kcal mol71, 353 5.200.06 348 3.150.06 343 1.840.01 338 104 k /s71 1.080.01 The kinetic parameters (298 K) of Z6,Z5-IHR in 2-acetyl, 9-methyl- and 9-phenylfluorenyl complexes of chromium were determined.64 R1 R1 k1 7 R2 R2 k2 7Cr(CO)3 Cr(CO)3 R1=H,R2=COMe; R1=Me, R2=H;R1=Ph, R2=H. R2 R1 Keq 104 k2 /s71 104 k1 /s71 0.63 1.02 1.9 4.0 5.82 5.76 6.37 5.73 3.03 HHCOMe HMe H McGlinchey and coworkers studied 70 the rearrangements of the zwitterionic tricarbonylmanganese complexes and the anionic tricarbonylchromium complexes containing a very interesting ligand, viz., 4H-cyclopenta(phenanthrene), or the corresponding hydrogenated ligand.The possible paths of these rearrangements and the transition states were investigated experimentally as well as theoretically by the extended HuÈ ckel method.The Z6,Z5-inter- ring haptotropic rearrangements both in the manganese and chromium complexes 33 with the completely fused ligand proceed much more readily than those in the anions 34 prepared from the corresponding hydrogenated complexes. The Z6-anions 33 are kinetically stable at temperatures below 740 8C. At higher temperatures, these anions undergo rearrangements into the Z5-anions 35, whereas the rearrangements of the hydrogenated anions 34 into anions 36 occur at high temperatures. The authors attributed this fact to the retention of the stable naphthalene system of the aromatic anion in the course of IHR. 689 7 + + M(CO)3 M(CO)3 35 33 M(CO)3 7 M(CO)3 M(CO)3 34 36 7M(CO)3 M=Mn, Cr.Relatively rapid Z6,Z5-inter-ring rearrangements in the fluo- renyl complexes of manganese,71 iron,72 molybdenum and tung- sten 73 were studied. The Ea values of these processes were demonstrated to depend substantially on the nature of transition metal. The DH= values are in the range of 24 ± 25 kcal mol71 and the DS= values vary from 2 to 71.6 e.u. The rearrangements in the manganese complexes are irreversible.71 7 + L(CO)2Mn L(CO)2Mn R R R=H, But; L=CO, PBun3 . 7 + Fe(Z5-C5H5) Fe(Z5-C5H5) Investigations of azafluorenyl systems by the INDO method demonstrated 43, 74 that the introduction of the electronegative nitrogen atom into the six-membered ring leads to a decrease in electron density in the five-membered ring compared to that in the fluorenyl system, i.e., the inter-ring haptotropic rearrangement results in a decrease in the portion of the Z5-isomer.N N 7 Cr(CO)3 (CO)3Cr7 The Z6,Z5-inter-ring haptotropic rearrangements do not pro- ceed in tricarbonylchromium complexes of carbazolyl and indolyl, which is difficult to explain within the framework of the above- decribed simplified approach because the electron density in the five-membered rings of these compounds is very high. Apparently, these rearrangements do not occur due to the fact that the electron pair at the nitrogen atom in the pyrrolyl, indolyl and carbazolyl anions occupies an MO orthogonal to the p-MO of the cyclic system, whereas the electron pair at the carbon atom in the cyclopentadienyl, indenyl and fluorenyl anions is involved in a p-MO participating in bonding with the metal atom.Experimental investigations of the structures and properties of isomeric Z6- and Z5-fluorenyl complexes and their analogues and examination of the mechanism of Z6,Z5-IHR have been summarised.74 ± 79 Analysis of the above-considered data demon- strates that the structure of the p-ligand, the nature of the metal atom and its ligand environment affect substantially the metal- lotropic equilibrium, the site of location of the organometallic group depending to a large extent on the electron density in the five-membered ring. Apparently, the determination of enthalpies of isomeric complexes and transition states using high-level quantum-chemical calculations can improve the reliability of the prediction of the results of IHP.80690 6.Z6,Z6-Rearrangements in tricarbonylchromium complexes of naphthalene and its analogues Deubzer 48 demonstrated that isomerisation of tricarbonylchro- mium complexes of 2,3-dimethylnaphthalene, which is the first example of Z6,Z6-IHR available in the literature, proceeds slowly and is characterised by a high activation barrier. These results were indirectly confirmed in studies performed by Nicholas,40 KuÈ ndig 46 and Albright 47 with their coworkers. The authors failed to detect the haptotropic rearrangement in tricarbonylchromium complexes of Z6-naphthalene and Z6-acenaphthylene by dynamic NMR spectroscopy upon heating of specimens in the probe of an NMR spectrometer.Analogously, inter-ring rearrangements, rapid within the NMR time scale, were not detected at relatively high temperatures in some other naphthalene complexes, for example, in [(Z6-C10H8)Ir(Z5-C5Me5)]2+ (70 8C, CF3CO2H) 77 and (Z6-C10H8)2Cr (130 8C, C6D6).46 Hence, the only possible strategy for studying Z6,Z6-rear- rangements in p-complexes of naphthalene involves the removal of degeneration in these systems, i.e., the use of unsymmetrically substituted compounds, and their study by stationary kinetics methods. This approach was developed in detail only in the mid- 1980s.81 ± 84 Initially, procedures have been proposed for the regioselective synthesis of tricarbonylchromium complexes of substituted naphthalenes containing a label at a particular posi- tion of the coordinated or uncoordinated ring of the naphthalene ligand.Substituted naphthalenes can form tricarbonylchromium complexes upon heating with Cr(CO)6,85 including reactions in the presence of polar additives 86 as catalysts. For example, boiling of Cr(CO)6 with naphthalene in decalin in the presence of ethyl formate afforded (Z6-C10H8)Cr(CO)3 in 88% yield.87 Tricarbo- nylchromium complexes of substituted naphthalenes were also formed in the reactions of the ligand with (NH3)3Cr(CO)3 (the Rausch reaction 88) and with the Py3Cr(CO)3±BF3 . OEt2 system (the OÈ fele reaction 89, 90). Thus the reaction of fluoranthene with Cr(CO)6 performed under conditions of thermodynamic control gave rise to Z6-complex 37 in which the Cr(CO)3 is located at the benzene ring.91 The synthesis with the use of the Py3Cr(CO)3±BF3 .OEt2 system yielded complex 38 in which the Cr(CO)3 is located at the naphthalene fragment. The latter reaction also afforded a small admixture (<5%) of the com- pound 37, which was readily separated from the complex 38 by recrystallisation. Cr(CO)3 Cr(CO)6 174 8C, 5 h 37 Py3Cr(CO)3, BF3 . OEt2 25 8C, 1 h +37 38 Cr(CO)3 A convenient alternative procedure was developed for the synthesis of substituted naphthalenes. This procedure involves the migration of the tricarbonylchromium group from readily acces- sible (Z6-C10H8)Cr(CO)3 to another polyaromatic ligand.46 The result of this complexation reaction depends on the conditions of control (either kinetic or thermodynamic), which can be used for preparative purposes.However, these methods for the synthesis of complexes of monosubstituted naphthalenes generally give rise to mixtures of both possible isomers 39 and 40 containing the tricarbonylchromium group either in the substituted or unsubsti- tuted rings. Yu F Oprunenko R R R Cr(CO)6 or + Py3Cr(CO)3±BF3 . OEt2 40 (CO)3Cr Cr(CO)3 39 For alkyl-substituted complexes, the 39 : 40 ratio is close to unity and the isomers can be separated by chromatography (TLC or HPLC92). It was demonstrated 93 that the very high regiose- lectivity of the OÈ fele reaction can be achieved if the substituent possesses pronounced electron-withdrawing or electron-donating properties.This behaviour agrees well with the results of calcu- lations of the relative energies of isomeric tricarbonylchromium complexes of monosubstituted naphthalenes within the frame- work of density functional theory (DFT).80 Thus under the conditions of the OÈ fele reaction, complexes of the type 39 are predominantly formed if R=Br, whereas com- plexes of the type 40 are obtained as the major products if R=SiMe3 and, particularly, if R=SnMe3. A thorough search for the reaction conditions enables one to achieve satisfactory yields (50% and higher) The most general and versatile procedure for the regioselective synthesis of substituted complexes 39 and 40 involves the follow- ing reaction sequence.Initially, 1- and 2-bromo- or 1- and 2-trimethylstannylnaphthalenes are converted into the corre- sponding tricarbonylchromium complexes 39a,b or 40a,b, respec- tively, by the OÈ fele reaction. R Br Br a b, c Cr(CO)3 39c ± l Cr(CO)3 39a,b b, d R SnMe3 SnMe3 a b, c 40c ± j 40a,b Cr(CO)3 Cr(CO)3 39: R=1-Br (a), 2-Br (b), 1-D (c), 2-D (d), 1-Me (e), 2-Me (f ), 1-SiMe3 (g), 2-SiMe3 (h), 1-SnMe3 ( i ), 2-SnMe3 ( j ), 1-Cl (k), 2-Cl ( l ); 40: R=1-SnMe3 (a), 2-SnMe3 (b), 1-D (c), 2-D (d), 1-Me (e), 2-Me ( f ), 1-SiMe3 (g), 2-SiMe3 (h), 1-Cl ( i ), 2-Cl ( j ); (a) Py3Cr(CO)3±Et2O. BF3; (b) BuLi; (c) RX; (d) Me3SnCl. In the reactions of these complexes with BuLi in ether at 770 8C, the Br or SnMe3 substituents are readily replaced by lithium. Under these conditions, the resulting organolithium compounds neither decompose nor undergo isomerisation.Sub- sequently, the latter compounds react with various electrophilic reagents to give the complexes 39c ± l or 40c ± j in yields varying from 50% (R=Cl) to 90% (R=D). It appeared that the reaction must be performed in ether because this reaction in THF is essentially complicated by intramolecular and intermolecular Li7H exchange (see below).94 One of procedures for the selective introduction of a substitu- ent into the coordinated ring involves direct metallation of tricarbonyl(Z6-naphthalene)chromium (4) followed by treatment of the resulting organolithium compound with an electrophilic reagent.93 It is known that coordination with the tricarbonylchro- mium group leads to a substantial increase in the acidity of the arene protons (the pKa values decrease by 6 ± 7 95).The reaction with the use of BuLi as the metallating agent in THF at 778 8C proceeded nonselectively both at positions 1 and 2 of the coordi- nated ring to form mixtures of complexes following reactions with electrophiles.Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands Li Li RX BuLi + THF,778 8C 4 (CO)3Cr (CO)3Cr (CO)3Cr R R + B A (CO)3Cr (CO)3Cr RX=D2O, MeI, Me3SiCl, Me3SnCl. If lithium salts are generated at a temperature below778 8C, the ratio between the isomers A and B in the mixture depends on the nature of the substituent.Total yield (%) The ratio A: B RX 0.80 0.56 0.44 0.31 87 66 74 45 D2O MeI Me3SnCl Me3SiCl The replacement proceeded regioselectively at position 2 (the yield was 20%) only in the reactions of lithium salts with I2.I 1) BuLi, THF,778 8C 2) I2 4 (CO)3Cr Cr(CO)3 KuÈ ndig et al. demonstrated 83 that the use of sterically hindered 2,2,6,6-tetramethylpiperidyllithium instead of BuLi made it possible to perform regioselective metallation of tricarbo- nyl(Z6-naphthalene)chromium. In this case, the replacement occurred exclusively at position 2, which may be due to steric hindrances to electrophilic attack at position 1. III. Studies of mechanisms and kinetics of rearrangements in tricarbonylchromium complexes of polycyclic arenes Elucidation of the mechanism is the vital problem in studies of chemical processes. For inter-ring haptotropic rearrangements in tricarbonylchromium complexes of polycyclic arenes, it is impor- tant to determine whether the reaction proceeds through an intramolecular or intermolecular dissociation mechanism. In the first case, the migration of the Cr(CO)3 group along the ligand plane occurs without the cleavage of bonds with the ligand.In the second case, the Cr(CO)3 group is eliminated from one ring of the ligand and passes into solution. Then this group is coordinated by solvent molecules containing a heteroatom followed by addition to another ring of the same ligand or to another ligand.Bimo- lecular reactions of the complexes can proceed as well (either with each other 96 or with free ligands formed upon thermal destruction of the complexes). The latter are accompanied by the migration of the tricarbonylchromium group in the ligand, i.e., IHR can occur through an intermolecular exchange mechanism. Previously, the intramolecular mechanism of Z6,Z5-IHR in isomeric fluorenyl complexes 2 and 3 was confirmed by cross experiments.42, 43 An analogous approach was used in studies of the mechanism of Z6,Z5-IHR in tricarbonylchromium complexes of naphthalene derivatives. For this purpose, the redistribution of the tricarbonylchromium group between naphthalene and per- deuterionaphthalene was examined.93 It was found that even prolonged heating (90 8C, >40 h) of the complex 39e with perdeuterionaphthalene in anhydrous decane did not lead to migration of the tricarbonylchromium group to perdeuterionaph- thalene.Although the 39e>40e inter-ring haptotropic migration 691 occurred under these conditions, the (Z6-C10D8)Cr(CO)3 complex was not formed. According to the data of mass spectrometry with the use of chemical ionisation, heating of (Z6-C10H8)Cr(CO)3 in the presence of perdeuterionaphthalene at 130 8C (i.e., at a temperature which is substantially higher than that required for IHR to proceed at a noticeable rate) for 6 h afforded only a small amount (< 5%) of (Z6-C10D8)Cr(CO)3.93 However, in the reac- tions with the use of benzene derivatives (which coordinate the tricarbonylchromium group much more readily than polycyclic aromatic compounds) as the solvent, the migration of the Cr(CO)3 group to monocyclic arene was observed even at 90 ± 130 8C and monocyclic complexes were obtained in yields of up to 20%.The complete replacement of naphthalene by monocyclic arene was achieved only upon heating of (Z6-C10H8)Cr(CO)3 in unsubsti- tuted benzene (140 8C, 6 h). Therefore, even before IHR were found to occur in naphtha- lene complexes, investigations of these compounds have demon- strated that isomerisation can proceed according to both intramolecular and intermolecular mechanisms depending on the nature of the solvent. Later, the most convincing evidence that IHR in tricarbonylchromium complexes of monosubstituted naphthalenes proceeds through an intramolecular mechanism was obtained with the use of enantiomerically pure compounds.Since chiral organometallic compounds are used in asymmet- ric synthesis and catalysis, they attract growing interest of researchers in organometallic chemistry.97 Although the use of chiral compounds for studying the mechanisms of purely organic reactions has long been a classical method, this approach to investigation of the mechanism of IHR has been used only recently.98 With the aim of elucidating the mechanism of Z6,Z6- IHR in tricarbonylchromium complexes of naphthalene deriva- tives possessing planar chirality, the ability of their individual enantiomers to retain or lose optical activity in the course of the rearrangements was examined.If Z6,Z6-IHR proceeds according to an intramolecular mechanism, the optical activity should be retained and isomeric complexes, for example, 39h and 40h (for which this process was studied in most detail), should be present in the equilibrium mixture only as either R or S forms. Me3Si SiMe3 Cr(CO)3 (CO)3Cr (S)-39h (R)-39h Me3Si SiMe3 (CO)3Cr Cr(CO)3 (S)-40h (R)-40h Mirror plane If the rearrangement proceeds through either a dissociation or exchange intermolecular mechanism, the individual enantiomer should undergo racemisation due to the equally probable attacks of the tricarbonylchromium group [free (`hot'), s-bonded to the solvent molecule containing a heteroatom or p-bonded to the solvent molecule, which does not contain a heteroatom (benzene or toluene)] on the arene from opposite sides.In this case, both R R (Solv)nCr(CO)3+ Cr(CO)3 RCr(CO)3 Solv=THF, PhH, PhMe.692 isomeric complexes containing the Cr(CO)3 group in either sub- stituted or unsubstituted ring should be present in the equilibrium mixture both as the R and S forms. Some planar-chiral tricarbonylchromium complexes of sub- stituted naphthalenes can be separated by HPLC on an optical Chiracel OD column (Diacel, Japan).99 Previously,100 it had been demonstrated that this column, with modified cellulose as the support, is very efficient for separation of enantiomers of tran- sition metal complexes, including tricarbonylchromium com- plexes of arenes.101 For a number of complexes of monosubstituted naphthalenes, very high selectivity factors of separation (a) a �� t2 ¡¦ t0 , t1 ¡¦ t0 were achieved, where t0 is the retention time of the compound not retained (3.35 min) and t1 and t2 are the retention times of the enantiomers, which are eluted successively.For example, the selectivity for the complex 40b was 5.23! The results of HPLC are given in Table 5. Table 5. Data on separation of enantiomers of tricarbonylchromium complexes of naphthalene derivatives by HPLC on a ChiracelODcolumn. t2 t1 a /min /min Com- plex 39 1.07 1.00 1.33 1.37 1.62 1.28 1.16 1.62 14.00 11.30 8.66 10.64 17.37 14.69 8.00 8.06 13.32 11.30 7.33 8.65 12.00 12.2 7.34 6.26 lfhjkeigNote. The rate of the mobile phase [9 : 1 (v/v) n-hexane ¡À propan-2-ol] was 10 ml min71.To elucidate the mechanism of IHR in the tricarbonylchro- mium complexes of monosubstituted naphthalenes, 39h and 40h, the possibility of their enantiomers retaining optical activity in the course of the rearrangements (85 8C,C6F6 or toluene orC6F6 with 5% of THF) was examined. The chromatogram (Fig. 1 a) of one b a 1 1 4 23100 0 5 10 min 0 5 10 min Figure 1. Chromatograms (HPLC on a Chiracel OD column) of the enantiomers of 39h in C6F6 without heating (a) and with heating at 85 8C for 40 h in C6F6 (b), toluene (c) and the C6F6 ¡ÀTHF system (95 : 5) (d). Peaks (1) and (4) correspond to enantiomers of 39h and peaks (2) and (3) correspond to enantiomers of 40h.The integrated intensities of peaks (1) and (4) and of peaks (2) and (3) on the chromatogram (d) are equal. t2 t1 a Com- plex 40 /min /min 10.35 1.11 12.32 1.09 14.66 4.27 20.00 5.23 14.30 1.65 18.00 1.76 15.20 2.29 1.81 9.65 11.53 6.00 6.53 10.00 11.66 8.53 6.66 jfhbieag 9.35 d c 1 1 4 2 2 3 4 3 0 5 10 min 0 5 10 min Yu F Oprunenko of the enantiomers of the complex 39h in C6F6 was obtained. [It should be noted that the absolute configuration was determined (for example, by X-ray diffraction analysis) for none of the complexes.] As the inter-ring haptotropic rearrangement proceeds (C6F6, 85 8C, 40 h), two intense peaks appear on the chromato- gram.One of them corresponds to an enantiomer of 39h (peak 1) and the second peak corresponds to an enantiomer of 40h (peak 2), i.e., the 39h>40h equilibrium occurs (Keq= [40h]/[39h]=17.26). The portions of enantiomers of 39h (peak 3) and 40h (peak 4) with another configurations were 41% (Fig. 1 b). Analogous results were obtained in the case of the 39h>40h rearrangement in decane. The absence of racemisation proves that IHR in inert non- solvatsolvents proceeds through an intramolecular mecha- nism without participation of dissociation and exchange processes. In this case, the Cr(CO)3 group `slips' along the plane of the p-ligand, bonding with the ligand being retained. However, the degree of racemisation in pure toluene under the same conditions is *20% (Fig. 1 c), i.e., IHP in aromatic solvents proceeds to some extent through an intermolecular exchange mechanism.Me SiMe3+ Cr(CO)3 39h Me SiMe3 SiMe3+ 40h Cr(CO)3 Cr(CO)3 In the reaction performed under the same conditions but in C6F6 with the addition of 5% of THF, the complexes 39h and 40h underwent virtually complete racemisation (575%) (Fig. 1 d). For some other complexes, analogous results were achieved upon addition of 10% of dibutyl ether to C6F6. Therefore, thermally induced Z6,Z6-IHR in tricarbonylchromium complexes of naph- thalene derivatives in the presence of solvating solvents proceeds primarily by the dissociation intermolecular mechanism. The use of chiral complexes shows considerable promise for elucidation of fine details of the mechanism of organometallic reactions, in particular, IHR.For example, studies of racemisa- tion of some complexes in the course of their rearrangements made it possible to determine the kinetic and thermodynamic parameters of IHP in the cases where these parameters cannot be obtained by other methods. Based on the results of calculations by the extended Hu�� ckel method, Albright et al.47 suggested that the presence of an exocyclic double bond should substantially promote Z6,Z6-IHR in tricarbonylchromium complex of acenaphthylene 5. The authors attempted to detect degenerate Z6,Z6-IHR by dynamic NMR spectroscopy upon heating of the complex 5 in deuteriode- calin to 160 8C.However, no broadening of lines caused by migration of the tricarbonylchromium group between the equiv- alent six-membered rings was observed. Apparently, the activa- tion barrier is such that this rearrangement cannot be studied by two-dimensional NMR spectroscopy. However, the rearrange- ment can be examined without introduction of a label into the coordinated ring. The enantiomers of the complex 5 can be separated and each isomer should undergo racemisation in the course of IHR, which, in principle, allows one to determine the rate constant of the rearrangement as the rate constant of racemisation. (S)-5 Cr(CO)3 (CO)3Cr (R)-5Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands 1.Z6,Z6-Rearrangements in tricarbonylchromium complexes of naphthalene derivatives The kinetics of IHR in a number of tricarbonylchromium com- plexes of monosubstituted naphthalenes was studied by NMR spectroscopy. The rearrangements were performed in C6F6 at 85 8C.94 The 2H{1H} NMR spectra of the deuterium-containing complexes 39c,d and 40c,d were measured. For the remaining complexes 39 and 40, the 1H NMR spectra were recorded. The signals in the NMR spectra were repeatedly integrated and averaged. Kinetic studies demonstrated that the parameters obtained are well reproduced for pairs of the complexes, which differ in the position of the Cr(CO)3 group, and are independent of a partic- ular isomer chosen as the starting one.The typical kinetic curves for a pair of the complexes 39g and 40g are shown in Fig. 2. c (%) 100 2 80 60 40 1 200 t /h 80 40 20 60 Figure 2. Kinetic curves of isomerisation of the tricarbonylchromium complexes of 1-trimethylsilylnaphthalene 39g (1) and 40g (2) (starting from 39g). The 2-iodonaphthalene complex 40k, unlike the remaining tricarbonylchromium complexes of substituted naphthalenes, was demonstrated to be irreversibly rearranged to the isomeric com- plex 39m. However, attempts to measure the kinetic parameters of the rearrangement failed due to instability of both complexes in all solvents under study. I I 39m 40k (CO)3Cr Cr(CO)3 It should be noted that no noticeable destruction of the remaining compounds accompanied by the appearance of free ligands in the reaction mixtures was observed.In all cases, the rearrangements Table 6. Rate constants, equilibrium constants and free activation energies of IHR in the complexes 39 (C6F6, 85 8C).94 Complex Keq 105 k2 105 k1 /s71 /s71 DG=(358) /kcal mol71 29.40.6 29.2 a 28.51.2 28.70.7 0.510.02 1.010.10 1.450.07 0.500.02 0.900.01 1.820.10 3.020.24 2.400.02 1.77 1.80 2.09 4.80 17.26 5.00 13.26 0.034 0.12 29.10.6 29.50.5 28.80.7 30.60.4 0.110.01 0.140.01 0.170.01 4.360.06 1.930.03 0.700.01 2.240.04 0.150.01 30.20.8 3.460.06 0.400.01 39e 39f 39f 39g 39h 39i 39j 39k 39l a Calculations were performed based on the data reported in Ref.83; cyclohexane-d12 was used as the solvent. 693 k1 40 39 k2 are described with a high degree of accuracy by a model of a reversible first-order reaction. The kinetic and thermodynamic parameters of the rearrangements of the complexes 39 in solutions in C6F6 are given in Table 6. 2. Z6,Z6-Rearrangements in tricarbonylchromium complexes of naphthalene-related ligands The kinetics of rearrangements in isomeric (Z6-dimethylsilaace- naphthene)tricarbonylchromium complexes 41 and 42, related to naphthalene complexes, was studied. The complexes 41 and 42 were prepared starting from dimethylsilaacenaphthene by the O�� fele reaction in 76% yield.102 SiMe2 SiMe2 SiMe2 Py3Cr(CO)3 + 42 41 Cr(CO)3 (CO)3Cr The 41 : 42 ratio was equal to 1 : 6.i.e., as discussed above, the tricarbonylchromium group tends to occupy the position in the ring containing an electron-donating substituent. Under condi- tions of TLC, the complexes 41 and 42 are characterised by virtually identical mobilities due to which they were separated by HPLC. At 75 ¡À 130 8C, the kinetically less stable complex 41 in C6F6 underwent rearrangement to the thermodynamically more stable complex 42. k1 41 42. k2 The content of the complex 42 in the equilibrium mixture was higher than 70%, which made it possible to measure a large number of points for the time dependence of the concentration, thus improving the accuracy and reliability of the experimental data. The kinetic data measured for the forward and back processes agree well with each other.103 The relative concentrations of the complexes 41 and 42 were determined by repeated integration followed by averaging of the well-resolved signals for the protons of the methyl groups in the 1H NMR spectra.The measurements were carried out in the ln =(k1+k2)Dt, temperature range of 105 ¡À 118 8C because the rate of IHR at temperatures below 100 8C is very low, whereas at temperatures above 118 8C the complexes decompose (in theNMRspectrum, a signal for the protons of the methyl groups of the ligand appears at d 0.52). The dependence of the concentration of the isomer 42 on the time (t) was calculated according to the equation for a reversible first-order reaction:104 A0 ¡¦ Ae At ¡¦ Ae where Ae, A0 and At are the equilibrium, initial and actual concentrations of the complex 42, respectively.The linear kinetic dependences were processed by the least-squares method. The kinetic dependences in the above-mentioned temperature range are shown in Fig. 3. The rate constants for the back and forward reactions 41>42 are given in Table 7. The activation parameters of this rearrangement are given below. Reaction Ea DS= /e.u. DG= /kcal mol71 DH= /kcal mol71 /kcal mol71 41?42 42?41 26.10.8 25.90.8 78.10.2 30.01.0 710.40.2 30.51.0 26.90.8 26.70.8 These results agree well with those reported in the literature for the related systems, in particular, for tricarbonylchromium complexes of methyl- 83 and methoxynaphthalenes.84 The results of studies are unambiguously indicative of the absence of the694 ln A0 ¡¦ Ae At ¡¦ Ae 3 321 4 Figure 3.Kinetic first-order dependences obtained for IHR in the tricarbonylchromium complexes of Z6-dimethylsilaacenaphthene 41 and 42 (starting from 41) at 105 (1), 110 (2) and 118 8C (3). Table 7. Rate constants of IHR in the tricarbonylchromium complexes of Z6-dimethylsilaacenaphthene 41 and 42. Rate constant Temperature /8C /s71 85 105 k1 1.150.10 a 11.80.4 14.01.2 37.31.1 105 k2 0.640.06 a 5.00.5 7.680.8 a The rate constants were calculated taking into account the thermody- namic parameters. contribution of the intermolecular dissociation mechanism to the IHR under examination.This is also confirmed by the relatively small negative entropies of activation both of the back and forward reactions because dissociation processes are generally characterised by higher negative DS= values.61 The Z6,Z6-inter-ring haptotropic rearrangements in a number of naphthalene-related tricarbonylchromium complexes in degassed anhydrous decane (an inert nonsolvating solvent) were examined.105, 106 For example, four possible isomeric tricarbonyl- chromium Z6-complexes 43 ¡À 46 (two isomers with the exo con- figuration of the methyl group and two isomers with the endo configuration) were obtained starting from 1-methylacenaph- thene under the conditions of the O�� fele reaction 105 in 37% total yield; the 43 : 44 : 45 : 46 ratio was 1 : 1 : 0.8 : 0.7. The complexes 43 and 44 were isolated in the individual state by HPLC.Attempts to separate the complexes 45 and 46 failed, but their mixture enriched with the isomer 45 was obtained (45 : 46=3 : 1). Me H Py3Cr(CO)3 (CO)3Cr H Me + + 45 (CO)3Cr The individual complex 43 was prepared by deprotonation of (Z6-acenaphthene)tricarbonylchromium (47) followed by the reaction of the resulting anion 48 with MeI. 2 1 16 103 t /s 12 8 118 110 105 Me Me H H + + 44 43 Cr(CO)3 H Me 46 Cr(CO)3 Yu F Oprunenko 7 MeI BuLi 43 48 47 (CO)3Cr (CO)3Cr The electron-withdrawing tricarbonylchromium group sub- stantially increases the acidity of the protons bound to the sp3- hybridised carbon atoms as a result of which deprotonation proceeds regioselectively only at the CH2 group that is located more closely to the substituted ring.The reaction of the anion 48 with MeI is stereoselective. The methyl group attacks the neg- atively charged carbon atom from the side opposite to the bulky tricarbonylchromium group. The low yield (23%) of the target product 43 indicates that the anion 48 is unstable and decomposes to form, apparently, the corresponding carbanion. This agrees with the fact that 1-methylacenaphthene was isolated from the reaction mixture in a substantial amount. Previously, the stereo- and regioselectivity of such reactions was established for deriva- tives of fluorene 78 and indene 79 by 1H NMR spectroscopy (the ASIS effect) and X-ray diffraction analysis.In the case of the complex 43, irradiation of the methyl multiplet in the nuclear Overhauser effect (NOE) experiment led to an increase in the intensity of the corresponding doublet for the protons of the coordinated ring, which proves that the methyl group is located in the vicinity of this ring. In addition, the signal for the protons of the methyl group in the complex 43 is observed at higher field than those in the complexes 44 ¡À 46. Based on the ASIS effect,70, 79 this fact is indicative of the exo structure of 43. The Z6,Z6-inter-ring haptotropic rearrangement k1 44 43 proceeded in solution in decane at 90 8C with a noticeable rate [k1=1.0061075 s71, Keq=1.13, DG=(363)=29.7 kcal mol71] with retention of the exo configuration of the methyl group.Under the same conditions, the rearrangements of the complexes 45 and 46 proceeded with retention of the endo configuration of the methyl group. Isomerisation was monitored by HPLC. The addition of 10% of THF to decane resulted in the formation of a mixture of all four possible stereoisomers without regard to the particular isomer used as the starting compound. This is unam- biguous evidence that the rearrangements in a nonsolvating medium proceed through an intramolecular mechanism, whereas the rearrangements in the presence of solvating solvents occur according to an intermolecular mechanism.105 These results agree with the data of more recent studies confirming the retention of optical activity of enantiomerically pure tricarbonylchromium complexes of naphthalene derivatives in the course of Z6,Z6-IHR.98 The Z6,Z6-inter-ring haptotropic rearrangements in isomeric tricarbonylchromium complexes of methylacenaphthylene (49 and 50) were examined.106 A mixture of these complexes was obtained by the O�� fele reaction in 78% yield (the 49 : 50 ratio was 45 : 55).The compounds 49 and 50 were isolated in pure form by HPLC. Me Me Me Py3Cr(CO)3 + 50 49 Cr(CO)3 (CO)3Cr The necessity of studying IHR in methylated derivatives of tricarbonylchromium complexes of acenaphthylene stems from the fact that attempts to introduce a label directly into the six- membered ring of (Z6-acenaphthylene)tricarbonylchromium (5) 107 [as in the case of tricarbonyl(Z6-naphthalene)chromium] failed. The course of isomerisationInter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands k1 50 49 in solution in decane at 90 8C was monitored by NMR spectro- scopy and HPLC.The rearrangement proceeded somewhat more [k1=2.2961075 s71, rapidly DG=(363)= Keq=1.18, 29.1 kcal mol71] than those in the case of analogous acenaph- thene or naphthalene complexes. This fact agrees with the results of theoretical analysis.47 In the cited study, potential energy surfaces were calculated by the extended HuÈ ckel method and it was demonstrated that the presence of an exocyclic double bond in the polyaromatic ligand should lead to a decrease in the activation barrier to IHR.1,1-Dimethyl-1-silaphenalene (51) also contains an exocyclic double bond and two different aromatic fragments suitable for coordination to the transition metal atom. Hence, one would expect that Z6,Z6-IHR will also readily proceed in the correspond- ing isomeric tricarbonylchromium complexes. The compositions of the products of the reaction of 1,1- dimethyl-1-silaphenalene (51) with chromium carbonyls per- formed under different conditions were examined.108 It was found that the reaction of the compound 51 with (NH3)3Cr(CO)3 in boiling dioxane or with (MeCN)3Cr(CO)3 in boiling THF afforded complex 52 as the only product. According to the data of 1H and 13C NMR spectroscopy, the tricarbonylchromium group in the complex 52 is coordinated at the silicon-substituted ring.Cr(CO)3 a or b SiMe2 SiMe2 52 51 (a) (NH3)3Cr(CO)3, dioxane, 100 8C (42%); (b) (MeCN)3Cr(CO)3, THF, 64 8C (6%). Both possible isomers 52 and 53 were formed only when the reaction temperature was decreased. The reaction of the com- pound 51 with (MeCN)3Cr(CO)3 at 34 8C gave rise to a mixture of the complexes 52 and 53 in a total yield of41%. The reaction of the compound 51 with Py3Cr(CO)3 at 0 8C afforded a mixture of the isomers in 54% yield with the complex 53 predominating. a or b 51 52 + (CO)3Cr SiMe2 53 (a) (MeCN)3Cr(CO)3, 34 8C (1%) (52 : 53=3 : 2); (b) (Py)3Cr(CO)3, 0 8C, 1 h (54%) (52 : 53=2 : 3). This is yet another example of the effect of thermodynamic and kinetic factors on the formation of tricarbonylchromium com- plexes with polycyclic aromatic ligands.The complexes 52 and 53 were identified by 1Hand 13C NMR spectroscopy and mass spectrometry. Since the complex 53, unlike the complex 52, is unstable under conditions of TLC, it was not isolated in pure form. The complex 52 decomposes upon storage of a mixture of the complexes even in the solid state. Heating of a mixture of the complexes 52 and 53 in the C6F6±C6D12 system or in heptane to 90 8C for 5 h resulted in the complete rearrangement of the complex 53 to form the thermodynamically more stable complex 52. The absence of traces of the free ligand in the reaction mixture and the fact that the rearrangement proceeds in nonpolar media are indicative of an intramolecular mechanism of IHR.The reverse 52?53 rear- rangement was observed in none of the experiments. The rearrangements of silaphenalene complexes proceed under milder conditions than the rearrangements of silaacenaph- 695 thene complexes. This is indicative of a decrease in the activation energy of the process, although the rate constant of this process was not estimated. It should be noted that the isomer in which the tricarbonylchromium group is coordinated at the silicon-substi- tuted ring is thermodynamically more stable ase of the silaphenalene complexes 52 and 53 and the silaacenaphthene complexes 41 and 42. The complex 52 acts as a catalyst of hydrosilylation of butadiene.The catalytic action is achieved upon addition of 0.2% of the compound 52 to an equimolar mixture of diene and silane.109 The charactersitic feature of the reaction is its regio- and stereospecificity. Hydrosilylation afforded (Z)-1-triethoxysilylbut-2-ene as the only product in a yield of590%. Its structure was established by 1Hand 13C NMR spectroscopy.109 +HSi(OEt)3 52 120 8C, 2 h CH2Si(OEt)3 The catalytic activity of tricarbonylchromium complexes of polycyclic arenes (naphthalene, phenanthrene and anthracene) in hydrogenation of conjugated dienes and acetylenes, isomerisation of dienes, Kharasch addition of CCl4 to alkynes, etc. was attributed to the ability of the tricarbonylchromium group, which is Z6-bonded to the arene ring, to change the mode of bonding to form an Z4-intermediate.110 3.Z6,Z6-Rearrangements in tricarbonylchromium complexes of biphenylene Tricarbonylchromium complexes of biphenylene are of consider- able interest from the viewpoint of studies of Z6,Z6-IHR. The Z6-mononuclear (54) and Z6-binuclear (55) tricarbonylchromium complexes of biphenylene were synthesised by the OÈ fele reaction in 66% total yield (54 : 55=15.5 : 1).106 The individual complexes were isolated from the reaction mixture by TLC. Py3Cr(CO)3 Cr(CO)3 + 55 Cr(CO)3 54 (CO)3Cr Previously, the complexes 54 and 55 were obtained in the reaction of biphenylene with (NH3)3Cr(CO)3 in boiling dioxane in a similar ratio.111, 112 It was assumed that the binuclear complex 55, like the structurally similar binuclear complex of fluorene,78 contains the Cr(CO)3 group in the trans position because such a structure was established for the biphenylene binuclear complex of molybdenum Z6,Z6-C12H8[Mo(CO)3]2 by X-ray diffraction analysis.113 The directed introduction of substituents into the coordinated ring was performed by a procedure analogous to that used in the case of tricarbonyl(Z6-naphthalene)chromium. Metallation of the complex 54 with butyllithium in THF proceeded regio- and stereoselectively to form complexes 56a,b in which the label is located exclusively at position 1 of the coordinated ring.R Li RX 54 56a,b Cr(CO)3 Cr(CO)3 R = D (a), Me (b). The Z6,Z6-inter-ring haptotropic rearrangements in tricarbo- nylchromium complexes of 1-substituted biphenylenes 56a,b con- taining the planar ligand (according to the data of X-ray diffraction analysis) proceed more slowly than the corresponding696 rearrangements in monosubstituted naphthalene complexes.Thus, the isomers 57a,b were detected in the reaction mixture obtained upon heating of the complexes 56a,b only at temper- atures above 100 8C, IHR proceeding with a noticeable rate at 120 8C. In the course of the rearrangement, the tricarbonylchro- mium group migrates between the nonequivalent six-membered rings. R R k1 57a,b 56a,b Cr(CO)3 (CO)3Cr R = D (a), Me (b). According to the data of 2H NMR spectroscopy, the equilibrium for the deuterium-containing complexes 56a and 57a in decane was established at 120 8C in 6 h [k1=4.7961075 s71, Keq= 1.00, DG=(393)=31.00 kcal mol71].The Z6,Z6-rearrangements in the methyl-substituted complexes 56b and 57b proceeded somewhat more slowly [k1=3.4161075 s71, Keq=1.10, DG=(393)=31.25 kcal mol71]. 4. Synthesis of tricarbonylchromium complexes of substituted biphenyls and studies of the mechanism and kinetics of rearrangements in these complexes Until the mid-1980s, the data on IHR in complexes with bi- or polycyclic open ligands were virtually lacking in the literature. Biphenyl complexes represent the simplest model systems.114 ± 116 More complex analogues, viz., tricarbonylchromium complexes of 1,3,5-triphenylbenzene,117 triphenylene,118 9-phenylanthra- cene,119 9,10-dihydrophenanthrene,120 silaoxaphenanthrene,121 9,9-dimethyl-9-silafluorene,122 fluorene,74 carbazole 74 and fluo- ranthene,91, 123 were also prepared. Procedures for the introduc- tion of a label into tricarbonylchromium complexes of biphenyl derivatives are scarce.[The only procedure for the introduction of a label into (Z6-biphenyl)tricarbonylchromium involving the reaction of its stannylated derivative with butyllithium has been reported.124] Only two examples of rearrangements in complexes of biphenyl and its derivatives with transition metals were described before 1985. The inter-ring haptotropic rearrangement in the ethylbiphenyl complex of iridium was described in the study 125 with reference to the Ph. D. Thesis of Mellea.126 + + 40 8C Et Et Ir(PPh3)3 Ir(PPh3)3 The reaction proceeded under mild conditions (40 8C) and the equilibrium ratio of the isomers was 1 : 1.Isomerisation of tricarbonylchromium complexes of 9-phe- nylanthracene 119 proceeded under more drastic conditions (dibu- tyl ether, 130 8C). The tricarbonylchromium group successively migrates first over the biphenyl system and then over the fused system to the peripheral rings.119 The mechanism of this rear- rangement is not entirely known. Although the authors assumed that the rearrangement proceeds through an intermolecular mechanism, this statement is questionable and calls for special studies. 130 8C Cr(CO)3 Cr(CO)3 Recently, the first example of Z6,Z6-IHR in sandwich chro- mium complexes was reported.127 The vapour-phase (350 8C) Yu F Oprunenko reaction of fluoranthene with metal vapour was demonstrated to afford complex 58 as the major product along with isomeric complex 59.In the former complex, the chromium atom is located at the benzene rings, whereas one ligand in the complex 59 is coordinated at the `naphthalene' fragment. Cr Cr + 58 59 Prolonged heating (for several days) of the isomer 59 in deuteriobenzene at 100 8C resulted in the reversible rearrange- ment to form complex 60 in which both ligands are coordinated at the `naphthalene' fragment, i.e., Z6,Z6-IHR of the biphenyl type occurred. 100 8C, PhH Cr Cr 60 59 Migrations of the Cr(CO)3 group in biphenyl-related open systems, viz., in tricarbonylchromium complexes of diphenylal- kanes [(Z6-Ph)Cr(CO)3](CH2)nPh (n=1 ± 4), unsymmetrically substituted cis- and trans-stilbenes and 1,1-diarylethylenes, were observed.128, 129 At 140 8C, the Cr(CO)3 group in complexes of substituted stilbenes (61 and 62) and diarylethylenes migrates between the rings containing the methyl and CD3 groups.140 8C Me CH CH D3C C6D6 cis, trans-61 Cr(CO)3 Me CH CH D3C cis, trans-62 Cr(CO)3 Me Me Me Me 140 8C C6D6 Me Me D3C D3C (CO)3Cr Cr(CO)3 The rate of the intramolecular rearrangement is noticeably higher than that of the rearrangement proceeding through the dissocia- tion mechanism, which is evidenced by an insignificant amount of products of replacement of the ligand in these complexes by the `trapping' ligand.The rate constants of IHR were determined for the cis- and trans-isomers of stilbene complexes at 140 8C. For the rearrange- ment 61?62 proceeding by an intramolecular mechanism, the rate constants are 4.2061075 s71 and 6.9561075 s71 for the trans- and cis-isomers, respectively. The rate constants of exchange through a dissociation mechanism are substantially smaller (0.1861075 s71 and 0.06661075 s71, respectively). a. Synthesis of tricarbonylchromium complexes of substituted biphenyls It was not until 1985 that progress was achieved in the synthesis of tricarbonylchromium complexes of biphenyl derivatives labelled at one of the rings. For this purpose, insignificantly modifiedInter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands procedures, which had been developed primarily for the synthesis of tricarbonylchromium complexes of monocyclic arenes, were used.The reactions of the ligands (biphenyl and 2-bromo, 4-bromo- and 4-trimethylstannylbiphenyls) with chromium car- bonyls L3Cr(CO)3 (L=CO,130 NH3 87 or Py;88 L3=C10H8 46) afforded complexes 63 ± 68 in moderate yields (30% ± 40%).131 R abcd (a) L=CO, R=H, Bu2O:THF=1 : 6, 140 8C, 4 h; (b) L=CO, R=2-Br, Bu2O:THF=1 : 5, 140 8C, 5 h; (c) L=CO, R=4-Br, Bu2O:THF=1 : 7, 140 8C, 7 h; (d) L = NH3, R=4-SnMe3, dioxane, 100 8C, 2 h. Either both possible isomers or only one isomer are formed depending on the nature of the substituent in the ligand, its position and the type of chromium carbonyl used in the reaction.Thus the reaction of 4-trimethylstannylbiphenyl with (NH3)3Cr(CO)3 gave rise exclusively to the complex 68. Due to the strong electron-donating properties of the trimethylstannyl group, the electron density at the substituted ring increases resulting in coordination of the Cr(CO)3 group at this ring. Analogously, the Rausch reactions of 2- and 4-bromobiphenyls with (NH3)3Cr(CO)3 (dioxane, 100 8C, 2 h) afforded only the complexes 65 or 67 containing the Cr(CO)3 group in the unsub- stituted ring due to the electron-withdrawing properties of bro- mine (kinetic control). Prolonged heating of the bromo derivatives 65 or 67 with Cr(CO)3 in a mixture of dibutyl ether and THF at 140 8C (thermodynamic control) yielded a mixture of the isomers 64 and 65 or 66 and 67, respectively, with the complex containing the bromine atom in the uncoordinated ring predominating.Although the isomers can be readily isolated by TLC, the above- considered procedure cannot be used for the preparative synthesis of complexes containing halogen in the coordinated ring. Based on analysis of the above-considered results, it can be stated with assurance that individual complexes containing a substituent either in the uncoordinated or coordinated rings can be selectively prepared starting from bromo- or trimethylstannyl derivatives of biphenyl by the Rausch reaction. Subsequently, the bromine atom or the SnMe3 group can be first replaced by lithium under the action of BuLi and then by the R substituent by the reactions of the lithium derivatives with electrophilic reagents RX.On the whole, the selective synthesis of tricarbonylchromium complexes of substituted biphenyls by the exchange reaction of bromine for lithium is analogous to the procedure used for the preparation of naphthalene derivatives. The reaction of the complex 67 containing the tricarbonylchromium group in the unsubstituted ring with BuLi in anhydrous ether at778 8C led to the replacement of Br by Li. The lithium derivative is stable at low temperatures and reacts rather smoothly with electrophiles (D2O, L3Cr(CO)3 63 Cr(CO)3 + 64 65 Br (CO)3Cr (CO)3Cr Br Br Br +(CO)3Cr 67 (CO)3Cr 66 Me3Sn(CO)3Cr 68 697 Me3SnCl, MeI, etc.) at 0 8C to form selectively labelled complexes 69 ± 71 containing the label at position 4 of the uncoordinated ring.131 BuLi RX Li Br Et2O 67 (CO)3Cr Cr(CO)3 R Cr(CO)3 69 ± 71 RX=D2O, Me3SnCl, MeI; R=D (69), Me (70), Me3Sn (71).Attempts to prepare the corresponding complexes containing a label at position 2 of the uncoordinated ring starting from the compound 65 have not met with success even when the reactions were performed at room temperature, which is, apparently, due to steric hindrance to the replacement of Br by Li. It should be noted that it is necessary to perform these reactions, like the synthesis of naphthalene complexes, in ether because direct metallation of the coordinated ring in THF proceeds more rapidly than the replacement of Br by Li. Thus the reaction of the derivative 65 with BuLi in THF followed by treatment of the reaction mixture with methyl iodide gave rise to the complexes 72a ± c (72a : 72b : 72c=13 : 57 : 30).The low con- tent of the complex 72a in the mixture is attributable to steric hindrances, which appear either in the stage of metallation or, what is less probable, in the stage of the reaction of the lithium derivatives with MeI. Li MeI BuLi THF Cr(CO)3 Br Br 65 Cr(CO)3 RCr(CO)3 Br 72a ± c R=2-Me (a), 3-Me (b), 4-Me (c). The replacement of the Me3Sn group by lithium in the biphenyl complexes 68 was not performed. However, by analogy with the corresponding naphthalene complexes, complexes selec- tively substituted at the coordinated ring can be apparently prepared according to this procedure.BuLi RX 68 Li Et2O (CO)3Cr R Cr(CO)3 RX=D2O, MeI, Me3SiCl, etc. The synthesis of tricarbonylchromium complexes of substi- tuted biphenyls by direct metallation of (Z6-biphenyl)tricarbonyl- chromium was described. Thus the reaction of the complex 63 with BuLi in THF was accompanied by metallation of the coordinated ring. Then the reaction of the resulting lithium derivatives with the electrophileRXat778 8Cafforded a mixture of all three possible isomers.131 Li RX BuLi THF Cr(CO)3 63 Cr(CO)3698 R R + + R 73a,b 75a,b 74a,b Cr(CO)3 Cr(CO)3 Cr(CO)3 R = D (a), Me (b). For example, the reactions of lithium derivatives with D2O or MeI at 778 8C gave rise (according to the data of 2H NMR spectroscopy) to a mixture of deuterated complexes 73a ± 75a in 84% total yield (73a : 74a : 75a=10 : 65 : 25) or a mixture of methyl derivatives 73b ± 75b in 17.5% total yield (73b : 74b : 75b=12 : 58 : 30), respectively. The ratios of the iso- mers of deuterio- and methyl-containing derivatives are rather similar, which does not allow one to consider the effects of kinetic and thermodynamic control on their formation.Storage of the reaction mixture at 0 8C for 1 h followed by treatment with MeI afforded only the isomer 73b. Apparently, the 2-lithium-substi- tuted derivative is the thermodynamically most stable compound and is predominantly formed through the replacement ofH by Li.On the whole, the procedure for the synthesis involving metal- lation of the coordinated ring in the complex 63 is inferior to the above-described procedures (see Section III.1.b) in the prepara- tive respect. However, the use of all available synthetic approaches enables one to prepare various tricarbonylchromium complexes of substituted biphenyls. It should be noted that in the general case these procedures are applicable to all polyaromatic arenes. b. Studies of rearrangements in tricarbonylchromium complexes of substituted biphenyls For all available tricarbonylchromium complexes of substituted biphenyls, the possibility of IHR was studied upon their heating in anhydrous oxygen-free decane in the temperature range of 140 ± 174 8C for 4 ± 20 h.131 R R k1 k2 Cr(CO)3 (CO)3Cr 65, 67, 69 ± 71 64, 66, 68, 75a,b R=4-SnMe3 (68, 71), 2-Br (64, 65), 4-Br (66, 67), 4-D (69, 75a), 4-Me (70, 75b).The course of the rearrangement was monitored by with- drawing samples every 30 min and recording the 1H NMR spectra (2H NMR spectra were measured for deuterated deriva- tives) to determine the isomer ratio. Heating was terminated if the isomer ratio ceased to change (according to the NMR data). In some cases, this approach made it possible to estimate the rate constants and equilibrium constants for IHR. The structures of the complexes were established by 1H and 13C NMR spectro- scopy. The inter-ring haptotropic rearrangements in the biphenyl complexes 64, 66 and 68 are irreversible. Heating of the stanny- lated derivative 71 in boiling decane (174 8C, 5.5 h) resulted in the complete migration of the Cr(CO)3 fragment from the unsubsti- tuted ring to the substituted fragment to form the complex 68 [k2=6.4261075 s71, DG=(447)=35.10 kcal mol71], whereas the corresponding isomeric complex 68 did not undergo rear- rangement under these conditions, i.e., the reaction is actually irreversible.At 174 8C, the tricarbonylchromium group in the 2- and 4-bromophenyl derivatives (64 and 66) migrates irreversibly from the substituted ring to the unsubstituted one in 13 and 6 h, respectively [k1=2.23 and 5.1361075 s71, DG=(447)=36.04 and 35.3 kcal mol71 for the complexes 64 and 66, respectively].Reverse migration of the tricarbonylchromium group was not observed, which is evidence in favour of the irreversibility of IHR. Analysis of the kinetic data demonstrated that the rearrangement in the 2-bromo derivative 64 proceeds much more slowly than that Yu F Oprunenko in the 4-bromo derivative 66. Apparently, this is due to steric interactions in the complex 64 existing between the bulky bromine atom, on the one hand, and the ortho-proton of the adjacent ring and the organometallic group, on the other hand, resulting in the appearance of a twisting angle (a nonzero dihedral angle) and distortion of conjugation of the phenyl rings (see, for example, Refs 132 and 133). Ceccon et al.134 also revealed the presence of the twisting angle and a noticeable degree of noncoplanarity of the five- and six-membered rings in the tricarbonylchromium com- plexes 29 and 30.For the thermodynamically more stable (and, correspondingly, more populated) conformation of the ortho- derivatives of (Z6-biphenyl)tricarbonylchromium, the degree of noncoplanarity of the six-membered rings is higher than that in the corresponding para-derivatives. This effect is manifested in a substantial decrease in NOE between the protons of the coordi- nated and uncoordinated rings due to an increase in the average distance between their ortho-protons. The appearance of the nonzero dihedral angle between the benzene rings agrees with the data obtained by the DFT method for tricarbonyl(Z6-2- chlorobiphenyl)chromium containing the Cr(CO)3 group in the unsubstituted ring, which was chosen instead of the 2-bromo derivative to simplify calculations.In this compound, the dihedral angle is 58.14 8 and the halogen and chromium atoms are oppositely directed.135 The results of quantum-chemical calculations with the use of the HyperChem program (the PM3 method) for uncoordinated biphenyl and 2-bromobiphenyl also confirm the above-considered experimental and theoretical data. According to these calcula- tions, the biphenyl molecule is virtually planar (the dihedral angle is 0.2 8), whereas the phenyl rings in 2-bromobiphenyl are essen- tially noncoplanar (the dihedral angle is 52.9 8). It should also be noted that the corresponding dihedral angle in the tricarbonylchromium complex of unsubstituted biphenyl (calculations by the DFT method) is 21.38 8, i.e., coordination to the transition metal atom leads to an increase in steric hin- drance.135 The inter-ring haptotropic rearrangements in 4-deuterated and 4-methyl derivatives of tricarbonylchromium complexes of biphenyl, unlike those in the corresponding trimethylstannyl and bromo derivatives, are reversible.Heating of the complexes 75a,b in decane at 174 8C for 5 and 8 h, respectively, resulted in reversible rearrangements to form the isomeric complexes 69 and 70, respectively (Keq=[75a]/[69]=1.0 and Keq=[75b]/ [70]=1.4). The binuclear complex 76 was isolated from a mixture of the 4-methylbiphenyl complexes 70 and 75b (which was formed as a result of IHR) by TLC in 5% yield.Cr(CO)3 Me 76 Cr(CO)3 The structure of the compound 76 was established by mass spectrometry and 1H NMR spectroscopy. The formation of 76 indicates that insignificant dissociation of the complexes with elimination of the tricarbonylchromium group that passes into solution did occur under drastic thermal conditions even in inert nonsolvating decane. An alternative explanation assumes that a bimolecular reaction proceeded giving rise to a molecule of the binuclear complex 76 and the ligand from two molecules of the complex 75b. 76+Me 2Me (CO)3Cr 75b Therefore, inter-ring haptotropic rearrangements in tricarbo- nylchromium complexes of biphenyl derivatives proceed through an intramolecular mechanism.However, due to the high temper-Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands atures of the reactions, an intermolecular mechanism is realised to a small extent. This conclusion is consistent with kinetic evidence that metallotropic rearrangements in tricarbonylchromium com- plexes of diphenylalkanes proceed predominantly by an intra- molecular mechanism.128, 129 Analogous results were obtained upon heating of a mixture of the biphenyl complexes 73a ± 75a containing deuterium in the coordinated ring in decane at 174 8C for 6 h. Under these conditions, the reversible rearrangement proceeded accompanied by migration of the tricarbonylchromium group from the deute- rium-labelled ring to the deuterium-free ring.The ratio between the isomers containing deuterium in the coordinated and uncoor- dinated rings is close to unity in all three pairs of 2-, 3- and 4-deuterium-substituted complexes (according to the 2H NMR data; acetone as the solvent). When a mixture of complexes containing 2-, 3- and 4-methyl- substituted coordinated rings (73b ± 75b) was heated under the same conditions, only isomers 74b and 75b underwent rearrange- ments to form complexes containing substituents in the uncoordi- nated ring (Keq is 1.6 and 1.4 for the isomers 74b and 75b, respectively). The fact that IHR did not occur in the 2-methyl-substituted complex 73b, as in the case of the corresponding bromo derivative, is attributable to the fact that the phenyl rings are essentially noncoplanar in the most populated rotamer.This is consistent with a decrease in NOE between the ortho-protons of the coordinated and uncoordinated rings in the complex 73b (3% ± 4%) compared to the corresponding values obtained for the complexes 74b and 75b (7% ± 9%). The dihedral angle in the 2-methylbiphenyl complex contain- ing the Cr(CO)3 group in the unsubstituted ring calculated by the DFT method is 62.01 8.135 Interestingly, a mixture of 2-methyl-substituted complexes containing the Cr(CO)3 group either in the coordinated or uncoordinated rings did form from 2-methylbiphenyl and Cr(CO)3 upon refluxing in the dibutyl ether ±THF system. The isomeric 2-methyl-substituted tricarbonylchromium complexes were separated by TLC.The isomer ratio (1.3 : 1) appeared to be similar to that obtained as a result of inter-ring haptotropic rearrangements in tricarbonylchromium complexes of 3- and 4-methyl-substituted biphenyls. Cr(CO)6 + (CO)3Cr Me Me Me (CO)3Cr The absence of IHR in 2-methyl-substituted tricarbonylchro- mium complexes of biphenyl and the fact that their mixture can be readily obtained by the reaction of the ligand with chromium hexacarbonyl is additional evidence in favour of the intramolec- ular character of IHR. Therefore, a necessary condition of IHR to occur in complexes of substituted biphenyls and related ligands is a small degree of noncoplanarity of the rings (the dihedral angle should be smaller than 22 8, as in the case of the biphenyl complex), which is impossible in 2-substituted complexes due to steric effects.5. Z6,Z6-Rearrangements in tricarbonylchromium complexes of biphenyl-related ligands Biphenyl-related 10,10-dimethyl-10-sila-9-oxa-9,10-dihydrophe- nanthrene, which contains substituents possessing different elec- tronic properties in both rings, is of interest. Tricarbonyl- chromium and -tungsten complexes of this compound were prepared by reactions with different chromium and tungsten carbonyls.121, 136 In the case of M=Cr, the bis(tricarbonyl)chro- mium complex m-Z6:Z6-C12H8OSiMe2[Cr(CO)3]2 (79) was formed in a small amount. 699 L3M(CO)3 O SiMe2 + O SiMe SiMe2 M(CO)3 O SiMe2 M(CO)3 78a,b 77a,b M=Cr (a),W(b).Cr(CO)3 (CO)3Cr O SiMe2 79 The six-membered rings in the tricarbonylchromium com- plexes 77a and 78a are noncoplanar and the dihedral angles between these rings are (according to the X-ray diffraction data) 20.4 8 and 15.5 8, respectively.137 Heating of the complex 77a in anhydrous degassed decane at 130 8C led to IHR yielding the isomer 78a. The equilibrium was established in 6 h. According to the data of 1H and 13C NMR spectroscopy, the 77a : 78a ratio was 12 : 1 (the same isomer ratio was obtained upon heating of the complex 78a). In both cases, the 1H NMR spectra of the equilibrium mixtures have low-intensity signals belonging to the ligand and the binuclear complex 79. However, their total amount was no higher than 5% even after heating of the reaction mixture at 174 8C for 20 h.The binuclear complex and the ligand were formed due to either thermal destruction or bimolecular reactions proceeding in the complexes 77a and 78a. Analogously (but under substantially milder con- ditions), the inter-ring haptotropic rearrangement proceeded in the tricarbonyltungsten complex 77b. After heating of 77b in decane at 100 8C for 6 h, the 1H NMR spectrum had additional signals, which were assigned to the isomer 78b (77b : 78b=15 : 1). Attempts to isolate the complex 78b from the reaction mixture by TLC failed and the 1H NMR spectrum of the pure complex was not measured.136 Fluoranthene is of considerable interest because Z6,Z6-IHR both of the `biphenyl' and `naphthalene' types can proceed in this compound. Cr(CO)3 37 (S)-38 (CO)3Cr (R)-38 Cr(CO)3 In this case, the complex 38 exists as two enantiomers, which are interconverted in the course of the `naphthalene' rearrangement.Apparently, the kinetic parameters of such degenerate rearrange- ments may be determined from the rates of racemisation of the corresponding enantiomers in the future. Heating of the Z6-fluoranthene complex 38 in which the Cr(CO)3 group is located at the naphthalene fragment even at rather low temperature (in decane or benzene at 100 8C for 10 h) led to IHR to form the complex 37 in which the Cr(CO)3 group is located at the benzene ring.91, 123 On the contrary, prolonged heating (>100 h) of the complex 38 under the same conditions did not give rise to the isomer 37.The inter-ring haptotropic700 rearrangement was not accompanied by substantial decomposi- tion of the complexes or migration of the Cr(CO)3 group to the benzene ring, which is consistent with the results of other studies.46, 90 It was demonstrated that when tricarbonyl(Z6-naph- thalene)chromium was heated in benzene, the migration of the tricarbonylchromium group to the benzene ring proceeded with a noticeable rate (ksub=1.7361074 s71) 46 only at 140 8C and was completed in 6 h. The rate constant of the irreversible Z6,Z6-inter-ring hapto- tropic rearrangement 38?37 was determined from the data of 1H NMR spectroscopy (C6F6 as the solvent). The corresponding kinetic curves are shown in Figs 4 and 5.The rate constant of the rearrangement is equal to (5.50.2)61076 s71; Ea=30.13 kcal mol71. The complexes 37 and 38 were separated by TLC and their structures were established by 1H and 13C NMR spectroscopy and X-ray diffraction analysis. Therefore, IHR in planar fluoranthene complexes proceed under substantially milder conditions (80 ± 90 8C) than those in open biphenyl complexes (even in the cases of 4-bromo- or 4-methyl-substituted complexes, IHP started at temperatures above 130 8C). At 100 8C, the replacement of the fluoranthene ligand by C6D6 in the course of Z6,Z6-IHR was not observed. This replacement was achieved only at 130 8C. After 10 h, the ligand :C6D6Cr(CO)3 ratio reached 1 : 1 and the replacement of fluoranthene by C6D6 was completed in 30 h (data from 1H and 2H NMR spectroscopy and TLC).In isomeric fluoranthene complexes, Z6,Z6-IHR were observed in melts for the first times, i.e., the rearrangements proceed in the absence of solvating interactions with solvent molecules, which can lead to elimination of the tricarbonylchro- mium group from the ligand. Thus when melted in a tube at 154 8C, the complex 38 was completely converted into the isomer 37, which was established by TLC and NMR spectroscopy. This fact is yet further evidence of the intramolecular character of Z6,Z6-IHR in tricarbonylchromium complexes of polycyclic aro- matic compounds. c (%) 80 1 40 2 0 20 15 10 5 104 t /s Figure 4. Kinetic curves for changes in the concentration (c) of the complexes 37 (1) and 38 (2) in the course of IHR. ln (c0/c) 1.6 1.2 0.8 0.40 104 t /s 20 15 10 5 Figure 5.Kinetic first-order dependence for irreversible IHR 38>37 at 90 8C; c0 and c (%) are the initial concentration of the complex 38 and the concentration at time t, respectively. Yu F Oprunenko The rearrangements in fluoranthene complexes proceed sub- stantially more readily than those in biphenyl derivatives for two reasons. First, the ligands in the complexes 37 and 38 are planar (according to the X-ray diffraction data). Second, the difference in thermodynamic stability of the complexes 37 and 38 is substan- tially larger than that in the case of analogous biphenyl pairs. The OÈ fele reaction of the complex 37 with an excess of Py3Cr(CO)3 afforded a binuclear fluoranthene complex (80) in 70% yield.The complex 80 has, apparently, a trans-structure. It should be noted that the complex 38 did not react with Py3Cr(CO)3 under the same conditions. (CO)3Cr Py3Cr(CO)3 Cr(CO)3 80 37 Cr(CO)3 Recently, Z6,Z6-IHR in tricarbonylchromium complexes of 2-acetylfluorene, which is a planar analogue of biphenyl, have been studied by 1H NMR spectroscopy (deuteriotoluene, 120 8C).138 k1k2 COMe COMe Cr(CO)3 Cr(CO)3 Under these conditions, the intramolecular inter-ring hapto- tropic rearrangement was accompanied by the replacement of the fluorenyl ligands by deuteriotoluene (ksub=0.1961075 s71). According to the kinetic and thermodynamic data [k1=0.1861075 s71, k2=0.02461075 s71, DG== 33.5 kcal mol71], the Z6,Z6-rearrangements in these complexes pro- ceed much more rapidly than those in biphenyl complexes because the ligand in the former complexes (as in the complexes 37 and 38) is planar.Therefore, IHR in tricarbonylchromium complexes with different polycyclic aromatic ligands have been studied in suffi- cient detail. The kinetic and thermodynamic parameters of these rearrangements, which depend on the structure of the ligand and the nature of the substituents, are summarised in Table 8. The Table 8. Kinetic and thermodynamic data for Z6,Z6-IHP in tricarbonyl- chromium complexes with polycyclic ligands. Ligand Ea b T /K Keq 105 k1 a DG= /kcal mol71 /s71 /kcal mol71 26.9 28.5 3.02 2.09 358 27.5 ± 26.8 29.7 1.00 1.13 27.8 ± 28.1 30.0 363 (see d) 0.73 26.9 ± 25.5 29.1 2.29 1.18 28.6 ± 27.8 31.0 4.79 1.00 32.4 ± 31.5 35.1 6.42 1.00 32.6 ± 31.72 35.3 (see d) 5.13 32.5 ± 33.4 36.04 (see d) 2.23 2-Methyl- naphthalene Methylace- 363 naphthene c Fluoran- thene Methylace- 363 naphthylene 1-Deuteriobi- 393 phenylene 4-Deuterio- 447 biphenyl 4-Bromobi- 447 phenyl 2-Bromobi- 447 phenyl a The rate constant k1 defines the rate of migration of the tricarbonylchro- mium group between the substituted and unsubstituted rings.b The value was calculated taking into account average DS=. c Data for the exo- complexes. d Irreversible IHR.Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands data on IHR in complexes of other transition metals with polyaromatic ligands are scarce.Nevertheless, these data indicate that the rates of rearrangements depend substantially on the nature of transition metal and the ligand environment. In conclusion, note that the first example of Z6,Z6-IHR in fused nonplanar polycycles, viz., in tricarbonylchromium com- plexes of substituted heptalene, which occur under rather drastic conditions, has been reported recently.139Me Me Me Me Me Me Cr(CO)3 (CO)3Cr Me Me 6. Theoretical approach to studies of inter-ring haptotropic rearrangements in transition metal complexes with polycyclic aromatic ligands During the progress of experimental investigations of IHR it became evident that further advances in studies of these processes could not be achieved without application of quantum-chemical methods.At present, ab initio high-level quantum-chemical cal- culations for heteroorganic and organometallic compounds are complicated. The restricted Hartree ± Fock ± Roothaan method with the use of small basis sets is well suited to studies of organic molecules. However, this method is inefficient in investigations of organometallic compounds. In the latter cases, it is necessary to perform calculations using larger basis sets with polarisation functions taking into account electronic correlations and relativ- istic corrections. As a result, theoretical studies were generally performed using semiempirical methods for calculating the struc- tures and properties of organometallic compounds and only qualitative results were obtained.Recently, the application of the DFT method has opened prospects for obtaining reliable quanti- tative data. Theoretical studies of the mechanisms and the major regular- ities of IHR in different systems were surveyed in a review.140 The pioneering investigations in this field were carried out by Albright et al.47 7. Studies of the mechanism of Z6,Z6-haptotropic rearrangements in naphthalenechromium tricarbonyls by the density functional method Recently, it has been demonstrated that time-saving calculations by the DFT method can give high-quality data competitive with the results of ab initio high-level quantum-chemical calcula- tions.141 The fact that the kinetic parameters of IHP in naphtha- lene complexes are weakly sensitive to the effect of the substituents in the ligand is somewhat unexpected (see Table 6) because the effects of the substituents are pronounced in the majority of other metallotropic rearrangements. Detailed investigations of the mechanism of the rearrangements by the DFT method provided reasonable explanations for this fact.Calculations were carried out with the use of the most-time-saving program with extended basis sets.141 The results reproduced adequately the geometric parameters of tricarbonyl(Z6-naphthalene)chromium (4) deter- mined by X-ray diffraction analysis,85 including such details as the differences between the Cr7C(1) and Cr7C(2) bond lengths and between the Cr7C(4a) and Cr7C(8a) bond lengths (see the note in Table 9).The results of calculations are given in Tables 9 and 10. These results agree qualitatively with the results obtained in another study.47 In particular, migration of the Cr(CO)3 group from one ring to another along the shortest path through the midpoint of the C(4a)7C(8a) bond is forbidden in the case of the basis statesA and B. Table 9. Comparison of the calculated geometric parameters of the model molecules with the experimental data. Bond Calculated values /A Cr(CO)6 (see b) 1.908 1.153 Cr7C C7O (Z6-C6H6)Cr(CO)3 (see c) 1.846 2.222 1.410, 1.426 Cr7CCO Cr7CPhH C7C (Z6-C10H8)Cr(CO)3 (see d) Cr7CCO 1.847 Cr7C(1) 2.218 Cr7C(2) 2.214 Cr7C(4a) 2.336 C(2)7C(3) 1.426 C(3)7C(4) 1.405 C(4)7C(4a) 1.439 C(4a)7C(8a) 1.442 Note.The atomic numbering scheme used for (Z6-C10H8)Cr(CO)3 is shown in Scheme 1. a The difference between the experimental and calculated values. b The experimental values were determined by electron diffraction study.142 c The experimental values were determined by electron diffraction study.143 d The experimental values were determined by X-ray diffraction analysis.85 5 3 4 4a 6 2 1 8a 8 7 A Table. 10. Relative energies of the isomeric complexes A and B and the transition states C and D (kcal mol71). A B C D R 000 H1-Cl 2-Cl 1-SiMe3 2-SiMe3 1-Me 2-Me 71.64 74.47 0.52 70.19 The transition state for (Z6-C10H8)Cr(CO)3 (3) has the struc- ture of the trimethylenemethane complex with the symmetry C2n in which the Cr(CO)3 group is coordinated through the C(4), C(4a), C(5) and C(8a) atoms and is slightly shifted to the periphery of the ligand.The potential energy surface does not have a stationary point (a local minimum) corresponding to an inter- 701 Deviation a Experimental values /A 1.924 1.165 70.016 70.012 1.863 2.208 1.417 70.017 0.014 0.001 0.025 0.018 0.001 0.015 0.051 0.019 0.014 0.003 1.815 ± 1.830 2.186, 2.214 2.213, 2.214 2.306, 2.337 1.375 1.389, 1.383 1.447, 1.404 1.439 Scheme 1 C B D 30.44 30.82 30.91 30.63 26.62 30.38 29.92 30.44 33.93 30.23 29.59 26.41 32.18 28.86 02.78 2.22 0000702 mediate of the Z3-allylic type, which had been postulated in another study.47 For tricarbonylchromium complexes of substituted naphtha- lenes, the tricarbonylchromium group can migrate between the substituted and unsubstituted rings via two paths through the transition states C and D.The results of calculations reproduce adequately the relative stabilities of the isomers A and B, which differ by the position of the Cr(CO)3 group in the unsymmetrically substituted naphthalene ligand. The presence of the substituent changes the energy of the transition state C in which this substituent is located in the immediate vicinity of the migrating group, whereas the energy of the transition state D changes only slightly (0.5 kcal mol71).Therefore, the existence of two reac- tion pathways one of which is virtually nondisturbed is, on the whole, responsible for the low sensitivity of the kinetic parameters of the rearrangement to the effects of the substituent in the ligand. Taking into account the assumptions made (the isolated-molecule approximation was used and the difference in the zero-point energy was ignored), the calculated activation barriers agree well with the experimental DG= values. IV. Conclusion Essentially all known data on IHR in p-complexes of transition metals with fused polycyclic ligands are surveyed in the present review. Metallotropic rearrangements revealed in new complexes and elucidation of the general regularities of these rearrangements are of considerable importance because they gave impetus to the development of new synthetic procedures.Presently, the general concepts of IHR are rather clearly stated. Further studies should be aimed at the more detailed theoretical description of these processes, in particular, it is necessary to have an idea of the structures of transition states in the course of IHR to perform the directed construction of new tautomeric systems. In this respect, calculations of the potential energy surfaces for the rearrangements with the use of more rigorous quantum-mechanical methods will be of paramount importance. References 1. MGielen, R Willem, B Wrackmeyer (Eds) Advanced Applications of NMR to Organometallic Chemistry (Physical Organometallic Chemistry) Vol 1 (New York: Wiley, 1996) 2.A N Nesmeyanov, D N Kravtsov Dokl. Akad. Nauk SSSR 135 331 (1960) a 3. N T Anh, M Elian, R Hoffmann J. Am. Chem. Soc. 100 110 (1978) 4. S F Waller J. Mol. Catal. 31 123 (1985) 5. N A Nesmeyanov J. Organomet. Chem. 100 161 (1975) 6. F A Cotton, in Dynamic Nuclear Magnetic Resonance Spectroscopy (Eds J L Jackman, F A Cotton) (New York: Academic Press, 1975) p. 377 7. C G Kreiter Adv. Organomet. Chem. 26 297 (1986) 8. B E Mann, B F Taylor, N A Taylor, R Wood J. Organomet. Chem. 162 137 (1978) 9. J A Gibson, B E Mann J. Chem. Soc., Dalton Trans. 1021 (1979) 10. J E Nordlander, J D Roberts J. Am. Chem.Soc. 81 1769 (1959) 11. B M Mikhailov, Yu N Bubnov Bororganicheskie Soedineniya v Organicheskom Sinteze (Organoboron Compounds in Organic Synthesis) (Moscow: Nauka, 1977) 12. H Kwart, J Slutsky J. Am. Chem. Soc. 94 2515 (1972) 13. E A Halevi Orbital Symmetry and Reaction Mechanism (Berlin: Springer, 1992) 14. I F Lutsenko, Yu I Baukov, O V Dudykina, E N Kramarova J. Organomet. Chem. 11 35 (1968) 15. I F Lutsenko, Yu I Baukov, I Yu Belavin, A N Tvorogov J. Organomet. Chem. 14 229 (1968) 16. T A Waldbach, P H van Rooyen, S Lotz Angew. Chem. 105 795 (1993) 17. P Jutzi Chem. Rev. 86 983 (1986) 18. R B Larrabee J. Organomet. Chem. 74 313 (1974) Yu F Oprunenko 19. A Bonny, S R Stobart J. Chem. Soc., Dalton Trans. 786 (1979) 20. J Sandstroem Dynamic NMR Spectroscopy (New York: Academic Press, 1982) 21.J P Jesson, E L Muetterties, in Dynamic Nuclear Magnetic Resonance Spectroscopy Ch. 8 (Eds J L Jackman, F A Cotton) (New York: Academic Press, 1975) 22. F A Cotton J. Organomet. Chem. 100 29 (1975) 23. B E Mann, in Comprehensive Organometallic Chemistry Vol. 3 (EdsGWilkinson, FGAStone, EWAbel) (Oxford: Pergamon Press, 1982) p. 89 24. B E Mann Chem. Soc. Rev. 15 167 (1986) 25. G Deganello Transition Metal Complexes of Cyclic Polyolefins (New York: Academic Press, 1979) 26. B Foxman, D Marten, A Rosan, S Raghu,M Rosenblum J. Am. Chem. Soc. 99 2160 (1977) 27. H W Whitlock Jr, C Reich,W D Woessner J. Am. Chem. Soc. 93 2483 (1971) 28. F A Cotton, D L Hunter J. Am. Chem.Soc. 98 1413 (1976) 29. M J Hails, B E Mann, C M Spencer J. Chem. Soc., Dalton Trans. 693 (1985) 30. A Carbonaro, A L Segre, A Greco, C Tosi, G Dall-Asta J. Am. Chem. Soc. 90 4453 (1968) 31. K J Karel, T A Albright,M Brookhart Organometallics 1 419 (1982) 32. M Brookhart, M L H Green J. Organomet. Chem. 250 395 (1983) 33. B F G Johnson, J Lewis, D J Yarrow J. Chem. Soc., Dalton Trans. 2084 (1972) 34. GAMMunro, P L Pauson J. Chem. Soc., Chem. Commun. 134 (1976) 35. R Benn, R Mynott, I Topalovic, F Scott Organometallics 8 2299 (1989) 36. A Stanger Organometallics 10 2979 (1991) 37. K Nakasuji,M Yamaguchi, I Murata J. Am. Chem. Soc. 108 325 (1986) 38. O I Trifonova, Yu F Oprunenko, Yu A Ustynyuk, N A Ustynyuk Metalloorg. Khim. 3 686 (1990) b 39.O I Trifonova, E A Ochertyanova, N G Akhmedov, V A Roznyatovsky, D N Laikov, N A Ustynyuk, Yu A Ustynyuk Inorg. Chim. Acta 280 328 (1998) 40. K M Nicholas, R C Kerber, E I Stiefel Inorg. Chem. 10 1519 (1971) 41. Yu A Ustynyuk Vestn. Mosk. Univ., Ser. 2, Khim. 23 605 (1982) c 42. N A Ustynyuk Metalloorg. Khim. 2 43 (1989) b 43. N A Ustynyuk, Doctoral Thesis in Chemical Sciences, Institute Organoelement Compounds, Academy of Sciences of the USSR, Moscow, 1990 44. A Ceccon, A Gambaro, G Agostini, A Venzo J. Organomet. Chem. 217 79 (1981) 45. T Thoma, V G Pleschakov, N S Prostakov, Yu A Ustynyuk, A N Nesmeyanov, N A Ustynyuk J. Organomet. Chem. 192 359 (1980) 46. E P KuÈ ndig, C Peret, S Spichiger, G Bernardinelli J. Organomet.Chem. 286 183 (1985) 47. T A Albright, P Hofmann, R Hoffmann, C P Lillya, P A Dobosh J. Am. Chem. Soc. 105 3396 (1983) 48. B Deubzer, Ph.D Thesis Technische Hochschule, MuÈ nchen, 1966 49. A Stanger, H Weismann J. Organomet. Chem. 515 183 (1996) 50. A Stanger, in The Xth IUPAC Conference on Physical Organic Chemistry (Abstracts of Reports), Haifa, 1990 p. 119 51. J Tsuji J. Organomet. Chem. 300 281 (1986) 52. K Nakasuji,M Yamaguschi, I Murata, K Tatsumi, A Nakamura Chem. Lett. 1489 (1983) 53. K Nakasuji,M Yamaguschi, I Murata, K Tatsumi, A Nakamura Organometallics 3 1257 (1984) 54. A Keasey, P M Bailey, P M Maitlis J. Chem. Soc., Chem. Commun. 142 (1978) 55. J B Woell, P Boudjouk J. Organomet. Chem. 172 C43 (1979) 56. L A Paquette, R Gree J.Organomet. Chem. 146 319 (1978) 57. Yu A Ustynyuk, O I Trifonova, Yu F Oprunenko, V I Mstislavsky, I P Gloriozov Organometallics 9 1707 (1990) 58. H Normant Angew. Chem., Int. Ed. Engl. 6 1046 (1967) 59. A N Nesmeyanov, Yu F Oprunenko, N A Ustynyuk, B V Lokshin, Yu A Ustynyuk Izv. Akad. Nauk SSSR, Ser. Khim. 1942 (1979) d 60. N A Ustynyuk, B V Lokshin, Yu F Oprunenko, V A Roznyatovsky, Yu N Luzikov, Yu A Ustynyuk J. Organomet. Chem. 202 279 (1980) 61. I Smid Ions and Ion Pairs in Organic Reactions Vol. 1 (New York: Wiley-Intersciences, 1972)Inter-ring haptotropic rearrangements in p-complexes of transition metals with polycyclic aromatic ligands 62. T E Hogen-Esch Adv. Phys. Org. Chem. 15 155 (1977) 63. Yu F Oprunenko, Yu N Luzikov, Yu A Ustynyuk, N A Ustynyuk J.Organomet. Chem. 231 137 (1982) 64. V A Mal'tseva, Degree, Department of Chemistry, Moscow State University, 1996 65. A N Nesmeyanov, N A Ustynyuk, T Thoma, N S Prostakov, A T Soldatenkov, V G Pleshakov, K Urga, Yu A Ustynyuk, 69. A Ceccon,A Gambaro, A Venzo J. Chem. Soc., Chem. Commun. 540 O I Trifonova, Yu F Oprunenko J. Organomet. Chem. 231 5 (1982) 66. N A Ustynyuk, Yu A Ustynyuk, A N Nesmeyanov Dokl. Akad. Nauk SSSR 255 127 (1980) a 67. O I Trifonova, Yu F Oprunenko, Yu A Ustynyuk Metalloorg. Khim. 3 465 (1990) b 68. A Ceccon, A Gambaro, F Gottardi, A Venzo, in The XIIIth International Congress on Organometallic Chemistry (Abstracts of Reports), Torino, 1988 p. 123 (1985) 70. A Decken, J F Britten,M J McGlinchey J.Am. Chem. Soc. 115 7275 (1993) 71. M E Rerek, F Basolo Organometallics 3 647 (1984) 72. N A Ustynyuk, N A Pomazanova, L N Novikova, D N Kravtsov, Yu A Ustynyuk Izv. Akad. Nauk SSSR, Ser. Khim. 1688 (1986) d 73. V E Zvorykin, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1990 74. Yu F Oprunenko, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1982 75. L N Novikova, Candidate Thesis in Chemical Sciences, Institute of Organoelement Compounds, Academy of Sciences of the USSR, Moscow, 1981 76. P M Treichel, J W Johnson J. Organomet. Chem. 88 207 (1975) 77. C White, J Thompson, P M Maitlis J. Chem. Soc., Dalton Trans. 1654 (1977) 78. N A Ustynyuk, L N Novikova, Yu F Oprunenko, S G Malyugina, Yu A Ustynyuk J.Organomet. Chem. 277 75 (1984) 79. N A Ustynyuk, L N Novikova, V K Bel'skii, Yu F Oprunenko, S G Malyugina, O I Trifonova, Yu A Ustynyuk J. Organomet. Chem. 294 31 (1985) 80. D N Laikov, Yu A Ustynyuk, in The XVIIIth International Conference on Organometallic Chemistry (Abstracts of Reports), 81. Yu F Oprunenko, S G Malyugina, Yu A Ustynyuk, N A Ustynyuk 82. Yu F Oprunenko, S G Malyugina, Yu A Ustynyuk, N A Ustynyuk Munich, 1998 B17 Izv. Akad. Nauk SSSR, Ser. Khim. 2405 (1984) d Izv. Akad. Nauk SSSR, Ser. Khim. 2156 (1985) d 83. E P KuÈ ndig, V Desorby, C Grivet, B Rudolph, S Spichiger Organometallics 6 1173 (1987) 84. R U Kirss, P M Treichel Jr J. Am. Chem. Soc. 108 853 (1986) 85. V Kunz, W Nowacki Helv.Chim. Acta 50 1052 (1967) 86. H P Fritz, E O Fischer J. Organomet. Chem. 7 121 (1967) 87. M Hudechek, S Toma J. Organomet. Chem. 393 115 (1990) 88. M D Rausch, G A Moser, E J Zaiko, A L Lipman Jr J. Organomet. Chem. 23 185 (1970) 89. K OÈ fele Chem. Ber. 99 1732 (1966) 90. Yu F Oprunenko, S G Malyugina, V A Roznyatovsky, V I Mstislavsky, N G Akhmedov, D N Laikov, Yu A Ustynyuk, in The XVIIIth International Conference on Organometallic Chemistry (Abstracts of Reports), Munich, 1998 B19 91. B Deubzer, E O Fischer, H P Fritz, C G Kreiter, N Kriebitzsch, H D Simmons Jr, B R Willeford Jr Chem. Ber. 100 3084 (1967) 92. E A Chernyshev, Yu F Oprunenko, O B Afanasova, V M Nosova, Yu E Zubarev, S G Malyugina, A V Kisin, V N Retunskii, N A Ustynyuk, Yu A Ustynyuk Izv.Akad. Nauk SSSR, Ser. Khim. 2324 (1986) d 93. Yu F Oprunenko, S G Malyugina, Yu A Ustynyuk, N A Ustynyuk, D N Kravtsov J. Organomet. Chem. 338 357 (1988) 94. Yu F Oprunenko, N G Akhmedov, D N Laikov, S G Malyugina, V I Mstislavsky, V A Roznyatovsky, Yu A Ustynyuk, N A Ustynyuk J. Organomet. Chem. 583 136 (1999) 95. G Jaouen, A Meyer, G Simonneaux J. Chem. Soc., Chem. Commun. 813 (1975) 96. T G Traylor, K Stewart Organometallics 3 325 (1984) 97. Y Kawajiri, N Motohashi J. Chem. Soc., Chem. Commun. 1336 (1989) 703 98. Yu Oprunenko, S Malyugina, P Nesterenko, D Mityuk, O Malyshev J. Organomet. Chem. 597 42 (2000) 99. E Yashima,H Fukaya,Y Okamoto J. Chromatogr. A 677 11 (1994) 100. J A Ramsden, C M Garner, J A Gladysz Organometallics 10 1631 (1991) 101. T E Bitterwolf, T L Hubler, R Todime J. Macromol. Sci., Chem. A27 1437 (1990) 102. O B Afanasova, Yu E Zubarev, V A Sharapov, N N Kirillova, A I Gusev, V M Nosova, N V Alekseev, E A Chernyshev, Yu T Struchkov Dokl. Akad. Nauk SSSR 279 904 (1984) a 103. Yu F Oprunenko, O K Sokolikova, S G Malyugina, O B Afanasova, Yu A Ustynyuk, N A Ustynyuk Metalloorg. Khim. 3 1396 (1990) b 104. H E Avery Basic Reaction Kinetic and Mechanisms (New York: Macmillan, 1974) 105. Yu F Oprunenko, S G Malyugina, O I Trifonova, O Ya Babushkina, A P Pisarevskii, Yu A Ustynyuk, N A Ustynyuk, D N Kravtsov, P N Nesterenko Izv. Akad. Nauk SSSR, Ser. Khim. 438 (1988) d 106. Yu F Oprunenko, S G Malyugina, O Ya Babushkina, Yu A Ustynyuk, P N Nesterenko, N A Ustynyuk, D N Kravtsov, V A Piven' Metalloorg. Khim. 1 384 (1988) b 107. P M Treichel, R U Kirss Organometallics 6 249 (1987) 108. O B Afanasova, O N Kalinina, Yu F Oprunenko, T M Sokolova, Yu A Ustynyuk, E A Chernyshev Metalloorg. Khim. 3 399 (1990) b 109. O B Afanasova, O N Kalinina, E A Chernyshev, Yu F Oprunenko, G N Turkel'taub, L P Khromushkina, Yu A Ustynyuk Metalloorg. Khim. 2 1435 (1989) b 110. M Sodeoka, M Shibasaki Synthesis 643 (1993) 111. G A Moser, M D Rausch Synth. React. Inorg. Met. Org. Chem. 4 37 (1974) 112. A Ceccon, A Gambaro, A M Romanin, A Venzo J. Organomet. Chem. 239 345 (1982) 113. S L Latesky, A K McMullen, G P Niccolai Organometallics 4 902 (1985) 114. E O Fischer, N Kriebitzsch, R D Fischer Chem. Ber. 92 3214 (1959) 115. P Corradini, G Allegra J. Am. Chem. Soc. 82 2075 (1960) 116. D A Brown, J R Radji J. Chem. Soc. A 1617 (1966) 117. J Deberitz, H NoÈ th J. Organomet. Chem. 61 271 (1973) 118. M Sato, Y Ishida, Y Nakamura, M Kajiwara Nippon Kagaku Zasshi 91 1188 (1970) 119. S D Cunningham, K OÈ fele, B R Willeford Jr J. Am. Chem. Soc. 105 3724 (1983) 120. K SchloÈ gl, A Werner, M Widhalm J. Chem. Soc., Perkin Trans. 1 1731 (1983) 121. N I Kirillova, A I Gusev, V A Sharapov, O B Afanasova, Yu E Zubarev, N V Alekseev, E A Chernyshev, Yu T Struchkov J. Organomet. Chem. 306 55 (1986) 122. O B Afanasova, Doctoral Thesis in Chemical Sciences, State Research Institute of Chemistry and Technology of Organoelement Compounds, Moscow, 1998 123. D Lemenovskii, Yu Oprunenko, Ch Elschenbroich, S Malyugina, D Mityuk, in The XIXth International Conference on Organo- metallic Chemistry (Abstracts of Reports), Shanghai, 2000 124. M Wright Organometallics 8 407 (1989) 125. R H Crabtree, C P Parnell Organometallics 3 1727 (1984) 126. M Mellea, Ph.D Thesis, Yale University, Yale, 1984 127. Ch Elschenbroich, R MoÈ ckel, A Vasil'kov, B Metz, K Harms Eur. J. Inorg. Chem. 1391 (1998) 128. T G Traylor, M J Goldberg Organometallics 6 2531 (1987) 129. T G Traylor, K J Stewart J. Am. Chem. Soc. 108 6977 (1986) 130. E O Fischer, K OÈ fele Z. Naturforsch. B 13 458 (1958) 131. Yu F Oprunenko, I A Shaposhnikova, Yu A Ustynyuk Metalloorg. Khim. 4 1377 (1991) b 132. Ch Elschenbroich, J Schneider, H Burdorf J. Organomet. Chem. 391 195 (1990) 133. H Kalchhauser, K SchloÈ gl,W Weissensteiner, A Werner J. Chem. Soc., Perkin Trans. 1 1723 (1983) 134. A Ceccon, A Gambaro, F Gottardi, S Santi, A Venzo, V Lucchini J. Organomet. Chem. 379 67 (1989) 135. Yu F Oprunenko, Doctoral Thesis in Chemical Sciences, Moscow State University, Moscow 1999Yu F Oprunenko 704 136. N I Kirillova, A S Zhdanov, A I Gusev, O B Afanasova, A S Frenkel', E V Bulycheva, G K-I Magomedov Metalloorg. Khim. 3 335 (1990) b 137. O B Afanasova, N I Kirillova, E A Chernyshev, Yu T Struchkov J. Organomet. Chem. 536 ± 537 31 (1997) 138. N G Akhmedov, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow 1999 139. Yu A Ustynyuk, O I Trifonova, A V Yatsenko, A A Borisenko, Kh Yu Khansen, P Ubel'khard Izv. Akad. Nauk, Ser. Khim. 2220 (1994) d 140. T Ziegler Chem. Rev. 91 651 (1991) 141. D N Laikov Chem. Phys. Lett. 281 151, 156 (1997) 142. K W Egger J. Am. Chem. Soc. 89 3688 (1967) 143. T Nozol, K Takahashi Bull. Chem. Soc. Jpn. 38 665 (1965) a�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) b�Russ. Organomet. Chem. (Engl. Transl.) c�Moscow Univ. Bull. (Engl. Transl.) d�Russ. Chem. Bull. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (8) 705 ± 716 (2000) Fluoroindate glasses P P Fedorov, RMZakalyukin, L N Ignat'eva, VMBouznik Contents I. Introduction II. Physicochemical properties of indium fluoride III. Optimisation of glass compositions IV. The structures of fluoroindate glasses V. Properties and prospects for the application of fluoroindate glasses VI. Conclusion Abstract. the of concepts and properties compositions, The The compositions, properties and concepts of the structure These considered. are glasses fluoroindate of structure of fluoroindate glasses are considered. These glasses glasses are very hence and energies phonon low by characterised are characterised by low phonon energies and hence very low low theoretical to According region. IR the in losses optical theoretical optical losses in the IR region.According to the the vibrational polymerisation of degree the data, spectroscopic vibrational spectroscopic data, the degree of polymerisation of of octahedra in than lower is glasses fluoroindate in octahedra in fluoroindate glasses is lower than in fluorogallate fluorogallate and fluoroaluminate glasses. The compositions of fluoroindate and fluoroaluminate glasses. The compositions of fluoroindate glasses regarded be cannot components nine to up containing glasses containing up to nine components cannot be regarded as as being completely optimised. Comparative analysis of the chemical being completely optimised. Comparative analysis of the chemical and thermal stability of fluoroindate and fluorozirconate glasses and thermal stability of fluoroindate and fluorozirconate glasses was of susceptibility high the that shown is It performed.was performed. It is shown that the high susceptibility of indium indium fluoride to hydrolysis accounts for the instability of glasses and fluoride to hydrolysis accounts for the instability of glasses and optical losses. The prospects for the use of fluoroindate glasses in optical losses. The prospects for the use of fluoroindate glasses in lasers, fibre optics and scintillation instruments are discussed. The lasers, fibre optics and scintillation instruments are discussed. The bibliography includes 157 references. bibliography includes 157 references. I. Introduction Intensive studies of fluoride glasses began in the mid-1970s.At that time, fluoroberyllate and fluoroaluminate glasses were well known. An unexpected discovery of fluorozirconate glasses with high values of coordination numbers (c.n.) of glass-forming atoms (c.n. 6 ± 8) stimulated the development of studies of glass forma- tion in multicomponent fluoride systems and revealed other fluorides as `progenitors of glass'. Fluoride glasses are promising optical materials, including their prospective use in telecommuni- cation systems, because they are characterised by a much wider P P Fedorov A V Shubnikov Institute of Crystallography, Russian Academy of Sciences, Leninsky prosp. 59, 117333 Moscow, Russian Federation. Fax (7-095) 135 10 11. E-mail: ppf@newmail.ru RMZakalyukin Research Centre `Kosmicheskoe Materialovedenie' of A V Shubnikov Institute of Crystallography, Russian Academy of Sciences, ul.Akademicheskaya 2, 248033 Kaluga, Russian Federation. Fax (7-084) 224 86 14. E-mail: ruslan@ns.cry.ras.ru L N Ignat'eva Institute of Chemistry, Far East Branch of the Russian Academy of Sciences, prosp. 100-letia Vladivostoka 159, 690022 Vladivostok, Russian Federation. Fax (7-423) 231 18 89. E-mail: chemi@online.ru VMBouznik Khabarovsk Research Centre, Far East Branch of the Russian Academy of Sciences, ul. Shevchenko 9, 680000 Khabarovsk, Russian Federation. Fax (7-421) 233 74 95. E-mail: bouznic@khsc.khv.ru Received 13 March 2000 Uspekhi Khimii 69 (8) 767 ± 779 (2000); translated by V D Gorokhov #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n08ABEH000582 705 706 707 711 713 714 transmission region in the IR range compared to silicate glasses. Various experimental and computational methods are employed in the studies of the structure of fluoride glasses.Experimental results are described using different models of glassy state (crystal, disordered network, stressed mixed clusters, chemical twinning, ionic glass, polymeric, colloidal, etc.).Anumber of criteria of glass formation have been proposed.1± 11 At present, the best studied glasses are based on zirconium fluoride of the type ZBLAN (53 ZrF4 . 20 BaF2 . 4 LaF3 . 3 AlF3 . 20 NaF).{ In order to prevent reduction of Zr4+ to Zr3+, the fluorozirconate glasses of ZBLAN type are doped with small amounts of indium fluoride (0.015 mol.% ± 0.25 mol.%).12, 13 In this case, no reduction occurs because the couple In3+/In+ (like, e.g., Sn4+/Sn2+ but not Ga3+/Ga+) has a larger electrochemical potential than the couple Zr4+/Zr3+ (see Refs 14 and 15).The following redox reaction takes place 2Zr 4+ + In+ . 2Zr 3+ + In3+ An analogous role seems to be played by the additive InF3 to fluorohafnate glasses: they prevent reduction of Hf4+ to Hf3+.16 ± 18 The presence of univalent indium in the glass leads to the appearance of a double absorption peak in theUVregion.14 Upon addition of up to 4 mol.% of InF3 to ZBLAN, the resistance of this glass to crystallisation is slightly increased. The refractive index is also increased but less essentially than upon addition of PbF2.19 Upon addition of up to 3 mol.% of InF3 to the fluorozirconate glass of ZBLALi type with a high (*8 mol.%) content of PbF2, no stabilisation effect was achieved.20 The increase in the InF3 concentration in the fluorohafnate glass containing 5 mol.% of CeF3 induced a decrease in the integral yield of Ce3+ luminescence.18 Attempts to stabilise the glass with composition 24 CdF2 .29 CdCl2 . 11 BaF2 . 36 NaF by adding up to 8 mol.% of InF3 were unsuccessful.21 In recent years, investigators have shown an appreciable interest in glasses with high content of indium fluoride, the so- called fluoroindate glasses. These glasses are characterised by lower mean values of the energy of phonons (*510 cm71) than conventional fluorozirconate glasses (*580 cm71).22 Thus, fluo- roindate glasses absorb in a farther IR region.The range of transparency extends from the UV region to *7 ±8 mm in the IR region (for fluorozirconate and fluoroaluminate glasses, this extends from the UV region to *5.5 mm).{ Consequently, the { Hereinafter, the compositions of glasses are given in mol.%. { The edges of transparency are determined approximately, since they depend not only on the sample composition, but also on its thickness.706 calculated values of minimum optical losses determined by the sum of the Rayleigh scattering and the multiphonon absorption are substantially smaller in fluoroindate glasses than in fluoro- zirconate glasses.For example, for the glass BIZYbT (30 BaF2 . 30 InF3 . 20 ZnF2 . 10 YbF3 . 10 ThF4) such losses reach *0.002 dB km71 at a wavelength of 3 mm,23 while for ZBLAN they are *0.044 dB km71 at a wavelength of 2.55 mm.24 There- fore, fluoroindate glasses may be regarded as promising materials for the IR instrumentation, in particular in systems of optical fibre communication for the transmission of information in the IR region. However, because of technological difficulties the real optical losses are far from the calculated values: they are larger than in the ZBLAN-based fibres and still larger than in the silicate optical fibres (see below). We are still far from the time when it will be possible to use the calculated characteristics of fluoroindate glasses for the development of many-kilometre fibre optical systems for the transmission of information in the range of 2.5 ¡À 3.7 mm.Characteristically, the impurities of 3d elements which affect the transmittance in fluoride glasses in the 0.7 ¡À 1.5 mm range, even if present in negligible amounts, have virtually no effect on the transmission in the 2.5 ¡À 3.2-mmrange. In the latter range, an essential contribution to the absorption is made by the OHgroups (the band at 2.9 mm),24 ¡À 26 which creates technological difficulties because of easy hydrolysis of indium fluoride. Other anionic impurities, in particular SO2¡¦ 4 , also influence essentially optical characteristics of fluoroindate glasses in this range.27 The prospects for the use of fluoroindate glasses in the communication optical fibre systems is related to specific features of real silicate fibre-based systems with losses below 1 dB km71.In the first-generation systems known since 1976, use is made of multi-mode fibres (the operating wavelength is 0.8 mm). Systems of the next generation developed in 1983 are based on the use of both single- and multi-mode fibres (the operating wavelength is 1.3 mm, the second telecommunication window). Modern third- generation systems operate at the wavelength 1.5 mm (the third communication window), which corresponds to the minimum optical losses of the silicate fibre.28 The systems of all the three generations require installation of laser amplifiers of the signal at intervals of a few tens of kilometres.The fluoride glasses and especially the fluoroindate glasses are characterised by low energies of phonons; therefore, they can be used as a base for the design of efficient lasers with high quantum yield (up to 6%¡À 7%).29 For different operating wavelengths, glasses are doped with ions of rare-earth elements, viz., with Tm3+, 0.8 mm; Pr3+ or Nd3+, 1.3 mm; Er3+, 1.5 mm (Ref. 30). Doping of fluoroindate glasses with ions of rare-earth elements is easier than that of fluorozirconate glasses.31 In recent years, the subject of most active development is the fluoride fibre doped with Pr3+ ions, which is required for the design of optical amplifiers used in telecommunication systems.17, 32, 33 The active element of fibre communication is the optical fibre composed of the core and the cladding.Anumber of requirements are set for the respective materials. The refractive index of the core (n1) should be larger than that of the cladding (n2), while the coefficients of thermal expansion and the temperatures of glass transition and crystallisation of the core and the cladding should be close. In order to achieve the maximum operation efficiency of laser amplifiers, it is necessary to use single-mode optical fibres (i.e., fibres with a small diameter of the core ranging from 3 to 5 mm) with a large (0.35 ¡À 0.4) aperture.20, 34 NA=(n21 7n22 )1/2 or the quantity Dn �� n1 ¡¦ n2 . n1 It is possible that fluoroindate glasses will also find application as optical waveguides (in particular in medicine) for the trans- mission of laser radiation in the IR region from Er :YAG solid P P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik (2.94 mm) and CO gas (4.8 ¡À 5.5 mm) lasers.The limiting factor is the threshold of glass damage.29 Chalcogenide (of the As2S3 type) and chalcohalide glasses [e.g., formed in the Ge ¡ÀAs ¡À Te ¡À I and Hg ¡À Pb ¡À SI systems,35 as well as halide (e.g., bromide 36 and mixed chloride-fluoride 21)] glasses are characterised by lower mean energies of phonons than fluoroindate glasses. Chalcogenide glasses are transparent much farther into the IR region; however, their transmission in the visible spectral region is limited.Mixed halide glasses are unstable in the ordinary atmosphere and undergo easy hydrolysis. Compared to the fluorozirconate glasses, the fluoroindate glasses possess enhanced thermal stability judging from the differ- ence in glass transition temperatures (Tg) and the onset of exothermic of crystallisation (Tx) on the heating curves DT=Tx7Tg or from the Hruby criterion 37 H �� Tx ¡¦ Tg , Tm ¡¦ Tx where Tm is the melting temperature. } In the opinion of some investigators (see, e.g., Ref. 39), the fluoroindate glasses are more stable than the fluorozirconate ones. The preparation of bulk castings of fluoroindate glasses requires lower cooling rates of glass-forming melts.40 Apparently, this characteristic depends strongly on the technology of preparation of glasses (oxygen-containing impurities induce their crystallisa- tion).41 The critical cooling rates (Rc) of glass-forming melts can be determined experimentally (e.g., by the method proposed by Barandiaran and Colmenero 42), viz., from the dependences of the observed supercooling DT1 (the difference between the tem- peratures of liquidus and the beginning of crystallisation) on the cooling rate R (linear extrapolation to the infinite supercooling using lnR71/DT 2 1 coordinates).Yet another method consists in calculations using the data derived from the TTT (temperature ¡À time ¡À transformation) curves.43, 44 Both methods require labori- ous experiments. The values of Rc reported from different studies show strong discrepancies.Thus the Rc values measured by the former method for the five-component glass (30 BaF2 . 30 InF3 . 20 ZnF2 . 10 YbF3 . 10 ThF4) differ by nearly two orders of magni- tude: 1.7 K min71 (Ref. 45) and 120 K min71 (Ref. 44) (for ZBLAN, these values were 0.7 and 3 K min71, respectively). Measurements by the last-mentioned method gave Rc= 102 K min71 (for ZBLAN, this value was 8 K min71).23 Such a difference is possibly explained by different conditions of prepa- ration of the samples. II. Physicochemical properties of indium fluoride At normal pressure, indium fluoride has the high-temperature cubic polymorph of the ReO3 type.46 As the temperature decreases, it passes into the trigonal polymorph of the VF3 type, space group R3c (Fig.1). The melting temperature of InF3 is 1320 8C.48, 49 Indium fluoride is characterised by high vapour pressure,50 it vaporises strongly upon heating above 900 8C, undergoes easy hydrolysis upon slight heating and absorbs moisture when stored in air. Its hydrate InF3 . 3H2O can hardly be dehydrated because its hydrolysis with water of crystallisation occurs even at room temperature 51 InOHF2 + HF + 2H2O . InF3 .3H2O The ionic radius of In3+ is substantially smaller than those of lanthanide cations, it is close to the radius of scandium ion and is } In the original paper,37 the temperature of the beginning of melting (solidus) is used in this formula; in a number of other papers (see, e.g., Ref. 38) the temperature of liquidus is used.The difference between them may be substantial.Fluoroindate glasses a Figure 1. Structural types ReO3 (a), b-YF3 (b) 47 and VF3 (c). a T /8C 1200 L 900 L+LiF 600 Li3InF6 LiF 20 60 mol.% cLiInF4 L 700 F 500 20 PbF2 60 mol.% Figure 2. Phase diagrams of the systems LiF ± InF3 (a), BaF2 ± InF3 (b), PbF2 ± InF3 (c) and YbF3 ± InF3 (d).48, 53 (1) DTA data; (2) single-phase samples; (3) two-phase samples from X-ray powder diffraction data; L, melt; F, solid solution with the fluorite structure; S, solid solution with the ReO3 structure; a, solid solution with the a-YF3 structure. b y zx FY ORe c a0 FV 0a p c ' 2 ÅÅÅ3 b T /8C L 1200 1000 F L+InF3 800 LiInF4 + +InF3 20 InF3 BaF2 InF3 60 mol.% d 1300 L 1100 S a 123 900 20 InF3 InF3 YbF3 60 mol.% 707 substantially larger than the radii of gallium and aluminium ions.52 Sc3+ Lu3+ In3+ Y3+ 1.01 1.159 1.117 1.06 La3+ 1.30 Ion r /A (c.n.8) A morphotropic transition occurs on going from InF3 to trifluorides of heavy rare-earth elements (Lu ± Er, Y). Indium fluoride is partially dissolved in high- and low-temperature RF3 polymorphs corresponding to the structural types a-YF3 and b-YF3, respectively.53 Being a strong Lewis acid, indium fluoride forms numerous compounds with fluorides of uni- and divalent metals (Fig. 2).46, 54 III. Optimisation of glass compositions Undoped indium fluoride is not vitrified.However, like other Lewis acids, such as AlF3, ZrF4, GaF3, ZnF2 (Fig. 3), it is the `progenitor' of glass. According to the proposed classification,11 indium fluoride is classed as a glass former of second degree, since at ordinary quenching rates (*103 K s71) glasses are formed in some binary systems, namely, InF3 ± BaF2,55, 56 InF3 ± PbF2 57 and InF3 ± SrF2.58 The regions of glass formation in the first two systems are in the vicinitectic points; the phase diagram of the third system has not been studied. First reports on the synthesis of fluoroindate glasses appeared in the early 1980s.55, 59Adouble glass-forming system BaF2 ± InF3 was found;55 however, the glass formed in this system was not sufficiently stable. Numerous stabilising components, viz., YF3, UF4,55 FeF3,60 ZnF2, ThF4,61 YbF3,23, 55, 62 ScF3, GaF3, CdF2,63 SrF2,58 MgF2,57, 64 BiF3 65 and In(PO3)3,66 were proposed.A second glass-forming system PbF2 ± InF3 was found almost simul- taneously which was doped with CdF2, ZnF2, MnF2 and SnF2 as stabilising ingredients. The regions of glass formation in some ternary systems are shown in Fig. 4. Note that the boundaries of the regions of glass formation are tentative. These depend on the experimental con- ditions and are in particular very strongly dependent on the cooling rate. The data from different studies on the regions of glass formation correlate but do not coincide. For example, glass formation was observed in the binary system BaF2 ± InF3,55, 61 whereas according to the results of another study,63 only three- component systems BaF2 ± InF3±MFn did undergo vitrification.The results of studies on the system BaF2 ± InF3±YF3 (cf. Refs 57 and 61) are at variance [two regions of glass formation were revealed (cf. Ref. 57, Fig. 4 b)]. It was stated that the system BaF2 ± InF3 ± GaF3 has an isolated region of glass formation,63 while Boutarfaia et al.57 revealed a `belt' region between the systems BaF2 ± InF3 and BaF2 ± GaF3 (see Fig. 4 d). Some dis- crepancies are found even within the framework of a single study. In particular, in a study,60 of all the studied systems PbF2 ± InF3 ± MFn the glass formation in a binary system PbF2 ± InF3 was revealed only in a single case (see Fig. 4 g), while in other cases the glass formation was observed only in three-component systems (Fig.4 h,i). In three-component systems, the rate of devitrification is great. Only in the system PbF2 ± BaF2 ± InF3 was a glass which has prospects for practical application found. Videau et al.55 have described the four-component glass 40 InF3 . 32 BaF2 . 17YF3 . 11 PbF2. A 2 mm-thick sample of this glass had the optical transmission edge of 13 mm. Auriault et al.60 obtained a bulk casting for the glass 30 PbF2 . 20 BaF2 . 50 InF3, the trans- mission edge in the IR region being *8.2 mm for a 3 mm-thick sample. Attempts were undertaken to vary the composition of this glass by doping it with ZnF2, SrF2, AlF3, YF3, CdF2 and PbF2 to enhance stabilisation. Optical characteristics of some 5 ± 8-com- ponent glasses (the refractive indices being in the range from 1.549 to 1.566, the edges of IR transmission varying from 7.7 to 8.9 mm) were determined.The possibility of replacement of InF3 by GaF3 as well as doping of glasses with rare-earth fluorides were also708 z/r 876 Ba2+ 54321 Na+ K+ Cs+ Rb+0.5 Figure 3. The positions of glass formers in the plot of electronegativity (X) vs. generalised moment of cation (z/r),10 where z is the oxidation state and r is the ionic radius; the open circles refer to the glass-forming cations in fluoride systems. noted. Of the glasses of this type, some prospects for practical uses have 19 PbF2 . 23.8 BaF2 . 47.6 InF3 .1.9 AlF3 . 4.9 SrF2 . 2.8YF3 67 and 48 InF3 . 24 PbF2 . 24 BaF2 . 3 LaF3 . 1 LuF3.68 Further search for optimum compositions was conducted for the Pb-free glasses (apparently, this was related to the idea of an increase in Tg and enhancement of the thermal stability of glasses). Gradual increase in the stability was achieved by increasing the number of components. For example, Bouaggad et al.23 found a glass with composition 30 BaF2 . 30 InF3 . 30 ZnF2 . 10 ThF4 in a four-component system (1 cm-thick samples were prepared) and a glass 30 BaF2 . 30 InF3 . 20 ZnF2 . 10 YbF3 . 10 ThF4 was found in a five-component system. Chiaruttini et al.43 used the latter glass (the sample with Rc=120 K min71) as the initial system in the experiments aimed at the increase in the number of glass compo- nents.Partial replacement of InF3 by GaF3 decreased Rc; the minimum values (20 ± 25 K min71) were observed for a glass containing 8 mol.% ± 14 mol.% of GaF3. Synthesis of several stable glasses was carried out by increasing the number of components of the system. These include 30 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 . 10 YbF3 . 6 ThF4 . 4 ZrF4 (glass BIGaZYbTZr, Rc=6 K min71) and 30 BaF2 . 18 InF3 . 12GaF3 . 16 ZnF2 . 10 YbF3 . 10 ThF4 . 4MnF4 (glass BIGaZYbTMn, Rc= 6 K min71).43, 69 These glasses may be regarded as the known glasses 15 BaF2 . 28.3 ZnF2 . 28.3 YbF3 . 28.3 ThF4 70 and 12.5 BaF2 . 26.25 YbF3 . 35 ThF4 . 26.25MnF2 71 modified in order to increase their stability and decrease the content of radioactive thorium.A rather stable composition 40 InF3 . 20 ZnF2 . 20 BaF2 . 20 SrF2 (glass IZBS) has become the `parent' of a separate family of glasses.58 Messaddeq et al.39 varied this composition by adding PbF2, CdF2, CaF2, NaF and GdF3. Taking into account the DT values, these authors have selected the most stable composition, Ti4+ Hf4+ Re4+ W4+ Mo4+ Ta4+ Zr4+ Tb4+ Al3+ Mn3+ Ce4+V3+ Mo3+ U4+ Nb3+ Sc3+ Ta3+ Be2+ Y3+ Lu3+ Fe2+ Cr2+ Zn2+ Gd3+ Nd3+ La3+ Ni2+ U3+ Mn2+ Cu2+ Mg2+ Nb2+ Ti2+ Ca2+ Sr2+ Hg2+Pb2+ Cd2+ Li+ Cu+ Au+ Tl+ Ag+ 1.5 1.0 P P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik As5+ Mo6+ W6+ Si4+ V5+ Re5+U6+ Sb5+ Ta5+ Ni4+ Mo5+ Nb5+ Se4+ Mn4+ Ge4+ Cr4+ Fe4+ Co4+ Bi5+ V4+ U5+ Rh4+Pt4+ Ru4+Os4+ Nb4+ Sn4+ Fe3+ Co3+ Pb4+ Ni3+ Ga3+ Cr3+ Ir3+ Th4+ Sb3+ In3+ Au3+ Tl3+ Bi3+ Co2+ Pt2+ 2.0 X viz., 40 InF3 .20 ZnF2 . 20 SrF2 . 16 BaF2 . 2GdF3 . 2 NaF, with the refractive index n equal to 1.4930. Isomorphic replacement of InF3 by GaF3 was carried out in a wide range (up to 30 mol.% of GaF3, with an optimum of 6 mol.% of GaF3) in order to decrease n.72 Further variations of compositions were undertaken with the aim of enhancement of the resistance of glasses to crystallisation, changes of physical properties (in particular, for selecting the appropriate core/cladding combinations in optical fibres), incor- poration of corresponding activators for the appearance of special properties of glasses which allow their use in generators of stimulated radiation and scintillation instruments.In order to remove optically active Yb3+ and Mn2+ ions, Adam et al.26 selected an eight-component composition 27BaF2 . 18InF3 . 12GaF3 . 20ZnF2 . 10LuF3 . 6ThF4 . 4ZrF4 . 3PbF2, which possessed however a somewhat higher critical cooling rate (16 K min71). It is of note that for some practical applications the presence of ytterbium is desirable (see, e.g., Ref. 16). Charron et al.73 studied the replacement of BaF2 by lithium, sodium and potassium fluorides in the glass 30 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 .10 YbF3 . 10 ThF4 (BIGaZYbT). The incorpo- ration of alkali metals decreased the refractive index; Tg was decreased upon addition of LiF and NaF and remained virtually unaltered upon introduction of KF. The critical cooling rates remained approximately at the same level (18 K min71 for the initial glass and 12, 7 and 24 K min71 for the glasses doped with 5 mol.% of LiF, 12.5 mol.% of NaF and 5 mol.% of KF, respectively). A family of glasses 30 BaF2 . 30 InF3 . 10 ThF4 . 9 ZnF2 . 20MF.1RF3 (M=Li, Na, K; R=Pr, Nd, Tb) with high contents of alkali metal fluorides has been described.74 In order to reduce the energy of phonons (and thus shift the absorption edge to the far IR region), Adam et al.75 attempted to decrease the content of InF3 (and fluorides of other trivalentFluoroindate glasses a InF3 0.8 YF3 0.4 0.4 ZnF2 0.8 ThF4 0.8 0.4 MFn BaF2 e InF3 0.8 0.4 0.4 0.8 0.4 MgF2 BaF2 0.8 i InF30.8 0.4 0.4 0.8 0.8 0.4 ZnF2 PbF2 Figure 4.Regions of glass formation in systems involving InF3: (a) BaF2 ± InF3±MFn (M=Y, Th, Zn);61 (b) BaF2 ± InF3±YF3;57 (c) BaF2 ± InF3 ± ZnF2;58 (d) BaF2 ± InF3 ±GaF3;57 (e) BaF2 ± InF3 ±MgF2;57 (f) BaF2 ± InF3 ± SrF2;58 (g) PbF2 ± InF3 ±MnF2;59, 60 (h) PbF2 ± InF3 ± BaF2;60 (i) PbF2 ± InF3 ± ZnF2;60 (j) SrF2 ± InF3 ± ZnF2;58 (k) PbF2 ± InF3±YF3;57 (l) (0.5 BaF2+0.5 SrF2) ± InF3 ± ZnF2;58 (1) quenched glasses; (2) cast glasses; (3) glass-ceramic. metals) below 10 mol.%. The modified glass IZBS with com- position 12.5 BaF2 .30 InF3 . 30 ZnF2 . 12.5 SrF2 . 8NaF . 5 CdF2 . 2 LaF3 58 was taken as the base. It turned out that upon sub- stitution of zinc for indium the stability of the glass is decreased. The problem of stability was solved in part by replacement of certain portion of InF3 by GaF3 and optimisation of the content of modifiers. Qui et al.76 studied the possibility of incorporation of SnF2 into a glass of the IZBS type with composition 15 BaF2 . 40 InF3 . 20 ZnF2 . 20 SrF2 . 2 NaF. 3GdF3. It was found that it is possible to replace BaF2, ZnF2 or SrF2 by SnF2 (up to 15 mol.% ± 20 mol.%); however, in all the cases Tg and the difference Tx7Tg decreased. With the SnF2 content above 20 mol.%, crystal phases appear, including that of InF3.Upon variation of the BaF2 to SrF2 ratio, a glass with intermediate composition was selected which exhibited maximum stability (judging from the DT value). Zhang and Poulain 31 studied the effect of additions of rare- earth fluorides on the properties of the glass with composition 30 PbF2 . 20 GaF3 . 15 InF3 . 20GdF2 . 15 ZnF2; up to 10 mol.% of RF3 (R=La, Ce, Pr, Nd, Gd, Er, Yb and Lu) was introduced. It was established that the stability of the glass first increased and then diminished passing through a maximum with 1 mol.% ± 2 mol.% of RF3. On the whole, rare-earth fluorides act as modifiers. Up to 20 mol.% of GdF3 may be incorporated into the glass-forming system BaF2 . InF3 . ZnF2;77 however, the com- positions including 10 mol.% of GdF3 and 10 mol.% of SrF2 are the most stable.Akella et al.78 studied the possibility of variation of the composition 28 InF3 . of the nine-component glass 20ZnF2 .16.5BaF2 . 12GaF3 . 9PbF2 .6NaCl . 5SrF2 . 3CaF2 . 0.5PrF3, which has the optical transmission edge at 10 mm. Sodium b c InF3 0.8 0.4 0.4 0.4 0.8 0.8 0.4 BaF2 0.8 YF3 BaF2 g f InF3 0.8 0.4 0.4 0.4 0.8 0.8 0.4 0.8 SrF2 BaF2 PbF2 k j InF30.8 0.4 0.4 0.4 0.8 0.8 0.4 0.8 ZnF2 PbF2 SrF2 chloride was introduced to increase the cross-section of absorp- tion at f ± f transitions of rare-earth elements. It was found that 10 mol.% of GaF3 can be replaced by InF3, 4 mol.% of GaF3 can be replaced by YF3, whereas PbF2 can be totally replaced by CaF2 or partially replaced by SrF2 or BaF2 without crystallisation.Delben et al.79 studied a glass of the type IZBS doped with 5 mol.% of NaCl or LiCl. Complex temperature dependence of viscosity was established and bifurcation of temperature Tg was noted. Taking this into account, a two-phase model of glass was proposed. Compositions and properties of some fluoroindate glasses are listed in Table 1, while the evolution of their compositions is reflected by Scheme 1. The stabilities of glasses could be enhanced upon increase in the number of components of the system. This method is known as `the confusion principle'.70 The efficacy of this approach is explained by the fact that it is difficult to attain an ordered structure (the obstruction effect) 76, 86 with great diversity of types of local ordering, i.e., polymorphism of clusters.It is of note that the glasses BIGaZYbTZr and BIGaZYbTMn 69 contain each six (!) glass formers per one modifier. In many cases, the increase in the number of components and isomorphous replace- ments of one glass component by another result in the appearance of a stabilising `polycationic effect', which consists in the enhance- ment of glass stability reaching a maximum for some intermediate composition.87 For fluoroindate glasses of different compositions, this effect is observed upon partial isovalent substitution of GaF3 for InF3 43, 57, 75, 88 and heterovalent substitution of ZnF2 for InF3 75 (isomorphism of glass formers) as well as strontium for barium 75 (isomorphism of modifiers).Destabilisation observed upon substitution of ZnF2 for InF3 is apparently explained by the fact that the initial glass has already been stabilised (the optimum 709 d InF3 InF3 0.8 0.8 123 0.4 0.4 0.4 0.8 0.8 0.4 0.8 0.4 GaF3 ZnF2 BaF2 h InF3 InF3 0.8 0.8 0.4 0.4 0.4 0.8 0.8 0.4 0.4 0.8 BaF2 MnF2 PbF2 l InF3 InF3 0.8 0.8 0.4 0.4 0.4 0.8 0.4 0.4 0.8 0.8 ZnF2 0.5BaF2 + 0.5 SrF2 YF3P P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik 710 Table 1. Some properties of fluoroindate glasses: temperatures (8C) of glass transition Tg, onset of the exothermic crystallisation Tx, maximum peak of the exothermic crystallisation Tc and onset of melting (solidus) Tm, refractive index for white radiation or at a wavelength of 584.9 nm (n), density (d /g cm73) and coefficient of thermal expansion (a /107 K71).Ref. Composition (mol.%) Tx Tg Tm Tc n d a 591 591 77 588 7 7 60 7 5.56 1.517 332 328 307 298 304 306 306 222 214 209 202 250 283 336 324 285 283 650 580 537 5.09 1.495 7 777386 287 1.5005 1.4990 77 7777777 333 323 328 339 327 332 5.14 5.05 7 7 4.96 7 7 5.09 7 7 5.10 7 7 5.07 7 7 5.10 7 7 5.02 5.14 77 1.5005 1.4980 1.5300 777 7 1.4870 1.5080 1.4930 4.99 777 7 40 InF3 . 43 BaF2 . 17YF3 41 InF3 . 43 BaF2 . 16 YbF3 40 InF3 . 44 BaF2 .16UF4 40 InF3 . 23 BaF2 . 21 CaF2 . 16UF4 40 InF3 . 32 BaF2 . 17YF3 . 11 PbF2 37 InF3 . 22 BaF2 . 21MnF2 . 15YF3 . 5 LaF3 37 InF3 . 22 BaF2 . 21 CdF2 . 15YF3 . 5 LaF3 20 PbF2 . 40 CdF2 . 40 InF3 48 PbF2 . 25 ZnF2 . 27 InF3 40 PbF2 . 30MnF2 . 30 InF3 40 PbF2 . 25MnF2 . 35 InF3 30 PbF2 . 20 BaF2 . 50 InF3 20 PbF2 . 30 BaF2 . 50 InF3 40 BaF2 . 40 FeF3 . 20 InF3 30 BaF2 . 30 InF3 . 20 ZnF2 . 10 YbF3 . 10 ThF4 50 CdF2 . 35 GaF3 . 15 InF3 40 InF3 . 25 BaF2 . 20 ZnF2 . 5 CdF2 . 10 NaF 55 InF3 . 22.5BaF2 . 22.5 SrF2 40 InF3 . 40 BaF2 . 20 ZnF2 40 InF3 . 20 BaF2 . 20 ZnF2 . 20 SrF2 30 InF3 . 15 BaF2 . 30 ZnF2 . 20 SrF2 . 5 CdF2 50 InF3 .40YF3 . 10 BaF2 50 InF3 .3YF3 . 33 BaF2 . 14 SrF2 50 InF3 .5YF3 . 31.5 BaF2 . 13.5 SrF2 45 InF3 . 10YF3 . 31.5 BaF2 . 13.5 SrF2 45 InF3 .5YF3 . 35 BaF2 . 15 SrF2 50 InF3 .5YF3 . 13.5 BaF2 . 31.5 SrF2 40 InF3 . 20 BaF2 . 20 ZnF2 . 20 SrF2 40 InF3 . 20 ZnF2 . 20 SrF2 . 15 BaF2 . 5 CdF2 40 InF3 . 20 ZnF2 . 5 SrF2 . 25 BaF2 . 10 PbF2 40 InF3 . 20 ZnF2 . 20 SrF2 . 15 BaF2 . 5 CaF2 30 InF3 . 30 ZnF2 . 20 SrF2 . 15 BaF2 . 5NaF 40 InF3 . 20 ZnF2 . 20 SrF2 . 17 BaF2 . 2GdF3 . LaF3 40 InF3 . 20 ZnF2 . 20 SrF2 . 16 BaF2 . 2GdF3 . 2 NaF 15 InF3 . 20 GaF3 . 30 PbF2 . 20 CdF2 . 15 ZnF2 6.017 5.44 5.56 1.611 1.5050 1.5090 7 7641 634 579 74.88 7 400 388 377 357 372 356 356 255 258 263 238 318 351 409 447 367 346 324.5 381 360 300.6 388 290.6 393 416 395 412 390 385 418 300.6 388 383 366 385 380 390 390 354 354 442 442 478 370 380 383 405 7 7 7 7 7 7 55 7 7 7 7 7 55 7 7 7 7 7 55 7 7 7 7 7 55 7 7 7 7 7 55 7 7 7 80 7 7 7 80 7 7 7 7 7 60 7 7 7 7 7 60 7 7 7 7 7 60 7 7 7 7 7 60 1.5492 7 7 7 7 7 60 7 7 7 7 7 60 171 23 7 7 7 63 63 7 7 7 7 58 7 7 7 7 7 58 187 58 81 57 57 57 57 57 57 187 58 182 39 189 39 7 7 7 186 39 192 39 7 7 39 39, 82 7 7 7 7 83 166 31 171 43 43 7 7 7 26 170 32 209 75 76 7 7 7 77 1.578 7 7 7 7 1.5091 396 401 425 401 419 394 394 425 396 400 372 390 389 410 401 364 374 7777 7 1.60 7 7 1.498 7 7 7 5.114 582 7 291 277 295 290 294 290 260 243 330 332 332 250 306 291 300 250 424 350 7 7 65 78 5.15 7 4.8 7 7 84 7 7 7 7 7 7 85 86 see a 7 7 7 7 1.49 320 236 265 1.61135 6.32 5.55 460 7 7 1.564 327 345 77 30 BaF2 .18 InF3 . 12 GaF3 . 20 ZnF2 . 10 YbF3 . 6 ThF4 . 4 ZrF4 30 BaF2 . 18 InF3 . 12 GaF3 . 16 ZnF2 . 10 YbF3 . 10 ThF4 . 4MnF4 27 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 . 10 LuF3 . 6 ThF4 . 4 ZrF4 . 3 PbF4 29 PbF2 . 3 PbCl2 . 15 InF3 . 20 GaF3 . 13 ZnF2 . 20 CdF2 10 BaF2 . 6I nF3 . 47 ZnF2 .26 SrF2 . 4GaF3 . 5 CdF2 . 2 LaF3 15 BaF2 . 40 InF3 . 20 ZnF2 . 20 SrF2 . 2NaF . 3GdF3 40 InF3 . 20 ZnF2 . 20 BaF2 . 10 SrF2 . 10GdF3 30 BaF2 . 30 InF3 . 15 ZnF2 . 25 BiF3 28 InF3 . 20 ZnF2 . 16.5 BaF2 . 12 GaF3 . 9 PbF2 . 6 NaCl . 5 SrF2 . 3 CaF2 . 0.5 PrF3 36 InF3 . 20 ZnF2 . 20 SrF2 . 16 BaF2 . 6GaF3 . 2 NaF 30 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 .9YF3 . 6 ThF4 . 4 ZrF4 .NdF3 19 ZnF2 . 17 InF3 . 17 GaF3 . 43 PbF2 . 4 LaF3 30 PbF2 . 20 GaF3 . 15 InF3 . 15 ZnF2 . 20 CaF2 a The data of R Lebullenger, L A O Nuces and A C Hernandes, in The XIIth International Symposium on Non-Oxide Glasses and Advanced Materials (Extended Abstracts), Floriano'polis, Brazil, 2000 p.439. ratio InF3 : ZnF2 has been reached). Upon partial substitution of AlF3 for InF3,38, 43 the stabilities of glasses did not change. Substitutions of SrF2 for PbF2 and of PbF2 for BaF2 have not been studied systematically. Substitution of PbF2 for BaF2 in the BIZYbT glass leads to its destabilisation.43 Considering the isomorphism concept, the substitutions observed are well expected and correspond to the so-called `geo- chemical star' of indium.89 In particular, a wide heterovalent isomorphism of indium and zinc is known among compounds with different ionicities. From the physicochemical viewpoint and in conformity with the Rawson principle, the essence of the `confusion principle' is reduced to the decrease in the temperatures of eutectics as the number of components in the system is increased.10 It is of note that Bouaggad et al.23 suggested that the five-component composition of BIZYbT with a melting temper- ature of 650 8C is close to the eutectic composition; however, the thermogram derived from the same study contradicts this sugges- tion.The thermogram of the IZSB glass does not correspond to the eutectic either.90 This discrepancy may be indicative of the potential reserves of the stability enhancement. However, theFluoroindate glasses BaIn k=2 (see a) BaInY BaInSr BaInZn BaInGa k=3 PbBaInY BaInZnSr BaInGaZn BaInZnTh k=4 PbBaInYSr BaInZnYbTh BaInZnSrGd k=5 k=6 BaInGaZnLuGd BaInGaZnYbTh BaInZnSrGdNa PbBaInYSrAl BaPbInZnSrGdAl PbBaInGaSrLaYLiNa PbInGaZnLaCdHf PbInGaZnCdGdNa, BaInGaZnYbThZr, BaInZnSrGaGdNa k57 BaInGaZnYbThMn a k is the number of components; only cationic composition of glasses is indicated. studies performed so far show that the stabilities of fluoroindate glasses virtually reach `saturation' with the number of compo- nents equal to 6 or 7.The attainment of saturation corresponds to the Berezhnoi rule 91 for the temperatures of multicomponent eutectics. IV. The structures of fluoroindate glasses In order to elucidate the structures of fluoroindate glasses, it is necessary to know the structures of crystalline compounds which are similar in their compositions to those of the glasses under study and to reveal the types of polymers favouring the glass formation (Stenwart's criterion 1).Relevant measurements showed that the temperature dependences of the viscosities of melts of a number of fluoroindate glasses (like those of fluoro- zirconate glasses) are substantially non-linear in the coordinates log Z*1/T. This is the indication of `fragile' glasses and suggests that the polymerisation processes occur during the cooling of melts. This also points to higher ionicities of fluoride glasses compared, e.g., with `strong' silicate glasses.92, 93 The crystal structures of indium fluoride compounds are very diverse and are similar to the structures of ScF3 compounds. The indium cation can have c.n. 6 or 7 (it is not ruled out that its c.n. is 8 in solid solutions). The most characteristic polyhedra are octahera, often dis- torted.The structure of InF3 is a three-dimensional framework of octahedra connected through their vertices (the octahedra are slightly deformed in the low-temperature polymorph). Successive addition of fluorides of uni- and divalent metals to indium fluoride leads to a typical structural depolymerisation of octahedral structures (Fig. 5) with the following sequence: a framework ± a layer ([InF4]7, compounds MInF4) or a corrugated layer (LiInF4) ± a framework ([In3F14]57, compounds M5In3F14 of the chiolite type) ± a chain ([InF5]27, BaInF5) ± isolated octahedra ([InF6]37, compounds M3InF6). In all the cases, the octahedra are linked by vertices through the bridging fluorine atoms.54, 94, 95 The structure of high-temperature b-Ba3In2F12 (the structural type Sr3Fe2F12) comprises chains of octahera, in addition to isolated octahedra.Ternary compounds with octahedral forma- tions are known (e.g., of the type of weberite Na2MIIInF7). Compounds with octahedra [InF6]37 are isostructural to the corresponding fluoroaluminates and fluorogallates.94 Yet another important structural motif, which can give polymeric structures, are seven-vertex pentagonal bipyramids [InF7]47. In compounds of the KIn2F7 type, such bipyramids bind to each other through equatorial edges, form layers which are in turn linked to form a three-dimensional framework through the vertices of bipyramids. In compounds M2In3F11 and MIn3F10 711 Scheme 1 PbIn PbInZn PbInGa PbBaIn PbInGaZn PbInZnLa PbBaInAl PbBaInLa PbBaInGa PbInGaZnLa PbInGaZnCd PbInGaZnCa PbBaInAlLi PbInGaZnCdHfNa (M=K, Rb, Cs), junctions of bipyramids with octahedra occur.Crystallochemical analysis showed that the structures containing [InF7]47 are related to the a-U3O8 structure, which in turn is a derivative of the a-UO3 structure.96 Thus, the corresponding fluoroindates may be regarded as the products of depolymerisa- tion of the virtual InF3 polymorph of the a-YF3 (a-UO3) type. According to the data of X-ray powder diffraction, the structural type inherent in high-temperature polymorphs of fluorides of heavy rare-earth elements is identical to a-UO3. It may also be described as anti-Li3N. The structure of hypothetical RF3 pro- posed by Aleonard et al.96 may also be one of the a-YF3 models. On the basis of the corresponding RF3 polymorphs, solid sol- utions with indium fluoride are formed (see Fig.2). The structures of BaR2F8 compounds of the type of ortho-rhombic BaTm2F8 (R=Dy ± Yb) and monoclinic BaLu2F8, in which isomorphous substitutions of indium for rare-earth elements are possible, are related to the structural type a-YF3.97 Yet another type of seven-vertex polyhedron which is found in fluoroindates may be described as trigonal prisms with one capped lateral face. Slightly distorted polyhedra of this type are present in the structures Sr2InF7 98 and Pb2InF7, which is isostructural to the former, whereas paired seven-vertex polyhedra linked along the edge are found in the low-temperature polymorph of a-Ba3In2F12.99 Judging from the interpretation of KYF4,100 which is isostructural to NaInF4, the structure of the latter contains chains of the above-mentioned seven-vertex polyhedra.They may be regarded as the products of polymerisation through collectivisation of the `caps' of dimeric groups identified in a-Ba3In2F12. The structure of ortho-rhombic polymorphs of b-YF3 (see Fig. 1) in which InF3 can also be dissolved (see Fig. 2) may also be represented as the result of polymerisation of such polyhedra to form a three-dimensional framework due to which they are transformed into eight-vertex polyhedra by acquir- ing one more `cap' (trigonal prisms with two centred faces). Thus, in fluoroindate (like in fluorozirconate) glasses different structural units can be responsible for glass formation.In con- formity with the known diagrams,57, 61 the regions of glass formation in the systems BaF2 ± InF3±MFn adjoin the region of primary crystallisation of compound BaInF5 (see Figs 2, 4), which testifies in favour of the octahedral structures of these glasses. However, the facts that the system BaF2 ± InF3±YFn con- tains, in addition to the region of glass formation adjoining the side BaF2 ± InF3, yet another region similar in composition to that of BaY2F8 57 and that two types of glasses exist with considerably different Tg allow one to suggest the presence of two structural motifs in fluoroindate glasses (octahedra and pentagonal bipyr- amids). However, this may simply be related to the existence of two eutectics (without structural differentiation).Grande et al.101712 ae Figure 5. Packings of [InF6]37 and [InF7]47 polyhedra in structures of fluoroindates: (a) [InF6]37 in theM3InF6 structure; (b) [InF5]27 in the BaInF5 structure; (c) [In3F14]57 in the M5In3F14 structure; (d) [InF4]7 in the MInF4 structure; (e) [InF4]7 in the LiInF4 structure; (f) InF3; (g) [In3F10]7 in the MIn3F10 structure; (h) [In3F11]27in theM2In3F11 structure; (i) [In2F7]7in the MIn2F7 structure; (j) [InF7]47in the Sr2InF7 structure; (k) [In2F12]67in the structure of low-temperature phase Ba3In2F12; (l) [In2F11]7 in NaInF4. noted the glass formation on the basis of lithium nitride, which is structurally related to BaY2F8.Wide isomorphism with Ga in fluoride glasses (cf., e.g., Refs 57, 75, 82) suggests the octahedral surrounding of indium. Isomorphism with aluminium is usually restricted;38 however, complete isomorphism of Al3+ and In3+ (though, with the conjugated replacement of barium and lead) in both the crystalline and glassy states occurs in the system PbF2 ± BaF2 ± AlF3 ± InF3.41, 102 Analysis of binary glasses containing 50 mol.% ± 67 mol.% of InF3 in combination with BaF2 or SrF2 by the EXAFS method showed that the nearest surrounding of indium is octahedral, the distance In ± F (2.06 ± 2.07 A) being virtually the same as in InF3 (2.05 A).103 According to the data from EXAFS, X-ray radial distribution function and molecular dynamics for the glass 45 InF3 .40 BaF3 . 15 ErF3, the mean c.n. values are 7.15, 8.86 and 7.83 for In, Ba and Er, respectively. According to the X-ray photoelectron spectroscopy, the energies of the F1s bond are virtually equal to the arithmetic mean of the energies of the bonds of glass components.104 Todotoki et al.68 studied the up- bf i j kP P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik cg conversion luminescence } of Er3+ in the glass 48 InF3 . 24 PbF2 . 24 BaF2 . 3 LaF3 . 1 ErF3 and found that the local fre- quency of vibrations in the vicinity of Er3+ ions (26010 cm71) is considerably lower than the mean frequency of the phonon spectrum of the matrix (4705 cm71) as determined by Raman scattering. The EPR data on transition metal ions (Cr3+, Ti3+, Fe3+, Mn2+, Co2+, Cu2+) in the glass 25.2 PbF2 . 22.5 BaF2 .44 InF3 . 3.7 SrF2 . 3.8YF3 . 0.8 AlF3 provide evidence of their octahedral surrounding.105 Important information on the structures of glasses is provided by spectroscopic methods. In the IR and Raman spectra of fluoroindate glasses, there are bands corresponding to the vibra- tional modes of isolated octahedra [InF6]37.106, 107 According to the interpretation of the data from vibrational spectroscopy for the IZBS glass,106 its structure contains isolated octahedra [InF6]37. Later, the octahedral coordination of In3+ was con- } The up-conversion is the ionic fluorescence at a shorter wavelength than that of the acting laser radiation. d h lFluoroindate glasses firmed for the series of glasses 60 InF3 .40 BaF2, 40 InF3 . 20 ZnF2 . 20 SrF2 . 15 BaF2 . 5 CdF2 and 20 InF3 . 40 ZnF2 . 20 SrF2 . 15 BaF2 . 5 CdF2. A suggestion was put forward about the high content of vertex-bound octahedra in the binary glasses and the transition to almost unbound structures in the multi- component glasses.108, 109 The vibrational modes of the octahedra [ZnF6]47 and [InF6]37 lie close to each other. Ignat'eva et al.107 studied glasses in the system RF3 ± PbF2±MF2 (R=Ga, In; M=Zn, Mn, Ba). Analysis by IR and Raman spectroscopy showed that the glasses contain octahedral groups [InF6]37 and [GaF6]37, the degree of their polymerisation being higher in the fluorogallate glasses. Based on the analysis of the IR spectra in the system PbF2 ± BaF2 ± AlF3 ± InF3 ± LiF, Ignat'eva et al.110 sug- gested that their structures consist of frameworks formed by the octahedra.Upon successive substitution of AlF3 for InF3, starting approximately from the ratio 1 : 1 (*15 mol.% of AlF3), alumi- nium fluoride forms an independent framework in the glass close to the structural motif in the crystalline Pb3Al2F12. Incorporation of lithium fluoride disturbs the degree of connectivity of polyhe- dra in the glass. The fluoroindate glass 15 BaF2 . 40 InF3 . 20 ZnF2 . 20 SrF2 . 2 NaF. 3GdF3 underwent destruction as the degree of polymerisation of octahedra was increased judging from the IR spectra (devitrification upon polymerisation).76 Bureau et al.111 used precision 19F NMR spectroscopy to investigate glasses in the system BaF2 ± PbF2 ± InF3 (PBI) and 48 PbF2 .25 ZnF2 . 27 InF3 (PZI). They found that three types of fluoride ions contribute to the chemical shift: the ions unbound to the glass former, the ions present in isolated octahedra around the glass-forming atom and the bridging fluoride atoms binding the octahedra to each other. The relative contents of these ions are 5%± 11%, 62%± 68% and 21% ±32% for the PBI glasses and 11%, 37% and 52% for the PZI glasses. The dimensionality of octahedral formations for PZI is 2D, whereas for PBI it is in the interval between 1D and 2D. Thus, according to the classification based on c.n. of the glass- forming cation,11 the fluoroindate glasses, like the aluminate ones, may be referred to the octahedral glasses.Upon replacement Al?Ga?In, the degree of the glass polymerisation decreased on the whole (it depends strongly on the particular glass composi- tion). A number of complex fluoroindate glasses include different glass formers. Presumably, these glasses have complex micro- heterogeneous structures. Azkargorta et al.112 studied properties of the glass 30 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 . 9 LuF3 . 6 ThF4 . 4MnF2 . 1NdF3. Emission of the stimulated radiation at two wavelengths, viz., 1050.2 and 1054.9 nm, was obtained for the laser transition 4F3/2?4I11/2, which suggests the existence of two spatially separated subsystems with different coordination sur- roundings of the neodymium ions.V. Properties and prospects for the application of fluoroindate glasses Refractive indices of fluoroindate glasses (n'1.5) are somewhat higher than those of the ZBLAN glasses (see Table 1); the values of Abbe numbers (n'80), which characterise the dispersion of refractive indices, correspond to the values of this parameter for the fluorozirconate glasses. In the generalised n ± n diagram, the fluoroindate glasses occupy approximately the same area as the fluorozirconate glasses.113 The dispersion of refractive indices of glasses of the types BIZYTh and IZBS was measured.72, 114 The refractive indices and dispersions, as well as the theoretical minima of optical losses,115 can be calculated from the molecular increments.113 Upon addition of Pb2+ and Bi3+, the refractive indices of these glasses increased, while upon substitution of Ga3+ for In3+ they decreased on the contrary 72 (like upon substitution of Hf4+ for Zr4+ in the fluorozirconate glasses). The values of heat and thermal conductivities in fluoroindate glasses are*20% higher than those in ZBLAN, but are approx- imately the same magnitude lower than those in majority of 713 silicate and phosphate glasses.116 Studies of the stress relaxations for the glass 40 InF3 .20 SrF2 . 16 BaF2 . 20 ZnF2 . 2GdF3 . 2 NaF at temperatures below Tg showed that this process is strongly influenced by the atmospheric moisture.117 Characteristics of some fluoroindate glasses are presented in Table 2.Table 2. Densities (d), heat capacities (C) and thermal conductivities (l) of some fluoroindate glasses.116 Composition (mol.%) 103 l /W K71 cm71 d C /g cm73 /J g71 K71 0.67 5.359 10.40.9 0.67 4.761 9.90.8 0.67 4.753 10.20.8 30 PbF2 . 15 InF3 . 15 ZnF2 . 20 GaF3 . 20 CaF2 34 InF3 . 20 ZnF2 . 20 SrF2 . 16 BaF2 . 6GaF3 . 4 NaF 40 InF3 . 20 ZnF2 . 20 SrF2 . 16 BaF2 . 2GdF3 . 2NaF As it was mentioned above, smooth pyrohydrolysis of indium fluoride complicates the preparation of high-quality glasses, the appearance of oxofluoride phases inducing devitrification. Jha et al.83 studied the influence of the fluorinating atmosphere on the quality and stability of the glass 15 InF3 .20 GaF3 . 30 PbF2 . 20 CdF2 . 15 ZnF2 in the course of its casting. The best samples were obtained when SF6 was added to nitrogen as the carrier gas, whereas the flow of hydrogen fluoride had a somewhat weaker effect. The poorest effect was produced by ammonium hydrogen fluoride. In studies of the possibility of the use of fluoroindate glasses as the laser matrices, spectroscopic data on the following dopant ions were analysed: UO2á 2 (Ref. 118), Er3+ (Refs 84, 119 ± 121), Nd3+ (Refs 85, 122 ± 125), Pr3+ (Refs 126 ± 128), Tm3+ (Refs 129 ± 131), Cr3+ (Ref. 132), Yb3+ (Ref. 133) and Eu3+ (Refs 125, 134). The non-linear effects appearing upon the laser radiation in fluoroindate glasses doped with erbium and neodymium were studied.121, 135 The up-conversion was studied in the glasses activated by Nd3+ (Ref.136), Tm3+ (Refs 131, 137), Yb3+ (Ref. 138), Yb3+±Er3+, Yb3+±Tm3+ or Yb3+±Pr3+ (Refs 139 ± 141). In order to increase the pumping efficiency of praseodymium lasers used for the fibre amplifiers at 1.3 mm, it was proposed to coactivate (sensitise) the fluoroindate glasses containing ytterbium with a third rare-earth element, viz., neodymium or erbium.16, 17 When selecting the cladding for the glass 30 BaF2 . 18 InF3 . 12 GaF3 . 20 ZnF2 . 10 YbF3 . 6 ThF4 . 4 ZrF4, Rigout et al.38 found that the addition of 10 mol.% of KF, 3 mol.% of NaF (instead of a part of BaF2) and 6 mol.% of AlF3 (instead of InF3/GaF3) led to Dn=1.53% with virtually the same coefficients of thermal expansion and Tg.The critical cooling rate is somewhat increased (20 instead of 6 K min71). Soufiane et al.72 selected the compositions of the core and claddings of glasses of the IZIBS type by varying n by way of incorporation of GaF3 and PbF2: Dn=0.24% was achieved. Jiang et al.32 described the selection of the cladding for the fluoroindate fibres. Good results were obtained for fluorophos- phate glasses in combination with the GIPCZ glass. The compo- nents 29 PbF2 . 3 PbCl3 . 15 InF3 . of the glass fibre, 20 GaF3 . 13 ZnF2 (the core) and 20NaPO3 . 10 (NH4)2HPO4 . 40 ZnF2 . 15 PbF2 . 15 BaF2 (the cladding), have close values of Tg and the coefficients of thermal expansion. Both glasses mix in unlimited proportions.The technology of manufacture of the optical fibres based on the fluoroindate glasses is making good progress. The first fibres based on the BIZYbT glass had the optical losses of *10 000 dB km71.142 For the fibres based on the BIGaZYbTMn glass this value was already 500 dB km71.26 Soufiane et al.72 reported the fabrication of an optical fibre with714 losses of 400 dB km71 at 1.2 mm. Kanamori et al.143 described the technology of preparation of single-mode fibres with a length up to 500 mfrom fluoride glasses of the system InF3 ± GaF3 ± PbF2 ± BaF2 ± SrF2 ± LaF3±YF3 ± LiF ± NaF. Based on the fluoroindate glass doped with Pr3+ with Dn=3.6%, the minimum level of optical losses was achieved (114 dB km71 at a wavelength of 1.23 mm).The coefficient of signal amplification at the wavelength 1.3 mm is 0.18 dB mW71.This is sufficient for these fibres to be used as laser amplifiers, although the level of optical losses is very far from the theoretical level and is fivefold higher than the level achieved for the optical fibre based on ZBLAN. The reason for such a difference seems to be latent crystallisation. In order to decrease the optical losses, it is necessary to enhance the resistance of glasses to crystallisation and to increase the degree of purifica- tion from oxygen-containing admixtures. Itoch et al.144 used a single-mode fibre with the core from the glass 18 InF3 . 20 GaF3 . 32 PbF2 . 10.5 ZnF2 . 10.5 CdF2 . 4 LaF3 . 5 HfF4 doped with Pr3+ and the cladding from the ZBLAN glass.The main characteristics of this fibre are: NA'0.55 at 0.6 mm, the optical losses are 150 ± 250 dB km71, the coefficient of signal amplification is 0.29 dB mW71 at 1.3 mm. Optical amplifiers based on fluoroindate glasses doped with Nd3+ (Ref. 145) and Pr3+ (Ref. 146) have been described. There are prospects for the use of planar and channel wave- guides based on the fluoride glasses, including the fluoroindate glasses, as elements of solid lasers. The idea is to create surface structures with high refractive indices. P de Melo et al.147 obtained such a structure by diffusion of silver ions from the AgF film deposited on the surface of fluoroindate glass. Ionic exchange was observed 148 (Na+ for Li+ and K+) in the glass 15 BaF2 .18 InF3 . 12 GaF3 . 20 ZnF2 . 10 YbF3 . 10 ThF4 . 15 NaF dipped in the melts of salts. An increase in the refractive index was found in both cases, although judging from the bulk samples,73 the incorporation of potassium instead of sodium should rather decrease the value of n. The effect observed was explained by the mechanical stresses. Josse et al.149 used the surface pyrohydrolysis (treatment with the water vapour): the refractive index increased owing to the ionic substitution of OH7 for F7. The surface structure was applied by the photolithography method. The fluoride films for planar waveguides can be obtained by physical vapour-phase deposition.80 Fluoroindate glasses, particularly those enriched in lead fluoride, are characterised by rather high fluorine ion conductiv- ities,150 which are however smaller than the conductivities of the crystalline phases of analogous compositions.High mobility of fluorine ions was detected in fluoroindate glasses by the 19F NMR spectroscopy.107, 151, 152 For the glasses of the type IZBS and 40RF3 . 40 BaF2 . 20 YbF3, the ionic conductivity was observed to decrease 82, 153 upon replacement of indium by gallium and then aluminium. Costa et al.154 studied the mechanical properties of fluoroindate glasses. VI. Conclusion In contrast to the fluorozirconate glasses, the compositions of fluoroindate glasses cannot be regarded as optimised. Apparently, in the multicomponent systems involving InF3 there are several concentration-dependent regions of the stabilities of glasses.It was found that the replacement of InF3 by GaF3, BaF2 or SrF2 induced a stabilising polycationic effect; however, the possibility of isomorphous substitutions of PbF2 for BaF2 has not been studied systematically. It should be noted that the following techniques have a potential in a search for the optimum compo- sition of glasses, viz., the multifactorial design of experiment, the addition of a new component until a new crystalline phase appears as described by Mukhin and Gutkina 10, 155 and physicochemical analysis including prognostication of the coordinates of multi- component eutectics.156 The statements about the higher chemical and thermal stabil- ities of fluoroindate glasses compared to the fluorozirconate P P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik glasses are questionable.Great discrepancies (by two orders of magnitude) in the determination of the critical cooling rates of multicomponent fluoroindate glasses suggest non-reproducibility of the technology of their preparation. The proneness of InF3 to hydrolysis determines the instability of these glasses and optical losses in the range of 2.0 ± 3.2 mm. According to the data from vibrational spectroscopy, fluo- roindate glasses are characterised by lower degrees of polymer- isation of octahedra than fluorogallate and fluoroaluminate glasses. The reasons for this phenomenon are unclear. The fields of possible applications of fluoroindate glasses have been little studied.Optical transparency over a wide spectral range, comparatively high density and the possibility of doping with different activators increase the prospects of the practical applications of fluoroindate glasses. It is expedient to investigate the possibility of the production of scintillation materials (includ- ing those doped with cerium) with the short decay time.157 Owing to high concentration of indium fluoride, fluoroindate glasses may be applied for the creation of scintillation media for the neutrino detection.96 In order to manufacture highly efficient readjustable lasers, it seems attractive to co-activate fluoroindate glasses with neodymium and chromium ions. References 1.H Rawson Inorganic Glass-Forming Systems (London: Academic Press, 1967) 2. G T Petrovskii Optiko-Mekhan. Promst. (7) 39 (1964) 3. J Lucas J. Mater. Sci. 24 1 (1989) 4. E M Dianov, L N Dmitruk, V G Plotnichenko, M F Churbanov 5. PWFrance (Ed.) Fluoride Glass Optical Fibres (London: Blackie, 6. I D Aggarwal, G Lu (Eds) Fluoride Glass Fiber Optics (New York: Vysokochist. Veshch. (3) 10 (1987) 1990) Academic Press, 1991) 7. V D Khalilev, V L Bogdanov Zh. Vses. Khim. O-va im. D I Mendeleeva 36 593 (1991) a 8. E G Rakov Zh. Neorg. Khim. 36 828 (1991) b 9. M M Shul'ts, O V Mazurin Sovremennye Predstavleniya o Stroenii Stekol i Ikh Svoistvakh (Modern Views on Glass Structure and Their Properties) (Leningrad: Nauka, 1988) 10. P P Fedorov Neorg.Mater. 33 1415 (1997) c 11. P P Fedorov Kristallografiya 42 1141 (1997) d 12. G F West, R Sen J. Non-Cryst. Solids 161 52 (1993) 13. L Anselm, G H Frischat J. Non-Cryst. Solids 213 ± 214 375 (1997) 14. L J Andrews, B T Hall, R C Foilweller, C T Moynihan Mater. Sci. 19. L Wetenkamp, T Westendorf, G West, A Kober Mater. Sci. Forum Forum 32 ± 33 43 (1988) 15. A J Bruce,M Chui-Sabourin, A E Neeves, M M Broer, W A Reed Mater. Sci. Forum 67 ± 68 109 (1991) 16. B I Galagan, B I Denker, L N Dmitruk, V V Motsartov, V V Osiko, S E Sverchkov Kvantovaya Elektron. 23 103 (1996) e 17. B I Galagan, B I Denker, V V Motsartov, V V Osiko, S E Sverchkov Kvantovaya Elektron. 23 109 (1996) e 18. K W Bell, R M Brown, D J A Cockerill, P S Flower, P R Hobson, B W Kennedy, A L Lintern, J M Parker, M Sproston J.Non-Cryst. Solids 256 ± 257 42 (1999) 32 ± 33 471 (1988) 20. M Braglia, D Guojun, G Cocito, J Kraus, S Mosso J. Non-Cryst. Solids 213 ± 214 44 (1997) 21. M Matecki,M Poulain J. Non-Cryst. Solids 140 82 (1992) 22. R F Bartholomew, B G Aitken,M A Newhouse J. Non-Cryst. Solids 184 229 (1995) 23. A Bouaggad, G Fonteneau, J Lucas Mater. Res. Bull. 22 685 (1987) 24. G F West, W Hofle J. Non-Cryst. Solids 213 ± 214 189 (1997) 25. P S Christensen, J Lucas, M Ferraris Mater. Sci. Forum 32 ± 33 511 (1988) 26. J L Adam, F Smektala, J Lucas J. Non-Cryst. Solids 140 31 (1992) 27. Y Messaddeq, M Poulain J. Non-Cryst. Solids 140 77 (1992) 28. Y Suematsu Mater. Sci. Forum 32 ± 33 279 (1988) 29.M Poulain J. Non-Cryst. Solids 184 103 (1995)Fluoroindate glasses 30. B J Ainslie, S T Davey, D Szebesta, J R Williams,M W Moore, T Whitley, R Wyatt J. Non-Cryst. Solids 184 225 (1995) 31. G Zhang, M Poulain J. Alloys Comp. 275 ± 277 15 (1998) 32. J Jiang,G Zhang,M Poulain J. Non-Cryst. Solids 213 ± 214 11 (1997) 33. R S Qumby, B G Aitken IEEE Photonics Technol. Lett. 11 3 (1999) 34. A Jha, S Jordery, M Naftaly J. Non-Cryst. Solids 184 268 (1995) 35. M Mitkova, P Boolchand J. Non-Cryst. Solids 240 1 (1998) 36. J Cheng, Z Jin J. Non-Cryst. Solids 184 213 (1995) 37. A Hruby Czech. J. Phys. B22 1187 (1972) 38. N Rigout, J L Adam, J Lucas J. Non-Cryst. Solids 161 161 (1993) 39. Y Messaddeq, A Delben,M Boscolo,M A Aegerter, A Soufiane, M Poulain J.Non-Cryst. Solids 161 210 (1993) 40. R E O Diniz, S J L Ribeiro, Y Messaddeq, G Ghiselli, L A Nunes J. Non-Cryst. Solids 219 187 (1997) 41. R M Zakalyukin, A I Boltalin, P P Fedorov, B P Sobolev Fiz. Khim. Stekla 35 355 (1999) 42. J M Barandiaran, J Colmenero J. Non-Cryst. Solids 46 277 (1981) 43. I Chiaruttini, G Fonteneau, J Lucas, P S Christensen, S Mitachi Mater. Sci. Forum 67 ± 68 245 (1991) 44. M Poulain J. Non-Cryst. Solids 140 1 (1992) 45. P A Tick Mater. Sci. Forum 32 ± 33 115 (1988) 46. P P Fedorov, P I Fedorov, I V Shishkin, I P Zibrov, B P Sobolev Ftorid Indiya i Ftorindaty (Preprint No. 10) [Indium Fluoride and Fluoroindates (Preprint No. 10)] (Moscow: Institute of Crystallo- graphy Academy of Sciences of the USSR, 1989) p.37 47. L S Garashina, B P Sobolev, V B Aleksandrov, Yu S Vishnyakov Kristallografiya 25 294 (1980) d 48. P P Fedorov, I P Zibrov, B P Sobolev, I V Shishkin, P I Fedorov, M F Sviderskii, N P Petranin Zh. Neorg. Khim. 34 741 (1989) b 49. M Samouel, A de Kozak, J Renaudin, G Ferey Z. Anorg. Allg. Chem. 569 169 (1989) 50. R P Steiger, J L Margrave High Temp.-High Press. 5 471 (1973) 51. P P Fedorov, O S Bondareva, V A Ctasyuk,M D Ckorokhvatov, B P Sobolev, P I Fedorov Zh. Neorg. Khim. 42 698 (1997) b 52. R A Shannon Acta Crystallogr., Sect. A, Fund. Crystallogr. 32 751 (1976) 53. P P Fedorov, I P Zibrov, E V Tarasova, L V Medvedeva, B P Sobolev, I V Shishkin, P I Fedorov Zh. Neorg. Khim. 34 3150 (1989) b 54. J-C Champarnaud-Mesjard, B Frit, B Gaudreau Rev.Chim. Miner. 15 328 (1978) 55. J J Videau, B Dubois, J Portier C.R. Seances Acad. Sci., Ser. 2 Math. 297 483 (1983) 56. A Soufiane, Y Messaddeq, M Poulain, B J Costa J. Non-Cryst. Solids 213 ± 214 85 (1997) 57. A Boutarfaia, M A Poulain, M J Poulain, S E Bouaoud J. Non-Cryst. Solids 213 ± 214 36 (1997) 58. Y Messaddeq, M Poulain Mater. Sci. Forum 67 ± 68 161 (1991) 59. C Jacoboni, Le Bail, R de Pape Glass Technol. 24 164 (1983) 60. M Auriault, G Guery, A M Mercier, C Jacoboni, R M de Pape Mater. Res. Bull. 20 309 (1985) 61. J J Videau, J-M Dance, J Portier, B Dubois Rev. Chim. Miner. 23 789 (1986) 62. I V Shishkin, V M Yanishevskii P P Fedorov, B P Sobolev, N P Petranin, V A Bardin Tez. Dokl. VIII Vsesoyuz.Simp. po Khimii Neorganicheskikh Ftoridov (Abstracts of Reports of the VIIIth All- Union Symposium on the Chemistry of Inorganic Fluorides) (Moscow: Nauka, 1987) p. 420 63. M Poulain,M Poulain Mater. Sci. Forum 32 ± 33 137 (1988) 64. M Poulain,M Poulain J. Non-Cryst. Solids 140 87 (1992) 65. E B Merkulov, V K Goncharuk Tez. Dokl. X Simp. po Khimii Neorganicheskikh Ftoridov, Moskva, 1998 (Abstracts of Reports of the Xth Symposium on the Chemistry of Inorganic Fluorides, Moscow, 1998) p. 106 66. Y Messaddeq, D Grando, P Melnikov, S J L Ribeiro, L F C Oliveira J. Alloys Compd. 275 ± 277 81 (1998) 67. B Boulard, C Jacoboni,MRousseau J. Solid State Chem. 80 17 (1989) 68. S Todotoki, K Hirao, N Soga J. Non-Cryst. Solids, 140 46 (1992) 69.J Lucas, I Chiaruttini, G Fonteneau, P Christensen, S Mitachi Proc. SPIE ± Int. Soc. Opt. Eng. 1228 56 (1990) 70. J Lucas Proc. SPIE ± Int. Soc. Opt. Eng. 320 22 (1982) 71. Y Le Pape, G Fonteneau, J Lucas Rev. Chim. Miner. 21 589 (1984) 72. A Soufiane, F Gan, H L Helgoualch, M Poulain J. Non-Cryst. Sol- ids 184 36 (1995) 715 73. C Charron, G Fonteneau, J Lucas J. Non-Cryst. Solids 161 213 (1993) 74. G Amoranath, S Buddhutu J. Non-Cryst. Solids 143 252 (1992) 75. J L Adam, C Ricordel, J Lucas J. Non-Cryst. Solids 213 ± 214 30 76. J Qiu, K Maeda, R Terai, H Wakabayashi J. Non-Cryst. Solids 77. Y Messaddeq, S Inoue, C T M Ribeiro J. Non-Cryst. Solids 78. A Akella, E A Downing, L Hesselink J. Non-Cryst. Solids 79. J R J Delben, A A S T Delben, S L Oliveira, Y Messaddeq 80.C Jacoboni, O Perrot, B Boulard J. Non-Cryst. Solids 184 184 (1997) 213 ± 214 363 (1997) 213 ± 214 179 (1997) 213 ± 214 1 (1997) J. Non-Cryst. Solids 247 14 (1999) (1995) 81. Y Messaddeq, M Matecki,M Poulain Mater. Sci. Forum 67 ± 68 251 (1991) 84. A Florez, Y Messaddeq, O L Malta, M A Aegerter J. Alloys 82. J Kincs, A Soufiane J. Non-Cryst. Solids 213 ± 214 388 (1997) 83. A Jha, B E Kinsman, E R Taylor, G Kakarantzas J. Non-Cryst. Solids 213 ± 214 101 (1997) Compd. 227 135 (1995) 85. J-L Adam, J-L Doualan, L Criscom, S Girard, R Moncorge J. Non-Cryst. Solids 256 ± 257 276 (1999) 86. J Nishii,Y Kaite, T Yamagishi Phys. Chem. Glasses 30 (2) 55 (1989) 87. J Qiu, K Maeda, A Konishi, R Terai, K Kadono J.Non-Cryst. Solids 184 109 (1995) 88. GZhang, B Friot,MPoulain J. Non-Cryst. Solids 213 ± 214 6 (1997) 89. A E Fersman Geokhimiya (Geochemistry) (Leningrad: Nauchno- Tekhnich. Izd. Khim. Lit., 1939) Vol. 4 90. RWA Franco, C C Tambelli, C J Magon, J P Donoso, Y Messaddeq, S J L Ribeiro, M Poulain J. Chem. Phys. 109 2432 (1998) 91. A S Berezhnoi Neorg. Mater. 6 1396 (1970) c 92. M Naftaly, A Jha, E R Taylor, K C Mills J. Non-Cryst. Solids 213 ± 214 106 (1997) 93. G Zhang, J Jiang,M Poulain, A S Delben, J R Delben J. Non- Cryst. Solids 256 ± 257 135 (1999) 94. D Babel, A Tressaud, in Inorganic Solid Fluorides (Ed. P Hagenmuller) (New York: Academic Press, 1985) p. 77 95. P Gravereau, J P Chaminade, T Gaewdang, J Grannec, M Pouchard, P Hagenmuller Acta Crystallogr., Sect. C, Cryst.Struct. Commun. 48 769 (1992) 96. S Aleonard, Y Le Fur, J-C Champarnaud-Mesjard, B Frit, M T Roux J. Solid State Chem. 46 87 (1983) 97. N V Machulin, S V Nikolaev, MD Skorokhvatov,MA Terekhin, A V Etenko, P P Fedorov, O S Bondareva, V A Stasyuk, B P Sobolev, P I Fedorov, V N Makhov Razrabotka Novykh Stsintillyatsionnykh Materialov dlya Registratsii Neitrino (Preprint IAE-6016/2) [Development of New Scintillation Materials for Registration of Neutrinos (Preprint IAE-6016/2)] (Moscow: Research Scientific Centre `Kurchatov Institute', 1996) p. 9 98. J Scheffler, R Hoppe J. Fluor. Chem. 25 27 (1984) 99. J Scheffler, R Hoppe Z. Anorg. Allg. Chem. 521 79 (1985) 100. Y Le Fur, N M Khaidukov, S Aleonard Acta Crystallogr, Sect.C, Cryst. Struct. Commun. 48 978 (1992) 101. T Grande, S Jacob, J R Holloway, P F McMillan, C A Angell J. Non-Cryst. Solids 184 151 (1995) 102. R M Zakalyukin, T Yu Glazunova, A I Boltalin, P P Fedorov, B P Sobolev Zh. Neorg. Khim. 44 1730 (1999) b 103. V Mastelaro, S Ribeiro, Y Messaddeq, M A Aegerter J. Mater. Sci. 31 3441 (1996) 104. Y Kawamoto, K Ogura,M Shojiya,M Takahashi, K Kadono J. Fluor. Chem. 96 135 (1999) 105. C Legein, J Y Buzare, C Jacoboni J. Non-Cryst. Solids 161 112 (1993) 106. R M Almeida, J C Pereira, R Alcala Phys. Chem. Glasses 33 187 (1992) 107. L N Ignat'eva, T F Antokhina, V Ya Kavun, S A Polishchuk, E A Stremousova, E B Petrova, V M Bouznik, S G Bakhvalov Fiz. Khim.Stekla 21 75 (1995) 108. R M Almeida, J C Pereira, Y Messaddeq,M A Aegerter J. Non-Cryst. Solids 161 105 (1993) 109. A M Efimov J. Non-Cryst. Solids 213 ± 214 205 (1997)716 110. L N Ignat'eva, R M Zakalyukin, P P Fedorov, V M Bouznik Fiz. Khim. Stekla 26 (2000) 111. B Bureau,G Silly, J Y Buzare, C Jacoboni J. Non-Cryst. Solids 258 110 (1999) 112. J Azkargorta, I Iparagguirre, R Balda, J Fernandes, J L Adam, E Denoue, J Lucas J. Non-Cryst. Solids 213 ± 214 271 (1997) 113. F Gan J. Non-Cryst. Solids 184 9 (1995) 114. G Fonteneau, I Chiaruttini, S Mitachi, J Lucas J. Non-Cryst. Solids 140 340 (1992) 115. M E Lines J. Non-Cryst. Solids 103 265 (1988) 116. S M Lima, J A Sampaio, T Catunda, R Lebullenger, A C Hernandes, M L Baesso, A C Bento, F C G Gandra J.Non-Cryst. Solids 256 ± 257 337 (1999) 117. D A B Filho, A M Ferraz, Y Messaddeq, M A Aegerter J. Non-Cryst. Solids 213 ± 214 336 (1997) 118. R Reisfeld, M Eyal, C K Jorgensen Chem. Phys. Lett. 132 252 (1986) 119. K Hirao, S Todoroki, N Soga J. Non-Cryst. Solids 143 40 (1992) 120. W Lozano, C B Dearaujo, L H Acioli, Y Messaddeq J. Appl. Phys. 84 2263 (1998) 121. C B de Araujo,G S Maciel,N Rakov, Y Messaddeq J. Non-Cryst. Solids 247 209 (1999) 122. P Le Gall, J L Adam, J Lucas Mater. Sci. Forum 67 ± 68 509 (1991) 123. E Macho, R Balda,M J Elejalde, J Fernandez, J L Adam J. Non- Cryst. Solids 161 245 (1993) 124. R M Macfarlane, B Jacquier J. Non-Cryst. Solids 161 254 (1993) 125. K Annapurna, N S Hussain, K R Reddy, S Buddhudu Ferroelectr.Lett. Sect. 21 111 (1996) 126. A Florez, O L Malta, Y Messaddeq, M A Aegerter J. Non-Cryst. Solids 213 ± 214 315 (1997) 127. C X Cardoso, Y Messaddeq, L A O Nunes,M A Aegerter J. Non- Cryst. Solids 161 277 (1993) 128. V K Tikhomirov,M Naftaly, A Jha J. Appl. Phys. 86 351 (1999) 129. J L Adam, C Guery, P Le Gall, J Lucas Mater. Sci. Forum 32 ± 33 517 (1988) 130. I R Martin, V D Rodriges, R Alcala J. Non-Cryst. Solids 161 294 (1993) 131. M F Joubert, S Guy, C Linares, B Jacquier, J L Adam J. Non- Cryst. Solids 184 98 (1995) 132. C R Mendosa, B J Costa, Y Messaddeq, S C Zilio Phys. Rev. B, Condens. Matter 56 2483 (1997) 133. I R Martin, V D Rodrigues, U R Rodrigesmendoza, V Lavin Phys. Rev. B, Condens. Matter 57 3396 (1998) 134. V D Rodrigues, V Lavin,U R Rodrigesmendoza, I R Martin Opt. Mater. 13 1 (1999) 135. T Catunda,M L Baesso, Y Messaddeq, M A Aegerter J. Non- Cryst. Solids 213 ± 214 225 (1997) 136. L S Menezes, C B Araujo, Y Messaddeq, M A Aegerter J. Non- Cryst. Solids 213 ± 214 256 (1997) 137. S Kishimoto, K Hirao J. Non-Cryst. Solids 213 ± 214 393 (1997) 138. M T Murtagh,G H Sigel, J C Fajardo, B C Edwards, R I Epstein J. Non-Cryst. Solids 253 50 (1999) 139. D F Desousa, L F C Zonetti,M J V Bell, R Lebullenger, A C Hernandes, L A O Nunes J. Appl. Phys. 85 2502 (1999) 140. D C Yeh,W A Sibley Mater. Sci. Forum 32 ± 33 545 (1988) 141. W Lozano, C B Dearaujo, C Egalon, A S L Gomes, B J Costa, Y Messaddeq Opt. Commun. 153 271 (1998) 142. P S Christensen, P Le Gall, G Fonteneau, J Lucas, J Y Boniort, J Leboucq, D Tregoat, H Poignant Mater. Sci. Forum 32 ± 33 467 (1988) 143. T Kanamori, Y Terunuma, Y Nishida, K Hoshino, K Nakagawa, Y Ohichi, S Sudo J. Non-Cryst. Solids 213 ± 214 121 (1997) 144. K Itoch, H Yanagita, H Tawarayama, K Yamanaka, E Ishikawa, K Okada, H Aoki, Y Matsumoto, A Shirakawa, Y Matsuoka, H Toratani J. Non-Cryst. Solids 256 ± 257 1 (1999) 145. G S Macial, L D Menezes, C B Dearaujo, Y Messaddeq J. Appl. Phys. 85 6782 (1999) 146. Y Nishida,M Yamada, T Kanamori, K Kobayashi, J Temmyo, S Sudo, Y Ohishi IEEE J. Quantum Electron. 34 1332 (1998) 147. R P de Melo, B J P da Silva, E L Falca-Filho Appl. Phys Lett. 67 886 (1995) 148. C Charron, E Forget, G Fonteneau, R Rimet, J Lucas J. Non- Cryst. Solids 184 222 (1995) P P Fedorov, R MZakalyukin, L N Ignat'eva, VMBouznik 149. E Josse, J E Broquin, G Fonteneau, R Rimet, J Lucas J. Non- Cryst. Solids 213 ± 214 152 (1997) 150. N I Sorokin, P P Fedorov, R M Zakalyukin, B P Sobolev, A I Boltalin, M D Val'kovskii Neorg. Mater. 35 88 (1999) c 151. S G Bakhvalov, E M Petrova, A I Livshits, A A Shubin, V M Bouznik, V M Denisov Zh. Strukt. Khim. 39 798 (1998) f 152. T Auler, J P Donoso, P L Frare, C J Magon, Y Messaddeq, A A S T Delben J. Non-Cryst. Solids 247 92 (1999) 153. N I Sorokin, P P Fedorov, B P Sobolev Kristallografiya 39 119 (1994) d 154. B J Costa, Y Messaddeq, S J L Ribeiro, in The XIIth International Symposium on Non-Oxide Glasses and Advanced Materials (Extended Abstracts), Floriano'polis, 2000 p. 453 155. E Ya Mukhin, N G Gutkina Kristallizatsiya Stekol i Metody ee Preduprezhdeniya (Crystallisation of Glasses and Methods of Its Prevention) (Moscow: Gos. Izd. Oboron. Promsti, 1960) p. 128 156. I K Garkushin, B V Anipchenko Zh. Neorg. Khim. 44 295 (1999) b 157. B P Sobolev, P P Fedorov Neorg. Mater. 29 458 (1993) c a�Mendeleev Chem. J. (Engl. Transl.) b�Russ. J. Inorg. Chem. (Engl. Transl.) c�Inorg. Mater. (Engl. Transl.) d�Crystallogr. Rep. (Engl. Transl.) e�Quantum Electron. (Engl. Transl.) f�Russ. J. Struct. Chem. (En

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (8) 717 ¡¾ 725 (2000) Solution-free methods for the determination of the molecular mass distribution of polymers V I Irzhak Contents I. Introduction II. The direct problem III. Relations between rheological properties and molecular mass distribution of polymers IV. Methods based on dynamic measurements V. Spin-spin relaxation technique (pulse NMR spectroscopy) VI. Thermomechanical analysis VII. Conclusion Abstract. molecular the of assessment the for methods known The The known methods for the assessment of the molecular mass These considered. are polymers bulk of distribution mass distribution of bulk polymers are considered. These are are based on the analysis of the melt flow curves, the frequency based on the analysis of the melt flow curves, the frequency dependence the of shape the modulus, relaxation the of dependence of the relaxation modulus, the shape of the free free induction resonance magnetic nuclear (pulse curve decay induction decay curve (pulse nuclear magnetic resonance spectro- spectro- scopy) in compliance of dependence temperature the and scopy) and the temperature dependence of compliance in the the region (thermomechanical plateau elasticity rubbery the of region of the rubbery elasticity plateau (thermomechanical anal- anal- ysis).121 includes bibliography The ysis). The bibliography includes 121 references. references. I. Introduction The molecular mass (MM) and molecular mass distribution (MMD) are important structural characteristics which determine many physical properties of polymers.Various methods are used for the determination of MM; however, only a few of them can be applied to assess MMD. Gel permeation chromatography (GPC)1 is the most popular, since it is simple, fast in operation, reliable and relatively cheap. Other methods, in particular, those based on fractionation of a polymer, are time-consuming. All methods are based on analysis of the properties of dilute solutions. Therefore,MMDof many insoluble polymers cannot be assessed at all. This holds for, e.g., polymer networks, where it is desired to know the size distribution of linear chain segments between cross-link points. On the other hand, a method based on destruction of polymers has been developed.2 However, this method has found no practical use since it is laborious and requires preliminary establishment of the destruc- tion mechanism, which itself is a rather complicated problem.Attempts at using the relaxation properties of bulk polymers for the determination of MMD are also known. To this end, studies aimed at finding the relation between MMD and the relaxation time spectrum of a polymer were carried out. In this review, the results of such studies carried out over the last three decades are analysed. V I Irzhak Institute of Problems of Chemical Physics, Russian Academy of Sciences, 142432 Chernogolovka, Moscow Region, Russian Federation. Fax (7-096) 515 35 88. Tel. (7-096) 517 16 90. E-mail: irzhak@icp.ac.ru Received 24 March 2000 Uspekhi Khimii 69 (8) 780 ¡¾ 788 (2000); translated by AMRaevsky #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n08ABEH000587 717 717 718 720 721 722 723 Several short reviews devoted to the studies on the assessment of MMD based on the analysis of the relaxation properties of polymers have been published.3 ¡¾10 However, the authors dwelled, as a rule, on one particular approach without comparing different methods of attacking the problem.II. The direct problem To find the MMD function from relaxation data, one should know the dependence of the relaxation properties of polymers on their MMD. This problem has rather long been posed;11 ¡¾ 17 however, no unambiguous solution has been found so far.Because of this, mixing rules of the contributions of polymer fractions to the relaxation properties are widely used. Theoretical prerequi- sites to the use of this approach can be found mainly in the reptation theory.14, 17, 18 For instance, it is suggested that the time dependence of the relaxation modulus, G(t), be described by the following expression18 G(t)=G0j(t) m(t), where G0 is the modulus of elasticity in the region of the rubbery elasticity plateau; j(t) is the function which takes into account the tube renewal, m(t) is the probability that motion of an arbitrarily chosen segment of a polymer chain is confined to a deformed tube. For a monodisperse system, p¢§2exp m(t)= 8 1 l ¢§p2t :m t , l p2 pa1; 3;::: X where p is the summation index (an odd number), t is the current time, l is the maximum relaxation time for the Doi ¡¾ Edwards model.17 For a binary polymer blend, a linear mixing rule of the weight fractions wi of each component is valid t t m(t)=w1m a w2m 2 1 .l l 2 w . G(t)= iG1i =2OtU i An analogous linear mixing rule has been used in other studies.19 ¡¾ 21 A more complex G(t) dependence based on the double reptation model has been proposed 22 ¡¾ 25 X718 Maier et al.26 suggested a generalised empirical formula for calculating the time dependence of the relaxation modulus b (1) G(t)=G0 F1=bOt M , UwOMUd lnM . O? lnMe Here, F (t,M) is the function which describes the relaxation behaviour of a polymer fraction with MM=M, t is the current time, b is the parameter characterising the mixing rule and Me is theMMof a chain segment between entanglements.The results obtained using different theoretical models were compared with the experimental data on the relaxation behaviour of binary polystyrene blends.26 It was shown that dependence (1) fits the experimental data best with b=3.84 virtually irrespective of the form of the function F (t,M). In addition to the above- mentioned theoretical models,18, 22 ¡¾ 25 the step 27 and exponen- tial 28 functions and an empirical function containing a finite number (up to 25) of exponential terms with an empirical relaxation time spectrum (the so-called BSW spectrum) 29 were also considered. The last-mentioned function was found from the relaxation data for monodisperse specimens.It was found that none of the models considered described adequately the relaxation curves of both polymer blends and even monodisperse polymeric systems. This is consistent with the conclusions reported in a recent study 30 in which the results of calculations using the Doi ¡¾ Edwards model were compared with the experimental results obtained for monodisperse polystyrene specimens. Empirical mixing rules found mainly in the studies on the analysis of relaxation properties of binary blends of monodisperse polymers are applied to the relaxation time spectrumH(l) (l is the relaxation time), which is defined as follows: G(t)= HOlUexp ¢§ t dl . l O lmax lmin The simplest relation proposed by Ninomiya 31 has the form , H(l)= wiHi l Ai ia1 XN where Hi (l) is the relaxation time spectrum of a monodisperse polymer; Ai is the time shift factor due to intermolecular inter- action between polymers (such an interaction facilitates the relaxation of the components with higher MM and hampers the relaxation of the components with lower MM).Attempts at taking into account these interactions by deriving generalised formulae lead to sophisticated expressions 32 ¡¾ 34 H(l)= wiwjHij l Aij , ia1 ja1 XN XN l H(l)= wiwjwkHijk Aijk , ia1 ja1 ka1 XN XN XN which seem to be hardly applicable in practice. No simpler is yet another relationship 35, 36 i (2) , H(l)= BiHi lAi ia1 XN where B is the form factor,37, 38 aijwj i a w2i a wi jaia1 i XN is the BSW spectrum, which is a function of MM, and Hi namely,29 lAi n li , 1 ¢§ li l l max max H(li)=n h n is a constant (n<1) and h is the Heaviside step function.V I Irzhak Table 1. Characteristics of components and parameters Ai and Bi of Eqn (2) for a binary polystyrene blend.39 w2 a A2 A1 B2 M2 M1 1.00 1.06 1.17 1.50 1.78 2.89 3.89 0.000 0.003 0.005 0.007 0.020 0.050 0.170 1.000 0.00100 0.00080 0.00055 0.00040 0.00 0.01 0.03 0.05 0.10 0.20 0.40 1.00 0.01 0.01 0.01 0.01 38900 38900 38900 38900 38900 38900 38900 38900 23400 38900 72400 124000 70.030 0.045 0.060 0.070 0.110 0.230 1.000 0.00023 0.00100 0.00650 0.04000 0.90 1.10 1.00 1.20 427000 427000 427000 427000 427000 427000 427000 427000 2810000 2810000 2810000 2810000 aw1+w2=1. As an example, Table 1 lists the values of the parameters of Eqn (2) taken from Ref.39. As can be seen, the parameters Ai and Bi depend on the composition andMMof the polymer blend and on the ratio of MM of the blend components. It was also found that all these characteristics affect the values of the maximum relaxation time of each component. This can be exemplified in the data of a study on the relaxation of polymer chains by infrared dichroism 40 (Table 2). Table 2. Maximum chain relaxation time (s) in a binary blend of deuterated and hydrogenated polyisoprene.40 lH lM (for H+L Content of a high- lL molecular-mass component (vol.%) blend) (for M+H blend) (for L+M blend) 2.8 3.5 4.5 6.4 6.7 7.0 20 60 120 150 180 210 50 70 130 160 186 210 10 20 30 50 75 100 Note. H, M and L denote the blend components with high (3706103), medium (1256103) and low (536103) molecular mass, respectively. No general rule exists which would allow passage from the relaxation properties of monodisperse polymers to those of polydisperse polymers. This is largely due to the lack of a theory which can give an adequate description of the dynamic behaviour of polymeric systems.41 Currently, there is no unique solution of the direct problem of obtaining a relaxation time spectrum based on a particular model or the relaxation properties of a known blend of monodisperse fractions. Obviously, this severely hampers the solution of the inverse problem, consisting of finding the MMD of a polydisperse polymer from its relaxation character- istics.III. Relations between rheological properties and molecular mass distribution of polymers Much attention has been paid to the studies on the relations between MMD and viscosity (Z) of polymer melts and concen- trated solutions.12, 42 Most often, this problem was solved by relating the viscosity to the number-average (Mn 7 ), or weight- average (Mw 7 ), or a high-level-average MM (Mz 7), or even to their combination (e.g.,Mn 7 Mz 7/Mw 7 ).Solution-free methods for the determination of the molecular mass distribution of polymers 1=a, Z! Mai i These approaches have not been supported by any theoretical concepts.Attempts to represent the relation between viscosity and MMD in the form X where a characterises the dependence of viscosity on theMMof a monodisperse polymer, Z!Ma are equally unjustified. An attempt to find a physically substantiated method relating the rheological properties and MMD of a polymeric system has been undertaken.43 This approach was based on the following assumptions: (i) the relaxation time spectrum is limited by the maximum relaxation time lmax(g .), where g .is the shear rate; (ii) the 1=3:4 m ma1 K OM ¢§ wi a ma1 ¢§MmU¢§1, (3) 1=3:4 ZK Z iama1 maximum relaxation time is inversely proportional to the shear rate; (iii) the maximum relaxation time of a given polymer is affected by the presence in the blend of another polymer with a lower MM and (iv) the relaxation time spectrum of a given polymer is independent of the type of polydispersity of the polymer.{ Based on these assumptions, the following formula relating the MMDfunction to the viscosities of the polymer fractions was derived 43 X w (3a) .¢§ Zm1=a3:14 ¢§ Zm1=3:4 m a where wi is the weight fraction of a polymer fraction with MM=Mi , Zm is the viscosity of a polymer with MM=Mm , K is the coefficient in the equation relating the viscosity of a monodisperse polymer to itsMM Z=KM3.4.Hence it follows that 1=3:4 Z1=3:4 ¢§ Z1=3:4 m m¢§1 1K Mma1 ¢§Mm Mm ¢§Mm¢§1 Expression (3a) was used for analysing a number of binary blends and commercial polystyrene specimens.44 It was shown that the polydispersity indices, Mw , g=Mn determined by rheological methods are close to those found by GPC. A physical substantiation of the rheological approach has been given by Malkin and Teishev 45 ¡¾ 47 who used the phenom- enon of flow separation in monodisperse polymers, discovered by Vinogradov,48 as the basis for their consideration. Mathemati- cally, this phenomenon can be described by the relation Z(g .)= ts / g .for t>ts. Z0 for t4ts Here Z0 is the highest Newtonian viscosity and ts is the stress at which flow separation occurs in the polymer.Based on the known relationship, Z0=KMa, Malkin and Teishev derived the formula 45, 46 { This assumption seems to be not universally true for systems with entanglements. 719 g . (4) Z(g .)= OKMaU1=afOMUdMa a fOMUdM , O MO U 1=a O0 ts g . 0 MO U g . where the integration limit is , g .M(g .)= Ks t 1=4 and f(M) is theMMfunction. Formula (4) can be used for unimodalMMDfunctions which include a small number of parameters (Malkin and Teishev 46 analysed the Bisley, lognormal and other two-parameter distribu- tion functions). In this case, the melt flow curve is sufficiently sensitive to the parameters of the distribution function.Gordon and Shaw 4 calculated the MMD function using another analytical expression for the flow curve, namely, Z(g .)= g .7v at t>ts, Z0 at t4ts where ts (and, correspondingly, g .s) defines the point of transition from Newtonian to a non-Newtonian flow and n is the exponent in the power law of the melt flow Z!g .7v. The position of this point depends on theMMof the polymer. The procedure for the calculation of the MMD function depends on which mixing rule for viscosities is used. Gordon and Shaw considered the following mixing rule: log Z=w1log Z1+w2log Z2 . Using this mixing rule, the relation between the integralMMD function F(M) and parameters of the melt flow curve has the form F(M)=1 a 1 d lnZ v d lng .g .agOMU, where the function g(M) relates the shear rate to the MM of the polymer and the parameter n.A more general mixing rule, Z1/a=w1Z1/a+w2Z1/a leads to the following formula, , F(M)=1 a 1v 0 sm¢§a=v s 1=a v=a d lnZ g . g .d lng .g .ag . Z Z where m= M . Mw (5) , a=b0+b1 Ml An algorithm and software for calculations of the MMD function using the last two formulae have been published in a monograph.4 It should be noted that the exponent a is not known with certainty. Recently,49 it was shown that the parameter a depends on the shear rate (and frequency) in such a manner that its magnitude decreases from values in the range between 3 and 5, which is typical of low shear rates (the highest Newtonian viscosity), down to 0 for high shear rates (the lowest Newtonian viscosity).Moreover, it was empirically found that the parameter a depends also on MMD of the polymer as follows Mh where b0=3.390.37, b1=0.030.04, and Mh and Ml are the MM of the high-molecular and low-molecular mass components of a binary polymer blend, respectively. The values of the720 coefficients of equation (5) also vary to a certain degree depending on the type of the polymer. With some grounds, it is believed 44, 46 that the rheological approach can hardly be used for characterisation of an arbitrary MMD function; however, the authors of a monograph (Ref. 4) have a more optimistic view on these things. IV. Methods based on dynamic measurements Attempts at using the frequency dependences of the real and imaginary parts of the relaxation modulus of elasticity for the determination of the MMD of polymers have long been under- taken.Ninomiya and Fujita seem to be the first who reported studies in this line of investigation as long ago as 1957.50 ± 52 This approach is based on theoretical or empirical relationships between the relaxation time spectrum and MM of poly- mers 12, 13, 53 (see also Section II). In the simplest form, the theory of polymer dynamics was reported by Rouse who developed the well-known single-chain model which is basic to the studies by Ninomiya.31, 50 ± 52, 54, 55 More recent investigations 56 were based on the Doi ± Edwards model in which interchain interaction is considered as interweav- ing of polymer chains, or entanglements, which confine mainly lateral displacements of the polymer chains.The results of studies carried out on monodisperse specimens of different polymers and their blends (for the most part, binary blends) 26, 39, 40, 57 ± 61 can be outlined as follows. For polymers with narrow MMD, the log-log plots of the frequency dependence of the real part of the relaxation elasticity modulus, G0(o), nearly coincide in the high-frequency region (the region of the rubbery elasticity plateau). Marked distinctions are observed in the region of low-frequency `tails' where G0(o) depends on o as G0(o)!o2. The `tails' can be matched by using a shift factor proportional to MM raised to a certain power; it is commonly accepted 12, 13, 62, 63 that the exponent value lies between 3.4 and 3.7.The frequency dependences of the imaginary part of the relaxation modulus of elasticity,G00(o), pass through a maximum.The position of the maximum depends onMMof a polymer in the same way as the low-frequency `tails' of the G0(o) and G00(o) curves. In binary polymer blends, the polymer fractions possess similar relaxation properties; however, no complete coincidence with the behaviour of monodisperse systems is achieved. That is, the low-frequency `tails' are shifted towards higher frequencies for the component with the higher MM and towards lower frequen- cies for the component with the lower MM. This phenomenon is associated with the interchain interaction (coupling effect, nematic interaction),39, 40, 64 ± 68 which also man- ifests itself as the orientation of low-molecular-mass chains (even oligomers) in the bulk of the oriented high-molecular-mass matrix.It should be noted that this frequency shift should occur even in the absence of intermolecular interaction. Indeed, let the relaxation properties be determined by the only relaxation time, which is of course dependent on the MM of the high-molecular- mass fraction. Then, for o?0, the real part of the relaxation modulus of elasticity of a binary polymer blend takes the form 2 1 1 á w2 ?o2Öw1l21 á w2l22 Ü. G0ÖoÜ à w 2 1 o2l2 1 á o2l2 o2l2 1 á o2l2 Obviously, it is this effect that should be observed if l1>l2 , namely, a shift of the relaxation moduli of elasticity of the high- molecular-mass and low-molecular-mass fractions towards high and low frequencies, respectively. An analogous relationship is also valid for G00(o), foro?0 á w2 G00ÖoÜ à w1 1 ol2 1 á o2l2 ol1 1 á o2l2 The aforesaid can be illustrated by the corresponding plots shown in Fig.1. On the other hand, a direct mutual effect of the polymer fractions (see Table 2) cannot also be ruled out. a logG0(o) 0.0 3 1 70.5 71.0 71.5 b logG00(o) 70.5 3 1 71.0 71.5 72.073 71 72 Figure 1. Frequency dependences of the real (a) and imaginary (b) parts of the relaxation modulus of elasticity for specimens with one relaxation time of 100 (1) and 1 arbitrary unit (2) and for their blend containing 25% of the first component (3).A salient feature of the G0(o) and G00(o) curves of binary blends is that the height of the rubbery elasticity plateau (in the first case) and that of the peak (in the second case) depend on the relative concentration of components. Similarly to the G(t) curve, a `step' appears on the G0(o) curves. The G00(o) curves are characterised by two peaks and the peak height ratio depends on the ratio of the blend components. However, data on the relation between the peak heights and the concentration of the blend components are contradictory. Different authors use different formulae. For instance, a linear dependence was suggested 69 á w2G2 GÖtÜ à w1G1 t ; d t d1 . wi à G0ÖoiÜ 1=2 G0 the parameter d is associated with the coupling effect.Other (mainly quadratic) types of mixing rules have also been suggested.13, 70 ± 72 For instance, based on the concept that each polymer fraction can be characterised by a frequency correspond- ing to the MM of the fraction and that at lower frequencies the contribution of this fraction to the relaxation modulus can be neglected, a procedure was developed 26 for the determination of theMMDfunction taking into account a quadratic dependence of the real part of the relaxation modulus of elasticity on the concentration of the high-molecular-mass component. Each frac- tion makes a contribution Comparison of the results obtained using this procedure with those obtained by GPC showed 27 that for binary polymer blends the dynamic approach is less accurate than the rheological method based on the analysis of the melt flow curve.?oÖw1l1 á w2l2Ü. 2 2 2 0 logo 2 V I IrzhakSolution-free methods for the determination of the molecular mass distribution of polymers . The complex viscosity, Z*(o), was also proposed 44, 49 to be used for assessing the MMD function. qAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA aG0OoUa2 a aG00OoUa2 ZOoU a o In the high-frequency region (for o??), the complex viscosity decreases nearly proportionally to o72, whereas in the low-frequency region (for o?0) its takes the limiting value corresponding to the highest Newtonian viscosity. It is in the low-frequency region that the strongest dependence of Z*(o) on MMD of a polymer melt is observed and the parameter a [see Eqns (4) and (5)] has the maximum value.Thus, the frequency dependence of viscosity is similar to the dependence of viscosity on the shear rate, which corresponds to the Cox ¡¾ Merz rule.12 Comparison of different approaches to the solution of the problem of determining the MMD function allowed Tuminello and Cudre-Mauroux 44 to conclude that the best results can be obtained by analysing the frequency dependence of the complex viscosity; it was also found that analysis of the real part of the relaxation modulus of elasticity gives overestimated, whereas analysis of the imaginary part gives underestimated values. Investigations on the relaxation of dielectric properties are widely used for studying the relaxation properties of polymers.The relation between dielectric properties and MMD has been analysed 73 ¡¾ 77 taking cis-polyisoprene as an example. It was shown that in this case the dielectric loss function can be predicted with ease if the autocorrelation function of a single chain, the mixing rule and the MMD function are known. Difficulties arise when considering the inverse problem. An attempt has been undertaken 77 to find the MMD function from the dielectric relaxation data using the Tikhonov regularisation, which was previously used for calculating the relaxation time spectra.78, 79 This study was carried out assuming a linear mixing rule and that the MMD function is described by the Flory ¡¾ Schulz equation fOMU!Mnexp ¢§ M .Mn Information on the width of MMD can be obtained from the analysis of the effective relaxation time 80 in the low-frequency region or, in other words, of the low-frequency `tails' of the G0(o) and G00(o) curves. Indeed, assuming that the relaxation time of a monodisperse polymer depends on its MM, l=KMa, foro?0 we get G0(o)&G i wiMai . 0o2Xwil2i a G0o2K2XwiM2i a, G00(o)&G0oK i Hence Xi 2 , 2: G0OoU hli a oKhMai G00OoU a oKPPwiMai a oKhMai1 . hli1: G00OoU a oK G0 i i wiM2i a i wiMai X The subscripts 2 and 1 on the mean values of l and MM correspond to a greater degree of averaging in the former case and to a lesser one in the latter.The ratio of the mean values can serve as a measure of the width ofMMD.This algorithm is analogous to that which is commonly accepted for assessing the width ofMMD and uses the mass-average and number-average MM and their ratio as the polydispersity index; the difference is that in the latter case a=1 (in the relaxation method, a&3 ¡¾ 4). Of course, it should be kept in mind that even monodisperse polymers possess a relaxation time spectrum; as a rule, the higher theMMthe broader the relaxation time spectrum.29 721 V. Spin-spin relaxation technique (pulse NMR spectroscopy) Both rheological and relaxation approaches have been used for the analysis of the MMD of linear polymers. For polymer networks, a steady flow is impossible.However, analysis of stress relaxation data (especially, of the frequency dependences of the relaxation modulus) can be used for extracting information on the MMD of chains of a polymer network. This was done using the spin-spin relaxation technique (pulse NMR spectroscopy).81 ¡¾ 89 At sufficiently high temperatures (T>Tg+100), the depend- ence of the free induction decay on time is given by the following expression (6) GOtU a ON pO=2 1 0 gON; y; tUPONUsinydy dN, where N is the number of segments (the chain length) and y is the angle between the vector connecting the chain ends and the direction of external magnetic field. g(N, y, t)&exp ¢§ol2O3cos2y ¢§ 1U2t 2 lnN, N2 where ol is the second moment of the free induction decay curve, expressed in frequency units.In the time domain, the free induction decay has a long slowly descending component. The initial portion of this component contains information on MMD, while the length of its `tail' is proportional to the average length of the chain segment between entanglements and inversely proportional to the current time t. Hence it follows that one can assess the MMD for those chains which are shorter than the length of the chain segment between entanglements (N0). As was shown experimentally,85 N0&105 segments. It should be noted that no independent determination of the chain segment length was performed, which introduces an error into the results obtained. However, even assuming that the segment is a unit of the polymer chain, this method can be applied to those linear polymers for whichMMis much higher than 106.An essential result obtained in a series of studies 85 ¡¾ 89 is that formula (6) can be used for assessing MMD of chain segments between entanglements in polymer networks. At T>Tg+100, intermolecular interaction has almost no effect on the relaxation of sufficiently long chains with the number of segments N>10. Hence the curve of the free induction decay characterises the MMDof chain segments between entanglements. In this case, the integral of the functionG(t) can be represented in a linearised form Ot F (t)= t 0 1GOt1Udt1& ¢§ 0:03hN2i a hNit pAAA 2 3 p . The coefficients of this equation define the polydispersity index of the chains of the polymer network, gn: Mnw gn=Mnn a hN2i hNi2 (the index n indicates that the chains of the polymer network are considered). Thus, the analysis of the curve of the free induction decay makes it possible to assess MMD of both linear polymers and chain segments between entanglements of the polymer network.The theoretical conclusions have been confirmed experimen- tally taking poly(tetrahydrofuran) with isocyanate end groups as an example.90 In this case, the polymer network resulted from a trimerisation reaction,91 which would ensure identical MMD for both the initial prepolymer and chain segments between entangle- ments. Comparison of the GPC data on MMD of poly(tetrahy- drofuran) macromolecules with the results of NMR studies on MMD of the chains of the polymer network made it possible to722 draw a conclusion thatNMRspectroscopy can be effectively used in this case.Pulse NMR spectroscopy was used for studying the MMDof the chains in different polymer networks depending on the conditions of the synthesis,88 as well as the evolution of MMD during ageing 92 and at thermal 92, 93 and mechanical 94 destruc- tion.Limitation on the use of NMR spectroscopy for assessing the MMD of the chains of polymer networks with low concentration of nodes is due to the fact that entanglements behave as chemical cross-link points, which leads to the anisotropy of the molecular motion of chain segments. This limitation is similar to that which appeared in the determination of the equilibrium modulus of elasticity of polymer networks: for networks with low node concentration, the number of nodes is equal to the sum of chemical (covalent) cross-link points and topological entangle- ments.95 ¡¾ 102 In highly cross-linked polymer networks, the effective dipole- dipole interaction is determined not only by the interaction of protons in the same chain, as is theoretised; an appreciable contribution comes also from the interaction between protons of different chains.This imposes some limitations on the use of NMR spectroscopy. Finally, one should keep in mind that the model network considered here has a simple topology (chains linked by tri- or tetrafunctional cross-link points), whereas the real network contains `tales', small rings, a sol fraction and other defects responsible for considerable uncertainty of the relaxation time spectra of a polymer.95, 103, 104 VI.Thermomechanical analysis Usually, thermomechanical curves of polymers are analysed based on the concepts of the relaxation time spectrum and the temperature dependence of relaxation times.12 It is this approach that has been used for the rationalisation of (i) the relation between the temperature dependence of the extent of the rubbery elasticity plateau and the MM of linear polymers discovered by Kargin and Slonimskii 105 and (ii) such an important relation as the Williams ¡¾ Landel ¡¾ Ferry formula.13 It is the relaxation nature of deformation phenomena that is responsible for the fact that the shape of a thermomechanical curve is `sensitive' to the polydisper- sity of a polymer.106 Recently,107 a procedure for solving the direct problem, which implies theoretical determination of the flow point (temperature) of a linear polymer, Tf , under conditions of thermomechanical analysis (TMA) from the structure and MM of the polymer, has been proposed.However, the authors were rather sceptical of the possibility of using TMA for the determination of MMD of polymers. On the other hand, such a possibility has been demon- strated.108, 109 Physical substantiation 110 ¡¾ 113 of the approach used was based on the model of a network of physical links.41, 114 In its most explicit form, the relaxation modulus (compliance or deformation at a given stress) can be related to the MMD of a polymer using the classical theory of rubbery elasticity of polymer networks and the temperature dependence of the concentration of physical nodes.110 However, an analysis 112 showed that the concentration of nodes formed by weak intermolecular bonds cannot vary substantially in a rather narrow temperature range between the glass transition temperature and the flow point (for linear polymers) or the temperature at which the equilibrium modulus is attained for covalent polymer networks (the extent of the rubbery elasticity plateau).Much stronger is the temperature dependence of the lifetime of the nodes as well as the dependence of the relaxation of the polymer chain conformations on the temperature ¡¾ time superposition.Irrespective of the details of the physical pattern of the phenomenon, the kinetic law can be to sufficient accuracy expressed by a step function.110 Relaxation V I Irzhak occurs in a narrow time range near the characteristic relaxation time, which depends on theMMof a polymer or a chain segment between entanglements of the polymer network. If mechanical properties of a system are considered according to the Kelvin ¡¾ Voigt model, the following formulae relating the function j(T) containing information on MMD to compliance (deformation) are valid: for a linear polymer (7) j(T)=JT ¢§ J0 , JT for a polymer network eq , j(T)=JT ¢§ J0 JT J Jeq ¢§ J0 where JT , Jeq and J0 is the compliance at a given temperature, the equilibrium compliance (for a polymer network) and the compli- ance of a hypothetical polymer with infinite MM, respectively.The function j(T) is in essence the integral MMD function expressed in terms of the flow point of a chain with a given MM rather than theMMof a polymer fraction: w j(T)= idiOt;T U,a b 0 Tube displacement (arb. units) ia1 X? di (t, T)=d[t7li (T)]= 0 for t<li (T ) 1 for t5li (T), li (T ) is relaxation time of the chain conformation. The J0 value characterises that part of the temperature dependence of the modulus of elasticity (compliance) which is unaffected by the MM and MMD of the polymer and serves as a basis for the calculations of the function j(T ) [see Eqn (7)].Schematically, the procedure for the determination of j(T ) using the temperature dependence of deformation (compliance) of a linear polymer is shown in Fig. 2. From the rod displacement with respect to the dilatometric curve 1 one can determine the J0 value (in rod displacement units, this is the value h0 calculated as the difference between the corresponding points on curves 2 and 1) and compliance J, which is proportional to h calculated as the difference between the corresponding points on curves 3 and 1. The main problem ofTMAstudies is to determine the relation between the temperature and MM, i.e., to pass from the function j(T ) to the integralMMDfunction j(M). Such a relation can be found after choosing a particular mixing rule.As in many cases considered above, it seems reasonable to assume that the relaxa- tion time of a chain conformation with a givenMMis independent h 30 710 3 20 720 h0 h0 2 730 h 10 740 1 0 Figure 2. Rod displacement (h) curves obtained in TMA studies. Experimental curves (a) and difference between the rod displacement and dilatometric curves (b). Dilatometric curve (1), hypothetical curve for a polymer with infinite MM(2) and the real rod displacement curve (3). Tube displacement (arb. units) 750 20 10 10 T(arb. units) 0Solution-free methods for the determination of the molecular mass distribution of polymers of whether it is a monodisperse or a polydisperse polymer. As can be seen, this problem is common to all methods for the determi- nation of the MMD function. At first glance, if this condition is met, the problem seems to be simple. That is, one needs to find a relation between the extent of the rubbery elasticity plateau, DT=Tf7Tg and MM.However, constructing this dependence faces some difficulties. The first problem is to determine the basic function, J0(T). This requires studying the temperature dependence of the deformation of model polymer specimens with narrowMMDand highMMin the region of the rubbery elasticity plateau. Here, it is encouraging that such a dependence is rather weak,110, 113 so to a sufficient accuracy the function J0(T ) can be considered constant. Yet another problem is associated with the fact that the temperature range corresponding to passage from the rubbery elasticity plateau to the flow of a polymer is rather wide and depends on both theMMandMMDof the polymer.That is why several formulae for calculating the function DT(M) were derived, e.g., the Kargin ¡À Slonimskii formula 105 and a number of similar formulae,109 e.g.: for polar polymers logM=1.6+200 �¢ DT , 20DT for non-polar polymers, logM=2.0+100 �¢ DT . 10DT In a recent study 113 of monodisperse specimens of poly(oxy- ethylene glycol) with low-molecular masses (1500 ¡À 12000), a rather sharp transition from the rubbery elasticity plateau to the flow of a polymer was observed and a calibration curve was constructed in the form 1.0 0.8 0.6 0.4 0.2 0.0 4 5 6 logM 1000/T /K71 3.6 3.4 3.2 D1 D2 D3 3.0 3 Figure 3.Integral MMD functions (a, b) and calibration curve (c) for amorphous polypropylene.115 Obtained from GPC (a) and TMA (b) studies. Rod load, g: 0.5 (1), 2.0 (2). 1073 MM: 180 (a, b); c: 52 (1), 189 (2) and 330 (3). Integral MMD function b a D1 D2 0.04 0.08 17Tg/T c 4 5 logM 6 723 (8) lnM=A+B 1 ¡¦ Tg . T It was found that for a wide array of polymers A&3 and B&17. A calibration procedure was proposed,115 which refers to a polymer with an arbitrary MMD characterised by, e.g., GPC. To this end, the integral curve j(M) is compared with the integral curve j(T) which can be constructed in the j(T ) ¡À Tg/T coordi- nates. By comparing the MM and T values corresponding to the same value of the integral fraction it is possible to find the relation between MM and temperature, i.e., to construct a calibration curve for a wide range of MM using the data for one polymeric specimen.The integral MMD curve of amorphous polypropylene is shown in Fig. 3 a and the dependence j(T ) in the coordinates of Eqn (8) is shown in Fig. 3 b. The calibration curve (Fig. 3 c) was constructed by comparing theMM and T values. As can be seen, the results obtained for three samples of polymers with different MMgive a versatile dependence of the type (8). Thermomechanical analysis can be applied to different types of loading of polymeric specimens, among which penetration and tension are the most popular.106, 113 VII.Conclusion Consideration of `solution-free' approaches to the determination ofMMDof polymers proposed by different authors shows that, in essence, there are no well-developed methods so far. First of all, this is due to the fact that the inverse problem, viz., determination of the relaxation time spectrum and, hence, the MMD of a polymer from the relaxation data, belongs to the so-called ill- posed mathematical problems.26, 47, 77 ¡À 79, 116 ¡À 118 Attempts at solving the problem require some a priori statements or exper- imental establishment of the relaxation time spectrum (or other property, e.g., viscosity or the flow point) with the MM of a polymer. In essence, it is this point to which all the approaches discussed above come. Regularisation methods for solving ill- posed problems,119 as applied to the problem of determining the relaxation time spectrum or, more particularly, the MMD of a polymer 77 ¡À 79, 117, 118 are also based on certain assumptions (see, e.g., Ref.77). Baumgarten et al.29 derived the relaxation time spectrum (the so-called BSW spectrum) of monodisperse polystyrene specimens without any preliminary postulates by analysing the relaxation data using IRIS software (see Ref. 120). More recently, the BSW spectrum was used for the description of the relaxation properties of polybutadiene.38, 39 Empirical constants characterising the BSW spectrum, as well as the relaxation time spectrum itself, have no theoretical substantiation and it is unclear to what extent the expansion of the function G(t) in exponents is unique.It is likely, the lack of a well-founded theory which takes into account mutual effects of polymers with differentMMis the most serious obstacle to the development of a relevant relaxation method for the determination of MMD of polymers. Similarly to the Rouse and Doi ¡À Edwards theories, such a theory would help to obtain an exact relation between the MM and the correspond- ing relaxation time. An empirical relation, l(M)!Ma, where a&3.4 casts some doubt since there are strong reasons (see, e.g., Refs 30, 41, 80,121) to believe that a=4. Thermomechanical analysis seems to be the most promising method. Probably, mutual effects of polymers are more pro- nounced in isothermal (rheological and frequency) experiments than in the studies on the temperature dependence of deformation parameters.The relaxation time spectrum uniquely characterises both the frequency and temperature dependences of the relaxation properties of a polymer. However, as was shown experimentally, the relation between the flow point and the MM of a polymer is724 much more pronounced than the relation between the frequency (time) of this transition and MM. Further development of `solution-free' methods will open prospects for the determination of MMD of polymer blends and composite matrices. This may give an impetus to scientific research in this area. This review has been written with the financial support by the Russian Foundation for Basic Research (Project No.96-03- 33358). References 1. B G Belen'kii, L Z Vilenchik Khromatografiya Polimerov (The Chromatography of Polymers) (Moscow: Khimiya, 1978) 2. A A Berlin, S A Vol'fson Kineticheskii Metod v Sinteze Polimerov (Kinetic Method in Synthesis of Polymers) (Moscow: Khimiya, 1973) 3. M Fleissner Makromol. Chem., Makromol. Symp. 61 324 (1992) 4. G V Gordon, M T Shaw Computer rams for Rheologists (Munich, Vienna, New York: Hauser Publisher, 1994) Ch. 4, p. 72 5. D W Mead J. Rheol. 38 1797 (1994) 6. S H Wasserman J. Rheol. 39 601 (1995) 7. HBraun,AEckstein,KFuchs, Ch Friedrich Appl. Rheol. 6 116 (1996) 8. Y Liu, T Shaw,W H Tuminello J. Rheol. 42 453 (1998) 9. V I Irzhak Vysokomol.Soedin., Ser. B 41 1063 (1999) a 10. V I Irzhak, in Heterophase Polymer Network Materials: Synthesis, Characterisation and Properties (Ed. B A Rozenberg) (London: Gordon and Breach, 2000) 11. F Bueche Physical Properties of Polymers (New York, London: Interscience, 1962) 12. G V Vinogradov, A Ya Malkin Reologiya Polimerov (Rheology of Polymers) (Moscow: Khimiya, 1977) 13. J D Ferry Viscoelastic Properties of Polymers (New York: Wiley, 1980) 14. P-G de Gennes Scaling Concept in Polymer Physics (Ithaca, London: Cornell University Press, 1079) 15. G M Bartenev, A G Barteneva Relaksatsionnye Svoistva Polimerov (Relaxation Properties of Polymers) (Moscow: Khimiya, 1992) 16. A Yu Grossberg, A R Khokhlov Statisticheskaya Fizika Makro- molekul (Statistical Physics of Macromolecules) (Moscow: Nauka, 1989) 17.M Doi, S F Edvards The Theory of Polymer Dynamics (Oxford: Clarendon Press, 1986) 18. M Doi,W W Graessley, E Hefland, D S Pearson Macromolecules 20 1900 (1987) 19. W Graessley,M Struglinski Macromolecules 19 1754 (1986) 20. M Rubinstein, E Hefland, D Pearson Macromolecules 20 882 (1987) 21. H Watanabe,M Tirrell Macromolecules 22 927 (1989) 22. C Tsenoglou Am. Chem. Soc. Polym. Prepr. 21 185 (1987) 23. J de Cloizeau Macromolecules 23 4678 (1990) 24. J de Cloizeau Macromolecules 25 835 (1992) 25. P Cassgnau, J P Monfort, G Marin, P Monge Rheol. Acta 32 156 (1993) 26. D Maier, A Eckstein, Cr Friedrich, J Honerkamp J. Rheol. 42 1153 (1998) 27. W H Tuminello Polym.Eng. Sci. 26 1339 (1986) 28. C Tsenoglou Macromolecules 24 1762 (1991) 29. B Baumgartel, ASchausberger,HHWinter Rheol. Acta 29 400 (1990) 30. V I Irzhak Vysokomol. Soedin., Ser. B 42 (9) (2000) a 31. K Ninomiya J. Colloid Sci. 14 49 (1959) 32. W Prest Polym. J. 4 163 (1973) 33. W Prest, R Porter Polym. J. 4 154 (1973) 34. M Kurata, K Osaki, Y Einaga, T Sugie Polym. J. 12 849 (1974) 35. D Bogue, T Masuda, Y Emaga, S Onogi Polym. J. 5 563 (1970) 36. J K Jackson, H H Winter Rheol. Acta 35 645 (1996) 37. A Schausberger Rheol. Acta 25 596 (1986) 38. J Jackson,M de Rosa, H Winter Macromolecules 27 2426 (1994) 39. J K Jackson, H H Winter Macromolecules 28 3146 (1995) 40. CMYlitalo, J A Kornfeld,G G Fuller Macromolecules 24 749 (1991) 41. V I Irzhak Usp.Khim. 69 283 (2000) [Russ. Chem. Rev. 69 261 (2000)] 42. Y H Zang, R Muller, D Froelich Polymer 28 1577 (1987) 43. B H Bersted, J D Slee J. Appl. Polym. Sci. 21 2631 (1977) 44. W H Tuminello, N Cudre-Mauroux Polym. Eng. Sci. 31 1496 (1991) V I Irzhak 45. AYa Malkin,AETeishev Vysokomol. Soedin., Ser.A29 2230 (1987) a 46. A Ya Malkin, A E Teishev Vysokomol. Soedin., Ser. A 30 175 (1988) a 47. A Ya Malkin, A E Teishev Polym. Eng. Sci. 31 1590 (1991) 48. G V Vinogradov Vysokomol. Soedin., Ser. A 13 294 (1971) a 49. Y Liu, M T Shaw J. Rheol. 42 267 (1998) 50. K Ninomiya, H Fujita J. Colloid Sci. 12 204 (1957) 51. K Ninomiya, H Fujita J. Polym. Sci. 24 233 (1957) 52. K Ninomiya, H Fujita J. Phys. Chem. 61 814 (1957) 53. S Ya Frenkel', G M Bartenev Fizika Polimerov (The Physics of Polymers) (Leningrad: Khimiya, 1990) 54.K Ninomiya J. Colloid Sci. 17 759 (1962) 55. K Ninomiya, J D Ferry J. Colloid Sci. 18 421 (1963) 56. S Wu Polym. Mater. Sci. 50 43 (1984) 57. S Onogi, T Masuda Kobunshi 17 640 (1968); Ref. Zh. Khim. 5(II) 5 S 82 (1969) 58. G V Vinogradov Mekhan. Polim. (1) 160 (1975) 59. M Rubinstein, R H Colby J. Chem. Phys. 89 5291 (1988) 60. C Hayes, L Bokobza, F Boue, E Mendes, L Monnerie Macromolecules 29 5036 (1996) 61. K Fuchs, C Friedrich, J Weese Macromolecules 29 5893 (1996) 62. J S Fodor, D A Hill Macromolecules 28 1271 (1995) 63. K Adachi, T Kotaka Macromolecules 17 120 (1984) 64. W W Merrill, M Tirrell, J-P Jarry, L Monnerie Macromolecules 22 896 (1989) 65.M Doi, D Pearson, J Kornfeld, G G Fuller Macromolecules 22 1488 (1989) 66. H Watanabe, T Kotaka, M Tirrell Macromolecules 24 201 (1991) 67. U Seidel, R Stadler, G G Fuller Macromolecules 27 2066 (1994) 68. U Seidel, R Stadler, G G Fuller Macromolecules 28 3739 (1995) 69. T Fujimoto,M Dzaki, M Nagasawa J. Polym. Sci., Polym. Chem. Ed. 6 129 (1968) 70. W W Graessley Farad. Symp. Chem. Soc. 18 7 (1983) 71. W W Graessley, in Physical Properties of Polymers (Eds J E Mark, A Eisenberg, W WGraessley, L Mandelkern, J L Koenig) (Washington, DC: American Chemical Society, 1984) 72. W H Tuminello Polym. Eng. Sci. 29 645 (1989) 73. Y Imanishi, K Adachi, T Kotaka J. Chem. Phys. 89 7585 (1988) 74. K Adachi, I Nishi, S Itoh, T Kotaka Macromolecules 23 2550 (1990) 75.K Adachi, S Itoh, I Nishi, T Kotaka Macromolecules 23 2554 (1990) 76. J S Fodor, D A Hill Macromolecules 26 5379 (1993) 77. J S Fodor, D A Hill J. Phys. Chem. 98 7674 (1994) 78. J Honerkamp, J Weese Macromolecules 22 4372 (1988) 79. C Elster, J Honerkamp Rheol. Acta 10 161 (1991) 80. T F Irzhak, O V Nikitina, L I Kuzub, V I Irzhak Vysokomol. Soedin., Ser. A 42 (8) (2000) a 81. M Andreis, J L Koenig Adv. Polymer Sci. 89 71 (1989) 82. V V Marchenkov, A K Khitrin Khim. Fiz. 3 1399 (1984) b 83. G E Karnaukh, A A Lundin, B N Provotorov, K T Summanen Zh. Eksp. Teor. Fiz. 91 2229 (1987) 84. E I Ivanova, B N Provotorov Zh. Eksp. Teor. Fiz. 107 473 (1995) 85. T P Kulagina, V V Marchenkov, B N Provotorov Vysokomol.Soedin., Ser. B 30 23 (1988) a 86. T P Kulagina, V V Marchenkov, B N Provotorov Vysokomol. Soedin., Ser. A 31 381 (1989) a 87. T P Kulagina, V M Litvinov, K T Summanen J. Polym. Sci., Part B, Polym. Phys. 31 241 (1993) 88. G I Sandakov, L P Smirnov, A I Sosikov, K T Summanen, N N Volkova J. Polym. Sci., Part B, Polym. Phys. 32 1585 (1994) 89. T P Kulagina, Doctoral Thesis in Physicomathematical Sciences, Institute of Chemical Physics in Chernogolovka, Russian Academy of Sciences, Chernogolovka, 1995 90. G I Sandakov, V P Tarasov, N N Volkova, Yu A Ol'khov, L P Smirnov, L N Erofeev, A K Khitrin Vysokomol. Soedin., Ser. B 31 821 (1989) a 91. A K Zhitinkina, N A Shibanova, O G Tarakanov Usp. Khim. 54 1866 (1985) [Russ. Chem. Rev. 54 1104 (1985)] 92.NNVolkova,GI Sandakov,AI Sosikov,YuAOl'khov,LP Smirnov, K T Summanen Vysokomol. Soedin., Ser. A 34 (3) 77 (1992) a 93. G I Sandakov, L P Smirnov, K T Summanen, N N Volkova Progr. Colloid Polym. Sci. 90 235 (1992) 94. L N Erofeev, G I Sandakov, L P Smirnov, A I Sosikov, K T Summanen, N N Volkova, in Proceedings of International Rubber Conference `IRC'94', Moscow, 1994 Vol. 2, p. 628725 Solution-free methods for the determination of the molecular mass distribution of polymers 95. V I Irzhak, B A Rozenberg, N S Enikolopyan Setchatye Polimery (Network Polymers) (Moscow: Nauka, 1979) 96. K Kremer, G S Grest, in Monte Carlo and Molecular Dynamics Simulation in Polymer Science (Ed. K Binder) (New York, Oxford: University Press, 1995) p. 194 97. L Mullins J. Polym. Sci. 19 225 (1956) 98. C C Moore, W F Watson J. Polym. Sci. 19 237 (1956) 99. E R DuÈ ring, K Kremer, G S Grest Phys. Rev. Lett. 67 3531 (1991) 100. G S Grest, K Kremer, E R DuÈ ring Europhys. Lett. 19 195 (1992) 101. E R DuÈ ring, K Kremer, G S Grest Macromolecules 26 3241 (1993) 102. E R DuÈ ring, K Kremer, G S Grest J. Chem. Phys. 101 8169 (1994) 103. I I Zakirov, V M Lantsov, V I Irzhak, B A Rozenberg Vysokomol. Soedin., Ser. A 30 915 (1988) a 104. V I Irzhak, V M Lantsov, B A Rozenberg, in Crosslinked Epoxies. (Eds B Sedlacek, J Kahovec) (Berlin, New York: Walter de Gruyter, 1987) 105. VAKargin,GL Slonimskii Dokl. Akad. Nauk SSSR 62 239 (1948) c 106. B Ya Teitel'baum Termomekhanicheskii Analiz Polimerov (Thermomechanical Analysis of Polymers) (Moscow: Nauka, 1979) 107. Yu I Matveev, A A Askadskii Vysokomol. Soedin., Ser. B 35 63 (1993) a 108. Russ. P. 1 763 952; Byull. Izobret. (35) 173 (1992) 109. Russ. P. 2 023 255; Byull. Izobret. (21) 142 (1994) 110. Yu A Ol'khov, S M Baturin, V I Irzhak Vysokomol. Soedin., Ser. A 38 849 (1996) a 111. V I Irzhak, S E Varyukhin, T F Irzhak Prog. Colloid Polym. Sci. 102 4 (1996) 112. T F Irzhak, S E Varyukhin, Yu A Ol'khov, S M Baturin, V I Irzhak Vysokomol. Soedin., Ser. A 39 671 (1997) a 113. Yu A Ol'khov, V I Irzhak Vysokomol. Soedin., Ser. B 40 1706 (1998) a 114. V I Irzhak, G V Korolev, M E Solov'ev Usp. Khim. 66 179 (1997) [Russ. Chem. Rev. 66 167 (1997)] 115. E E Al'yanova, N M Bravaya, T I Ponomareva, V I Tsvetkova, P M Nedorezova, V I Irzhak Vysokomol. Soedin., Ser. B 40 1691 (1998) a 116. J Honerkamp, J Weese Contemp. Mech. Thermodyn. 2 17 (1990) 117. J Honerkamp, J Weese Rheol. Acta 32 65 (1993) 118. J Weese Comput. Phys. Commun. 77 429 (1993) 119. V A Morozov Methods for Solving Incorrectly Posed Problems (Berlin: Springer, 1984) 120. M Baumgaertel, H H Winter Rheol. Acta 28 511 (1989) 121. A V Tobolsky, J J Aklonis, G Akovali J. Chem. Phys. 42 723 (1965) a�Polym. Sci. (Engl. Transl.) b�Russ. J. Chem. Phys. (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC