摘要:

Russian Chemical Reviews 71 (3) 175 ± 194 (2002) Non-carbon nanotubes: synthesis and simulation A L Ivanovskii Contents I. Introduction II. Methods for the synthesis of non-carbon nanotubes III. Simulation of the structures and prediction of the properties of nanotubulenes based on boron, silicon, germanium and phosphorus IV. Theoretical models of nanotubes based on p- and d-element compounds V. Models of hybrid nanotubular structures VI. Conclusion Abstract. carbon, of form allotropic new a of discovery The The discovery of a new allotropic form of carbon, extended tubular quasi-unidimensional nanometre-sized extended nanometre-sized quasi-unidimensional tubular struc- struc- tures the for prospects broad as well as nanotubes), (carbon tures (carbon nanotubes), as well as broad prospects for the use use of in studies numerous initiated them on based nanomaterials of nanomaterials based on them initiated numerous studies in the the search in based structures nanotubular of, design and for, search for, and design of, nanotubular structures based in other other compounds.for methods main the and properties Some compounds. Some properties and the main methods for the the synthesis of non-carbon nanotubes are considered. Studies on synthesis of non-carbon nanotubes are considered. Studies on the simulation of the electronic structures of these unique objects the simulation of the electronic structures of these unique objects are theoretical and experimental of Results analysed. are analysed. Results of experimental and theoretical studies studies along these lines are discussed. The bibliography includes 328 along these lines are discussed.The bibliography includes 328 references. I. Introduction Intense studies on the nanotubular form of matter [tubulenes or nanotubes (NT)] have started following the discovery in 1991 of hollow cylindrical carbon-based structures, the lengths of which were several orders greater than their diameters, in cathodic condensates formed upon electrical arc discharge between graph- ite electrodes.1 The new quasi-unidimensional carbon cluster was called `fullerene tubulene' and its electronic spectrum (ES) was calculated.2 It was found that the electronic spectrum of the carbon tube studied was characteristic of metals. From the very beginning, considerable attention by experi- menters and theoreticians was given to nanotubes as representa- tives of a new quasi-unidimensional allotropic modification of carbon in the series of previously known modifications, viz., 3D (diamond)?2D (graphite)?1D (carbyne)?0D (fullerene). With time, nanotubes, which were exotic objects of unique experi- ments and theoretical calculations, became the subject of numer- ous physicochemical studies; their unusual properties became the basis for many daring technological solutions.At present,NTfind broad practical application; now they are commercial products and are subject to marketing studies. A L Ivanovskii Institute of Solid State Chemistry, Urals Branch of the Russian Academy of Sciences, ul.Pervomaiskaya 91, 620219 Ekaterinburg, Russian Federation. Fax (7-343) 274 44 95. Tel. (7-343) 274 53 31. E-mail: ivanovskii@ihim.uran.ru Received 3 January 2002 Uspekhi Khimii 71 (3) 203 ± 224 (2002); translated by S S Veselyi #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n03ABEH000709 175 177 182 184 187 189 Data on the structures and properties of carbon nanotubes, on the methods of their synthesis and on their uses can be found in many publications; they have been generalised in a number of monographs (see, e.g., Refs 3 ± 7) and in many reviews (see, e.g., Refs 8 ± 32). In a number of reviews recently published in Russian scientific journals,15, 16, 21, 24 ± 27 the methods of synthesis and study of carbon NT were discussed.15, 16 The main principles of the theoretical simulation of the geometrical and electronic structures of NT were described in a monograph 7 and in papers.21, 26 The methods for the synthesis of such NT were described in detail by Rakov.24 He also reported the most recent data on the chemistry and application of carbon NT considered as promising materials for modern technologies.Tarasov et al. 25 discussed the possibility of use of NT as `tanks' for the storage of hydrogen. Basic information about carbon NT can be found elsewhere. 33, 34 Shortly after the paper by Iijima 1 was published, scientists were challenged by a number of questions that are of major importance to the development of the physics and chemistry of NT and for the prospects of their practical application.1. What is the nature of NT and what is the mechanism according to which they are formed? Are there any physical or chemical restrictions limiting the scope of compounds which can be obtained in a nanotubular form? 2. What physicochemical properties should a compound possess to be a potential initial substance for the synthesis of NT? 3. What are the specific features of methods that can be used to obtain non-carbon NT? What is their value in materials science and technology? From the very start, the concept of non-carbon NT was developed due to combined efforts of experimenters (syntheses of NT and studies of their functional characteristics) and theoret- icians (simulations of new nanotube forms, prediction of their structures and properties).In 1992, the first non-carbon NT based on lamellar molybde- num and tungsten disulfides were synthesised.35 In 1994, the possibility of existence of tubulenes based on hexagonal boron nitride (BN-nanotubulenes) was predicted.36 ± 38 Their dielectric properties were presumed to remain stable upon changes in their geometrical characteristics. This prediction was very important for the development of nanoelectronics; it initiated numerous studies on the synthesis of such NT. A large number of non-carbon NT have been synthesised to date or their existence has been predicted. In addition to the NT mentioned above, others based on germanium silicide and tran-176 sition metal dichalcogenides, sulfides and chlorides have been synthesised.The molecular and electronic structures of various hypothetical NT containing boron, phosphorus, silicon, germa- nium and transition metal diborides were simulated; the mecha- nisms of their growth were discussed and the properties of certain nanotubular structures were studied. Boron ¡¾ nitrogen and dichalcogen-containing NT were the subject of special analysis.12, 29 Some data on the synthesis and properties of non-carbon NT can be found in monographs 5, 7, 8 and reviews 11, 14, 19, 21, 24 devoted to carbon tubulenes. The present review discusses the general state of studies on the synthesis and simulation of the molecular and electronic struc- tures of non-carbon NT which were developed most intensely in the last years.Along with the `purely non-carbon' NT listed above, we also consider the class of `transitional' nanotubular structures con- taining both carbon and other p-elements (boron, nitrogen and silicon). These NT possessing diverse properties can either be obtained together with carbon NT within the same synthetic procedure or synthesised using methods specifically developed for this purpose. Of these, chemical substitution is the most promising method in which carbon NT are used as the starting material. The overwhelming majority of non-carbon NT have been obtained (or their synthesis is anticipated) on the basis of compounds that, like carbon, have lamellar (quasi-two-dimen- sional) crystal structures.Therefore, classifications and descrip- tions of atomic structures of non-carbon NT are performed, and their geometrical models are built,7, 12, 21, 29 on the basis of the concepts and methods developed for carbon NT.3 ¡¾ 10, 15 ¡¾ 32 Let us summarise briefly those to be employed below. Ideal carbon NT have the form of cylinders whose walls are formed by hexagons, with carbon atoms in the vertices. The process of formation of a nanotubulene can be visualised as the rolling of a `band' of carbon atoms cut out of a graphite monolayer (graphene network). A graphite layer can be `cut' in different ways: along lines which are perpendicular to the C7C bonds or pass through these bonds, or else at a definite angle, namely the so-called chiral angle y (Fig.1 a). As the resulting `bands' roll and their edges are linked together, two mainNTtypes appear, viz., non-chiral and chiral. The so-called zigzag and dented non-chiral NT are distinguishable. The structures of the main NT types are shown in Fig. 1 b. Using the base vectors a1 and a2 of the graphene network, the following vector can be defined: ch=na1+ma2 . The main geometrical parameters of a tubulene, i.e., the diameter (D) and the chiral angle (y ), are related unambiguously to the base vectors of the graphene network through the following relationships involving the indices n and m: p D=j ch j a a 3On2 a m2 a mnU , rAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA p, y=arctan ¢§2n a m 3AAAAAA m p A where a indicates the interatomic distances in the planar network.In terms of these indices, the corresponding NT are denoted as (n,m)-NT. All (n,m)-NT with 0<y <308 are regarded as chiral (spiral). Non-chiral NT have y=0 and 30 8. At y=0, a family of zigzag-shaped (n,0)-NT is formed; these were named according to the type of tube `cross-section' (see Fig. 1). The chiral angle for the dented (n,n)-NT equals 30 8. The NT ends may be open or closed by `caps' formed by fullerene hemispheres. Other (e.g., conical) `cap' types are also possible. Single-layer and multilayer NT are known. The latter can consist of a system of concentric tubulenes or have a scroll- type structure. Nanotubes can be single or form diverse aggregates A L Ivanovskii a a1 a2 y b 3 2 1 Figure 1.Model for theNT formation upon graphene layer rolling into a cylinder (a) and nanotube types (b). NT: (1) non-chiral dented (n,n); (2) chiral (n,m); (3) nonchiral zigzag-like (n,0). (braids, splices, etc.) that may involve other nanospecies (full- erenes). Many possible aggregates formed by carbon NT are described in a review by Rakov.24 The majority of NT contain various defects (topological defects, rehybridisation defects, `bro- ken' bonds). Directed introduction of defects into NT makes it possible to adjust flexibly their functional properties. More detailed information about NT types, their geometry, and their properties mentioned above can be found in a number of papers.3¡¾32 The material of this review is organised according to the following scheme.First, modern methods for the synthesis of non-carbon NT are considered. Further, we describe methods for the preparation of nanotubulenes from p-element compounds (BN, BCx, CNx, BxCyNz), which are most structurally similar to carbon NT, and other classes of chemical compounds (transition metal dichalcogenides and oxides). Finally, we characterise meth- ods (template synthesis and film `folding') which are rather versatile and can be suitable for the synthesis of NT based on various melts and inorganic compounds. After that, we discuss the results of theoretical simulation of the molecular and electronic structures of non-carbon NT by the methods of quantum chemistry and molecular dynamics.Data on the structural features and possible methods for the synthesis of hypothetical NT based on boron, silicon, phosphorus, CNx , GaN, transition metal dichalcogenides and diborides and `hybrid' nanotubular structures are considered. Special attention is paid to predictions of NT stability and properties. Data on the electronic properties of a number of NT formed in the B7C7N system are briefly discussed.7, 21 Data about the physicochemical properties of non-carbon NT are far from being exhaustive yet, and their analysis is beyond the scope of this review. Some of the most interesting parameters of certain groups of non-carbon NT are briefly mentioned in the corresponding sections.Non-carbon nanotubes: synthesis and simulation II.Methods for the synthesis of non-carbon nanotubes 1. Synthesis of nanotubes in the B7C7N system The structures of nanotubes formed in the B7C7N system are similar to those of carbon tubulenes. The review of the methods for the synthesis of such non-carbon NT is given using the sequence in which the methods for the synthesis of carbon NT were considered by Rakov.24 This will allow us to compare the potential of `traditional' (for carbon NT) synthetic methods to that of new methods suggested for the synthesis of non-carbon NT. a. Electric arc synthesis Electric arc synthesis is the simplest method widely used for the synthesis of carbon NT (see Refs 9, 11, 15, 16, 24, 32). Numerous attempts to use this method for the synthesis of nanotubulenes in the B7C7N system have been reported.Of course, the standard method 11, 15, 24 should be modified so as to ensure the introduc- tion of not only carbon but also boron and nitrogen into the reaction zone. As a rule, this is done using an appropriate composition of air (e.g., the synthesis is performed in a flow of nitrogen) and/or of the electrode. Since boron nitrides are insulators,7, 36 ± 38 in pure form they are unsuitable for the preparation of electrodes. BN-Nanotubu- lenes were first obtained in an electric arc discharge (in a helium atmosphere) between a copper cathode and an anode made of boron nitride enclosed in a tungsten shell.39 The cathodic deposit was found to contain a wide variety of boron ± nitrogen nano- structures with diverse morphology, including BN-nanotubu- lenes.The tubes consisted of a number of layers (the inner and outer diameters varied from 1 to 3 and from 6 to 8 nm, respec- tively, and the distance between the walls was *0.33 nm); their lengths exceeded 200 nm. The ends of NT were `capped' with tungsten-containing nanoparticles (which presumably consisted of tungsten borides or nitrides). The growth mechanism of BN- nanotubulenes was assumed39 to involve catalysis by metal atoms. A number of boron-containing phases, e.g., ZrB2 and HfB2 , have been studied as electrodes for the electric arc synthesis of NT in a flow of nitrogen.40 ± 43 In the latter case, the plasma contained metal atoms, which could serve as growth catalysts for BN- nanotubulenes.Cumings 44 reported a high-performance method for the syn- thesis of NT based on BN; the electrodes consisted of elementary boron (99 at.%) and admixtures of cobalt or nickel (up to 1 at.%), which served as catalysts. The pressure of nitrogen in the chamber was kept at 380 Torr; the current was 60 A. The residue on the cathode arranged on the bottom of the chamber was grey; a thin layer of the product was also found on the chamber walls. Analysis of the structure of the product obtained showed that it contained only double-layer NT (with diameters of the outer and inner BN-cylinders equal to 2.9 and 2.2 nm, respectively, and with a distance of *0.37 nm between their walls). These tubes formed bunches (bundles).No metal admix- tures were found in the NT; the stoichiometric composition of the tubes matched the ratio B :N=1 : 1. It should be noted that double-layer NT were also obtained from BN using HfB2 cath- odes.40 The growth mechanism and kinetics of such double-layer NT have not been ascertained. Using scanning electron microscopy (SEM), it was found that NT bundles are often joined to form `knots' that have their own structures. These junctions consist of boron nanocrystals coated by multilayer shells, i.e., the specific cocoons. Nano-cocoon layers are formed by graphite-like boron nitride. Cumings 44 suggested a method for removal of boron nanocrystallites from nanococoons; as a result, the latter become empty and resemble the well-known carbon `onion-bulb' structures, the so-called onions.3±6 It was suggested that hollow BN-nanococoons should be considered as a new nanomaterial for chemical and electrochemical industries.It 177 is possible to vary the properties of such material deliberately by filling the cocoons with definite atoms or molecules. Vaporisation of porous BC4N was accompanied by simulta- neous formation of carbon, BN- and mixed BNC-nanotubu- lenes;45 however, homogeneous boron carbonitride NT could not be obtained, probably, because of differences between the vapour pressures of gaseous boron nitrides and carbon. b. Laser synthesis In addition to such methods to obtain carbon NT as vaporisation of the target (graphite) under electron and ion beams, sunlight, Joule heat (resistance vaporisation), yet another method became popular,24 viz., ablation of targets under laser irradiation.Attempts to use laser irradiation for the synthesis of nanotubu- lenes containing B, C and N have been reported.46, 47 Hexagonal BN served as the target and high pressure of nitrogen was maintained in the chamber. The products contained BN-nano- tubulenes and particles of cubical boron nitrides. A CO2 laser was used to synthesise single-layer BN-nanotubulenes on a multigram scale.48 It was found that the zigzag-shaped tubes form bunches; furthermore, small amounts of double-layer BN-nanotubulenes and fullerene-like clusters of boron nitride were detected. Nanotubes (and nanoparticles) containing B, C and N atoms were also obtained by other methods of target irradiation.For example, electron microscopy allowed Golberg et al.49 to observe in situ the formation of single- and multi-shell fullerene-type nanoparticles of boron carbide upon exposure of lamellar BC3 to an electron beam; it cannot be ruled out that a densely packed structure, viz., diamond doped with boron, was formed inside these particles. Proton ± ion bombardment of boron pyronitride resulted in a microcomposite material which incorporated blocks of BN-nanotubulenes and possessed enhanced plasticity.50 c. Pyrolytic synthesis Catalytic pyrolysis of hydrocarbons in the presence of metal catalysts is becoming one of the most widespread and promising methods for the synthesis of carbon NT.24 Pyrolysis of aCH3CN± BCl3 system in the presence of a cobalt powder at*1270K was first used 51 to obtain nanotubulenes and nanofibres with diverse morphologies and compositions.More detailed studies 52 of the reaction D BC2N+3 HCl CH3CN± BCl3 showed that its products contained NT whose composition was close to BC2N. However, the composition homogeneity was violated considerably in the NT walls and spontaneous separation occurred (presumably, during the NT growth) to give C/BN `islets'. The most stable NT contained BN layers separated by graphite-like carbon layers.52 Sen et al. 53 found systems suitable for the synthesis of BCN- and CN-nanotubes using the pyrolytic method. These were obtained using the 1 : 1 (CH3)3N±BH3 adduct and pyridine, respectively.The reactions were carried out at *1270K with cobalt as a catalyst. The products contained nanotubes with diverse chemical compositions and morphologies. Pyrolysis of a mixture of acetylene with diborane (in an atmosphere of helium and hydrogen) 54 gave boron-containing carbon NT whose compositions varied from C35B to C50B. The boron content in such tubes was close to the nitrogen content in CN-nanotubulenes.53 Nitrogen-containing carbon NT were also found in products of catalytic pyrolysis of triazine.55Maet al. 56, 57 used B4N3O2H as the precursor for the synthesis of BN-nanotubulenes. d. Substitution reactions Han et al. 58 ± 60 described a new chemical method for the synthesis of NT based on replacement of carbon atoms by nitrogen and/or boron atoms in carbon NT.Carbon tubes obtained by one of the known technologies were used as the starting materials. This178 method was employed to obtain a wide variety of boron ± carbon (BxC17x), boron ± carbon ± nitrogen (BxC17x7yNy) and boron ± nitrogen (BN) multi- and single-layer NT, as well as their bundles.58 ± 66 For example, Han et al. 60 synthesised (BxC17x7yNy)-nano- tubes using a mixture of B2O3 with a product of arc synthesis which contained up to 70% carbon NT, together with fullerene- like nanoclusters and amorphous carbon with an admixture of particles of the metal catalysts (Fe/Ni). The reactions were carried out in a flow of nitrogen in the temperature range 1523 ± 1803 K.The products were analysed using transmission electron micro- scopy and electron energy loss spectroscopy (EELS) (Fig. 2). aC p* I (rel.u.) 16 12840 c B 4 p* 2 N 0 250 150 350 E /eV Figure 2. Electron energy loss spectra for the original carbon NT (a), NT with the composition BxC17x (x&0.1) (b), `stoichiometric' BN-nano- tubulenes (c), nanotubes BxC17x7yNy (x=y*0.1) (d ) obtained by chemical substitution.60 The characteristic peaks in the spectra correspond to the presence of the respective elements, viz., B, C and N. Heating at 1523K for 30 min resulted exclusively in boron ± carbon NT (with the ratio B :C&0.1) joined as bundles. A mixture of single-layer (BxC17x)-, (BxC17x7yNy)- and BN- nanotubulenes was obtained at higher temperatures (*1623 K).On subsequent increase in temperature (to 1803 K), the original carbon NT was destroyed completely; the resulting major fraction was a mixture of hexagonal, rhombohedral and turbostatic boron nitride with insignificant admixtures of nano-sized BN-structures: nanocones, polyhedral nanoclusters and nanofibres. No forma- tion of BN-nanotubulenes was detected. The morphology of the BCN-nanotubulenes obtained resembled that of the original carbon NT in many respects. (BxC17x7yNy)-Nanotubulenes were also obtained 67 as reaction products of carbon nanotubu- lenes and CNx-nanotubulenes with B2O3 and ammonia. The reactions leading to BC- and BN-nanotubulenes can be described 58, 59 by the following equations: x B2O3+(2+3x)C (nanotubes) 2BxC (nanotubes)+3x CO, B2O3+3C (nanotubes)+N2 2 BN (nanotubes)+3 CO.The maximum boron content in (BxC17x)-nanotubulenes did not exceed *10 at.%. The bundles in (BxC17x7yNy)-nanotubu- lenes differed in the C: (B+N) ratios (Fig. 3). In certain cases, the `stoichiometry' of N: B&1.0 was reached. This is due to the selective character of substitution:60 the probability that `pure' BN-nanotubulenes (or, at least, boron ± nitrogen clusters in the b I (rel.u.) C p*B 840 d B 4 C N 20 250 150 350 E /eV a Number of NT 20 100 b 20 100 3 2 1 Figure 3. Composition histogram of multilayer BCN-nanotubulenes synthesised by chemical substitution from carbon NT (at 1773 K in a flow of nitrogen);63 (a) without oxides, (b) in the presence of MoO2 .walls of the original carbon NT) are formed is much higher than the probability of NT formation if the replacement of carbon atoms by boron and nitrogen is purely statistical. An interesting result was obtained by Han et al.60 Inside single-layer BN-nanotubulenes, endohedral polyhedral BN-nano- clusters, described as four-shell octahedral BN-fullerenes with the composition B12N12@B76N76@B208N208@B412N412 (resembling carbon onion structures), were detected.68 Similar hybrid nano- tubular structures consisting of fullerenes located inside carbon NT (Cn@C) and carbon NT incorporating endofullerene chains (M@Cn@C) were also reported.69, 70 Attractive features of the method suggested by Han et al.58 ± 61 are its ease and efficiency. The potential of the method is only limited by the amount of the original material available, viz., carbon NT. Yet another important factor is the fixed size of the NT obtained in the system B7C7N, which is determined by the parameters of the matrices, i.e., the carbon NT. The obvious drawback of the method is that it is difficult to synthesise NT with homogeneous chemical compositions. This method was further developed 63, 65, 66 and used for the synthesis of bundles of multilayer BN-nanotubulenes. A mixture of carbon NT and boric anhydride was complemented withMoO3 (see Ref. 63) or a mixture of MoO3 with PbO (see Ref. 65) or MoO3 with V2O5 ;66 the process was conducted in a flow of nitrogen at 1503 ± 1773 K.Analysis of the reaction product involving molybdenum trioxide by transmission electron micro- scopy (TEM) showed that it contained bundles of multilayer BN- nanotubulenes, the major fraction of which (up to *82%) had non-chiral zigzag geometry.66 The ends of such NT were open in contrast to the NT obtained 58 ± 60 in the absence of metal oxides. The following scenario of the growth of BN-nanotubulenes was suggested,66 taking into consideration that transition metal oxides are efficient oxidants for carbon. In the first stage, the terminal structures (`caps'), which close the carbon NT, are removed. 2 B+3 CO, 3C+B2O3 Mo+3 CO, 3C+MoO3 2V+5 CO. 5C+V2O5 As a result, the gas phase can access the intra- and interlayer space of multilayer carbon NT.Further, carbon is simultaneously replaced outside and inside the carbon NT (Fig. 4). These A L Ivanovskii C: (B+N) 4Non-carbon nanotubes: synthesis and simulation a (B2O3;MoO3; V2O5)+N2 C (B2O3;MoO3; V2O5)+N2 C b c B2O3; N2 C C CO C C C C C C C CO C C B2O3; N2 B N N B N B N B N B N B Figure 4. Scheme of the reaction of oxides andN2 with multilayer carbon NT (a) in the reaction with its `open' end (b) and its side surface (c) during chemical substitution in the synthesis of BN-nanotubulenes.66 processes mostly involve the diffusion mechanism;71 as a result, growth of BN-nanotubulenes occurs. High-resolution TEM was used to study the structure of BN- nanotubulene bundles obtained in the presence of PbO as the activator.72 All NT had zigzag geometry, their diameters ranged from 20 to 30 nm and their axes were arranged strictly parallel.It was suggested to consider this product as a new nanomaterial. It is characterised by invariable dielectric parameters and low chemical reactivity (in comparison with carbon NT). High-resolution TEM analysis of the structures of multilayer BN-nanotubulenes showed that `packs' of adjacent layers with hexagonal and rhombohedral types coexist, the wall structure changes from dented to zigzag-shaped and noticeable polygonisa- tion of NT cross-sections takes place.73 The substitution method was also used to obtain `nanocables', i.e., BN-nanotubulenes filled with an alloy of Ni (40%) and Fe (60%).74 In the first stage, carbon NT with nanoparticles of the alloy at the ends were obtained. These NT were then placed in a mixture of gaseous B2O3 and N2 and kept for 30 min at the melting point of the alloy (1723 K).Two processes occurred simultaneously, viz., filling of the cavities with the iron ± nickel alloy (due to the capillary effect), and change in the composition of the tube (`shell') due to chemical substitution. It should be noted that the first SiC-nanotubulenes were obtained using the substitution reaction.75 Large-diameter carbon NT and gaseous SiO were used as the starting materials. e. Miscellaneous methods A number of methods for the synthesis of BCN-nanotubulenes based on thermal reactions have been suggested.Carbothermal reduction of amorphous boron oxide with simultaneous nitridation (at 1373 ± 1723 K) 76 gave a product containing cylindrical BN-nanotubulenes with various configu- rations (including bent or kinked ones, T-shaped, `bamboo- shaped') and nanostructures in the form of concentric truncated cones. The overall reaction can be described by the following equation: B2O3(amorph.)+3B4C+7N2=14BN+3CO. 179 Cylindrical (closed) multilayer NT formed bundles; their junctions contained multi-shell onion-type BN-particles which served as growth centres. The tube diameters ranged from 10 to 500 nm; the length-to-diameter ratios ranged from 3 to 30. Nanotubes grew both in the gas and in the solid phase. The growth of bamboo-shaped NT involved alternation of cylindrical and onion-type fragments: NT± onion ±NT ± onion.It was assumed that certain structures incorporated B3N4 heptagons, which ensured contacts at the junctions of cylindrical fragments. The obvious drawback of this synthetic method is that the forms of the nanotubulenes obtained are too diverse and that no method to control the formation of NT with required param- eters is available. Nanotubes and nanocones based onBNwere also obtained by thermal treatment of rhombohedral b-boron at 1470 K in the presence of graphite-like boron nitride and lithium vapour.77 Growth of zigzag-shaped NT was observed at the edges of graph- ite-like BN particles. BN-Nanofibres were found to appear upon annealing (*2000 ± 2370 K) of BN powder in the presence of iron.78 2. Dichalcogenide nanotubulenes Transition metals dichalcogenides MX2 (X=S, Se, Te) were the first inorganic compounds to attract the attention of researchers as compounds that are potentially suitable for the synthesis of tubular nanostructures.These compounds have a distinct quasi- two-dimensional structure. For example, molybdenum and tung- sten disulfides have lamellar structures (of the MoS2 type, space group P63/mmc) consisting of packs of S ± (Mo,W) ± S layers; the metal atoms are in a trigonal-prismatic environment 79 and the layers are packed due to van der Waals forces. Numerous NT based on tungsten, molybdenum and niobium dichalcogenides have been synthesised to date, mostly by chemical methods.The formation of fullerene-like species and fragments of cylindrical nanostructures consisting of WS2 and MoS2 was first observed upon thermal treatment of tungsten or molybdenum films on quartz substrates in a flow of H2S/H2 .35, 80 Remskar et al.81, 82 used chemical transport reactions to obtain MoS2 andWS2 cylindrical nanocrystals and tubular structures (MoS2 microtubes several millimetres long with wall thickness smaller than 0.1 mm). Galvan et al.83 ± 85 proposed a low-temperature (<100 8C) acti- vation method to obtain WS2-nanotubulenes involving treatment of condensed tungsten disulfide with concentrated nitric acid. A new material consisting of WS2-based NT was obtained using WOx nanoparticles.85 In the synthesis of fullerene-like molybdenum disulfide par- ticles, a gas phase reaction of MoO37x with H2S at 1073 ± 1223 K gave MoS2-nanotubulenes as side products.86, 87 The tubes con- sisted of 5 ± 10 MoS2 layers on average, were several micrometres long and were mixed with MoS2 (2H-polytype).The methods for the preparation of dichalcogenide nanotubes (and nanofibres) and for the synthesis of fullerene-like polyhedral particles of these compounds are being developed in parallel.88 Both NT and polyhedral hollow or filled nanoparticles of d-metal dichalcogenides are often obtained in a single experiment. Molybdenum disulfide particles with 3 ± 5 nm diameters were synthesised 89 by laser ablation of MoS2 targets.Laser-induced vaporisation of tungsten and molybdenum disulfides in argon at 720 ± 1320 K gave 90 various hollow and metal-filled multilayer nanoparticles. For example, the major fraction obtained upon vaporisation of WS2 at 1320 K consisted of nanoparticles with 10 ± 15 nm diameters containing 4 ± 8 concentric layers of WS2 . The bulk of filled nanoparticles was found to contain a stabilised b-W phase, which is unstable under ordinary conditions, as the filler metal. In addition, the products contained tubular nano- structures.180 Certain properties of dichalcogenide NT were summarised in a review by Zhu et al.88 Hollow WS2-onions are promising as solid lubricants;89, 91 their tribological properties can be optimised by filling the cavities with tungsten carbide.The synthesis of such `composite' particles is based on the pyrolysis of WC nano- particles in hydrogen sulfide.92 The properties of MS2-nanotubulenes have not been studied in sufficient detail. It was found that disulfide nanotubes are semiconductors; they possess low chemical reactivities.93 Nor- mally, multilayerNTcontain numerous defects in the outer layers, whereas the inner walls generally have nearly ideal structures, which is reflected in the conducting properties of NT.94 ± 96 The coalescence and the growth mechanism of WS2-nanotubulenes to give `braids' have been studied;97 a method for changing the conducting properties of theseNTunder the effect of the substrate was suggested.98 The application of WS2-nanotubulenes in scan- ning microscopy was discussed;99 several versions of the growth mechanism of these NT were suggested 100 and the adsorption of methane on their surfaces was studied.101 In addition to the syntheses of nanotubulenes based on transition metal disulfides, attempts were undertaken to obtain nanostructures of other dichalcogenides as well.For example, bombardment of niobium diselenide with electrons resulted in NbSe2-nanotubulenes, a few nanometres long, closed on one end.101, 102 By varying the accelerating voltage and the current density, NbSe2-onions were obtained as well. A similar technique was used successfully to synthesise MoTe2- (Refs 103 ± 106) and WSe2-nanotubulenes.102 Nath and Rao 107 considered a method to obtain NT by reduction of tungsten and niobium triselenides or by decomposition of ammonium selenometallates in an atmos- phere of hydrogen.Recently, methods for the preparation of CuInS2-nanotubulenes 108 and Bi3Se4-nanorods 109 were pro- posed. 3. Oxide nanotubes Transition metal oxides are widely used to construct diverse nano- structured functional materials in the form of films, nanorods, and mesoporous materials.110 ± 122 Some of these nano-objects are obtained with the use of carbon NT. For example, filling the internal cavities of carbon NT with V2O5 , MoO3 , PbO, or Bi2O3 particles followed by chemical removal of the carbon shell gave oxide nanofibres and nanotubular ceramics.123 ± 125 Nanotubulenes of transition metal (vanadium and titanium) oxides have not been synthesised until recently.126 ± 133 The sol ± gel method was found to be efficient for this synthesis 126 ± 129 (for more detail about colloid methods to obtain nanostructures, see Refs 134 and 135).Krumeich et al.126 used the sol ± gel method to obtain multi- layer vanadium oxide NT (VOx-nanotubulenes). Ethanolic solutions of vanadium(V) triisopropoxide and an amine CnH2n+1NH2 (44n422) or a,o-diamine H2N[CH2]nNH2 (144n420) in the molar ratio 2 : 1 were stirred for 1 h in an inert atmosphere and hydrolysed. Ageing of the system for 12 ± 96 h resulted in an orange composite, which gave a black product upon hydrothermal treatment for 2 ± 7 days (at 450 K) in an autoclave. The composition, structure and certain properties of the product obtained were examined by X-ray diffraction, XPS and TEM and by magnetic susceptibility measurements. It was found that the major components of the composites were nanostructures with tubular morphologies, whose outer and inner diameters were 5 ± 50 and 15 ± 150 nm, respectively, and whose lengths were up to 15 mm.Special experiments to study the cross-sections of the resulting VOx-nanotubulenes showed them to be multilayered.127, 128 Both spiral (scroll-shaped)NTand those in the form of a set of concentric cylinders with organic molecules intercalated between the cylinders were detected. Spiral NT are formed most commonly. Mixed structures were also observed. Figure 5 shows the cross-section types of multilayer NT.The figure shows the defects of NT packs. Certain NT contain regions with parallel arrangement ofVOx-layers, both inside the tubes and 4 1 5 23 67 Figure 5. Models (1 ± 3) and different cross-section structures (4 ± 11) of multilayer NT;127 (1) concentric cylinders (`Russian matreshka'), (2) scroll, (3) `papier- mache'; (4 ± 11) typical cross-sections of vanadium-oxide NT. in the vicinity of junctions between the adjacent tubes. The number of NT layers varies from 2 to 30; they are built of tetrahedra (VO4) and of square pyramids (VO5).126 These poly- hedra also serve as structural elements of the crystalline phases a-V1.08P0.92O5 , (NH4)2V3O8, K2V3O8 and BaV7O16 . nH2O. The distance between the adjacent NT layers (1.6 ± 3.8 nm, Table 1) changes proportionally to the length of the alkylamine molecule serving as the intercalant.The overall molecular formula of NT can be written asVOxOy/2[CnH2n+4N]y . Attempts to obtain `pure' VOx-nanotubulenes by removal of organic molecules upon ther- mal treatment are probably hopeless, as the tubes undergo complete degradation at temperatures >520 K. It is more likely that amines can be replaced by cations of alkali or alkaline-earth metals (e.g., Na+, K+, Mg2+, Ca2+, Sr2+). This method may become efficient for modifying the properties of VOx-nanotubu- lenes. Replacement of monoamines by diamines in NT results in a noticeable decrease in the interlayer distances (from 1.6 ± 3.8 to 0.9 ± 1.0 A).Specimens of the composite manifest semiconducting properties. Table 1. Chemical composition, distances between the VOx-layers, mag- netic moments of vanadium atoms (MM) and relative concentrations of V4+ ions in VOxOy/2[CnH2n+4N]y-nanotubulenes.126 Composition VO2.40[C4H12N]0.25 VO2.30[C6H16N]0.27 VO2.47[C10H24N]0.27 VO2.40[C11H26N]0.27 VO2.40[C12H28N]0.26 VO2.42[C14H32N]0.27 VO2.45[C16H36N]0.26 VO2.37[C18H40N]0.26 VO2.43[C22H47N]0.30 45 66 33 48 45 42 35 52 46 a Data obtained by X-ray (XRD) and electron (ED) diffraction. A L Ivanovskii 8910 11 [V4+] (%) Interlayer distance MM/mB /nm EDa XRDa 1.16 1.41 0.99 1.20 1.17 1.13 1.03 1.25 1.18 1.58 1.62 2.17 2.16 2.21 2.43 2.63 2.86 3.16 1.66 1.96 2.50 2.59 2.77 3.02 3.20 3.48 3.80Non-carbon nanotubes: synthesis and simulation Two new methods for the preparation of VOx-nanotubulenes from vanadium oxotrichloride and pentoxide as the starting compounds have recently been proposed.130 Both processes are carried out in two stages.In the first stage, a solution of VOCl3 (or V2O5) with the amine CnH2n+1NH2 (114n420) is hydrolysed to give a gel; in the second stage, hydrothermal treatment is conducted to yield NT. If V2O5 is used, amines are directly intercalated between the pentoxide layers. The resulting NT can be described by the general formula VO2.420.03. .(CnH2n+4N)0.270.01 (114n420). If VOCl3 is used, the general formula of the NT is VO2.450.05(CnH2n+4N)0.270.01 (114n416).The NT morphology is similar to that described above. The low cost of the product obtained is worth a special note.130 Pillai et al.131 succeeded in synthesising a unique type of multilayer vanadium oxide NT with periodically alternating distances between the VOx-layers by changing pH of the medium. Ammonia was added to the reaction mixture obtained upon hydrolysis of an ethanolic solution of vanadium(V) triisoprop- oxide to pH 10. The final product was obtained in accordance with the procedure described above (see Ref. 126). It was found that the outer diameters of the resulting NT (>250 nm) were much larger than those of the NT obtained without ammonia (at pH 4 ± 8); however, the lengths of the former (2 ± 5 mm) were much smaller than those of the latter.The walls of the NT obtained in the presence of NH3 are composed of regularly alternating VOx layers; between these, amine molecules and NHá4 are intercalated alternately (the VOx±VOx interlayer dis- tances are 2.0 and 0.9 nm, respectively). The NT structures contain various defects, for example, violation of periodic alter- nations of the interlayer distances; in certain cases, breakdown of layers (both inner and outer ones) occurred. The fact that NT contain defects was interpreted 131 as an indication of their high stability. The overall chemical formula of the new NT could be expressed as [(NH4)yVOx] [(CnNH3)zVOx]; the defects are present in the [(CnNH3)zVOx] layers.It is possible to control the structure of vanadium-oxide NT by changing the pH of the gel. It is assumed 131 that `pure' [(NH4)yVOx]-nanotubulenes with small interlayer distances can be synthesised at pH *12. On the other hand, it is known 127 that it is ammonium vanadate [(NH4)2V3O8] that is formed under these conditions. In addition to particles with well-shaped tubular morpholo- gies, products with ribbon-like structures and intermediate forms could be revealed by SEM in these composites. Syntheses of TiO2-nanotubulenes from solution 132 and with the use of mesoporous Al2O3 (see Ref. 133) were reported. 4. NiCl2-Nanotubulenes Hacohen et al.136 were the first to synthesise halogen-containing nanotubular structures, namely, NiCl2-nanotubulenes.Nickel dichloride has lamellar structure (of the CdCl2 type, space group R3m); its specific feature is that it has ferro- and antiferromagnetic types of spin ordering inside the layers and between them, respectively. Multilayer NT were obtained; their cross-sections were up to *7 nm, while the lengths were several micrometres. The tubular structures remained stable for a few days. It is assumed 136 that the most interesting applications of such struc- tures may be related to their unusual magnetic characteristics. 5. Miscellaneous methods for the synthesis of non-carbon nanotubes a. Template synthesis The template method of synthesis is used to obtain nanotubular composites, i.e., multilayer NT containing layers with different chemical compositions.Stable NT (normally, carbon nanotubes) are used as the support, which is then coated with layers of various metals or compounds.124, 137 ± 141 The interest in nanotubular composites is due both to the possibility of preparation of tubular forms of those materials which are as yet unknown as isolated NT, and to specific proper- 181 ties of the composites themselves (e.g., they behave as `interfaces' between the adjacent cylinders of different chemical nature). It should be noted that the method for the application of coatings on carbon NT is similar to that for the chemical modification of their surfaces, in which the composition of the outer layers of multilayer carbon NT is carried out by doping (substitution).142 ± 146 Seeger et al.142 proposed a method for the application of SiOx - coatings on multilayer carbon NT at room temperature (in addition to the high-temperature method 138).The low-temper- ature (colloid) method is based on the formation of positively charged centres on the surface of multilayer carbon NT followed by deposition of negatively charged colloidal SiOx particles onto this surface. The high-temperature method 138 is based on the thermal decomposition (T>503 K) of tetraethoxysilane SiO2+2C2H5OH+2C2H4 , Si(OC2H5)4 adsorbed on multilayer carbon NT. The low-temperature SiOx coatings had the form of continu- ous amorphous layers 3 ± 10 nm thick with included nanocrystal- lites; the high-temperature process resulted in a non-uniform SiOx layer, *10 nm thick (complete covering of the support surface could not be reached).In both cases, no chemical interaction occurred between the contacting SiOx layers and the multilayer carbon NT; the oxide film coated the NT directly without formation of any intermediate (e.g., carbide) layers. The method used to obtain ZnS-nanotubulenes 147 may be regarded as a combination of the template and substitution approaches. Highly structured semiconducting zinc oxide films which incorporated single-crystal ZnO columns obtained by non- equilibrium electrodeposition on a conducting substrate (glassy SnO2) were used as the support.148 ± 150 The columns (several micrometres high, with diameters of 100 ± 300 nm) were oriented orthogonally to the substrate.The process of formation of ZnS-nanotubulenes is based on the ion-exchange reaction Ar, 673K ZnS+H2O. ZnO+H2S After exposure for 15 min, the thickness of the ZnS layer on the ZnO columns was *15 nm; complete transformation of ZnO- columns into ZnS-nanotubulenes did not occur. The untrans- formed column matrix (ZnO) was removed by treatment with dilute H2SO4 . The resulting hollow ZnS tubes (open or closed) with wall thickness of *10 nm maintained the surface morphol- ogy of the original crystallites; their cross-sections had polygonal shapes. It is expected that other types of substitution may result in a wide range of NT with diverse chemical compositions, e.g., tubular forms of semiconducting CuO, CuS, CdS, CdTe, ZnTe, etc.b. Film `rolling' As noted above, it is compounds which form phases with quasi-2D structures under equilibrium conditions that are usually consid- ered as potential candidates for the synthesis of NT. A fundamentally new approach to the problem of obtaining the nanotubular form of matter has been developed.151, 152 It was proposed to formNTfrom thin films. The principle of this method (which is subdivided 151 into the `general' and `special' methods) is demonstrated in Fig. 6. `General' method. A thin film is applied on a double-layer support the top layer of which can be removed by selective etching. Following the etching, an edge of the film is detached from the support, bent and placed atop the film. The tubular structure is formed in the bent region (see Fig.6). For instance, this method was used to obtain silicon ± germanium NT up to 12 mm long and with a diameter of *230 nm. The thickness of one layer was 16 nm and was determined by the conditions of the film deposi-A L Ivanovskii 182 b a 1 0 1 1 00 2 2 3 3 so-called strain energy is often calculated (Estr). This parameter is defined as the difference between the energy of the flat atomic layer and the corresponding NT. It is considered as an estimate of the energy required to transform a compound in crystalline (lamellar) modification to the tubular form. By comparing the Estr for the known (carbon) NT and hypothetical non-carbon NT, it is possible to make conclusions on the probability of accessing the latter.Compounds that have stable or metastable lamellar phases become the most common subjects of a theoretical search for non- carbon NT. In general, the methods used to build geometrical models of such NT resemble those for the carbon NT.7, 21, 26, 30 Their classification is also carried out by analogy with carbon NT (using n, m indices). 1. Boron-based nanotubulenes Figure 6. General (a) and special (b) film `rolling' methods for obtaining NT;151 (1) single-layer (double-layer) film that forms nanotubes; (2) a top support layer removed upon selective etching; (3) a bottom support layer. Arrows show the force directions when a double-layer film is rolled. A number of studies 156 ± 165 deal with prediction of the tubular structures of boron, which are electron-deficient in comparison with carbon NT.A number of techniques were used to simulate the boron NT formation process. The formation of B-nanotubulenes was simulated by the `cluster assembling' method. In the first calculation stage, the Hartree ± Fock andDFTmethods were used to determine the sizes and configurations of small (several atoms) isolated boron clus- ters; after that, their association was considered. For example, starting from a B7 hexagonal pyramid, a large number of topologically different Bn clusters were constructed: quasiplanar (quasi-2D), convex and three-dimensional. Planar structures were found to be most stable (see also Refs 166 and 167). Convex structures are less stable.The stability of these structures increases 156, 159 if they incorporate a combination of two base clusters, viz., hexagonal and pentagonal pyramids. tion. Yet another procedure of the synthesis of SiGe films 151 gave NT with up to 50 nm diameter and a wall thickness of*6 nm. `Special' method. Double-layer films made of different materi- als are used; the interatomic distances in the material of the outer film layer (1 0) should be larger than those in the inner film layer (1 00) (see Fig. 6). After etching and detaching a film edge from the substrate, each material tends to restore its original structure; owing to the resulting mechanical stress, the free film edge bends upwards, makes a full turn and forms a tube. In the case of prolonged etching, the film edge can make several turns to give specific rolls that resemble the familiar `scroll'-type carbon tubes.This method made it possible to obtain tubes up to 20 mm long and with a diameter of 530 nm from double-layer InAs/GaAs,152 InGaAs/GaAs 153 and SiGe/Si films 154 and from epitaxial SiGe films.155 Boustani et al. 164 and Gindulyte et al. 165 simulated boron- based NT. They evaluated the stability of the isomers of the tubular form of the B32 cluster, the structures of which are shown in Fig. 7. It was found that the stability of these isomers decreases with a decrease in the diameter. In contrast, as the diameter increases (the curvature of the cylinder walls decreases), their energies approach that of the most stable planar form of boron.a b It is important that the thickness ofNT`walls' obtained by this method, i.e., the number of layers in the multilayer NT, depends exclusively on the film technology used and can vary over a wide range, down to a monoatomic layer. The approach sug- gested 151, 152 does not impose fundamental restrictions on the chemical compositions of individual NT layers, which allows tubular hetero-structures to be formed from multilayer films. Yet another important advantage is that it is possible to set exactly the location of the NT on the support: it is defined by the position of the film edge and by the etching time. Thus, the main factors in the manufacturing of NT within this approach are the choice of the film deposition method (which defines the size and chemical composition of the resulting NT), the time of chemical etching and the elastic properties of the material of the NT-forming film.It is expected 151 that this approach will provide single-crystal NT of many semiconductors. c d III. Simulation of the structures and prediction of the properties of nanotubulenes based on boron, silicon, germanium and phosphorus Figure 7. Model structures with C2h (a), C2 (b), D4 (c, d ) symmetry types for cylindrical B32 clusters with different diameters.164 The search for new methods for the synthesis of non-carbon NT is performed in parallel with active studies on theoretical simulation of their configuration and electronic structure. Generally, the majority of studies employ zone or cluster models, as well as modern methods of density functional theory (DFT).Further- more, structural features and kinetics ofNTformation are studied by molecular dynamics methods. Details on specific approaches and computational schemes can be found in the original studies discussed below. In addition to the conventional energy and electronic charac- teristics of NT, such as energy bands, electronic state densities (SD), interatomic interaction parameters and charge states, theNon-carbon nanotubes: synthesis and simulation Acompetition between the following factors affecting the stability of the isomers was noted:164 first, the high deformation energy of the planar layer, which increases abruptly with an increase in the curvature and prevents the formation of tubes with small diame- ters; second, the energy gain due to the coupling of broken (side) bonds in planar layers when they are rolling into a cylinder.Thus, it is quite possible to synthesise boron NT, but their diameters should be relatively large. Examples of such extended boron NT built like carbon NT were considered by Gindulyte et al. 165 The formation of NT in the bulk of a-boron quasi-crystals may be expected.158 Calculations of local structural deformations for a `compressed' boron quasi-crystal with rhombohedral lattice showed that closed hollow structures considered as precursors of the growth of extended NT are possible in this system.168 The formation of boronNTmay be critically affected by admixtures of various elements, e.g., carbon or nitrogen. 2.Nanotubulenes based on silicon and germanium The possibility of preparing NT consisting of silicon, which is a carbon analogue, still remains debatable. This is explained by great differences in the energy states of the atoms of these elements. It is known 169 that the most stable electronic config- uration for the carbon atom is sp2, whereas for silicon it is sp3. Therefore, graphite is the most energetically favourable crystalline form for carbon, whereas a diamond-like structure is the most favourable for silicon. In the nano-state, carbon exists as atomic chains, hollow fullerene-like clusters and NT, whereas silicon does not form any polyhedral nanoparticles.170 The first calculations of the energy and electronic states of hypothetical silicon-based NTformed by rolling of ribbon-shaped graphite-like structures of a layer composed of silicon atoms [nonchiral (6,6)- and (10,0)-NT and chiral (8,2)-NT] were carried out by the DFT method.171 The dependence of the electronic properties of such NTon their types is similar (see the correspond- ing reviews 7, 21, 26, 30) to that of similar carbon NT.This is evident from the state densities for silicon NT shown in Fig. 8: the zigzag (10,0)-NT is a semiconductor with narrow energy gap, whereas the dented (6,6)-NT has metallic conductivity. Of interest are comparative energy evaluations of various carbon and silicon structures. The total energies (Etot) of a silicon atom within a graphite-like network and in an NT are 0.79 and 0.83 eV lower than in crystalline silicon.The energy of cohesion (Ecoh) of silicon atom in NT is only 82% of Ecoh in a diamond-like crystal, whereas the Ecoh of carbon atom in NT and in crystal (graphite) differ by no more than 1%. According to estimates,171 the energies required for rolling of networks consisting of carbon and silicon atoms into tubular structures are very similar (0.04 ± 0.05 eV per atom). Presumably, the main problem in the synthesis of silicon NT is that it is difficult to obtain silicon in a lamellar (graphite-like) form. b Density of states (rel.u.) Density of states (rel.u.) a 0.5 0.5 E /eV E /eV 71.5 70.5 71.5 70.5 Figure 8. Full densities of states for silicon (6,6)- (a) and (10,0)- (b) nanotubulenes.171 Vertical lines show the Fermi energy.183 Calculations on mixed carbon ± silicon NT showed 172 that the Si?C replacement energy within carbon NT is *3.1 eV per atom. If a silicon admixture is present, it generates resonance states*0.7 eV above the Fermi level in the electronic spectrum of a metallic carbon (6,6)-NT. A silicon atom within a semiconduct- ing carbon (10,0)-NT gives rise to a local electronic level arranged *0.6 eV above the valence band. Seifert et al.173, 174 calculated the electronic states and analysed the stability of silicon-containing NT based on silanes and siloxanes with complex composition. In such NT, the silicon atoms forming the tube frame are surrounded by hydrogen atoms or by hydrogen atoms and OH groups.Like silicon, crystalline germanium has a diamond-like struc- ture with sp3 bond type; its quasi-2D modifications (of the CaSi2 type 175 or CaGe2 type 176) are also known. Recently, polymers containing networks of germanium atoms (GeH hexagonal layers) were synthesised.177 According to calculations,178 these com- pounds should be semiconductors (the width of the forbidden band is Eg&1.4 eV). The DFT method was used 179 to simulate the electronic properties of hypothetical (n,0)- and (n,n)-NT (n=6 ± 10, 20) based on GeH. The cross-sections of (8,0)- and (8,8)-GeH-nanotubulenes, which are the most stable of those described above, are shown in Fig. 9. It was shown 7, 21, 26, 30 that the dependence of the energy deformation of NT on its diameter (D) has the form: (1) Estr& 1 D2 .All of these tubes are semiconductors; the forbidden band width varies considerably depending on the NT diameter [from 1.33 to 1.16 eV for (20,20)- and (7,0)-NT, respectively]. Possible methods for the synthesis of these NT include 179 topochemical reactions and synthesis from molecular GeH4 . a b HGe Figure 9. Structures of cross-sections (8,0)- (a) and (8,8)- (b) for GeH- nanotubulenes.179 3. Phosphorus nanotubulenes Orthorhombic black phosphorus is the most stable allotropic form of elementary phosphorus. The crystals of black phosphorus consist of graphite-like networks, the distance between which (0.36 nm) is much longer than the length of the P7P bonds in the network (0.223 nm).Unlike the planar graphene layers, the net- works of phosphorus atoms are `bulged'; this is explained by the repulsion of unshared electron pairs in the neighbouring atoms. Seifert and Hernandez 180 compared the electronic and energy characteristics of an isolated network of black phosphorus with those of a number of hypothetical single-layer NT. In the initial stage, the structures of P-nanotubulenes (of zigzag and dented types) were constructed from ribbons of planar graphite-like layer of phosphorus atoms rolled to make a cylinder. After that, their geometry was optimised with respect to the axial displacements of atoms. Generally, relaxed structures of P-nanotubulenes are184 similar to the structures of carbonNTand only differ in the type of six-membered rings.In the case of P-nanotubulenes, these rings have the form of cyclohexane rings consisting of phosphorus atoms. The strain energy (Estr) of phosphorus NT is considerably higher than the Estr of carbon NT with similar diameters. This difference increases as the tube diameter decreases, which is explained 180 by changes in the electronic configurations of the carbon and phosphorus atoms in the networks. As a graphene network consisting of sp2 carbon atoms folds, the increase in the wall curvature increases the contribution of the sp3 orbitals.181 For a phosphorus network consisting of sp3 phosphorus atoms, minimisation of the repulsion energy of unshared electron pairs of adjacent atoms occurs due to the maximum increase in the distances between adjacent phosphorus atoms upon their dis- placement from the network plane (the network `bulging' men- tioned above).As such a `bulged' network is folded, the distances between the unshared electron pairs of certain phosphorus atoms decrease, resulting in rapid increase in Estr . However, the Estr of phosphorus NT at D>1.23 nm is no higher than 0.1 eV per atom, which is comparable with the deformation energies of the corresponding carbon NT. These energy estimates suggest that there are no obstacles to the existence of P-nanotubulenes. However, it should be expected that NT with rather large diameters will be formed. The following methods may be used to obtain P-nanotubulenes: the template method, a modification of the method for the synthesis of thin phosphorus films upon luminescent decomposition of PH3 182 and sublimation of phosphorus in the presence of catalysts.It was noted 180 that phosphorus NT should have semicon- ducting characteristics invariable with respect to diameter changes (Eg changed in the range from 0.9 to 1.99 eV). Considering that the properties of phosphorus and arsenic are similar, the possi- bility that As-nanotubulenes may exist cannot be ruled out. IV. Theoretical models of nanotubes based on p- and d-element compounds 1. Models of nanotubes consisting of B, C and N atoms It was found in early studies 4±8 of carbonNTthat their properties could be affected by doping.The cyclic cluster model and the semiempirical quantum- chemical MNDO method were used to study the charge distribu- tion in carbon (6,6)-NT doped (one atom per unit cell) with one of the atoms: B, N or Si, in neutral, anionic and cationic state.183 It was found that the distortions introduced by the dopants propa- gated along the NT to longer distances than across the NT. Distortions from charged dopants have a longer range than those from neutral atoms. The local structural deformations of carbon NT appearing upon doping with single boron or nitrogen atoms were esti- mated.184 In the case of nitrogen, the effects of structure relaxation are weak. Doping with boron resulted in displacement of the nearest carbon atoms by*0.011 nm along the B7C bonds.Scanning tunnelling spectroscopy and calculations within the DFT were used to study the effect of boron atoms on the proper- ties and growth of carbon NT.185, 186 Doping of multilayer carbon NT with boron confers on them metallic properties; the boron atoms are segregated to form separate islets in NT walls.185 A boron admixture prevents the formation of `caps' and favours the preferable growth of (n,0) type NT.187 This mechanism does not apply to dented NT; this fact can be used for growing selectively NT of certain geometries. Electrophysical studies 188 showed that doping of carbon NT with boron increases the paramagnetic susceptibility of the latter. Furthermore, their thermal factors of electrical conductivity are positive.a. Boron nitride Boron nitride was the first compound suggested for the synthesis of non-carbon NT. Quantum-chemical methods 36 ± 38 revealed a A L Ivanovskii number of interesting features of BN-nanotubulenes. Unlike carbon NT, the conducting properties of which are determined by their diameters and chiralities,4± 8 BN-nanotubulenes are wide- gap semiconductors (Eg&5.5 ± 6.0 eV), which is important for the development of materials with predetermined properties. More detailed information about the electronic and energy characteristics of `ideal' single- and multilayer BN-nanotubulenes can be found in a monograph 7 and in a number of reviews.21, 29, 30 In recent years, BN-nanotubulene structures were mostly simulated using the molecular dynamics method.76, 189 The prop- erties of the following tube types were compared:76 chiral, zigzag- shaped, dented (built of graphite-like networks of B3N3 hexagons) and those built of four- (B2N2) and eight-membered (B4N4) rings.The latter type of BN-nanotubulenes were found to be the less stable. The elastic properties of BN-nanotubulenes were esti- mated:190 their Young modulus (Y) was 14 times higher than that for crystalline graphite-like BN, and comparable to the extremely high Y predicted 191, 192 for carbon NT (1 ± 6 TPa). This agrees with the subsequent measurements 193 of the Young modulus for multilayer and single-layer carbon NT (0.4 ± 4.0 TPa). The experimental studies 194 on the elastic proper- ties of mutlilayer BN-nanotubulenes give values which are in good agreement with calculations 190 (Y&1.220.24 TPa).Of the currently known elongated nanostructures with dielectric proper- ties, BN-nanotubulenes are characterised by extreme values of the Young modulus comparable to those of glassy fibres.195, 196 The growth mechanisms of BN-nanotubulenes depend on their structures and differ considerably for (n,0)- and (n,n)-nano- tubes.197, 198 The determining role in the growth process belongs to the specifics of the formation of terminal structures by virtue of B7N binding, these bonds being much more favourable than the B7B and N7N bonds. Zigzag NT stop to grow once `caps' have formed. Dented NT grow by addition of extra atoms near the NT ends.The role of BN-rings of different types (in particular, five- membered rings) in the formation of BN-nanotubulene `caps' and polyhedral boron nitride clusters, which are heteroatomic full- erene analogues, was discussed.199 ± 207 b. Boron carbides Unlike BN-nanotubulenes, the type of electronic properties of BC3-nanotubulenes changes from metallic to semiconductor-type with the increase in their diameters.207 It is possible to increase the conductivity (hole concentration) of such NT by partial replace- ment of carbon atoms by `superstoichiometric' boron atoms. Halogen atoms can serve as electron donors, as is observed for a number of intercalated graphites. The value of Estr for BC3- nanotubulenes is lower than that for carbon NT, which is in agreement with experimental data on their syntheses (see above). c.Carbon nitrides The existence of condensed carbon nitride (CNx) and the extreme (comparable to that of diamond) hardness of b-C3N4 (the b-Si3N4 structure type) were predicted.208, 209 Subsequently, CNx was synthesised successfully (in amorphous, film and nanocrystalline states 210 ± 217) and both various polymorphous modifications of crystalline CNx and its tubular nanostructures were simulated theoretically. Fundamental electronic properties of CNx with the structures of the a-Si3N4 ,218 ± 220 Zn2SiO4 and a-CdIn2Se4 219, 220 types and of a series of its quasi-2D modifications 219 ± 222 were considered. A lamellar structure containing graphite-like networks formed by a combination of triazine rings with nitrogen bridges was sug- gested for C3N4 .221 Each carbon atom in the network is bound to three nitrogen atoms; each of the latter has only two closest neighbours.Planar C3N4-networks can be packed according to ABAB... type (the Bernal structure) or ABCABC... type (the rhombohedral graphite structure).222 The mean C7N bond length is *0.136 nm (which is close to the sum of the Sleyter radii for the carbon and nitrogen atoms), the C7N7C bondNon-carbon nanotubes: synthesis and simulation ab CN Figure 10. Planar networks used to simulate NT with compositions CN (a) and C3N4 (b).225 angle is 116 8 (see Ref. 221). It was found 223 that vacancies appearing in the N-sublattice of cubic C3N4 favour its trans- formation into a pseudoplanar form.A number of other densely packed and lamellar modifications of carbon nitride have also been suggested (see the review 224). The structures of non-chiral single-layer (n,n)- and (n,0)-CN- nanotubulenes with different stoichiometry are described by net- works similar to those of the C3N4 layers mentioned above and a hypothetical graphite-like layer with the composition CN (Fig. 10).225, 226 All CNx-nanotubulenes, irrespective of their structures and diameters, have metallic conductivity. Their defor- mation energy increases with a decrease in diameter, but it is smaller than that of carbon NT. In turn, Estr for CN-nano- tubulenes is smaller than Estr for NT based on C3N4 (Fig.11). It is assumed 225 that the interlayer distances in multilayer CNx- nanotubulenes will be*0.04 nm smaller than those in multilayer carbon NT. The sp2-configuration of nitrogen atoms favours the stability of NT. Estr /eV 0.4 123 0.3 0.2 0.1 8 6 4 D /A Figure 11. Deformation energy ± diameter relationships for C- (1), C3N4- (2) and CN-nanotubulenes (3).225 The formation of fullerene-like molecules, NT and CNx crystals was simulated by the molecular dynamics method.227 The simulation started from a planar fragment of a graphite network forming a cluster of 96 carbon atoms; in this system, the structural deformation upon replacement C?N in the N:C range 0.07 ± 1.00 was studied. The CNx layer remained planar up to the nitrogen content of*20%; then it became `corrugated' due to deviation of atoms (up to*0.07 nm) away from the plane.It is believed 228 that this results from occupying the antibonding p*-orbitals (which is visible in photoelectron spectra) and for- mation of new, rather stable N7N bonds. The CNx crystals may have various types of the layered packing of such deformed nitrogen ± carbon networks. In should be noted that when a critical nitrogen content (*20%) is reached, a different type of structural evolution of this system is possible to result in NT or a fullerene-like molecule.227 It is important that stabilisation of the tubular and molecular states is attained at different system stoichiometry, viz., CN and C3N4 , respectively.It is of note that onlyCNx-nanotubulenes and nanofibres with relatively small x values (x40.04 ± 0.05) are known.229 ± 236 Their electronic properties are interpreted on the basis of calculations for carbon NT containing nitrogen admixtures.236, 237 In the general case, replacement of some carbon atoms in an NT by nitrogen atoms (electron dopants) increases the population of the energy bands and `metallisation' of the NT. Admixtures of nitro- gen create donor levels near the edge of the conduction band in carbon NT.236, 237 Introduction of nitrogen atoms results in the formation of local pyridine-like fragments within NT walls; this explains the subtle features of electronic state distribution near the Fermi level, which may be important when creating nano-sized devices with hetero-transitions based on these NT.236 d.Boron carbonitrides Theoretical construction of nanotubular structures was carried out based on a lamellar form of boron carbonitride BC2N the electronic spectrum of which has been studied previously.238 ± 240 It was assumed that the intralayer charge anisotropy, which causes the formation of `channels' with increased and decreased electron density in BC2N networks,238 ± 243 would be preserved when these networks are rolled into cylinders (according to the chiral type) and would result in spiral motifs. As a result, circular currents would appear in BC2N-nanotubulenes, and suchNTmay be considered as nano-solenoids (Fig. 12). a Ez BCN Figure 12.A variant of the atomic structure of a BCN-tube (a) and the scheme of circular currents therein (b).241 Ez is the direction of the external field. 185 b Jz Jx186 Nanotubes with small diameters maintain the conductivity type characteristic of the original BC2N monolayers. The metallic type of conductivity of (n,m)-NT can change to the semiconductor type at certain diameters and chirality. The most stable are those (n,n)-NT which contain the maximum number of strong C7C and B7N bonds. Miyamoto et al.243 analysed the conditions required for the appearance of circular currents in such NT. Attempts were made to describe the role of defects in the properties of BC2N-nanotubulenes.241, 243 Partial replacement of carbon atoms by boron (or nitrogen) resulted in donor (or acceptor) admixture states for B1+xC27xN (or BC27xN1+x) in the electronic spectra.Miyamoto et al.243 considered the role of electron (or hole) carriers in the formation of chiral currents in BC2N-nanotubulenes.243 Analysis of the general trends in the formation of the atomic structure of BC2N-nanotubulenes performed on the basis of results of theoretical studies showed that such nanotubulenes are characterised by the formation of BN- and C-islets.244, 245 The BCx-nanotubulenes should undergo similar spontaneous segrega- tion into regions consisting of carbon atoms and BC3 fragments. These phenomena can give rise to quantum dots and hetero- transitions. 2. Nanotubes based on gallium nitride and selenide The possibility of obtaining tubular structures for other com- pounds of Group III ±V elements, which do not form lamellar structures under equilibrium conditions, was analysed for GaN (Ref.246) and GaSe.247 In the first stage, the structures of a hypothetical crystalline graphite-like GaN were simulated by the DFT method and the equilibrium distances Ga7N in the layer were determined (0.178 nm).246 The difference between the bond energies of sphalerite-like (s) and graphite-like (g) phases of gallium nitride was found to be small, namely, 0.36 eV per atom. This value is smaller than the energy difference between s-GaN and its meta- stable cubic (sodium chloride-like) modification (0.5 eV per atom), which is considered 246 to be evidence of the possibility of obtaining a lamellar allotrope of GaN.The next simulation stage involved calculations of the elec- tronic structure and Estr of gallium ± nitrogen (n,n)- and (n,0)-NT as a function of their diameters. Sawada and Hamada 248 esti- mated the minimum possible NT sizes corresponding to zigzag- like (5,0)- and dented (2,2)-NT. It was found that Estr changes inversely to the square of the NT diameter; the geometry type of the walls of such tubes (unlike carbon ones 249) only slightly affects Estr . On the contrary, this factor becomes important for the dielectric properties of GaN-nanotubulenes: although both tube types are wide-gap semiconductors [e.g., the values of Eg for (5,5)- and (9,0)-GaN-nanotubulenes are 2.15 and 2.16 eV, respectively], dented (n,n)-NT, like the hypothetical g-GaN, have indirect type of inter-band transitions.Zigzag-like (n,0)-NT have direct type of the transitions. The Eg of (n,0)-tubes decreases most significantly with a decrease in diameter. The GaN-nanotubulenes (like BN- nanotubulenes 7) were found to manifest a relaxation effect: the atoms of the cationic (Ga) and anionic (N) sublattices are displaced by *0.005 nm from the surface of an ideal cylinder: the gallium atoms are displaced towards the tube axis, whereas the nitrogen atoms are displaced in the opposite direction. Similar properties are also typical of GaSe-nanotubulenes.247 It is assumed that GaN-nanotubulenes can be obtained by template synthesis using carbon NT as the substrate.246 This will probably result in multilayer tubes with various chemical compo- sitions (C/GaN nanotubular composites) with a metal ± semicon- ductor interface. To date, tubular composites which are carbon NT partially filled with GaN, as well as separate GaN rods with 7 ± 10 nm diameters and up to 40 mm long, have been synthes- ised.250 These specimens were obtained in an electric arc in an atmosphere of nitrogen with a graphite cathode and an anode containing a mixture of gallium nitride, graphite and nickel powder.A L Ivanovskii `Pure' GaN-nanotubulenes [with Eg values *0.45 eV higher than those of the cubic phase of GaN (see Refs 246, 251, 252)] can be used in nanosized optoelectronic devices that are necessary, e.g., in colour displays.3. Sulfide nanotubes Nanotubes based on lamellar molybdenum and tungsten disul- fides were constructed 253, 254 from fragments of triple-layer pack- ages S ± (Mo,W) ± S rolled into cylinders along certain chosen directions of the two-dimensional lattice. The tubes were classified (by analogy with the carbon NT) as non-chiral zigzag-like (n,n)- and dented (n,0)-NT. The theoretical electron diffraction patterns built for these structures 253 well describe the experimental data.255 Seifert et al.253, 254 used the DFT method in the strong bond approximation to study the dependence of the electronic proper- ties and stability of a series of disulfide NT on their diameters. For example, the diameters of (n,0)- and (n,n)-WS2-nanotubulenes (n=6, ..., 22) ranged 254 from *1.1 [for (6,6)- and (10,0)-NT] to *3.1 nm [for (18,18-NT)]. For MoS2-nanotubulenes, the diame- ters ranged from 0.8 to 2.6 nm.253 The maximum diameters of model NT used in these studies were close to the minimum diameters of the disulfide NT synthesised. The energy and electronic characteristics of all disulfide NT have certain common features,253, 254 which are also characteristic of crystalline phases.255, 256 Three bands are observed in the valence band of WS2-nanotubulenes, viz., A, B and C (Fig. 13), which result from contributions of quasi-framework S3s- (band A), hybridised S3p±W5d- (band B) and W5d-states (band C). The W5d-states dominate in the vicinity of the lower edge of the conduction band.All (W,Mo)S2-nanotubulenes are narrow-gap semiconductors. However, (n,0)-NT are semiconductors with direct transition in point G, whereas the family of (n,n)-NT is characterised by indirect transition type (Fig. 14). The forbidden band width increases with a decrease in the tube diameter, but it B A C E /eV 0 710 720 Figure 13. Full density of states for (18,18)-WS2-nanotubulene. EF=0.254 E /eV a b 21 EF EF 0 71 72G GX X Figure 14. Energy zones for (22,0)- (a) and (18,18)- (b) WS2-nanotubu- lenes.254 Horizontal lines show the Fermi energies. Density of states (rel.u.)Non-carbon nanotubes: synthesis and simulation remains smaller than the Eg for three planar S ± (W,Mo) ± S layers. This result agrees with optical experiments 256 which showed that Eg in the electronic spectrum of WS2-nanotubulenes was smaller than that for crystalline tungsten disulfide.Charge transfer in disulfide NT occurs in the direction W(Mo)?S.253, 254 For MoS2-nanotubulenes, the effective charges are*0.9 e (0.44 e per sulfur atom), which is much smaller than the formal ionic charges in the crystal (Mo4+, S27). The deformation energies for disulfide NT are higher than those of carbon NT with the same diameters. In turn, the Estr of zigzag-like (Mo,W)S2-nanotubulenes is somewhat higher than the Estr for dented ones. Seifert et al.253 studied the effect of topological defects (in the form of four- and eight-membered rings) on the structure of the `caps' on closed NT and polyhedral nanoparticles (Mo,W)S2 (see also Ref.257). Milosevoc et al.258 considered the general geo- metrical aspects of the structure of dichalcogenide NT, deter- mined the symmetry groups of single-layer NT based on 2H- and 3R-polytypes of MoS2 and WS2 and discussed some of their properties that depend on the system symmetry (in particular, such as splitting of the energy bands, rules of selection and structures of the phonon spectra). It is assumed 253 that the semiconductor conductivity type of (Mo,W)S2-nanotubulenes can be changed to the metallic type if the constituent molybdenum or tungsten atoms are partially replaced by niobium atoms. Recently,259 annealing of Nb2O5- coated W18O49 nanorods at 1100 8C in H2S atmosphere gave the first `mixed' NT with the composition W17xNbxS2 (x40.2). Their properties have not been studied so far.The DFT method was used to evaluate the stabilities and simulate the electronic properties of hypothetical (n,n)- and (n,0)-NbS2-nanotubulenes 260 {n=6, ... , 22, the NT diameters ranged from *1.0 [for (6,6)- or (10,0)-NT] to *2.4 nm [for (14,14)-NT]}. It was found that the strain energy Estr of NbS2- nanotubulenes obeys Eqn (1), and that it is higher than that of carbon NT with the same diameters. This fact was explained by a higher energy consumption for rolling of a fragment of an S7Nb7S triple-layer sheet into a tube, in comparison with similar consumption required for rolling of a graphene mono- layer. The energy bands of the tubes considered above differ only slightly from each other and from the energy bands of planar niobium disulfide layers.261 The states of these bands near the Fermi level (predominantly of Nb4d-type) have high density, which determines their metallic properties; this, in turn, gives a reason to expect 260 that NbS2- and NbSe2-nanotubulenes may manifest superconductivity. It was noted 260 that the onset of superconductivity may be due to an increase in the electron ± phonon interaction and an increase in the critical temperature (Tc) in tubular niobium disulfide structures in comparison with the Tc for its crystal [*5.6 ± 6 K (see Ref.262)]. It was suggested that NbS2-tubes could be obtained by thermal treatment of NbOx nanorods in an H2S atmosphere or by using template synthesis on carbon NT as the templates. The latter method gave 263 long nanostructures, viz., niobium disulfide tubes and rods.4. Boride nanotubes A broad class of lamellar materials comprises metal diborides (MB2) and other MX2-phases (X=Be, Si, Ga, Hg, Zn, Cd, Al, Cu, Ag, Au) with the AlB2 type structure (space group D16h7 P6/mmm).264, 265 These materials aroused particular interest due to the discovery, in 2001, of a superconducting transition in MgB2 (Tc&40 K).266 Since then, superconducting MgB2 has been obtained as single crystals, dense and porous ceramics, films, long wires and bands, as well as in a nano-structured state; numerous studies of its properties have been carried out (see the reviews 267, 268).Chernozatonskii 269 analysed whether MgB2 can exist as an NT. Double-layer MgB2-nanotubulenes were simulated by rolling 187 (by the chiral and non-chiral types) ribbon-like fragments cut from two adjacent layers of a MgB2 crystal. It was assumed that stabilisation of the electron-deficient boron layer occurred due to partial electron transfer (Mg?B); the bond between the coaxial, oppositely charged Mg- and B-cylinders is partially ionic, and the distance between them is similar to the interlayer distance in an MgB2 crystal. According to molecular dynamics estimates, dibor- ide (n,n)-NT, which have a minimum deformation energy if the diameter of the `outer' B-tube is *19 A [for (11,11)-MgB2-nano- tubulenes], are more stable. The author's conclusions are based on estimates of the lattice instability of the MgB2 and AlB2 type structures 270 formed by alternating (in the sequence ...MXMX...) planar hexagonal metal networks (M) and graphite-like non- metal layers (X).Synthesis of magnesium-diboride NT is most likely to give tubes of a dented type. Considering the results of calculation of the electronic structures of crystalline MgB2 (see Refs 267, 271 ± 274), it is assumed 269 that such tubes should have metallic conductivity. It is also assumed that they may have superconducting properties (see Refs 7 and 272 on the superconductivity in carbon NT). Anumber ofNTstructures based on ZrB2 have been reported. It is assumed 269 that the `caps' of closed MB2-nanotubulenes can have the form of structures with defects in the M- and B-shells as five- and four-membered rings.Probably, diborides may also exist as multilayer structures, such as multilayer NT, or as onions. Such systems as (7,7)@(11,11)@(15,15)-NT and (MgB2)180@(MgB2)288 , the so-called bifullerene, may serve as examples. The DFT method was used to calculate several structures,275 the prototypes of AlB2-nanotubulenes. It was found that all diboride NT should have metallic conductivity. V. Models of hybrid nanotubular structures 1. Intercalated BN-nanotubulenes Particular interest is attracted to the control of the properties of NT and synthesis of new nanocomposite materials with nontrivial functional properties based on these.Chemical modification of NT is one of the most effective methods to solve these problems. The properties of NT can be changed in various ways, for example, (1) by introducing compounds � intercalants (atoms or molecules)�into the internal cavity of single-layer or multilayer NT, into the space between layers (neighbouring walls) of multi- layer NT or into cavities between the adjacent NT in NT braids or in nanotubular crystals; (2) by adding foreign atoms or molecules to the broken bonds of atoms belonging to the terminal groups of open NT, to atoms of the `caps' of closed NT or to atoms of the outer surface of NT walls (adsorption on NT surface); (3) by replacing partially the NT atoms by atoms of some other kind (NT `doping').These modification methods became most popular for carbon NT (see the reviews 15 ± 17, 21 ± 28). For non-carbon NT, simulations of intercalated nanotubulenes and `symbiotic' tubular structures are used most frequently. In the latter case, this idea was used for the possible formation of various 1D-heterostructures [the so- called symbiotic boron carbonitrides, viz., (C+BN)-nanotubu- lenes] in addition to `ideal' NT with a regular distribution of different kinds of atoms over the nodes in a B7C7Nsystem.276 It is expected that the conducting properties of such NT would vary widely. Certain experimental data on their synthesis were men- tioned above. Chernozatonskii et al.277 considered hypothetical BCN-nano- tubular structures consisting of alternating C- and BN-fragments as representatives of `symbiotic' NT.Considering the similarity of the bond lengths C7C (0.142), C7N (0.150) and B7N (0.145 nm in graphite-like planes), such heterostructures can be formed,276 e.g., as a result of the growth of carbon NT on188 substrates containing BN nanocrystallites, or as a result of the formation of BN-nanotubulenes on diamond crystallites, or due to both processes occurring simultaneously. Yet another way to obtain these kinds of structures is chemical doping; as the carbon atoms are replaced by B and N, the latter form BN-`islets' within the original carbon NT. Simulation of symbiotic tubular structures in the B7C7N system made it possible to predict a number of interesting proper- ties for them.277, 278 It was suggested to consider C± BN-nano- tubulenes containing rings of carbon atoms as a new class of quasi-one dimensional systems with quantum dots in which the local disturbance of the energy of electronic states exceed a certain characteristic value (usually kTroom&0.024 eV).The crystal orbital method in one-orbital approximation was used to calculate the electronic spectrum of zigzag-like (n,0)-BN-nanotubulenes (n=5, 6, 9, 27, 72) in which a number of (B,N)-atomic rings in the bulk or at the ends of open tubes were replaced by rings of carbon atoms. The states of the terminal carbon rings form a number of discrete levels around the dielectric gap of BN-nano- tubulenes; as the number of carbon atoms increases, these levels are transformed into mini-bands.For example, the spectrum of carbon states (in the dielectric gap) for (27,0)-CBN-nanotubu- lenes consists of two groups of levels the widths of which are 0.7 and 1.7 eV. A similar effect (Fig. 15) was noted for the states of bridging carbon rings between the fragments of neighbouring NT. It is assumed that the existence of `quantum dots' (or `quantum rings',278, 279 i.e., conducting carbon rings at the ends or between dielectric BN-nanotubulenes) can result in unusual properties of these structures, e.g., lead to resonance effects in external mag- netic fields. Lammert et al.280 discussed the results of doping of carbon NT with boron and nitrogen admixtures carried out in order to obtain structures containing chains of quantum dots.Such structures can be regarded as models of novel quantum nano-devices. c b a E /eV 642 EF EF EF 0 72 74 76 78 2 2 2 0 0 0 p p p Figure 15. Energy zones for (BN)3C2(BN)3 fragments of `symbiotic' (n,0)- nanotubes;277 n=9 (a), 24 (b), 72 (c). The electronic characteristics of the hetero-transitions in NT consisting of the C/BN, BC2N/BN segments (the superlattice model) were found 279 to be stable against changes in the radii, chirality type and the number of layers in these NT. Intercalated boron-nitride NT present yet another type of `tubular composites'. Rubio et al.281 calculated the electronic spectra of (4,4)-BN-nanotubulenes which incorporated potassium or aluminium `fibres' arranged along the cylinder axis.The evaluated intercalation energies (the differences between the full energies of the composite and its constituents, namely, the NTand the metal `fibre') were much smaller than the intercalation energies for similar systems with carbon NT. In the system K@(4,4)-BN- nanotubulene, no new states appear in the electron energy spectrum. In the system Al@(4,4)-BN-nanotubulene, additional A L Ivanovskii levels of Als, p-type appear in the energy gap, and the electronic spectrum of this system becomes similar to a superposition of the spectrum of a monoatomic chain of aluminium atoms and the spectrum of (4,4)-BN-nanotubulene. The density of electron carriers, which define the conductivity of intercalated BN-tubes, is concentrated inside the cylinder, hence such NT are preferable (as compared to carbon NT) when making nano-wires with predefined parameters. It is deemed possible to intercalate silver and gold atoms into BN-nanotubulenes. Hernandes et al.282 presented a new idea that a molecular diode can be created on the basis of carbon NT intercalated with boron and nitrogen. A nano-diode consists of a carbon (10,0)-NT (with metallic conductivity). A `fibre' along the axis of this tube has two parts, viz., one formed by boron atoms (p-dopants) and the other formed by nitrogen atoms (n-dopants).According to calculations, n ± p-doping results in a potential barrier (*1 eV) between the adjacent sections of the tube containing the boron and nitrogen parts of the `fibre', respectively.Jishi et al.283 discussed the electronic states of (n,n)-BC3-nanotubulenes (n=2, 3, 4) intercalated with a chain of bromine atoms. 2. `Symbiotic' nanostructures: metal ± carbon cages (met- carbs) in nanotubes The effect of intercalation on the properties of NT for more complex systems was considered by Ivanovskii et al.284, 285 It was noted 284 that the formation processes, stabilities and properties of M@C type nanosystems obtained upon intercalation of NT with atoms of various metals depends on the nature of the intercalant and the atoms that form the NT. For example, lead, zinc, nickel and copper atoms readily penetrate the carbon NT (these metals are either inert to carbon under equilibrium conditions or form unstable carbide phases 286).On the contrary, introduction of atoms of Group IV ±VI d-metals, which readily form carbides, is hindered: their strong interactions with the carbon atoms in the NT walls would result in the deformation and destruction of the latter. Obviously, similar reasons should prevent the synthesis of M@NT type nanostructures (where M are atoms of the Group IV ±VI d-elements) based on BN-, BC- and BCN-nanotubulenes. It is assumed 284, 285 that new nanostructures incorporating Group IV ±VI d-metals can be prepared if the latter are intro- duced into NT as their `own' stable nanostructures. For example, these are represented by polyhedral clusters containing metal atoms along with carbon atoms, the so-called met-carbs, discov- ered not so long ago (in 1992).287, 288 Metallocarbohedranes with the composition M8C12 are the most stable of them.Metal ± carbon nanoparticles with other compositions and structures have also been obtained (see, e.g., the reviews 289 ± 293). Ivanovskii et al.284, 285 calculated the energy and electronic characteristics of hybrid quasi-one-dimensional nantems (M8C12@NT), i.e., crystals formed by regular chains of met- carbs {[M8C12]?, where M=Sc, Ti, V, Zr, Nb} arranged along the axis of single-layer C-, BN-, BCx-, BxC17x7yNy , Si- and GaN-nanotubulenes. Analysis of the electronic structures of a number of systems belonging to quasi-unidimensional M8C12@(12,0)-NT type (Fig. 16) was carried out with consideration of the chemical composition of the NT, mutual arrangement of met-carbs and NT (displacement of the met-carbs along the axis and rotation around the tubulene axis were simulated) and the electron concentration (EC) in the system defined by the nature of the transition metal.284 1D-Crystals incorporating the following met- carbs were considered: Sc8C12, Ti8 C12, V8C12, Zr8 C12 and Nb8C12 .It was found that the electronic spectrum of a quasi-1D Ti8C12@BN crystal consists of five occupied bands formed by the N2s-, C2s,2p-, (C2s,2p+N2p)-, (N2p-B2s,2p+Ti3d )-, C2p- and Ti3d-states. The first four bands contain the states responsible for the chemical bonds in Ti8C12 nanostructures and in BN-nano- tubulenes. The states of transition metals around the Fermi level are responsible for the bonds of the neighbouring met-carbs in the chain and for the chain ±NT interactions.Non-carbon nanotubes: synthesis and simulation a b It is difficult at the moment to compare the methods existing for the synthesis of non-carbon NT with respect to their produc- tivity and effectiveness. It is possible that, even in the future, it will be hard to prefer some specific `versatile' method, because the efficiency of each synthetic method depends on the chemical nature of the non-carbon NT.CTi BN c d e ical and functional properties are only occasional. These are only reported in selected studies and are far from being comprehensive. Although it is possible to speak now of inorganic, physical, colloid, polymer and analytical chemistry of carbon NT,27 it is too early to distinguish any particular areas of chemistry of non- carbon NT.The materials science of non-carbon NT is also in its infancy. Many suggestions to use them in practice are based on theoretical simulations and are far from technological solutions. The wide-scale production of non-carbonNTis also a problem yet to be solved. f Figure 16. Models of 1D-crystals of Ti8C12@C- (a) and Ti8C12@BN-nanotubulene (b);284 (c ± f ) possible mutual orientations for met- carb ± nanotube. It should be noted that stabilisation of a [Ti8C12]? polymeric chain occurs in a 1D-crystal of Ti8C12@BN-nanotubulene. How- ever, the interatomic bonds in the NT and met-carbs change by no more than 1%± 3%.In other words, the main types of the bonds which make these structures stable in the `isolated' state are preserved in the crystal, which is evidence in favour of the stability of the simulated `symbiotic' nanosystems. Ivanovskii et al.284, 285 analysed the stability of this type of 1D-crystal depending on the position of the met-carbs and the NT and on the nature of the metal atom. Calculations for 1D-crystals incorporating Sc8C12 and BN-nanotubulene in which the number of electrons per met- carb is less than 8 [as compared to the number of electrons per met- carb for Ti8C12@(12,0)-BN-nanotubulene] showed that there are no bonds between the met-carbs. In contrast, the increase in ECon going to the V8C12@(12,0)-BN-nanotubulene system favours a noticeable increase in the population of these bonds (by *54%).Thus, it is possible to control both the bond strength and the stability of a system as a whole, as well as its dielectric properties, by changing the EC in the system (for example, by changing the chemical composition of the met-carbs by `doping' them 293). It should be noted that a number of hybrid systems incorpo- rating two types of stable nanostructures, viz., fullerenes enclosed into an NT (Cn@C type) 69 and NT incorporating endofullerene chains (M@Cn@C type),70 have recently been synthesised. VI. Conclusion Development of methods for the synthesis of non-carbon NT is now the leading direction in experimental studies on these unique subjects.Along with attempts to apply the methods known for the synthesis of carbon NT to the synthesis of other inorganic compounds in nanotubular form, elaboration of many `special' synthetic methods is currently in progress. In addition to physical methods, such as electric arc synthesis and laser ablation,3 ± 6, 15, 16, 24 which are particularly popular for producing carbon NT, chemical synthesis is becoming more and more important. In many cases, carbon NT serve as starting materials for the synthesis of non-carbon NT (chemical substitu- tion, template synthesis). 189 The stage at which experimental studies are today is charac- terised by research into the synthesis of as wide a range of non- carbon NT as possible. The majority of studies describe the synthetic procedures and report the chemical compositions and morphology of the products.However, studies on physicochem- Theoretical models have been developed for those non-carbon NT which have (or assumed to have) the closest similarity to carbon tubulenes. Ideal models of non-carbon NT in the shape of ideal cylinders rolled from a monoatomic sheet (or a polyatomic layer) remain the basis for describing these structures. Problems of describing various defects in non-carbon NT, processes of their growth and association into `braids', films and crystals still need to be solved. There are no published data on the electronic and energy characteristics and properties of composite NT (e.g., oxide NT that comprise oxide layers and an organic component).Problems of theoretical simulation also include NT reactivity and deviation of non-carbon NT from ideal morphology. * * * When this review was being prepared for publication, new interesting results appeared dealing with the synthesis and study of the properties of new non-carbon NT and materials based on them, as well as with their theoretical simulation. The most important studies are briefly considered below. Tang et al. 294 proposed a simple method for the synthesis of boron ± nitrogen NT by thermal treatment of a mixture of elementary boron and iron oxide (in the ratio 15 : 1) in a flow of gaseous NH3 . Chemical deposition from the gas phase onto a support containing Al2O3 particles with sizes of several micro- metres gave multilayer (containing from two to five concentric layers) BN-nanotubulenes with unusually large inner diameters (8 ± 10 nm).295 Guo et al.296 developed a two-stage process for chemical deposition from the gas phase, which allowed them to obtain Y-shaped splices of boron ± carbon-nitride NT.Maet al.297 obtained a `coaxial cable', i.e., a carbon NT filled with iron `core' and containing dielectric boron ± nitrogen nanotubulenes as the housing. It was confirmed by EELS that the carbon and boron ± nitrogen layers in such heterostructures are separated from each other.Studies on dichalcogenide NT were continued. Thermal decomposition of (NH4)Mo17xWxS4 in an inert atmosphere or in hydrogen gave single-layer and multilayer NT with `mixed' composition Mo17xWxS2 (x<0.1) 298 containing new defect types.A method for the synthesis of sulfide NT with the composition Cu5.5FeS6.5 was suggested.299 Pyrolysis of a mixture of WO37x with carbon NT in an N2+H2S atmosphere gave new composite structures, viz., single-layer carbon NT encapsulated in multilayer WS2-nanotubulenes.300 Problems of hydrogen accu- mulation in MoS2-nanotubulenes were discussed.301 Synthesis of silicon NT, both open and closed on the ends, with inner diameters of *60 ± 80 nm was reported.302 They were obtained by deposition of a mixture of SiCl4 with hydrogen onto a NixMgyO support that served simultaneously as a catalyst.190 Interesting results were reported by Bao et al.303 who obtained nano-sized nickel tubes by electrochemical deposition of metallic nickel in pores of an alumina membrane. These nanostructures manifest magnetic properties.Li et al.304 were the first to obtain gallium-nitride NT. Progress was also achieved in methods for the synthesis of oxide nanotubulenes. New schemes for the preparation of vana- dium-oxide NT were discussed.305, 306 Dobley et al.306 were the first to obtain mixed NT with the composition Mn0.1VO2.5+d by ion-exchange reactions. A method for the synthesis of TiO2 , ZrO2, Nb2 O5 and Ta2O5 metal-oxide nanotubulenes using organic crystallites as templates was discussed.307 Reactions of crystalline TiO2 (anatase or rutile) with a NaOH solution gave a multilayer titanium-oxide NT with distances of *0.75 nm between the adjacent layers.308 The chemical compositions of the tubes (viz., the Ti :O ratio) vary widely.A compacted set of titanium oxide-based NT was found in a product of anodic oxidation of pure titanium.309 The NT diameters range within 25 ± 65 nm; their ends closed on the support side form a barrier layer resembling porous Al2O3 . GeO2 nanorods of 50 ± 150 nm diameters and a few hundred micrometres long were obtained.310 Studies of the photolumines- cent properties of these nanostructures showed that they are promising as materials for optoelectronic nanodevices. The sol ± gel method was used to obtain semiconducting Ga2O3 and In2O3 nanotubulenes with*100 nm diameters.311 Maurin et al.312 studied the electrochemical doping of carbon electrodes built ofNTwith lithium; a number of other issues of the electrochemistry of lithium-doped carbon NT were considered elsewhere.313, 314 The positron annihilation method was used to study multilayer carbon NT doped with potassium, rubidium and cesium.315 A review paper 316 discussed methods for the synthesis of composites containing metal atoms intercalated in carbon NT and the role of metal atoms as NT growth catalysts, as well as the results of studies of these composites by molecular dynamics and quantum-chemical methods.A new composite material consisting of multilayer carbon NT in amorphous SiOx has been obtained.317 Furthermore, the SiOx/NT structure was simulated by a strong-bond DFT method. The sol ± gel method was used to obtain yet another new compo- site containing vanadium pentoxide and carbon NT, which may be promising as lithium battery electrodes.318 Studies on Cn@C-nanotubulene hybrid systems (fullerenes in nanotubes) were continued.319 ± 321 The local electronic states of the C60@C-nanotubulene system were studied by scanning tun- nelling microscopy 319 and EELS.320 It was concluded 320 that fullerene ±NT interactions occur due to weak van der Waals forces.The methodological aspects of the use of EELS spectro- scopy for studies on the optical and electronic structures of Cn@C-nanotubulenes were analysed.321 Interesting results were obtained in theoretical simulation of the systems in question.322 For example, calculations for a set of hypothetical SiC-nanotubulenes by DFT made it possible to suggest that they can be synthesised by injection of holes.323 The type of the interband transition in these metastable tubular structures (i.e., whether direct or indirect transitions occur) depend on their chirality.The Estr of these NT were found to be smaller than the Estr of the corresponding carbon NT. Much attention is paid to the development of microscopic models of `hybrid' nanostructures. Based on calculations of the potential energies for systems containing spherical and ellipsoid fullerenes inside carbon NT, the `optimum' sizes for the fullerenes to be incorporated were discussed.324 The molecular dynamics method was used to calculate the energy states and the elastic and tribological parameters of fullerenes in (n,n)-C-nanotubulenes (where n=5 ± 10).325 A very interesting problem was considered by Liu et al.:326 they optimised model `hybrid' nanostructures in the form of carbon NT containing a monoatomic `fibre' of carbon atoms inside (along the axis).It was shown that for the NT with A L Ivanovskii diameters larger than*0.43 nm the fibre turns into a spiral along the NT axis. Non-empirical DFT calculations of the energy states of C60 fullerene in boron ± nitrogen NT were carried out.327 It was estimated that incorporation of fullerenes into (9,9)- and (10,10)-NT is exothermic, which implies that such `hybrid' struc- tures can actually be synthesised. It was shown that they would be semiconductors.The formation mechanism of yet another type of `hybrid' structure, viz., small-size ionic KI crystallites inside carbon NT, was studied by molecular dynamics methods.328 This study was partially supported by the Russian Foundation for Basic Research (Project No. 01-03-32513). References 1. S Iijima Nature (London) 354 56 (1991) 2. J W Mintmire, B I Dunlap, C T White Phys. Rev. Lett. 68 631 (1992) 3. M S Dresselhaus, G Dresselhaus, P Eklund (Eds) Science of Fuller- enes and Carbon Nanotubes (San-Diego, CA: Academic Press, 1996) 4. T W Ebbessen (Ed.) Carbon Nanotubes. Preparation and Properties (New York: CRC Press, 1996) 5. R Saito, G Dresselhaus, M S Dresselhaus Physical Properties of Carbon Nanotubes (London: Imperial College Press, 1998) 6.K Tanaka, T Yamabe, K Fuku (Eds) The Science and Technology of Carbon Nanotubes (Oxford: Elsevier, 1999) 7. A L Ivanovskii Kvantovaya Khimiya v Materialovedenii. Nanotubu- lyarnye Formy Veshchestva (Quantum Chemistry in Materials Tech- nology. Nanotubular Forms of Substance) (Ekaterinburg: Urals Branch of the Russian Academy of Sciences, 1999) 8. P J F Harris Carbon Nanotubes and Related Structures: New Materials for the Twenty-First Century (Cambridge: Cambridge University Press, 1999) 9. T W Ebbesen Ann. Rev. Mater. 24 235 (1994) 10. R Tenne Adv. Mater. 7 965 (1995) 11. T W Ebbesen Phys. Today 49 26 (1996) 12. A Zettl Adv. Mater. 8 443 (1996) 13. P M Ajayan Prog. Cryst. Growth Charact. Mater. 34 37 (1997) 14. P M Ajayan, T W Ebbesen Rep.Prog. Phys. 60 1025 (1997) 15. A V Eletskii Usp. Fiz. Nauk 167 945 (1997) a 16. Yu E Lozovik, A M Popov Usp. Fiz. Nauk 167 751 (1997) a 17. A Nikolic, V Radmilovic, M Simicic, D Koruga Adv. Mater. Proc. 282 83 (1998) 18. O M Kepp, P N D'yachkov Chem. Phys. Rep. 17 1179 (1998) 19. S Subramoney Adv. Mater. 10 1157 (1998) 20. B Coq, J M Planteix, V Brotons Appl. Catal. A 178 175 (1998) 21. A L Ivanovskii Usp. Khim. 68 119 (1999) [Russ. Chem. Rev. 68 103 (1999)] 22. M Terrones, W K Hsu, H W Kroto, D R M Walton Fullerene Relat. Struct. 199 198 (1999) 23. T Braun, A P Schubert, R N Kostoff Chem. Rev. 100 23 (2000) 24. E G Rakov Usp. Khim. 69 41 (2000) [Russ. Chem. Rev. 69 35 (2000)] 25. B P Tarasov, N F Goldshleger, A P Moravsky Usp.Khim. 70 149 (2001) [Russ. Chem. Rev. 70 131 (2001)] 26. P N D'yachkov Zh. Neorg. Khim. 46 93 (2001) b 27. E G Rakov Usp. Khim. 70 934 (2001) [Russ. Chem. Rev. 70 827 (2001)] 28. R Tenne Prog. Inorg. Chem. 50 269 (2001) 29. R Tenne, A Zettl Carbon Nanotubes 80 81 (2001) 30. M L Cohen Mater. Sci. Eng. C 15 1 (2001) 31. N F Goldshleger Fullerene Sci. Technol. 9 255 (2001) 32. E T Thostenson, Z F Ren, T W Chou Compos. Sci. Technol. 61 1899 (2001) 33. I V Zolotukhin Sorosovskii Obrazovatel'nyi Zh. (3) 111 (1999) 34. A V Eletskii Sorosovskii Obrazovatel'nyi Zh. (4) 86 (1999) 35. R Tenne, L Margulis,M Genut, G Hodes Nature (London) 360 444 (1992) 36. A Rubio, J L Corkill,M L Cohen Phys. Rev. B 49 5081 (1994) 37. X Blase X,A Rubio, S G Louie,M L Cohen Eur.Lett. 28 335 (1994) 38. Y Miyamoto, A Rubio, S G Louie Phys. Rev. B 50 18360 (1994) 39. N G Copra, R J Luyken, K Cherrey, V H Crespi,M L Cohen, S G Louie Science 269 966 (1995)Non-carbon nanotubes: synthesis and simulation 40. A Loiseau, F Willaime, N Demoncy, G Hug, H Pascard Phys. Rev. Lett. 76 4737 (1996) 41. A Loiseau, F Willaime, N Demoncy, N Schramchenko, G Hug, C Colliex, H Pascard Carbon 36 743 (1998) 42. Y Saito,M Maida J. Phys. Chem. A 103 1291 (1999) 43. Y Saito,M Maida, T Matsumoto Jpn. J. Appl. Phys. 38 159 (1999) 44. J Cumings, A Zettl Chem. Phys. Lett. 316 211 (2000) 45. Y Schimizu, Y Moriyoshi, S Komatsu, T Ikegami, T Ishigaki, T Sato, Y Bando Thin Solid Films 316 178 (1998) 46. D Golberg, Y Bando, M Eremets,M Takemura, K Kurashima, H Yusa Appl.Phys. Lett. 69 2045 (1996) 47. D P Yu, X C Sun, C S Lee, I Bello, S T Lee, H D Gu, K M Leung, G W Zhou, Z F Dong, Z Zhang Appl. Phys. Lett. 72 1966 (1997) 48. R S Lee, J Gavillett, M L de la Chapelle, A Loiseau, J-L Cochon, D Pigache, J Thibault, F Willaime Phys. Rev. B 64 1405 (2001) 49. D Golberg, Y Bando, K Kurashima, T Sasaki Carbon 37 293 (1999) 50. V I Vereshchagin,M A Sergeev, B S Semukhin, Y V Borodin Ref. Ind. Ceramics 41 440 (2001) 51. MTerrones, AMBenito, C Manteca-Diego, WK Hsu, O I Osman, J P Hare, D G Reid, H Terrones, A K Cheetham, K Prassides, H W Kroto, D R M Walton Chem. Phys. Lett. 257 576 (1996) 52. Ph Kohler-Redlich, M Terrones, C Manteca-Diego, W K Hsu, H Terrones, M Ruhle, H W Kroto, D R M Walton Chem.Phys. Lett. 310 459 (1999) 53. R Sen, B C Satishkumar, A Govindaraj, K R Harikumar, G Raina, J-P Zhang, A K Cheetham, C N R Rao Chem. Phys. Lett. 287 671 (1998) 54. B C Satishkumar, A Govindaraj, K R Harikumar, J-P Zhang, A K Cheetham, C N R Rao Chem. Phys. Lett. 300 473 (1999) 55. M Terrones, N Grobert, J Olivares, J P Zhang, H Terrones, K Kondratos,W K Hsu, J P Hare, P D Townsend, K Prassides, A K Cheetham, H W Kroto, D R M Walton Nature (London) 388 52 (1997) 56. RZ Ma,YBando, T Sato,KKurashima Chem. Mater. 13 2965 (2001) 57. R Z Ma, Y Bando, T Sato Chem. Phys. Lett. 337 61 (2001) 58. W Han, Y Bando, K Kurashima, T Sato Appl. Phys. Lett. 73 3085 (1998) 59. W Han, Y Bando, K Kurashima, T Sato Chem.Phys. Lett. 299 368 (1999) 60. D Golberg, Y Bando, W Han, K Kurashima, T Sato Chem. Phys. Lett. 308 337 (1999) 61. W Han, Y Bando, K Kurashima, T Sato Jpn. J. Appl. Phys. 2, Lett. 38 L755 (1999) 62. D Golberg, W Han, Y Bando, L Bourgeois, K Kurashima, T Sato J. Appl. Phys. 86 2364 (1999) 63. D Golberg, Y Bando, K Kurashima, T Sato Chem. Phys. Lett. 323 185 (2000) 64. D Golberg, Y Bando, L Bourgeois,K Kurashima, T Sato Carbon 38 2017 (2000) 65. D Golberg, Y Bando, K Kurashima, T Sato Diam. Relat. Mater. 10 63 (2001) 66. D Golberg, Y Bando, K Kurashima, T Sato Solid State Commun. 116 1 (2000) 67. W Han, J Cumings, X S Huang, K Bradley, A Zettl Chem. Phys. Lett. 346 368 (2001) 68. O Stefan, Y Bando, A Loiseau, F Willaime, N Shramachenko, T Tamiya, T Sato Appl.Phys. A 67 107 (1998) 69. BWSmith,MMonthioux,DE Luzzi Chem. Phys. Lett. 315 31 (1999) 70. K Hirahara, K Suenaga, S Bandow, H Kato, T Okazaki, H Shinohara, S Iijima Phys. Rev. Lett. 85 5384 (2000) 71. O Louchev, Y Sato Appl. Phys. Lett. 74 194 (1999) 72. D Golberg, Y Bando, K Kurashima, T Sato Scr. Mater. 44 1561 (2001) 73. D Golberg, Y Bando, L Bourgeois, K Kurashima, T Sato Appl. Phys. Lett. 77 1979 (2000) 74. Y Bando, K Ogawa, D Golberg Chem. Phys. Lett. 347 349 (2001) 75. C Pham-Huu, N Keller, G Ehret,M J Ledoux J. Catal. 200 400 (2001) 76. T S Bartnitskaya, G S Oleinik, A V Pokropivnyi, V V Pokropivnyi Pis'ma Zh. Eksp. Teor. Fiz. 69 145 (1999) c 77. M Terauchi, M Tanaka, K Suzuki, A Ogino, K Kimura Chem.Phys. Lett. 324 359 (2000) 78. L Bourgeois, Y Bando, T Sato J. Phys. D 33 1902 (2000) 191 79. N V Podberezskaya, S A Magrill, N V Pervukhin, S V Borisov Zh. Strukt. Khim. 42 783 (2001) d 80. L Margulis, G Salitra, R Tenne,M Talianker Nature (London) 365 113 (1993) 81. M Remskar, Z Skraba, F Cleton, R Sanjines, F Levy Appl. Phys. Lett. 69 351 (1996) 82. M Remskar, Z Skraba, F Cleton, R Sanjines, F Levy Surf. Rev. Lett. 5 423 (1998) 83. D H Galvan, R Rangel, G Alonso Fullerene Sci. Technol. 6 1025 (1998) 84. D H Galvan, R Rangel, E Adem Fullerene Sci. Technol. 7 805 (1999) 85. E B Mackie, D H Galvan, E Adem, S Talapatra, G Yang, A D Migone Adv. Mater. 12 495 (2000) 86. Y Feldman, E Wasserman,D J Srolovitz,R Tenne Science 267 222 (1995) 87.Y Feldman, G L Fray,M Homyonfer, V Lyakhovitskaya, L Margulis, H Cohen, G Hodes, J L Hutchinson, R Tenne J. Am. Chem. Soc. 118 5362 (1996) 88. Y Q Zhu, W K Hsu, N Grobert, B H Chang,M Terrones, H Terrones, H W Kroto, D R M Walton, B Q Wei Chem. Mater. 12 1190 (2000) 89. L Rapoport, Y Bilik, Y Feldman,M Homoyonfer, S R Cohen, R Tenne Nature (London) 387 791 (1997) 90. R Sen, A Govindaraj, K Suenaga, S Suzuki, H Kataura, S Iijima, Y Achiba Chem. Phys. Lett. 340 242 (2001) 91. M Chhowalla, G A J Amaratuga Nature (London) 407 164 (2000) 92. A Rothschild, J Sloan, A P E York, ML H Green, J L Hutchison, R Tenne Chem. Commun. 363 (1999) 93. G L Fray, S Elani,M Homoyonfer, Y Feldman, R Tenne Phys. Rev.B 57 6666 (1998) 94. M Remskar, Z Skraba,M Regula, C Ballif, R Sanjines, F Levy Adv. Mater. 10 246 (1998) 95. A Rothschild, R Popovitz-Biro, O Lourie, R Tenne J. Phys. Chem. B 104 8976 (2000) 96. Y Q Zhu, W K Hsu, H Terrones, N Grobert, B H Chang, MTerrones, B Q Wei, HWKroto, D RMWalton, C B Boothroyd, I Kinlock, G Z Chen, A H Windle, D J Fray J. Mater. Chem. 10 2570 (2000) 97. M Remskar, Z Skraba, R Sanjines, F Levy Appl. Phys. Lett. 74 3633 (1999) 98. O Tal,M Remskar, R Tenne, G Haase Chem. Phys. Lett. 344 434 (2001) 99. A Rothschild, S R Cohen, R Tenne Appl. Phys. Lett. 75 4025 (1999) 100. A Rothschild, J Sloan, R Tenne J. Am. Chem. Soc. 122 5169 (2000) 101. E B Mackie, D H Galvan, A D Migone Adsorbtion ± J. Intern. Adsorb. Soc.6 169 (2000) 102. D G Galvan, R Rangel, E Adem Fullerene Sci. Technol. 8 9 (2000) 103. D G Galvan, J H Kim, M B Maple, E Adem Fullerene Sci. Technol. 9 225 (2001) 104. D G Galvan, R Rangel, E Adem Fullerene Sci. Technol. 7 421 (1999) 105. D G Galvan, J H Kim, M B Maple, M Avalos-Borja, E Adem Fullerene Sci. Technol. 8 143 (2000) 106. E Flores, A Tlahuice, E Adem, D H Galvan Fullerene Sci. Technol. 9 9 (2001) 107. M Nath, C N R Rao Chem. Commun. 2236 (2001) 108. Y Jiang, Y Qu, S W Yuan, B Xie, S Y Zhang, Y T Qian J. Mater. Res. 16 2805 (2001) 109. Y F Liu, J H Zeng, W H Zhang, W C Yu, Y T Qian, J B Cao, W Q Zhang J. Mater. Res. 16 361 (2001) 110. D M Antonelli, J Y Ying Chem. Mater. 8 874 (1996) 111. D M Antonelli, A Nakamura, J Y Ying Inorg.Chem. 35 3126 (1996) 112. M S Wong, J Y Ying Chem. Mater. 10 2067 (1998) 113. V Luca, J M Hook Chem. Mater. 9 2731 (1997) 114. P Liu, I L Moudrakovski, J Liu, A Sayari Chem. Mater. 9 2513 (1997) 115. T Chirayil, P Y Zavalij, M S Whittingham Chem. Mater. 10 2629 (1997) 116. A Ayral, C Guizard Mater. Trans. 42 1641 (2001) 117. S Nakade, S Kambe, T Kitamura, Y Wada, S Yanagida J. Phys. Chem. B 105 9150 (2001)192 118. M Kruk, M Jaroniec Chem. Mater. 13 3169 (2001) 119. J van de Lagemaat, A J Frank J. Phys. Chem. B 105 11194 (2001) 120. H Imai, H Hirashima J. Am. Ceram. Soc. 82 2301 (1999) 121. H Imai,M Matsuda, K Shimizu, H Hirashima, N Negishi J. Mater. Chem. 10 2005 (2000) 122. M M Yusuf, H Imai, H Hirashima J. Non-Cryst. Solids 285 90 (2001) 123.H Hirashima, H Imai, V Balek J. Non-Cryst. Solids 285 96 (2001) 124. P M Ajayan, O Stephan, P Redlich, C Colliex Nature (London) 375 564 (1995) 125. D Urgate, T Stockli, J M Bonard, A Chatelain, W A De Heer Appl. Phys. A 67 101 (1998) 126. F Krumeich, H-J Muhr, M Niederberger, F Bieri, B Schnyder, R Nesper J. Am. Chem. Soc. 121 8324 (1999) 127. F Krumeich, H-J Muhr, M Niederberger, F Bieri, R Nesper Z. Anorg. Allg. Chem. 626 2208 (2000) 128. E Muller, F Krumeich Ultramicroscopy 84 143 (2000) 129. H-J Muhr, F Krumeich, U P Schonholzer, F Bieri, M Niederberger, L J Gauckler, R Nesper Adv. Mater. 12 231 (2000) 130. M Niederberger, H-J Muhr, F Krumeich, F Bieri, D Gunter, R Nesper Chem. Mater. 12 1995 (2000) 131. K S Pillai, F Krumeich, H-J Muhr, M Niederberger, R Nesper Solid State Ion. 141 ± 142 185 (2001) 132.T Kasuga,M Hiramatsu, A Hoson, T Sekino, K Niihara Adv. Mater. 11 1307 (1999) 133. H Imai, Y Takei, K Shimitsu,M Matsuda, H Hirashima J. Mater. Chem. 9 2971 (1999) 134. B D Summ, N I Ivanova Usp. Khim. 69 995 (2000) [Russ. Chem. Rev. 69 911 (2000)] 135. V L Volkov, G S Zakharova, I M Bondarenko Kserogeli Prostykh i Slozhnykh Polivanadatov (Xerogels of Simple and Complex Poly- vanadates) (Ekaterinburg: Urals Branch of the Russian Academy of Sciences, 2001) 136. Y R Hacohen, E Grunbaum, R Tenne, J Sloan, J L Hutchison Nature (London) 395 336 (1998) 137. B C Satishkumar, E M Vogl, A Govindaraj, C N R Rao J. Phys. D 29 3173 (1996) 138.B C Satishkumar, A Govindaraj, E M Vogl, L Basumallick, C N R Rao J. Mater. Res. 12 604 (1997) 139. B C Satishkumar, A Govindaraj, M Nath, C N R Rao J. Mater. Chem. 10 2115 (2000) 140. Q Q Li, S H Fan,W Q Han, C H Sun,W J Liang Jpn. J. Appl. Phys. 36 L501 (1997) 141. X H Chen, J T Xia, J C Peng, W Z Li, S S Xie Compos. Sci. Technol. 60 301 (2000) 142. T Seeger, P Redlich, N Grobert, M Terrones, D R M Walton, H W Kroto, M Ruhle Chem. Phys. Lett. 339 41 (2001) 143. T W Ebbesen J. Phys. Chem. Solids 57 951 (1996) 144. T Nakajiama, S Kasamatsu, Y Matsuo Eur. J. Solid State Inorg. Chem. 33 831 (1996) 145. E T Mickelson, C B Huffman, A G Rinzler, R E Smalley, R H Hauge, J L Margrave Chem. Phys. Lett. 296 188 (1998) 146. P J Boul, J Liu, E T Mickelson, C B Huffman, L M Ericson, I W Chiang,K A Smith,D T Colbert,R H Hauge, J L Margrave, R E Smalley Chem.Phys. Lett. 310 367 (1999) 147. L Dloczik, R Engelhardt, K Ernst, S Liechter, I Sieber, R Konen- kamp Appl. Phys. Lett. 78 3687 (2001) 148. S Peulon, D Lincot J. Electrochem. Soc. 145 864 (1998) 149. T Pauporte, D Lincot Appl. Phys. Lett. 75 3817 (1999) 150. R KoÈ nenkamp,K Boedecker,M C Lux-Steiner,M Poschenrieder, F Zenia, C Levy-Clement, S Wagner Appl. Phys. Lett. 77 2575 (2000) 151. O G Schmidt, K Eberl Nature (London) 410 168 (2001) 152. V M Osadchii, V Y Prinz JETP Lett. 72 312 (2000) 153. V Y Prinz, V A Seleznev, A K Gutakovsky, A V Chehovskiy, V V Preobrazhenskii,M A Putyato, T A Gavriliva Physica E 6 828 (2000) 154.S P Golod, V Y Prinz, V I Mashanov, A K Gutakovsky Semicond. Sci. Technol. 16 181 (2001) 155. O G Schmidt, N Y Jin-Phillipp Appl. Phys. Lett. 78 3310 (2001) 156. I Boustani Inter. J. Quant. Chem. 52 1081 (1994) 157. I Boustani Surf. Sci. 370 355 (1997) 158. I Boustani Phys. Rev. B 55 16 426 (1997) A L Ivanovskii 159. I Boustani, A Quandt Eur. Lett. 39 527 (1997) 160. I Boustani J. Solid State Chem. 133 182 (1997) 161. I Boustani, A Quandt Comput. Mater. Sci. 11 132 (1998) 162. M K Sabra, I Boustani Eur. Lett. 42 611 (1998) 163. I Boustani, A Quandt,A Rubio J. Solid State Chem. 154 269 (2000) 164. I Boustani, A Rubio, J A Alonso Chem. Phys. Lett. 311 21 (1999) 165. A Gindulyte,W N Lipscomb, L Massa Inorg. Chem. 37 6544 (1998) 166.A Ricca, C W Bauschlicher Chem. Phys. 208 233 (1996) 167. F L Gu, X Yang, A C Tang, H Jiao, P Schleyer J. Comput. Chem. 52 1081 (1998) 168. A L Ivanovskii, G P Shveikin Kvantovaya Khimiya v Materialove- denii. Bor ego Splavy i Soedineniya (Quantum Chemistry for Materials Technology. Boron, Its Alloys and Compounds) (Ekaterinburg: Izd. `Ekaterinburg', 1998) 169. U RoÈ thlisberger,W Andreoni,M Parrinello Phys. Rev. Lett. 72 665 (1994) 170. B I Yakobson, R E Smalley Am. Sci. 85 324 (1997) 171. S F Fagan, R J Baierle, R Mota, A J R da Silva, A Fazzio Phys. Rev. B 61 9994 (2000) 172. R J Baierle, S B Fagan, R Mota, A J R da Silva, A Fazzio Phys. Rev. B 64 5413 (2001) 173. G Seifert, Th Frauenheim, Th Kohler,H M Urbassek Phys.Status Solidi, B 228 393 (2001) 174. G Seifert, Th Kohler, H M Urbassek, E Hernandez, Th Frauenheim Phys. Rev. B 63 193409 (2001) 175. K H Janzon, H Schafer, A Weiss Z. Anorg. Allg. Chem. 372 87 (1970) 176. H J Wallbaum Naturwissenschaften 32 76 (1944) 177. G Vogg,M S Brandt, M Stutzmann Adv. Mater. 12 1278 (2000) 178. Z Hajnal, G Vogg, L J P Meyer, B Szucs,M S Brandt, Th Frauenheim Phys. Rev. B 64 3311 (2001) 179. G Seifert, Th Kohler, Z Hajnal, Th Frauenheim Solid State Commun. 119 653 (2001) 180. G Seifert, E Hernandez Chem. Phys. Lett. 318 355 (2000) 181. J-C Charlier, P Lambin, T W Ebbesen Phys. Rev. B 54 R8377 (1996) 182. S Veprek, H R Oswald Z. Anorg. Allg. Chemie 412 190 (1975) 183. I V Zaporotskova, Candidate Thesis in Physical and Mathematic Sciences, Volgograd State University, Volgograd, 1997 184.J Yi, J Bernholc Phys. Rev. B 47 1708 (1993) 185. D L Carroll, P Redlich, X Blase, J C Charlier, S Curran, P Ajayan, S Roth,M Ruhle Phys. Rev. Lett. 81 2332 (1998) 186. X Blase, J C Charlier, A de Vita, R Car, P Redlich, M Terrones, W K Hsu, H Terrones, D L Carroll, P M Ajayan Phys. Rev. Lett. 83 5078 (1999) 187. E Hernandez, P Ordejon, I Boustani, A Rubio, J A Alonso J. Chem. Phys. 113 3814 (2000) 188. WK Hsu, S Y Chu, E Minoz-Picone, J L Boldu, S Firth, P Franchi, B P Roberts, A Schilder, H Terrones, N Grobert, Y Q Zhu, M Terrones, M E McHenry, H W Kroto, D R M Walton Chem. Phys. Lett. 323 572 (2000) 189. B C Wang,M H Tsai, Y M Chou Synth. Met. 86 2379 (1997) 190.L Vaccarini, C Groze, L Henrard, E Hernandez, P Bernier, A Rubio Carbon 38 1681 (2000) 191. G Overney,W Zhong, D Tomanek Z. Phys. D 27 93 (1992) 192. B I Yakobson, C J Brabek, J Bernholc Phys. Rev. Lett. 76 2511 (1996) 193. M M Treacy, T W Ebbesen, J M Gibson Nature (London) 381 678 (1996) 194. N G Chopra, A Zettl Solid State Commun. 105 297 (1998) 195. M W Barsoum, P Kangutkar, A S D Wang Compos. Sci. Technol. 44 257 (1992) 196. C Marotzke Compos. Sci. Technol. 50 393 (1994) 197. X Blase, A De Vita, J C Charlier, R Car Phys. Rev. Lett. 80 1666 (1998) 198. J C Charlier, X Blase, A De Vita, R Car Appl. Phys. A 68 267 (1999) 199. F Jensen, H Toftlund Chem. Phys. Lett. 201 89 (1993) 200. M L Sun, Z Slanina, S L Lee Chem.Phys. Lett. 233 279 (1995) 201. I Silaghi-Dumitrescu, F Lara-Ochoa, P Bishof, I Haiduc J. Mol. Struct. (THEOCHEM) 367 47 (1996) 202. P W Fowler, T Heine, D Mitchell, R Schmidt, G Seifert J. Chem. Soc., Farday Trans. 92 2197 (1996)Non-carbon nanotubes: synthesis and simulation 203. G Seifert, P W Fowler, D Mitchell, D Porezag, T Frauenbheim Chem. Phys. Lett. 268 352 (1997) 204. P W Fowler, K M Rogers, G Seifert, M Terrones, H Terrones Chem. Phys. Lett. 299 359 (1999) 205. K M Rogers, P W Fowler, G Seifert Chem. Phys. Lett. 332 43 (2000) 206. T Hirano, T Oku, K Suganuma Duam. Relat. Struct. 9 625 (2000) 207. H F Bettinger, T Dumitrica, G E Scuseria, B I Yakobson Phys. Rev. B 65 1406 (2002) 208. A Y Liu,M L Cohen Science 245 841 (1989) 209.M L Cohen Phys. Rev. B 32 7988 (1985) 210. B L Korsunskii, V I Pepekin Usp. Khim. 66 1003 (1997) [Russ. Chem. Rev. 66 901 (1997)] 211. F Weich, J Winday, Th Franenheim Phys. Rev. Lett. 78 3326 (1997) 212. J M Bonard, R Kurt, C Klinke Chem. Phys. Lett. 343 21 (2001) 213. A Karimi, R Kurt Surf. Eng. 17 99 (2001) 214. R Kurt, A Karimi Thin Solid Films 377 163 (2000) 215. R Kurt, J M Bonard,A Karimi Diam. Relat. Mater. 10 1962 (2001) 216. R Kurt, J M Bonard, A Karimi Carbon 39 1723 (2001) 217. M Terrones, P Redlich, N Grobert, S Trasobares, W K Hsu, H Terrones, Y Q Zhu, J P Hare, C L Reeves, A K Cheetham, M Ruhle, H W Kroto, D R M Walton Adv. Mater. 11 655 (1999) 218. Y Guo,W A Goddart Chem. Phys. Lett. 237 72 (1995) 219. D M Teter, R J Hemley Science 271 53 (1996) 220.N R Widany, F Wiech, T Kohler, D Porezag Diam. Relat. Mater. 5 1031 (1996) 221. A Y Liu, R M Wentzcovitch Phys. Rev. B 50 10 362 (1994) 222. J Ortega, O F Sabkey Phys. Rev. B 51 2624 (1995) 223. J E Lowther Phys. Rev. B 57 5724 (1998) 224. A L Ivanovskii Russ. J. Inorg. Chem. 45 (Suppl. 1) S1 (2000) 225. Y Miyamoto,M L Cohen, S G Louie Solid State Commun. 102 605 (1997) 226. M Cote, J C Crossman, M L Cohen, S G Louie Phys. Rev. B 58 664 (1998) 227. M C dos Santos, F Alvarez Phys. Rev. B 58 13 918 (1998) 228. S Suoto,M Pickholz, M C dos Santos, F Alvarez Phys. Rev. B 57 2536 (1998) 229. E G Wang J. Am. Ceram. Soc. 85 105 (2002) 230. M Terrones, N Grobert, J P Zhang, J Olivares, H Terrones, W K Hsu, J P Hare,A K Cheetham,H W Kroto,D R M Walton Chem.Phys. Lett. 285 299 (1998) 231. R Sen, B C Satishkumar, S Govindaraj, K H Harikumar, M K Renganathan, C N R Rao J. Mater. Chem. 7 2335 (1997) 232. J Kong, J Cao, H J Dai, E Anderson Appl. Phys. Lett. 80 73 (2002) 233. Z W Pan, S S Xie, B H Chang, L F Sun,W Y Zhou, G Wang Chem. Phys. Lett. 298 97 (1998) 234. M Nath, B C Satishkumar, A Govindaraj, C P Vinod, C N R Rao Chem. Phys. Lett. 322 333 (2000) 235. R Kurt,C Klinke, J M Bonard,K Kern,A Karimi Carbon 39 2163 (2001) 236. R Czerw,M Terrones, J -C Charlier, X Blase, B Foley, R Kamalakaran, N Grobert, H Terrones, P M Ajayan, W Blau, D Tekeleab,M Ruhle, D L Carroll Nano Lett. 1 457 (2001) 237. H J Choi, J Ihm, S G Louie,M L Cohen Phys.Rev. Lett. 84 2917 (2000) 238. Y Miyamoto,M L Cohen, S G Louie Phys. Rev. B 52 14 971 (1995) 239. A Y Liu, R M Wentzcovitch,M L Cohen Phys. Rev. B 39 1760 (1989) 240. J P LaFemina J. Phys. Chem. 94 4346 (1990) 241. A Rubio, Y Miyamoto, X Blase,M L Cohen, S G Louie Phys. Rev. B 53 4023 (1996) 242. H Zhu, D J Klein,W A Seitz Inorg. Chem. 34 1377 (1995) 243. Y Miyamoto, A Rubio,M L Cohen, S G Louie Phys. Rev. B 50 4976 (1994) 244. X Blase, J C Charlier, A De Vita, R Car Appl. Phys. A 68 293 (1999) 245. X Blase Comput. Mater. Sci. 17 107 (2000) 246. S M Lee, Y H Lee, Y G Hwang, J Elsner, D Porezag, T Frauenheim Phys. Rev. B 60 7788 (1999) 247. M Cote,M L Cohen, D J Chadi Phys. Rev. B 58 R4277 (1998) 248. S Sawada, N Hamada Solid State Commun.83 917 (1992) 249. D H Oh, Y H Lee Phys. Rev. B 58 7407 (1998) 250. W Q Han, P Redlich, F Ernst,M Ruhle Appl. Phys. Lett. 76 652 (2000) 193 251. S M Lee, Y H Lee, Y G Hwang, C J Lee J. Korean Phys. Soc. 34 S253 (1999) 252. S M Lee, Y H Lee, Y G Hwang,, J Elsner, D Porezag, T Frauenheim MRS Internet J. Nitride Semicond. 4 U568 (1999) 253. G Seifert, H Terrones,M Terrones, G Jungnickel, Th Frauenheim Phys. Rev. Lett. 85 146 (2000) 254. G Seifert, H Terrones,M Terrones, G Jungnickel, Th Frauenheim Solid State Commun. 114 245 (2000) 255. K Kobayashi, J Yamauchi Phys. Rev. B 51 17085 (1995) 256. J D Fuhr, J O Sofo, A Saul Phys. Rev. B 60 8343 (1999) 257. P A Parilla, A C Dillon, K M Jones, G Riker, D L Schulz, D S Ginley, M J Heben Nature (London) 397 114 (1999) 258.I Milosevoc, T Vukovic,M Damnjanovic, B Nikplic Eur. Phys. J. B 17 707 (2000) 259. Y Q Zhu, W K Hsu, S Firth,M Terrones, R J H Clark, H W Kroto, D R M Walton Chem. Phys. Lett. 342 15 (2001) 260. G Seifert, H Terrones,M Terrones, T Frauenheim Solid State Commun. 115 635 (2000) 261. N J Doran Physica B 99 227 (1980) 262. A Nadar, A Briggs, Y Gotoh Solid State Commun. 101 149 (1997) 263. Y Q Zhu, W K Hsu, H W Kroto, D R M Walton Chem. Commun. 2184 (2001) 264. G V Samsonov, T I Serebryakova, V A Neronov Boridy. Atomizdat Moscow 1975 265. Yu B Kuz'ma Kristallokhimiya Boridov (Crystal Chemistry of Borides) (L'vov: Izd. `Vishcha Shkola', 1983) 266. J Nagamatsu, N Nakagawa, T Muranaka, Y Zenitani, J Akimitsu Nature (London) 410 63 (2001) 267.A L Ivanovskii Usp. Khim. 70 811 (2001) [Russ. Chem. Rev. 70 717 (2001)] 268. C Buzea, T Yamashita Supercond. Sci. Technol. 14 R115 (2001) 269. L A Chernozatonskii JETP Lett. 74 335 (2001) 270. W B Pearson The Crystal Chemistry and Physics of Metals and Alloys (New York; London; Sidney; Toronto: Wiley-Interscience, 1972) 271. J Kortus, I I Mazin, K D Belaschenko, V P Antropov, L L Boyer Phys. Rev. Lett. 86 4656 (2001) 272. J M An,W E Pickett Phys. Rev. Lett. 86 4366 (2001) 273. N I Medvedeva, Yu E Medvedeva, A L Ivanovskii, V G Zubkov, A Friman Pis'ma Zh. Eksp. Teor. Fiz. 73 378 (2001) c 274. E Wilson Chem. Eng. News 79 6 (2001) 275. A Quandt, A Y Liu, I Boustani Phys. Rev. B 64 5422 (2001) 276. L A Chernozatonskii Khim. Fiz. 16 78 (1997) e 277. L A Chernozatonskii, Ya K Shimkus, I V Stankevich Phys. Lett. A 240 105 (1998) 278. L A Chernozatonskii, E G Gal'pern, I V Stankevich, Ya K Shimkus Carbon 37 117 (1999) 279. X Blase, J C Charlier, A De Vita, R Car Appl. Phys. Lett. 70 197 (1997) 280. P E Lammert, V H Crespi, A Rubio Phys. Rev. Lett. 87 6402 (2001) 281. A Rubio, Y Miyamoto, X Blase,M L Cohen, S G Louie Phys. Rev. B 53 4023 (1996) 282. E Hernandez, C Goze, P Bernier, A Rubio Appl. Phys. A 68 287 (1999) 283. R A Jishi, C T White, J W Mintmire J. Phys. Chem. B 102 1568 (1998) 284. A A Sofronov, V V Ivanovskaya, Yu N Makurin, A L Ivanovskii Chem. Phys. Lett. 351 35 (2002) 285. V V Ivanovskaya, A A Sofronov, A L Ivanovskii Teor. Eksp. Khim. 37 331 (2001) f 286. H J Goldschmidt Interstitial Alloys (London: Butterworth, 1967) 287. B C Guo, K P Kerns, A W Castleman Science 255 1411 (1992) 288. B C Guo, S Wei, J Purnell, S Buzza, A W Castleman Science 256 515 (1992) 289. T Pradeep, P T Manoharan Curr. Sci. 68 1017 (1995) 290. H T Deng, K P Kern, A W Castleman J. Am. Chem. Soc. 118 446 (1996) 291. R Selvan, T Pradeep Curr. Sci. 74 665 (1998) 292. M Rohmer,M Bernard, J-M Poblet Chem. Rev. 10 495 (2000) 293. A A Sofronov, V V Ivanovskaya, Yu N Makurin, A L Ivanovskii Koord. Khim. 27 857 (2001) g 294. C C Tang, S S Fan, P Li, Y H Liu, H Y Dang Mater. Lett. 51 315 (2001)A L Ivanovskii 194 295. R Z Ma, Y Bando, T Sato, K Kurashima Chem. Phys. Lett. 350 434 (2001) 296. J D Guo, C Y Zhi, X D Bai, E G Wang Appl. Phys. Lett. 80 124 (2002) 297. R Z Ma, Y Bando, T Sato Chem. Phys. Lett. 350 1 (2001) 298. M Nath, K Mukhopadhaya, C N R Rao Chem. Phys. Lett. 352 163 (2002) 299. Y Y Peng, Z Y Meng, C Zhong, J Lu, J Q Xu, S Y Zhang, Y T Qian New J. Chem. 25 1359 (2001) 300. R L D Whitby, W K Hsu, P Watts, H W Kroto, D R M Walton, C B Boothroyd Appl. Phys. Lett. 79 4574 (2001) 301. J Chen, N Kuriyama, H Yuan, H T Takeshita, T Sakai J. Am. Chem. Soc. 123 11813 (2001) 302. H B Chen, J D Liu, J Yi, G Wei, Y Z Lin, D W Liao Chin. Chem. Lett. 12 1139 (2001) 303. J C Bao, C Y Tie, Z Xu, Q F Zhou,D Shen, Q Ma Adv. Mater. 13 1631 (2001) 304. J Y Li, X L Chen, Z Y Qiao, Y G Cao, H Li J. Mater. Sci. Lett. 20 1987 (2001) 305. F Bieri, F Krumeich, H J Muhr, R Nesper Helv. Chim. Acta 84 3015 (2001) 306. A Dobley, K Ngala, S F Yang, P Y Zavalij, M S Whittingham Chem. Mater. 13 4382 (2001) 307. F Miyaji, Y Tatematsu, Y Suyama J. Ceram. Soc. Jpn. 109 924 (2001) 308. G H Du, Q Chen, P C Che, Z Y Yuan, L M Peng Appl. Phys. Lett. 79 3702 (2001) 309. D Gong, C A Grimes, O K Varghese, W C Hsu, R S Singh, Z Chen, E C Dickey J. Mater. Res. 16 3331 (2001) 310. X C Wu,W H Song, B Zhao, Y P Sun, J J Du Chem. Phys. Lett. 349 210 (2001) 311. B Cheng, E T Samulski J. Mater. Chem. 11 2901 (2001) 312. G Maurin, F Henn, B Simon, J F Colomer, J B Nagy Nano Lett. 1 75 (2001) 313. H Shimoda, B Gao, X P Tang, A Kleinhammers, L Fleming, Y Wu, O Zhou Phys. Rev. Lett. 88 5502 (2002) 314. N Bendiab, E Anglaret, J L Bantigies, A Zahab, J L Sauvajol, P Petit, C Mathis, S Lefrant Phys. Rev. B 642 5424 (2001) 315. H Murakami, M Sano J. Phys. Soc. Jpn. 71 125 (2002) 316. F Banhart, N Grobert,M Terrones, J C Charlier, P M Ajayan Int. J. Mod. Phys. B 15 4037 (2001) 317. T Seeger, T KoÈ hler, T Frauenheim, N Grobert,M RuÈ hle, M Terrones, G Seifert Chem. Commun. 34 (2002) 318. J S Sakamoto, B Dunn J. Electrochem. Soc. 149 A26 (2002) 319. D J Hornbaker, S J Kahng, S Misra, B W Smith, A T Johnson, E J Mele, D E Luzzi, A Yazdani Science 295 828 (2002) 320. X Liu, T Pichler, A Knupfer,M S Golden, J Fink, H Kataura, Y Achiba, K Hirahara, S Iijima Phys. Rev. B 65 5419 (2002) 321. T Pichler New Diam. Frontier Carbon Technol. 11 375 (2001) 322. Y Miyamoto, B D Yu Appl. Phys. Lett. 80 586 (2002) 323. J M Schon, Ch Kloc, B Batlogg Nature (London) 406 702 (2000) 324. M Hodak, L A Girifalco Chem. Phys. Lett. 350 405 (2001) 325. D Qian, W K Liu, R S Ruoff J. Phys. Chem. B 105 10 753 (2001) 326. D Liu, J M Dong, X G Wan Chin. Phys. Lett. 19 98 (2002) 327. S Okada, S Saito, A Oshiyama Phys. Rev. B 642 1303 (2001) 328. M Wilson J. Chem. Phys. 116 3027 (2002) a�Physics-Uspekhi (Engl. Transl.) b�Russ. J. Inorg. Chem. (Engl. Transl.) c�J. Exp. Theor. Phys. Lett. (Engl. Transl.) d�Russ. J. Struct. Chem. (Engl. Transl.) e�Chem. Phys. Rep. (Engl. Transl.) f�Theor. Exp. Chem. (Engl. Transl.) g�Russ. J. Coord. Chem. (

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (3) 195 ± 221 (2002) Captodative aminoalkenes A Yu Rulev Contents I. Introduction II. Synthesis of captodative aminoalkenes III. Structures and spectroscopic characteristics of captodative aminoalkenes IV. Reactions of captodative aminoalkenes V. Methylidene- and arylmethylidenequinuclidin-3-ones VI. Conclusion Abstract. amino- captodative of synthesis the for Procedures Procedures for the synthesis of captodative amino- alkenes, 15 last the over primarily developed were which alkenes, which were developed primarily over the last 15 years, years, are and structural the of influence The systematised. are systematised. The influence of the structural and stereoelec- stereoelec- tronic and atom nitrogen the at both substituents the of effects tronic effects of the substituents both at the nitrogen atom and at at the double bond on the reactivity of these systems is considered. the double bond on the reactivity of these systems is considered.Particular stereoselectivity and regio- to given is emphasis Particular emphasis is given to regio- and stereoselectivity of of electrophilic, of reactions radical and nucleophilic electrophilic, nucleophilic and radical reactions of captodative captodative aminoalkenes of construction the of processes to as well as aminoalkenes as well as to processes of the construction of carbo- carbo- and aminoalkenes. these on based compounds heterocyclic and heterocyclic compounds based on these aminoalkenes. A special methylidene- of chemistry the to devoted is section special section is devoted to the chemistry of methylidene- and and arylmethylidenequinuclidin-3-ones.The bibliography includes arylmethylidenequinuclidin-3-ones. The bibliography includes 327 references 327 references. I. Introduction Two decades ago, Viehe 1± 4 proposed the concept of the capto- dative effect. Later on, the notion of captodative alkenes was introduced for compounds characterised by the simultaneous presence of electron-withdrawing (captive) and electron-donating (dative) substituents at the same carbon atom of the double bond. Alkenes containing the electron-donating and electron-withdraw- ing substituents in the vicinal position are termed push-pull. Among captodative alkenes, enamines containing the geminal electron-withdrawing group (EWG) attract the most attention.Since these alkene derivatives possess special chemical properties, they were regarded as a separate group of organic compounds. Recent studies demonstrated that cross-conjugated isomers exhibit specific and much more versatile reactivities as compared to fully conjugated push-pull aminoalkenes. Many captodative aminoalkenes have already found wide application. It will suffice to mention that these compounds are structural fragments of bioorganic molecules and serve as building blocks in the synthesis of various biologically active heterocycles and analogues of natural compounds. In the solution of some natural-science problems of the prebiotic chemistry and philo- sophic questions about the origin of life on Earth, much attention has been focused on the simplest representatives of captodative AYu RulevA E Favorskii Irkutsk Institute of Chemistry, Siberian Branch of the Russian Academy of Sciences, ul.Favorskogo 1, 664033 Irkutsk, Russian Federation. Fax (7-395) 239 60 46. Tel. (7-395) 246 29 11. E-mail: rulev@irioch.irk.ru Received 10 December 2001 Uspekhi Khimii 71 (3) 225 ± 254 (2002); translated by T N Safonova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n03ABEH000705 195 196 202 204 215 217 alkenes (see, for example, the studies 5± 9 devoted to 2-amino- acrylonitrile). However, while the chemistry of push-pull aminoalkenes 1 has been studied in sufficient detail and its various aspects were surveyed in a series of fundamental reviews,10 ± 16 captodative aminoalkenes 2, on the contrary, remained poorly studied until recent years.EWG EWG N N1a ± d 2a ± d EWG =CN (a), CO2R (b), C(O)R (c), CHO (d). The first effort to systematise the main procedures for the preparation of 2-aminocycloalk-2-enones and consider their most remarkable properties was made in the review,17 which covered the data published up to 1983. In the review 11 devoted to the chemistry of b-aminoenones, a special section was concerned with the data on the chemical behaviour of captodative aminoalke- nones (mainly, cyclic), which have been reported only in 30 scientific sources. In the last 15 years, some important aspects of the chemistry of captodative aminoalkenes were developed sub- stantially owing, primarily, to elaboration of simple and efficient procedures for the synthesis of such systems.The present review generalises the data on the chemistry of captodative amino- alkenes. The first section of the review is dedicated to procedures for the preparation of captodative alkenes containing an unsub- stituted, mono- or disubstituted amino group and one electron- withdrawing group (carbonyl, alkoxycarbonyl or cyano group) at the same centre. In the second section, the spectroscopic charac- teristics of captodative aminoalkenes are analysed, including the data on their structures. In the third section, the major types of transformations characteristic of these compounds are classified. Aspecial section is devoted to the chemistry of 2-methylidene- and 2-arylmethylidenequinuclidin-3-ones belonging to bicyclic capto- dative aminoalkenes, which do not exhibit properties of enamines due to their structural characteristic features.Quinoid structures, which formally contain the C=C(NR2)CO fragment, and acti- vated alkenes bearing the geminal acylamino group are beyond the scope of the present review. Data on some compounds of this type (in particular, dehydroamino acids) were surveyed in a number of publications (see, for example, the review 18).196 II. Synthesis of captodative aminoalkenes 1. Insertion of the double bond into systems containing geminal amino and electron-withdrawing groups a.Olefination of carbonyl compounds One of the classical ways for inserting a double bond into a molecule involves Wittig-type olefination of aldehydes and ketones. Various captodative aminoalkenes were successfully prepared according to this procedure. Costisella and Gross demonstrated that cyanophosphonate 3 reacted with aromatic or aliphatic aldehydes and ketones in the presence of a strong base (BuLi, THF;19 NaH, dioxane or DMSO20) to form 1-cyanoen- amines 2a in moderate yields. CN CN R1 B R1 O + NMe2 (EtO)2P NMe2 R2 O 3 R2 2a (22% ± 71%) O O R1=H, Me, Et, Pri, n-C6H13, Ph, 4-NO2C6H4, (EtO)2P(O)CH2, ; R2=H, Ph; B is a base. Later on, it was found that this reaction could be performed by electrolysis of phosphonates 3 on platinum or glassy-carbon electrodes,21 the carbanion being generated on the cathode.This carbanion reacts with the carbonyl compound to give finally the target alkenes 2a in satisfactory yields. In addition to the aminonitriles 3, the reactions were success- fully performed with esters. Thus methyl (dimethoxyphosphor- yl)piperidinoacetate (4) reacted with benzaldehyde to form unstable methyl 3-phenyl-2-piperidinoacrylate (5), which was transformed into b-ketophosphonate 6 without isolation.22 CO2Me (MeO)2P(O) PhCHO N NaH, THF 4 Ph Ph CH2P(O)(OMe)2 CO2Me O LiCH2P(O)(OMe)2 N N 5 6 (25%) The 1-cyanoenamines 2a were also prepared from amino- acetonitriles 7 and aliphatic (including a,b-unsaturated) or aromatic aldehydes by the Peterson reaction 23 ± 27 or condensa- tion in the presence of a strong base (KH, KOH±K2CO3).24, 28 Both procedures appeared to be extremely sensitive to the nature of the amino group in the starting substrate 7.In the case of dimethylamino-, diethylamino- or pyrrolidinoacetonitriles, the reactions either did not take place at all or proceeded in very low CN R3 1) R3R4CO 2) B NR1R2 R4 2a CN R1R2N 7 SiMe3 1) B 2) R3R4CO 1) B 2) Me3SiCl 2a CN R1R2N R1=Me, CH2=CHCH2, MeCH=CHCH2; R2=Ph, Bn; R3=H, Me, Prn, Pri, n-C5H11, Ph, 4-MeOC6H4, MeCH=CH, EtCH=CH, PhCH=CH, 2-thienyl, 1-naphthyl; R4=H, Me, Et, Ph; B=KH, LDA or KOH±K2CO3±Bu4NBr. A Yu Rulev yields.24 (N-Methylanilino)acetonitrile [7, NR1R2=N(Me)Ph] proved to be the only substrate, which was almost quantitatively transformed into the corresponding 1-cyanoenamine.Aldol condensation of a-aminoketones bearing the active methylene group with aromatic aldehydes afforded the corre- sponding captodative aminoalkenes 2c.29, 30 b. Elimination A wide range of captodative aminoalkenes were prepared by elimination reactions, which very readily proceed in the case of b-heterosubstituted ketones, aldehydes, nitriles and esters. Elimination of hydrogen halide from a-amino-b-haloketones or derivatives of carboxylic acids is a valuable approach, provided that the corresponding starting compounds are readily accessible. The idea of the use of haloaldehydes as building blocks in the synthesis of captodative aminoalkenes has long been known.More than 40 years ago, Temin 31 developed an original synthesis of a-morpholinoacrylonitrile from chloroacetaldehyde (8, R1=R2=H, X=Cl), which was generated in situ from dime- thylacetal.32, 33 This classical procedure (with small modifications) continues to be used for the preparation of the 1-cyanoenamines 2a.34 ± 36 CN B R1 R1 1) HNR3R4 2) KCN (NaCN) O NR3R4 R2 R2 X X 10 8 CN R1 NR3R4 R2 2a (39% ± 94%) CN B KCN, MeOH, D R1 R1 NR3 NHR3 R2 R2 Cl Cl 9 CN R1 NHR3 R2 2a (39% ± 94%) R1=H:R2=Prn, Pri, Ph; R1=Me: R2=Et, Ph; R1=R2=H, Me, Et, Cl; R1±R2=(CH2)4, (CH2)5, CH2CH=CHCH2; X=Cl, Br; NR N N 3R4=NH2, NMe2, NEt2, , ,NO, N(Me)Ph, N(Me)C6H11-cyclo; R4=H:R3=Me, Et, Pri, But, cyclo-C6H11, CH2Bn; B=Et3N or NaOH.De Kimpe et. al.37, 38 extended the scope of application of this method using a-chloroaldimines 9 as the starting compounds. The latter are prepared from aliphatic aldehydes by successive imina- tion with primary amines and chlorination with N-chlorosuccin- imide. The reactions of the aldimines 9 with KCN in methanol afford the 1-cyanoenamines 2a in high yields. It was assumed 37, 38 that the synthesis proceeded through 2-(N-alkylamino)-3-chloro- substituted nitriles as intermediates, which were readily dehydro- chlorinated under the reaction conditions. There are also other examples where the resulting 2-amino-3- halopropiononitriles 10 were subjected to dehydrohalogenation to give the corresponding 1-cyanoenamines 2a, including those with the unsubstituted,39 mono- 40 or disubstituted amino group.41, 42 For these reactions, Et3N32 ± 36, 39 ± 41 or NaOH31 are the reagents of choice.An elegant preparative procedure proposed recently for the synthesis of various 1-cyanoenamines 2a is based on the involve- ment of (N-methylanilino)phenylthioacetonitrile (11) in the tan- dem alkylation ± elimination process.43, 44 The reaction proceeds stereospecifically to form the E isomer, which is accounted for byCaptodative aminoalkenes anti-elimination of PhSH from intermediate 12. The latter was isolated and characterised by spectroscopic methods.43 CN ButOK, RCH2X PhS N(Me)Ph 11 SPh R R H CN 7PhSH Ph(Me)N CN N(Me)Ph 2a (61%± 83%) H 12 R=H, Me, Pr, Br(CH2)2, I(CH2)4, Ph, CH2=CH, PhCH=CH, MeCH=CH, BrCH=CH, PhSCH2CH=CH, PhSCH=CH; X=Cl, Br, I.An attempt to reduce ester 13 with Raney nickel led to unexpected elimination of PhSH giving rise to methyl 2-amino- 3-phenylacrylate (14).45 NH2 NH2 Ni/Ra Ph Ph CO2Me CO2Me 20 8C, 24 h PhS 13 14 (62%) In my opinion, the above-considered reactions are of consid- erable synthetic value. In addition, there are other examples of the synthesis of captodative aminoalkenes resulting in elimination of hydrogen halide,46 alkanethiol,47 amine,48 hydrogen cyanide,49 water 50 or alcohol.51, 52 Dehydrogenation of N,N-disubstituted 2-aminocycloalkanones is also worthy of mention. Thus 2-amino- cyclopent-2-enones were obtained in high yields by passing gas- eous 2-aminocyclopentanones over Pd/C.53 NR2 NR2 Pd/C (5%) O O (80% ± 85%) NR2=NEt2, , N N O.c. Rearrangements In some unsaturated compounds, double-bond migration readily occurs under the action of strong bases. The driving force for this process is the formation of a thermodynamically more stable isomer due to interaction between the p-system of the double bond and the electron-withdrawing group. Treatment of b,g-unsaturated a-aminonitriles 15 with strong bases (ButOK or MeONa) under mild conditions afforded the corresponding 1-cyanoenamines 2a.26, 27, 54 ± 56 It should be noted that the stereochemistry of the reaction product depends both on the conditions of its preparation and the structure of the starting substrate.In addition, thermally initiated isomerisation can take place in these systems, as, for example, upon distillation of the nonconjugated aminonitriles 15.34 CN CN R1 R1 B or D NR3R4 R2 NR3R4 R2 2a 15 R1=H:R2=Me, Pr, Ph; R1=R2=H; NR3R4=N(Me)Ph, N(All)Ph, N(Me)C6H4Br-2; B=ButOK, MeONa. Under the action of strong bases, the b,g-unsaturated N,N- disubstituted a-aminonitriles 15 produce ambident anions, which can undergo alkylation either at the a or g position.57 This property was used for the preparation of various 1-cyanoen- 197 amines 2a.54, 58 ± 63 It was established that regioselectivity of alkylation is determined by different factors, among which are the steric parameters of the electrophilic reagent, the amino group and the g-substituent in the substrate.58 The formation of the captodative aminoalkenes 2a is favoured by the presence of a bulky amino group in the substrate molecule.For example, alkylation of the nitriles 15 containing the N(Me)Ph or N(cyclo-C6H11)Ph groups always proceeds regiospecifically to form the 1-cyanoenamines 2a.61 The stereochemistry of addition depends also on the reaction conditions. Thus the reactions performed at low temperatures (778 8C) afforded predominantly or even exclusively the Z isomers of the compounds 2a.60, 61 CN R1 1) B 2) R5X NR3R4 R2 15 CN R1 CN R1 R5 R5 + NR3R4 R2 NR3R4 R2 2a R1=H, Me, Ph, MeCH=CH; R2=H, Me; NR3R4=N(Me)Ph, NMe2, N N , O ; R5=Me, Et, Pri, Bu, Bn, All, Br(CH2)3, Cl(CH2)3, 2-BrC6H4CH2; X=I, Br; B=LDA or ButOK.In some cases, the initially formed b,g-unsaturated a-amino- nitriles or esters also undergo double-bond migration to the a,b position with respect to the electron-withdrawing group.64 ± 67 For instance, the reaction of methyl 2-chloro-3-methylbut-3-enoate (16) with piperidine started with the attack on the allylic carbon atom to give finally the conjugated amino ester 17.65 It is reasonable to assume that intermediate 18 generated through the replacement of the chlorine atom is isomerised to the more stable alkene 17 under the action of the amine. NH CO2Me CO2Me CO2Me N N Cl 16 18 17 (70%) When developing a general procedure for the synthesis of esters of b,g-unsaturated a-amino acids using organozinc com- pounds, the French research team 66, 67 found that the reaction of aminoacetal 19 with divinylzinc afforded amino ester 20, which underwent isomerisation to give methyl 2-diethylaminobut-2- enoate.CO2Me CO2Me CO2Me (CH2 CH)2Zn Me MeO NEt2 NEt2 19 NEt2 (28%) 20 Prototropic isomerisation of b-allenic amino esters, ketones or nitriles 21 under the action of strong bases [MeONa ±DMSO,68 ButOK±HMPA (HMPA is hexamethylphosphoramide) 69 or ButOK±THF70 ] proceeded rapidly to give conjugated amino- alkenes 22 in high yields.68 ± 70 EWG R1 B EWG NR2 R1 2 R22 N 21 22 (48% ± 100%) R1=H, Me, Et, Pr, Bu, Ph; NR22 N =NMe2, NEt2, ,N ,NO; EWG =CO2Me, CO2Et, C(O)Me, CN.198 Generally, tautomeric transformations are not considered as preparative reactions.However, unlike simple enamines with unsubstituted or monosubstituted amino groups, which exist predominantly in the imino form, captodative aminoalkenes 2 are much more stable in the enamine form. This allows their isolation either in the individual form or as a tautomeric mixture in which the enamine form dominates. Examples of such syntheses of aminoalkenes containing the activating carbonyl,71, 72 alkoxy- carbonyl 73 or cyano group 74 are available in the literature. A series of captodative aminoalkenes were prepared from the corresponding substituted aziridines.45, 75 ± 77 In some cases, their formation was absolutely unexpected. For example, treatment of aziridinecarboxylate 23 with sodium azide in methanol gave rise to the unsaturated ester 14 instead of the expected methyl 2-amino-3- azido-3-phenylpropionate.45 NH2 HN NaN3, NH4Cl Ph MeOH, D, 18 h CO2Me Ph CO2Me 23 14 (25%) It was reported that 2-amino-3-methoxy-1,3-diphenylpropa- none was transformed into a-aminochalcone under the action of strong bases.78 The authors assumed that this reaction proceeded through intermediate aziridinoketones.However, the formation of unsaturated a-aminoketones has not subsequently been con- firmed.79 Recently, a-aminoenone 24 was synthesised by photooxida- tion of cyclopropanol 25.80 Photolysis was carried out in the presence of photosensitising agents, viz., 9,10-dicyanoanthracene or triphenylpyrylium tetrafluoroborate, for 0.5 ± 5.2 min.It was noted that the enone 24 decomposed when the reaction time was increased. O Ph HO Ph Ph 3O2 , hn N N Ph O O 25 24 (26% ± 28%) 2. Insertion of the amino group into the conjugated C=C7EWG system The reaction of acrylonitrile with formaldehyde dimethylhydra- zone produced 2-dimethylaminopropenonitrile regardless of the reaction conditions.81 CN CN MeCN, 100 8C, 10 h N + NMe2 NMe2 Apparently, this is the only example of the formal synthesis of captodative aminoalkene from unsubstituted activated alkene. a. Nucleophilic substitution Nucleophilic substitution of a particular leaving group is most commonly used for the insertion of the amino group into the conjugated C=C7EWG fragment.Although the replacement of the functional group X in the conjugated C=C(X)EWG system by the amino group seems at first glance to be the shortest pathway to captodative amino- alkenes, the reactions of N-nucleophiles with haloalkenes acti- vated with the geminal electron-withdrawing group are complicated processes and do not necessarily afford the desired products. The results of these studies were critically analysed in the review.82 All the aforesaid is clearly exemplified by the data from the recent study,83 which demonstrated that the results of the reac- tions of primary amines with bromo(cycloalkylidene)acetates 26a,b depend substantially on the reaction conditions, the nature of the amine and the size of ring in the starting substrate.Under superhigh pressure, these reactions afforded spiroaziridines 27a,b in high yields (and with high diastereoselectivity in the case of the derivatives 27b). Under the same conditions, amines containing an aromatic or bulky substituent at the nitrogen atom produced predominantly a-amino-b,g-unsaturated esters 28a,b. Thermal activation of the compounds 26a,b was accompanied by migration of the exocyclic double bond followed by the replacement of the allylic bromine atom to form exclusively the esters 28a,b. Only in the case of the derivatives 28a containing the five-membered ring, back migration of the endocyclic double bond was observed giving rise formally to ipso-substitution product 29. R2( )n R3NH2 Br CO2R1 26a,b n=0, R1=Et, R2 = H (a); n=1, R1=Me, R2=But (b); R3=Prn, Pri, Ph, Bn, PhCHMe.In the reactions with N,N- or N,O-binucleophiles, compounds 26a,b underwent a cascade of transformations. Both upon thermal activation and under high pressure, the reactions involved double- bond migration and substitution of the halogen atom followed by condensation to form oxazinones (in the case of N,O-dinucleo- philes) or piperazinones (in the case of N,N-dinucleophiles).84 As in the case of reactions involving primary amines, the captodative system can be constructed only with the use of the bromo(cyclo- pentylidene)acetate 26a and aminoethanol derivatives.84 OH RHN26a R N 26b R N R=Me, Bn. Since the halide ion is one of the best nucleofuges, it is not surprising that this ion is most generally used as the leaving group.Nevertheless, examples of the replacement of other groups at the sp2-carbon atom are also available in the literature. These reac- tions often proceed very readily. A Yu Rulev R2( )n R1=Et 11 kbar N CO2Et R3 27a,b R2 n=0 D ( )n CO2Et R3HN CO2R1 R1=Et, R2=HR3HN 29 (17% ± 50%) 28a,b + + O R R N N CO2Et CO2Et O OH OH But O OCaptodative aminoalkenes In spite of the ambiguous results of the reactions of 2-amino- cycloalk-2-enones with N-nucleophiles,85 2-morpholinocyclohex- 2-enone (30) is recognised as a convenient substrate in trans- amination.17 O O O N NHAr ArNH2 30 Ar=Ph, 4-MeOC6H4, 4-BrC6H4, 2-HOC6H4, 2-HO2CC6H4. 2-Morpholinocyclohex-2-enone (30) was prepared by the reaction of 3-nitrocyclohex-2-enone with morpholine, isomeric b-aminoketone being obtained as the minor product.86 O O O O O HN N + N 30 NO2 O Refluxing of an equimolar mixture of 2-(p-tolylsulfonyloxy)- tropone (31) and azacrown ether 32 in benzene in the presence of pyridine afforded the corresponding substitution product.87 O O O O O O OTs O PhH, Py N O + O O NH32 31 (85%) Competitive ipso- and tele-substitution of various leaving groups in other troponoid systems has been well studied as well.88 ± 90 For example, isocolchicides 33a ± f readily reacted with piperidine, the ratio between the ipso (34) and tele isomers (35) being determined by the nature of the leaving group X.The reactions of the compounds 33a ± c afforded exclusively the isomers 34, whereas the compounds 33d ± f produced the tele isomers 35 in 10%± 25% yields. In the case of colchicides containing a substituent at position 10, only ipso-substitution was observed regardless of the nature of the leaving group.90 MeO NH MeO NHAc MeO X O 33a ± f MeO MeO MeO MeO NHAc NHAc + MeO MeO N N O O 35 34 X=Cl (a), SOMe (b), SMe (c), OMe (d), F (e), OTs (f). Finally, it should be noted that none of the attempts to replace the alkoxy group in a-alkoxyacroleins afforded the a-formyl- substituted enamines 2d91 although the reactions of amines with 3-alkoxy- or 3-phenoxypropenals are accepted to be a procedure of choice for the synthesis of their push-pull isomers 1d.12, 92, 93 199 b.Synthesis of captodative aminoalkenes bearing the unsubstituted amino group Undoubtedly, captodative aminoalkenes containing an unsubsti- tuted amino group are of particular interest. The classical approach to their synthesis is based on reduction of the corre- sponding nitro and azido derivatives. It was recommended to use hydrogen on a metal catalyst,94 hydrogen sulfide 94 or aluminium amalgam 94 ± 96 for reduction of the azido group. In addition, a-amino-a,b-unsaturated ketones and esters were prepared by electrolysis of solutions of ene azides activated with geminal carbonyl or ester groups.97 ± 100 Refluxing of azidoesters in toluene in the presence of catalytic amounts of iodine also afforded esters of unsaturated a-amino acids.18, 101, 102 EWG EWG R R NH2 N3 (15% ± 88%) EWG =CO2Me, CO2Et, MeCO, 4-MeC6H4CO, 4-ClC6H4CO; R=Me, Et, Prn, Pri, Ph, PhCH=CH, 4-MeC6H4, 4-HOC6H4, 4-MeOC6H4, 4-ClC6H4, 2,4,6-Me3C6H2, 2-BnC6H4, 2-MeC6H4, 2-PriC6H4, 2-furyl, 2-thienyl, 3-thienyl, 3-indolyl.An unusual synthesis of a-aminoenone 36 from a-bromo- ketone 37 and sodium azide has been reported.71 In the first step, the bromine atom was replaced to give azidoketone 38. Subse- quent elimination of N2 afforded imino ketone, which underwent rapid isomerisation into the unstable alkene 36. O O O NaN3, Et3N Ph Ph Ph 7N2 PhMe, D Br NH2 36 (35%) N3 38 37 Nitroalkenes are readily reduced with aluminium amalgam to the corresponding amino derivatives.103 ± 107 However, this proce- dure allows the preparation only of esters of b,b-disubstituted a-amino acids 39.Reduction of b-methoxy-b-alkyl- or b-alkyl- substituted substrates gave rise predominantly to hydroxyimino esters 40.104, 107 Me CO2R3 R1=R2=Me Me NH2 R1 Al ±Hg 39 (8% ± 62%) CO2R3 R2 R2 NO2 R1=Alk, R2=H, OMe CO2R3 Alk NOH 40 (44% ± 49%) R1=Me, Et, Prn, Pri; R2=H, Me, MeO; R3=Me, Et, Bun. Careful removal of various protective groups also made it possible to prepare the captodative aminoalkenes 2a ± c contain- ing the unsubstituted amino group (see, for example, Refs 108 ± 112). Procedures were developed for the synthesis of captodative aminoalkenes containing the unsubstituted amino group along with the keto, ester or cyano group.The exceptions are the formyl derivatives 2d. Recent attempts to synthesise these compounds based on 2-haloalk-2-enals and hexamethyldisilazane or (diphe- nylmethylidene)amine failed.113 Admittedly, the question about the possibility of the existence of the aminoalkenes 2d containing the unsubstituted amino group as well as of all aminoaldehydes with the unprotected amino group 114, 115 remains open.200 3. Simultaneous insertion of the double bond and the amino group a. Syntheses based on diketones and ketoesters One of the most general methods for the synthesis of captodative aminoalkenes is based on condensation of primary and secondary amines with readily enolysable 1,2-dicarbonyl compounds. As a rule, the reactions are carried out at room temperature or on refluxing of equimolar amounts of the reagents in an appropriate solvent (benzene, toluene, more rarely, chloroform or alcohol) in the presence or absence of a base or acid catalyst.The liberated water is removed either by azeotropic distillation or using molec- ular sieves. Examples of the syntheses of 2-aminocycloalk-2-enones con- taining the disubstituted amino group were given in the review 17 and in a number of recent publications.116 ± 119 The use of chiral pyrrolidines in reactions with five ± eight-membered cyclic 1,2- diketones allowed the preparation of optically active a-amino- enones.120 Primary amines can also produce a-aminoenones in the reactions with cyclic diketones.121 ± 124 Along with usual alkyl- and arylalkylamines [BnNH2, cyclo-C6H11NH2, Ar(CH2)2NH2], esters of some a-amino acids, such as glycine, alanine and tyrosine, were also successfully used in these reactions.125, 126 O O HNR1R2 ( )n ( )n n=0, 1, 2, 3; R1=H, Alk, All, Ar; R2=Alk, All, CH2CO2Et, CH(Alk)CO2Et.Since acyclic analogues of aminoketones are much less stable, the structures of the starting substrate and the reaction conditions are of particular importance in their synthesis. Acyclic a-amino- ketones containing the internal or terminal double bond were prepared from a-diketones according to a classical procedure,127 the yields being at most 30%. All attempts to purify the products led to their resinification.128, 129 Conventional procedures for the preparation of enamines containing the geminal alkoxycarbonyl group offered consider- able promise.It appeared that prolonged refluxing of 2,2-diethoxyethyl- amine with ethyl (3-chloro-2-nitrophenyl)pyruvate in toluene actually afforded the corresponding enamine in quantitative yield.130 Taking into account the high sensitivity of alkyl pyru- vates both to acidic and basic reagents, Arnold 131 improved this procedure. He proposed that TiCl4, which is traditionally used in the synthesis of enamines, should be changed for softer AsCl3. In essence, this modification allowed the first synthesis of the terminal aminoalkenes 2b (R1=H). In spite of low stability, all products were prepared in the pure form in high yields.EWG NHR2R3 R1 Cat O NR2R3 EWG R1 O N H MeCO O N EtCO H O N Me EtCO O N Ph PhCO H NMe2 CO2Me O N H CO2Me O NR1R2EWG R1 NR2R3 2b,c Yield (%) Ref. Catalyst 128, 129 TiCl4 7 129 30 TiCl4 129 25 TiCl4 129 TiCl4 7 131 82 AsCl3 131 73 AsCl3 A Yu Rulev EWG R1 Yield (%) Ref. Catalyst NR2R3 N 131 75 H AsCl3 CO2Me 131 68 H AsCl3 NMe2 CO2But OMe 133 100 MeCO H AsCl3 N The sole exception is the pyrrolidine derivative, which was not isolated by the author. It should be noted that methyl a-piper- idinoacrylate has been obtained previously from the correspond- ing chlorine-substituted ester and piperidine 132 in very low yield (5% ± 10%).Later on, the procedure proposed by Arnold 131 was successfully used for the synthesis of optically active acyclic a-aminoenone, viz., 3-[(S)-2-(methoxymethyl)pyrrolidino]but-3- en-2-one.133 The reaction of o-phenylenediamine with pyruvic acid in a 0.1 M HCl solution at 20 8C gave rise to a mixture of the corresponding a-amino derivative of acrylic acid and its tautomer with the imine structure.134 Both isomers can be isolated in the individual form. An attempt to prepare the captodative enamines 2d by direct condensation of methylglyoxal with two equivalents of secondary amine was doomed to failure because the reaction proceeded with the involvement of the more reactive aldehyde group.135 Not only amines but also their derivatives serve as N-nucleo- philes in condensation with dicarbonyl compounds giving rise to the aminoenones 2c.136, 137 For example, trialkylgermylamines were found to react at one of the carbonyl groups of a-diketones 41a,b to form a-amino-a-triethylgermyloxy ketones 42a,b.The latter compounds appeared to be thermally unstable. Thus the compound 42a decomposed to give aminoenone 43.136 O O O R=Me Et3GeNMe2 R R Me R R Me2N OGeEt3 42a,b NMe2 43 O41a,b R=Me (a), Ph (b). Recently, an ingenious procedure was developed for the syn- thesis of esters of unsaturated amino acids 2b from esters of hetarylpyruvic acids 44.138 The esters 44 exist only in the enol form and are very reactive. These compounds are readily transformed to give the unsaturated amino esters 2b in high yields upon refluxing with ammonium formate in ethanol.However, all attempts to obtain the compounds 2b from the esters 44 under the action of other aminating reagents (aniline, benzylamine, ammonium chloride or ammonium acetate) failed. CO2Et CO2Et HCO2NH4 Het Het EtOH, D OH NH2 2b (71% ± 85%) 44 N S N N N Me , . , , , Het= N N N N b. Oxirane-ring opening The oxirane-ring opening under the action of amines is widely used only for the preparation of cyclic aminoenones. These reactions have been surveyed in the review.17 Since then, no modifications of this procedure were proposed. Noteworthy only are the studies,139, 140 which were not included in the cited review, and the investigations141, 142 reported at a later time.Captodative aminoalkenes 4.Formation of an electron-withdrawing group in aminoalkene Examples of the insertion or generation of an electron-withdraw- ing group for the construction of captodative systems are few in number. In addition, these reactions are of limited application and of no particular preparative value. The formation of the aminoalkenes 2c containing the geminal carbonyl group was observed upon hydrolysis of enedi- amines,143, 144 oxidation of enamines 145, 146 or aminoalcohols 147 and removal of protective groups.148, 149 Thus analogues of one of the natural components of the smell of freshly backed bread, viz., compounds 45a,b, were synthesised by careful acid hydrolysis of the corresponding ketal 46 148 or azomethine derivative 47.149 2N HCl, D, 14 h Me N 16% Pri OMe OMe R 46 N O Pri HO2CCO2H, D, 2 h 45a,b Et N 79% N Pri Pri 47 R=Me (a), Et (b).A more general and efficient procedure for the synthesis of the 1-cyanoenamines 2a is based on the nucleophilic substitution of the halogen atom by the cyanide ion.150, 151 For example, pro- longed refluxing of 1-chloro-1-(dimethylamino)alkenes 48 with potassium cyanide in acetonitrile or with zinc cyanide in chloro- form afforded the 1-cyanoenamines 2a in good yields.150 Unfortu- nately, this procedure is inapplicable to the synthesis of b-monosubstituted analogues of the compounds 2a (R1=H) because the starting chloro derivatives are much less accessible and, in addition, they readily undergo dechlorination under the action of bases or upon heating.R1 R1 KCN, MeCN, D CN Cl R2 R2 or Zn(CN)2, CHCl3, D NR3R4 NR3R4 2a (60% ± 80%) 48 R1=Me, Et; R2=Me, Ph, Cl; R3=R4=Me. Finally, one activating function in compounds of the type 2 can be transformed into another one. Thus 2-(N-methylanilino)- pent-2-enonitrile was transformed into the enamines 2c contain- ing the geminal carbonyl group under the action of organolithium, -magnesium or -zinc compounds.54 RX C(O)R CN Et Et Et2O or THF N(Me)Ph N(Me)Ph 2c (36% ± 65%) R=Me, Bu, Bn, All; X =MgCl, MgI, Li, AllZn. 5. Other procedures Other reactions giving rise to the captodative aminoalkenes 2a ± d have also been described in the literature.The results of inves- tigations for a series of activated alkynes were the basis of several publications. It is known that the reactions of ethynylcarbonyl compounds with N-nucleophiles are among procedures most commonly used for the preparation of amino derivatives of the push-pull type 1.12 However, it was unexpectedly found that dialkylamines containing bulky substituents added not only at the b position but also at the a position with respect to the aldehyde group of 4,4-dimethylpent-2-ynal (49) to give a mixture of the unsaturated aminoaldehydes 1d and 2d.152 Thus the 201 proportion of the isomer 2d reached 50% and 10% in the reactions of Pri2NH and Bun2 NH, respectively. The authors related the regioselectivity of the addition to the spatial structure of the nucleophile.NR2 HNR2 CHO But ButC CCHO CHO + But NR2 49 1d 2d R2=Et2, Pri2, Bun2 , Bn2, Ph2. Analogously, the reactions of methyl 4-nitrophenylpropiolate (50) with some secondary amines afforded the corresponding a-amino derivatives 51.153 CO2Me HNR2 C CCO2Me O2N NR2 O2N 51 (27% ± 88%) 50 R=(CH2)2OH, (CH2)2OSO2Me, (CH2)2Cl. In the synthesis of pyrimidine derivatives from triazines and 1-diethylaminoprop-1-yne (52), HCN, which was eliminated in the course of the reaction, added to the starting alkyne to produce 2-diethylaminocrotononitrile (53) in yields of up to 76%.154 HCN CN Me MeC CNEt2 52 NEt2 53 Cyclic a,a 0-dibromoketones 54a ± c reacted with primary 155 and secondary amines 17, 156 to give either Favorskii rearrange- ment products 55b,c or the corresponding 2-aminocycloalk-2- enones 56a ± c depending on the ring size of the ketone and the structure of the nucleophile.The ratio of the reaction products is substantially affected by the nature of the solvent. The formation of the enamines 56a ± c is favoured by polar aprotic solvents. O O O NR1R2 Br Br HNR1R2 + NR1R2 ( )n ( )n ( )n 56a ± c 54a ± c 55b,c n = 0 (a), 1 (b), 2 (c); R1=H, R2=Prn, Pri, But; NR1R2=N ,N ,N O. In spite of the fact that treatment of a,a-dihaloketones with amines generally afforded the corresponding diamino deriva- tives,157 the reactions of ethyl 1-alkyl-3,3-dibromo-2-oxocyclo- hexanecarboxylates 57a,b with a tenfold excess of morpholine gave rise to aminoalkenes 58a,b.158 O O O O HN R Br R N Br CO2Et CO2Et 57a,b 58a,b R=Me (a), Et (b).Recently, 2-aminocyclohex-2-enones 59a ± c were prepared 126 by oxidation of saturated ketones 60a ± c. Oxidation proceeded O O HN HN CO2Et CO2Et O2 R R 59a ± c 60a ± c R = H (a), Me (b), 4-HOC6H4CH2 (c). even upon storage of the starting substrates without a solvent in air. The same results were obtained by bubbling oxygen through a solution of the ketones 60 in chloroform.202 The most reactive captodative aminoalkenes 2d were first nucleophilicity. In contrast, if the structure 66 predominates, captodative aminoalkenes must approach the corresponding activated alkenes in their chemical properties. synthesised by L Duhamel and co-workers.159, 160 The reactions of Grignard reagents with formylated or acylated enediamines 61 produced the corresponding aminoalkenes 2c,d in good yields.C(O)R1 C(O)R1 R22 N R3 R3MgX N N61 2c,d (43% ± 90%) , NMe2; R1=H, Me, Ph; NR22 = N R3=Me, Et, Prn, Pri, Bun, But, Bn, Ph. It is well known that the degree of conjugation between the p-electrons of the double bond and the lone electron pair of the nitrogen atom or the p-system of the activating group is man- ifested by the chemical shift of the olefinic proton in the 1H NMR spectrum.165 ± 167 Apparently, this correlation can be made only if the steric parameters of the substituents at the nitrogen atoms are identical or similar to those at the b-carbon atom, whereas the differences in their chemical shifts are rather large. The 13C NMR spectra are less dependent on the influence of the medium and the anisotropic effects and directly reflect the state of both the p and s electron density in the molecule.For a series of push-pull alkenes, the difference in the chemical shifts of two adjacent carbon nuclei was used for estimating the degree of polarisation of the double bond.165 ± 167 However, it should be emphasised that 13C NMR spectroscopy provides only qualitative estimates of the degree of polarisation of the double bond in the captodative aminoalkenes 2. An effort to perform spectral analysis of 2-aminocycloalk-2- enones was made in the studies.17, 168 The spectroscopic characteristics of captodative aminoal- kenes containing the terminal double bond are given in Table 1.In the 1H NMR spectra of these compounds, the signals for the Table 1. Chemical shifts of the vinyl protons (in CDCl3) in captodative Ingenious methods were proposed for the synthesis of esters of dehydroamino acids with the free amino group.161 ± 164 Alkyl 2-aminoalk-2-enoates 62a ± e were prepared by alcoholysis of anhydrides 63a ± e under mild conditions.161, 162 In the stud- ies,163, 164 aldimines 64e ± l (which were prepared by condensation of glycine esters with aldehydes) were used as the starting reagents, which were transformed into the target compounds 62e ± l upon refluxing in toluene in the presence of triethylamine or upon storage in the MeCN±DBU system (DBU is 1,8-diazabi- cyclo[5.4.0]undec-7-ene) at room temperature.Spectral analysis demonstrated that the esters 62a ± l were always generated only as the Z isomers. O aminoalkenes R2OH R1 O HN CO2R2 O Com- pound R1 63a ± e NH2 62a ± l PhMe, Et3N, D R1 CO2R2 or MeCN±DBU N 64e ± l abc R1=Me (a), Et (b), Prn (c), Pri (d), Ph (e), Ph2C=CH (f), Me2C=CH (g), d 3-NO2C6H4 (h), 4-NO2C6H4 (i), 4-MeOC6H4 (j), 4-ClC6H4 (k), (4-ClC6H4)2C=CH (l); R2=Me, Et. ef Other reactions giving rise to captodative aminoalkenes have also been reported. However, they are of limited application, cannot be used for preparative purposes, and hence, are beyond the scope of the present review. g III.Structures and spectroscopic characteristics of captodative aminoalkenes hijkl In the analysis of the structures of captodative aminoalkenes, one should first of all answer the question of how the combination of the electron-withdrawing and electron-donating groups is reflected in the degree of p,p- and p,p-conjugation in such systems. Captodative aminoalkenes can be represented by superposi- tion of the structural fragments of enamines, on the one hand, and of a,b-unsaturated aldehydes, ketones or derivatives of carboxylic acids, on the other hand. This approach assumes that three canonical structures must be taken into account in the description of the reactivities of these compounds. m 7 EWG EWG EWG + 7 + n NR1R2 NR1R2 NR1R2 o 65 2 66 a In CCl4 .b The author's own data. If the structure 65 makes a larger contribution to the total electronic state of the molecule compared to that of the resonance form 66, the b-carbon atom would be expected to possess higher EWG (a ± o). NR1R2 EWG NR1R2 CN NH2 CN NMe2 CN N CN N S CN N CN Ph N CN NPh N S CN N CF3 N N CN N(Me)Ph CN CO2Me NMe2 CO2Me N CO2Me O N C(O)Me N C(O)Me O N A Yu Rulev Chemical shifts Ref. of the vinyl protons /ppm 39 4.67, 4.74 20 4.33, 4.53 a 33, 169 4.55, 4.75 170 4.63, 4.83 33 4.15, 4.42 170 4.62, 4.80 170 4.67, 4.87 170 4.63, 4.83 170 4.71, 4.92 23 4.64, 4.80 4.43, 5.04 131 131 4.54, 5.11 131 4.63, 5.26 4.48, 4.87 see b 128 4.80, 5.20Captodative aminoalkenes protons of the H2C= group are observed in the `olefinic region' between the signals typical of vinylamines (d=3.40 ± 3.78) 171 and the corresponding activated alkenes (d=5.75 ± 6.36).172 ± 174 It should be noted that the chemical shifts of the olefinic protons depend on the nature of the captodative substituents.For exam- ple, the shifts d(H2C=) in the spectra of a-aminoacrylonitriles bearing the nitrogen-containing heterocycles with similar steric parameters (see Table 1, the compounds c ± i) are changed as the electron-donating ability of the amino group changes. As in the series of unsubstituted enamines,175 the strongest orbital inter- action between the lone electron pair of the nitrogen atom and the double bond is observed in the pyrrolidine derivative e. The insertion of the alkyl or aryl substituent at the b position enhances deshielding of the olefinic proton and provides the possibility for the existence of the geometric isomers (Table 2).Table 2. Chemical shifts of the olefinic protons and the carbon nuclei (in CDCl3) in captodative aminoalkenes EWG Com- pound a CHO b CHO c CHO d CHO e CHO f CHO g C(O)Me h C(O)Et i C(O)Me j C(O)Ph k C(O)Ph l C(O)Ph m CO2Me CO2Me CO2Me CO2Me CN nopqr CN s CN a In CCl4. b The assignment to the E and Z isomers was made based on the author's own data. c The author's own data. R3 NR1R2 Me N O N Me Me NEt2 Ph N Ph N O Ph NEt2 Me O N Me N O Et N(Me)Ph Ph N Ph N O Ph NEt2 Me N Me Et Pr Me N(Me)C6H11-cyclo N(Me)Ph N(Me)Ph NMe2 Ph NMe2 Et N(Me)Ph As in the case of unsubstituted enamines,175 the signal for the olefinic proton of the E isomers of captodative aminoalkenes is always observed at higher field.The results of the study 48 showed that this regularity is violated by some formyl-substituted enam- ines. However, I thoroughly checked the molecular geometry by NMR spectroscopy using the Overhauser effect and found that the assignment of the signals made in the cited study 48 was incorrect. To a first approximation, the larger value of d(HC=) for the Z isomers can be considered a result of strengthening of p,p-conjugation with a simultaneous decrease in p,p-conjugation due to the steric interaction between the b-substituent and the tertiary amino group.For this reason, the Z isomer is generally thermodynamically more stable. Factors other than the structural characteristics of the mole- cule are also responsible for the ratio between the E andZisomers. Isomer Chemical shifts of the olefinic protons /ppm 5.15 5.48 EZEZZ 5.80 6.20 6.50 6.50 a 6.75 a EZ 6.10 6.50 6.66 6.77 Eb ZEZE 4.80 E 4.90 Z 6.55 Eb 5.70 5.90 6.06 5.57 6.12 EZb EZ 4.54 6.08 6.50 6.94 6.89 5.08 5.82 5.82 6.18 5.81 6.15 EZZZZEZEZEZ 203 EWG H (a ± s). C2 C1NR1R2 R3 Ref. Chemical shifts of the carbon nuclei /ppm C2 C1 176 148.8 151.8 122.7 140.4 176 147.4 150.3 149.2 124.2 141.8 146.6 176 48 7 7 7 7 48 7 7 7 7 7 7 see c 146.9 139.37 7 129 129 150.6 106.9 54 147.2 116.97 7 177 146.9 107.0 129 177 see c 103.40 145.75 7 7 178 178 178 178 20 20 179 140.4 7133.9 135.9 136.6 7777121.3 146.3 102.2 127.8 141.9 143.9 142.5 113.5 133.3 115.3 129.8 135.6 146.4204 Example of E,Z isomerisation of captodative aminoalkenes, which proceeded rather readily either in the course of the reactions 48, 128, 129, 177 or upon subsequent purification (for exam- ple, upon distillation) 179 and even upon storage over a short period at room temperature,129 are available in the literature. For captodative aminoalkenes bearing the internal double bond, the chemical shift of the olefinic proton is also sensitive to variations in the nature of the electron-donating and electron- withdrawing substituents.Generally (although not necessarily), the stronger are the electron-withdrawing properties of the activating group and the weaker are the electron-donating proper- ties of the amino group, the more substantial is the downfield shift of the signal for the olefinic proton. If the nitrogen atom possesses substituents capable of being involved in conjugation with the lone electron pair of the nitrogen atom, the chemical shift of the olefinic proton is increased further by 0.3 ± 0.8 ppm. This fact also indicates that in these compounds conjugation between the double bond and the electron-withdrawing group typical of activated alkenes is virtually completely restored.In the 13C NMR spectra of captodative N,N-disubstituted aminoalkenes, the signal for the C2 atom is observed at a higher field as compared to the signal for the C1 atom. It should be noted that this regularity is also true for the series of enamines bearing such a powerful electron-withdrawing substituent as the formyl group.176 The difference in the chemical shifts of the adjacent carbon atoms of the alkene fragment observed in these systems is comparable with that for unsubstituted enamines (30 ± 45 ppm). These data are unambiguously indicative of high polarisation of the double bond and are a strong argument in favour of the predominant contribution of the structure 65 to the ground electronic state of the molecule.In the case of N-methylaniline derivatives containing the geminal electron-withdrawing group, an inversion of the usual order of the chemical shifts of the C1 and C2 nuclei is observed. The difference d(C2)7d(C1) reaches +14.3 ppm (see Table 2, compounds o, p and s). The exceptions are enamines bearing the carbonyl group for which the positions of the resonance signals for the carbon nuclei are less dependent on the nature of the amino group (see Table 2, compounds h, i, k and l). A satisfactory explanation for this phenomenon is lacking. The presence of conjugation in theC=C7EWGsystem of the carbonyl-containing captodative aminoalkenes 2d is evidenced by the presence of intense absorption bands at 1655 ± 1690 (C=O) and 1585 ± 1620 cm71 (C=C) in their IR spectra.180 The former spectral region is close to the stretching vibration frequencies characteristic of the conjugated carbonyl group, whereas the frequencies belonging to the latter region are somewhat lower than the vibration frequencies of the double bond in the corre- sponding unsubstituted enals and enones and are noticeably lower than the vibration frequencies of the C=C bond in simple enamines.Analysis of the IR spectra of captodative carbonyl-containing enamines by comparing the intensities of the absorption bands of theC=CandC=Ogroups led Arnould et al.177 to the conclusion that the E isomer in the ground state adopts the s-cis conforma- tion, whereas the Z isomer has the s-trans conformation. Finally, the efficiency of p,p-conjugation in 2-aminocycloalk- 2-enones depends dramatically on the ring size (Table 3).For example, the shifts d(HC=) in the 1H NMR spectra of piperidine and morpholine derivatives of cycloalkenones increase substan- tially on going from seven- to six- and then to five-membered rings. As in the spectra of acyclic enamines bearing the geminal carbonyl group, the resonance signals for the C3 atom in the spectra of 2-aminocycloalk-2-enones containing mono- and di- substituted amino groups are shifted upfield as compared to the signals for the C2 atom. The difference between the chemical shifts Table 3.Chemical shifts of the olefinic protons in 2-aminocycloalk-2- enones. Compound R Chemical shifts of the olefinic protons /ppm O1 5.98 a 5.83 b NHPri NHBut R 2 3 5.91 N 6.42 O N O 5.70 c N 1 R 23 5.86 d O N O 5.23 N 1 R 23 5.13 O N The chemical shifts of the carbon nuclei: a d(C3)=120.61, d(C2)=145.13 (see Ref. 155); b d(C3)=121.04, d(C2)=142.62 (see Ref. 155); c d(C3)= 126.1, d(C2)=148.1 (see Ref. 146); d d(C3)=133.7, d(C2)=150.8 (see Ref. 146). of the b- and a-sp2-hybridised carbon atoms varies between 722 and 727 ppm, whereas this difference for unsubstituted cyclic a,b-unsaturated ketones ranges from 21 to 33 ppm. Undoubtedly, this effect should be related to the fact that p,p-conjugation between the nitrogen atom and the double bond dominates over the p,p-interaction between the double bond and the carbonyl group. All the aforesaid send one in the critical review of the conclusion made previously 17 that the canonical structure 66 is more appropriate for the description of the ground state of 2-aminocycloalk-2-enones.Hence, the spectroscopic data provide evidence for the pres- ence of competitive p,p- and p,p-conjugation in the captodative aminoalkenes 2a ± d, the efficiency of the conjugation being determined by the nature of the functional groups. In the general case, the compounds under consideration are intermediate between unsubstituted enamines and a,b-unsaturated aldehydes, ketones or derivatives of carboxylic acids.What this means is these systems would be expected to exhibit dual reactivity with respect to nucleophilic and electrophilic reagents. IV. Reactions of captodative aminoalkenes Being polyfunctional systems, the captodative aminoalkenes 2a ± d are not only valuable starting reagents for organic synthesis but are also excellent models for solution of some problems of alternative reaction centres. Analysis of the data available in the literature shows that insignificant (at first glance) changes in the structure of the substrate or the reagent and variations in the reaction conditions can give rise to radically different reaction products. In addition, the polyfunctional properties of the com- pounds under consideration are often responsible for their unex- pected and surprising cascade transformations. A Yu Rulev Ref.Solvent 155 155 CDCl3 CDCl3 181 CCl4 182 CDCl3 156 CCl4 86 CDCl3 183 CCl4 183 CCl4Captodative aminoalkenes 1. Reactions with electrophilic reagents a. Protonation and hydrolysis Apparently, the nitrogen atom, the b-carbon atom and the heteroatom of the activating group serve as the nucleophilic centres in captodative aminoalkenes. Consequently, the regiose- lectivity of the electrophilic attack is the central problem of the reactions of the compounds 2a ± d with electrophilic reagents. EWG E N Protonation serves as the simplest test for the reactivity of such systems. Unlike simple enamines generating stable immo- nium salts 175 and the push-pull aminoenones 1c, which are protonated predominantly at the oxygen atom,10 the captodative aminoalkenes 2a,c,d can add the proton at each of three major nucleophilic centres depending on the structure of the substrate, the nature of the protonating reagent and the reaction conditions. Thus the reactions of mineral or strong carboxylic acids with the aminoalkenes 2, which are activated by the geminal cyano,184 carbonyl,185 formyl 186, 187 or azomethine group,188 generally afford the corresponding enammonium salts 67a,c ± e.It should be emphasised that the hydrochlorides 67a,c,d (X=Cl) were isolated and completely characterised. EWG EWG HX + R1 R1 NHR3R4 X7 NR3R4 R2 R2 67a,c ± e 2a,c ± e EWG =CN (a), C(O)Ph (c), CHO (d), CH=NMe (e); R1=H,R2=Me, Ph; R1±R2=(CH2)5; NR3R4=NHBut, NEt2,N ,N O; X=Cl, CF3CO2, CCl3CO2.The regioselectivity of protonation of the 2-aminocycloalk-2- enones 56a ± i is determined by the ring size and the nature of the amino group.121, 189 For example, the reaction of a solution of trifluoroacetic acid in CDCl3 with morpholinocyclohexenone 30 afforded exclusively enammonium salt 68b (NR1R2 is morpho- lino), whereas the reaction with the seven-membered homologue 56c (NR1R2 is morpholino) performed under the analogous conditions led to protonation both at the heteroatom to give compound 68c (NR1R2 is morpholino) and at the b-carbon atom to generate the corresponding immonium salt 69c.121 Protonation of aminocyclopentenones (56a) as well of the cyclohexenone 56b containing the monosubstituted amino group (NR1R2= NHC6H4Me-4) afforded immonium salts 69a,b, which underwent subsequent enolisation.121 The dependence of the regioselectivity O X7 HNR1R2 + Y O ( )n 68b,c NR1R2 HX Y O + ( )n X7 NR1R2 56a ± i Y( )n 69a ± i Y=CH2: n = 0 (a), 1 (b), 2 (c); Y=CHMe: n=0 (d), 1 (e), 2 (f); Y=NH: n = 0 (g), 1 (h), 2 (i); NR1R2=NHBn, NHC6H4Me-4, N ; X = CF3CO2, Cl.O, NMe2, N 205 of protonation on the ring size was also observed in the case of the enaminolactams 56h,i.190, 191 The formation of nitrogen-protonated captodative amino- alkenes is a rare example of the existence of enammonium salts, which, as a rule, readily undergo isomerisation to the correspond- ing immonium salts.175 A rather stable enammonium salt was first synthesised in 1976.192, 193 However, really stable compounds with the enammonium structure were prepared only upon protonation of the captodative aminoalkenes 2c,d.Apparently, their stability is associated with the powerful influence of the electron-withdraw- ing geminal substituent and the stabilising effect of the intra- molecular hydrogen bond that is formed.{ This assumption is supported by the fact that protonation of (s-cis-diethylamino)- benzylideneacetophenone gave rise to the enammonium salt adopting the s-trans conformation.185 Ph O Ph H+ Ph O Ph Et2N H + NEt2 The salts 67d are so stable that they were not transformed into immonium salts at all.186, 187 As shown in the previous section, a decrease in the electron-withdrawing ability of the activating group in the captodative aminoalkenes 2 leads to an increase in the proportion of the resonance form 65 bearing a negative charge on the b-carbon atom.As a consequence, these aminoalkenes exhibit properties of simple enamines. Actually, unlike the formyl- substituted enamines 2d, their azomethine analogues 2e generated the enammonium salts 67e, which underwent slow isomerisation in CDCl3 solution to give the more stable C-protonated deriva- tives 70e.188 NMe CH NMe CH NR2=N HX + 720 8C 710 8C Me NR2 Me NHR2 X7 2e 67e NMe CH + 7 Me N X 70e ; O, N NR2= N X=CF3CO2, CCl3CO2. The result of the reaction of the acyclic aminoenones 2c with acids depends substantially on the nature of the protonating reagent and the reaction conditions. Thus the reaction with anhydrous HCl was accompanied predominantly by protonation at the nitrogen atom, whereas the formation exclusively of immonium salts was observed in a solution of trifluoroacetic acid.185 The difference in the behaviour of inorganic and carbox- ylic acids is accounted for by the bifunctional catalytic action of carboxylic acids in the course of proton transfer from the nitrogen atom to the carbon atom.194 In the enammonium salts 67, p,p-conjugation becomes impos- sible.Moreover, the donor fragment NR3R4 is transformed into the acceptor N+HR3R4 upon quaternisation. The formation of the system with the geminal electron-withdrawing substituents results in the essential electron density redistribution.This is the reason that the multiplicity of the C=C bond is somewhat decreased and, as a consequence, the barrier to rotation about this bond is substantially lowered. Examples of E,Z isomerisation of enammonium salts (67c ± e), which proceeded very readily, were cited in Refs 185 ± 188. In spite of the fact that enammonium salts do not necessarily undergo tautomeric transformations into immonium salts, the { The author's own data.A Yu Rulev 206 stronger alkylating reagents, for example, of dialkyl sulfates, alkyl fluorosulfonates or triethyloxonium tetrafluoroborate.196 In the latter case, alkylation proceeded under mild conditions and was not accompanied by side processes.captodative aminoalkenes 2b ± d are hydrolysed to give the corresponding a-dicarbonyl compounds. For example, the ami- noesters 62a ± e were virtually quantitatively transformed into esters of 2-oxocarboxylic acids 71a ± e in a 1 M solution of hydro- chloric acid at room temperature during 0.5 h.162 CN CN Et3O+BF¡4CO2R2 CO2R2 R1 R1 H3O+ R1 R1 NEtR3 NHR3 0.5 h, 20 8C O NH2 R2 71a ± e 62a ± e R2 (62% ± 95%) R1=Me (a), Et (b), Prn (c), Pri (d), Ph (e); R2=Me, Et. R1=Me, Et; R2=Me; R1±R2=(CH2)5; R3=Me, Pri, But, cyclo-C6H11. The acyl- and formyl-substituted enamines 2c,d are also hydrolysed in an acidic medium.160, 186 Various a-diketones can be prepared from the compounds 2c by varying the substituents in the starting substrate.C(O)R2 `Magic methyl' was successfully used for quantitative O-alky- lation of 2-aminotropone (75a) and 2-methylaminotropone (75b).197 It is reasonable to assume that the driving force for this reaction is the formation of the aromatic system of tropylium cation 76a,b. R1 C(O)R2 N OMe O R1 NH. HCl + NHR NHR O FSO3Me + FSO¡32c,d R1=Alk, Ar; R2=H, Alk, Ar. 76a,b (100%) 75a,b R = H (a), Me (b). The primary, secondary and tertiary captodative aminoal- Protonation and hydrolysis of the captodative aminoalkenes 2 are used in a number of simple and efficient synthetic procedures. An example is hydrolysis of the 1-cyanoenamines 2a containing the unsubstituted, mono- or disubstituted amino group to give the corresponding carboxylic acids 72.20, 110, 150 The reactions of the cyanoenamines 2a with Lewis acids (AlCl3 or ZnCl2) in ethanol followed by hydrolysis gave rise to monosubstituted amides 73.184, 195 O R1 a kenes 2 are difficult to acylate.Thus attempts to perform the reactions of the primary aminoalkenes 2b with acetyl chloride or acetic anhydride failed.138 However, some reactions of the com- pounds 2b with acid chlorides or anhydrides of strong carboxylic acids produced N-acylated products in moderate yields.95, 96, 104, 105, 107, 198 ± 200 These reactions were used in the peptide synthesis.200 ± 204 OH R3=R4=Me CN R2 R2 R2 72 R4X R1 CO2R3 CO2R3 NR3R4 R1 R1 O R2 NHR4 NH2 2a R1 b, c (8% ± 60%) 2b NHR3 R4=H R2 73 R1=H, Me; R2=Me, Et, Prn, Pri, Ph, MeOCH2OC6H4, Het; R3=Me, Et; R4=ClCH2CO, BnOCO, 4-MeC6H4SO2, CF3CO, R1=H, Me, Et; R2=Me, Et, Ph, 4-MeOC6H4; R1±R2=(CH2)5; O R3=R4=Me; R4=H: R3=H, Me, Pri, But, cyclo-C6H11. (a) H2O, H+, (b) AlCl3, EtOH or HCl, PhH; (c) H2O.NCH2CO; X=Cl, CF3CO2 . O b. Alkylation and acylation Alkylation of captodative aminoalkenes requires more drastic conditions and affords the corresponding enammonium salts regardless of the nature of the alkylating reagent. It is not surprising that the nature of the amino group exerts a pronounced effect on the ease of the reaction. Thus prolonged refluxing of 2-aminocycloalk-2-enones 56a,b with alkyl halides afforded N-alkyl derivatives 74a,b.181, 189 R3 O O N-Acylation of 3-aminocoumarin derivatives was successfully carried out under the action of phthalic anhydride and 2-chloro- nicotinoyl chloride.205, 206 Perfluoroacyl fluorides proved to be excellent acylating reagents for the formyl-substituted amino- alkenes 2d.207 The reaction of 2-tert-butylaminocyclohex-2- enone with ethyl chlorocarbonylacetate in the presence of pyridine and 4-dimethylaminopyridine afforded a mixture of N-acylated derivative 77 and the product of its subsequent intramolecular condensation 78.208 NR1R2 X7 NR1R2 R3X O + O CO2Et NHBut Cl ( )n ( )n 74a,b 56a,b O EtO2C O n = 0 (a), 1 (b); NR1R2=NMe2,N ,N O; R3=Me, Et; NBut NBut X=I, Cl, BF4.CO2Et + O 78 (32%) 77 (53%) The cyanoenamines 2a containing a bulky substituent on the monosubstituted amino group do not react with alkyl halides or alkyl p-toluenesulfonates.However, aminoalkenes bearing the disubstituted amino group can be prepared with the use ofCaptodative aminoalkenes The formation of diester 79 in the reaction of methyl 2-dime- thylaminoacrylate with methoxalyl chloride is a rare example of C-acylation of captodative aminoalkenes. The compound 79 is readily hydrolysed to the symmetrical ester of dioxo acid 80.131 O O NMe2 NMe2 Cl CO2Me H2O 1N HCl CO2Me MeO2C CO2Me Et3N, Et2O 79 (70%) H O O CO2Me MeO2C 80 (80%) Unsymmetrical amino groups in captodative aminoalkenes are sufficiently basic and nucleophilic that they can be subjected to condensation with carbonyl compounds. These aminoalkenes smoothly react with aldehydes or dimethylformamide dimethyl acetal to form the corresponding azomethines.94, 106, 110, 138, 209 Even the unstable cyanoenamines 2a were readily transformed into 2-aza-3-cyanobuta-1,3-dienes 81 and 82 upon refluxing in benzene with azeotropic distillation of water.110 CN PhCHO CHPh Ar N CN PhH, D 81 (80%) Ar NH2 CN 2a (MeO)2CHNMe2 CHNMe2 Ar N 82 (73% ± 80%) Ar=Ph, 4-MeOC6H4. The reactions of the aminoalkenes 2b with 5-formylpyran-2- one, which begin with the ring opening, provide the convenient means for preparing dihydroazepines 83.210 O EtO2C NH CO2Et O + Ar 7CO2 CHO Ar NH2 83 2b CHO Ar=Ph, 4-ClC6H4.In reactions with some carbonyl compounds, captodative aminoalkenes containing mono- and disubstituted amino groups act as C-nucleophiles. For example, the reactions of the 2-amino- cycloalk-2-enones 56b,c with quinones afforded various hetero- cyclic compounds.122 Their structures depend on the ring size and the nature of the amino group of the substrate.It was believed 11, 122, 211, 212 that the first step of the reaction gave rise to derivative 84. Cl ( )n Cl ( )n O HO O + O NR1R2 OH O Cl Cl NR1R2 56b,c 84 207 ( )n Cl HO n=1, 2 O N O NR1R2=N O Cl O (*50%) Cl n=1 HO NR1R2=NHBn O NBn O(12%) Cl The first investigations into the involvement of the captoda- tive aminoalkenes 2 in aminomethylation demonstrated that the structure of the resulting Mannich base depends dramatically on the ring size of the substrate and the reaction conditions.181, 213 Thus the 2-aminocyclopent-2-enones 56a produced bis-amino- methylated compounds 85 in moderate yields,181 whereas the reaction of the aminoalkene 30 gave rise only to monoaminome- thylated product 86 even in the presence of a twofold excess of paraformaldehyde and morpholine hydrochloride.213 O HCl NR2 +CH2O+HN( )n X 56a O OH N( )n X N X ( )n 85 n=1, X = CH2; n=2, X = CH2, O.O O O OH N HCl O O +CH2O+HN N 30 86 (83%) c. Reactions of electrophilic reagents with carbanions generated from captodative aminoalkenes As mentioned above, the b-substituted captodative aminoalkenes 2a,b are capable of being deprotonated under the action of strong bases to give allylic anions 87, which can react with various electrophilic reagents (alkyl halides, aryl halides and carbonyl compounds) either exclusively at the g-carbon atom or simulta- neously at the a- and g-carbon atoms.34, 57, 151, 178, 214 ± 218 The fact that the attack proceeds predominantly at the g-position is attributed primarily to the steric characteristics of the substrate and the electrophile.These reactions are preparatively valuable because they allow modifications of captodative aminoalkenes to obtain compounds, which are used in the synthesis of various polyfunctional systems.208 EWG R1 B NR3R4 R2 2a,b R1 R5X R5 R2 R6C(O)R7 (R3=Me, R4=Ph) EWG =CN (a), CO2Me (b); R1=H, Me, PhS, PhSO2; R2=H, Me; NR3R4=NMe2, NEt2, N R5=Me, Et, Pri, Bun, n-C5H11, Ph, Bn, CH2=CHCH2,MeCH=CHCH2, Me2C=CHCH2; X=Hal; R6=H, Me; R7=Me, Pri, Ph; R6±R7=(CH2)5; B = BunLi, LDA, ButOK. Some tertiary enamines 2a,b react with aldehydes in the presence of strong bases to give five- or six-membered lactones depending on the nature of the activating group.20, 214 CN 1) LDA 2) R2CHO R1H2C NMe2 2a OO R1 R2 (18% ± 46%) R1=H, Me; R2=Ph, Me(CH2)3, Me(CH2)5, MeCH=CH, O, .O CO2Me 1) LDA 2) PhCHO Me NMe2 2bPh O (38%) The Michael addition of various anions to 2-(N-methylanili- no)acrylonitrile afforded initially anions 88.23, 35, 58 Their subse- quent reactions with electrophiles gave rise to bifunctional CN R1Li N(Me)Ph CN R1 89 R1=But, Ph, PhCOCH2CH2, PhCH=CHCHPh, ButC(O)CH2; R2=H, Me, Et, Bn; X=OH, I.Br. EWG R1 7 NR3R4 R2 87 R1 EWG + R2 EWGR5 NR3R4 NR3R4 EWG R1 R2 R6 N(Me)Ph R7 OH , N(Me)C6H11-cyclo, N(Me)Ph; R1 CN H2O, H+ R2 NMe2 OH , CO2Me Ph 7MeOH NMe2 OH NMe2 O CN R2X Li+ 7 R1 N(Me)Ph 88 O H3O+ R2 R1 R2 N(Me)Ph 90 (60% ± 92%) A Yu Rulev derivatives 89, which were hydrolysed to the corresponding carbonyl compounds 90.23, 35, 58 When developing this methodology and varying the nature of the electrophile and the organolithium derivative, Ahlbrecht and co-workers unveiled the considerable synthetic potential of one of the simplest captodative aminoalkenes, viz., 2-(N-methylanili- no)acrylonitrile.This compound proved to be a versatile and readily accessible starting reagent for the one-pot synthesis of ketones,23 b-, g- and d-diketones,179, 219 ± 221 g- and d-oxoalde- hydes,221, 222 b-phenylthioketones, which were transformed into a,b-unsaturated carbonyl compounds 223 upon pyrolysis, as well as of derivatives of oxocarboxylic acids (nitriles,224, 225 esters and amides 226). The efficient syntheses of bicyclic diketones 220 and pyrroles 179 based on 2-(N-methylanilino)acrylonitrile are worthy of notice. O O Me O Me Me Me Me Me PhS O CN O CN Me N(Me)Ph O Me O N O O Me Me Me X Me Y O Me Me X=H, Alk, Ar, OR, NR2; Y=H, Alk, Ar. 2. Reactions with nucleophilic reagents Taking into account the electronic structure of captodative aminoalkenes, one would expect that these compounds will act as poor Michael acceptors and the attack of nucleophilic reagents will proceed predominantly either at the electrophilic centre of the activating group or at the a-carbon atom. EWG Nu N Actually, attempts to perform the 1,4-addition of some C- and N-nucleophiles to 2-aminocycloalk-2-enone have not met with success. Only nitroalkanes produced Michael adducts in moderate yields (see Ref.17). In the study on the cyanoenamines 2a,54 it was concluded that the direction of the addition of nucleophiles generally follows the rule of hard and soft acids and bases (HSAB). However, some of nucleophilic reactions of captodative aminoalkenes do not obey the HSAB principle.The 1,2-addition of organometallic compounds to form a new C7C bond is the most extensively used nucleophilic reaction of captodative aminoalkenes. The reactions of the 2-aminocycloalk- 2-enones 56a ± d with Grignard reagents proceeded regiospecifi- cally to produce 2-hydroxycycloalkanones 91a ± d in good yields.120, 227 ± 230Captodative aminoalkenes O R3 OHO R1 R1 NR22 1) R3MgBr 2) H2O ( )n ( )n 91a ± c, j (35% ± 88%) 56a ± c, j n = 0 (a), 1 (b), 2 (c), 3 (j); R1=H, Me; MeO OMe; OMe, NR22= , N N N R3=Me, Et, Ph, CH2=CH, CH2=CMe. 2 1,4-Adducts were detected in no cases. The reactions with the use of chiral substrates afford hydroxyketones with high enantio- selectivity, which depends on the solvent used.Thus the enantio- meric excess (ee) achieved in the reaction of the aminoenone 56a [R1=H, NR2 is (2S )-2-methoxymethylpyrrolidino] with CH2=CHMgBr in tetrahydrofuran was only 33%, whereas it reached 84% in a Et2O±THF mixture (9 : 1).230 Interestingly, the absolute configuration of the optically active a-hydroxycycloal- kanones 91 depends on the nature of the substituent R2. For example, the aminoenone 56b [R1=H, NR22 is (2S )-2-methoxy- methylpyrrolidino] reacted with methyl- or ethylmagnesium bro- mide to give the compounds (2R)-91b (R3=Me or Et) with an optical purity of 92%± 95%, whereas the hydroxyketones 91b (R3=CH2=CH,Ph), which were obtained by the reactions of the same substrate with vinyl- or phenylmagnesium bromide, adopt an opposite configuration (ee 58% ± 95%).120 The authors attrib- uted these stereochemical results to the difference in the structure of an intermediate complex of the organometallic compound with the aminoenone. Like 2-aminocycloalk-2-enones, acyclic captodative amino- alkenes react with organometallic compounds to form predom- inantly 1,2-adducts.Thus the formyl-substituted enamines 2d are readily transformed into acyloins 92, which provides the basis for a new efficient procedure for the preparation of these synthetically important compounds.{ OH 1) R3MgBr 2) H2O R1 R3 R1 O O 2 NR2 2d 92 (50% ± 54%) R1=Me, Ph; R2=Alk, Het; R3=Et, Ph. Analogously, the reactions of 2-(N-methylanilino)pent-2-eno- nitrile with organomagnesium and organozinc reagents proceed as the 1,2-addition.However, the reaction of the same substrate with benzylmagnesium bromide gave rise to 1,4-addition product 94b along with 1,2-adducts 93b in comparable amounts.54 R R Me 1) RMX 2) H2O Me CN Me CN O + N(Me)Ph N(Me)Ph N(Me)Ph 93a ± c (5% ± 60%) 94b (19%) RMX=MeMgI, BnMgCl, AllZnBr; R=Me (a), Bn (b), All (c). In the case of the secondary cyanoenamines 2a, Grignard reagents serve as strong bases. These reactions afford trialkylke- tenimines 95 as the major reaction products.231 { The author's own data. 209 R1 1) MeMgI 2) H2O CN R2 NHR3 2a R1 R1 R1 NHR3 CN NR3+ R2 + R2 R2 O NR3 95 (27% ± 61%) R1=R2=Me, Et; R3=Pri, But. In the reactions with captodative aminoalkenes, N-nucleo- philes attack either the electrophilic centre of the activating group or the a-olefinic carbon atom.For example, 2-amino-3,3- dichloroacrylonitrile (96) reacted under mild conditions with primary and secondary amines to give amides 97 and 98.232 The mechanism of formation of these compounds involves the nucleo- philic attack at the carbon atom bound to the cyano group followed by hydrolysis of the amidines generated in the first step. Curiously, the chlorine atoms were not replaced even in the presence of a fourfold excess of the amine (cf. Ref. 82). At the same time, the reaction of the substrate 96 with more nucleophilic thiophenol proceeded regiospecifically with the replacement of both chlorine atoms.232 O O 1) HNR1R2 2) H2SO4, D Cl Cl NR1R2+ NH2 Cl Cl CN 98 (24% ± 32%) 97 (12% ± 34%) Cl NH2 CN Cl PhS PhSH 96 NH2 SPh(82%) NHPr.O, NR1R2=N ,N Condensation of primary amines and hydrazine derivatives with the captodative aminoalkenes 2c,d proceeded exclusively at the carbonyl group. Under the action of the liberated water, the initially formed azomethinoenamines 99c were hydrolysed in a basic medium to give oxo derivatives 100.233 Under these con- ditions, the compounds 99d were converted into bis-azomethines 101. The ratio 99d : 101 depends on the reaction conditions and the nature of the starting reagents.234 Under analogous conditions, guanidine and thiourea did not react with the aminoalkenes 2c.233 NR4 C(O)R2 R4NH2 R2 R1 R1 NR3 2 2 NR3 99c,d (32% ± 59%) 2c,d NR4 H2O Ph R1 (R2=Ph) O 100 (40% ± 90%) R4NH2 NR4 (R2=H) R1 NR4 101 (2% ± 31%) R1=Me, Ph; R2=Ph (c), H (d); NR32 =NEt2, ; N N O R4=Me, Bu, NH2, NHEt.210 2-Morpholinocyclohex-2-enone (30) undergoes condensation with ethanolamine acting as a N-nucleophile to give the corre- sponding azomethine derivative.235 As mentioned in Section II.2.a, 2-aminocycloalkenones are involved in ipso-sub- stitution reactions under the action of primary aromatic amines and even ammonia.17, 88 Unlike N-nucleophiles, S-nucleophiles can be involved in the intra- or intermolecular conjugated addition to captodative ami- noalkenes.For example, prolonged refluxing of 2-(N-methylani- lino)pent-2-enonitrile with sodium thiophenoxide and thiophenol in THF gave rise to 1,4-addition product 102.54SPh Me PhSNa, PhSH Me CN CN THF, D N(Me)Ph N(Me)Ph 102 (70%) The reactions of the aminoalkenes 2d with alkane- or arene- thiols or dithiols afforded neither 1,2- nor 1,4-adducts; instead, the compounds 2d were unexpectedly transformed into the corre- sponding thiol esters of a-amino acids 103 in yields of up to 87%.236 Neither changes in the structures of the reagents nor variations in the reaction conditions have an effect on the direction of the reaction.O HSR3 SR3 R1 O R1 PhH or THF NR2 NR2 2 2 2d 103 (40% ± 87%) R1=Me, Ph; NR22 =NEt2, ;R3=Et, Bu, Ph, CH2CH2SH. N This methodology, which allows the construction of the CH(NR2)CO fragment, was further developed in the synthesis of esters of a-amino acids containing the disubstituted amino group.It is well known that these compounds are important precursors for the design of new pharmaceuticals and for the preparation of polyfunctional compounds. It appeared that the reactions of the formyl-substituted enamines 2d with an equimolar amount of dialkyl phosphonate in the presence of sodium alkoxide afforded esters of N,N- disubstituted a-amino acids 105 rather than the expected phos- phonates 104.237 ± 239 The moderate yields of the esters 105 were accounted for by competitive anionic polymerisation of the starting substrate or one of the intermediates under the action of sodium alkoxide. In the absence of the catalyst, the reaction of the phosphonate with the aminoalkene 2d did not take place.R3ONa, R3OH R1 O + (R3O)2P(O)H NR22 2d OH R1 P(O)(OR3)3 NR22 104 O OR3 R1 NR22 105 (30% ± 37%) R1=Me, Pr, Ph; NR22 N =NEt2, ,N O;R3=Me, Et. Unlike the formyl-substituted enamines 2d, the aminoenones 2c containing the terminal double bond are involved in the regiospecific 1,4-addition to dimethyl phosphonate. Thus 3-piper- idinobut-3-en-2-one smoothly produced the Michael adduct.239 A Yu Rulev The 1,2-adduct was not detected even in trace amounts. 2-Di- ethylaminobenzylideneacetophenone did not react with dimethyl phosphonate. O O (MeO)2P(O) MeONa, MeOH N N + (MeO)2P(O)H (73%) The scheme providing an explanation for the generation of the amino acid derivatives 103 and 105 was proposed in the stud- ies.236 ± 240 Undoubtedly, the first step involves the addition of the nucleophile to the formyl group of the aminoalkene 2d. The key step of the scheme involves isomerisation of the initially formed 1,2-adduct to give thermodynamically more stable tetrasubsti- tuted alkene 106, which is rapidly transformed into the thiol ester 103 (in the case of S-nucleophiles) or ketophosphonate 107 (in the case of P-nucleophiles). Heterolysis of the C7P bond in the intermediate 107 readily proceeds under the action of alkoxides to give the ester of a-amino acid 105. OH OH Nu Nu 2d+HNu R1 R1 NR2 NR2 2 2 106 Nu=SR3 103 O O Nu=P(O)(OR4)2 R1 OR4 P(O)(OR4)2 R1 R4ONa, R4OH NR2 NR2 2 2 107 105 The proposed hypothesis was strongly supported by isolation of the enol 106 [Nu=P(O)(OMe)2] as a stable silyl ether gener- ated in the reaction of the a-formyl-substituted enamine 2d (R1=Ph, NR22 =NEt2) with dimethyl trimethylsilyl phos- phite.239 The above-described reaction of the substrates 2d with dialkyl phosphonates is one of the first examples of the vicarious nucleophilic addition.From the practical standpoint, the reac- tions of soft nucleophiles with the aminoalkenes 2d provide a new convenient procedure for the one-pot synthesis of derivatives of N,N-disubstituted a-amino acids. 3. Radical reactions Aminoalkenes containing a geminal activating group are excellent radicalophiles.The stabilising effect of captodative substituents on the radical centre has been postulated more than 20 years ago 1 and was subsequently confirmed by numerous experiments. Ito et al.241 carried out the kinetic study of the addition of arylthiyl radicals to acrylonitriles bearing the electron-donating groupXat the a position. It appeared that the rate constant of this reaction decreases in the series NR2>OR>Cl>OAc> Me>H. This sequence correlates well with their electron-donat- ing ability and indicates that captodative substituents exert the synergistic effect on stability of the radical. It should be noted that the relative activity of amino derivatives is*20 times higher than that of their alkoxy analogues and *550 times higher than the reactivity of unsubstituted acrylonitrile. Consequently, the ami- noalkenes 2 proved to be the most efficient radical traps of all the known compounds.CN CN + 4-YC6H4S 4-YC6H4S X XO, OEt, Cl, OAc, Me, H; Y=H, Me, But, OMe, Cl, Br. X=NCaptodative aminoalkenes This ability of captodative systems finds versatile synthetic applications.4, 132, 242 Below are given the possible pathways of transformations of the radical intermediate giving rise to various polyfunctional compounds with high selectivity. R2N EWGX XR2N EWG EWG X EWG EWG X NR2 +X X NR2 EWG NR2 2a ± d X NR2 EWG X NR2 EWG =CN (a), CO2R (b), C(O)R (c), CHO (d); R=Alk, Ar. The behaviour of captodative aminoalkenes in photolysis is very sensitive to insignificant structural changes in the substrate and primarily to the nature of the substituents at the nitrogen atom. Conceivably, that is why it is difficult to make general- isations and predict the direction of the reaction based on the available data.In most cases, photolysis of the 2-aminocyclohex-2-enones 56b [R1=Et, R2=Me; R1±R2=(CH2)4, (CH2)2OCH2] con- taining the a-hydrogen atoms in the substituents at the nitrogen atom is accompanied by elimination of one of these hydrogen atoms. Subsequent cyclisation of the resulting biradical inter- mediate affords bicyclic azetidine derivatives 108.189, 243 Ana- logues of the compounds 56b bearing the methyl group at position 3 produced the azetidines 108 in low yields.189 In some cases, the reactions of the aminoalkenes 56b performed under the same conditions unexpectedly gave rise to spiroaziridines 109, which were isolated as alcohols 110 after reduction of the carbonyl group.119, 244 It is believed that the aziridines 109 were formed through a-hydrogen abstraction followed by the rearrangement of unstable azabicyclooctenol.119 O NR1 R2 O 108 (30% ± 65%) RN1 hn R2 OH O R1 N R1 N NaBH4 56b R2 R2 110 (35% ± 80%) 109 (40%) 108: R1=Et, R2=Me; R1±R2=(CH2)4, (CH2)2OCH2; 109, 110: R1=H, R2=Ph; R1=Bn, R2=Ph; R1=Me, All, R2=CH=CH2 . Photochemical cyclisation of N-arylaminocycloalkenones afforded carbazole derivatives 111 and 112.140 O OH O RN RN N(Ph)R hn + 112 (10%) 111 (40% ± 50%) 56b R=Me, Et.211 Radical cyclisation of 2-(haloarylethyl)aminocycloalk-2- enones proceeds both at the C2 atom of the enone fragment to form aminoalkenes, which are azepine derivatives, and at the nitrogen atom to produce compounds of the indole series.245 The reactions of 2-aminotropones with palladium acetate, which are considered to be a catalytic radical process, differ essentially from the reactions of other troponoid compounds performed under the same conditions.246 Depending on the substituents at the nitrogen atom, either acetoxylation (NR2=NMe2, NEt2) or arylation [NR2=N(Me)Ph] were observed. O AcO NR2 NR2=NMe2, NEt2 O NR2 (20% ± 35%) Pd(OAc)2 PhH, D O N(Me)Ph Ph NR2=N(Me)Ph (5%) The photochemical behaviour of the acyclic aminoalkenones 2c differs from that of the 2-aminocyclohex-2-enones 56b.Thus in photolysis of aminoalkenone 113 in ether or pentane, allylic hydrogen abstraction proceeded more efficiently to give isomeric nonconjugated alkene 114 as the only reaction product.177 In the absence of the allylic hydrogen atom, for example, in aminoalkene 115, the substrate underwent rapid E,Z isomerisation followed by slow formation of a mixture of compounds consisting primarily of cyclisation product 116 and elimination product 117.177 O O Me CH2 Me Me Me Me N N hn 113 114 (50%) O O Ph O Ph Ph H Ph Ph + R1N N R1 117 (10% ± 30%) R2 115 R2 116 (40% ± 55%) R1=Et, R2=Me; R1±R2=(CH2)4, (CH2)2OCH2. Intramolecular photocyclisation of 2-anilinoalkenonitriles gave rise to 2-cyano-1-methyl-3-R-indoles (R=Me, Ph, CH=CHMe).Under the action of Bu3SnH in the presence of azoisobutyronitrile, radical cyclisation of 2-anilinoacrylonitriles containing the haloalkyl (alkenyl) substituent at position 3 gave rise to substituted cycloalkanes or cycloalkenes.63, 247 4. Captodative aminoalkenes in the synthesis of carbo- and heterocyclic compounds Due to the presence of multiple carbon ± carbon and carbon ± heteroatom bonds, captodative alkenes can be used in the synthesis of carbo- and heterocyclic compounds. Cycloaddi- tion reactions are apparently the most important transformations of captodative aminoalkenes. These reactions have been exten- sively studied in the last decade. The aminoalkenes 2a ± c can be involved in [2+2]-, [3+2]- and [4+2]-cycloaddition reactions, the latter reactions being most common.It is well known that dienophiles containing electron-with- drawing substituents are readily involved in the Diels ± Alder212 reactions with electron-rich dienes. It is not surprising that [4+2]- cycloaddition of simple enamines to nonactivated dienes occurs very rarely. Unlike simple enamines, captodative aminoalkenes successfully react with various dienes, including nonactivated compounds. The reactions of the 2-aminocycloalk-2-enones 56a,b with electrophilic diazenes 118 are formally classified as [4+2]-cyclo- addition.248 ± 253 Stability of 1,3,4-oxadiazines 119 thus formed depends both on the nature of the amino group and the ring size of the substrate as well as on the presence of the functional groups in the diazene 118.Generally, aminoenones bearing a disubstituted amino group form rather stable cyclic derivatives 119, which are slowly transformed into acyclic isomers 120 at room temperature or upon heating.248 ± 250 Analogous [4+2]-adducts were also obtained from N-aryl-substituted aminocycloalkenones.252 In other cases, the formation of the heterocycles 119 is not necessarily detected and the enediamines 120 are the only reaction products, which are readily rearranged into more stable isomers.251 ± 253 Benedetti et al. 251 assumed that the compounds 120 can also be formed immediately from the starting substrates.O O NR1R2 C(O)Y NR1R2 Y O N + N N ( )n ( )n X NX118 56a,b 119 (21%± 95%) X O O NR2 NR1R2 R1=H NHC(O)Y NHC(O)Y ( )n ( )n NH NX 120 (21% ± 100%) O; n = 0 (a), 1 (b); R1=H; R2=Bun, But, Ph; NR1R2=N ,N X=Ph, 4-NO2C6H4, PhCO, CO2Me, CO2Et; Y=Ph, 4-NO2C6H4, OEt. The reactions of the cyclic and acyclic aminoalkenones 56b and 2c (as dienophiles) with nitroalkenes (as heterodienes) were used for the synthesis of various carbo- and heterocyclic systems, viz., tetrahydroindoles,141 polyquinanes 254, 255 and penta- lenes.129, 256 ± 258 In most cases, the authors of the cited studies postulated that 1,2-oxazine N-oxides 121 were formed as inter- mediates whose subsequent transformations depend primarily on the nature of the substituents and the reaction conditions. Some heterocyclic derivatives 121 were so stable that they can be isolated in quantitative yields.141, 257 ± 259 In a number of studies,129, 141 [2+2]-cycloaddition products or Michael adducts were detected along with the compounds given below.O R3R4N O NR3R4 R5 O2N N O R1 R1C(O) + R5 R2 R2 R6 2c R6 121 (70% ± 100%) R1=Et, Ph; R2=H, Me, Ph; R1±R2=(CH2)3; NR3R4=NHBun, NHBut, NHPh,N , N ,N O; R5, R6=H, Me, Ph. R5 R5 R6 R6 O2N O2N R1 R1 R2 R2 HO HO 121 (R1=Et, Ph; R2=H, Me, Ph) O NR3R4 (62% ± 96%) (70% ± 96%) R5=H, Me, Ph; R6=H, Me. A Yu Rulev O R4R3N R5 R3 6àH NO2 (*100%) R6 O 121 [R1±R2= (CH2)3] R4 N R3=H R5 (65% ± 80%) R6 R5=H, Me, Ph; R6=Me, Ph.DoÈ pp and co-workers 260 ± 264 performed extensive studies on cycloaddition of various captodative alkenes to unsaturated systems over many years. The Diels ± Alder reactions of the cyanoenamines 2a with 1-naphthaldehyde (122a), 1-acetonaph- thone (122b), 1-naphthophenone (122c) or methyl 1-naphthoate (122d) proceeded with high regio- and stereoselectivity to form adducts 123a ± d.36, 170, 261 ± 263 By-products of [2+2]-cycloaddi- tion were obtained in negligible yields (<1%) (in some cases, these compounds were not detected at all). In contrast, the cyanoenamine 2a (NR2 is morpholino) is involved in [2+2]- cycloaddition with naphthalene-1-carbonitrile 122e to give a mixture of isomeric polycyclic compounds 124 and 125 in low yields.264 X CN hn + NR2 122a ± e 2a X CN NR2 123a ± d (19% ± 72%) NR2 NR2 CN NC CN CN X=CN + 125 (9%) 124 (14%) X=CHO (a), C(O)Me (b), C(O)Ph (c), CO2Me (d), CN (e); NPh, N Ph, N S, N O, N N NC6H4CF3-3, NR2= N HN OMe.N, S The reactions of the aminoalkene 2a (NR2 is morpholino) with substituted 1-acetonaphthones afford the [2+2]- and [4+2]- cycloaddition products in a ratio which depends on the position of the substituent. The reaction with 2-methoxy-1-acetonaph- thone produced only the [2+4]-adduct, whereas derivatives bear- ing a substituent at position 4 gave rise to [2+2]-cycloaddition products.265 Under the conditions of photolysis, 2-morpholino- acrylonitrile was involved in the [2+2]-cycloaddition reaction with 3-acetylthianaphthene.266 Unlike unsubstituted enamines, the cyanoenamines 2a react with a number of nonactivated conjugated dienes to give [4+2]- cycloaddition products 126 and 127 (Scheme 1).The acrylonitrile 2a (NR2 is morpholino) proved to be more reactive as a dienophile than its analogue containing the aromatic amino group, the regioselectivity of the addition being dependent on the nature of the substituents in the diene (see Scheme 1).Captodative aminoalkenes Scheme 1 R5 CN R6 R4 + NR12 R3R2 2a R2 R3 R4 R6 R5 + CN CN NR1 R5 R6 2 NR12R4 R3 R2 126 127 Ref. R6 R5 R3 R4 R2 NR12 Total yield (%) (ratio 126 : 127) 32, 267 62 ± 72 (8 : 2) H Me H H H N O 32 24 (10 : <1) H Me H H H N(Me)Ph 32, 267 81 ± 92 Me Me H H H N O 32 18 Me Me H H H N(Me)Ph 32 H 54 H H CH2 N O N(Me)Ph 32 H 55 H H CH2 32 H 92 H H (CH2)2 N O N(Me)Ph 32 H 65 H H (CH2)2 F F H 268 81 (10 : <1) OSiPh3 H O N The above-considered reactions clearly demonstrate that captodative aminoalkenes are good dienophiles, which are much superior to the corresponding alkoxy and thio analogues in reactivity.269 These compounds are equally efficient as dienes provided that the b-substituent contains one or several multiple bonds.25, 26, 270 The reactions of dienes 128a,b with derivatives of acrylic acid afforded unexpected cyclisation products.69, 271 Aminonorborne- nones 129a,b were formed through the [4+2]-cycloaddition of methyl acrylate or acrylonitrile to aminocyclopentadienones 130a,b generated upon intramolecular condensation of the dien- amino esters 128a,b.The [2+2]-cycloaddition products 131a,b were obtained as by-products. The proportion of the [4+2]- cycloaddition products 129a,b depends on the nature of the EWG NR22 CO2Me R1 NR22O R1 130a,b 128a,b O NR22 R1 EWG 129a,b (26% ± 75%) O EWG R22 N R1 131a,b (10% ± 30%) EWG=CO2Me, CN; R1=Me (a), Et (b); NR22 =NMe2, NEt2,N . 213 amino and activating groups, viz., the highest yields were achieved in the case of the dienes 128a,b containing the piperidine sub- stituent. The captodative alkenes 2a ± d are excellent substrates in [2+2]-cycloaddition. The reactions of these compounds with substituted alkenes or carbonyl compounds produced respectively cyclobutanes 272 or oxetanes (the Paterno ±BuÈ chi reaction).273, 274 Thus the addition of symmetrical 1,2-diketones to the cyanoen- amines 2a proceeded regio- and stereospecifically to give head-to- head [2+2]-adducts with the cis-oriented amino and aryl groups (the configuration of the resulting diastereomer determined in the study 273 is, apparently, in error).Under irradiation of a mixture of unsymmetrical diones and the cyanoenamines 2a, the addition at the carbonyl groups proceeded with comparable rates to give a mixture of isomeric oxetanes in 9%± 90% yields.274 O CN hn Ar2 + Ar1 PhH NR2 O 2a O O + NC C(O)Ar1 NC C(O)Ar2 Ar2 Ar1 R2N R2N Ar1=Ar2=Ph, 4-MeC6H4, 4-MeOC6H4, 4-ClC6H4, 4-CF3C6H4, 2-C10H7; Ar1=Ph: Ar2=4-MeOC6H4; 2-C10H7; NR2=N ,N , N N ,N , O, N Ph.Examples of head-to-head and head-to-tail photochemical dimerisation of captodative aminoalkenes are available in the literature,5, 6, 275 the type of dimerisation being dependent on the structure of the substrate. The result of the reaction depends also on the reaction conditions. For instance, 2-aminoacrylonitrile underwent dimerisation in the presence of benzophenone as a photosensitising agent to form a mixture of diastereomers 132.5 However, the formation of the cyclobutanes 132 was not observed in the absence of Ph2CO. In the latter case, the main reaction involved fragmentation of the starting substrate to yieldHCNand MeCN accompanied by competitive isomerisation into aziridine- 2-carbonitrile.6 NH2 Ph2CO CN CN CN NH2 132 (74%) NH2 H 2a MeCN +HCN+ N CN (10% ± 20%) The captodative aminoalkenes 2a are much more reactive than their push-pull isomers 1a.These compounds react with dimethyl acetylenedicarboxylate or dibenzoylacetylene by being involved in the so-called `a-cyclisation of tertiary amines'.276 ± 278 Depending on the nature of the amino group in the aminoalkenes 2a and the activating group in the acetylene derivative, either furans 133a,b and 134c ± f or dihydropyrroles 135a ± g were obtained as the final reaction products. The preparation of dihydropyrroles, which were formed with high stereoselectivity, required more drastic conditions (refluxing in acetonitrile or DMSO).276, 279 The reaction mechanism includes [2+2]-cyclo- addition and rapid recyclisation of adducts 136 into dienes 137, which are the key intermediates for all products.The dienes 137a,c are involved in 1,3-dipolar addition reactions with dipolarophiles, such as acrylonitrile, yielding finally a mixture of cycloadducts 138a,c and the dihydropyrroles 135a,c (Scheme 2).278214 R2 R1 R1 R2 X C C X N N CN CN X X 2a 136a ± f Ph O Ph X=C(O)Ph Me O 137a,b R2 NC N R1 Ph O Ph X=C(O)Ph Me O 137c ± f NC N R2 R1 Me MeO2C MeO2C X=CO2Me 137a ± g N NC R2 R1 135a ± g (37%± 91%) CO2Me CN CO2Me Me 137a,c NC X=CO2Me N R2 R1 138a,c (49% ± 91%) R1=R2=H(a), Me (b); R1±R2=(CH2)2 (c), (CH2)3 (d), (CH2)4 (e), (CH2)5 (f), CH2OCH2 (g); X=C(O)Ph, CO2Me.In [3+2]-cycloaddition, captodative aminoalkenes behave analogously to common enamines. Like enamines,175 these com- pounds add aryl azides to give the corresponding dihydrotria- zoles.11, 33, 280, 281 However, some of these cyclic derivatives are unstable and cannot be isolated. For example, dihydrotriazoles 139 formed in the reactions of the cyanoenamines 2a underwent CN R2N + XC6H4N3 NR2 XC6H4 2a 139 R2N N N XC6H4 N 140 (75% ± 92%) NR2=N ,N ,N O; X=H, 4-Me, 3-NO2, 4-Br, 3-Cl. Scheme 2 R1 R2 N CN X X 137a ± f Ph O Me Ph N NC R1 133a,b (26% ± 55%) Ph O Me Ph N NC CHO 134c ± f (24% ± 73%) H+ 135a,c (0% ± 8%) NC N N N A Yu Rulev spontaneous HCN elimination to generate 1,2,3-triazoles 140.3, 280 The reactions always proceed regiospecifically to yield the only isomer.Like enamines, the b-oxophosphonate 6 reacted with tosyl azide to give diazo compound 141, which underwent cyclisation into pyridazine 142.22 P(O)(OMe)2 N2 P(O)(OMe)2 TsN3 O O Ph Ph N N 141 6 N P(O)(OMe)2 N OH Ph N 142 In spite of the fact that the reactions of 2-morpholinoacrylo- nitrile with aldonitrones 143 afforded morpholides 144, DoÈ pp and Walter 282 considered these reactions as 1,3-dipolar cycloaddition. Actually, the formation of the amides 144 can be attributed to elimination ofHCNfrom intermediate isoxazolidine 145 followed by the ring opening.Hence, this cycloaddition reaction is also regiospecific, the terminal carbon atom of the starting amino- nitrile being nucleophilic. Keto nitrones did not react with 2-morpholinoacrylonitrile.282 R2 CN R2 N N R1 + R1 N O N 7HCN O O O7 143 CN 145 R2 O N R1 HN O 144 (32% ± 98%) Cycloaddition of 1,3-dipolar compounds 146a,b to 2-pyrroli- dinocyclohept-2-enone afforded initially compound 147b, which lost amine upon acid treatment to form pyrazolone derivative 148a or isoxazolone derivative 148b.283 Ph PhH H+ N PhC N X + X N 146a,b N O O 147b (89%) Ph N X O 148a,b (55% ± 73%) X=NPh (a), O (b). 5. Other reactions most typical of captodative aminoalkenes The captodative aminoalkenes 2a ± c are involved in some reac- tions, which cannot be unambiguously classified with the above- considered types.The most typical reactions, which are notCaptodative aminoalkenes associated with the presence of a particular specific substituent, are considered in this section. The aminoalkenes 2a ± c are rather readily subjected to hydro- genation. Under traditional conditions of heterogeneous cataly- sis, selective reduction of the C=C double bond generally takes place.17, 126, 284 In some cases, the reactions are accompanied by simultaneous reduction of the activating group. For example, the addition of hydrogen to the enamines 59a,b afforded the corre- sponding 2-aminocyclohexanones 60a,b. However, hydrogena- tion of the substrate 59c performed under the same conditions gave rise to a mixture of diastereomers of cis-1,2-aminoalcohols 149 in quantitative yield.126 O NH CO2Et H2 , 5% Pd/C R 59a ± c O NH CO2Et R =H, Me R 60a,b OH NH CO2Et R=4-HOC6H4CH2 CH2C6H4OH-4 149 R = H (a), Me (b), 4-HOC6H4CH2 (c).Of a wide variety of different activated alkenes, only the captodative aminoalkenes 2b appeared to react with hydrosilanes with reduction of the C=C double bond.285 CO2Et CO2Et ButHgCl + RnSiH47n DMSO Me NR2 NR2 2b (60% ± 87%) O; n=1±3. NR2=NEt2, N The reactions with the use of metal aluminium hydride or borohydride as reducing agents either proceeded only at the activating group (C=O or CN) 17, 150 or resulted in simultaneous reduction of both functional groups.159 Wittig olefination of optically active 2-(methoxymethyl)pyr- rolidinobut-3-en-2-one afforded chiral 2-amino-substituted buta- 1,3-diene 150.This compound is the first representative of the compounds, which have previously been difficult to access.133 OMe OMe PPh3 CH2 N N THF, 20 8C O 150 (59%) Noteworthy are the palladium-catalysed reactions of 2-(N- methylanilino)alkenonitriles with iodobenzene.286 Owing to the studies performed by De Kimpe and co-workers,184, 195 this trans- formation of cyanoenamines became an efficient procedure for the synthesis of the corresponding amides. O CN Pd(OAc)2 R + PhI R N(Me)Ph N(Me)Ph (50% ± 86%) 215 Captodative aminoalkenes readily undergo isomerisation. As mentioned above, the aminoenones 2c and the aminoenals 2d can be very readily subjected to Z,E isomerisation, which either occurs in the course of their preparation or subsequent transformations or takes place upon storage.In addition to Z,E isomerisation, examples of rearrangements of these compounds under the action of both acidic 287 and basic reagents 183 are available in the literature. Analysis of the most typical transformations of the captoda- tive aminoalkenes 2a ± d shows that these compounds can exhibit the properties of either enamines or activated alkenes depending on various factors. The balance in favour of the canonical structure of enamine can be rather easily disturbed upon insignif- icant changes in the nature of the activating or amino group.V. Methylidene- and arylmethylidene- quinuclidin-3-ones Methylidene- and arylmethylidenequinuclidin-3-ones 151 not only serve as the key intermediates in the design of new pharma- ceuticals but are also of interest in theoretical organic chemistry. These compounds can formally be considered as bicyclic capto- dative aminoalkenes. However, they cannot exhibit properties of enamines because of the fixed spatial orientation of the lone electron pair of the nitrogen atom. This is the essential distinction between the compounds 151 and the above-considered systems. The 2-arylmethylidenequinuclidin-3-ones 151 are generally prepared by base-catalysed condensation of quinuclidin-3-one 152 with various arene- or hetarenecarbaldehydes.288 ± 299 The compounds 151 are obtained as individual isomers or as mixtures of geometric isomers in yields from moderate to quantitative. The same methodology was successfully applied to the synthesis of ferrocenylmethylidene-substituted 300 and tricyclic analogues of quinuclidinone.301 O O RCHO R N N 152 151 (37%� *100%) R=Ph, 2-PhC6H4, 2-HOC6H4, 2-MeOC6H4, 2-FC6H4, 2-ClC6H4, , , 3-HOC6H4, 4-HOC6H4, 4-MeOC6H4, 4-ClC6H4, 4-NO2C6H4, 4-Me2NC6H4, 4-PhOCH2CH2OC6H4, 2,4-(MeO)2C6H3, 3,4-(MeO)2C6H3, 3,4,5-(MeO)3C6H2, BnOC6H4, C5H5FeC5H4, O, .O S O The cyclopropane-ring opening in spirocyclic quinuclidinone derivative 153, which occurs under the action of C-, N- or O- nucleophiles, provides an interesting example of the preparation of b,b-disubstituted methylidenequinuclidones.302 Thus the reac- tion of the compound 153 with lithium diphenylcuprate afforded a mixture of isomers 154 and 155.The compound 154 was com- pletely transformed into the quinuclidinone 155 during 20 h (the ratio Z:E=3 : 1).302 O O O LiCuPh2 NO2 Ph Ph 20 8C, 20 h N N N 153 Ph Ph 154 Bn 155 The quinuclidinone 151 (R=H) containing the terminal double bond was smoothly generated from the ketone 152 under the conditions of the Mannich reaction 303 as well as by dehydra-216 tion of 2-hydroxymethyl-3,3-dihydroxyquinuclidine hydrochlor- ide (156).304O Me2NH + CH2 O N O 152 N CH2 OH K2CO3 151 (93% ± 96%) . HCl OHOH N156 The orthogonal arrangement of the lone electron pair of the nitrogen atom hinders p,p-conjugation as a result of which the amino group exerts only the inductive effect on the double bond.This characteristic feature of the arylmethylidenequinuclidinones 151 is reflected in the observed downfield shifts of the signals for the olefinic proton (d=6.80 ± 7.50) in the 1H NMR spectra and for the b-carbon atom (d=125 ± 146) in the 13C NMR spectra as compared to the signals of the above-considered captodative aminoalkenes 2c or 56b. Hence, the quinuclidinones 151 would be expected to exhibit properties of strong amines in reactions with electrophilic reagents and properties of Michael acceptors in reactions with nucleophiles. O Nu E N R Like captodative aminoalkenes, the arylmethylidene- and ferrocenylmethylidenequinuclidinones 151 are protonated and alkylated at the nitrogen atom.299, 305 ± 307 However, the same regioselectivity of the electrophilic attack is accounted for by different reasons.Thus enammonium salts of captodative amino- alkenes are highly stable due primarily to the stabilising effect of an intramolecular hydrogen bond, whereas N-protonation (alky- lation) of the quinuclidinones 151 is associated only with the basicity of the nitrogen atom. Bubbng of dry HCl through a solution of methylidenequi- nuclidinone led to quaternisation of the nitrogen atom to form compound 157 (R=H) along with the anti-Markovnikoff addi- tion of hydrogen halide to the double bond.305 In the case of the ferrocenylmethylidenequinuclidinones 151 (R=Fc), two hetero- atoms are simultaneously protonated.The formation of the dications 158 was confirmed by NMR spectroscopy.299, 307O O R=H,X=Cl Cl7 X7 + + HCl NH O R HX NH 157 Cl (*100%) N OH R=Fc, X=BPh4, CF3CO2 R 151 + 2X7 +NH Fc 158 R=H, Ph, 4-FC6H4, 2,4-(MeO)2C6H3, Fc; X=Cl, BPh4, PhCO2, CF3CO2. Under the action of gaseous HCl in chloroform 288, 290, 295, 308 or of NaBPh4 in acetic acid,299, 307 the compounds 151 underwent reversible Z,E isomerisation. For example, Z-2-arylmethylidene- and Z-2-ferrocenylmethylidenequinuclidinones are readily (often A Yu Rulev quantitatively in a matter of minutes) transformed into the corresponding E isomers.295, 299, 307 In this case, there is the evident analogy with enammonium salts of captodative amino- alkenes, viz., isomerisation becomes possible due to a decrease in the double-bond character of the alkene fragment.299 These transformations of the quinuclidinones 151 provide the means of performing the selective synthesis of their individual geometric isomers, which is of importance in studying the structure ± pro- perties relationships (including, the structure ± biological activity relationships). OH O O + 2X7 + R N N NH R R E-151 Z-151 158 The fact that the quinuclidinones 151 cannot exhibit the properties of enamines is reflected in their reactions with nucleo- philic reagents.For example, the majority of the reactions of the compounds 151 and their tricyclic analogues with various C-nucleophiles proceed as conjugated 1,4-addi- tion.294, 301, 302, 309 ± 314 In some cases, Michael adducts 159 ± 162 are obtained in moderate yields, which is attributable to compet- itive 1,2-addition giving rise to compounds 163.The latter com- pounds were obtained as the major products in the reactions of the quinuclidinones 151 with aryl- or alkyllithium derivatives (Scheme 3).299, 300, 307, 315, 316 As mentioned above, numerous attempts to perform the Michael addition to captodative amino- alkenes failed almost without exception (see Section IV.2). Alcohols serve as O-nucleophiles in the Michael addition reactions with methylidenequinuclidin-3-one, the reactions pro- Scheme 3 O PhMgBr Ph N R1 159 (38% ± 72%) O R2CH2R3 R2 B R3 N O R1 160 (25% ± 70%) N R4 (R4=H, Ph) R1 151 O NH NH EtONa or AcOH N R1 R4 161 (19%� *100%) OH O R5Li R5 + (R5=H, Ph) R5 N N R1 R1 162 163 R1=H, Ph, 2-ClC6H4, 4-MeOC6H4, Fc; R2=CO2Me, R3=CN; R2=Ph, R3=CN; R2=Ph, R3=N=CHPh; R2=H,R3=NO2; R2=C(O)Me, R3=CO2Et; R2=C(O)Me, R3=C(O)Ph; O R2=R3=C(O)Me, R2R3CH2= Me; B=NaOH, MeOH, BuLi.Me OCaptodative aminoalkenes ceeding slowly.306 As expected, aniline forms the corresponding Schiff's base with the latter compound.317 The presence of two electrophilic centres, viz., the carbon atom of the carbonyl group and the b-carbon atom of the double bond, allows the use of the quinuclidinones 151 for the construc- tion of various heterocyclic systems. For example, the reactions of the ketones 151 (R=H, Ph or Fc) with hydrazine afforded quinuclidine derivatives 164 or 165 depending on the reaction conditions.Under the action of basic reagents (KOH or an excess of hydrazine), the latter compounds were decomposed upon heating to give pyrazoles 166.290 ± 292, 296, 297, 300, 317, 318 Analogous transformations were observed in the reactions of the compounds 151 with thiosemicarbazide.318 Phenylhydrazine reacts analo- gously to primary amines to produce phenylhydrazone.290 O H2NNH2 OHNHNH N N R 164 (65% ± 90%) 151 HN N NNH N R 165 (21% ± 90%) 166 (68% ± 88%) R=H, Ph, Fc. The reactions of hydroxylamines with the arylmethylidene- quinuclidinones 151 bearing a substituent at the double bond afforded oximes 167,298, 319, 320 whereas the reaction of the unsub- stituted methylidenequinuclidin-3-one 151 (R1=H) with hydrox- ylamine gave rise to oxime 168, which was isolated as dihydrochloride.321 It was assumed that this reaction proceeded through the intermediate formation of 4a-hydroxyisoxazoli- dino[4,5-b]quinuclidine 169.The N-substituted derivatives of the isoxazolidine 169 were prepared in high yields by the reactions of quinuclidinone 151 (R1=H) with N-arylhydroxylamines.321 O R2NHOH N R1 151 N OH R1 6à H N R1 167 (18% ± 99%) OHO R1=H NR2 N 169 (70% ± 90%) R1=H, Ph, 2-HOC6H4, 3-HOC6H4, 4-HOC6H4, 2-MeOC6H4, 4-MeOC6H4, 4-Me2NC6H4, 4-NO2C6H4; R2=H, Ph, 4-MeC6H4, 4-ClC6H4, 3-ClC6H4. The compounds 151 equally readily reacted with thiourea and phenylthiourea to produce pyrimidinethiones 170 318 or 171.290, 300, 322 Finally, cyclisation of the ketones 151 with cyano- thioacetamide gave rise to the corresponding thiones 172.323 R R NH OH N OHO R2=H N N N 168 (83%) 217 S HO HN R1=H NR2 O N 170 (43% ± 84%) H2NC(S)NHR2 N S NH R1 151 R1 6à H NR2 N R1 171 (36% ± 92%) R1=H, Ph, 4-MeOC6H4, Fc; R2=H, Ph.HN O S H2NC(S)CH2CN N N CN R1 151 R1 172 (58% ± 82%) R1=Ph, 4-FC6H4 . As in the case of captodative aminoalkenes, reduction of methylidene- and arylmethylidenequinuclidinones can be directed with high selectivity toward either the C=C double bond or the carbonyl group by varying the nature of the reducing agent.When subjected to platinum- or palladium-catalysed hydrogenation, methylidene- and arylmethylidenequinuclidinones were trans- formed into saturated ketones 173.288, 291, 292, 298, 305 The corre- sponding allylic alcohols 174 were selectively obtained with the use of NaBH4.288, 308, 324, 325 Stereospecific reduction of both functional groups in the benzylidenequinuclidinone 151 (R=Ph) with lithium aluminium hydride afforded benzylquinu- clidinol trans-175 (R=Ph).326 Catalytic hydrogenation of the same compound 151 (R=Ph) followed by reduction with NaBH4 gave rise to cis-175 (R=Ph).308O OH R=Ph H2, Pd/C NaBH4 N N Ph R 173 (43% ± 92%) cis-175 (78%) O OH NaBH4 N N R R 151 174 (75%790%) OH R=Ph LiAlH4 N Ph trans-175 (78%) R=H, Ph, 2-ClC6H4, 4-MeOC6H4, 3,4-(MeO)2C6H3. VI.Conclusion Recent progress in the chemistry of captodative alkenes allows the preparation of many aminoalkenes containing the geminal acti- vating group, which can be considered as polyfunctional building blocks. Various transformations characteristic of this type of compounds, such as electrophilic or nucleophilic addition and, in particular, cyclisation, can be used in organic synthesis. Unlike the push-pull enaminocarbonyl compounds 1c for which a special type of non-cyclic aromatic stabilisation was assumed,327 the cross-conjugated aminoalkenes 2a ± d act as `chemical chame-218 leons'.The latter compounds are able to exhibit the properties of either enamines or activated alkenes resulting from insignificant changes in the structure of the substrate and the reaction con- ditions. Owing to these properties, captodative aminoalkenes are unique building blocks and there is reason to hope that they will find increasing use in organic synthesis. References 1. H G Viehe, R Mere'nyi, Z Janousek Pure Appl. Chem. 60 1635 (1988) 2. H G Viehe, Z Janousek, R Mere'nyi, L Stella Acc. Chem. Res. 18 148 (1985) 3. H G Viehe, R Mere'nyi, L Stella, Z Janousek Angew. Chem. 91 982 (1979) 4. L Stella, Z Janousek, R Mere'nyi, H G Viehe Angew. Chem. 90 741 (1978) 5. G Ksander, G Bold, R Lattmann, C Lehmann, T FruÈ h, Y-B Xiang, K Inomata, H-P Buser, J Schreiber, E Zass, A Eschenmoser Helv.Chim. Acta 70 1115 (1987) 6. S Drenkard, J Ferris,A Eschenmoser Helv. Chim. Acta 73 1373 (1990) 7. E Wagner, Y-B Xiang, K Baumann, J GuÈ ck, A Eschenmoser Helv. Chim. Acta 73 1391 (1990) 8. D MuÈ ller, S Pitsch, A Kittaka, E Wagner, C E Wintner, A Eschenmoser Helv. Chim. Acta 73 1410 (1990) 9. Y-B Xiang, S Drenkard, K Baumann, D Hickey, A Eschenmoser Helv. Chim. Acta 77 2209 (1994) 10. J V Greenhill Chem. Soc. Rev. 6 277 (1977) 11. U KucklaÈ nder, in The Chemistry of Enamines (Ed. Rappoport) (New York: Wiley, 1994) p. 523 12. Ya F Freimanis Khimiya Enaminoketonov, Enaminoiminov, Enaminotionov (The Chemistry of Enamino Ketones, Enamino Imines and Enamino Thiones) (Riga: Zinatne, 1974) 13.H E Gottlieb, in The Chemistry of Enones (Eds S Patai Z Rappoport) (New York: Wiley, 1989) p. 129 14. C Cimarelli, G Palmieri Recent Res. Dev. Org. Chem. 179 (1997); Chem. Abstr. 130 22 2781 (1999) 15. Yu V Smirnova, Zh A Krasnaya Usp. Khim. 69 1111 (2000) [Russ. Chem. Rev. 69 1021 (2000)] 16. P Lue, J V Greenhill Adv. Heterocycl. Chem. 67 207 (1997) 17. G I Polozov, I G Tishchenko Vestn. Belorus. Univ., Cep. 2 (3) 3 (1984) 18. U Schmidt, A Lieberknecht, J Wild Synthesis 159 (1988) 19. B Costisella, H Gross Z. Chem. 27 143 (1987) 20. B Costisella, H Gross Tetrahedron 38 139 (1982) 21. M E Niyazymbetov, V A Petrosyan, I Kaitel', B Kostizella, K Kh Shvarts Izv. Akad. Nauk SSSR, Ser. Khim.172 (1989) a 22. D Collomb, C Deshayes, A Doutheau Tetrahedron 52 6665 (1996) 23. H Ahlbrecht, K Pfaff Synthesis 897 (1978) 24. K Takahashi, K Shibasaki, K Ogura, H Iida J. Org. Chem. 48 3566 (1983) 25. J-M Fang, C-C Yang J. Chem. Soc., Chem. Commun. 1356 (1985) 26. J-M Fang, C-C Yang, Y-W Wang J. Org. Chem. 54 477 (1989) 27. J-M Fang, C-C Yang, Y-W Wang J. Org. Chem. 54 481 (1989) 28. A Jonczyk, Z Owczarczyk Synthesis 297 (1986) 29. B Leseche, J Gilbert, C Viel J. Heterocycl. Chem. 18 143 (1981) 30. Z Miura, Y Iwasaki, K Tatsuta J. Antibiot. 47 1171 (1994) 31. S C Temin J. Org. Chem. 22 1714 (1957) 32. H Ahlbrecht, K Pfaff Synthesis 413 (1980) 33. D DoÈ pp, M Pies J. Chem. Soc., Chem. Commun. 1734 (1987) 34. J-L Boucher, L Stella Tetrahedron 41 875 (1985) 35.A Derdour, T Benabdallah, B Merah, F Texier Bull. Soc. Chim. Fr. 127 69 (1990) 36. J-M Fang, L-F Liao, C-C Yang Proc. Natl. Sci. Counc., Part A 9 1 (1985) 37. N De Kimpe, R Verhe', L De Buyck, H Hasma, N Schamp Tetrahedron 32 3063 (1976) 38. N De Kimpe, R Verhe', L De Buyck, J Chys, N Schamp Bull. Soc. Chim. Belg. 88 695 (1979) 39. K JaÈ hnisch, E Weigt, E Bosies Synthesis 1211 (1992) 40. R Verhe', N De Kimpe, L De Buyck,M Tilley, N Schamp Bull. Soc. Chim. Belg. 86 879 (1977) 41. N De Kimpe, R Verhe', L De Buyck, N Schamp Chem. Ber. 116 3846 (1983) A Yu Rulev 42. H Ahlbrecht, D Liesching Synthesis 495 (1977) 43. J-M Fang, C-C Chen. J. Chem. Soc., Perkin Trans. 1 3365 (1990) 44. C-C Chen, S-T Chen, T-H Chuang, J-M Fang.J. Chem. Soc., Perkin Trans. 1 2217 (1994) 45. J Legters, L Thijs, B Zwanenburg Recl. Trav. Chim. Pays-Bas 111 16 (1992) 46. A K Sharma, A K Saha, V S Chauhan Indian J. Chem. B24 7 (1985) 47. A Josczyk, Z Owczarczyk,M Makosza, J Winiarski Bull. Soc. Chim. Belg. 96 303 (1987) 48. J-L Klein, J-C Combret Bull. Soc. Chim. Fr., Part II 28 (1983) 49. H Plieninger, R El-Berins, H Mah Chem. Ber. 104 3983 (1971) 50. E Hardegger, F Szabo, P Liechti, Ch Rostetter, W Zankowska-Jasinska Helv. Chim. Acta 51 78 (1968) 51. H Poisel, U Schmidt Angew. Chem. 88 295 (1976) 52. H Poisel Chem. Ber. 110 942 (1977) 53. R A Karakhanov,MM Vartanyan, R B Apandiev, N P Karzhavina Izv. Akad. Nauk SSSR, Ser. Khim. 1905 (1982) a 54. J-M Fang, H-T Chang J.Chem. Soc., Perkin Trans. 1 1945 (1988) 55. C-C Yang, P-J Sun, J-M Fang J. Chem. Soc., Chem. Commun. 2629 (1994) 56. C-C Yang, H-M Tai, P-J Sun J. Chem. Soc., Perkin Trans. 1 2843 (1997) 57. H Ahlbrecht, C Vonderheid Synthesis 512 (1975) 58. J D Albright Tetrahedron 39 3207 (1983) 59. J-M Fang, H-T Chang, C-C Lin J. Chem. Soc., Chem. Commun. 1385 (1988) 60. J-M Fang, C-J Chang J. Chem. Soc., Chem. Commun. 1787 (1989) 61. C-J Chang, J-M Fang, L-F Liao J. Org. Chem. 58 1754 (1993) 62. C-C Yang, J-M Fang J. Chem. Soc., Perkin Trans. 1 3085 (1992) 63. C-C Yang, H-T Chang, J-M Fang J. Org. Chem. 58 3100 (1993) 64. G M Zhdankina, G V Kryshtal', V S Bogdanov, V I Kadentsev, L A Yanovskaya Izv. Akad. Nauk SSSR, Ser. Khim. 346 (1982) a 65.J Mathew, B Alink J. Org. Chem. 55 3880 (1990) 66. R Golse,M Bourhis, J-J Bosc C.R. Hebd. Seances Acad. Sci., Ser. C 287 585 (1978) 67. M Bourhis, J-J Bosc, R Golse J. Organomet. Chem. 256 193 (1983) 68. S Mageswaran, W D Ollis, D A Southam, I O Sutherland, Y Thebtaranonth J. Chem. Soc., Perkin Trans. 1 1969 (1981) 69. M Bourhis, R Golse, M Goursolle, P Picard Tetrahedron Lett. 26 3445 (1985) 70. N Ste'venart-De Mesmaeker, R Mere'nyi, H G Viehe' Tetrahedron Lett. 28 2591 (1987) 71. K Van Sant, M S South Tetrahedron Lett. 28 6019 (1987) 72. N De Kimpe, C Stevens J. Org. Chem. 58 2904 (1993) 73. U Schmidt, E OÈ hler Angew. Chem. 89 344 (1977) 74. F Palacios, I P de Heredia, G Rubiales J. Org. Chem. 60 2384 (1995) 75. E Vedejs, J W Grissom J.Org. Chem. 53 1882 (1988) 76. Y Gelas-Mialhe, E Touraud, R Vessiere Can. J. Chem. 60 2830 (1982) 77. Y Gelas-Mialhe, G Mabiala, R Vessiere J. Org. Chem. 52 5395 (1987) 78. L Reichel, P Pritze Liebigs Ann. Chem. 120 (1974) 79. S Seko, N Tani Tetrahedron Lett. 39 8117 (1998) 80. W Weigel, H-G Henning Chem. Commun. 1893 (1997) 81. A B Koldobskii, Yu N Luzikov, V V Lunin Zh. Org. Khim. 22 636 (1986) b 82. AYu Rulev Usp. Khim. 67 317 (1998) [Russ. Chem. Rev. 67 279 (1998)] 83. A Yu Rulev, J Maddaluno Eur. J. Org. Chem. 2569 (2001) 84. A Yu Rulev, J Maddaluno J. Phys. Org. Chem. 15 (9) (2002) (in the press) 85. G I Polozov, I G Tishchenko Vestn. Akad. Nauk Bel. SSR, Ser. Khim. Nauk (3) 62 (1978) 86. Y D Vankar, A Bawa, G Kumaravel Tetrahedron 47 2027 (1991) 87.H Takeshita, Q F Wang, K Kubo, A Mori Chem. Lett. 993 (1995) 88. G Biggi, F Del Cima, F Pietra J. Am. Chem. Soc. 95 7101 (1973) 89. M Cavazza, F Pietra J. Chem. Soc., Chem. Commun. 897 (1994) 90. M Cavazza, F Pietra J. Chem. Soc., Perkin Trans. 1 2657 (1995) 91. N A Keiko, A Yu Rulev, I D Kalikhman, M G Voronkov Izv. Akad. Nauk SSSR, Ser. Khim. 2610 (1985) a 92. V T Klimko, T V Protopopova, A P Skoldinov Dokl. Akad. Nauk SSSR 146 1084 (1962) c 93. R Gelin, D Makula Bull. Soc. Chim. Fr. 1129 (1968) 94. J SmodisÏ , R Zupet, A PetricÆ , B Stanovnik,M TisÏ ler Heterocycles 30 393 (1990)Captodative aminoalkenes 95. C-g Shin, Y Yonezawa, K Unoki, J Yoshimura Bull. Chem. Soc. Jpn. 52 1657 (1979) 96.C-g Shin, Y Yonezawa, T Obara, H Nishio Bull. Chem. Soc. Jpn. 61 885 (1988) 97. D Knittel, V S Rao Monatsh. Chem. 117 1185 (1986) 98. D Knittel, V S Rao Monatsh. Chem. 119 223 (1988) 99. B Geist, D Knittel Monatsh. Chem. 119 571 (1988) 100. D Knittel Monatsh. Chem. 116 1133 (1985) 101. D M B Hickey, C J Moody, C W Rees J. Chem. Soc., Chem. Commun. 3 (1982) 102. L Henn, D M B Hickey, C J Moody, C W Rees J. Chem. Soc., Perkin Trans. 1 2189 (1984) 103. N D Heindel, N Foster,M Choudhuri J. Org. Chem. 48 3817 (1983) 104. C-g Shin,M Masaki, M Ohta J. Org. Chem. 32 1860 (1967) 105. A G Brown, T C Smale J. Chem. Soc., Perkin Trans. 1 65 (1972) 106. C L Branch, M J Pearson J. Chem. Soc., Perkin Trans. 1 2123 (1982) 107. C-g Shin,M Masaki,M Ohta Bull.Chem. Soc. Jpn. 43 3219 (1970) 108. T Moriya, K Matsumoto,M Miyoshi Synthesis 915 (1981) 109. T Nozoe, K Takase, H Saito, H Yamamoto, K Imafuku Chem. Lett. 1577 (1986) 110. I Jaafar, G Francis, R Danion-Bougot, D Danion Synthesis 56 (1994) 111. T M Ibrahim, F S M Ahmed, S A Shedid Phosphorus Sulfur Silicon Relat. Elem. 86 263 (1994) 112. P Molina, E Aller, A Lo'pez-La'zaro, M AlajarõÂ n, A Lorenzo Tetrahedron Lett. 35 3817 (1994) 113. A Yu Rulev, T A Kuznetsova, L I Larina, L V Sherstyannikova, N A Keiko, M G Voronkov Zh. Org. Khim. 35 1622 (1999) b 114. L E Fisher, J M Muchowski Org. Prep. Proced. Int. 22 399 (1990) 115. M T Reetz Chem. Rev. 99 1121 (1999) 116. Y Matsumoto, R Tsuzuki, A Matsuhisa, K Takayama, T Yoden, W Uchida,M Asano, S Fujita, I Yanagisawa, T Fujikura Chem.Pharm. Bull. 44 103 (1996) 117. U KucklaÈ nder, B Schneider Chem Ber. 119 3487 (1986) 118. S Massa, G Stefancich, M Artico, F Corelli, R Silvestri Farmaco 42 567 (1987) 119. C Meyer, J-P Pete, O Piva Recl. Trav. Chim. Pays-Bas 114 492 (1995) 120. T Fujisawa,M Watanabe, T Sato Chem. Lett. 2055 (1984) 121. U KucklaÈ nder, B Schneider Arch. Pharm. 326 287 (1993) 122. U KucklaÈ nder, K Kuna, B Schneider Arch. Pharm. 326 415 (1993) 123. Y Zhang, S Takeda, T Kitagawa, H Irie Heterocycles 24 2151 (1986) 124. G Bobowski J. Heterocycl. Chem. 18 1179 (1981) 125. P Nitti, G Pitacco, A Pizzioli, E Valentin J. Heterocycl. Chem. 34 33 (1997) 126. S Bozzini, F Felluga, G Nardin, A Pizzioli, G Pitacco, E Valentin J.Chem. Soc., Perkin Trans. 1 1961 (1996) 127. W A White, H Weingarten J. Org. Chem. 32 213 (1967) 128. F Felluga, P Nitti,G Pitacco, E Valentin Tetrahedron Lett. 29 4165 (1988) 129. F Felluga, P Nitti, G Pitacco, E Valentin J. Chem. Soc., Perkin Trans. 1, 1645 (1991) 130. K Tanaka, K Kariyne, S Umio Chem. Pharm. Bull. 17 611 (1969) 131. Z Arnold Synthesis 39 (1990) 132. S Mignani, Z Janousek, R Merenyi, H G Viehe, J Riga, J Verbist Tetrahedron Lett. 25 1571 (1984) 133. D Enders, O Meyer, G Raabe Synthesis 1242 (1992) 134. M I Abasolo, C H Gaozza, B M Ferna'ndez J. Heterocycl. Chem. 24 1771 (1987) 135. H BoÈ hme, Y S Sadanandam Arch. Pharm. 306 227 (1973) 136. M RivieÁ re-Baudet, J Satge' Recl. Trav. Chim. Pays-Bas 94 19 (1975) 137.L Yu Sandalova, L I Mizrakh, V P Evdakov Zh. Obshch. Khim. 36 1451 (1966) d 138. R Zupet,M TisÏ ler J. Org. Chem. 59 507 (1994) 139. M A Tobias, J G Strong, R P Napier J. Org. Chem. 35 1709 (1970) 140. J C Arnould, J Cossy, J P Pete Tetrahedron 36 1585 (1980) 141. F Benedetti, F Berti, P Nitti, G Pitacco, E Valentin Gazz. Chim. Ital. 120 25 (1990) 142. J Cossy, C Poitevin, L Salle', D Gomez Pardo Tetrahedron Lett. 37 6709 (1996) 219 143. E Ballaben, M Forchiassin, P Nitti, C Russo Gazz. Chim. Ital. 123 387 (1993) 144. F Benedetti, M Forchiassin, C Russo, A Risaliti Gazz. Chim. Ital. 115 663 (1985) 145. R A Jerussi J. Org. Chem. 34 3648 (1969) 146. K Blau, V Voerckel J. Prakt. Chem. 331 285 (1989) 147. G BuÈ chi, H Wmest J.Org. Chem. 36 609 (1971) 148. N De Kimpe, C Stevens Tetrahedron 51 2387 (1995) 149. N De Kimpe, L D'Hondt, E Stanoeva Tetrahedron Lett. 32 3879 (1991) 150. J Toye, L Ghosez J. Am. Chem. Soc. 97 2276 (1975) 151. J B Schwarz, P N Devine, A I Meyers Tetrahedron 53 8795 (1997) 152. A S Medvedeva, A I Borisova, I D Kalikhman, N S Vyazankin Izv. Akad. Nauk SSSR, Ser. Khim. 1347 (1987) a 153. Z B Papanastassiou, R J Bruni, E V White J. Med. Chem. 10 701 (1967); Ref. Zh. Khim. 12 Zh 555 (1968) 154. H Neunhoeffer, H-W FruÈ hauf Liebigs Ann. Chem. 758 125 (1972) 155. N De Kimpe, L D'Hondt, L Moens Tetrahedron 48 3183 (1992) 156. K Sato, S Inoue, S-i Kuranami, M Ohashi J. Chem. Soc., Perkin Trans. 1 1666 (1977) 157. N De Kimpe, R Verhe', in The Chemistry of a-Haloketones a-Haloaldehydes and a-Haloimines (New York: Wiley, 1998) p.66 158. K Sato, S Inoue, M Ohashi Bull. Chem. Soc. Jpn. 47 2519 (1974) 159. L Duhamel, G Ple', P Commare C.R. Hebd. Seances Acad. Sci., Ser. C 278 1113 (1974) 160. P Duhamel, L Duhamel, V Truxillo Tetrahedron Lett. 51 (1974) 161. C-g Shin, Y Yonezawa, J Yoshimura Chem. Lett. 1635 (1981) 162. C-g Shin, Y Yonezawa, T Yamada Chem. Pharm. Bull. 32 3934 (1984) 163. P W Groundwater, T Sharif, A Arany, D E Hibbs, M B Hursthouse, I Garnett,M Nyerges J. Chem. Soc., Perkin Trans. 1 2837 (1998) 164. P W Groundwater, T Sharif, A Arany, D E Hibbs, M B Hursthouse, M Nyerges Tetrahedron Lett., 39 1433 (1998) 165. E Kleinpeter, St Tomas, G Uhlig, W-D Rudolf Magn.Reson. Chem. 31 714 (1993) 166. G Fisher, W-D Rudolf, E Kleinpeter Magn. Reson. Chem. 29 212 (1991) 167. D Tourwei, G Van Bist, S A G De Graaf, U K Pandit Magn. Reson. Chem. 7 433 (1975) 168. E D Cone, R H Garner, A W Hayes J. Org. Chem. 37 4436 (1972) 169. K K Balasubramanian, S Selvaraj J. Org. Chem. 45 3726 (1980) 170. D DoÈ pp, B Mlinaric Bull. Soc. Chim. Belg. 103 449 (1994) 171. D R MuÈ ller Diss. Dokt. Naturwiss., Univ. Stuttgart 1977; Ref. Zh. Khim. 10 B312 (1979) 172. L A Lee, J W Wheeler J. Org. Chem. 37 497 (1972) 173. B P Nosov, V I Brel', Yu I Kheruze, B I Ionin, L I Mashlyakovskii, A A Petrov Zh. Obshch. Khim. 52 811 (1982) d 174. I Naito, A Kinoshita, T Yonemitsu Bull. Chem. Soc. Jpn. 49 339 (1976) 175.P W Hickmott Tetrahedron 38 1975 (1982) 176. A Yu Rulev, A S Mokov, L I Krivdin, N A Keiko, M G Voronkov Magn. Reson. Chem. 35 533 (1997) 177. J C Arnould, A Enger, A Feigenbaum, J P Pete Tetrahedron 35 2501 (1979) 178. H Ahlbrecht, H Simon Synthesis 58 (1983) 179. H Ahlbrecht, A von Daacke Synthesis 610 (1984) 180. N A Keiko, A Yu Rulev, I D Kalikhman, M G Voronkov Izv. Akad. Nauk SSSR, Ser. Khim. 2031 (1991) a 181. K Sato, S Inoue, T Kitagawa, T Takahashi J. Org. Chem. 38 551 (1973) 182. K Sato, Y Kojima, H Sato J. Org. Chem. 35 2374 (1970) 183. E S Balenkova,M A Gorokhova Zh. Org. Khim. 13 896 (1977) b 184. N De Kimpe, R Verhe', L De Buyck, J Chys, N Schamp Bull. Soc. Chim. Belg. 88 59 (1979) 185. A Yu Rulev, S V Zinchenko Mendeleev Commun.70 (2001) 186. N A Keiko, A Yu Rulev, I D Kalikhman, M G Voronkov Izv. Akad. Nauk SSSR, Ser. Khim. 2471 (1986) a 187. N A Keiko, A Yu Rulev, I D Kalikhman, N I Shergina, L V Sherstyannikova,M G Voronkov Izv. Akad. Nauk SSSR, Ser. Khim. 1093 (1988) a 188. A Yu Rulev, A S Mokov,M G Voronkov Mendeleev Commun. 53 (1995) 189. J C Arnould, J Cossy, J P Pete Tetrahedron 37 1921 (1981)220 190. V G Granik, N P Kostyuchenko, V G Smirnova, Yu N Sheinker, R G Glushkov Zh. Org. Khim. 9 2299 (1973) b 191. A B Grigor'ev, M K Polievktov, V G Smirnova, V G Granik, R G Glushkov Zh. Org. Khim. 9 2332 (1973) b 192. H Matsushita, Y Tsujino, M Noguchi, S Yoshikawa Chem. Lett. 1087 (1976) 193. H Matsushita, Y Tsujino, M Noguchi, S Yoshikawa Bull. Chem.Soc. Jpn. 50 1513 (1977) 194. L Nilsson, R Carlson, C Rappe Acta Chem. Scand., Ser. B. 30 271 (1976) 195. N De Kimpe, R Verhe', L De Buyck, J Chys, N Schamp Org. Prep. Proc. Int. 10 149 (1978) 196. N De Kimpe, R Verhe', L De Buyck, N Schamp Synthesis 741 (1979) 197. T Machiguchi, T Takeno, T Hasegawa, Y Kimura Chem. Lett. 1821 (1992) 198. C-g Shin, Y Yonezawa, T Obara Heterocycles 24 1561 (1986) 199. C-g Shin, Y Yonezawa, K Watanabe, J Yoshimura Bull. Chem. Soc. Jpn. 54 3811 (1981) 200. Y Yonezawa, C-g Shin, Y Ono, J Yoshimura Bull. Chem. Soc. Jpn. 53 2905 (1980) 201. C-g Shin,Y Yonezawa, J Yoshimura Tetrahedron Lett. 4085 (1979) 202. C-g Shin, Y Yonezawa,M Takahashi, J Yoshimura Bull. Chem. Soc. Jpn. 54 1132 (1981) 203.R J Cregge, T T Curran,W A Metz J. Fluorine Chem. 88 71 (1998); Chem. Abstr. 128 321 900 (1998) 204. F McCapra,M Roth J. Chem. Soc., Chem. Commun. 894 (1972) 205. K R Prasad, M Darbarwar Synth. Commun. 20 1379 (1990) 206. K R Prasad, M Darbarwar Synth. Commun. 22 2479 (1992) 207. A Yu Rulev, Ya V Zachinyaev, A I Ginak, in Novye Dostizheniya YaMR v Strukturnykh Issledovaniyakh (Tez. Dokl. II Vseros. Semi- nara), Kazan' 1995 [The new Achievements of NMR in Structural Investigations (Abstracts of Reports of the All-Russian Seminar), Kazan', 1995] p. 94 208. M Ikeda, T Uchino, K Maruyama, A Sato Heterocycles 27 2349 (1988) 209. K R Prasad, M Darbarwar Org. Prep.Proced. Int. 7 547 (1995); Ref. Zh. Khim. 15 Zh 168 (1996) 210. V Kvita, H Sauter, G Rihs Helv.Chim. Acta 72 457 (1989) 211. U KucklaÈ nder, K Kuna, B Schneider, A Steigel, B Mayer J. Prakt. Chem. 335 345 (1993) 212. U KucklaÈ nder, B Schneider Chem. Ber. 121 577 (1988) 213. M Ohashi, T Takahashi, S Inoue, K Sato Bull. Chem. Soc. Jpn. 48 1892 (1975) 214. B Costisella, H Gross, H Schick Tetrahedron 40 733 (1984) 215. H Ahlbrecht, H Simon Synthesis 61 (1983) 216. S De Lombaert, B Lesur, L Ghosez Tetrahedron Lett. 23 4251 (1982) 217. S De Lombaert, L Ghosez Tetrahedron Lett. 25 3475 (1984) 218. B Lesur, J Toye,M Chantrenne, L Ghosez Tetrahedron Lett. 2835 (1979) 219. H Ahlbrecht, H-M Kompter Synthesis 645 (1983) 220. H Ahlbrecht,M Dietz, C Schln, V Baumann Synthesis 133 (1991) 221. H Ahlbrecht,M Dietz, L Weber Synthesis 251 (1987) 222.H Ahlbrecht, A von Daacke Synthesis 24 (1987) 223. H Ahlbrecht,M Ibe Synthesis 210 (1988) 224. H Ahlbrecht,M Ibe Synthesis 929 (1987) 225. H Ahlbrecht,M Ibe Synthesis 421 (1985) 226. H Ahlbrecht,M Dietz Synthesis 417 (1985) 227. C Maignan, F Rouessac Bull. Soc. Chim. Fr. 1454 (1973) 228. C Maignan, F Rouessac Bull. Soc. Chim. Fr. 2035 (1974) 229. C Alexandre, F Rouessac Bull. Soc. Chim. Fr. 117 (1977) 230. M J Brown, T Harrison, L E Overman J. Am. Chem. Soc. 113 5378 (1991) 231. N De Kimpe, R Verhe', L De Buyck, J Chys, N Schamp J. Org. Chem. 43 2670 (1978) 232. K Matsumura, T Saraie, N Hashimoto Chem. Pharm. Bull. 24 912 (1976) 233. D Simon, O Lafont, C C Farnoux,M Miocque J. Heterocycl. Chem. 22 1551 (1985) 234.N A Keiko, A Yu Rulev, I D Kalikhman, M G Voronkov Izv. Akad. Nauk SSSR, Ser. Khim. 361 (1987) a 235. N V Kuznetsov, N K Makhnovskii, R A Myrsina, A K Shurubura Ukr. Khim. Zh. 611 (1976) A Yu Rulev 236. A Yu Rulev, L I Larina, N A Keiko,M G Voronkov J. Chem. Soc., Perkin Trans. 1 1567 (1999) 237. A Yu Rulev, L I Larina, M G Voronkov Tetrahedron Lett. 41 10211 (2000) 238. A Yu Rulev, L I Larina, M G Voronkov Molecules 6 892 (2001) 239. A Yu Rulev, L I Larina, M G Voronkov Zh. Obshch. Khim. 71 1996 (2001) d 240. A Yu Rulev, L I Larina, M G Voronkov, in Sovremennye Prob- lemy Organicheskoi Khimii, Ekologii i Biotekhnologii (Tez. Dokl. I Mezhdunarodnoi Nauchnoi Konferentsii), Luga, 2001 [Modern Problems of Organic Chemistry, Ecology and Biotechnology (Abstracts of Reports of the First International Conference), Luga, 2001] Vol.1, p. 103 241. O Ito, Y Arito, M Matsuda J. Chem. Soc., Perkin Trans. 1 869 (1988) 242. S Mignani,R Mere'nyi, Z Janousek,H G Viehe Tetrahedron 41 769 (1985) 243. J C Arnould, J P Pete Tetrahedron Lett. 2463 (1975) 244. J Cossy, J P Pete Tetrahedron Lett. 21 2947 (1980) 245. M A Brumfield, P S Mariano, U C Yoon Tetrahedron Lett. 24 5567 (1983) 246. K Saito,M Kozaki, K Uenishi, N Abe, K Takahashi Chem. Pharm. Bull. 39 1843 (1991) 247. C-C Yang, J-M Fang J. Chem. Soc., Perkin Trans. 1 879 (1995) 248. M Forchiassin, G Pitacco, C Russo, E Valentin Gazz. Chim. Ital. 112 335 (1982) 249. M Forchiassin, G Pitacco, A Risaliti, C Russo, E Valentin J.Heterocycl. Chem. 20 305 (1983) 250. F Felluga, P Nitti, E Ruocco, C Russo Gazz. Chim. Ital. 127 387 (1997) 251. F Benedetti, M Forchiassin, G Pispisa, P Nitti, G Pitacco, C Russo, E Valentin Gazz. Chim. Ital. 120 327 (1990) 252. F Felluga, P Nitti, A Pizzioli,M Prodan, C Russo Gazz. Chim. Ital. 126 297 (1996) 253. F Felluga, P Nitti, A Pizzioli,M Prodan, C Russo Gazz. Chim. Ital. 127 31 (1997) 254. J W Huffman, M M Cooper, B B Miburo,W T Pennington Tetrahedron 48 8213 (1992) 255. M M Cooper, J W Huffman J. Chem. Soc., Chem. Commun. 348 (1987) 256. G Barbarella,G Pitacco, C Russo, E Valentin Tetrahedron Lett. 24 1621 (1983) 257. G Barbarella, S BruÈ ckner, G Pitacco, E Valentin Tetrahedron 40 2441 (1984) 258.F Felluga, G Nardin, P Nitti, G Pitacco, E Valentin Tetrahedron 44 6921 (1988) 259. D Brussa, F Felluga, P Nitti, G Pitacco, E Valentin Gazz. Chim. Ital. 122 85 (1992) 260. D DoÈ pp Mol. Supramol. Photochem. 6 101 (2000); Chem. Abstr. 134 178 109 (2001) 261. D DoÈ pp, C KruÈ ger, H R Memarian, Y-H Tsay Angew. Chem. Int. Ed. 24 1048 (1985) 262. D DoÈ pp, H R Memarian Chem. Ber. 123 315 (1990) 263. D DoÈ pp, B Mlinaric Ach. Mod. Chem. 131 377 (1994); Chem. Abstr. 122 132 715 (1995) 264. D DoÈ pp, A W Erian, G Henkel Chem. Ber. 126 239 (1993) 265. H R Memarian,M Nasr-Esfahani, R Boese, D DoÈ pp Liebigs Ann. Chem. 1023 (1997) 266. D DoÈ pp, A A Hassan, G Henkel Liebigs Ann. Chem. 697 (1996) 267. L Stella, J-L Boucher Tetrahedron Lett.23 953 (1982) 268. F Jin, Y Xu,W Huang J. Chem. Soc., Chem. Commun. 814 (1993) 269. A Meier, J Sauer Tetrahedron Lett. 31 6855 (1990) 270. Y-W Wang, J-M Fang, Y-K Wang, M-H Wang, T-Y Ko, Y-J Cherng J. Chem. Soc., Perkin Trans. 1 1209 (1992) 271. M Bourhis, R Golse, E Adjanohoun, J-J Bosc,M Goursolle, P Picard Tetrahedron Lett. 29 1139 (1988) 272. Ch De Cock, S Piettre, F Lahousse, Z Janousek, R Mergnyi, H G Viehe Tetrahedron 41 4183 (1985) 273. D DoÈ pp, H R Memarian, M A Fisher, A M J van Eijk, C A G O Varma Chem. Ber. 125 983 (1992) 274. D DoÈ pp, M A Fischer Recl. Trav. Chim. Pays-Bas 114 498 (1995) 275. Y-L Liang, J-M Fang, T Chow, T-I Ho, C-R Lee, Y Wang J. Org. Chem. 59 5742 (1994)Captodative aminoalkenes 276. B De Boeck, S Jiang, Z Janousek, H G Viehe Tetrahedron 50 7075 (1994) 277.B De Boeck, Z Janousek, H G Viehe Tetrahedron 51 13239 (1995) 278. B De Boeck, H G Viehe Tetrahedron 54 513 (1998) 279. B Tinant, J Feneau-Dupont, J P Declercq, B De Boeck, S Jiang, Z Janousek, H G Viehe Bull. Soc. Chim. Belg. 104 397 (1995) 280. F Texier, A Derdour, H Benhaoua, T Benabdellah, O Yebdri Tetrahedron Lett. 23 1893 (1982) 281. F Clerici, G Marazzi,M Taglietti Tetrahedron 48 3227 (1992) 282. D DoÈ pp, J Walter Heterocycles 20 1055 (1983) 283. M Bonadeo, C De Micheli, R Gandolfi J. Chem. Soc., Perkin Trans. 1 939 (1977) 284. R Damico, J M Nicholson J. Org. Chem. 38 3057 (1973) 285. G A Russell, B Z Shi,W Jiang, S Hu, B H Kim,W Baik J. Am. Chem. Soc. 117 3952 (1995) 286. C-C Yang, H-M Tai, P-J Sun Synlett 812 (1997) 287. D K Wall, M C McMaster, N H Cromwell J. Org. Chem. 34 1124 (1969) 288. E J Warawa, J R Campbell J. Org. Chem. 39 3511 (1983) 289. R Madhav Synthesis 27 (1982) 290. O I Gorbyleva, T Ya Filipenko, E E Mikhlina, K F Turchin, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 793 (1982) e 291. K F Turchin, A D Yanina, T Ya Filipenko, E E Mikhlina, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 1248 (1985) e 292. A D Yanina, T K Trubitsyna, E E Mikhlina, L N Yakhontov Khim.-Farm. Zh. 808 (1987) f 293. G Caliendo, G Greco, E Perissutti, V Santagada, C Silipo, A Vittoria, L Turbaniti, A R Renzetti, E Benedetti, C Pedone, A Santini Farmaco 48 1359 (1993) 294. G Viti, D Giannotti, R Nannicini, G Balacco, V Pestellini Tetrahedron Lett. 35 5939 (1994) 295. A Santini, E Benedetti, C Pedone,M Giordano, G Caliendo, V Santagada, P Grieco, G Greco Tetrahedron 51 1995 (1995) 296. M Rowley, H B Broughton, I Collins, R Baker, F Emms, R Marwood, S Patel, S Patel, C I Ragan, S B Freedman, P D Leeson J. Med. Chem. 39 1943 (1996) 297. M Rowley, I Collins, H B Broughton, W B Davey, R Baker, F Emms, R Marwood, S Patel, S Patel, C I Ragan, S B Freed- man, R Ball, P D Leeson J. Med. Chem. 40 2374 (1997) 298. L N Koikov, N V Alekseeva, N B Grigor'ev, V I Levina, K F Turchin, T Ya Filipenko,M D Mashkovskii,M E Kaminka, V B Nikitin, G N Engalycheva,M A Kalinkina, I S Severina, I K Ryaposova, V G Granik Khim.-Farm. Zh. 26 (1997) f 299. E I Klimova, L Ruiz Ramirez, T Klimova,M Martinez Garcia J. Organomet. Chem. 559 43 (1998) 300. E I Klimova, L Ruiz Ramires,M Garsia Martines, R G Espinosa, N N Meleshonkova Izv. Akad. Nauk, Ser. Khim. 2743 (1996) a 301. J A Lowe III, S E Drozda, S McLean, D K Bryce, R T Crawford, R M Snider, K P Longo, A Nagahisa, M Tsuchiya J. Med. Chem. 37 2831 (1994) 302. Y Besidsky, K Luthman, U Hacksell J. Heterocycl. Chem. 31 1321 (1994) 303. A R Hansen, H Bader J. Heterocycl. Chem. 3 109 (1966); Ref. Zh. Khim. 3 Zh 316 (1967) 304. T K Morgan Jr, R Lis, A J Marisca, T M Argentieri, M E Sulli- van, S S Wong J. Med. Chem. 30 2259 (1987) 305. V Ya Vorob'eva, V A Bondarenko, E E Mikhlina, K F Turchin, L F Linberg, L N Yakhontov Khim. Geterotsikl. Soedin. 1370 (1977) e 306. V A Bondarenko, K F Turchin, E E Mikhlina, L N Yakhontov Khim. Geterotsikl. Soedin. 948 (1981) e 307. E I Klimova, T B Klimova,MMartinez Garcia, N N Meleshonkova, L Ruiz Ramirez Mendeleev Commun. 233 (1997) 308. C J Swain, E M Seward,M A Cascieri, T M Fong, R Herbert, D U MacIntyre, K J Merchant, S N Owen, A P Owen, V Sabin, M Teall, M B VanNiel, B J Williams, S Sadowski, C Strader, R G Ball, R Baker J. Med. Chem. 38 4793 (1995) 309. Y Besidsky, K Luthman, A Claesson, C J Fowler, I CsoÈ regh, U Hacksell J. Chem. Soc., Perkin Trans. 1 465 (1995) 221 310. K F Turchin, V A Bondarenko, E E Mikhlina, L N Yakhontov Dokl. Akad. Nauk SSSR 259 383 (1981) c 311. V A Bondarenko, E E Mikhlina, T Ya Filipenko, K F Turchin, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 1387 (1981) e 312. V Ya Vorob'eva, K F Turchin, E E Mikhlina, V A Bondarenko, A I Ermakov, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 1377 (1977) e 313. V A Bondarenko, T K Trubitsyna, E E Mikhlina, M D Mashkovskii, L N Yakhontov Khim.-Farm. Zh. 67 (1981) f 314. V A Bondarenko, E E Mikhlina, T Ya Filipenko, K F Turchin, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 371 (1979) e 315. V A Bondarenko, T K Trubitsyna, O S Anisimova, E E Mikh- lina,M D Mashkovskii, L N Yakhontov Khim.-Farm. Zh. 51 (1982) f 316. E I Klimova, L Ruiz Ramirez, M Martinez Garcia, N N Meleshonkova J. Organomet. Chem. 532 181 (1997) 317. O I Gorbyleva, T Ya Filipenko, E E Mikhlina, K F Turchin, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 1232 (1982) e 318. V A Bondarenko, E E Mikhlina, T Ya Filipenko, K F Turchin, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 1393 (1979) e 319. L N Koikov, N V Alexeeva, N B Grigoryev, V I Levina, K F Turchin, T Ya Filipenko, I S Severina, I K Ryaposova, V G Granik Mendeleev Commun. 94 (1996) 320. T Ya Filipenko, O I Gorbyleva, K F Turchin, O S Anisimova, O M Peresleni, E E Mikhlina, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 666 (1981) e 321. K F Turchin, V A Bondarenko, T Ya Filipenko, E E Mikhlina, Yu N Sheinker, L N Yakhontov Khim. Geterotsikl. Soedin. 118 (1993) e 322. V N Postnov, A V Goncharov, I Hocke, D P Krut'ko J. Organomet. Chem. 456 235 (1993) 323. A M Shestopalov, V Yu Mortikov, Yu A Sharanin, A V Turov, V P Litvinov Zh. Org. Khim. 25 1980 (1989) b 324. A D Yanina, V Ya Vorob'eva, T K Trubitsyna, E E Mikhlina, M D Mashkovskii, L N Yakhontov Khim.-Farm. Zh. 1202 (1986) f 325. T Rosen, K J Guarino Tetrahedron 47 5391 (1991) 326. K Koch, J H Smitrovich Tetrahedron Lett. 35 1137 (1994) 327. I G Ostroumov, Doctoral Thesis in Chemical Sciences, Moscow State University, Moscow, 1999 a�Russ. Chem. Bull., Int. Ed. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Transl.) c�Dokl. Chem. (Engl. Transl.) d�Russ. J. Gen. Chem. (Engl. Transl.) e�Chem. Heterocycl. Compd. (Engl. Transl.) f�Pharm.-Chem. J. (En

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (3) 223 ± 238 (2002) Electrochemical activation of reactions involving organometallic compounds T V Magdesieva, K P Butin Contents I. Introduction II. Reactions of odd-electron complexes of transition metals III. Electrocatalysis: combination of catalysis by metal complexes with electrochemically induced electron transfer IV. The formation and cleavage of carbon ± metal s-bonds V. Electrochemically activated reactions involving `electron sponges' Abstract. various of activation electrochemical the on Data Data on the electrochemical activation of various reactions are compounds organometallic involving reactions involving organometallic compounds are generalised. generalised. Primary attention is devoted to the key types of transformation Primary attention is devoted to the key types of transformation that electron electrochemical using performed be can that can be performed using electrochemical electron transfer, transfer, namely, redox activation of 16- and 18-electron complexes of namely, redox activation of 16- and 18-electron complexes of transition of range broad a of step first the as metals transition metals as the first step of a broad range of reactions, reactions, electrocatalysis, of electrosynthesis and processes, mediator electrocatalysis, mediator processes, and electrosynthesis of com- com- pounds bibliography The containing pounds containing carbon carbon7metal s-bonds.-bonds. The bibliography includes 188 references includes 188 references. I. Introduction The problem of increasing the reactivity of compounds arose as soon as chemists became engaged in targeted synthesis.Numerous methods for the increase in the reactivity are currently known, ranging from simple methods (introduction of an appropriate substituent into the molecule) to more complex ones that have become classical (for example, catalysis) and, finally, to the most advanced techniques (in particular, photochemically or electro- chemically induced electron transfer). Figure 1 presents classifi- cation of the activation processes proposed by Chanon.1 The diagram axis can be continued without limit because new ways of solving this problem would certainly appear with the progress of science. Activation through electron transfer is rather effective.In this case, very high process selectivity can be attained because it is possible to specify precisely the potential needed for the gener- ation of a particular active species. The transformation of a molecule into a reactive radical cation or radical anion (Scheme 1) initiates a wide range of chemical reactions. This is due to the following factors: (1) usually, charged species react with nucleophiles or electrophiles more readily than neutral molecules; (2) the presence of the unpaired electron activates homolytic reaction pathways; (3) further redox proc- esses, disproportionation and so on are also possible. T V Magdesieva, K P Butin Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119992 Moscow, Russian Federation.Fax (7-095) 939 55 46. Tel. (7-095) 939 30 65. E-mail: tvm@org.chem.msu.ru (T V Magdesieva), butin@org.chem.msu.ru (K P Butin). Received 26 November 2001 Uspekhi Khimii 71 (3) 255 ± 272 (2002); translated by Z P Bobkova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n03ABEH000704 223 224 227 231 233 Scheme 1 E+ S7. 7e7 R. a b e7 P S 7e7 Nu7 e7 S+. Ra b S is the substrate, Nu is the nucleophile, E is the electrophile, P is the product; a is a redox reaction; b is dissociation, etc. Activation of reactions through electron transfer was first employed by Taube 2 and Basolo and Pearson 3±6 for inorganic systems. In organic chemistry, the development of this line of research was triggered by Kornblum's discovery 7, 8 of radical nucleophilic substitution (SRN1) the mechanism of which has been later studied in detail by Bunnett 9 and Russel.10, 11 Saveant 12 was the first to use electrochemical electron transfer for the activation of organic reactions.Since then, the French electrochemical Electrochemical electron transfer Photochemical electron transfer Enantioselective Catalysis Stereoselective Regioselective Solvation Reversible structural changes Irreversible structural changes (introduction of an appropriate substituent and so on) Figure 1. Multilevel diagram of the activation methods of reactions.1224 school has held a leading position in the research along this line.A great number of examples in which diverse organic reactions are activated by electrochemical electron transfer have been reported (see, for example, Refs 1, 13). The intense use of electrochemical electron transfer in reac- tions involving organometallic compounds started approximately in the early 1980s. Most of the transition metal p-complexes are able to undergo reversible electrochemical redox transitions to give active species with odd numbers of electrons; therefore, electrochemically induced electron transfer can be used efficiently to facilitate ligand exchange, isomerisation, insertion and other reactions.14 ± 19 It is of interest to compare the behaviours of mononuclear metal complexes and supramolecular species containing these complexes as structural units.20 Systems of this type are able to undergo electrochemically induced reversible structural changes, and they can be regarded as prototypes of `molecular machines.' In this case, the electron transfer involving one structural unit entails changes over the whole supramolecular system (for exam- ple, electrically induced charge separation, migration of structural fragments, etc.).20 A relatively new field in the use of electrochemical electron transfer for changing the reactivity of transition metal complexes is the design of redox-switchable ligands.The introduction into the ligand of ferrocenyl, cobaltocenyl, anthraquinone or other groups prone to reversible redox transitions allows the formation of complexes with switching reactivity.21 A change in the redox state of the `switch' induces a change in the electronic properties of the ligand and, hence, the nature of the metal ± ligand bond. This promotes various types of reaction, for example, ligand exchange. This approach appears highly promising for the design of `redox-switchable' catalysts.Chiral metal complexes with `redox-switchable' ligands can be used in asymmetric synthesis, for the separation of enantiomers, etc.21 Yet another extensive field of application of the electrochem- ical electron transfer in reactions involving organometallic com- pounds is electrochemical generation of labile complexes of transition metals in low oxidation states and involvement of these species in subsequent chemical transformations in situ, most often, as catalysts.The combination of catalysis by metal complexes with the electrochemical electron transfer has opened up almost unlimited scope for performing various mediator processes (see, for example, Refs 22 ± 24). Numerous examples of reactions activated by electrochemical electron transfer have now been reported in the literature. Since it is impossible to embrace them all within a single review, we intended first of all to show what types of transformations can be accomplished by virtue of the electrochemical electron transfer and how the use of electrochemical methods extends the synthetic capacity of a chemist in the field of organometallic chemistry. Primary attention has been devoted to the studies published during the last decade and not covered in earlier reviews.II. Reactions of odd-electron complexes of transition metals The transfer from relatively stable 16- and 18-electron configu- rations to 17- and 19-electron configurations of transition metal complexes can be attained via electrochemical oxidation or reduction. The odd-electron complexes thus formed are much more reactive; therefore, they can be involved in reactions that either do not proceed at all or proceed too slowly for complexes with 16- and 18-electron configurations. Research into the chem- ical transformations of odd-electron complexes has been the subject of numerous studies (see, for example, reviews 15, 18, 22). If the transformations are classified in terms of the reaction type, the main types are ligand exchange, isomerisation, insertion, extru- sion, chelation/dechelation and dimerisation.T V Magdesieva, K P Butin 1. Ligand exchange Ligand exchange reactions can be initiated by both oxidation and reduction of mono- or polynuclear metal complexes. Initiation by reduction Initiation by oxidation 19 e 17 e ML+e7 ? ML7. 18 e ML7e7 ? ML+. 18 e 17 e 19 e ML7.+L0 ? M(L0 )7.+L 19 e ML+.+L0 ? M(L0 )+.+L 17 e M(L0 )+.+ML ? ML0+ML+. M(L0 )7.+ML ? ML0+ML7. 19 e 18 e 18 e 19 e 17 e 18 e 18 e 17 e Upon reductive activation, an electron passes to an unoccu- pied orbital; this is accompanied by weakening of theM7L bond. In essence, a similar reaction takes place during oxidation, because an associative mechanism of ligand exchange including the intermediate formation of a 19-electron complex with an additional ligand has been proven for 17-electron complexes by Kochi and coworkers.23, 25 [MeCpMn(CO)2(MeCN)]++PPh3 17 e [MeCpMn(CO)2(MeCN)(PPh3)]+ 19 e [MeCpMn(CO)2(PPh3)]++MeCN 17 e Cp=C5H5 .A second ligand present in the solution is coordinated to the 17-electron complex to give a 19-electron intermediate (as in the case of reductive activation) in which cleavage of theM7L bond is facilitated. A large number of examples of electrochemically activated ligand exchange have been described (see, for example, the reviews 15 ± 18, 26 and the monographs 27, 28). We shall dwell only on some publications that appeared during the last three to four years.An interesting example of replacement of the arene ligand in binuclear fulvalene (Fv) complexes of iron has been reported.29 The doubly reduced 38-electron complex readily reacts withCOor with different phosphine ligands. 2 e7 [Fe2(m2-Z5:Z5-Fv)(C6H6)2](PF6)2 72PF¡ Fe2(m2-Z5:Z5-Fv)(C6H6)2 6 PMe3 Fe2(m2-Z5:Z5-Fv)(PMe3)4 CO Fe2(m2-Z4:Z4-Fv)(CO)6 715 8C, 1 atm Preparative electrolysis of the fluorenyl complex of cobalt (at a potential of 72.56 V vs. Fc/Fc+, where Fc is ferrocene) in the presence of various phosphine or phosphite ligands also results in ligand exchange.30 P(OMe)3 e7 [CpCo(PPh3)(C13H9)]7. CpCo(PPh3)(C13H9) 7PPh3 CpCo[P(OMe)3](C13H9). {CpCo[P(OMe)3](C13H9)}7.7e7 An unusual ligand exchange in the metal coordination sphere accomplished by virtue of a `redox switch' connected to the reaction centre of the molecule has been reported.31 The role of the `redox switch' is played by a ferrocenyl fragment linked by a covalent bond to the arene ligand of the chromiumtricarbonyl complex. Since the ferrocene fragment is oxidised first, the reaction takes place at an anodic overpotential substantially lower than that required for oxidation of the chromiumtricar- bonyl group. Thus the partial destruction of the complex is avoided and the process selectivity is increased. Although theElectrochemical activation of reactions involving organometallic compounds chromium atom has formally an 18-electron configuration, the adjacency of the strong electron acceptor results in easy replace- ment of the carbonyl group by a phosphorus-containing ligand.[(C6H5Fc)Cr(CO)2P(OEt)3]++CO. (C6H5Fc)Cr(CO)3+P(OEt)3 7e7 The subsequent transfer of an electron from the Cr atom to the Fe atom turns off the `redox switch'. Upon further displacement of the potential to the anodic region, the `redox switch' is actuated once more, the Cr atom is activated again, and the reaction gives a purple complex, which was isolated. [(C6H5Fc)Cr(CO)2P(OEt)3]++P(OEt)3 7e7 {(C6H5Fc)Cr(CO)2[P(OEt)3]2}2+. Electrochemical cleavage of the metal ± halogen covalent bond with subsequent replacement of the halogen atom by another ligand has been described in detail.32 Upon single- electron reduction of the rhenium complex with diphenylposphi- noethane (dppe) [ReCl2(dppe)2]+ [the Re(d 4)?Re(d5) transi- tion], the Re7Cl bond length increases by 0.1A; the transfer of one more electron induces cleavage of this bond, and another ligand present in the solution enters the coordination sphere of the metal.e7 e7 Re(dppe)2Cl2 [Re(dppe)2Cl2]7 [Re(dppe)2Cl2]+ 7Cl7 L Re(dppe)2Cl(L) Re(dppe)2Cl L=ButCN. Oxidation of the central metal atom in the complex results in strengthening of bonds with s-donor ligands (ammonia, amines, pyridine, etc.) and weakening of bonds with p-acceptor ligands (CO, nitriles, cyanides, etc.); this stipulates the required type of electrochemical activation (cathodic or anodic) in each particular case.Actually, chelation/dechelation reactions can also be classi- fied as ligand exchange. Some organic groups, for example, the dithiocarbamate group, act as chelating ligands in the coordina- tion sphere of unsaturated complexes. However, they can also occur in the metal coordination sphere as monodentate ligands. It was shown33 that the monodentate dithiocarbamate ligand can be readily converted into a chelating ligand upon oxidative activa- tion. S S 7e7 + Fe Fe C C OC OC S S NR2 NR2 CO CO 7CO e7 + Fe Fe OC OC S S S S C C NR2 NR2 In principle, the inverse problem can also be solved. If it is required to change electrochemically the oxidation state of the central metal atom in a complex but to avoid ligand exchange or coordination of additional ligands, it is expedient to use a so- called cage ligand.20 For example, oxidation of 1,10-phenanthro- line complexes of Cu(I) is irreversible because it is accompanied by a structural rearrangement with coordination of solvent molecules to the copper atom giving rise to an octahedral geometry typical of Cu(II) complexes.If 2,9-dimethyl-1,10-phenanthroline is used as the ligand instead of 1,10-phenanthroline, the oxidation of the 225 Cu(I) complex is completely reversible because the presence of substituents at the a-position relative to nitrogen prevents the coordination of solvent molecules to the metal and, hence, the geometry does not change.34 2. Isomerisation When considering the electrochemically activated isomerisation of transition metal complexes, one can distinguish three main types of transformations encountered most often, namely, cis ± trans isomerisation, change in the hapto-coordination of aromatic p-ligands to the metal in various sandwich and half-sandwich structures and metal migration in complexes with polydentate aromatic ligands.The first example of electrochemically activated cis ± trans isomerisation was described by Rieke et al.35 The reaction follows the scheme N N NMe MeN NMe MeN OC CO OC CO Mo Mo (DG=70.35 eV) OC NMe OC CO NMe MeN N CO MeN cis Ntrans (DG=0.53 eV) 7e7 (DG=0.36 eV) 7e7[cis]+ [trans]+ (DG=70.18 eV) Since then, numerous examples of both cathodically and anodically activated cis ± trans isomerisations have been reported.27 In particular, the rate constants for anodic iso- merisation of the thiolate isocyanate complexes cis-[M(SC6H2Pri3-2,4,6)2(NCR)4] (M=Mo, W; R=Me, But) have been calculated from the data of cyclic voltammetry.36 Isomerisation can be accompanied by a change in the type of metal bonding to the ligand: the terminal ligand passes to a bridging position or vice versa.For example, the 48-electron cluster Cp*IrCp2Co2(CO)3 (Cp*=C5Me5) can exist as two isomers, one containing a terminal and two bridging carbonyl groups (A) and one with three bridging CO ligands (B), the second isomer being thermodynamically more stable. CO CO CO CO CO Co Ir Co Ir Co Co CO B A The rate of isomerisation depends appreciably on the number of electrons in the cluster 37 and varies in the following sequence: 48 e 55 49 e<47 e (the ratio of the corresponding rate constants is 1 : 104 : 108).Thus, anodic activation is the most efficient. Anodically activated isomerisation of diene ligands, for exam- ple, cyclooctadiene (cod), is known.38 [(C5Ph5)Rh(1,3-cod)]+ (C5Ph5)Rh(1,3-cod) 7e7 e7 (C5Ph5)Rh(1,5-cod). [(C5Ph5)Rh(1,5-cod)]+ The trimetallic carbonyl clusters M3(CO)7(dppm)(tolan) (M=Os, Ru; dppm is diphenylposphinomethane; tolan is diphe- nylacetylene) can also exist as two isomers which differ in the position of the alkyne ligand with respect to the triangle formed by the metal atoms.39 The interconversion of the two isomers can be performed through electrochemical redox isomerisation because one of them is more stable in the reduced form and the other one, in the oxidised form.226 Numerous examples of the change in the type of hapto- coordination of the p-ligand to the metal atom induced by electrochemical electron transfer are known.For example, two- electron oxidation of a trinuclear ruthenium complex induces haptotropic isomerisation giving rise to an unusual m3-Z3:Z3- coordination of the benzene ring.40, 41 (Cp*Ru)3(m-H)3(m3-Z2:Z2:Z2-C6H6) 72 e7 [(Cp*Ru)3(m-H)3(m3-Z3:Z3-C6H6)]2+. The structure of the resulting complex was confirmed by X-ray diffraction data. The change in the hapto-coordination upon oxidative or reductive electrochemical activation has also been reported for indenyl, cyclopentadienyl and some other aromatic ligands 41 (see also the review 26 and references therein).Of special interest are redox-induced reactions of complexes with polydentate ligands having, at least, two possible sites of coordination to the metal. In this case, redox switching of the metal from one coordination site to the other has been observed.26 7 +MLn MLn MLn=FeCp, FeCp*, Mn(CO)3, Cr(CO)¡3 . The ring ± ring haptotropic isomerisation of 18-electron fluo- renyl complexes with the d 6 metal configuration proceeds very slowly even at high temperatures, and the equilibrium is shifted towards the Z5-isomer. In the case of 19-electron complexes, isomerisation rapidly takes place at 780 8C, the equilibrium being shifted towards the Z6-isomer: e7 [(Z6-C13H9)FeCp]7.(Z6-C13H9)FeCp (Z5-C13H9)FeCp. [(Z5-C13H9)FeCp]7. 7e7 Detailed analysis of the mechanism of ring ± ring haptotropic tautomerism and a survey of published examples of these proc- esses have been reported.26 3. Insertion and extrusion The electrochemical migrational insertion of a carbonyl group into the Fe7C(sp3) bond can be used to synthesise the corre- sponding acyl complexes; this reaction is carried out with either reductive 42 or oxidative 43 activation. e7 [CpFe(CO)2Me]7. PPh3 CpFe(CO)2Me CpFe(CO)(COMe)PPh3; [CpFe(CO)(COMe)PPh3]7. 7e7 [CpFe(CO)(PPh3)Me]+. MeCN CpFe(CO)(PPh3)Me 7e7 [CpFe(CO)(PPh3)(MeCN)Me]+. e7 [CpFe(PPh3)(MeCN)(COMe)]+. CpFe(PPh3)(MeCN)(COMe).If both anodic and cathodic types of activation are possible, one should prefer the method for which the ratio of the currents of the inverse and forward electrochemical processes (Iinv/Ifor) in the cyclic voltammogram of the product is closer to unity, i.e., a classical electrochemically reversible process should be the method of choice.15 The Kochi cathodic activation has provided the possibility of the back reaction, i.e., CO extrusion from formyl metal com- plexes.44 T V Magdesieva, K P Butin e7 [(CO)5ReRe(CO)4(CHO)]7 7CO (E=72.1 V) [(CO)5ReRe(CO)4H]7. [(CO)5ReRe(CO)4H]27 7e7 4. Electrochemically activated reactions in the ligand sphere of the metal Transition metal complexes tend to undergo reactions accompa- nied by electrochemical activation of p-bonded ligands, resulting in ligand modification.The activation of C7Hbonds followed by substitution in p-coordinated aromatic ligands provides an exam- ple.45 CN CN7 7 CN 7H+,72 e7 H M M M 7 + e7 e7 CO2 Co(Cp) Co(Cp) Co(Cp) MeI CO2Me CO27 7I7 7H+,72 e7 Co(Cp) Co(Cp) + CO2Me . Co(Cp) Examples of reactions of this type can also be found in Walder's monograph.28 A convenient method for building-up the hydrocarbon side chain during electrochemical reduction of acetyl derivatives of ferrocene, cymanthrene and benzenetricar- bonylchromium has been proposed.46 Me MeCN COMe CHCH2CN (E=72.2 V) Mn(CO)3 Mn(CO)3 Of recent publications devoted to this topic, a study by Gusev et al.47 deserves attention.The researchers identified the main regularities of the redox-induced activation of the C7H bonds in Pd and Pt Z4-pentamethylcyclopentadiene complexes. Single- electron oxidation of these compounds proceeds selectively as homolysis of the C7Hbond at the sp3-hybridised carbon atom of H +Pt H H H e7 + 7H. H2C CH2 Pt Pt Pt 19 e H 2+ Pt 7H. 2+ 7e7 Pt 17 eElectrochemical activation of reactions involving organometallic compounds the Cp ring, while its single-electron reduction results in activation of the C7H bonds of the methyl groups. Dimerisation of organometallic radicals according to the `ligand ± ligand' type can also be regarded as a reaction in the ligand sphere of the metal.Reductive dimerisation of the [Mn(Z6-C6H6)(CO)3]+ cationic complex, resulting in the forma- tion of a new C7C bond, is an example.48 [Mn(Z6-C6H6)(CO)3]+ e7 0.5 [Mn(CO)3]2-m-Z5:Z5-C6H67C6H6 (83%) 2 e7 0.5 {[Mn(CO)3]2-m-Z4:Z4-C6H67C6H6}27 (49%) Dimerisation giving a newC7Cbond in the ligand sphere of a metal was observed in a two-electron reduction of sandwich iridium complexes.49 2 e7 2 [(Z5-C5Me5)Ir(Z5-C5Me5)]+ (Z5-C5Me5)Ir(m-Z4:Z4-C5Me5 A similar approach has been used to prepare `molecular wires' consisting of polyyne chains with ferrocenyl and ruthenocenyl groups at the ends.50 In solutions, 19-electron arenecyclopentadienylruthenium radicals either undergo dimerisation involving the six-membered ring (Scheme 2, pathway b) or capture a hydrogen atom from the reaction medium (Scheme 2, pathway a).It was found that a decrease in the electron-donating capacity of ligands or an increase in the ligand bulk enhances the stability of radicals and dimerisation becomes the preferential process over the addition of a hydrogen atom.51, 52 a Ru + H e7 Ru Ru b H 5. Disproportionation Disproportionation can also take place with intermediate forma- tion of 19-electron complexes. For example, the Mn2(CO)10 dimeric complex can disproportionate in the presence of highly electron-donating nitrogen-containing tridentate ligands. This reaction is subject to both electrochemical and photochemical initiation 53 hn 2 [Mn(CO)5] ., Mn2(CO)10 [Mn(CO)3N3] .+2 CO, [Mn(CO)5] .+N3 [Mn(CO)3N3]++[Mn2(CO)10]7., [Mn(CO)3N3] .+Mn2(CO)10 C5Me5)Ir(Z5-C5Me5).Scheme 2 HHRu 227 [Mn [Mn(CO)5]7+[Mn(CO)5] .. 2(CO)10]7. The overall reaction is as follows: [Mn(CO)3N3]++[Mn(CO)5]7+2 CO. Mn2(CO)10+N3 III. Electrocatalysis: combination of catalysis by metal complexes with electrochemically induced electron transfer An important application of the electrochemical electron transfer in reactions involving metal complexes and proceeding via orga- nometallic intermediates is electrocatalysis. This line of research has been vigorously developing during the last 15 years and combines the advantages of catalysis by metal complexes and activation through electrochemical electron transfer.When the reaction is carried out at an electrode, electron transfer does not require additional oxidants or reducing agents, and the process is highly selective. Modification of the electrode surface by metal complexes (or the addition of complexes as mediators to the solution) provides the possibility of conducting electrochemically activated reaction at a substantially lower overpotential. The processes that combine catalysis by metal complexes with electrochemical transfer of an electron can be conventionally classified into two types. One type includes so-called mediator processes in which a complex is first reversibly reduced (or oxidised) at an electrode and then it passes the electron to an organic substrate (S) (or withdraws the electron from it in the case of oxidative activation).In this case, the electron transfer usually follows an outer-sphere mechanism, i.e., no bond between the metal complex and the organic substrate is formed. The mediator electron transfer yields an active form of the organic substrate (either reduced or oxidised form), which is then converted into product (P); as this takes place, the mediator (A) is regenerated. Oxidation Reduction A+. A7. A + e7 A7e7 A+S+. A+S7. A+.+S A7.+S P+. S7. P7. S+. P P+.+ e7 P P7.7e7 In the second type of process, an electron transfer from the catalyst (Cat) to the organic substrate is an inner-sphere process, i.e., it proceeds via the formation of an intermediate activated complex, decomposition and subsequent transformations of which afford product P.Oxidation Reduction Cat+e7 Cat+. Cat7. Cat7e7 CatS+. CatS7. Cat7.+S Cat+.+S P+.+Cat CatS7. P7.+Cat CatS+. P P+.+ e7 P P7.7e7 or or P+Cat+. P+.+Cat P+Cat7. P7.+Cat An advantage of mediator processes 54 is a decrease in the kinetic barrier to electrochemical reactions; the potential approaches the thermodynamic value. This is valid for processes that include a fast chemical step, S+(7)?P+(7). The mediator participating in these reactions should satisfy the following con- ditions: (1) the ability to undergo reversible oxidation or reduc- tion; (2) jEmj55jEsj (where Em and Es are the potentials of the mediator and substrate, respectively), the optimal potential differ- ence is usually 0.3 ± 0.5 V; (3) structural correspondence between the mediator and the substrate to ensure the electron transfer.The228 mediator can occur in the solution or, alternatively, be chemically bound to the electrode (chemically modified electrode). The inner-sphere electron transfer has a number of advantages over the outer-sphere process.27 1. The redox process does not depend on the potential differ- ence between the mediator and the substrate; therefore, a much more pronounced decrease in the overpotential is possible. 2. The inner-sphere transfer is characterised by a lower activation barrier. 3. An outer-sphere process allows the transfer of only one electron, while in the case of the inner-sphere mechanism, transfer of two or more electrons is possible.(This is especially important for bioelectrocatalysis and for activation of small molecules.) In real processes, the consumption of electricity is determined by the contributions of various competing reactions involving radical cations or radical anions derived from the catalyst. More- over, apart from the classical cases in which an inner-sphere or outer-sphere mechanism of electron transfer is regarded as pro- ven, there exist a large number of intermediate processes for which it is difficult to draw a conclusion about the nature of the transition state. In the case of metal complexes with an inner- sphere or a similar mechanism of electron transfer, the activated molecule enters the metal coordination sphere, which increases its reactivity. Numerous examples of electrocatalytic processes involving transition metal complexes are known.55 ± 61 The greatest number of studies have been devoted to the activation of small molecules and to various reactions of C7C bond formation (homo- and cross-coupling, carboxylation and carbonylation, the addition to multiple bonds).1. Electrocatalytic reduction of small molecules A series of reviews and original papers (see, for example, the studies 22, 56 ± 61) have been devoted to electrocatalytic reduction of nitrogen and its oxides, carbon oxides, and molecules or ions with triple bonds (acetylene, cyanide ion). A common feature of all the above compounds is that they are electrochemically inactive and contain very strong bonds that are fairly difficult to cleave.Therefore, activation of such molecules through introducing them into the metal coordination sphere followed (or preceded) by electrochemical electron transfer is the most promising way of involving these compounds in various chemical transformations. In addition, considerable attention to electrocatalytic processes with participation of metal complexes is inspired by the idea of utilisation of the enormous reserves of N2, CO2 and O2 present in the biosphere as a cheap and readily available feedstock for various chemical transformations. Since it is impossible to cover the tremendous number of publications devoted to various processes of this type, in this review, we consider only examples of electrocatalytic binding of CO2.These data, first, provide an idea of the main principles and features of electrocatalytic reactions and, second, they are fairly important from the practical standpoint. Direct electroreduction of CO2 at a metallic (for example, mercury) cathode requires a high overpotential: ECO2/CO¡2 .= 72.21 V vs. a saturated calomel electrode (SCE) in DMF.62 Other metallic cathodes, for example, copper or gold ones, are able to reduce CO2 in an aqueous solution at somewhat less cathodic potentials (71.3 to71.7 Vvs. SCE), but these values are also too great for practical application of this process 63, 64 because water is also reduced at these potentials with hydrogen evolution. This could be avoided by using non-aqueous solvents;65 ± 71 however, such a process is unfavourable due to a high energy expenditure and undesirable reduction of components present in the solution.A lot of data on the use of transition metal complexes for the electrocatalytic reduction of CO2 can be found in the literature (see, for example, studies 7, 58, 72, 73). As a rule, these are complexes with nitrogen-containing macrocyclic ligands such as porphyrins and their analogues, well-known as biocatalysts. Synthetic metal T V Magdesieva, K P Butin complexes with nitrogen-containing macrocyclic ligands, for example, phthalocyanines (Pc) or 1,4,8,11-tetraazacyclotetrade- cane (cyclam) derivatives, are often not inferior to enzymes in their catalytic activity.Metal phthalocyanines 3, 8 ± 11, 15, 74 ± 84 and metal porphyr- ins,2, 12, 13, 19, 58, 73, 85 ± 87 Ni(II) complexes with cyclam deriva- tives,88 ± 94 bis- and polypyridine complexes of Rh(II),95 Re(I),96 Cu(I),97 Co(II), Fe(II) and Ni(II),76, 84, 98 ± 101 iron and cobalt complexes with the 4,5-dihydroxybenzene-1,3-disulfonate and 2-hydroxy-1-nitrosonaphthalene-3,6-disulfonate ligands 102, 103 and other complexes can serve as catalysts for CO2 reduction. An important advantage of using metal complexes as catalysts for the reduction of CO2 is the possibility of changing their reactivity by modification of the macrocyclic ligand either by introducing electron-donating or -withdrawing substituents or by replacing the central metal atom.104 Indeed, a Ni(II) complex with a cyclam analogue, 3,10-dimethyl-1,3,5,8,10,12-hexaazacyclote- tradecane, exhibits a much higher catalytic activity than the unsubstituted Ni(II)(cyclam) complex.105 Surprisingly, the intro- duction of s-acceptor substituents, for example, fluorine, into the nitrogen-containing ring also increases the catalytic activity and the selectivity of CO formation (with respect to hydrogen evolu- tion).The Ni(II) complex with 3,3,10,10-tetrafluorocyclam proved to be the most efficient. Further increase in the number of fluorine atoms at the macrocycle leads to a lower catalytic activity.106 The influence of the nature of the central metal atom on the catalytic efficiency of CO2 electroreduction has been studied. In the case of metal complexes with meso-tetraphenylporphyrin, the highest current yield of CO was observed for Co, Fe and Zn complexes, while the catalytic activity of Ni, Cu, Mg and Mn compounds was fairly low.19 A wide range of monophthalocyanine complexes of various metals have been tested in the electrocatalytic reduction of CO2 at a gas-diffusion electrode in an aqueous solution.3, 15 The electro- reduction of carbon dioxide gave CO, HCOOH and CH4, the ratios and the current yields of which depended appreciably on the properties of the central metal atom in the phthalocyanine complex.With Group VIIIA transition metals such as Co, Ni, Fe and Pd, the reaction usually gives CO as the major product.The electrolysis of CO2 in the presence of Sn, Pb, In and Al phthalocyanine complexes affords mainly HCOOH, and in the case of Zn phthalocyanine, approximately equal amounts of HCOOHand CO are formed. Under the action of phthalocyanine complexes of Group IIIB and IVB elements, In and Sn, formic acid is produced with the highest selectivity, and in the case of Cu, Ga and Ti, methane is formed, although in a low yield. Metal-free phthalocyanine proved to be almost inefficient in the reduction of CO2. Thus, coordination of CO2 to the metal atom in the complex plays an important role. The data on the use of metal phthalocya- nines in reactions of CO2 reduction (see Refs 3 and 15) are listed in Table 1. It is of interest that the cathodic electrocatalytic activity of cobalt complexes with macrocyclic ligands increases if the Co atom is bound to the axial electron-donating ligand, for example, pyridine or imidazole.107 ± 109 Thus CO2 electroreduction in water at a glass-carbon electrode coated by Co tetraphenylporphyrin with the axial pyridine ligand occurs with a high selectivity.107 Recently,110 investigations of diphthalocyanine complexes as catalysts for carbon dioxide electroreduction in protic media have been started.The major reaction products were CO and HCOOH (or methyl formate if electroreduction is carried out in methanol). It is worthy of note that the formation of a metal ±CO2 complex is postulated as the key step in almost every study devoted to the catalytic fixing of CO2 by transition metal complexes.3, 9, 109 ± 114 In most cases, this is supported only by the data of IR or NMR spectroscopy because, first, the lifetime of these intermediates is short and, second, the presence of solvent molecules, also capable of coordination, hampers detailed analy- sis of the structure of the complex.Electrochemical activation of reactions involving organometallic compounds Table 1.Results of cathodic reduction of CO2 in water in the presence of phthalocyanine complexes of transition metals (PcM).3, 15 Current efficiency (%) E /V (vs. SCE) M CO CH4 28.1 0.8 002.0 0.2 4.2 1.3 000 28.0 0000.3 15.8 10.0 33.0 10.0 7.0 6.2 46.0 100.0 05.0 5.0 14.0 4.0 81.4 98.0 47.0 Ti In Zn Sn Mn Al Cd Ni Mo Cr Pb Cu Pt Pd Co Fe 71.98 71.96 71.94 71.93 71.90 71.90 71.84 71.75 71.72 71.68 71.64 71.63 71.59 71.50 71.30 71.21 Anumber of calculations for ligand-free complexesMCO2 (M is an alkali or transition metal) have been carried out (see Ref.115 and references therein). The first systematic study into the geometry, configurational interactions and electron density dis- tribution for a wide range ofMCO2 complexes (whereMis a metal of the 3-rd period: Ca ± Mn, Cu, Zn) showed the occurrence of electron transfer from the metal atom to the CO2 group, the electron density being concentrated on the carbon atom and the M7C bond being nearly ionic. The electron transfer from the metal to CO2 is accompanied by redistribution of electrons in the internal levels and, in particular, results in a higher electron density on the oxygen atom.Thus, coordination to the metal increased the basicity of the oxygen atoms in the CO2 group. Mechanisms for the electrocatalytic reduction of CO2 have been proposed in a few publications.79, 109 These studies deal with the reduction of CO2 in an aqueous solution at a graphite electrode coated by PcCo bound to poly-4-vinylpyridine (PVP), the latter acting as a donor axial ligand (Scheme 3). H2 H+ H+ e7 Co(I)Pc27 Co(II)Pc27 Co(I)Pc27H+ CO H O O O C C H2O H+ Co(II)Pc27 Co(II)Pc27 It can be seen from Scheme 3 that the catalytic process starts only in the second step of PcCo reduction because the first reduction is followed by the fast addition of H+ to one of the peripheral nitrogens of the phthalocyanine ring, and only the second step gives an active species capable of reducing both H+ and CO2 to H2 and CO, respectively.Nevertheless, research- ers 79, 83 who studied electrocatalysis by pure PcCo adsorbed on graphite, were unable to observe directly the PcCo7CO2 adduct. An interesting sequence of steps taking place during CO2 reduction catalysed by iron porphyrins has been proposed and confirmed by kinetic data.112, 113 The key step of this process is the formation of the Fe7C bond (Scheme 4, the porphyrin ligands are omitted, AH is a Brùnsted acid). HCOOH H2 40.4 47.7 40.0 20.0 88.0 69.9 45.5 2.0 98.0 90.2 30.0 49.2 95.6 9.5 3.0 47.3 5.0 26.8 27.0 71.0 3.0 16.9 4.8 0.4 2.0 5.0 65.0 3.5 01.0 0.4 5.0 Scheme 3 e7 Co(I)Pc37H+CO2 O O7 CCo(II)Pc37H+ 229 Scheme 4 Fe(I)7 + e7 Fe(0)27 O7 O AH Fe(II):C Fe(I)7.C Fe(0)27+CO2 O7 O7 O O7 Fe(I)7.C Fe(II):C O7 O7 H A H A O7 H A AH Fe(II):C O7 H A O Fe(I)7+.C O7 H A O O Fe(II) + 7C Fe(I)7+. C H A O 7O7 H A O7 H A Fe(II)CO+H2O+2A7 Fe(II):C O7 H A O 7C HCOOH +2A7 +AH H A O7 Fe(II)CO +Fe(0)27 Fe(I)CO7+Fe(I)7 2 Fe(I)7 Fe(II)+Fe(0)27 CO+Fe(I)7 Fe(I)CO7 It has been shown116 that a rotating ring ± disk electrode, in which the disk electrode is represented by graphite coated by cobalt(II) N,N0,N00,N000-tetramethyltetra-3,4-pyridoporphyrazine and protected by a polymeric film (Nafion) and the ring electrode is represented by platinum, can be used for quantitative determi- nation of CO2 in aqueous solutions (the concentration range is 1074 ± 3.261072 mol litre71).The method is based on the fact that CO2 is reduced quantitatively at the disk to give CO (at 71.15 V vs. SCE), while the latter is detected on the basis of the typical oxidation peak at the platinum ring (0.4 V vs. SCE): [CO]ad+H2O CO2+2H+ +2e7. 2. Elctroreductive coupling, addition, carboxylation and carbonylation Electroreductive coupling reactions are extensively used in syn- thesis. Phosphine, tetraazamacrocyclic and polypyridyl com- plexes of transition metals (mainly, Ni, Pd and Pt) are used most often as catalysts.Electrochemical electron transfer can play different roles in these reactions � an electron can participate directly in the catalytic cycle, or/and labile complexes of transition metals in low oxidation states can be generated electrochemically and participate in the reaction in situ. The general scheme of the elctroreductive coupling can be represented in the following way:27, 55, 117, 118 2 e7, ML2Cl2 Ar Ar+2 X7 2 ArX X=I, Br, Cl, OTf; M=Ni, Pd; L=PPh3, dppe, bipyridyl (bipy), etc.230 The activity of Ni(0) and Pd(0) complexes in coupling reac- tions of organohalogen compounds was shown 119, 120 to depend on a number of factors: the size of the electrochemical gap, the strength of the metal7ligand bond, and electronic and steric effects of the ligands.The first step of the catalytic cycle is the oxidative addition of an organic substrate to a zerovalent metal complex, resulting in the formation of a s-arylnickel or s-aryl- palladium intermediate. M(II)L2X2 2 e7 72X7 ArX 7X7 M(0)L2 e7 M(II)ArXL2 M(I)XL2 e7 Ar Ar M(I)ArL2 7X7 ArX M=Ni, Pd. Numerous reactions of this type have now been described (see the monograph 27 and the papers 55, 118, 121, 122). Among these, electrochemical synthesis of substituted 3-phenylthiophenes is noteworthy.123 The reduction of a mixture of 3-bromothiophene and ZnBr2 in the presence of the bipyridine nickel complex gives rise to 3-thienylzinc bromide, which enters into cross-coupling reactions with various aryl halides (PhI, p-NCC6H4Br, o-MeOC6H4Br, p-MeOC6H4Br).Y, Br ZnBr X e7, ZnBr2, NiBr2(bipy) Pd(PPh3)4 S SThe synthesis of optically active 1,3-dienes containing chiral substituents from alkenylstannanes by electrooxidation and homocoupling with participation of CuCl2 has been reported.124 Phosphine and bipyridine Ni complexes serve as the catalysts of the electroreductive coupling of aryl and hetaryl halides with mono- and dichlorophosphines, which allows the synthesis of tertiary phosphines in high yields.125, 126 2 e7 RBr Ni(II)RBrLm Ni(0)Lm Ni(II)Ln Ph2PCl 7Br7,7Cl7 Ni(II)Lm+Ph2PR L=PPh3. Under the action of electrochemically generated Ni(0) or Sm(II) compounds, arylation of white phosphorus takes place.127 e7 SmCl2+Cl7, SmCl3 Cl7 [P¡ P¡ SmClá SmCl2+P4 4 4 2 ] 7SmCl3 The intramolecular reductive coupling of allyl 2-halophenyl ethers catalysed by the Ni(cyclam)Br2 complex provides the preparation of 3-methyldihydrobenzofuran in 90% yield.128 Y S PhBr Ph3P.T V Magdesieva, K P Butin Me X e7, Ni(cyclam)Br2 O O X=Cl, Br, I. If CO2 or CO is introduced in the reaction mixture containing an organic halide and the electrochemically generated Ni(0) complex, the reaction gives carboxylic acids or carbonyl com- pounds, respectively, instead of the homocoupling products. Ni(II)L2X2 2 e7 72X7 ArCOO7 ArX Ni(0)L2 e7 Ni(II)ArXL2 Ni(II)Ar(CO2)L2 e7 CO2 Ni(I)ArL2 7X7 Carbon dioxide adds to Ni(I)ArL2 approximately twice as fast as ArX; therefore, the side reaction giving biaryls scarcely takes place.129 ± 133 In the presence of CO2 and catalytic amounts of electrochemically generated Pd(0) complex, aryl and nyl triflates are converted into aromatic and a,b-unsaturated carboxylic acids, respectively.134 Carboxylic acids can also be prepared from allyl and benzyl halides using Co complexes as catalysts.135 e7 RX e7, CO2 CoR(salen) [Co(salen)]7 Co(salen) 7X7 Co(salen) +RCOO7 R=PhCH2, Ph(Me)CH, MeCH=CHCH2; salen is bis(salicylidene)ethylenediamine.Recently,136 electrosynthesis of ketones from organic halides and CO catalysed by Ni(0) bipyridine complex was carried out.Study of the process mechanism showed that the reaction involves the intermediate formation of the acyl nickel complex. 2 e7, Ni(bipy) RCOR+2X7. 2RX+CO The use of electrochemical electron transfer makes it possible to `invert' the polarity of aroyl halides ArCOX and introduce them into reactions with electrophiles.137 2 e7, Pd(0) ArCOE+X7. ArX+CO+E+ The process includes the formation of the PdArX complex and the subsequent CO insertion, resulting in the intermediate PdArCOX complex, which is reduced more readily than the initial aryl halides; after activation by electrochemical electron transfer, this species can react with electrophiles, for example, with H+. Electrocatalytic arylation of alkenes is documented. Although the process mechanism is not entirely clear, it is assumed that coordinatively unsaturated Ni(0) complexes serve as the basis for the spatial proximity of halobenzenes and alkenes; the subsequent treatment with a base yields arylalkenes.138Electrochemical activation of reactions involving organometallic compounds IV.The formation and cleavage of carbon ± metal s-bonds Reactions giving carbon ± metal s-bonds are very important from the theoretical and practical standpoints because they allow the synthesis of new organometallic compounds with various useful properties. We shall consider two main types of these reactions, namely, electrochemical alkylation and electrochemical arylation of metal-containing nucleophiles. 1. Electrochemical alkylation The methods used most widely for the introduction of alkyl groups are based on SN2 reactions of various organic compounds with metal-containing nucleopiles. The latter can be synthesised conveniently by an electrochemical procedure using reductive activation, which yields metal complexes in low oxidation states.This method was used to prepare the alkyl derivatives of Co,139 Fe 140 and Rh porphyrins 141, 142 as well as complexes with other nitrogen-containing macrocyclic ligands, in particular, dimethyl- glyoximate 143 and phthalocyanine.144L L N N N N e7 e7, L CoII CoIII 7L 7Hal7 N N N N L Hal R L N N N N RHal CoIII CoI 7Hal7 N N N N L L is a Lewis base. A comparative study of alkylation of electrochemically generated nucleophilic Co, Rh, Ir, Ni and Pd complexes with various nitrogen- and sulfur-containing ligands on treatment with butyl bromide was carried out.143 Electrochemically activated reactions with transfer of the methyl group from dimethylglyoximate complexes of methylcobalt onto Au(I) 145 and Ni(I) 146 complexes and onto mercury 147 and germanium salts148 are also known.2. Electrochemical arylation and vinylation Using the approach described above, only alkyl fragments can be introduced into the metal coordination sphere. The introduction of aryl and vinyl substituents requires not only electrochemical activation of a metal-containing nucleophile but also an addi- tional activation of the C(sp2)7Hal bond (SRN1 reactions). Theoretical investigation into the electrochemically activated SRN1 type reactions was started by Saveant's group.12, 149 ± 157 The stabilities of the radical anions formed initially, the relative rates of each step of the SRN1 process, the most important competing reactions and the types of chain termination have been studied. A number of procedures for preparative electrosyn- thesis of various organic compounds underlain by SRN1 processes have been developed.153, 158 ± 160 However, until recently, the authors of only two studies 161, 162 have described examples of electrochemically activated SRN1 reactions involving organometallic compounds, namely, the syn- thesis of aryl and vinyl derivatives of iron porphyrin 161 and the reactions of the [CpM(CO)3]7 anions (M=Mo, W) with iodoni- trobenzene.162 Apparently, this is due to the fact that the double electro- chemical activation of both reagents requires that a specified potential be strictly maintained during the reaction.This creates some difficulties for practical implementation of these processes because the reaction product can be subsequently destroyed (compounds containing a carbon ± metal s-bond are usually oxidised and reduced irreversibly with cleavage of this bond). If 231 the product is reduced at more cathodic potentials than the reaction begins (the activation potential of the initial reactants), the product destruction at the electrode can be avoided.161 Systematic investigations of electrochemically activated arylation in which the reaction product is irreversibly reduced at potentials less cathodic than the activation potential of the reactants have been started 163 ± 167 in relation to electrochemical arylation and vinylation of anionic iron, molybdenum and tungsten carbonyl complexes.The general scheme of the reaction is as follows: e7 RHal+0.5 [CpM(CO)n]2 MR(CO)nCp+Hal7 M=Fe, Mo, W. Electrochemical electron transfer activates the organic halide by transforming it into a reactive radical and leads to generation of metal-containing nucleophiles as a result of cleavage of the M7M bond in the dimer. The active species that arise react with each other in situ to give a s-organometallic compound. Activation of the first reactant e7 RHal7.R.+Hal7. RHal Activation of the second reactant e7 0.5 [CpM(CO)n]2 [CpM(CO)n]7 M=Fe, n=2; M=Mo, W, n=3. The formation of the s-organometallic compound [MR(CO)nCp]7., R.+[CpM(CO)n]7 MR(CO)nCp. [MR(CO)nCp]7. 7e7 These reactions have been carried out for both aryl halides (p-XC6H4Hal: X=H, Me, Cl, I, COMe, CN, NO2; Hal=I, Br; C6F5X: X=H, F, Cl, Br; C5F5N) and vinyl halides (Z-, E-PhCH=CXHal:X=H, Ph;Hal=I, Br; Z-, E-PhCF=CFHal, Z-, E-ButCF=CFHal: Hal=F, Cl, Br). Direct electrochemical activation of this reaction is possible only for those aryl halides that are reduced more readily than the reaction product, a cyclopentadienyldicarbonyliron derivative. This is due to the fact that electrochemical reduction of s-organo- metallic compounds of this type is irreversible and is accompanied by C7M bond cleavage.For any other compound, the product resulting from electrochemical activation decomposes immedi- ately at the electrode. It can be seen from Fig. 2 that the region of ERed /V 3 H Me 1 Cl 2 2 COMe CN 3 1 NO2 0.2 0 0.6 0.4 0.8 sp Figure 2. Reduction potentials (ERed) (Pt cathode) of aryl iodides p-IC6H4X (1) and the products p-CpFe(CO)2C6H4X (2) vs. Hammett sp-constants of substituents. Reduction potentials of initial aryl iodides at aHg electrode (3). The region of possible electrochemical activation of the reaction is hatched.163232 possible electrochemical activation of the reaction is rather narrow. It can be extended by decreasing the potential of the onset of the reaction, i.e., the potential at which active species are generated.Three approaches have been proposed to solve this problem.One approach includes fast potential cycling toward a less cathodic region and back; thus, at least some of the product formed at the electrode has time to diffuse into the solution.164 According to the second approach,163 the electrode material is replaced (Hg is used instead of Pt); this allows a significant decrease in the potential at which the reaction starts because this facilitates the reduction of aryl and vinyl halides (see Fig. 2, dashed line). In this case, direct electrochemical activation can be used for iodobenzene, p-bromobenzonitrile and other aryl halides (see Fig. 2), the reduction of which at mercury falls into the required potential range, while electrochemical activation of these compounds at platinum is impossible.For vinyl halides studied by Magdesieva et al., 163 except for b-iodostyrene, the mere replace- ment of the electrode material does not suffice for electrochemical activation of the nucleophilic vinyl substitution because s-vinyl derivatives of Z5-cyclopentadienyldicarbonyliron are reduced more readily than the initial substrate even at mercury. The third approach 165 implies the use of homogeneous redox catalysis in which a decrease (in magnitude) in the potential of the reaction onset is attained by adding an appropriate reversible A/A7. redox pair, which acts as a mediator { by carrying an electron from the electrode to the aryl or vinyl halide.electrode 2 [CpFe(CO)2]7, [CpFe(CO)2]2+2e7 electrode A7., A + e7 A7.+RX solution (RX)7.+A, (RX)7. solution R.+X7, solution [FeR(CO)2Cp]7., R.+[CpFe(CO)2]7 [FeR(CO)2Cp]7.+RX solution FeR(CO)2Cp+RX7., [FeR(CO)2Cp]7.+A solution FeR(CO)2Cp+A7. R is aryl, vinyl. Owing to the homogeneous redox catalysis, these reactions can be carried out for such substrates for which the replacement of electrode material alone does not result in a positive value of DE=ERed RHal7ERed FeRÖCOÜ2Cp (for example, p-iodotoluene, bromostyrenes and Z-bromostil- bene). In addition, by using this approach, one can increase the yield of nucleophilic substitution products for substrates that formally fall into the region of possible reaction activation upon the replacement of the electrode material but still have too small DE values, for example, for iodobenzene (Table 2).Thus, Magdesieva et al. 165 were the first to show experimen- tally that the use of homogeneous redox catalysis allows the electrochemically activated reactions of organometallic com- pounds to be performed under conditions where the products are redox-unstable upon direct electrochemical activation. This implies selection of a mediator whose activities with respect to the reactant and the product are appreciably different. This has markedly extended the scope of applicability of electrochemical activation, first of all, for the synthesis of s-organometallic compounds.{ Study of a wide range of conceivable mediators showed that phthalo- nitrile is the compound of choice for this reaction.165 Table 2. Yields of products of direct and mediator preparative electro- lysis a of aryl and vinyl halides in the presence of [CpFe(CO)2]2.163 ± 165 Substrate Potential of electrolysis /V p-MeC6H4I p-MeC6H4I 71.95 cycling from71.95 to71.55 71.80 p-MeC6H4I in the presence of o-NCC6H4CNc PhI PhI 71.90 cycling from71.90 to71.55 PhI in the presence 71.75 of o-NCC6H4CNc PhCH=CHI PhCH=CHI 71.55 cycling from71.90 to71.55 71.75 PhCH=CHI in the presence of o-NCC6H4CNc a Electrolysis conditions: Hg, Bun4 NBF4 (0.25 mol litre71), MeCN, Ag+/AgCl/KCl, 20 8C. b Fp=CpFe(CO)2. c The phthalodinitrile con- centration is 561074 mol litre71.This approach has permitted the preparative synthesis of about twenty aryl and vinyl derivatives of CpFe(CO)2 containing the C(sp2)7metal s-bond, particularly high product yields being attained when using homogeneous redox catalysis. Study of the stereochemistry of the reaction products (Table 3) showed that, irrespective of the configuration of the initial vinyl halide, the product with the trans-arrangement of the organometallic group and the organic group always predomi- nates. This is due to the fact that generation of the radical anion is accompanied by isomerisation to give the most stable configura- tion (Scheme 5). Table 3. Results of preparative electrolysis a of vinyl halides RCF=CFX in the presence of the [CpFe(CO)2]2 binuclear complex.167 R X Sub- strate config- uration F (CF3)3C (CF3)3C Cl Z (CF3)3C Br Z F Ph Z Cl Ph E Cl Ph a Electrolysis conditions: Hg, Bun4 NBF4 (0.25 mol litre71), MeCN, Ag+/AgCl/KCl, 20 8C.T V Magdesieva, K P Butin Product Prepa- rative yield (%) 0± 1 2 ± 3 p-MeC6H4Fp b p-MeC6H4Fp 43.0 p-MeC6H4Fp 5.0 15.0 C6H5Fp C6H5Fp 45.0 C6H5Fp 12.5 19.0 PhCH=CHFp PhCH=CHFp 28.0 PhCH=CHFp Nucleophilic substitution product config- uration Product Prepa- rative yield (%) (CF3)3CCF=CFFp (CF3)3CCF=CFFp (CF3)3CCF=CFFp PhCF=CFFp PhCF=CFFp PhCF=CFFp 40 traces 15 traces 14 traces 425 323 105 ZEZEZEZEZEZEElectrochemical activation of reactions involving organometallic compounds Ph H e7 Ph H 7e7 Ph Br Br Ph H Ph Br Ph7Br7 HPh Ph Fp7 Fp Ph H Ph 7e7 H Fp Ph Ph Using electrochemical activation, it is possible to arylate 168 Co(II) and Co(III) complexes with chelating ligands containing nitrogen and sulfur atoms, for example: H O O Py Me Me N N Co N N Me Me X O O H X=Py, Hal.The essence of the `iodonium method' proposed in one study 168 is to use diphenyliodonium salts including those containing various substituents in the benzene ring as arylating agents. e7 Ph.+PhI, Ph2I+ Co(III)Ph(Chel)2 Ph.+Co(II)(Chel)2 Chel is a chelating ligand. Not only Co(II) complexes but also Co(III) complexes, which are normally more stable and more readily available, can be introduced in the arylation.However, in this case, electrochemical activation of both reactants is needed. e7 Ph.+PhI+BF¡ Ph2IBF4 e7 Co(II)(Chel)2, Co(III)(Chel)2 Co(III)Ph(Chel)2. Ph.+Co(II)(Chel)2 It is noteworthy that the use of diphenyliodonium salts as arylating agents has obvious advantages over halobenzenes used Scheme 5 Br H H Ph 7e7 e7 Ph Ph Br Ph Ph Br Ph H7Br7 Ph Ph H Fp7 Fp Ph Ph H 7e7 Ph Fp H Ph R N C N S Me Co Me N S N C R R=Me, OMe. 4 , 233 traditionally in SRN1 reactions, because the former provide generation of phenyl radicals upon electrochemical activation at much lower cathodic overpotentials. 3.Electrosynthesis with sacrificial electrodes Yet another widely used electrochemical approach to the synthesis of organometallic compounds is the use of sacrificial (this is the name for soluble) anodes. In an undivided cell, metal dissolution takes place at the anode and the organic substrate is reduced at the cathode to give anions which should be stable enough to avoid side reactions with the solvent.28, 169 ± 171 (nX7) nRH (or RX) n e7 (or 2n e7) M7n e7 n/2H2 (or n X7) + nR7 Mn+(X¡)n MRn Examples of syntheses with anodes made of Ti, Zr, Hf, V, Cr, Mn, Fe, Co, Ni, Cu, Al, Pb, Sn and various ligands are docu- mented.172 ± 175 Kharisov et al.176 have compared the traditional chemical and electrochemical methods for the synthesis of transition metal complexes with nitrogen-containing ligands.An advantage ensured by the use of sacrificial anodes is the possibility of preparing Cu, Ni, Co, Pd, Zn and Cd complexes in one step in high yields. Electrochemical cleavage of the carbon ± metal s-bond gives in most cases organic radicals the further transformations of which are determined 177 by the applied potential and the solution composition. Fe(IV)R2(bipy)2 Fe(III)R2(bipy)2 Fe(II)R2(bipy)2 7e7 7e7 b-elimination (alkenes, alkanes) concerted elimination (R R) homolytic cleavage (R. and the products of their reactions) V. Electrochemically activated reactions involving `electron sponges' When considering electrochemical activation of reactions involv- ing organometallic compounds, two types of processes can be distinguished: (1) electrochemical activation of various types of bonds (C7Hal,C7M,M7Hal,M7M, etc.), resulting in their cleavage to give highly reactive species (cations, anions, radicals); (2) activation through transfer of one or several electrons accompanied by the change in the electronic structure and, hence, properties of the molecule, but involving no bond cleavage.In the reactions considered above, activation of the former type takes place. Now we shall give some examples of the latter type of processes. Systems capable of reversible stepwise addition or release of several (three or more) electrons are referred to as `electron sponges'.The transfer of an electron onto such systems results in activation of the whole molecule and in the change in its properties (for example, catalytic, spectroscopic and other properties) but does not induce cleavage of any bond. A fairly large number of such systems have been reported;178, 179 however, here we restrict ourselves to consideration of two of them, namely, phthalocya- nine complexes and fullerene derivatives, for which electrochemi- cally activated reactions involving organometallic compounds are known.234 1. Electrosynthesis of exohedral metal complexes of C60 and C70 A unique feature of fullerenes and their derivatives is the capa- bility of reversible stepwise addition of up to six electrons, which makes them fairly promising participants of redox processes.Recently,180, 181 we proposed an electrochemical method for the synthesis of exohedral complexes of C60 and C70, namely, C60ML2 [M=Pd, Pt; L2=(Ph3P)2, bis(diphenylposphino)ferrocene (dppf) or diphenylposphinocymanthrene] and C70Pt(PPh3)2 (a mixture of two isomers) from simple and stable inorganic salts (MCl2) or complexes of platinum group metals (ML2Cl2). M(PPh3)2 M=Pd, Pt. Pt(PPh3)2 Reactive complexes of zerovalent metals sensitive to atmos- pheric oxygen can be generated in situ using electrochemically generated C2¡ 60 or C270¡ dianions as reducing agents. In this case, fullerene plays a dual role acting simultaneously as an electron transporter (mediator) and as one of structure-forming reagents.2 e7 C60 C2¡ 60 , C2¡ C60+ML2+2Cl7, 60 +ML2Cl2 C60ML2. C60+ML2 An advantage of this method is the possibility of preparing a pure monometal complex containing no polymetallated fuller- enes. This is related to the fact that the electron transfer is followed by the fast reaction of C60 with theML2 compound formed in situ, which is formed in an exactly equimolar amount. This approach has also been used to prepare new exohedral heteropolymetallic complexes of [60]fullerene containing Pd, Fe and Mn atoms.182, 183 2 e7 C60 C260¡ PdCl2+dppf dba is dibenzylideneacetone; 2 e7 C60 C2¡ 60 , C2¡ 60 +PdCl2+2Ph2POC Pt(PPh3)2 Pd2(dba)3 .C6H6+dppf+C60 Ph Ph P Fe Pd P Ph Ph 72Cl7 Mn CO CO 2. Electrochemically activated reactions of phthalocyanines Typical examples of `electron sponges' are mono- and diphthalo- cyanine metal complexes.R N N N N M N N N R N R R N N N N N R N R MN N N N R N N NN R Recently, it has been found 184 that some diphthalocyanine complexes are able to undergo eight reversible and quasi-rever- sible redox transitions in the anodic and cathodic regions. Owing to this feature, they are rather promising participants of various redox processes, in particular, electrocatalytic ones. In Section III.1, we presented examples of electrocatalytic reduction of CO2 by various complexes with macrocyclic nitrogen-containing ligands, including phthalocyanine complexes. Now we consider one more example of an electrochemically activated reaction of phthalocyanines in which the electrochemical transfer of an electron is used for catalyst activation.The binding of CO2 by epoxides resulting in the formation of alkylene carbonates is known185 to be catalysed by metal phtha- locyanines. However, this reaction proceeds under rigorous con- ditions, i.e., in an autoclave at high pressure and temperature. With electrochemically reduced forms of metal phthalocyanines used as catalysts, this reaction was found74 to proceed at 20 8C and under atmospheric pressure, in conformity with the scheme R Cat CO2+ O R=Me, CH2Cl. T V Magdesieva, K P Butin COCO OC Ph Mn Ph P Pd P Ph Ph Mn OC CO CO R X R N R N N N M N N N R N R R N NR RR O O OElectrochemical activation of reactions involving organometallic compounds The electrochemical activation of the catalyst and the subse- quent nucleophilic attack occur in the following way: e7 (PcM)7., PcM (PcM .CO2)7., (PcM)7.+CO2 O R N C + O7 O O N C O 7O R N=PcM. Thus, electrochemically reduced forms of metal phthalo- cyanines exhibit much higher catalytic activities than neutral forms. Both mono- and diphthalocyanine complexes can be used as catalysts in this reaction; however, the latter offer an N N N NH HN N N N CO2 O N N N + NH HN N N N E B O HO H N HN O O R SO OHO7 N C HN 7 + OH R S ADP is adenosine diphosphate, A is adenosine; R=(CH2)4COOH.R O O 7 N7 O N e7 NH N O7 NNH NO N C HN HA E R S O O O O O P P P OA O O7 O7 O7 Mg2+ advantage, namely, in this case, the electrochemically activated binding of CO2 can be carried out at a low cathodic polarisation (the potentials of the first redox transition of Pc2M range from +0.01 to 70.65 V vs. Ag/AgCl/KCl).74, 104, 186 A study of the influence of the catalyst immobilisation mode on the efficiency of the catalytic process showed that the highest efficiency is observed in the case of amino-substituted mono- and diphthalocyanine complexes (NH2PcM and NH2Pc2M, respectively) electropolymerised on smooth electro- des. Turnover numbers as high as 104 ± 105 have been achieved for these catalysts, which is close to the efficiency of natural enzymes.187 Monomer Catalyst turnover number in 75 min It is of interest that carboxy-containing adducts formed under the action of phthalocyanine complexes and biotin (the co-enzyme which bindsCO2 and transports it to organic molecules in vivo) are structurally similar: in both cases, they contain a fragment similar to carbamic acid (Scheme 6).N7 N HN N N CO2 O O7 N N HN N N OHO7 OH 7ADP, 7Mg2+(HOPO2¡ 3 ),7H+ NH2PcCu NH2PcCo 2.656105 4.76103 1.056103 1.46104 N7 N N e7 NH HN 7N N N CO2 O N N N NH HN N7 N N O O N C HN O7 R S 235 NH2 NH2PcPd Pc2Lu NH2Pc2Er 7.06103 Scheme 6 O7236 The above data indicate that the simulation of functioning of the active sites of enzymes operating by the proton transfer mechanism by processes occurring by a redox mechanism is possible but only at the level `the same intermediate (in this particular case, the resonance-stabilised planar N anion) � the same product,'188 * * * The information presented here demonstrates that, although numerous studies on electrochemically activated reactions involv- ing organometallic compounds have been performed, the list of types of these reactions is quite limited.The vast majority of publications are devoted to transformations in the coordination sphere of transition metal complexes and electrocatalysis. There- fore, development of new electrochemical approaches to the synthesis of organometallic compounds and conduction of reac- tions involving them remains to be quite a topical task.In this review, we attempted to demonstrate the extensive potential of organic electrochemistry and thus facilitate implementation of electrochemical methods in the routine practice of chemists work- ing with organometallic compounds. This work was supported by the Russian Foundation for Basic Research (Project Nos. 01-03-33147 and 00-03-32888) and Science and Engineering Programme `Fullerenes and Atomic Clusters'. References 1. M Chanon Bull. Soc. Chim. Fr. 209 (1985) 2. R L Rich, H Taube J. Am. Chem. Soc. 76 2608 (1954) 3. F Basolo, M L Morris, R G Pearson Discuss. Faraday Soc. 29 80 (1960) 4. F Basolo, P H Wilks, R G Pearson, R G Wilkins J.Inorg. Nucl. Chem. 6 161 (1958) 5. F Basolo, A F Messing, P H Wilks, R G Wilkins, R G Pearson J. Inorg. Nucl. Chem. 6 203 (1958) 6. F Basolo, R G Pearson Mechanisms of Inorganic Reactions (New York: Wiley, 1967) 7. N Kornblum, R E Michel, R C Kerber J. Am. Chem. Soc. 88 5662 (1966) 8. N Kornblum Angew. Chem., Int. Ed. 14 734 (1975) 9. J F Bunnett Acc. Chem. Res. 11 413 (1978) 10. G A Russel, W C Danen J. Am. Chem. Soc. 88 5663 (1966) 11. G A Russel Spec. Publ. Chem. Soc. 24 271 (1970) 12. J-M Saveant Acc. Chem. Res. 13 323 (1980) 13. M Julliard, M Chanon Chem. Rev. 83 425 (1983) 14. A J L Pombeiro New J. Chem. 21 649 (1997) 15. D Astruc Angew. Chem., Int. Ed. Engl. 27 643 (1988) 16. D Astruc Chem.Rev. 88 1189 (1988) 17. D Astruc Acta Chim. 7 69 (1996) 18. M C Baird Chem. Rev. 88 1217 (1988) 19. F Battaglini, E J Calvo, F Doctorovich J. Organomet. Chem. 547 1 (1997) 20. MVenturi, A Credi, V Balzani Coord. Chem. Rev. 185 ± 186 233 (1999) 21. A M Allgeier, C A Mirkin Angew. Chem., Int. Ed. 37 894 (1998) 22. O N Efimov, V V Strelets Usp. Khim. 57 228 (1988) [Russ. Chem. Rev. 57 129 (1988)] 23. J K Kochi Organometallic Mechanisms and Catalysis (New York: Academic Press, 1978) 24. Topics in Organic Electrochemistry (New York; London: Plenum, 1986) 25. JW Hershberger, J K Kochi J. Chem. Soc., Chem. Commun. 212 (1982) 26. V V Strelets Elektrokhimiya 32 6 (1996) a 27. D Asturc Electron Transfer and Radical Processes in Transition Metal Chemtry (New York: VCH, 1995) 28.L Walder, in Organic Electrochemistry (Eds H Lund,M M Baizer) (New York; Basel; Hong Kong: Marcel Dekker, 1991) p. 809 29. M Lacoste, M H Delville-Desbois, N Ardoin, D Astruc Organometallics 16 2343 (1997) T V Magdesieva, K P Butin A J L Pombeiro, R L Richards J. Chem. Soc., Dalton Trans. 3015 30. B T Donovan-Merkert, P H Rieger,W E Geiger Organometallics 18 3194 (1999) 31. L K Yeung, J E Kim, Y K Chung, P H Rieger, D A Sweigart Organometallics 15 3891 (1996) 32. T Al Salih, M T Duarte, J J R Frausto de Silva, A M Galvao, M F C Guedes da Silva, P B Hitchcock, D L Hughes, C J Pickett, (1993) 33. C Amatore, J-N Verpeaux, A Madonik, M-H Desbois, D Astruc J. Chem. Soc., Chem. Commun. 200 (1988) 34.P Federlin, J-M Kern, A Rastegar, P A Marnot, J-P Sauvage New J. Chem. 14 9 (1990) 35. R D Rieke, H Kojima, K Ofele J. Am. Chem. Soc. 98 6735 (1976) 36. M F C Guedes da Silva, P B Hitchcock, D L Hughes, K Marjani, A J L Pombeiro, R L Richards J. Chem. Soc., Dalton Trans. 3725 (1997) 37. W E Geiger, M J Shaw,M Wmnsch, C E Barnes, F H Foersterling J. Am. Chem. Soc. 119 2804 (1997) 38. M J Shaw,W E Geiger, J Hyde, C White Organometallics 17 5486 (1998) 39. L Pospishil, J Fiedler, D Osella, C Nervi J. Electroanal. Chem. 412 147 (1996) 40. A Inagaki, Y Takaya, T Takemori, H Suzuki, M Tanaka,M Haga J. Am. Chem. Soc. 119 625 (1997) 41. C Amatore, A Ceccon, S Santi, J -N Verpeaux Chem. Eur. J. 3 279 (1997) 42. R H Magnuson, R Meirowitz, S J Zulu,W P Giering Organometallics 2 460 (1983) 43.D Miholova', A A VlcÆ ek J. Organomet. Chem. 240 413 (1982) 44. B A Narayanan, C Amatore, C P Casey, J K Kochi J. Am. Chem. Soc. 105 6351 (1983) 45. W E Silverthorn J. Am. Chem. Soc. 102 842 (1980) 46. M G Peterleitner, L I Denisovich, D N Kravtsov Izv. Akad. Nauk SSSR, Ser. Khim. 442 (1988) a 47. O V Gusev, L N Morozova, T A Peganova,M G Peterleitner, SMPeregudova, L I Denisovich, P V Petrovskii, Yu F Oprunenko, N A Ustynyuk J. Organomet. Chem. 493 181 (1995) 48. S Lee, S R Lovelace,D J Arford, S J Geib, S G Weber,N J Cooper J. Am. Chem. Soc. 118 4190 (1996) 49. O V Gusev, M G Peterleitner,M A Ievlev, A M Kal'sin, P V Petrovskii, L I Denisovich,N A Ustynyuk J. Organomet. Chem.531 95 (1997) 50. Y Hayashi, M Osawa, Y Wakatsuki, in First International Forum on Hyper-Molecular Structure, Amsterdam, 1996 p. 35 51. O V Gusev, M A Ievlev, M G Peterleitner, S M Peregudova, L I Denisovich, P V Petrovskii,N A Ustynyuk J. Organomet. Chem. 534 57 (1997) 52. O V Gusev, M A Ievlev, M G Peterleitner, S M Peregudova, L I Denisovich, P V Petrovskii,N A Ustynyuk Izv. Akad. Nauk, Ser. Khim. 1691 (1996) b 53. A E Stiegman, A S Goldman, C E Philbin, D R Tyler Inorg. Chem. 25 2976 (1986) 54. Yu G Budnikova,G K Budnikov Zh. Obshch. Khim. 65 1517 (1995) c 55. M C Chakravorti, R Subramanian Coord. Chem. Rev. 135 65 (1994) 56. A E Shilov, G B Shul'pin Chem. Rev. 97 2879 (1997) 57. J Costamagna, G Ferraudi, B Matsuhino, M Campos-Vallette, J Canales,M Villagran, J Vargas, M F Aguirre Coord.Chem. Rev. 196 125 (2000) 58. P Vasudevan, N Phougat, A K Shukla Appl. Organomet. Chem. 10 591 (1996) 59. T A Bazhenova, A E Shilov Coord. Chem. Rev. 144 69 (1995) 60. H B Mark Jr, J F Rubinson, J Krotine,W Vaughn, M Goldshmidt Electrochim. Acta 45 4309 (2000) 61. M Shibata,K Murase,N Furuja J. Appl. Electrochem. 28 1121 (1998) 62. E Lamy, L Nadjo, J-M Saveant J. Electroanal. Chem. 78 403 (1977) 63. H Noda, S Ikeda, Y Oda,K Imai,M Maeda,K Ito Bull. Chem. Soc. Jpn. 63 2459 (1990) 64. Y Hori, K Kikuchi, A Murata, S Suzuki Chem. Lett. 897 (1986) 65. T R O'Toole, B P Sullivan,M R M Bruce, L D Margerum, R W Murray, T J Meyer J. Electroanal. Chem. 259 217 (1989) 66. A Gennaro, A A Isse, E Vianello J.Electroanal. Chem. 289 203 (1990) 67. E Fujita, B S Brunschwig, T Ogata, S Yanagida Coord. Chem. Rev. 132 195 (1994) 68. J-P Collin, J-P Sauvage Coord. Chem. Rev. 93 245 (1989)Electrochemical activation of reactions involving organometallic compounds 69. P Christensen, A Hamnett, A V G Muir, J A Timney J. Chem. Soc., Dalton Trans. 1455 (1992) 70. P Christensen, A Hamnett, A V G Muir, J A Timney, S Higgins J. Chem. Soc., Faraday Trans. 90 459 (1994) 71. W J Albery, P Barron J. Electroanal. Chem. 138 79 (1982) 72. Electrochemical and Electrocatalytic Reduction of Carbon Dioxide (Amsterdam: Elsevier, 1993) 73. J Costamagna, G Ferraudi, J Canales, J Vargas Coord. Chem. Rev. 148 221 (1996) 74. T V Magdesieva, S V Milovanov, B V Lokshin, Z S Klemenkova, L G Tomilova, K P Butin, N S Zefirov Izv.Akad. Nauk, Ser. Khim. 2205 (1998) b 75. J Zagal,M Paez, C Fierro, in Electrode Materials and Processes for Energy Conversion and Storage (Eds S Srinivanasan, S Wagner, H Wrobloba) (Pennington, NJ: The Electrochemical Society, 1987) p. 198 76. T Yoshida, T Iida, T Shirasaji, R Lin,M Kaneko J. Electroanal. Chem. 344 355 (1993) 77. H Tanabe, K Ohno Electrochim. Acta 32 1121 (1987) 78. E R Savinova, S A Yashnik, E N Savinov, V N Parmon React. Kinet. Catal. Lett. 46 249 (1992) 79. T Yoshida, K Kamato,M Tsukamoto, T Iida, D Schlettwein, D WoÈ hrle, M Koneko J. Electroanal. Chem. 385 209 (1995) 80. S Meshitsuka, M Ichikawa, K Tamaru J. Chem. Soc., Chem. Commun.158 (1974) 81. D Masheder, K P J Williams J. Raman Spectrosc. 18 391 (1987) 82. M N Mahmood, D Masheder, C J Harty J. Appl. Electrochem. 17 1223 (1987) 83. C M Lieber, N S Lewis J. Am. Chem. Soc. 106 5033 (1984) 84. P Christensen, S Higgins J. Electroanal. Chem. 387 127 (1995) 85. M Tezuka,M Iwasaki Chem. Lett. 241 (1993) 86. M Hammouche, D Lexa, J-M Saveant J. Electroanal. Chem. 249 347 (1988) 87. N Furuya, S Koide Electrochim. Acta 36 1309 (1991) 88. M Fujihira, Y Hirata, K Suga J. Electroanal. Chem. 292 199 (1990) 89. C de Alwis, J A Crayston, T Cromie, T EisenblaÈ tter, R W Hay, Y D Lampeka, L V Tsymbal Electrochim. Acta 45 2061 (2000) 90. J-P Collin, A Jouaiti, J-P Sauvage Inorg. Chem. 27 1986 (1988) 91. M Beley, J-P Collin, R Ruppert, J-P Sauvage J.Chem. Soc., Chem. Commun. 1315 (1984) 92. M Beley, J-P Collin, R Ruppert, J-P Sauvage J. Am. Chem. Soc. 108 7461 (1986) 93. C B Balasz, F C Anson J. Electroanal. Chem. 322 325 (1992) 94. F Abba, G De Santis, L Fabbrizzi, M Licchelli, A M M Lanfredi, P Pallavicini, A Poggi, F Ugozzoli Inorg. Chem. 33 1366 (1994) 95. M N C D Sauthier, A Deronzier, R Ziessel Inorg. Chem. 33 2961 (1994) 96. S C Rasmussen, M M Richter, E Yi, H Place, K J Brewer Inorg. Chem. 29 3926 (1990) 97. R J Haines, R E Wittrig, C P Kubiak Inorg. Chem. 33 4723 (1994) 98. J A R Sende, C R Arana, L Hernandez, K T Potts, M Keshavarz-K, H D AbrunÄ a Inorg. Chem. 34 3339 (1995) 99. H C Hurrell, A-L Mogstad, D A Usifer, K T Potts, H D AbrunÄ a Inorg.Chem. 28 1080 (1989) 100. C Arana, S Yan, M Keshavarz-K, K T Potts, H D AbrunÄ a Inorg. Chem. 31 3680 (1992) 101. C Arana,M Keshavarz-K, K T Potts, H D AbrunÄ a Inorg. Chim. Acta 225 285 (1994) 102. K Ogura,M Higasa, J Yano, N Endo J. Electroanal. Chem. 379 373 (1994) 103. K Ogura, H Sugihara, J Yano,M Higasa J. Electrochem. Soc. 141 419 (1994) 104. T V Magdesieva, I V Zhukov, L G Tomilova, O V Korenchenko, I P Kalashnikova, K P Butin Izv. Akad. Nauk, Ser. Khim. 379 (2001) b 105. C I Smith, J A Crayston, R W Hay J. Chem. Soc., Dalton Trans. 3267 (1993) 106. M Shionoya, E Kimura, Y Iitaka J. Am. Chem. Soc. 112 9237 (1990) 107. T Atoguchi, A Aramata, A Kazusaka,M Enyo J. Electroanal. Chem. 318 309 (1991) 108. Z Y Zeng, S I Gupta, H Huang, E B Yeager J.Appl. Electrochem. 21 973 (1991) 237 109. T Abe, H Imaya, T Yoshida, S Tokita, D Schlettwein, D Wlhrle,M Koneko J. Porphyrins Phthalocyanines 1 315 (1997) 110. T V Magdesieva, I V Zhukov, D N Kravchuk, O A Semenikhin, L G Tomilova, K P Butin Izv. Akad. Nauk, Ser. Khim. 742 (2002) b 111. K Koga, N Morohuma Chem. Phys. Lett. 202 330 (1993) 112. I Bhugun, D Lexa, J-M Saveant J. Am. Chem. Soc. 116 5015 (1994) 113. I Bhugun, D Lexa, J-M Saveant J. Am. Chem. Soc. 118 1769 (1996) 114. K P Butin, Yu M Kiselev, T V Magdesieva, O A Reutov J. Organomet. Chem. 235 127 (1982) 115. M Maggini, A Karlsson, G Scorrano, G Sandona, G Farnia, M Prato J. Chem. Soc., Chem. Commun. 589 (1994) 116. J Zhang, W J Pietro, A B P Lever J.Electroanal. Chem. 403 93 (1996) 117. A Jutand, A Mosleh J. Org. Chem. 62 261 (1997) 118. P W Jennings, D G Pilsbury, J L Hall, V T Brice J. Org. Chem. 41 719 (1976) 119. Yu G Budnikova, Yu M Kargin Zh. Obshch. Khim. 65 1536 (1995) c 120. Yu G Budnikova, D G Yakhvarov, Yu M Kargin Zh. Obshch. Khim. 68 1123 (1998) c 121. C Amatore, A Jutand Organometallics 7 2203 (1988) 122. Novel Trends in Electroorganic Synthesis (Tokyo: Kodansha, 1995) 123. C Gosmini, J Y Nedelec, J Perichon Tetrahedron Lett. 38 1941 (1997) 124. I Toshiyuki, E Sachie, K Michiyo, O Hiroyuki, T Hideo, T Shigeru Electrochim. Acta 42 2133 (1997) 125. Yu G Budnikova, Yu M Kargin, J-Y Nedelec, J Perichon J. Organomet. Chem. 575 63 (1999) 126. Yu G Budnikova, Yu M Kargin Zh.Obshch. Khim. 65 1660 (1995) c 127. Yu G Budnikova, D G Yakhvarov, Yu M Kargin Zh. Obshch. Khim. 68 603 (1998) c 128. S Olivero, J-P Rolland, E DunÄ ach Organometallics 17 3747 (1998) 129. C Amatore, A Jutand Acta Chem. Scand. 44 755 (1990) 130. C Amatore, A Jutand J. Am. Chem. Soc. 113 2819 (1991) 131. J F Fauvarque, C Chevrot, A Jutand,M FrancË ois, J Perichon J. Organomet. Chem. 264 273 (1984) 132. L Garnier, Y Rollin, J Perichon J. Organomet. Chem. 367 347 (1989) 133. J Y Nedelec, J Perichon, M Troupel J. Electrochem. Soc. 137 150 (1991) 134. A Jutand, S Negri Eur. J. Org. Chem. 1811 (1998) 135. J-G Folest, J-M Duprilot, J Perichon Tetrahedron Lett. 26 2633 (1985) 136. M OcË afrain, E Dolhem, J Y Nedelec,M Troupel J.Organomet. Chem. 571 37 (1998) 137. C Amatore, E Carre, A Jutand, H Tanaka, S Torii, I Chiarotto, I Carelli Electrochim. Acta 42 2143 (1997) 138. M Troupel, J Robin, G Meyer, J Perichon Nouv. J. Chim. 9 480 (1985) 139. D Lexa, J-M Saveant, J P Saufflet J. Electroanal. Chem. 100 159 (1979) 140. D Lexa, J-M Saveant, D L Wang Organometallics 5 1428 (1986) 141. J E Anderson, C-L Yao, K M Kadish Inorg. Chem. 25 718 (1986) 142. Y Aogama, T Yoshida, K Sakurai, H Ogoshi Organometallics 5 168 (1986) 143. K P Butin, R D Rakhimov, I G Il'ina Izv. Akad. Nauk, Ser. Khim. 71 (1999) c 144. R Taube, H Drevs, D Steinborn Z. Chem. 12 425 (1978) 145. R D Rakhimov, K I Grandberg, K P Butin Vestn. Mosk. Univ., Ser. 2, Khimiya 40 1 (1999) d 146. R D Rakhimov, K P Butin Izv.Akad. Nauk, Ser. Khim. 55 (2000) b 147. R D Rakhimov, E R Milaeva, O V Polyakova, K P Butin Izv. Akad. Nauk, Ser. Khim. 309 (1994) b 148. R D Rakhimov, K P Butin Izv. Akad. Nauk, Ser. Khim. 2157 (1997) b 149. J-M Saveant, M-G Severin, A A Isse J. Electroanal. Chem. 402 195 (1996) 150. J-M Saveant, M-G Severin, A A Isse J. Electroanal. Chem. 399 157 (1995) 151. J-M Saveant Acc. Chem. Res. 26 455 (1993) 152. J-M Saveant Tetrahedron 50 10117 (1994) 153. C P Andrieux, C Blocman, J M Dumas-Bouchiat, J-M Saveant J. Am. Chem. Soc. 101 3431 (1979)T V Magdesieva, K P Butin 238 154. C Amatore, J-M Saveant, A Thiebault J. Electroanal. Chem. 103 155. C Amatore, J Pinson, J-M Saveant, A Thiebault J. Electroanal. 156. C Amatore, J Pinson, J-M Saveant, A Thiebault J.Electroanal. a�Russ. J. Electrochem. (Engl. Transl.) b�Russ. Chem. Bull., Int. Ed. (Engl. Transl.) c�Russ. J. Gen. Chem. (Engl. Transl.) d�Moscow Univ. Bull. (Engl. Transl.) e�Russ. J. Org. Chem. (Engl. Transl.) f�Mendeleev Chem. J. (Engl. Transl.) 157. C P Andrieux, J M Dumas-Bouchiat, J-M Saveant 303 (1979) Chem. 107 59 (1980) Chem. 107 75 (1980) J. Electroanal. Chem. 113 1 (1980) 158. C Amatore, C Combellas, N-E Lebbar, A Thiebault, J-N Verpeaux J. Org. Chem. 60 18 (1995) 159. C Degrand, R Prest J. Org. Chem. 55 5242 (1990) 160. C Degrand Tetrahedron 46 5237 (1990) 161. D Lexa, J-M Saveant J. Am. Chem. Soc. 104 3503 (1982) 162. L I Denisovich, N A Ustynyuk, M G Peterleitner, V N Vinogradova, D N Kravtsov Izv. Akad. Nauk SSSR, Ser. Khim. 2635 (1987) b 163. T V Magdesieva, I I Kukhareva, G A Artamkina, K P Butin, I P Beletskaya J. Organomet. Chem. 468 213 (1994) 164. T V Magdesieva, I I Kukhareva, G A Artamkina, I P Beletskaya, K P Butin J. Organomet. Chem. 487 163 (1995) 165. T V Magdesieva, I I Kukhareva, E N Shaposhnikova, G A Artamkina, I P Beletskaya, K P Butin J. Organomet. Chem. 526 51 (1996) 166. T V Magdesieva, I I Kukhareva, G A Artamkina, K P Butin, I P Beletskaya Zh. Org. Khim. 30 591 (1994) e 167. I I Kukhareva, T V Magdesieva, G A Artamkina, I P Beletskaya, K P Butin Izv. Akad. Nauk, Ser. Khim. 1523 (1996) b 168. T V Magdesieva, D N Kravchuk, K P Butin Izv. Akad. Nauk, Ser. Khim. 83 (2002) b 169. J Grobe Comments Inorg. Chem. 9 149 (1990) 170. F F Said, D G Tuck Can. J. Chem. 58 1673 (1980) 171. D G Tuck, in Molecular Electrochemistry of Inorganic, Bioorganic and Organometallic Compounds (Eds A J L Pombeiro, J A McCleverty) (Dordrecht, Netherlands: Kluwer Academic, 1993) p. 15 172. J J Habeeb, D G Tuck, F H Walters J. Coord. Chem. 8 27 (1978) 173. D G Tuck Pure Appl. Chem. 51 2005 (1979) 174. A P Tomilov, I N Chernykh, Yu M Kargin Elektrokhimiya Elementoorganicheskikh Soedinenii. Elementy I ± III Grupp (Electrochemistry of Organoelement Compounds. Elements of Groups I ± III) (Moscow: Nauka, 1985) 175. A P Tomilov, Yu M Kargin, I N Chernykh Elektpokhimiya Elementoorganicheskikh Soedinenii. Elementy IV, V, VI Grupp (Electrochemistry of Organoelement Compounds. Elements of Groups IV, V, VI) (Moscow: Nauka, 1986) 176. B I Kharisov, L M Blanko, A D Garnovskii, A S Burlov, L I Kuznetsova, L V Korovina, D A Garnovskii, T Dieck Polyhedron 17 381 (1997) 177. W Lau, J C Huffman, J K Kochi Organometallics 1 155 (1982) 178. D Astruc Acc. Chem. Res. 33 287 (2000) 179. M-H Delville Inorg. Chim. Acta 291 1 (1999) 180. T V Magdesieva, V V Bashilov, D N Kravchuk, P V Petrovskii, V I Sokolov, K P Butin, in Fullerenes: Recent Advances in Chemistry and Physics of Fullerenes and Related Materials Vol. 6 (Eds K M Kadish, R S Ruoff) (Pennington, NJ: The Electro- chemical Society, 1998) p. 1322 181. T V Magdesieva, V V Bashilov, D N Kravchuk, V I Sokolov, K P Butin Elektrokhimiya 35 1125 (1999) a 182. T V Magdesieva, V V Bashilov, D N Kravchuk, V I Sokolov, K P Butin Izv. Akad. Nauk, Ser. Khim. (2002) (in the press) b 183. V V Bashilov, T V Magdesieva, D N Kravchuk, P V Petrovskii, A G Ginzburg, K P Butin, V I Sokolov J. Organomet. Chem. 599 37 (2000) 184. T V Magdesieva, I V Zhukov, L G Tomilova, E V Chernykh, K P Butin Izv. Akad. Nauk, Ser. Khim. 2149 (1997) b 185. Russ. P. 2 100 355; Byull. Izobr. (36) (1997) 186. Russ. P. 2 141 470; Byull. Izobr. (32) (1999) 187. Russ. P. 2 154 052; Byull. Izobr. (22) (2000) 188. K P Butin, T V Magdesieva Ross. Khim. Zh. 44

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (3) 239 ± 264 (2002) Modern methods for the synthesis of peptide ± oligonucleotide conjugates { E M Zubin, E A Romanova, T S Oretskaya Contents I. Introduction II. Synthesis of peptide ± oligonucleotide conjugates in solutions III. Solid-phase synthesis of peptide ± oligonucleotide conjugates IV. Conclusion Abstract. solution chemical of methods the on data published The The published data on the methods of chemical solution and solid-phase synthesis of peptide ± oligonucleotide conjugates and solid-phase synthesis of peptide ± oligonucleotide conjugates are and systematised are methods known The reviewed. are reviewed. The known methods are systematised and their their advantages approaches The considered. are disadvantages and advantages and disadvantages are considered.The approaches to to the conjugates oligonucleotide ± peptide of synthesis solution the solution synthesis of peptide ± oligonucleotide conjugates are are systematised between bonds chemical of type the to according systematised according to the type of chemical bonds between the the fragments, are synthesis solid-phase the to those whereas fragments, whereas those to the solid-phase synthesis are classified classified according of preparation the for used procedure the to according to the procedure used for the preparation of conjugates, conjugates, viz on chains peptide and oligonucleotide of elongation stepwise ., ., stepwise elongation of oligonucleotide and peptide chains on the of condensation solid-phase or support polymeric same the same polymeric support or solid-phase condensation of two two presynthesised 141 includes bibliography The fragments. presynthesised fragments.The bibliography includes 141 referen- referen- ces. I. Introduction Complexes of nucleic acids with proteins play an important role in the storage and transmission of genetic information. These supra- molecular structures include ribosomes, chromatin, viruses and complexes of nucleic acids with enzymes and regulatory proteins. It is of note that these structures maintain their stability by virtue of protein ± nucleic acid (NA) non-covalent interactions. Protein ±NA complexes the components of which are linked by covalent bonds have been known since long ago 1 ±3 and became the subject for detailed investigations aimed at elucidation of their chemical structures and functions in the cell.The most convenient tools for the study of such complexes are model peptide ± nucleotide conjugates. The first peptide ± oligonucleotide conjugates with phosphor- amide, phosphodiester, amide and ester bonds between the frag- ments were synthesised by researchers at the Department of Chemistry of M V Lomonosov Moscow State University led by M A Prokof'ev and Z A Shabarova.3, 4 In the late 1960's ± early 1970's, these scientists performed the synthesis of the first oligo- E M Zubin, T S Oretskaya Department of Chemistry,M V Lomonosov Moscow State University, Leninskie Gory, 119992 Moscow, Russian Federation.Fax (7-095) 939 31 81. Tel. (7-095) 939 31 48. E-mail: zubin@bioorg.chem.msu.ru (E M Zubin) Tel. (7-095) 939 54 11. E-mail: oretskaya@belozersky.msu.ru (T S Oretskaya) E A Romanova A N Belozersky Institute of Physicochemical Biology, M V Lomonosov Moscow State University, Leninskie Gory, 119992 Moscow, Russian Federation. Fax (7-095) 939 31 81. Tel. (7-095) 939 31 48. E-mail: romanova@belozersky.msu.ru Received 13 December 2001 Uspekhi Khimii 71 (3) 273 ± 302 (2002); translated by R L Birnova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n03ABEH000707 239 240 254 262 nucleotide peptides, viz., mixed biopolymers containing oligonu- cleotide and peptide fragments.5, 6 However, at that time these studies were not extended further, primarily because of the lack of clearcut definitions of molecular-biological tasks which might be solved through the use of peptide ± oligonucleotide conjugates.The situation changed dramatically in the mid-1980's when it was first suggested to improve the intracellular transport of antisense oligonucleotides through their covalent attachment to various peptides able to penetrate the cell membrane. Note- worthy, the interest in peptide ± oligonucleotide conjugates arose simultaneously with notable improvement of existing techniques for the formation of peptide and internucleotide bonds and the development of the solid-phase methodologies which eventually led to complete automation of oligonucleotide and peptide syn- thesis.A vast number of papers on peptide ± oligonucleotide con- jugates published in the last 10 ± 15 years brought forth the necessity to systematise the existing techniques of their synthesis. Previous studies on the synthesis of peptide ± oligonucleotide conjugates were briefly described in the reviews 7±10 published in the early 1990's and devoted to a broader field, viz., the synthesis of oligonucleotide derivatives. The present review includes two main sections describing the synthesis of peptide ± oligonucleotide conjugates in solution and on insoluble solid supports. The methods for the preparation of oligonucleotide conjugates with amino acids and proteins fell outside the scope of the review, since each of these groups of methods is a subject for independent reviews.In considering the methods for the synthesis of peptide ± oli- gonucleotide conjugates in solution, the main criterion for the systematisation of the published data was the type of the chemical bond between the two components of the conjugate molecule, while the procedure for the preparation of conjugates (stepwise elongation of oligonucleotide and peptide chains on the same polymeric support or solid-phase condensation of two presyn- thesised fragments) was used as a criterion of choice in the analysis of solid-phase methods. Here we do not touch upon the problems relevant to the structures of the intermediate and target compounds, for in the absolute majority of studies correct analysis has been carried out and the most advanced physicochemical approaches, particularly, standard procedures for the isolation of oligonucleotides and { This review is dedicated to the memory of M A Prokof'ev and Z A Shabarova, Professors of M V Lomonosov Moscow State Univer- sity and outstanding pedagogues and scientists who have laid the ground- work for the synthesis of compounds considered.240 peptide ± oligonucleotide conjugates have been employed.Nor did we consider the prospects of the application of peptide ± oligonu- cleotide conjugates in antisense biotechnology, because these problems were analysed in the reviews published by Gait et al.11 and Tung and Stein 12 in 2000. The latter publication describes the preparation of peptide ± oligonucleotide conjugates, albeit in a very concise form, without going into details about the chemical reactions involved in the synthesis of hybrid molecules.Such a limitation of the range of problems under consider- ation has made it possible to focus the main attention on the description of various approaches to the synthesis of peptide ± oligonucleotide conjugates and the formation of cova- lent bonds between the oligonucleotide and peptide fragments of the conjugates and to perform a detailed analysis of the problems involved. The principles underlying the classification and nomenclature of peptide ± oligonucleotide conjugates have not been finally established yet. There exist several types of naming for the peptide ± oligonucleotide conjugates.Some authors propose to consider these compounds as peptide derivatives containing an oligonucleotide substituent in the side chain. Based on this principle, it was proposed to term compounds comprising amino acid (peptide) and nucleoside (nucleotide) residues as `nucleo- amino acids' and `nucleopeptides'.13 In the abbreviated form, the name of the oligonucleotide is indicated in parentheses immedi- ately after the symbol designating the corresponding amino acid, e.g., H-Ala-Ser(pATAT)-Ala-OH. This approach is often used for compilation of names of conjugates where the peptide compo- nents are linked to oligonucleotides through a phosphodiester bond formed by the OH group of the hydroxy amino acid residue and the terminal phosphate of the oligonucleotide.An alternative approach to the nomenclature of this class of mixed biopolymers was suggested by Z A Shabarova.4 All com- pounds containing amino acid (peptide) and nucleotide (oligonuc- leotide) fragments were termed as peptide ± nucleotide conjugates. The name of a concrete compound is made up of two components, viz., the name of a nucleotide residue and the name of an amino acid (peptide). Between these names, the designation of the type of the linkage is placed in parentheses. Thus a compound in which the phenylalanine residue is linked to uridine 50-phosphate by a phosphoramide bond will be termed as uridilyl-(50?N)-phenyl- alanine. Such nomenclature has been extensively used to describe peptide ± nucleotide conjugates with a phosphodiester or a phos- phoramide bond between the fragments.The main difficulties arise in the compilation of names for those conjugates where the oligonucleotide and peptide compo- nents are connected through a special linker possessing a complex structure. In this case, the conjugate actually consists of three components. The authors of the overwhelming majority of publications prefer to give more general names to such com- pounds. In foreign literature, it is the term `peptide ± oligonucleo- tide conjugates' (`hybrids') that has become especially popular. II. Synthesis of peptide ± oligonucleotide conjugates in solutions 1. Conjugates with phosphodiester bonds between the fragments Studies of covalent complexes isolated from adenoviruses, polyo- viruses and certain bacteriophages 2, 3 have demonstrated that the protein component of these compounds is linked to the nucleic acid through a phosphodiester bond formed by the OH group of the hydroxy amino acid and the 50-terminal phosphate group of the polynucleotide. E M Zubin, E A Romanova, T S Oretskaya NH O O O O P X HN O7 O RO OH(H) X=7CH27,7CH2CH27,7CH(CH3)7,7CH2C6H47; R, a polynucleotide chain; B, hereinafter, a heterocyclic base.It has been found that topoisomerase I-induced transforma- tions of DNA involve the formation of a DNA± protein structure in which the protein is covalently bound to the cleavage site of a double helix chain through a tyrosine residue.14 In order to simplify the analysis of complex NA± protein structures, it seems expedient to use model compounds, e.g., peptide ± oligonucleotide conjugates of the phosphodiester type.In this context, the main attention is given to the development of efficient procedures for the synthesis of these compounds. Thus van Boom et al.15 have proposed to this end phosphorylation of the hydroxy group of a peptide by the 50-phosphate group of an the N-blocked dipeptide amide, Nps-Ala-Tyr-NH2 oligonucleotide. Partially protected dinucleoside phosphate 1 and (2) (Scheme 1), were used as the key compounds.15 An alternative HO Ura O O OTHP O P 2-ClC6H4O Ura O O AcO OTHP 1 NH2 O CH2 HN O Me HNNps NH2 O CH2 HN O Me NH2 O2N , Nps= THP= S O O B Scheme 1 2-ClC6H4O N N N N N O P O N 1) O NH2 O N CH2 HN OH, 2) O Me N MeNpsNH 2 OP O O Ura O 2-ClC6H4O OTHP O 2-ClC6H4O P O Ura O O 3 AcO OTHP OP O O Ura O OH OH O HO P O Ura O O 4 HO OH .241 Modern methods for the synthesis of peptide ± oligonucleotide conjugates jugates is associated with the high lability of the phosphodiester bond.This circumstance significantly complicates the deprotec- tion and restricts the range of the reagents used. approach to the synthesis of peptide ± oligonucleotide conjugates of the phosphodiester type is based on the reaction of an O-phosphorylated peptide and an oligonucleotide with a free 50-hydroxy group. In both cases, the investigators had to solve at least two problems, viz., to choose an optimum method for the activation of the phosphate residue and to search for adequate procedures for the protection of all functional groups of the peptide and the oligonucleotide which are not to be phosphorylated or involved in the condensation. It is of note that the removal of the protective group should be carried out under mild conditions to prevent side reactions.The analysis of published data 3, 4 prompts a conclusion that the stabilities of the phosphodiester bonds in oligonucleotidyl- (50?O) ± peptides depend on the nature of the hydroxy amino acid due to the difference in the mechanisms of cleavage of the phosphodiester bonds. If the tyrosine residue is involved, the bond is stable in moderately acidic and alkaline media, but is cleaved under strongly alkaline conditions in exactly the same way as in ordinary phosphoric acid diesters.If the phosphodiester bond involves a serine or a threonine residue, this undergoes b-elimination in alkaline media with the liberation of a phospho- monoester group, i.e., of an oligonucleotide containing 50-termi- nal phosphate. In the case of serine, the peptide fragment formed as a result of this reaction contains a dehydroalanine residue (Scheme 2). Scheme 2 O NHR1 O OH7 O P O CH O B HN O7 O R2O 2-Chlorophenyl bis(benzotriazolido) phosphate in pyridine was used for the phosphorylation and simultaneous activation of dinucleoside phosphate 1.The dipeptide 2 was added to the activated derivative formed in the presence of N-methylimidazole to give a fully protected peptide ± oligonucleotide conjugate 3 in 80% yield. The latter was isolated by column chromatography on silica gel. Of extreme importance is the fact that the condensation of the oligonucleotide and peptide components leads to the formation of a phosphotriester bond between them. In an alter- native procedure for the synthesis of compound 3, the hydroxy group of the tyrosine constituent of the peptide is phosphorylated by the bifunctional phosphorylating reagent after which dinucleo- side phosphate 1 is added to the reaction mixture. The yield of the peptide ± oligonucleotide conjugate 3 prepared by this method was 78%.NH O O 7O P O CHR1+ O B HN O7 O R2O R1=H, Me; R2, polynucleotide chain. Deprotection of the functional groups in compound 3 giving the targed product 4 was carried out using the following reaction sequence.15 First, both 2-chlorophenyl phosphate-protective groups were removed by treatment with the oximate ion. The main difficulty in this deprotection is connected with the presence of an additional aryloxy group in one of the phosphotriesters formed through the tyrosine side chain. Since in this case the pKa values of the leaving groups are close to that of the remaining aryloxy group, the possibility of selective scission of the 2-chlo- rophenyl group under the action of syn-4-nitrobenzaldoxime and N1,N1,N3,N3-tetramethylguanidine was problematic.The hydrolytic stabilities of serine and threonine derivatives are different,3 the latter being more resistant to alkaline treatment. The conjugates 6a,b NH2 R O O In order to examine the possibility of selective removal of the 2-chlorophenyl protective group, the authors performed the syn- thesis of a model compound 5. Thy O P O CH O NH2 HN 4-MeC6H4S O O O Ura O P O CH2 Me O O HN NpsNH 2-ClC6H4O O P O 2-ClC6H4O Me Thy O O O O NpsNH 5 H OMe 6a,b THPO R = H (a), Me (b) The reaction products formed upon its deprotection did not contain dipeptide 2. The structure of H-Ala-Tyr(pU)NH2 was confirmed by 31P NMRspectroscopy. Thus, it was concluded that the 2-chlorophenyl phosphate-protective group could be selec- tively removed with syn-4-nitrobenzaldoxime and N1,N1,N3,N3- tetramethylguanidine.15 were prepared 16 using a procedure similar to that described above.S-p-Tolyl bis(benzotriazolido) phosphothioate (7) was used as a phosphorylating reagent, since the p-tolylsulfenyl group (Mps) is split off under mild conditions.16 4-MeC6H4S N N N O P O N N N O The removal of the 30-O-acetyl group from compound 3 was effected by treatment with triethylamine in aqueous methanol. The 2-nitrophenylsulfenyl (Nps) and the 20-O-tetrahydropyranyl (THP) groups were cleaved at room temperature by treatment with hydrogen chloride in methanol for 3 h and with aqueous HCl (pH 2) for 14 h. 7 The yield of compound 6a was 80%.Compound 6b was prepared in a relatively low yield (50%), which can be attributed to low solubility of the dipeptide Nps-Ala-Thr-NH2 in a dioxane ± pyridine mixture. The study by the same group of Dutch investigators 16 is considered to be a logical continuation of their previous work.15 An attempt was undertaken to develop a reliable procedure for the preparation of peptide ± oligonucleotide conjugates of the phos- phodiester type with the DNA fragment linked to the peptide through a serine or a threonine residue. As a matter of fact, the problem in the synthesis of such peptide ± oligonucleotide con-242 the recourse to the protective groups commonly employed in peptide synthesis was made in the synthesis of peptide ± oligonu- cleotide conjugates.Removal of the Mps group from the completely protected peptide ± oligonucleotide conjugate 6a was performed by treat- ment with an excess of silver acetate in aqueous pyridine. The splitting of the Mps group was completed in 48 h at room temperature. However, the deprotection of the internucleotide phosphate by the oximate ion led to the cleavage of the phospho- diester bond between the components of the hybrid molecule 6a by the b-elimination mechanism (31P NMR spectroscopic data). It is of note that the removal of the phosphate-protective groups in compound 6b using the same reagents did not affect the phospho- diester bond. Tetrabutylammonium fluoride has proved to be a milder reagent for the splitting of the 2-chlorophenyl protective group.The removal of Nps and THP groups was performed with 0.01 M HCl (pH 2) (0 ± 5 8C, 16 h). Van Boom et al.15 have protected the NH2 groups of adenine and cytosine with the Nps group. The latter is relatively easily introduced, is stable under conditions of phosphorylation and increases the stability of N-glycosidic bond of purine nucleosides in acid media. However, an attempt to synthesise N2-nitrophe- nylsulfenyldeoxyguanosine failed.18 Therefore, advantage was taken of N,N-di-n-butylformamidine protection (N=CHNBun DNB) of the exocyclic NH2 group of guanine, which withstands column chromatography on silica gel and can be removed by mild treatment with hydrazine. It should be noted that it was exclu- sively amides of N-protected peptides that have been used as starting components for the synthesis of peptide ± oligonucleotide conjugates in early studies,15, 16 while later Dutch investigators used peptides with terminal carboxy groups protected as allyl (All) 17 or anthraquinon-2-ylmethyl (Maq) esters.18 The stabilities of the peptide ± oligonucleotide conjugates H-Ala-Ser(pTT)-NH2 and H-Ala-Thr(pTT)-NH2 in alkaline media have been studied.16 Studies by 31P NMR spectroscopy have demonstrated that H-Ala-Ser(pTT)-NH2 was completely hydrolysed by 0.15 M NaOH (20 8C, 25 min) to form the dinu- cleotide pTT, whereas no complete hydrolysis of H-AlaThr(pTT)- NH2 occurred after 9 h treatment with alkali under identical conditions.In further studies, the same group of investigators used this approach to obtain conjugates with a heterooligonucleotide component.17, 18 The problem arising in such a synthesis relates to the necessity of protection of exocyclic amino groups of the heterocyclic bases.The authors had to abandon the use of acyl protective groups employed in routine oligonucleotide syntheses, because their removal by treatment with concentrated ammonia leads to the cleavage of the phosphodiester bond between the oligonucleotide and the peptide formed with the involvement of the hydroxy group of serine or threonine. This could result in the base-catalysed racemisation of amino acid residues. Therefore, The approaches to the synthesis of the peptide ± oligo- nucleotide conjugates H-Phe-Tyr(pATAT)-NH2, H-Phe- Tyr(pGC)-NH2 and H-Phe-Ser(pGC)-Ala-OH (see Refs 17, 18) are, on the whole, similar to those proposed in earlier studies,15, 16 but the procedures used for the deprotection of the intermediate fully protected peptide ± oligonucleotide conjugates 8 ± 10 have been modified: Nps-protection was cleaved with 2-mercapto- pyridine, while 30-levulinoyl (Lev) and DNB protections were removed by hydrazine hydrate.O CH2Ph CH2Ph NH2 NH NH NpsNH NpsNH O O CH2 O CH2 P O 2-ClC6H4O AdeNps O O OP 2-ClC6H4O O O O P 2-ClC6H4O Thy O O 2-ClC6H4O O O 2-ClC6H4O P AdeNps O O O O P 2-ClC6H4O Thy O O O 8 LevO . Lev= Me O E M Zubin, E A Romanova, T S Oretskaya 2 , O CH2 (Maq) O O Me O CH2Ph OMaq NH2 NH NpsNH HN O O CH2 O O P 2-ClC6H4O GuaDNB O O O GuaDNB O O O P 2-ClC6H4O CytNps O O O O P CytNps O O LevO 10 LevO 9Modern methods for the synthesis of peptide ± oligonucleotide conjugates The phosphotriester method has initially been used to prepare H-Ala-Ser(pATAT)-Ala-OAll.17 The yield of the fully blocked peptide ± oligonucleotide conjugate was 64%.The removal of protective groups was carried out by the same method as that used for compound 8. An alternative approach consisting of phosphorylation of the tripeptide Nps-Ala-Ser-Ala-OAll by 2-chlorophenyl bis(benzotriazolido) phosphate and subsequent reaction of the activated tripeptide derivative with the partially protected tetramer ATAT was inefficient. The phosphorylation of the hydroxy group of serine led to the formation of the undesired symmetrical product.These difficulties were successfully overcome through the use of the phosphoramidite method.17 Me O Me HN OAll AllOP(NPri2)2 (12) NpsNH HN O O CH2 11 OH Me O Me OAll HN NH NpsNH O O CH2 OP NPri 13 AllO 2 The tripeptide 11 was first treated with equimolar amounts of 1H-tetrazole and the phosphitylating reagent 12 in acetonitrile after which equimolar amounts of the tetramer 14 and 1H- tetrazole were added to the phosphite 13 formed. The duration of the reaction was 18 h. The phosphite unit in the resulting conjugate was oxidised to the phosphate by treatment with tert- butyl hydroperoxide.HO AdeNps O O O 2-ClC6H4O P Thy O O O 1) 1H-tetrazole 2) ButOOH O P 2-ClC6H4O 13+ AdeNps O O O O 2-ClC6H4O P Thy O O 14 LevO Me O Me OAll HN NH NpsNH O O CH2 O O AllO P AdeNps O O OP O 2-ClC6H4O Thy O O O O 2-ClC6H4O PO O OP O 2-ClC6H4O O 15 LevO The notable advantage of this method is the use of the allyl protective group in the phosphotriester bridge between the tripeptide and the oligonucleotide in the completely protected peptide ± oligonucleotide conjugate 15, which is easily removed with aqueous pyridine. The only disadvantage of the phospho- triester and phosphite versions of H-Ala-Ser(pATAT)-Ala-OAll synthesis is that the carboxy group of the C-terminal amino acid could not be recovered by transition metal complexes.Of special interest are the studies by Japanese investiga- tors 19, 20 who have developed an original procedure for the preparation of peptide ± oligonucleotide conjugates of the phos- phodiester type, viz., H-Ala-Tyr(pUU)-OH (16) and H-Ala- Ser(pTT)-Phe-OH (17) based on the chemistry of phospho- thioates. Synthesis of the phosphorylated dipeptide component 18 was carried out using the Boc-strategy (Scheme 3).19 In the first step, the partially protected tyrosine derivative 19 was phosphorylated by S,S-diphenyl phosphodithioate (PSS) in the presence of isoduroldisulfonyl chloride (DDS). N-protected alanine was added to the phosphorylated tyrosine derivative 20 after deprotection of the amino group.N,N0-Dicyclohexylcarbo- diimide and 1-hydroxybenzotriazole were used as the condensa- tion reagent in the formation of the peptide bond. Selective splitting of one phenylthio group in the derivative 21 was achieved by treatment with triethylammonium hypophosphate in pyri- dine.19 The phosphorylated tripeptide 22 was prepared in a similar way.20 In this case, the hydroxy group of serine was phosphor- ylated by S,S-bis(p-methoxyphenyl) phosphodithioate (MPSS). 243 AdeNps Thy O244 O O H BocN H BocN OPhac CH2 CH2 a O OH O P 19 SPh 20 O Me HNOPhac TcBocNH O CH2 d O SPh P O 21 SPh O O C Ph, TcBoc= C OBut , Phac= Boc= CH2Me ClO2S 7 SPh (PSS), (a) O Me OPSPh Me SO2Cl (b) 4 M HCl ± dioxane; (c) TcBoc-Ala-OH, cyclo-C6H11N=C=NC6H11-cyclo (DCC), + (d ) NEt3H H2PO¡2 , C5H5N.UraTcBoc HO OOTHP O O 18+ PhS P UraTcBoc O OOTHP TcEocO 25 . TcEoc = CCl3CH2OCO Scheme 3 OPhac b, c SPh O Me HNOPhac TcBocNH O CH2 O SPh P O + 18 O7 NEt3H Me OC O C CCl3 ; Me Me (DDS), 1H-tetrazole; NN (HOBT); NOH O Me NH OPhac TcBocNH O CH2 O O PhS P UraTcBoc O OOTHP O O PhS PO O 27 TcEocO E M Zubin, E A Romanova, T S Oretskaya However, treatment of compound 22 with a solution of triethylammonium hypophosphate in pyridine yielded MPSS rather than compound 23.20 Presumably, this was due to b-elim- ination of phosphodithioate.Therefore, selective removal of one p-methoxyphenylthio group was achieved by milder treatment with bis(tributyltin) oxide which afforded compound 24. O Me CH2PhOPhac HN TcBocNH NH O O CH2 O O P SC6H4OMe-p 22 ± 24 X + X=SC6H4OMe-p (22), O7 HNEt3 (23), OSnBun3 (24). The syntheses of partly protected dinucleotide components 25 and 26 (Schemes 4 and 5) made use of PSS and DDS as bifunc- tional condensation reagents. The uracil and thymine N3-TcBoc protective group can be removed by reduction with zinc in acetylacetone under neutral conditions. In the synthesis of the target peptide ± oligonucleotide con- jugate 16, compounds 18 and 25 were introduced into the reaction with DDS and 3-nitro-1,2,4-triazole (Scheme 4).The fully pro- tected peptide ± oligonucleotide conjugate 27 was obtained in 73% yield.19 The use of the same reagents in the synthesis of the pepti- de ± oligonucleotide conjugate 17 (Scheme 5) results in the for- mation of both the target compound 17 and the tripeptide containing a dehydroalanine residue. Therefore, the reaction was carried out in the presence of N-methylimidazole and an excess of triisopropylbenzenesulfonyl chloride (TPSCl).20 Successive deprotection of compounds 27 and 28 afforded the conjugates 16 and 17, respectively. The phosphate groups were deprotected with an excess of (Bun3 Sn)2O. The TcBoc, Phac and TcEoc groups were cleaved by reduction with zinc in acetylacetone. The 2 0-O-tetra- hydropyranyl groups were removed from compound 27 using a standard procedure (0.01 M HCl, pH 2).Scheme 4 O Me NH OH H2N O CH2 O O HO P Ura O OOH O O HO P UraTcBoc Ura O O 16 OH HO OTHPModern methods for the synthesis of peptide ± oligonucleotide conjugates ThyTcBoc HO O O O 24+PhS P ThyTcBoc O O TcBocO 26 2. Conjugates with phosphoramide bonds between the fragments From the historical point of view, mixed biopolymers with the phosphoramide bonds between the oligonucleotide and the pep- tide fragments were the first to be synthesised. The possibility of modification of the internucleotide phosphate group was demon- strated by Prokof'ev, Shabarova et al.5, 21 in the example of dinucleoside phosphates (Scheme 6).B AcO O O a O OH PO O B 29 AcOO HN HO CH2Ph O (a) (PhO)2P(O)Cl; (b) H-Ala-Phe-OH. O 7 oligonucleotide O P O OH O peptide X C X=OMe, NH2 . O Me HNTcBocNH O CH2 O O P p-MeOC6H4S O O PhS PO 28 TcBocO Scheme 6 B AcO O O O b O O P (PhO)2P O O B 30 AcO B AcO O Me O O P NH O O B 31 AcOPh3P/Py2S2 OH MeN N Me N O oligonucleotide O HN P OH O7 245 Scheme 5 O Me CH2Ph CH2Ph OPhac OH HN H2N NH HN O O O CH2 O O HO P ThyTcBoc Thy O O O O O O HO P ThyTcBoc Thy O O O 17 HO Treatment of dinucleoside phosphate 29 containing protected hydroxy groups with diphenylphosphochloridate yielded the corresponding non-symmetrical pyrophosphate 30.The latter represented an active phosphorylating reagent. Reaction of pyro- phosphate 30 with the dipeptide H-Ala-Phe-OH is accompanied by nucleophilic substitution at the internucleotide phosphorus atom; the role of the leaving group is played by the diphenyl phosphate anion. The advantages of this method are as follows: (i ) the synthesis of pyrophosphate does not affect heterocyclic base residues; (ii ) pyrophosphate can be introduced into a reaction with the nucleophile without preliminary isolation. The yield of the target product 31 was 75%. However, this method 5 gave satisfactory results only in the case of deoxyribo derivatives; in the presence of a protected 2 0-hydroxy group, the degree of conversion was only 10%. An alternative approach to the synthesis of peptide ± oligonu- cleotide conjugates of the phosphoramide type was developed by a group of Novosibirsk investigators.22 ± 25 The a-amino group of the peptide reacts with preactivated 3 0- or 50-terminal phosphate of the oligonucleotide.The advantage of this approach is the possibility to use non-protected fragments of nucleic acids. This, in turn, is due to the fact that the phosphate residue is selectively activated with a mixture of triphenylphosphine and 2,2 0-dipyridyl disulfide in the presence of the nucleophilic catalyst N-methylimi- dazole (Scheme 7). Under these conditions, the internucleotide phosphate groups and the reactive centres of heterocyclic bases are not involved. The mechanism of this activation of phosphoric acid monoesters is considered in detail in the review by Zary- tova.26 In the study,22, 23 the 5 0-terminal phosphate group was acti- vated in a mixture of dimethylformamide and dimethyl sulfoxide.The oligonucleotides were solubilised in organic solvents as cetyltrimethylammonium (cetavlon) salts. The activation of the phosphate residues of the oligonucleotide gives rise to N-methyl- imidazolide which reacts with the amines with pKa values in a very Scheme 7 O O peptide C X H2N + oligonucleotide OH O O N PO7246 broad range; with an increase in pKa, the rate of the phosphor- amide bond formation increases. The guanidine group of arginine in the peptides H-Pro-Arg-Val-OMe and H-(Leu-Arg)n-Gly-NH2 (n=2 ± 4) is basic.In order to prevent the reaction of the latter with the activated phosphate group of the oligonucleotide, the peptides were introduced into the reaction in the form of trifluoro- acetate salts. Under these conditions, the guanidine group remains protonated throughout the reaction and the role of the main nucleophilic centre after selective deprotonation by triethylamine is played by the a-amino group of the peptide. The presence of a P7N bond in the peptide ± oligonucleotide conjugates formed was confirmed by 31P NMR spectroscopy and acid hydrolysis under conditions of cleavage of the phosphora- mide bond (0.1 MHCl, 18 h). The target peptide ± oligonucleotide conjugates were isolated by reversed-phase HPLC in 70%± 85% yields.It was noted 23 that the addition of the peptide to the oligonucleotide increases the hydrophobicity of the latter. In the reversed-phase HPLC, the retention time of the reaction product increases in comparison with that of the original oligonucleotide. The method described above was used for the conjugation of a synthetic analogue of the peptide antibiotic netropsin with oligo- thymidylic and oligodeoxyadenylic acids.24 In this case, the peptide was attached to the 3 0-end of the oligonucleotide. The yields of the conjugates 32 were 30%± 50%. HN R O 4 NH O HN O 4 Me N O n SN N 32 n=2 ± 5; R, fragment of the oligonucleotide chain. A somewhat different approach was proposed for the syn- thesis of covalent complexes of non-modified oligonucleotides and their thiophosphate analogues with netropsin (Nt) or yet another peptide antibiotic, viz., distamycin (Dst).25 +NH2 O Me N C O NH2 NH HN R NH Me N NH H NH2 + (distamycin, Dst).R=O (netropsin, Nt), H2N HN O NMe O Automated phosphoramidite oligodeoxyribonucleotide syn- thesis was used to prepare oligonucleotides bearing, at their 3 0- or 5 0-ends, hexaethylene glycol linkers with phosphate groups. This was followed by the activation of the terminal phosphate groups by redox reagents of the types Ph3P and Py2S2 in the presence of 4-dimethylaminopyridine (DMAP), after which a peptide anti- biotic (Nt or Dst) was added to the activated oligonucleotide together with an excess of DMAP.The latter is thought 25 to be necessary for neutralisation of the positive charges of the amidine and guanidine groups. E M Zubin, E A Romanova, T S Oretskaya O O 1) Ph3P/Py2S2, DMAP 2) Nt (Dst), DMAP P O7 RO P O(CH2CH2O)6 O7 O7 O O P Nt(Dst) RO P O(CH2CH2O)6 O7 O7 R, fragment of the oligonucleotide chain. OH7 R2 7R2NH2 Presumably,25 a covalent P7N bond formed between the antibiotic and the NA fragment involves the amidine group of the antibiotic. Indirect evidence can be derived from the fact that the rates of hydrolysis of oligonucleotide netropsin and acetamidine derivatives under the action of concentrated aqueous ammonia are identical. Apparently, the hydrolysis is accompanied by deamination: O R1O PO7 NH C CH2 +NH2 O NH C Me R1O P O O7 R1, fragment of the oligonucleotide chain, R2, residue of the antibiotic.A group of investigators representing the Moscow research team led by Z A Shabarova 27 have developed a method for the carbodiimide-induced activation of terminal phosphate groups of oligonucleotides by N-hydroxybenzotriazole (HOBT). HOBT esters are rather stable in aqueous media and can effectively phosphorylate nucleophilic reagents with the formation of cova- lent bonds (Scheme 8). Amino acids and peptides can be used as nucleophilic reagents. Scheme 8 O7 30 50 N oligonucleotide O O P OH N N 7HOBT O peptide H2N O oligonucleotide peptide HNP O OH O7 This was applied to the synthesis of the Gly-Gly-(5 0?N)- (pTCTAG) conjugate 28 and to the synthesis of oligodeoxyribo- nucleotides with the antibiotic polymixin B1, which represents a polycationic peptide attached to the 5 0-terminal phosphate group.29 The yields of the product reached 80%.An undecanu- cleotide conjugate has been designed and successfully prepared with 3 0-terminal phosphate attached to the antibiotic daunomycin and the 5 0-terminal phosphate attached to polymixin B1, the latter serving to facilitate the penetration of the oligonucleotide through the cell membrane. This conjugate was isolated by polyacrylamide gel electrophoresis (PAGE); its total yield was 20%. The extension of these methods 27 ± 29 to RNA derivatives was a logical continuation of the studies.30 Conditions for effective synthesis ofHOBTesters of 5 0-oligoribonucleotides were selected.The formation of a specific complex of this duplex with the peptide Tat was demonstrated in the example of a syntheticRNAcomplex possessing a binding site for the protein Tat HIV-1. The covalent coupling of the activated RNA duplex to the peptide proceeded with 20%± 30% efficiency and did not require any additional activation. Thus, the peptide Tat ±TAR RNA complex is an example of effective covalent coupling of a protein toRNA within a specific complex.Modern methods for the synthesis of peptide ± oligonucleotide conjugates Synthesis of a conjugate of a triplex-forming oligonucleotide with a functionally important peptide has recently been described. Such a conjugate is a convenient tool for the study of modulation of gene expression.31, 32 3.Conjugates with amide bonds between the fragments The studies 33 ± 40 are devoted to the synthesis of peptide ± oligo- nucleotide conjugates with the amide bond between the fragments formed as a result of the reaction of the activated a-carboxy group of the peptide with the aliphatic NH2 group incorporated into the oligonucleotide. The aliphatic amino group is usually attached to the 5 0- or the 3 0-end of the oligonucleotide through a polymethylene spacer in the course of automated phosphoramidite synthesis. Commer- cially available 6-aminohexyl 2-cyanoethyl N,N-diisopropylphos- phoramidite, the amino group of which is protected by the 4-monomethoxytrityl (MMTr) group, a convenient reagent for the preparation of DNA fragments with 5 0-terminal NH2 groups.O(CH2)2CN H MMTrN(CH2)6OP NPri2 The synthesis of oligonucleotides containing a 3 0-aminoalkyl group is carried out with the use of a special UV-cleavable polymeric support with a photolabile anchor containing a termi- nal aminohexanol group (Scheme 9). The attachment of the first nucleotide is performed in an automated DNA synthesiser, whereas the splitting of the oligonucleotide with the 3 0-terminal amino group incorporated is effected by UV irradiation.41 The main advantage of this method is that the presence of a photo- labile anchor group facilitates the synthesis of protected oligo- nucleotides with only the aliphatic amino group free.Such a partially protected oligonucleotide was introduced into a condensation with one of the two N-blocked tripeptides, viz., Fmoc-Gly-Gly-Gly-OH (Fmoc is 9-fluorenylmethoxycarbonyl), or Fmoc-Gly-Gly-His-OH (see Scheme 9). The activation of the carboxy groups in the tripeptides was carried out by a Ph3P±Py2S2 system in the presence of DMAP. The reaction was carried out in dimethylformamide at room temperature because of the possibility of cleavage of the Fmoc group and concomitant oligomerisation of the tripeptide under more drastic conditions. The yield of the peptide ± oligonucleotide conjugates after depro- tection and isolation by ion-exchange HPLC or PAAG electro- phoresis was 89% ± 99%. The condensation of short peptides and oligonucleotides using redox reagents was shown 42 to be accompanied by epimerisation of the C-terminal amino acid residue.Model experiments showed O DMTrO NH 6 1) Removal of the DMTr group 2) Oligonucleotide synthesis 3) hn O NO2 DMTrO MeO CPG HN O 2 O O DMTrO NH O P O (CH2)6 protected oligonucleotide O(CH2)2CN O oligonucleotide NH O P O HO (CH2)6 O7 GPG, controlled pore glass; DMTr, 4,4 0-dimethoxytrityl protective group; PyBOP, (benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate. 247 that the degree of epimerisation reached 30% under these con- ditions (1H NMR spectroscopic data). An attempt has been made 34 to prepare peptide ± oligonucleo- tide conjugates with long peptide chains.In order to activate the terminal carboxy group of the peptide, the authors used (benzo- triazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate (PyBOP).34 However, mass spectrometric analysis performed after deprotection of the peptide ± oligonucleotide 33 formed and isolation of the target product revealed that its peptide component contains only two amino acid residues (compound 34). Fmoc-Asn-Phe-Leu-Ala-Gly-Val-Gly-Ser(Ac)-Gln O NH 1) NH3 . aq 2) 80% AcOH DMTrO O P O (CH2)6 protected oligonucleotide 33 O(CH2)2CN O oligonucleotide NH Gln-Ser-OH HO O P O (CH2)6 34 O7 It is suggested 34 that the cleavage of the peptide component at the amide bond between the serine and glycine residues was due to N?O acyl migration.Apparently, the free OH group of serine liberated upon treatment with ammonia attacks the amide bond between the serine and glycine residues, which is followed by the scission of the peptide chain due to hydrolysis of the ester bonds. The use of milder procedures for the removal of protective groups (0.4 M NaOH in methanol, 17 h) did not give any positive results either. This problem could be overcome by replacing the glycine residue by the alanine residue in order to create steric hindrances in the site of possible nucleophilic attack. Some other changes necessary for further formation of a stable helical complex between the peptide ± oligonucleotide conjugate and the comple- mentary NA target were also effected, viz., the a-amino group of the peptide was protected by the acetyl group and an additional tetraethylene glycol bridge was introduced between the 30-termi- nal nucleotide and the alkylamine in order to separate the peptide fragment from the DNA chain.In this case, the yield of the conjugate 35 was 72%. Ac-Asn-Phe-Leu-Ala-Gly-Val-Ala-Ser-Gln NH O O oligonucleotide HO O O P (OCH2CH2)4 P O (CH2)6 35 O7 O7 Scheme 9 O N-protected peptide, Ph3P/Py2S2 , DMAP or PyBOP O P O (CH2)6NH2 protected oligonucleotide O(CH2)2CN 1) NH3 . aq 2) 80% AcOH N-protected peptide peptide248 The peptide ± oligonucleotide 3 0,5 0-bisconjugates 36 were successfully obtained using the polymeric support mentioned above in the first step.The peptide ± oligonucleotide 3 0-conjugates were synthesised according to Scheme 9 after which the 5 0-ter- minalHOgroup was made to react with commercially available 6- aminohexyl 2-cyanoethyl N,N0-diisopropylphosphoramidite to which another peptide was then attached.35 O oligonucleotide O P R1C NH(CH2)6O O(CH2)6NH CR2 36 O O O7 R1,R2, peptide residues. The use of protected oligonucleotides with 3 0-terminal car- boxy groups for the synthesis of peptide ± oligonucleotide conjugates was a natural continuation of these studies.36 A UV- cleavable polymeric support was also employed. O DMTrO 3 O NO2 MeO CPG O HN 2 O The covalent bond between the 3 0-end of the oligonucleotide and the N-end of the peptide is easily formed even in the case of spatial hindrance created by the N-terminal amino acid.Excellent results were obtained in the coupling of tetrapeptides; the con- jugates were obtained in 89%± 95% yields. Special mention should also be made of a synthesis of a series of peptide conjugates by coupling them to non-protected oligo- nucleotides T6 and T10 containing 5 0-terminal NH2 groups using PyBOP as the condensation reagent.37 Synthesis of covalent oligonucleotide complexes with a syn- thetic analogue of the peptide antibiotic CC-1065 has been described.38 O O O R1O P 5 NH O7 N HN R2 1± 3 R1, oligonucleotide chain, R2 =CONH2, Boc. This method includes condensation of 2,3,5,6-tetrafluoro- phenyl ester of a peptide and a deprotected oligonucleotide bearing a 5 0- or 30-terminal amino group.The reaction is carried out in dimethyl sulfoxide. No detailed description of the synthesis of theDNAfragments containing a 5 0- or 3 0-terminal aminohexyl residue is provided. The oligonucleotides were introduced into the reaction as cetavlon salts. The yields of the condensation products were 60%± 90% (HPLC data). It is noted that precipitation of the reaction products by the addition of a 2% solution of LiClO4 in acetone and their further isolation by reversed-phase HPLC decreases the total yield to 15%± 50%. An interesting approach to the synthesis of peptide ± oligonu- cleotide conjugates containing an amide bond was proposed by Swiss investigators.39 Their method is based on the Wolff rear- rangement.Diazo ketone 37 was used as a peptide component, and an oligothymidylate, as an oligonucleotide. The reaction was carried out in dimethylformamide. The diazo ketone 37 is decom- posed upon UV irradiation with abstraction of the nitrogen molecule to give carbene 38 which undergoes the Wolff rearrange- ment to be further converted into ketene 39 (Scheme 10). The latter reacts with oligothymidylate 40 containing a 5 0-terminal amino group. The yield of the peptide ± oligonucleotide conjugate E M Zubin, E A Romanova, T S Oretskaya Scheme 10 Me O HN hn N +N CH 7 7N2 ZNH 37 Me O Me O Me C+ O CH 7 HN HN CH ZNH ZNH Me O Me O 38 Me C O CH NH ZNH 39 Me O O oligonucleotide OH 39+ O P O H2N(CH2)6 40 O7 Me O HN HN oligonucleotide OH (CH2)6O P O ZNH O Me O O7 41 Z=PhCH2OCO.41 is 35%. The attempts to introduce diazo ketone 37 into the reaction with heteropolymeric DNAmolecules were unsuccessful. The authors attribute this phenomenon to the poor solubilities of the heteropolymers in dimethylformamide. The presence in the reaction mixture of water molecules competing with oligonucleo- tides for the reaction with ketenes is an alternative explanation for the low yields. The approaches discussed above to the synthesis of oligonu- cleotide peptides largely entail the use of reactions occurring in organic solvents. An original procedure for the formation of covalent bonds between peptides and oligonucleotides in aqueous solutions based on the chemical ligation method has been devel- oped.40 This consists in approaching the reacting groups in the terminal fragments of the oligonucleotides I and II as a result of complementary interactions of these oligonucleotides with the complementary adjacent sites of the oligonucleotide III (Scheme 11).Scheme 11 30 50 Thy O O II NH2 III OC Ura peptide S O I 50 OMe O 30 DNA duplex 50 7 Ura Thy O S O O + peptide NH OMe O C 30 OModern methods for the synthesis of peptide ± oligonucleotide conjugates The peptides used in the study 40 were prepared in the form of carboxythionic acids by solid-phase synthesis on a polymeric support. Treatment of these peptides with Ellman's reagent [5,5 0-dithiobis(2-nitrobenzoic acid)] gave the carboxythionic acid S-ester.The latter reacted with an oligonucleotide containing a 2 0-O-methyl-5 0-deoxy-5 0-thiouridine to give the oligonucleotide peptide 42 with the thioester bond (Scheme 12). The reaction was carried out for 30 min at 25 8C in a buffer solution (pH 8.0) in the presence of spermidine. The product formed 42 is stable at room temperature for 24 h, but is easily hydrolysed at 37 8C over the same period.40 Scheme 12 Ura HS O O O R2O OMe R1SSR1 peptide peptide C SH C SR1 O peptide C Ura S O 42 R2O OMe NO2 O2N COOH, R1SSR1= HOOC S S R2, oligonucleotide chain. The condensation of the peptide linked to a single-stranded fragment of the oligonucleotide I with the oligonucleotide II containing 3 0-terminal 3 0-amino-3 0-deoxythymidine within the DNA duplex is the key step in the synthesis of peptide ± oligo- nucleotide conjugates possessing amide bonds (see Scheme 11).A DNA duplex represents a highly ordered system; in its cleavage site, the terminal 3 0-amino group of one oligonucleotide and the thioester group of the other peptide are located in close proximity to one another and are spatially oriented in a definite manner owing to the interactions stabilising the double helix. As a consequence, the chemical reaction is characterised by high selectivity and high rate, which are unattainable for this type of conversion occurring outside the helical complex. Under optimum conditions (2 ± 4 mmol of spermidine, pH 8, 37 8C), the conden- sation was completed in 15 ± 30 h; the peptide ± oligonucleotide conjugates were obtained in good yields (60% ± 85%).Owing to the weakly alkaline medium, the 3 0-amino group remained non- protonated, while the presence of spermidine in the reaction mixture favoured additional stabilisation of the DNA duplex. This approach allows simultaneous synthesis of several pep- tide ± oligonucleotide conjugates possessing different structures. This is possible when an oligonucleotide with a 3 0-amino group is involved in the simultaneous formation of two different DNA duplexes as has been demonstrated in the study under discussion. Peptides of various lengths underwent condensation with oligonucleotides.However, reactions with short peptides encoun- tered serious difficulties because of possible formation of cyclic structures. This is due to the fact that the a-amino group of the peptide, along with the 3 0-amino group of theDNAfragment, can attack the carbonyl carbon atom resulting in the formation of a cyclic peptide. Moreover, yet another prerequisite for efficient condensation is that the C-terminal amino acid residue of the peptide did not contain a branched side chain, i.e., its C-end should be devoid of isoleucine or threonine residues. The use of a non-modified oligonucleotide with a 3 0-terminal hydroxy group in the condensation is also noteworthy.40 Here, the formation of an ester bond between the peptide and the DNA fragment occurs; the yields of peptide ± oligonucleotide conju- gates with ester bonds are lower than those with amide bonds.In 249 contrast to the amide bond, the ester bond is easily hydrolysed in alkaline media (0.01 M NaOH) at 25 8C for 1 h. A more recent report 43 describes selective non-matrix chem- ical ligation of modified RNAcontaining the 5 0-S-thioester group and polypeptides with N-terminal cysteine residues which affords peptide ± 5 0-RNA conjugates (Scheme 13). Scheme 13 O O O NHá 7 3 peptide RNA 30 + OOC S P O RS O7 SH O 7 O peptide NH2 O OOC RNA S P O 30 S O7 O O O O S P RNA 30 NH 7 O7 peptide OOC SH This reaction included transthioesterification and subsequent spontaneous intramolecular rearrangement resulting in the for- mation of a peptide bond.The authors observed high selectivity of ligation (50% yield of the conjugate of 102-membered RNA with a 13-membered peptide) at low (micromolar) concentrations of the reactants. This reaction proceeds under non-denaturing con- ditions favouring the preparation ofNA± protein conjugates with simultaneous preservation of the native structure of the protein. The synthesis of peptides carrying N-terminal cysteine residues does not present a problem. The 5 0-terminal phosphothioate residue can be easily introduced into a ribo- or deoxyribonucleic acid either chemically or enzymatically. The term `native ligation' (by analogy with condensation of two unprotected peptide fragments, one of which contains a C-terminal S-carboxythioate group and the other one contains an N-terminal cysteine, which is routinely used in peptide chem- istry 44) was used 45, 46 to describe a new procedure for the preparation of peptide ± oligonucleotide conjugates 43 (in the absence of a template).HS O O P O O R1 R2 HN OH NH 43 O O R1, fragment of an oligonucleotide chain, R2, fragment of a peptide chain. This method is based on the reaction of the peptide 44 modified by the introduction of the S-thioester group in its N- end with an oligonucleotide containing a 5 0-terminal cysteine residue which is introduced using compound 45. O R3 O R1 BnS HNNH2 HN NH O O R2 O n 44 R1,R2,R3, side chains of amino acids.ButS OCH2CH2CN O P HN FmocHN NPri2 45 O250 S-Benzyl thiosuccinate 44 was obtained in the last step of a standard automated peptide synthesis (Fmoc strategy). The O-trans-4-(Na-Fmoc-S-tert-butyl-L-cysteinyl)aminocyclo- hexanol phosphoramidite 45 was attached to the 5 0-end of the oligonucleotide in the course of a standard solid-phase phosphor- amidite synthesis. The preparation of the phosphoramidite syn- thon 46 based on 4-hydroxypiperidine has also been described.46 ButS CH2CH2CN N FmocHN O P O NPri2 46 The peptide 44 and the oligonucleotide modified by com- pound 45 (or 46) after deprotection and detachment from the resin were condensed without additional purification in water ± organic solvents to which tris(2-carboxyethyl)phosphine (in order to remove the S-tert-butyl group in situ) and thiophenol (in order to increase the efficiency of thiol exchange) were added.Some reactions between peptides and oligonucleotides of various sequences and lengths have been described and the ways to increase the efficiency of their synthesis have been considered.45 The peptide ± oligonucleotide conjugate 43 was the main product in all cases; sometimes, its yield reached 75%. The side reaction, i.e., intramolecular cyclisation involving the succinyl fragment of the peptide, took place only with glycine at the N-end. No doubt, further experiments are necessary in order to optimise the compo- sition of the solvent, reaction conditions (particularly, for long- chain peptides with a secondary structure).However, even now it is quite clear that the method in question 45 can be used for the preparation of oligonucleotide conjugates with a very broad range of peptides with small restrictions for amino acid sequences. This approach has obvious advantages over the method considered above for the synthesis of peptide ±RNA conjugates,43 since it allows the use of peptides with thioesters at their C- and N-ends and a 5 0-cysteinyl-substituted oligonucleotide prepared by a standard phosphoramidite procedure. The method 43 allows the use of exclusively N-terminal cysteine-containing peptides and RNA modified with 5 0-phosphothioate. 4. Conjugates with sulfide bonds between the fragments The optimum variant of peptide ± oligonucleotide synthesis seems to be the one in which the reactive group of the oligonucleotide reacts selectively with the reactive group of the peptide, whereas other functional groups of both macromolecules are not involved.For example, the conjugation can be effected by the addition of thiols to maleimides. O O S R1 R1 SH+ N R2 N R2 O O R1,R2, an oligonucleotide or a peptide chain fragments. The sulfhydryl or the maleimide active groups are first introduced into the oligonucleotide or the peptide using various reagents. It should be noted, however, that the role of a reaction centre can be played by the HS group of the cysteine residue initially present in the peptide. It is preferable, as a rule, that the sulfhydryl group is localised in the oligonucleotide, whereas the maleimide group, in the peptide.47 ± 58 An alternative variant 59 ± 63 is that the HS group of the N- or C-terminal cysteine residue reacts with the maleimide group incorporated into the oligonucleotide chain.In both cases, the reaction is carried out in aqueous or water ± organic media (pH 6.0 ± 7.2) at room temperature. The peptide is almost always taken in an excess (from 4 to 15 equiv.); in only one study, the ratio between the two components of the reaction mixture was 1 : 1.48 The reaction time varies from 1 to E M Zubin, E A Romanova, T S Oretskaya 20 h. The yields of the conjugates vary within a sufficiently broad range, viz., from 30% to 90%. The formation of a thioether bond between the oligonucleo- tide and the peptide can also take place as a result of a nucleophilic attack of the HS group at the carbon atom of the halogenoacetyl derivative.For this purpose, the halogenoacetic acid residue is incorporated into oligonucleotides or peptides.64 ± 72 O O R2 SH Hal Hal R1 NH2+ X HN R1 O R2 S HN R1 O O ; , X= N Hal O O R1,R2, an oligonucleotide or a peptide chain fragments. As a rule, the oligonucleotide containing an iodoacetyl group reacts with an excess of the cysteine-containing peptide in neutral media at room temperature.64 ± 67 The reaction time is 12 ± 20 h. The yields of the conjugates usually vary from 55% to 98%. Reed et al.68 performed the synthesis in a weakly alkaline medium (pH 8.3).The authors cite interesting results concerning the dependence of the reaction rate on the net charge of the peptide. Thus the reaction of a modified oligonucleotide with a peptide containing one arginine residue and four lysine residues (the net charge is +5) is complete within a few minutes. In the case of the other two peptides with the net charges of +1 and zero, the reaction time is 3 and 20 h, respectively. The authors presume that the increase in the condensation rate is caused by electrostatic interactions of the negatively charged phosphates in the DNA molecule with the positively charged side groups of lysine and arginine of peptides. The synthesis of peptide ± oligonucleotide conjugates in which two DNA fragments are linked through a peptide bridge has been described (Scheme 14) 64 with H-Cys(SBut)-(Arg)n-Cys-NH2 (n=3, 5, 7) as a bridge.Scheme 14 O I H-Cys(SBut)-(Arg)n-Cys-NH2+ R1 NH O 1) Bu3P 2) R2NHC(O)CH2I S R1 NH H-Cys(SBut)-(Arg)n-Cys-NH2 O O S S NH HN R1 R2 HCys-(Arg)n-CysNH2 R1,R2, fragments of oligonucleotide chains; n=3, 5, 7. Initially, the free HS-group of the C-terminal cysteine residue reacted with the iodoacetyl group localised at the 5 0-end of one oligonucleotide. In order to prevent the removal of the S-tert- butyl group, it was necessary to carry out the reaction in a neutral medium. After termination and chromatographic isolation of the peptide ± oligonucleotide conjugate, the second thiol group was deprotected by treatment with a large excess of tributylphosphine at room temperature for 4 h.Taking into consideration low solubility of tributylphosphine in water, the removal of the protective group was carried out in a biphasic water ± dichloro- methane system. The authors presume that tributylphosphine has a salient advantage over dithiothreitol which is widely used for the reduc- tive cleavage of the S7S bond. On termination of the reaction, the excess of dithiothreitol should be removed, whereas in reactions with tertiary phosphine the separation of the organic phase fromModern methods for the synthesis of peptide ± oligonucleotide conjugates the aqueous phase is sometimes sufficient. It was found that complete cleavage of the S-tert-butyl group should be carried out first, and only after this has been done, another oligonucleo- tide containing an iodoacetyl group at its 3 0-end is to be added to the reaction mixture. Otherwise, the removal of the S-protective group occurs at a slow rate, since iodoacetic acid and its amide can form quaternary phosphonium salts with Bu3P.The reaction of the HS group of the N-terminal cysteine residue of the peptide ± oligonucleotide conjugate with the electro- philic centre of the second oligonucleotide proceeded in a weakly alkaline medium, which made it possible to increase the yield of the target product. It should be noted also that all the reactions were carried out in a lithium chloride solution.64 The reason lies in poor solubility of the peptide ± oligonucleotide conjugate in water due to the interactions of the phosphate groups of DNA with positively charged guanidine groups of arginine, which are atte- nuated by the presence of lithium chloride.At the same time, the oligonucleotide pairs containing a peptide bridge are readily soluble in water because they contain an excessive (with respect to the guanidine groups) amount of negatively charged phos- phates. This approach was also used to prepare the conjugate 47 the two peptide fragments in which were attached to the 3 0- and 5 0-ends of the oligonucleotide. O O S Cys-(Orn)3 (Orn)3-Cys S oligonucleotide NH NH 47 In the studies,69 ± 72 the bromoacetic acid residue was intro- duced into the peptide fragment prior to the condensation with an oligonucleotide containing an HS group.The bromoacetyl deriv- ative of the peptide was taken in excess (from 5 to 25 equiv.). The reaction was carried out at room temperature for 5 ± 20 h. The yield of the target product was 70% ± 100%. The effect of the medium acidity (pH 7.0, 7.5, 7.9 and 8.5) on the completeness of the reaction and the dependence of the reaction rate on temper- ature have been studied.69 It was found that the peptide ± oligo- nucleotide conjugate was formed in quantitative yield in less than 3 h at 40 8C and pH 7.0 and 7.5. The publication 72 devoted to the synthesis of the complex fluorescently labelled peptide ± oligonucleotide conjugate 48 pro- vides an illustrative example of a successful application of the above-described procedure to the synthesis of peptide ± oligonu- cleotide conjugates.This was preceded by the synthesis of 19 ± 25- membered oligodeoxyribonucleotide phosphothioates containing a 50-terminal protected thiol group and an amino group at the end of the linker attached to its 3 0-end. This bisfunctionalised oligo- nucleotide was introduced into the reaction with a fluorescein isothiocyanate (FITC) fluorescent label at the 3 0-amino group; the thio group was then deprotected and reacted with the N-bromoacetyl derivative of the peptide. S7 peptide P S O (CH2CH2O)3 oligonucleotide phosphothioate O O HO (CH2)6 S NH C O NH COOH O 48 Special emphasis is laid on the possibility of performing large- scale synthesis of the conjugate 48. The success of this synthesis depended on the development of a new procedure for deprotection 251 of oligonucleotides in which the oligomers were treated with a mixture of dipyridyl disulfide and concentrated ammonia in the presence of phenol and methanol.This deprotecting mixture enables splitting of the oligonucleotide from the polymeric sup- port, liberation of the 3 0-terminal amino group, deprotection of the phosphate and amino groups of heterocyclic bases and the conversion of the 5 0-terminal thioacetyl protective group into the pyridyl disulfide group. The presence of phenol in the deprotect- ing mixture makes it possible to avoid the side reaction associated with a large-scale synthesis.Usually, the deprotection of internu- cleotide phosphates affords an electrophile, viz., acrylonitrile (CH2=CHCN), which reacts with the nucleophilic sulfur of the (oligonucleotide ± linker ± SH) system to give the undesirable product oligonucleotide ± linker ±SCH2CH2CN. The synthesis of peptide ± oligonucleotide conjugates based on the reaction of peptidyl chloromethyl ketone with an oligonculeo- tide containing a 3 0-terminal thio group has been described (Scheme 15).73 Scheme 15 oligonucleotide L L L SH O MeO Ala-Ala-Pro-Val CH2Cl O oligonucleotide L L O LSCH2 Val MeO Ala-Ala-Pro O O O O P . L= 6 O7 The conjugation of the oligonucleotide with the peptide was carried out in a water ± organic solution (pH 7.5) at room temper- ature for 12 ± 16 h.The yield of the target product was 60%. Activated esters of e-maleimidohexanoic (49),47, 48 3-maleimi- dobenzoic (50),51, 53, 54, 59, 60 b-maleimidopropionic (51) 55, 56 and 4-maleimidomethylcyclohexanecarboxylic (52) 57, 61 ± 63 acids were used as heterobifunctional reagents for the incorporation of the maleimide group into peptides and oligonucleotides. O O N O N O O 49 O O O O O R O C N O N O N N O O O O 51 50 O O O R N C O N CH2 O 52 O O R=H, SO3Na.252 In the study,48 the a-amino group of the peptide was acylated by the reagent 49 in water at pH 7.0 for the incorporation of the maleimide group into the peptide. This reaction was also carried out in methanol in the presence of triethylamine.47 In the majority of cases, the peptide was prepared by solid-phase synthesis; after deprotection, the free a-amino group was acylated by one of the above-mentioned bifunctional reagents in the presence of 1-hydroxybenzotriazole.51, 53 ± 56 Low storage stability of maleimide peptide derivatives con- taining lysine residues in aqueous media was noted.51 A complex mixture of reaction products is formed after a period of 3 ± 4 days as a result of the reaction of the e-amino groups of lysine residues with the maleimide group.To prevent side reactions, it is proposed to introduce the modified peptide into the reaction immediately after its detachment from the polymeric support.It was recommended to incorporate the bromoacetic acid residue into peptides by performing the acylation of the peptide e-group by bromoacetic acid anhydride in the course of the solid- phase synthesis.69 ± 71 The synthesis of maleimide derivatives of oligonucleotides includes two steps, viz., the automated oligonucleotide synthesis followed by modification of the oligomeric chain formed by 6-(4- monomethoxytrityl)aminohexyl 2-cyanoethyl N,N-diisopropyl- phosphoramidite.59 ± 62 The aliphatic amino group was introduced into the predetermined position of the DNA fragment using a commercially available 3 0-phosphoramidite derivative of the modified nucleoside 53.63 O O HN CF3 HN NH O N O DMTrO O OP 53 NPri NCCH2CH2O 2 After splitting of the oligonucleotide from the polymeric support and removal of protective groups, the aliphatic amino group reacted with one of the heterobifunctional reagents 49 ± 52 containing a maleimide group (pH 7 ± 8).The average yields of the oligonucleotide maleimide derivatives are 60%± 70%. The methods for the preparation of oligonucleotides contain- ing iodoacetic acid are similar. The amino group attached to the 3 0- or 50-end of the oligonucleotide through a hexamethylene bridge is acylated by N-hydroxysuccinimide ester 64 ± 66 or iodo- acetic anhydride in a water ± organic mixture.68 5. Conjugates with disulfide bonds between the fragments To this point, we have dealt with various procedures for the synthesis of conjugates, the oligonucleotide and peptide compo- nents of which were linked by stable covalent bonds.However, the development of techniques for the synthesis of peptide ± oligonu- cleotide conjugates with easily cleavable bonds between the frag- ments is a no less important task. Such a bond can be represented by an S7S bond, which can be formed and cleaved under conditions acceptable for various biological processes. The synthesis of conjugates with disulfide bonds between their peptide and oligonucleotide fragments can be carried out using three main approaches. First, this is in the oxidation of the HS groups of the oligonucleotide and the peptide with atmospheric oxygen with the formation of a disulfide bridge. In this case, the peptide is taken in an excess, the yield of the conjugate is 60%.61 However, according to Azhayev et al.74 the yields of peptide ± oligonucleotide conjugates prepared by this method are extremely low (0.4% ± 23%).E M Zubin, E A Romanova, T S Oretskaya The other two approaches consist of the activation of the HS group of the oligonucleotide or the peptide with 2,2 0-dipyridyl disulfide.61, 74 ± 81 SH+ R1 S S N N R2SH R1 S S R2 R1 N S S R1,R2, oligonucleotide or peptide chain fragments. The reaction is carried out under mild conditions at room temper- ature. Unprotected peptides and oligonucleotides can form non- covalent complexes due to electrostatic interactions of the neg- atively charged phosphates of DNA with positively charged side groups of the peptides.Such complexes are precipitated from solutions. Having regard to this circumstance, potassium chloride and acetonitrile, which neutralise molecular charges and facilitate their dissolution, are added deliberately to the reaction mixture.76 It is also recommended to carry out the synthesis of the conjugates in 90% formamide.74 The introduction of the S-(3-nitro-2-pyridylsulfenyl)cysteine residue (CysNpys), which is Boc-protected at the a-amino group, is a distinctive feature of a conjugation method.51, 69 It is noteworthy that the Npys-protective group is resistant to acid treatment. After cleavage from the polymeric support and removal of protective groups (with the exception of the Npys group), the peptide formed in the solid-phase synthesis is introduced into the reaction with a DNA fragment containing an HS-group under mild conditions.The reaction of the HS group of the oligonucleotide with the Npys-protected thiol group of the cysteine of the polypeptide yields a conjugate with the S7S bond between the peptide and the oligonucleotide fragments. NO2 NO2 peptide S + peptide N S S oligonucleotide S S NH oligonucleotide HS It should be noted that the HS group can be introduced into oligonucleotides in several ways. First, it can be attached to the 3 0- (see Refs 54 and 56) or 5 0-end (see Refs 51, 53, 55) of the DNA fragment using linkers of various lengths. Second, the synthesis of a protected nucleoside 3 0-phosphoramidite containing a thiol group can be followed by its introduction into any predetermined position of the oligomeric chain by the automated oligonucleotide synthesis. Earlier it was mentioned that it is cysteine-containing peptides that are usually coupled to oligonucleotides.However, some methods permit the introduction of HS groups into peptides devoid of the cysteine residues. Thus b-mercaptopropionic acid was attached to the N-terminal amino acid residue in the course of solid-phase peptide synthesis.74 6. Synthesis of conjugates using a carbonyl derivative of one of the fragments Syntheses of conjugates considered below are based on the reaction of carbonyl compounds with nucleophiles. The indisput- able advantage of the carbonyl groups is that they do not require additional activation and their reactions proceed with high selectivities.Dialdehyde derivatives of nucleic acids were extensively used in previous studies, e.g., in the structural elucidation of the monomeric components of RNA and DNA. The methods for the introduction of aldehyde groups into carbohydrate fragmentsModern methods for the synthesis of peptide ± oligonucleotide conjugates of nucleic acids and the properties of the thus modified derivatives have recently been reviewed.82, 83 Lebleu et al.84 ± 89 have performed a successful synthesis of peptide ± oligonucleotide conjugates based on dialdehyde oligo- nucleotide derivatives. After periodate oxidation, the oligonucleo- tide with a 3 0-terminal ribo fragment was subjected to condensation with poly(L-lysine) followed by reduction by sodium cyanoborohydride.The linkage between the peptide and the oligonucleotide fragments was effected through the morpho- line ring formed. O7 50 oligonucleotide HO O P O O B O 1) NaIO4 2) poly(L-Lys) 3) NaBH3CN HO OH O7 50 oligonucleotide O O P HO B O O N (CH2)4 poly(L-Lys) A (20?5 0)(A)n-polyribonucleotide prepared by enzymatic synthesis was used as an oligonucleotide with a 3 0-terminal ribo fragment;84 in the case of oligodeoxyribonucleotide, this was introduced with the help of T4 RNA ligase;85 the ribo fragment was the first to be attached to a polymeric support and the oligonucleotide was prepared by the automated synthesis.87 The early works of Lebleu et al.84 ± 87 reported the antisense properties of conjugates of dialdehyde oligonucleotide derivatives with peptides and their possible applications in the study of various biological systems.The results of testing of the complexes for the ability to inhibit replication of HIV-1 were published later (the oligonucleotide fragment of the complex was found to be complementary to the translation initiation centre of the Tat- protein).88, 89 Some sequence-specific antiviral effects of these compounds were observed. In the study,90 the electrophilic centres of oligonucleotides were generated enzymatically: the uracil residues were cut off from the oligonucleotide chain using uracil ±DNA glycosylase to obtain the apurine/apyrimidine fragments. This was followed by modification of the aldehyde-containing oligonucleotides by various ligands carrying an amino group.In particular, the attachment of tripeptides to such oligonucleotides has been described (Scheme 16). Scheme 16 R1 R1 O O OH O OH O 1) NH2-peptide 2) NaBH3CN O O R2 R2 R1 O OH peptide NH O R2 R1,R2, oligonucleotide chain fragments. A highly efficient selective procedure for oligonucleotide syn- thesis involving the formation of oximes and thiazolidines has recently been proposed.91 In their search for an `ideal' conjugation 253 procedure, i.e., the simplest one-step reaction between a free peptide and an oligonucleotide under physiological conditions, the authors made their choice in favour of two novel types of binding groups, viz., the oxime and the thiazolidine groups.The synthesis of conjugates involving the formation of an oxime (reaction of an aldehyde with an aminooxy derivative) and thiazolidine (reaction of the aldehyde with 1,2-aminothiol) is shown in Scheme 17. The aldehyde or the aminooxy group were incorporated into the oligonucleotide fragment at the 5 0-end, while the peptide contained the corresponding complementary group, viz., the aminooxy- or the 1,2-aminothiol groups for the reaction with the aldehyde-containing oligonucleotide or the a-N-glyoxyloyl group for the reaction with the aminooxy oligonucleotide derivative. 5 0-Modified oligonucleotides were prepared by the standard automated synthesis using the corresponding modified 3 0-phos- phoramidite derivatives in the final step.91 In turn, the N-terminal position or the side chain of lysine in the peptides was functional- ised by introducing the aldehyde or the aminooxy groups by means of oxidative cleavage of the serine residue or using N-Boc- O-(carboxymethyl)hydroxylamine, respectively.The thiazolidine ring between the peptide and the oligonucleotide is formed in Scheme 17 O HO O P R1 C O O (CH2)4 O B O ONH2+H NH OR2 O HO O P R1 O O N O (CH2)4 O B NH OR2 O HO O P R1 O O H+H2NO(CH2)4 O B NH O OR2 O HO HN N R1 P O O O O (CH2)4 O B OR2 O HO O P O R1 SH+ O O C (CH2)4 O B H HN NH2O OR2 O HO R1 NH P HN O O (CH2)4 O B S OR2 R1, fragment of the peptide chain, R2, fragment of the oligonucleotide chain.254 those cases where the original peptide is modified by attachment of cysteine at the N-end or at the side chain of lysine.Oxime can also be formed in theDNAduplex. In this case, the 5 0-aldehyde-containing oligonucleotide is hybridised with the complementary DNA chain and the DNA duplex formed is introduced into the reaction with a peptide carrying the aminooxy group. This strategy opens up fresh opportunities for post- synthetic modifications of single- and double-stranded DNA. The formation of an oxime bond between the oligonucleotide and the peptide fragments occurs at a high rate and with high selectivity.However, oligonucleotides possessing aminooxy groups are seldom introduced into reactions with peptides carry- ing aldehyde groups, since such oligonucleotides easily react with any carbonyl groups and are therefore not very convenient. Those peptide ± oligonucleotides in which the oligonucleotide and the peptide fragments are linked through the thiazolidine ring are less stable than those containing an oxime bond. Later, the same authors 92 reported the synthesis of conjugates using oligonucleotides with 3 0-terminal carbonyl groups. The 3 0-terminal aldehyde group was generated by mild periodate oxidation of a 1,2-aminoalcohol attached to the 3 0-end of the oligonucleotide in the automated synthesis on a standard com- mercially available polymer.The oligonucleotide thus prepared reacted with a peptide containing an aminooxy group similarly to the 5 0-aldehyde-containing oligonucleotide.91 Simultaneously, another group of investigators 93 published a paper devoted to the synthesis of peptide ± oligonucleotide con- jugates from oligonucleotides containing 5 0-terminal aminooxy groups and peptides containing N-terminal lysine groups with the side chains modified by the reaction with levulinic anhydride, i.e., peptides with N-terminal keto groups. In this case, the binding of the oligonucleotide and the peptide fragments occurred due to formation of the oxime bond. O peptide HN (CH2)4 + O O 50 oligonucleotide O O P O H2N (CH2)10 + O7 peptide HN (CH2)4 N O 50 oligonucleotide O O O (CH2)10 OPO7 The preparation of conjugates involving the formation of oximes has been described by Dey and Sheppard.94 The keto group was introduced into the C(5)-position of the oligonucleo- tide uracil through a spacer.III. Solid-phase synthesis of peptide ± oligonucleotide conjugates The performance of chemical reactions on insoluble polymeric supports is one of the main techniques in modern organic syn- thesis. The idea underlying this method is virtually versatile and applicable to a wide variety of biopolymers and their analogues. In the past few decades, the solid-phase method has become a routine procedure in the synthesis of peptides and oligonucleotides; therefore, investigators give considerable attention to the develop- ment of strategies and tactics of chemical synthesis of peptide ± oligonucleotide conjugates which is partly or completely per- formed on a solid phase.This strategy is based on the construction of amino acid and nucleotide sequences which includes either stepwise elongation of peptide and oligonucleotide chains on the same solid support or E M Zubin, E A Romanova, T S Oretskaya condensation of the two preformed fragments. The second route implies the attachment of a preformed peptide to an oligonucleo- tide synthesised on a solid phase and immobilised on a polymeric support. In development of the strategy of chemical synthesis of peptide ± oligonucleotide conjugates, special emphasis is laid on the choice of optimum methods for the formation of amide and phosphodiester bonds and a search for the most adequate combi- nations of protective groups designed for peptides and nucleic acid fragments and conditions for the deprotection (the latter should be selected with due regard to the stability of the target pepti- de ± oligonucleotide conjugates).The choice of method for the formation of a bond between the peptide and the oligonucleotide fragments is also of great importance. 1. Stepwise synthesis of peptide and nucleotide fragments on a single solid support The synthesis of oligonucleotides and peptides includes a series of repeated operations involving structurally similar monomeric components. The sequential synthesis of the oligonucleotide and the peptide fragments of the hybrid molecule on a single solid support can raise serious problems connected with the impossi- bility, in principle, of combining some standard procedures inherent in the solid-phase oligonucleotide and peptide syntheses.Therefore, meticulous selection of methods and optimisation of experimental protocols are necessary in order to ensure acceptable yields of target products and to avoid the formation of by- products. In 1972, M A Prokof'ev, Z A Shabarova et al. published a paper of fundamental importance 6 in which they gave the first description of a solid-phase synthesis of peptide ± oligonucleotide conjugates (Scheme 18). The authors used a classical Merrifield support, which repre- sents a chloromethylated styrene ± divinylbenzene (2%) copoly- mer.The synthesis of the peptide fragment was carried out using Boc-protected amino acids and N,N0-dicyclohexylcarbodiimide (DCC) as a condensation reagent. The choice of amino acids was determined so as to avoid complications resulting from the splitting of the target product off the polymer. The first nucleotide was attached to the peptidyl polymer by the so-called pyrophos- phate method, which consists in the activation of the phosphate group of the nucleotide owing to the formation of mixed anhy- dride with diphenyl phosphate. This reaction was carried out in dry DMF in the presence of a tenfold excess of thymidine 5 0-phosphate ± diphenyl phosphate mixed anhydride (20 8C, 100 h).The elongation of the oligonucleotide chain was carried out by the phosphodiester method. The condensation was per- formed in dry pyridine in the presence of a five- or a tenfold excess of 3 0-O-acetylthymidine 5 0-phosphate and a fivefold excess of dicyclohexylcarbodiimide with respect to thymidylyl-(5 0?N)- Phe-Gly-Phe. The deprotection of the 3 0-hydroxy group of the nucleotide was carried out in alkaline aqueous dioxane (pH 9). Such a treatment ensured a practically complete removal of acetyl protection without splitting of the dinucleotidyl-(5 0?N)-peptide off the polymer. The final step included hydrolysis of the ester bond between the polymer and the trinucleotidyl-(5 0?N)-peptide. The authors encountered serious difficulties because of both steric hindrances and the impossibility of using strong acids, which might result in cleavage of the phosphoramide bond.The maximum extent of cleavage of the ester bond (10%) is achieved by treating the oligonucleotidyl-(5 0?N)-peptide with NaOH in aqueous dioxane. In its modern modification, this approach combines stepwise synthesis of the peptide from the C-end and subsequent elongation of the oligonucleotide chain on the peptidyl polymer by the phosphoroamidite method (Scheme 19). If the peptide is devoid of hydroxyamino acid residues, the synthesis of the DNA frag- ment should be preceded by incorporation of a hydroxy group into the peptidyl polymer. For this purpose, the a-amino group is acylated by a bifunctional reagent which contains an activatedModern methods for the synthesis of peptide ± oligonucleotide conjugatesO (PhO)2P O CH2Ph CH2Ph P O CH2 H2N HN HN O O O CH2Ph CH2Ph P O CH2 HN HN HN O O O HO P 1) pT(Ac), DCC 2) OH7 Thy O O HO carboxy and a protected hydroxy group.This promotes the insertion of a linker between the N-end of the peptide and the oligonucleotide. Scheme 19 Introduction of an anchor group P P anchor groupSynthesis of the peptide fragment P anchor group protected peptide Introduction of a linker linker P anchor group protected peptide Synthesis of the oligonucleotide fragment pro- tected protected oligonucleo- P anchor group linker peptide tide Splitting of the peptide ± oligonucleotide conjugate and deprotection of functional groups linker oligonucleotide peptide a.Solid support The success of a solid-phase synthesis largely depends on the correct choice of solid support. The most popular support is controlled pore glass (CPG).95 ± 102 It is common practice to use CPG with a pore diameter of 500 A. The synthesis of peptide ± oligonucleotide conjugates on silica gel (Fractosil) has also been described.100 However, in a more recent publication 102 the same authors had to admit that the use ofCPGwas preferable, since this solid support allows the synthesis of the target product in higher yields even at lower specific loading with the first monomeric unit. At the same time, some investigators (see, e.g., Refs 103 ± 105) call into question the expediency of CPG application, since the synthesis of the peptide fragment on this support affords products with shorter chains in addition to the target oligomer.In the authors' opinion,53, 103 the formation of short peptide chains can 255 Scheme 18 OP OH O O O Thy HO twice P pTpTpT-Phe-Gly-Phe O CH2 be avoided through the use of a copolymer of polyethylene glycol with polystyrene (PEG-PS). In a series of publications by Spanish investigators,104 ± 113 the use of a copolymer of styrene with 1% divinylbenzene has proved to be especially efficient in peptide synthesis. b. Introduction of an anchor group The first amino acid, which is to become a C-terminal residue in the peptide fragment to be synthesised, is attached to the support through the ester bond.In order to effect the esterification reaction between the amino acid and the solid support, the latter should contain reactive groups. Special methods for chemical modification of the support are being developed. The use of polystyrene modified by aminomethyl 106 ± 109 or p-methylbenzhydrylamine 109, 110 groups as a polymeric support is reported. In the first step, the NH2 group of the support was acylated by N-protected leucine used as an internal standard. After removal of the amino protective group, the leucyl polymer was treated with a bifunctional reagent, viz., 3-nitro-4-(2-hydr- oxyethyl)benzoic acid (54) 106 ± 110, 113 orN-(9-hydroxymethylfluo- ren-2-yl) succinamide (55).109 HO X OH (54 or 55) P H Leu NH CH R P CH HO Leu NH X R O (54), C Me; X = R=H, CH2 CH2O2N O (55).HN O CH2 The presence of a bridge between the support and the C-terminal amino acid residue allows splitting of the target product from the solid phase under very mild conditions under the action of base. The linkers can be attached directly to the support, e.g., to PEG-PS containing an amino group.53, 103 The use of N-protected 6-aminohexyl succinate 56 as a linker has been described.53 O O OH MMTrNH 56 O Commercially available CPG containing 3-aminopropyl groups was used 95 ± 97, 100 ± 102 for the synthesis of peptide ± oligo-256 nucleotide conjugates. In this case, chemical modification of the solid support was reduced to the acylation of 3-aminopropyl groups by activated esters of 4-hydroxybutyric 95 ± 97 or 10- hydroxydecanoic acid 100 ± 102 with the hydroxy groups protected as acid-labile dimethoxytrityl and pixyl (9-phenyl)xanthen-9-yl derivatives. Two additional e-aminocaproic acid molecules were introduced between the 4-hydroxybutyric acid residue and the aminopropyl support.97 Lukhtanov et al.99 used porous glass modified by an extended linker containing a terminal NH2 group.The latter was acylated with the activated ester 57. F F O H MMTrN O O O 57 F F It is of note that the addition of the first amino acid to the support and the formation of the amide bond in the course of solid-phase synthesis of peptide fragments are usually carried out by the methods involving carbodiimide, activated esters or sym- metrical anhydrides. c.Protective groups Temporary protection of a-amino groups of amino acids is mainly accomplished using Boc 53, 99, 101 ± 103, 106-110 or Fmoc 95 ± 98, 100 groups. Sequential solid-phase synthesis of peptide and oligonu- cleotide fragments demands protection of reactive side groups of certain amino acid residues. Protective groups are selected in order to ensure selective removal of other groups. The Fmoc 53, 101, 103, 104 or the Boc group 95, 96 are most commonly used for the protection of e-amino groups of lysine; trifluoroacetyl (Tfa) 104 and 1-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl (Dde) 100 protective groups are used less frequently.The latter withstands splitting of theNa-Fmoc group by treatment with 20% piperidine in DMF. The b-carboxy group of aspartic acid is protected by conversion into 9-fluorenylmethyl ester.53, 103, 106, 108 The possibilities for the protection of the indole (in) ring of tryptophan are rather limited. Nin-Formyl protection is one of them. It is stressed 104 that the use of this protection in the synthesis of peptide ± oligonucleotide conjugates prevents the formation of side product. Marchan et al.104 have succeeded in obtaining Boc-Arg(Fmoc)2-OH which can be used in the solid- phase synthesis of the peptide fragment of the conjugate. It is known111 that unprotected amide groups of asparagine or glutamine residues in the peptide fragments of conjugates can undergo phosphitylation during subsequent elongation of the oligonucleotide chain.However, this side reaction proceeds at a very low rate which excludes the necessity of blocking side groups of asparagine and glutamine.53, 104 Grandas et al.109, 110, 112 protected the hydroxy group of homoserine by the dimethoxytrityl group leaving the HO groups of serine, threonine and tyrosine unprotected.106 ± 110 Other authors 101 recommend protecting the hydroxy groups of serine as allyl carbonates. Assuming that the use of the Fmoc-strategy of solid-phase peptide synthesis may result in partial cleavage of this protective group, the authors gave preference to the Boc-strategy. Despite the vast variety of techniques for the protection of the imidazole (im) ring of histidine, no ideal protective group has been O R2 R1 Oligonucleotide synthesis P O HN anchor group Y NH NHX O O n m OH Y=Ac, Pac; X=7CH27,7CH2C6H47,7CH2CH27,7CH(CH3)7; R1, R2, side chains of amino acids.E M Zubin, E A Romanova, T S Oretskaya found for this compound. In the synthesis of peptides containing this amino acid, Nim-benzyloxymethyl protection was used and deprotection was effected by catalytic hydrogenolysis over palla- dium black in the presence of cyclohexa-1,4-diene. According to Truffert et al.,102 such a treatment should not have any negative effect on the stability of the oligonucleotide fragment of the conjugate. However, this method can hardly be considered satisfactory, since the authors failed to completely remove the peptide ± oligonucleotide conjugate formed from the catalyst's surface. The use of the Boc or the Fmoc groups for the protection of the imidazole ring of histidine cannot be recognised as very convenient either.The reason is that treatment of the peptide ± oligonucleotide conjugate with trifluoroacetic acid to remove the Nim-Boc group can result in hydrolysis of the N-glycosidic bonds in purine nucleotides, whereas the Nim-Fmoc group is rather labile under conditions of the solid-phase syn- thesis. In the authors' opinion,102 the tosyl (Tos) group is especially convenient for the protection of the imidazole ring of histidine. This group is removed under the action of 1-hydroxy- benzotriazole commonly added to the reaction mixture in order to inhibit racemisation in the formation of the peptide bond pro- moted by DCC.Two procedures for the protection of the imidazole ring of histidine by the tosyl and 2,4-dinitrophenyl groups have been developed by Beltra n et al.110 These authors share the opinion that 1-hydroxybenzotriazole can be replaced by tetrazole which has proved to be an efficient reagent when combined with DCC. The analysis of reversed-phase HPLC data led the authors to conclude that the use of Boc-His(Tos)-OH and deprotection of the imidazole ring immediately prior to the elongation of the oligonucleotide chain is the most efficient approach to the solid-phase synthesis of the peptide fragment. d.Introduction of the linker After termination of solid-phase synthesis of the peptide compo- nent of the conjugate containing a hydroxy amino acid residue,{ deprotection of the a-amino group is followed by its repeated selective protection by introducing the phenylacetyl (Pac) 106 ± 108, 113 or the acetyl (Ac) group 109, 110 after which the oligonucleotide synthesis is performed (Scheme 20). If the peptide fragment of the conjugate is devoid of hydroxy amino acids, deprotection is followed by acylation of the free a-amino group by a compound containing an activated carboxy group and a protected hydroxy group.53, 95 ± 103 p-Nitrophenyl O-pixyl-4-hydroxybutyrate,97, 99 p-nitrophenyl 100, 101 and penta- fluorophenyl 102 O-dimethoxytrityl-10-hydroxydecanoates and compound 5853, 95, 96, 103 are used as acylating reagents. O NH DMTrO O NO2 58 O The incorporation of the aliphatic OH group into the peptide and its further selective deprotection make the synthesis of the DNA fragment possible.{ For incorporation of homoserine into the peptide chain, Na-Pac-O- DMTr-Hse is introduced into the condensation reaction in the form of the triethylammonium salt.109, 110 Scheme 20 O R2 R1 O HN P anchor group Y NH NH X O O n m O protected oligonucleotideModern methods for the synthesis of peptide ± oligonucleotide conjugates e. Synthesis of the oligonucleotide fragment In the studies 95 ± 102 where CPG was used as a solid support, automated solid-phase phosphoramidite synthesis of an oligonu- cleotide component of a conjugate was carried out according to a standard protocol.If a copolymer of styrene with ethylene glycol 53, 103 or a copolymer of styrene with 1% divinylben- zene 106 ± 113 are used as polymeric supports, the standard proce- dures had to be modified in order to increase the yields in each step of the oligonucleotide synthesis. These modifications include a severalfold increase in the time required for implementation of all the main steps of internucleotide bond formation, the use of 5-(2- nitrophenyl)tetrazole instead of tetrazole and a significant increase in the concentration of the nucleotide solution in dichloromethane. In the overwhelming majority of studies, the elongation of the oligomeric chain was carried out in the direction 3 0?5 0 except for the cases where 5 0-phosphoramidites of protected deoxynucleo- sides were used as monomeric components.106 ± 108 Labile protec- tive groups 106, 108, 109, 113 were used together with acyl ones for the protection of exocyclic amino groups of heterocyclic bases.For instance, the N,N-dimethylformamidine and tert-butylphenoxy- acetyl groups were used for the protection of the NH2 groups of adenine and guanine, while the isobutyryl and the tert-butylphe- noxyacetyl groups were used for cytosine. f. Cleavage of conjugates from the solid support and deprotection of functional groups The problem which has to be solved for the successful solid-phase synthesis of peptide ± oligonucleotide conjugates by stepwise elongation of peptide and oligonucleotide chains is to find optimal conditions for the deprotection of functional groups and cleavage of the bonds between the conjugate and the support.Thus it is not recommended to liberate the conjugate from the support by treatment with strong acids or by using acid-labile protective groups because of the risk of partial apurinisation of DNA. For this reason, in the majority of studies post-synthetic treatments were carried out under mild basic conditions. The apprehensions that alkaline treatment might lead to racemisation of the amino acid residues at the a-carbon atom were not confirmed.53 It was noted also that alkaline treatment of conjugates in which the peptide is linked to the oligonucleotide through a serine residue favours the cleavage of the phosphodiester bond between the two fragments.3, 4 This reaction includes cleavage of the C7O bond and occurs as base-catalysed b-elimination (see Scheme 2).How- ever, a detailed analysis of this problem showed 108 that treatment of conjugates with a concentrated aqueous ammonia ± dioxane mixture (1 : 1) at room temperature for 15 ± 24 h did not cause the cleavage of the bond between the oligonucleotide fragment and the serine residue. A convenient procedure was developed 106 ± 109 for the libera- tion of peptide ± oligonucleotide conjugates from the solid phase by treatment with 0.05 M tetrabutylammonium fluoride in tetra- hydrofuran at room temperature for 30 ± 60 min.Such treatment ensures simultaneous removal of b-cyanoethyl protective groups from internucleotide phosphates and splitting of the 9-fluorenyl- methyl ester of aspartic acid. This is followed by deprotection of exocyclic amino groups of heterocyclic bases under the action of a mixture of concentrated aqueous ammonia and dioxane (1 : 1) at room temperature for 15 ± 18 h. The latter procedure was also carried out at 55 8C (see Ref. 110). Apparently, more drastic conditions were required for the removal of the isobutyryl protective group from the exocyclic amino group of the guanine base in the DNA fragment. The same procedure was used for the removal of some other protective groups, e.g., the tosyl and the 2,4-dinitrophenyl groups (from histidine).However, an attempt of de la Torre et al.53 to liberate a peptide ± oligonucleotide conjugate from the polymeric support (PEG-PS) using tetrabutylammonium fluoride failed. It was assumed that polyethylene glycol, which is a constituitive element of the support, prevents b-elimination catalysed by the fluoride 257 ion. The use of tetrabutylammonium fluoride is also unacceptable for the liberation of those peptide ± oligonucleotide conjugates in which the components of the conjugate are linked through the hydroxy group of tyrosine. In this case, the two parts of the hybrid molecule are linked through a phosphotriester bond. It is there- fore quite probable that the nucleophilic attack of the fluoride ion at the phosphorus atom will be accompanied by the cleavage of the P7O bond with the formation of a stable phenoxide ion.Therefore, Robles et al.109 proposed to use a mixture of concen- trated aqueous ammonia and dioxane (1 : 1) at 55 8C for 18 h for the liberation of the peptide ± oligonucleotide conjugate from the solid support. It should be noted that the carboxy group of the C-terminal amino acid is converted into the amide group under these conditions. The use of concentrated aqueous ammonia for the cleavage of peptide ± oligonucleotide conjugates from the solid support and concomitant removal of protective groups yielded a mixture of products.53, 96 This can be explained by the fact that the cleavage of the ester bond under the action of concentrated ammonia is accompanied by the formation of both an amide and an ammo- nium salt.Prior to ammonolysis, the oligonucleotide ± peptidyl polymer was treated with a solution of 0.5 M 1,8-diazabicy- clo[5.4.0]undec-7-ene (DBU) in acetonitrile.53 This made it possi- ble to remove the Ne-Fmoc and b-cyanoethyl groups and to cleave the 9-fluorenylmethyl ester of aspartic acid. In the authors' opinion, this reaction order should exclude the conversion of the ester into the amide. In order to liberate the protected peptide ± oligonucleotide conjugate from the support, Truffert et al.100 used a 50% solution of ethanolamine in absolute ethanol (60 8C, 30 h). The final product was obtained in the form of monoethanolamide in a very low yield.To increase the yield, Truffert et al.101, 102 used 0.1 M NaOH instead of ethanolamine. Under these conditions, hydrolysis of the ester bond between the peptide ± oligonucleotide conjugate and the support is completed in 2 h at room temper- ature and is accompanied by a simultaneous removal of other protective groups, e.g., the b-cyanoethyl groups from internucleo- tide phosphates, the tosyl group from histidine, the allyloxycar- bonyl group from serine and the Fmoc or Dde groups from lysine. The authors note that deprotection of exocyclic amino groups of adenine, cytosine and guanine requires no less than 24 h. The use of the Boc-protection for the side amino group of lysine is undesirable, since its removal proceeds under rather drastic conditions [treatment with trifluoroacetic acid and 1,2- ethanedithiol (9 : 1) for 5 min 95, 96] which may lead to hydrolysis of N-glycosidic bonds in the oligonucleotide fragment and sub- sequent destruction of the target product.2. Sequential synthesis of oligonucleotide and peptide fragments on the same solid support In the studies considered above, peptide ± oligonucleotide con- jugates were prepared by stepwise synthesis, viz., the peptide fragment of the hybrid molecule was first synthesised on a solid phase after which it was possible to proceed to the synthesis of the oligonucleotide chain. In an alternative approach, the DNA fragment is synthesised first. The main limitation of this method is the impossibility of using the Boc-strategy for the peptide synthesis.The solid-phase synthesis of oligonucleotides with a 5 0-termi- nal reactiveNH2 group has been described.114 ± 117 To this end, the 5 0-amino-5 0-deoxythymidine residue was incorporated into the oligomeric chain in the final step of automated phosphoramidite synthesis. The nucleoside thus modified plays the role of a linker between the oligonucleotide and the peptide fragments of the conjugate. The Fmoc-strategy was used for the elongation of the peptide chain by Bergmann and Bannwarth 114 (Scheme 21); the activation of the carboxy group was carried out in situ using O-(benzotriazol-1-yl)-N,N,N0N0-tetramethyluronium hexaflu- orophosphate (HBTU). The allylic group was selected as the phosphate-protective group instead of the b-cyanoethyl group258 Thy Thy Thy O Oligonucleotide synthesis Peptide synthesis R O P O O OR DMTrO H2N OAll 5 Thy Thy Thy Me O O NH P O O P O HN O 6 NC(CH2)2O H O NC(CH2)2O 5 Thy Thy Me O O O P O P O NH O 6 O7 NHá4 H O O7 NHá45 Me O O N R= CPG .HN O commonly used for this purpose, because in this study the Fmoc- group was removed by a strong base (DBU) instead of piperidine. For the same reason, the solid-phase synthesis was carried out on modified CPG. To ensure the stability of the bond between the oligonucleotide and the support during DBU treatment, a sarco- sine residue was incorporated into the anchor group. After termination of the synthesis of the peptide fragment, synthesis of another oligonucleotide chain was begun with the use of 2-cyanoethyl N,N-diisopropylphosphoramidites of 2 0-deoxy- nucleosides.This procedure afforded conjugates in which two DNA fragments were linked through a peptide fragment. These were liberated from the solid support by consecutive treatment with Pd[PPh3]4 and morpholine in a DMSO±THF± dioxa- ne ± 0.5 M HCl mixture (2 : 2 : 2 : 1) at 60 8C for 3 h and then with 30% aqueous ammonia at room temperature for 2 h. It should be noted that the bond between the peptide and the second oligonucleotide is of a phosphoramide type, which is not cleaved in the post-synthetic deprotection of the peptide ± oligonucleotide conjugate and its liberation from the solid support.An analogous synthetic scheme was developed for the prepa- ration of conjugates in which the oligonucleotide fragment is a heteropolymer.115, 116 Special mention should be made of one of the most recent studies carried out by Bleczinski and Richter 117 which describe a procedure for the preparation of cyclic peptide ± oligonucleotide conjugates. Fmoc-protected amino acids and All-protected nucleoside phosphoramidites were used as the starting com- pounds. This approach was used to obtain the cyclic hybrid Glu-Leu-T*T-DP-Lys (59) (T* is 5 0-amino-5 0-deoxythymidine, DP is 3-hydroxy-2,2-dimethylpropionic acid). The amide bond of the `head-to-tail' type was formed by the e-amino group of lysine and the g-carboxy group of glutamic acid.The structure of compound 59 was confirmed by 1H NMR spectroscopy and MALDI-TOF mass spectrometry. The cyclic conjugate appeared to be resistant to exo- and endonucleases. By varying the peptide fragment, it was possible to obtain an analogous 39-membered ring. 3. Synthesis of peptide ± oligonucleotide conjugates on solid supports with a bifunctional branched anchor group An approach to the synthesis of peptide ± oligonucleotide con- jugates proposed in the early 1990's 118 ± 121 and further success- fully developed in the studies carried out in the past three years 122 ± 125 is of considerable interest. The oligonucleotide E M Zubin, E A Romanova, T S Oretskaya Scheme 21 Thy Thy O Oligonucleotide synthesis Me R O P O O HN H OAll 5 NH O 6 Thy O Ammonolysis R O O P O OAll 5 Thy Thy O H O O P O NH O7 NHá45 syntheses were performed on solid supports with branched anchor groups containing HO and NH2 groups.The presence of this group allows sequential attachment of monomeric units, first, of one type and then, of another type, on the same support. Post- synthetic treatments result in a hybrid molecule in which the oligonucleotide and the peptide fragments are connected through a linker. The distinctive feature of this method is that the bifunc- tional } linker is attached directly to the solid phase rather than to one of the fragments of the synthetic conjugate. 3-Aminopropane- 1,2-diol 118 and 6-amino-2-hydroxymethylhexan-1-ol 119 can be used as linkers.The hydroxy and amino functions are protected by the DMTr and Fmoc groups commonly used for this purpose. In these studies,118, 119 the peptide fragment was synthesised first. In the peptide synthesis, the e-amino group of lysine was protected by the Boc group; the hydroxy group of serine and the thiol group of cysteine were protected by the tert-butyldimethyl- silyl (TBDMS) and trityl (Tr) groups, respectively, whereas the amino groups of adenine and guanine were protected by the 2-(acetyloxymethyl)benzoyl group, while those of cytosine were protected by the acetyl group. The main difference in the approaches proposed by Cana- dian 118 and American 119 investigators consists in the choice of an insoluble support and the method for its chemical modification.The polymer employed by the Canadian investigators Juby et al.118 was commercially available Teflon to which 5 0-O-dimeth- oxytrityl-2 0,3 0-di-O-acetyladenosine was attached through an extended linker. The first step was the selective removal of the DMTr protective group from the 5 0-hydroxy group followed by the introduction of a protected 3-aminopropane-1,2-diol residue using the amidite strategy to yield derivative 60 (Scheme 22). This was employed for the peptide and then for the oligonucleotide synthesis. After completion of the peptide and oligonucleotide syntheses, the functional groups were deprotected under mild conditions by treatment with ethylenediamine in absolute ethanol (1 : 1) at 55 8C for 1 h.(This procedure for the removal of protective groups was specially developed in the synthesis of oligonucleotide methylphosphonate derivatives.) The cleavage of the target product from the solid phase was carried out in two steps. At first, the cis-glycol group of the 2 0,3 0-di-O-acetyl } It would be more correct to speak about a tri- rather than about a bifunctional linker, since it contains one more reactive group involved in the formation of a covalent bond with a solid support.Modern methods for the synthesis of peptide ± oligonucleotide conjugates Teflon HN N N DMTrO N N O DMTrO AcO OAc NH protected peptide O O P NC(CH2)2O O protected oligonucleotide H FmocN ODMTr O O HN OHN protected peptide peptide HN OH adenosine linker was oxidised by sodium periodate after which b-elimination of the peptide ± oligonucleotide conjugate [cleavage of the C(5 0)7Obond] under the action of PrNH2 was performed.The American investigators Basu and Wickstrom 119 used a linker-modified copolymer of styrene with ethylene glycol as a support (Scheme 23). The post-synthetic deprotection of functional groups in the target product was performed under mild basic conditions (0.05 MK2CO3 in methanol). It was noted that complete cleavage of the bond between the peptide ± oligonucleotide conjugate and the polymer required additional 30-min treatment with ammonia at room temperature. Successful application of the solid support 61 with a branched linker for the synthesis of peptide ± oligonucleotide conjugates has been reported by Nielsen et al.121 O O O NH O O HN 2 O O O O (CH2)6 N DMTrO H HN O 61 Me De Napoli et al.122 used a nucleoside residue containing cytosine as a heterocyclic base for the modification of the polymeric support.This residue was attached to the spacer group of the polymeric support through the exocyclic amino group of cytosine (Scheme 24). The differently protected 3 0- and 5 0-hydroxy groups of the deoxyribonucleoside residue on the polymeric support were subject to subsequent modifications by phosphoramidite derivatives. The 5 0-hydroxy group was used for FmocNH O O P NC(CH2)2O 60 HN NN O N OOAc AcO 1) Peptide synthesis 2) Oligonucleotide synthesis P protected oligonucleotide O O O HN O oligonucleotide O NHFmoc 2O CPG HN 259 Scheme 22 Teflon HN N N 1) Peptide synthesis 2) Oligonucleotide synthesis N O N OOAc AcO Teflon peptide N NH O 1) NH2CH2CH2NH2 2) NaIO4 3) PrNH2 O7 O PO7 oligonucleotide O Scheme 23 1) CF3COOH, HSCH2CH2SH 2) 0.05 M K2CO3 in CH3OH 3) NH3 .aq Pconventional construction of the oligonucleotide chain in the direction 3 0?5 0, whereas the 3 0-hydroxy group was phosphity- lated by a commercially available standard amino linker and subsequently used for peptide synthesis. Scheme 24 O P NH NH O N ... O N DMTrO O OTBDMS O 30 50 HO NH peptide NH2 O P O (CH2)6 oligo- nucleotide OH The design, synthesis and testing of a novel polymeric support 62 for the synthesis of peptide ± oligonucleotide conjugates and their phosphothioate analogues have been described.123, 124 The deblocked hydroxy group of this support was involved in standard oligonucleotide syntheses, while the deblocked amino group is involved in peptide synthesis. As a result, the 3 0-end of the oligonucleotide becomes bound to the C-end of the peptide through a phosphodiester (or a thiophosphodiester) bond and a short linker.260 ODMTr H FmocN O O ...O NH P 62 OH 30 50 peptide HN H OH O phosphothioate- containing oligonucleotide Later, the same authors 125 performed the synthesis of the peptide ± oligonucleotide conjugate 63 using a polymeric support with a branched linker. peptide NH2 HN 50 O O HO OH phosphothioate- containing oligonucleotide HN 63 O 4.Synthesis of oligonucleotide peptides by fragment condensation The essence of this method is in the formation of a covalent bond between the presynthesised peptide and the oligonucleotide after completion of the automated oligonucleotide synthesis but prior to the cleavage of the nucleic acid fragment from the solid support. The main advantage of this approach is that one component of the conjugate is not exposed to the effect of chemical reagents used for the synthesis of the other component as is the case with sequential solid-phase synthesis of peptide ± oligonucleotide conjugates.Very often, peptide ± oligonucleotide conjugates are prepared by the reaction of an oligonucleotide containing free 5 0-hydroxy groups and immobilised on an insoluble support with a peptide phosphoramidite derivative in the presence of tetra- zole.108, 111, 126 ± 133 The peptide phosphoramidite derivatives can be prepared in good yields by the reaction of the hydroxy group of serine or tyrosine with a standard phosphitylating reagent.108, 126 ± 132 It was also proposed to phosphitylate the C-terminal amide group of the peptide.111, 133 As in the case of other methods for the synthesis of pepti- de ± oligonucleotide conjugates, the main efforts of investigators were directed at a search for the mildest conditions for post- synthetic treatment.This is especially important in those cases where the formation of the phosphodiester bond between the oligonucleotide and the peptide fragments of the conjugate involves the serine residue. Such compounds destroy under alkaline conditions. In contrast, those molecules, which contain phosphoramide bonds, are unstable in strongly acidic media. van Boom et al.128 protected the a-amino group of the peptide by the p-nitrobenzyloxycarbonyl group and the o-carboxy groups of aspartic and glutamic acids were converted into p-nitrobenzyl esters. These protective groups are removed simultaneously with a sodium dithionite ± sodium hydrogencarbonate mixture. The exo- 1) N,N0-Carbonyldiimidazole 2) H2N(CH2)nNH2 H2N CPG HO protected oligonucleotide O O peptide n O protected oligonucleotide FmocNH NH NH n=3, 6, 12.E M Zubin, E A Romanova, T S Oretskaya cyclic amino groups of the heterocycles were protected by the 2-(tert-butyldiphenylsilyloxymethyl)benzoyl group. Oxalic acid served as a linker between the first nucleoside and the polymer.128 Liberation of the target product from the solid phase with simultaneous deprotection of heterocyclic bases and internucleo- tide phosphates was carried out using tetrabutylammonium fluoride in aqueous pyridine. It is known that allyloxycarbonyl and allyl protective groups are split off under mild conditions. These groups were used 129 ± 131 for the protection of the functional groups of the oligonucleotide (exocyclic amino groups of heterocyclic bases, internucleotide phosphates) and the peptide (terminal NH2 and COOH groups) fragments.These protective groups are easily removed by a palladium complex in the presence of diethylammonium formate. The synthesis of peptide ± oligonucleotide conjugates contain- ing nucleotidyl-(5 0?O)-serine fragments was performed using protection of the lateral carboxy group of aspartic acid by any of the three esters, viz., 9-fluorenylmethyl, benzyl or 2-(2-nitro-4- acetylphenyl)ethyl esters.108 The exocyclic amino groups of adenine residues were protected by benzoyl or N,N-dimethylfor- mamidino group. The cleavage of the peptide ± oligonucleotide conjugate from the solid support and removal of protective groups were carried out using three different methods, viz., by treatment with concentrated aqueous ammonia, potassium carbonate or lithium hydroxide in a methanol ± dioxane mixture. It is of note that the authors 108 failed to correctly characterise the target product and to confirm its structure by mass spectrometry and amino acid analysis data.The methods for the synthesis of conjugates containing an amide bond between the oligonucleotide and the peptide fragments were considered in the section devoted to the solution synthesis of peptide ± oligonucleotide conjugates (see Section II.3). The difficulties reported in the publications 38, 39 could be largely overcome by performing the solid-phase con- densation of the oligonucleotide and the peptide components, which significantly increased the yield of the target product.The decomposition of diazoketone 37 (see Scheme 10) was carried out in the presence of silver benzoate rather than underUVirradiation in the apprehension that in the latter case the reaction rate would decrease due to partial absorption of energy by the solid support (polystyrene or CPG).39 The condensation of the pentapeptide Ac-Ala-Tyr-Gln-Val- Phe with tri- and tetranucleotides containing 5 0-terminal 5 0-amino-5 0-deoxythymidine residues and immobilised on a solid phase was reported by Tetzlaff et al.115 However, these authors drew attention to the low efficiency of this reaction, which they attributed to the low solubility of the peptide in dimethyl- formamide.The recently published data deserve special mentioning. Thus, Gait et al.134, 135 used the following approach for the incorpora- tion of an aliphatic amino group into the DNA fragment: the 5 0-hydroxy group of the CPG-linked oligonucleotide following removal of the terminal DMTr protective group was activated by N,N-carbonyldiimidazole and introduced into the reaction with a diamine (Scheme 25). When PyBop was used as a condensation reagent in the presence of 1-hydroxybenzotriazole, the degree of racemisation at the a-carbon atom of the C-terminal amino acid was very small. Nevertheless, peptides containing C-terminal Scheme 25 O O peptide OH FmocNH n CPG O protected oligonucleotide NH PyBOP CPGModern methods for the synthesis of peptide ± oligonucleotide conjugates glycine were selected for conjugation to oligonucleotides, since in this case the a-carboxy group of the peptides could be activated without any risk of racemisation.A tetrapeptide containing non- polar amino acid residues and an octapeptide built up of polar amino acids were used as peptide fragments, while various homo- (e.g., T12) and heterooligonucleotides were used as oligonucleo- tides. The conjugation of both peptides to T12 proceeded with high efficiency.134, 135 However, the yields of peptide ± oligonucleotide conjugates prepared by condensation with heterooligonucleotides were very low. The authors failed to explain the reason for this phenomenon.Neither did the changes in experimental conditions, i.e., the use of PEG-PS instead of CPG, a larger peptide excess (10 instead of 5 equiv.) and higher temperature (40 8C), give any positive result. It was suggested that the main problem is related to poor solubility of protected heterooligonucleotides in dimethylform- amide due to their high hydrophobicity. This problem was over- come by selective removal of b-cyanoethyl protective groups by triethylamine in acetonitrile which caused some increase in the yields of the target products. In the majority of cases, the elongation of the linker containing the amino group also had a beneficial effect on the conjugate yields. A modification of this reaction was employed 136 in the condensation of a presynthesised peptide with a blocked oligo- nucleotide immobilised on a solid phase and containing a linker attached to the C(2 0)-atom of the carbohydrate fragment (Scheme 26).Oligodeoxyribonucleotides employed contained either three modified nucleosides, viz., 2 0-(3-aminopropionyl)- amino-2 0-deoxy-arabino-adenosine, 2 0-(3-aminopropionyl)ami- no-2 0-deoxyuridine or 2 0-amino-2 0-deoxy-arabino-adenosine. Scheme 26 O HN CPG DMTrO O protected oligonucleotide HNFmoc OH protected peptide FmocNH O protected oligonuc- CPG DMTrO O HBTU, EtNPri leotide 2 NH2 protected oligonuc- CPG O DMTrO leotide 1) NH3 . aq NH 2) 80% AcOH protected peptide FmocNH O oligonucleotide HO OH peptide NH H2N O First, the Fmoc-protected amino group attached to the C(2 0)- atom of the carbohydrate fragment of one of nucleotide residues either directly or through a linker was selectively deprotected with 50% morpholine in dimethylformamide for 1 h.Then the oligo- nucleotide with the free NH2 group reacted with a protected peptide fragment. Two short peptides, viz., Na-Fmoc-Leu-Gly and Na-Fmoc- Tyr-D-Ala-Phe-Gly, were selected as peptide fragments. In both cases, the C-terminal amino acid residue was represented by a glycine residue. The latter circumstance is very important, since in this case the activation of the carboxy group of the peptides was not accompanied by racemisation due to the absence of a chiral carbon atom in the glycine molecule.261 The reaction of the Na-blocked peptide with the oligodeoxy- ribonucleotide immobilised on CPG-500 was carried out in an- hydrous dimethylformamide in the presence of HBTU and ethyldiisopropylamine. The reaction was complete within 1 ± 2 h and gave high yields (80% ± 90%). In the final step, the conjugate was cleaved from the solid support and the functional groups of the oligonucleotide and the peptide components were depro- tected. The use of the solid-phase method requires large excesses of the peptides and the condensation reagent. This increases the efficiency of formation of the amide bond; the excesses of dimethylformamide-soluble reagents are easily removed by thorough washing of the support. The main advantage of Fmoc protection of amino groups is that this can be removed in two ways.The first approach entails post-synthetic treatment of the oligonucleotide with saturated aqueous ammonia. The second approach consists of the removal of the Fmoc group with morpholine in an anhydrous medium, which allows selective deprotection of the aliphatic amino group of the oligonucleotide. Under these conditions, the oligonucleo- tide is not split from the support, and its other functional groups remain protected. An approach proposed by Zubin et al.136 allows simultaneous attachment of several peptide fragments to the oligonucleotide, since the modified nucleoside fragment containing an aliphatic amino group can be inserted into any predetermined position of the oligomeric chain and the number of inserts is not limited in principle.Another salient advantage of this approach is that the 5 0- and 3 0-ends of the oligonucleotide fragment of the conjugate remain free, which allows the attachment of various reporter groups, e.g., radioactive or fluorescent labels. Certain difficulties arise when this method is used for the synthesis of oligonucleotide conjugates with larger peptides. The problem is associated with the activation of carboxy groups. Moreover, this method demands the use of large excesses of peptides, which is hardly justified from the economical point of view. The approach based on successive attachment of small peptide fragments to oligonucleotides is an alternative.After attachment of the first fragment, the Na-Fmoc group has to be removed, after which the next fragment is introduced into the reaction. Na-Fmoc-Leu-Gly and Na-Fmoc-Tyr-D-Ala-Phe-Gly were used for the synthesis of oligodeoxyribonucleotide conjugates with peptides containing four and eight amino acid residues. A similar ideology is shared by Hwang and Greenberg 137, 138 who have succeeded in incorporating 2 0-modified phosphor- amidite synthons into different positions of the oligonucleotide chain in the course of automated synthesis. The amino group at the C(2 0) atom of the synthon was protected by the photolabile protective group (X). 50 oligonucleotide HO Ura O O 30 oligonucleotide O CPG O NH X 64 O 50 oligonucleotide Ura O HO O 30 O oligonucleotide CPG O O X O 3 HN 65 OMe MeO .X= NO2262 Me CH FmocNH CH2 O peptide H2N peptide marker DIPEA is diisopropylethylamine. P O linker N O C C O2N O linker O P O R C Linker is7(CH2)67,7Gly7NHCO(CH2)67,7Gly-Gly-Gly7NHCO(CH2)67; (a) EtNPri2; (b) RH (RH is H-Gly-Gly-Gly-NEt2 or H-Leu-Ala-Lys(Tfa)-Leu-NEt2); (c) NH3 . aq. The attachment of tripeptides to the modified oligonucleo- tides 64 and 65 afforded the corresponding conjugates with satisfactory (70% ± 80%) yields. The synthesis of photodegradable peptide ± oligonucleotide conjugates designed for detection of target NA sequences by the MALDI mass spectrometry methods has been described (Scheme 27).139 The photodegradable conjugate contains an oligonucleotide probe for target DNA or RNA sequences and a peptide mass marker easily detectable by mass spectrometry after hybridisation and subsequent photodegradation of the conjugate.A special polymer with ketoxime functional groups was proposed for oligonucleotide synthesis (Scheme 28).140, 141 The anchor group (linker) was attached to the support through an ester bond. The use of such supports facilitates the cleavage of the target product under the action of nucleophiles, the role of a nucleophile can be played by the peptide fragment. IV. Conclusion The analysis of published data suggests that no versatile method for the synthesis of peptide ± oligonucleotide conjugates of any structure exists.This is due to the fact that several factors play crucial roles in the synthesis of peptide ± oligonucleotide conju- gates, viz., the type of modified nucleic acid fragment, the amino acid sequence, the size of the peptide, the conjugation procedure and the type of bond connecting the two parts of the hybrid molecule. The possibility of varying these parameters within a sufficiently broad range and of using their different combinations have given a strong impetus to the appearance of numerous publications devoted to novel methods for the synthesis of peptide ± oligonucleotide conjugates. However, because of serious limitations of each of these methods, very few of them have found O 50 30 oligonucleotide O P O O(CH2)2CN NO2 O Me O P O CH O(CH2)2CN NH NO2 CH2 50 photodegradable linker oligonucleotide probe P Oligonucleotide synthesis ODMTr O2N O oligonucleotide ODMTr O7 CPG 1) Tetramethylguanidine 2) FmocNH7peptide7COOH, PyBOP7HOBT7DIPEA 3) NH3 .aq 50 30 oligonucleotide 30 O linker N O C C NC(CH2)2O wide acceptance in the preparative synthesis of peptide ± oligonu- cleotide conjugates. In this connection, a search for optimal approaches to the synthesis of peptide ± oligonucleotide conju- gates still remains a very urgent task. The authors thank A V Kachalova for her valuable assistance in the preparation of this review. This work has been carried out within the framework of the program `Leading scientific schools' (Grant No.00-15-97944) and `The universities of Russia' (Grant No. 015.07.02.027) and with the financial support of the Grant `Wellcome Trust' (Grant No. 057361). References 1. A A Bogdanov Usp. Sovr. Biol. 55 321 (1963) 2. B A Yuodka Bioorg. Khim. 6 1445 (1980) a 3. B A Yuodka Kovalentnye Nukleinovo-Belkovye Struktury i ikh 5. N I Sokolova, G I Gurova, Z A Shabarova, M A Prokof'ev Vestn. Khimicheskoe Modelirovanie (Covalent Nucleic Protein Structures and Their Chemical Modelling) (Vil'nyus: Mokslas, 1985) 4. Z A Shabarova, in Progress in Nucleic Acid Researches and Molecular Biology (Eds J N Davidson, W E Cohn) (New York: Academic Press, 1970) Mosk. Univ., Ser. 2, Khimiya 24 104 (1969) b 6. V D Smirnov, T N Bocharova, Z A Shabarova, M A Prokof'ev Vestn.Mosk. Univ., Ser. 2, Khimiya 27 3 (1972) b 7. J Goodchild Bioconjugate Chem. 1 165 (1990) 8. E Uhlmann, A Peyman Chem. Rev. 90 543 (1990) 9. S L Beaucage, R P Iyer Tetrahedron 49 10441 (1993) 10. M Manoharan, in Antisense Research and Applications (Eds S L Crooke, B Lebleu) (Boca Raton, FL: CRC Press, 1993) 11. D A Stetsenko, A A Arzumanov, V A Korshun,M J Gait Mol. Biol. 34 998 (2000) c E M Zubin, E A Romanova, T S Oretskaya Scheme 27 CPG Scheme 28 O a, b, c protected oligonuc- O P O ODMTr leotideModern methods for the synthesis of peptide ± oligonucleotide conjugates 12. C-H Tung, S Stein Bioconjugate Chem. 11 605 (2000) 13. E SchroÈ der, K LuÈ bke The Peptides (New York; London: Wiley, 1967) 14.J C Wang Ann. Rev. Biochem. 54 665 (1985) 15. C Schattenkerk, C T J Wreesmann, M J de Graaf, G A van der Marel, J H van Boom Tetrahedron Lett. 25 5197 (1984) 16. E Kuyl-Yeheskiely, P A M van der Klein, G A Visser, G A van der Marel, J H van Boom Recl. Trav. Chim. Pays-Bas 105 69 (1986) 17. E Kuyl-Yeheskiely, C M Tromp, A W M Lefeber, G A van der Marel, J H van Boom Tetrahedron 44 6515 (1988) 18. E Kuyl-Yeheskiely, C M Dreef-Tromp, A Geluk, G A van der Marel, J H van Boom Nucl. Acids Res. 17 2897 (1989) 19. H Hotoda, Y Ueno, M Sekine, T Hata Tetrahedron Lett. 30 2117 (1989) 20. Y Ueno, R Saito, T Hata Nucl. Acids Res. 21 4451 (1993) 21. Z A Shabarova, A A Bogdanov Khimiya Nukleinovykh Kislot i ikh Komponentov (The Chemistry of Nucleic Acids and Their Components) (Moscow: Khimiya, 1978) 22.V F Zarytova, E M Ivanova, S N Yarmolyuk, I V Alekseeva Biopolim. Kletka 4 220 (1988) 23. D V Pyshnyi,M N Repkova, S G Lokhov, E M Ivanova, A G Ven'yaminova, V F Zarytova Bioorg. Khim. 23 497 (1997) a 24. A N Sinyakov, S G Lokhov, I V Kutyavin, H B Gamper, R B Meyer J. Am. Chem. Soc. 117 4995 (1995) 25. A S Levina, V G Metelev, A S Cohen, P C Zamecnik Antisense and Nucl. Acid Drug Dev. 6 75 (1996) 26. V F Zarytova, in Bioorganicheskaya Khimiya (Itogi Nauki i Tekhniki) [Bioorganic Chemistry (Advances in Science and Engineering Series)]; article deposited at the VINITI, Academy of Sciences of the USSR, Moscow, 1984, p. 85 27. M G Ivanovskaya,M B Gottih, Z A Shabarova Nucleosides, Nucleotides 6 913 (1987) 28.Z A Shabarova, M G Ivanovskaya,M B Gottih, in Nucleic Acids Chemistry Improved and New Synthetic Procedure. Methods and Techniques (Eds L B Townsend, R S Tipson) (London: Wiley, 1991) p. 386 29. M B Gottikh,M G Ivanovskaya, E A Skripkin, Z A Shabarova Bioorg. Khim. 16 514 (1990) a 30. E A Taran,M G Ivanovskaya,M J Gait, Z A Shabarova Mol. Biol. 32 832 (1998) c 31. S Kuznetsova, S Ait-Si-Ali, I Nagibneva, F Troalen, J-P Le Villain, A Harel-Bellan, F Svinarchuk Nucl. Acids Res. 27 3995 (1999) 32. S A Kuznetsova, N V Sumbatyan, C Malvy, J-R Bertrand, A Harel-Bellan, G A Korshunova, F P Svinarchuk Vestn. Mosk. Univ., Ser. 2, Khimiya 42 281 (2001) b 33. D L McMinn, T J Matray, M M Greenberg J. Org. Chem. 62 7074 (1997) 34.D L McMinn, M M Greenberg Bioorg. Med. Chem. Lett. 9 547 (1999) 35. J D Kahl, D L McMinn, M M Greenberg J. Org. Chem. 63 4870 (1998) 36. J D Kahl, M M Greenberg J. Org. Chem. 64 507 (1999) 37. O N Jensen, S Kulkarni, J V Aldrich, D F Barofsky Nucl. Acids Res. 24 3866 (1996) 38. E A Lukhtanov, I V Kutyavin, H B Gamper, R B Meyer Bioconjugate Chem. 6 418 (1995) 39. C Guibourdenche, D Seebach Helv. Chim. Acta. 80 1 (1997) 40. R K Bruick, P E Dawson, S B H Kent, N Usman, G F Joyce Chem. Biol. 3 49 (1996) 41. D L McMinn, M M Greenberg Tetrahedron 52 3827 (1996) 42. D L McMinn, M M Greenberg J. Am. Chem. Soc. 120 3289 (1998) 43. M McPherson, M C Wright, P A Lohse Synlett S1 978 (1999) 44. P E Dawson, T W Muir, I Clark-Lewis, S B H Kent Science 266 776 (1994) 45.D A Stetsenko,M J Gait J. Org. Chem. 65 4900 (2000) 46. D A Stetsenko,M J Gait Nucleosides, Nucleotides 19 1751 (2000) 47. N D Sinha, R M Cook Nucl. Acids Res. 16 2659 (1988) 48. K Arar,M Monsigny, R Mayer Tetrahedron Lett. 34 8087 (1993) 49. C Pichon, K Arar, A J Stewart,M D Dodon, L Gazzolo, P J Courtoy, R Mayer,M Monsigny, A-C Roche Mol. Pharm. 51 431 (1997) 50. C Neves, G Byk, D Scherman, P Wils FEBS Lett. 453 41 (1999) 263 51. R Eritja, A Pons, M Escarceller, E Giralt, F Albericio Tetrahedron 47 4113 (1991) 52. B G de la Torre, A M Avino, M Escarceller,M Royo, F Albericio, R Eritja Nucleosides, Nucleotides 12 993 (1993) 53. B G de la Torre, F Albericio, E Saison-Behmoaras, A Bachi, R Eritja Bioconjugate Chem.10 1005 (1999) 54. D Gottschling, H Seliger, G Tarrason, J Piulats, R Eritja Bioconjugate Chem. 9 831 (1998) 55. N J Ede, G W Tregear, J Haralambidis Bioconjugate Chem. 5 373 (1994) 56. S Soukchareun, J Haralambidis, G W Tregear Bioconjugate Chem. 9 466 (1998) 57. J J Hangeland, J T Levis, Y C Lee, P O P Ts'o Bioconjugate Chem. 6 695 (1995) 58. W Mier, R Eritja, A Mohammed, U Haberkorn, M Eisenhut Bioconjugate Chem. 11 855 (2000) 59. C-H Tung,M J Rudolph, S Stein Bioconjugate Chem. 2 464 (1991) 60. T Zhu, Z Wei, C-H Tung,W A Dickerhof, J Breslauer, D E Georgopoulos,M J Leibowitz, S Stein Antisense Res. Dev. 3 265 (1993) 61. J-P Bongartz, A-M Aubertin, P G Milhaud, B Lebleu Nucl. Acids Res. 22 4681 (1994) 62. J G Harrison, S Balasubramanian Nucl.Acids Res. 26 3136 (1998) 63. M A Zanta, P Belguise-Valladier, J-P Behr Proc. Natl. Acad. Sci. USA 96 91 (1999) 64. Z Wei, C-H Tung, T Zhu, S Stein Bioconjugate Chem. 5 468 (1994) 65. T Zhu, C-H Tung, K J Breslauer, W A Dickerhof, S Stein Antisense Res. Dev. 3 349 (1993) 66. T Zhu, S Stein Bioconjugate Chem. 5 312 (1994) 67. S Stein, T Zhu Methods Enzymol. 280 51 (1997) 68. M W Reed, D Fraga, D E Schwartz, J Scholler, R D Hinrichsen Bioconjugate Chem. 6 101 (1995) 69. K Arar, A-M Aubertin, A-C Roche, M Monsigny, R Mayer Bioconjugate Chem. 6 573 (1995) 70. L Meunier, R Mayer,M Monsigny, A-C Roche Nucl. Acids Res. 27 2730 (1999) 71. L Meunier, S Bourgerie, R Mayer, A-C Roche, M Monsigny Bioconjugate Chem. 10 206 (1999) 72.Y Aubert, S Bourgerie, L Meunier, R Mayer, A-C Roche, M Monsigny, N T Thuong, U Asseline Nucl. Acids Res. 28 818 (2000) 73. Y Lin, A Padmapriya, K M Morden, S D Jayasena Proc. Natl. Acad. Sci. USA 92 11 044 (1995) 74. M Antopolsky, E Azhayeva, U Tengvall, S Auriola, I JaÈ aÈ skelaÈ inen, S RoÈ nkkoÈ , P Honkakoski, A Urtti, H LoÈ nnberg, A Azhayev Bioconjugate Chem. 10 598 (1999) 75. D R Corey J. Am. Chem. Soc. 117 9373 (1995) 76. E ViveÁ s, B Lebleu Tetrahedron Lett. 38 1183 (1997) 77. G Sengle, A Jenne, P S Arora, B Seelig, J S Nowick, A Jaschka, M Famulok Bioorg. Med. Chem. 8 1317 (2000) 78. Z Kupihar, Z Schmel, Z Kele, B Penke, L Kovacs Bioorg. Med. Chem. 9 1241 (2001) 79. T Ishihara, D R Corey Nucl. Acids Symp. Ser. 141 (1999) 80. T Ishihara, D R Corey J. Am. Chem. Soc. 121 2012 (1999) 81. A Astriab-Fisher, D S Sergueev,M Fisher, B R Shaw, R L Juliano Biochem. Pharmacol. 60 83 (2000) 82. O M Gritsenko, E S Gromova Usp. Khim. 68 267 (1999) [Russ. Chem. Rev. 68 241 (1999)] 83. B S Ermolinskii, S N Mikhailov Bioorg. Khim. 26 483 (2000) a 84. B Bayard, C Bisbal, B Lebleu Biochemistry 25 3730 (1986) 85. M Lemaitre, B Bayard, B Lebleu Proc. Natl. Acad. Sci. USA 84 648 (1987) 86. M Lemaitre, C Bisbal, B Bayard, B Lebleu Nucleosides, Nucleotides 6 311 (1987) 87. J P Leonetti, B Rayner,M Lemaitre, C Gagnor, P G Milhaud, J-L Imbach, B Lebleu Gene 72 323 (1988) 88. G Degols, J-P Leonetti, M Benkirane, C Devaux, B Lebleu Antisense Res. Dev. 2 293 (1992) 89. G Degols, C Devaux, B Lebleu Bioconjugate Chem. 5 8 (1994) 90. M Manoharan, L K Andrade, V Mohan, S M Freier, P Dan Cook Nucleosides, Nucleotides 16 1741 (1997) 91. D Forget, D Boturyn, E Defrancq, J Lhomme, P Dumy Chem. Eur. J. 7 3976 (2001)E M Zubin, E A Romanova, T S Oretskaya 264 96. J Haralambidis, L Duncan, K Angus, G W Tregear Nucl. Acids 136. E M Zubin, E A Romanova, E M Volkov, V N Tashlitsky, G A Korshunova, Z A Shabarova, T S Oretskaya FEBS Lett. 456 59 (1999) 137. J-T Hwang, M M Greenberg Org. Lett. 1 2021 (1999) 138. J-T Hwang, M M Greenberg J. Org. Chem. 66 363 (2001) 139. J Olejnik, H-C Ludemann, E Krzymanska-Olejnik, S Berkenkamp, K J Rothschild Nucl. Acids Res. 27 4626 (1999) 140. M Fujii, T Hasegava, I Koujima Nucl. Acids Symp. Ser. 37 71 (1997) 141. M Fujii, T Hasegava, I Koujima Nucleosides, Nucleotides 18 1487 (1999) 92. D Forget, O Renaudet, D Boturyn, E Defrancq, P Dumy Tetrahedron Lett. 42 91 (2001) 93. B Cebon, J N Lambert, D Leung, H Mackie, K L McCluskey, X Nguyen, C Tassone Aust. J. Chem. 53 333 (2000) 94. S Dey, T L Sheppard Org. Lett. 3 3983 (2001) 95. J Haralambidis, L Duncan, G W Tregear Tetrahedron Lett. 28 5199 (1987) Res. 18 493 (1990) 97. S Soukchareun, G W Tregear, J Haralambidis Bioconjugate Chem. 6 43 (1995) 98. N Guzzo-Pernell, G W Tregear Aust. J. Chem. 53 699 (2000) 99. E A Lukhtanov, I V Kutyavin, R B Meyer Bioconjugate Chem. 7 a�Russ. J. Bioorg. Chem. (Engl. Transl.) b�Moscow Univ. Bull. (Engl. Transl.) c�Mol. Biol. (Engl. Transl.) 564 (1996) 100. J-C Truffert, O Lorthioir, U Asseline, N T Thuong, A Brack Tetrahedron Lett. 35 2353 (1994) 101. J-C Truffert, U Asseline, N T Thuong, A Brack Protein Peptide Lett. 2 419 (1995) 102. J-C Truffert, U Asseline, A Brack, N T Thuong Tetrahedron 52 3005 (1996) 103. B G de la Torre, A AvinÄ o', G Tarrason, J Piulats, F Albericio, R Eritja Tetrahedron Lett. 35 2733 (1994) 104. V Marchan, L Debethune, M Beltran, J Robles, I Travesset, G Fabregas, E Pedroso, A Grandas Nucleosides, Nucleotides 18 1493 (1999) 105. J Robles,M Beltra'n, V Marcha'n, Y Pe'rez, I Travesset, E Pedroso, A Grandas Tetrahedron 55 13 251 (1999) 106. J Robles, E Pedroso, A Grandas Tetrahedron Lett. 35 4449 (1994) 107. J Robles, E Pedroso, A Grandas J. Org. Chem. 59 2482 (1994) 108. J Robles, E Pedroso, A Grandas Nucl. Acids Res. 23 4151 (1995) 109. J Robles, M Maseda,M Beltran, M Concernau, E Pedroso, A Grandas Bioconjugate Chem. 8 785 (1997) 110. M Beltra'n, E Pedroso, A Grandas Tetrahedron Lett. 39 4115 (1998) 111. J Robles, E Pedroso, A Grandas J. Org. Chem. 60 4856 (1995) 112. M Beltran, M Maseda, Y Perez, J Robles, E Pedroso, A Grandas Nucleosides, Nucleotides 16 1487 (1997) 113. M Beltran, M Maseda, J Robles, E Pedroso, A Grandas Lett. Peptide Sci. 4 147 (1997) 114. F Bergmann,W Bannwarth Tetrahedron Lett. 36 1839 (1995) 115. C N Tetzlaff, I Schwope, C F Bleczinski, J A Steinberg, C Richert Tetrahedron Lett. 39 4215 (1998) 116. I Schwope, C F Bleczinski, C Richert J. Org. Chem. 64 4749 (1999) 117. C F Bleczinski, C Richter Org. Lett. 2 1697 (2000) 118. C D Juby, C D Richardson, R Brousseau Tetrahedron Lett. 32 879 (1991) 119. S Basu, E Wickstrom Tetrahedron Lett. 36 4943 (1995) 120. T Y-K Chow, C Juby,R Brousseau Antisense Res. Dev. 4 81 (1994) 121. J Nielsen, S Brenner,K D Janda J. Am. Chem. Soc. 115 9812 (1993) 122. L De Napoli, A Messere, D Montesarchio, G Piccialli, E Benedetti, E Bucci, F Rossi Bioorg. Med. Chem. 7 395 (1999) 123. M Antopolsky, A Azhayev Helv. Chim. Acta 82 2130 (1999) 124. M Antopolsky, A Azhayev Tetrahedron Lett. 41 9113 (2000) 125. M Antopolsky, E Azhaeva, U Tengvall, A Azhayev Tetrahedron Lett. 43 527 (2002) 126. C M Dreef-Tromp, E M A van Dam, H van den Elst, G A van der Marel, J H van Boom Nucl. Acids Res. 18 6491 (1990) 127. C M Dreef-Tromp, H van den Elst, J E van der Boogaart, G A van der Marel, J H van Boom Nucl. Acids Res. 20 2435 (1992) 128. C M Dreef-Tromp, J C M van der Maarel, H van den Elst, G A van der Marel, J H van Boom Nucl. Acids Res. 20 4015 (1992) 129. A Sakakura, Y Hayakawa, H Harada, M Hirose, R Noyori Tetrahedron Lett. 40 4359 (1999) 130. Y Hayakawa, A Sakakura, S Heidenhain,M Kataoka Collect. Symp. Ser. 2 105 (1999) 131. A Sakakura, Y Hayakawa Tetrahedron 56 4427 (2000) 132. J Robles, E Pedroso, A Grandas Tetrahedron Lett. 32 4389 (1991) 133. A Grandas, J Robles, E Pedroso Nucleosides, Nucleotides 14 825 134. S Peyrottes, B Mestre, F Burlina,M J Gait Tetrahedron 54 12 513 135. S Peyrottes, B Mestre, F Burlina,M J Gait Nucleosides, (1995) (1998) Nucleotides 18 1443

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC