摘要:

Russian Chemical Reviews 68 (2) 85 ± 102 (1999) Chemistry on the border of two centuries�achievements and prospects { A L Buchachenko Contents I. Introduction II. Novel chemical structures and materials III. Chemistry in micro- and macro-reactors IV. Coherent chemistry V. Spin chemistry and chemical radiophysics VI. Chemistry under extreme and exotic conditions VII. Once again about cold fusion VIII. Physics of chemical reactions IX. Atomic-resolution spectroscopy and chemistry X. Conclusion Abstract. Chemistry is on the threshold of a new flight; new ideas and lines of research are originating now; new directions arise, which form a new face of chemistry. The bibliography includes 79 references. I. Introduction Chemistry ultimately formed as a fundamental science only in the early XXth century when it got a reliable and solid physical basis made of three main concepts of quantum mechanics, namely � the SchroÈ dinger equation, which is the quantum successor of the classical mechanics equations (Hamilton ± Jacobi equa- tions); � Pauli's exclusion principle, which governs the distribution of electrons among spin states and energy levels and �wave function, which carries information on the charge and spin distribution.It is these concepts that have imparted physical meaning to the D. I. Mendeleev Periodic Table, an outstanding discovery of the XIX century the significance of which goes far beyond chemistry. From the viewpoint of these three concepts, a chemical reaction should be regarded as a physical process involving rearrangement of electron shells and nuclei.The understanding and realisation of the significance of these quantum-mechanics principles for chemistry makes chemical science clear and predictable as regards its main features; they give rise to the diversity and richness as well as to the elegant logic and beauty of chemistry. In the XX century chemistry became an exact science; many quantitative regularities and precise laws have been established (including the `electron periodicity' of the Mendeleev law); a very high metrological level of determination of atomic, molecular, A L Buchachenko N N Semenov Institute of Chemical Physics, Russian Academy of Sciences, ul. Kosygina 4, 117977 Moscow, Russian Federation.Fax (7-095) 938 24 84. Tel. (7-095) 939 74 90 Received 26 October 1998 Uspekhi Khimii 68 (2) 99 ± 118 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 54 85 86 89 91 94 96 98 98 100 101 thermodynamic and kinetic constants, characterising both sub- stances and chemical processes, has been attained. During this century, chemistry has developed into a branched science. Nowadays, many fields of chemistry exist as separate branches of science�inorganic, organic and physical chemistry; radiochemistry, biochemistry, analytical chemistry, geochemistry, etc. Each of them has its own subject and the scope of inves- tigations as well as its own problems and experimental techniques. However, by the 80s, the professional separation of the chemistry branches has started to be replaced by attempts to tackle jointly the general fundamental problems of chemical science.The first attempt to identify the most important `integration' lines in chemistry was made in a paper by Legasov and Bucha- chenko,1 in which chemistry was classified at a new level. The chemical science was structured according to the targets and problems rather than according to the names of various `chem- istries,' which now number more than four dozen. This classifica- tion follows the logic inherent in chemistry, which does not separate chemistry into `chemical counties' but instead organises it as a unified science and also unites chemists as a community.Nowadays, at the end of the XXth century, the classification proposed in the publication mentioned above 1 has not changed much, and the hierarchy of general problems of chemistry can be presented as follows. �the art of chemical synthesis; �chemical structure and function; �control of chemical processes; �chemical materials science; �chemical technology; �chemical energetics; �chemical analytics and diagnostics; �chemistry of living matter. These are the main `arterial roads' which direct the develop- ment of modern chemistry and reflect its achievements and progress. { This paper is an extended version of the report presented by the author at the General Meeting of the Division of General and Technical Chemistry of the Russian Academy of Sciences and at the XVIth Mendeleev Congress on General and Applied Chemistry (St.Petersburg, 1998).86 Chemical synthesis is the key direction in the development of chemistry and the source of all its treasures. This route makes chemistry the most constructive branch of science. Chemistry provides other branches of science and industry with materials; in this sense, one can say that it stands in the middle of natural sciences. It is especially intriguing that, despite the existence of theoretical principles of chemical synthesis, there still remains scope for the freak of imagination and intuition. This brings chemical synthesis closer to an art. The atomic and molecular architecture and the electronic structures of newly synthesised compounds are infinitely diverse, so are their physical and chemical properties and, therefore, functions.The relationship between the structure and functional behaviour of a substance is the subject of the second line of research. The control of chemical processes and their molecular mecha- nisms and the use of chemical factors (complex formation, solvation, molecular organisation, catalysis) and physical treat- ment (ranging from light to mechanical treatment) for controlling chemical processes constitute the essence of the third line. A substance is not a material yet but only the precursor of a material. People have to `teach' substances to work as materials and to estimate their operation characteristics and ranges of applicability.These are the tasks of the chemical materials science. The aim of chemical technology is to design, optimise and scale a technological process as well as to ensure its low energy expenditures and safety and to make it environmentally clean. The development of efficient ways of transformation of chemical energy into other kinds of energy, accumulation of energy in energy-storing substances and materials (including lasers with chemical and solar pumping), transformation of solar energy, chemical cells and coupling of energy-producing and energy-consuming processes constitute the subject of chemical energetics. The progress in chemical materials science and chemical technology would be impossible without reliable chemical ana- lytics and diagnostics.This is a vigorously developing line of research (which includes chemical sensors and the chemistry of odour) having links to lots of branches, ranging from technical monitoring systems to medicine or ecology. There is no question that all these lines of research are related not only logically. They are intrinsically connected by the method- ology of chemical research itself�elements of several lines can be found in a good scientific article. The remarkable combination of differentiation and integration makes up the efficient and creative style of modern chemistry. Finally, the chemistry of living matter is an enormous chemical galaxy, which is still to be explored. It assimilates the results of biochemistry, chemistry of natural products, phytochemistry, enzyme science, medical and pharmaceutical chemistry, genetic engineering, biotechnology and many other fields of chemistry.This line of research is associated with great expectations; it undoubtedly possesses an exceptionally high potential, indisput- able prospects and a highly promising future. Even today, its outlines and significance can be traced by looking at the transgene technology. II. Novel chemical structures and materials Each arterial road of chemistry has its own events and discoveries, which can be both regular and unexpected. However, the most significant events still occur in theevelopment of methods for the synthesis and design of new substances and materials are the main goals of chemistry.It should be noted that the number of compounds that have been created chemically (artificially and skilfully) is permanently increasing. The molecular architecture of newly formed com- pounds is infinitely variable and fabulously abundant. For exam- ple, the following unusual structures have been obtained: rhombic molecules � structural units of one-dimensional metals; proton A L Buchachenko `sponges' and `tubes' (molecularly organised proton-carrying reservoirs and channels); toroidal molecules; polymetal rotaxanes; catenanes; crowns and anticrowns, capable of separating cations and anions; hypervalent radicals such asNH4, CH5, H3O, some of which are long-lived species, for example, (CH2)2NH2 (lifetime *3 ms);2 high-spin molecules containing tens of unpaired elec- trons within a single species [for example, the complex Mn2á 6 (R2NO.)6 has 36 unpaired electrons]; multi-deck polyaro- matic molecules; molecules with large numbers of chiral centres, etc.Some of the compounds obtained artificially may remain exotic species for good, but some others would trigger new lines of research. The adoption of new principles of radial synthesis in which molecules unite according to the fractal pattern to give an enormous molecule, dendrimer, has been an important event in chemistry. Nature has used this principle to produce glycogen, amylopectin and some other polysaccharides and proteins. Many remarkable discoveries have been made in the chemistry of dendrimers, especially in the polymer chemistry.3 The architec- tures of dendrimer structures are virtually infinite; the same is true for their functions and applications.In particular, it has been predicted that dendrimers based on polyconjugated molecules could serve as energetic antennas collecting the energy of optical (including solar) excitation and transforming it into photoelectric current. Impressive results have been attained in the synthesis of organometallic polymers. The preparation of ladder, tape and rod-like organometallic polymers can be formally classified as `one-dimensional' or `two-dimensional' radial synthesis. Thus Fig. 1 shows two projections of an unusual organometallic ladder polymer [(CuL2)PF6]?, prepared by polymerisation of 2,7-diaza- pyrene (L) and [Cu(MeCN)4]PF6 in theMeCN±PhCN system.4 It can be seen from the Figure that a monomer unit of this polymer is ab Figure 1.Two projections of the organometallic ladder-type polymer [(CuL2)PF6]? (L=2,7-pyrene) with polycatenane and adamantane struc- tural units.4Chemistry on the border of two centuries �achievements and prospects formed by three interpenetrating adamantane networks arranged at angles of 90 8 relative to one another. Examples of three- dimensional organometallic dendrimers have been reported (see below). The discovery of superconducting ceramics has stirred up not only physicists and engineers but also chemists. These compounds had been prepared well before their superconducting properties were discovered, but it was this intriguing and unexpected feature of ceramics that aroused most enthusiasm and hope. The achieve- ments in this field proved to be modest compared to the expect- ations; although the progress along this line has now substantially slowed, some pleasant surprises are still possible (note the startling publication 5 dealing with superconductivity of inorganic compo- sites at room temperature, the value of which is still unclear).It is about this type of situation that Niels Bohr said that problems are more important than solutions. In fact, solutions can become obsolete to leave the problems, which stimulate the search for new solutions. After a great number of paramagnetic organic compounds (stable radicals and high-spin molecules) had been synthesised, attempts to prepare molecular ferromagnets were undertaken.The principle of design of organic ferromagnets had been formulated long before they were discovered in 1989.6, 7 The exchange interaction of the unpaired electrons in paramagnetic organic molecules is known to be negative; it accounts for the antiparallel arrangement of the magnetic moments of the electrons. It has been suggested that macroscopic ferromagnetism in ferromagnets is attained via pairwise antiferromagnetic interactions. This hypoth- esis immediately received experimental support.8 Figure 2 shows the crystal structure of 2-(4-nitrophenyl)- 4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazolyl-N0-oxyl N-oxide. c a 0 Figure 2.Two-dimensional representation of the crystal lattice of 2-(4-nitrophenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazolyl-N0-oxyl N-oxide. The NO. groups (areas with high spin density) are coloured black, while the NO2 groups (areas with small negative spin density) are designated by crosses.8 87 ON NO2 N Me Me Me Me O It can be seen that the nitroxides are arranged in the crystal in such a way that fragments with high spin density (NO. groups) are located close to fragments where the spin density is low and negative (NO2 groups). Owing to the pairwise exchange inter- action of the neighbouring spins with positive and negative electron density, spins with positive density align in one direction, while those with negative density align in the opposite direction; this creates macroscopic ferromagnetism in the crystal.This is only one of the numerous examples of organic ferromagnets.9 One should not, however, create illusions about this class of compound. The specific spin density and the degree of magnet- isation in these substances are very low; correspondingly, the temperatures of ferromagnetic phase transitions are also very low (*1 K). Mixed organometallic ferromagnets are more promising (and therefore more attractive). Figure 3 presents the structure of an organometallic ferromagnet, {[FeCr(ox)3]. .(NBu4)}? (ox is oxalate anion). The structure consists of the Cr(ox)3 fragments being joined together by the Fe ions to form a huge ferromagnetic molecule.In the crystal, the compound {[FeCr(ox)3](NBu4)}?has a layered structure.10, 11 Fe O Fe O O O O O O O O O Cr Fe OO Cr OO O O O O O O O O O Fe Fe O organometallic the of Figure 3. Structure ferromagnet {[FeCr(ox)3](NBu4)}? (Tc=12 K). For the sake of simplicity, the tetra- butylammonium cation is not shown.10 Fullerenes have become a real treasure for chemistry. The most daring and optimistic forecasts are associated with fullerenes. FullerenesC60, C70, etc. can be prepared in macroscopic quantities by the arc method. However, the mechanism of their formation still remains obscure. The perception of this fact stimulates experimental and theoretical (molecular dynamics) studies of the formation of fullerenes from atoms, pairs of atoms and clusters under the conditions of gas discharge, plasma, electric arc, etc.The main goal of these studies is to understand the mechanism of polyhedral clustering and the arrangement of atoms to give a fullerene structure (see, for example, Ref. 12). The search for catalysts of this process is also perfectly reasonable. It is beyond doubt that this route would bring both predicted discoveries and godsends. At present, in addition to traditional chemistry, targeted chemistry of these compounds is in progress. In particular, it is of interest to perform polymerisation of fullerenes with retention of the spherical cage units (via cross-linking). According to predictions, these polymers would possess high electron `capacity,' which would permit them to be used in chemical and storage cells. Thermal polymerisation of the most readily avail- able C60 fullerene is known to be forbidden from the symmetry viewpoint.Therefore, it requires some sort of treatment which would stimulate distortion of the icosahedral symmetry ofC60 and88 eliminate the prohibition (high pressure,{ shear deformations, shock waves, etc.). This is a key problem. Among other lines of research in the targeted chemistry of fullerenes, mention should be made of fullerene modification aimed at producing materials for microelectronics (the easy construction of integral circuisimple and clean nano-technolo- gies and other advantages) and synthesis of fullerene-based organometallic and organic complexes to be used as supercon- ductors.Thus it was found that M3C60 and M2M0C60 fullerides (M,M0= K, Na, Rb, Cs) exhibit superconducting properties. The compound RbCs2C60 still remains the record holder, Tc = 33 K. A remarkable feature of the superconducting fullerides is the dependence of the temperature of superconductivity transition Tc on the crystal lattice parameter (Fig. 4).13 This dependence proved to be opposite to that observed for organic superconduc- tors based on tetrathiafulvalene and its derivatives, which were famous in the 60s ± 70s (for details, see Ref. 1). In the latter case, compression of the crystal lattice by external pressure results in higher Tc values; conversely, in the case of fullerides, compression decreases Tc.This unusual behaviour has stimulated the search for new superconducting fullerides based on the expansion of the fullerene crystal lattice (by intercalation by NaN3 and NaH, introduction of NH3, etc). In particular, the compound (NaH)4C60 obtained by intercalation of C60 by NaH exists as two crystal phases, cubic and orthorhombic. The cubic b-phase conducts electricity, while the orthorhombic a-phase does not.14 Within the framework of the same idea (crystal lattice expansion), synthesis of charge transfer complexes based of fullerenes and tetrathiafulvalene and its derivatives (the latter served as `bricks' in the construction of organic superconductors) is being devel- oped.Tc /K 30 20 100 14.2 14.4 14.0 a /A 13.8 Figure 4.Dependence of the superconductivity transition temperature Tc on the crystal lattice parameter a in the series of fulleridesM3C60 with a face-centered cubic structure.13 The mechanism of superconductivity still remains an intrigu- ing question. Does the consideration of this problem go beyond the canonical electron pair theory, according to which the move- ment of electron pairs is governed by the lattice phonon spectrum (the Bardeen, Cooper, Shrieffer�BCS�theory)? This question has stimulated the synthesis of 13C60 fullerene (with 99% replace- ment of 12Cby 13C) and the corresponding fullerides,K312C60 and K313C60.15 In qualitative agreement with the BCS theory, the superconduction point decreased from 19.2 to 18.8 K on passing from K312C60 to K313C60.This important result confirms partic- { The report that at high pressures and temperatures the hardness of the fullerene polymerisation product exceeds that of diamond has not been supported. A L Buchachenko ipation of phonons in the formation and movement of the Cooper pairs. According to the canonical BCS theory, the superconduction point Tc ! M7a, where M is the mass of the particle responsible for the phonon spectrum (in this particular case, 12C60 and 13C60), and a is a characteristic parameter, equal to 0.5. Taking account of the isotope effect and proceeding from the value Tc = 18.8 K, one can easily find that a=0.3 (rather than 0.5 as predicted by the BCS theory) for the fulleride K312C60.This could prompt the conclusion that the BCS theory requires modification. However, this conclusion is not ultimate. In particular, there are no grounds for claiming that the phonons responsible for the Cooper pairing are generated by vibrations of C60 as a whole; the pairing might be due to the part of the phonon spectrum created by internal vibrations of C60 (for example, the breathing vibration of the sphere). In this case, a mass smaller than the mass of C60 should be substituted into the expression Tc!M7a and thus the resulting a would move towards 0.5. These conclusions are consistent with the results obtained for Rb3C60 with partial substitution of 13C atoms for 12C.The a value found for this fullerene (0.37) is closer to 0.5. Endofullerenes constitute one of the most exotic and remark- able classes of chemical compounds created by chemistry. They are formed upon insertion and capture of atoms or ions inside the fullerene sphere. However, it is not the insertion into the fullerene sphere itself that makes these compounds unique but rather the fact that capture of these species by fullerenes is a virtually irreversible process (unlike, for example, the capture of species by cyclodextrins, which is reversible). A great number of endofullerenes have already been synthes- ised: T@C60 (tritium atoms are formed from 6Li ions on irradi- ation with neutrons in the presence of C60), (3He,4He)@C60, K@C59B, La@(C60,C70,C74,C82), etc.; endofullerenes with two or even three captured atoms have been obtained, for example, He2@C70, La2 @(C72,C74,C80), (Gd2,Y2)@C90, (Sc2,Er2)@C82, Sc3@C82, etc.16 It should be noted that in the synthesis of La2@(C72,C74) in the arc of graphite implanted with La2O3, only endofullerenes La2@C72 and La2@C74 are intensively formed, no `empty' C72 and C74 being observed.This fact may prove significant from the viewpoint of catalysis of fullerene synthesis. Some answers to the question of what is the mechanism of the endofullerene synthesis have already been found. As was to be expected, the early idea that the carbon shell is permeable for high- energy atoms proved to be naive. This idea might (although not for sure) be valid in the case of hydrogen atoms; however, even for He direct insertion into C60 (without violation of the cage structure) would require *10 eV, and He+ would be able to penetrate through a six-membered ring upon an orthogonal collision only if it has an energy of 35 ± 45 eV.This is much larger than the energy required for dissociation of the C±C bonds in the fullerene shell (70 ± 80 kcal mol71). It is clear that rupture of the C±C bonds to open a `window' in the shell followed by insertion of the guest atom and closure of the `window' upon restoration of the C±C bonds is energetically much cheaper than direct inser- tion.This mechanism is not at variance with the experimental results. Thus helium slowly (over a period of several weeks) enters the C60 sphere at a pressure of *3000 atm and a temperature of 600 ± 650 8C.17 The endofullerene Ne2@C70 is obtained by heating C70 under a high pressure of neon, the C70 :Ne@C70 : Ne2@C70 ratio being dependent on the number of the synthesis cycles and reaching a limiting value of 1000 : 1 : 0.02.18 These findings led to the important conclusion that there exists a promoter or catalyst which facilitates the insertion of He atoms into C70.This might be a radical, which adds to fullerene and thus weakens the adjacent chemical bonds and promotes the opening of a `window'. Cer- tainly, discovery of a catalyst able to reversibly open and close a `window' in the fullerene sphere would ensure a breakthrough into this new class of compound.Chemistry on the border of two centuries �achievements and prospects Probably, a reversibly appearing 'window' as a capture mechanism is important for small neutral atoms (such as He and Ne). In the case of large atoms (e.g., lanthanides), trapping of atoms (or ions) during the formation of the carbon sphere is apparently more significant.The search for appropriate catalysts which would facilitate this process is also worth efforts. Both mechanisms can act simultaneously; apparently, this is the case for trapping of an N atom by C60 (when C60 is heated in a nitrogen plasma).19 In the N@C60 molecule, the nitrogen atom with three unpaired electrons (one electron in each p orbital) is chemically inert (this is also true for hydrogen in H@C60 and for phosphorus in P@C60).The unique chemical inertness of an inner p-electron shell in fullerenes is in sharp contrast with the high reactivity of the outer p shell. Thus reactions involving the outer shell (Fig. 5) barely affect the inner nitrogen atom; only slight distortion of its EPR spectrum is observed, which is manifested as weak satellite lines, indicating that the spherical symmetry of C60 has been somewhat distorted. This makes the nitrogen p orbitals energetically slightly non-equivalent (by *0.003 eV) and stim- ulates dipole spin splitting of the Zeeman levels. COOEt E N BrCH(COOEt)2 , NaH toluene N@C60C(COOEt)2 N@C60 Figure 5. Scheme of the N@C60 molecule and the product of its cyclo- propanation. The corresponding EPR spectra are shown below.Yet another feature of endofullerenes, which is of fundamen- tal importance, is noteworthy, namely, contraction of the electron shell inside the fullerene sphere. The constant of hyperfine (electron ± nucleus) coupling in a free N atom is 3.7 ê, that in the KN3 crystal is 5.0 êand the same value for N@C60 is 5.7 ê. This indicates that atoms inside the fullerene sphere experience substantial pressure, which results in contraction of the p orbitals of valence electrons. In terms of quantum chemistry, this means perturbation of the wave function resulting in an increase in the contribution of the s orbital to the superposition of atomic orbitals. Contraction of the electron shell of the inner atom has also been found in the endofullerenes (Sc,La,Y)@C60.In these compounds, anisotropy of the g-factor of the unpaired electron was observed,20 which points unambiguously to hindered rotation of the atoms inside the sphere. It is not difficult to find a relationship between the electron shell contraction and the behaviour of `onion' fullerenes such as C60@C240@C540..., which consist of several spherical molecules arranged one inside the other, the distance between the sphere surfaces being 50.34 nm. This is a new class of endofullerenes. The inner spheres experience enormous pressure of the outer spheres, and this accounts for the fact that the inner spheres are converted into diamond even upon weak photo or radiation treatment. Synthesis of cylindrical carbon nanotubes (*100 A Ê in diame- ter), which are formed according to the same principle as full- erenes, has become a prominent event in modern chemistry.They can be doped with metals (both on the inside and on the outside) 89 and are characterised by high solubility of hydrogen (owing to capillary condensation); therefore, they can be used in chemical cells. Young's moduli of carbon nanotubes are close to that of graphite (in one plane). The first experimental data on the use of nanotubes as probing tips in scanning force tunnelling microscopy were reported. Development of methods for the preparation of oriented nanotubes on thin metallic films using laser etching is in progress. Polymerisation of nanotubes opens up ways to new porous materials such as organic zeolites.Multilayer nanotubes consisting of molecular cylinders enclosed one into another have been obtained. Technologies that enable manipulation of nanotubes by virtue of atomic force tunnelling microscopy are being developed.21 One can lay, bend, cut, straighten, cross and move nanotubes thus arranging molecular electronic devices. Thus, it becomes possible to pass from fine words to impressive deeds. Preparation of metallic hydrogen represents one more impor- tant achievement of recent years. It can be debated whether it was done by chemists or physicists; however, the transformation of hydrogen into metallic hydrogen is undoubtedly a chemical process, resulting in a rearrangement of the electron shells of hydrogen molecules.Metallic hydrogen is obtained by shock compression of a 0.5-mm thick liquid film of molecular hydrogen placed between single-crystal Al2O3 anvils at a pressure of 2 Mbar.22 The electrical conductivity of this material (*2000 O71 cm71) is close to those of molten cesium and rubidium; in this sense, metallic hydrogen resembles liquid alkali metals. The mechanism of its formation is unknown; it involves either immediate complete release of the electrons from a hydro- gen molecule or dissociation of the molecule into atoms followed by the discharge of electrons to the conduction band. In order to decide between these variants, non-stationary reaction regimes need to be studied. III. Chemistry in micro- and macro-reactors It should be admitted that the chemistry created by the mind and hands of Man is only a minor fraction of chemistry created by Nature.The vast majority of reactions implemented by humans are `non-organised' reactions in which species (molecules, ions, atoms or radicals) react when they meet by chance (in space and time). Meanwhile, the `natural' chemistry is highly organised, i.e. almost all the chemical transformations occur in systems with molecular or supermolecular order. Large sets of biochemical reactions are organised both in space and in time. Thus, during photoreception each light quantum absorbed by the retina of an eye triggers a cascade of reactions yielding ultimately cyclic guanosine monophosphate (quantum yield 104), which induces an electric potential on the membrane.This signal is then recorded in the brain as the signal for light reception (other examples can be found in a review1). It is owing to the high degree of organisation that the selectivity and productivity of biochemical reactions reach very high levels, which are yet inaccessible in conventional chemistry. The realisation of this fact and the turning of chemistry as a science to molecular and supermolecular organisation of reactants started in the second half of this century; now this is a leading line in the intrinsic self-organisation of chemistry directed towards the future. The simplest microreactor in which reactants can be organised in space is a van der Waals molecule.A large number of micro- reactors of various natures and sizes and various degees of molecular ordering exist. They include complexes, crystal sol- vates, gas (for example, methane) hydrates, inclusion compounds (in which `hosts', represented by cyclodextrins, semispherands, calixarenes, etc., organise reactions of `guest' molecules), macro- molecules, (helices or rigid zigzag-like rods), the cavities of zeolites and other porous media, micelles and vesicles. Microreactors change the molecular dynamics of reactants, mechanisms and rates of chemical transformations, the pK values of acids and90 bases, local charges and charge distributions, ionisation poten- tials, electron affinities, conformations and reactivities. Reactions in microreactors follow new `forced' pathways, and other anom- alous features can also be observed.Two-dimensional microreactors are represented by the elec- tric double layer, monomolecular layers (of the Langmuir ± Blodgett type), membranes (both biological and artificial), inter- faces (they are used in phase transfer catalysis, especially in reactions involving ordered and directed exchange by substances and electric charges between the phases), adsorption layers of reactants on solids (`two-dimensional' zone reactions) and so on. A classical two-dimensional microreactor � electric double layer � has now a new lease on life, associated with high time resolution regimes (*10712 ±1079 s), which are brought about by the laser-induced chemistry at solid ± liquid interfaces.The main character in such a microreactor is an electron, either solvated or free. Its chemical behaviour is the subject of special interest regarding generation of electric potentials and modelling of the potentials existing in `natural' chemistry. Potentials at a liquid ± liquid interface also deserve attention. The interest in 2D monolayers on solids is due not only to their use as heterogeneous catalysts (this topic is already out of date) but also to the new opportunities offered by synthetic chemistry on the surface. For example, 1,5-cyclooctadiene, which binds to the surface of an Si(001) single crystal upon rupture of one C=C bond, forms a two-dimensional molecular lattice.23 The highly ordered monolayer consisting of grafted molecules exhibits high reactivity and can participate in further chemical functionalisa- tion leading to any chemical architecture.Cluster microreactors are highly diverse. They can be atomic, molecular, ionic, ion-molecular and organometallic, either charged or neutral. Clusters are formed in gases (for example, in expanding supersonic atomic-molecular beams), in zeolite chan- nels (for example, the In8 clusters), in liquids, on solid surfaces, ilid matrices (such as solid argon or xenon), etc. They play an important role in the chemical materials science (cluster materials with unusual physics and mechanics), in chemical analytics and diagnostics (detectors and sensors) and in other fields of applied chemistry.The main, fundamental problem in the chemistry of clusters is associated with dimension effects. This is an intriguing question� how do the properties of separate species evolve on their combi- nation into the properties of a new phase; how are the bridges between the world of a separate molecule and the world of macroscopic matter being built? The dimension effects are so diverse and unexpected that no general solution of this problem exists. Quantum chemistry and molecular dynamics methods provide partial answers only to some questions. Clusters still remain enigmatic objects (for unusual properties of clusters, see for example, Ref. 24). Thus Re atom is inert towards CH4, whereas the Re3 cluster is active; the linear clusters Cu3 and Ni3 do not react with H2 or CH4, whereas triangular clusters react readily and rapidly; the Con clusters are reactive towards H2 at n = 3, 10 and 12 and are inert at any other n; the Nbn clusters with n = 3, 4 and 5 react with H2, whereas other Nb clusters are relatively inert; Nb clusters with n = 5, 6 and 11 are active in dehydrogenation of benzene; the Agn clusters exhibit metallic properties starting from n = 6, whereas Nan do not exhibit any signs of a metal phase even at n = 14.The I2(C6H6)m cluster is polyhedral and Vn(C6H6)m has a multi-deck sandwich structure. The MoO+ cation does not react with NH3, whereas the Mo3Oá8 and Mo3Oá9 clusters dehydrogenate NH3 to give com- plexes with the nitrogen formed upon dehydrogenation. The same clusters oxidise CO to CO2.The V3Oá7 and V3Oá12 clusters are efficient in dehydrogenation of butadiene and destruction of butane, while V3Oá6 and V3Oá11 are inert in these reactions but they are active in the addition and oxidative cyclisation reac- tions.25 A L Buchachenko Mg+(CH3OH)n (MgOCH3)+(CH3OH)n71+H, Mg+(H2O)n (MgOH)+(H2O)n71+H In the cluster of a-naphthol with (H2O)n (these compounds are called crystal hydrates), proton transfer from excited a-naphthol to water has been detected only for n520; in the case of a-naphthol clusters with (NH3)n this process occurs even at n54. In the clusters of univalent magnesium (or calcium) with (H2O)n and (CH3OH)n, the electron transfer and hydrogen reduction occur for 6 4n415; at any other n the reaction does not occur, and hydrogen elimination is not observed.26 The rate of photoinduced isomerisation of trans-stilbene placed in the (hexane)n cluster was measured in molecular beams by picosecond spectroscopy.27 The rate constant for this process decreases in the sequence 9.76109 (n=0, free molecule), 36109 (n=1), 16109 (n=2), 0.86109 (n=3), 0.56109 s71 (n=4).Unexpectedly, all these constants proved to be smaller than the observed isomerisation constant of trans-stilbene in pure liquid hexane at the same temperatures and energies (11.56109 s71); i.e. immersion of trans-stilbene into the (hexane)n cluster decreases its reactivity. The cluster shell fixes the molecule even more strongly than a `liquid-phase' shell; it abnormally increases the kinetic (Kramers) friction, which accompanies the movement of the reacting molecule along the reaction coordinate.Special interest is aroused by the dimension effects in semi- conducting nanocrystals the size of which is smaller than the Bohr radius of an exciton. Quantum capture of an electron and a hole (generated by photochemical or radiation-chemical treatment) in the confined space of these nanoclusters (called quantum dots) results in an increase of the band gap following a decrease in the cluster size; hence, recombination luminescence shifts to the blue region (to higher energies). A striking example is provided by nanocrystals of CdSe `coated' by one or two ZnS monolayers. When the diameter of the inner CdSe nucleus increases stepwise from 23 to 55 A Ê , the bright blue luminescence passes into green, yellow, orange and, finally, bright red luminescence.28 In addition, cluster chemistry opens up the way to a new strategy in heterogeneous catalysis, especially when combined with scanning tunnelling microscopy. The tip of the microscope is able to apply any number of any atoms on the faces, edges or terraces of any crystal, thus creating diverse catalytic micro- reactors and providing conditions for testing various reactions on them. This is a new wind in the fundamental science of catalysis and its future. Clusters are much more widespread and, hence, much more significant than it has been believed until recently.Thus it has been found experimentally (by cold-neutron and X-ray scattering) that liquid ammonia is clustered � it consists of the (NH3)7 clusters containing one molecule in the centre and six molecules at the periphery.29 The cluster structure of liquid water is well known: water molecules are united into hexa-, penta- and tetramers, having close energies and forming prismatic, cage, `open-book' and adamantane structures, and also into dodecahedra and other big clusters.Liquid aqueous solutions of ethanol are also non- uniform; they consist of water and ethanol clusters.30 It is note- worthy that when the content of the alcohol in water is*40%, the proportions of ethanol and water clusters are equal. Perhaps, this accounts for the special receptive-gustatory properties of the well- known and popular water-ethanolic drink.Indications of cluster- ing have also been found in liquid benzene. This might prove to be a general phenomenon; if this is really so, all liquid-phase reactions should be interpreted as occurring in microreactors. Furthermore, a chemical reaction can itself stimulate molec- ular organisation of a microreactor. Thus a colourless molecule of spiropyran having a long hydrocarbon `tail' is converted upon photochemical treatment into a coloured bipolar form with separated charges; it is in this form that spiropyran moleculesChemistry on the border of two centuries �achievements and prospects unite to form micelles. Here photochemical reaction acts as the trigger for micelle formation.Brain is a macroreactor of enormous complexity. There occur a tremendous number of chemical reactions, responsible for the synthesis of the memorising molecular structures, which form memory and the whole control system of a living organism. Brain is of key importance in the entire chemistry of living matter; chemistry of brain is the chemistry of the XXI century. The main purpose of brain is to transform chemical energy into electric energy; this task is accomplished by neuromediators. They stimulate enzymic synthesis in synaptic membranes and ensure communication between neurons, formation of electric potentials and transfer of electric signals via recharge of the membrane. They control a cobweb of impulses and potentials, which govern all functions of a living organism. Neuroprotectors act via chemical mechanisms.Many neuromediators have long been known. Thus dopamine and its chemical functions have been studied comprehensively; however, only recently it has been found that its deficiency in an organism is responsible for dotage. One reason for the reduction of the dopamine level with advance in age is that it is destroyed under the action of the enzyme monoamino oxidase B. Another reason is that the number of dopamine-receiving neurons in an organism decreases with time (a low level of these neurons is also characteristic of those who suffer from parkinsonism). In addi- tion, it has been found that some narcotics trap neurons and remove them from the functioning of brain.It is clear that the understanding of the chemical mechanisms of the operation of neuromediators opens up ways for eliminating disorders in the work of the brain macroreactor. Other neuromediators are also known; among these, mention should be made of versatile nitrogen oxide with a broad spectrum of chemical and physiological action and Semaks ± a neuropeptide cong of seven amino acids. Study of a macroreactor such as the human brain requires delicacy, which sharply restricts the range of possible direct experiments. Nevertheless, these studies are carried out within the framework of neurophysiology in combination with medicine and clinical surgery. The studies widely use microelectrode equip- ment, which permits estimation of the behaviour of the electric potentials formed and their response to chemotherapy and physiological treatment.Enormous opportunities are also pro- vided by positron-emission tomography, in which the 11C isotope and 11C-labelled chemical compounds being tested are used as the positron sources. At present, researchers have realised that wide use of experimental modelling of brain elements can help in the understanding of the operation of this macroreactor. The main goal of modelling is to answer the question how does the brain perform its primary task, i.e. how does it transform chemical energy into electric energy. By the way, as a `side result' of this modelling, one could expect design of molecular devices charac- terised by high coefficients of conversion of chemical energy into electric energy (within the framework of chemical energetics).The Earth is a huge geochemical macroreactor. This mecha- nochemical reactor is the source of earthquakes. It covers the whole of mechanochemistry and its consequences: chemical bond rupture induced by mechanical treatment, formation and migra- tion of dislocations, generation and multiplication of cracks, shears and shear waves, generation of electric potentials and their gradients, radio waves, magnetic fields and chemical emis- sion. Now scientists have understood that the key to the diagnos- tics and prediction of earthquakes lies in mechanochemistry. All events that take place in the geochemical macroreactor echo in the ionosphere as electic field gradients and are detected there as forerunners of earthquakes.Mechanochemistry, a field of science in which chemistry occupies a prominent place, is directed towards the XXI century. 91 IV. Coherent chemistry Coherent chemistry is a `new face' of chemistry. Coherence is the ability of chemical systems to generate oscillation regimes of reactions. The coherence (time synchronisation) of a chemical reaction is manifested as cyclic variations of the reaction rate and can be detected as oscillating product yields, luminescence emis- sion, electrochemical currents or potentials, etc. Chemical coherence exists at two levels, quantum and macro- scopic levels. In the former case, the coherence refers to the reactivity of an assembly of reacting particles generated by a laser pulse. Such an assembly oscillates between states with different reactivities, these oscillations modulating the yields of the reaction products.The vibrational and spin types of coherence are of the quantum origin. In the case of macroscopic coherence, the concentrations of active reagents (or intermediates) periodi- cally change. The best known example of macroscopic coherence is the Belousov ± Zhabotinsky reaction. The coherence introduces new notions into chemistry, such as wave packet, phase, loss of phase coherence (dephasing), interfer- ence, bifurcation and bifurcation diagrams, phase portrait, strange attractor and phase turbulence. This is not simply a new language of chemistry; this reflects a new level of thinking and a new level of chemical research practice.In the coherent chemistry, the random, statistical behaviour of molecules is replaced by organised, ordered and synchronous behaviour; thus, chaos turns into order. A short (10714 ±10713 s) laser pulse (with a duration shorter than the period of vibrations of atoms) `excites' a molecule and `throws' it onto a new potential. In this new potential, an assembly of molecules prepared by the laser pulse behaves coherently, i.e. vibrations of atoms of all of the members of the assembly are synchronised, the whole assembly being a wave packet. When moving on the potential surface, a wave packet can disperse into several smaller packets (with different vibration amplitudes and phases); some of them can go out of phase (lose the coherence) and disappear, while some other can interfere to restore partially the initial packet, etc.A classical illustration of the coherent behaviour of a wave packet is provided by an assembly of NaI molecules (Fig. 6).31 Molecules of NaI excited by a short (*50 fs) pulse are `trans- ferred' from the surface of the ground (ionic) state to the surface of an excited `covalent' state; this creates a wave packet localised on a potential surface. This wave packet slides in an adiabatic potential (Fig. 6 a, the upper curve), in which it oscillates between the covalent and ionic states and crosses at some intervals the point of intersection of the ionic and covalent terms (when the intera- tomic distance is 6.93 A Ê ).Due to the non-adiabatic connection of these two terms, there is a finite probability of penetration of the packet from the upper curve to the lower one. As this takes place, a part of the penetrated packet dissipates to generate Na atoms (they can be detected experimentally) and the rest becomes a new wave packet, which moves coherently in the lower valley. The point of intersection of the ionic and covalent terms (where the upper and the lower potential curves contact with each other) is the bifurcation point in which the wave packet splits into two packets, one of which is partially irreversibly destroyed. In principle, it is clear that the packet situated in the lower potential curve can return to the upper potential curve via the same bifurcation point and restore the initial packet (interference), although in this particular case, the probability of this event is low.For other systems characterised by a different configuration of the potential energy surface, this probability can be substantial and the packet interference can be manifested fairly clearly. These qualitative considerations can be converted into a quantitative picture by calculating the probabilities of dissipation of the packet at the bifurcation point (according to Landau ± Zener). Fig. 6 b shows the pattern of the packet movement on the time and space coordinates. It can be readily seen that the coherent movement of the packet in the upper valley generates a stream ofNa 0.2 0 Signal (rel.units) Potential energy 92 INaI* 0 0.6 0.472 Na atoms (reaction product) at the bifurcation point. The density of this stream oscillates with time, the oscillation frequency being equal to the frequency of packet oscillation in the upper valley; the density amplitude attenuates due to the packet dissipation, dephasing and falling into the deep lower valley (Fig. 6 c). The coherence created initially is retained during the reactions and is transferred to the products. This indicates that the molec- ular assembly undergoes a collective, synchronous and phase coherent chemical transformation. This coherence can be found not only in simple reactions such as dissociation of the NaI, I2, HgI2, etc.molecules but also in complex systems. Thus phototransformation of rhodopsin and its analogues occurs in parallel via a coherent and a non-coherent } channel over a wide spectral range (400 ± 1100 nm).32 The key process in the phototransformation of rhodopsin � a molecular system consisting of an apoprotein (opsin) and 11-cis-retinene attached to it�is the cis ± trans-isomerisation of retinene around the C(11)=C(12) bond. The direct coherent formation of the first intermediate (bathorhodopsin from rhodopsin and hRK from halorhodopsin hR) occurs over a period of*200 fs, the coherence being retained in the reaction product. The fact that even in a complex molecular system such as rhodopsin, the degree of } Phototransformation via a non-coherent channel implies statistical thermalisation of the excited state followed by slow (2 ± 4 ps) statistical formation of the product.a Na Na+ I I7 The ionic state: Na++I7 The covalent state: Na+I Na+I [Na...I]= 20 15 5Rx=6.93A Ê 10 Internuclear distance/ A Ê cNa Na+ I7 I 4 6 2 Delay time /ps 500 fs Time /fs b 160 fs 200 fs 700 fs 1300 fs 0 5 10 15 Internuclear distance/ A Ê Figure 6. Femtosecond dynamics of photodissociation of NaI; (a) potential curves for the covalent and ionic states; (b) dynamics of the ionic and covalent wave packets; (c) experimental dynamics of wave packets detected based on the activated complex NaI* and from the yield of free sodium atoms (the upper curve).31 coherence remains remarkably high (*20%) merits attention.When the hydrogen atoms at C(11) and C(12) are replaced by deuterium, the isotope effect is not observed over a short period of time (up to 110 fs); it appears only in the range 110 ± 170 fs, when rocking around the C(11)=C(12) bond is characterised by a large amplitude and leads to full rotation. This means that the coher- ence of the whole process (it takes *200 fs) is a stable character- istic, which is retained at any amplitude along the coordinate of isomerisation.33 Vibrational coherence can also be detected in the chemical transformations of such a bulky molecular system as the bacterial photosynthetic centre.34 Moreover, coherence has been found in the antennae of photosynthesising centres, its degree and `length' (i.e.the length of the `coherent' accumulation of energy) being dependent on the molecular order and spatial arrangement of the light-collecting complexes near the the energy-collecting photo- synthesising centre. It is not the coherence itself that is the most remarkable feature of the coherent chemistry but rather new and non-traditional methods for controlling chemical reactivity. When the first pulse throws a wave packet onto a potential surface, the second pulse can influence its evolution: either throw it further onto a new surface or bring it back onto the initial surface (the Raman process). This can be done either before or after the wave packet has passed the bifurcation point by changing the delay between the first (pump) pulse and the second (probe) control pulse.In A L Buchachenko State:covalent ionic20Chemistry on the border of two centuries �achievements and prospects other words, the dynamics of wave packets and their chemical fate can be controlled; the ratio of the channels of their transformation and the yields of the reaction products can be altered. Thus a new tool for controlling a chemical reaction appears, i.e. the phase of a coherent assembly. By affecting a wave packet in various phases of its evolution, one can change its chemistry. For example, in the electronically coherent assembly of HD+ molecules, prepared from HD by a short ionising laser pulse (pump pulse), the electron oscillates between the H and D atoms.Using the second (control) pulse, one can attain the coherent control of the HD+ dissociation, which gives either H+D+ or H++D, depending on the instant when the second pulse acts. To induce the former dissociation pathway, the second pulse should be switched on when the electron resides on the hydrogen atom, whereas for the latter pathway, it should occur when the electron is on deuterium. This is the highest level of coherence in chemistry, the prospects of which are being discussed even now.35 An even finer and a more elegant way of controlling the chemical fate of a wave packet is embodied in the phase relation- ship between the packet and the probe pulse. The current state of optics makes it possible to control the frequency-dependent phase of laser pulses.The delay t(o) of the group velocity of a particular frequency component of the optical spectrum is known to depend on the phase of this component F by the relation t(o) =qF(o)/ qo. For a pulse with a given spectrum, all the frequency compo- nents should be in phase and should arrive simultaneously at the destination point, i.e. t(o) does not depend on frequency. In this case, F(o) includes no frequency-dependent quadratic terms (or terms with higher exponents). However, one can change F(o) in a specified way (by virtue of optical grids and prisms) in order to introduce frequency-dependent quadratic terms (they are called chirps). In these chirped pulses, phase velocities become frequency dependent.Thus when the chirps are negative, the red component of the pulse lags behind and arrives at the destination point (e.g., at a wave packet) later than the high-energy blue component. Conversely, in the case of positive chirps, the red light is ahead of the blue one. It is clear that by changing the chirp of an optical pulse, one can control the relationship between the phases of the wave packet and the control pulse and thus `catch' the packet in various positions on the potential surface by either red, or blue, or another appropriate light. Certainly, this is a simplified picture illustrating in a clear form the main principles and chemical effects of the vibrational quantum coherence. The main idea is that the coherent chemistry brings about a new factor controlling chemical reactions, namely, the phase. By changing the phase (this can be attained by delaying the control pulse relative to the generating pulse or by chirping), one can manipulate the chemical behaviour of assemblies of reacting species without changing the energy or angular momen- tum.The next level of coherence is quantum periodicity in a system consisting of two species, each carrying an electron spin (for example, a pair of radicals).Such a spin pair can exist in two states � singlet (reactive) and triplet (chemically inert). An assembly of these pairs, prepared by a pulsed method (photolysis or radiolysis) in a specified spin state (e.g., in the triplet state), oscillates synchronously between the two states.The driving force of this oscillation is the difference between the Zeeman energies of the two partner spins or the hyperfine electron ± nuclear coupling (if the radicals contain magnetic nuclei). The magnitude of these interactions is of the order of several megahertz (it is with this frequency that the assembly oscillates between the signlet and triplet states). It is clear that the yield of the products of a chemical reaction (which are formed only when the species occur in the singlet state) is time-modulated at the same frequency (Fig. 7). The periodicity in the product formation is a consequence of the coherence of the behaviour of spin pairs; this is an example of electron-spin quantum beats in chemical reactions (as opposed to the vibrational quantum beats, considered in the previous Sec- tion).I (t) a 1.2 1.0 0.8 0.6 rss(t) b 1.0 0.75 0.5 0.250 20 40 60 80 t/ns Figure 7. Quantum beats in the recombination of the (diphenyl sul- fide)+/(p-terphenyl-d14)7 radical-ion pair in cis-decalin atH=1200 (a, b) and 3000 G (c, d). The product of recombination of p-terphenyl-d14 was detected by fluorescence. (a, c) Experimental data, (b, d) theoretical calculations; I(t) is the fluo- rescence intensity, rss(t) is the content of the singlet state in the radical-ion pair.The electron-spin coherence has been found in photochemical and radiation-induced chemical reactions of radical pairs 36 and in the primary photochemical event of charge (and electron spin) separation in the photosynthetic reaction centres.In these systems coherence originates due to the synchronous oscillation of spin pairs between the singlet and triplet states. Moreover, in addition to the electron-spin coherence, nuclear-spin coherence has been found in these systems (ordered in-phase precession of the system of polarised nuclei occurs).37 The advantage of spin coherence is that it provides new methods of investigation and identification of fine and accurate details of the structure and dynamics of chemically generated spin systems. Furthermore, even when the chemical reaction produc- ing spin systems occurs statistically (non-coherently), spin coher- ence exists in these systems, which opens up a way to the new field of science, chemical radiophysics (see Section V).The oscillating regimes of chemical reactions were first observed rather long ago. At that time, oscillations were consid- ered to be an exotic rather than a regular feature. Nowadays the Belousov ± Zhabsky reaction and oscillations of pH and electrochemical potentials in heterogeneous systems such as oil ± water systems, Liesegang rings and wave combustion, have become classical examples. However, the fact that macroscopic coherence is a fundamental property has been realised only recently. The science of chemical oscillators has been revived owing to two important facts. Firstly, it became clear that coherent regimes could increase the yields of reaction products and the process selectivity and could lead to self-cleaning of surfaces from catalytic poisons, etc.These expectations have been confirmed, especially for chemical oscillators with forced oscillations.38 Secondly, the interest in chemical oscillators has been aroused again owing to biochemical oscillating processes occurring in nerve cells, muscles, and mitochondria. This resulted in active assimilation of new systems of oscillators, i.e. in combination of several chemical oscillators into a common unified system. It is believed that such a system would be a prototype (although yet primitive) of the future models of neuron networks.39 However, even now, studies are in progress on the oscillator excitation thresholds in such a combined system and on the influence of the 93 I (t) c 1.2 1.0 0.8 0.6 rss(t) d 1.0 0.75 0.5 0.250 20 40 60 80 t/ns94 E/ V 71.2 70.6 71.2 70.6 71.2 70.6 80 0 40 Figure 8.Periodic potential oscillations during the reduction of bromate ions in a solution of NaOH at a silver electrode and at the current 1.5 (a), 2.5 (b) and 3.0 mA (c).42 number of oscillators, the way of their connection (a linear or a circulating system) and the connection parameters (mass exchange, control of electrochemical potential and currents) on the excitation of collective oscillations, etc. New chemical oscillators have been discovered. Thus the potential of a platinum electrode in the ClO2 ± I2±CH3C(O)CH3 system varies as a periodic function of time.When formaldehyde and methanol are oxidised under galvanostatic conditions, the oscillation frequency repeatedly doubles, resulting in the break- down to chaos at the bifurcation points.40 This type of behaviour resembles that observed in some enzymic reactions and biological systems.41 Oxidation of bromate ions in alkaline solutions is also an oscillating process (Fig. 8). Oscillations are due to the concen- tration effects at the electrode surface and, therefore, they vary as a function of the surface geometry (roughness, presence of islets, etc.).42 Another bromate oscillator catalysed by tris(bipyri-dine)r- uthenium is, in addition, a photo-controlled system displaying different bifurcation diagrams in the light and in the dark.A strange bifurcation behaviour of the order and chaos regimes and doubling of frequency are also observed in the peroxidase- catalysed oxidation of dihydro(nicotinamide)adenine dinucleo- tide (NADN).43 In heterogeneous catalytic oscillators, oscillations may arise due to a non-linear relationship between the rates of chemical reaction and heat exchange, concentration gradients, periodic modulations of the composition of a `two-dimensional' reactant, `breathing' of the surface (when atoms go down and up, similarly to waves in a two-dimensional lattice), etc. Examples of chemi- cally oscillating processes are provided by the reduction of NO with hydrogen (when the pressure of H2 decreases, the period of oscillations repeatedly doubles until break-down to chaos occurs), oxidation of CO with oxygen on platinum, accompanied by generation of time-and-space helical waves (Fig. 9) 44 and some other processes.The brain is a striking and a unique example of a system self- organised in space and time. The chemical activity of enzymes and, as a consequence, the electric potentials in the system of synaptic membranes and neurons in brain are remarkably abc t /s 160 200 120 A L Buchachenko f c e b d a Figure 9. Helical wave on the Pt(110) surface which develops during oxidation of CO with oxygen (434 K). The light sections are areas occupied by CO; the dark sections are areas occupied by oxygen. The photographs were taken consecutively at times 0 (a), 10 (b), 21 (c), 39 (d ), 56 (e) and 74 s ( f ) on a 0.260.2 mm2 area.44 synchronised (coherent).It is also clear that the scales of coher- ence (i.e. the sizes of synchronised areas in brain) are different for different levels of brain functioning. In other words, the brain consists of biochemical oscillators of various sizes connected to one another and combined into a unified control system. Order is the normal state of this system, while chaos implies severe pathologies (such as Alzheimer's disease). Perfect order is the property of a brilliant genius brain (this idea is only hypo- thetical but looks like being true). V. Spin chemistry and chemical radiophysics Yet another new field of modern chemistry is spin chemistry concerned with the behaviour of angular momenta (spins) of electrons and nuclei during chemical reactions. Spin chemistry is based on the fundamental law: in adiabatic chemical reactions the electron and nuclear spins are strictly retained.Only those reactions are allowed which do not require spin change. In other words, all the chemical reactions are spin-selective, i.e. they are allowed only for those spin states of reactants the total spin of which is equal to the total spin of products, and they are completely forbidden when the reactant spin is not equal to the product spin. A simple example is provided by a radical pair comprising methyl and hydroxyl radicals. Recombination of these radicals to give methanol occurs only in the singlet state, because this ensures that the electron spin of the pair (S = 0) is identical to that of CH3OH.However, disproportionation of the same radicals in the singlet state to yield H2O and triplet methylene (the total spin of the products S = 1) is forbidden for spin reasons. For the triplet radical pair (S = 1), on the contrary, recombination is forbidden, whereas disproportionation is allowed, although its activation energy is relatively high,*6 kcal mol71. One should be conscious of the fact that chemistry is governed by two fundamental factors � energy and spin (as shown in the previous Section, one more factor, phase, appears in the coherent chemistry). Unlike the prohibition for energy reasons (it appears when the energy of reactants is lower than the energy barrier to theChemistry on the border of two centuries �achievements and prospects reaction), which can be overcome via tunnelling under the barrier, the spin prohibition is insurmountable.Spin can be changed upon non-chemical magnetic interac- tions; only they can transform spin-forbidden (non-reactive) states of reactants (e.g., radical pairs) into spin-allowed (reactive) states. Magnetic interactions, the energy of which is negligibly small, switch reaction channels � they can open closed channels and, conversely, close the open (allowed) ones, depending on the initial state of the reactants. Actually, they write a new, magnetic script of a chemical reaction. The spin selectivity and, hence, the magnetic sensitivity of chemical reactions has served as the source of three generations of magnetic effects discovered during the last two decades.45 Fig.10 illustrates the origin of these effects in relation to a particular example, which is, however, frequently encountered in chemistry, namely, a radical pair able to exist in either the singlet or the triplet spin state. In the case where triplet ± singlet transitions are induced by static magnetic fields (external fields or internal fields of magnetic nuclei), first-generation magnetic effects arise. Magnetic effects of the second generation appear when the spin conversion of pairs occurs on exposure to microwave fields. Finally, when the conversion is induced by the influence of a third paramagnetic species, a remarkable phenomenon, spin catalysis, can be observed, which represents the third generation of magnetic effects. The third species (a radical or a paramagnetic ion) acts as a spin catalyst. First-generation magnetic effects.An external magnetic field exwo types of influence on the spin conversion of radical pairs and, hence, on the chemical reactivity. Firstly, the field switches off two of the three channels of spin conversion, namely, S ± T+ and S ±T7, so that only the S ± T0 channel is left (T+ , T0 and T7 are triplet states which differ in the projection of the total spin). Secondly, it accelerates the S ± T0 conversion if the partners of the pair have different Zeeman energies. Depending on the initial spin state and magnetic parameters (hyperfine coupling energy, g-factors, dipole interaction) of the reactants, chemical reactions can be accelerated or decelerated by an external mag- netic field.Magnetic field effects have been found in a great number of chemical and biochemical reactions, their range constantly extending. A reliable theory capable of predicting the significance SPIN CHEMISTRY Singlet radical pair Triplet radical pair 123 1 2 3 Exchange interaction Microwave and radiofrequency Zeeman and Fermi interaction irradiation Third-generation magnetic effects spin catalysis Second-generation magnetic effects RYDMR, RIMIE, SNP First-generation magnetic effects MFE, MIE CIDNP, CIESP Figure 10. Spin evolution of radical pairs.The following designations are used: MFE is magnetic-field effect, MIE is magnetic isotope effect, CIDNP and CIESP are chemically induced dynamic nuclear polarisation and electron spin polarisation, RYDMR is reaction yield detected magnetic resonance, RIMIE is radio-induced magnetic isotope effect and SNP is stimulated nuclear polarisation. 95 105M/g 54321 6 8 H/ kG 0 2 4 Figure 11. Molecular mass of the polymer obtained by emulsion polymer- isation of styrene (dibenzyl ketone as a photoinitiator; sodium dodecyl- sulfate as an emulsifier; constant polymerisation time) vs magnetic field intensity.46 and behaviour of magnetic effects has been developed. The magnitudes of these effects are often quite large.Thus the rate of photoinduced aqueous-emulsion polymerisation of styrene and the molecular mass of the resulting polymer increase 6 ± 8-fold even in weak magnetic fields, *100 G (Fig. 11).46 An even more pronounced effect (20 ± 25-fold increase) is attained in the light scattering by colloidal particles subjected to photochemical cross- linking.47 This strong effect can be used for visualisation of magnetic fields and their gradients. Being selective regarding the electron spin, chemical reactions between spin carriers (radicals, paramagnetic ions and molecules, carbenes, etc.) are also selective as regards the nuclear spin. If the two spin subsystems (electronic and nuclear) are related by the Fermi hyperfine coupling (FHC), the nuclear subsystem can affect the behaviour of the electronic subsystem via the FHC and thus modify the chemical reactivity of spin carriers.The nuclear-spin selectivity is manifested as different rates of spin-selective reac- tions of radicals (or other spin carriers) with magnetic and non- magnetic nuclei. This is a new phenomenon referred to as magnetic isotope effect (MIE), which differs fundamentally from the classical isotope effect (CIE), resulting from nuclear-mass selectivity of reactions. Both effects sort the isotope nuclei: the CIE selects nuclei according to their weights, while the MIE selects them according to their spins and magnetic moments. The discovery of the magnetic isotope effect is a prominent event of the last quarter of the XX century, its significance being comparable with that of the discovery of the classical isotope effect.In addition, the magnitude of this effect substantially exceeds that of the CIE (Fig. 12) and, unlike the magnitude of the CIE, it depends on the magnetic field, temperature, molecular and chemical dynamics and the spin state of the reactants. The magnetic isotope effect results in fractionation of magnetic and non-magnetic isotopes in chemical, biochemical, geochemical and interstellar processes.48 The third effect belonging to the group of first-generation magnetic effects is chemically induced dynamic nuclear polar- isation (CIDNP). Unlike the MIE, in this case the nuclei are sorted not only according to their magnetic moments but also according to their orientations.During a chemical reaction, nuclei with different orientations are sent into different products and this creates non-equilibrium populations of the Zeeman nuclear levels in these products. An excessive population of the lower Zeeman level corresponds to the positive nuclear polarisation, whereas overpopulation of the upper level implies negative polarisation. The latter case is especially remarkable: when the overpopulation of the upper level exceeds some permissible limit, the populations are inverted; this provides the basis for chemical radiophysics. An assembly of product molecules with inverted populations accumulates energy in the Zeeman storage reservoir; the energy can either be consumed as heat (via spin ± lattice magnetic relaxation) or be converted into stimulated emission at a Zeeman96 C 1.2 OSi SGe U 1.0 Figure 12.Comparison of the magnitudes of the magnetic (white areas) and classical isotope effects (black areas) in the separation of the 13C/12C, 17O/16O, 29Si/28Si, 33S/32S, 73Ge/72Ge and 235U/238U isotope pairs. The coefficient a of one-stage isotope separation (a universal measure of isotope effect) is plotted as the abscissa. nuclear frequency. In this case, the reaction becomes a radio- frequency emitter, i.e. a quantum generator with chemical pump- ing (similar to chemical lasers). This new phenomenon � radio- frequency emission of chemical reactions � was first predicted theoretically and later found experimentally.49 It appears when the energy stored in the Zeeman storage reservoir exceeds the generation threshold, the movement of nuclear spins spontane- ously becomes coherent, and this coherent set of nuclei functions as a quantum generator.Chemically induced electron spin polarisation (CIESP) is the fourth effect. It arises due to the electron-spin selection during the reaction and results in non-equilibrium population of the Zeeman electron levels in radicals and paramagnetic molecules (in this respect, CIESP resembles CIDNP). The two latter effects are widely used in chemistry and biochemistry as new investigation methods (diagnostics of reac- tion mechanisms, detection of radicals, and chemical kinetics and chemical physics techniques).Second-generation magnetic effects. The microwave emission of chemical reactions is only one aspect of chemical radiophysics. Achemical reaction can be not only a generator but also a receiver of microwaves. The reception at the chemical level follows from the principles of spin chemistry: the resonance microwave radia- tion stimulates the triplet ± singlet conversion of radical pairs (or pairs of other spin carriers) and changes the yields of chemical products, which can be detected by chromatography, lumines- cence, electrical conductivity, etc. Thus, owing to the second- generation magnetic-spin effects, a reaction becomes a chemical receiver of microwave radiation.50 Moreover, this expedient can be performed selectively.If the microwave pumping affects all the radical pairs, the final outcome reduces to the change in the product yield at resonance frequencies. This effect has been called reaction yield detected magnetic resonance (RYDMR). Selective pumping involving only radical pairs with magnetic nuclei brings about a remarkable phenomenon, radio-induced magnetic iso- tope effect (RIMIE). Finally, when the microwave pumping is also selective regarding the orientation of nuclear spins (involves only assemblies of radical pairs with selected orientations of nuclear spins), stimulated nuclear polarisation (SNP) appears. Third-generation magnetic effects. Spin catalysis is the first example of purely physical catalysis.41 This phenomenon is notable because the spin conversion of reactants is induced by a 1.4 1.8 1.6 1.08 1.04 1.08 1.04 1.02 1.01 1.01 1.02 1.04 1.02 a A L Buchachenko paramagnetic species, a spin catalyst.Conversion occurs as a result of exchange interaction of the catalyst with thactants. Spin catalysis accelerates radical recombination, cis ± trans iso- merisation of compounds with double bonds (by seven to eight orders of magnitude), recombination of spin-polarised atoms, etc. Spin catalysis might also be effective in biochemical processes. The spin chemistry and the chemical radiophysics are based on electron- and nuclear-spin manipulation. When manipulation is accomplished by the chemical reaction itself, interesting first- generation magnetic-spin effects arise; they include microwave generation, which transforms a chemical reaction into a molecular radio station.When spin manipulation is induced by microwaves, even more elegant second-generation magnetic effects originate. They serve as an indicator of microwave reception. The spin chemistry and the chemical radiophysics are tightly connected; nevertheless, each of them solves independent prob- lems. The former develops new principles of the control of chemical reactions (including that by virtue of microwaves), while the latter is significant practically for biomedical purposes. VI. Chemistry under extreme and exotic conditions Nowadays, the scope of chemistry is expanding, modern chem- istry is actively encroaching upon fields that have not been of interest or have been inaccessible for classical chemistry.Tran- sition from the `explored' reaction conditions to extreme, non- classical and even exotic conditions occurs more and more swiftly. These conditions include intense electric and magnetic fields, superhigh pressures, shear deformations, light fields with inten- sities comparable with those of the electric fields inside the molecules, supercritical conditions, powerful gravitation, sonic and microwave fields, etc.52 Ultrashort laser pulses (duration 10 fs or shorter) carry enor- mously powerful optical radiation and electric fields. This imme- diately gave impetus to the search for new possible effects. In fact, the interaction of optical and electric fields with the electron shells of molecules gives rise to lots of unusual effects.Picosecond pulses with a power of up to 109 W cm72 induce `orbital' excitations in molecules, which lead mainly to the molecule fragmentation. However, when the power is higher (about 1014 ±1015Wcm72), multielectron ionisation of molecules occurs, which is followed by Coulomb explosion of the atomic- nuclear core. This type of behaviour is observed, for example, for clusters: an intense laser pulse (120 fs, 1015 W cm72) induces efficient ionisation of the NH3 molecular clusters, which is accompanied by Coulomb explosion, yielding charged nitrogen atoms (N2+, N3+, N4+).53 Fullerene C60 is also ionised by an intense laser beam (100 fs, 1016Wcm72) to give C4á 60 , which is split into fragment ions via Coulomb explosion.54 The Lorentz forces, which accompany the passage of a laser pulse through a substance, induce several other effects.Thus they ensure the formation of microparticles in polymer solutions. The photon pressure of the Lorentz forces in a concentrated beam captures polymer chains and condenses them into microparticles with diameters of *10 nm or larger.55 In addition, laser beams cause ablation of atoms and atomic-molecular clusters from solid surfaces, thus performing molecular etching. Each of these effects can provide the basis for not only new chemical research methods but also new high-end chemical technologies. High-power laser pulses are an excellent means for generating powerful short shock waves (in particular, they are used in physics for laser compression of deuterium aimed at attaining controlled nuclear fusion).In chemistry, laser shock waves are employed to study the behaviour of substances under extreme conditions (high temperatures, high pressures, shear deformations). Thus laser- induced shock waves ensure a pressure of up to 5 GPa in the front; the duration of the leading front can reach several hundred picoseconds for reversible compression and *20 ± 25 ps for irreversible compression. The temperature drop behind the shock front occurs at enormous rates (about 1011 K s71).Chemistry on the border of two centuries �achievements and prospects Laser-stimulated shock waves provide great scope for `extreme' chemistry; they actually give rise to a `new wave' in the use of shock waves. Onion fullerenes and the transformation of fullerene spheres into diamond have been already mentioned.The same occurs with multilayer nanotubes: at pressures of*50 GPa, their outer shells are torn and twisted to give graphite structures (diamond nanocrystals have also been detected).57 The synthesis of diamond according to the known explosion technology might occur (at least, partly) via onion fullerenes and nanotubes, which are then converted into diamond under the action of a shock wave. The supercritical state of matter is a source of unexpected and, hence, `anomalous' effects. For example, near the critical point, highly developed density fluctuations are observed, which implies fast reversible clustering of the substance.Perhaps, it is this feature that accounts for the considerable technological advantages of the supercritical state for extraction and other processes.58 Synthesis of metallic hydrogen (which has been mentioned above) and the reaction of tritium with hydrogen and deuterium in normal liquid helium and in superfluid quantum helium should be noted as the latest achievements of the `extreme' chemistry.59 It was found that the exceptionally large isotope effects observed in this reaction (this is predictable) are different for normal and quantum helium (this is unexpected). If this finding is confirmed, new and unusual evidence for chemical coherence will be thus obtained. Magnetic-field effects in chemistry have been already dis- cussed within the framework of spin chemistry. All the effects observed in spin chemistry arise as a response of the electron and nuclear magnetic moments to magnetic treatment (a magneto- static field or the magnetic component of a microwave field).In a system of moving electric charges (solvated electrons or ions), the magnetic fields create Lorentz forces, which, in turn, induce a number of other effects. Therefore, the range of magnetic-field effects may prove to be even broader than predicted by pure spin chemistry. Magnetic fields have considerable influence on the rate of migration of dislocations in ionic and atomic crystals such as NaCl and Si (the rate increases 3 ± 6-fold in 4 ± 5 ê magnetic fields),60 on the plastic deformation rate and on the strength of ionic crystals 61 (Fig.13). These effects are of interest for chem- istry and solid-state mechanics. Despite the fact that they can find a reasonable and non-contradictory explanation in terms of the concepts of spin chemistry (spin-dependent recombination of dislocations on paramagnetic stoppers, affecting the rate of migration of dislocations), a contribution of Lorentz forces also a ef . /e0 . e b 2 1.5 s 1072 MPa 1 1.0 0.5 0 e (%) Figure 13. Rate of plastic deformation of a single crystal of NaCl e .f in a 0.7 T magnetic field (in relation to the rate of deformation e .0 in a zero field) as a function of deformation e (a), and deformation e as a function of stress s (b) (the up arrow shows the instant when the magnetic field is switched on, and the down arrow shows the instant of switching off the field).1073 97 cannot be ruled out, even more so because the effects themselves depend on low-frequency (about hundreds of hertz) electric fields. The influence of magnetic fields on the electrodeposition of silver on copper is apparently also associated with Lorentz forces (the yield of the reaction in a *80 ê field at a field gradient of 46103 kê m71 increases to 45%; in addition the density of the dendrites of the deposited metal increases and their quality improves).62 In this case, the magnetic field influences the micro- dynamics and microturbulence of the electric double layer at the electrode ± solution interface via the Lorentz force.(It would be pertinent to evaluate the prospects of this new phenomenon). Unlikchemical radiophysics, based on the magnetic compo- nent of a microwave field, microwave chemistry utilises the electric component of this field. In microheterogeneous systems with localised polar and non-polar regions (e.g. when polar reagents are concentrated in microreactors), microwaves are absorbed by polar and polarisable areas. Actually, these areas function as miniature microwave ovens with addressed supply of heating energy; therefore, reactions which occur in these reactors are highly selective.63 Chemical reactions initiated by ultrasound also occur in microreactors (cavitations).Although the chemical effects in them are specific, they largely resemble those induced by low- temperature plasma and shock waves. Both microwave and ultrasound chemistry are regarded (for good reason) as providing new means for synthetic chemistry. Finally, the chemistry in intense gravitation fields (as well as the chemistry in the state of weightlessness) undoubtedly belongs to the `extreme' chemistry. A sharp increase in the gravitational force of molecules, clusters and associates in these fields should generate new effects; the magnitudes and signs of the concen- tration gradients change, equilibria shift, the arrangement of phases in accordance with density is inverted and the rates and competition of processes alter.64 The opportunities provided by these effects are virtually unlimited; the problem lies only in the availability of technical means needed to implement them in practice.Of course, this would be possible only for high-end technology rather than for large-scale chemical production. The low-temperature region (near 4 K) was assimilated by chemistry relatively long ago. The most outstanding result is the discovery of the quantum mechanism of chemical reactions, i.e. tunnelling through the barrier, and its consequences (enormous isotope effects, temperature-independent limiting rates of reac- tions). Certainly, this also can be classified as `extreme' chemistry. The chemistry dealing with temperatures of 1074 ±1076 K should be regarded as `exotic'. The preparation of ultracold atoms is based on the fact that the velocity of their movement changes upon the absorption of an optical quantum (laser cooling of atoms).If atoms and laser photons are selected in such a way that the absorption occurs in the low-frequency region of the spectrum (`red' edge), then the resonance absorption by an atom moving towards the photons shifts to the line centre and is enhanced, due to the Doppler shift. In the case of `accompanying' atoms, the Doppler effect shifts the resonance away from the centre and attenuates the absorption; consequently, the atoms experience a decelerating force directed along the photon stream. Atoms placed in orthogonal laser beams are decelerated in all the three directions; this creates an optically viscous medium in which the movement of atoms stops, their kinetic temperature being 1074 ± 1076 K (even temperatures of 10710 K can be attained).From ultracold 85Rb atoms, a crystal lattice has been con- structed (it proved to be a body-centred cubic lattice), the lattice parameters were measured by optical diffraction and the frequen- cies of collective lattice vibrations were determined.65 In other words, scientists succeeded in creating a new state of matter � crystalline gas. Ultracold atoms, devoid of kinetic energy, are of interest for precise spectroscopy and metrology, for probing the atom ± atom and atom ± surface potentials, for experimental verification of the postulates of quantum electrodynamics of a single-atomic maser.98 Optical excitation of atoms in the crystalline gas affords electroni- cally excited atoms, which react with other atoms to give excimer molecules implanted into the crystalline gas.The first steps in the chemistry of the energy-free cold atoms and molecules have already been made; its future starts today. Furthermore, the possibility of laser cooling of molecules in liquids is being discussed.66 VII. Once again about cold fusion In the late 80s of the outgoing century, the scientific world was overwrought by a dramatic event�discovery of nuclear reactions accompanying electrochemical synthesis. The brilliant horizons of the cold nuclear fusion immediately became visible; indirect evidence supporting this finding (neutrons, g-radiation, excess thermal effects) was even obtained.However, the euphoria has soon come to an end; experimental mistakes and non-reproduci- bility of the observed effects were found. Consequently, `cold fusion' was wittily renamed to `confusion'. At present, both experimental research and debates concerning cold nuclear fusion have transformed into sluggish processes continued by a small group of enthusiasts. However, the curiosity of this `discovery' has not disappeared. The questions of whether a chemical reaction can induce nuclear fusion and whether transformations of the electron shell can stimulate nuclear transformations still remain open. Probably, the answer is that generation of neutrons can accompany a chemical process; however, neutrons do not result directly from this process but are formed as a secondary product.They can arise upon nuclear decay induced by g- and X-radiation emitted by the electron shell, i.e. they have a chemical nature. Although direct chemonuclear fusion was not accomplished, it has given rise to a new strategy of chemical energetics � coupling of a mechanochemical reaction with a non-branched (or slightly branched) chain photonuclear reaction. This strategy is based on the following idea: mechanically stimulated reactions result in excited electron shells and give rise to g- and X-rays, which are captured by nuclei (photonuclear reaction); the nuclei excited in this way decay to generate new g-quanta and/or neutrons.Achain (or partially branched) chemo- nuclear reaction with energy evolution can occur.The problem is that the mechanical treatment must excite inner electrons; only in this case, conversion of outer electrons to inner vacancies (such as the Auger process) would generate hard X-rays or g-rays. It is clear that shock waves would be the most suitable type of mechanochemical treatment for this purpose. It is also necessary to elucidate theoretically the appropriate type of electron shell compression (this has already been mentioned in relation to endofullereners and diamond synthesis) that would enable exci- tation of high-lying electronic levels of inner electrons (excitation of the outer electrons followed by ionisation would imply escape and inefficient consumption of the mechanical energy).Another problem is to choose an atomic composition of molecules (or mixtures of molecules) that would permit complete capture of g- and X-rays by the nuclei. The capture cross-section in photonuclear reactions is known to be relatively large, its spectrum being fairly broad. This provides grounds for believing that the second problem could be solved more easily than the first one (effective mechanochemical generation of hard radiation). Obviously, this is a strategic goal; on the way to its solution, one can face insurmountable and yet unpredictable obstacles. However, the aim deserves development (at least theoretically, to start with). VIII. Physics of chemical reactions Chemistry, like any other highly developed branch of science, has become a high-risk science.This means that the problems of chemistry are complicated and it is not always possible to solve A L Buchachenko them by existing methods; therefore, the probability of failures and wasted labour, time and resources becomes fairy high. The risks can be reduced only on the basis of deep insight into the physical aspect of chemical reactions, which is the most important and elitist part of chemical physics. The physics of chemical reactions is based on four types of dynamics: �molecular dynamics and molecular organisation; � energy dynamics (energy dynamics in gases, energy migra- tion in clusters, energy transformation in liquids, energy relaxa- tion in solids); �chemical dynamics (potential energy surfaces, n the chemical dynamics, theory of chemical reactions); �angular momentum dynamics and spin dynamics.The physics of chemical reactions is a field of science involving both experimental and theoretical research; its advances are natural and the problems are traditionally clear (they are analysed in Ref. 24). The main point in this field of science is chemical dynamics, the most important aspect of which is the theory of chemical reactions. Chemical dynamics have grown out of chemical kinetics; the latter include a general knowledge that has penetrated into all the branches of chemistry and has promoted the development of universal chemical criteria for reaction mechanisms. However, chemical kinetics deal with the number of chemical events (i.e.the interval between the events), whereas chemical dynamics are concerned with the course of the events themselves (movement of the reactants over the potential energy surface). This is the central event in the circle of which the whole chemistry moves and which lies at the heart of a chemical reaction. Chemical dynamics are the `cardiogram' of this `chemical' heart. The advances in the theory of chemical reactions constitute an individual and broad subject.67, 68 Actually, two large problems need to be solved: the construction of potential energy surfaces and the calculation of the movement of the reactant nucleus in the calculated potential fields (the proper dynamics of a chemical event).Both problems can be solved in various modifications, with several levels of approximation and with different degrees to which the quantum nature of the movement is taken into account. In principle, it is now possible to construct a potential surface for any reaction with any degree of accuracy (taking account of the capabilities of modern computers, one has to be able to reduce a reaction of any desired complexity to a simple one without losing its physical or chemical contents). Reliable methods for the calculation of the dynamic trajectories of the reactant movement over the potential surfaces have now been developed. They include methods of classical trajectories (based on the Hamilton ± New- ton ± Lagrange laws of classical mechanics), semiclassical trajec- tories (taking account of quantum effects via superposition of the initial quantum states of the reactants) and purely quantum trajectories (based on the solution of the SchroÈ dinger equation, which presents the probabilities of reactant transformations via all reaction channels as an S-matrix or scattering matrix). The resulting set of trajectories is used to calculate the canonical k(T) and microcanonical k(E) rate constants (T and E are temperature and energy).Two examples of such calculations are presented in Figs 14 and 15. General methods for the solution of the dynamic problem have been developed: the method of the reaction path Hamilto- nian, the method of the trajectory curvature, multidimensional quantum dynamics and quantum transition state theory.The main idea of the dynamical methods is to distinguish the move- ment of the reactants along the reaction coordinate and to take into account the influence of all the other types of movement (both `assisting' and retarding). The interaction of all the types of movement (both those directed along the coordinate and others) is considered to be `dynamic friction' and, among other tasks, the modern theory should provide calculation of its value.Chemistry on the border of two centuries �achievements and prospects a 2 3 4 r(I7I)/ A 5 5 b r(I7I)/ A 65 v(BrI)=2 J=35 4 v(I2)=0; ET=145 cm71 3 3 4 5 6 2 Figure 14. Potential energy surface for the reaction Br+I2!BrI+I (a) and one of the trajectories on this surface (b) with a duration of 17.6 ps; n are vibrational states, J is the rotation moment and ET is the kinetic energy.However, the great theoretical power of chemistry has an unpleasant aspect. Owing to the broad and successful application of quantum chemistry and molecular dynamics methods, researchers tend to avoid science-dependent and expensive experi- ments, this tendency becoming more and more pronounced. This is objective reality and the general trend of development of world's chemistry (perhaps, not only chemistry). The theory of chemical reactions is closely related to the theory of nuclear reactions (this is true not only for the diffusion regime in which the kinetic equations and their solutions are absolutely identical).A system that undergoes a chemical reaction in a liquid or in a solid is similar to the reacting nucleons surrounded by the medium, i.e. other nucleons in a nucleus. In this case, too, chemistry occupies an advantageous position: it is simple in the sense that only one type of interaction operates, namely, the Coulomb interaction with a known energy ± distance relationship. This is a fundamental statement, which follows from quantum mechanics (the accuracy of its laws is beyond contro- versy, like that of the laws of classical mechanics). However, there is also direct experimental evidence supporting this statement. The case in point is the experimental probing of atomic wave func- tions, which are the basis for the entire chemistry.2 3 4 r(I7Br)/ A r(I7Br)/ A E/ kJ mol71 150 100 500 750500 7 E/ kJ mol71 150 100 500 750 500 Figure 15. Free energy surfaces for the reaction Cu2+?Cu+ at an electrode for adiabatic (a) and nonadiabatic (b) transformation; z is the distance of the ion from the electrode, E is the energy of solvent reorganisation. This experiment is very simple: an atomic (Ca or Cu) beam is excited by polarised light to the pz or dz2 state; after that, this beam with spatially oriented pz or dz2 orbitals is exposed to a crossed beam of He atoms; scattering of the He beam is detected, the scattering pattern being used to reconstruct the scattering poten- tial. The result proved to be very vivid: the scattering profile exactly reproduced the shapes of the atomic orbitals.One can argue that this result is trivial (in the sense that no other result was to be expected); however, it leads to the under- standing of the fact that only the Coulomb potential operates in chemistry (a part of it, the exchange interaction, is spin-depend- ent). Moreover, if we recall the fact that the atomic wave functions have originated on the tip of a pen from the quantum-mechanics postulates and the SchroÈ dinger equation with a centrosymmet- rical Coulomb potential, we can feel the triumph of the human mind. The magic and fascination of chemistry are due to the fact that the Coulomb potential gives rise to a variety of chemical bonds (ionic, covalent, donor ± acceptor, dative, hydrogen and van der Waals bonds), chemical species (atoms, carbenes, ions, molecules, exciplexes, Rydbergs, excimers, van der Waals molecules, com- plexes, radicals, etc.), states (ground, excited, charge, spin, orbital, differing in symmetry, etc.), mechanisms of chemical transforma- tions and controlling factors (energy, spin, phase, orbital symme- try).With this diversity and wealth, chemistry is always new. The discovery of new states of hydrogen molecules can serve as an example (Fig. 16). They are matched by adiabatic potential minima located in the vicinity of the points of intersection of the binding term H++H7(1s2) and the repulsive terms H(1s)+H(np), where n = 1, 2, 3,.. . These species are not Rydberg molecules having `plump' electron shells with principal quantum numbers reaching hundreds but `normal' molecules with a 0 7500 8 b 0 7500 8 99 Cu+ Cu2+ z /A 3 4 5 6 Cu+ Cu2+z /A 3 4 5 6 7100 H+, H7 H, H (3p) H, H (2p) H, H (1p) 0.7A Ê 6AÊ 106A Ê 13A Ê Firstly, the time periods of 10 ± 100 fs are so short in relation to the period of nuclear vibrations that the nuclei remain virtually fixed at their positions during this period.For example, the interatomic distances in a iodine molecule in the ground state oscillate between 2.5 A Ê (contracted bond) and 5 A Ê (stretched bond). A femtosecond pulse `catches' a molecular assembly with a particular interatomic distance and a particular mo phase and throws it almost instantly onto a new potential surface.On this potential, atoms start a new movement, which is now synchronous (coherent). Thus the femtosecond pulse has created a wave packet � an assembly of molecular oscillators with a fixed vibration phase, a given interatomic distance and a definite energy. In other words, femtochemistry opens up the way for production of coherent wave packets. Secondly, a time resolution of 1 ± 100 fs corresponds to a coordinate resolution of 0.1 ± 0.01 A Ê . This means that the move- ment of the nuclei over the potential surface, including that on the top of the barrier and near it, is monitored with this coordinate resolution. It is clear that we are dealing with spectroscopy and chemistry of the transition state, which has always been an object for investigation by theorists but now has become the subject of experimental research.Femtochemistry deals with the time of movement of reaction systems on the potential surface and introduces into chemistry the experimental chemical dynamics. Thirdly, by monitoring the movement of nuclei over the potential surface, one can interfere in the dynamics of the transition state by its energy or phase pumping and thus change the transformation pathways of the transition state itself. In this sense, one can say that femtochemistry is the chemistry of the transition state. Potential energy R/ A Figure 16. Terms of metastable hydrogen molecules with interatomic distances 6, 13 and 106 A.They are located in adiabatic valleys at the intersection of the ionic term (H+, H) with the terms corresponding to the orbitally excited states H (1p), H (2p) and H (3p). equilibrium interatomic distances of 6, 13, 106 A Ê , etc. character- ised by their own vibrational-rotational spectra. Similar molecules can also exist for other atoms. They can be detected in low- temperature or rarefied plasma or in the huge interstellar chemical reactor. IX. Atomic-resolution spectroscopy and chemistry This field of modern chemistry is being vigorously developed; it assumes detection of chemical reactions with high time resolution (about410 fs, which is shorter than the period of vibrations of an atom in a molecule) or with high space resolution (*1 ± 5A Ê , which is about the size of a single atom or a molecule).Both values correspond to atomic resolution. The adoption of short laser pulses has extended the scope of chemistry and introduced into chemistry new ideas, which have been embodied in femtochemistry.69 A L Buchachenko As an example of the control of chemical reaction channels, let us consider decomposition of a chlorine dioxide molecule OClO. When the pumping energy is*0.361012Wcm72, decomposition follows predominantly the O+ClO pathway; when the energy exceeds 0.861012 W cm72, the second channel leading to O2 + Cl, starts to predominate and at an energy of 1.461012 W cm72, the rate of decomposition via the second channel is 12 times greater than that according to the first channel.70 Chemical approaches also permit complete switching of the reaction from the first channel to the second one, for example, this can be attained by adding a water molecule to chlorine dioxide or by placing chlorine dioxide in an (H2O)n cluster.71 Although the chemical method leads to the same result as femtochemistry, this result is achieved by an absolutely different mechanism.In fact, when the initial state of the reactants changes, the reaction switches to a new potential surface; thus, the chemical approaches represent a classical example of changing the reactivity upon complex formation or solvation. Fourthly, femtochemistry has opened a new chapter in the structural analysis of short-lived species including transition states.Instruments which permit ultrafast diffraction of electrons on these species have been designed. The first femtosecond pulse from a photocathode generates an ultrashort electron beam accelerated by an intense electric field (*30 kW cm71). After that, the second femtosecond pulse (with a controlled delay) generates a wave packet or a transition state for a chemical system under study. After that, electron beam diffraction on the tran- sition state or on the reaction product is recorded. In the future, this remarkable equipment would permit probing of the transition state geometry.72 A resolving power of *1 ps with respect to the electron beam (which is the factor limiting the capability of the instruments) has already been attained, and this limit can undoubtedly be overcome. This would enable not only time (spectroscopic) but also geometrical monitoring of the transition state (based on the electron diffrac- tion).Finally, fifthly, femtochemistry provides great possibilities for investigating ultrafast reactions such as proton (electron) transfer in acid ± base (donor ± acceptor) pairs; tautomerisation; cis ± trans-isomerisation with detection of twist-conformations; photo- generation and relaxation of an electron ± hole pair in semicon- ductors; elementary steps of photosynthesis; decomposition of molecules (e.g., decomposition of iodine in clusters near the gas ± liquid phase transition or under supercritical conditions, etc.). Femtochemistry reveals important new facts concerning het- erogeneous processes, in particular, desorption of molecules and desorption mechanism.Afemtosecond pulse heats an electron gas to a temperature of several thousand degrees but leaves phonons in metal almost cold (heated to a temperature of only several hundred degrees). The low phonon temperature cannot account for either the high rates of desorption or the high vibrational temperature (*2000 K) of diatomic molecules. It is clear that the fast desorption of molecules (not more than 10 vibrations on the surface after a femtosecond pulse) is stimulated by the hot electron gas (apparently, via generation of an electron ± hole pair followed by its fast annihilation at the adsorption site).73 Independent experiments with generation of a surface electromagnetic wave, which probes the surface and its adsorption coverage, has con- firmed that, after a femtosecond pulse has been initiated, a molecule remains on the surface only for 3 ± 5 vibrations.Apparently, femtochemistry would enable detection of fast electronic events such as electron exchange between the protons in the Há2 cation (these events take *2 fs), p-electron migration in the Kekule structures of benzene, electron revolution round the nucleus, electron transfer to upper orbitals upon excitation, etc. However, this would require lasers of the next generation with a resolving power in the attosecond range (1 attosecond=10718 s). The progress in the design of photodetectors and scanning optical microscopes has given access to observation of singleChemistry on the border of two centuries �achievements and prospects molecules (such as rhodamine 6G, and terrylenediimine).By exciting and monitoring stationary luminescence of single mole- cules (molecular `fireflies'), one can study lateral diffusion of molecules on solid surfaces and bulk diffusion in solids (such as polymers). When polarised light is used for excitation and polarised luminescence is recorded, it becomes possible to meas- ure reorientation of molecules and monitor their trajectories. The first wonderful phenomena have already been discovered. Thus an unusual type of behaviour of rhodamine 6G molecules deposited on glass was found; some of the molecules remain still, some other move without rotating and some other rotate without moving.74 The behaviour of terrylenediimine on an SiO2 surface proved to be even more surprising: luminescence of some molecules was found twice modulated by a low frequency (10 and 100 Hz), the luminescence bands of `modulated' molecules being shifted to higher frequency.75 Even these two examples demonstrate that probing of a surface by molecular `fireflies' promises a lot of new discoveries.The capability of optical detection using single molecules is not limited to investigations of molecular dynamics. A `flickering' (non-istical) behaviour of single molecules associated with spectral diffusion has been found; it is also possible to measure directly the rate of energy transfer to neighbouring molecules and to elucidate the dependence of the energy on the distance (by changing the number of `fireflies' per unit surface; normally several molecules are located on one square micron).76 It is obvious that the potentialities of detection using single molecules are substantially increased by using pulse excitation.This brings about new opportunities in the investigation of ergodicity and new methods for direct measurement of quantum yields in the photodecay of single molecules, etc. The discovery of tunnelling vibrational spectroscopy of single molecules is a new great breakthrough in chemistry.77 The tunnelling current, which flows between the tip of a tunnelling scanning microscope and a solid surface on which an adsorbed molecule `sits' under the tip, has a clear-cut resonance character. The resonance always occurs in the case where the tip potential (and, hence, the energy of tunnelling electrons) corresponds to the vibronic levels of an adsorbed molecule.Examples of vibrational spectra for several single molecules are shown in Fig. 17. Vibrational spectroscopy of a single molecule exactly repro- duces its portrait and permits one to follow its dynamics (rest time) and chemical fate.78 This opens up new prospects for the science of a I/ nA 74 76 710 9.0 10.0 c I/ nA 75 710 715 720 8.0 E/ V 9.0 b I/ nA 0 710 720 E/ V 7.6 E/ V 7.2 Figure 17. Tunnel current as a function of the potential for alu- mina (a), oxygen on titanium (b) and a water molecule on rutile (c).A clear-cut vibrational structure (vibrational spectrum of single molecules) is manifested. 101 catalysis. The ways for detecting electron paramagnetic resonance of a single spin are clear and have already been theoretically substantiated.79 A great number of new ideas are originating in atomic-resolution chemistry. X. Conclusion Chemistry as a fundamental science has no limits; it is impossible to represent even its outline within one paper. The face of chemistry is being permanently renewed; new beautiful features appear, which include coherence, atomic resolution (in time and in space), magnetic script of chemical reactions, chemical radio- physics, new ideas in molecular architecture and new molecular assemblies.A special part in the modern chemistry is assigned to the coherent chemistry, which introduces a third controlling factor, the phase (in addition to the energy and spin). In this review, we have discussed the quantum vibrational coherence, the quantum spin coherence and the macroscopic coherence with characteristic times of 10715 ± 10713, 1079 ± 1077 and 1071 ± 1073 s, respectively. A new level of this field of science would be associated with the generation of electron coherence, i.e. creation of electronically coherent wave packets with characteristic times of 10718 ± 10715 s. The intrinsic order and the structure of chemistry and also everything created by chemistry display the charm of strict logic, substantiation and perfection.This is yet another feature of chemistry, which makes it a fundamental science. The scope of chemical research is constantly expanding; it is possible (and useful) to make predictions only regarding the tendencies of development. We cannot foresee the particular events in science; this is clearly demonstrated by the achievements of the last three decades (fullerenes, telomerase, superconducting ceramics, etc.). However, we can make use of our scientific intuition, foresight and professional sense of understanding of the general movement of science and thought. The author is grateful to the Russian Foundation for Basic Research (Projects Nos 96-3-34193, 96-15-97285), the INTAS (Grant 96-1269) and the Federal programme `Integration' (Grant K-0093) for financial support.References 1. V A Legasov, A L Buchachenko Usp. Khim. 55 1949 (1986) [Russ. Chem. Rev. 55 1113 (1986)] 2. V Q Nguyen,M Sadilek, J Ferrier, A J Frank, F TurecÏ ek J. Phys. Chem. A 101 3789 (1997) 3. M Trollsas, J Hedrick J. Am. Chem. Soc. 120 4644 (1998) 4. A Blake, N Champness, A Khlobystov, D Lemenovskii, W-S Li, M SchroÈ der J. Chem. Soc., Chem. Commun. 1339 (1997) 5. D Chung, in International Conference on Composites, Las Vegas, 1998 6. H M McConnell J. Chem. Phys. 39 1910 (1963) 7. A L Buchachenko Mol. Cryst. Liq. Cryst. 176 307 (1989) 8. K Awaga, Y Maruyama Chem. Phys. Lett. 158 556 (1989); J. Chem. Phys. 91 2743 (1989) 9. A L Buchachenko Usp.Khim. 59 529 (1990) [Russ. Chem. Rev. 59 307 (1990)] 10. H Tamaki, Z Zhorg, N Matsumoto, S Kida, M Koikawa, N Achiwa, Y Hashimoto, H Okawa J. Am. Chem. Soc. 114 6974 (1992) 11. L O Atovmyan, G V Shilov, R N Lyubovskaya, N S Ovanesyan, Yu G Morozov, S I Pirumova, E I Zhilyaeva, I G Gusakovskaya Pis'ma Zh. Eksp. Teor. Fiz. 58 818 (1993) 12. Ya Yamaguchi, S Maruyama Chem. Phys. Lett. 286 337; 343 (1998) 13. T Yildirim, J Fischer, R Dinnebier, P Stephens, C Lin Solid State Commun. 93 269 (1995) 14. K Imaeda, J KroÈ ber, Ch Nakano, H Inokuchi, K Ichimura J. Phys. Chem. B 101 10 136 (1997) 15. Ch-Ch Chen, Ch Lieber J. Am. Chem. Soc. 114 3141 (1992) 16. S Stevenson, P Burbank, K Harich, Z Sun, H C Dorn, P H M van Loosdrecht, M S de Vries, J R Salem, C-H Kiang, R D Johnson, D S Bethune J.Phys. Chem. A 102 2833 (1998)102 17. F Hensel, P Edwards Science 271 1693 (1996) 18. J Laskin, T Peres, Ch Lifshitz, M Saunders, R J Cross, A Khong Chem. Phys. Lett. 285 7 (1998) 19. B Pietzak,M Waiblinger, T Almeida Murphy, A Weidinger, M HoÈ hne, E Dietel, A Hirsch Chem. Phys. Lett. 279 259 (1997) 20. S Knorr, A Grupp, M Mehring, U Kirbach, A Bartl, L Dunsch Bruker Rep. 145 40 (1998) 21. T Hertel, R Martel, P Avouris J. Phys. Chem. B 102 910 (1998) 22. M Saunders, R J Cross, H Jimenez-Vazquez, R Shimshi, A Khong Science 271 1692 (1996) 23. J Hovis, R Hamers J. Phys. Chem. B 101 9581 (1997) 24. A L Buchachenko Usp. Khim. 56 1593 (1987) [Russ. Chem. Rev. 56 907 (1987)] 25.R C Bell, K A Zemski, P K Kerns, H T Dehg, AWCastleman Jr J. Phys. Chem. A 102 1733 (1998) 26. W Lu, Sh Yang J. Phys. Chem. A 102 825 (1998) 27. A Heikal, S Chong, J Baskin, A Zewail Chem. Phys. Lett. 242 380 (1995) 28. B Dabbousi, J Rodriguez-Viejo, F Mikulec, J Heine, H Mattoussi, R Ober, K F Jensen, M Bawendi J. Phys. Chem. B 101 9463 (1997) 29. S Sarkar, A K Karmakar, R N Joarder J. Phys. Chem. A 101 3702 (1997) 30. K Egashira, N Nishi J. Phys. Chem. B 102 4054 (1998) 31. A H Zewail J. Phys. Chem. 100 12 701 (1996) 32. T Kobayashi,M Kim,M Taiji, T Iwasa, M Nakagawa,M Tsuda J. Phys. Chem. B 102 272 (1998) 33. T Kakitani, R Akiyama, Y Hatano, Y Imamoto, Y Shichida, P Verdegem, J Lugtenburg J. Phys. Chem. B 102 1334 (1998) 34. G Fleming, R van Grondelle Phys. Today 47 48 (1994) 35. M GroÈ nager, N Henriksen J. Phys. Chem. A 102 4277 (1998) 36. O Anisimov, V Bizyaev,N Lukzen, V Grigoryants, Yu Molin Chem. Phys. Lett. 101 131 (1983) 37. G Kothe,M Bechtold, G Link, E Ohmes, J-U Weidner Chem. Phys. Lett. 283 51 (1998) 38. W Hohmann, D Lebender, J Mmller, N Schinor, F Schneider J. Phys. Chem. A 101 9132 (1997) 39. V Votrubova, P Hasal, L Schreiberova,M Marek J. Phys. Chem. A 102 1318 (1998) 40. H Okamoto, N Tanaka,M Naito J. Phys. Chem. A 101 8480 (1997) 41. H Degn, A Holden, L Olsen (Eds) Chaos in Biological Systems (NATO ASI Ser. A) Vol. 138 (New York: Plenum, 1987) 42. Z Li, J Cai, Sh Zhou J. Phys. Chem. B 102 1539 (1998) 43. T Bronnikova,W Schaffer,M Hauser, L Olsen J. Phys. Chem. B 102 632 (1998) 44. S Jakubith,H H Rotermund,W Engel, A von Oertzen,G Ertl Phys. Rev. Lett. 65 3013 (1990) 45. K M Salikhov, Yu N Molin, R Z Sagdeev, A L Buchachenko Spin Polarization and Magnetic Effects in Radical Reactions (Amsterdam: Elsevier, 1984) 46. N J Turro, M-F Chow, C-J Chung, C-H Tung J. Am. Chem. Soc. 105 1572 (1983) 47. V B Ivanov Vysokomol. Soedin., Ser. A 33 1811 (1991) a 48. A L Buchachenko Chem. Rev. 95 2507 (1995) 49. A L Buchachenko, V L Berdinskii Usp. Khim. 52 3 (1983) [Russ. Chem. Rev. 52 1 (1983)] 50. A Buchachenko, E Frankevich Chemical Generation and Reception of Radio- and Microwaves (New York: VCH, 1994) 51. A L Buchachenko, V L Berdinsky J. Phys. Chem. 100 18 292 (1996) 52. R van Eldik, C Hubbard (Eds) Chemistry under Extreme or Non-Classical Conditions (New York: Wiley, 1997) 53. D Card, D Folmer, S Sato, S Buzza, A Castleman Jr J. Phys. Chem. A 101 3417 (1997) 54. J Kou, N Nakamishima, S Sakable Chem. Phys. Lett. 289 334 (1998) 55. P Borowicz, J Hotta, K Sasaki, H Masuhara J. Phys. Chem. B 102 1896 (1998) 56. D Dlott, S Hambir, J Franken J. Phys. Chem. B 102 2121 (1998) 57. Y Zhu, T Sekine, T Kobayashi, E Takazawa, M Terrones, 58. A Yu Shaulov, N I Andreeva, A L Buchachenko, N S Enikolopov 59. Y Aratono, T Matsumoto, T Takayanagi, T Kumada, 60. V I Al'shits, E V Darinskaya, E Yu Mikhina, E A Petrzhak H Terrones Chem. Phys. Lett. 287 689 (1998) Zh. Eksp. Teor. Fiz. 63 157 (1972) K Komaguchi, T Miyazaki J. Phys. Chem. A 102 1501 (1998) Fiz. Tv. Tela 38 2426 (1996) A L Buchachenko 61. Yu I Golovin, R B Morgunov Chemistry Rev. (Russia) 24 1 (1998) 62. Y Tanimoto, A Katsuki, H Yano, Sh Watanabe J. Phys. Chem. A 101 7359 (1997) 63. Microwave-Induced Chemistry, Fundamentals and Applications (New York: American Chemical Society, 1997) 64. G A Abakumov, V B Fedoseev Ross. Khim. Zh. 42 36 (1998) b 65. A GoÈ rlitz, M WeidemuÈ ller, T HaÈ nsch, A Hemmerich Phys. Rev. Lett. 78 2096 (1997) 66. J Clark, P Miller, G Rumbles J. Phys. Chem. A 102 4428 (1998) 67. A I Voronin, V I Osherov Dinamika Molekulyarnykh Reaktsii (Dynamics of Molecular Reactions) (Moscow: Nauka, 1990) 68. V Benderskii, D Makarov, C Wight Chemical Dynamics at Low Temperatures (New York: Wiley, 1994) 69. V SundstroÈ m (Ed.) Femtochemistry and Femtobiology (Singapore: World Scientific, 1998) 70. M Blackwell, P Ludowise, Y Chen J. Chem. Phys. 107 283 (1997) 71. J ThoÈ gersen, C Thomsen, J Poulsen, S Keiding J. Phys. Chem. A 102 4186 (1998) 72. D Zhong, A Zewail J. Phys. Chem. A 102 4031 (1998) 73. J Prybyla, T Heinz, J Misevich, M Loy, J Glownia Phys. Rev. Lett. 64 1537 (1990) 74. A Ruiter, J Veerman,M Garcia-Parajo, N van Hulst J. Phys. Chem. A 101 7318 (1997) 75. K Weston, S Buratto J. Phys. Chem. A 102 3635 (1998) 76. S Mais, J Tittel, Th Basche', C BraÈ uchle, W GoÈ hde, H Fuchs, G MuÈ ller, K MuÈ llen J. Phys. Chem. A 101 8435 (1997) 77. F Dalidchik, S Kovalevsky, M Grishin, B Shub Surf. Sci. 387 50 (1997) 78. F I Dalidchik, S A Kovalevskii,M V Grishin, B R Shub Pis'ma Zh. Eksp. Teor. Fiz. 66 37 (1997) 79. A L Buchachenko, M A Kozhushner, B R Shub Izv. Akad. Nauk, Ser. Khim. 1732 (1998) c a�Polym. Sci. (Engl. Transl.) b�Mendeleev Chem. J. (Engl. Transl.) c�Russ. Chem. Bull. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (2) 103 ± 118 (1999) Simulation of nanotubular forms of matter AL Ivanovskii Contents I. Introduction 103 II. Theoretical models of electronic properties of ideal carbon nanotubulenes 104 III. Defect structure of carbon nanotubulenes. Nanotubular composites 106 IV. Electronic properties of several non-metal compounds in the nanotubular form 109 V. Theoretical models of the growth and morphology of nanotubes 112 VI. Nanotube crystals 114 VII. Conclusion 116 Abstract. Data on the electronic and chemical structure of a new quasi-one-dimensional form of matter, viz., nanotubulenes, are generalised and systematised. Methods and approaches used in modern quantum chemistry for the simulation of the composition, structure, and properties of isolated tubulenes based on layered phases (graphite, boron nitride, boron carbide and boron car- bonitride), nanotubular composites and nanotube crystals are described.The role of quantum theory in the development of the concepts of fundamental properties of substances in the nano- tubular form and methods of their targeted modification is discussed. Prognostic potentials of theoretical models in solving material science problems are considered. The bibliography includes 197 references. I. Introduction The study of nanomaterials (NM) is one of the leading areas of investigation in modern chemistry. Nanomaterials that are cur- rently produced in the form of isolated clusters, powders, films, and consolidated systems with a characteristic size of crystallites not more than 100 nm possess many unusual properties that are promising for the development of high technologies.1 ±8 In addi- tion, they form a unique class of objects used for studying fundamental problems of the metastable state of matter and the prospects of producing its new forms that have no analogues in nature.2, 6, 8 The greatest progress in this area has been achieved in the field of synthesis and comprehensive studies of a new allotropic modification of carbon, fullerenes, and first of all, `classical' spheroidal carbon clusters (C60 and C70), endohedral complexes, as well as filmand crystalline materials based on these substances (fullerites).1, 2, 8 The topology and physicochemical properties of nonspherical carbon clusters built of a large number of atoms (the so-called giant fullerenes Cn, n>70) have become the subject of numerous A L Ivanovskii Institute of Solid State Chemistry, Ural Branch of the Russian Academy of Sciences, ul.Pervomaiskaya 91, 620219 Ekaterinburg, Russian Federation. Fax (7-343) 274 44 95. Tel. (7-343) 249 30 82 Received 21 April 1998 Uspekhi Khimii 68 (2) 119 ± 135 (1999); translated byAMRaevsky #1999 Russian Academy of Sciences and Turpion Ltd UDC 541.16 recent studies in which different types of closed carbon structures (capsulenes, torrenes, barrellenes, etc.) were discussed.1, 2 Iijima 9 has discovered quasi-one-dimensional (quasi-1D) hol- low tubular carbon fibres, viz., nanoscale carbon tubes (nano- tubulenes), which are extended cylinders whose walls are formed by rolled graphite sheets.Carbon nanotubes with diameters ranging fromfive to several tens of nanometres and consisting of 2 to 50 coaxial graphite cylinders [the so-called multilayered nanotubes (MNT)] were detected 9 by transmission electron microscopy (TEM) in the soot obtained in arc-discharge experi- ments using graphite electrodes. More recently, both multilayered and single-layered nanotubes (SNT) were obtained.10± 16 It was suggested8,10 to consider nanoscale tubular structures as a new, quasi-1D form of carbon that occupies a `vacant' position in the series of the known forms of carbon of different topological dimension 3D (diamond), 2D (graphite), and 0D (fullerene). Almost simultaneously with the first reports on the syntheses of carbon NT,10 ±16 several studies devoted to simulation of their electronic structure were carried out,17± 24 and unexpected and extremely interesting results were obtained.It was established that, depending on the NT diameter and the type of helical ordering of carbon atoms constituting the cylinder walls, tubu- lenes can possess both metallic and semiconducting properties. The predicted possibilities of targeted modification of carbon NT conductivity have at once become the subject of daring hypotheses for their use in diverse exotic devices and elements of electronic circuits of uniquely small sizes (nanocoils, nanowires, etc.).10, 16 The list of possible `nanoscale devices' in which elec- tronic, optical, catalytic, capillary, and other properties ofNTcan be used, is increasing rapidly.25± 31 In the first stage of experimental studies, methods of produc- tion of carbon NTand a number of nanotubular composites were developed.The results of there studies were generalised in a monograph.8 Recently, stable nanotubular structures based on several layered compounds, viz., BN,32 ±36 BC3,37 BC2N,37, 38 and MoS2 39, 40 have been synthesised and it was suggested that nano- tubular forms of silicon 41, 42 and boron 43 as well as consolidated nanotubular materials 44, 45 can be obtained. Research in the field of the synthesis and study of the proper- ties of nanotubular forms of substances is intimately associated withthe development of quantumchemistry of tubulenes, whichis caused by several reasons. The interest of investigators in quasi- 1Dcarbon NT and initiation of large-scale projects in the field of104 synthesis, physical chemistry, and material science of NT is to a great extent due to predicted unusual electronic properties of these objects.17 ¡¾ 24 Nanotubulenes are obtained by vapour deposition in various media (see, e.g., Refs 6, 8).For instance, carbon tubulenes are synthesised in arc-discharge apparatus in the presence of catalysts. As a rule, the final product is a mixture of NT and other carbon- containing species (e.g., metal carbides), which hampers substan- tially the studies of the properties of NT themselves.Suffice it to say that the isolation of NT possessing semiconducting and metallic properties predicted in 1992 17 ¡¾ 24 was reported only recently.46, 47 Theoretical simulation of NT plays an extremely important role in the solution of the problem of obtaining the nanotubular substances `in pure form': almost all the known substances in the nanotubular form and many of their properties were predicted from the results of calculations. In the present review, the main stages of the development of ideas of the electronic properties of substances in the nanotubular form calculated first of all using modern quantum-chemical methods are considered. Particular attention is paid to the application of theory to the simulation of the structure and description of physicochemical properties of non-ideal nanotubu- lenes, nanotubular composites, and nanotubular crystals that are most interesting from the viewpoint of their practical use. II.Theoretical models of electronic properties of ideal carbon nanotubulenes The principles of theoretical description of ideal carbon single- layered NT were developed in 1992.17 ¡¾ 24 Schemes of NT classi- fication were proposed and fundamental dependences between the atomic structure of cylindrical nanostructures and their electronic properties were established. Saito et al.18 considered the principles of the formation of two types of NT taking closed tubulenes as an example. Fullerene C60 hemispheres were taken as the end groups; they can be obtained as a result of dissecting the C60 framework in such a manner that the section plane contained 9 or 10 carbon atoms (Fig.1). This respectively corresponds to the so-called `arm-chair' or `zigzag' cross sections. Since the type of ordering of the atoms constituting the cylinder walls closed by the above-mentioned end groups will be determined uniquely, the corresponding terminology is also used for the description of the NT obtained, and `arm-chair' or `zigzag' NT are considered (Fig. 2). `Open' NT can also be constructed by rolling a planar atomic graphite sheet. In this case each cylinder is characterised by its diameter and the type of helical atomic ordering (chirality).18, 19 ab Figure 1. `Arm-chair' (a) and `zigzag' (b) cross sections of fullerene C60.A L Ivanovskii a b Figure 2. Structure of `arm-chair' (a) and `zigzag' (b) tubulenes. Using the basis vectors of the graphite sheet a1 and a2 (Fig. 3), the diameter (d), and the pitch angle (y), the vector cn=na1+ma2 is given [the so-called (n, m)-NT]. The parameters d and y are defined by the following relationships: p d a jcnj p a apAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 3On2 a m2 a mnU , where a is the interatomic distance in the planar sheet, and y a arctan ¢§2n a m 3AAAAAA m p A . cn A0 y A a1 a2 Figure 3. The method for selecting the vector cn that defines the diameter and the pitch angle y for an `open' nanotube that can be constructed by rolling a graphite sheet (by matching the points A and A0).Depending on the indices n and m, chiral (n, m) and achiral (n, 0) nanotubes are distinguished. Calculations of electronic properties for a large number of SNT with variable indices n andmwere performed.17 ¡¾ 24 In Fig. 4, the distributions of the density of states (DOS) calculated by the tight-binding (TB) method for the (10, 0) and (9, 0) `zigzag' carbon NT19 are shown as an example. The energy bands are formed by mixed p(2pz) and s(2s,2px,y) carbon states. Depending on the radius of curvature, carbon nanotubulenes can possess either metallic or semiconducting properties. According to calculations,17 ¡¾ 24 there is a clear dependence between the structural indices (n, m) and conducting properties of ideal cylindrical single-layered tubulenes.All carbon (n, n)-NT possess a metallic type of conductivity. If m=0 and n=m+3q, where q is an integer, then the corresponding tubulenes are semiconductors with narrow band gaps, for which the band gap (DEg) depends on the NT radius DEg^R72 . All other NT possess properties of semiconductors with wide band gaps, for which DEg^R71 .Simulation of nanotubular forms of matter a DOS (rel. units) 1.0 0.5 b 1.0 0.5 0 72 074 Figure 4. The total density of states (TDOS) for the `zigzag' (10, 0)- (a) and (9, 0)-carbon nanotubulenes (b).19 The TDOS for a planar graphite sheet is shown by dashed lines. Calculations by the TB method. It is natural that the relationships listed above are valid for the tubes of relatively small diameters, whereas at d?? the band structure ofNTis transformed into that of a planar graphite sheet.This fundamental dependence of electronic properties of NT on their structure17 ± 24 was confirmed experimentally.46, 47 Two groups of NT with varied chirality were isolated and their current vs. voltage characteristics were measured. In accordance with theoretical predictions, a metallic or semiconducting type of conductivity was observed for these groups of NT depending on their structure. Refinement of the details of electronic structure of carbon SNT was reported.48 ± 64 For instance, it was pointed out that s ± p-hybridisation should be taken into account in the case of semiconducting tubulenes of small radius.50 A detailed study of this effect for NT with wide band gaps in the framework of the TB method made it possible to suggest an analytical dependence of DEg on the tube radius.34 The results of some calculations of (n, m)-NT with wide band gaps are listed in Table 1.Table 1. Dependence of DEg for carbon (n, m)-NT with wide band gaps on their radii.26, 54 R /A (n, m) DEg /eV see a 0.79 1.12 0.65 0.80 2.7 3.1 3.9 4.3 (7, 0) (8, 0) (10, 0) (11, 0) 1.00 1.22 0.85 0.88 a Calculations 54 by the TB method in the parametrisation by Tomanek and Schluter.59 b Calculations 54 by the TB method in the parametrisation by Su et al.65 c Data taken from Ref. 26. 2 E (rel. units) see c see b 1.11 1.33 0.87 0.96 105 b a XG d c DOS /eV atom71 EF EF 0.4 12 0.20 2 E/eV 2 E/eV 0 0 74 72 74 72 Figure 5.Dispersion of the energy bands (a, b) and the density of p (solid line) and s (dashed line) states of NT, (c) and (d), respectively; (a) and (c) (10, 5)-tubulene; (b) and (d) (9, 0)-tubulene. The d-shaped DOS resonances near the Fermi level are shown by arrows. The dependence of the DOS distributions on the NT radii and pitch angles was considered taking a group of metallic NT as an example.50 The presence of a DOS `plateau' near the Fermi level between the d-shaped peaks (Fig. 5) separated by the E interval is a characteristic feature of carbon nanotubulenes with metallic (and quasi-metallic) types of conductivity [e.g., the (10, 5) and (9, 0)-NT, respectively]. A general relation describing the E(y, R) dependence was derived and numerical calculations of this spec- tral parameter were performed (Fig.6). As can be seen, the E value (the width of the DOS `plateau' near the Fermi level) is determined by the tube size, viz., E decreases substantially as R increases. For this reason, the parameters of the electronic states of tubulenes possessing metallic and semiconducting properties depend explicitly on their geometric indices. E/Vp ± p 1 �1 �2 �3 �4 0.5 4 3 2 R/d0 Figure 6. Dependence of the energy interval (E/Vp± p) between the d-shaped DOS resonances near the Fermi level (see Fig. 5) for NT with metallic (1), (2) and quasi-metallic (3), (4) types of conductivity on their radii (R) referred to the C7C interatomic distance (d0); Vp± p is the energy parameter used in TB calculations and determining the interaction between p ± p orbitals; (1), (2) y=308; (3), (4) 04y430 8.The Peierls type of instability effects for metallic 1D-tubulenes was considered.17 ± 19 Based on the results of ab initio self-consis- tent calculations, it was shown17 that, unlike, e.g., trans-polyace- tylene,58 the temperature of the Peierls transition for the metal-like NT should be much higher than room temperature.59106 The dependence of the stability of carbon tubulenes on their sizes was evaluated 60 and compared to similar data for other carbon nanoscale structures and macrostructures. The analysis was based on calculations of carbon atom binding energies (Eb) in NT.The calculations were carried out using two empirical po- tentials 61, 62 for a total of 169 nanotubes with R<9 A and pitch angles in the range 0 8<y<308 (the C7C distances were assumed to be equal to 1.44 A). Several highly symmetric (n, n)- NT (n=3, 4, 5, 6, 7, 9) were also calculated in the framework of the ab initio local electron density functional (LEDF) method.63, 64 Figure 7 illustrates a clear dependence of the Eb of carbon atom on the NT geometric size (Eb*R72), according to which the binding energy of a carbon atom in NT approaches that of a carbon atom in graphite as the NT radius increases. As can also be seen, the Eb values of carbon atoms in fullerenes C60, C180, and C240 are higher than those of carbon atoms in the NT of corresponding radii.It was concluded that the formation of tubular or spheroidal nanoscale forms of carbon will be controlled not only by the energy state of a system, but also by the kinetics for the conditions of growth. Eb/ eV atom71 0.6 12345 0.4 0.20 C180 C240 C60 8 R /A 6 4 2 Figure 7. Binding energies of carbon atoms for nanotubes of various radii; (1)LDAcalculations; (2) and (3) calculations by the pseudopotential method using data taken from Refs 27 and 28, respectively. The dashed line corresponds to the binding energy of carbon atoms in graphite. The binding energies of carbon atoms in fullerenes C60, C180, and C240 are also shown [the data were taken from Refs 27 (4) and 28 (5)].The electron energy spectra (EES) of multilayered NT are much less studied than those of SNT. A double-layered NT (the radii of the `outer' and `inner' cylinders were 3.4 and 6.8 A, respectively) was considered using the approximate density-func- tional computational scheme.66 The calculations were carried out assuming a uniform surface electron density distribution without taking into account specific features of individual atomic geome- try of the cylinders. It was found that the total denstates (TDOS) distribution for the double-layeredNTwas similar to that of SNT with characteristic oscillations reflecting the van Hove DOS singularities for 1D systems (Fig. 8). It is believed 66 that an DOS/ eV71 1 0.02 2 5 6 3 4 0 E/ eV EF 710 730 720 Figure 8.Decomposition of the TDOS for a double-layered NT into partial s- (1), (2), p- (3), (4), and d-components (5), (6) for the inner (1), (3), (5) and outer (2), (4), (6) nanotubes.29 A L Ivanovskii increase in the number of layers will lead to the DOS of MNT approaching that of graphite. On the whole, the results of quantum-chemical simulation of ideal cylindrical SNT made it possible to obtain a detailed description of their electronic properties and establish clear correlations between the parameters of the electron energy spec- tra, topology, andNTsize. These results have become the basis for further calculations and predictions of some physicochemical properties of NT.III. Defect structure of carbon nanotubulenes. Nanotubular composites Investigation of defect NT is a certain stage of the development of the quantum chemistry of nanotubular forms of carbon. Both `intrinsic' structural defects of tubulenes and various hetero- structures involving NT (the so-called nanotubular composites) are considered.6, 8 Three main types of `intrinsic' defects of NT are distinguished, viz., topological defects, rehybridisation defects, and defects of unsaturated (or `dangling') bonds. In the first case, it is assumed that the rolled graphite sheet contains pentagons and heptagons (or their combinations). Rehybridisation defects are associated with the possibility of changing the standard sp2 hybridisation of carbon atoms (e.g., sp2>sp3).The third type of defects is due to vacancies, impurities, dislocations, and to the structure of the end groups (open or closed) of NT. 1. Structural defects and relaxation processes The effects of quasi-1D structural relaxation capable of distorting the ideal cylindrical geometry of model NT were assessed 67 by the molecular-dynamics (MD) Car ± Parrinello method 68 for a num- ber of chiral and achiral NT. The `optimised' structure of the (10, 0)- and (11, 5)-tubulenes considered 67 virtually coincides with the initial structure of the model rolled graphite sheet (the difference in the interatomic distances does not exceed*0.01 A). A possible explanation 67 is that the substance in the tubular form retains the main types of interatomic interactions characteristic of graphitic sheets.Donor (nitrogen) or acceptor (boron) substitutional impur- ities 67 cause perturbations of the electron density and can affect the conducting properties of carbon NT. Local relaxation rear- rangements upon N?C substitution are insignificant. When NT are doped with boron, the carbon atoms closest to the boron atoms are shifted along the bond axis by*0.11 A. Charge distributions in the presence of substitutional impur- ities (individual B, N, and Si atoms, anions, and cations) were investigated 57 in the framework of embedded cyclic cluster model 69 by the MNDO/PM3 method taking a (6, 6)-NT as an example. It was pointed out that discrete donor or acceptor levels appear in the band gap of the tubulene.Defect-induced perturba- tions are propagated along the NT axis to a larger number of neighbouring spheres than those propagated along the circum- ference of the cylinder. Charged defects produce larger perturba- tion areas than neutral ones. Atopological type of defect associated with the formation of a carbon vacancy was considered.70 Taking a metal-like (4, 4)-NT as an example, it was found that the number of conductivity channels (according to terminology 71, 72) decreases in the presence of carbon vacancies, which leads to suppression of conducting properties of the NT. 2. Radial deformations of nanotubulenes Charlier et al.73 discussed the effect of change in the curvature of atomic layers along the circumference of the carbon cylinder on the electronic properties of NT.The possibility for non-cylindrical carbon NT with pronounced `cut' of the walls to be formed was confirmed by three-dimensional holographic TEM images.74 ± 76 It should be noted that the nature of microdeformations of carbon NT walls (the `ripples' observed on the NT surface 77) is associated with the rehybridisation effect, i.e., with the change inSimulation of nanotubular forms of matter the effective hybridisation of the carbon atom from sp2 to sp3 (diamond). In this case, local protrusions of the surface that form `structural waves' (folds) with variable curvature will be charac- terised by averaged sp2+a carbon hybridisation, where the param- eter a depends on the curvature of the microarea of the NT surface.The edges of prismatic NT can be considered as visual lines of rehybridisation that occurs upon the change in the atomic configurations from sp2 to sp3, or as lines of hybridisation defects.77 Tight-binding DOS calculations for a cylindrical (10, 0)-tube and (10, 0)n-tubes with n-gonal cross sections showed that in the latter case there is a number of peculiarities on the DOS profile such as clearly seen effects of s* ± p* hybridisation and appreci- able changes in the energy band structure. These peculiarities are characteristic of small-diameter NT. In Fig. 9, the energy band structures of the (10, 0)- and (10, 0)5-NT are shown as an example. The p*-band energy of the (10, 0)5-NT decreases substantially (the so-called NT `metallisa- tion').The nature of this phenomenon becomes clear from comparing the charge density distributions (r) in the cross- sectional planes of cylindrical (10, 0)- and pentagonal (10, 0)5- NT for the G-states near the Fermi level. The clearly seen r anisotropy at the vertices of the pentagon, partial outward shift of the charge density (the formation of s-like states), and the reduction of the symmetry of the system lead to a substantial decrease in the band gap and to a change in the type of conductivity from semiconducting for the (10, 0)-NT to metallic for the (10, 0)5-NT. a E/ eV 2.0 1.0 FF FF 0 71.0 72.0 X G Figure 9. The energy band structures of (10, 0)- (a) and (10, 0)5-NT (c), and a tubulene with `intermediate' type of cross section (b).Changes in the energy band structure of metal-like NT upon `polygonisation' of their cross sections were considered taking (12, 0)n-NT (n=6, 4, 3) as examples.73 It was established that the conducting properties of the NT change in the following order: metallic [(12, 0)-NT]?semiconducting [(12, 0)6-NT with a band gap of about 0.5 eV]?metallic [(12, 0)4- and (12, 0)3-NT], which is mainly determined by the energies of the p ± p* bands. 3. Uniaxial compression (tension) of nanotubes Along with radial deformations in the cross-sectional plane, nanotubes can undergo longitudinal compressions (in particular, axial ones). This type of effect becomes of particular interest in connection with the investigation of mechanical properties of NT.6, 8, 78, 79 The effect of uniaxial stress on the density of states and type of conductivity of SNT was studied by the TB method taking `zigzag' c b FF X X G G 107 (n, 0) and `arm-chair' (n, m) carbon nanotubulenes as examples.80 A homogeneous uniaxial compression (s<0) and tension (s>0) was simulated by displacements of carbon atoms. Uniaxial stress will lead to the change in the NT chirality, which determines the energy characteristics of the cylindrical structures.It was found 80 that, depending on the index n and considering it as (3q71), 3q, or (3q+1), where q is an integer, the (n, 0)-NT can be arbitrarily classed into three groups. The band gaps of (3q+1)-NT will increase while those of (3q71)-NT will decrease upon compres- sion of (n, 0)-NT.The type of conductivity of the (3q, 0)-NT changes from metallic to a semiconducting one; for this reason, these NT are the most convenient objects for experimental testing of the dependence predicted.80 For all (n, 0)-NT considered,80 the band gaps depend linearly on the modulus of compression (s). According to the estimates 80 the band gap sensitivity to stress, |dEg/ds|, is about 10.7 meV GPa71 in the range |s|<10 GPa. All metal-like `arm- chair' (n, m)-NT retain their properties upon uniaxial stress. The effect of compression on the electronic characteristics and other properties (in particular, on chiral currents) of nanotubu- lenes was studied by the self-consistent electron density functional method taking a group of carbon NT as an example (Table 2).81 It was found that the major effect of compression of (9, 7)-NT consists in an appreciable increase in the band gap due to a decrease in the dispersity of p and p* bands near the Fermi level.It is possible to make the system conducting by introducing hole or electron carriers. This can be done by doping NT with atoms of alkali metals (AM) or halogens. Miyamoto 81 simulated this effect by fixed shift of EF. By solving the Boltzmann equation the components of the conductivity tensors of NT were found and the pitch angles for charge carrier currents were determined (see Table 2). It was stated81 that violation of the symmetry of the hexagonal atomic ordering of the cylinder walls upon its com- pression can cause chiral currents to flow.This effect is most pronounced for the `arm-chair' NT. Table 2. Geometric parameters of compressed (n, m)-NT and pitch angles of chiral currents (y 0) in the case of doping with electrons (e) and holes (h). Geometric parameters of (n, m)-NT y 0 /deg (see a) degree of compression (%) d /A (n, m) y /deg 10 4.1 10.78 (13, 1) 1.6(h) 8.8(e) 10 15.5 10.60 (11.4) 0.5(h) 12.4(e) 4.2(h) 24.9 10.97 (9, 7) 5 a The calculations were carried out assuming that EF is shifted by 50 meV from the bottom of the conduction band and from the top of the valence band upon doping of NT with electrons and holes, respectively. TEM images 82 ± 84 indicate that carbon NT can undergo diverse changes in their shapes and spatial configurations and that the types of NT deformations considered (radial and uniax- ial) are a particular case of such changes.Studies of arbitrarily deformed NT are scarce due to the complexity of the description of their electronic properties. 4. Deformations of carbon nanotubulenes As a rule, the problems of deformational behaviour of NT are studied in the context of the development of `nanomechanics'. Most of experimental studies in this line 6, 8, 78, 79, 85 ± 90 have been devoted to investigation of the brittle ± elastic properties of NT under different conditions of external strain. The microscopic simulation of the energy characteristics of mechanically deformed NT was performed by MD methods 88, 91108 b a E (rel.units) (dE/dj)/E00 0.6 0.8 0.4 0.3 0 0 1 j/ rad Figure 10. Dependence of the strain energy (E) and its derivative (dE/dj) on the bending angle (j) for the (13, 0)-NT (a) and the NT shape upon `kink' (b). using many-body interatomic potentials 61, 62 and energy param- eters obtained in band structure calculations of `slightly deformed' NT.60, 92, 93 Typical of this investigation line are MD calculations of the brittle ± elastic properties of NT exposed to classical types of deformation (compression, bending, and torsion).88, 91 Figure 10 illustrates the results obtained fromMDcalculations of the strain energy dependence of a (13, 0)-NT, 8 nm in length and 1 nm in diameter, on the tube bending angle.External strain was applied to both tube ends. The strain energy smoothly increased before the tube `kink' (or inelastic bending) and sharply decreased after- wards. The results of calculations 88, 91 reproduce the observed singu- larities on the strain ± stress curves of NT of different dimension- ality and are in good agreement with the experimental results.8, 85 ± 89 This makes it possible to point to the development of theoretical foundations of the nanomechanics of nanotubular objects. Quantum-chemical calculations of electron states were carried out only for certain possible statistical configurations of deformed NT, e.g., for uniform and non-uniform tubular helices (Fig. 11) 94 both predicted theoretically 95 and observed using the TEM technique.96, 97 In addition to hexagons, the walls of helices include various heptagonpentagon combinations.Helical NT are an extensive class of objects related to toroidal carbon nano- structures.8, 20, 98 ± 102 The topology of helical NT is It should be noted that the above-mentioned polygons (pen- tagons and heptagons) can be constituents of the ends of `open' nanotubes and impart to them a conical shape of variable conicity, which can become of great importance for the development of nanoscale electron-emitting devices based on NT.105 Different types of matching and the properties of carbon NT of different diameter involving a heptagon ± pentagon pair were simulated by the TB method taking (12, 0)/(9, 0)- and (12, 0)/(8, 0)- a b tubulenes as examples (see Fig.12).106 According to calculations Figure 11. Models of helically deformed uniform (C360) (a) and non- uniform (C540) (b) tubular structures.94 A L Ivanovskii described 96, 103 using the Kekule structures and the concept of `phason lines'.104 Tight-binding calculations 94 showed that the energy band structure (in particular, the change in the band gap DEg) depends strongly on the topology of the NT helices: an interrelation has been found between the arrangement of the `phason lines' in the helices and the type of their conductivity. 5. Connection of nanotubulenes Targeted introduction of topological defects into an NT, i.e., replacement of a part of hexagons in the rolled graphite sheet that forms tubulene walls by other polygons (heptagons or pentagons) can change substantially its electronic properties.In this case, the theoretically predicted possibility of the formation of hetero- junctions based on NT6, 8 opens wide prospects for using carbon NT as the elements of integrated circuits of uniquely small size. This situation can be considered from the viewpoint of longitudinal connection of two different tubulenes in which heptagons, pentagons, and their various combinations play a determining role in the formation of matched structures of NT (Fig. 12), thus opening new possibilities of producing electronic n ± p transitions. ab Figure 12. Variants of longitudinal connection of `zigzag' carbon tubu- lenes; (a) (12, 0)/(9, 0)- and (b) (12, 0)/(8, 0)-tubulene.Carbon atoms participating in the formation of heptagons and pentagons in the structure of intertubular connectors are shown as bold solid circles. performed taking into account only p-states, there are delocalised states for each of the tubulenes and localised states corresponding to open ends ofNT(`dangling' bonds) in the spectrum of the entire system. The estimates 106 of the dependence of conducting proper- ties on the voltage applied (U) showed that the current I increases as U increases for matched (12, 0)/(9, 0)-nanotubes with metallic type of conductivity and that oscillations corresponding to resonances of tunnelling probability between the two NT are observed on the current vs.voltage plot. The conductivity of the system of connected metallic (12, 0)- and semiconducting (8, 0)- nanotubes will be determined by the contribution from the (12, 0)- tubulene to the region near the Fermi level in the spectrum of the entire system. A more detailed consideration of the EES of carbon NT with variable number of topological defects (pentagon ± heptagon pairs, or 5/7-defects) of the wall surfaces of `zigzag' (n, 0)-tubu-Simulation of nanotubular forms of matter lenes was carried out.107 A system of (12, 0), (11, 0), and (9, 0)- nanotubulenes matched in pairs using one, two, and three pentagon ± heptagon pairs arranged either along the cylindrical axis or along the cylinder circumference was calculated.The data listed in Table 3 illustrate the change in the energy of NT with topological defects (relative to that of `perfect' tubulenes). As can be seen, in the general case, an axial arrangement of the 5/7-defect is the most energetically favourable. Even more energetically favourable is the formation of a specific type of defect in which a pentagon and a heptagon are separated by rings of hexagons (the 5/6/7-defect, see Table 3). Table 3. Energy changes (relative to the energies of `perfect' tubulenes) for different types of nanotube connection depending on the type of the arrangement of topological defects. NT system DE / eV atom71 Number of defects 5/7 defects along the circumference 0.132 0.258 0.405 one 5/7 two 5/7 three 5/7 (12, 0)/(11, 0) (12, 0)/(11, 0) (12, 0)/(9, 0) a 5/6/7-Defect, see text.Regularities of the rearrangement of electron states near the Fermi level in the region of hetero-junction can be traced taking a matched (12, 0)/(11, 0) metallic ± semiconducting NT with opti- mised geometry as an example (Fig. 13). On going from the last segment of the (12, 0)-NT to the first segment of the connection structure, the DOS oscillations relative to the DOS `plateau' of isolated (12, 0)-NT are observed near EF. SimilarDOSoscillations of gradually decreasing amplitude are observed when moving further along the connection structure toward the first segment of the (11, 0)-NT. The first segment of the (11, 0)-NT is gapless, which manifests itself as a smooth metal to semiconductor transition.The DOS of the hetero-junction structure approaches those of individual nanotubulenes at a distance of 20 A from the connection structure on each side of the cylinders. a DOS 18 17 16 15 14 13 12 11 10 987654321 2 E/ eV 0 72 Figure 13. The electron density of p states (a) and the cross sections (1) ± (18) of the intertubular connector structure for the (12, 0)/(11, 0)- NT (b). specific defect a axial 5/7 defects 7 70.208 7 0.249 0.291 b 18 17 16 15 14 13 12 11 10 987654321 109 6. Nanotubular composites: intercalation of carbon nanotubulenes Intercalation of NT with atoms of various elements makes it possible to obtain quasi-1D nanomaterials with varied properties. For instance, it is possible to obtain composites by incorporating metal wires into NT.These can be used as nanowires of uniquely small cross section in which the tubulene simultaneously acts as both an insulator and a protective shell (e.g., in oxidation processes). The properties of the components can change substan- tially upon interaction between the metal nanowire and NT (e.g., new phases of a metal `encapsulated' into the NT can be formed). This interaction can also cause a non-trivial modification of the properties of the entire hetero-system. For instance, its electron- phonon spectra and superconducting properties can be changed appreciably by intercalating the NT with alkali metals (AM), as is observed for fullerenes doped with AM or for intercalated graph- ite (see, e.g., Ref.8). In most of experiments on NT intercalation, multilayered tubulenes of large diameter (> 100 A) were obtained, which were filled (either directly in the synthesis or using the capillary effect) with Mg, Bi, Pb, several transition metals, lanthanides, and their carbides.108 ± 117 These systems are considered in the framework of several computational schemes. Semiempirical MNDO calculations 57 of (n, 0)-tubulenes (n=6, 8) with one or two encapsulated mona- tomic potassium wires oriented along the cylinder axis were performed using the embedded cyclic cluster model. It was found that the band gaps characteristic of initial NT disappear in the intercalated systems because of the appearance of newK4p states.As a result, the K@(n, 0) systems are 1D conductors. A quasi-1D hetero-structure consisting of a carbon nanotube (R=3.4 A) filled with a Pb nanowire (r=2.4 A) was consid- ered.66 Taking into account the approximations used in the calculations (a numerical method 118 was employed), OÈ stling et al.66 could only suggest that the electron states of the Pb nanowire will hybridise mainly with the near-the-Fermi-level p-states of the tubulene. More rigorous calculations of (4, 4)-NT intercalated with 1D potassium and aluminum wires were carried out by the ab initio pseudopotential method.119, 120 It was assumed that the metal wires and the (4, 4)-NT constitute commensurate `phases' and that the interatomic distances in the wires are the same along the whole tube length.For the K@(n, n)-NT system, the energy effect of intercalation (heat of formation) was also calculated depending on the tubulene diameter (*1 eV atom71 K71).119 The corre- sponding value for the Al@(4, 4)-NT system was found to be negative120 and it was assumed that intercalation of Al into the (4, 4)-NT can occur only in the gas phase under high pressure. Under normal conditions, only carbon nanotubes of large diameter can be intercalated with aluminum. The results obtained 119, 120 show that quantum-chemical methods can be effectively used for solving important problems related to search for compositions of nanotubular composites and conditions of their production.Taking into account promising technological applications of these materials, one can expect the development of the material science of this class of substances. IV. Electronic properties of several non-metal compounds in the nanotubular form The progress in the synthesis and the study of the properties of carbon NT has initiated investigations on the search for tubular nanostructures of other substances. Hexagonal boron nitride (h-BN), an isoelectronic and isostructural graphite analogue, was proposed 121, 122 as the first candidate. It should be noted that the syntheses of the nanotubular form of BN and some other boron compounds 32 ± 38 were performed on the basis of predic- tions made in the first theoretical studies on their structure and properties.121 ± 124110 Based on similarity in the structural and electronic properties of the known 2D boron compounds (h-BN, BC3, and BC2N125, 126) and graphite, the theoretical approaches successful in calculations on the tubular forms of carbon were used.121 ± 124 1. Boron nitride-based nanotubulenes Electronic properties of nanotubes based on h-BN were consid- ered.121, 122, 127 The dependence of the EES of chiral and achiral (n, m)-NT on their diameters was studied using a simple s,p TB model.122 It was found that all the boron nitride nanotubulenes (BNT) considered are semiconductors in which the band gap (DEg) decreases as the tube radius decreases.Chiral (n, m)-NT are less sensitive to variations of the tube radius than (n, 0)-NT.In addition, (n, 0)-NT are semiconductors with direct gaps (G± G), whereas (n, m)-NT have indirect band gaps (D ± G). Changes in DEg upon varying pitch angles y are most pronounced for the tubes of small diameter. Nanotubes with d>12 A have wide band gaps which depend slightly on y. The following qualitative explanation can be proposed: in NT of small radii, the number of nonadjacent atoms is larger, and the distances between them appear to be comparable with interlayer separations in h-BN. MNDO calculations of the electronic properties of (n, 0)- and (n, n)-BNT (n=6, 7, 10, 12) were performed.57 All NT were referred to dielectrics (DEg ^4.8 eV). The position of the bottom of the conduction band is determined mainly by boron s, p states, whereas that of the top of the valence band is determined by nitrogen states.More rigorous LDA calculations 121 of (n, 0)- and (n, n)-BNT with diameters in the range from 4.0 to 12.0 A indicate that they are semiconductors with wide band gaps (DEg ^5.5 eV). Because of this, the dependence of the electronic properties of BNT on their size and structure is less pronounced than in the case of graphite-based tubes. This fact is of importance for the prepara- tion of nanomaterials with preset parameters. When analysing the bands near the Fermi level, it was found that the states of the bottom of the conduction band are similar to those of `quasi-free' electrons.121 These states are concentrated in the inner part of the tube (Fig.14). This peculiarity may have important technological applications in the doping of BNT with electron carriers, which would make it possible to vary their electrophysical properties over a wide range. a b Figure 14. The total charge density distribution in the longitudinal section (including nitrogen atoms) (a) and in the cross section (b) of the (6, 6) BNT (LDA calculations). Interesting results were obtained in the optimisation of the geometry of BNT.121 It was shown that this tubular structure can be considered as a `double-layered' structure consisting of oppo- sitely charged coaxial cylinders that form a specific quasi-1D dipole. The boron ions constitute the `outer' cylinder, while the nitrogen ions constitute the `inner' one.The `intercylinder' dis- tance depends slightly on the tube chirality and decreases as its diameter increases. The total energy (Etot) of BNT appeared to be lower than that of carbon NT of the corresponding diameter (Fig. 15). Comparison of the total energies of a BNT and an h-BN `strip' obtained upon the `development' of the corresponding cylinder showed that the `strip' form of h-BN is much less stable than its tubular form. It was established that if the distance between the tube walls in double-layered BNT [e.g., a (4, 4)@(9, 9)-NT] is comparable with A L Ivanovskii 0.4 �1 �2 �3 0.20 10 d /A 8 the interlayer separation in crystalline h-BN, the intertubular interaction has virtually no effect on the structure of electron states of individual NT.Single-layered and multilayered BN tubulenes are stable semiconductors with wide band gaps.121 Simulation of hetero-structures based on boron nitride tubu- lenes was first performed by Rubio et al.128 who considered (4, 4)- BN-tubulenes intercalated with K and Al. It was assumed that, taking into account the peculiarities of the electronic structure of BNT,121, 122 the interaction between the metal atoms and the NT walls will be appreciably weaker than for carbon tubulenes. The estimated intercalation energies (the difference between the total energies of the composite and its components, viz., the tubulene and the metallic wire) of these hetero-structures appeared to be much lower than those of analogous systems based on carbon NT. The K 4s states of the K@(4, 4)-BNT were found to interact with the tubulene band near the Fermi level (Fig.16). For an Al@(4, 4)-BNT, it was found that additional Al s and pz levels d c b a 64 EF EF 20 72 74 X X X G G G G g Xf h e Figure 16. Energy band structures (a) ± (d) and charge distributions (e) ± (h) for the (4, 4)-BNT and for the composite K@(4, 4)-BNT and Al@(4, 4)-BNT. The energy bands: (a) K@(4, 4)-BNT; (b) and (c) (4, 4)-BNT for single and double unit cells, respectively; and (d) Al@(4, 4)-BNT. Charge distribu- tions for the states near the Fermi level are shown for longitudinal sections of the tubes. For the composite Al@(4, 4)-BNT, the contributions from Al s states are shown by solid lines.E/ eV Etot/ eV atom71 0.6 6 4 Figure 15. Dependence of the total energy (Etot) on the diameter (d) for carbon (1) and BN (2) nanotubes. The Etot value for a `strip' obtained by `developing' the (6, 0)-BN nanotube (the point 3) is shown.Simulation of nanotubular forms of matter appear in the band gap. The EES of the Al@(4, 4)-BNT appeared to be close to a simple superposition of the spectra of a monatomic Al wire and a (4, 4)-tubulene, which indicates a weak interaction between them. The plots of the charge density distributions for the near-the-Fermi level states in the K(Al)@(4, 4)-BNT (see Fig. 16) show that the density of electron carriers that control the conductivity of intercalated BNT is localised inside the cylinder.This differs these systems from composites based on carbon NT in which the contribution to the conductivity from the tubulene walls dominates. For this reason,BNTcan be preferred (as compared to carbon tubulenes) for production of the nanowires with preset properties. The possibility of intercalation of BNT with Ag and Au atoms for production of continuous nanowires 127 cannot also be ruled out. 2. Boron carbide-based nanotubulenes Electronic characteristics of some hypothetical (n, 0)-NT based on layered BC3 were considered.124 The conductivity type changes from metallic to semiconducting as the NT diameter increases. This effect is associated with noticeable deviations of the bond angles (for NT of small diameter) from 120 8.With an increase in the NT diameter, the bond angles approach the ideal value for planar sheets (120 8), and DEg for NT approaches the band gap of single BC3 sheets (see, e.g., Refs 125 and 126). The analysis of the DOS distribution for a BC3-tubulene shows that the lowest unfilled bands are composed of p- and p*-states and the highest filled bands are composed of s-states. Near the Fermi level, the density of s-states is considerably higher than that of p-states. It is assumed that the conductivity of a BC3-NT of a fixed diameter can be varied over a wide range upon doping by varying the degree of filling of the energy bands. In this case, `metallisation' of the s-band, i.e., a decrease in the concentration of electrons, will be a more efficient way of increasing the NT conductivity.Additional holes can appear upon partial replacement of carbon atoms by `superstoichiometric' boron atoms. Atoms of halogens (Cl, Br, I) can also act as hole donors, as is the case for, e.g., intercalated graphite-based compounds.124 Strain energies of BC3 nanotubes were calculated.124 The total energies of these tubulenes appeared to be lower than those of graphite-based NT. Further studies of BC3-nanotubes should be carried out taking into account not only single-, but also multi- layered nanotubes, since it is known that the interlayer interaction plays a decisive role in the formation of conducting properties of crystalline BC3.125 3. Boron carbonitride-based nanotubulenes Depending on the ordering type of atoms (B, C, N) constituting the layers and on the type of layer-by-layer packing, graphite-like boron carbonitrides can possess either metallic or semiconducting properties.125, 126 In addition, the strong `intralayer' anisotropy of charge distributions suggests that circular chiral currents can arise upon rolling the single layers into cylinders. In this case the NT can be considered as nanoscale solenoids.The geometry of several (n, n)-BC2N nanotubes was calcu- lated by the TB method.123 Nanotubulenes of small diameters (n=2) retain the conductivity type intrinsic in the initial BC2N single layers.125 However, for (n, m)-NT the change in the conductivity type from metallic to semiconducting is possible depending on their diameters and chirality (like for graphite-based NT).It is assumed that (n, n)-NT should be the most stable, since their atomic configuration provides the maximum number of the strongest C7C and B7N bonds. According to Miyamoto et al.,123 partial replacement of carbon atoms by `extra' B or N atoms leads to the appearance of donor or acceptor impurity levels, respectively, in the band energy spectrum of carbonitride NT. This occurs due to the increase in the concentration of electrons or holes in B17xC2+xN or BC2+xN17x nanotubes, respectively. Calculations of chiral cur- rents for doped BC2N-tubulenes were carried out and the magni- tude of induced magnetic field was evaluated.129 111 4.`Mixed' boron carbonitride-based nanotubulenes Hypothetical `mixed' single-layered NT built of segments of carbon and BN hexagons that alternate either along the tube circumference or along the tube axis are considered as an example of possible quasi-1D systems possessing interesting quantum properties. They were studied by the MNDO method.57 The following combinations of packing of carbon and BN hexagons were considered: alternation of C6 and B3N3 `belts' { along the circumference of a (6, 6)-NT (kB3N3+lC6; k=3, 2, 1; and l=1, 2, 3, respectively) and alternation of strips { of hexagons along the tube axis (kB3N3+lC6; k=1, 2, 3, 4, 5; and l=5, 4, 3, 2, 1, respectively). We present here only the most important conclu- sions.57 The introduction of a `belt' or a strip of carbon hexagons substantially reduces DEg of the boron nitride NT.An increase in the number of carbon fragments is accompanied by further decrease in the DEg value. The introduction of `belts&ap strips of carbon hexagons between dielectric B3N3 `belts' or strips leads to the appearance of additional states in the band gap of the (6, 6)- BN tubulene, with the major contribution from the `frontier' atoms of contacting different strips or `belts'. For a `belt' structure of mixed tubulenes, the contribution of carbon states to the bands near the Fermi level dominates, thus controlling the conducting and magnetic properties of the C6 `belts'. 5. On the possible existence of nanotubular forms of boron The results of calculations 43, 130 ± 134 of elemental boron clusters have led to unexpected and non-trivial conclusions about possible directions of structural evolution of isolated boron microparticles upon their aggregation into larger associates.Calculations 43, 130 ± 134 were carried out using the Hartree ± - Fock methods with different basis sets (STO-3G, 3-21G, and SCF/DZ) and the local electron density functional method. The energy band structure was calculated for boron clusters with optimised geometry. A large number of topologically different Bn forms (starting from B7) was designed. Three main possible types of their structural evolution were revealed, which lead to the formation of stable forms of boron clusters: (1) three-dimensional (3D) due to the growth of individual fragments of the structure of crystalline boron; (2) convex clusters (segments of a sphere or a cylinder); and (3) quasi-planar clusters.126 Nanotubes can also be formed in the bulk of a-boron quasi- crystals.132, 133 This follows from calculations of possible local structural transformations in the bulk of an a-boron quasi-crystal with a `compressed' rhombohedral unit cell for one of its frag- ments, viz., joined B12 icosahedra, lying on the minor diagonal of the rhombohedron. The structure of the B24 group changes substantially in the course of optimisation. The stages of struc- tural evolution of the B24 fragment are illustrated in Fig.17. `Diffusion' of the boron atoms located between the B12 icosahedra leads to the formation of a closed pseudotubular group (see a b c Figure 17.Structural evolution of the B24 fragment in the bulk of rhombohedral a-boron quasi-crystal in the course of geometry optimisa- tion. { According to terminology used in Ref. 57, a `belt' is a tube segment `cut' perpendicular to the tube axis, whereas a `strip' is `cut' along the tube axis.112 Fig. 17 c), which can become a `protoform' of extended nano- tubular boron structures. It was emphasised 43, 132 ± 134 that the structure and properties of proposed nanostructural forms of elemental boron (including a nanotubular form of boron) can differ fundamentally in the presence of impurities of other elements (e.g., carbon, nitrogen, silicon, etc.). However, neither detailed studies along this line nor calculations of the electronic structure of the nanotubular form of boron have been carried out so far.V. Theoretical models of the growth and morphology of nanotubes One of the fundamental challenges of theoretical investigations is to describe the nature, the kinetics of nucleation and growth, and the morphology of nanotubular structures on the microscopic level. Thermodynamic aspects of the formation of tubular and spheroidal carbon structures have been discussed.8, 11, 14 It was emphasised 135, 136 that a symmetry axis determining the direction of the structure growth is required in the reaction zone in order to form tubulenes. In the synthesis of cylindrical structures, such a direction can be specified, e.g., by a beam of accelerated carbon ions 135, 136 or catalysts.8 The most widely used method for the NT synthesis is their growth in an electric arc discharge apparatus between graphite electrodes in the presence of various catalysts.It is not possible, in the present state-of-the-art in the field of NT synthesis, to control their size and chirality exactly. Along with NT (or NT bundles whose diameters vary over a wide range), the end product contains fullerenes, other nanoparticles, and soot.8, 11, 14, 135 ± 138 The nature and mechanism of NT growth became the subject of discussion in the first studies in which the preparation of NT was reported.8 Fullerene hemispheres were considered to be precursors,18, 19, 139, 140 and the growth of tube walls was explained by successive addition of C2 dimers to the carbon atoms at the hemisphere circumference.Yet another type of mechanism of SNT growth involving cyclic structures as precursors was proposed.141 These structures were found in the studies of laser ablation of graphite.142 ± 144 The most typical structures of Cn species (104n440) are planar cyclic structures,143 ± 145 whereas spheroidal structures are formed at n>40. Two mechanisms of growth of the structures built of carbon rings were suggested,141 viz., the formation of either a fullerene or an NT. According to the model proposed by Kiang and Goddard,141 tubular structures are formed due to the pres- ence of catalysts that activate the process.138 The model of NT growth from a nucleus (a carbon ring) is illustrated in Fig.18. In the first stage, a planar ring is distorted in arc plasma in the presence of a catalyst (e.g., cobalt carbide, ComCn) to adopt a cis- or a trans-form analogously to crystalline carbyne rings.146 The cis-form of the ring can act as a precursor in the synthesis of achiral NT by addition of C2 dimers. The trans- form of the carbon ring may give rise to a larger number of cylindrical structures. The formation of the NT geometry occurs without involving the catalyst species.141 In other models of NT growth, the primary emphasis is placed on the role of the substrate and the catalyst. A phenomenological model of mechanism of SNT growth on diamond-like crystallites was proposed.147 The (111) type of the surface was considered and it was found that the prerequisite for the beginning of the process is the presence of a nanoscale surface defect, e.g., an atom of another element that `shades' one of the carbon atoms of the upper monatomic layer.A`nanoisland' that formed is surrounded along its circumference by carbon atoms that have free bonds oriented normal to the surface and interact with the ions in the vapour phase.The ions that precipitate form a ring, an NT `root' capable of further independent growth. The formation of the first layer according to the model proposed by Chernozatonskii 147 is A L Ivanovskii b a ComCn d c ComCn ComCn f e ComCn ComCn Figure 18. A model of the formation of seeds of nanotubular structures based on cyclic structures: (a) initial planar ring; (b) local deformation involving the planar ring with the formation of cis- and trans-forms; (c) nucleation of an achiral nanotube based on the cis-form; (d) nucleation of an achiral nanotube based on the trans-form; and (e), (f) nucleation of chiral nanotubes based on the trans-form.depicted schematically in Fig. 19 taking a (6, 0)-NT as an exam- ple.The above-mentioned model 147 is versatile and provides a qualitative explanation for the NT growth under any conditions of their synthesis. An important point is that the size of the nucleus (a surface defect) or that of metal particles (catalysts) should be smaller than the characteristic diameter of the SNT.Then, taking into account that each preceding carbon atom interacts with the `surface' states of the nanostructure to form its own `free' bond and becomes an active adsorption centre for the succeeding atom, the process can be interpreted simply.148 In the case of metal particles whose size is larger than the characteristic SNT diameter, a carbon (carbide) layer is formed on the surface in the first stage of precipitation, which is then followed by growth of the SNT nuclei (roots). For several metals (e.g., lanthanides) it was found 8, 149, 150 that, depending on the surface morphology, SNT can grow in the form of compact bundles radiating from the nucleus. In this case it is assumed that nanoscale inhomogeneities (protrusions) of the surface of metal grains play the role of `active nuclei' in the formation of SNT roots.ABBC �1 �2 �3 �4 �5 CA Figure 19. A scheme of the formation of the (6, 0)-NT on the diamond (111) surface. The substrate atoms (1), the tube `root' atoms (2) surround- ing the surface atom (3) `shaded' by a univalent ion (4), and the atoms of tubulene walls (5) are shown. The surface (A) and near-surface (B), (C) atomic layers are shown.Simulation of nanotubular forms of matter A microscopic model of growth of carbon NT on the surface of a metal catalyst was developed.151 The process on a planar metal surface covered with a hexagonal network of carbon atoms formed upon primary precipitation was considered. Special emphasis was placed on the energy effects upon the rearrangement of the initial layer geometry in the region of curving the elementary polygons that form the SNT base (Fig.20). It was shown that such a rearrangement results in the formation of pentagons and heptagons. The carbon atoms that precipitated successively on the metal carbide surface migrate toward the SNT base to form `belts' of hexagons constituting the SNT stem. This provides the possibility for the tube wall surrounding the protrusion on the substrate surface to grow. The model 151 makes it also possible to explain why multilayered NT do not grow. 1 E D C A B 2 Figure 20. A scheme of growth of single-layered carbon nanotube on a protrusion (1) of a surface layer (2) of a metal catalyst.(A) ± (E) are characteristic positions of carbon atoms migrating from the SNT base toward its stem (according to the model by Maiti et al.151). Starting from the first study by Iijima,9 the MNT were considered as a system of ideal coaxial cylinders formed by rolled graphitic sheets. The conditions of theMNTgrowth depending on the substrate (diamond) structure were considered.147 It was argued that the presence of the (111) crystallite terraces is a prerequisite for the MNT growth. In this case, `nucleation ring seeds' consisting of the atoms separated from one another by a distance of several AngstroÈ m can be distinguished. It has been pointed out that the MNT growth can give rise to intercylindrical interactions and narrowing of the walls of individual NT.As a result, the MNT can be `shrunk' into cones,152 thus forming a peculiar kind of `graphite caps' that close the protrusions of diamond crystallites. Mention may be made of yet another structural model of a `quasi-tubular' form of carbon. The conditions under which `hangars' (multilayer `arches' of graphene sheets `resting' on the substrate surface) are formed have been discussed.147 There are open `hangars' and those closed at the ends with truncated spheroidal structures. A growth model of conical multilayered tubular structures was proposed 153 and the possibility of the formation of a helical MNT structure was considered.154, 155 Successive stages of the growth of helical MNTare shown in Fig. 21. The NT nucleus was assumed to consist of a fullerene hemisphere formed at the substrate surface.Depending on the type of the hemisphere cross section, the `primary' tube stem can grow in different ways. Ideally, this is a right circular cylinder;153 however, if the tube base contains pentagons (a `zigzag' C60 cross section), the NT walls can be tilted with respect to the normal to the surface. The first tube type and the corresponding energy of interaction between the pair of graphitic sheets clearly defines the nucleation template (achiral or chiral) of the graphite layer of the next tube walls. The structure of the third sheet (the outer one with respect to the first and second sheets) is not unambiguously defined. This allows mutual reorientations of the system of coaxial tubes (e.g., rotations about a common axis).Effects of NT sliding over one another in MNT have been discussed.8 113 f e d c b a 3 3 3 2 2 1 1 1 Figure 21. The major stages of growth of helical NT (a) ± (f).153 (1) ± (3) graphitic layers in the cross sections of corresponding structures; the edges of dislocations are shown. Arrows indicate the direction ofMNT growth. The increase in the number of graphitic sheets (see Fig. 21) can lead to the formation of lateral dislocations that are considered as energy stabiliser of the system.153 As a result, theMNTstructure is transformed into a helical one and theMNTgrows by building up a helix rather than by nucleating the next graphitic sheet. It is of importance that a pair of lateral dislocations of opposite signs should be present in each helical nanotube.The presence of dislocations will lead to some increase in the tube volume. The MNT growth is terminated when dislocations become oriented parallel to the tube axis.153 The authors of the above-mentioned models of MNT growth used qualitative ideas of a specific type of interatomic interactions in the systems in question. Kwon et al.156 studied growth con- ditions, morphology, and stability of achiral double-layered (5, 5)@(10, 10)- and (9, 0)@(18, 0)-tubulenes consisting of coaxial cylinders with respectively `arm-chair' and `zigzag' edges by performing rigorous ab initio calculations in the framework of the electron density functional formalism (a cluster model with geometry optimisation according to the DMol version). During the NT growth process, adatoms can either build up the open end of either tube or connect these ends to become bridging atoms (see respectively configurations 1 and 2, 3 in Fig.22). For each of the configurations considered, ab initio calculations of energy parameters and electronic structure were carried out. The tube growth was analysed with an increment of 10 extra atoms arranged as described above. Relaxation effects were taken into account by the MD method. Let us illustrate the results of calculations156 taking a double- layered (5, 5)@(10, 10)-MNT as an example. Placing extra carbon atoms exactly on the tube edges leads to the formation of a system of `dangling' bonds.However, in this case the cohesive energies of MNT remain virtually unchanged. For bridging atoms, the a 1 2 3 b Figure 22. Cross (a) and longitudinal (b) section of possible growth structures (1) ± (3) of a two-layer (5, 5)@(10, 10)-NT.156114 situation is markedly different. For a fully optimised structure, the energy gain per each extra atom is rather large (DE ^0.42 eV), which confirms the assumption 158 of stabilising contribution of `lip-lip' interactions between the walls of neighbouring tubes to the total energy of the system. The stabilising effect of bridging atoms depends on their distance from the tube interior and on the type of the polygons formed by these bridging atoms and the atoms of tube walls in the NT cross section.Thus, the config- uration 3 in Fig. 22 is unstable (DE ^73.52 eV). Analogous conclusions can be drawn from the analysis of the (9, 0)@(18, 0)- MNT. The optimum NT morphology is determined by a combi- nation of the above-mentioned factors. Ab initio studies of MNT growth and termination depending on the number of extra atoms have also been carried out.156 The main parameter used in the case of a growing cylindrical MNT was the degree of the edge coverage e a N , Na where N is the total number of carbon atoms on the tube edge and Na is the number of added (extra) atoms. Then the energy parameter was calculated as a function of e Ec a EtotOCNU ¢§ NEtotOC?U , Na where Etot(CN) is the total energy of the truncated NT of a given length and Etot(C?) is the total energy of an infinite tube.{ It was assumed that NT growth is favoured to the greatest extent at a minimum Ec value that can be achieved at a certain concentration and for particular configurations of adatoms on the tube edge when the `growing' tube has a minimum energy as compared to the `end reaction product', a perfect infinite MNT.The results of calculations for a (5, 5)@(10, 10)-NT are shown in Fig. 23. Stabilisation of the structure is at e = 2/3. The `lip-lip' interactions play an important role in the structure stabilisation. Stable structures of MNT growth contain carbon hexagons and pentagons, but no squares es. In the range 0<e42, the initial tube growth is simulated.The interval 2<e44.5 corre- sponds to an `intense edge coverage' of the tubulene. Changes in the relative energies of dome closure, which terminates the process of nanotube growth, are also shown. The simplest types of the domes for closing the inner and outer tubes of the (5, 5)@(10, 10)- NT are hemispheres of the C60 and C240 fullerenes, respectively. To form closed noninteracting domes, the degree of the edge coverage should lie in the range 8/3<e<14/3. E/ eV 1 1.0 2 3 0.50 4.0 3.5 2.5 3.0 2.0 1.5 1.0 0.5 e Figure 23. Dependence of the energy (E) of edge groups of the (5, 5)@(10, 10)-NT on the number of adatoms e (see text). The interval 0<e42 corresponds to the MNT growth; curves (1) ¡¾ (3) in the region 2<e44.6 describe how E changes for continuous NT growth (1) and for the dome closure of the nanotube without (2) and with (3) account of `lip-lip' interactions.156 The mechanism of double-layered NT growth and conditions of its termination 156 can be extended toMNTwith any number of layers.However, problems can arise in the case of odd-walled { An idealised model of a quasi-1D structure is an infinite tube. A L Ivanovskii MNT. It is much more difficult to determine optimum conditions of the saturation of `dangling bonds' in this case. Probably, MNT with an odd number of layers are less stable. This is indirectly confirmed by larger abundance of MNT with an even number of layers.158 At the same time, the proposed growth mechanism can be corrected to a certain extent for the presence of `passivating' elements as impurities (`covalent spot welds') capable of saturat- ing one or several free (`dangling') bonds of the carbon atoms of outer cylinders.VI. Nanotube crystals Nanotube crystals (NTC) are regular 3D structures formed by individual NT. To date, theoretical models of crystalline nanotubular forms of matter and predictions of their characteristics have been reported only for certain possible crystalline forms of nanotubular carbon. It was suggested that NTC be considered as a new phase of crystalline carbon with quasi-one-dimensional, strongly aniso- tropic properties. Progress in the synthesis of single-layered NT and, in partic- ular, the possibility of producing NT of most plausible diame- ters 6, 8 under specified conditions of plasmochemical process provides an experimental basis for performing theoretical studies.Production of compact nanotubular associates, e.g., NT bundles and ropes, as well as nanotube films with regular arrangement of individual cylinders has been reported.6, 8, 159 ¡¾ 169 Let us consider the main lines of theoretical investigations of nanotube crystals. 1. Intertubular interactions in nanotube crystals In the studies of NT packed into regular 3D structures, the problem arises of establishing the form of their packing and the nature of interaction between individual NT, and of determining the energy effect of this process. It is assumed 166, 167 that the intertubular bonding is due to the van der Waals interactions (as is the case, e.g., for fullerites 8).Optimum ways of packing individual NT into a periodic 3D structure were studied.168 Nanotube crystals formed by (6, 6)- nanotubulenes with parallel cylindrical axes were calculated. Depending on the type of mutual arrangement of individual NT in the nanotube crystal, a tetragonal and a hexagonal unit cell was investigated. In the latter case, calculations were performed for two types of packing of single-layered NT, which correspond to the hypothetical graphite AAA... stacking and to the conventional graphite ABAB... stacking. It was found that the system in which carbon atoms of neighbouring tubes are packed similarly to graphite ABAB... stacks is the most stable (Fig.24). In addition, the `intertubular' Ecoh/ eV cell71 0 70.1 123 70.2 70.3 3.2 l /A 3.0 3.4 Figure 24. Dependence of the cohesive energy of an NTC formed by (6, 0)-NT on the distance (l) between individual NT for tetragonal (1) and hexagonal (2), (3) unit cells. Type of the NT packing in hexagonal unit cells: (AAA...) (2), and (ABAB...) (3).Simulation of nanotubular forms of matter cohesive energy Ecoh inNTC(9.76 meV atom71) is lower than the interlayer energy in graphite (20 meV atom71) 170 and compara- ble with the energies of intercluster interactions in condensed fullerenes (*10 meV atom71). It can be assumed that the con- ditions of the formation of 3D NTC from tubes of different diameters control the degree of close packing of individual NT and the possibility of the AB type of packing of the carbon atoms in neighbouring tubes.Any violation of the optimum combination of these factors will lead to destruction of crystal structure and to orientational disorder in the NT arrangement. The energy band structure of a hexagonal carbon NTC is characteristic of a mixed 1D± 3D electronic system.168 The main dispersive effects are due to `intratubular' interactions; at the same time, intertubular interactions also make a noticeable contribu- tion. The NTC analysed possess a metallic type of conductivity, which means that packing of individual (6, 6)-NT into a crystal does not change the nature of their conductivity. 2. Radial deformations of nanotubulenes in crystals Charlier et al.168 considered an NTC in which the shape of NT constituting the crystal was assumed to be preset and unchanged upon packing into the 3D structure.The problem of possible deformations of SNT in the interaction was posed.8, 166, 167 The original concepts of the effect of the interaction between two NT on their cross section geometry were developed by Ruoff et al.166 who calculated structural deformations for a pair of nanotubes using a many-body interaction model. It was estab- lished that, first, the walls of neighbouring NT become `flattened' so that the respective regions of NT reproduce the geometry of planar graphite sheets (which corresponds to a maximum inter- action energy) and, second, the intertubular volume is decreased (Fig.25). It was also pointed out 166 that the degree of deforma- tion depends strongly on the NT diameter and increases as the diameter increases. a y /A 300 730 b r (rel. units) 100 40 x /A 0 740 Figure 25. Radial deformations of a pair of contacting NT upon their interaction (a) and the charge density r distribution (b). The above-mentioned peculiarities of deformation of two individual cylinders upon their packing into a 3D structure are clearly illustrated in Fig. 26. The principles of calculations of structural distortions for a pair of contacting cylinders166 were extended to 3D NTC.167 In Fig. 26, one can clearly see the formation of a `honeycomb' structure of the NTC cross section with gradual transformation of the cross sections of cylindrical NT into those of hexagonal prisms as the tube diameter increases.In the framework of the model used,167 the `flattening' effects of NT walls are detected starting from d ^15 A. These structural deformations will play an important role in the formation of cohesive and elastic properties of NTC. In the general case, one can expect that the increase in the NT diameter 115 c b a Figure 26. NT deformations in the NTC as a function of tubulene diameter; (a) d=10 A; (b) d=40 A; and (c) hexagonal distortions of a nanotube cross section in the crystal with an increase in the NT diameter (for anNTwith d=10, 15, 20, 30, and 40 A). The points correspond to the carbon atoms in the NT cross section. will be accompanied by weakening of interparticle interactions (in the XY plane of the NTC) and reduction of its 2D cohesive characteristics.According to calculations,167 the dependence of the cohesive energy and the modulus of elasticity on the NT diameter is noticeably nonmonotonic (Fig. 27). The formation of a `honeycomb' structure imparts an anomalous hardness to the NTC that is maintained upon further increase in the diameter (starting from d ^25 A) of NT constituting the crystal. It is suggested that NTC be divided arbitrarily into two groups with `small' (or `large') diameters of NT constituting the crystal, possessing respectively a nonmonotonic and a monotonic depend- ence of the cohesive energy on theNTdiameter.Presumably, these groups of NTC will have substantially different physicochemical properties and will respond differently to the effects of doping (e.g., upon introduction of alkali metal ions into intertubular channels) upon the formation of composites, etc. b M/ eV A73 a 7Ecoh /meV atom71 15 0.2 20 0.10 25 60 40 20 40 20 d /A Figure 27. Dependence of the cohesive energy (Ecoh) (a) and the modulus of elasticity (M) (b) for a nanotube crystal on the diameter (d) of nanotubes constituting theNTC. Filled circles correspond to analytical extrapolation of the results of calculations of NTC consisting of nanotubes of small diameter. 3. Hetero-junctions in nanotube crystals In the formation of 3D structures of regularly packed NT it may happen that the geometric and electronic parameters of a pair of neighbouring NT are different or an NTC contains the domains consisting of different NT.It is obvious that in this case one can suggest the existence of `interphase domains' (or `interface domains') between nanotube `polycrystals'. Chico et al.70 consid- ered hetero-junctions in pairs of (12, 0)/(6, 6)- and (9, 0)/(6, 3)-NT and possible differences in electronic properties between the `interphase domain' and the NT constituents using Green's functions formalism. This study can be considered as the first attempt to describe the above-mentioned systems by semiempir- ical quantum-chemical methods. The development of the ideas of condensed regular nano- tubular forms of matter as 3D periodic structures, viz., nanotube crystals, makes it possible to consider these solids taking into account all factors determining the `composition ± structure ± properties' interrelations inherent in all `conventional' solids.Assuming that isolated nanotubes whose properties can be varied over a wide range can be considered as `structural elements'116 of NTC, one can accept the idea of a defect (impurity, vacancy) structure of NTC and consider the order ± disorder problem therein. Such an approach makes it possible to consider NTC in the context of general problems of the description of specific properties of crystals (including electronic ones) and to pose the problems of surface states of NTC, interfaces between nanotube polycrystals, etc.The above-mentioned lines of the development of the ideas concerning fundamental properties of NTC are still in a primitive state even for carbon NTC. One can expect that they will be useful for the development of the problems of quantum materials science and their application to NTC forms of other substances including those considered in this review. VII. Conclusion The discovery of the nanotubular form of matter was one of the major achievements of physical chemistry in the last decade. Unique properties of nanotubulenes offer extremely attractive prospects of their use in technology and are of considerable interest for fundamental research. Only the principal results of the studies of electronic structure of NT by methods of solid state quantum chemistry have been covered in this review.The most considerable progress has been achieved in the studies of single-layered isolated nanotubes and prediction of their properties. Recent studies on the synthesis of NT and nanotubular materials 171 ± 174 are intimately associated with the development of theoretical models. One of the most important theoretical problems is to describe the `life cycle' of tubulenes (nucleation ± - growth ± disintegration). Currently, considerable interest is dem- onstrated in `intermediate' nanoscale forms (barrellenes,175 ± 177 onions,178 calabashes,179 and other nanoscale structures and mesostructures 180, 181) which can act as intermediates in the formation or disintegration of NT.182 Topical are investigations of various nanotubular structures (tori,183 ± 185 cones,186 nanotubes with negative Gaussian curva- ture,187 quasi-2D NT associates, and nanotubular layers and films 188 ± 191), nanotubular superstructures,191 ± 194 and `mixed' nanoscale forms and composites (e.g., NT-fullerene,195 amor- phous nanotubular carbon 196) which can possess unusual proper- ties.The theory of adsorptive, capillary, electromagnetic, catalytic, mechanical, and other properties of NT that are of interest for researchers in the field of materials science is in the early stage.The problem of whether the nanotubular form of matter is a universal form for the whole known diversity of inorganic substances or are there certain physical or chemical criteria clearly reducing the number of potential objects whose nanotubular forms can be obtained under certain (which?) conditions, is a basic problem which remains open so far.The solution of the above-mentioned problems is one of the priorities in the quantum chemistry of nanotubulenes. It provides a basis for the development of the tubulene materials science, thus favouring their effective use in nanoscale electronics, in instru- ments and chemical engineering, and in other areas. References 1. H W Kroto, A W Allaf, S P Balm Chem. Rev. 91 1213 (1991) 2. V I Sokolov, I V Stankevich Usp. Khim. 62 455 (1993) [Russ. Chem. Rev. 62 419 (1993)] 3. R A Andrievskii Usp. Khim. 63 431 (1994) [Russ. Chem. Rev.63 411 (1994)] 4. T W Ebbesen Ann. Rev. Mater. Sci. 24 235 (1994) 5. Y Tenne Adv. Mater. 7 965 (1995) 6. T W Ebbesen Phys. Today 273 26 (1996) 7. V M Smirnov Khimiya Nanostruktur. Sintez, Stroenie, Svoistva (The Chemistry of Nanostructures. Synthesis, Structure, Properties) (St.-Petersburg: St.-Petersburg State University, 1996) A L Ivanovskii 8. M S Dresselhaus, G Dresselhaus, P C Eklund Science of Fullerenes and Carbon Nanotubes (New York: Academic Press, 1996) 9. S Iijima Nature (London) 354 56 (1991) 10. T W Ebbesen, P M Ajayan Nature (London) 358 220 (1992) 11. S Iijima, T Ichihashi Nature (London) 363 603 (1993) 12. S Iijima, P M Ajayan, T Ichihashi Phys. Rev. B, Condens. Matter 47 5485 (1993) 13. S Iijima, P M Ajayan, T Ichihashi Phys. Rev.Lett. 69 3100 (1992) 14. D S Bethune, C H Kiang,M S de Vries, G Gorman, R Savoy, J Vazquez, R Beyers Nature (London) 363 605 (1993) 15. P M Ajayan, J M Lambert, P Bernier, L Barbedette, C Colliex, J M Planeix Chem. Phys. Lett. 215 509 (1993) 16. P Calvert Nature (London) 357 365 (1992) 17. J W Mintmire, B I Dunlap, C T White Phys. Rev. Lett. 68 631 (1992) 18. R Saito,M Fujita, G Dresselhaus, M S Dresselhaus Phys. Rev. B, Condens. Matter 46 1804 (1992) 19. R Saito,M Fujita, G Dresselhaus, M S Dresselhaus Appl. Phys. Lett. 60 2204 (1992) 20. B I Dunlap Phys. Rev. B, Condens. Matter 46 1933 (1992) 21. N Hamada, S Sawada, A Oshiyama Phys. Rev. Lett. 68 1579 (1992) 22. M S Dresselhaus, G Dresselhaus, R Saito Solid State Commun.84 201 (1992) 23. M S Dresselhaus, G Dresselhaus, R Saito Phys. Rev. B, Condens. Matter 45 6234 (1992) 24. A Oshiyama, S Saito, N Hamada, Y Miyamoto J. Phys. Chem. Solids 53 1457 (1992) 25. M S Dresselhaus Phys. World 9 18 (1996) 26. M Menon, D Srivastava Phys. Rev. Lett. 79 4453 (1997) 27. R Saito Science 278 77 (1997) 28. P G Collins, A Zettl, H Bando, A Thess, R E Smalley Science 278 100 (1997) 29. A Lorezoni, H E Roman, F Alasia, R A Broglia Chem. Phys. Lett. 276 237 (1997) 30. E W Wong, P E Sheehan, C M Lieber Science 277 1971 (1997) 31. M S Dresselhaus Nature (London) 391 19 (1998) 32. P Gleize, M C Schouler, P Gadelle, M Caillet J. Mater. Sci. 29 1575 (1994) 33. P Gleize, S Herreyre, P Gadelle, H Mermoux,M C Cheynet, L Abello J.Mater. Sci. Lett. 13 1413 (1994) 34. N G Chopra, R J Luyken, K Cherrey, V H Crespi,M L Cohen, S G Louie, A Zettl Science 269 966 (1995) 35. A Loiseau, F Willaime, N Demoncy, G Hug, H Pracard Phys. Rev. Lett. 76 4737 (1996) 36. A Zettl Adv. Mater. 8 443 (1996) 37. Z Weng-Sieh, K Cherrey, N G Chopra, X Blase, Y Miyamoto, A Rubio,M L Cohen, S G Louie, A Zettl, R Gronsky Phys. Rev. B, Condens. Matter 51 1129 (1995) 38. O Stephan, P M Ajayan, C Colliex, P Redlich, J Lambert, P Bernier, P Lefin Science 266 1683 (1994) 39. Y Feldman, E Wasserman, D J Srolovitz, R Tenne Science 267 222 (1995) 40. M Remskar, Z Skraba, F Cleton, R Sanjines, F Levy Appl. Phys. Lett. 69 351 (1996) 41. H Nakamura, Y Matsui Adv. Mater. 7 871 (1995) 42.H L Frisch, J E Mark Chem. Mater. 8 1735 (1996) 43. I Boustani, A Quandt Europhys. Lett. 39 527 (1997) 44. W K Hsu, J P Hare, M Terrones, H W Kroto, D Walton, P Harris Nature (London) 377 687 (1995) 45. M Freemantle Chem. Eng. News 73 9 (1995) 46. J W Wildler, L C Venema, A Rinzler, R E Smalley, C Dekker Nature (London) 391 59 (1998) 47. T W Odom, J-L Huang, P Kim, C M Lieber Nature (London) 391 62 (1998) 48. J W Mintmire, D H Robertson, C T White J. Phys. Chem. Solids 54 1835 (1993) 49. H Yorikawa, S Muramatsu Phys. Rev. B, Condens. Matter 50 12203 (1994) 50. J W Mintmire, C T White Synth. Met. 77 231 (1996) 51. H Yorikawa, S Muramatsu Solid State Commun. 94 435 (1996) 52. R Saito, G Dresselhaus, M S Dresselhaus Phys. Rev. B, Condens.Matter 50 14698 (1994) 53. H Ajiki, T Ando J. Phys. Soc. Jpn. 62 1255 (1993) 54. H Yorikawa, S Muramatsu Phys. Rev. B, Condens. Matter 52 2723 (1995)Simulation of nanotubular forms of matter 55. H Yorikawa, S Muramatsu Solid State Commun. 97 115 (1996) 56. I V Zaporotskova, A O Litinskii Vest. Volgogradsk. Gos. Univ. (1) 145 (1997) 57. I V Zaporotskova, Candidate Thesis in Physicomathematical Sci- ences, Volgograd State University, Volgograd, 1997 58. M S Dresselhaus, G Dresselhaus Adv. Phys. 30 139 (1981) 59. D Tomanek, M A Schluter Phys. Rev. Lett. 67 2331 (1991) 60. D H Robertson, D W Brenner, J W Mintmire Phys. Rev. B, Condens. Matter 45 12592 (1992) 61. J Tersoff Phys. Rev. B, Condens. Matter 37 6991 (1988) 62. D W Brenner Phys.Rev. B, Condens. Matter 42 9458 (1990) 63. J W Mintmire, C T White Phys. Rev. B, Condens. Matter 28 3283 (1983) 64. J W Mintmire Phys. Rev. Lett. 50 101 (1983) 65. C N R Rao, R Seshadri, A Govindaraj, R Sen Mater. Sci. Eng. R15 209 (1995) 66. D OÈ stling, D Tomanek, A Rosen Phys. Rev. B, Condens. Matter 55 13980 (1997) 67. J Yi, J Bernholc Phys. Rev. B, Condens. Matter 47 1708 (1993) 68. R Car,M Parrinello Phys. Rev. Lett. 55 2471 (1985) 69. A O Litinskii, N G Lebedev, I V Zaporotskova Zh. Fiz. Khim. 69 189 (1995) a 70. L Chico, L X Benedict, S G Louie,M L Cohen Phys. Rev. B, Solid State 54 2600 (1996) 71. R Landauer Philos. Mag. 21 863 (1970) 72. D S Fisher, P A Lee Phys. Rev. B, Condens. Matter 23 6851 (1981) 73. J C Charlier, P Lambin, T W Ebbesen Phys.Rev. B, Condens. Matter 54 R8377 (1996) 74. S Seraphin, D Zhou, J Jiao Acta Microscop. 3 45 (1994) 75. M Liu, J M Cowley Ultramicroscopy 53 333 (1994) 76. M Liu, J M Cowley Carbon 32 393 (1994) 77. H Hiura, T W Ebbesen, J Fujita, K Tanigaki, T Takada Nature (London) 367 148 (1994) 78. M S Dresselhaus Nature (London) 358 195 (1992) 79. M S Dresselhaus, G Dresselhaus, R Saito Carbon 33 883 (1995) 80. R Heyd, A Charlier, E McRae Phys. Rev. B, Condens. Matter 55 6820 (1997) 81. Y Miyamoto Phys. Rev. B, Condens. Matter 54 R11149 (1996) 82. J Despers, E Daguerre, K Ladfi Carbon 33 87 (1995) 83. N Chopra, L Benedict, V Crespi, M Cohen, S Louie, A Zettl Nature (London) 377 135 (1995) 84. R S Ruoff, D C Lorents Bull.Am. Phys. Soc. 40 173 (1995) 85. C Lieber Bull. Am. Phys. Soc. 42 591 (1997) 86. S Iijima, C Brabec, A Maiti, J Bernholc J. Chem. Phys. 104 2089 (1996) 87. T W Ebbesen, in Carbon Nanotubes. Preparation and Properties (Ed. TW Ebbesen) (Boca Raton, FL: CRC Press, 1997) p. 225 88. B I Yakobson, M B Campbell, C J Brabec, J Bernholc Comput. Mater. Sci. 8 341 (1997) 89. J Lu Phys. Rev. Lett. 79 1297 (1997) 90. M R Falko, C H Clary, R M Taylor, V Chi, F P Brooks, S Washburn, R Superfine Nature (London) 389 582 (1997) 91. B I Yakobson, C J Brabec, J Bernholc Phys. Rev. Lett. 76 2511 (1996) 92. G Adams Science 256 1792 (1992) 93. A Lucas, P Lambin, R Smalley J. Phys. Chem. Solids 54 587 (1993) 94. K Akagi, R Tamura, M Tsukada, S Itoh, S Ihara Phys.Rev. B, Condens. Matter 53 2116 (1996) 95. S Ihara, S Itoh, J Kitakami Phys. Rev. B, Condens. Matter 48 5643 (1993) 96. X B Zhang, X F Zhang, D Bernaerts, G van Tendeloo, S Amelinckx, J van Landuyt, V Ivanov, J B Nagy, Ph Lambin, A A Lucas Europhys. Lett. 27 141 (1994) 97. V Ivanov, J B Nagy, Ph Lambin, A Lucas, X B Zhang, X F Zhang, D Bernaerts, G van Tendeloo, S Amelinckx, J van Landuyt Chem. Phys. Lett. 223 329 (1994) 98. S Ichara, S Itoh, J Kitakami Phys. Rev. B, Condens. Matter 47 12908 (1993) 99. B I Dunlap Phys. Rev. B, Condens. Matter 49 5643 (1994) 100. B I Dunlap Phys. Rev. B, Condens. Matter 50 8134 (1994) 101. S Itoh, S Ichara Phys. Rev. B, Condens. Matter 48 8323 (1993) 102. S Itoh, S Ichara Phys. Rev.B, Condens. Matter 47 1703 (1993) 103. K Akagi, R Tamura, M Tsukada, S Ihara, S Itoh Phys. Rev. Lett. 74 2307 (1995) 117 104. MFujita Fullerene Sci. Technol. 1 365 (1993) 105. Y Saito, K Hamaguchi, K Hata, K Uchida, Y Tasaka, F Ikazaki, M Yumura, A Kasuya, Y Nishina Nature (London) 389 27 (1997) 106. R Saito, G Dresselhaus, M S Dresselhaus Phys. Rev. B, Condens. Matter 53 2044 (1996) 107. J C Charlier, T W Ebbesen, P Lambin Phys. Rev. B, Condens. Matter 53 11108 (1996) 108. R A Jiski, M S Dresselhaus Phys. Rev. B, Condens. Matter 45 12465 (1992) 109. P M Ajayan, S Iijima Nature (London) 361 333 (1993) 110. E Dujardin, T W Ebbesen, H Hiura, K Tanigaki Science 265 1850 (1994) 111. C Guerret-Plecourt, Y Le Bouar, A Loiseau, H Prasacard Nature (London) 372 761 (1994) 112.P M Ajayan, C Colliex, J M Lambert, P Bernier, L Barbedette, M Tence, O Stephan Phys. Rev. Lett. 72 1722 (1994) 113. K Tanaka, T Sato, T Yamabe, K Okahara, K Uchida, M Yumura, H Niino, S Ohshima, Y Kuriki, K Yase, F Ikazaki Chem. Phys. Lett. 223 65 (1994) 114. P M Ajayan, T W Ebbesen, T Ichihashi, S Iijima, K Tanigaki, H Hiura Nature (London) 362 522 (1993) 115. S Seraphin, D Zhou, J Jiao, J C Withers, R Loutfy Nature (London) 362 503 (1993) 116. S Subramoney, R S Ruoff, D C Lorents, B Chan, R Malhotra, M J Dyer, K Parvin Carbon 32 507 (1994) 117. S C Tsang, Y K Chen, P Harris, M Green Nature (London) 372 159 (1994) 118. S Salomonson, P Oster Phys. Rev. A, Gen. Phys., Ser. 3 40 5559 (1989) 119.Y Miyamoto, A Rubio, X Blase,M L Cohen, S G Louie Phys. Rev. Lett. 74 2992 (1995) 120. J A Wu,M L Wang,R Lu,W H Duan Int. J. Mod. Phys. B12 1601 (1998) 121. X Blase, L X Benedict, E L Shirley, S G Louie Phys. Rev. Lett. 72 1878 (1994) 122. A Rubio, J L Corkill,M L Cohen Phys. Rev. B, Condens. Matter. 49 5081 (1994) 123. Y Miyamoto, A Rubio,M L Cohen, S G Louie Phys. Rev. B, Condens. Matter 50 4976 (1994) 124. Y Miyamoto, A Rubio, S G Louie,M L Cohen Phys. Rev. B, Condens. Matter 50 18360 (1994) 125. A L Ivanovskii Usp. Khim. 66 511 (1997) [Russ. Chem. Rev. 66 459 (1997)] 126. 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', 1997) 127. H Zhu, D J Klein,W A Seitz, N H March Inorg. Chem. 34 1377 (1995) 128. A Rubio, Y Miyamoto, X Blase,M L Cohen, S G Louie Phys. Rev. B, Condens. Matter 53 4023 (1996) 129. Y Miyamoto, S G Louie, M L Cohen, Phys. Rev. Lett. 76 2121 (1996) 130. I Boustani, Int. J. Quantum. Chem. 52 1081 (1994) 131. I Boustani Chem. Phys. Lett. 240 135 (1995) 132. I Boustani, A Quandt, P Kramer Europhys. Lett. 36 583 (1996) 133. I Boustani, in Proceedings of the 12th International Symposium on Boron, Borides, Related Compounds (Abstracts of Reports), Vienna, 1996 p. 14 134. I Boustani Phys. Rev. B, Condens. Matter 55 101 (1997) 135. C-H Kiang,W A Goddard III, R Beyers, J R Salem, D S Bethune J.Chem. Phys. 98 6612 (1994) 136. S Seraphin, D Zhou Appl. Phys. Lett. 64 2087 (1994) 137. Y Saito, MOkuda, N Fujimoto, T Yoshikawa,M Tomita, T Hayashi Jpn. J. Appl. Phys., Part 2 33 L526 (1994) 138. C-H Kiang, W A Goddord, R Beyers, J R Salem, D S Bethune J. Phys. Chem. Solids 57 35 (1996) 139. R Saito,M Fujita, G Dresselhaus, M S Dresselhaus Mater. Sci. Eng. 19B 185 (1993) 140. M Endo, H W Kroto J. Phys. Chem. 96 6941 (1992) 141. C-H Kiang, W A Goddard Phys. Rev. Lett. 76 2515 (1996) 142. G von Helden, N G Gotts, M T Bowers Nature (London) 363 60 (1993) 143. J Hunter, J Fye,M F Jarrold Science 260 784 (1993)A L Ivanovskii 118 145. G von Helden, P R Kemper, N G Gotts,M T Bowers Science 259 144. J Hunter, J Fye, E J Roskamp, M F Jarrold J.Phys. Chem. 98 1810 (1994) 1300 (1993) 146. R B Heinman, J Kleiman, N M Salansky Nature (London) 306 164 (1983) 147. L A Chernozatonskii Khim. Fiz. 16 78 (1997) b 148. C J Brabec, A Maiti, C Roland, J Bernholc Chem. Phys. Lett. 236 150 (1995) 149. D Zhou, S Seraphin, S Wang Appl. Phys. Lett. 65 1593 (1994) 150. S Subramoney, R S Ruoff, D C Lorents, R Malhotra Nature 192. E G Gal'pern, I V Stankevich Mol. Mater C8 169 (1996) 193. L A Chernozatonskii Mol. Mater C7 41 (1996) 194. E G Gal'pern, I V Stankevich, A L Chistyakov, L A Chernozatonskii Fullerene Sci. Technol. 6 499 (1998) 195. T Kukumaki, T Hayashi, K Miyazawa, H Ichinose, K Ito, Y Ishida Mater. Trans. JIM 39 574 ( 1998) 196. G van Tendeloo, D Bernaerts, S Amelinckx Carbon 36 487 (1998) 197. A L Ivanovskii Kvantovaya Khimiya v Materialovedenii. Nanotubu- lyarnye Formy Veshchestva (Quantum Chemistry in Materials Technology. Nanotubular Forms of Substance) (Ekaterinburg: Ural Brunch Russian Academy of Sciences, 1999) a�Russ. J. Phys. Chem. (Engl. Transl.) b�J. Chem. Phys. (Engl. Transl.) c�Physics-Uspekhi (Engl. Transl.) (London) 366 637 (1993) 151. A Maiti, C J Brabec, J Bernholc Phys. Rev. B, Condens. Matter 55 R6097 (1997) 152. M Ge, K Sattler Chem. Phys. Lett. 220 192 (1994) 153. S Amelinckx, D Bernaerts, X B Zhang, G van Tendeloo, J van Landuyt Science 267 1334 (1995) 154. X F Zhang, X B Zhang, G van Tendeloo, S Amelinckx, M Op de Beek, J van Landuyt J. Cryst. Growth 130 368 (1993) 155. X B Zhang, X F Zhang, S Amelinckx, G van Tendeloo, J van Landuyt Ultramicroscopy 54 237 (1994) 156. Y K Kwon, Y H Lee, S G Kim, P Jund, D Tomanek, R E Smalley Phys. Rev. Lett. 79 2065 (1997) 157. B Delley J. Chem. Phys. 92 508 (1990) 158. D Tomanek NATO ASI, Ser. B 313 405 (1996) 159. W A de Heer, M S Basca, A Chatelain, T Gerfin, R Humphrey- Baker, L Forro, D Ugarte Science 268 845 (1995) 160. O Chauvet, L Forro,W Bacsa, D Ugarte, B Doudin, W A Heer Phys. Rev. B, Condens. Matter 52 R6963 (1995) 161. J C Charlier, J-P Issi J. Phys. Chem. Solids 57 957 (1996) 162. J E Fisher, H Dai, A Thess, R Lee, N M Hanjani, D L Dehaas, R Smalley Phys. Rev. B, Condens. Matter 55 R4921 (1997) 163. H Dai, E Wong, C Lieber Science 272 523 (1996) 164. L Langer, V Bayot, E Grivel, J-P Issi, J P Heremanas, C H Olk, 165. G Baumgartner,M Carrard, L Zuppiroli,M S Basca Phys. Rev. B, L Stocman, C van Haesendonck, Y Bruynseraede Phys. Rev. Lett. 76 479 (1996) Condens. Matter 55 6704 (1997) 166. R S Ruoff, J Tersoff,D C Lorents, S Subramoney, B Chan Nature (London) 364 514 (1993) 167. J Tersoff, R S Ruoff Phys. Rev. Lett. 73 676 (1994) 168. J C Charlier, X Gonze, J P Michenaud Europhys. Lett. 29 43 (1995) 169. J C Charlier, X Gonze, J P Michenaud Carbon 32 289 (1994) 170. P M Ajayan, S Iijima Nature (London) 358 23 (1992) 171. Yu E Lozovik, A M Popov Usp. Fiz. Nauk 167 751 (1997) c 172. P M Ajayan, T W Ebbesen Rep. Prog. Phys. 60 1025 (1997) 173. A Nikolic, V Radomolivic, M Simicic, D Koruga Adv. Mater. Process. 282 83 (1998) 174. A V Eletskii Usp. Fiz. Nauk 167 945 (1997) c 175. E G Gal'pern, I V Stankevich, L A Chernozatonskii, A L Chistyakov Pis'ma Zh. Eksp. Teor. Fiz. 56 483 (1992) 176. Z Ya Korsakovskaya, L A Chernozatonskii, E A Fedorov Pis'ma Zh. Eksp. Teor. Fiz. 56 26 (1992) 177. L A Chernozatonskii Phys. Lett. A166 55 (1992) 178. E D Obraztsova, M Fujii, S Hyashi, V L Kuznetsov, Y V Butenko Carbon 36 821 (1998) 179. Z L Wang, J S Yin Chem. Phys. Lett. 289 189 (1998) 180. P J F Harris Int. Mater. Rev. 42 206 (1998) 181. P Schraff Carbon 36 481 (1998) 182. B Wei, J Zhang, J Liang, D Wu Carbon 36 997 (1998) 183. L A Chernozatonskii Phys. Lett. A170 37 (1992) 184. E G Gal'pern, I V Stankevich, L A Chernozatonskii, A L Chistyakov Fullerene Sci. Technol. 2 1 (1994) 185. V Meunier, Ph Lambin, A A Lucas Phys. Rev. B, Solid State 57 14 886 (1998) 186. J Han, R Jaffe J. Chem. Phys. 108 2817 (1998) 187. H Terrones, M Terrones Carbon 36 725 (1998) 188. L A Chernozatonskii Phys. Lett. A172 173 (1992) 189. L A Chernozatonskii, E F Kukovitskii, A L Musatov, A B Ormont, S G L'vov Carbon 36 713 (1998) 190. J Lefebvre, R Antonov, A T Johnson J. Appl. Phys. A67 71 (1998) 191. B An, S Fukuyama, K Yokogawa, M Yoshimura Jpn. J. Appl. Phys. 37 3809

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (2) 119 ± 135 (1999) Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems V K Belsky, BMBulychev Contents I. Introduction II. Systems with halides of the first-row elements III. Systems with halides of the second-row elements IV. Systems with halides of the third-row elements V. Systems with halides of the fourth-row elements VI. Systems with halides of the fifth-row elements VII. Systems with actinide halides VIII. Conclusions Abstract. The review surveys data on over 200 crystal and molecular structures of metal halide complexes with oxygen- containing crown ethers. The major types of the complexes are distinguished. The characteristic features of formation of these complexes and their coordination fragments are discussed.The bibliography includes 143 references. I. Introduction Structural aspects of the chemistry of macrocyclic polyether complexes with metal salts have been discussed in several reviews and monographs (see, for example, Refs 1 ± 3). These data were most completely surveyed in the monograph 2 containing a bibliography up to 1989. In succeeding years, a large number of publications have been devoted to structural studies of crown compounds with metal halides prepared from organic media, including those devoid of water molecules, along with studies of salts of alkali and alkaline-earth metals (which comprise the major body of structural data throughout the period of studies of these compounds). The results obtained confirmed on the whole the existence of several structural types of crown complexes, which were suggested for the first time by Dalley 4 and more recently by Poonia.5 The first group comprises complexes in which all the oxygen atoms of the crown ether (CE) molecule are coordinated to the metal atom, which is located within or outside the CE cavity (type A). The second group includes complexes in which only some of the oxygen atoms are involved in the coordination sphere of the metal atom (type B).The third group contains complexes in which the metal atom is located outside the cavity of the crown ether and is not directly bound to the oxygen atoms of the CE molecule (type C). Finally, complexes of the `half-sandwich' type in which the metal atom deviates substantially from the mean V K Belsky State Research Centre of the Russian Federation `L Ya Karpov Research Physicochemical Institute,' ul.Vorontsovo Pole 10, 117120 Moscow, Russian Federation. Fax (7-095) 975 24 50. Tel. (7-095) 916 31 46 BMBulychev Department of Chemistry,MV Lomonosov Moscow State University, Leninskie gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 31 81. Tel. (7-095) 939 36 91 Received 22 May 1998 Uspekhi Khimii 68 (2) 136 ± 153 (1999); translated by T N Safonova #1999 Russian Academy of Sciences and Turpion Ltd UDC 548.737 119 119 120 121 127 129 131 132 plane through the oxygen atoms of the CE molecule (type D) are documented. The structural classification of crown compounds has been based on CE as such.2, 3 At that time, this was quite reasonable because the data on structurally characterised crown ethers were scarce in the literature.Presently, it seems difficult to use this classification. In the present review, we consider structural fea- tures and compositions of crown compounds with metals in the order of increasing atomic numbers. We deliberately exclude CE complexes with sodium halides and with halides of the heaviest alkali metals from consideration since their structures are rather trivial and have been reported many times. II. Systems with halides of the first-row elements 1. Lithium halides The reaction of anhydrous lithium chloride with 12-crown-4 (12C4) afforded a molecular complex with composition [(LiCl) .(12C4)] (1).6 The lithium atom is coordinated by the chlorine atom (Li ±Cl = 2.290 A) and four oxygen atoms (Li ±O=2.128 A) and deviates from the mean plane through the oxygen atoms by 0.929 A (Fig. 1). Li COCl Figure 1. Structure of the complex 1. The reaction of lithium chloride dihydrate with 18-crown-6 (18C6) in dry diethyl ether gave a molecular binuclear complex with composition [(H2O)2 . (LiCl)2 . (18C6)] (2).7 Therefore, the CE molecule with a large inner cavity serves as a `host' simulta- neously for two lithium atoms. The coordination environment about each lithium atom is a strongly distorted tetrahedron formed by two oxygen atoms of the polyether, one oxygen atom of the water molecule and one chlorine atom.This coordination120 results in small five-membered rings with the following distances: 2.065 A, Li ±Ocr=1.980 Li ±Ow=1.865 A and and Li ±Cl=2.277 A (Fig. 2). The five-membered metallocycles in complex 2 adopt an envelope conformation. Li COCl Figure 2. Structure of the complex 2. By contrast, this reaction in a typical solvent, which have not been dehydrated,8 gave structurally different ionic complexes with compositions {[(Li)2 . (H2O)4 . (18C6)]2+ + 2X7}, where X=Cl (3) or Br (4), with a unique structure. In these compounds, two lithium atoms are also `guests' but they are arranged nonsym- metrically with respect to the centre of the CE molecule. Both cations have a trigonal-bipyramidal environment (Fig.3), one of them being coordinated by three oxygen atoms of the ether molecule and two oxygen atoms of water molecules, while the second cation, on the contrary, being coordinated by two oxygen atoms of the ether molecule and three oxygen atoms of water molecules. The bipyramids are linked to each other through a shared vertex formed by the bridging oxygen atoms. The Li ±Ocr bond lengths vary from 1.88 to 2.12 A and the Li ±Ow bond lengths are in the range 1.90 to 2.00 A. In the crystal structure, an extensive three-dimensional network of OH. . .O and OH. . .X intermolecular hydrogen bonds exists. Li CO Figure 3. Structure of the cation in the complex 3. 2. Beryllium chloride Only one complex of BeCl2 with crown ethers, viz., [(BeCl2) .(15C5)] (5), which was prepared by the reaction of anhydrous beryllium chloride with the polyether in anhydrous diethyl ether, was structurally characterised.9 In this complex, the beryllium atom retains the typical tetrahedral coordination formed by two oxygen atoms of the crown ether molecule Be COCl Figure 4. Structure of the complex 5. V K Belsky, BMBulychev (Be ±Ocr=1.64 and 1.73 A) and two chlorine atoms (Be ±Cl=1.88 and 2.0 A) (Fig. 4). Therefore, complex 5 is a typical representative of type B compounds. The bond angles in theO2BeCl2 tetrahedron vary from 90 to 118 8, which is associated with steric hindrances due to coordination of two closely arranged oxygen atoms of the polyether.3. Boron trifluoride Presently, two complexes of BF3 with 18C6, which have been prepared by the reaction of boron trifluoride etherate with the polyether, are known, viz., the complex with composition {[(BF3) . (OH2)] + (18C6)} (6) solvated with toluene 10 and the complex {2[(BF3) . (H2O)]+(18C6)+(H2O)2} (7) obtained when BF3 and CE were taken in a 2 : 1 ratio.11 Both compounds belong to type C complexes because the tetrahedral boron-containing fragments interact with the CE molecule through hydrogen bonds between the coordinated water molecules and the oxygen atoms of the crown ether molecule. In structure 6, only two oxygen atoms of the ether molecule are involved in `guest ± host' interactions, while in structure 7 all six oxygen atoms of the ether molecule participate in these interactions. The reaction of the complex BF3 .Et2O with dicyclohexyl-18- crown-6 (DCHex18C6) in the presence of atmospheric moisture 11 afforded a product with composition {[(BF3) . (H2O)2] + (DCHex18C6)} with intramolecular H-bonds analogous to those observed in structure 7. III. Systems with halides of the second-row elements 1. Magnesium chlorides In the compound {[(Mg) . (H2O)6]2++(12C4)+2Cl7} (8) (Fig. 5), which was prepared by the reaction of MgCl2 . 6H2O with 12C4, the metal atom is not involved in direct bonding with the oxygen atoms of the CE molecule.12 Hence, this complex should be assigned to type C. In the crystal structure of 8, the [(Mg) . (H2O)6] octahedra, the CE molecules and the chloride ions are linked to each other through a complex system ofOH.. .Oand OH. . . Cl hydrogen bonds. The Mg±Ow distances in the cation vary from 2.052 to 2.082 A. Mg COCl Figure 5. Structure of the complex 8. The reactions of slightly hydrated magnesium chloride with 15C5 in acetonitrile or chloroform gave related compounds with compositions {[(H2O) . (MgCl) . (15C5)]++Cl7} . (L), where L=CH3CN (9) or CHCl3 (10), respectively. The magnesium atom is located in the cavity of the CE molecule (type A complex) and has a pentagonal-bipyramidal coordination.13 Five oxygen atoms of the crown ether molecule occupy the equatorial plane of the polyhedron (Mg ±Ocr=2.11 ± 2.23 A). The water molecules (Mg±Ow=2.05 and 2.06 A) and the chlorine atoms (Mg ±Cl=2.42 and 2.44 A) are located in the apical positions (Fig.6). The coordinated water molecule forms intermolecular hydrogen bonds with the chloride ions (OH . . . Cl7=3.10 ± 3.11 A).Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems Mg COCl Figure 6. Structure of the complex 9. In the case of complete hydration of the above-described complexes, which is achieved by keeping these complexes in an atmosphere with 100% humidity,14 the apical chlorine atom is replaced by the second water molecule and the composition of the complex corresponds to the formula {[(Mg) . (H2O)2 . (15C5)]2+ +2Cl7+(H2O)} (11). TheMg±Ocr distances vary from 2.086 to 2.225 A and Mg±Ow=2.045 A.In this crystal, the system of OwH. . .Ow and OH. . . Cl hydrogen bonds is more complex than those observed in crystals of 9 and 10. The reaction of anhydrous magnesium chloride with 18C6 in anhydrous tetrahydrofuran 15 afforded a molecular complex [(MgCl2) . (18C6)] (12) (Fig. 7). The coordination polyhedron Mg COCl Figure 7. Structure of the complex 12. about the magnesium atom is a pentagonal pyramid. The coordi- nation number of the Mg atom is 7, which is very untypical of this metal. The Mg±O distances in complex 12 vary from 2.220 to 2.331 A. The CE molecule also adopts an unusual conformation due to coordination of the metal atom by only five out of the six oxygen atoms. The magnesium atom has analogous coordination in the complex {[(MgCl2) .(18C6)] . (HCl)2} (13) (Fig. 8),16 which was prepared by the reaction of MgCl2 with the clathrate of composition {[(H3O) . (18C6)]+ + [HCl2]7}. This complex is considered 16 as a purely ionic compound with composition {[(Mg) . (18C6)]2++2[HCl2]7}. However, the Mg±Cl distances in 13 (2.428 and 2.454 A) and in 12 (2.429 A) are close to the sum of the covalent radii of both elements and, in our opinion, it is more reasonable to describe both compounds as molecular com- plexes. Mg COCl H Figure 8. Structure of the complex 13. 121 2. Aluminium chloride A complex with composition {[(AlCl2) . (15C5)]+ + [AlCl4]7} (14) was produced by the reaction of anhydrous AlCl3 with 15C5 in anhydrous ether.17 The coordination number of the aluminium atom in the cation is 7.The coordination polyhedron about the metal atom is a pentagonal bipyramid with the oxygen atoms of the crown ether molecule in the equatorial plane and the chlorine atoms in the axial positions. This type of the polyhedron and the coordination number are untypical of aluminium compounds. The Al ±O distances vary from 2.081 to 2.142 A and the Al ±Cl distances are 2.188 and 2.210 A. The statistically disordered crown ether molecule in 14 is located on the mirror plane (Fig. 9). The complex of AlEtCl2 with benzo-15-crown-5 (15) has an analogous structure.18 Al COCl Figure 9. Structure of the complex 14. Complete hydration of complex 14 gave a compound with composition {[(Al) .(H2O)6]3+ + (15C5) + 3Cl7 + (H2 O)4} (16)7 in which the metal atom is not coordinated by the oxygen atoms of the crown ether molecule and the individual structural units are linked to each other through a complex network of OH. . .O and OH. . . Cl hydrogen bonds. 3. Silicon tetrafluoride The only silicon-containing complex with CE known to date was prepared by the reaction of SiF4 with 18C6.19 The compound has the composition {[(SiF4) . (H2O)2] + (18C6) + (H2O)2} (17) and belongs to type C complexes. The silicon atom is in an octahedral environment. The isolated structural fragments are linked to each other through intermolecular OH. . .O hydrogen bonds with the participation of water and CE molecules. IV. Systems with halides of the third-row elements 1.Calcium chloride The reaction of crown ether 12C4 with an aqueous solution of calcium chloride afforded a compound with composition {[(Ca) . (12C4) . (H2O)4]2++2Cl7+(H2O)2} (18).20 The calcium atom is located above the mean plane of the macrocycle and is coordinated by all the four oxygen atoms of CE and the four water molecules (Fig. 10). The Ca ±O distances vary from 2.380 to Ca COCl Figure 10. Structure of the complex 18.122 2.540 A. In the crystal structure of 18, a complex network of OH. . .O and OH. . . Cl intermolecular hydrogen bonds with the participation of all water molecules is realised. The reaction of molten CaCl2 with 18C6 in acetone yielded 7 a complex with composition {[(Ca) .(18C6) . (H2O)2]2+ + 2Cl7 + (H2O)2} (19), which is a typical representative of type A crown compounds. The calcium atom is located exactly in the centre of the cavity of the CE molecule (Ca ±Ocr=2.555 ± 2.604 A) and is additionally coordinated by two water molecules (Ca ±Ow=2.29 A) which are located above and below the mean plane of the macrocycle. 2. Scandium chloride Data on homometallic complexes of scandium halides with CE molecules are unavailable in the literature. The structures of heterometallic halides have been studied.21 ± 23 These complexes were prepared by the following reactions: 2ScCl3+CuCl2+2 15C5 {2[(ScCl2) . (15C5)]++[CuCl4]27} , 20 ScCl3+SbCl5+18C6 {[(ScCl2) . (18C6)]++[SbCl6]7} , 21 ScCl3+SbCl5+DB24C8 {[(ScCl2) .(DB24C8)]++[SbCl6]7} , 22 ScCl3+SbCl5+DB30C10 {[(ScCl2) . (DB30C10)]++[SbCl6]7}. 23 In compounds 20 ± 23, the scandium atom is located in the centre of the cavity of the CE molecule and has a bipyramidal coordination. The Sc ±O distances vary from 2.09 to 2.30 A. 3. Titanium chlorides The compositions and structures of complexes of TiCl4 with various CE reflect the general tendency of interactions between metal halides in the highest oxidation states and polyethers, which is associated with destruction of the latter and the formation of various reactive oxygen-containing fragments. Only one complex of TiCl4 with 12C4, which was prepared by the reaction of a mixture of TiCl4 and SbCl5 in acetonitrile, was reported in the literature.24 This reaction yielded a complex with composition {[(TiCl3) .(CH3CN)3]+ + [SbCl6]7} as an inter- mediate. Treatment of the latter with 12C4 in dichloromethane afforded a compound with composition {[(TiCl)2(m2-O) . . (12C4)]2+ + [SbCl6]7 + (CH2Cl2)} (24). The bridging oxygen atoms in this structure occur apparently due to destruction of a portion of CE molecules. The structure of the Ti-containing fragment is shown in Fig. 11. The Ti ±Om distances (1.81 and 1.82 A) are intermediate between the length of the double bond in titanyl chloride and the coordination bond. The Ti ±Ocr distances vary from 2.12 to 2.38 A, i.e., these bonds are nonequivalent. The coordination number of the Ti atom is 7. The coordination polyhedron about the metal atom is a distorted pentagonal Ti COCl Figure 11.Structure of the [(TiCl)(m2-O) . (12C4)]á2 fragment in the com- plex 24. V K Belsky, BMBulychev bipyramid with the oxygen and chlorine atoms in the apical positions. Apparently, analogous destruction of CE to form titanyl chloride as an intermediate occurs in the reaction of TiCl4 with 15C5 in acetonitrile.25 The resulting complex has the composition [(TiCl) . (15C5) . (m2-O)(TiCl5)] (25) (Fig. 12). Unlike the above- described compound, the Ti ±O distances in 25 are essentially nonequivalent (1.691 and 2.014 A), which is indicative of the predominant formation of the Ti(1) ±O double bond and the Ti(2) ±O coordination interaction due to the lone electron pair of the bridging oxygen atom.The Ti ±Ocr distances vary from 2.073 to 2.140 A. The coordination polyhedra about the Ti(1) and Ti(2) atoms are a pentagonal bipyramid and a typical octahedron, respectively. Ti COCl Figure 12. Structure of the complex 25. It was demonstrated 25 that destruction of CE is suppressed upon introduction of metal halides with readily ionisable M±X bonds which are not prone to autocomplexation. Thus the reaction of TiCl4 with 15C5 in the presence of MgCl2 in acetoni- trile afforded the complex {[(Mg) . (CH3CN)2 . (15C5)]2+ + [TiCl6]27} (26) in virtually quantitative yield. As in the com- pounds 9 ± 11, the magnesium atom in the complex 26 has a pentagonal-bipyramidal environment. The anion has a usual octahedral structure with the Ti ±Cl distances varying from 2.324 to 2.354 A. The reaction of TiCl3 and AlCl3 with 15C5 in tetrahydrofuran 25 gave a complex with composition {[(TiCl2) .(15C5)]+ + [AlCl4]7} (27). In 27, the titanium atom is located in the centre of the cavity of the CE molecule. The Ti ±O distances vary from 2.12 to 2.20 A and Ti ±Cl=2.38 A. The structure of the anion is analogous to that of the anion in the complex 14. According to the published data, the reaction of TiCl4 with 18C6 did not lead to destruction of the CE molecule. Thus a distinct type B complex with composition [(TiCl4) . (18C6)] (28) was prepared.26 The coordination polyhedron about the titanium atom in this compound is an octahedron formed by four chlorine atoms and two oxygen atoms of the CE molecule (Ti ±O=2.125 and 2.145 A).The CEmolecule is strongly distorted (Fig. 13). The reaction of TiCl4 with 18C6 in ethanol 27 afforded a type C Ti COCl Figure 13. Structure of the complex 28.Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems complex with composition [(TiCl3) . (OC2H5) . (C2H5OH)2 . . (18C6)2] (29). The environment about the titanium atom is an octahedron (Fig. 14) formed by three chlorine atoms (Ti ±Cl= 2.337 ± 2.387 A), the oxygen atom of the ethoxy group (Ti ±O=1.726 A) and two coordinated ethanol molecules (Ti ±O=2.030 and 2.033 A). The coordinated ethanol molecules interact with the macrocycle molecules through OH. . .O hydro- gen bonds (2.63 and 2.64 A).Ti COCl Figure 14. Structure of the complex 29. 4. Vanadium chlorides The reaction of VCl4 with 15C5 in acetonitrile in the presence of water was accompanied by disproportionation of the halide to form a complex with composition {[(VIIICl2) . (15C5)]+ + [VVOCl4]7} (30).28 The coordination polyhedron about the vanadium atom in the cation is a pentagonal bipyramid formed by the five oxygen atoms of the macrocycle (V ±O=2.08 ± 2.11 A) and the two apical chlorine atoms (V ±Cl=2.34 and 2.35 A). The coordination environment about the vanadium atom in the anion is a tetragonal pyramid formed by four chlorine atoms (V ±Cl=2.23 ± 2.26 A) and one oxygen atom (V ±O=1.548 A). The structure of the complex prepared by the reaction of VOCl2 with 18C6 in aqueous ethanol was reported 29, 30 to belong to type C and it is described by the formula {[(VOCl2) .(H2O)2] + [(18C6) . (H2O)]} (31). The coordination polyhedron about the vanadium atom is a slightly distorted tetragonal pyramid (V ±O=1.573 A, V±Ow=1.964 2.024 A and and V± Cl=2.215 and 2.362 A). Both fragments of the structure are linked to each other through OH. . .O intermolecular hydrogen bonds with the participation of water and CE molecules. 5. Chromium trichloride Presently, only one structure of the chromium-containing com- plex with composition [(CrCl3) . (H2O) . (15C5)] (32) (Fig. 15) is known.31 This complex was prepared by the reaction of CrCl3 with 15C5 in aqueous acetonitrile.In the monomolecular complex of type B, the coordination environment about the chromium Cr COCl Figure 15. Structure of the complex 32. 123 atom is an octahedron formed by three chlorine atoms, one oxygen atom of the water molecule and two (of five) oxygen atoms of the CE molecule (Cr ±Ocr=2.05 ± 2.10 A). 6. Manganese halides The reaction of manganese bis(tribromide) with 12C4 gave the sandwich complex {[(Mn) . (12C4)2]2++2[Br3]7} (33) (Fig. 16).32 TheMn ±O distances vary from 2.241 to 2.369 A. Mn CO Figure 16. Structure of the [(Mn) . (12C4)2]2+ cation in the complex 33. One product of the reaction of MnCl2 with 18C6 in tetrahy- drofuran is a compound with composition {2[(H3O) . (18C6)]++ [MnCl4]27} (34). Its formation can only be attributed to partial hydrolysis in the solution or to the use of the reagents containing water.33 The oxonium ion in 34 is located in the centre of the cavity of the CE molecule.In the nearly regular tetrahedral anion, the Mn±Cl distances vary from 2.35 to 2.37 A. Interestingly, hydrol- ysis products were not detected in the analogous reaction per- formed in the presence of an excess of water. In this case, the salt underwent complete ionisation and the cation was hydrated.34, 35 The structure of the Mn-containing fragment is more complex and the CE molecule is located in the second coordination sphere. 2MnCl2+18C6+8H2O= ={[(H2O)4 . Mn(m-Cl2)Mn . (H2O)4]2++2Cl7+(18C6)}. 35 Individual fragments of the structure 35 (Fig.17) are linked to each other through a complex system of OH. . .O and OH. . . Cl hydrogen bonds. Mn COCl Figure 17. Structure of the [(H2O)4 . Mn(m-Cl2)Mn . (H2O)4]2+ cation in the complex 35. 7. Iron halides The reaction of FeBr2 with 15C5 in dry dichloromethane afforded a complex of type A with composition {[(FeBr2) . (15C5)]+ (CH2Cl2)] (36) with the iron atom in the pentagonal-bipyramidal124 environment. The Fe ±O distances in the equatorial plane vary from 2.18 to 2.26 A and the Fe ± Br distances (the Br atoms are located in the apical positions) are in the range of 2.623 ± 2.628 A.36 The reaction of FeCl3 with 15C5 in dry tetrahydrofuran proceeded analogously to the reaction of aluminium chloride with this CE (see the preparation of 14).The process yielded an ionic complex with composition {[(FeCl2) . (15C5)]++[FeCl4]7} (37) as the final product.37 The Fe ±Ocr distances in the cation vary from 2.112 to 2.192 A. The reaction of FeCl3 or FeBr3 with 18C6 in ordinary terahydrofuran containing water resembles the above-considered reaction of MnCl2 with 18C6, i.e., this reaction proceeded as hydrolysis to form type C complexes with composition {[FeX4]7 + [(H3O) . (18C6)]+} [X=Cl (38) or Br (39)]. In the resulting complexes, the oxonium ion is a `guest.' The cations have a flattened tetrahedral structure.38, 39 An analogous reaction per- formed in a mixture of dry tetrahydrofuran and diethyl ether afforded an unusual sandwich complex of type B with composi- tion {[(FeCl) .(18C6)2]2+ + 2[FeCl4]7} (40) (Fig. 18).40 Crystals of this complex contain the dications [(FeCl) . (18C6)2]2+ in which the five-coordinate iron atom exists in two equally probable orientations. The coordination environment about the iron atom in the cation is a distorted octahedron formed by two chlorine atoms [statistically disordered positions (50%)] and four of the six oxygen atoms of theCE molecule (the Fe ±Ocr distances vary from 2.156 to 2.209 A). The anion has a flattened tetrahedral structure. Fe COCl Figure 18. Structure of the [(FeCl) . (18C6)2]2+ cation in the complex 40. The reaction of an equimolar mixture of FeCl2 and FeCl3 with 18C6 in tetrahydrofuran gave a product with an interesting structure.41 In the complex of type B, the FeII atoms form complex cations {2[(Fe) .(H2O) . (C4H8O) . (18C6)](m2-Cl2)}2+ (41) (Fig. 19). The coordination polyhedron about the metal atom is an octahedron. The equatorial plane is formed by two chlorine atoms (Fe ±Cl=2.43 ± 2.47 A) and two oxygen atoms of the CE molecule (Fe ±Ocr=2.14 ± 2.18 A). The apical positions are Fe COCl Figure 19. Structure of the cation {2[(FeCl) . (H2O) . (C4H8O) . (18C6) (m-Cl2)]}2+ in the complex 41. V K Belsky, BMBulychev occupied by the oxygen atom of the water molecule (2.06 A) and the oxygen atom of the tetrahydrofuran molecule (2.13 A). The FeIII atoms form tetrahedral anions [FeCl4]7. 8. Cobalt chlorides The reactions of CoCl2 with 15C5 were studied in a series of works.42 ± 47 This reaction in acetonitrile always afforded the inner-cavity cation [(Co) .(CH3CN)2 . (15C5)]2+ but the composi- tion of the anion in the complex varied from [(CoCl3) . (CH3CN)]7 (42) and [CoCl4]27 (43) (mononuclear tetrahedral) to [Co2Cl6]27 (44) (binuclear bridging) depending on the concentration of CoCl2. The presence of water in organic solvents resulted only in hydration of the complex rather than in hydrolysis of the salt. In this case, the acetonitrile molecules located in the apical positions of the pentagonal bipyramid (the coordination polyhedron about the cobalt atom) were replaced by water molecules with retention of the structural type of the cation [(Co) . (H2O)2 . (15C5)]2+ (see Ref. 44).Only free Cl7 ions (45) or tetrahedral anions [CoCl4]27 (46) act as anions. In all the structures under consideration, the coordination polyhedron about the cobalt atom is a pentagonal bipyramid. The Co ±Ocr distances vary from 2.07 to 2.40 A. The reaction of anhydrous CoCl2 with 15C5 in ethanol yielded a complex of type A with composition {[(Co) . (EtOH)2 . . (15C5)]2++[CoCl4]27} (47).45 The Co ±OEt distances vary from 2.11 to 2.13 A. The reaction of CoCl2 with 15C5 in the presence of an equimolar amount of MgCl2 in acetonitrile afforded an inner- cavity magnesium complex with CE with composition {[(Mg) . (CH3CN)2 . (15C5)]2+ + [CoCl4]27 + (CH3CN)} (48).42 One product of hydration of this compound is a homometallic complex with composition {[(Co) .(H2O)2 . (15C5)]2++2Cl7+ (H2O)2.5} (49). The cobalt atom is in the typical pentagonal- bipyramidal coordination. The structural fragments are linked to each other through a system of OH. . .O and OH. . . Cl intermo- lecular hydrogen bonds.42 The reaction of a mixture of cobalt chloride and copper chloride with 15C5 in acetonitrile gave a complex with composi- tion {[(Co) . (CH3CN)2 . (15C5)]2+ + [Cu2Cl6]27} (50). The Co ±Ocr distances in this compound vary from 2.103 to 2.269 A, Co ±N=2.097 A. Complete hydration of this complex led to changes in the composition of the cation and in the structure of the anion. The resulting compound has the composition {[(Co) . (H2O)2 . (15C5)]2++[CuCl4]27} (51). The reaction of anhydrous CoCl2 with 18C6 in acetonitrile yielded a compound with composition {[(Co) .(CH3CN) . . (18C6)]2++[CoCl4]27} (52). In this complex, the cobalt atom has a rather untypical hexagonal-pyramidal coordination (the coordination number is 7). The metal atom is coordinated to all the six oxygen atoms of the CE molecule (Co ±Ocr=2.26 A) and to the nitrogen atom (Co ±N=1.83 A) and deviates from the mean plane through the oxygen atoms by 0.5 A. The values of the torsion angles indicate that the CE molecule is highly strained.45 The reaction of CoCl2 . 6H2O with 18C6 in acetone afforded a complex with composition {[(Co) . (H2O)6]2++[CoCl4]27+ (18C6)+[(CH3)2CO]} (53). All structural units of this complex are linked to each other through a branched system of OH.. .O and OH. . . Cl hydrogen bonds.48 The reaction of a mixture of CoCl2 and BaCl2 with 18C6 gave a compound {[(Ba) . (H2O)2 . (18C6)]2++[CoCl4]27} (54).49 9. Nickel chloride The reaction of anhydrous NiCl2 and NaCl with 15C5 gave a complex with composition {2[(Na) . (15C5)]+ + [NiCl4]27} (55) in which the tetrahedral anion and the complex cation are linked through an Na . . . Cl interaction (2.65 A).50 The reaction of nickel chloride dihydrate with 18C6 in meth- anol afforded a compound {[(H2O)4 . Ni(m-Cl2)Ni . (H2O)4]2++ (18C6)+2Cl7} (56) the structure and composition of which are analogous to those of the manganese complex 35.51Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems 10.Copper halides The products of reactions of a mixture of monovalent copper iodide and KI with various CE in acetone were studied by Rath and Holt.52 Complexes containing potassium cations, CE mole- cules (dibenzo-24-crown-8, DB24C8, or dibenzo-18-crown-6, DB18C6) and [CuxIy]n7 anions of different compositions were prepared according to the following reactions: 12CuI+7KI+6 12C4 [Cu4I6]27+[Cu8I13]57+7K++6 12C4 , 57 4CuI+2KI+2 15C5 [Cu4I6]27+2K++2 15C5 , 58 3CuI+KI+DB24C8 [Cu3I4]7+K++DB24C8 . 59 The reaction of CuI with DB18C6 in acetonitrile gave a molecular complex of composition {[(Cu4I4) . (CH3CN)4] + (DB18C6)} (60) with fragments that are not linked to each other.53 Only one structure of a monomolecular complex of CuCl2 with 12C4 (61) (Fig.20) has been reported.54 In this complex, the copper atom has an octahedral coordination formed by four oxygen atoms of the CE molecule and two chlorine atoms (Cu ±Ocr=2.113 ± 2.403 A and Cu ±Cl=2.214 ± 2.228 A).Cu COCl Figure 20. Structure of the complex 61. The reaction of CuCl2 with 15C5 in acetonitrile was studied in several works.55 ± 57 Ionic complexes of the type {[(Cu) . . (CH3CN)2 ±x . Clx . (15C5)] + A}, where x=0 or 1 and A is an anion in which the coordination environment about the copper atom is a pentagonal bipyramid, were synthesised. The following complexes were obtained depending on the halide : CE : solvent ratio: 4CuCl2+2 15C5+2CH3CN {2[(CuCl) . (CH3CN) . (15C5)]++[Cu2Cl6]27} , 62 4CuCl2+15C5+2CH3CN {[(Cu) .(CH3CN)2 . (15C5)]2++[Cu3Cl8]27} . 63 The Cu ±Odistances in these compounds vary from 2.16 to 2.34 A and the Cu ±N distances are in the range of 1.95 ± 1.99 A. In the complex 63, the anion exists as an infinite chain of alternating tetragonal-bipyramidal and tetragonal-pyramidal fragments linked through bridging chlorine atoms (Fig. 21). The Cu ±Cl distances are in the range of 2.255 ± 2.787 A. Hydration of compounds 62 and 63 with water vapour resulted in the removal of the CE molecules from the coordination sphere of the metal atom and in the change in the type of the complex from A to C.57 In this case, compounds with compositions {[(Cu) . (H2O)4]2+ + [(Cu3Cl8) . (H2O)2]27+(15C5)2} (64) and {2[(CuCl2) .(H2O)2] + (15C5)} (65) were formed along with CuCl2 . 2H2O. Separate fragments in 64 and 65 are linked to each other through OHw . . .Ocr hydrogen bonds (Figs 22 and 23, respectively). 125 Cu Cl Figure 21. Structure of the [Cu3Cl8]27 anion in the complex 63. Cu COCl Figure 22. Structure of the complex 64. Cu COCl Figure 23. Structure of the complex 65. The reaction of anhydrous CuCl2 with 15C5 in aqueous methanol gave a complex of type C with composition {[(CuCl2) . (- H2O)2 . (CH3OH)] + (15C5)} (66).58 The coordination environ- ment about the copper atom is a square pyramid. In the crystal structure of the complex,OH. . .Ointermolecular hydrogen bonds with the participation of water, methanol and CE molecules are realised.The reaction of CuBr2 with 15C5 yielded a complex with composition {[(H2O)2 . (CuBr2)]+ (15C5)} (67). The coordina- tion polyhedron about the copper atom is a strongly flattened tetrahedron.59 The structures of a number of products which were formed in the reactions of a mixture of CuCl2 and MgCl2 with 15C5 in acetonitrile where the reagents were taken in different ratios and at various concentrations were reported.60 CuCl2+MgCl2+15C5+CH3CN {[(CH3CN) . (Mg) . (15C5) . (m2-Cl) . (CuCl3)]} , 68 2CuCl2+MgCl2+15C5+2CH3CN {[(CH3CN)2 . (Mg) . (15C5)]2++[Cu2Cl6]27} , 69 3CuCl2+MgCl2+15C5+2CH3CN {[(CH3CN)2 . (Mg) . (15C5)]2++[Cu3Cl8]27} . 70 The structure 68 consists of monomolecular fragments (Fig.24), while structures 69 and 70 are ionic. In all the three compounds, the magnesium atom is located in the centre of the cavity of the CE molecule and has a pentagonal-bipyramidal126 Cu Mg COCl N Figure 24. Structure of the complex 68. coordination. In compound 70, the chain anion is analogous to that observed in the structure 63. Hydration of compounds 68 ± 70 with water vapour gave complexes with compositions {[(H2O)2 . (Mg) . (15C5)]2++ [CuCl4]27} (71) and {[(Cu3Cl8) . . (H2O)2]27 + [(Mg) . (H2O)6]2+ + (15C5)2} (72). The structure of the cation in 71 is analogous to that in 11. The metal atom in the cation of 72 has an octahedral coordination. The central copper atom in the anion has an octahedral coordination formed by two oxygen atoms of the water molecules and four chlorine atoms.The terminal copper atoms have a tetrahedral environment. The fragments are linked to each other through OHw . . .Ocr and OHw . . . Cl hydrogen bonds. The reactions of copper chloride with 18C6 in the presence of halides of alkali or alkaline-earth metals gave complexes in which the metal atoms are coordinated by CE molecules and the copper atoms form the [Cu2Cl6]27 anions.61 2CuCl2+2RbCl+2 18C6 2[(Rb) . (18C6)]++[Cu2Cl6]27 , 73 2CuCl2+BaCl2+18C6 [(Ba) . (18C6)]2++[Cu2Cl6]27 . 74 The product of the reaction of CuCl2 with benzo-15-crown-5 (B15C5) in chloroform was also structurally characterised.62 The product has the composition {[(CuCl2) . (B15C5)] + (CHCl3)} (75).The copper atom is located in the centre of the cavity of the CE molecule and has a pentagonal-bipyramidal coordination. This reaction in acetonitrile in the presence of NaCl gave a complex {2[(Na) . (B15C5)]++[Cu2Cl6]27} (76).63 An analogous product with composition {2[(Na) . (DB18C6)]+ + [Cu2Cl6]27} (77) was formed when DB18C6 was used as a polyether.63 The reaction with KCl and naphtho-15-crown-5 (N15C5) yielded a compound with a similar composition, {2[(K) . (N15C5)]+ + [Cu2Cl6]27 + (H2 O)} (78), containing a water molecule of crystallisation.64 11. Zinc halides The reactions of anhydrous ZnCl2 with 15C5 in different solvents gave complexes of type A of the general formula {[(Zn) . (L1) . (L2) . (15C5)]2+ + [Zn2Cl6]27}, where L1=Cl and L2=(CH3)2CO (79), L1=Cl and L2=H2O (80) (both prepared in acetone), L1=L2=C4H8O (81) (in tetrahydrofuran) and L1=L2=CH3CN (82) (in acetonitrile).65, 66 In all the structures under consideration, the zinc atom is located in the cavity of the CE molecule and has a pentagonal-bipyramidal coordination.The Zn ±Ocr distances vary from 2.13 to 2.27 A. The reaction of zinc chloride dihydrate with 15C5 in methanol yielded a typical complex of type C with composition {[(H2O)2 . (ZnCl2)] + (15C5)} (83).67 In this complex, the zinc atom has a tetrahedral coordination. The structural fragments are linked to each other through OH. . .O hydrogen bonds. In going from 15C5 to 18C6, the compositions and structures of zinc chloride complexes become more diversified.Thus the reaction in tetrahydrofuran 66, 68 gave a complex of type B with composition [(H2O) . (ZnCl2) . (18C6)] (84). The coordination pol- V K Belsky, BMBulychev yhedron about the zinc atom is a tetrahedron formed by two chlorine atoms (Zn ±Cl=2.20 and 2.23 A), one water molecule (Zn ±Ow=1.97 A) and one oxygen atom of the CE molecule (Zn ±Ocr=2.08 A). The water molecule, in turn, forms an intra- molecular hydrogen bond with another Ocr atom (2.77 A) result- ing in strong distortion of the macrocycle (Fig. 25). Zn COCl Figure 25. Structure of the complex 84. The reaction of ZnCl2 with 18C6 in diethyl ether afforded a complex of type B with composition {2(H2O)2 . (ZnCl) . . (18C6)]++[Zn2Cl6]27} (85).69 The zinc atom is coordinated by one chlorine atom, one oxygen atom of the water molecule and three (of six) oxygen atoms of the CE molecule.Recrystallisation of this compound from a mixture of acetone and CCl4 gave a complex of type C with composition {[(Zn) . (H2O)6]2++ 2[(ZnCl3) . (H2O)]7+ (18C6) + [(CH3)2CO]} (86). In the crystal structure of this complex, an extensive system of intermolecular hydrogen bonds exists. The reaction of ZnCl2 with 18C6 in terahydrofuran yielded a product with composition {[ZnCl4]27+2[(H3O) . (18C6)]7} (87) with the oxonium ion occupying the centre of the cavity of the CE molecule.32 Under analogous conditions, the reaction of ZnI2 with 18C6 afforded the compound {[(H2O)2 . (ZnI2)] + (18C6)2 + (H2O)} (88) in which the tetrahedral zinc-containing fragments are located between two CE molecules and are bound to the CE molecules through OH.. .O hydrogen bonds (Fig. 26). Zn COI Figure 26. Structure of the complex 88. 12. Gallium halides It was shown 70 that the complexes prepared by reactions of gallium halides with 15C5 have the composition {[(GaX2) . (15C5)]+ + [GaX4]7}, where X=Cl (89) or Br (90). These complexes are isostructural to the aluminium complex 14. The reaction of GaCl3 with 18C6 in acetonitrile afforded the compounds {[GaCl4]7 + [(H3O) . (18C6)]+} (91) with the oxo- nium ion occupying the centre of the CE molecule. The structure of this complex is analogous to that of the above-described manganese complex 34.Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems 13.Arsenic halides The structures of the molecular complexes [(AsCl3) . (12C4)] (92) (Fig. 27) and [(AsCl3) . (15C5)] (93) were reported.71 The As ±Ocr distances in these compounds vary from 2.776 to 2.915 A and from 2.944 to 3.156 A, respectively, which are only slightly smaller than the sum of the van der Waals radii. In general, a complex of arsenic bromide with 15C5 has an analogous structure.72 How- ever, the As . . .O bonds in the latter complex are nonequivalent: four bonds vary from 3.12 to 3.29 A and the fifth bond is substantially longer (3.85 A). As COI Figure 27. Structure of the complex 92. V. Systems with halides of the fourth-row elements 1.Zirconium tetrachloride The reactions of metal halides in the highest oxidation states with CE were accompanied by destruction of the latter. The reaction of ZrCl4 with 18C6 in tetrahydrofuran is a typical example.73 The resulting product (Fig. 28) has the composition {[(ZrCl5) . . (C4H8O)]7+[(ZrCl2) . (O5C8H16(CH2)2O(CH2)2Cl)]+} (94). The Zr ±Odistances (1.93 and 2.24 ± 2.33 A) in the cation indicate that both covalent and coordination interactions exist between the zirconium and oxygen atoms. Zr COCl Figure 28. Structure of [(ZrCl2) . (O4C8H16(CH2)2O(CH2)2Cl)]+ cation in the complex 94. 2. Molybdenum halides The reaction of MoCl5 with 15C5 in 1,2-dimethoxyethane (L) in the presence of sodium silicide was accompanied by reduction of MoV to MoIV (see Ref.74). The resulting complex of type C has the composition {[(MoCl4) . (L)] + (15C5)} (95). The coordina- tion polyhedron about the molybdenum atom is a distorted octahedron. The Mo±Ocr distances vary from 2.14 to 2.18 A. The reaction of MoCl5 with 15C5 in acetonitrile in the presence of AlCl3 gave the compound {[(MoOCl4) . (CH3CN)]7 + [(AlCl2) . (15C5)]+ + (CH3CN)} (96). It was assumed 75 that the oxychloride was formed as a result of decomposition of the CE molecule under the action of the strong Lewis acid, viz., MoCl5. The coordination polyhedron about the molybdenum atom in the 127 anion of 96 is an octahedron. The structure of the cation of 96 is analogous to that of the complex 14.The reactions of molybdenum oxychloride MoVOCl3 with 12C4 or 18C6 yielded complexes containing the octahedral cations [(MoOCl4) . (H2O)]7 and the oxonium-containing cations of the type [(H3O) . (CE)]+, viz., {[(MoOCl4) . (H2O)]7 + [(H3O) . . (12C4)]+} (97) and {[(MoOCl4) . (H2O)]7 + [(H3O) . (18C6)]+} (98), respectively.76, 77 The Mo±O and Mo±Ow distances in these complexes are 1.641 (1.654) and 2.285 (2.355) A, respectively. Separate structural fragments are linked to each other through a system of OH. . .O and OH. . . Cl hydrogen bonds. The reaction with MoOBr3 gave a complex with an analogous structure {[(MoOBr3) . (H2O)]7+[(H3O) . (18C6)]+} (99).78 3. Palladium chloride The reaction of PdCl2 with 18C6 in a tetrahydrofuran solution of hydrochloric acid afforded a binuclear complex {[Pd2Cl6]27 + 2[(H3O) .(18C6)]+} (100) of type C.38 The palladium atoms in the planar anion have a square coordination. The Pd ±Cl distances are 2.326 A (for the terminal bond) and 2.668 A (for the bridging bond). 4. Silver halides The reactions of silver halides with DB18C6 in dimethylforma- mide in the presence of alkali metal halides were studied by Helgesson and Jagner.79 AgCl+3KCl+3 DB18C6 AgCl+3RbCl+3 DB18C6 AgBr+3KBr+3DB18C6 AgBr+3RbBr+3 DB18C6 {3[(K) . (DB18C6)]++[AgCl3]27+Cl7} , {3[(Rb) . (DB18C6)]++[AgCl3]27+Cl7} , {3[(K) . (DB18C6)]++[AgBr3]27+Br7} , {3[(Rb) . (DB18C6)]++[AgBr3]27+Br7} . 101 102 103 104 In all cases, complexes containing planar anions of silver halides were detected (Fig.29). Rb Ag COCl Figure 29. Structure of the complex 102.128 5. Cadmium halides The reactions of cadmium halides with 18C6 gave typical com- plexes of type A with composition [(CdCl2) . (18C6)] (105), [(CdBr2) . (18C6)] (106) and [(CdI2) . (18C6)] (107) in which the cadmium atom is located in the centre of the cavity of the CE molecule and has a hexagonal-bipyramidal coordination.80, 81 The Cd ±Ocr distances in the complexes vary from 2.36 to 2.75 A. 6. Indium halides The reactions of indium(III) halide hydrates with 15C5 or 18C6 always yielded complexes of type C in which direct In ±Ocr contacts were absent. Thus the reaction of InCl3 . 4H2O with 15C5 gave the complex {[(InCl3) . (H2O)2] + (15C5)} (108).82 The coordination environment about the indium atom in this complex is a slightly distorted trigonal bipyramid.The In ±Ow distances are in the range of 2.201 ± 2.218 A. The fragments are linked to each other through OwH. . .Ocr hydrogen bonds. An analogous reaction with 18C6 afforded the complex {[(InCl3) . (H2O)3]+(18C6)} (109). In this compound, the indium atom has an octahedral environment. The In ±Ow distances vary from 2.23 to 2.25 A. The reaction of InCl3 with 18C6 in tetrahydrofuran afforded the compound {[InCl4]7+[(H3O) . (18C6)]+} (110) isostructural to the complex 38.38 The formation of the tetrahedral anions of the same type was also observed in the reactions of indium mixed halides with alkali metal halides and alkyl derivatives.83 ± 85 InCl3+CH3Li+15C5 {[(Li) .(15C5)]++[(InCl3) . (CH3)]7} , 111 InClI2+KI+18C6 {[(K) . (18C6)]++[InClI3]7} , 112 InClPri2+KCl+15C5 {[(K) . (15C5)]++[InCl2Pri2]7} . 113 7. Tin chlorides Only one example of complexation of SnCl2 with CE (18C6) in methanol saturated with hydrogen chloride is documented.86 This reaction gave the complex {[(SnCl) . (18C6)]+ + [SnCl3]7} (114) with a very unusual structure (Fig. 30). The tin atom in the cation has a nontypical hexagonal-pyramidal coordination with the Sn ±O distances varying from 2.59 to 2.88 A. The tin atom in the anion has a pyramidal coordination and is located in the mean plane of the base of the pyramid. Oxidation of SnCl2 in aqueous methanol containing 15C5 with atmospheric oxygen at 120 K yielded the complex of type C with composition {[(SnIV- Cl4) .(H2O)2] + [15C5]} (115).86 The coordination polyhedron about the tin atom is an octahedron. The Sn ±Ow distances vary from 2.107 to 2.125 A. The fragments in this compound are linked to each other through OwH. . .Ocr hydrogen bonds. The reaction of SnCl4 with 15C5 in acetonitrile resulted in decomposition of the CE molecule.88 In this case, two C±Obonds were cleaved, one ±CH2±CH2 ± unit was eliminated in the form of dichloroethane and two tin atoms were m2-coordinated by two Sn COCl Figure 30. Structure of the complex 114. V K Belsky, BMBulychev Sn COCl Sn(1) Sn(2) Figure 31. Structure of the complex 116. terminal oxygen atoms (Fig.31). The resulting complex has the composition {[(SnCl4) . (O5C8H16SnCl2)]+(CH3CN)} (116). The coordination environment about the Sn(1) atom is a pentagonal bipyramid. The Sn(1) ±Ocr coordination bonds are in the range of 2.237 ± 2.239 A. The Sn(1) ±Om covalent bond is 2.074 A. The Sn(2) atom has an octahedral coordination with the Sn(2) ±Om distances of 2.149 A. The reaction of SnCl4 with 18C6 in chloroform gave a complex of type C with composition {[(SnCl4) . (H2O)2] + (18C6) + (CHCl3) + (H2 O)} (117).89, 90 The octahedral tin- containing fragment is linked to the CE molecule through OwH. . .Ocr hydrogen bonds. In addition, there are H-bonds between the coordinated water molecules and the water molecules of crystallisation.8. Antimony halides The complex [(SbCl3) . (12C4)] (118) 71, 91 has a structure analo- gous to that of the complex 92. The reaction of a mixture of SbCl3 and SbCl5 with 12C4 in acetonitrile gave a compound with composition {[(SbIII) . (12C4)2]3+ + 3[SbVCl6]7 + (CH3CN)} (119).92 The Sb ±Ocr distances in the sandwich cation vary from 2.325 to 2.530 A. The reaction of SbCl3 with 15C5 afforded the monomolecular complex [(SbCl3) . (15C5)] (120).93 The pyramidal SbCl3 molecule is hanging over the mean plane of the CE molecule and the lone electron pair of the antimony atom is directed toward the poly- ether molecule. The Sb ±Ocr distances vary from 2.787 to 2.997 A. The reaction of SbBr3 with 15C5 and the reaction of SbCl3 with DB18C6 in chloroform afforded analogous structures [121 94 and 122 (Fig.32), respectively].{ Sb COCl Figure 32. Structure of the complex 122. The reaction of a mixture of tri- and pentavalent antimony chlorides with 18C6 in acetonitrile afforded a compound with composition {[(SbIIICl2) . (18C6)]++[SbCl6]7} (123) 95 in which the antimony(III) atom is coordinated by two chlorine atoms and only two oxygen atoms of the CE molecule. The coordination polyhedron about this atom (taking into account the lone electron pair) is a trigonal bipyramid. The anion in the complex 123 has the typical octahedral structure. {V K Belsky. Unpublished data.Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems 9. Tellurium tetrachloride The reaction of TeCl4 with a mixture of SbCl5 and 15C5 gave a bimetallic ionic complex with composition {[(TeCl3) .(15C5)]++ [SbCl6]7} (124). The Te ±Ocr distances vary from 2.66 to 2.79 A.96 VI. Systems with halides of the fifth-row elements 1. Barium bromide Only one structure of the complex formed by the reaction of barium bromide hydrate with 15C5 was reported in the litera- ture.97 The resulting compound has the composition {[(Ba) . (15C5)2]2+ + 2Br7 + (H2O)2} (125). The barium ion is located in the centre of the sandwich formed by CE molecules. The Ba . . .O distances are in the range of 2.749 ± 2.880 A. In the crystal, the Br7 anions and the water molecules are involved in the formation of a network of OH.. . Br intermolecular hydrogen bonds (3.326 ± 3.493 A). 2. Lanthanide and yttrium halides The structures of complexes of lanthanide halide hydrates with 12C4, which were crystallised from a mixture of methanol with acetonitrile, are known.98, 99 Compounds of three different struc- tural types were obtained depending on the nature of the metal atom and the reaction temperature. In the case of lanthanide and yttrium chlorides, the compositions and structures of the resulting complexes are similar, viz., {[(Ln) . (H2O)5 . (12C4)]3+ + 3Cl7 + (H2O)2}, where Ln=Ce (126) (Fig. 33), Nd (127), Sm (128), Eu (129), Gd (130), Dy (131), Y (132), Ho (133), Er (134) orYb (135). In these complexes, the coordination environment about the metal atom is a monocapped tetragonal antiprism formed by four oxygen atoms of the CE molecule and five oxygen atoms of the water molecules (Ln ±Ocr=2.42 ± 2.63 A and Ln ±Ow= 2.23 ± 2.51 A).All water molecules (both coordinated and of crystallisation) and chloride ions are involved in the formation of a complex system of OH. . . Cl intermolecular hydrogen bonds. Ce COCl Figure 33. Structure of the complex 126. In complexes with composition {[(H2O)2 . (LnCl2) . (12C4)]+ +Cl7}, where Ln=Ho (136), Er (137),Tm(138), Yb (139) or Lu (140) (Fig. 34), the coordination environment about the metal atom is a bicapped trigonal prism formed by four oxygen atoms of the CE molecule, two oxygen atoms of the water molecules and two chlorine atoms. The structural fragments are linked to each other through a system of OH.. . Cl hydrogen bonds. The reaction of yttrium chloride with 12C4 afforded a com- plex with composition {[(H2O) . (YCl2) . (CH3OH) . (12C4)]+ + Cl7} (141) in which one water molecule is replaced by the methanol molecule. Though the yttrium atom retains the coordi- nation number of 8, it has a different coordination environment (a square antiprism, Fig. 35). with complexes praseodymium The compositions {[(H2O) . (PrCl3) . (12C4)] + (12C4) + (CH3OH)} (142) and [(CH3OH) . (PrCl3) . (12C4)] (143) were also described in the literature.100 In both structures, the coordination environment Figure 34. Structure of the complex 136. Figure 35. Structure of the complex 141. cr=2.56 ± 2.63 A, Pr ±OCH about the metal atom is a tetragonal antiprism with Pr ±O- 3OH=2.51 A and Pr ±Ow= 2.48 A.In the complex 142, the structural fragments are linked to each other through OwH. . .Ocr hydrogen bonds (Fig. 36). Figure 36. Structure of the complex 142. Complete hydration of complexes of lutetium chloride with 12C4 gave eight-coordinate (square antiprism) octaaqualutetium ions.101, 102 {[(Lu) . (H2O)8]3++3Cl7+ LuCl3+12C4+9H2O +(12C4)+(H2O)} (at7150 8C) , 144 LuCl3+NaCl+2 12C4+9H2O + [(Na) . (12C4)2]++4Cl7+(H2O)} . 145 The reactions of LnCl3 with 15C5 generally gave complexes of two types. The complexes of the first type, viz., [(LnCl3) . (15C5)], were formed when anhydrous chlorides of praseodymium (146), neodymium (147) or samarium (148) were used.103, 104 The struc- tures of these complexes are analogous to those of trihalides of Group V elements (see, for example, structure 93).The pyramidal LnCl3 molecule is located above the mean plane of the CE molecule. The Ln ±Ocr distances vary from 2.50 to 2.75 A. The stabilities of these complexes were calculated by the INDO method.105 129 Ho COCl YCOCl Pr COCl {[(Lu) . (H2O)8]3++130 Complexes of the second type have the composition {[(Ln) . (H2O)8]3+ + (15C5) + 3Cl7 + (H2 O)n}. These com- pounds were obtained by the reactions with metal chloride hydrates in aqueous solvents or by hydration of lanthanide salts with water vapour. The complexes with Ln=Y (149), Gd (150) and Lu (151) were prepared.106 ± 109 In all cases, the coordination polyhedron about the metal atom is a distorted dodecahedron.Complex systems of OH. . .O and OH. . . Cl hydrogen bonds are observed in all the structures under consideration. Products of complexation of chlorides of rare-earth elements with 18C6 are characterised by different types of crystal struc- tures.110, 111 The reactions performed in a 1 : 3 mixture of methanol and acetonitrile afforded complexes with compositions {[(H2O)2 . (LnCl) . (18C6)]2++2Cl7+(H2O)2}, where Ln=Nd (152), Sm (153), Gd (154), Tb (155), Pr (156) or Eu (157). The metal atom is coordinated by six oxygen atoms of the CE molecule, one chlorine atom (located `above' the mean plane of CE) and two water molecules (located `below' the mean plane of CE) (Fig.37). Tb COCl Figure 37. Structure of the complex 155. The Ln ±Ocr and Ln ±Ow distances are in the ranges of 2.53 ± 2.55 and 2.36 ± 2.45 A, respectively. In all cases, the coordi- nation polyhedron about the lanthanide atom is a tricapped trigonal prism. The lanthanide atoms have the same coordination in another type of complexes in which one coordinated water molecule is replaced by the chlorine atom or the methanol molecule:111 {[(CH3OH) . (LaCl2) . (18C6)]+ + Cl7 + (H2O)1.5} (158), {[(H2O) . (LaCl2) . (18C6)]+ + Cl7} (159) and {[(H2O) . (CeCl2) . (18C6)]++Cl7 + (H2O)2} (160). An anhydrous lanthanum complex [(LaCl3) . (18C6)] (161) was prepared by crystallisation at 70 8C.111 The coordination polyhedron in the complex 161 is also similar to a tricapped trigonal prism (formed by six oxygen atoms of the CE molecule and three chlorine atoms).The addition of the salting-out reagent, LiCl, to a solution of [(GdCl3) . (H2O)6 + (18C6)] in methanol afforded a compound with a more complex composition, {[GdCl6]37 + 3[(H2O) . (GdCl2) . (18C6)]++2[CH3OH]} (162).111 The chlorine atoms and water molecules in the cation are statistically disor- dered. The structure of the complex 162 is analogous to those of compounds 159 and 160. The reactions of lanthanide chloride hexahydrates LnCl3 . 6H2O with 18C6 in a methanol : acetonitrile mixture under argon 112 gave complexes with composition {[(H2O)7 . (Ln) . (CH3OH)]3+ + 2[(H2O)2 .(LnCl) . (18C6)]2+ + 7Cl7 +(H2O)2}, where Ln=Y (163) or Dy (164) (X-ray data were collected at7150 8C). The structures of the double-charged cations are analogous to those observed in the complexes 152 ± 157. The trivalent cation has an octahedral environment. All the water and methanol molecules and noncoordinated chloride ions are involved in intermolecular hydrogen bonding. In addition, two more types of complexes of DyCl3 with 18C6 were reported. One of them was formed in acetonitrile and has the V K Belsky, BMBulychev composition {[(Dy) . (H2O)8]3+ + (18C6) + 3Cl7} (165).113 The coordination polyhedron about the dysprosium atom is a bicapped trigonal prism. In a complex with triethylene glycol (L) with composition {[(DyCl3) .(L)] + (18C6)} (166), the dyspro- sium atom has a pentagonal-bipyramidal environment.114 3. Tantalum chlorides As mentioned above, the reaction of CE with strong Lewis acids (halides of metals in the highest oxidation states) are often accompanied by opening and decomposition of the macrocycle to produce large amounts of water, hydrogen chloride (due to hydrolysis), ±O±CH2±CH2 ±O± fragments and products of their interactions. The products of the reaction of anhydrous TaCl5 with 15C5 have been studied most thoroughly.115 The reaction in dry acetonitrile afforded a mixture of several com- pounds. Thus crystals of an ionic complex with composition {[(TaCl2) . (C8H16O5)]++ [TaCl6]7} (167) were isolated at room temperature. The structure of the cation of this compound (Fig.38) is analogous to that of the complex 116. The Ta(1) is located in the centre of the mean plane formed by five oxygen atoms of the CE molecule. However, as in the case of the complex 116, the metal atom is bound to three oxygen atoms [O(1), O(4) and O(4a)] through donor-acceptor bonds (Ta ±O= 2.24 ± 2.27 A) and to two oxygen atoms [O(7) and O(7a)] through covalent bonds (Ta ±O=1.89 A). O(7a) O(4) Ta(1) O(7) Ta COHCl O(1) O(4a) Figure 38. Structure of the complex 167. The coordination polyhedron about the tantalum atom is a slightly distorted pentagonal bipyramid. The Cl ± Ta ± Cl angle is 162 8. Heating of the reaction mixture to 35 ± 40 8C afforded a complex with composition {[TaCl6]7 + [(H3O) . (18C6)]+} (168).115 The structure of the cation in this compound is analogous to those, for example, of the complexes 34, 87 and 91.The appearance of a new CE (18C6) in the TaCl5 ± 15C5 ±CH3CN system was attributed to processes of profound destruction of 15C5, chain extension of the open-chain polyether and closure of the latter. Apparently, these processes are accompanied by the formation of water molecules and hydrolysis of TaCl5 as follows from the presence of the oxonium ion in the complex 168 and the concurrent formation of complexes with compositions {[(H2Cl) . (15C5)2]++[TaCl6]7} (169) and {[(15C5) . . (CH3C(Cl)=NH2)]+ + [TaCl6]7} (170). The fragments in the cation of the last-mentioned compound are linked to each other through NH.. .O hydrogen bonds (Fig. 39). Ta NCOCl Figure 39. Structure of the complex 170.Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems Introduction of an additional component, namely, a tertiary amine or a quaternary ammonium salt, into the TaCl5 ± 15C5 ± a- cetonitrile system afforded the complex {[((C2H5)3NH) . (15C5)]++[TaCl6]7} (171) in which the fragments of the cation interact through NH. . .O hydrogen bonds and a complex tantalum oxychloride with composition {[Ta3O3Cl12]37+3[N(CH3)4]++(CH3CN)} (172), which con- tains no CE molecules. The structure of the cation in the complex 172 is shown in Fig. 40. Ta OCl Figure 40. Structure of the Ta3O3Cl12 fragment in the complex 172.The reaction of TaCl5 with 15C5 in acetonitrile in the presence of LiCl gave complexes in which the lithium atom is a `guest' for the CE molecule and the tantalum atom forms anions of different compositions depending on the ratio of halide molecules.116 Two products with compositions {[(TaCl5)2O]27+ 2[(Li) . (CH3CN) . . (15C5)]+} (173) (Fig. 41) and {[TaCl6]7+ [(Li) . (CH3CN) . . (15C5)]+} (174) were identified. The structures of the cations in both compounds are identical. The lithium atom has a pentago- nal-pyramidal coordination and deviates from the mean plane through the oxygen atoms of the CE molecule by 0.45 A. The Li ±Ocr distances vary from 2.08 to 2.39 A. The Li ±N distance is 2.11 A. Li Ta COCl N Figure 41.Structure of the complex 173. 4. Rhenium halides The reactions of ReOCl3 and ReOBr3 with 12C4 and 18C6 in the presence of water gave complexes with compositions {[(ReOX4) . (H2O)]7+[(H3O) . (12C4)]+}, where X=Cl (175) and Br (176), respectively. The reaction of ReOCl2 with 18C6 yielded the complex {[(ReOCl4) . (H2O)]7+[(H3O) . (18C6)]+} (177).76, 117 The structures of the anions in the compounds 175 ± 177 are analogous to that in the complex 98. The reaction of a mixture of ReCl4 and LiCl with 12C4 in acetonitrile afforded a product with composition {(ReCl5) . . (CH3CN)]7+[(Li) . (12C4)2]+ + [(CH3CN)] (178) containing the octahedral anion.117 5. Gold trichloride The reaction of a mixture of gold trichloride and potassium chloride with 15C5 yielded a compound with composition {[(K) .(15C5)2]+ + [AuCl4]7} (178) containing the tetrahedral anion [AuCl4]7. 131 6. Mercury halides The reactions of HgCl2 and HgBr2 with 18C6 gave typical complexes of type A with composition [(HgX2) . (18C6)], where X=Cl (180) and Br (181), respectively. The mercury atom is located in the cavity of the CE molecule and has a hexagonal- bipyramidal coordination. The Hg ±Ocr distances vary from 2.31 to 2.86 A.81, 118 7. Thallium halides The reaction of a mixture of TlCl and SbCl5 with 12C4 afforded a bimolecular complex with composition {[SbCl6]7 + [(Tl) . (12C4)2]+} (182).119 The thallium atom in the sandwich cation is located between two CE molecules. The Tl ±Ocr distances vary from 2.865 to 2.904 A.The reaction of an equimolar mixture of TlBr3 and RbBr with 15C5 yielded a complex with composition {[TlBr4]7+[(Rb) . (15C5)2]+} (183) with the thallium atom in the tetrahedral anion and the rubidium atom in the sandwich cation.49 The reactions of mixtures of TlX and TlX3 with 18C6 in the presence of copper or manganese dihalides gave complexes with the analogous composition {4[(TlI) . (18C6)]+ + 2[TlIIIX4]7 + [MX4]27}, where M=Cu and X=Cl (184) or Br (185); or M=Mn, X=Cl (186).49, 61, 120 In all the structures, the thal- lium(I) atom deviates from the mean plane through the oxygen atoms of the CE molecule by*0.9 A. The Tl ±Ocr distances vary from 2.93 to 3.01 A. 8. Lead halides The reaction of lead dibromide with 18C6 in a hot aqueous solution gave a complex of type A with composition {[(PbBr2) .(18C6)] + (H2O)} (187). The coordination number of the lead atom is 8 (the Pb ±Ocr distances with the six oxygen atoms of the CE molecule vary from 2.77 to 2.80 A; Pb ± Br=2.924 and 2.927 A).121 In the crystal, the complex molecules are linked to each other through OH. . . Br intermolecular hydrogen bonds (3.37 A) with the participation of water molecules. The complexes [(PbI2) . (18C6)] (188)122 and [(PbCl2) . (DCHex18C6)], where DCHex18C6 is dicyclohexyl-18-crown-6, (189) 123 have analogous structures. The reaction of a 1 : 2 mixture of PbCl2 and SbCl5 with 18C6 in acetonitrile afforded a compound with composition {[(CH3CN)3 . (Pb) .(18C6)]2+ + 2[SbCl6]7} (190) in which the lead atom is also located in the centre of the CE cavity and is additionally coordinated by three acetonitrile molecules (one molecule is located `above' the mean plane of the macrocycle and two molecules are located `below' this plane).124 9. Bismuth halides Complexes of BiCl3 with 12C4 (191) and of BiBr3 with 12C4 (192) are structurally similar to the corresponding arsenic complex (92).125, 126 A 1 : 1 mixture of BiCl3 and SbCl3 reacted with 12C4 in acetonitrile to yield a product with composition {[(Bi) . (12C4)2]3+ + 3[SbCl6]7 + [CH3CN]} (193),92 which is isostructural to the antimony complex 119. The reaction of a 1 : 1 mixture of BiCl3 and SbCl5 with 15C5 in acetonitrile afforded the complex {[(CH3CN) .(BiCl2) . (15C5)]+ + [SbCl6]7} (194) in which the [(BiCl2) . (CH3CN)] fragment is located above the mean plane through the oxygen atoms of the CE molecule.127 The analogous reaction of a 1 : 2 mixture of tri- and pentachlorides of bismuth and antimony with 18C6 gave a product with composition {[(CH3CN)2 . (BiCl) . (18C6)]2+ + 2[SbCl6]7} (195) in which the bismuth atom is coordinated by one chlorine atom, two nitrogen atoms and four (of six) oxygen atoms of the CE molecule.127 VII. Systems with actinide halides The reaction of uranium(VI) oxychloride with 12C4 in tetrahy- drofuran containing water yielded a compound with composition {[(H2O)2 . (UO2Cl2) . (12C4)] + [12C4]} (196).128, 129 The coordi- nation environment about the uranium atom is a pentagonal bipyramid formed by two oxygen atoms of the water molecules132 (U ±Ow=2.41 A), two chlorine atoms (U ±Cl=2.68 ± 2.71 A), two oxygen atoms of the uranyl fragment (U ±O=1.76 A) and one oxygen atom of theCEmolecule (U ±Ocr=2.55 A) (Fig.42). UCOCl Figure 42. Structure of the complex 196. The reaction of uranyl chloride with 12C4 in the presence of NaCl afforded the complex {[UO2Cl4]27 + 2[(Na) . (12C4)]+} (197) containing the octahedral anion.130, 131 The reaction of uranyl chloride with benzo-15-crown-5 (B15C5) proceeded anal- ogously to form a complex with composition {[UO2Cl4]27 + 2[(Na) . (B15C5)]+} (198).132 Thorium tetrachlorides reacted with various CE in solvents containing water to produce complexes of type C in which the thorium atom is not directly bound to oxygen atoms of the CE molecules.133 ± 135 ThCl4+18C6+9H2O {[(ThCl2) .(H2O)7]2++2Cl7+(18C6)+(H2O)2} , 199 ThCl4+18C6+3CH3OH+2H2O {[(ThCl4) . (CH3OH)3 . (H2O)]+(18C6)+(H2O)} , 200 ThCl4+15C5+2CH3OH+CH3CN+2H2O {[(ThCl4) . (CH3OH)2 . (H2O)2]+(15C5)+(CH3CN)}. 201 The coordination polyhedron about the thorium atom in the complexes 199, 200 and 201 is a tricapped trigonal prism, a square antiprism and a distorted dodecahedron, respectively. The reactions of UCl4 with crown ethers 15C5, DB18C6 and DCHex18C6 in air proceeded with oxidation of U4+ to U6+ (see Refs 131 and 136 ± 138). UCl4+CaCl2+15C5+H2O UCl4+2NH4Cl+4DB18C6+CH3CN UCl4+CH3OH+2DB18C6+H2O UCl4+2NH4Cl+15C5+CH3CN {[(H2O)3 .(Ca) . (15C5)]2++[UO2Cl4]27} , {[UO2Cl4]27+2[(NH4) . (DB18C6)2]++(CH3CN)} , {[UO2Cl4]27+2[(H3O) . (DB18C6)]++(CH3OH)} , {[UO2Cl4]27+2[(NH4) . (15C5)2]++2(CH3CN)} . 202 203 204 205 In the absence of air, the reaction of UCl4 with B15C5 in the presence of NH4Cl afforded the octahedral anion [UCl6]27 and the cation [(NH4) . (B15C5)]+ (206).131 The reaction of uranium tetrachloride with DCHex18C6 gave compounds with composi- tions {[UCl6]27 + 2[(H3O) . (DCHex18C6)]+} (207) 137 and {[UCl6]27 + 2[(UCl3) . (DCHex18C6)]+} (208) 139 depending on the reaction conditions. The coordination number of the uranium V K Belsky, BMBulychev atom in the cation of 208 is equal to 9. The metal atom is coordinated by three chlorine atoms and all the six oxygen atoms of the CE molecule.The reaction of a 1 : 1 mixture of UCl4 and NH4Cl with 18C6 in acetonitrile yielded a compound with composition {[UCl6]27+ 2[(NH4) . (18C6)]++(CH3CN)2} (209).140 An interesting example of a redox process that occurred in the reaction of UCl4 with 18C6 in tetrahydrofuran was reported.141 The reaction product contained the uranium(IV) and uranium(VI) atoms. H2O 3UCl4+18C6 {2[(UIVCl3) . (18C6)]++[(UVIO2Cl3) . (OH) . (H2O)]27} . 210 The uranium(VI) atom in the anion has a pentagonal-bipyr- amidal environment and the uranium(IV) atom in the cation is coordinated by all the six oxygen atoms of the CE molecule (U ±Ocr=2.48 ± 2.61 A) and three chlorine atoms. VIII.Conclusions To date, crown compounds of the majority of metal and semi- metal halides in the highest oxidation states have been prepared and structurally characterised. However, data on the structures of this class of compounds with metal halides in the lowest oxidation states are virtually unavailable. As has been mentioned in the Introduction, the structures of the great majority of crown compounds can be actually classed into three major types, viz., A, B and C, and their variations. Homonuclear molecular com- plexes of crown ethers with halides of semimetals, such as As, Sb and Bi, as well as complexes, the nature of bonding in which is still not clearly understood, should be assigned to exceptions or to the individual type D. The geometrical correspondence between the dimension of the metal atom (dm) and the dimension of the cavity (dc) of the crown ether molecule is a necessary but not nearly sufficient condition for the formation of inner-cavity crown compounds.The principle of this correspondence virtually forbids the synthesis of these compounds only if dm > dc. Moderate-sized metal atoms in the cavity are displaced toward two or three oxygen atoms or find alternative possibilities for the formation of complexes (see, for example, complexation of halides of lithium, beryllium, metals of the arsenic subgroup and lanthanides).142, 143 In these cases, the metal atoms are generally located outside the cavity of the CE molecule. It should be noted that unusual and nontypical coordi- nation numbers of metal atoms located in the cavity of the CE molecule (for example, of the aluminium atom in the complex with the crown ether 15C5) should not be regarded as a violation of the well-known structural principles because the above-mentioned deviations of the metal atoms are induced by the macrocyclic nature of the polyethers. Apparently, the possibility of ionisation and/or autocomplex- ation of a salt in a solution should be regarded as a prerequisite for the preparation of crown-ether compounds containing the metal atom in the cavity. All the above-considered inner-cavity crown compounds were prepared in solvents of the donor type.Since the processes of ionisation and autocomplexation in these solvents occur with difficulty, complexes with metal halides in the lowest oxidation states are very poorly studied.It should be noted that these metals generally form rather stable polymeric structures. Acceptors of halide ions are introduced into salt solutions in order to direct the reactions toward inner-cavity complexes. However, in these cases the reactions always afford hetero- or homonuclear different-valence ionic complexes in which the inner-cavity cation is formed by the metal ion with the lower affinity for the halogen atom (a weaker Lewis acid). In this respect, the conversion of the complexes based on SbX3 from type D to type A in the presence of SbX5 is a prominent example. The introduction of a versatile solvent, viz., water, into the reaction system prevents, as a rule, the formation of inner-cavityStructural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems crown compounds or causes their destruction, i.e., leads to the passage from type A to type B or C in which the fragments are linked to each other through a system of hydrogen bonds.We are aware of only three examples where the metal atoms remain in the cavity of the macrocycle in the presence of water, viz., the complexes prepared by the reactions of 15C5 with CoCl2 (see Ref. 42) and with MgCl2 (See Ref. 14) and by the reaction of CaCl2 with 18C6.7 The presence of water in a salt ±CE ± organic solvent system sharply changes the character of interactions between the compo- nents.In this case, the process involves either hydration of the cation accompanied by displacement of molecules of the organic solvent from the coordination sphere of this cation or hydrolysis accompanied by the appearance of hydrogen halide, oxonium ions, oxides, hydroxides and a number of other products in the solution. However, the reasons for the appearance of water in the reaction zone have not been a matter of discussion in the majority of works. Undoubtedly, the experiments performed with the use of crystal hydrates or ordinary (not dehydrated) solvents are quite understandable the more so that hydrated compounds are formed under these conditions in rather high yields. The instances where the reactions which were carried out with the use of anhydrous solvents and thoroughly dehydrated salts afforded single crystals of hydrated complexes are not so unam- biguous.Since the yields of these salts were no more than5%± 8% (see, for example, Refs 7, 14 and 41), hydration of salts may be attributed to residual traces of moisture in the initial compounds and in solvents as well as to the anomalous hydrophilicity of salts of some divalent metals. Finally, the appearance of water in reactions of crown ethers with some anhydrous metal halides in the highest oxidation states, for example, with MoCl5 75 and TaCl5,115, 116 is still unclear. The complexes prepared by these reactions never have and apparently cannot have the structures of classical intra-cavity crown com- pounds because these structures can be formed only if the charge of the cation is no less than 3+, which is observed in none of the above-described structures.However, these reactions always afforded a wide range of various compounds, including hydrolysis products in yields of up to 20%± 30%. These yields are substan- tially larger than those calculated taking into account all possible sources of residual moisture that could exist in the reaction zone. In this connection, we believe that hydrolysis can be explained only by profound decomposition of the polyether molecule accompanied by the formation of large amounts of water. In this case, residual moisture, which is introduced into the reaction zone with the initial compounds and the solvent, can act as an initiator of cleavage of the macrocycle.Taking into account the above- mentioned characteristic features of the salt behaviour, not only the results of experiments on extraction and separation of metal halides (and, apparently, compounds of other types as well) with the use of crown ethers but also the ideology of these studies are highly questionable. Reactions of solvation and complexation are the major sub- ject matter of coordination chemistry. However, these reactions are explicitly and implicitly involved in many processes which are far from this field of chemistry. In this respect, the chemistry of macrocyclic compounds (in particular its structural aspect) is undeniably helpful in developing general model concepts of interactions between components of complex systems which can be used in various divisions of chemistry. For example, products of intercalation of salts of the acceptor type (generally, metal halides, Lewis acids) into the interlayer space of graphite (the so-called intercalation graphite compounds, IGC) are known.Most of these products are compounds formed by metal halides MXn which are intercalated into graphite only in the presence of free halogen or metal halides in the highest oxidation states M*Xm (hetero-IGC). If the graphite network is considered as a polyligand, this behaviour of salts MXn is readily explained from the standpoint of formation of heterocrown compounds, i.e., on the assumption that ionisation, autocomplex- 133 ation and, finally, intercalation of ionic salts of the type {[M*Xm71]+[MXn+1]7} or {[X+[MXn+1]7} into the interlayer space occur.Yet another example is associated with the chemistry of Ziegler ± Natta catalytic systems applied onto mineral supports whose surface layers, in principle, can be considered as polyli- gands. It is known that introduction of magnesium salts, for example, of MgCl2, into catalysts AlRR0R00 ± TiXX0X00, where R=Alk, R0=H, R00=Hal; X=Hal, X0=Alk, X00=C5H5, etc., increases their activity in polymerisation of alkenes by several orders of magnitude. Aluminium and titanium salts can form both cationic and anionic components of complexes, while magnesium is involved only in cations of CE complexes. Therefore, magne- sium efficiently competes with the first two metals in formation of cations thus causing them to form anions.One would expect that, like MgCl2, salts of other metals which can undergo ionisation and are not prone to autocomplexation (for example, ScCl3), are efficient promoters of polymerisation. References 1. M Hiraoka Crown Compounds. The Characteristics and Applications (Amsterdam: Elsevier, 1982) 2. A Yu Tsivadze, A A Varnek, V E Khutorskii Koordinatsionnye Soedineniya Metallov s Kraun Ligandami (Coordination Compounds of Metals with Crown Ligands) (Moscow: Nauka, 1991) 3. N R Strel'tsova, A A Varnek, A S Glebov, V K Belsky, in Problemy Kristallokhimii (Problems of Crystal Chemistry) (Moscow: Nauka, 1989) p. 47 4. N K Dalley, in Synthetic Multidentate Macrocyclic Compounds (Ed. RMIzatt) (New York: Academic Press, 1978) p.200 5. N S Poonia, in Progress in Macrocyclic Chemistry Vol. 1 (Ed. RMIzatt) (New York: Interscience, 1979) p. 115 6. F Gingl,W Hiller, J Strahle, H Borgholte, K Dehnicke Z. Anorg. Allgem. Chem. 606 91 (1991) 7. V K Belsky, N R Strel'tsova, B M Bulychev, L V Ivakina, P A Storozhenko Zh. Neorg. Khim. 37 2676 (1992) a 8. T Chang, M Zhao, N Hu, Z Jin Rev. Chim. Miner. 24 382 (1987) 9. V K Belsky, B M Bulychev, N R Strel'tsova, L V Ivakina, P A Storozhenko Zh. Obshch. Khim. 59 1806 (1989) b 10. S G Bott, A Alvanipour, J L Atwood J. Inclus. Phen. 10 153 (1991) 11. Yu A Simonov, V O Gel'mbol'dt, E V Ganin, A A Dvorkin, M S Fonar', L V Ostapchuk, Ya Lipkovskii Koord.Khim. 21 760 12. M S Fonar', Yu A Simonov,M D Mazus, E V Ganin, 13. M Newman, E C Steiner, F P van Remoortere, F P Boer Inorg. 14. N R Strel'tsova, B M Bulychev, V K Belsky, O K Kireeva 15. N R Strel'tsova, L V Ivakina, P A Storozhenko, B M Bulychev, 16. J L Atwood, S G Bott, C M Means, A W Coleman, H Zhang, 17. N R Strel'tsova, V K Belsky, L V Ivakina, P A Storozhenko, 18. S G Bott, H Elgamal, J L Atwood J. Am. Chem. Soc. 107 1796 19. V O Gel'mbol'dt, E V Ganin, Yu A Simonov, L V Ostapchuk, (1995) c V O Gel'mbol'dt Kristallografiya 42 858 (1997) d Chem. 14 734 (1975) Zh. Obshch. Khim. 61 795 (1991) b V K Belsky Dokl. Akad. Nauk SSSR 291 1373 (1986) e M T May Inorg. Chem. 29 467 (1990) B M Bulychev Koord. Khim. 13 1101 (1987) c (1985) M S Fonar', Ya Lipkovskii, A A Dvorkin, A A Ennan 20.P P North, E C Steiner, F P van Remoortere, F P Boer Acta 21. G R Willey, M T Lakin, N W Alcock J. Chem. Soc., Chem. 22. N R Strel'tsova, V K Belsky, B M Bulychev, O K Kireeva 23. G R Willey, P R Meehan, M D Rudd, M G B Drew J. Chem. Soc., 24. G R Willey, J Palin, N W Alcock J. Chem. Soc., Dalton Trans. 1117 Koord. Khim. 21 183 (1995) c Crystallogr., Sect. B 32 370 (1976) Commun. 1619 (1992) Zh. Neorg. Khim. 37 1822 (1992) a Dalton Trans. 3173 (1995) (1992)134 25. V K Belsky, B M Bulychev, N R Strel'tsova Zh. Neorg. Khim. 37 1531 (1992) a 26. S G Bott, H Prinz, A Alvanipour, J L Atwood J. Coord. Chem. 16 303 (1987) 27. N R Strel'tsova, L V Ivakina, V K Belsky, P A Storozhenko, B M Bulychev, A I Gorbunov Zh.Obshch. Khim. 58 861 (1988) b 28. G Frenzen,W Massa, T Ernst, K Dehnicke Z. Naturforsch., B Chem. Sci. 45 1393 (1990) 29. V S Sergienko, V K Borzunov Koord. Khim. 17 1072 (1991) c 30. X Jin, Z Pan,M Shao, D Huang, Z Tai, J Zhang Kexue Tongbao(Chin.) 29 319 (1984); Chem. Abstr. 101 31 470 (1984) 31. T Ernst, K Dehnicke, H Goessmann, D Fenske Z. Naturforsch., B Chem. Sci. 45 967 (1990) 32. B B Hughes, R C Haltiwanger, C G Pierpont,M Hampton, G L Blackmer Inorg. Chem. 19 1801 (1980) 33. R Chenevert, D Chamberland,M Simard, F Brisse Can. J. Chem. 68 797 (1990) 34. X Jin, Z Pan, Y Tang, D Huang, Z Tai, J Zhang Kexue Tongbao(Chin.) 27 1044 (1982); Chem. Abstr. 97 206 031 (1983) 35. X Jin, Z Pan, Y Tang, D Huang, Z Tai, J Zhang Kexue Tongbao(Chin.) 28 1334 (1983) 36.S B Larson, S H Simonsen, J N Ramsden, J J Lagowski Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 46 1930 (1990) 37. L V Ivakina, N R Strel'tsova, V K Belsky, P A Storozhenko, B M Bulychev, A B Tarasov Zh. Obshch. Khim. 57 1600 (1987) b 38. R Chenevert, D Chamberland,M Simard, F Brisse Can. J. Chem. 67 32 (1989) 39. U Russo, G Valle, G J Long, E O Schlemper Inorg. Chem. 26 665 (1987) 40. L V Ivakina, N R Strel'tsova, V K Belsky, P A Storozhenko, B M Bulychev, A I Gorbunov Zh. Obshch. Khim. 58 349 (1988) b 41. N R Strel'tsova, A B Tarasov, V K Belsky, B M Bulychev Zh. Obshch. Khim. 60 2337 (1990) b 42. N R Strel'tsova, V K Belsky, B M Bulychev, O K Kireeva Zh. Neorg.Khim. 37 1815 (1992) a 43. O K Kireeva, B M Bulychev, N R Streltsova, V K Belsky, A G Dunin Polyhedron 11 1801 (1992) 44. N R Streltsova, V K Belsky, B M Bulychev, O K Kireeva Inorg. Chim. Acta 189 111 (1991) 45. N R Strel'tsova, L V Ivakina, P A Storozhenko, V K Belsky, B M Bulychev, A I Gorbunov Zh. Obshch. Khim. 59 40 (1989) b 46. B M Bulychev, O K Kireeva, V K Belsky, N R Streltsova Polyhedron 11 1809 (1992) 47. W Massa, T Ernst, K Dehnicke Z. Naturforsch., B Chem. Sci. 45 563 (1990) 48. T B Vance Jr, E M Holt, C G Pierpont, S L Holt Acta Crystallogr., Sect. B 36 150 (1980) 49. N S Fender, S S Finegan, D Miller, M Mitchell, I A Kahwa, F R Fronczek Inorg. Chem. 33 4002 (1994) 50. K Ruhland-Senge, U Mueller Z. Naturforsch., B Chem. Sci.45 995 (1990) 51. J Jarrin, F Dawans, F Robert, Y Jeannin Polyhedron 1 409 (1982) 52. N R Rath, E M Holt J. Chem. Soc., Chem. Commun. 665 (1985) 53. J P Jasinski, N R Rath, E M Holt Inorg. Chim. Acta 97 91 (1985) 54. F P van Remoortere, F P Boer, E C Steiner Acta Crystallogr., Sect. B 31 1420 (1975) 55. D Fenske, H Goesmann, T Ernst, K Dehnicke Z. Naturforsch., B Chem. Sci. 45 101 (1990) 56. V K Belsky, N R Streltsova, O K Kireeva, B M Bulychev, T A Sokolova Inorg. Chim. Acta 183 189 (1991) 57. N R Strel'tsova, V K Belsky, B M Bulychev, O K Kireeva Zh. Neorg. Khim. 36 2024 (1991) a 58. F Dejehet, R Debuyst, M Spirlet, J-P Declercq, M van Meerssche J. Chim. Phys. Phys-Chim. Biol. 80 819 (1983) 59. J Feneau-Dupont, J-P Declercq, G Germain,M van Meerssche, E Arte Acta Crystallogr., Sect.B 35 1215 (1979) 60. T B Rubtsova, O K Kireeva, B M Bulychev, N R Streltsova, V K Belsky, B P Tarasov Polyhedron 11 1929 (1992) 61. I A Kahwa, D Miller, M Mitchel, F R Fronczek, R G Goodrich, D J Williams, C A O'Mahoney, A M Z Slawin, S V Ley, C J Groombridge Inorg. Chem. 31 3963 (1992) 62. T Sakurai, K Kobayashi, S Tsuboyama Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 39 206 (1983) V K Belsky, BMBulychev 63. W Dreissig, Z Dauter, A Cygan, J F Biernat Inorg. Chim. Acta 96 21 (1985) 64. J F Biernat, A Cygan, E Luboch, Yu A Simonov, T I Malinovski, V K Belsky, N F Bolotina J. Inclus. Phen. 15 369 (1993) 65. L V Ivakina, V K Belsky, N R Strel'tsova, P A Storozhenko, B M Bulychev Zh. Struct.Khim. 30 (3) 161 (1989) f 66. V K Belsky, N R Streltsova, B M Bulychev, P A Storozhenko, L V Ivakina, A I Gorbunov Inorg. Chim. Acta 164 211 (1989) 67. F Dejehet, R Debuyst, J-P Declercq J. Chim. Phys. Phys-Chim. Biol. 83 85 (1986) 68. V K Belsky, L V Ivakina, N R Strel'tsova, P A Storozhenko, B M Bulychev Zh. Struct. Khim. 30 (4) 188 (1989) f 69. K M Doxsee, J R Hagadorn, T J R Weakley Inorg. Chem. 33 2600 (1994) 70. N R Strel'tsova,M G Ivanov, S D Vashchenko, V K Belsky, I I Kalinichenko Koord. Khim. 17 786 (1991) c 71. N W Alcock, M Ravindran, G R Willey Acta Crystallogr., Sect. B 49 507 (1993) 72. B Borgsen, F Weller, K Dehnicke Chem.-Ztg. 114 111 (1990) 73. H Prinz, S G Bott, J L Atwood J. Am. Chem.Soc. 108 2113 (1986) 74. E Hey-Hawkins, H G von Schnering Z. Naturforsch., B Chem. Sci. 46 307 (1991) 75. B M Bulychev, V K Belsky, N R Strel'tsova Zh. Neorg. Khim. 38 72 (1993) a 76. V S Sergienko, A B Ilyukhin, A V Luzikova, M A Porai-Koshits Koord. Khim. 17 1489 (1991) c 77. V S Sergienko, L Kh Minacheva, N Kh Ashurova,M A Porai- Koshits, K G Yakubov, V K Sakharova Koord. Khim. 17 381 (1991) c 78. V S Sergienko, L Kh Minacheva, G G Sadikov, N Kh Ashurova, V V Minin, K G Yakubov, V M Shcherbakov Koord. Khim. 18 346 (1992) c 79. G Helgesson, S Jagner J. Chem. Soc., Dalton Trans. 1069 (1993) 80. C R Paige, M F Richardson Can. J. Chem. 62 332 (1984) 81. A C Hazell, in The 14th International Congress on Crystallography (Abstracts of Reports), Netherlands, 1988 p.180 82. N R Strel'tsova,M G Ivanov, S D Vashchenko, V K Belsky, I I Kalinichenko Koord. Khim. 17 646 (1991) c 83. P C Stark,M Huff, E A Babaian, L M Barden, D C Hrneir, S G Bott, J L Atwood J. Inclus. Phen. 5 683 (1987) 84. E A Babaian, L M Barden, D C Hrneir, W E Hunter, J L Atwood J. Inclus. Phen. 5 605 (1987) 85. B Neumueller, F Gahlmann, M Schaefer, S Magull J. Organomet. Chem. 440 263 (1992) 86. M G B Drew, D G Nicholson J. Chem. Soc., Dalton Trans. 1543 (1986) 87. E Hough, D G Nicholson, A K Vasudevan J. Chem. Soc., Dalton Trans. 2335 (1986) 88. V K Belsky, B M Bulychev, N R Strel'tsova, P A Storozhenko, L V Ivakina Dokl. Akad. Nauk SSSR 303 1137 (1988) e 89. P A Cusack, B N Patel, P J Smith, D W Allen, I W Nowell J.Chem. Soc., Dalton Trans. 1239 (1984) 90. G Valle, A Cassol, U Russo Inorg. Chim. Acta 82 81 (1984) 91. M Takahashi, T Kitazawa,M Takeda J. Chem. Soc., Chem. Commun. 1779 (1993) 92. R Garbe, B Vollmer, B Neumuller, J Pebler, K Dehnicke Z. Anorg. Allgem. Chem. 619 271 (1993) 93. E Hough, D G Nicholson, A K Vasudevan J. Chem. Soc., Dalton Trans. 427 (1987) 94. M Schaefer, J Pebler,K Dehnicke Z. Anorg. Allgem. Chem. 611 149 (1992) 95. A Neuhaus, G Frenzen, J Pebler, K Dehnicke Z. Anorg. Allgem. Chem. 618 93 (1992) 96. M Schaefer, K Dehnicke Z. Anorg. Allgem. Chem. 614 21 1992 97. J Feneau-Dupont, E Arte, J-P Declercq, G Germain, M van Meerssche Acta Crystallogr., Sect. B 35 1217 (1979) 98. R D Rogers Inorg. Chim.Acta 133 175 (1987) 99. R D Rogers, A N Rollins, M M Benning Inorg. Chem. 27 3826 (1988) 100. R D Rogers, L Nunez Inorg. Chim. Acta 172 173 (1990) 101. R D Rogers J. Coord. Chem. 16 415 (1988) 102. R D Rogers, E J Voss Inorg. Chim. Acta 133 181 (1987) 103. L Nunez, R D Rogers J. Cryst. Spectr. Res. 22 265 (1992)Structural-chemical aspects of complexation in metal halide ± macrocyclic polyether systems 104. N Hu, Z Li,Q Zhou, B Yang, Z Jin, J Ni Yingyong Huaxue 4 (6) 22 (1987); Chem. Abstr. 108 141 082 (1987) 105. Z Li, J Ni,G Xu, J Ren Kexue Tongbao 31 956 (1986); Chem. Abstr. 106 9525 (1987) 106. R D Rogers, L K Kurihara Inorg. Chim. Acta 129 277 (1987) 107. R D Rogers, L K Kurihara Inorg. Chim. Acta 130 131 (1987) 108. R D Rogers, L K Kurihara Inorg. Chim. Acta 116 171 (1986) 109. R D Rogers, L K Kurihara J. Less-Common Met. 127 268 (1987) 110. R D Rogers, L K Kurihara Inorg. Chem. 26 1498 (1987) 111. R D Rogers,A N Rollins,R D Etzenhouser, E J Voss, C B Bauer Inorg. Chem. 32 3451 (1993) 112. R D Rogers, L K Kurihara, E J Voss Inorg. Chem. 26 2360 (1987) 113. R D Rogers Inorg. Chim. Acta 133 347 (1987) 114. R D Rogers, R D Etzenhouser Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 44 1533 (1988) 115. B M Bulychev, V K Belsky Zh. Neorg. Khim. 40 1834 (1995) a 116. B M Bulychev, V K Belsky Zh. Neorg. Khim. 42 260 (1997) a 117. H Stenger, F Weller, K Dehnicke Z. Anorg. Allgem. Chem. 606 109 (1991) 118. D A Pears, S J Fraser, J Crosby, B L Allwood, D J Williams Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 42 51 (1986) 119. H von Arnim, K Dehnicke, K Maczek, D Fenske Z. Naturforsch., B Chem. Sci. 48 1331 (1993) 120. I A Kahwa, D Miller, M Mitchell, F R Fronczek Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 49 320 (1993) 121. Yu A Simonov, V Kh Kravtsov, M S Fonar', A Yu Nazarenko Zh. Neorg. Khim. 41 388 (1996) a 122. A Yu Nazarenko, O I Kronikovski,M S Fonari, Yu A Simonov Supramol. Chem. 3 777 (1994) 123. A Yu Nazarenko, E B Rusanov Pol. J. Chem. 68 1233 (1994) 124. H von Arnim, K Dehnicke, K Maczek, D Fenske Z. Anorg. Allgem. Chem. 619 1704 (1993) 125. N W Alcock, M Ravindran, G R Willey J. Chem. Soc., Chem. Commun. 1063 (1989) 126. R D Rogers, A H Bond, S Aguinaga, A Reyes J. Am. Chem. Soc. 114 2967 (1992) 127. M Schaefer, G Frenzen, B Neumuller, K Dehnicke Angew. Chem. 104 354 (1992) 128. R D Rogers,M M Benning, R D Etzenhouser, A N Rollins J. Chem. Soc., Chem. Commun. 1586 (1989) 129. R D Rogers,M M Benning, R D Etzenhouser, A N Rollins J. Coord. Chem. 26 299 (1992) 130. R D Rogers Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 44 638 (1988) 131. R D Rogers, L K Kurihara, M M Benning Inorg. Chem. 26 4346 (1987) 132. D C Moody, R R Ryan Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 8 933 (1979) 133. R D Rogers Lanthanide Actinide Res. 3 71 (1989) 134. R D Rogers,M M Benning Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 44 641 (1988) 135. R D Rogers, L K Kurihara, M M Benning J. Chem. Soc., Dalton Trans. 13 (1988) 136. M M Benning, L K Kurihara, R D Rogers J. Less-Common Met. 127 269 (1987) 137. P Zheng,M Wang, B Wang, J Wu,W Wang, J Lin Acta Chim. Sin. 46 837 (1988); Ref. Zh. Khim. 6 B 2086 (1989) 138. R D Rogers, A H Bond, W G Hipple J. Crystallogr. Spectrosc. Res. 20 611 (1990) 139. G C de Villardi, P Charpin, R-M Costes, G Folcher, P Plurien, P Rigny, C De Rango J. Chem. Soc., Chem. Commun. 90 (1978) 140. R D Rogers,M M Benning Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 44 1397 (1988) 141. G Bombieri, G De Paoli J. Inorg. Nucl. Chem. 40 1889 (1978) 142. Z Mao, Z Zhou, Z Hong, H Yang,M Zhang, B Ruan Jiegou Huaxue 12 266 (1993); Chem. Abstr. 120 31 230 (1994) 143. R D Rogers, A N Rollins, R F Henry, J S Murdoch, R D Etzenhouser, S E Huggins, L Nunez Inorg. Chem. 30 4946 (1991) 135 a�Russ. J. Inorg. Chem. (Engl. Transl.) b�Russ. J. Gen. Chem. (Engl. Transl.) c�Russ. J. Coord. Chem. (Engl. Transl.) d�Russ. Crystallogr. (Engl. Transl.) e�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) f�J. Struct. Chem. (En

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (2) 137 ± 148 (1999) The chemistry of furazans fused to six- and seven-membered heterocycles with one heteroatom A B Sheremetev Contents I. Introduction II. Furazan ring as a fragment of a bicyclic system III. Furazanopyridines IV. Furazanopyrans V. Furazanothiopyrans VI. Furazanoazepines VII. Furazanothiepines VIII. Conclusion Abstract. The data on the synthesis and properties of furazan derivatives fused with pyridine, pyran, thiopyran, azepine and thiepine rings are surveyed and described systematically. The bibliography includes 85 references. I. Introduction Furazan (1,2,5-oxadiazole) derivatives are heterocyclic com- pounds of a non-natural origin. The first representatives of furazans were synthesised more than a hundred years ago.The subsequent development of the chemistry of this class of com- pound has been stimulated both by the diversity of their chemical transformations and by a number of their interesting properties, which provide the possibility of using them in medicine, agricul- ture and some fields of engineering. In this decade alone, the peculiar behaviour of furazans and their N-oxides (furoxans) and the large body of information, which permanently extends as a result of studies of numerous groups of researchers, have led to the publication of several reviews 1 ±10 and monographs.11, 12 Fused systems consisting of two or more heterocycles, of which at least one heterocycle is a furazan ring, deserve close attention both from the theoretical viewpoint and due to the possible application of these compounds as intermediates in synthesis.Some derivatives of this class have been proposed as herbicides, rubber modifiers, liquid crystals, luminescent dyes and explosives. Systems in which a furazan ring is fused to a five-membered heterocycle have been considered in a review.6 The present review is devoted to the least studied derivatives, furazans fused to six- and seven-membered heterocycles with one heteroatom. A B Sheremetev N D Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky prosp. 47, 117913 Moscow, Russian Federation. Fax (7-095) 135 53 28. Tel. (7-095) 938 36 51. E-mail: sab@cacr.ioc.ac.ru Received 23 September 1998 Uspekhi Khimii 68 (2) 154 ± 166 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.793.2 137 137 138 145 146 146 147 147 II.Furazan ring as a fragment of a bicyclic system Furazan can be regarded as a hetero-analogue of cyclopentadiene. On passing from electron-releasing cyclopentadiene to electron- withdrawing furazan, chemical properties are virtually reversed. One can say that furazan is a sort of antipode of cyclopentadiene. Successive replacement of carbon atoms by heteroatoms results in redistribution of electron density and in enhancement of p-acceptor properties of the system in the following series: cyclopentadiene, furan, isoxazole, furazan. In all of these rings, electronic effects of substituents are transferred solely through the (hetero)diene system of bonds.13 The CH2 group (in cyclopentadiene) and oxygen atoms (in the other three rings) act as a sort of insulator for this transfer.As a consequence, the degrees of aromaticity of these four cyclic systems are approximately equal. N N N Compound O O O Aromaticity index (I) 14 43 47 43 45 It should be noted that the degree of bond order fixation in the heterodiene system of furazan is extremely high; derivatives of this heterocycles are not susceptible to annular-group tautomerism. The chain consisting of three heteroatoms, present in the furazan molecule, accounts for the substantial electron-withdraw- ing properties of this heterocycle.15 ± 17 The Taft induction con- stants s* of 4-R-furazanyl groups (as substituents) fall in the range 2.55 ± 2.88 (these values correspond to R=NH2 and R=NO2, respectively).18 For comparison, the s* value of the 3-furyl group is 0.62.The well-known electron-withdrawing picryl group, 2,4,6-(NO2)3C6H2 has a s* value of only 1.62. The following substituents are characterised by s* constants close to that of furazan: CH=C(CN)2 (s* = 2.56), CF3 (s* = 2.6); CBr(NO2)2 (s* = 2.8); 5-tetrazolyl group (s* = 2.82); nitrile group (s* = 3.25).19 The reactivity of annelated furazan derivatives is also signifi- cantly affected by the nature of the second ring � its size, the number of heteroatoms and their type and positions. A typical feature of annelated furazans is bond order fixation over the whole contour of the bicyclic system.It is known that the degree of delocalisation of p electrons in benzoannelated heterocycles can be estimated using the Gunther method 20 based on the ratio of spin-spin coupling constants, Jbc/Jab. When the bonds are fully138 equivalent, as in benzene, this ratio is equal to unity, whereas completely localised bonds correspond to a ratio of 0.5. According to 1H NMR spectroscopy, the Jbc/Jab ratio for benzofurazan is 0.706; in other words the bonds are substantially localised.21 Ha Hb NO N Hc The replacement of one =CH± unit in the benzene molecule by an isoelectronic fragment, =N±, is known to result in a less uniform distribution of electron density and more pronounced localisation.14 In fact, on passing from benzofurazan to fura- zano[3,4-b]pyridine, the spin-spin coupling constant for protons a, b changes only slightly (J=9.34 Hz for benzofurazan 22 and J=9.00 Hz for furazanopyridine 23), whereas the constant for protons b, c changes more significantly (J=6.50 and 3.70 Hz, respectively).Evidently, this points to a decrease in the order of the b ± c bond and, hence, to an increase in the degree of local- isation upon the replacement of the =CH± fragment by =N±. On passing from benzene to pyridine and then from benzofur- azan 22 to furazanopyridine,23 the 1H NMR signal shifts down- field, which is also due to a decrease in the electron density in this series. However, it should be mentioned that the influence of the furazan ring is still weaker than that of two ortho-nitro groups.H H H H H N N NO2 Com- pound O O N N N N N NO2 8.92 8.25 7.96 7.46 7.27 d, ppm The N-oxide fragment acts as a resonance electron-releasing group of medium strength;14 therefore, on passing to the N-oxide, furoxanopyridine, the electron-withdrawing influence of the fur- azan ring becomes weaker. In particular, the signals of the protons at C(7) in the 1H NMR spectra of both isomers of furoxanopyr- idine are shifted upfield in relation to corresponding signal in the spectrum of furazanopyridine.23 O 8.11 7.87 8.25 N N N 7.46 7.30 7.40 O O O 8.68 8.88 8.97 N N N N N N O Theoretical calculations 24 and polarographic studies 25 ± 29 have also been used to describe the influence of the furazan ring on the properties of molecules incorporating it.Acid ± base properties of some annelated furazan derivatives have also been studied.30 III. Furazanopyridines Derivatives of all three possible types of bicyclic systems based on furazan and pyridine rings � furazano[3,4-b]pyridine (A), fur- azano[3,4-c]pyridine (B) and furazano[3,4-a]pyridine (C) � have been described in the literature. 4 4 4 3 3 N 5 5 5 N 3N N + N2 O2 O2 6 6 6 N O N 8 N 7 1 1 7 1 7 B C A 1. Furazano[3,4-b]pyridine The heterocyclic system of furazano[3,4-b]pyridine (A) is also referred to as 1,2,5-oxadiazolo[3,4-b]pyridine and the correspond- A B Sheremetev ing 1(3)-oxides are also called furoxano[3,4-b]pyridines and pyr- idofuroxans.Derivatives of the bicyclic system A can be obtained from both furazan and pyridine synthons. Thus the reaction of 3-amino- furazan-4-carbonitrile 1 with b-dicarbonyl compounds in the presence of a catalytic amount of nickel acetylacetonate gives rise to labile enamines 2a ± e. On heating with acetic acid in ethanol, they undergo intramolecular condensation to give the corresponding amino-derivatives 3a ± e. The overall yields of the compounds 3a ± e are 80%± 95%.31, 32 O R2 R1 R1 H2N CN H2N O O EtOH, AcOH H2N R2 Ni(acac)2 N N O N N O1 O 2a ± e O NH2 N R1 O N R2 N 3a ± e R1=R2=Me (a); R1=Me, R2=Ph (b); R1=R2=; R1=OEt, R2=Me (d); R1=OEt, R2=Ph (e). Even in the case where R1 = R2, as in the compound 2b, the reaction occurs regioselectively giving only one product 3b in which the more bulky substituent is attached directly to the pyridine ring.The reaction of the nitrile 1 with dimedone in the presence of nickel acetate affords tricyclic compound 4 in 67% yield. In this case, the intermediate enamine was not isolated.32 O O NH2 Ni(OAc)2 N 1+ O N O N 4 Dehydration of 1,2-dioximes (glyoximes) is used most widely for the synthesis of furazans.1, 6 An unusual version of dehydra- tion has been employed in order to prepare furazan 5. 2,3- Dihydroxyiminopyridine 6 adsorbed on silica gel undergoes spontaneous dehydration over a period of 12 h to give the compound 5 in 14% yield.33 NOH NO SiO2 N NOH N N N N O O 5 6 Readily available furoxanopyridine 7a can be reduced to the corresponding furazan 8 by a short-term heating to 100 8C in trimethyl phosphite.23 Benzo-annelated derivative 9 can be obtained in a similar way from the furoxan 10.34 O N N P(OMe)3 O O N N N N 7a 8 (40%) O N N ON ON P(OMe)3 N N Me 9 (62%) 10 MeThe chemistry of furazans fused to six- and seven-membered heterocycles with one heteroatom The above furazanopyridines have been synthesised only recently.Meanwhile, the 1(3)-oxides of furazanopyridines of type A (furoxans) were first prepared more than 40 years ago.35 The synthesis of these compounds is based on the ability of o-azido nitro aromatic compounds to undergo thermally induced intramolecular cyclocondensation.The mechanism of this trans- formation has been studied comprehensively by measuring the kinetics of this reaction for benzene derivatives.11 It is believed that an oxygen atom of the nitro group attacks a nitrogen atom of the azide group. Cyclisation can be represented as electron trans- fer occurring synchronously with elimination of a nitrogen molecule. Thus heating of tetrazolopyridine 11a, which is formed upon treatment of chloropyridine 12a with sodium azide, in dichlor- obenzene leads to elimination of a nitrogen molecule from the intermediate azide 13a and closure of the furoxan ring yielding the bicyclic compound 7a.35 Later it has been shown that the compound 7a is formed in approximately the same yield when thermolysis is carried out in some other solvents or without a solvent (Table 1).NO2 NO2 NO2 NaN3 N D Cl N N N N3 13a 12a NN 11a O O N N D O O 13a 7N2 N N N 7N +N 7a N Table 1. Synthesis of furoxano[3,4-b]pyridines 7a ± i by thermolysis of the corresponding tetrazolopyridines 11a ± i. O R2 R2 NO2 NO N R1 R1 N N7a ± i N N N 11a ± iR2 Time/ Yield Ref. (%) Sol- vent h Tem- pera- ture/ 8C Com- R1 po- und 11a 35 0.03 97 170 none H H 36 37 38 75 96 0.66 7 175 110 130 PhOPh PhMe none 39 40 dioxane none 97 0.03 93 11b Me H D155 COOH 11c 11d 11e HHH NO2 11f H N3 440.03 86 0.03 97 7 0.03 792 88 85 75 40 40 37 38 37 37 37 37 7 161 106 130 95 95 82 80 110 AcOH none COOMe """PhMe MeCN C6H6 PhMe NO2 NPPh3 H NHPic a 118 11g 11h 41 11i 42, 43 95 1.5 80 NH2 C6H6 NO2 139 This reaction is general.The presence of any substituents, especially electron-withdrawing ones, in the initial nitrotetrazolo- pyridines 11a ± i facilitates their transformation into tautomeric 2-azido-3-nitropyridines and is thus favourable for the formation of the furoxan ring upon thermolysis (see Table 1). O It should be noted that furoxanopyridine 7f exists as the more stable tricyclic tautomer 14.37 O N N O O N N N N3 NN N N7f 14 A particular modification of this method is the reaction of 2-chloro-3,5-dinitropyridine with sodium azide in the presence of phase transfer catalysts in dichloroethane followed by keeping the reaction mixture at 60 8C.The furoxanopyridine 7e is prepared in this way in 92% yield without isolation of the intermediate azide or tetrazolopyridine.44 The thermolysis of azide 15, isomeric to azide 13a, gives the furoxanopyridine 7a, which is thermodynamically more stable than isomer 16.45, 46 It is noteworthy that for the azide 15, azido ± tetrazole tautomerism is impossible, and in this case, the solvent determines not only the yield of the bicyclic compound 7a but also the reaction pathway itself. A similar situation is observed in the transformation of o-azidonitrobenzenes into benzofuroxans.11 Thus at 87 8C in decalin, the yield of the compound 7a is 94%, while in ethanol, it is equal to 80%.Thermolysis of the azide 15 in o-dichlorobenzene or acetic anhydride does not give a furoxan derivative but yields instead azo compound 17.46 An attempt to synthesise isoquinolinofuroxan 18 by thermol- ysis of azide 19 failed.47 O N N O O N N N N N3 7a O 16NO2 NO2 N N 15 N N N 17 NO2 O N3 N ON NO2 D N N 18 19 Yet another method for the synthesis of furoxano-[3,4-b]pyr- idines consists in oxidative condensation of 2(3)-amino-3(2)- nitropyridines. The most widely used oxidant which permits successful preparation of this bicyclic system is diacetoxy-l3- iodanylbenzene.34, 36, 39, 48, 49 In fact, when amines 20a ± e or isomeric amines 21a ± e, or bicyclic amines 22 and 23 are merely stirred with this reagent in acetic acid, acetone or benzene at room temperature or at 50 ± 60 8C, the corresponding furoxans 7 or 10a,b are formed in 70%± 90% yields.The use of sodium hypochlorite as the oxidant in the synthesis of furoxano[3,4-b]pyridine derivatives usually does not lead to the desired products, unlike its use in the synthesis of benzofurox- ans 11 from the corresponding o-amino nitro compounds.36 Only one example of successful employment of this oxidant in the reaction in question is known. On treatment with sodium hypo- chlorite in a methanolic solution of KOH, the amine 23 undergoes140 R3 R2 NO2 gives rise to the corresponding acids 29 (yield 72%) and 30 (yield 65%).The acid 30 was converted into amide 31.34 (O)n (O)n R3 O R1 NH2 N N R2 N N 20a ± e ON PhI(OAc)2 SeO2 O R3 N N R1 N NH2 R2 7a,b, j ± l 9, 10 27, 28 CHO Me R1 NO2 (O)n N 21a ± e N ON a, b, c R1=R2=R3=H(a); R1=R3=H, R2=Me (b); R1=R2=H, R3=Me (20c, 21c, 7j); R1=Me, R2=R3=H(20d, 21d, 7k); R1=R3=Me, R2=H(20e, 21e, 7l). N O O 31 29, 30 COOH N NO2 ON NH2 PhI(OAc)2 (a) Et3N, BuOCOCl; (b) H2N(CH2)2NMe2; (c) HClO4; n=0 (9, 27, 29), n=1 (10, 28, 30). N N 10a,b R 22, 23 R R=Me (22, 10a), OC6H4Cl-p (23, 10b). Interestingly, when sodium chlorite is employed as the oxidant in this reaction, the acid is not formed but instead the aldehyde group is eliminated and the C(5) atom in the ring is oxidised.The corresponding isoquinolinones 32 and 33 are produced in 75%± 79% yields. The compound 33 (yield 54%) can also be obtained by alkaline hydrolysis of the methoxy-derivative 24. oxidative condensation with simultaneous replacement of the p-chlorophenoxy group by a methoxy group to give tricyclic compound 24 in 64% yield.34 O (O)n N O N ON N NaClO2, NaH2PO4 NaOCl 23 MeOH N N OMe 24 27, 28 CHO O N ON KOH 33 N 24 OMe n=0 (27, 32), 1 (28, 33). Furazano[3,4-b]pyridines enter into two types of reactions. The first type comprises transformations that do not affect the bicyclic cage of the initial molecule. Reactions in which this cage is destroyed belong to the second type. The first type of reactions includes transformations of the functional groups present in the pyridine moiety of the molecule.Thus refluxing of the tetrazole 14 in ethanol in the presence of triphenylphosphine results in the formation of the iminophosphorane 11g in 90% yield. Hydrolysis of this product occurs quantitatively and affords amine 25, which is readily acylated with acetic anhydride to give amide 26.37 O O N N PPh3 O O N N N N3 Treatment of the N-oxide 23 with excess (N,N-dimethyl- amino)ethylamine at 75 8C leads to substitution of the p-chlor- ophenoxy group. The resulting amine 34 was isolated as the corresponding perchlorate in 42% yield.34 O O N11f NN N 14 O N O O N 2. HClO4 1. H2N(CH2)2NMe2 N N AcOH Ac2O N O O HCl N N N H2N Ph3P 23 OC6H4Cl-p N25 N 11g O NO N AcHN N26 The second type of reaction typical of furazano[3,4-b]pyridine derivatives is transformation of the furoxan ring.For example, when the bicyclic derivatives 7a,e are heated in dimethyl sulfoxide, retro-cleavage of the furoxan ring at the N±O bond occurs giving rise to a nitro group and a nitrene. The nitrene reacts with the solvent yielding sulfimine S-oxides 35a,e.38 Irradiation of the N- oxide 7a with a UV lamp in aqueous acetonitrile for 30 h affords nitropyridone 36 in 18% yield.33 The furoxan 7a is not oxidised with m-chloroperbenzoic acid. However, treatment of 7a with concentrated hydrogen peroxide in oleum results in oxidation of the furoxan ring giving 2,3-dinitropyridine 37 in 60% yield.No products resulting from the oxidation of the pyridine nitrogen Refluxing of the tricyclic compounds 9 (n=0) or 10 (n=1) with selenium dioxide in dioxane is accompanied by oxidation of the methyl group. This gives aldehydes 27 and 28 in 56% and 75% yields, respectively.34 Further oxidation of these products with bromine in a solution of sodium bicarbonate in aqueous butanol A B Sheremetev ON NaHCO3, Br2 NO N ON N NMe2 . HClO4 NH (O)n N ON NH 32, 33 O N ON N 34 NH(CH2)2NMe2 . HClO4The chemistry of furazans fused to six- and seven-membered heterocycles with one heteroatom atom were detected. An additional nitro group present in the initial furoxan, for example in 7e, makes the furoxan ring resistant to oxidation.37 O R NO N N7a,e R=H(a), NO2 (e).hn MeCN, H2O H2O2, SO3 7a Na2S Diamine 38 has been obtained in 17% yield on treatment of the furoxan 7a with sodium sulfide. It was suggested that the furazan 8a could also arise in this reaction; however, it was not isolated in this study.35 On treatment with sodium borohydride at 50 8C in methanol, the furoxanisoquinoxaline 10 is reduced to give glyoxime 39 in a low yield.34 The reduction of a furoxan ring giving two oxime groups occurs more efficiently under the action of hydroxylamine. Thus the reaction of the N-oxide 7b with hydroxylamine affords glyoxime 40 in 54% yield. It should be noted that in addition to the reduction of the furoxan moiety, this reaction involves amination of the pyridine ring.40 O N ON NaBH4 N 10 Me O Me NO N N7b Photolysis of the compound 7a (l=254 nm) carried out for 15 h in dichloromethane in the presence of morpholine under argon at room temperature gives a series of morpholine-contain- ing products such as the glyoxime 6 (5%), the furazan 5 (9%), monomorpholylpyridine 41a (0.3%) and bis(morpholyl)pyridine 42 (0.3%).33O hn NO MorH Mor N N 7a NH2 + N Mor NO2 41a Mor is morpholin-1-yl.R O2N DMSO Me N N S Me O 35a,e NO2 O NH 36 NO2 N NO2 37 NH2 N NH2 38 NOH NOH N 39 Me NOH Me NH2OH NOH N H2N 40 NOH NO+ + N Mor NOH N5 N 6 Mor NH2 + Mor NO2 N42 141 However, the compound 41 as well as a large number of its analogues can be prepared in a higher yield by conventional nucleophilic substitution.Treatment of the compound 7a with N- and O-nucleophiles in water or ethanol at 18 ± 25 8C induces opening of the furoxan ring accompanied by nucleophilic sub- stitution of the hydrogen atom at C(6) in the pyridine ring. This reaction has been used to prepare a group of diverse 3-amino-2-nitropyridines 41 and 43. The yields of the products are 70%± 80%.45, 50 NH2 R1R2NH O N R1R2N NO2 41 N NH2 O RONa N N 7a R3O NO2 N 43 R1±R2=(CH2)2O(CH2)2, (CH2)2NH(CH2)2, (CH2)4; R1=R2=H, Me, Et; R1=H: R2=Me, Et, Bu; R3=H, Me, Et. It is noteworthy that aminonitropyridines such as 41 and 43 as well as some other products of transformation of furoxanopyr- idines are potential precursors of polyfunctional derivatives of the same bicyclic system (which can be synthesised, for example, by oxidation). Unfortunately, this line of research is not represented in original publications.The reaction of furoxanopyridines and amines in the presence of carbonyl compounds having an a-methylene unit follows a totally different pathway. In this case, the so-called Beirut reaction occurs,11 in which enamine formed in the reaction medium functions as the active nucleophile, attacking the furoxan ring rather than the C(6) atom of the pyridine ring. The amine merely catalyses this transformation and is not incorporated in the reaction products, which are azaquinoxaline bis-N,N0-oxides 44.39, 51, 52 The mechanism proposed for this process is presented below.R3 R3 O R3 N R2 + NH R2 R1 R3 O R1 O N N R32 NCR1=CHR2 O N+ N 7 N N O 7a O7R1 R3 N N + R3 + R2 N N O7 O O R1 R3 R1 N N N R3 7HNR32H N R2 N N N R2 O 44 O Me Me Pri HH Me CO2Me CH(OMe)2 H HMe Me CH2CO2Me CH(OMe)2 Me H CO2Me Me Me R1 R2 R1 R2 CONMe2 Unlike benzofuroxans, the furoxanopyridine 7a gives rise to the Beirut reaction products in low yields (20% ± 30%). The standard procedure is very simple and consists in stirring the furoxan 7a with excess carbonyl compound in ethanol in the presence of an amine (ammonia, morpholine, pyrrolidine, etc.) or sodium acetate.The resulting azaquinoxaline bis-N,N0-oxides 44 are of interest for pharmacology.142 Detailed analysis of the 1H and 13C NMR spectra has provided full assignment of all the 13C and 1H signals for furazano[3,4-b]pyridine derivatives.23, 31, 37 NMR spectroscopy is especially useful for the investigation of isomeric N-oxides of this bicyclic system.23, 49 The acid ± base properties of the compound 7h have been estimated based on the data of potentiometric titration in various solvents.41 X-Ray diffraction analysis of the bicyclic furazan 3a 31 and the tricyclic furoxan 14 37 has shown that the bond lengths and bond angles in these compounds are close to the standard values. Kinetic studies led to the assumption that the initial events in the thermal decomposition of the compounds 7e and 7i and benzofuroxan and pyridine N-oxide derivatives are similar. These reactions were assumed to involve the formation of three-membered activated intermediates of the oxaziridine type.53 ± 55 The furoxan 7a has been patented as a modifier (vulcaniser) for unsaturated polymers.56 ± 59 Nitro-derivatives of this bicyclic system are thermally stable explosives.42, 53, 54 Biological studies have shown that some furazanoquinoxalines exert moderate inhibitory effect on c-AMP-dependent protein kinase.60 2.Furazano[3,4-c]pyridines Derivatives of furazano[3,4-c]pyridine (B) are also referred to as 1,2,5-oxadiazolo[3,4-c]pyridines, and the corresponding N-oxides are also called furoxano[3,4-c]pyridines.The methods used to synthesise this class of compounds are more diverse, while the procedures employed to prepare the N-oxides largely resemble those outlined above. The first representative of furazano[3,4-c]pyridines � cage furazan 47 � was prepared 100 years ago by treatment of a,a 0- dihydroxyiminotropinone hydrochloride 45 with an alkaline solution of hydroxylamine. Obviously, the reaction starts with oximation of the free carbonyl group, which is followed by dehydration of two neighbouring oxime groups in the trioxime 46 thus formed. Compound 47 was isolated in 90% yield.61 NOH NOH NH2OH NMe NOH NMe O 7H2O NOH NOH 46 45 N O N NMe 47 NOH A series of fused furazan derivatives 48 have been obtained in high yields from 2,2,6,6-tetrasubstituted 62 and 2,6-disubsti- tuted 63 piperazin-4-ones 49.O N O N NOH NOH HON NH2OH R1 R3 R1 R3 N N R1 R2 R1 R2 R4 R4 48a ± k 49a ± k R4 R3 R2 Compounds 48 and 49 R1 Me Me Me Me Me Me t HOH H Me Me Me H Me Me H H Me Me Bu H Ph Ph Me H Ph Me Me OMe abcdefgh Me Me H i Ph H H S A B Sheremetev Compounds 48 and 49 R1 R4 R3 R2 t jk H H H Bu H H H H When asymmetrical initial compounds are used, for example 49g,i, mixtures of isomers are produced. Nitrosation of quinolizinium salt 50 with sodium nitrite in 2 M hydrochloric acid affords o-amino nitroso compound 51, which was reported 64 to be converted into tricyclic derivative 52 during recrystallisation from ethanol.H H NH2 O N HCl, NaNO2 D N + + N N 7Cl 50 51 HN ON N52 Furazano[3,4-c]pyridine derivatives have also been synthes- ised from furazan precursors. Thus condensation of 3,4-diben- zoylfurazan 53 with benzylamine in boiling toluene catalysed by diazabicycloundecene (DBU) results in the formation of com- pound 54a in 13% yield.65 However, condensation of dibenzoyl- furazan with salts of amines having electron-withdrawing substituents at the a-carbon atom is morficient and gives derivatives 54b ± e in higher yields.66 Ph Ph N PhCH2NH2 O N DBU, D N O O Ph Ph 54a Ph Ph N N + R 7 N RCH2NH3X O53 O N N 54b ± e Ph Yield (%) X R Compound 54b 54c 54d 54e Cl Cl HSO4 Cl CO2Et CO2Me CN COPh 41 45 73 44 The reaction of trioxime 55a with potassium ferricyanide occurs as oxidative cyclocondensation of two neighbouring oxime groups yielding furoxan 56a.A similar treatment of the trioxime 55b is accompanied by oxidation of the hydroxylamine function and gives rise to the furoxan nitroxide 57 stable under ambient conditions.62 N ON O HON R=H Me NOH Me Me Me NOH HON NHK3Fe(CN)6 56a Me Me Me N N Me ON O HON R=OH R 55a,b Me Me Me N Me57 O R=H(a), OH (b).The chemistry of furazans fused to six- and seven-membered heterocycles with one heteroatom Oxidation of an alkaline solution of cyclic glyoxime 58 with sodium hypochlorite involves closure of the furoxan ring and gives lactam 59 in 87% yield.67 On heating in toluene, this product partly rearranges to the isomeric 3-oxide 60.The rearrangement gives a mixture of the isomers 59 and 60 in 1 : 4 ratio. O NOH NCH2Ph 58 Benzoannelated analogues of the lactams 59 and 60, for example 61, have been synthesised by thermolysis of 4-azido-3- nitroquinolones 62 in boiling bromobenzene. The compounds 61 exist as unseparable mixtures of 3-oxides (61a) and 1-oxides (61b); this is clearly manifested in the 1H NMR spectra.68 R1 Thermolysis of 4-azido-3-nitropyridine 63 in boiling ethanol or decalin at 80 8C gives furoxan 64 in 74% and 95% yields, respectively.46 It is worth noting that the earlier attempts to perform this reaction in boiling toluene have failed.69 N3 N63 NH2 N65 N3 NO 66 NO2 NO 67 When amine 65 is oxidised with diacetoxy-l3-iodanylbenzene, only traces of the N-oxide 64 are detected; when sodium hypo- chlorite is used as the oxidant, the product 64 is not formed at all.36 143 The attempts to annelate a furoxan ring to pyridine N-oxides have been unsuccessful.Thus thermolysis of N-oxides 66 and 67 gave only high-melting compounds, the structures of which were not identified.69 N N O ON ON NOH D NaOCl O N O N O An additional nitro group present in the initial azidonitropyr- idine facilitates the formation of the furoxan ring. In this case, an unusual solvent effect is manifested. Thermolysis of compound 68 in decalin gives furoxan 69 in 97% yield.46 When the process is carried out in benzene, the yield becomes 45% to 70%, the reaction product being produced as a complex with the solvent.The same reaction performed in acetic acid gives rise to hydrated compound 70.69 CH2Ph 60 N ON CH2Ph 59 O O2N D decalin N3 O2N NO2 O O2N D N68 N3 N MeCO2H ON O NO2 N69 N ONOH H N 156 8C 70 H 7N2 O N N O Being an analogue of dinitrobenzofuroxan,11 the compound R1 R2 61a R2 62 O 69 might also exhibit superelectrophilic properties. However, no studies considering its reactivity from this viewpoint have been reported to date. N Yield of 61 (%) R2 R1 On heating in decalin, 4-azido-3-nitroquinoline 71 is smoothly ON N O converted into furoxan 72 in 94% yield.46, 47 The same compound is formed as a minor product, together with 3-nitroquinoline, when quinoline is treated with nitrogen oxides.70 H Me 94 H Ph 9292 (CH2)3 R1 N3 R2 61b NO2 D N ON O N 71 N NxOy 72 N NO2 D N ON O N NO2 PhI(OAc)2 64 NO2 N Since the electron-withdrawing furazan ring in furazano[3,4-c]- pyridines is separated from the pyridine nitrogen, the N atom exhibits basic properties; hence, these compounds can form salts with acids.69 Compounds in which the pyridine ring is hydro- genated show the highest basicity.61 Conversely, the nitro deriv- ative 69 is an acid; it forms salts with bases, hydrazine and transition metal cations and charge transfer complexes with naphthalene derivatives and pyrene (no data about the structure of complexes formed by 69 are presented in the study cited).69 The oximes 47 and 48 can be acylated on treatment with acyl chlorides in the presence of bases according to a conventional proce- dure.61, 63 Deoximation of the nitroxide 57 is accompanied by simultaneous migration of the N-oxide oxygen atom in the furoxan ring to give ketone 73.62 ON O O D N N N O HON O N3 Me Me O ONMe Me ONMe Me Me Me N N 57 O 73 O Oxidation of hydroxylamino oxime 48b is accompanied by deoximation with generation of a radical site.62 The furazan nitroxide 74 thus formed is stable under ambient conditions.144 N N ON HON O Me Me Me ONMe Me Me Me N N Me48b OH 74 O Treatment of the oximes 48c ± i with sodium hypochlorite in alkaline solutions affords nitro derivatives of dihydrogenated bicyclic compounds 75c ± i.However, the yields of these products are normally relatively low (12% ± 43%). The oximes 48c ± i are chlorinated with nitrosyl chloride in organic solvents to give chloro nitroso compounds, which have been isolated as salts 76c ± h in 60%± 70% yields. The reaction of the compound 48i with nitrosyl chloride affords the chloro nitroso derivative 76i, its yield being only 7%, due to the side reactions caused by the presence of the thienyl substituent.63 N ON O2N NaOCl R2 R1 N N ON HON 75c ± i R3 N R2 R1 N ON ON HCl . HCl Cl NOCl R3 48c ± i R2 R1 N 76c ± i R3 Nitration of the bicyclic compounds 48k (R1=R2=R3=H) and 48j (R1=R2=H,R3=But) withN2O4, N2O5, BF4NO2 or a mixture of HNO3 with sulfuric or trifluoroacetic acid either gives trinitro derivative 77 or involves aromatisation giving rise to mono (78a) or di-nitro compound (78b).The reaction can be directed at the predominant formation of either of the products, their yields reaching 80%.63 N ON O2N O2N N N ON HON 77 NO2 N N ON O2N 48k, j R1 R2 N 78a,b R1=H(48k), But (48j); R2=H(78a), NO2 (78b). The reduction of the compound 54b with sodium borohydride involves three reaction sites � the ester group (to give hydrox- ymethyl derivative 79), the pyridine ring (with the formation of dihydro derivative 80) and the N±O bond (this leads to cleavage of the furazan ring resulting in the formation of diamines 81 and 82).Depending on the reaction conditions (ratio of the reactants, temperature, reaction duration), a mixture of compounds enriched in a particular reduction product can be obtained.71 The yields of individual compounds are rather low. When LiAlH4 is used as the reducing agent, the reaction carried out at room temperature affords alcohol 79 in 62% yield. A B Sheremetev Ph Ph HOCH2 EtO2C N N NaBH4, EtOH + O O N N N N 79 Ph 54b Ph Ph Ph NH2 EtO2C EtO2C N + + O + N HN N NH2 Ph H 80 81 Ph Ph NH2 HOCH2 + N NH2 82 Ph Only two types of products can be formed in a similar reaction of the cyano derivative 54d. Stirring this compound with sodium borohydride in ethanol for 1 h at 20 8C gives furazan 83 in 65% yield and diamine 84 in a yield of only 4%.If the reaction mixture is refluxed for 10 min, the diamine 84 becomes the sole reaction product (yield 79%).71 Ph NC N NaBH4, EtOH O N N 54d Ph Ph Ph NH2 NC NC N + O N HN N NH2 Ph 83 H 84 Ph Treatment of the dihydro derivatives 80 and 83 with NaOH in ethanol (1.5 h, 20 8C) or with DBU in benzene (4 h) induces oxidative aromatisation giving rise to the compounds 54b,d in good yields.71 The structures of furazano[3,4-c] pyridine derivatives have been studied by mass spectrometry,72, 73 IR 71 and NMR67 spec- troscopy and UV spectrometry.71 The compound 54d has been patented as a luminescent dye.74 3. Furazano[2,3-a]pyridines Furazano[2,3-a]pyridines (C) are also referred to as 1,2,5-oxadia- zolo[2,3-a]pyridines.The only method for the synthesis of these compounds known to date is based on pyridine precursors. The reaction of the E-oxime of 2-acetylpyridine N-oxide 85a with toluene-p-sulfonyl chloride in the presence of alkali leads on closure to a furazan ring and gives the bicyclic derivative 87 in 67% yield.75 Me Me TsCl N N N Me NaOH O N N N OTs O O OTs OH 87 86a 85a Presumably, this transformation involves the intermediate formation of the tosylate 86a. A similar reaction of the corre- sponding Z-isomer 85b does not give a bicyclic product but stops at the stage of formation of the tosylate 86b.The chemistry of furazans fused to six- and seven-membed heterocycles with one heteroatom Me Ac2O 85a N N Me N N OAc O O90a 89 sigmatropic rearrangement N Me N O N N Me O N O 91 Treatment of the E- and Z-isomers of the methylquinolyl and methylisoquinolyl ketoxime N-oxides did not result in closure to give a furazan ring.76 Me Me TsCl N N NaOH N N OHO85b OTsO86b Heating of the E-isomer 85a in acetic anhydride at 130 8C for 2 h affords furazan derivative 88 in 88% yield.At temperatures below 100 8C, the reaction stops at the stage of formation of acetyl derivative 89. The Z-isomer 85b is not converted into a furazan derivative even at high temperatures. The compound 88 is formed by a radical mechanism. This was confirmed experimentally by conducting the reaction in the presence of radical sources or radical traps.The course of the reaction can be described by a scheme with intermediates 90 ± 92 (Scheme 1).77 The compound 88 remains unchanged when refluxed with 20% hydrochloric acid or alkali. However, it is readily bromi- nated in chloroform to give a mixture of di- (93) and tetra- brominated (94) compounds in 52% and 33% yields, respectively. On refluxing in ether for 4 h, the compound 94 eliminates a bromine molecule, being quantitatively converted into the dibro- mide 93.77 The dimer 88 is stable against catalytic hydrogenation (H2, Raney nickel).77 O Br N Br2, CHCl3 + 88 N Br Me N Me93 O N Br O Br Br N + N Me N Br O N Me94 The tosylate 87 is reduced with hydrogen over Pd to give, depending on the reaction conditions, either hydrogenated bicy- clic compound 95 (in a yield of up to 25%) or monocyclic furazan 96 (in a yield of up to 89%).75 Me Me(CH2)3 H2, Pd N N Me Me + N N O N O N O 96 87 95 OTs OTs The structure of these derivatives, which are unusual for the furazan series, was established based on thorough analysis of their 145 Scheme 1 dimerisation N Me N N Me Me N N O N 90d O90c O90b H O O N N H oxidation N N N Me Me Me Me O N N 92 88 1H NMR, 13C NMR, UV, IR and mass spectra; 2D NMR techniques have also been used for this purpose.75, 77 IV.Furazanopyrans Of the two theoretically possible types of bicyclic compounds comprising furazan and pyran rings (D and E), only furazano- [3,4-c]pyran derivatives (type E) are known.4 4 3 3 5 5 N N +O O+ O2 O2 6 6 N17 N17 E D Furazano[3,4-c]pyran derivatives can be obtained both from furazan and pyran synthons. Thus treatment of furazan-3-car- boxylic acid 97 with acetic anhydride or dicyclohexylcarbodiimide (DCC) induces intramolecular esterification. This gives lactone 98 in 80% ±85% yield.78 O OHCOOH Ac2O or DCC O N N N N O 98 O97 Dehydration of glyoxime 99 is accompanied by the formation of a furazan ring. This reaction gives cage furazan compound 100.79 Me Me NOH N O O Me O Me N Me Me NOH 100 99 Refluxing of the pyridinium salt 101, containing a carbohy- drate fragment, with sodium nitrite in the presence of pyridine in aqueous THF leads to furoxan 103, which is formed in 58% yield, apparently, via the intermediate formation of nitro derivative 102.80 O O O O Ph Ph NaNO2 NH2OH, Py O O OMe O OTs OMe HON Py+TsO7 101 O O O Ph O Ph OMe O OMe N N O O HON NO2 102 103 O146 The isomeric furoxan 105 can be prepared by successive treatment of 2-keto-D-glucopyranoside 104 with hydroxylamine in aqueous pyridine and with a solution of sodium nitrite in aqueous dioxane followed by keeping the reaction mixture for 3 h at 90 8C.The yield of the furoxan 105 is 40%.80 O O Ph OMe O O OTs O OMe Ph O N N O O O 104 105 Data on the reactivity of furazano[3,4-c]pyran derivatives are very scarce.It is only known that the lactone ring in the compound 98 can be opened by amines to give amides of furazan-3-carbox- ylic acid 106.78 O OHO HNRR0 O NRR0 N N N N O O 106a ± c 98 R=R0=Me (a), Et (b); R=H, R0=Ph (c). Spirocyclic lactone 107 is cleaved on treatment with methanol, being thus converted into monocyclic furazan derivative 108.81 O O OH CO2Me N O MeOH CH2 O N N N O O 108 107 Compounds of this group have been characterised by the data of IR and UV spectroscopy and mass spectrometry.80 V. Furazanothiopyrans Two types of furazanothiopyrans can be conceived, furazano- [3,4-b]thiopyrans (F) and furazano[3,4-c]thiopyrans (G). How- ever, no derivatives of structure F are known. 4 4 3 3 5 5 N N +S S+ O2 O2 6 6 N17 N17 G F Only one study concerning furazano[3,4-c]thiopyran deriva- tives has been published.63 The dehydration of trioximes 109a ± d induced by heating with alkali or DCC is accompanied by closure to give a furazan ring, resulting in the formation of oximes 109a ± d in 67%± 85% yields.The formation of a furazan ring with simultaneous acylation of the third oxime group was observed when trioximes 109b and 109d were treated with acetic anhydride. The acetates of bicyclic oximes 110d and 110e were isolated in 88%± 93% yields.63 NOH N ON NOH R2ON HON R1 S R1 R1 R1 S 110a ± e 109a ± d R1=R2=H(a); R1=Ph, R2=H(b); R1=Me, R2=H(c); R1=2-furyl, R2=H(109d), Ac (110d); R1=Ph, R2=Ac (110e). No data concerning the reactivity of this bicyclic system can be found in the literature.VI. Furazanoazepines Four types of annelation of furazan and azepine rings (H, I, J and K) are possible. H 3 3 5 N N N 4 6 2 O 2 O 1 N N 7 1 8 H I H 4 3 5 3 N 6 2 N 2 O N 1 O N9 N 7 1 8 H K J Only few derivatives of furazano[3,4-c]azepine (I) have now been described. The heterocyclic system of furazano[3,4-c]azepine has been mentioned in the literature as 1,2,5-oxadiazolo[3,4-c]azepine. The Beckmann rearrangement of Z-oxime 111 in the presence of PCl5 in ether at 20 8Cgives rise to azepin-4-one 112 (yield 29%). Treatment of E-oxime 113 under similar conditions gives the same product 112 in a yield of only 5%. The main bulk of the initial compound is converted into a polymer rather than into the expected isomeric azepine 114.82 HO N PCl5 N Me ON Me 111 OH N PCl5 N Me ON Me 113 O NaN3, H2SO4 N Me ON Me 115 H O N N 113 ON Me Me 114 Azepine 112 is also formed upon the Curtius rearrangement of ketone 115, induced by treatment with sodium azide in a mixture of chloroform and sulfuric acid.Note that the Curtius rearrange- ment gives the azepine 112 in a higher yield (55%) than the Beckmann rearrangement.83 The reaction of 112 with acetyl chloride in the presence of sodium hydride in toluene gives the corresponding N-acetylation product in 67% yield.82 The structure of this azepine has been confirmed by the data of its IR, 1H NMR and mass spectra.The azepine 112 exhibits moderate antimicrobial activity.83 VII. Furazanothiepines Of the three types of annelation of a furazan ring to a thiepine ring possible theoretically (L, M and N), only type N derivatives, furazano[3,4-d]thiepines, have been synthesised. A B Sheremetev 4 5 N 6 7 84 5 6 7 8 O H N N ON Me Me 112The chemistry of furazans fused to six- and seven-membered heterocycles with one heteroatom 4 4 4 5 3 3 3 5 5 S S N N N 6 6 2 O 2 O 2 O S 6 1 1 N N N 7 7 7 1 8 8 8 M N L Heating of a benzene solution of bis-N,N-oxide 117, prepared by treatment of dioxime 116 with an alkaline solution of sodium hypochlorite, affords furoxan 118 in 90% yield.84 O O NOH HON N N O O NaOCl D O O S S 116 117 O O S N N O O 118 Thieno-annelated compounds 119a,b have been synthesised in a similar way.85 However, the replacement of a benzene ring in the initial molecule by a thiophene ring results in somewhat lower yields of the corresponding tetracyclic compounds 119a,b (72% ± 74% yield).O O N N NOH HON O O O O NaOCl D S S S S R R O O S S R N N O 119a,b O R=H(a), Me (b). Furoxans 119a,b exist as unseparable mixtures of isomers. VIII. Conclusion The data presented in this review demonstrate on the one hand, the great synthetic potential of the class of compounds considered and, on the other hand, the scarcity of investigations along this line. The latter is apparently due to the facts that the required initial compounds are relatively difficult to obtain and that the methods for their transformations are not adequately developed.Meanwhile, some existing efficient methods for the construction of the title bicyclic systems have not yet found wide use. We hope that systematisation of the published data presented in this review will serve as the basis for the further development of this field of chemistry. References 1. V G Andrianov, A V Eremeev Khim. Geterotsikl. Soedin. 1443 (1990) a 2. S-K Lin Youji Huaxue 11 106 (1991); Chem. Abstr. 114 143 173 (1991) 3. W Sliwa, A Thomas, N Zelichowicz Collect. Czech. Chem. Commun. 57 978 (1992) 147 4. M E Niyazymbetov, E V Ul'yanova, V A Petrosyan Elektrokhimiya 28 555 (1992) b 5.W Friedrichsen, in Houben-Weyl, Methoden der Organischen Chemie Vol. E8c (Ed. E Schaumann) (Stuttgart: Thieme Verlag, 1994) p. 648 6. A B Sheremetev J. Heterocycl. Chem. 32 371 (1995) 7. N Vivona, S Buscemi Heterocycles 41 2095 (1995) 8. R M Paton, in Comprehensive Heterocyclic Chemistry II Vol. 4 (Eds A R Katritzky, CWRees, E F V Scriven) (Oxford: Pergamon Press, 1996) p. 229 9. A B Sheremetev Ross. Khim. Zh. 41 43 (1997) c 10. N N Makhova, T I Godovikova Ross. Khim. Zh. 41 54 (1997) c 11. L I Khmel'nitskii, S S Novikov, T I Godovikova Khimiya Furoksa- nov. Stroenie i Sintez (The Chemistry of Furoxans. The Structure and Synthesis) (Moscow: Nauka, 1996) 12. L I Khmel'nitskii, S S Novikov, T I Godovikova Khimiya Furoksa- nov. Reaktsii i Primenenie (The Chemistry of Furoxans.The Reac- tions and Use) (Moscow: Nauka, 1996) 13. A F Pozharskii Teoreticheskie Osnovy Khimii Geterotsiklov (Theo- retical Fundamentals of the Chemistry of Heterocycles) (Moscow: Khimiya, 1985) p. 23; 103 14. C W Bird Tetrahedron 41 1409 (1985) 15. M A Shokhen, V G Andrianov, A V Eremeev Khim. Geterotsikl. Soedin. 1355 (1979) a 16. I V Tselinskii, S F Mel'nikova, S N Vergizov, G M Frolova Khim. Geterotsikl. Soedin. 35 (1981) a 17. I V Tselinskii, S F Mel'nikova, S N Vergizov Khim. Geterotsikl. Soedin. 321 (1981) a 18. I V Tselinskii, S F Mel'nikova, M P Zelenov Zh. Org. Khim. 32 766 (1996) d 19. A N Vereshchagin Induktsionnyi Effekt. Konstanty Zamestitelei dlya Korrelyatsionnogo Analiza (Induction Effect.Substituent Constants for Correlation Analysis) (Moscow: Nauka, 1988) 20. H Gunther Tetrahedron Lett. 2967 (1967) 21. M J Cook, A R Katritzky, P Linda, in Advances in Heterocyclic Chemistry Vol. 17 (Eds A R Katritzky, A J Boulton) (New York: Academic Press, 1974) p. 256 22. R K Harris, A R Katritzky, S Oeksne, A S Bailey, W G Paterson J. Chem. Soc. 197 (1963); Chem. Abstr. 58 6820 (1963) 23. L W Deady, N H Quazi Spectrosc. Lett. 28 1033 (1995) 24. V G Andrianov, M A Shokhen, A V Eremeev Khim. Geterotsikl. Soedin. 1261 (1989) a 25. E S Levin, Z I Fodiman, Z V Todres Elektrokhimiya 2 175 (1966) b 26. NMAtherton, J N Ockwell, R Dietz J. Chem. Soc., A 771 (1967) 27. V Sh Tsveniashvili, S I Zhdanov, Z V Todres Khim.Geterotsikl. Soedin. 712 (1968) a 28. Z V Todres, Z I Fodiman, E S Levin Khim. Geterotsikl. Soedin. 604 (1974) a 29. V Sh Tsveniashvili, M V Malashkhiya, N S Khavtasi, S F Mel'nikova Zh. Obshch. Khim. 53 2511 (1983) e 30. V M Tsentovskii, L V Petukhova,M I Evgen'ev Kislotno-Osnovnye Ravnovesiya Benzofurazanov v Dipolyarnykh Protonnykh i Aproton- 36. J J Eatough, L S Fuller, R H Good, R K Smalley J. Chem. Soc., C nykh Rastvoritelyakh (Acid ± Base Equlibria of Benzofurazans in Dipolar Protic and Aprotic Solvents); article deposited at the ONI- TEKhim, Cherkassy, 1984. No. 446khp-D84 31. L S Vasilev, A B Sheremetev, V A Dorokhov, N Kiu Khoa, M O Dekaprilevich, Yu T Struchkov, L I Khmelnitskii Mendeleev Commun. 57 (1994) 32. L S Vasil'ev, A B Sheremetev, V A Dorokhov, N Khua, Z K Dem'yanets, D E Dmitriev Izv.Akad. Nauk, Ser. Khim. (1999) (in the press) f 33. T Miyazawa, T Takabatake,M Hasegawa J. Heterocycl. Chem. 32 387 (1995) 34. L W Deady, N H Quazi Aust. J. Chem. 46 411 (1993) 35. J H Boyer, D I McCane, W J McCarville, A T Tweedie J. Am. Chem. Soc. 75 5298 (1953) 1874 (1970) 37. C K Lowe-Ma, R A Nissan, W S Wilson J. Org. Chem. 55 3755 (1990) 38. B Stanovnik,M Tisler Chimia 25 272 (1971) 39. D Binder, C R Noe, J Nussbaumer, B C Prager Monatsh. Chem. 111 407 (1980) 40. J H Boyer,W Schoen J. Am. Chem. Soc. 78 423 (1956)A B Sheremetev 148 85. O O Mamaeva,M M Krayushkin, F M Stoyanovich Izv. Akad. Nauk SSSR, Ser. Khim. 913 (1990) f 41. V M Tsentovskii, M I Evgen'ev, I F Falyakhov, I I Evgen'eva, I V Shulaeva, A V Knyazev Zh.Obshch. Khim. 57 660 (1987) e 42. H H Licht, H Ritter, in Proceedings of the 24th International Annual Conference ICT. Energetic Materials�Insensitivity and Environ- mental Awareness, Karlsruhe, 1993 p. 6/1-8 43. H Ritter, H H Licht J. Heterocycl. Chem. 32 585 (1995) 44. N R Ayyangar, S Madan Kumar, K V Srinivasan Synthesis 616 (1987) 45. S K Kotovskaya, G A Mokrushina, I Ya Postovskii, M V Polyakova Khim. Geterotsikl. Soedin. 654 (1981) a 46. L W Deady,M W Wong Aust. J. Chem. 38 1045 (1985) 47. Altaf-ur-Rahman, A J Boulton, D P Clifford, G J T Tiddy a�Chem. Heterocycl. Compd. (Engl. Transl.) b�Russ. J. Electrochem. (Engl. Transl.) c�Mendeleev Chem. J. (Engl. Transl.) d�Russ.J. Org. Chem. (Engl. Transl.) e�Russ. J. Gen. Chem. (Engl. Transl.) f�Russ. Chem. Bull. (Engl. Transl.) g�Kinet. Catal. (Engl. Transl.) h�Pharm. Chem. J. (Engl. Transl.) J. Chem. Soc., B 1516 (1968) 48. L W Deady, J J Harvey, T B Jarman Aust. J. Chem. 45 371 (1992) 49. A J Boulton, P J Halls, A R Katritzky J. Chem. Soc., B 636 (1970) 50. I Ya Postovskii, S K Kotovskaya, G A Mokrushina, in Teoreti- cheskie i Prikladnye Aspekty Khimii Aromaticheskikh Soedinenii. (Tez. Dokl. III Vsesoyuz. Simp. po Organicheskomu Sintezu) [Theo- retical and Applied Aspects of the Chemistry of Aromatic Com- pounds (Abstracts of Reports of the Third Simposium on Organic Synthesis)] (Moscow: Nauka, 1981) p. 39 51. D Binder, A Georgopoulos, C R Noe, J Nussbaumer, B C Prager, F Turnowsky Arzneim.-Forsch.32 10 (1982) 52. BRD P. 2 703 369; Chem. Abstr. 87 152 274 (1977) 53. Yu Yu Nikishev, I Sh Saifullin, O R Klyuchnikov Kinet. Katal. 34 969 (1993) g 54. I F Falyakhov, G P Sharnin, N M Safin, I Sh Saifullin, R Kh Fassakhov, in Proceedings of the XXVth International Pyrotechnics Seminar, Moscow, 1995 p. 187 55. O R Klyuchnikov, Yu Yu Nikishev Khim. Geterotsikl. Soedin. 1573 (1995) a 56. US P. 4 557 306; Chem. Abstr. 104 131 206 (1986) 57. Eur. P. 175 062; Chem. Abstr. 105 61 965 (1986) 58. US P. 4 570 690; Chem. Abstr. 105 7773 (1986) 59. Eur. P. 195 927; Chem. Abstr. 106 139 650 (1987) 60. Z X Lu, N H Quazi, L W Deady, G M Polya Biol. Chem. Hoppe-Seyler 377 373 (1996) 61. R Willstatter Chem. Ber. 30 2679 (1897) 62. L B Volodarsky Janssen Chim. Acta 8 (3) 12 (1990) 63. A B Sheremetev, in Proceedings of the 16th International Congress on Heterocyclic Chemistry (Abstracts of Reports), Bozeman, 1997 POII-239 64. T L Hough, G Jones J. Chem. Soc., C 1088 (1968) 65. S Mataka, K Takahashi, M Tashiro Synthesis 687 (1979) 66. S Mataka, K Takahashi, M Tashiro Synthesis 842 (1980) 67. BRD P. 4 218 977; Chem. Abstr. 120 270 408 (1994) 68. P Roschger, W Fiala,W Stadlbauer J. Heterocycl. Chem. 29 225 (1992) 69. A S Bailey, M W Heaton, J I Murphy J. Chem. Soc., C 1211 (1971) 70. M J S Dewar, P M Maitlis J. Chem. Soc. 944 (1957) 71. S Mataka, K Takahashi, T Imura,M Tashiro J. Heterocycl. Chem. 19 1481 (1982) 72. R T Aplin,W T Pike Chem. Ind. 2009 (1966) 73. L K Dyall Org. Mass. Spectrom. 24 465 (1989) 74. Eur. P. 406 762; Chem. Abstr. 115 146 246 (1991) 75. Y Tagawa, N Honjo, Y Goto, T Chiba, T Kato Chem. Pharm. Bull. 31 2269 (1983) 76. Y Tagawa, N Honjo, Y Goto Chem. Pharm. Bull. 34 564 (1986) 77. Y Tagawa, N Honjo, Y Goto, T Kato Chem. Pharm. Bull. 34 4984 (1986) 78. M Giannella, F Gualtieri, W Fedeli, S Cerrini, E Gavuzzo J. Heterocycl. Chem. 20 385 (1983) 79. G Cusmano Gazz. Chim. Ital. 55 218 (1925) 80. C S Wu,W A Szarek, J K N Jones J. Chem. Soc., Chem. Commun. 1117 (1972) 81. A Gasco,A J Boulton, in Advances in Heterocyclic Chemistry Vol. 29 (Ed. A R Katritzky) (New York: Academic Press, 1981) p. 295 82. S Mitkido, J Stephanidou-Stephanatov, C A Tsoleridis, N E Alexandrou Collect. Czech. Chem. Commun. 55 245 (1990) 83. E I Ivanov, I P Konul, L A Konul, D E Stepanov, L V Grishchuk, V V Vysotskaya Khim.-Farm. Zh. 37 (1993) h 84. O O Mamaeva, F M Stoyanovich, M M Krayushkin Izv. Akad. Nauk SSSR, Ser. K

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (2) 149 ± 164 (1999) Reactions of halophosphines with conjugated heterodienes E Ya Levina, AMKibardin Contents I. Introduction II. Reactions with oxadienes III. Reactions with azadienes IV. Reactions with oxazadienes V. Conclusion Abstract. Data on reactions of halophosphines PHal3, 7PHal2, and PHal with hetero-1,3-dienes resulting mostly in the forma- tion of phosphorus-containing heterocycles are generalised and systematised. The influence of the nature, the number and positions of heteroatoms (O, N) in the heterodiene system, and of the nature of the substituents at the phosphorus atom on the reaction pathways and mechanisms is analysed. The bibliography includes 155 references. I. Introduction Reactions of tricoordinate phosphorus derivatives with diene systems is an important route to cyclic phosphorus-containing compounds, first of all following the [4+1]-cycloaddition mech- anism.In this regard, reactions of halophosphines (HP) PHal3, 7PHal2, and PHal are most numerous. Their reactions with 1,3- dienes containing no heteroatoms are well studied. The reaction of [4+1]-cycloaddition of HP to 1,3-dienes was discovered by McCormack in 1953 1, 2 and most intensively studied in the 1960's and early 1970's simultaneously by several groups of investigators. Reactions of HP with conjugated dienes have been adequately covered in reviews 3, 4 and monographs.5±7 Despite a wide variety of compounds obtained, they all belong to phosphol- 2(3)-enes ([4+1]-cycloaddition reaction), to polymers or oligom- ers containing a phosphorus atom in the chain (1,4-addition reaction), or to the products of cyclisation occurring with reten- tion of the system of conjugated bonds.The studies in this field 8, 9 reported after publication of a book by Quin 6 in 1981 contain no radically new data, except for a series of studies dedicated to investigation of reactions of 1,3-dienes with HP of the X2CHPX2 (X=Cl, Br, I) type in the presence of triethylamine.10 ± 16 Under the action of excess triethylamine, this type of HP undergo dehydrohalogenation to form unstable phosphaethenes XP=CX2. The latter enter in situ into the Diels ± Alder reaction with dienes following the [4+2]-cycloaddition mechanism to give aromatic compounds, viz., 2-halophosphorinines containing a E Ya Levina, AMKibardinAE Arbuzov Institute of Organic and Physical Chemistry, Kazan Scientific Centre of Russian Academy of Sciences, ul.Acad. Arbuzova 8, 420083 Kazan, Russian Federation, Fax (7-843) 275 22 53. Tel. (7-843) 276 73 34. E-mail: lina@iopc.kcn.ru (E Ya Levina) Received 29 April 1998 Uspekhi Khimii 68 (2) 167 ± 183 (1999); translated by AMRaevsky #1999 Russian Academy of Sciences and Turpion Ltd UDC 66.095.252:547.241:547.77:547.8 149 149 157 161 161 dicoordinate phosphorus atom, via intermediate six-membered 1,2,2-trihalo-1,2,3,6-tetrahydrophosphorinines. The study of reactions ofHPaddition to hetero-1,3-dienes was begun by Conant 17 in 1917 taking unsaturated carbonyl com- pounds as an example and is being developed successfully at present. These reactions are very diverse and result in the formation of not only five-, but also six-membered rinds, and open-chain structures as well.Recently, the range of heterodienes under study has been considerably extended, which has led, in particular, to obtaining of novel classes of phosphorus-containing heterocycles, which are not only of theoretical, but also of considerable practical interest. The aim of the present review is to systematise the available information on the reactions of HP with heterodienes and to reveal peculiarities of the influence of heteroatoms in the diene system on the direction and mechanisms of these reactions. We do not discuss here the reactions of compounds that formally contain a heterodiene system, if one or both double bonds of a diene are fragments of arene or hetarene systems.II. Reactions with oxadienes 1. Reactions with a-dicarbonyl compounds Investigation of the reactions of P(III) derivatives with a-dicar- bonyl compounds was carried out almost simultaneously with the study of analogous reactions of dienic hydrocarbons. However, it should be noted that in most cases it was phosphites that have been used as dienophiles, whereas reactions with HP have been the subject of an insignificant number of publications; moreover, all of them involved a-diketones only. Like dienic hydrocarbons, a-diketones 1 react with dialkyl- phosphorochloridites by the [4+1]-cycloaddition mechanism to form adducts 2 containing a pentacoordinate phosphorus atom.Further elimination of alkyl chloride from the adducts 2 according to the mechanism of the second stage of the Arbuzov reaction (see Refs 18, 19) gives 2-alkoxy-2-oxo-1,3,2-dioxaphospholes 3.20, 21 O R1 R1 O + (R2O)2PCl P 7R2Cl OR2 OR2 Cl R1 O O R1 1 2 R1 O OR2 P O R1 O3 R1=Me, Ph; R2=Et, Prn, Bui, Bun.150 Dialkyl phosphorochloridites are less reactive than trialkyl phosphites and phosphoramidites.20 ± 22 It is believed 22 that the closure of the dioxaphosphole ring occurs as a result of a concerted or almost concerted passage of the lone electron pair of the phosphorus atom to the conjugated p-system of the a-diketone and transfer of the p-electrons of this system to orbitals of the phosphorus atom.The reaction of 2-chloro-1,3,2-dioxaphospholane (4) with biacetyl occurs to give initially spirophosphorane 5 containing a P(V)7Cl bond and is followed by its conversion into 4,5- dimethyl-2-oxo-2-(2-chloroethoxy)-1,3,2-dioxaphosphole (6) by the mechanism of the second stage of the Arbuzov reaction.23 Me O O O Me O D P +Cl P O O Me Me O Cl O 4 5 Me Me O O O D O +PP O(CH2)2Cl O Me Me OCl7O 6 Biacetyl reacts with dichlorophosphines in a different way, viz., with the formation of 4-chloro-3H-5-methyl-2-oxo-1,2-oxa- phosphole 7, following a mechanism involving recyclisation.24, 25 Me O Me O +RPCl2 PRCl2 O Me O Me Me O O O Cl Me P 7HCl MeC C OPRCl R Cl 7 R=Me, Et, Ph.A phosphorane-type product of [4+1]-cycloaddition is obtained in the reaction of phosphorus trichloride with 3,5-di- tert-butyl-o-quinone (8).26 Based on detection of signals in the EPR spectra recorded during this reaction, it was argued 26 that its mechanism involves a stage of one-electron transfer. But But O O +PCl3 PClá3O But O But 8 But O PCl3 O But Similar signals in the EPR spectra were also observed in the reactions of PCl3 with o-chloranil, o-naphthoquinone, and 9,10- phenanthrenequinone; however, the final products have not been isolated.26 It should be noted that the appearance of paramagnetic intermediates, detected by the EPR technique, in the reaction mixture is not the unambiguous evidence for the formation of the final phosphorane as a result of one-electron transfer.Phosphenium salts, which are closest analogues of HP, react with o-quinone 8 similarly to HP, following the [4+1]-cyclo- addition mechanism to form 1,3,2-dioxaphospholenium salts 9.27 But R1 O +P8+R1R2P+X7 X7 R2 O But 9 R1=Pri2N: R2=Cl, N=PPh3; R1, R2=Alk; R1±R2=MeN(CH2)2NMe; X=AlCl4, CF3SO3. In the reactions with phosphenium salts the o-quinone 8 behaves analogously to 1,3-dienes.28 ± 30 At the same time, (Pri2N)2PCl in which the phosphorus atom is less electrophilic than in the phosphenium cation, does not enter into this reac- tion.27 Thus, as in the case of dienic hydrocarbons, symmetric heterodiene systems of diketones and o-quinone add a phospho- rus-containing dienophile at positions 1 and 4 with the formation of a cycloadduct containing a pentacoordinate phosphorus atom in the first stage.The final reaction products are derivatives of 1,3,2-dioxaoxaphospholes or 3H-1,2-oxaphospholes. No indica- tions of the existence of intermediate bipolar ions were found. The intermediate 2 was detected by 31P NMR spectroscopy.22 2. Reactions with a,b-unsaturated carbonyl compounds Mesityl oxide and benzylideneacetophenone were the first 1,3- heterodienes used to study the reactions with HP taking PCl3 and PBr3 as an example.18 The assumption that the reactions with phosphorus trihalides follow the [4+1]-cycloaddition mechanism with the formation of five-membered cycles, viz., 3H-1,2-oxa- phospholes,17 was confirmed in further studies.A brief summary of the results obtained in these earlier studies has been reviewed,3 so we do not discuss them here. After the review 3 was published, it was established that passage of hydrogen sulfide through a mixture of phosphorus trihalide and a,b-unsaturated ketones 10 ± 12 afforded 3H-1,2-oxaphosphole sulfides 13 ± 15 in good yields.31, 32 This is evidence in favour of [4+1]-cycloaddition. O Me R3 +PCl3 R110 R2Me S O P Cl R1 R3 R213 R1, R2, R3=H, Me. O +PCl3 11 S O P Cl 14 O Ph Me +PX3 12 PhPh S O P X Me Ph15 X=Cl, Br. It was also shown later that heating phosphorus trichloride and mesityl oxide in a sealed tube in the absence of a solvent gave 2-chloro-3,3,5-trimethyl-2-oxo-3H-1,2-oxaphosphole.33 We believe that the formation of this compound can be explained by the replacement of two chlorine atoms in the intermediate [4+1]- cycloaddition product by the carbonyl oxygen atom of the second mesityl oxide molecule.E Ya Levina, AMKibardin Me O H2S PCl3 72 HCl R1 R2 R3 O PCl3 H2S 72 HCl Ph O H2S PX3 72HX Me PhReactions of halophosphines with conjugated heterodienes O Me Me O O P +PCl3 Me H Cl Me Me Me The reactions of methyl phosphorodichloridite with ketones 10 also follow the Diels ± Alder reaction mechanism, subsequent elimination of methyl chloride from the adduct 16 results in 2-chloro-2-oxo-3H-1,2-oxaphospholes 17.34 Me Me Cl O O O P P 10+MeOPCl2 7MeCl Cl OMe Cl R1 R1 R3 R3 R2 R217 16 R1, R2, R3=H, Me.Reactions of dichlorophosphines with a,b-unsaturated ketones 10 in acetic anhydride give first the adducts of the type 16 and then 3H-1,2-oxaphospholes 18.35 ± 37Me O O Ac2O P [adduct] 10+R4PCl2 72 AcCl R1 R4 R3 R218 R1, R2, R3=H, Me; R4=Et, Ph, 2-thienyl. Benzylideneacetone and phenyl(dichloro)phosphine do not enter into this reaction in acetic anhydride.35 However, the reactions of dichlorophosphines and phenyl phosphorodichloridites with mesityl oxide (heating in sealed tubes in the absence of a solvent) afford acyclic b,g-unsaturated phos- phorus acid chlorides 19 in which the coordination number of P atom is 4.The latter result from a ring ± chain isomerisation of the adducts of the type 16 initially formed in the Diels ± Alder reaction.33 O R Me Me [adduct] +RPCl2 C CHCMe2P Me Cl Cl 19 O Me R=Me, Et, Ph, PhO. Analogous products are obtained in the reaction of dichloro- phosphines with a,b-unsaturated aldehydes 20. In this case the reaction occurs under milder conditions (in solution).38 O PR2Cl2 R1CH CHCHO+R2PCl2 20 R1 H CHCHR1PR2Cl C Cl O R1=Me, H; R2=Et, Ph. Aldehydes 20 react with phenyl(dichloro)phosphine or PCl3 in acetic acid to give a-hydroxy-b,g-unsaturated phosphorus acid dichlorides 21 in which the coordination number of the P atom is 4.38 H AcOH CHCH(OH)PR2Cl C 20+R2PCl2 7AcCl R1 21 O R1=Et, Ph; R2=Ph, Cl.151 Compounds 4, 22 and 23 react with a,b-unsaturated ketones 10 (heterodienes) by mechanisms involving cycloaddition fol- lowed by dealkylation, which is analogous to the second stage of the Arbuzov rearrangement. This results in the formation of 3H- 1,2-oxaphosphole derivatives,39, 40 as is shown below taking the reaction with benzylideneacetone as an example. Cl Me Me O O O O (CH2)n (CH2)n +ClP P Ph O H H R R O Ph H 4, 22, 23 H O Me O P O CH(CH2)nCl H R H Ph 4: R=H, n=1; 22: R=Me, n=1; 23: R=H, Me; n=2. As was mentioned in Section II.1, biacetyl reacts with phos- pholane 4 following a conventional mechanism of cycloaddition to the O=C7C=O bond system. However, it was reported 41 that the reactions of biacetyl with compounds 4, 22 and 23 under virtually the same conditions resulted in the formation of quite different products, viz., 1,2-oxaphospholan-4-ones 24. This was rationalised by a [4+1]-cycloaddition of compounds 4, 22 and 23 to the enol form of biacetyl as the heterodiene.Cl Me O O Me O O (CH2)n (CH2)n +ClP P O HO CH2 R O R HO 4, 22, 23 Me O P O CH(CH2)nCl O R 24 O Condensation of ketones 10 with 2-chloro-1,3,2-oxathiaphos- pholane occurs analogously to the reaction with compound 4. The second stage of the Arbuzov rearrangement of the intermediate adduct occurs with the cleavage of the C7O bond (not the C7S bond, as in the reaction with butadiene) to give derivatives of 2-oxo-3H-1,2-oxaphospholes 25 containing the P7S bond.42 Cl Me O O O P P 10+Cl S R1 S R3 R2 O Me O P S(CH2)2Cl R1 25 R2 R3 R1, R2, R3=H, Me.The structure of the final product of reaction between 2-chloro-1,3,2-dithiaphospholane (26) and ketones 10 (R1=H, Me; R2=R3=H) is determined by the nature of the substituent R1. 2-Thioxo-3H-1,2-oxaphosphole 27 and 2-oxo-3H-1,2- thiaphosphole 28 are respectively formed at R1=Me and H.43 The reaction of compound 26 with ketone 10 (R1=H) occurs either with the rearrangement of an intermediate adduct or with the initial formation of thioxo derivative of the type 27, which after the Pishchimuka rearrangement 44 is converted into the phosphoryl compound 28.152 O Me S +Cl P R3 S 26 R110 R2 Me S(CH2)2Cl O R1=Me P S Me 27 Me S(CH2)2Cl S R1=H P O 28 R1=H, Me; R2=R3=H.Alkylphosphonochloridothioites 30 react with ketones 10 in a similar way; however, in this case, alkyl halide is eliminated from the intermediate [4+1]-cycloaddition product 29 to give 2-ethyl(- phenyl)-2-thioxo-3H-1,2-oxaphospholes 31 (Scheme 1, route a).45 In the presence of proton-donor reagents (acetic acid or water) this reaction results in the formation of S-alkyl phosphinothioates 32 46 (Scheme 1, route b). The formation of products 32 is explained as follows.46 The ambident ion 33 formed in the first reaction stage as a result of nucleophilic C-addition of alkylphos- phonochloridothioite 30 to the unsaturated ketone 10 reacts with acetic acid 46 to give an intermediate acyclic adduct 34.The latter undergoes the Arbuzov rearrangement with the elimination of acyl chloride and the formation of the final product 32. If the reaction is performed in the presence of water, alkylphosphono- chloridothioite 30 is initially hydrolysed into the corresponding acid 35, which then is added to the ketone 10 to give the ester 32 (Scheme 1, route c). We believe that it is possible to combine the routes a, b and c into a unified scheme assuming that betaine 33 can be a common intermediate to routes a and b and that in the presence of water the process can follow not only route c, but also route b (the latter can be considered as a route common to any proton-donor reagents). Under the action of proton-donor reagents, the halogen atom in the phosphorane 29 can be replaced by an OX group to form a product 36.The latter adds hydrogen halide to give an ion 34, a R2 SR5 Me O SR5 + MeC C C PR4 R3 +Cl P R1 R3 R1 7O Cl R4 30 33 R2 10 b XOH SR5 H P 30 H2O 7HCl R4 O35 R1, R2, R3=H, Me; R1=R2=H, R3=Ph; R4=Et, Ph; R5=Et, Prn, Pri, Bun, Bui; X=H, Ac. E Ya Levina, AMKibardin which is in equilibrium with an ion 37, and then the final product 32. Heating of diethylchlorophosphine and mesityl oxide in a sealed tube in the absence of a solvent resulted in a crystalline adduct 38, which was converted into phosphine oxide 39 upon treatment with alcohols following the scheme shown below:47 O Me Me O ROH P Me+Et2PCl Et Et Cl 7HCl H Me Me Me 38 Me Me Cl7 O + HCl P MeC CHC PEt2 Et Et OR 7RCl OH Me Me Me OR 41 40 Me MeCCH2C PEt2 O Me O 39 R=Me, Bu.Apparently, the intermediates 3, 40 and 41 are the analogues of intermediates 29, 36 and 34, respectively, while the product 39 is an analogue of the products 32. It is quite possible that both products 39 and 32 are formed via intermediates 33. Thus, a conclusion can be drawn that the reactions of P(III) acid mono- chlorides with unsaturated ketones in the presence of proton- donor reagents follow a common mechanism irrespective of the reaction conditions. A peculiar type of reaction is that of HP with b-diketones in the presence of organic base. It is believed that b-diketones 42 enter into reaction in the enol form, i.e., that they can be considered as a particular case of a,b-unsaturated ketones 10 (R2=Me, R3=OH).Reactions of P(III) dichlorides with b-dike- tones give 4-methylene-4H-1,3,2-dioxaphosphorinines 43 with a semicyclic homodiene system of bonds.48 ± 51 Scheme 1 Me S O P R4 7R5Cl R1 Me O R3 R231 P SR5 Cl R4 R1 Me O R3 R2 29 P SR5 OX R4 XOH 7HCl R1 R3 R2 36 HCl R2 R2 SR5 SR5 + + MeC MeC C C PR4 7OX C C PR4 Cl7 R3 R3 OH OH R1 R1 Cl OX 37 347XCl R2 SR5 MeC CH C PR4 c 10 R3 O O R132153 Reactions of halophosphines with conjugated heterodienes O Me Mn(CO)2Cp O H2C Et3N PPh 2Et3N OH+R2PCl2 7Et3N. HCl R1 Cp(CO)2MnPPhCl2+acac O 72Et3N. HCl H 42 Me Me O Me O O H2C H2C PClR2 PR2 P(Ph)W(CO)5 Et3N acac O O O 7Et3N.HCl H R1 R1 2Et3N Me (CO)5WPPhCl2 72Et3N. HCl Me 43 Me 44 CHAc)2 (CO)5WP(OC 2 acac Ph Me R1=H, Et, Cl, Ac; R2=Et, Ph, MeO, EtO, PrnO, PriO, Me2N, Et2N, EtS. This type of reactions was studied in more detail taking the reaction of pentacarbonylchromium complex 46 with acetylacet- one as an example.57 acac (CO)5CrPPhOC CHAc 7HCl 7HCl Cl Me The reaction proceeds via intermediate heterodienes, viz., a,b- unsaturated ketones 44, which are formed upon O-phosphoryl- ation of the hydroxyl group of the ketones 42, which leaves the diene system intact. Subsequent intramolecular O-phosphoryl- ation leads to the closure of the six-membered ring and the formation of dioxaphosphorinines 43 containing an exocyclic methylene group.49 ± 51 acac 7HCl Et3N (CO)5CrPPh(OCMe CHAc)2 2 acac 72 HCl (CO)5CrPPhCl2 46 46 7HCl If P(III) monochlorides are used instead of P(III) dichlorides in the reaction with ketones 42, in most cases the process is completed by the formation of acyclic products 45, viz., O-phos- phorylated derivatives of the ketones 42, which are analogous to the ketones 44.51 ± 55 Cr(CO)5 O H2C Thus, unlike all reactions considered above, this reaction occurs without involvement of the system of conjugated bonds of the heterodiene at all.acac O Me O Me PPh O 72 HCl H Et3N Me OH+ClPR2R3 Me 7Et3N. HCl R1 R1 OPR2R3 Me 45 42 Finally, it should be noted that ethyl(dichloro)phosphine reacts in acetic acid with acetylacetone in its diketone form rather than in its enol form.58 R1=H, Ac, Cl; R2=EtO, PrnO, Me2N, Et2N; R3=Me, MeO, EtO, PriO, BuiO, Ph, Et2N; R2±R3=O(CH2)2O, OCH2CH(Me)O, o-O2C6H4.Reactions of enamino ketones 47 (nitrogen analogues of the diketones 42) with alkyl phosphorodichloridites yield azaana- logues of dioxaphosphorinines 43, viz., 4-alkylidene-4H-1,3,2- oxazaphosphorinines 48.51, 59 OR5 O R1 O R1 In the case of acetylacetone, it was established that the reaction product 45 is a mixture of Z- and E-isomers with a considerable predominance of the E-isomer.51, 54 P 2Et3N N NHR4+R5OPCl2 72Et3N. HCl R4 R2 R2 48 CHR3 47 CH2R3 R1=H, Me; R2=R3=H; R4=Me, Ph; R1±R2=(CH2)3; R2±R3=(CH2)2; R5=Me, Et, Pr, Bu.The reaction of enamino ketones 47 with dialkyl phosphoro- chloridites halts in the stage of N-phosphorylation of the amino group and the formation of acyclic dialkyl phosphoramidites 49.51, 60 R4 CH2R3 O Et3N 47 + (R5O)2PCl (R5O)2PN 7Et3N. HCl C CR2CR1 49 If P(III) monochlorides containing a dialkylamino group are introduced into the reaction with acetylacetone (42, R1=H), the linear heterodienes 45 formed initially undergo cyclisation into dioxaphosphorinines 43 with the elimination of dialkylamine. Dioxaphosphorinines 43 are isolated as final products after distillation. The reaction occurs successfully in the presence of acidic admixtures, which behave as catalysts and favour the conversion of the dialkylamino group into a good leaving group.Cyclisation occurs with particular ease if the intermediate linear heterodienes 45 contain two dialkylamino groups at the phospho- rus atom. An interesting peculiarity was observed in those cases where it was possible to isolate the intermediates 45 or to obtain them by independent synthesis, viz., the yields of the cyclisation products 43 were higher than the portion of Z-isomer in the intermediates 45. This indicates that Z,E-isomerisation of hetero- dienes 45 occurs under the cyclisation conditions.51, 54 R1=H, Me; R2=R3=H; R4=Me; R5=Et, Pr. Et3N Et3N. HCl [45] 42+ClPR2R3 43 7Et3N. HCl 7Et2NH R1=H; R2=EtO, PrnO, PriO, Et2N; R3=Me2N, Et2N. However, as in the case of diketones 42, the reaction of enamino ketone 47 (R1=R4=Me; R2=R3=H) with bis(die- thyl)phosphorodiamidochloridite occurs in a more complex way.Observation of weak signals along with other signals in the 31P NMR spectrum indicates that the reaction mixture contains 4- methylene-4H-1,3,2-oxazaphosphorinene 50 and/or isomeric 6- methylenephosphorinene 51. Metal complexes of phenyl(dichloro)phosphine react with acetylacetone in the presence of triethylamine analogously to uncoordinated dichlorophosphines, leaving the metal-containing groups intact.56, 57154 NEt2 NEt2 O O Me H2C P PN N Me Me Me 51 CH2 50 However, only 5-methylene-2-oxo-5H-1,2-azaphosphole 52 was isolated after distillation 51, 59 due to the vinyl phosphite ± vinylphosphonate rearrangement.50, 51, 53 Me NEt2 Et3N P 47+ClP(NEt2)2 O 7Et3N.HCl,7Et2NH N H2C 52 Me Reactions of phosphorochloridites and crotonaldehyde in the presence of triethylamine occur with proton abstraction from the methyl group at the double bond and the formation of butadienyl phosphites.61 We believe that the initial attack of phosphoro- chloridite on the oxygen atom of the carbonyl group is followed by proton abstraction from the methyl group. MeCH CHCHO+ClP(OR)2 + Et3N CH Cl7 CH3CH CHOP(OR)2 7Et3N. HCl CH2 CH CH CHOP(OR)2 R=Et, Ph; (OR)2=O(CH2)2O. 2-Chloro-1,3,2-dioxaphospholane (4) reacts with a,b-unsatu- rated aldehydes only on the carbonyl group to form phosphorus- containing oligomers 53 with retention of the C=C bond.39 O P RCH CHCHO+Cl O 4 O O O P CH2 CH2 P RCH CHCHO RCH CHCH O O n Cl 53 Cl R=H, Me.The distinction between the reactions of functionalised ketones 42 and 47 with HP and those of aldehydes 20 with phosphoromonochloridites from the reactions of other hetero- dienes with HP is that these a,b-unsaturated carbonyl compounds react analogously to saturated carbonyl compounds,62 i.e., they enter into O-phosphorylation reactions rather than into the [4+1]-cycloaddition reactions. In turn, HP act simply as the corresponding acid chlorides rather than as P(III) derivatives, the coordination number of the phosphorus atom in which is increased immediately in the reaction with (hetero)dienes (as in the case of [4+1]-cycloaddition reaction). It should be noted that in the formation of oligomers 53 the coordination number of the phosphorus atom increases to four only upon subsequent dioxa- phospholane ring opening.The system of the C=C7C=O bonds can be a fragment of a more extended conjugated cyclic system, e.g., p-benzoquinone and its anthraquinone type analogues. A few reactions of these compounds with HP have been documented; however, since they occur with involvement of the entire quinone system, we do not discuss them here and restrict ourselves to referring to the corresponding reviews.63, 64 Numerous studies have been dedicated to reactions ofHPwith a,b-unsaturated acids and their derivatives. Several early studies have been reviewed.3 E Ya Levina, AMKibardin Reactions of HP with acids 54 can follow several pathways.Dihalides most often add to the acids 54 to form b-phos- phinicocarboxylic acid dichlorides 55.65 ± 77 R1CH CR2COOH+R3PCl2 ClPR3CHR1CHCOCl 54 55 R2 O R1, R2=H, Me; R3=Me, Et, Ph, p-MeC6H4, 2-thienyl, EtO, Me2(CCl3)CO, CH2Cl. Two different mechanisms of this reaction have been pro- posed. According to the first of them, HP react with unsaturated acids 54 following the [4+1]-cycloaddition pathway, which most often is considered as stepwise, to give linear phosphoranes, which undergo the intramolecular Arbuzov rearrangement resulting in the final products 55 (Scheme 2).65, 66, 68, 71, 73, 76 Scheme 2 O + Cl2PR3CHR1CR2 COH R3PCl2+CHR1 CR2C OH 54 O7 O OH O O R3Cl2P R3Cl2P R2 R2 R1 R1 Cl7 O O + ClPR3CHR1CHCOCl R3ClP R2 55 R2 O R1 Here the P(III) atom acts as a nucleophile, since the higher the nucleophilicity of the R3 substituent the more readily the reaction proceeds.76 However, this is also consistent with an alternative scheme involving no [4+1]-cycloaddition. According to the alternative scheme,73, 75 ± 79 (Scheme 3) the chlorine atom in the HP molecule and the hydroxyl group in the molecule of acid 54 exchange in the initial step.Then the hydro- phosphoryl compound 56 that formed is added to the unsaturated acid chloride 57 according to the Pudovik reaction.80, 81 Scheme 3 55 54+R3PCl2 R1HC CR2COCl+R3ClP(O)H 56 57 It is believed 82 that this scheme involves a preliminary stage of P-protonation of HP by the unsaturated acid 54.+ R1HC CR2COO7+HPCl2R3 54+R3PCl2 + R1HC CR2C O PHR3 [56+57] 55 7 O Cl Cl The presence of the hydrophosphoryl compound 56 and unsaturated acid chloride 57 in the reaction mixture is detected by NMR spectroscopy; sometimes it is possible to isolate these compounds. Moreover, compounds 56 and 57 do add to each other to give the final product 55, which has been proved in a model reaction.77, 82 Scheme 3 is also consistent with increase in the reaction rate with the increase in the electron-donor ability of the substituents at the phosphorus atom. Finally, which is also proved by kinetic studies,78 hydrogen chloride, which is always present in the reaction mixture, plays an important role in these reactions.In this connection it is noteworthy that diethyl acryloyl phosphite (58) protonated under the action of HCl rearranges into b-diethoxyphosphorylpropionyl chloride (59) in good yield. It is most likely that in this case the reaction follows the same sequence of stages as that discussed for the preceding scheme.82Reactions of halophosphines with conjugated heterodienes O O 7Cl H+ O CCH CH2 (EtO)2P (EtO)2POCCH CH2+HCl 58 [(EtO)2PHO+ClC(O)CH CH2] (EtO)2P(CH2)2COCl 59 O Thus, it can be considered as proved that P(III) acid dichlor- ides react with unsaturated acids 54 according to Scheme 3. However, this does not imply that the [4+1]-cycloaddition scheme is wrong and makes no contribution to the reaction under consideration. The rate of the reaction of the a,b-unsaturated acids 54 with HP decreases if acids 54 contain a- and/or b-substituents.71, 76, 79 Moreover, only acid chlorides 57 can be isolated as the reaction products of ethyl(dichloro)phosphine with several b-substituted a,b-unsaturated acids.76, 83, 84 57+EtClP(O)H EtPCl2+R1CH CR2COOH 54 R1=Ph, HOOC, trans-Cl, 2-furyl; R2=H.This fact is consistent with both reaction schemes proposed. According to Scheme 2, the phosphorus atom attack on the b-carbon atom of the acid 54 is the key reaction step. b-Sub- stituents in the acid molecule produce steric hindrances, which only results in the formation of acid chlorides 57 from the carboxylic acid 54. In the framework of Scheme 3, the attack on the carboxylic group in all cases is the initial and necessary stage.Obviously, if acid chlorides 57 contain substituents indicated in the scheme, the addition of ethyl(dichloro)phosphine to com- pounds 57 occurs very slowly due to both spatial shielding of the b-carbon atom and delocalisation of the electron density on the b-substituent and only these compounds can be isolated from the reaction mixtures.82 The reaction of dibutyl phosphorochloridite with acrylic acid at a reduced pressure gave only acryoyl chloride.77 The acids 54 react with methyl phosphorodichloridite to give unsaturated carboxylic acid chlorides 57 and oligomers 60.34 O O D 57+ MeOPCl2+54 n ClPCHR1CHR2CO 60 R1, R2=H, Me. Only oligomeric products were isolated in the reactions of acids 54 with compounds 4, 22 and 23.39, 85 O O O D (CH2)n PCHR1CHR2CO R1CH CR2COOH +ClP 54 R3 OCHR3(CH2)nCl m O 4, 22, 23 R1, R2, R3=H, Me; n=1, 2.It was assumed 34, 39, 85 that the reactions of the acids 54 and other (hetero)dienes with HP occur according to the [4+1]-cyclo- addition mechanism and are followed by the Arbuzov rearrange- ment of the adduct containing a pentacoordinate phosphorus atom with the formation of oligomers. However, it was shown later that no adduct containing a P(V) atom is formed in the reaction of ethyl phosphorodichloridite with acrylic acid without heating the reaction mixture at very high temperatures. Therefore, the formation of oligomeric products in the reactions of HP with a,b-unsaturated acids was explained by thermal conversions of acid dichlorides 55 that were formed.77 It should be noted that the formation of oligomers together with distillable addition prod- ucts 65, 68, 79 was also considered as a result of thermolysis of primary reaction products upon distillation.155 2,5-Dioxo-1,2-oxaphospholanes 61 are occasionally isolated along with the products 55 in the reactions of P(III) acid dichlor- ides with acids 54.86 ± 89 These originate from mixed anhydrides formed as byproducts from acid dichlorides 55 and acids 54 and are produced on heating during distillation.71, 86 R1CH CR2COOH +ClPR3CHR1CHR2COCl 7HCl 55 54 O R1CH CR2COPR3CHR1CHR2COCl O O 7R1CH CR2COCl O O O P R3 R2 61 R1 R1, R2=H, Me; R3=ClCH2, 2-thienyl, 5-chlorothienyl-2, Et, p-MeC6H4.Oxaphospholanes 61 were also obtained in the reactions of phosphorochloridites and phosphonochloridites with the acids 54. Their formation (as a result of elimination of alkyl chloride from intermediate addition products) competes with the forma- tion of acyclic compounds 62 analogous to acid dichlorides 55.75, 79, 82, 90 ± 92 Either or both linear acid chlorides 62 and oxaphospholanes 61 were isolated under different conditions. Since cyclisation of acid chlorides 62 in the course of distillation can also result in the formation of the phospholanes 61, it is unclear whether the latter are formed as a result of either or both direct reaction of HP with the acids 54 and thermolysis of the products 62.R4 R2 R4 R1 O7 + R1CH R3OP C CHCH CCOOH+R3OPCl 54 OH R2 Cl R4 PCHR1CHCOCl O R2 R3O R2 62 O O D 7R3Cl R1 P R4 R3O O R2 Cl O R1 P 7R3Cl O R4 61 R1=H; R2=H, Me; R3=Et, Me2(CCl3)C, (ClCH2)2CH, ClCH2CH2, CCl3 ; R4=Et, Ph, MeO, PhO, Me2(CCl3)CO, MeEt(CCl3)CO. Only linear products analogous to compounds 62 are formed in the reactions of alkyl phosphorochloridothioites with acrylic acid.93 This is associated with the lesser tendency of the AlkS7P group to undergo dealkylation as compared to the AlkO7P group (see monograph 94). Et Et CHCOOH P(CH2)2COCl PCl+CH2 RS RS O R=Et, Prn, Pri, Bun, Bui. Reaction of dichlorophosphines with cis-b-chloroacrylic acid (63) occurs with elimination of HCl from intermediate com- pounds 64 to give unsaturated acid dichlorides 65.76156 R H H RPCl2+ PCHClCH2COCl COOH Cl Cl 63 O 64 R PCH CHCOCl Cl O 65 R=Me, Et, Ph.The products of reactions of dialkyl(chloro)phosphines with a,b-unsaturated acids and their derivatives were not isolated and characterised after treatment with alcohols. For instance, dieth- yl(chloro)phosphine reacts with a,b-unsaturated acids or their chlorides to form crystalline adducts 66, the structure of which is not established. Treatment of these adducts with alcohols gave phosphine oxides 67.47, 73, 76, 95 ± 97 R4OH [66] Et2PCl+R1R2C CR3COX R1 O C Et2PCR2CHR3 OR4 67 O R1, R2, R3=H, Me; R4=Me, Et, Pr, Bu, C5H11; X=OH, Cl. Treatment of the adducts 66 formed from acids (X=OH) and acid chlorides (X=Cl) with ethanethiol instead of alcohols gives phosphine oxides 68 containing a thioester group and phosphine sulfides 69, respectively.96 CR2COX Et2PCl+R1HC [66]O X=OH Et2PCHR1CHC SEt EtSH 68 R2 O O X=Cl Et2PCHR1CHC OEt 69 R2 S R1, R2=H, Me; X=OH, Cl.Reactions of alkyl acrylates 70 with diethyl(chloro)phosphine result in different products depending on the reaction conditions. For instance, polymers or oligomers 71 containing a quaternary phosphonium unit in the main chain are formed quantitatively in a hexane solution after 30 h. However, monomeric phosphine sulfides 72 can be isolated in high yields from the reaction mixtures after 7 h following treatment with butanethiol.The adducts formed in the reaction conducted in ethyl bromide (or in hexane, in the presence of HCl) and then treated with methanol decom- pose to give phosphine oxides 73.98 These contradictory results require additional proof. Et COOR + n Cl7 C6H14 PCH2CH 30 h Et n 71 Et2PCl+CH2 CHCOOR 70 Et2P(CH2)2COOR 1. C6H14, 7 h 2. BuSH 72 S + EtBr or C6H14, HCl Et2ClP(CH2)2COOR Cl7 Et2PCl MeOH 70 Et2P(CH2)2COOR MeOH 73 O R=Me, Et, Bu. E Ya Levina, AMKibardin It is believed that the mechanisms of reactions of unsaturated acids and their derivatives with diethyl(chloro)phosphine are the same or similar to those of their reactions with P(III) acid dichlorides.47, 73, 76, 95 ± 97 However, it is likely that higher nucleo- philicity of the phosphorus atom in diethyl(chloro)phosphine than in P(III) acid dichlorides favours [1+4]-cycloaddition. This assumption is confirmed by the fact that methyl acrylate readily reacts with diethyl(chloro)phosphine 98 where the formation of an intermediate hydrophosphoryl compound is impossible.A similar mechanism was proposed 97 for the reaction of diethyl(chloro)phosphine with acryloyl chloride. O7 + Et2PCl +H2C CHCOCl C Et2PCH2CH Cl Cl Cl EtOH Et2P(CH2)2COOEt + O Et2PCl7 O It is confirmed by preparative isolation of 5-chloro-2-ethoxy- 2-oxo-3H-1,2-oxaphosphole from the reaction of diethylphos- phorochloridite with acryloyl chloride.97 Cl O EtO P (EtO)2PCl+H2C CHCOCl 7EtCl O Other monochlorophosphines react with the acids 54 analo- gously to diethyl(chloro)phosphine.Treatment of initially formed adducts of unknown structure with alcohols results in their decomposition and the formation of phosphine oxides 74.99, 100 R4OH [adduct] EtPR3CHR1CHCOOR4 R3EtPCl +54 O 74 R2 R1=H, Ph; R2=H, Me; R3=(CH2)2CN, CH2CH(Me)CN, CH2CH=CH2, 2-thienyl; R4=Me, Et. Acrylamide and methacrylamide (compounds 75a and 75b, respectively) react with various HP analogously to the corre- sponding acids. Nitriles 76a,b resulting from dehydrochlorination of imidoyl chlorides 77a,b are the major reaction prod- ucts.76, 99, 101 ± 110 O R1 C PCl+CH2 CNH2 R2 R3 75a,bO7 NH R1 ClO R1 + CNH2 P(Cl)CH2CR3 P R2 R2 R3 Cl O O R1 R1 D N NH PCH2CHR3C PCH2CHR3C 7HCl R2 R2 76a,b 77a,b R1, R2=Me, Et, Ph, Bui, (CH2)2CN, CH2CH(Me)CN, CH2CH=CH2, O ; OC(CCl3)Me2, OBu, OCH(CH2Cl)2, p-MeC6H4, CH2Cl, Cl, O R3=H(a), Me (b).In the presence of acetic acid, HP react with primary and tertiary acrylamides to form compounds 78 with the amide group remaining intact.99, 100, 111, 112Reactions of halophosphines with conjugated heterodienes AcOH O R1R2PCl +H2C CR3C R1R2PCH2CHR3C O 7AcCl NR4R5 78 O NR4R5 R1=Me, Et, Ph, p-MeC6H4; R2=Cl, Et, 2-thienyl, EtOOC(CH2)2, Me2(CCl3)CO; R3=H, Me; R4, R5=H, Et, Pri. In some instances, distillation of products 78 is accompanied by elimination of hydrogen chloride (probably, due to thermo- lysis) and the formation of 2,5-dioxo-1,2-azaphospholanes 79.112 R4 O D N O 78 P 7HCl R1 R3 79 R1=Et, Ph;R2=Cl; R3=H, Me;R4=H, Bu,CH2=CHCH2; R5=H.Halophosphines react with several other functional deriva- tives of acids in a somewhat different pathway. Reactions of dialkyl(chloro)phosphines with alkyl perfluoromethacrylates 80 occur with a characteristic replacement of a fluorine atom at the C=C bond to form an intermediate 81. Then, the chlorine atom and a fluorine atom in compound 81 are exchanged. The reaction results in the formation of phosphines, which usually undergo cyclisation into 5-oxo-5H-1,2-oxaphospholes 82 in the isola- tion.113 Cl C C(CF3)CO2R2 R12 PCl+F2C R12 P C(CF3)CO2R2 80 F F 81 R1 F O O P C C(CF3)CO2R2 R12 P 7R2F F Cl R1FCl CF3 82 R1=Et, R2=Me; R1=Pri, R2=Et.b-Hydroxy derivatives 83 52 and b-amino derivatives 51, 60, 114 of esters of a,b-unsaturated carboxylic acids 84 (enol tautomers of acetoacetic esters and b-aminocrotonates, respectively) react with P(III) monochlorides in the presence of triethylamine to undergo O- or N-phosphorylation without affecting the double bonds. The reaction occurs due to the presence of an acidic proton of the hydroxyl or amino group in the heterodiene and results in unsaturated phosphites 85 and 86, respectively. Et3N CHCOOMe R1R2PCl +HOCMe 7Et3N. HCl 83 CHCOOMe R1R2POCMe85 R1, R2=EtO, Et2N. Et3N 7Et3N. HCl (R1O)2PCl +R2HNCR3 Ê R4COOEt 84 CR4COOEt (R1O)2PNR2CR3 86 R1=Et, Pr; R2, R3=H, Me; R4=H, CN.Acrylates and methacrylates react with dialkyl phosphoro- chloridites in a similar way, i.e., without affecting the heterodiene system. The reactions result in mixed anhydrides 87, which are the products of O-phosphorylation.115 CH2 (R1O)2PCl+KOCCR2 CH2 (R1O)2POCCR2 7KCl O 87 O R1=Et, Pr, Bu; R2=H, Me. 157 a,b-Acetylenic acids also react with HP analogously to a,b- unsaturated acids. For instance, propiolic acid (88) 70, 76, 116 ± 118 reacts with dichlorophosphines to form unsaturated acid dichlor- ides 89 similar to the acid dichlorides 55. RClPCH CHCOCl RPCl2+HC CCOOH 88 O 89 R=Et, ClCH2, Ph. When alkylphosphonochloridites were used in this reaction instead of dichlorophosphines, the products 90 were not isolated but rather were directly converted into diesters 91 by treatment with ethanol or into thioesters 92 by treatment with butanethiol.An attempt at distilling compound 90 [R1=Ph, R2= Me2(CCl3)C] resulted in isolation of a product of its cyclisation, viz., 2,5-dioxo-5H-1,2-oxaphosphole 93, which is an unsaturated analogue of phospholane 61.119 O R1(R2O)PCl +88 CHCOCl R1(R2O)PCH90 O EtOH CHCOOEt R1(R2O)PCH91 O 7HCl BuSH R1(R2O)PCH CHCOSBu 92 CCl3 . R1=Et, Ph, p-MeC6H4; R2=Me2(CCl3)C, O D 90 O P 7R2Cl Ph O 93 R1=Ph, R2=Me2(CCl3)C. EtPCl2+MeC Treatment of an intermediate product 94, obtained in the reaction of tetrolic acid (95) with ethyl(dichloro)phosphine, with acetic anhydride resulted in the 5H-1,2-oxaphosphole 96.120 CCOOH 95 O O Et Ac2O O PCMe CHCOCl Me P 72 AcCl Cl Et 94 O 96 III.Reactions with azadienes Investigation of reactions of HP with dienes containing one or several nitrogen atoms in the conjugation chain was begun recently. Most studies in this field were carried out in the late 1980's and the early 1990's. Currently, it is azadienes (among different types of heterodienes that react with HP) that attract the greatest attention of researchers. However, whereas the reactions of azadienes with dienophiles containing no phosphorus atom have long been studied and are well known (see, e.g., a review 121), almost no overview information concerning their reactions with phosphorus-containing dienophiles is available.1. Reactions with a-diimines Characteristic of a-diimines capable of adopting s-cis-conforma- tion 122 are [4+1]-cycloaddition reactions.121, 123 Diimines 97a,b react with ethyl(dichloro)phosphine to form cyclic diaminophosphonium salts 98a,b. It is convenient to isolate158 them as the corresponding tetrafluoroborates 99a124 or phosphine oxides 100a,b.125, 126 R1 NR2 +EtPCl2 R1 NR2 97a,b R1 Et3O+BF¡4 7Et2O, EtCl R1 R1 H2O, 2 Et3N 72Et3N. HCl R1 100a,b R2 R1=Me, R2=Bu (a); R1±R1=(CH2)4 (b). Reactions of HP with N,N0-dibutylbiacetyldiimine 97a occur to give 1,3-dihydro-1,3,2-diazaphospholes 101.127 The reaction occurs with dealkylation, as in the case of dienic hydrocarbons and a-dicarbonyl compounds. NBu Me +EtOPRCl 7EtCl NBu Me 97a R=Cl, EtO, Et2N.Reaction of 2-chlorobenzo-1,3,2-dioxaphosphole 102 with the diimine 97a resulted in the phosphorane 103.128 An analogous product was obtained in the reactions of compound 102 with 1,3- dienes. O 97a+Cl P O102 The reaction of the diimine 97a with phosphorus trichloride results in resinification only.128 However, its reaction with phos- phorus tribromide in the presence of an organic base follows the [4+1]-cycloaddition pathway, and the reaction product is a tricoordinate phosphorus derivative, viz., 2-bromo-1,3-dihydro- 1,3,2-diazaphosphole 104.128 Me Et3N 97a+PBr3 72 [Br] Me 104 Bu 2,3-Diphenyl-5,6-dihydropyrazine (105) reacts with PBr3 in an analogous manner to give a bridged product 106.129 PBr Ph N Et3N +PBr3 72 [Br] N Ph 105 R2 R1 Cl N+P N R1 Et Cl7 98a,b R2 R2 N Cl +P N Et BF¡499a R2 R2 N O P Et N Bu Me N P N Me 101 Bu Bu Me N O P N Me O Cl Bu 103 Bu NPBr N N Ph N Ph 106 OR Reactions of glyoxal diimines 107 with PCl3 in the presence of a base also follow the [4+1]-cycloaddition pathway to form tricoordinate phosphorus derivatives, viz., 2,4-dichloro-1,3-dihy- dro-1,3,2-diazaphospholes 108 as the reactions products.Their formation can be rationalised by elimination of HCl from an intermediate cyclic adduct.128, 130, 131 R NR N +PCl3 PCl3 N NR 107 R R=cyclo-C6H11, But. Diimine 97a reacts with tetraethylphosphorodiamidochlori- dite in the presence of triethylamine 132 to give a mixture of two 1,3-dienes, viz., 1,3,2-diazaphospholane 109 with an exocyclic system of conjugated double bonds and bicyclic compound 110. It was assumed132 that N-phosphorylation of both imino groups followed by dehydrochlorination results in an intermedi- ate diene 111 as a mixture of s-cis- (111a) and s-trans-conformers (111b) (analogously to the formation of phosphorylated enamines from imines of saturated aldehydes 128, 133).Then, the reactions of cis- and trans-forms of compound 111 follow different pathways. Elimination of tris(diethylamino)phosphine from the s-cis-form 111a results in monocyclic diazaphospholane 109. s-trans-Isomer 111b is converted into diazadiphosphabicyclooctadiene 110 as a result of elimination of two molecules of diethylamine.The bicyclic compound 110 was isolated in an individual form, while the formation of phospholane 109 was proved by spectroscopic methods.132 2Et3N 97a+2 (Et2N)2PCl 72Et3N. HCl Bu H2C NP(NEt2)2 H2C NP(NEt2)2 111a Bu7P(NEt2)3 Bu H2C N PNEt2 N H2C109 Bu Reactions of PCl3 and phosphorodichloridites with diimines containing an exocyclic system of double bonds in the presence of a base result in bicyclic 1,3,2-diazaphospholane derivatives con- taining a cyclohexadiene fragment. These reaction products are bicyclic analogues of the phospholanes 109.126 2B 72B. HCl NR1+R2PCl2 NR1 R2=Cl, AlkO, Alk2N. Reaction of 2,3-dimethyl-5,6-dihydropyrazine (112, a cyclic analogue of diimine 97) with dialkyl(chloro)phosphines (alkyls were not specified) in the presence of a base occurs with the formation of monophosphorylated tetrahydropyrazine 113 con- taining one exocyclic methylene group,129 which means that the reaction proceeded only on one C=N bond.E Ya Levina, AMKibardin R H N Et3N PCl 7Et3N. HCl N Cl 108 R Bu NP(NEt2)2 H2C CH2 (Et2N)2PNBu 111b 72Et2NH Bu N PNEt2 Et2NP N 110 BuR1 NPR2 N R1Reactions of halophosphines with conjugated heterodienes Me N N Me Me N B + +R2PCl H2C Me N N 7B . HCl Me Cl7 113 PR2 PR2 N 112 R=Alk. Reactions of dialkyloxalimidates 114 with dichlorophos- phines and phosphorodichloridites occur with elimination of two mobile protons to form derivatives of 1,3,2-diazaphospholes 115,134 i.e., compounds 114 undergo cyclisation without partic- ipation of the heterodiene system as such.NH R1O R1O N Et3N PR2 +R2PCl2 7Et3N. HCl N R1O R1O NH 115 114 R1=Me, Et; R2=Bu, Ph, p-MeC6H4, p-ClC6H4, PhO, p-MeC6H4O, p-ClC6H4O. 2. Reactions with azines It was argued 39, 135 that the reactions of cyclic phosphorochlor- idites with azines and with other conjugated systems proceed as [4+1]-cycloadditions; the 1,3,2-dioxaphospholane that forms undergoes ring opening according to the second stage of the Arbuzov reaction to give 3,5-dihydro-1,2,4-diazaphospholes. However, it was established later that dioxaphospholanes 4 and benzodioxaphospholes 102 react with acetonazine (116) in the presence of triethylamine to form 3,5,5-trimethyl-4,5-dihydropyr- azolophosphites 117.Initially, electrophilic attack of the phos- phorus atom on the nitrogen atom of ketazine 116 occurs, as in reactions of HP with imines.128, 133 The immonium salt 118 that forms undergoes the pyrazoline condensation to give the final product 117.136 O R PCl+Me2C N N CMe2 116 O4, 102 Me N CMe2 O N Et3N O + R P N P R 7Et3N. HCl O O N CMe2 Cl7 118 Me Me 117 R=(CH2)2 (4), o-C6H4 (102). Reaction of diethyl phosphorochloridiite with acetonazine 116 in the presence of triethyl phosphite only resulted in com- pound 119, the product of addition of diethyl phosphite to one of the C=N bonds.The formation of pyrazolinophosphite 117 was detected only by spectroscopic methods.137 N CMe2 Me2CHN 116+(EtO)2PCl+(EtO)3P O (EtO)2P 119 Reactions of asymmetrical azines 120 with the phosphite 102 occur with symmetrisation of the azine in the initial step.138 This process is accelerated by diethyl phosphorochloridite (a Lewis acid) present in the reaction mixture. The symmetrical azine 121 that forms then reacts with phosphite 102 analogously to aceto- nazine 116 to form pyrazolinophosphites 122.137 159 PhCH N CR1CHR22 N 120 [PhCH N N CHPh + N CR1CHR22 ] R22 CHCR1 N121R1 N O Et3N R2 N P 121+102 7Et3N. HCl O 122 R2 CHR2 R1 2 R1, R2=H, Me. 3. Reactions with a-enimines It was postulated 135 that a-enimines, as well as other 1,3-dienes and heterodienes, react with 1,3,2-dioxaphospholanes according to the [4+1]-cycloaddition pathway.However, the absence of any specific information on the products of these reactions makes one treat such an argument with precaution. Reactions of enimines 123 with phosphorus trichloride or ethyl(dichloro)phosphine in the presence of an organic base 139 ± 141 result in 1,2-dihydro-1,2-azaphosphorinines 124. R2 R3 R2 R3 + PCl N N PCl2 Et3N CH2R1 Cl7 CH2R1 7Et3N. HCl R1 R1 H 123 125 H R2 R3 N Et3N R2 N+ P R3 Cl7 PCl CHR1 7Et3N. HCl R1 R1 R1 H 127 126 H R2 R3 P N R1 R1 124 R1=Et, R2=Bun, R3=Cl, Et; R1=H, R2=But, R3=Cl. The reaction begins with electrophilic attack of the phospho- rus trichloride or ethyl(dichloro)phosphine on the lone electron pair of the nitrogen atom of enimine 123, which results in an immonium salt 125 containing a P7Nbond and a positive charge delocalised over the system of conjugated bonds.Then the first triethylamine molecule eliminates an acidic proton from the d- carbon atom of the salt 125 to give an intermediate adduct 126. This reaction is analogous to that leading to phosphorylated enamines from imines of saturated aldehydes and is similar to reactions resulting in the formation of heterodienes 111 and 118.128, 133 Then, the closure of the six-membered ring containing the P7C bond occurs to form the intermediate 127 as a result of the intramolecular interaction of the phosphorus atom with the d-carbon atom on which partial negative charge is localised due to conjugation with the lone electron pair of the nitrogen atom.The second molecule of the base eliminates hydrogen chloride from the intermediate 127, which results in 1,2-dihydro-1,2-azaphosphor- inine 124 as the final product. It was assumed 141 that in this case the [4+1]-cycloaddition reaction cannot compete with the attack on the nitrogen atom due to kinetic control because of asymmetry of the enimine molecule. In addition, the nitrogen atom in a-enimines is more nucleophilic than that in a-diimines, since two electronegative azomethine groups in the latter interact with each other to weaken the nucleophilicity of the nitrogen atom in each of them.160 Butadienylamine 123 0, the possible existence of which has been discussed,121 is the minor tautomer of the enimine 123 and can also contribute to the formation of azaphosphorinine 124. R1CH2CH CR1CH NR2 123 R3PCl2, 2Et3N 124 CHNHR2 R1CH 72Et3N.HCl CHCR1 123 0 The similarity of behaviour of the enimines 123 and their oxygen analogue, crotonaldehyde, in reactions with HP in the presence of an organic base is obvious. These asymmetrical heterodienes can be represented by a general formula RCH2CH=CRCH=Y (Y=O, NAlk). In contrast to their sym- metrical analogues, a-diimines and a-diketones, the above-men- tioned heterodienes undergo phosphorylation at the nucleophilic centre Y rather than enter into [4+1]-cycloaddition reactions leading to the increase in the coordination number of the phosphorus atom.However, reactions of functionalised enimines, e.g., cinna- maldehyde o-hydroxyanils 128, with P(III) chlorides proceed by [4+1]-cycloaddition, which precedes O-phosphorylation of the hydroxyl group or follows it. This reaction results in fused five- membered heterocyclic compounds 129 containing the pentacoor- dinate phosphorus atom.142 R3 OH Ph R2 O P N Ph R2R3PCl, Et3N 7Et3N. HCl N R1 128 R1 129 R1=Me: R2±R3=MeN(CH2)2NMe; R1=H, Me; R2=R3=Me, Ph, OMe; R2±R3=O(CH2)2O, o-C6H4O2. As in the case of reactions of dienes with HP,4± 7 tricyclophos- phoranes 129 can have structural isomers as a result of the double bond migration in azaphosphole (prototropism), and geometric isomers due to pseudorotation.143 For instance, if amino groups are bonded to the phosphorus atom, the primary reaction product 129 undergoes partial [for R27R3=MeN(CH2)2NMe] or complete (for R2=R3=NMe2) rearrangement into a thermody- namically more stable isomer 130.The latter is isolated as the final reaction product of compounds 128 with bis(dimethylamino)- chlorophosphine. R3 Ph R2 O P 129 N R1 130 R1=Me: R2±R3=MeN(CH2)2NMe, R2=R3=NMe2. Reaction of diisopropylamino(chloro)phosphenium tetra- chloroaluminate with crotonaldehyde dimethylhydrazone (131) 144 proceeds as a classical [4+1]-cycloaddition and is analogous to reactions with 1,3-dienes 28 ± 30 and a-dicarbonyl compounds;27 this results in cyclic phosphonium salt 132, which is hydrolysed to give 2-oxo-1,2-dihydro-3H-1,2-azaphosphole 133.+ MeHC CHCH N NMe 131 2+Pri2NPCl AlCl¡4 NMe2 NMe2 N O Cl H2O P N+PNPri NPri 2 2 AlCl¡4Me 133 Me132 E Ya Levina, AMKibardin Aminoenimines 134 react with P(III) chlorides in the presence of triethylamine. Their cyclisation occurring with elimination of two HCl molecules results in 1,2-dihydro-1,3,2-diazaphosphor- inines 135,145 like the reaction of dialkyloxalimidates with dichlor- ophosphines leading to derivatives of 1,3,2-diazaphospholes (see Section III.3). R5 NH R1 R1 2Et3N P N +R5PCl2 72Et3N. HCl NHR4 N R2 R2 R4 R3 135 R3 134 R1=H, Me; R2, R4=Ph, p-MeC6H4, cyclo-C6H11; R3=H, Ph; R5=Ph, Cl.This is an example of synthetic procedures used for the preparation of nitrogen heterocycles based on azadienes, which has been discussed in detail.146 ± 148 Reactions of HP with enamino ketones 47, which are nitrogen analogues of b-diketones, were considered in Section II.2.a; however, these can also be discussed in this Section. Indeed, compounds 47 can also exist in the imine form 136.149 R1C R1CCR2 CNHR4 CR2C NR4 OH O CH2R3 CH2R3 136 47 Tautomers 47 and 136 can react with HP concurrently to give the same final products. Finally, reactions of HP with heteroenyne C=C7C:N systems are formally similar to the reactions of HP with enimine C=C7C=N systems considered in this Section. Acrylo- and methacrylonitriles 137 react with alkylphosphonochloridothioites in the presence of proton donors according to the scheme:46 R2P(SR3)CH2CHR1CN YXH 7YCl H2C CR1CN+R2ClPSR3 137 X R1=H, Me; R2=Et, Ph; R3=Et, Pr, Bu; X=O, S; Y=H, Ac.These reactions are analogous to those of compounds 30 with a,b-unsaturated ketones in the presence of proton donors and follow the same mechanism (see Section II.2.a). 4. Reactions with azoalkenes 3-oxo-2,3-dihydro-4H-1,2,3-diazaphospholes Dichlorophosphines react with azoalkenes 138 to give [1+4]- cycloaddition products, viz., cyclic phosphonium salts 139. Hydrolysis of compounds 139 results in a mixture of stereo- isomers, viz., 140.143, 150 R3 NR3 N H2O Cl N +R4PCl2 N +PCl7 7HCl R4 R1 R1 R2 138 R2 139 R3 R3 OH N N O N N P P 7H2O R4 OH R4 R1 R1 R2 R2 140 R1=Me, Ph, Bn; R2=H, Me, Ph; R3=Me, Ph; R4=Me, Ph, But, Bn, o-ClC6H4, p-MeC6H4, p-MeOC6H4. Reaction of the azoalkene 138 (R1=R3=Me, R2=H) with PhPCl2 results in a mixture of products with different position ofReactions of halophosphines with conjugated heterodienes the double bond, viz., 2,3-dihydro-4H- and 1H-1,2,3-diazaphosp- holes (141 and 142), as is observed for isomerisation of phosphor- anes 129 into phosphoranes 130.Me Me H N N O O N N 138+PhPCl2 + P P Ph Ph Me Me 142 141 An analogous double bond migration in the ring is also often observed for phospholes formed in reactions of HP with dienic hydrocarbons.5, 6 The azoalkenes 138 react with PCl3 to give diazaphosphole intermediates; treatment of the latter with alcohols causes elimi- nation of a hydrazine fragment to form b-oxo phosphonates 143.151 R3 N 3R4OH N R1CCHR2P(OR4)2 138+PCl3 PCl3 7H2N NHR3 O R1 R2 O 143 R1=R2=R3=Me, Ph; R4=Me, Et, Ph; R1±R2=(CH2)4; R3=R4=Me.IV. Reactions with oxazadienes Reactions of HP with 1,3-oxaza-1,3-dienes (N-acylketimines) 144 follow the [4+1]-cycloaddition pathway to give 5-oxo-4,5-di- hydro-1,3,5-oxazaphospholes 145.152 CF3 F3C SO2 or Ac2O O N P R1CN C(CF3)2+R2PCl2 144 R2 O O R1 145 R1=CF3, OEt; R2=Me, OEt. The reaction proceeds only in the presence of acetic anhydride or sulfur dioxide. Trialkyl phosphites 153 react according to the same scheme without these additives.Halophosphines can be added to 1,4-oxaza-1,3-dienes (mono- imines of a-diketones) as has been exemplified in benzyl o-hydrox- yanil (146).154, 155 The latter reacts with different HP with the formation of intermediate compounds 147 and 148, which are then transformed into polycyclic phosphoranes 149 or diphosph- ethanes 150 and 151. The key stage of this reaction is [4+1]-cycloaddition, which precedes phosphorylation of the hydroxyl group and/or follows it. Ph Ph +R1R2PCl O NC6H4OH-o 146 Ph Ph Et3N 7Et3N. HCl O NC6H4OPR1R2-o 147 C6H4OH-o Ph N P Cl R2 R1 Ph O148 161 [147] Ph N Et3N [148] P OR2 7Et3N. HCl O Ph R1 149 R1, R2=Me, Ph, OMe, OPh, OCH2Ph, NMe2; R1±R2=O(CH2)2O, O(CH2)3O, S(CH2)3S, o-C6H4O2, MeN(CH2)2NMe.Ph Cl N Ph P P O O Et3N N N Ph 146+Ph N Ph N Ph 7Et3N. HCl P POEt 150 OEt Ph N Cl Ph O P O P 2Et3N N N Ph Ph 2 146+Ph N N Ph 72Et3N. HCl O O P P Ph N Cl Ph 151 Biacetyl acylmonohydrazones 152 react analogously. As in the case of oxazadiene 146, the final bicyclic dioxadiazaphosphor- anes 153 result from [4+1]-cycloaddition and O-phosphoryla- tion.154 R1 N NHCOR1 N Me Me Et3N N +R22 PCl O P 7Et3N. HCl Me R2 O Me O152 R2 153 R1=Me, Ph, Bn; R2=Me, Ph. V. Conclusion Though certain judgements on the mechanisms of the reactions considered in this review cannot be called indisputable, summing up the above information makes is possible to state that, despite the variety of types of HP, heterodienes, and the products that formed, the initial steps of these reactions belong to three types: (1) 1,4-addition reaction, in particular, [4+1]-cycloaddition reaction (concerted or stepwise); (2) reaction involving one of the multiple bonds of a hetero- diene; and (3) reaction without particiption of multiple bonds.Differences in the structure of the final products formed in each of these three types of reactions are determined by further transformations of the primary reaction products. Most often the reactions under consideration follow the [4+1]-cycloaddition pathway to give usually cycloadducts con- taining the pentacoordinate phosphorus atom. In some instances, it is possible to isolate them or to detect them by spectroscopic methods.However, [4+1]-cycloaddition reactions of HP with a- diketones, their diimines and mono-o-hydroxyanils, a,b-unsatu- rated ketones, azoalkenes and 1,2-oxazadienes result in various five-membered phosphole-type heterocycles containing P(III), P(IV) or P(V) atoms as the final products. a,b-Unsaturated carboxylic acids and their derivatives also react withHPaccording162 to this scheme (not always) to give acyclic compounds. In some instances, linear products are also formed from unsaturated carbonyl compounds. If only one of two double bonds of the heterodiene system is involved in the reaction with HP in the initial stage, the second double bond can be either involved in further transformations or left intact.This occurs in the reactions of HP with different types of heterodienes in the presence of organic bases, e.g., in the reactions with crotonaldehyde and enimines, which are nitrogen analogues of a,b-unsaturated aldehydes. Unlike their symmetrical analogues (a-diketones, a-diimines), enimines do not enter into [4+1]-cycloaddition reactions, which result in an increase in the coordination number of the phosphorus atom, sinceHPreact with their nucleophilic centre (the oxygen or nitrogen atom) thus phosphorylating it. Reactions of several cyclic a-diimines (dihy- dropyrazines) and azines proceed analogously involving one C=N bond. The same occurs in the reaction of tetraethylphos- phorodiamidochloridite with acyclic a-diimine: each of the C=N bonds reacts independently.Usually, the alkyl group adjacent to the double bond is inert and left intact in the first reaction stage; the involvement of such a group in the reaction is one of the peculiarities of the reactions of HP wuth heterodienes in the presence of an organic base. Under the action of a base, proton abstraction from these groups occurs, which results in the for- mation of a new double bond and/or in the closure of phosphorus- containing rings. Reactions involving one of the double bonds can also occur in the absence of a base and, correspondingly, without dehydrohalogenation. For instance, unsaturated aldehydes react with 2-chloro-1,3,2-dioxaphospholane at the C=O bond, which results in the formation of phosphorus-containing oligomers.Perfluoromethacrylates react with dichlorophosphines only at the C=C bond leaving the ester group intact. It is possible to consider the reactions of a,b-unsaturated acids with HP as those occurring at the C=C bond if they proceed via exchange of the chlorine atom by the hydroxyl of the carboxy group followed by reaction between hydrophosphoryl compound and unsaturated acid chloride that formed according to the Pudovik reaction. Alkylphosphonochloridothioites react analogously with a,b- unsaturated ketones in the presence of proton donors. Finally, reactions of several heterodienes (capable of eliminat- ing one or two protons in the presence of an organic base) and salts of a,b-unsaturated carboxylic acids with HP proceed without participation of conjugated bonds at all.These reactions result in the formation of phosphorylation products. In these cases, HP behave as ordinary acid halides. If the phosphorus atom carries appropriate substituents, the phosphorylated heterodiene that formed can enter into further intramolecular reactions. These involve the diene system and result in its cyclisation. Acyclic P(III) derivatives initially obtained from HP and enol tautomers of b-diketones or their nitrogen analogues react in such a manner. References 1. US P. 2 663 738; Chem. Abstr. 49 7602 (1955) 2. US P. 2 663 737; Chem. Abstr. 49 7601 (1955) 3. A N Pudovik, V K Khairullin Usp. Khim. 37 745 (1968) [Russ. Chem. Rev. 37 317 (1968)] 4. K D Berlin,D M Hellwege, in Topics in Phosphorus Chemistry Vol.6 (EdsMGrayson, E J Grillith) (New York: Interscience 1969) p. 1 5. L D Quin Trivalent Phosphorus Compounds as Dienophiles in 1,4-Cycloaddition Reactions (New York: Academic Press, 1967) 6. L D Quin The Heterocyclic Chemistry of Phosphorus: Systems, Based on the Phosphorus7Carbon Bond (New York: Wiley, 1981) 7. F G Mann The Heterocyclic Derivatives of Phosphorus, Arsenic, Antimony and Bismuth (New York: Wiley-Interscience, 1970) 8. N S Isaacs, C N El-Din Synthesis 967 (1989) 9. R Boukherroub, E Garrigues, G Manuel Phosphorus, Sulphur, Silicon Relat. Elem. 105 101 (1995) 10. P Le Floch, F Mathey Tetrahedron Lett. 30 817 (1989) 11. P Le Floch, L Ricard, F Mathey Polyhedron 9 991 (1990) E Ya Levina, AMKibardin 12.P Le Floch,D Carmichael, F Mathey Organometallics 10 2432 (1991) 13. H T Teunissen, F Bickelhaupt Tetrahedron Lett. 33 3537 (1992) 14. H T Teunissen, F Bickelhaupt Phosphorus, Sulphur, Silicon Relat. Elem. 76 335 (1993) 15. H T Teunissen, J Hollebeek, P J Nieuwenhuizen, B L M van Baar, F J J de Kanter, F Bickelhaupt J. Org. Chem. 60 7439 (1995) 16. H T Teunissen, F Bickelhaupt Organometallics 15 794 (1996) 17. J B Conant J. Am. Chem. Soc. 39 2683 (1917) 18. R S Edmundson, in Comprehensive Organic Chemistry Vol. 3 (Eds D B Barton,WD Ollis) (Oxfoed: Pergamon Press, 1979) 19. T Kh Gazizov Reaktsiya Arbuzova (Arbuzov Reaction) (Kazan: Kazan State University, 1991) 20. V A Kukhtin, I P Gozman Dokl. Akad. Nauk SSSR 158 157 (1964) a 21. I P Gozman, V A Kukhtin Zh.Obshch. Khim. 37 881 (1967) b 22. Ya A Levin, I P Gozman, S G Salikhov Izv. Akad. Nauk SSSR, Ser. Khim. 2609 (1968) c 23. G Meggendorfer, R Schwarz, I Ugi Tetrahedron Lett. 21 2493 (1980) 24. A N Pudovik, V K Khairullin, Yu Yu Samitov, R R Shagidullin Zh. Obshch. Khim. 37 865 (1967) b 25. A N Pudovik, V K Khairullin, I I Kharitonova Izv. Akad. Nauk SSSR, Ser. Khim. 466 (1969) c 26. E S Klimov, A A Bumber, O Yu Okhlobystin Zh. Obshch. Khim. 53 1739 (1983) b 27. M R Maziere's, T C Kim, R Wolf, M Sanchez Bull. Soc. Chim. Fr. 127 79 (1990) 28. C K SooHoo, S G Baxter J. Am. Chem. Soc. 105 7443 (1983) 29. A H Cowley, R A Kemp, J G Lasch, N C Norman, C A Stewart J. Am. Chem. Soc. 105 7444 (1983) 30.A H Cowley, R A Kemp Chem. Rev. 85 367 (1985) 31. Br. P. 801 568; Chem. Abstr. 53 9251 (1959) 32. B A Arbuzov, A O Vizel', L I Shchukina, T A Zyablikova, R R Shagidullin Zh. Obshch. Khim. 46 2501 (1976) b 33. S Kh Nurtdinov, N V Dmitrieva, V S Tsivunin, T V Zykova, G Kh Kamai Zh. Obshch. Khim. 40 2189 (1970) b 34. K I Novitskii, N A Razumova, A A Petrov, in Khimiya Organi- cheskikh Soedinenii Fosfora (The Chemistry of Organic Compounds of Phosphorus) (Ed. S N Danilov) (Leningrad: Nauka, 1967) p. 248 35. K Bergesen Acta Chem. Scand. 19 1784 (1965) 36. A N Pudovik, V K Khairullin, R M Kondrat'eva Izv. Akad. Nauk SSSR, Ser. Khim. 2548 (1970) c 37. R Z Aliev, V K Khairullin Izv. Akad. Nauk SSSR, Ser. Khim. 2785 (1973) c 38. S Kh Nurtdinov, V I Savran, T V Litsova, T V Zykova, V S Tsivunin, in Khimiya i Tekhnologiya Elementoorganicheskikh Soedinenii i Polimerov, Kazan, 1980 (The Chemistry and Technology of Organoelement Compounds and Polymers, Kazan, 1980); article deposited at the ONITEKhim., Cherkassy, No.13KhP-D81 (1981) 39. N A Razumova, A A Petrov, A Kh Voznesenskaya, K I Novitskii Zh. Obshch. Khim. 36 244 (1966) b 40. N A Razumova, A A Petrov Dokl. Akad. Nauk SSSR 158 907 (1964) a 41. A Kh Voznesenskaya, N A Razumova Zh. Obshch. Khim. 38 1553 (1968) b 42. N A Razumova, L S Kovalev, A A Petrov Zh. Obshch. Khim. 38 126 (1968) b 43. L S Kovalev, N A Razumova, A A Petrov Zh. Obshch. Khim. 38 2277 (1968) b 44. K V Vatsuro, G V Mishchenko Imennye Reaktsii v Organicheskoi Khimii (Named Reactions in Organic Chemistry) (Moscow: Khimiya, 1976) 45.N I Rizpolozhenskii, V D Akamsin, R M Eliseenkova, R R Shagidullin Izv. Akad. Nauk SSSR, Ser. Khim. 2765 (1972) c 46. V D Akamsin, R M Eliseenkova, N I Rizpolozhenskii Izv. Akad. Nauk SSSR, Ser. Khim. 80 (1973) c 47. S Kh Nurtdinov, V S Tsivunin, T V Zykova, G Kh Kamai Zh. Obshch. Khim. 37 692 (1967) b 48. J von Seyerl, G Huttner Z. Naturforsch., B Chem. Sci. 35 1373 (1980) 49. F S Mukhametov, L V Stepashkina, E E Korshin,R R Shagidullin, N I Rizpolozhenskii Zh. Obshch. Khim. 53 1274 (1983) b 50. F S Mukhametov, E E Korshin Zh. Obshch. Khim. 54 2795 (1984) b 51. F S Mukhametov Zh. Obshch. Khim. 61 10 (1991) b 52. M I Kabachnik, P A Rossiiskaya,M P Shabanova, D M Paikin, L F Efimova, N M Gamper Zh. Obshch.Khim. 30 2218 (1960) bReactions of halophosphines with conjugated heterodienes 53. D M Malenko, L A Repina, A D Sinitsa Zh. Obshch. Khim. 54 2402 (1984) b 54. F S Mukhametov, E E Korshin, R R Shagidullin, N I Rizpolozhenskii Zh. Obshch. Khim. 53 2703 (1983) b 55. E E Korshin, F S Mukhametov Izv. Akad. Nauk SSSR, Ser. Khim. 1918 (1984) c 56. J von Seyerl, D Neugebauer, G Huttner Angew. Chem. 89 896 (1977) 57. J von Seyerl,D Neugebauer,G Huttner, C KruÈ ger, Y-H Tsay Chem. Ber. 112 3637 (1979) 58. S Kh Nurtdinov, I V Tsivunina, V I Savran, N M Ismagilova, T V Zykova, V S Tsivunin Zh. Obshch. Khim. 51 1549 (1981) b 59. F S Mukhametov, R M Eliseenkova, E E Korshin Zh. Obshch. Khim. 59 321 (1989) b 60.F S Mukhametov, R M Eliseenkova, E E Korshin Zh. Obshch. Khim. 56 2054 (1986) b 61. D M Malenko, N V Simurova, A D Sinitsa Zh. Obshch. Khim. 621426 (1992) b 62. E L Gefter,M I Kabachnik Dokl. Akad. Nauk SSSR 114 541 (1957) a 63. A A Kutyrev, V V Moskva Usp. Khim. 56 1798 (1987) [Russ. Chem. Rev. 56 1028 (1987)] 64. A A Kutyrev Zh. Obshch. Khim. 66 474 (1996) b 65. V K Khairullin Dokl. Akad. Nauk SSSR 162 827 (1965) a 66. V K Khairullin, R R Shagidullin Zh. Obshch. Khim. 66 289 (1996) b 67. V K Khairullin, A N Pudovik Zh. Obshch. Khim. 36 494 (1966) b 68. A N Pudovik, V K Khairullin, V N Eliseenkov Zh. Obshch. Khim. 37 455 (1967) b 69. V K Khairullin, T I Sobchuk, A N Pudovik Zh. Obshch. Khim. 37 710 (1967) b 70. V K Khairullin, V N Eliseenkov, A N Pudovik Zh.Obshch. Khim. 37 871 (1967) b 71. V K Khairullin, G V Dmitrieva, A N Pudovik, in Khimiya Organichskikh Soedinenii Fosfora (The Chemistry of Organic Compounds of Phosphorus) (Ed. S N Danilov) (Leningrad: Nauka, 1967) p. 40 72. V K Khairullin,R M Kondrat'eva,A N Pudovik Zh. Obshch. Khim. 38 288 (1968) b 73. T Kh Gazizov, A P Pashinkin, G V Dmitrieva, L L Tuzova, V K Khairullin, A N Pudovik Zh. Obshch. Khim. 42 1730 (1972) b 74. V K Khairullin, G V Dmitrieva, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1166 (1969) c 75. V N Eliseenkov, A N Pudovik, V K Khairullin Izv. Akad. Nauk SSSR, Ser. Khim. 1582 (1966) c 76. V K Khairullin, in Problemy Organicheskoi i Fizicheskoi Khimii (The Problems in Organic and Physical Chemistry) (Ed.Ya A Levin) (Kazan: Institute Organic and Physical Chemistry, Academy of Sciences of the USSR, 1971) p. 109 77. A N Pudovik, T Kh Gazizov, A P Pashinkin, V A Kharlamov Zh. Obshch. Khim. 45 2123 (1975) b 78. T Kh Gazizov, A P Pashinkin, V A Kharlamov, V I Kovalenko, A N Pudovik Zh. Obshch. Khim. 47 1226 (1977) b 79. A N Pudovik, V K Khairullin, M A Vasyanina Zh. Obshch. Khim. 37 411 (1967) b 80. I V Konovalova, L A Burnaeva Reaktsiya Pudovika (Pudovik Reaction) (Kazan: Kazan State University, 1991) 81. A N Pudovic, I V Konovalova Synthesis 81 (1979) 82. T Kh Gazizov, M A Vasyanina, A P Pashinkin, N P Anoshina, Z I Gol'dfarb, A N Pudovik Zh. Obshch. Khim. 41 1957 (1971) b 83. A N Pudovik, V K Khairullin, T I Sobchuk Izv.Akad. Nauk SSSR, Ser. Khim. 932 (1966) c 84. V K Khairullin,G V Dmitrieva, A N Pudovik Zh. Obshch. Khim. 37 1838 (1967) b 85. K I Novitskii, N A Razumova, A A Petrov Zh. Obshch. Khim. 36 1649 (1966) b 86. V K Khairullin, R R Shagidullin, T I Sobchuk, A N Pudovik, in Khimiya Organichskikh Soedinenii Fosfora (The Chemistry of Organic Compounds of Phosphorus) (Ed. S N Danilov) (Leningrad: Nauka, 1967) p. 35 87. V K Khairullin, A N Pudovik, N I Kharitonova Zh. Obshch. Khim. 39 608 (1969) b 88. V K Khairullin, R Z Aliev Zh. Obshch. Khim. 44 1683 (1974) b 89. V K Khairullin, L I Nesterenko Zh. Obshch. Khim. 46 795 (1976) b 163 91. V K Khairullin,M A Vasyanina, A N Pudovik Izv. Akad. Nauk 90. V K Khairullin,M A Vasyanina, A N Pudovik, Yu Yu Samitov, in Khimiya Organichskikh Soedinenii Fosfora (The Chemistry of Organic Compounds of Phosphorus) (Ed.S N Danilov) (Leningrad: Nauka, 1967) p. 29 SSSR, Ser. Khim. 950 (1967) c 92. A N Pudovik, V K Khairullin, R P Shagidullin, T I Sobchuk, V N Eliseenkov,M A Vasyanina, in Khimiya i Primenenie 93. V D Akamsin, N I Rizpolozhenskii Izv. Akad. Nauk SSSR, Ser. Fosfororganicheskikh Soedinenii (The Chemistry and Applications of Organophosphorus Compounds) (Ed. MI Kabachnik) (Moscow: Nauka, 1972) p. 220 Khim. 1976 (1967) c 94. A N Pudovik, E S Batyeva, O G Sinyashin Tioproizvodnye Kislot Trekhvalentnogo Fosfora (Thio-Derivatives of Trivalent 102. A N Pudovik, V K Khairullin, G V Dmitrieva Dokl. Akad. Nauk Phosphorus Acids) (Moscow: Nauka, 1990) 95.V S Tsivunin, G Kh Kamai, S Kh Nurtdinov Zh. Obshch. Khim. 36 1223 (1966) b 96. V S Tsivunin, G Kh Kamai, S Kh Nurtdinov Zh. Obshch. Khim. 36 1827 (1966) b 97. S Kh Nurtdinov, V S Tsivunin, T V Zykova, G Kh Kamai Zh. Obshch. Khim. 38 2517 (1968) b 98. V S Tsivunin, S Kh Nurtdinov, R R Shagidullin, G Kh Kamai Zh. Obshch. Khim. 39 1561 (1969) b 99. V K Khairullin, G V Dmitrieva, I A Aleksandrova, M A Vasyanina Izv. Akad. Nauk SSSR, Ser. Khim. 2744 (1973) c 100. V K Khairullin, R Z Aliev Zh. Obshch. Khim. 43 2165 (1973) b 101. A N Pudovik, V K Khairullin, T I Sobchuk Dokl. Akad. Nauk SSSR 174 104 (1967) a SSSR 174 372 (1967) a 103. V K Khairullin, R M Kondrat'eva, A N Pudovik Izv. Akad. Nauk SSSR, Ser.Khim. 2097 (1967) c 106. V K Khairullin,M A Basyanina, A N Pudovik Zh. Obshch. Khim. 104. V K Khairullin, A N Pudovik Zh. Obshch. Khim. 37 2742 (1967) b 105. V K Khairullin, T I Sobchuk, A N Pudovik Zh. Obshch. Khim. 38 584 (1968) b 38 858 (1968) b 107. V K Khairullin, R M Kondrat'eva, A N Pudovik Zh. Obshch. Khim. 38 2071 (1968) b 108. A N Pudovik, V K Khairullin Zh. Obshch. Khim. 39 1724 (1969) b 109. V K Khairullin,M A Vasyanina, A N Pudovik Izv. Akad. Nauk 110. V K Khairullin, R M Kondrat'eva, A N Pudovik Izv. Akad Nauk 111. A N Pudovik,M A Vasyanina, V K Khairullin Zh. Obshch. Khim. 112. A N Pudovik, V K Khairullin, G V Dmitrieva Zh. Obshch. Khim. 113. I L Knunyants, U Utebaev, E M Rokhlin, E P Lur'e, E I Mysov 114. A M Kibardin, I N Belova, M A Pudovik, A N Pudovik 115.A N Pudovik, E I Kashevarova, V M Gorchakova Zh. Obshch. 116. V K Khairullin, T I Sobchuk, A N Pudovik Zh. Obshch. Khim. 36 117. V K Khairullin, G V Dmitrieva, A N Pudovik Izv. Akad. Nauk 118. V K Khairullin, G V Dmitrieva, A N Pudovik Izv. Akad. Nauk 119. M A Vasyanina, V K Khairullin, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1603 (1967) c SSSR, Ser. Khim. 2334 (1969) c 40 1030 (1970) b 40 1034 (1970) b Izv. Akad. Nauk SSSR, Ser. Khim. 875 (1976) c Izv. Akad. Nauk SSSR, Ser. Khim. 2654 (1985) c Khim. 34 2213 (1964) b 296 (1966) b SSSR, Ser. Khim. 468 (1970) c SSSR, Ser. Khim. 1254 (1971) c SSSR, Ser. Khim. 452 (1970) c 120. V K Khairullin, G V Dmitrieva, A N Pudovik Izv. Akad. Nauk SSSR, Ser.Khim. 1249 (1971) c 121. D L Boger Tetrahedron 39 2869 (1983) 122. J M Kliegman, R K Barnes Tetrahedron 26 2555 (1970) 123. F Orsini, G Sala Tetrahedron 45 6531 (1989) 124. A M Kibardin, Yu B Mikhailov, T V Gryaznova, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1684 (1986) c 125. A M Kibardin, P I Gryaznov, T V Gryaznova, V G Lodochnikov, in The XIVth Conference on Phosphorus Chemistry (Abstracts of Reports), Cincinnati, 1998 p. 109E Ya Levina, AMKibardin 164 126. A M Kibardin, Yu B Mikhailov, T V Gryaznova, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 2655 (1985) c 127. A M Kibardin, T Kh Gazizov, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1452 (1980) c 128. A M Kibardin, A N Pudovik Zh. Obshch. Khim. 63 2430 (1993) b 129. T V Gryaznova, V G Lodochnikov, A M Kibardin, A N Pudovik, in The XIth International Conference on Chemistry of Phosphorus Compounds (Abstracts of Reports), Kazan, 1996 p. 114 130. A M Kibardin, I A Litvinov, V A Naumov, Yu T Struchkov, T V Gryaznova, Yu B Mikhailov,A N Pudovik Dokl. Akad. Nauk SSSR 298 369 (1988) a 131. A M Kibardin, T V Gryaznova, K M Enikeev, Yu B Mikhailov, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1684 (1987) c 132. A M Kibardin, I A Litvinov, V A Naumov, Yu Ya Efremov, T V Gryaznova, A N Pudovik Dokl. Akad. Nauk SSSR 297 376 (1987) a 133. A M Kibardin, T Kh Gazizov, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 2186 (1980) c 134. G V Dregval', G I Derkach Zh. Obshch. Khim. 33 2952 (1963) b 135. A A Petrov, N A Razumova, A Kh Voznesenskaya Zh. Obshch. Khim. 34 3512 (1964) b 136. A M Kibardin, R L Yanilkina, R Z Musin, A N Pudovik Zh. Obshch. Khim. 62 1981 (1992) b 137. P I Gryaznov, S N Kurochkina, R Z Musin, A N Pudovik, A M Kibardin Zh. Obshch. Khim. 66 383 (1996) b 138. P G Ugryumov Zh. Obshch. Khim. 29 4091 (1959) b 139. A M Kibardin, E Ya Levina, V B Mikhailov, Yu B Mikhailov, A N Pudovik Izv. Akad. Nauk SSSR, Ser. Khim. 1943 (1988) c 140. E Ya Levina, A N Pudovik, Sh K Latypov, R Z Musin, A M Kibardin Izv. Akad. Nauk SSSR, Ser. Khim. 490 (1989) c 141. E Ya Levina, A N Pudovik, A M Kibardin Zh. Obshch. Khim. 60 759 (1990) b 142. A Schmidpeter, J H Weinmaier, W S Sheldrick, D Schomburg Z. Naturforsch., B Anorg. Chem., Org. Chem. 34 906 (1979) 143. G Baccolini, R Dalpozzo Heteroatom. Chem. 1 333 (1990) 144. A M Arif, A H Cowley, R M Kren, D L Westmoreland Heteroatom. Chem. 1 117 (1990) 145. J Barluenga, J Jordon, F Palacios, V Gotor Synthesis 309 (1985) 146. D L Boger Chem. Rev. 86 781 (1986) 147. J S Sandhu, B Sain Heterocycles 26 777 (1987) 148. J Barluenga,M Tomas, in Advances in Heterocyclic Chemistry Vol. 57 (Ed. A R Katrizky) (San Diego: Academic Press, 1993) 149. Ya F Freimanis Khimiya Enaminoketonov, Enaminoiminov, Enaminotionov (The Chemistry of Enamino Ketones, Enamino Imines, Enamino Thiones) (Riga: Zinatne, 1974) 150. G Baccolini, P E Todesko J. Org. Chem. 39 2650 (1974) 151. G Baccolini, P E Todesko, R Dalpozzo, A De Nino Gazz. Chim. Ital. 126 272 (1996) 152. O V Korenchenko, A Yu Aksinenko, V B Sokolov, I V Martynov, in 6-ya Vsesoyuzn. Konf. po Khimii Ftororga- nicheskikh Soedinenii (Tez. Dokl.), Novosibirsk, 1990 [The Sixth All-Union Conference on the Chemistry of Organofluorine Compounds (Abstracts of Reports), Novosibirsk, 1990] p. 252 153. K Burger, K Geith, D HuÈ bl Synthesis 189 (1988) 154. A Schmidpeter, J H Weinmaier Angew. Chem. 87 517 (1975) 155. A Schmidpeter, J H Weinmaier Chem. Ber. 111 2086 (1978) a�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) b�Russ. J. Gen. Chem. (Engl. Transl.) c�Russ. Chem. Bull. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC