摘要:

Russian Chemical Reviews 70 (4) 265 ± 284 (2001) Size effects in chemistry of heterogeneous systems N F Uvarov, V V Boldyrev Contents I. Introduction II. Size effects in single-component systems III. Heterogeneous multicomponent solid-phase systems (composites) IV. Mechanochemical synthesis of nanocomposites and nanoparticles V. Chemical properties of nanoscale systems VI. Conclusion Abstract. multicomponent and single-component in effects Size Size effects in single-component and multicomponent solid-phase systems are discussed. Peculiarities of the size effects in solid-phase systems are discussed. Peculiarities of the size effects in microcrystalline and nanoscale systems are analysed. Methods for microcrystalline and nanoscale systems are analysed. Methods for mechanochemical synthesis of nanoparticles are considered. mechanochemical synthesis of nanoparticles are considered.Chemical properties of nanoscale systems are analysed. Particular Chemical properties of nanoscale systems are analysed. Particular attention is paid to the results of studies of the properties of attention is paid to the results of studies of the properties of nanocomposites and nanoparticles. Problems concerning their nanocomposites and nanoparticles. Problems concerning their stability and the possibility of mechanochemical synthesis of these stability and the possibility of mechanochemical synthesis of these systems are discussed. The bibliography includes 175 references systems are discussed. The bibliography includes 175 references.I. Introduction The understanding of basic principles and the development of new methods for the synthesis of functional materials with unusual and valuable properties is a key problem of solid state chemistry. Currently, `dry' (i.e., partially or completely solvent-free) tech- nologies are known to be the most efficient and ecologically safe way to carry out chemical reactions.1 Mechanochemical processes meet this requirement. A mecha- nochemical reaction is a multistage process proceeding by a complex mechanism, which requires that many factors be taken into account when optimising the conditions for mechanical activation of reactants. The mechanism of a mechanochemical reaction involves the following main stages: initial deformation of the crystal structures of reactants; formation and accumulation of both point and linear defects and interaction between them; dispersion of the reactants into blocks; formation of intermediate metastable states in the regions of phase contacts and, finally, chemical homogenisation of the reaction product followed by its relaxation to reach a thermodynamically equilibrium state.These processes have been best studied for metals where mechanochemical reactions proceed with relative ease. However, the results of recent investigations showed that ionic and molec- ular crystals are also promising objects for studying mechano- N F Uvarov Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch of the Russian Academy of Sciences, ul.Kutateladze 18, 630128 Novosibirsk, Russian Federation. Fax (7-383) 232 28 47. Tel. (7-383) 232 56 45. E-mail: uvarov@solid.nsk.su V V Boldyrev Novosibirsk State University, ul. Pirogova 2, 630090 Novosibirsk, Russian Federation. Fax (7-383) 239 71 01. Tel. (7-383) 232 15 50 Received 20 November 2000 Uspekhi Khimii 70 (4) 307 ± 329 (2001); translated by AMRaevsky #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n04ABEH000638 265 266 271 278 280 282 chemical processes. In ionic crystals, the charge and potential of the double electrical layers on the phase boundaries have a pronounced effect on the interphase interaction. In molecular crystals, there is a broad spectrum of chemical bonds differing in nature, direction and energy and each type of these bonds is characterised by a specific response to mechanical treatment.By varying the type and positions of functional groups in the molecules of the crystal one can control the mechanism of mechanochemical reactions, which is of great importance for the synthesis of advanced materials. Mechanical activation of heterogeneous mixtures allows preparation of nanocomposites with unusual properties, includ- ing supramolecular systems. Supramolecular nanoscale systems comprise chemically active components (substrates) bound to particular atoms or molecular groups of the carrier matrices (receptors) by non-covalent interactions. Such systems have been poorly studied as yet; however, a novel class of materials based on them has been designed.Unique properties of these materials are due to both specific types of intramolecular and intermolecular interactions in the bulk of the substrate and the character of the receptor ± substrate interactions. Here, the role of receptor can be played by either the molecular fragments or by a particular sublattice in the crystal structure of the reactant (intercalated compounds or `guest ± host' inclusion compounds), or by active groups on the surface of heterogeneous inclusions, which favour the formation of, e.g., Langmuir films or other surface layers. A salient feature of supramolecular heterogeneous systems is the strong interface interaction, which depends on many interre- lated factors including chemical, structural, morphological and thermodynamic ones.This interaction is responsible for the structural rearrangement in the substrate and for the changes in its thermodynamic, transport and chemical properties. In certain systems, the cationic or protonic conductivity of an ionic salt (substrate) can increase by several orders of magnitude and unusual phases can be stabilised in the receptor ± substrate contact region. For instance, composites of the ionic salt ± oxide type are candidates for the synthesis of supramolecular solid electrolytes based on them. Of considerable interest are also heterogeneous systems of the molecular crystal ± oxide type. Theoretical analysis of the mechanism of the surface inter- action between different phases is a complicated problem.Cur- rently, the properties of intergrain and phase boundaries in metals have been fairly well investigated. However, the ionic crystal ± oxide or molecular crystal ± oxide interfaces are still inadequately explored both experimentally and theoretically. Experimental studies are hampered by the problems of controlling the morphol-266 ogy of the composite, which leads to irreproducibility of the results obtained. Modern calculations require the inclusion of the contributions of not only covalent, hydrogen and van der Waals interactions but also long-range Coulomb and dipole ¡¾ dipole interactions, which makes them rather difficult. Moreover, consideration of characteristic features of real interface structures is also of great importance.Nevertheless, one can carry out trial calculations of the energetically favourable surface configurations for systems of the molecular crystal (substrate) ¡¾ oxide (receptor) type using the known model interaction potentials. Such estimates would be useful for targeted search of a substrate for a particular receptor. Since the action range of surface forces is confined to a thin near-surface layer, one can expect that nanocomposites will hold the greatest promise for the synthesis of supramolecular systems. Nanocomposites are heterophasic systems, which represent dense aggregates of nanoparticles of both phases, which are uniformly distributed in the bulk. The particle size of a nanocomposite does not exceed 10 nm.In nanocomposites, virtually the whole bulk of the substrate is in the interface region, which allows experimental investigations of its physicochemical properties at the nanoscale level. To date, the properties of nanocomposites have been poorly studied, since the preparation of these systems by conventional methods is an extremely complicated task. Mechanical activation facilitates considerably the grinding of the active component and its spreading over the oxide (receptor) surface, thus favouring strong adhesive interaction. It is known that the physical and chemical properties of solids change as the particle size decreases. These types of phenomena are usually called size effects. Numerous papers dealing with theoretical and experimental studies of size effects have appeared (see, e.g., reviews 2¡¾13).Currently, it has been reliably shown that size effects are inherent in all solids; some phenomenological models have been proposed for qualitative or semiquantitative description of these effects. Size effects have been best studied for metals and to a lesser degree for ionic crystals. For many materials, size effects are observed at relatively large grain size. For instance, the diffusion coefficient, conductivity and mechanical properties of a material become dependent on the average grain size (L) for L*103 ¡¾ 104 nm; as L reduces to 10 ¡¾ 102 nm, the melting temperature decreases and a change in the crystal lattice parameter of the material is observed.Size effect is most pronounced in nanoscale dispersed systems with a characteristic particle size in the range from 1 to 10 nm. In this case, the properties of the material can change basically, e.g., it can undergo a transformation into a high- temperature phase or another high-energy state, which cannot be realised under standard conditions. In heterogeneous systems, size effects depend on some addi- tional factors such as the chemical nature of the components of the heterogeneous mixture, the particle morphology (spatial distribu- tion of the components in the bulk of the composite) and the interface interaction. Consideration of these factors is a compli- cated problem and no satisfactory theoretical model for the description of size effects in composites has been proposed so far. The most pronounced changes in the properties are observed in both pure substances (undoped homogeneous individual com- pounds) and nanocomposites if the average particle size in these systems is less than 10 nm.Recently, the interest increased in the studies of unusual effects observed in nanoscale systems. This is not surprising since the processes proceeding in such systems serve as the basis for future solid-state nanotechnologies. However, chemical reactions in solid-phase mixtures are still the least investigated aspect of solid state chemistry. This is to a great extent due to the necessity of taking into account specific features of small particles involved or formed in the reactions.For instance, the reaction system can become a nanocomposite where the properties of reactants differ substantially from those at equilibrium in the stage of nucleation of the reaction product phase. In this review, we tried to describe N F Uvarov, V V Boldyrev the most plausible mechanisms of the influence of size effects on the course of physicochemical (including mechanochemical) proc- esses in heterogeneous solid-phase systems. In presenting the material, we pass from simple systems to more complex ones. First, size effects in isolated particles of simple substances are considered. Then, we discuss size effects in ensembles of small particles, in heterogeneous systems of different morphologies and in the nanoparticles formed in the initial stages of heterogeneous chemical reactions and prepared by mechanical activation.II. Size effects in single-component systems Size effects can be arbitrarily divided into two types. Weak effects where insignificant changes in the physicochemical properties of a substance are observed with an increase in the specific surface area (i.e., with reduction of the particle size) belong to the first type of size effects. Here, all the changes observed can be explained by surface effects. As a rule, these effects are usually observed in crystals with a characteristic size exceeding 10 nm. These systems will be referred to as microcrystalline ones. Strong effects where dramatic changes in the properties of a substance cannot be interpreted in terms of conventional surface phenomena belong to the second type of size effects.Most often, they are observed for very fine particles with a characteristic size of less than 10 nm. These systems will be referred to as nanoscale ones. In this review, we will compare these two types of size effects and follow a gradual transformation of microcrystalline systems into nanoscale ones. 1. Microcrystalline systems It has been experimentally shown that physicochemical properties of substances change noticeably as the particle size changes in the range from 10 to 100 nm. As mentioned above, the effects observed in this case are explained by the influence of the surface on the properties of crystals. As L decreases, the fraction of surface atoms increases.A 10 nm isotropic crystal of a simple substance has *1%¡¾ 5% of the atoms on the surface. The classical Gibbs thermodynamics treats all the excess thermody- namic characteristics of such fine particles, which differ from the corresponding bulk values, as being due to surface effects. Let us consider a small particle of a substance. Ignoring adsorption, the excess Gibbs free energy, DG, per unit area of the tension surface is the surface tension g of the substance. For an isotropic medium (e.g., a liquid), g equals the surface energy of the substance, s. If the particle size (here, the diameter of a drop of the liquid) is much larger than the characteristic molecular size, the tension surface virtually matches the geometric surface of the particle.For a spherical particle, the excess energy is G (1) s a gA a sA a s 6V L , where A, V and L are the surface area, volume and diameter of the particle, respectively. Generally, the surface tension, g^, of a phase in an anisotropic medium is a tensor quantity. For a slightly curved surface, it is defined as the difference between the proper local value of the stress tensor, EA, and its bulk value, EAa, extrapolated to the same point and is found by integration (2) g^ a OE^ ¢§ Ea ^ Udz , O? ¢§? where z is the coordinate normal to the surface of the layer. The absolute value of g is defined as half the spur (trace) of the surface tension tensor (3) g a 12 Trhg^i a 12 Og11 a g22U.For a liquid or an isotropic medium, g=g11=g22 and expression (1) holds. In the crystalline phase, the surface areaSize effects in chemistry of heterogeneous systems can be increased in two independent ways, viz., by increasing the number of surface atoms and by elastic deformation of a crystal.A specific feature of a solid is the presence of an immobile structural component, the chemical potential of which is considered to be a tensor quantity.14, 15 Nonuniformity of the chemical potential distribution over the surface layer of the crystal can lead to substantial difference between the surface tension at and surface energy of different surface areas. Owing to anisotropy of the crystal lattice, each face of the crystal is characterised by its own values of the surface tension and surface energy.In addition to the crystal faces, some contribution to the excess energy of the crystal also comes from the excess energy due to the presence of edges and vertices. Assuming an ideally smooth surface, the surface tension of ith crystal face is related to the surface energy by the following relationship i (4) ds gi a si a Ai dA . i The effective surface energy of a crystal with N faces,Medges and Z vertices equals 16 (5) A¢§1, s= ek kjlj a siAi a j i k XM XZ XN where kj is the energy of formation of the jth edge, lj is the length of parameters depend on the crystal shape) and A=PAi is the total the jth edge, ek is the energy of formation of the kth vertex (these tension surface area.The terms in formula (5) describe the contributions of crystal faces, edges and vertices to the excess energy of the crystal ignoring surface stress. For a crystal with an equilibrium shape, s&3V si , A hi where hi is the distance from the Wulf point to the ith crystal face. It is appropriate to choose the parameter 16 (6) L=6V A as the characteristic crystal size. If the parameters s and L of a crystal are defined by formulae (5) and (6), an additional term can be entered into Eqn (1) and it can be rewritten in the form G (7) s a gA a sA a DGe a s 6V L a DGe , i DGe& ds Ai dA . i where DGe is the elastic strain energy X In real microcrystals, the DGe value depends on many factors such as the crystal shape, the tension surface area A, the type of lattice symmetry, presence of defects in the crystal, etc.Estimation of the elastic strain energy in real crystals is a complicated problem. Expression (7) holds for both the liquid and crystalline phases (note that the contribution of elastic strain energy is nonzero only in crystals, since for liquid DGe=0). (8) As follows from the general thermodynamic expression for the Gibbs free energy, G(P,T,A)=U+PV7TS+sA+DGe(A), where U, V, S are the internal energy, volume and entropy of a crystal at a pressure P and temperature T, the surface area of the crystal plays the role of an additional external parameter that defines the thermodynamic properties of the material.If the substance is highly dispersed, the contributions of the surface-dependent terms become noticeable. This leads to specific size effects such as changes in the crystal lattice parameters, low- temperature (at temperatures close to 0 K) heat capacity, the Debye temperature, phase transition temperatures and stabilisa- 267 tion of unusual phases. The electronic structure of the substance also changes, which is responsible for unusual optical, electrical and magnetic properties. Among all the size effects, the effect of a decrease in the phase transition (e.g., melting) temperature, Tt, of small particles has been studied in most detail. Transition temperatures are deter- mined assuming that the corresponding phases have equal Gibbs energies.In the framework of the classical Thomson model for an isolated particle, the relative change of Tt is a linear function of 1/L (see Refs 17 ¡¾ 19). (9) Tt T 0t a 1 ¢§ 6OgaVa ¢§ gbVbU 1 H L , t where T 0 t and Ht are the temperature and enthalpy of the phase transition from the high-temperature a-phase to the low-temper- ature b-phase, respectively; ga and gb are the surface tensions of the a- and b-phases at temperature Tt , respectively; andVa andVb are the molar volumes of the a- and b-phases, respectively. Deviation of the experimental dependences of Tt/T 0 t on 1/L from linear ones can be explained by the formation of a thin film of the liquid phase on the surface of a crystal owing to melting of small particles,20 by the dependence of ga and gb on the particle size 21 or by the contribution of elastic strain energy DGe(A).11 In the last-men- tioned case, Eqn (9) can be represented in the form (10) Tt T 0t a 1 ¢§ 6OsaVa ¢§ sbVbUL¢§1 a DGeOLU , Ht where sa and sb are the specific surface energies of the a- and b-phases, respectively.From Eqn (10) it follows that the transi- tion temperature depends not only on the particle size and specific surface energy, but also on the change in the molar volume, the elastic strain energy difference between two phases, DGe(L), and the enthalpy of the phase transition. In the case of melting, the term DGe(L) includes the contribution of elastic strain energy of the crystalline phase at the melting temperature. Neglecting the DGe contribution allows construction of the phase diagrams of a substance in the T¡¾ (1/L) coordinates.They can be used not only for the estimation of changes in the transition temperatures, but also for the prediction of the phase size effect with a decrease in the crystal size below a certain critical value.6, 7, 13 This effect manifests itself in, e.g., stabilisation of new phases which are atypical of the macrocrystal. Currently, numerous experimental studies have confirmed the existence of phase size effects in both films 6, 7 and polycrystalline samples.11, 13 For instance, the equilibrium crystal structure of Al2O3 was found to be dependent on particle size. If the radius of alumina particles does not exceed 35 nm, the g-phase is stable; if the particle size increases, g-Al2O3 undergoes transformations into b-, e-, y- and a-Al2O3.8 The formation of amorphous phases in a number of systems can also be due to size effects.7, 22, 23 In the early studies on size effects, changes in the crystal structure of small particles were explained by the effect of excess pressure, Dp=4g L , produced owing to surface tension.2, 4 Indeed, isolated particles of many metals, alkali metal halides, oxides and nitrides exhibit a monotonic decrease in the lattice parameter with a decrease in L, which corresponds to positive Dp values. However, the unit cell parameters of nanocrystals (e.g., Si, Se, Cr and Pd) increase as the particle size decreases.3 Most often, a decrease in L leads to stabilisation of high-temperature modifications rather than high- pressure phases.An explanation 24 of this fact involves the assumption of the effective negative pressure in small particles, so that the classical concept of the Laplacian pressure cannot be generally used to describe their properties.268 Recently, intensive research has been carried aimed at the synthesis of highly dispersed materials such as metals, oxides, ceramics and high-melting compounds. Currently, substances with the characteristic particle size of less than 100 nm are usually classified as nanocrystalline materials.25, 26 Most of the systems studied to date are characterised by a grain size, L, ranging from 10 to 100 nm and their properties differ insignificantly from those of large crystallites.Hence, they should be considered as micro- crystalline rather than nanoscale systems. As will be shown below, the systems with a particle size smaller than 10 nm can be classified as nanoscale ones. Diverse morphologies of, e.g., the particles of ceramic materials complicate the problem of descrip- tion of their physical properties using a unified approach. Partic- ular effects, such as the appearance of additional contribution to the heat capacity, a decrease in the Debye temperature, an increase in the coefficients of thermal expansion and enhanced diffusion depend on the state of intergrain boundaries and can be described in the framework of an additive model which takes into account the contributions of the bulk and the surface to the physical properties of a material.In this case, the intergrain boundaries and pores in the material can be considered as specific metastable phases, the concentrations of which increase as L decreases. On the other hand, such characteristics as the magnitude of micro- stresses or elastic properties of the materials under study are strongly dependent on the prehistory of a sample (viz., on the procedure for its preparation and heat or mechanical pretreat- ment). For samples with relatively large grain size, these proper- ties can be satisfactorily described using the Hall ± Petch model. As the grain size becomes less than 10 nm, the response of the dislocation structure, formed owing to the presence of intergrain boundaries, to mechanical action becomes hardly predict- able;8, 12, 27 however, this can also be due to inadequate exper- imental procedures employed in studies of these materials.2. Nanoscale systems Numerous experimental studies have shown that if the particle size is less than the critical size Lc*10 nm, the bulk properties of materials change noticeably. A particle of size L&10 nm con- tains 104± 105 atoms, of which 1%± 5% are on the surface of the particle and contribute largely to the physicochemical properties. Ensembles of such particles form nanoscale (nanocrystalline) systems in which the following characteristic effects are observed. 1. Structural changes. If L<Lc, the average interatomic distance in the crystal changes dramatically and the crystal lattice parameter a becomes dependent on L.In the near-surface layer, the interplanar spacings along the normal to the surface are smaller than in the bulk of the crystal,3, 4 which is due to the non- symmetrical action of interatomic forces on the near-surface atoms. High-resolution electron microscopy (HREM) studies have shown that the region of noticeable changes in the lattice parameter is confined to five or six atomic planes and has a characteristic size ls*1 ± 3 nm. The same values have been obtained in calculations on alkali halide crystals.28, 29 The ls value can be considered as the action range of surface forces or the effective thickness of the crystal surface.The entire near- surface region of thickness ls is in the state that is characteristic of the surface and the crystal structure of this region depends on particular characteristics of the surface. In this connection, mention may be made that the structure of, e.g., nanoscale particles of metal glasses, which is experimentally determined from the radial distribution functions, differs from the structure of the same glasses in bulk specimens.10 2. Changes in thermodynamic properties. In nanocrystalline systems, the Debye temperature decreases and an additional contribution to the low-temperature heat capacity of the material appears, which increases as the particle size decreases.2, 11, 12, 30 These effects can be explained by changes in the vibrational states of the nanocrystal,3, 4, 12 namely, by `softening' of the spectrum (an increase in the number of low-frequency modes owing to a decrease in the number of high-frequency modes) with a decrease N F Uvarov, V V Boldyrev in L.Qualitatively, `softening' of spectra can be explained by the action of the pure surface effects: the atoms on the surface of a bulk crystal are characterised by increased amplitudes and lower vibration frequencies. However, analysis of the behaviour of the atomic Debye ± Waller factors using the data of X-ray diffraction study has shown 3 that changes in the Debye temperature in nanocrystals cannot be explained only by the contribution of the vibrations of the surface atoms since a decrease in L causes a stronger lattice `softening' than it could be expected.This means that not only the vibrational frequencies of the surface atoms, but also those of the atoms in the bulk of the nanocrystal are lowered. Yet another interesting feature of nanocrystals is that both the entropy of melting and the melting temperature decrease as L decreases.11 Since changes in the vibrational entropy makes the major contribution to the entropy of melting, this effect is also explained by the `softening' of the lattice of the nanocrystal with the decrease in its size. 3. Phase size effect. This is an example of a specific type of manifestation of structural and thermodynamic changes in mate- rials. Phase size effect is most pronounced in nanoscale systems. A decrease in L down to 10 ± 100 nm leads to stabilisation of high- temperature phases in both nanocrystals and microcrystals.In some instances, phases that are not characteristic of a given substance (including the amorphous phase) can be formed.2, 6, 7, 8 Structural data for even smaller particles are lacking; however, numerous theoretical studies suggest that simple substances (metals, inert gases) occur in the form of polyatomic clusters as the particle size reduces to several nanometres.3, 4, 31, 32 4. Changes in the electronic properties of materials. Among the effects experimentally observed in nanoscale systems, the quantum size effect 4, 7, 33, 34 seems to be the most interesting for chemistry. If the size of a crystal is comparable with the de Broglie wavelength of elementary excitations, the quantisation conditions of electron energy are changed and the energy bands split into a system of energy levels (Fig.1). In semimetals, this effect leads to the appearance of low-temperature oscillations of the kinetic coefficients and optical absorption with a decrease in the thickness of specimens (films).7 In wide-band semiconductor crystals, electronic excitation results in the formation of loosely bound electron ± hole pairs. In this case, the quantum size effect manifests itself markedly as changes in the optical spectra, namely, reduc- tion of L leads to an increase in both the energy at the maximum of the absorption band and the exciton peak intensity and to a shift of the luminescence spectrum towards the short-wavelength region.The pattern of the absorption spectrum becomes similar to that of the spectra of individual macromolecules. Semiconduc- tor nanocrystals possessing these properties are called `quantum dots'; they can be used in optoelectronics and are currently Energy Crystal Nanoparticle Molecule EF Density of states Figure 1. Electronic spectra of an isolated molecule, a nanoparticle and a semiconductor crystal. EF is the Fermi level.Size effects in chemistry of heterogeneous systems intensively studied.33, 34 As the size of ferromagnetic crystals reduces to *10 ¡¾ 100 nm, the coercive force increases and a decrease in the Curie temperature is observed. However, further decrease in the particle size causes a decrease in the coercive force and disappearance of ferromagnetic properties owing to the effect of superparamagnetism.4 Ferromagnetic clusters with sizes of the order of several nanometres or layered structures with the layer thicknesses of the same order of magnitude possess abnormally high magnetoresistance coefficients.35 ¡¾ 37 The above-mentioned effects can hardly be explained in the framework of the commonly accepted theory of surface phenom- ena.Acorrect mention has been made 3 that the concept of surface tension loses its meaning in the case of nanocrystals since all atoms can be treated as surface ones.Changes in the interatomic distances in small particles are mainly due to the decrease in the number of atoms in the particle,3 which means that the size effect is a collective property of all the atoms constituting the particle.Therefore, it is helpful to follow the changes in the properties of materials on going from molecules to clusters, then to nano- crystals and, finally, to conventional bulk crystals in order to gain a deeper insight into the phenomena observed. Let us consider pure substances with a Lennard-Jones poten- tial of the interatomic interaction. If the number of atoms in a particle, N, is small (N47), regular polyhedral clusters are formed. If 7<N<13, a large number of isomeric forms of the clusters exists, the energies of the isomers being only slightly different. If N=13, a relatively stable icosahedral cluster is formed.Using this cluster as a `seed', one can obtain larger icosahedral clusters with N=55, 147, 309, 561, etc., by building up the atomic layers on the cluster faces.3, 4, 31, 32 Seldom, can large crystals of the icosahedral phase (quasicrystals) be obtained.38, 39 In addition to icosahedral clusters, pentagondodecahedral ones are also relatively stable. Owing to the presence of the five-fold symmetry axes, an increase in the cluster size should inevitably result in the development of elastic strain. According to theoret- ical estimates, the icosahedral and dodecahedral structures become unstable for N&500, which corresponds to the particle size L>*2 nm. In this case, the particle is broken into tetrahe- dral blocks with a close-packed, face-centred cubic (fcc) structure separated by defect domains such as systems of twin boundaries, dislocations and disclinations.4, 40 These defects provide the thermodynamic stability of the particle; in other words, small crystalline particles with imperfect structure are in the thermody- namically equilibrium state.Further increase in N leads to stabilsation of the defect-free Wulf polyhedra, the shape of which is determined from the minimum condition for the surface energy of the crystal. Thus, each substance is characterised by two `threshold' lengths denoted as L1 and L2. If the particle size of a substance is less than L1, the particle spontaneously breaks down into nano- crystalline blocks. If L<L2, a phase transition into an X-ray amorphous phase (ensemble of clusters) occurs in the particle.It should be mentioned that for small particles there is a great number of different atomic configurations with close energies.41 This makes the concept of the equilibrium state of a particle ambiguous and is responsible for a gradual, diffuse character of phase transitions in the nanoscale systems. In particular, melting of clusters is considered as a dynamic equilibrium between the `solidlike' and `liquidlike' clusters that coexist at any temper- ature.41, 42 On going from individual clusters to ensemble of nanopar- ticles, the effect of intergrain boundaries on the physicochemical properties of the ensemble becomes noticeable. Since the number of surface atoms in a nanoparticle is comparable with that in the bulk, in the case of random packing of particles the structure of each crystallite is `tuned' to correspond to the local environment, which leads to effective broadening of the intergrain boundaries.Information on the structure of intergrain boundaries in nano- scale materials is contradictory. Both strongly disordered boun- 269 daries with a loose amorphous structure 9, 10 and the boundaries with relatively high concentration of atoms (typical of bulk polycrystals) are observed.12 In a 10 nm nanocrystal with an intergrain boundary width of 0.5 ¡¾ 1 nm, the fraction of the atoms situated on these boundaries reaches 15%¡¾ 30% of the total number of atoms.10, 12 Depending on the conditions for the preparation and annealing of ceramic nanoparticles, the struc- tures of the intergrain boundaries can be different.Freshly prepared specimens have more disordered structures compared to either sintered or simply stored specimens in which the boundary atoms are ordered to form conventional intergrain boundaries. This is accompanied by the growth of crystallites owing to a decrease in the concentration of disordered boundary atoms. According to molecular dynamics calculations of model ensembles using the Coulomb or Lennard-Jones interatomic potentials,43, 44 these relaxation processes should be observed after intense mechanical deformation of nanoscale systems. 3. Transport properties of microcrystals and nanocrystals(11) aVMa a aVMa?exp ¢§ 2gu RkT a aVMa? exp ¢§ Q , kT Polycrystalline materials are characterised by increased mobility of atoms.This has been confirmed in studies on the kinetics of sintering and recrystallisation,45 diffusion of radioactive iso- topes 46 and ionic conductivity.47, 48 In bulk crystals, diffusion and mass transfer occur by migration of point defects which has an activation character. Earlier, enhanced diffusion in fine crys- tallites was thought to be due to changes in the concentration of vacancies caused by curvature of the crystallite surface. In fact, the expression for the excess concentration of point defects in a metal near the curved surface, [VM], obtained using the Thomson ¡¾ Kelvin equation for the equilibrium of vacancies in crystal, has the form 45 (12) , Msa a aMsa a exp ¢§ gs 2kT aV where [VM]?is the concentration of defects in the bulk andRis the radius of curvature.If g=100 erg cm72 and u=10 cm3 mol71, the curvature of the surface makes an additional contribution, Q, of 0.003 eV (R=10 nm) and 0.03 eV (R=1 nm) to the energy of formation of defects. This effect can lead to a noticeable increase in the concentration of defects only in very fine particles with the size of about 1 nm. Currently, it is accepted that the activation energy of grain boundary diffusion in metals is lower than that of bulk diffusion; this is explained by the relatively loose local grain boundary structure and by high concentration of boundary defects.10, 46 The formation of a defect in the bulk of a metal involves two stages,49 namely, the formation of a pair comprising an excess atom above the surface and a vacancy in the surface layer and detachment of the vacancy from the surface followed by its diffusion into the interior of the metal.This is illustrated in Fig. 2 in which the plots of changes in the free energy of the crystal upon the formation of a surface vacancy and its diffusion into the bulk of the metal are also presented. In the first stage, surface disordering occurs: the metal atom M goes out of the surface layer on the outer surface of the crystal, which leads to the formation of a pair of defects, viz., the adatom,Ms, and a vacancy in the surface layer, VMs.The concentration of defects on the surface, [VMs], and in the bulk, [VM]?, are determined by the corresponding energies of formation (gs and g0 , respectively) (13) aVMa? a exp ¢§ g0 . kT Surface diffusion can occur owing to both migration of atoms over the outer surface and migration of vacancies within the surface layer. In polycrystals, the role of the outer surface is played270 G G0 / 2 Gs / 2 Figure 2. Schematic representation of the elementary act of surface disordering (a) and the formation of a Schottky defect in the bulk of a crystal (b) as well as change in the Gibbs free energy of the crystal upon the formation of a defect (c) and the spatial distribution of defects in the near- surface layer of the ionic crystal of the MX type (d). [VM 0] is the concentration of cationic vacancies (1), [VX.] is the concen- tration of anionic vacancies (2); i is the order number of the near-surface layer along the normal to the surface directed into the interior of the crystal (i=1 corresponds to the outer surface of the crystal). aVMsa a exp by the intergrain boundaries; in this case, [Ms]&1 and surface disordering is described by the equation The energy of formation of a defect and, hence, the concen- tration of defects in the bulk of the crystal are determined by the fundamental properties of matter rather than surface disordering. On the contrary, the gs value can vary between 0 and g0 depending of many factors such as orientation of the surface, the type of intergrain boundary, adsorption of impurities on the surface, etc.50 Hence, the activation energy of surface diffusion should depend on the procedure for the preparation of a given material.In ionic crystals of the MX type, surface disordering is accompanied by the formation of the electrical double layer since both cationic (V0Ms) and anionic (VXs . ) vacancies within the surface layer can be formed independently in accordance with the equations of quasichemical reactions 0>V0Ms+Ms. , 0>VXs . +X0 (here, zero corresponds to a perfect crystal). As in the case of metals, the concentration of uncharged surface defects should be completely determined by the energy of formation of these defects gs=(gVM)s+(gVX)s.However, (gVM)s 6a (gVX)s , ab c d ln[V 0M], ln[VX. ] 12 3 4 i 2 1 1 2 3 4 5 (14) ¢§ gs . kT (15) (16) s N F Uvarov, V V Boldyrev hence, an excess surface charge and the corresponding surface potential js appear. The magnitude of js depends on the energy difference between the quasichemical surface disordering reac- tions:50 (17) js=OgVX Us ¢§ OgVMUs . 2e The concentration of the Schottky defects in the bulk of a crystal is determined by the equilibrium constant of a quasichem- ical reaction involving the formation energy of bulk defects of g0, (18) 0>V0M? +VX? . . (19) The concentrations of defects on the surface, g0 ¢§ gs a ejs [V 0Ms]=[V0M?] exp , 2kT kT (20) [VXs . ]=[VX? .] exp g0 ¢§ gs ¢§ ejs 2kT kT depend on both the surface potential (17) and the additional contribution of the surface disordering. In the absence of the latter, (i.e., in a particular case g0=gs), Eqns (19) and (20) coincide with the expressions obtained in the framework of Frenkel's model.51, 52 Analogous expressions can also be obtained for crystals with Frenkel defects. The parameter js can be considered as the potential jump at the crystal ¡¾ vacuum interface; this takes into account the proper- ties of the surface, in particular, specific adsorption of charged particles (vacancies) on the surface. The defects localised on the surface can form a dense diffusion layer. At low concentrations of defects, the properties of this layer (or the space charge region) are described in the framework of the Gouy ¡¾ Chapman model.The concentration of defects in the diffusion layer is calculated using the Poisson ¡¾ Boltzmann equation which relates the space charge to the potential. The simplified expression for the ionic conduc- tivity due to the electrical double layer has the form 53 (21) K a BLT ¢§1=2exp ¢§h40 a ej2 s ¢§ Em kT¢§1, ¢§ where B is the constant which includes all temperature-independ- ent terms, h0 is the enthalpy of formation of the defect, Em is the migration energy of the defects that form the electrical double layer; for simplicity, the entropy contribution to js is taken as zero. The expression for the conductivity of the diffusion layer written in the coordinates of the Arrhenius equation has the form: log OKT1=2U a log BL ¢§ kT Ea , where the activation energy of conductivity is Ea= h0 4 ¢§ ej2 s a Em.Two regions corresponding to the high-temperature and low- temperature conductivities can be seen on the plot of experimental temperature dependence of the conductivity of a polycrystalline ionic salt. The former corresponds to the intrinsic conductivity, while the latter is due to the contribution of surface defects and is described by Eqn (21). The mobility and transport properties of atoms in small crystals or in clusters have not been studied so far. Nevertheless, computer simulation predicts high mobilities of the atoms and points out that a distinctive feature of the mechanism of migration is its collective character.3, 54 This has been confirmed by the results of experimental studies of nanocrystalline powders of pure substances.55 ¡¾ 57 For instance, the coefficient of self-diffusion inSize effects in chemistry of heterogeneous systems copper nanocrystals at 271 K is 14 to 20 orders of magnitude greater than in conventional bulk copper.55 An analogous effect was also found in studies of bulk diffusion in nickel nanocrys- tals.56 However, one should treat the interpretation of the results obtained in studies of self-diffusion with caution.A critical analysis 58 of the experimental results obtained in a study of diffusion in Cu, TiO2 and Pd nanocrystals 57 has shown that correct consideration of the diffusion `suction' into the bulk of the crystal makes the coefficient of surface diffusion in Cu nano- crystals typical of polycrystalline copper rather than abnormally high.It has also been mentioned 58 that the coefficients of oxygen diffusion in TiO2 nanocrystals are 5 ± 6 orders of magnitude greater than in bulk crystals and that the coefficients of diffusion of the dissolved hydrogen in palladium nanocrystals are greater than in the bulk specimen. III. Heterogeneous multicomponent solid-phase systems (composites) Generally, a solid-phase composite is a complex, heterogeneous multicomponent system which comprises several real solid phases considered as components. By a real solid phase is meant a totality of single crystals of different sizes, which contain impurities, point lattice defects, dislocations and pores as well as free surfaces (e.g., cracks, intergrain and outer boundaries).Many properties of composites depend on their morphologies, namely, the particle size distribution of each phase and mutual spatial arrangement of monophasic domains. From the standpoint of connectivity of their components, composites can be divided into (i) three-dimen- sional random mixtures of different phases, (ii) layered and (iii) columnar structures.59, 60 Film and porous structures can also be considered as composites. Real heterogeneous systems are usually characterised by mixed morphologies. When studying composites, one should take into account both the properties of each individual phase and those of interphase boundaries. If different phases interact with one another, the interfaces can exhibit specific properties that are characteristic of neither of the individual phases.It is appropriate to use the thermodynamic approach to describe the properties of heteroge- neous systems and find the conditions for their stability. We will restrict ourselves to consideration of two possible types of the surface interaction (a weak interaction and a strong one) between components of a simple system taking a two-phase composite as an example. 1. Interface interaction and sintering of composite For a two-phase composite composed of macrocrystalline phases (phase 1 and phase 2) with weak interaction between them, the Gibbs energy, G, can be thought to be approximately equal to the sum of the Gibbs energies of the pure components: G=G1+G2 .In real polycrystalline specimens, the energies G1 and G2 depend not only on the bulk parameters but also on a number of additional factors such as the presence of (i) dislocations and corresponding elastic strain, (ii) point defects, (iii) interfaces and outer surfaces, etc. Assuming that the largest contribution to the total excess Gibbs energy of the real pure components comes from their surface energies, one can write: (22) G=G1+G2+Gs=(G1 +G2 ) + (Gs1+Gs2Ü. Here, G1 and G2 are the standard Gibbs energies and Gs1 and1 Gs2 are the surface energies of the components, which are defined for each component by expression (7).Long-term heat treatment of such a composite leads to conventional sintering accompanied by growth of crystallites of both phases and annealing of dislocations.45 After sintering, the Gibbs energy of each phase becomes close to its bulk value, G and G2 , and the system reaches a thermodynamic equilibrium. If a After sintering Initial mixture b G Composite G1 á G2 á Gs1 á Gs2 Component 1 G1 á Gs1 Component 2 G2 á Gs2 G2 Figure 3. Schemes of sintering (a) and changes in the Gibbs energy of a composite (b) in the absence of adhesion between the components (dashed arrows) and in the case of strong adhesion (solid arrows). the mixture was mechanically treated in such a manner that the particle sizes of one or both components decreased substantially, a post-annealing phase separation will occur in the composite to give a mixture of the initial macrocrystalline phases (Fig.3 a). Let us consider the case where the rates of recrystallisation and annealing of the component 1 are much higher than those of the component 2. The expression for the Gibbs energy of the compo- site annealed at the sintering temperature of the component 1 can be written in the form G=G1 +G2 +s2A2+DGe2 . (23) After sintering, the component 1 reaches an equilibrium state (G1=G1 ), though the entire system remains thermodynamically nonequilibrium since the component 2 remains dispersed. Changes in the Gibbs energy of the composite are schematically shown in Fig.3 b. If surface interaction occurs between the constituent phases of the composite, an additional term, Gs12, appears in the expression for the Gibbs energy; this term represents the total change in the Gibbs energy of a heterogeneous system owing to the appearance of interphase contacts G=G1+G2+Gs12. The Gs12 is calculated as follows Gs12=s12A12 , where s12 is the specific surface energy of the interface and A12 is the interface area between the components 1 and 2. The contri- bution of the elastic strain energy should be taken into account by introducing additional terms, DGe1 and DGe2. The total Gibbs energy of the composite with inclusion of surface interaction is G=(G1 +s1A1+DGe1)+(G2 +s2A2+DGe2)+s12A12.(25) The expression for the Gibbs energy of the component 2 of the composite can be obtained from Eqn (25) 271 G1 á G2 á Gs2 G1 á G2 á s12A2 G1 á s12A2=2 G1 G2 á Gs2 G2 á s12A2=2 (24)272 (26) G2=G2 +s2A2+s12 A12 2 +DGe2 , assuming that the interface energy is evenly distributed between the components 1 and 2. The contribution of the elastic strain energy, DGe2, to the total energy of the phase 2 is mainly determined by the elastic strain energy due to the crystal lattice misfit at the phase 1 ¡À phase 2 interface and by the energy of misfit dislocations formed in the near-interface regions.6¡À 8 Calculations of Ge values are very complicated even for simple thin film morphology, which make estimation of this contribution in composites difficult.Nevertheless, irrespective of the character of elastic strain (a dilation or compression strain), the contribu- tion DGe2 is always positive and should increase with an increase in A12. Thus, the third and fourth terms in expression (26), which determine the energy of the phase 2, depend on the interface area in the composite. Sintering of a mixture of a low-melting component 1 with the greater self-diffusion coefficient with nanoparticles of a high- melting compound 2 can lead to both a decrease and an increase in the interface area A12. In the former case, sintering results in conventional recrystallisation of the phases. In the latter case, which is of particular interest for solid state chemistry, the formation of the interface is thermodynamically favourable if < 0 .dG dA12 Sintering is accompanied by an increase in the interface area A12. Neglecting the dependences of the specific surface energies and elastic strain energy on the surface area (or particle size) and expressing s12 in terms of the free energy of adhesion (sa), s12=s1+s27sa , one can write (27) 2 dG dA12 dA2 dA12 dA1 dA12 �¢ s1 �¢ s2 ¡¦ sa. �¢ s �� s1 The formation and growth of the interface (the so-called `sticking') have been considered in detail.45, 61, 62 The mechanism and kinetics of `sticking' are different in two cases defined by the conditions sa<2s1, (28) sa>2s1. (29) If condition (28) is met, a `neck' is formed between two particles of the composite.The shape and size of the `neck' depend on the shapes and sizes of the particles and on s1, s2 and sa. Inequality (29) is known as the Gibbs ¡À Smith condition for ideal wetting of films. If this condition is met, phase 1 completely covers the particle of the phase 2 after sintering. Inequalities (28) and (29) also determine the growth mecha- nism of the films of the phase 1 on a support formed by the phase 2. Growth of island films according to the Volmer ¡À Weber mechanism is observed in the former case while a steady layer-by- layer growth of films according to the Franck ¡À van der Merve mechanism is observed in the latter case.6, 63 The inequality (29) follows from expression (27) if <0, &1, dG dA12 dA1 dA12 dA2 dA �� ¡¦1, 12 i.e.if a decrease in the free surface area of the component 2 (the support) leads to the formation of an equivalent free surface area of the component 1 (the film). Considering the general case of an arbitrary morphology of the initial mixture of the components 1 and 2, one can see that the newly formed free surfaces of the component 1 can overlap if the distance between the surfaces of the phase 2 is sufficiently short (see Fig. 3 b). Complete overlap of N F Uvarov, V V Boldyrev these free surfaces makes the derivative dA1/dA12 close to zero and, assuming that the following conditions are met, =0, < 0, &71 dG dA12 dA1 dA12 dA2 dA12 from Eqn (27) it follows that (30) sa>s1 .This inequality determines the condition for spreading of the component 1 over the surfaces of the phase 2 in the composite. Comparison of inequalities (30) and (29) (the condition for ideal wetting or epitaxy in films) shows that the surface interaction in composites must be much stronger and requires a lower adhesion energy than in conventional film structures. It is known that roughness of a surface improves its wetting and that the roughness factor can vary between 0 and 1. Inequality (30) corresponds to the limiting case where the coefficient of roughness of the surface of the phase 2 is equal to unity. If condition (29) or (30) is met, a spontaneous increase in the interface area or `wetting' of the well-developed surface of the component 2 with the component 1 occurs in the system.If compound 2 is in the nanocrystalline state, the effective particle size of the phase 1 can be substantially reduced. In other words, an unusual effect, viz., self-dispersion of the component 1 to give a two-phase nanocomposite, will be observed. Self-dispersion is an indicator of a strong interface interaction in the system. This interaction will occur during the long-term sintering of the composite until the entire surface of the component 2 will be covered with the component 1. In this limiting case, A12&A2 and the system reaches the state of metastable equilibrium, where(31) G=(G1 +G2 )+ s12A2+(DGe1+DGe2) , (32) G1=G1 +s12A2 2 +DGe1(A2).The effect of self-dispersion was experimentally observed in, e.g., solid-phase sintering of composites AgCl ¡ÀAl2O3 (see Ref. 64), AgI ¡ÀAl2O3 (see Ref. 65) and Li2SO4¡ÀAl2O3 (see Ref. 66). An analogous effect was detected in the heteroge- neous systems based on alkali nitrates and halides such as LiNO3¡ÀAl2O3 and NaNO3¡ÀAl2O3 (see Ref. 67), RbNO3¡ÀAl2O3 and CsNO3¡ÀAl2O3 (see Ref. 68), NaCl ¡ÀAl2O3, KCl ¡ÀAl2O3 and RbCl ¡ÀAl2O3 (see Ref. 53). This is indicative of strong adhesion of ionic salts to the alumina surface. Analysis of expression (32) shows that the properties of an ionic salt depend not only on the standard thermodynamic parameters of the material, but also on the contribution of the energy of interface interaction, which increases as the oxide surface area increases.If the components are taken in a 50 : 50 (vol.%) ratio and mixed uniformly and an ideal wetting of the entire surface of the oxide with the MX component is achieved, the effective MX grain size is nearly equal to half the particle size of the oxide. In the composites with nanoscale oxide particles (*10 nm), the size of the MX particles becomes suffi- ciently small to provide noticeable size effects. 2. Adhesion energy and formation of point defects at the interface Stable interphase contacts require that the energy of adhesion be sufficiently high. Generally, the mechanism of adhesion depends on the character and strength of the interaction between the atoms arranged on the contacting surfaces. The authors of classical monographs on adhesion 69 ¡À 71 have studied the systems in which the major contribution to the surface interaction comes from the dispersion forces.In this case, the energy of adhesion can be calculated using the Hamaker ¡À de Buhr theory (33) sa=2(sd1 sd2)1/2+s 012,Size effects in chemistry of heterogeneous systems where sd1 and sd2 are the contributions of the dispersion forces to the surface energies of the components 1 and 2, respectively, and the last term includes the contributions of the non-dispersion interactions, viz., the ion ¡À ion, donor ¡À acceptor, dipole ¡À dipole interaction, etc. In composites of the ionic salt ¡À oxide type, it is the last term of expression (33) that makes the largest contribution to the energy of adhesion.In the absence of the electrical double layers, the sa value can be approximately estimated using the model of `weak boundary layers',72 according to which the work of adhesion is equal to the geometric mean of the works of cohesion of the components (34) sa&(W1W2)1/2=2(s1 s2)1/2. From relationships (34), (29) and (30) it follows that self- dispersion should occur if s2>s1 and be enhanced with a decrease in s1. Among alkali halides, lithium salts are character- ised by the highest specific surface energies s, while cesium salts are characterised by the lowest ones, the s values being monotoni- cally decreased on going from fluorides to iodides.29 Therefore, self-dispersion should proceed most easily in the composites based on cesium iodide and have the lowest rate in the systems based on lithium fluoride.However, it was noted 70 that the model of `weak boundary layers' 72 does not take into consideration specific characteristics of the interface in composites (its morphology and the presence of surface-active centres), which prevents a correct separation of the contributions of different types of interactions to the surface energies of the components, and does not assume the formation of the double surface layers.72 There- fore, this model allows only timation. To carry out more reliable calculations, one should take into account the structure of the contact surface and the potentials of the ion ¡À ion interactions at the ionic crystal ¡À oxide interface, which represents an extremely complicated problem that has not been solved so far.Approximate estimates of the surface energy using the Harkins `broken bond' method have shown 13, 69, 73 that s increases in proportion to the crystal lattice energy for isostructural com- pounds, the coefficient of proportionality being dependent on the type of the corresponding crystallographic plane. Comparison of the s values for alkali halides 29 with the corresponding values for oxides Al2O3 , FeO, MgO and SiO2 (100 ¡À 300 and 600 ¡À 1200 erg cm72, respectively) 13, 69 shows that the wetting condi- tion (30) should be met for the composites based on ionic halides and high-melting oxides including alumina.Physically, a tendency of the components of composites of the ionic salt ¡À oxide type to decrease the surface energy owing to the ion ¡À ion interaction at the interface is the reason for rerrangement of surface layers. To a first approximation, the MX¡ÀA interface energy (MX is the ionic salt and A=Al2O3) GsMX¡¦A �� sMX¡¦A AMX¡¦A& (35) &EM¡¦Al �¢ EAl¡¦X �¢ EM¡¦O �¢ EX¡¦O is determined by the sum of the energies of the cation ¡À cation (M+¡ÀAl3+), cation ¡À anion (Al3+¡ÀX7 and M+¡ÀO27) and anion ¡À anion (X7¡ÀO27) interactions. For the sake of simplicity, let us assume that strong interfacial interaction provides a structural correspondence between the crystal lattices of the MX andA phases. Since the energies EM7Al, EX7Al, EM7O and EX7O differ from one another, an ideal structure of the boundary layer of each phase will be distorted to provide an additional gain in the surface energy owing to the mutual approach (or removal) of different surface ions.The magnitudes of the displacements of the ions from ideal positions are determined by the balance of energies Ei. Since the sizes of the anions of alumina and most of MX salts are larger than those of the cations, one can expect that, provided that both contacting phases are close-packed, the cations at the interface are displaced to longer distances than anions. As a result, an intermediate, positively charged layer enriched with cations is formed between the surface layers of the components (Fig.4) and 273 a AMX b (M7A)s. V0Ms Figure 4. A scheme for the formation of defects at the interface between two binary ionic compounds. Initial surfaces (a) and interface (b). a layer of cationic vacancies (V0Ms). Such a structural relaxation can be represented as the following quasichemical reactions: (36) 0 > V0Ms +(M7A)s. , (37) 0 > VAl 000 +(Al7MX)s... , which describe the surface disordering in MX (36) and Al2O3 (37) at the MX¡ÀA interface. In addition to reaction (36), interstitial cations also can be transferred into the space charge layer 48, 74(38) Mi. > (M7A) s. . The formation of surface defects at the MX¡ÀA interface can also proceed by other mechanisms. If the energy of the interaction between aluminum cations and X anions is sufficiently high, the following reactions can proceed at the interface:75 (39) 0 > VX .+(X7A) 0s , (40) 0 > Mi.+(X7A) 0s .Reaction (39) is analogous to reaction (36) except that the role of cations is played by the anions; this can occur in composites based on fluorides, e.g., PbF2 ¡ÀA (A=CeO2, ZrO2, SiO2, Al2O3).75 Reaction (40) can be observed in AgF ¡ÀAl2O3 compo- sites. Since the binding energy of Al ¡ÀX dramatically decreases on going from fluoride anions to other halide anions, the reac- tions (39) and (40) in MX¡ÀAl2O3 systems are unlikely. More probable is the reaction (41) 0 > X0i +(M7A)s. , which can occur in the salt-basedMF2 ¡À SrCl2 composites with the anti-Frenkel defects.Shukla et al.75 proposed to use the pE value of the isoelectric point of an oxide as an indicator of the surface activity of the (42) oxide. Indeed, equations analogous to Eqns (36) and (39) can also be written for the surface interaction of an oxide with water H2O > OH0 + (H7A)s. , (43) 2H2O > H3O.+(OH7A) 0s , The former reaction dominates for oxides with pE>7 (e.g., Al2O3, CeO2) and the latter is predominant for oxides with pE<7 (ZrO2, SiO2). By analogy with aqueous solutions one can expect that the surface reactions (36), (38) and (41) are typical of composites with basic oxides (pE>7), while reactions (39) and (40) are typical of acid oxides (pE<7). Currently, the lack of experimental data makes it impossible to assess the transferability of the model of acid-base equilibria in aqueous solutions to274 composites.Nevertheless, the competition between the proc- esses (36) and (42) (44) V0M + (M7A)s .+H2O > OH0+ (H7A)s.+MM , where MM is a cation with a zero effective charge, can be responsible for deterioration of the ionic conductivity of the composites AgCl ¡¾Al2O3 (see Ref. 64) and AgI ¡¾Al2O3 (see Ref. 65) and for the appearance of protonic conductivity in the composites Li2SO4¡¾Al2O3 (see Ref. 76) and M(NO3)n¡¾Al2O3 (see Refs 77 and 78) in a humid atmosphere. 3. Ionic conductivity of composites In essence, reactions (36) ¡¾ (41) represent the formation of an electrical double layer at theMX¡¾Al2O3 interface.The formation of this layer affects the properties of the oxide; e.g., a decrease in the NeA el temperature of iron(III) oxide was observed in AgI ¡¾ Fe2O3 composites.79 The effect of the electrical double layer on the properties of the ionic salt is even more pronounced. A change in the energy upon the displacement of the ion from its normal (equilibrium) position characteristic of isolated MX crystal to the space charge layer (Dgi ¡¾A) can be considered as the work of charging of the surface; therefore, Dgi ¡¾A is related to the surface potential at the MX¡¾A interface (45) jAs*Dgi¢§A , qi where qi is the charge of the displaced ion. The exact jAs value is determined not only by the Dgi ¡¾A value, but also by the contribu- tion of the spatial relaxation of the oxide ions; therefore, relation- ship (45) should be considered only as a correlation between Dgi ¡¾A and jAs.If adhesive binding is strong, the Dgi ¡¾A value is always negative; hence jAs>0 for qi<0 andjAs<0 for qi>0. For T=0, there is a dense layer on the surface of the MX particle contacting the oxide. This layer is analogous to the Helmholtz layer. An increase in temperature leads to the forma- tion of a diffusion space charge layer built of cationic vacancies, which is responsible for enhanced conductivity of the composite since the ion mobility in the dense layer is low. The mobilities of defects in the diffusion layer are taken to be equal to their mobilities in the bulk of the crystal 48, 74 since the concentration of defects in this layer is low and the crystal structure remains unchanged (no deformations, surface phase transitions, etc.).In this case, the conductivity of the composite is given by expres- sion (21) with jAs introduced instead of js and the migration energy Em corresponding to those defects which form the diffu- sion layer; for the sake of simplicity, the entropy contribution to jAs is taken as zero. For low potential jAs, which corresponds to a relatively weak surface interaction, the conductivity is low and the activation energy of conductivity is Ea a h0 4 ¢§ ej2As a Em . Here, Ea ranges from h0/4+Em7ejAs/2 to h0/4+Em and depends on the state of the MX¡¾A interface (on the presence of structural defects, adsorbed impurities, etc.).Therefore, the con- ducameters measured in the experiments are poorly reproducible, which is sometimes observed in poorly sintered composites. If the surface potential jAs is high, the energy of adhesion is rather high, which favours the formation of a dense interphase contact and a substantial increase in the charge density in the diffusion layer. As jAs increases, the concentration of surface defects increases and at a surface potential equal to N F Uvarov, V V Boldyrev (46) jAs=h0 2e , K a AL T ¢§1=2exp ¢§EkTm , a full coverage of the surface with the similarly charged ions or defects is attained. If ejAs>h0/2, several layers of similarly charged ions should be on the surface, which is unlikely. Thus, the condition (46) determines the maximum value of the surface potential ejAs in a composite.53 In composites with high surface potential, the concentration of surface defects is close to the limiting value.In this case, the conductivity and activation energy of conductivity are given by the relationships (47) Ea=Em . The results obtained using these expressions are in good agreement with the experimental data for composite systems AgCl ¡¾Al2O3, AgBr ¡¾Al2O3, TlCl ¡¾Al2O3 and LiI ¡¾Al2O3.48 Certain new factors affecting the transport properties of ionic salts appear on going to nanoscale systems. The assumption of constant mobilities of defects within the diffusion layer and in the bulk of the crystal is acceptable if the thickness of the space charge region, 2lD, does not exceed the size of the MX crystal. It is reasonable to assume that the mobilities of defects within the diffusion layer of the crystal depend on the position of the defect owing to the drag effect of the ion `atmosphere'.Near theMX¡¾A interface, the defect mobility is low; it gradually increases with distance from the surface and reaches a maximum deep in the interior of the crystal. Size reduction of theMXcrystal (this can be achieved by increasing the concentration of A) leads to a decrease in the defect mobilities and to an increase in the activation energy of conductivity, which is equal to the migration energy of the defect. This effect is noticeable if the crystal size L is less than 2lD.In fact, the activation energy of conductivity in AgI ¡¾Al2O3 nanocomposites monotonically increases as the oxide concentra- tion increases.65 If L<lD, the total concentration of defects in the bulk of the crystal becomes high. The defect ¡¾ defect interactions in crystals can lead to phase transitions into the superionic state.80 ¡¾ 82 It cannot be ruled out that the superionic phases can be formed in the near-surface region of the crystal. The formation of the superionic phase on the surface of the pure ionic crystal has been reported;83 in composites, this seems to be even more probable. (48) e2a2v0 6kT k ¢§Em , kT If the concentration of defects, n, is temperature-independent, the expression for conductivity has the form K a n Sm exp exp where a is the jump length of the defect, n0 is the vibrational frequency of the mobile ion and Sm is the entropy of migration, which can be calculated using the formula for the compensation effect Sm a Em T , observed in alkali halide crystals.In this case, T *=2.16103 K.84 The formula given below 85 relates the conductivity of the compo- site at 25 8C to the activation energy of conductivity and to the concentration of defects in the MX particles; it includes the compensation effect and uses two empirical parameters, a=0.3 nm and n0=3610712 s71, typical of ionic crystals, log K25 (S cm71)=log n (cm73)714.5 Ea (eV )720.55. (49) Figure 5 presents the experimental data for different compo- sites.The straight lines were plotted using Eqn (49) with different defect concentrations. The experimental data for a number of superionic conductors are also shown in Fig. 5. As can be seen, theSize effects in chemistry of heterogeneous systems log K25 (S cm71 ) 0 III III 75 5 4 710 3 2 1 715 0 0.6 0.4 0.8 0.2 Ea /eV Figure 5. Theoretical (straight lines) and experimental (points) depend- ences of the conductivity of superionic conductors (I ), composites (II ) and AgCl single crystal (III ) at 25 8C on the activation energy for conductivity at different concentrations of defects: 1017 (1), 1018 (2), 1019 (3), 1021 (4) and 1022 cm73 (5).85 concentration of defects in the composites with the highest conductivity is also high and close to the corresponding values for superionic conductors.This indicates the possibility of stabi- lisation of disordered phases in composites. 4. Size effects in heterophasic systems In studies of thin films it has been shown that the structures of solids in the near-interface regions depend strongly on peculiar- ities of the morphology and energetics of the contacting planes. As a rule, the epitaxial interphase contact is the most energetically favourable. In the case of strong adhesive binding, building up the layer of phase 2 on a support proceeds by the van der Merve mechanism. Owing to the crystal lattice misfit between the adjacent phases, elastic strain is accumulated in the bulk of the component 2 (the elastic strain energy increases as the thickness of the layer of phase 2 increases).6, 86, 87 According to calculations,88 the continuous layer is destroyed after reaching a certain critical thickness hc and a network of misfit dislocations is formed at the interface.If the lattice misfit parameter lies between 0.1% and 1%, the calculated thickness of the epitaxial layer, hc, is *5 ¡À 25 nm (see Ref. 88), which is consistent with the experimental data. If the rate of film precipitation is high, thin films with an amorphous structure are deposited on the support instead of the epitaxial layer. As the temperature or the thickness of the layer increases, the amorphous phase undergoes transformation into a crystalline film.7 The formation of thin epitaxial or amorphous films can be considered as manifestation of the phase size effect in heteroge- neous systems, viz., the reduction of the characteristic size (film thickness) leads to the formation of new surface phases.In composites, the role of the thickness of the epitaxial layer is played by the average distance between the grains of the inert component 2 (r22). At low concentrations of the phase 2, the distance between the grains of this phase exceeds hc and the properties of the phase 1 are similar to those of the island thick films oriented parallel to the faces that confine the intergrain space. On the contrary, at high concentrations of the compo- nent 2, the phase 1 is distributed in the micropores of the phase 2 after sintering.In this case, if the grain size of the phase 2 is sufficiently small (less than 10 nm), one gets r22<hc and the phase 1 occurs in the form of a thin layer or a surface phase. Since the properties of substances in the surface phase differ substan- tially from the bulk properties of the same substance, two phases 275 can coexist in composites, viz, the surface phase with anomalous properties and a conventional bulk polycrystalline phase with a high concentration of dislocations. Thermodynamic properties of substances coated on solid supports in the form of adsorbed layers or thin films differ from their bulk properties.2 ¡À 7, 18, 19, 89 Noticeable distinctions between the thermodynamic properties appear at a characteristic crystal size of *10 nm.2¡À7 Changes in the temperatures of phase tran- sitions that occur in the bulk of the component 1 can be estimated using the formula Tt T 0t sa1Aa1 �¢ sb1Ab1 �¢ �� 1 ¡¦ (50) 2 �¢ sa12Aa12 ¡¦ sb12Ab12 �¢ DGe H¡¦t 1, where the subscripts a and b correspond to the a- and b-phases at temperature Tt, respectively, and DGe is the excess elastic strain energy difference between these phases.If the changes in the volume and DGe upon the phase transition are small (Aa1&Ab1*L71; Aa12*Ab12*L71; DGe&0), the temper- ature of the phase transition into the a-phas which is usually observed in thin films. Some examples of the phase diagrams constructed in the coordinates Tt ¡À (T, L71) for different supported metal films are available in the literature (see Refs 6 ¡À 8, 18, 19).According to these phase diagrams, transition temper- atures can decrease by hundreds of degrees upon reduction of L to 10 nm.6, 7 Generally, the temperatures of phase transitions in composites can either decrease or increase depending on the particular form of the functions G=f (T), s=f (T) and DGe=f (T) for a given component in different phases. The results of studies on the physicochemical properties of composites AgI ¡ÀAl2O3 (see Refs 65, 90 ¡À 94), Li2SO4¡ÀAl2O3 (see Ref. 66), LiNO3¡ÀAl2O3, NaNO3¡ÀAl2O3, KNO3¡ÀAl2O3 (see Ref. 67), RbNO3¡ÀAl2O3 (see Refs 68 and 95), CsNO3 ¡À Al2O3 (see Ref. 68), CsHSO4 ¡À SiO2 (see Refs 96 and 97), CsCl ¡ÀAl2O3 (see Ref.98) with the specific surface area of Al2O3 particles, s, of 270 m2 g71 show that nanocomposites prepared by self-dispersion technique exhibit unusual structural, thermody- namic and transport properties. Thermodynamic calculations for all the above-mentioned systems showed that bulk chemical interaction at the temperatures of synthesis of nanocomposites is thermodynamically unfavourable. This is also confirmed by the results of special experiments on the sintering of the initial mixtures under `severe' conditions (high temperatures and long- term heat treatment). Thermodynamic properties of silver iodide in (17 x)AgI ¡À xAl2O3 nanocomposites differ substantially from those of pure AgI. This can be illustrated by nearly complete absence of the thermal effect at 147 8Cupon the phase transition into the a-phase of AgI in the nanocomposite with high Al2O3 concentrations (x>0.8).Presumably, the superionic a-phase is stabilised in the composite; however, the absence of AgI reflections in the X-ray diffraction patterns (including those typical of the a-phase) suggests that silver iodide is in the amorphous state. This is also confirmed by the absence of a thermal effect at the melting transition in AgI (see Ref. 93) and by the unusual luminescence spectra observed upon UV irradiation of the nanocomposite.94 More detailed studies have shown that an increase in the Al2O3 content in the composite leads to a gradual change in the proper- ties of silver iodide.Namely, two phase transitions are observed instead of one a?b transition, which gradually disappear as x increases. Both transitions are characterised by distinct hysteresis loops on the plots of the temperature dependences of conductivity and heat capacity, the temperature of the high-temperature phase transition depends slightly on x, while the temperature of the low- temperature phase transition decreases appreciably as the mole fraction of Al2O3 increases (see Refs 92 and 93). This suggests that silver iodide can occur in two crystalline states during sintering of the composites under study. These are276 the bulk state with the properties of pure AgI and the surface state localised at the AgI ¡¾Al2O3 interface. The properties of the latter are affected by the interfacial interaction.Like the bulk state, the surface crystalline state can be in the high-temperature and low- temperature disordered phases. As the Al2O3 content increases, the effective intercrystallite distance in the oxide (and the effective particle size of AgI) decreases and the conductivities of both phases approach each other, thus indicating a gradual levelling of distinctions between the phases. In addition to this effect, the increase in the concentration of Al2O3 leads to a dramatic decrease in the enthalpies of melting and b?a phase transition in AgI, which points to the appearance of the interface amorphous AgI phase in the composites (Fig. 6). At sufficiently high concentration of alumina, nearly all the silver iodide can be stabilised in the interface amorphous state.91, 93, 94 It was shown that doping of AgI with Al2O3 leads to a decrease in the temperature and enthalpy of the phase transition in AgI.99, 100 This is in excellent agreement with the results of our investigations of composites with low Al2O3 content.However, the authors of the papers cited (cf. Refs 99 and 100) have observed only minor changes in the transition temperature and enthalpy, since the heterogeneous additive used in their experiments was a relatively coarse-grained (0.06 mm) Al2O3. The use of alumina with a grain size of 0.01 mm allowed us to observe much more pronounced effects. Composites MNO3¡¾Al2O3 (M is an alkali metal) (see Refs 67, 68, 95) and the CsHSO4 ¡¾ SiO2 system 96, 97 behave analogously.DTA studies of composites (17x)RbNO3 ¡¾ xAl2O3 have shown 95 that all the phase transition temperatures of RbNO3 including the melting temperature remain virtually unchanged upon sintering, whereas the enthalpies of the phase transitions decrease dramatically. For x50.7, the DTA curves exhibit no peaks indicating the presence of crystalline RbNO3 in the composite (Fig. 7). Instead, a weak broad peak appears near 250 8C with the maximum intensity for 0.4<x<0.6. These effects are not observed in the composites containing coarse-grained Al2O3 (with grain sizes of several micrometres), a log KT (S cm71 K) 20 72 1 741 2 3 b DQ(arb. u.) 440 400 420 Figure 6. Temperature dependences of conductivity (a) and DTA curves (b) for pure AgI and the (17x)AgI ¡¾ xAl2O3 composites.93 AgI (1), x=0.6 (2), x=0.8 (3).3 2 1000/T /K71 123 460 T /K a DQ(arb. u.) 0.1 0.3 0.5 0.7 b DQ(arb. u.) 12 200 250 150 Figure 7. DTA curves for the (17x)RbNO3 ¡¾ xAl2O3 composites with a specific surface area of 270 m2 g71 (a) and for the 0.7RbNO3 ¡¾ 0.3 Al2O3 composites (b) with a specific surface area of 10 (1) and 270 m2 g71 (2). Figures near the curves shown in Fig. 7 a represent x values.95 which means that they are due to the influence of the oxide grain surface. Electron diffraction studies and conductivity measure- ments revealed the formation of an amorphous phase of RbNO3 in the RbNO3¡¾Al2O3 composites. The thermal effect at 250 8C is due to the glass transition in the amorphous phase and the transition temperature is much higher than the calculated value (755 8C).101 However, it should be taken into account that in this case the amorphous RbNO3 phase is formed at the RbNO3¡¾Al2O3 interface and its properties can be very different from those of pure amorphous rubidium nitrate. The results of thermodynamic studies of RbNO3¡¾Al2O3 composites are consis- tent with those obtained by the ionic conductivity measurements, Raman spectroscopy and X-ray diffraction analysis.95 The concentration and thickness of the interface phase in composites can be estimated using a simple brick-wall model for cubic particles.85, 93, 95 The model is based on the assumption that the brick size is equal to the particle size of the dispersed additive (LA) and uses the following formulae for the calculation of the volume fraction, fs , or the mole fraction, xs , of the interface phase fs a b 2l f O1 ¢§ fU, LA LA xs a b 2l g xO1 ¢§ xU 1 a xOg ¢§ 1U , where b is the geometric factor, l is the thickness of the surface layer, f and x are the volume fraction and molar concentration of the dispersed phase 2 (the oxide A), respectively, and the coef- ficient g depends on the densities (d) and molecular masses (m) of the components and is defined as follows g a mAd1 .m1dA N F Uvarov, V V Boldyrev T /8C (51) (52)Size effects in chemistry of heterogeneous systems The fraction of the ionic salt ( f1 or x1) that remained in the bulk phase can be calculated using the expressions f1 a 1 ¢§ f ¢§ fs a 1 ¢§ fb 2l O1 ¢§ fU, LA x1 a 1 ¢§ x ¢§ xs a 1 ¢§ x ¢§ b 2l gxO1 ¢§ xU 1 a xOg ¢§ 1U .calculations, the amorphous layer in AgI ¡¾Al2O3 These formulae are convenient for estimating the thickness of the interface phase provided that the particle size of the dispersed component, LA, is known. The ratio l/LA and the thickness of the interface phase, l, can be estimated from the concentration dependence of any extensive characteristic of the phase 1 in the composite using expressions (53) or (54) (Fig. 8). According to and RbNO3¡¾Al2O3 composites is respectively 3 93 and 4 nm95 thick. Studies of the conductivity of the nitrate-based composites revealed a substantial increase in the low-temperature conductiv- ities and disappearance of the conductivity jumps observed upon phase transitions. Note that the Arrhenius plots of the temper- ature dependences of conductivity are nonlinear (Fig.9). Among xs , xc 1.0 0.8 2 0.6 0.4 1 0.20 0.6 0.4 0.2 Figure 8. Dependences of the mole fraction of the interface amorphous phase (xs) (curve 1) and crystalline phase (xc) (curve 2) of AgI on the Al2O3 content in the AgI ¡¾Al2O3 composites according to DSC measure- ments of the enthalpy of the a ¡¾ b phase transition in AgI. Solid lines represent the results of calculations with l&4 nm and LA&10 nm.93 log K (S cm71) 72 73 74 75 76 77 781.6 2.0 2.4 Figure 9.Temperature dependences of conductivities of nitrates MNO3 (1 ¡¾ 5) and composites 0.5MNO3 ¡¾ 0.5Al2O3 (6 ¡¾ 10);M=Li (1, 6), Na (2, 7), K (3, 8), Rb (4, 9) and Cs (5, 10).67, 68 (53) (54) LA 0.8 xAl2O3 12345678910 1000 / T /K71 2.8 277 the composites based on alkali nitrates, the greatest increase in conductivity is observed for the LiNO3-based ones. At 100 8C, the conductivity of the composite 0.5 LiNO3 ¡¾ 0.5 Al2O3 is 7 orders of magnitude higher than that of pure lithium nitrate.67 The smallest increase in conductivity is observed for the CsNO3¡¾Al2O3 sys- tem.68 Similar effects were also found in CsHSO4 ¡¾ SiO2 compo- sites.96, 97 Here, the physicochemical characteristics of cesium hydrosulfate depend on the pore size of silica gel used as a dispersed additive, that is, there is an optimum pore size (*2 nm) at which the thermodynamic, structural and transport properties of CsHSO4 change to the greatest extent.97 Physicochemical properties of lithium sulfate change substan- tially on going from the bulk crystal to (17x)Li2SO4 ¡¾ xAl2O3 nanocomposites.66, 85 The low-temperature ionic conductivity of the composite substantially increases while the well-defined (for pure Li2SO4) phase transition becomes diffuse in the composite (Fig. 10).The results of DTA studies of these composites showed that lithium sulfate occurs in two states at low oxide concentra- tions. One state exhibits the properties of the initial Li2SO4 while the other state possesses the properties of the phase localised near the Li2SO4 ¡¾ g-Al2O3 interface.If the content of Al2O3 additive, x, exceeds 0.5, virtually the whole lithium sulfate is stabilised in the interface phase after sintering. In composites with the oxide particle size of the order of several nanometres the concentration of the interface phase is very low and no additional thermal effect is observed. Compared to the pure substance, the phase transition in the interface phase of Li2SO4 occurs at a much lower temper- ature (470 8C). log K (S cm71) 71 73 75 2 77 79 1 711 1000 /T /K71 2.0 1.0 Figure 10. Temperature dependences of the ionic conductivities of pure Li2SO4 (1) and the (17x)Li2SO4 ¡¾ xAl2O3 nanocomposite (2).66 In this case, the mechanism of adhesion involves a stage of surface chemical reaction between Li2SO4 and Al2O3 resulting in the formation of a thin interface lithium aluminate layer. The thickness of this layer was estimated at 1 ¡¾ 2 monolayers on the surface of the Al2O3 grains.Analysis of the X-ray and electron diffraction patterns of the nanocomposite points to epitaxial character of the lithium sulfate layers, which is due to the close (111) interplanar spacing in Li2SO4 and Al2O3 and to strong adhesion of Li2SO4 to the alumina surface. Lithium aluminate forms an intermediate layer and represents a specific `glue', which provides a denser epitaxial contact between the Li2SO4 and Al2O3 nanocrystals. The nanocomposite 0.33 Li2SO4 ¡¾ 0.67 Al2O3 can be prepared in two ways, viz., by sintering of the starting materials and by decomposition of the precursor, Li2SO4 .4 Al(OH)3 .mH2O.85 The second method allows one to synthesise a nanocomposite with uniform distribution of the components and an average particle size of less than 10 nm for both Li2SO4 and Al2O3 (Fig. 11). If the alumina particles are uniform in size, both synthetic procedures278 10 nm Figure 11. Morphology of the Li2SO4±Al2O3 nanocomposite obtained by thermal decomposition of its precursor, Li2SO4 . 2Al(OH)3 .mH2O.66 a 110 110 100 100 200 111 200 111 bc 200111 220 311 20 10 20 10 Y /deg Composites Initial mixtures Figure 12. X-Ray diffraction patterns of the initial mixtures and (17x)CsCl ± xAl2O3 composites after long-term sintering at 600 8C.Arrows denote the reflections corresponding to the fcc phase of CsCl, figures denote the crystallographic plane indices. x=0.3 (a), 0.6 (b), 0.8 (c). result in the formation of nanocomposites with identical struc- tural, transport and thermodynamic characteristics after suffi- ciently long-term heat treatment of the mixtures. This indicates N F Uvarov, V V Boldyrev that the composites reached metastable thermodynamic equili- brium. Decomposition of precursors, LinX. 2nAl(OH)3 .mH2O, allows the preparation of nanocomposites with different lithium salts LinX (X=F, Cl, Br, I, SO4, PO4); these systems exhibit high ionic conductivities.85 X-Ray diffraction and DTA studies of (17x)CsCl ± xAl2O3 composites (with Al2O3 with a specific surface area, s, of 270 m2 g71) showed that long-term sintering leads to stabilisation of a high-temperature CsCl phase localised at the interface (Fig.12).53, 98 No changes are observed in the X-ray diffraction patterns as the particle size of Al2O3 increased up to*1 mm; this is in agreement with the known results.102 The size effect, i.e., stabilisation of the high-temperature CsCl phase, seems to be due to the interface interaction, which is most pronounced in the composites with nanocrystalline additives. In the ionic salt ±Al2O3 systems, lithium salts form nano- composites somewhat more easily than rubidium and cesium salts and iodides form nanocomposites more easily than chlorides or fluorides.This points to important role of the polarising power of the cation and the polarisability of the anion in the mechanism of surface interaction. IV. Mechanochemical synthesis of nanocomposites and nanoparticles Mechanochemical processes proceed at low temperatures where the formation of a perfect crystal structure is difficult. Therefore, the use of mechanochemistry for the synthesis of metastable products, in particular, nanoparticles and nanocomposites, is of considerable interest.103 ± 109 Mechanochemical methods for the preparation of nanocrystalline materials have been intensively developed recently 105 ± 108 and investigations into the fundamen- tal problems associated with the formation, relaxation and stabilisation of metastable states during mechanical activation seem to be of great importance.1. Mechanical action on a solid and the formation of nanoscale systems Mechanical action leads to the development of elastic stress in the crystal. For different reasons (crystal lattice anisotropy, specific features of electronic and ionic properties, different energies of chemical bonds), elastic stress relaxation can follow different mechanisms such as vibrational and electronic excitations, ionisa- tion of chemical bonds, bond cleavage, rearrangements of atoms and migration of atoms and ions.109 By varying the intensity and character of mechanical action one can control the properties of materials obtained by mechanical activation.104 One of the most often encountered effects accompanying mechanical activation is failure which leads to grinding of the particles of a substance.Failure is preceded by accumulation, interaction and concentration of defects as well as by crack formation and propagation into the interior of the crystal. However, direct mechanical grinding of a solid usually does not allow preparation of nanoparticles since mechanical activation leads to acceleration of mass transfer owing to the formation of metastable defects. In addition, the stored elastic strain energy is partly transformed into thermal energy and the temperature in the impact zone can be significantly increased, thus favouring both recrystallisation of the substance and `healing' of defects that prevent grinding.106 Mechanical activation of heterogeneous mixtures seems to be more promising for the preparation of nanoparticles.Initially, the harder component acts as a crusher, thus favouring finer grinding of the softer component. Dispersion in deeper activation stages is due to the interfacial interaction between the components; here, the soft component plays the role of surfactant and facilitates the grinding of the harder component owing to the Rehbinder effect. Mechanochemical reactions in heterogeneous mixtures repre- sent the most efficient method for the preparation of nano- composites.106 ± 108 In this case, nanoparticles of a new phase areSize effects in chemistry of heterogeneous systems formed at the interface or as a result of decomposition of metastable states stabilised upon mechanical activation of the mixture.A unique feature of mechanochemical processes is the possibility of `deformation mixing' (mechanical alloying) of the components of a mixture, i.e., mixing of the starting components at the atomic level.105 ± 109 Mechanical alloying occurs at low temperatures whereupon diffusion is retarded. This allows stabi- lisation of various metastable intermediates including nanopar- ticles. 2. Methods for mechanochemical synthesis of nanocomposites and nanoparticles Mechanochemical synthesis of nanocomposites in metal systems is fairly well studied. The main classes of mechanochemical reactions used for the preparation of metal nanocomposites are as follows.110 1.Conventional mechanical treatment of a composite result- ing in transformation of one of the components into the nano- crystalline state. This is the simplest and most widely used procedure for the preparation of nanocomposites by exposing mixtures of solids to high-power mechanical impulses. If one component of a mixture is hard while the other is soft, mechanical treatment will lead to comminution of the former and deformation of particles of the latter. This can result in the formation of a composite comprising plate-like particles of the soft component with a thickness of several nanometres separated by large nanoparticles of the hard component. This type of nanocomposites has been obtained in, e.g., bimetallic systems Cr ± Fe,111 Ni ± Ti 112 and Ce ±Yb 113 (for a detailed list of systems, see the review by Matteazzi et al.114).Mechanically alloyed Cu30Cr170 has been obtained by treatment of a Cu ± Cr mixture 115 and studied by EXAFS, DSC and anomalous scattering of synchrotron radiation. It was shown that the alloy represents a nanocomposite rather than a solid solution of copper in chromium and that the Cu nanoparticles occur in the metastable phase. Mechanical treatment of a Cu ± Co mixture, 80 : 20 (at.%), allowed the preparation of a nanocompo- site containing nanoscale cobalt particles.116 Thorough field-ion microscopy studies of the structure of the nanocomposite showed that the material contains both individual 15-nm Co nanopar- ticles and 1 ± 3-nm cluster inclusions of Co atoms in the copper matrix.117 The presence of small Co clusters is responsible for anomalous magnetoresistance of the nanocomposite.Layered Co ±Cu structures obtained by vapour deposition or using pulse electrochemical deposition also possess analogous properties.118 2. Transformation of an amorphous phase into a nanocrystal owing to the elastic strain energy stored during mechanical treat- ment, e.g., preparation of nanocomposites from amorphous alloys based on iron and aluminium:119 AaBb (nanocrystal). AaBb (amorphous solid) Yet another example is the transformation of amorphous Fe ± Si ± B alloys into nanocrystals upon mechanical treatment. The process can be either accelerated by addition of cobalt or slowed down by addition of niobium (this can inhibit crystallisa- tion and retain the amorphous alloy). The reasons for the trans- formation of the amorphous phase into a crystalline one upon mechanical treatment have been discussed.120 ± 122 If mechanical treatment is carried out in air, the formation of nanocomposite is often favoured by the formation of an oxide film on the metal surface.119 3. Decomposition of an amorphous phase (usually, an inter- metallide) to give one of the components in the form of nano- particles: AaBb (amorphous solid) cA (nanocrystal)+Aa7cBb (amorphous solid).279 4. More complex mechanochemical reactions: aA+bB AaBb (nanocrystals), AaBbC (nanocrystals), AaBb+C AaBbCcDd (nanocrystals).AaBb+CcDd Mechanical alloying of the mixtures of different metals with graphite, silicon or tin is used for the preparation of carbide,114, 123 silicide and stannide nanoparticles.124 Nanocomposites can also be obtained by mechanochemical solid-phase displacement and exchange reactions, e.g., by the interaction of iron nitride with aluminium or silicon or by the reaction of iron carbide with chromium 125: AlN (nanocrystals)+2.5 Fe, Fe2.5N +Al Si3N4 (nanocrystals)+10 Fe, 4Fe2.5N+3Si (Fe,Cr)3C (nanocrystals). Fe3C+3Cr Hydrogenation of an intermetallide Pr(Co17xCux)5 results in the formation of copper and cobalt nanoparticles with sizes 10 ± 20 nm incorporated into the matrix of hydrogenation prod- ucts Pr(Co17yCuy)5Hx and PrH2 (see Ref.126). Specific methods are used if one needs to obtain nanoparticles of a single component of a nanocomposite rather than the nano- composite itself. For instance, nickel ± zinc ferrite nanoparticles with sizes 8 ± 50 nm were obtained by long-term (200 h) heat treatment of a mixture of the ferrite and quartz.127 To separate the metals from quartz and other components, the mixture was treated with 2M HClO4 for 1 h. This kind of treatment makes the particles on the ferrite surface charged, whereas other particles remain uncharged; this phenomenon is used for the separation of the mixture. Ferrite particles form a sol that can be destroyed by controlled change in the pH. An efficient method for preparation and isolation of nano- particles from mechanically alloyed composites has been pro- posed recently.128 ± 130 Nanoparticles were formed by solid-phase exchange reactions chosen in such a way that the by-products be readily soluble while the desired product, i.e., nanoparticles, be insoluble in conventional solvents.For instance, ZnS nanopar- ticles were prepared by the solid-phase mechanochemical reaction ZnS +CaCl2 . ZnCl2+CaS The newly formed zinc sulfide particles with an average size of 500 nm were found to represent aggregates comprised of smaller particles with characteristic sizes of 10 ± 12 nm. The formation of aggregates can be prevented by adding calcium chloride to the initial mixture; in this case, the desired product is obtained as separated ZnS particles of size 7 ± 9 nm.Calcium chloride is removed from the mixture of reaction products by selective dissolution in methanol followed by centrifugation. A similar procedure is used for the synthesis of CdS nanoparticles following the reaction CdS+2 NaCl . CdCl2+Na2S The starting mixture was additionally diluted with NaCl. After mechanical treatment for 1 h the mixture was washed with water to remove NaCl. The sizes of the CdS particles formed were from 4 to 8 nm (i.e., less than the Bohr radius of an exciton), which allowed the mechanochemical synthesis of `quantum dots'. Mechanochemical reduction of oxides is also known. The use of a peroxide instead of oxide allows transformation of this reaction into self-propagating high-temperature synthesis.131 By performing the reaction in the `smouldering' (incomplete combus- tion) rather than conventional combustion regime (this can be achieved by changing the heat treatment regime of the mixture or by diluting it with an inert compound), one can obtain the reaction product in the form of nanoparticles with sizes 40 ± 50 nm.Using this regime in the reactions of BaO2 with metals, mixed titanium- molybdenum oxide, barium aluminate and barium stannate nano- particles were obtained.131280 Preparation of zinc ferrite nanoparticles by mechanical treat- ment of a mixture of zinc and iron oxides represents yet another example of mechanochemical synthesis of nanoparticles of com- plex oxides from simple ones. However, in this case heating leads to rapid growth of stable large ferrite crystals.132 Mechanical treatment of the preliminarily prepared zinc ferrite, a-Fe2O3 and g-Al2O3 nanoparticles also leads to their growth.133 Therefore, optimisation of reaction conditions (in particular, mechanical treatment) should receive due attention when performing mecha- nochemical synthesis of nanoscale materials.A specific case of mechanical activation of a topochemical reaction has been studied.134 Intercalation of g-Al(OH)3 (gibbsite) with lithium salts from solution was shown to result in the development of wedging stress in the gibbsite lattice. This leads to an increase in the interlayer distances in the gibbsite structure, thus improving the intercalation conditions.Relaxation of mechanical stress can lead to the appearance of lattice defects. This will also favour an enhancement of the reactivity of gibbsite and can be used for carrying out various chemical reactions in the interlayer space of gibbsite.135 ± 138 A salient feature of this type of reactions is the possibility of crystallographic control of mutual orientation of the reactants, intermolecular distances and their mobilities. Nanoparticles can be prepared by intercalating the reactants into the interlayer space. They can be formed either if both reactants are present within the interlayer space or if the second reactant (e.g., oxygen) is supplied from the outside to oxidise the first reactant in the interlayer space. Nanoparticles thus formed can sometimes be retained within the interlayer space, as is the case of [MnOx(OH)4]472x7y nanoparticles formed by the reaction of the intercalated permanganate ions and unsaturated organic acids.139 In other cases, the products can form a separate phase and break down the layered matrix structure, as in the thermal decomposi- tion of complexes intercalated into the structure of hydrargillite (monoclinic gibbsite).140 V.Chemical properties of nanoscale systems In studies of nanoparticles and heterogeneous nanoscale systems it is of great importance to take into account a `feedback' between the reactivities of solid reactants and the character of the reaction between them.1 There are different forms of such a `feedback'. Sometimes, it can arise owing to the development of mechanical stress due to the absence of structural correspondence between the starting sub- stances and the reaction product (the so-called misfit).141, 142 In other cases, the reaction product can affect the electron ± hole equilibrium or the concentration of ionic defects in the region adjacent to the starting point of the reaction.Finally, the `feed- back' can manifest itself in transformation of a kinetically controlled reaction into a diffusion-controlled one if the molec- ular volume of the product is larger than the molecular volumes of the starting materials. It should also be kept in mind that the `feedback' can depend on the particle size of the solid reactants and products. 1. Properties of particles formed in the initial stages of heterogeneous chemical reactions Let us consider the oxidation of lower iron and manganese oxides as an example.Here, the size effect can manifest itself in the dependence of the composition of the reaction products on the sizes of their particles. In some instances, the mechanism of a solid-phase reaction can change upon change in the particle sizes of the reactants. This is thought to be due to transformation of a homogeneous system into a heterogeneous one, as is the case of, e.g., oxidation of Fe3O4 into Fe2O3.143, 144 Oxidation of fine Fe3O4 particles with a size less than 300 nm is a homogeneous process which involves the formation of a solid solution of oxygen in the magnetite lattice accompanied by gradual transformation of the Fe3O4 crystal lattice into the g-Fe2O3 lattice.Oxidation of larger N F Uvarov, V V Boldyrev 60 nm Figure 13. A HREM image of the AgBr ± silver stearate interface.153 Fe3O4 particles is a heterogeneous process since the Fe ions have no time to diffuse through the whole bulk of the particle, so it is only the surface layer that is oxidised. The appearance of the oxide concentration gradients and the development of stress in the surface layer of the particles facilitate the nucleation of rhombo- hedral Fe2O3. Analogously, fineMn3O4 particles are oxidised into Mn5O8, whereas oxidation of large particles gives a-Mn2O3.145 Similar effects seem to be expected for other systems. Size effect manifests itself in a somewhat different manner in the chemical and physical development of the latent photographic image, which begins as the Ag particle size exceeds a certain minimum value.146 ± 148 The usual chemical development involves treatment of silver bromide with a solution of a reducing agent and is due to the catalytic activity of Ag clusters formed upon aggregation of the surface silver atoms and comprising at least five atoms.149, 150 Physical development, or the appearance of an image owing to the deposition of the metal atoms from a solution on the potential crystallisation centres, occurs as the Ag crystal- lites become larger than the critical size of*4 nm.Palladium-free metallisation of dielectrics is similar to phys- ical development and proceeds analogously.151, 152 The process involves deposition of particles on a copper catalyst (obtained by thermal decomposition of copper hypophosphite) from a solution containing the copper salt and a reducing agent.The best results are obtained when the particle size of the copper catalyst is *25 nm. An increase in the particle size up to 80 nm leads to substantial reduction of the catalytic activity. Treatment of silver stearate with bromides results in the formation of nanoscale AgBr particles which serve as sensitisers for silver carboxylates; the latter are candidates for modern photographic processes.153 The case in point is as follows. Treat- ment of silver carboxylate with a KBr solution results in the formation of an AgBr particle and leads to the carboxylate lattice distortion at the AgBr ± silver carboxylate interface owing to misfit.A HREM image of the AgBr ± silver stearate interface is presented in Fig. 13. If the electron beam is oriented parallel to the basal plane of silver stearate, the layers of silver ions form rows of dark bands. As can be seen, these layers are disordered in the interface region and form an interfacial defect zone several nano- metres wide. This zone seems to play a significant role in the sensitisation of silver stearate contacting the AgBr particles. 2. Catalytic properties As mentioned above, the chemical properties of a substance including catalytic ones change on going from a bulk specimenSize effects in chemistry of heterogeneous systems to a nanocrystal.34, 154 For instance, the shifts of the energies of the conduction and valence bands in opposite directions cause changes in the photocatalytic properties of semiconductor par- ticles. Khairutdinov 34 has reported numerous examples of the effect of particle sizes on the optical and redox properties as well as on the kinetics of photochemical processes in various oxide and sulfide systems.In particular, it has been shown that the specific photocatalytic activity of TiO2 depends on its particle size. Unusual chemical properties of a-Fe2O3 ± SnO2 nanocomposites have also been reported.155 A salient feature of this system obtained by mechanical alloying of a mixture of oxides in high- energy mechanical activators is the high sensitivity and selectivity towards ethanol vapours and a moderate sensitivity to the accompanying gases, CO and CH4.This is mainly due to the very large area of the contact surface between the two oxides and to the presence of tin ions on the surface of a-Fe2O3 particles. As a result, the oxide a-Fe2O3, initially assumed to be insensitive to ethanol vapours, acquires sensitivity and a gas sensor based on it allows detection of 1000 ppm of ethanol vapour in air. A correlation was found between the particle size of a platinum catalyst and its catalytic activity in the reaction of hydrogenation of p-chloronitrobenzene, which is often used as a test for catalytic activity in hydrogenation reactions.156 Using commercial catalysts, the process can proceed up to the stage of formation of chloroaniline.The use of nanoscale platinum par- ticles with sizes 3 ± 4 nm allows the hydrogenation to reach deeper stages with the formation of aniline and cylohexylamine. Nano- crystalline platinum catalyst also appeared to be suitable for further hydrogenation of cylohexylamine to give dicylohexyl- amine as the end product. Cl Cl H2 H2 H2 7HCl 7H2O NH2 NH2 NO2 C6H5NH2, H2 7NH3 NH NH2 Finally, zinc ferrite nanoparticles were found to exhibit better sorption properties towardH2S as compared to conventional pure ferrite.125 3. Hydrogenation of metal systems Hydrogenation and dehydrogenation of magnesium have been studied.157 The use of nanocomposites formed upon mechanical treatment of mixtures of magnesium with nickel and other metals leads to an increase in the rate of hydrogenation and in the yields of the end products. The rate of hydrogen absorption and the hydrogen capacity of the mechanically alloyed composite with stoichiometryMg2Ni also increased.Taking LaNi5 as an example, it was shown that the addition of a hydrogenation catalyst (e.g., palladium) to the system together with mechanically treated absorbent leads to even greater increase in the rate of hydrogen intake. Nanocomposites obtained by mechanical treatment of Fe40Ti60, Fe50Ti50 and Fe67Ti33 alloys and by direct mechano- chemical alloying of elemental Fe and Ti were found to have a high hydrogen capacity. Compared to the starting alloys, these nano- composites absorb hydrogen at lower pressures and are stable towards the phase transition of the materials into the g-phase.Unusual properties of the alloys were explained by the presence of mechanical stress in the crystallites constituting the nanocompo- sites. 4. Electrochemical and anti-corrosion properties Electrodes made of mechanically alloyed Ti2Ni nanocomposite were found to be more efficient in the electrochemical production 281 of hydrogen gas than those produced by conventional proce- dures.158 A distinguishing feature of mechanically alloyed nanomate- rials is improved corrosion resistance.159 Nanostructured material with stoichiometry Fe32Ni36Cr14P12B6 obtained by crystallisation of an amorphous alloy exhibits a higher corrosion resistance than the initial alloy.160 This was explained by the possibility of fast diffusion of Cr atoms towards the surface along numerous block boundaries in the composite followed by the formation of a protective layer. A similar effect has also been observed.161 However, it does not always happen that the corrosion resistance increases.In studies of the behaviour of electrochemically depos- ited Ni ± P alloys in 0.1M H2SO4 (see Ref. 162) it was shown that nanocomposites exhibited a reduced corrosion resistance com- pared to bulk nickel owing to enhanced activities at block boundaries and the possibility of accelerated propagation of the reaction along these boundaries. Nanocomposites with stoichi- ometry FeAl8 also exhibit a reduced corrosion resistance towards H2SO4 and Na2SO4 solutions.163 Overvoltage (up to 900 mV) on steel cathodes is one of the main sources of energy loss in the electrochemical production of chlorates.It was proposed to use nanocrystalline cathodes pre- pared by mechanochemical activation of a RuO2 ± Fe ± Ti mix- ture.164 The mechanically alloyed nanocomposite appeared to be an excellent material for cathode production. At a current density of 250 mA cm72 and 70 8C, the overvoltage on the nanocompo- site cathodes decreases by 300 mV, which allows a 15% reduction of the expenditure of electrical energy in the production of chlorates. A nanostructured material obtained by mechanical alloying of Ti and TiO in an Ar atmosphere exhibits similar properties.165 The nanocomposite with oxygen content between 18.5 at.% and 33 at.% can be used as a cathode material in electrochemical production of sodium chlorate. At the operating current density of 250 mA cm72 and solution temperature of 70 8C, the overvoltage on this cathode is 600 mV, which is nearly half as large as the overvoltage on the cathode made of metallic titanium (1107 mV).5. Size effects in the chemistry of solid organic compounds Back in the late 1930s, it was found that stearic acid, which usually crystallises in stable b-form, undergoes transformation into the a-form as the crystal size reduces.166 As the size of molecular crystals decreases, the temperature of the a ± b-transition and the heat of melting also decrease.A change in the ratio between the surface area and volume of molecular crystals leads to changes in both dissolution rate and solubility. Reduction of the size of molecular crystals of chemical compounds used as medicinal drugs to submicron level enhances their biological assimilation and therapeutic efficiency. Since microscopic and nanoscale molecular crystals are prone to aggregation, it is desirable to place nanoparticles obtained by certain procedures on a support or to prepare nanocomposites of the molecular crystal ± support type. One of the well-known and popular methods for the preparation of this sort of nanocompo- sites involves treatment of the molecular crystal ± support mixture in a mechanical activator.This procedure can be conveniently used to prevent transformations of highly active, metastable polymorphous crystalline modifications formed during mechano- chemical activation into stable, low activity ones.167, 168 Here, it is appropriate to refer to a number of studies on the preparation of highly active, readily soluble forms of aspirin, sulfathiazole and other medicinal drugs in order to eliminate their major drawback of low solubility.169 ± 173 6. Specific chemical properties of nanocrystals of ionic compounds Incomplete compensation of the anion and cation charges in ionic compounds leads to the appearance of charged nanoparticles; this is a reason for changes in their chemical properties on going from bulk crystals to nanocrystals.174 This can be followed taking alkali282 2 1 3 4 6 5 X7 M+, Figure 14.Structures of a (3, 3, 3) nanocrystal of an ionic saltMX(1) and clusters that are reactive (2, 3) and passive (4 ± 6) towards chemisorption of NH3. [M14X13]+(1), [Na13F12]+(2), [Na22F21]+ (3), [M5X4]+ (4), [M14F13]+ (5) and [M23X22]+ (6).174 halides as an example. It is known that alkali halide vapours contain charged clusters comprised of, in particular, a small odd number of ions.175 In larger particles, electroneutrality can also be violated. Figure 14 presents the structure of a cubic nanocrystal of the (3, 3, 3) type (figures denote the number of ions along three edges). As compared to the ideal cubic lattice of ionic salt MX, the nanocrystal lattice is built in such a way that the X7 anions protrude out while the metal cations are recessed.Charged nanocrystals possess some characteristic properties. For instance, adsorption of polar molecules such asNH3 andH2O occurs in different manner depending on the number of lattice sites on the crystal edges. Perfect crystals of the (3, 3, 3) or (3, 3, 5) types do not adsorb polar molecules at room temperature. The loss of two ions (M+ and X7, or a single MX unit) leads to the appearance of pits on the surface and to enhancement of the reactivity of crystals. In Fig. 14, the structures of clusters [MnXn71]+ (M=Na, X=F) with such pits are shown. These clusters are highly reactive towards NH3. The structures of [MnXn71]+ clusters that are passive towards ammonia are also shown in Fig.14. It should be mentioned that the cluster structure remains unchanged after the adsorption of a polar molecule. The reaction of [Na53F52]+ nanoparticles with the surface of silicon occurring in their collisions is yet another example of the reactions of ionic nanoscale clusters. Upon collision, a single fluorine atom is transferred to the silicon surface to form a F7Si bond. The reaction occurs only if the silicon surface is pure. If the silicon surface is preliminarily passivated with hydrogen, the collision event will only result in the cleavage of the nanoparticle. Condensation of metal atoms on the surface of a nanoparticle leads to the formation of the structures similar to F-centres, where an electron is mainly localised at the anionic vacancy site.These centres are characterised by very low ionisation potentials, e.g., 3 eV for (Na+)n(F7)n71e7 and 5 eV for (Cs+)13(I7)11e7. As the temperature increases, the electron can wander onto the metal cation.Aneutral metal atom thus formed leaves the surface. In the case of a perfect nanocrystal without anionic vacancies, an electron that is loosely bound to the surface forms a surface centre, which is a strong electron donor. Photolysis of nano- crystals leads to transfer of the electron from the anion to the cation followed by desorption of a halogen atom from the surface. The transfer energy lies between the energies of analogous transfer in diatomic molecules MX and in bulk (MX)? crystals. By and large, studies of alkali halide nanocrystals showed that they are more passive towards many reagents as compared to bulk crystals.On the other hand, nanocrystals are prone to self- aggregation; therefore, the search for protective coatings for stabilisation of nanocrystals is topical. N F Uvarov, V V Boldyrev VI. Conclusion Size effects play an important role in chemical processes including mechanochemical reactions. One should distinguish two groups of size effects. Some of them, inherent in microcrystalline systems, are due to the purely surface phenomena. Stronger phase size effects manifest themselves if the particle size is less than 10 nm. In the former case, the properties of a substance can be interpreted in terms of the classical thermodynamic approach, which includes standard models for taking into account the contribution of surface effects.A basically new situation appears on going to nanoparticles since virtually the whole bulk of the sample under study falls within the action range of surface forces and classi- fication of the atoms into the surface atoms and bulk ones becomes unjustified. In this case, a nanoparticle should be considered to be a cluster or a macromolecule, the structural, thermodynamic and transport properties of which are strongly different from those of the macrocrystal. Various size effects are observed in heterogeneous systems represented by microcrystalline and nanoscale composites, the properties of which are affected by the interface interaction.Particularly strong are the interface effects typical of heteroge- neous systems with the ionic type of bonding. Here, the interface interaction involves the stage of formation of surface defects, which leads to abrupt increase in the ionic conductivity. Similarly to isolated nanoparticles, the material nanocomposites are made of occurs in unusual states and its properties (like those of the adsorption layers) are to a great extent determined by the characteristics of the interface, i.e., by the interatomic interactions within the adjacent phases in the near-surface region. Practically, nanocomposites represent a promising class of advanced materials with unique electronic properties and high molecular, atomic and ionic mobilities. Mechanochemical proc- esses in heterogeneous mixtures are strongly influenced by size effects, which is of great importance for solving practical tasks associated with controlling the process intensity and obtaining the desired reaction products.Problems which have arisen in this case are directly related to studies of the effect of particle size on the catalytic, electrochemical and specific chemical properties of materials. Therefore, further investigations of size effects in heterogeneous systems are of great theoretical and practical importance. This review has been written with the financial support of the Programme `Fundamental Problems of Modern Chemistry. Chemical Transformations Involving Nanoscale and Supramo- lecular Systems' of the Russian Federation Ministry of Industry, Science and Technology (Grant No. 3) and the US Civilian Research and Development Foundation (Grant REC-008). References 1.V Boldyrev,M Boulens, B Delmon The Control of Reactivity of Solids (Amsterdam: Elsevier, 1979) 2. I D Morokhov, L I Trusov, S P Chizhik Ul'tradispersnye Metallicheskie Sredy (Ultradispersed Metallic Media) (Moscow: Atomizdat, 1977) 3. Yu I Petrov Klastery i Malye Chastitsy (Clusters and Small Particles) (Moscow: Nauka, 1986) 4. S A Nepiiko Fizicheskie Svoistva Malykh Metallicheskikh Chastits (Physical Properties of Small Metallic Particles) (Kiev: Naukova Dumka, 1985) 5. Yu I Vesnin Vtorichnaya Struktura i Svoistva Kristallov (Secondary Structure and Properties of Crystals) (Novosibirsk: Institute of Organic and Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, 1997) 6.L S Palatnik, I I Papirov Epitaksial'nye Plenki (Epitaxial Films) (Moscow: Nauka, 1971) 7. Yu F Komnik Fizika Metallicheskikh Plenok. Razmernye i Strukturnye Effekty (The Physics of Metallic Films. Size and Structural Effects) (Moscow: Atomizdat, 1979)Size effects in chemistry of heterogeneous systems 8. V G Gryaznov, L I Trusov Prog. Mater. Sci. 37 289 (1993) 9. H Gleiter Prog. Mater. Sci. 33 223 (1989) 10. H Gleiter J. Appl. Cryst. 24 79 (1991) 11. R A Andrievskii, AMGlezer Fiz. Met. and Metalloved. 88 50 (1999) a 12. A I Gusev Usp. Fiz. Nauk 168 55 (1988) b 13. V L Tauson, M G Abramovich Fiziko-Khimicheskie Prevrashche- niya Real'nykh Kristallov vMineral'nykh Sistemakh (Physicochemical Transformations of Real Crystals in Mineral Systems) (Novosibirsk: Nauka, 1988) 14.A I Rusanov Solid State Ion. 75 275 (1995) 15. A I Rusanov Zh. Obshch. Khim. 70 353 (2000) c 16. A I Rusanov Fazovye Ravnovesiya i Poverkhnostnye Yavleniya (Phase Equilibria and Surface Effects) (Leningrad: Khimiya, 1967) 17. P Pawlow Z. Phys. Chem. 65 545 (1909) 18. A I Bublik, B Ya Pines Dokl. Akad. Nauk SSSR 87 215 (1952) d 19. B Ya Pines Ocherki o Metallofizike (Essays on Metallophysics) (Khar'kov: Khar'kov University, 1961) 20. C R M Wronski Br. J. Appl. Phys., J. Phys. D 18 1731 (1967) 21. M Hasegawa, K Hoshino, M Watabe J. Phys.F., Met. Phys. 10 619 (1980) 22. V A Pavlov Fiz. Met. Metalloved. 59 629 (1985) a 23. Yu G Sidorov, E M Trukhanov Poverkhnost' (6) 106 (1992) 24. I D Morokhov, S P Chizhik, N T Gladkikh, L K Grigor'eva, S V Stepanova Dokl. Akad. Nauk SSSR 248 603 (1979) d 25. H Gleiter Nanostruct. Mater. 1 1 (1992) 26. R W Siegel Phys. New Mater. 27 65 (1994) 27. D G Morris Mechanical Behaviour of Nanostructured Materials Vol. 2 (EdsM Magini, F H Wohlbier) (Materials Science Foundations Trans. Technology Publications, Zuerich: 1998) 28. G C Benson, T A Claxton J. Chem. Phys. 48 1356 (1968) 29. P W Tasker Philos. Mag. A Phys. Condens. Matter Defects Mech. Prop. 39 119 (1979) 30. N X Sun, K Lu Phys. Rev. B, Condens. Matter 54 6058 (1997) 31. J Friedel Ann.Phys. 1 257 (1976) 32. M R Hoare, P Pal Adv. Phys. 20 161 (1971) 33. A P Alivisatos Science 271 933 (1996) 34. R F Khairutdinov Usp. Khim. 67 125 (1998) [Russ. Chem. Rev. 67 109 (1998)] 35. C L Chien, J Q Xiao, J S Jiang J. Appl. Phys. 73 5309 (1993) 36. D S Lashmore,M P Dariel J. Electrochem. Soc. 135 1221 (1988) 37. S S Parkin, Z C Li, D J Smith Appl. Phys. Lett. 58 2710 (1991) 38. E Yu Ivanov, I G Konstanchuk, B B Bokhonov, V V Boldyrev React. Solids 7 167 (1989) 39. J Eckert, L Schultz, E Hellstern Appl. Phys. Lett. 55 117 (1989) 40. S Ino J. Phys. Soc. Jpn. 27 941 (1969) 41. D J Wales Science 271 925 (1996) 42. T L Hill Thermodynamics of Small Systems Part I (New York: Benjamin, 1963) 43. I I Gainutdinov, Yu T Pavlyukhin, V V Boldyrev Dokl.Akad. Nauk 344 189 (1995) d 44. I I Gainutdinov, Yu T Pavlyukhin Dokl. Akad. Nauk 350 206 (1996) d 45. Ya E Geguzin Fizika Spekaniya (The Physics of Caking) (Moscow: Nauka, 1984) 46. B S Bokshtein, Ch V Kopetskii, L S Shvindlerman Termodinamika i Kinetika Granits Zeren v Metallakh (Thermodynamics and Kinetics of Grain Boundaries in Metals) (Moscow: Metallurgiya, 1986) 47. J Maier Ber. Bunsenges. Phys. Chem. 90 26 (1986) 48. J Maier Prog. Solid State Chem. 23 171 (1995) 49. V N Chebotin Fizicheskaya Khimiya Tverdogo Tela (Physical Chemistry of Solids) (Moscow: Khimiya, 1982) 50. I M Lifshits, A M Kosevich, Ya E Geguzin, in Poverkhnostnaya Diffuziya i Rastekanie (Surface Diffusion and Spreading) (Ed. Ya E Geguzin) (Moscow: Nauka, 1969) p.243 51. Ya I Frenkel' Kineticheskaya Teoriya Zhidkostei (Kinetic Theory of Liquids) (Moscow: Academy of Sciences of the USSR, 1945) 52. K L Kliewer, K S Koehler Phys. Rev. A., At. Mol. Opt. Phys. 140 1226 (1965) 53. N F Uvarov, Doctoral Thesis in Chemical Sciences, Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 1998 54. J Horvath, U Landman Phys. Rev. Lett. 82 3835 (1999) 55. J Horvath, R Birringer,H Gleiter Solid State Commun. 62 319 (1987) 56. B S Bokshtein, G D Breze, P V Kurkin Izv. Vyssh. Uchebn. Zaved., Chernaya Metallurgiya 1 49 (1994) 283 57. H J Hoffler, R S Averback, H Hahn, H Gleiter J. Appl. Phys. 74 3832 (1993) 58. S M Klotsman Fiz.Met. Metalloved. 75 5 (1993) a 59. R E Newnham, D P Skinner, L E Cross Mater. Res. Bull. 13 525 (1978) 60. U Stuhr,H Wipf, T J Udovic, J Weissmueller,H Gleiter Nanostruct. Mater. 6 555 (1995) 61. B Ya Pines, N I Sukhinin Zh. Tekhn. Fiz. 26 2100 (1956) e 62. A A Chernov, E I Givargizov, Kh S Bagdasarov, L N Dem'yanets, V A Kuznetsov,A N Lobachev (Eds) Sovremennaya Kristallografiya (Modern Crystallography) (Moscow: Nauka, 1980) Vol. 3 63. N F Uvarov, J Maier Solid State Ion. 61 251 (1993) 64. N F Uvarov, V G Ponomareva Dokl. Aksd. Nauk 351 358 (1996) d 65. N F Uvarov, E F Hairetdinov, B B Bokhonov, N B Bratel Solid State Ion. 86 ± 88 (Pt I) 573 (1996) 66. N F Uvarov, B B Bokhonov, V P Isupov, E F Hairetdinov Solid State Ion. 74 15 (1994) 67.N F Uvarov, E F Hairetdinov, I V Skobelev Solid State Ion. 86 ± 88 577 (1996) 68. N F Uvarov, I V Skobelev, B B Bokhonov, E F Hairetdinov J. Mater. Synth. Proc. 4 391 (1996) 69. A Adamson Physical Chemistry of Surfaces (New York: Wiley, 1974) 70. A D Zimon Adgeziya Plenok i Pokrytii (Adhesion of Films and Coatings) (Moscow: Khimiya, 1977) 71. M J Jaycock, G D Parfitt Chemistry of Interfaces (New York: Ellis Horwood, 1981) 72. Ya O Bikerman Usp. Khim. 41 1431 (1972) [Russ. Chem. Rev. 41 651 (1972)] 73. A I Rusanov Dokl. Akad. Nauk SSSR 261 700 (1981) d 74. J Maier J. Phys. Chem. Solids 46 309 (1985) 75. A K Shukla, R Manoharan, J B Goodenough Solid State Ion. 26 5 (1988) 76. B Zhu, B-E Mellander Solid State Ion. 77 244 (1995) 77.B Zhu Solid State Ion. 125 397 (1999) 78. B Zhu, B-E Mellander Solid State Ion. 70 (Pt 1) 285 (1994) 79. Z-Y Zhao, C-Y Wang, S-Y Dai, L-Q Chen Solid State Ion. 9/10 1175 (1983) 80. Yu Ya Gurevich Dokl. Akad. Nauk SSSR 222 143 (1975) d 81. J B Boyce, B A Huberman Phys. Rep. 51 190 (1979) 82. N G Hainovsky, J Maier Phys. Rev. B, Condens. Matter 51 15789 (1995) 83. V N Bondarev, A B Kuklov Elektrokhimiya 26 1397 (1990) f 84. N F Uvarov, E F Hairetdinov J. Solid State Chem. 62 3 (1985) 85. N F Uvarov, V P Isupov, V Sharma, A K Shukla Solid State Ion. 51 41 (1992) 86. V M Kosevich, V M Ievlev, L S Palatnik, A I Fedorenko Struktura Mezhkristallitnykh i Mezhfaznykh Granits (Structure of Crystallite and Phase Boundaries) (Moscow: Metallurgiya, 1980) 87.L S Palatnik,M Ya Fuks, V M Kosevich Mekhanizm Obrazovaniya i Substruktura Kondensirovannykh Plenok (The Mechanism of Formation and Substructure of Condensed Films) (Moscow: Nauka, 1972) 88. W A Jesser, D Kulmann-Wilsdorf Phys. Status Solidi B 19 95 (1967) 89. V A Gusev, V A Gagarina, E M Moroz, E A Levitskii Kinet. Katal. 17 500 (1976) g 90. N F Uvarov, M C R Shastry, K J Rao Rev. Solid State Sci. 4 61 (1990) 91. E F Hairetdinov, N F Uvarov Mater. Sci. Forum 88 ± 90 723 (1992) 92. N F Uvarov, E F Khairetdinov,N B Bratel' Elektrokhimiya 29 1406 (1993) f 93. N F Uvarov, P Vanek,M Savinov, V Zelezny, J Studnicka, J Petzelt Solid State Ion. 127 253 (2000) 94. N F Uvarov, A A Politov, in Abstracts of Reports of the 12th International Conference on Solid State Ionics, Thessaloniki, 1999 M-06-P 95.N F Uvarov, P Vanek, Yu I Yuzyuk, V Zelezny, J Studnicka, B B Bokhonov, V E Dulepov, J Petzelt Solid State Ion. 90 201 (1996) 96. V G Ponomareva, N F Uvarov, G V Lavrova, E F Hairetdinov Solid State Ion. 90 161 (1996) 97. V G Ponomareva,G V Lavrova, L G Simonova Solid State Ion. 118 317 (1999) 98. N F Uvarov, L I Brezhneva, E F Hairetdinov Solid State Ion. 136 ± 138 1273 (2000)284 99. P Chowdary, V B Tare, J B Wagner J. Electrochem. Soc. 132 123 (1985) 100. A C Khandkhar, J B Wagner Solid State Ion. 20 267 (1986) 101. C A Angell, D B Helphrey J. Phys. Chem. 75 2306 (1971) 102. A Kumar, K Shahi Solid State Ion. 68 71 (1994) 103. G Heinike Tribochemistry (Berlin: Akademie-Verlag, 1984) 104.V V Boldyrev Eksperimental'nye Metody v Mekhanokhimii Tverdykh Neorganicheskikh Veshchestv (Experimental Methods in Mechanochemistry of Solid Inorganic Substances) (Novosibirsk: Nauka, 1983) 105. C C Koch Annu. Rev. Mater. Sci. 19 121 (1989) 106. E G Avvakumov Mekhanicheskie Metody Aktivatsii Khimicheskikh Protsessov (Mechanical Methods of Activation of Chemical Processes) (Novosibirsk: Nauka, 1986) 107. Mekhanokhimicheskii Sintez v Neorganicheskoi Khimii (Mechano- chemical Synthesis in Inorganic Chemistry) (Novosibirsk: Nauka, 1991) 108. Proceedings of the 1st International Symposium on Mechanical Alloying, Kyoto, 1991 109. P Yu Butyagin Usp. Khim. 63 1031 (1994) [Russ. Chem. Rev. 63 965 (1994)] 110.E Gaffet, M Abdellaoui, N N Malhourux-Gaffet Mater. Trans. JIM 36 198 (1995) 111. M Dadras, D S Morris Scr. Metall. Mater. 28 1245 (1993) 112. T Itsukaichi, M Umemoto, J G Cabanas-Moreno Scr. Metall. Mater. 29 583 (1993) 113. E Ivanov, K Sumiyamann, H Yamauchi, K Jamai, K Suzuki J. Alloys Compd. 198 195 (1993) 114. P Matteazzi, G Le Caer, A Mocellin Ceram. Int. 23 39 (1997) 115. K B Gerasimov, S V Pavlov, S V Mytnichenko, V A Chernov, S G Nikitenko J. Alloys Compd. 252 179 (1997) 116. M A Uimin, A Ye Yermakov, V V Serikov, A Yu Korobeinikov, A K Shtolz Phys. Status Solidi A 165 337 (1998) 117. V A Ivchenko, M A Uimin, A Ye Yermakov, A Yu Korobeinikov Surf. Sci. 440 420 (1999) 118. A V Demenko, A I Maslii, V V Boldyrev J. Mater.Synth. Proc. 3 303 (1995) 119. A R Yavary, O Dabrohlav Mater. Sci. Forum 225 ± 227 295 (1996) 120. M L Trudeau,R Schulz,D Dussault,A Van Neste Phys. Rev. Lett. 64 99 (1990) 121. R Schulz, M Simoneau, J Lantegne Physica C 233 113 (1994) 122. R Schulz, M L Trudeau, D Dussault, A Van Neste, L Dignard- Bailey Mater. Sci. Eng. A, Struct. Mater. 179 516 (1994) 123. P Matteazzi, G LeCaer J. Am. Ceram. Soc. 74 1382 (1991) 124. P Matteazzi, G LeCaer Proc. Adv. Mater. 1 135 20 (1991) 125. P Matteazzi, G LeCaer J. Alloys Compd. 187 305 (1992) 126. A Ye Yermakov, N V Mushnikov,M A Uimin, N K Zajkov, A Yu Korobeinikov, A K Shtolz, V V Serikov Mater. Sci. Forum 269 ± 272 907 (1998) 127. G Nicoara, D Fratiloiu, M Nogues, J L Dormann, F Vasiliu Mater. Sci. Forum 235 ± 238 145 (1997) 128. P G McCormick, in Handbook on the Physics and Chemistry of Rare Earth Vol. 24 (Amsterdam: Elsevier, 1997) p. 47 129. T Tsuzuki,W T A Harrison, P G McCormick J. Alloys Compd. 281 146 (1998) 130. T Tsuzuki, J Ding, P G MacCormick Physika B 239 378 (1997) 131. T F Grigorjeva, A P Barinova, G N Kryukova, V V Boldyrev, in Proceedings of the 5th European Conference on Advanced Materials, Processes and Applications, Dodrecht, 1997 Vol. 2, p. 333 132. V Sepelak, A Yu Rogachev, U Steinike, D-Chr Uecker, F Krumeich, S Wissman, K D Becker Mater. Sci. Forum 235 ± 238 139 (1997) 133. B B Bokhonov, I G Konstanchuk, V V Boldyrev Mater. Res. Bull. 30 1277 (1995) 134. A P Nemudry React. Solids 3 317 (1987) 135. A V Besserguenev, A M Fogg, R J Francis, S J Price, D O'Hare, V P Isupov, B P Tolochko Chem. Mater. 9 241 (1997) 136. V P Isupov, Doctoral Thesis in Chemical Sciences, Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 1998 137. V Rives, M A Ulibarri Coord. Chem. Rev. 181 61 (1999) 138. V P Isupov, L E Chupakhina, R P Mitrofanova, K A Tarasov Zh. Strukt. Khim. 39 453 (1998) h N F Uvarov, V V Boldyrev 139. E I Bakchinova, V P Isupov, V A Poluboyarov, V V Boldyrev Dokl. Akad. Nauk 324 592 (1992) d 140. V P Isupov, K A Tarasov, L E Chupakhina Dokl. Akad. Nauk 324 1217 (1992) d 141. V V Boldyrev, E V Boldyreva Mater. Sci. Forum 88 ± 90 711 (1992) 142. E V Boldyreva J. Therm. Anal. 38 89 (1992) 143. W Feitknecht Pure Appl. Chem. 9 423 (1964) 144. B Gillot, A Rousset, G Dupre J. Solid State Chem. 25 263 (1978) 145. B Gillot, J Tyranowic, A Rousset Mater. Res. Bull. 10 775 (1975) 146. H Reinder, L Hamburger Recl. Trav. Chim. Pays-Bas 50 475 (1931) 147. E A Galashin Dokl. Akad. Nauk SSSR 184 639 (1969) d 148. C E Mees, T H James The Theory of Photographic Process (New York: Macmillan, 1966) 149. J F Hamilton, P C Logel Photogr. Sci. Eng. 18 507 (1974) 150. J W Mitchell J. Imaging Sci. Technol. 38 193 (1995) 151. O I Lomovskii, Doctoral Thesis in Chemical Sciences, Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch, Academy of Sciences of the USSR, Novosibirsk, 1990 152. A Ya Lushnikov, O I Lomovskii Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (2) 18 (1989) 153. B B Bokhonov, L P Burleva,WC Frank, JRMiller,MRV Sahyun, D R Whitcomb, J M Winslow, C Zou J. Imaging Sci. Technol. 40 85 (1996) 154. J F Hamilton J. Vac. Sci. Technol. 13 319 (1976) 155. J Z Jiang, R Lin,W Lin, K Nielsen, S Morup, K Dam-Johansen, R Clasen J. Phys. D Appl. Phys. 30 1459 (1997) 156. G Vitulli, E Pitzalis, A Verrazani, P Pertici, P Salvadori, G Marta Mater. Sci. Forum 235 ± 238 929 (1997) 157. I G Konstanchuk, E Yu Ivanov, V V Boldyrev Usp. Khim. 67 75 (1998) [Russ. Chem. Rev. 67 69 (1998)] 158. T Benameur, B Rezgui, A R Yavary, R Durand Mater. Sci. Forum 235 ± 238 917 (1997) 159. E M Gutman Mekhanokhimiya i Zashchita Metallov ot Korrozii (Mechanochemistry and Protection of Metals from Corrosion) (Moscow: Metallurgizdat, 1981) 160. S J Trope, B Ramaswami, K T Aust J. Electrochem. Soc. 135 2162 (1988) 161. O El Kedim,M Tachikart, E Gaffet Mater. Sci. Forum 225 ± 227 825 (1996) 162. R Rofagha, R Langer, A M El-Sherik, U Erb, G Palumbo, K T Aust Scr. Metall. Mater. 25 2867 (1991) 163. M Schneider,W Zeiger, D Scharnweber, H Worch Mater. Sci. Forum 225 ± 227 819 (1996) 164. A Van Neste, S H Yip, S Jin, S Boily, E Ghali, D Guay, R Schulz Mater. Sci. Forum 225 ± 227 795 (1996) 165. E Tremblay, O Savadogo, S Boily, A Van Neste, R Schulz Mater. Sci. Forum 225 ± 227 813 (1996) 166. P A Thiessen Angew. Chem. 51 318 (1938) 167. A Ikekawa, S Hoyakawa Bull. Chem. Soc. Jpn. 54 2587 (1981) 168. A M Dubinskaya Khim.-Farm. Zh. 23 755 (1989) 169. T Shakhtshneider, V V Boldyrev, in Reactivity of Molecular Solids (Eds E Boldyreva, V V Boldyrev) (Chichester: Wiley, 1999) p. 271 170. A V Dushkin, Z U Rykova, T P Shakhtshneider, V V Boldyrev Int. J. Mechanochem. Mech. Alloying 1 48 (1994) 171. A Yu Yagodin, A V Dushkin, V V Boldyrev Pharmazie 3 69 (1991) 172. V V Boldyrev, T P Shahtshneider, L P Burleva, V A Severzev Drug Develop. Industr. Pharm. 20 1103 (1994) 173. T P Shakhtshneider, M A Vasilchenko, A A Politov, V V Boldyrev Int. J. Pharm. 130 25 (1996) 174. R L Whetten Acc. Chem. Res. 26 49 (1993) 175. P Davidovits, D L McFadden (Eds) Alkali Halide Vapours (New York: Academic Press, 1979) a�Phys. Met. Metallogr. (Engl. Transl.) b�Physics-Uspekhi (Engl. Transl.) c�Russ. J. Gen. Chem. (Engl. Transl.) d�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) e�Tech. Phys. (Engl. Transl.) f�Russ. J. Electrochem. (Engl. Transl.) g�Kinet. Catal. (Engl. Transl.) h�Russ. J. Struct. Chem. (Engl. Transl.) i�Chem. Pharm.

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (4) 285 ± 298 (2001) Size effects in electrochemistry O A Petrii, G A Tsirlina Contents I. Introduction II. Size effects in diffusion kinetics. Microelectrodes and microcells. III. Electrical double layer as a nanostructure IV. Size effects in electrocatalysis V. Size effects in the elementary act of charge transfer VI. Size effects in processes of new phase formation in electrochemical systems VII. Electrochemical nanotechnologies VIII. Conclusion Abstract. in phenomena size-dependent of characteristics General General characteristics of size-dependent phenomena in electrochemical systems are given. Primary attention is paid to electrochemical systems are given. Primary attention is paid to methodical a is which nanoelectrochemistry, of achievements methodical achievements of nanoelectrochemistry, which is a line line of of development The years.15 last the over created research of research created over the last 15 years. The development of the the main stream the by initiated electrochemistry of concepts main concepts of electrochemistry initiated by the stream of of nanoscopic for prospects The considered. is information nanoscopic information is considered. The prospects for local local studies steps elementary interfaces, charged on processes of studies of processes on charged interfaces, elementary steps of of these and nanoelectrodes of application and processes these processes and application of nanoelectrodes and related related systems The discussed. are fields interdisciplinary in systems in interdisciplinary fields are discussed.The bibliography bibliography includes 198 includes 198 references. references. I. Introduction Electrochemistry studies homogeneous condensed ionic systems (solutions, melts, solid electrolytes) and heterogeneous systems which include electrified interfaces. The spatial scales of hetero- geneous electrochemical systems vary over a wide range (Table 1). Correspondingly, the size effects observed in them are numerous and diversified, and quite a number of them are analogous to those in homogeneous electrochemical systems. However, the latter are beyond the scope of this review. Characteristic dimensions of conventional components of electrochemical systems (electrodes, electrolytes and membranes) cover 5 ± 10 orders of magnitude.The lower boundaries of the corresponding intervals approach typical sizes of specific space regions (layers) that arise only at the electrode ± electrolyte, electrode ± membrane and electrolyte ± membrane junctions. In electrochemical systems, the concentration distribution of com- ponents near the electrode differs from that in the bulk. As the distance from the interface increases, this distribution changes, i.e., the system contains nonuniform regions. The structure of such a region is considered in detail in Refs 1 ± 3. Below, we only briefly comment on the concepts employed in this article. O A Petrii, G A Tsirlina Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, Moscow 119899, Russian Federation.Fax (7-095) 939 01 71. Tel. (7-095) 939 55 01. E-mail: petrii@elch.chem.msu.ru (O A Petrii) Tel. (7-095) 939 13 21 (G A Tsirlina) Received 21 November 2000 Uspekhi Khimii 70 (4) 330 ± 344 (2001); translated by T Ya Safonova #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n04ABEH000639 Table 1. Characteristic scales of elements of electrochemical systems and devices. Elements Electrodes for practical applications for research microelectrodes ultramicroelectrodes thin-film electrodes Electrolyte layers electrolysers research cells thin-layer cells microcells Membranes Diffusion layers Chemisorption layers `Electronic tails' and near-surface layers in semiconductors When an electrode process occurs on an interface, the boun- dary layer is depleted of the corresponding reactant and(or) enriched with the products formed.Insofar as diffusion rates are finite, a region always exists in which the concentration of one or several components depends on the distance. We call this region the diffusion layer and usually consider it as an electrolyte component. A uniform distribution of components can also be disturbed under equilibrium conditions. Boundary regions that arise in the process are called the `electrical double layer' (EDL).{ Charged species are distributed in the EDL under the effect of different electrostatic interactions, the concentration of ions either decreas- ing or increasing with an increase in the distance from the { This name is associated with the concept of metallic and ionic plates of a hypothetical capacitor commonly used as the interface model.285 286 288 289 291 292 293 295 Size intervals Components /m metals, semicon- ductors, composi- tion materials 0.1 ± 10 1073± 1072 1075± 1074 1078± 1076 <1078 solutions, melts, solid electrolytes 0.1 ± 10 1073 ± 0.1 1076± 1075 1077± 1076 1077± 1072 1076± 1074 polymers, porous inorganic materials all electrolyte components 10710 ±1079 the same <10711 electrons and other charge carriers286 boundary. If the interaction of an ion with a charged electrode is much stronger than its interaction with induced charges, then the concentration distribution much resembles the diffuse distribu- tion of charges in an ionic atmosphere and is described by a parameter having the meaning of the Debye screening length and equivalent to the effective thickness of the diffuse part of EDL, namely, the ionic diffuse layer.If certain components of a system specifically interact with the electrode material, then a chemisorption layer (the dense part of EDL) is formed between the surface and the diffuse layer. For the vast majority of electrode ± solution systems, dense layers are formed involving solution molecules and often exclusively by such molecules. According to modern notions of the interface structure, the adsorbate layers and the `edges' of surface atoms on metal electrodes are spatially separated with a gap in between attributed to the emerging electronic gas (`electronic tail').In electrochemical systems, the properties of this gap depend on the electrode potential (charge). For semiconducting electrodes, the spatial dispersion of charge carriers is also characteristic of near-to- surface solid-phase layers. We can assess the properties of electrochemical systems from electrochemical responses, viz., different dependences of electrical signals (current, potential, voltage) on time and also on polar- isation conditions (for instance, modulation frequency and ampli- tude during ac polarisation). Size effects arise when the system's parameters are commensurate with the characteristic quantities that determine a particular response.We can distinguish several basic groups of size effects caused either by the scale commensurability of two and more subsystems, or `intrinsic' size-dependent properties of individual subsystems. Each of these groups merits special consideration (e.g., see Refs 4 ± 30). In this review, we consider the general features of size effects in electrochemical systems and give them a sort of classification. The prospects of using size effects for intensifying electrode processes and also for technological (including nano- technological) applications are discussed. The presented list of references does not exhaust all the literature (otherwise, it would be extremely large) and contains, apart from reviews, only representative papers of recent years.Size-dependent thermodynamic properties of electrochemical systems related to equilibria involving two-dimensional adsorp- tion layers and nanocrystals are the subject of an interdisciplinary research field that shares its frontiers with colloid chemistry, surface science and solid-state physical chemistry. Studies of other electrochemical size effects also show a trend for integration and are combined by a recently formed separate research field, viz., nanoelectrochemistry. Along with inherent electrochemical techniques, this actively developing new scientific discipline extensively uses probe techniques for local surface stud- ies,14, 16, 17, 20, 21, 31 ± 34 diffraction 35 techniques, and also opti- cal 36, 37 and small-angle X-ray methods 38 (including those in situ 39).Originally, nanoelectrochemistry was developed on the basis of studying classical size effects (for instance, the dependences of surface conductivity on film thickness 8). However, during the past decade, it has begun to encroach more actively on the province of quantum size-dependent phenomena. The develop- ment of nanoelectrochemistry is also stimulated by the progress in electronics; the latter branch, in turn, diversifies its elemental basis due, in particular, to the introduction of micro- and nanoelec- trochemical technologies. The reason for such a `feedback' resides in the fact that by extending the scope of electrochemical systems to the nanoscale, one substantially strengthens the material basis of electrochemistry and sharply enhances its experimental poten- tialities. The possibilities of experimental studies of micro- and nano- electrochemical systems are largely determined by the facilities for measuring small currents and charges.The characteristic settling times usually lie within hundreds of milliseconds, and, even if we O A Petrii, G A Tsirlina pass to the femtoampere range, the lower limit of measured charge will fall within 10716 and 10715 C. These values exceed the charges consumed in electrochemical reactions of single ions and molecules approximately by three orders of magnitude. At present, the unique measurements with nanosecond resolution have severe restrictions on the lower limit of the measured current.The crucial obstacle that prevents lowering of the limit mentioned results from the fact that characteristic polarisation capacities of micro- and nanoelectrochemical systems are commensurable with the inherent capacities of the measuring devices. II. Size effects in diffusion kinetics. Microelectrodes and microcells. The uppermost limit on the scale of size-dependent properties of electrochemical systems is occupied by those manifesting them- selves in electrochemical processes with diffusion-controlled rates. Such properties are studied in detail for the systems in which the dimensions of either electrodes or layers of electrolyte solutions are commensurable with the diffusion layer thickness (micrometer range that may be upscaled, for instance, due to gravitation effects).That is why, when separating microelectrodes and micro- cells into a special class of electrochemical systems, the upper limits of sizes (volumes) cannot always be determined unambigu- ously, being dependent on the configuration and convection conditions. For the majority of microelectrodes, such a limit is *10 mm. For disk, spherical, semispherical, conical and microdrop liquid electrodes exact relationships for the dependences of the limiting diffusion current Id on the characteristic size r were obtained.40 ± 42 Similar size limits for microcells are determined by a condition of the overlapping of diffusion layers for two electrodes between which the current flows.43 The corresponding quantitative expres- sions are very complicated; hence, the applications of microcells are largely related to in situ surface spectroscopy (the so-called thin-layer cells for minimising absorption in the solution bulk) and microtechnologies.In those cases, the cell volume contains actually no regions with constant concentrations of reactants, and different non-steady-state size-dependent distributions of solution components can arise. The most important basic application of microelectrodes deals with the determination of kinetic parameters of electrode proc- esses and is limited by the ratio between the corresponding rate constants and the limiting currents of reactant delivery. As an example, we consider the limiting diffusion current density id, which corresponds to the rate of an electrode process at a slow delivery of a reactant.For an electrode of a common (macroscopic) size r exceeding substantially the diffusion layer thickness, id is independent of the electrode configuration and determined only by the diffusion coefficient of the reactant D and its concentration c. Correspondingly, for an electrode of a geo- metric surface area S, the diffusion current (1) Id=idS=const cDmr2, where the power index m<1. The values of the constant and m are determined by the parameters that characterise the convection conditions (rotation velocity of a solid electrode, streaming velocity of a liquid metal from a capillary, etc.) and by the electrode shape and described by the rigorous relationships deduced within the framework of the planar diffusion model.Inasmuch as id and the rate of electron transfer stage are proportional to c, it is difficult to overcome the diffusion limitations and determine the rate constant for the electron transfer on common-size electrodes for quick processes. When the diffusion layer thickness is commensurable with electrode dimensions, the planar diffusion model does not operate due to a significant role of edge effects. For example, for a disk microelectrode, (2) id=const0 c D pAAAAr¡1.Size effects in electrochemistry log id, log Id , log IdR (arb. u.) 2 VI 4 1 V 0 IV 74 3 III II 78 I log r (cm) 72 76 710 Figure 1.Size dependences of electrochemical responses (1 ± 3) and typical scales of various nanotechnological processes (I ± VI); (1) id, (2) Id, (3) IdR; (I) nanostructurisation in STM configuration, (II) electron-beam technologies, (III) ion-beam technologies, (IV) holographic lithography, (V) X-ray lithography, (VI) photolithography. Plotted on the basis of the results taken from Refs 19, 24. Correspondingly, Id turns out to be proportional to r rather than to r2, which affects the dependence of the ohmic potential drop IdR on r (whereRis the resistance of the solution layer between the working electrode and a point in which the potential is measured by a reference electrode). Figure 1 illustrates how the most important characteristics of electrochemical systems tend to vary. In the case of fast delivery of a reactant, regardless of the electrode size, the rate of an electrode process is proportional to the surface area (in the absence of other types of size effects, see Sections IV and V).Hence, with a decrease in r, it becomes possible to overcome the diffusion limitations for still faster reactions. Microelectrodes allow quantitative measurement of the rate constants of an order of magnitude of centimetres per second, whereas the limits of conventional rotating and dropping electrodes do not exceed 1072 cm s71. To determine the rate constants of fast reactions, the arrays of microelectrodes which include flow-through channel systems with ensembles of microband electrodes are also widely used.44 A rotating ring-disk electrode can be considered as the first array electrode which allowed measurement of the kinetics of separate steps of electrode processes, owing to its optimum hydrodynam- ics.45 The electrochemical responses of electrode arrays are characterised by special types of size dependences determined by the relative positions of their fragments (for instance, by the spaces between them).Microelectrodes are widely used in electroanalysis,46, 47 espe- cially for solving environmental and biological problems (for example, in electrochemical measurements in vivo in the monitor- ing of the composition of biological media, diffusion of drugs and poisons, accumulation of glucose, etc.). These applications stimu- lated further miniaturisation of electrodes. As a result, a large class of the so-called ultramicroelectrodes with surface areas of less than 1 mm2 was developed.48, 49 Modern methods of insulation of the surfaces of metal wires and other items provide electrodes of characteristic sizes of *10 nm (sometimes named nanodes).Standardisation of the geometry (shape) of nanodes presents a particular problem which can be solved by using nanosize materials of known structures, for example, carbon nanotubes as the electrodes.50 The term `nanoelectrodes' is sometimes used in a different context when considering the phenomena in solutions of mono- dispersed colloidal metals. Thus when stabilising colloidal gold particles in dichloroethane by ferrocene-containing organic mol- ecules, the responses of redox transformations of the latter could 287 be detected in the single contacts of particles (nanoelectrodes) with the polarising macroelectrodes.51 In this case, an analogy can be traced with well known suspension electrodes in which macro- electrodes play the role of current collectors.52 A new impact for the development of microelectrode technol- ogy was made in the end of the 1980s when Bard et al.15 proposed a basically new method for carrying out local studies of surface processes and phenomena, namely, the technique of scanning electrochemical microscopy (SECM).This technique is based on the possibility of measuring the current on a probe microelectrode as a function of the distance h between the probe and the surface of an electrode under study of arbitrary dimensions dipped into a solution containing a redox system with a high electron-transfer rate constant and known diffusion coefficients of redox forms.When the diffusion layers of the probe and the electrode under study overlap (which occurs if h is commensurable or less than r), the processes on both electrodes can either be accelerated or decelerated, depending on the potentials of the probe and the electrode. The corresponding relationships for steady-state and non-steady-state currents in SECM configuration were derived as two-dimensional integral equations (for example, see Ref. 26) for systems with different geometry of probes and surfaces under study as applied to both heterogeneous and homogeneous proc- esses. Inasmuch as the diffusion layer boundary repeats to a certain extent the surface relief, the SECM technique allows one to visualise the latter with submicron resolution.Scanning tunnelling microscopy (STM) in air may be considered as a version of SECM, because the surface under study always bears an ultrathin film of condensed moisture playing the role of a liquid electrolyte.53 The resolution that can be achieved depends largely on the curvature radius of the probe tip (even for tips of uncertain general geometry). Moreover, the size of the tip exerts the most critical effect on distortions of the images of surface fragments with small curvature radii. The corresponding size effects can be simulated numerically,54 and there is a prospect of using advanced models for different convolution procedures (e.g., for recovery of real images of the surface studied and for the determination of the tip geometry from the distortions of standard surface images).The development of the STM method provided special procedures for the pretreatment of well characterised tips, which are common for all probe techniques including SECM.55, 56 The resolution in the SECM technique is enhanced if a non- uniform surface contains areas of different nature where the rates of local processes differ. Various specialised versions of SECMare developed, for instance, those for the studies of enzymic proc- esses,57 combined technique (SECM ± probe microscopy 43) and related techniques for investigations on the boundaries of two immiscible liquids.A unique result achieved in SECM configuration is the measurement of responses of an electrochemical transformation of a single molecule.58 ± 60 This became possible owing to the accumulation of the signals from multiple successive reversible transformations of a reactant and its product in a limited solution volume between a deepened probe ultramicroelectrode (r&10 nm) and a planar surface (Fig. 2). It is the localisation of small solution portions in micro- and nanocells the volumes of which can be reduced down to 10 nl and realisation of systems with two ultramicroelectrodes arranged along a common axis at a small distance from one another that will determine further progress in this field.61 Certain basic problems of microelectrochemistry formulated only recently 24, 26, 48 still remain unsolved.They pertain to the analysis of the potential distribution for overlapped diffusion and diffuse layers (including the case of diffuse layers of two closely spaced electrodes). This is observed, for example, in the config- uration of electrochemical tunnelling microscopes (in situ STM). Under these conditions, certain anomalies are observed which have not yet found unambiguous explanations.62 Thus the catho- dic deposition of a metal phase can occur at electrode potentials288 1 2 e7 A B e7 Conducting planar electrode Figure 2. A scheme of a microcell for measuring electrochemical trans- formations of single molecules.15 (1) Ultramicroelectrode, (2) insulation; A and B are oxidised and reduced reactant forms, respectively. banned by formal thermodynamics.Hence, one cannot rule out the possibility that distortions in the surface equipotentiality can occur in the vicinity of the tunnelling gap. It is still more probable that the configurations used do not provide correct measurement of the potential drop between the working and reference elect- rodes. III. Electrical double layer as a nanostructure As noted above, an electrochemical interface demonstrates a sufficiently complicated space distribution of components, due to chemisorption and electrostatic factors.23, 63, 64 Being the most important EDL characteristic, the zero-charge potential is directly related to the physical properties of electrodes (for example, the electronic work function).63 Its value depends on both the electrode material nature and the crystallographic orientation of the surface.For a thin-film (especially, semicon- ducting) electrode, such a thickness can be reached that permits the work function to manifest size-dependent effects;65 corre- spondingly, variations of film thickness can cause changes in the electrode charge vs. potential dependence.7 The models of EDL structure deal with space distributions of the potential in the dense and diffuse parts. For electrodes with pronounced heterogeneous structures and compositions, the potential can vary not only along a normal to the surface but also in parallel to it, both in the dense and diffuse parts.Comprehensive phenomenological models are developed for size-dependent parameters of EDL on polycrystalline surfaces constructed from different lattice faces with different values of zero-charge potential.66 For small heterogeneities, EDL structure is well approximated by a model with a uniform diffuse layer. With an increase in the heterogeneity size, transition layers appear between neighbouring regions of diffuse layers of different struc- tures. Probe techniques make it possible to study the heterogene- ities of the potential distribution near the surface and describe quantitatively the interaction between the diffuse layers of the electrode and the probe.62 This may be considered as further development of the approaches dealing with interaction of crossed filament electrodes.63 In the general case, the EDL is a multicomponent system in which not only concentrations but also the states of species may differ from those in the bulk.For example, the lateral Coulomb repulsion of ions adsorbed with charge transfer 67 weakens so that their surface coverage can reach a complete layer. The phenom- enon of the formation of adatomic submonolayers stabilised by the interaction with the support is called underpotential deposition (UPD), because it provides deposition before the redox system involved reaches its equilibrium potential (for example, the deposition of adatoms M at potentials more positive than the equilibrium potential of anM/Mn+ system).The different rates of formation and dissolution of adatomic monolayers compared Surface coverage by CO molecules O A Petrii, G A Tsirlina with metal phases also point to the differences between these systems. Apparently, the properties of a layer start to approach those of a phase only after the deposition of the third or fourth monolayer. Like adatoms, the molecules of organic adsorbates form structurally non-uniform adsorption layers. In the case of strong lateral interactions, different domains appear on a supporting surface, their structures being dependent on the surface geometry and charge as well as on the presence of other adsorbates.Two- dimensional condensation 68 typical of adamantane and camphor derivatives and of certain amino acids is a spectacular manifes- tation of lateral interactions. For adsorption with partial charge transfer (or without it at all), the co-adsorption of anions and cations can be responsible for the formation of closely packed monolayers. For example, the adsorption of multicharged poly- oxometalate anions is accompanied by the formation of the so- called salt-like layers.69 ± 71 Many adsorbates were reliably shown to preferentially adsorb on defective surface sites and behave specifically with respect to surface areas of different coordination.72, 73 Figure 3 shows, as an example, the results on CO adsorption on high-index faces of a platinum single crystal.A size effect manifests itself in an increase in the surface coverage by CO molecules on areas with a high density of steps (i.e., with a decrease in the terrace length). For well characterised systems, the general thermodynamic approach was applied for the analysis of the contributions made by terraces and steps into reversible adsorption processes and, in particular, for the elucidation of the effects of surface crystallography on the observed zero total charge potential value.72 Moreover, we can assume that the electronic work function is different in different surface points and, apparently, it makes sense to invoke the concept of the local work function (and, correspondingly, the local zero free charge potential 74).An argument for such an assumption may be the difference in adsorbabilities of the surfaces of terraces and steps. 0.90 0.70 0.50 30 10 0 20 Step density /106 cm71 Figure 3. Surface coverage by adsorbed CO species determined from the coulometric data in 100 mM HClO4 as a function of atomic step density on Pt[n(111)6(111)] single-crystal faces.72 It should be noted that, along with equilibrium properties of adlayers, the characteristic sizes of areas of given configurations also strongly affect the dynamic characteristics of adsorption processes. For instance, STM data at high time resolution showed that during the formation of the second layer of silver adatoms on a gold electrode surface diffusion on terraces proceeds more quickly than that on sites near the steps.75 As a result, the adatoms are stabilised near the steps, which is beneficial for the smoothing of the relief in the early deposition stages (Fig.4). The dynamics of adatomic-layer fragments can be followed in STM experiments with high time resolution.24 Unusual nanosize structures with barrier properties are formed by alkanethiol molecules adsorbed on metal atomsSize effects in electrochemistry ab 123 Figure 4. Schematic representation of a mechanism of silver deposition on an Au(100) surface.75 (a) Initial stages, (b) formation of multilayer deposit; (1) gold atoms, (2) silver atoms in equilibrium states, (3) mobile adatoms (marked by arrows) and silver adatoms incorporated into the lattice.owing to chemical (virtually irreversible) interaction of terminal thiol groups with metals.76 ± 78 For example, Hg(SR)+H+. RSH+Hg7e7 By changing the alkyl chain length, one can obtain, with high precision, a sufficiently closely packed interlayer of molecules arranged in parallel (the area occupied by a thiol centre is *20 A2) with a thickness invariant over the surface. The capaci- tances of alkanethiol-modified electrodes do not exceed several microfarads per square centimetre and decrease for longer alkyl chains.79 This suggests that the solvent is practically completely forced out from the EDL dense part, and a uniform dielectric interlayer is formed on the interface. Alkyl chains are shown not to be arranged along a normal to the surface but to be bent at a substantial angle (up to 40 8).Peculiarities of the formation and properties of alkanethiol layers are sensitive to surface defects,77 which offers a unique possibility of diagnosing the latter, for example, by following different manifestations of the barrier properties of alkanethiol layers during charge-transfer processes. These systems can also be applied for studying the discrete charging of double layers formed on small particles that operate as nanoelectrodes.80 Considerable progress in studies of nanosize structures became evident with the possibility of visualisation of adsorbates in situ by probe techniques.17, 76, 81, 82 Special attention should be drawn to the potentialities of these techniques for studying adsorption-induced phenomena, such as the appearance of sur- face stresses and the surface reconstruction.83, 84 Thus the latter can be simulated on the atomic level according to the results for faces of single crystals of gold and platinum (Fig 5).However, the problem of interpretation of STM data still remains sufficiently complicated, because the electrical double layers of a probe electrode and a scanned surface tend to overlap and interact with one another.62 Studies of the EDL structure on an atomic level is important for quantitative description of the interaction of reactants with electrified interfaces, i.e., for the solution of problems of electro- chemical kinetics. These studies can also shed light on the micro- Step Figure 5.A scheme of island formation on a Pt(111) surface induced by potential cycling in a region that includes oxygen adsorption.84 289 scopic origin of discreteness-of-charge effects arising in the specific adsorption of ions, the effects exerted by metal hydro- philicity on the dense-layer capacitance, etc., which were consid- ered so far within the framework of phenomenological approaches. Modern approaches to building physical models of charged interfaces are surveyed in Ref. 23. IV. Size effects in electrocatalysis Size-dependent electrochemical phenomena occur in chemisorp- tive and electrocatalytic processes that take place on highly dispersed, small (for instance, ultrathin) and surface-heterogene- ous electrodes. In such processes, the heterogeneities of character- istic sizes commensurable with either the thicknesses or lateral periods of adsorption layers (nanosize range) play a significant role.Many active electrocatalysts represent the so-called nano- structurised materials that exhibit periodic inhomogeinities not only on the surface but also in the bulk.22, 85 Nanostructures are at equilibrium only in rare cases, most commonly they are metasta- ble, but long-lived.Nanosize materials were introduced into the practice of electrochemical studies long before the methods of their struc- tural, morphological and electronic characterisation became available. First of all, these were the electrodes with well devel- oped surfaces (the ratio of the real surface area Sreal to the geometric area S being equal to 1000 and higher) of platinum- group metals consisting of nanocrystals.4, 6, 7 For them, the experimentally determined rates of electrocatalytic processes, the surface coverages by adsorbates and the polarisation capacitances were referred not to the individual characteristics of small particles but to the values of Sreal.86, 87 It is the dependence of the specific (calculated per Sreal) activity on the fabrication method that was the subject of numerous studies aimed at intensification of practically important processes applied mainly in fuel cells.The concepts of the related structural (size) effects in hetero- geneous catalysis 88 as applied to dispersed electrodes have also generated certain simplified notions based on the analysis of `electronic' and `structural' factors.For example, the latter factor was considered in the mitohedral approach,89 which was devel- oped further in the 1980s ± 1990s. During this period, well char- acterised electrode materials of similar compositions but substantially differing in surface coordination became available, viz., faces of single crystals and polycrystalline materials with pronounced preferential crystallographic orientations.10 Numer- ous attempts were undertaken (for example, see Refs 4, 12) to describe the activities of small metal particles in terms of an additive approach to the contributions of low-index faces. Usu- ally, the ratio of faces on the surface was calculated for ideal equilibrium crystals (in the nanosize range, it strongly depends on the grain size).Although a number of activity ± surface coordination rela- tionships have been revealed, the attempts to substantiate the decisive role of crystallographic orientation in electrocatalysis by dispersed materials have failed for virtually all actively studied processes of oxidation of organic compounds and hydrogen, as well as of oxygen reduction. Undeniably, the main reason for the deviations between experimental and model dependences lies in the non-equilibrium nature of small particles of real catalysts. It is also significant that the particle sizes used in the calculations were not necessarily determined by direct methods. The technique of high-resolution electron microscopy 90 ± 92 made it possible to refine the data on the real geometry of nanosize particles for a number of materials.However, the potentialities of this technique for the analysis of multilayer deposits and dispersed electrodes are limited. Reliable results are obtained mainly for smooth carbon supports with low surface coverages by catalysts. It should be noted that the interactions of small particles with the support substantially complicate the behaviour of these sys- tems;6, 7 no adequate models exist for the description of these290 interactions. The activity ± particle size relationships for different electrocatalytic processes are analysed in several papers (cf. Refs 92 ± 98). However, these relationships pertain to relatively narrow intervals of particle sizes.Moreover, the technological methods of fabricating catalysts with different particle sizes do not provide permanence of several important structural character- istics. In different studies, the specific activity was shown either to increase or to decrease with an increase in the particle size not only in the nanometre range but also in submicron range for which one can hardy expect any manifestations of equilibrium `structural' (and, all the more, `electronic') size effects. To date, different dispersed materials based on platinum and palladium have been examined by means of the STM technique with a nanometre resolution.99 ± 101 A comparative analysis of the data on the real surface area, porosity 102 and characteristic sizes of surface fragments (independently determined by different techniques) allows highly reliable conclusions to be drawn on the real structures of nanosize electrocatalysts: the latter include globules constituted by quasispherical particles.Sufficiently wide size distributions of particles and globules are revealed for electro- deposited dispersed catalysts. This result requires that a great number of property ± size relationships be reconsidered for those cases where the sizes were estimated from Sreal: the particle size thus determined certainly exceeds that corresponding to the distribution maximum. At a high degree of screening (or coales- cence) of particles in the globules, the aforementioned differences can reach an order of magnitude, as in the case of certain materials based on electrodeposited palladium.101 At present, the controversial views on the structure of differ- ent boundary regions in nanosize electrocatalysts, first of all, of intergrain boundaries in globules of metal crystals pose a problem which is extremely difficult to solve experimentally.Probably, as in the case of other nanostructurised materials,22, 85 it is these highly defective regions that play a significant role in adsorption and electrocatalytic phenomena. For a number of cases, a phenomenological analysis 103, 104 based on an assumption of high surface defectiveness of these electrocatalysts makes it possible to find a self-consistent set of experimental results.However, it fails to rationalise the nature of defective surface regions. Yet another aspect of size-dependent electrocatalytic effects that manifest themselves on the scale of molecular sizes is associated with the structure of chemisorption layers. The mech- anisms of the vast majority of electrocatalytic processes involve strongly bound adsorbates for which co-adsorption and compet- itive adsorption play important roles.9 A rather universal type of size effect occurs within the frame- work of the mechanism known as bifunctional catalysis.9 This mechanism is realised on bi- and milticomponent surfaces where the process is controlled by the chemical reaction of two different adsorbates. Electrocatalytic hydrogenation of organic substances involving hydrogen atoms and their oxidation involving adsorbed hydroxyl radicals may serve as examples of these processes.The activity of a catalyst proves to be highest where its surface contains simultaneously two components which selectively chemisorb the species involved in the slow stage of the reaction. In this case, both the mutual arrangement of components and the ratio of their surface concentrations are of importance. Particularly, there exists an optimum (depending on the reactant structures) size of surface areas with one or another composition which provides the greatest number of reaction sites. Such an area is always commen- surable with the size of the adsorbate species. Homogeneous alloys which exhibit no pronounced segregation and adatomic submonolayers are typical bifunctional catalysts.Figure 6 shows the rate of electrocatalytic oxidation of ethylene glycol on plati- num covered by different amounts of tin adatoms as a function of the electrode potential.105 Tin adatoms are the adsorption centres for active oxygen species and exhibit the maximum activity at medium surface coverages. log i (mA cm72) 10 71 0.3 0.2 0.1 Figure 6. Steady-state polarisation curves of ethylene glycol oxidation (0.5 M, in the presence of 1 M NaOH) on platinised platinum modified by tin adatoms.105 Surface coverage by adatoms: (1) 0, (2) 0.76, (3) 0.48; Er is the potential referred to the reversible hydrogen electrode. If a process is controlled by the recombination of atoms, then, conversely, optimisation of the electrocatalyst requires an increase in the number of neighbouring active adsorption centres of the same nature. For this mechanism, the dependence of the process rate on the geometry and sizes of reaction sites can be well simulated by hydrogen evolution on platinum-group metals modified by different adatoms.Depending on their sizes, adatoms can occupy one, two or three neighbouring positions otherwise available for adsorbed hydrogen. In real systems, we cannot always provide a uniform distribution of a modifying additive over the surface; therefore, the observed effects of recombination inhibition appear to be smaller than is expected. This situation can be used for solving the inverse problem, viz., studying the geometry of island structures.Yet another typical mechanism of electrocatalysis (first of all, for processes involving organic substances on platinum-group metals, Fig. 7) occurs for two and more parallel processes of log i (mA cm72) 321 0.7 0.5 0.3 Figure 7. Steady-state polarisation curves of formic acid oxidation (1 M, in the presence of 0.5 M H2SO4) on platinised platinum modified by thallium adatoms.106 Concentration of added Tl+/mM: (1) 0.01 (2) 0.033, (3) 0.1, (4) 0.33, (5) 0.48, (6) 0.96, (7) 1.9, (8) 4.8, (9) 0. In the vicinity of the maximum, current fluctuations occur with the amplitude shown by the dashed area. O A Petrii, G A Tsirlina 123 0.5 0.4 Er /V 123456789 0.9 Er /VSize effects in electrochemistry reactant chemisorption, one of which produces non-reactive strongly bound adsorbates that partly block the surface.For such self-inhibited catalytic processes, the selective suppression of the inhibitor adsorption is an important size-dependent factor. In systems with organic reactants, the blocking adsorbate occupies usually several neighbouring surface sites; hence, its formation is hindered by foreign adatoms that modify the surface (the so-called `third-body effect' 9). This is exemplified in Fig. 7 by the data on the acceleration of formic acid oxidation on platinum modified by thallium adatoms each blocking three neighbouring adsorption centres, according to estimates.The `third-body' effect is man- ifested only in the ascending branch of the curve in the potential region where the process is inhibited by strongly adsorbed CO-like species. Blocking adsorbates of this and nearby components are formed in both the destructive adsorption of various organic molecules and immediately in the adsorption of CO molecules. Thus the problem of designing systems with a pronounced `third- body effect' is of prime significance for the solution of an urgent applied problem, viz., preparation of CO-tolerant catalysts. In principle, macrokinetic effects in systems with porous catalysts can be also assigned to size effects. These are discussed in detail by Chizmadzhev et al.107 V. Size effects in the elementary act of charge transfer At present, subnanometre-scale size effects are actively studied on the molecular level within the framework of theoretical descrip- tions of electrochemical interfaces and attendant processes of the transfer of charged species (for example, see Refs 108 ± 112).In such a consideration, both the sizes of molecular subsystems (reactant and product species involved in the charge transfer, solvent molecules adsorbed on electrodes) and the characteristic scales of electron density distribution are of substantial impor- tance. Charge transfer processes cannot be studied without invoking quantum-mechanical approaches, which in turn requires a changeover from spatial scales to spatial ± temporal scales. The probability of electron transfer in an electrode ± reactant system is determined by a number of factors that depend on the characteristic sizes of reaction layers.108 The most `large-scale' and complex factor is the distance of the closest approach of a reactant a Fe HS O b P*/P+ (70.96 V) Au hn Fc/Fc+ (+0.61 V) P/P+ (+0.92 V) Figure 8.A scheme of a model system with two fixed redox centres, viz., those based on ferrocene (Fc/Fc+) and porphyrin (P/P+) (a) and the sequence of charge transfer steps to different redox centres (b).116 A self-assembled layer of substituted thiols is immobilised on the surface of an illuminated gold electrode in a solution of methyl viologen (redox pair MV+./MV2+); the equilibrium potentials of the corresponding systems are shown in parentheses.291 (3) K=const00 exp ¡ z , b to the surface zlim , which is determined by steric restrictions and by chemisorptive and electrostatic interactions. For ionic and molecular reactants, such distances do not usually exceed fractions of several tenths a nanometre. If zlim>1 nm, we speak of the `long-range' electron transfer. The limiting value zlim is arbitrary and is determined by a substantial attenuation of the electron overlap, which exponentially depends on the electrode ± reactant distance. By introducing the trans- mission coefficient K, one may represent this dependence as follows:108 where the parameter b depends on the degree of electron delocal- isation and is of an order of magnitude of a hundredth part of a nanometre.Usually, the delocalisation of reactant electrons plays the decisive role, in particular, the greatest b values (< 0.1 nm) are known for transition-metal complexes in which the charge trans- fer proceeds to orbitals of a mixed nature. In principle, all reactant species located at any distance from the surface contribute to the observed rate; however, the contri- butions of species located in a thin near-surface layer not exceed- ing zlim prevail. In place of the parameter zlim, another parameter of the same order of magnitude dz known as the reaction layer volume is used, for example, in the expression for the current density (rate of the process) (4) i=const000 (zlim ) dz exp ¡ Ea , RT where Ea is the activation energy,R is the universal gas constant, T is the temperature.In modelling the electron transfer, dz is determined by integrating the expression (3) within the limits from zlim to the infinity. The `long-range' transfer can be observed in systems with barrier layers on the electrode surfaces 79, 113, 114 and is studied most comprehensively for thiol systems. In particular, unambig- uous dependences of the transfer rate on the barrier layer thick- ness are established. Thiol layers with redox-active terminal substituents that provide rigid localisation of the reactant are widely used.115 The introduction of two redox centres, with a variable distance between them, into a thiol molecule is an interesting extension of this approach (Fig.8). N NHO N HN MV+./MV2+ (70.63 V)292 `Long-range' charge transfer is also observed for reactant species of considerable intrinsic sizes and with a pronounced degree of localisation of the transfer site (metalloproteins,112 polynuclear complexes 71). It should be noted that the character- istic length of the `electronic tail' (and, correspondingly, electron overlap) depends on the electrode charge; however, the theoretical analysis of these dependences is possible so far only in terms of simplified models.109 The quantity zlim affects not only the ovelap, which determines the prexponential factor in the equation for the rate constant of electrode reactions, but also different components of Ea: the work terms of reactants and products (if we deal with charged species, the Coulomb interactions weaken with the distance from the interface as the potential changes) and the solvent reorganisation energy.A detailed theoretical consideration of this problem is given in Refs 117, 118. Pronounced size effects can be observed in photoelectrochem- ical processes involving semiconducting particles separated from the metal surface by barrier layers. The size effects are caused by a shift of absorption edge when the diameter of particles ranges from 10 to 20 nm.119, 120 Substantial acceleration of photoelec- trochemical processes occurs on passing from massive to nanosize semiconductors. This phenomenon has already found a practical application, viz., in photoelectrolysis of water.121 Similar `intrin- sic' size effects of metal particles should be weaker and manifest themselves for particles with sizes below 5 ± 7 nm.However, in principle, there exist no methodical obstacles from discovering them in the nearest future. Prediction of the kinetics of electron transfer by taking into account all size-dependent factors becomes possible only when adequate ion ± molecular models of reaction layers are built. For a number of systems, this problem can successfully be solved by invoking quantum-chemical methods 122 based on the quantum- mechanical theory of the charge-transfer elementary act.123 Along with the classical effects of the cation size which manifest themselves in the reduction of anions on a negatively charged surface,124 we must especially mention the inhibition of electron transfer with increase in the ligand size for a number of transition- metal complexes.125 For non-symmetrical reactants, this depend- ence gives rise to various orientation effects in reaction layers.126 The independent (additional) possibilities of experimental simulation of the size effects of a quantum origin are provided by carrying out electron transfer processes in the STM config- uration at fixed tunnelling voltages, which govern the location of energy levels of transferred electrons in the metal.110, 127 Interest- ing similarities between STM and electrochemical responses can be traced for polyatomic reactants with dense outer barrier layers and for gold clusters, particularly those stabilised by alkane- thiols.128 In both cases, the current responses reveal the appear- ance of Coulomb blocking (up to six steps separated by potential intervals of*0.3 V).The latter system, which allows one to trace the dependence of electrochemical responses on the cluster size, is considered in the literature as `transition': for a small number of atoms in a cluster, the differences between the electrochemical charging of the gold cluster (`nanoelectrode') and an electrode process involving this cluster as a reactant in solution actually vanish.129 VI. Size effects in processes of new phase formation in electrochemical systems A complex combination of size effects of different origins is realised in electrochemical processes with the formation of new phases, for instance, in electrocrystallisation processes.25 In the nucleation on electrode surfaces (as in the formation of a new phase in a homogeneous medium), the key role is played by supersaturation determined by the electrode potential deviation from the equilibrium value for a given redox system (overpotential Z).A redox-system component formed as a result of a transfer of n electrons takes part in the nucleation process. A thermodynamic O A Petrii, G A Tsirlina description of this phenomenon relies on the analysis of the probabilities of formation and dissolution of new-phase nuclei 5 and gives the following expression for the critical nucleus radius rc (the minimum size, for which the probability of dissolution of a nucleus formed does not exceed the probability of its further growth): rc= gA nFZ or rc=nF 2suZ , where g is the surface tension of a two-dimensional nucleus, A is the area occupied by a mole of the substance in a monoatomic layer, F is the Faraday number, s is the surface tension of a three- dimensional nucleus, and u is the molar volume.When estimating the size of an extremely small critical nucleus, one cannot use the value s characteristic of macroscopic fragments of the same solid, because, in principle, this parameter is size-dependent. For the analysis to be correct, the EDL structure and involve- ment of chemisorbed solvent molecules and other solution com- ponents, as well as adatoms should also be taken into account.All these factors affect the surface energy of a nucleus and, hence, its critical size. Along with thermodynamic models, non-thermody- namic models of nucleation on the electrode surface have also been developed.25 The latter are based on either a statistical dynamic approach or on quantum-chemical studies of different- size clusters. The mentioned factors also play an important role in the description of steady-state and non-steady-state kinetics of nucle- ation and growth of new-phase particles on electrodes: in many cases, these factors can change the rate and the limiting stage of a process. Largely pertaining to potentiostatic models, the expres- sions for the non-steady-state current of nucleation ± growth of a unit nucleus and nuclei ensembles are described in detail in numerous reviews (for example, see Refs 5, 14, 130).The time- dependent nucleus size plays a key role in deriving such expres- sions. The other important parameters include the number of active nucleation centres, the nucleation rate constant and a parameter characterising the kinetics of the limiting stage (the diffusion coefficient or the rate constant for electron transfer). Moreover, even the simplest approximations (for example, for the formation of spherical or semispherical nuclei on a uniform plane) allow the criteria for diagnosing the nucleation type: two-dimen- sional or three-dimensional growth, diffusion or kinetic control, the ratios of nucleation and growth rates, to be found.The nucleation type is elucidated from experimental current transients (for instance, power dependences at small times) by analysing the maximum current corresponding to the hindrance of growth due to the coalescence of neighbouring nuclei. From electrochemical data, by using the corresponding algo- rithms, one can indirectly characterise the morphology of a system at any moment of deposition (even for a crystal size commensu- rable with the atomic size). However, the differences between the properties of real and model systems cause substantial deviations of the calculated values from the experimental results. For example, the calculated concentrations of active centres are often several orders of magnitude lower than those found from electron microscopy data (direct visualisation of grown crystals).Model systems for which the active centres represent certain surface areas (for example, dislocations) and the growth rate of a new phase is limited by a process localised immediately on the surface, viz., the surface diffusion, are of special interest 25, 130 In this case, the key size-dependent factor is the distance between the growth steps which incorporate the adatoms slowly diffusing over the surface. However, it is difficult to fabricate such systems, especially when using real polycrystalline materials the active centres (and point defects) of which may be represented by any surface areas with small radii of curvature. The exchange of adatoms on the growth steps can be observed by means of STM.24 The methods of independent determination of the number and sizes of nuclei at different moments are of prime interest, especiallySize effects in electrochemistry those that do not require the electrode to be withdrawn from solution.These requirements are fulfilled by the STM technique realised with an atomic resolution in situ in electrolyte solu- tions.16, 17, 20, 73, 131 ± 133 However, at present, local changes in the deposition conditions that occur near the probe, especially, for nucleation under diffusion control cannot be reliably eliminated. It was also reported that nanosize crystals are unstable in the electric field around the probe if adhesion is low (for example, this concerns the important group of model processes of metal deposition on highly oriented pyrographite). An efficient approach to the determination of the number and sizes of nuclei consists in substitution of the background solution for the deposition solution without opening the circuit.73 The probe is brought to the surface after all the deposition stages studied are completed.This, however, dramatically reduces the time resolu- tion.The techniques of atomic-force microscopy in a noncontact mode progressively gain importance.32, 134 The advantage of such methods is the possibility of studying objects of low conductiv- ities, for example, in electroplating on wide-band semiconduc- tors,31, 32 in structural transformations of polymer deposits during dedoping 135 and in the course of formation of nanosize barrier oxide layers.136 Combined photo-assisted probe techniques are being actively developed.21, 33 It should be noted that quantitative probe-microscopic data on nanosize particles of electrodeposits (i.e., where the radius of curvature of a probe is commensurable with that of the particles) can be obtained only when special convolution procedures are used, the algorithms of which are being extensively developed.54, 137, 138 A quantitative description of electrochemical responses in the formation of a new phase is complicated by the appearance of the so-called screening zones caused by the overlap of either the neighbouring nuclei or the corresponding diffusion layers.Certain effects of this sort can be simulated, which however always increases the number of fitting parameters, thus reducing the diagnostic possibilities of the model. Experimentally, the overlap can be avoided if the distances between the active centres are of an order of a micrometre: in this case a single nucleus can be grown on an ultramicroelectrode.139, 140 In fact, such experiments allow one to observe the elementary events that constitute the nucleation process and then to use the obtained information for the analysis of the growth of ensembles. Screening under diffusion control can also be used for obtain- ing the so-called fractal deposits (for example, see Ref. 141). Such deposits are formed in thin-layer cells, which includes the boun- daries of two immiscible liquids and the solution ± air interfaces.The main contribution to the growth of a new phase is made by the secondary nucleation processes which can be intensified by pulsed deposition modes (Fig. 9). Simulation of secondary nucle- ation which plays a significant role in many systems is a very Figure 9. SEM image of platinum deposit on a carbon surface electro- plated from a solution of 40 mM H2PtCl6 by square-wave pulses.142 The horizontal measuring interval is 1 mm. 293 complicated task, which defies analytical solution unless being considerably simplified. In this case, numerical techniques should be invoked.143 The technological possibilities of electrocrystallisation are extremely high and today are only partly used.The systems in which the crystals are formed on the surface of preliminarily developed adatomic submonolayers allow variation of the geom- etry of deposits on a nanolevel. In these layers, various two- dimensional phase transitions occur, and three-dimensional growth of nuclei is initiated by different mechanisms on domains of different structures.144 Tuning the fine structures by using the accompanying adsorp- tion phenomena is also possible in the dissolution of metals: the processes of this group can be generally considered as electro- crystallisation with the formation and growth of cavity-type nuclei. For the studies on the nanoscale level, a more precise control over the processes of new phase growth in the presence of organic additives traditionally used in industrial electroplating as the brighteners is possible by varying the ratio of the rates of nano- crystal growth in different directions.145 Deposition of unusually long-lived metastable highly defective solids deserves mention as a promising approach.146 ± 148 The deposition becomes possible due to the formation of crystals on electrodes under substantially non- equilibrium conditions.A related direction is associated with the controlled variation of the stoichiometry of multicomponent phases that are deposited.149 ± 151 In connection with the development of nanotechnology (see the next Section), it seems reasonable to outline the range of materials which can be deposited or formed by electrochemical mechanisms.On inert supports, one can carry out cathodic deposition of metals, alloys and certain semiconducting materials based on sulfur, selenium and tellurium and of solid complex compounds (for example, Prussian Blue and its analogues). The latter are of interest for electrochromic and sensor devices. Using anodic deposition on inert supports, one can obtain oxides and hydroxides,152 different polymers 153, 154 and salt phases with variable electrophysical properties.155 Anodising of metals in specially chosen media results in the formation of nanosize layers of passivating oxides many of which exhibit high photochemical activities.156 The formation of intercalates is an extremely important group of electrocrystallisation processes.Crystallisation of a new phase proceeds in the electrode bulk and, as a rule, is controlled by solid- state diffusion. Strains caused by intercalation can be observed by probe techniques.157 Rapid intercalation processes are typical of a number of oxohydroxide phases and are accompanied by changes in the oxygen stoichiometry, which often proceed simultaneously with the proton transfer. High intercalation rates can be reached, for example, in materials based on oxides of ruthenium 158 and tin 159 for lattices of certain oxygen deficiencies. In such materials, as in metal hydrides, intercalation seems to be accompanied by the formation of nanosize regions with different compositions. The structure of intergrain boundaries is of principal significance for the reversibility and rate of intercalation, in particular for pre- vention of cracking by changing the local composition of grains.22 VII.Electrochemical nanotechnologies Electrochemical fabrication of small-size articles occupies an important place in the system of chemical materials science and design of new materials having special properties. The mecha- nisms of formation or removal of individual small portions of solids play a decisive role in the field of electrochemical nano- technology. These mechanisms are not quite clear which prevents the use of the most important advantage of electrochemical methods of fabricating nanoobjects, viz., their controllability and the possibility of on-line quantitative monitoring.The con- trollability is provided by carrying out a process at a fixed potential (similar modes of deviation from equilibrium are practi-294 cally unrealisable in beam and in CVD and PVD technologies). In the general case, monitoring is provided by measuring current and charge. The sensitivity to the rate of product accumulation is determined by the aforementioned limits of measuring these parameters. Additional possibilities of monitoring will arise with the use of a quartz-crystal microbalance technique, which can easily be realised in electrochemical systems, but have not yet found any practical application. The existing nanotechnologies can be classified as local 31 and non-local.19, 160 Most often, the latter represent different versions of lithographic techniques, i.e., multistage combined technologies using photoinduced processes.Non-local nanotechnologies based on different self-organisa- tion processes in electrochemical systems represent an alternative to lithography. A typical example is the anodic electrosynthesis of alumina in which the ordered cylindrical pores with diameters of several tens nanometres are characterised by an extremely narrow size distribution.145, 161 Being of interest for catalysis, this nano- tube material 162 is widely used in combined technologies as a nanosize matrix. Metals 163, 164 and carbon materials 165 can be deposited into the pores of such an alumina matrix, and then alumina is removed at room temperature by dissolving it in alkali or HF.Similar processes can be realised for certain polymeric matrices (Fig. 10). Figure 10. SEM image of an array of gold nanowires fabricated by electroplating into a polycarbonate matrix (after removal of the matrix).166 Electrodeposition of a metal into a porous matrix (templating) and co-deposition of a metal and a matrix (quasitemplating) form the basis of technologies for production of metallised conducting polymers.167, 168 Under certain conditions, the phenomenon of the formation of ordered new-phase nuclei in the absence of a matrix can also be used in nanotechnologies, for instance, in the preparation of catalytically active materials with particles of fixed sizes and also in combined low-temperature technologies for the preparation of semiconducting particles.169 ± 172 Electrodeposition of lamellar bimetallic structures which is carried out either by different step-wise modes from a complex electrolyte or by consecutive deposition from different solutions constitute a separate group of processes.Continuous alternating 2-nm layers can be reproducibly obtained in systems characterised by two-dimensional growth. Layer-by-layer deposition finds wide application in the technologies of preparation of magnetic bimet- allic materials.173 ± 176 Layer-by-layer deposition of single layers of semiconducting oxide phases 177 and of special polymeric compo- sitions 178 is promising for the preparation of semiconducting superlattices. Cathodic deposition of Cu/Cu2O structures in the presence of organic complexones in a weakly alkaline medium, which is carried out in a simplest galvanostatic mode, is an example of a nanoelectrochemical technology for fabricating lamellar compo- sitions.179 ± 181 Deposition of alternating layers of nanosize thick- nesses is accomplished owing to a spontaneous initiation of potential oscillations caused by periodic changes in the pH and copper ion concentration in the near-electrode layer.In future, this method can be extended to other oscillating electrode proc- esses. The methods of dispersing metals by anodic dissolution are also worth being mentioned. Bach and coworkers 182 ± 185 appear to be the first to apply this approach for the preparation of platinum colloids in aqueous media in the electric arc.In less severe conditions, a similar method is possible for palladium in non-aqueous solvents.186 However, the problem of stabilisation of colloids is demanding, which requires introduction of organic additives in most cases. We also assign the methods of immobilisation of molecular ensembles, clusters and metal particles on electrode surfaces to electrochemical nanotechnologies. The particles are bound with the surface by adsorption induced by the potential (charge) of an electrode. In a number of cases, this process occurs under electro- phoretic supply and often involves surfaces preliminarily func- tionalised, for example, by alkanethiols. The best developed methods are those of colloid immobilisation 54, 187 ± 189 which provide virtually monodispersed coatings.When using weak stabilisers, for example, citrates, it is possible to simultaneously provide a narrow size distribution of particles and their high adsorbabilities (Fig. 11). I /mA 0.20 70.2 1.0 0.5 0 Figure 11. Size distribution of particles of immobilised platinum citrate colloid on polycrystalline gold (a) and potentiodynamic curve of the immobilised colloid (b). The distribution is plotted from STM data (1); for comparison, a wider distribution (2) corresponding to platinum electro- deposition is shown.54 Various local nanotechnologies can be realised in the config- uration of probe microscopes 18 based on diverse processes. Thus one can locally deposit polymers and metals either by supplying reagents from the solution bulk 190 or by generating them on the probe tip.191 It is also possible to modify a surface by transferring metal to it from the probe during short mechanical contact.192 In this process, no electrochemical reaction occurs, and adsorption phenomena on the charged surface appear to play a significant role in the stabilisation of clusters formed.Figure 12 shows a scheme of this procedure. The methods of local etching near the tip, which are realised in STM configuration (both in situ 193 and ex situ 194) where the role of an electrolyte is played by the moisture concentrated from air, are also gaining acceptance. Fraction of particles of diameter da 0.6 0.4 0.20 9 6 3 b O A Petrii, G A Tsirlina 12 107 d /cm Er /VSize effects in electrochemistry STM probe Cu2+ Gold surface Figure 12.A procedure for local modification of a gold surface by copper clusters in an STM configuration.20 The adsorption and related combined nanotechnologies based on self-organisation in two-dimensional layers and spontaneous chemical deposition employ efficiently electrochemical methods to govern the process.195 Local modification of a surface in a scanning mode with the formation of sufficiently stable cluster ensembles is sometimes called electrochemical nanolithography.192 ± 193 The response time of probe microscopes leaves only few chances for nanolithography in fabricating items on the nanosize scale.However, with the progress in nanoelectronics, the scale of these electrochemical technologies will clearly increase. We can offer quite a number of examples where island, lamellar and other nanostructures formed by applying electrochemical methods are successfully used as quantum wells, superlattices, tunnelling diodes, etc. Apparently, one of the main directions is associated with the achievement of quantum-size and resonance effects on the tips of STM probes modified electrochemically or by adsorption.196, 197 In particular, this involves the quest for systems with a negative differential resistance. Figure 13 shows an example of such a response for a local quasi-one-dimensional nanostructure. Probe Pt ± Ir 420 72 74 71.5 71.0 70.5 Tunnelling voltage /V Figure 13.Tunnelling current vs. voltage characteristic demonstrating the negative differential resistance effect using a probe modified by heteropoly molybdate.197 VIII. Conclusion The material reported in this review does not exhaust all the approaches to the research and application of size effects in electrochemistry. At present, prospects for the development of nanoelectrochemistry appear to be somewhat remarkable. This is associated with the fact that, in principle, possibilities of govern- Tunnelling current /nA Cu2+ 1.5 1.0 0.5 0 295 ing the operations on an atomic level have already been demon- strated. However, such achievements are realisable only at the lower limits of the possibilities of the available equipment.An on-line observation of elementary steps of electrochemical charge transfer, even if for relatively slow reactions, seems to be most impressive.24 However, this will require a substantial decrease in the operation time of probe techniques (to date, scanning of 50 images per second with an atomic resolution is apparently a record, however, not as yet for liquid phases). Conceptually, electrochemistry arrives at quite a novel stage of development when we must reconsider many common concepts and relationships. This reconsideration concerns now, particu- larly, the modification of the traditional Nernst equation of the equilibrium potential when applied to nanosize materials and the relationships of the classical theory of electrocrystallisation.Which of the predicted issues will remain in the status of science fiction and which will enter common research and technological practice and when this will occur, depends on many factors. However, the fast progress of nanochemistry in the recent years lends optimism. Great expectations are associated with the integrating trends of different branches of physics and chemistry. Carrying out their experiments, modern scientists actually deal not only with ele- mentary events that accompany transformations of single atoms and molecules, but also with transition states.198 In the near future it may become possible to transform the experimental basis of these studies so as to adapt it to electrochemical systems. This work was supported by the Ministry of Industry, Science and Technologies of the Russian Federation and the Foundation `Integration' (United Educational and Research Centre A0114).We are grateful to the Corresponding member of the RAS Professor Yu M Polukarov for discussing the topical problems of nanochemistry. References 1. J Goodisman Electrochemistry: Theoretical Foundation (Chichester: Wiley, 1987) 2. W Schmickler Interfacial Electrochemistry (Oxford: Oxford University Press, 1995) 3. B B Damaskin, O A Petrii Vvedenie v Elektrokhimicheskuyu Kinetiku (Introduction to Electrochemical Kinetics) (Moscow: Vysshaya Shkola, 1983) 4. K Kinoshita, P Stonehart, in Modern Aspects of Electrochemistry Vol. 12 (Eds J O'M Bockris, B E Conway) (New York: Plenum, 1979) 5.A I Danilov, Yu M Polukarov Usp. Khim. 56 1082 (1987) [Russ. Chem. Rev. 56 619 (1987)] 6. V S Bagotsky, A M Skundin Electrochim. Acta 29 757 (1984) 7. V S Bagotsky, A M Skundin Electrochim. Acta 29 951 (1984) 8. G N Mansurov, O A Petrii, V P Nedoshivin, A S Bluvshtein, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1988) Vol. 27, p. 62 9. O A Petrii, G A Tsirlina, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1988) Vol. 27, p. 3 10. O A Petrii, G A Tsirlina, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd.VINITI, 1991) Vol. 34, p. 3 11. S Makerjee J. Appl. Electrochem. 20 537 (1990) 12. R R Adzis, in Advances in Electrochemistry and Electrochemical Engineering Vol. 13 (Ed. H Gerischer) (New York: Wiley- Interscience, 1984) p. 159 13. A Hamelin, in Modern Aspects of Electrochemistry Vol. 16 (Eds J O'M Bockris, B E Conway) (New York: Plenum, 1985) p. 1 14. A I Danilov Usp. Khim. 64 818 (1995) [Russ. Chem. Rev. 64 767 (1995)] 15. A J Bard, D E Cliffel, C Demaille, F F Fan,M Tsionsky Ann. Chim. 87 15 (1997) 16. E T Yu Chem. Rev. 97 1017 (1997)296 17. A A Gewirth, B K Niece Chem. Rev. 97 1129 (1997) 18. R M Nyffenegger, R M Penner Chem. Rev. 97 1195 (1997) 19. J W Schultze, V Tsakova Electrochim.Acta 44 3605 (1999) 20. D M Kolb, M A Schneeweiss Interface (1) 26 (1999) 21. L F Garfias-Mesias, W H Smyrl Electrochim. Acta 44 3651 (1999) 22. H Gleiter Acta Mater. 48 1 (2000) 23. R Guidelli,W Schmickler Electrochim. Acta 45 2317 (2000) 24. D M Kolb Electrochim. Acta 45 2387 (2000) 25. E Budevski,G Staikov,W J Lorenz Electrochim. Acta 45 2559 (2000) 26. M V Mirkin, B R Horrocks Anal. Chim. Acta 406 119 (2000) 27. A Kaifer, M Gomez-Kaifer Supramolecular Electrochemistry (Weinheim: Wiley-VCH, 1999) 28. W J Lorenz,W Plieth (Eds) Electrochemical Nanotechnology. In situ Local Probe Techniques at Electrochemical Interfaces (Weinheim: Wiley-VCH, 1998) 29. R Schuster, V Kirchner, P Allongue, G Ertl Science 289 98 (2000) 30.K Krischer, in Modern Aspects of Electrochemistry Vol. 32 (Eds B E Conway, J O'M Bockris, R E White) (New York: Kluwer Academic Plenum, 1999) p. 1 31. M Cerisier, C Van Haesendonck, J-P Celis J. Electrochem. Soc. 146 1829 (1999) 32. J V Zoval, R M Stiger, P R Biernacki, R M Penner J. Phys. Chem. 100 837 (1996) 33. T Miki, H Yanagi Langmuir 14 3405 (1998) 34. J Li, E Wang Electroanalysis 8 107 (1996) 35. G Lehmpfuhl, Y Uchida,M S Zei, D M Kolb, in Imaging of Surfaces and Interfaces. Frontiers in Electrochemistry Vol. 5 (Eds J Lipkowski, P N Ross) (New York: Wiley-VCH, 1999) p. 57 36. M A Lovell, M J Walters, D Roy Phys. Chem. Chem. Phys. 1 1985 (1999) 37. P Christensen, A Hamnett Electrochim. Acta 45 2443 (2000) 38. D I Kochubei EXAFS-Spektroskopiya Katalizatorov (EXAFS Spectroscopy of Catalysts) (Novosibirsk: Nauka, 1992) 39.M Fleischmann, B W Mao J. Electroanal. Chem. 229 125 (1987) 40. Q Fulian, A C Fisher, G Denuault J. Phys. Chem. B 103 4387 (1999) 41. R D Martin, P R Unwin J. Electroanal. Chem. 439 123 (1997) 42. L Koene, M Sluyters-Rehbach, J H Sluyters J. Electroanal. Chem. 402 57 (1996) 43. C E Jones, J V Macpherson, Z H Barber, R E Somekh, P R Unwin Electrochem. Commun. 1 55 (1999) 44. P B Mills, W J Aixill, F Prieto, J A Alden, R G Compton, M Rueda J. Phys. Chem. B 102 6573 (1998) 45. A N Frumkin, L N Nekrasov Dokl. Akad. Nauk SSSR 126 115 (1959) a 46. M Koudelka-Hep, P D van der Wal Electrochim. Acta 45 2437 (2000) 47. M J SchoÈ ning, P Schroth, S SchuÈ tz Electroanalysis 12 645 (2000) 48.C Amatore C.R. Acad. Sci. 323 757 (1996) 49. M Fleischmann, B S Pons, D R Rolison, P P Schmidt Ultramicroelectrodes (Morganton, NC: Datatech. Systems, 1987) 50. J K Campbell, L Sun, R M Crooks J. Am. Chem. Soc. 121 3779 (1999) 51. S J Green, J J Stokes, M J Hostetler, J Pietron, R W Murray J. Phys. Chem. B 101 2663 (1997) 52. O A Petrii, A V Losev, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1979) Vol. 14, p. 120 53. S Yu Vassiliev, A V Denisov Zh. Tekhn. Fiz. 70 100 (2000) b 54. S Yu Vassiliev, S N Pron'kin, G A Tsirlina, O A Petrii Elektrokhimiya 37 523 (2001) c 55. A J Nam, A Teraen, T A Lusby, A J Melmed J.Vac. Sci. Technol. B 13 1556 (1995) 56. B Zhang, E Wang Electrochim. Acta 39 103 (1994) 57. C Kranz, G Wittstock, H Wohlschlager,W Schuhmann Electrochim. Acta 42 3105 (1997) 58. F F Fan, A J Bard Science 267 871 (1995) 59. A J Bard, F-R F Fan Acc. Chem. Res. 29 572 (1996) 60. F-R F Fan, A J Bard Science 277 1791 (1997) 61. R Schuster, V Kirchner, X H Xia, A M Bittner, G Ertl Phys. Rev. Lett. 80 5599 (1998) 62. Z-X Xie, D M Kolb J. Electroanal. Chem. 481 177 (2000) 63. A N Frumkin Potentsialy Nulevogo Zaryada (Zero-Charge Potentials) (Moscow: Nauka, 1982) O A Petrii, G A Tsirlina 64. M A Vorotyntsev, A A Kornyshev Elektrostatika Sred s Pros- transtvennoi Dispersiei (Electrostatics of Media with Space Dispersion) (Moscow: Nauka, 1993) 65.E L Nagaev Phys. Rep. 222 199 (1992) 66. M D Levi, B B Damaskin, I A Bagotskaya, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1983) Vol. 19, p. 3 67. O A Petrii, A S Lapa, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1987) Vol. 24, p. 3 68. A N Frumkin, N V Fedorovich, E V Stenina, in Elektrokhimiya (Itogi Nauki i Tekhniki) [Electrochemistry (Advances in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1978) Vol. 13, p. 5 69. H Matsumoto, I Oda, J Inukai,M Ito J. Electroanal. Chem. 356 275 (1993) 70. Y-G Kim, S-L Yau, K Itaya J. Am. Chem. Soc. 118 393 (1996) 71.M I Borzenko, G A Tsirlina, O A Petrii Elektrokhimiya 36 509 (2000) 72. V Climent, R Gomez, J M Feliu Electrochim. Acta 45 629 (1999) 73. S Strobac, O M Magnussen, R J Behm Phys. Rev. Lett. 83 3246 (1999) 74. P N Ross J. Chim. Phys. 88 1353 (1991) 75. N Ikemiya, K Yamada, S Hara Surf. Sci. 348 253 (1996) 76. G E Poirier Chem. Rev. 97 1117 (1997) 77. R Yamada, K Uosaki Langmuir 14 855 (1998) 78. D Hobara, K Miyake, S Imabayashi, K Niki, T Kakiuchi Langmuir 14 3590 (1998) 79. K Slowinski, R V Chamberlain, C J Miller, M Majda J. Am. Chem. Soc. 119 11 910 (1997) 80. S Chen, R W Murray, S W Feldberg J. Phys. Chem. B 102 9898 (1998) 81. P Sautet Chem. Rev. 97 1097 (1997) 82. T C Ta, V Konda,M T McDermott J. Phys. Chem. B 103 1295 (1999) 83.W Haiss, J K Sass J. Electroanal. Chem. 386 267 (1995) 84. P Christensen Chem. Rev. 92 197 (1992) 85. H Gleiter Nanostruct. Mater. 6 3 (1995) 86. S Trasatti, O A Petrii J. Electroanal. Chem. 327 353 (1992) 87. T D Gladysheva, B I Podlovchenko Vest. Mosk. Univ., Ser. 2, Khim. 38 3 (1997) d 88. S Z Roginskii Geterogennyi Kataliz. Nekotorye Voprosy Teorii (Heterogeneous Catalysis. Some Theoretical Problems) (Moscow: Nauka, 1979) 89. O M Poltorak, V S Boronin Zh. Fiz. Khim. 40 153 (1969) e 90. K Shimazu, K Uosaki, H Kita, Yo Nodosaka J. Electroanal. Chem. 256 481 (1988) 91. K Shimazu, D Weishaar, T Kuwana J. Electroanal. Chem. 233 223 (1987) 92. Y Takasu, N Ohashi, X-G Chang, Y Murakami, H Minagawa, S Sato, K Yahikozawa Electrochim.Acta 41 2595 (1996) 93. K Yahikozawa, Y Yamashita, Y Matsuda, Y Takasu Chem. Express 7 597 (1992) 94. T Frelink,W Visscher, J A R van Veen J. Electroanal. Chem. 382 65 (1995) 95. A Kabbadi, F Gloaguen, F Andolfatto, R Durand J. Electroanal. Chem. 373 251 (1994) 96. I G Gasella, E Desmondi Electroanalysis 8 447 (1996) 97. M Watanabe, I Salyusa, P Stonehart J. Electroanal. Chem. 271 213 (1989) 98. G Kokkinidis, A Papoutsis, D Stoychev, A Milchev J. Electroanal. Chem. 486 48 (2000) 99. M E Martins, C F Zinola, G Andreasen, R C Salvarezza, A J Arvia J. Electroanal. Chem. 445 135 (1998) 100. O A Petrii, G A Tsirlina, S N Pron'kin, F M Spiridonov, M L Khrushcheva Elektrokhimiya 35 12 (1999) c 101. M Yu Rusanova, G A Tsirlina, O A Petrii, T Ya Safonova, S Yu Vassiliev Elektrokhimiya 36 517 (2000) c 102.T D Gladysheva, E I Shkol'nikov, Yu M Vol'fkovich, B I Podlovchenko Elektrokhimiya 18 435 (1982) c 103. M L Khrushcheva, G A Tsirlina, O A Petrii Elektrokhimiya 30 755 (1994) c 104. S N Pron'kin, O A Petrii, G A Tsirlina, D J Schiffrin J. Electroanal. Chem. 480 112 (2000) 105. N V Smirnova, O A Petrii Elektrokhimiya 24 522 (1988) cSize effects in electrochemistry 106. G A Tsirlina, V N Andreev, Yu B Vassiliev, V E Kazarinov Elektrokhimiya 23 556 (1987) c 107. Yu A Chizmadzhev, V S Markin,MR Tarasevich, Yu G Chirkov Makrokinetika Protsessov v Poristykh Sredakh (Macrokinetics of Processes in Porous Media) (Moscow: Nauka, 1971) 108. A M Kuznetsov Charge Transfer in Physics, Chemistry and Biology, Mechanisms of Elementary Processes and Introduction to the Theory (Berkshire: Gordon and Breach, 1995) 109.A A Kornyshev, A M Kuznetsov, J Ulstrup J. Phys. Chem. 98 3832 (1994) 110. J E T Andersen, A A Kornyshev, A M Kuznetsov, L L Madsen, P Muller, J Ulstrup Electrochim. Acta 42 819 (1997) 111. A M Kuznetsov, J Ulstrup J. Inclusion Phenom. Macrocycl. Chem. 35 45 (1999) 112. Q Chi, J Zhang, E P Friis, J E T Andersen, J Ulstrup Electrochem. Commun. 1 91 (1999) 113. K Slowinski, K U Slowinska,M Majda J. Phys. Chem. B 103 8544 (1999) 114. K Slowinski, H K Y Fong, M Majda J. Am. Chem. Soc. 121 7257 (1999) 115. V Chechik, C J M Stirling Langmuir 14 99 (1998) 116. K Uosaki, T Kondo, T Kanai, K Iso-o Z.Phys. Chem. 212 23 (1999) 117. Yu I Kharkats Elektrokhimiya 9 881 (1973) c 118. A M Kuznetsov, E D German Elektrokhimiya 26 931(1990) c 119. M Miyake, T Torimoto, T Sakata, H Mori, H Yoneyama Langmuir 15 1503 (1999) 120. S Drouard, S G Hickey, D J Riley Chem. Commun. 67 (1999) 121. D A Tryk, A Fujishima, K Honda Electrochim. Acta 45 2363 (2000) 122. R R Nazmutdinov, G A Tsirlina, O A Petrii, Yu I Kharkats, A M Kuznetsov Electrochim. Acta 45 3521 (2000) 123. A M Kuznetsov, J Ulstrup Electrochim. Acta 45 2339 (2000) 124. W R Fawcett, in Electrocatalysis (Eds J Lipkowski, P N Ross) (New York: Wiley-VCH, 1998) p. 323 125. V Yu Kotov, G A Tsirlina, O A Petrii Elektrokhimiya 34 365 (1998) c 126. G A Tsirlina, Yu I Kharkats, R R Nazmutdinov, O A Petrii Elektrokhimiya 35 19 (1999) c 127.C Im, K M Engenhardt, S Gregory Phys. Rev. B 59 3153 (1999) 128. R C Ingram,M J Hostetler, R W Murray, T G Schaaff, J Khoury, R L Whetten, T P Bigioni, D K Guthrie, P N First J. Am. Chem. Soc. 119 9279 (1997) 129. J J Pietron, J F Hicks, R W Murray J. Am. Chem. Soc. 121 5565 (1999) 130. J A Harrison, H R Thirsk, in Electroanalytical Chemistry Vol. 5 (Ed. A J Bard) (New York: Marcel Decker, 1971) p. 67 131. R Q Hwang, M C Bartelt Chem. Rev. 97 1063 (1997) 132. J V Zoval, J Lee, S Gorer, R M Penner J. Phys. Chem. B 102 1166 (1998) 133. F Gloaguen, J M Leger, C Lamy, A Marmann, U Stimming, R Vogel Electrochim. Acta 44 1805 (1999) 134. Q Wu, D Barkey J. Electrochem. Soc. 144 L261 (1997) 135.Y Kurima, J Walton, G E Thompson, R C Newman, G C Wood, K Shimizu Phylos. Mag. B, Phys. Condens. Matter Electron. Opt. Magn. Prop. 70 1 (1994) 136. M Costa, F Chao, E Museux, C Tian J. Chim. Phys. 92 939 (1995) 137. P M Williams, K M Shakesheff,M C Davies, D E Jackson, C J Roberts, S J B Tendler J. Vac. Sci. Technol. B 14 1557 (1996) 138. J S Villarubia J. Res. Natl. Inst. Stand. Technol. 102 425 (1997) 139. A Milchev, E Michailova, R Lacmann, B Muller-Zulow Electrochim. Acta 38 535 (1993) 140. J P Sousa, S Pons, M Fleischmann J. Chem. Soc., Faraday Trans. 90 1923 (1994) 141. L Zeiri,O Younes, S Efrima,M Deutsch J. Phys. Chem. B 101 9299 (1997) 142. J L Zubimendi, G Andreasen,W E Triaca Electrochim. Acta 40 1305 (1995) 143. J Strutwolf,M WuÈ nsche, H Meyer, R Schumacher Z.Phys. Chem. 208 239 (1999) 144. T A Sorenson, D W Suggs, I Nandhakumar, J L Stickney J. Electroanal. Chem. 467 270 (1999) 145. H Natter, R Hempelmann J. Phys. Chem. B 100 19525 (1996) 146. J V Zoval, P R Biernacki, R M Penner Anal. Chem. 68 1585 (1996) 297 O A Petrii, T Ya Safonova J. Solid State Electrochem. 5 (3) (2001) 147. M Yu Rusanova,M Grden, A Czerwinski, G A Tsirlina, 148. O A Petrii, T Ya Safonova, G A Tsirlina,M Yu Rusanova Electrochim. Acta 45 4117 (2000) 149. Y Mastai, G Hodes J. Phys. Chem. B 101 2685 (1997) 150. G A Tsirlina, O A Petrii, S Yu Vassiliev J. Electroanal. Chem. 414 41 (1996) 151. G A Tsirlina, O A Petrii J. Electroanal. Chem. 404 33 (1996) 152. O A Petrii, G A Tsirlina, in Advances in Electrochemistry Sciences Electrochemical Engineering Vol.5 (New York: Wiley-VCH, 1998) p. 61 153. M E G Lyons (Ed.) Electroactive Polymer Electrochemistry Pt 1 (New York: Plenum, 1994) 154. M E G Lyons (Ed.) Electroactive Polymer Electrochemistry Pt 2 (New York: Plenum, 1996) 155. P Batail, K Boubekeur, M Fourmigue, J-C P Gabriel Chem. Mater. 10 3005 (1998) 156. J W Schultze,M M Lohrengel Electrochim. Acta 45 2499 (2000) 157. D Alliata, P HaÈ ring,O Haas,R KoÈ tz,H Siegenthaler Electrochem. Commun. 1 5 (1999) 158. J M Miller, B Dunn Langmuir 15 799 (1999) 159. G A Tsirlina, Yu E Roginskaya, G G Postovalova, S Yu Vassiliev Elektrokhimiya 35 1385 (2000) c 160. M Datta, D Landolt Electrochim. Acta 45 2535 (2000) 161. A I Vorobyova, V A Sokol, E A Outkina Appl. Phys. A, Solids Surf. 67 487 (1998) 162. G Patermarakis, K Moussoutzanis, J Chandrinos Appl. Catal. A 180 345 (1999) 163. M L Sandrock, C D Pibel, F M Geiger, C A Foss J. Phys. Chem. B 103 2668 (1999) 164. H Matsuda, H Tanaka, N Baba Chem. Lett. 621 (1990) 165. T Kyotani, B K Pradhan, A Tomita Bull. Chem. Soc. Jpn. 72 1957 (1999) 166. W Kautek, S Reetz, S Pentzien Electrochim. Acta 40 1461 (1995) 167. Yu M Maksimov, B I Podlovchenko, T L Azarchenko Electrochim. Acta 43 1053 (1998) 168. C-H Yang, T-C Wen Electrochim. Acta 43 207 (1998) 169. G S Hsiao, M G Anderson, S Gorer, D Harris, R M Penner J. Am. Chem. Soc. 119 1439 (1997) 170. M A Anderson, S Gorer, R M Penner J. Phys. Chem. B 101 5895 (1997) 171. S Gorer, G S Hsiao, M G Anderson, R M Stiger, J Lee, R M Penner Electrochim. Acta 43 2799 (1998) 172. S Gorer, J A Ganske, J C Hemminger, R M Penner J. Am. Chem. Soc. 120 9584 (1998) 173. T Szczurek, T Rausch,M Schlesinger, D D Snyder, C H Olk J. Electrochem. Soc. 146 1777 (1999) 174. Y Jyoko, S Kashiwabara, Y Hayashi,W Schwarzacher Electrochem. Solid-State Lett. 2 67 (1999) 175. G Wouters,M Bratoeva, J-P Celis, J R Roos J. Electrochem. Soc. 141 397 (1994) 176. A Cziraki, I Gerocs, B Fogarassy, B Arnold, M Reibold, K Wetzig, E Toth-Kadar, I Bakonyi Z. Met.kd. 88 781 (1997) 177. J R Switzer,M Shane, R Phillips Science 247 444 (1990) 178. G Inzelt,M Pineri, J W Schultze,M A Vorotyntsev Electrochim. Acta 45 2403 (2000) 179. A Bonnefont, R Kostecki, F McLarnon, J C Arrayet, L Servant, F Argoul J. Electrochem. Soc. 146 4101 (1999) 180. E W Bohannan, L-Y Huang, F S Miller, M G Shumsky, J A Switzer Langmuir 15 813 (1999) 181. J A Switzer, B M Maune, E R Raub, E W Bohannan J. Phys. Chem. B 103 395 (1999) 182. N A Bach, N A Balashova Zh. Fiz. Khim. 6 43 (1936) e 183. N A Bach, A A Rakov Zh. Fiz. Khim. 10 18 (1937) e 184. N A Balashova, N A Bakh Zh. Fiz. Khim. 10 575 (1937) e 185. N A Bach, A V Zimin Zh. Fiz. Khim. 13 1064 (1939) e 186. M J Reetz, W Helbig J. Am. Chem. Soc. 116 7401 (1994) 187. R S Ingram, R W Murray Langmuir 14 4115 (1998) 188. M Giersig, P Mulvaney Langmuir 9 3408 (1993) 189. Sh K Shaikhutdinov, F A Muller, G Meste, R J Behm J. Catal. 163 492 (1996) 190. R M Nyffenegger, R M Penner J. Phys. Chem. B 100 17041 (1996) 191. W Li, G S Hsiao, D Harris, R M Nyffenegger, J A Virtanen, R M Penner J. Phys. Chem. B 100 20103 (1996)O A Petrii, G A Tsirlina 298 192. D M Kolb, J C Ziegler, G E Engelmann Z. Phys. Chem. 208 151 (1999) 193. H Sugimura,N Nakagiri J. Vac. Sci. Technol. A, Vac. Surf. Films 14 1223 (1996) 194. F P Zamborini, R M Crooks J. Am. Chem. Soc. 120 9700 (1998) 195. K Seshadri, A M Wilson, A Guiseppi-Elie, D L Allara Langmuir 15 742 (1999) 196. G A Tsirlina, O A Petrii, S Yu Vassiliev Electrochim. Acta 41 1887 (1996) 197. A M Dykhne, S Yu Vassiliev, O A Petrii, A G Rudavets, G A Tsirlina Dokl. Akad. Nauk 368 467 (1999) a 198. A L Buchachenko Vest. Akad. Nauk 69 833 (1999) f a�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) b�Tech. Fiz. (Engl. Transl.) c�Russ. J. Electrochem. (Engl. Transl.) d�Moscow Univ. Bull. (Engl. Transl.) e�Russ. J. Phys. Chem. (Engl. Transl.) f�Herald Russ. Acad. Sci. (En

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (4) 299 ± 320 (2001) #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n04ABEH000617 Pyridopyridines V P Litvinov, S V Roman, V D Dyachenko Contents I. Introduction II. 1,5-Naphthyridines III. 1,6-Naphthyridines IV. 1,7-Naphthyridines V. 2,6-Naphthyridines VI. 2,7-Naphthyridines 299 299 302 311 313 314 of pyridine,2 indole,3 ±6 and quinuclidine,7 oximes of the isoquino- line series 8± 10 and substituted anilines 11 were successfully used as substrates in these reactions. Abstract. syn- the for procedures of development the on Studies Studies on the development of procedures for the syn- thesis of five isomeric pyridopyridines, 2,6- 1,7-, 1,6-, 1,5-, thesis of five isomeric pyridopyridines, viz., ., 1,5-, 1,6-, 1,7-, 2,6- and and properties chemical their 2,7-naphthyridines, and 2,7-naphthyridines, their chemical properties and biological biological activities surveyed are years 15 last the over performed activities performed over the last 15 years are surveyed and and systematised. references 223 includes bibliography The systematised.The bibliography includes 223 references. I. Introduction Unlike the above-listed reactions, condensation of cycloalka- none diethyl acetals 1 with 3-amino-2-chloropyridine (2) afforded azomethines 3 whose successive treatment with lithium diisopro- pylamide (LDA) and thioesters 4 gave rise to aryl(hetaryl)cyclo- (penta ± octa)[b][1,5]naphthyridines 5.2 (CH2)n NH2 120 8C, 2 h + Cl N EtO OEt 1 2 N N 1) LDA (CH2)n 2) RC(S)OEt (4) (CH2)n N Cl N 3 R 5 n=1 ± 4; R=Ar, Het.Pyridopyridines (diazanaphthalenes, napthyridines) are fused heterocyclic systems containing two nitrogen atoms in the adja- cent rings. Naphthyridine derivatives attract interest because of the broad spectrum of their biological activities. These com- pounds are used in diagnostics and treatment of different human diseases (including HIV infection), agriculture, animal husbandry for external- and internal-parasite control, in industry as preser- vatives and components of lubricating coolants for metal process- ing, in analytical chemistry as ligands, etc. Previously, we have considered general methods for the syn- thesis of naphthyridines, their structures and properties.1 In the present review, we survey procedures for the synthesis of five isomeric pyridopyridines, viz., 1,5-, 1,6-, 1,7-, 2,6- and 2,7-naph- thyridines, their chemical properties and biological activities.For convenience, the data are systematised according to the types of isomeric naphthyridines. II. 1,5-Naphthyridines Tryptamine, tryptophan and their derivatives are widely used in the synthesis of indolo[3,2,1-de][1,5]naphthyridine derivatives. Thus refluxing of tryptamine (6) with 2-formylbenzoic acid (7) in alcohol followed by the addition of concentrated hydrochloric acid afforded hexahydrobenzo[h]indolo[3,2,1-de][1,5]naphthyri- dine hydrochloride (8) used as the starting compound in the synthesis of compounds 9 and 10.The latter have an application in the treatment of oxygen deficiency.3 CHO (CH2)2NH2 NH. HCl N COOH a, b c + The major synthetic approach to the construction of the 1,5- naphthyridine system involves condensation of 3-aminopyridine derivatives with dicarbonyl compounds followed by the intra- molecular reaction of the pyridine ring with the terminal carbonyl group under the action of acid catalysts to form a C7C bond. This approach serves as the basis for well-known syntheses according to Skraup, FriedlaÈ nder, Knorr, etc.1 Amino derivatives O HN 6 8 7 NMe NH d N N V P Litvinov N D Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky prosp. 47, 119991 Moscow, Russian Federation. Fax (7-095) 135 53 28.Tel. (7-095) 135 88 37 S V Roman, V D Dyachenko T G Shevchenko Lugansk State Pedagogical Institute, ul. Oboronnaya 2, 348011 Lugansk, Ukraine. Fax (38-064) 255 31 27. Tel. (38-064) 253 83 94 9 10 Received 27 July 2000 Uspekhi Khimii 70 (4) 343 ± 367 (2001); translated by T N Safonova (a) EtOH, D, 2.5 h; (b) HCl; (c) LiAlH4, AlCl3; (d ) 37% CH2O, NaBH3CN, AcOH, MeCN, 20 8C, 2 h.300 1,5-Naphthyridine derivatives 11 and 12 prepared from a,a- dimethyltryptamine (13) and ethyl 3-chlorocarbonylpropionate are also used in the case of cerebral-vascular insufficiency and oxygen deficiency. Cyclisation of the intermediate 14 was carried out under the action of POCl3 and KBH4.4 CH2C(Me)2NH2 a +EtO2C(CH2)2COCl NH 13 CH2C(Me)2NH b, c CO(CH2)2CO2Et HN 14 Me Me Me Me d NH NH N N O 12 11 (a) Py, 20 8C, 1 h; (b) POCl3; (c) KBH4, EtOH ±AcOH (5 : 1); (d) LiAlH4, AlCl3.MeOH, HCl, D Indolonaphthyridine 15 was prepared by cyclocondensation of tryptophan (16) with 2-oxoglutaric acid (17) upon refluxing in methanol in the presence of HCl.10 CH2CHCO2H NH2 +HO2CC(CH2)2CO2H O NH 16 17 CO2Me NH N O 15 The chemical properties of 1,5-naphthyridines and the possi- bilities of their use for the preparation of biologically active compounds have been studied extensively. In particular, bromi- nation of three isomeric thieno[c][1,5]naphthyridines 18a ± c and their N-oxides with tetrabutylammonium perbromide or bromine in the presence of SOCl2 was investigated.12 The reactions involv- ing N-oxides proceed with higher regioselectivity.Thus bromina- tion of thienonaphthyridine 18b with tetrabutylammonium perbromide afforded a mixture of halogeno derivatives 19a ± c, whereas bromination of its N-oxide 20 under the same conditions gave rise to bromide 21 in 55% yield (23% of the initial compound remained unconsumed).12 S S N N S N N 18c N 18b N 18a Br S S N N a Br + + 18b N 19b (10%) N 19a (41%) Br S N Br + +18b (18%) N 19c (5%) V P Litvinov, S V Roman, V D Dyachenko S S N N a Br +20 (23%) N N 20 O 21 (55%) O (a) 1.2 equiv. Bu4N+Br¡3 , NaHCO3, CH2Cl2, 20 8C. One of the recent major procedures used in the synthesis of biologically active 1,5-naphthyridines is based on the replacement of the halogen atom at position 4 of the naphthyridine ring.13 ± 23 Thus substituted 1,5-naphthyridin-4-ylamines 22 were prepared by the reactions of 4-chloro-1,5-naphthyridines 23 with amino- phenols 24 ± 26 13 ± 15 or by the Mannich reaction of 8-(4-hydroxy- phenylamino)naphthyridines 27.16, 17 NHR2 N a or b Cl N R1 N 22 NHC6H4OH-4 N R1 N 23 d c 22 N R1 27 R1=H, Me, Cl, Br, CF3; HO CH2NR32 CH2NR3 R4 2 CH2NR32 (26), OH (25), R2= (24), OH Cl CH2NR32Me Me, , N ,N NR32 =NEt2, NPr2, NHBut, N , N Bn, N NBn; R4=H, Me; N (a) H2NR2, MeOH (EtOH), HCl, D; (b) AcNHR2, MeOH (EtOH), HCl, D; (c) 4-HOC6H4NH2 .HCl, MeOH, D; (d) CH2O, HNR32 , EtOH, D.The compounds 22 exhibit antimalarial activity both in vitro and in vivo. This activity is less pronounced in the case of R1=Br or CF3 and R2=Me or Et.14, 15 (7-Bromo-1,5-naphthyridin-4-yl)(piperidin-2 0-yl)methanol, which is a potential antimalarial drug of a new type, was synthesised in four steps starting from 4,7-dibromo-1,5-naphthyr- idine.22 The reaction of 4-chloro-1,5-naphthyridine with 4-fluorophe- nol (28) afforded 4-(4-fluorophenoxy)-1,5-naphthyridine (29) exhibiting fungicidal and insecticidal activities.23 Cl OC6H4F-4 N N 160 8C, 1 h +4-HOC6H4F 28 N N 29 Cross-coupling of 1,5-naphthyridines 30 containing the bro- mine or iodine atom at position 4 with organotin, -silicon or -boron compounds 31 in the presence of Pd(PPh3)4 gave rise to aryl-substituted 1,5-naphthyridines 32 covered by a patent as herbicides.24 Y R4 N R1 Pd(PPh3)4 + R5 R3 N R2 R6 X 31 30Pyridopyridines R4 N R1 R3 R2 N R5 32 R6 R1=Alk(C1±C4), OAlk, SAlk; R2=H, Me, F, Cl; R3±R6=H, Alk(C1±C4), OAlk, CN, Hal; X=Br, I; Y=SnMe3, SiMe3, B(OH)2.The chlorine atom in 2- or 4-chloronaphthyridines was replaced by the iodine atom under the action of NaI in the presence of hydroiodic acid yielding 2- or 4-iodo-1,5-naphthyr- idines, respectively.25 It was noted25 that treatment of 2-iodo-1,5- naphthyridine with NaNO2 in DMF gave rise to 2,2 0-bis(1,5- naphthyridyl) ether (33) in 44% yield, whereas an analogous reaction with 4-iodo-1,5-naphthyridine did not proceed. N N NaI, HI MeCOEt, 8 h Cl N N N N N N O33 In 6-chloronaphtho[1 0,2 0:4,5]thieno[2,3-c] [1,5]naphthyridine (34), the chlorine atom is readily replaced by the amino or hydrazinoamino group and also is reduced to yield the corre- sponding 1,5-naphthyridine derivatives 35 ± 37.Under the action of ethyl orthoformate, the hydrazine 37 underwent cyclisation to give triazole 38 26 (Scheme 1). Reductive cyclisation of esters 39 afforded 1,5-naphthyridine derivatives 40, which were dimerised under the action of acids.27 LiAlH4 NR Et2O N NH HO (CH2)2CO2Me 39 H R=Me, Bn. Methyl ()-3-ethyl-2,3,3a,4-tetrahydro-1H-indolo[3,2,1-de]- [1,5]naphthyridine-6-carboxylate (41) possessing psychostimulat- ing activity was prepared by alkylation of ester 42 with iodoethane in DMSO or by dehydration of hydroxy derivative 43.28 NH3, MeOH 180 8C, 24 h N NH2NH2 Py N S Cl 34 H2, Pd/C PhH, MeOH, KOH MeO2C HO MeO2C 1,5-Naphthyridine derivatives, like other isomeric pyridopyr- idines, were found in many natural substances and were used for the construction of a series of efficient medicines.NaNO2, DMF Recently, the previously unknown dimeric indole alkaloid cimiciduphytine (44) containing the 1,5-naphthyridine frag- ment 29 and derivatives of eburnane alkaloids 45 exhibiting hypotensive and pain-relieving activities have been isolated from naturally occurring sources. These compounds are suitable for the treatment of cerebral circulation disturbance.30 ± 35 140 ± 150 8C, 3 h I H NR R1=H:R2=H, CO2Me, CO2Et, CH2OH; R1±R2=O; R3=H, Cl, NO2.40 A derivative of penicillanic acid 46 was covered by a patent as an antibiotic.36 Derivatives of hydroxy acid 47 were covered by a patent as anthelmintics.37 N N S NH2 35 N N S NHNH2 36 N N S 37 EtI, DMSO NH N 1% NaOH 42 HCO2H NEt N43 Me N O N HO N HO H OMe CO2Et R3 N N R1R2 Et 45 Ph O O NH N HN O O NH 46 HC(OEt)3 S 38 301 NEt N MeO2C 41 NO O 44S Me N Me COOH Scheme 1 N N N N302 OH COOH N R N 47 , NH. N, N O, N R=OEt, OPr, N Derivatives of annelated benzoindolo-1,5-naphthyridines suitable for improvement of memory attenuated by drugs were described.3, 5 These are benzo[b ][1,5]naphthyridines, which are analogues of inhibitors of neurokininNK1-receptors.Various 1,5- naphthyridine derivatives, in particular, 2-naphthyridinecarbox- amide, possessing antiviral activity were also reported.38. Complexes 48 of Group IB or VIII metals with 1,5-naphthyr- idine derivatives 49 were synthesised. 39 The compounds 48 are used for the preparation of singlet oxygen. These compounds serve as sensitisers in photoinitiation of polymerases and electro- photography and as redox catalysts in organic reactions.39 X L1L2M XH N R1 R1 N R1 R1 (ML1L2Hal)2, THF N R2 R2 N R2 R2 XH ML1L2 49 X48 R1, R2=H, Me, Et, cyclo-Alk(C3±C7); R1±R2=(CH2)n, n=3±6; X=O or S,Mis Group IB or VIII metal; L1, L2=Hal, NO, PPh3, CN, CO.The electronic structures and the UV spectral parameters of six isomeric formyl-substituted 4,6-benzo-1,5- and -1,6-naphthyr- idines were calculated.40 III. 1,6-Naphthyridines Chemistry and biological activities of 1,6-naphthyridine deriva- tives were studied in much more detail than those of 1,5-, 1,7- and 2,7-naphthyridine derivatives. A wide variety of approaches to their syntheses are available. One of the promising procedures for the synthesis of 1,6- naphthyridine derivatives involves functionally substituted pyr- idones as the starting compounds. Thus enamines 50 prepared from 6-formylpyrid-2-ones underwent cyclisation upon refluxing with ammonium acetate in DMF to give 1,6-naphthyridin-2-ones R1 R2 N AcONH4 O HN 51 R1 R1 O R2 R2 NH2OH.HCl N O R1=Me, Et O O Me2N HN HN 52 50 Me R3 N R3NH2 O R1=Me R2=H HN 53 ; R1=Me, Et, Pr, Bu, Ph, O, S, N R2=H, Br, CN; R3=4-MeOC6H4, NH2, NMe2 . V P Litvinov, S V Roman, V D Dyachenko 51 in good yields.41 ± 47 N-Oxides 52 were prepared by cyclisation of the enamines 50 under the action of hydroxylamine hydro- chloride.42, 45 The reactions of the compounds 50 with amines in methanol or DMF afforded 6-substituted 1,2,5,6-tetrahydro-1,6- naphthyridin-2-ones 53.48 Treatment of enamine 54 derived from 5-cyano-6-formyl-2- pyridone with gaseous HBr in AcOH gave rise to 5-bromo-1,6- naphthyridin-2-one (55). The bromine atom in the compound 55 is readily replaced by nitrogen nucleophiles.This fact was used in the synthesis of 1,6-naphthyridine derivatives 56 ± 58.49 ± 52 The compounds 51, 52 and 55 ± 58 exhibit cardiotonic activity.41 ± 52 Br NC HBr N AcOH O N O Me2N 54 NH 55 N N N NH N NaH, DMF O N 56 H N N NHNH2 (HOCH2)2, 170 8C N N NH2NH2 .H2O [(MeO)2CH]2 O O N 58 H N 57 H 1,6-Naphthyridin-5(6H)-ones 59a ± g were prepared by the reactions of pyridines 60a ± g with 1,3,5-triazine upon refluxing in ethanol in the presence of sodium ethoxide. The reaction of the compound 60a with dimethylformamide diethyl acetal gave rise to enamine 61. The reactions of the latter with amines upon refluxing in ethanol, DMF, m-xylene or pyridine afforded 6-substituted naphthyridinone derivatives 62.53, 54 Refluxing of vicinal diamines 63 with enamine 61 yielded naphthyridinones 64, which were converted into tetracyclic derivatives 65 under the action of POCl3.55 O R R N CO2Et CO2Et EtO2C EtONa, EtOH HN + N N Me Me Me N 60a ± g N 59a ± g R = H (a), Ph (b), 4-ClC6H4 (c), 3,4-(MeO)2C6H3 (d), 2-NO2C6H4 (e), 4-NO2C6H4 (f), 3-pyridyl (g).CO2Et EtO2C RNH2 59a+Me2NCH(OEt)2 Me N Me2N 61 O CO2Et RN Me N 62 R=H, Ph, 2-FC6H4, 4-FC6H4, 3-F3CC6H4, 2-NH2C6H4. A A NH2 POCl3 B NH2O 61 + BC C CO2Et N D D NH2 63 Me N 64Pyridopyridines A N BC CO2Et N D Me N 65 A=B=C=D=CH; A=B=C=CH, D=N; A=N, B=C= D=CH; A=C=D=CH, B=N; A=C=N, B=D=CH; A=C=CH, B=D=N.The reactions of 1,4-dihydropyridines with 1,3,5-triazine in the presence of a base afforded 1,4-dihydro-1,6-naphthyridines 66a ± h. These compounds can also be prepared in two stages by the reactions of 1,4-dihydropyridines with dimethylformamide diethyl acetal followed by treatment of aminovinyl intermediates 67 with ammonia.56 The compounds 66a ± h were covered by patents as agents promising in the treatment of cardiovascular diseases.57 ± 61 Ar N Me NH , NaH, DMF RO2C CO2R N N HN CO2R 20 to 110 8C Me Me NH Ar O 66a ± h Ar CO2R RO2C (EtO)2CHNMe2 NH3, EtOH 66a ± h DMF Me NH CH CHNMe2 67 Ar=Ph (a), 4-ClC6H4 (b), 2-NO2C6H4 (c), 2-pyridyl (d), 2-thienyl (e), 2-FC6H4 (f), 2-CF3C6H4 (g), 2-F-3-ClC6H3 (h); R=Me, Et, But.Substituted pyridones are widely used as the starting com- pounds in the synthesis of substituted 1,6-naphthyridines. Thus the reactions of 1,3,5-triazine with pyridones 68a,b yielded anne- lated 1,6-naphthyridines 69a,b. The tricyclic compounds 69 con- taining the cyclopropyl substituent were covered by patents as efficient antibiotics.62 Pyrido[2,3-b] [1,6]naphthyridine 70 pre- pared from 1,4-dihydropyridine 71 possesses spasmolytic activity and is used in the treatment of cerebral, cardiac and peripheral artery diseases.63 O N , NaH, THF F CO2Et N N Me N F X 68a,b O O F NH N F X 69a,b X = N (a), CF (b). C6H4CF3-2 C6H4CF3-2 O O N CO2Et EtO2C , NaH, DMF NH HN N N Me Me NH NH 71 70 The reactions of 5-alkanoyl(aroyl)-6-methylpyridin-2(1H)- ones with N-bromosuccinimide in CCl4 afforded bromides 72.Successive treatment of the latter with (Me2N)2CHOBut (Breder- eck's reagent) and ammonium acetate gave rise to 5-alkyl(aryl)-3- bromo-1,6-naphthyridin-2(1H)-ones 73 in high yields.64 The com- 303 pounds 73 were converted under the successive action of POCl3 and ammonia into 5-alkyl(aryl)-2-amino-3-bromo-1,6-naphthyr- idines 74, which are promising starting compounds for the preparation of new biologically active compounds. O O O Br Br HN b, c d, e a HN HN Me Me COR R N COR 72 73 NH2 Br N R N 74 R=Me, Et, Prn, Bui, Ph, 4-NO2C6H4; (a) NBS (N-bromosuccinimide), CCl4; (b) (Me2N)2CHOBut, dioxane; (c) NH4OAc, DMF; (d ) POCl3; (e) NH3, EtOH.The Knoevenagel reaction of substituted 4-amino-3-formyl- pyridin-2-ones with CH-acids in the presence of piperidine was used in the synthesis of 2-imino-1,6-naphthyridin-5(6H)-ones (75) possessing antitubercular activity.65 O Py, HN CHO R1N +NCCH2X Me NHR2 O X R1N NH Me NR2 75 R1, R2=PhCH2, Ph(CH2)2; X=CN, CO2H, CO2Me, CO2Et, 4-NO2C6H4. When heated with POCl3, substituted pyridone 76 underwent cyclisation with simultaneous replacement of the 2-oxo group by the chlorine atom to give 3-chloro-4-cyanobenzo[b] [1,6]naphthyr- idine 77.66, 67 CHO N HN POCl3 N Cl NH O CN CN 77 76 The reaction of pyridone 78 with carbon disulfide and sodium hydride in benzene followed by methylation with iodomethane produced the 5,7-bis(methylthio)-1,6-naphthyridinone derivative 79.68 SMe NC CN NC 1) CS2, NaH N 2) MeI SMe N O Me O N C6H4Me-4 C6H4Me-4 79 78 5,6,7,8-Tetrahydro-1,6-naphthyridin-5-ones 80 were synthes- ised by the reactions of 4-amino-5,6-dihydropyridin-2(1H)-one with amino ketone hydrochlorides.69304 NH2 PrOH, D Me +RCO(CH2)2NMe2 .HCl O Me R NH N NH2 (CH2)2COR Me Me O O Me Me NH NH 80 R=Ph, 4-MeOC6H4, 4-ClC6H4. Piperidones are used as starting compounds in the synthesis of 1,6-naphthyridines. Thus the reactions of N-substituted piperi- dones 81a,b with 2-amino-4-chlorobenzoic acid afforded 2-aryl- (heteryl)tetrahydrobenzo[b][1,6]naphthyridines 82a,b, which serve as inhibitors of interleukin 1 and were covered by patents as antiinflammatory agents.70 ± 72 The reaction of the compound 82a with phenylhydrazine in the presence of concentrated hydrochloric acid gave rise to pyrazolo[4,3-c]quinoline (83) exhibiting antiinflammatory activ- ity.70, 72 Cl POCl3 NR COOH+RN O Cl D, 3 hCl NH2 81a,b N 82a,b NPh N PhNHNH2 82a HCl, EtOH, D N Cl (CH2)2NHC6H4CN-4 83 R=4-NCC6H4 (a), pyrimidin-2-yl (b).The reactions of N-benzylpiperidin-4-one (84) with 3-amino- acrolein and its homologues were used for the preparation of 5,6,7,8-tetrahydro-1,6-naphthyridines 85a ± c.73 O R2 a BnN +H2NC(R1) C(R2)CHO R1 Bn N N 85a ± c 84 R1=R2=H(a); R1=Me, R2= H (b); R1=H, R2=Me (c); + (a) NH2 AcO7, Et3N, 120 8C, 24 h.Partially hydrogenated 1,6-naphthyridines 86 were synthes- ised by the Hunch condensation of aromatic aldehydes with 2,4- dioxopiperidine (87) and derivatives of 3-aminocrotonic acid.74 The compounds 86 were covered by a patent as drugs for the treatment of cardiovascular diseases.75 O Ar O COR COR MeOH HN + ArCHO+HN D Me Me O H2N NH 86 87 R=MeO, EtO, MeNH, Et2N, NH(CH2)2OH. The reactions of 6-amino-5-methylisoquinolin-1(2H)-one (88) with 4-chloronicotinic acid (89a) or ethyl 4-chloro-2,6-dimethyl- nicotinate (89b) afforded amino-substituted 6-aminoisoquino- lines 90a,b, which underwent cyclisation to form 5H-iso- quinolino[6,7-b] [1,6]naphthyridin-12-ones 91a,b, respectively.76 The latter were used in the synthesis of 10-substituted 1,6-naphthyridine derivatives 92a,b and 93a,b, which are ana- logues of 9-azaellipticine possessing antitumour activity.How- ever, it appeared that the replacement of the pyrrole fragment in 9-azaellipticine by the 4-pyridone fragment led to a loss of biological activity.76 H2N Me R1 R1 R1 R1 O NH(CH2)3NEt2 N N R1 NH Me 93a,b R1=R2=H(a), R1=Me, R2=Et (b); (a) H2SO4, CF3CO2H; (b) H2SO4 or CF3SO3H; (c) POCl3, PCl5; (d) NH2(CH2)3NEt2 . Pyridinecarboxylamides are widely used in the synthesis of 1,6-naphthyridines. Thus heating of a mixture of 2-methylnicoti- namide with N,N-dimethylacetamide dimethyl acetal in the pres- ence of NaH gave rise to 7-methyl-1,6-naphthyridin-5(6H)-one (94), which is used for eliminating harmful effects caused by ionising radiation or chemotherapeutic agents during tumour treatment.77CONH2 Me N When heated in polyphosphoric acid (PPA), 6-phenyl-2- styrylnicotinamides (95) underwent cyclisation to form substi- tuted 5-oxo-5,6,7,8-tetrahydro-1,6-naphthyridines 96.78 Analo- gously, benzo[b] [1,6]naphthyridines 98 were prepared from amides of the quinoline series 97.78 ± 80 Naphthyridine 98 (R1=R2=H) was also synthesised by cyclisation of 3-cyano-2- styrylquinoline 99 under the action of a mixture of polyphos- phoric and sulfuric acids.80 Ph N 95 R=CH2=CHCH2, iso-C5H11, Ph, 3-MeC6H4.V P Litvinov, S V Roman, V D Dyachenko O R1 CO2R2 a N NH + Cl R1 88 89a,bO R1 CO2H b NH N NH Me 90a,b R1 OH O c N N NH Me 91a,b Cl O R1 d N N NH 92a,b Me NaH +(MeO)2C(Me)NMe2 N O O PPA NR NHR D Ph Ph N 96 O NHMe 94 PhPyridopyridines O O NR1 PPA NHR1 D N C6H4R2 N C6H4R2 98 97 R1=H, Ph, Bn; R2=H, 4-MeO, 3-Br.CN PPA, H2SO4 98 D Ph N 99 The reactions of N-substituted 2-chloro- or 4-chloronicotina- mides with nitriles containing the active methylene group in the presence of bases were used for the preparation of amino derivatives of 1,6-naphthyridin-5(6H)-one or 2,7-naphthyridin- 1(2H)-one, respectively.81, 82 A procedure was developed for the synthesis of dodecahy- droisoquinolino[2,1-g][1,6]naphthyridines 100 based on diethyl-2- methylnicotinamide (101) and 3,4-dihydroisoquinolines 102a,b.The compounds 100 are highly efficient a2-adrenoreceptor antag- onists 83, 84 and were covered by patents as drugs for the treatment of hypertonia, depression and diabetes, for inhibition of throm- bocyte aggregation and for weight reduction.85± 90 It should be noted that the structural analogue of a-yohim- bine, viz., 6-methoxy-N-methylsulfonyl-6H-isoquinolino[2,1-g] [1,6]naphthyridine (100, R1=MeO, R2=MeSO2 , the prepara- tion RS-15385),91 proved to be a very efficient drug in the treatment of impotence. A laboratory procedure for its prepa- ration on a kilogram scale was developed.92, 93 R1 R1 CONEt2 N O b a + N Me N 101 102a,b N R1 R1 O N N c d H H H H H H HN HN R1 N H H HR2N 100 R1 = H (102a), MeO (102b); R2=MeSO2, MeO(CH2)2SO2 , HO(CH2)2SO2, ButNHSO2, H2NSO2, Me2NCO, Cl(CH2)3SO2 ; (a) Pri2NLi, THF,740 8C; (b) H2, Rh/Al2O3, AcOH; (c) LiAlH4, THF, D; (d) R2X (X=Cl, Br), CH2Cl2, Et3N.Reductive cyclisation of substituted isoquinoline 103, which was prepared from 6,7-dimethoxy-1-methylisoquinoline (104) in four steps, was successfully used in the synthesis of 8,9-dimethoxy- 1H-benzo[d,e] [1,6]naphthyridine (aaptamine) (105). The latter compound was isolated from sea sponge Aaptos aaptos and exhibits sympatholytic and hypotensive activities.94, 95 MeO MeO a, b, c, d e, f N N MeO MeO Me 104 NO2 CH CHNO2 103 305 MeO N MeO HN 105 (a) HNO3,740 8C; (b) SeO2, dioxane; (c) MeNO2, Al2O3; (d) Al2O3, PhH, D; (e) Fe, AcOH, 20 8C, 2 h; ( f ) 90 8C, 45 min. Pyranopyridines are convenient starting compounds in the synthesis of 1,6-naphthyridine derivatives.Thus, the reaction of 7-phenylpyrano[4,3-b ]pyridin-5-one (106) with gaseous ammonia in ethanol was accompanied by the replacement of the oxygen atom of the pyran ring by the nitrogen atom to form hydroxy- naphthyridine 107. The latter was dehydrated with an alcoholic solution of HCl to give 7-phenyl-1,6-naphthyridin-5(6H)-one (108a). Treatment of pyranopyridine with propylamine led to the pyran ring opening to form 2-phenacyl-N-propylpyridine-3-car- boxamide (109), which underwent cyclisation under the action of a 5% HCl solution giving rise to N-propyl-substituted naphthyr- idinone 108b.96 O O Ph N O O 106 NH3 (gas) NH HCl NHOH EtOH EtOH N N Ph Ph 108a 107 O O PrNH2 HCl NPr NHPr EtOH N N Ph CH2COPh 109 108b Analogous reactions of 7-methylpyrano[4,3-b ]pyridines 110 with ammonia or aliphatic primary amines afforded 1,6-naph- thyridinones 111.97 O O R2NH2 NR2 O Me N Me N R1 111 R1 110 R1=Ac, CO2Et; R2=H, Me, Bu, (CH2)2NH2, (CH2)2OH, CH2=CHCH2.1,6-Naphthyridin-2(1H)-ones 113 were prepared from pyr- ano[4,3-b ]pyridine-2,7-diones 112 by successive treatment with NH3 , HBr and Zn in acetic acid.98 It was believed that the attack of the ammonia molecule on the C(5) atom led to the pyran ring opening. Subsequent decarboxylation afforded a tautomeric mix- ture of substituted tetrahydropyridines 114a,b.Dinitrile 115 generated by elimination of the ammonia molecule underwent cyclisation under the action of HBr to form 1,6-naphthyridine derivative 116. Dehydrobromination with Zn in acetic acid gave rise to the final products 113. O O R1 R1 NH NH NH3 CN CN R2 R2 7CO2 + COO7 NH2 O H3N H2N O 112306 O R1 NH CN R2H2N H NH2 114a O R1 R1 NH HBr R2 R2 Br CN CN 115 R1, R2=H, Me, Ph. Successive treatment of N-substituted 4-aminopyridines 117 with tert-butyllithium and 1,3-chloroiodopropane yielded tetra- hydro-1,6-naphthyridines 118.99 NHR NHR Li ButLi N N 117 R=COBut, CO2But. The reaction of chlorocarbonic ester with the lithium deriva- tive of enamine 119, which was prepared by the reaction of 2,4,6- trimethylpyridine with benzonitrile and phenyllithium, afforded a mixture of 1,6-naphthyridin-5(6H)-ones 120 and 121.100 Me PhCN, PhLi Et2O Me Me Me N O Me NH Me N 120 Aconvenient procedure was developed 101, 102 for the synthesis of 1,6-naphthyridine derivatives 122 containing the alkylselanyl substituent at position 7.Thus multicomponent condensation of cyanoselenoacetamide (123), 2-furfurylideneacetoacetic ester (124) and alkyl halides was carried out under the action of a twofold excess of N-methylmorpholine. The reaction involves the Thorpe dimerisation of the amide 123 yielding compound 125, its reaction with a molecule of furfurylideneacetoacetic ester (124) to form adduct 126, regioselective cyclocondensation of the adduct 126 producing substituted 1,4-dihydro-1,6-naphthyridine-7-sele- nolate (127) and alkylation of the selenolate 127 with alkyl halides giving rise to naphthyridines 122.Se MeN O NCCH2CNH2 HN 20 8C, EtOH 123 V P Litvinov, S V Roman, V D Dyachenko O O R1 NH2 NH EtO2C CN 7NH3 R2 SeNH2 Me O NH Se NH2 H2N114b CN 126 O O R1 NH NH Zn O NH2 AcOH R2 EtO2C N NH2 NH2 N 113 N 116 Me Se7 HN CN 127 O NH2 EtO2C NR N I(CH2)3Cl Me SeCH2R HN CN 122 N 118 R=H, Me, Et, Pr, C5H11, CH2Br, (CH2)3Br, C(Me)=CH2, HC=CH2, CO2Et; Hal=I, Br, Cl. This procedure also allows one to construct the naphthyridine system from other a,b-unsaturated carbonyl compounds under mild conditions.103 R1 Me R2 NHLi ClCO2Et +R4CH2Cl NCCH2CSeNH2+ O R3 Ph N 119 R1 NH2 O Me R2 N NH + R3 SeCH2R4 HN CN Ph N Me Ph 121 CO2Et R1±R3=H, Alk, Ar, Het, CN, COOH, COOAlk, COAr, COHet, CONH2, CONHAr; R4=Alk.Multicomponent heterocyclisation of benzaldehyde with malononitrile and its dimer 128 in the presence of piperidine or morpholine was used for the construction of the 1,6-naphthyr- idine system 129. Heterocyclisation was accompanied by the insertion of the cyclic-amine fragment into the molecule to form the compounds 129.104 NC NH2 PhCHO+CH2(CN)2+ NC CH2CN 128 H2N CH CCO2Et O O. N R=N , 124 COMe SeNH2 Se NC125 An ingenious method for the construction of the 1,6-naph- thyridine system with the use of the `chelate' methodology was developed.105 Thus the reaction of 3-acetyl-4-amino-5,5,5-tri- fluoro-3-penten-2-one (130) with Ph2BOBu afforded chelate com- plex 131 in which the oxygen atom was replaced by the nitrogen atom upon heating with primary amines.The resulting complexes O MeN 7H2O,7H2Se RCH2Hal H(Me)N+ O EtOH, MeN O, argon 20 8C Ph NH2 NC N RH R N H2N CN 129Pyridopyridines 132 reacted with two equivalents of dimethylformamide dimethyl acetal to give 5-trifluoromethyl-1,6-naphthyridin-4(1H)-ones 133 in 77%± 87% yields. It was believed 105 that the complexes 132 reacted with dimethylformamide dimethyl acetal to give conden- sation products at the methyl groups (134), which underwent deborylation with simultaneous cyclisation of intermediates 135 under the action of the methanol eliminated. F3C Ac CF3 Ph2BOBu 7BuOH Ac Ac NH2 130 HN F3C BPh2 (MeO)2CHNMe2 NR Ac Me 132 Me2NCH F3C OHC NHR CH NMe2 R=Bu, Bn.Ethyl (4,6-diamino-3,5-dicyano-2-pyridyl)acetate 136 or (4,6- diamino-3,5-dicyano-2-pyridyl)acetanilide 137 were converted into 1,6-naphthyridine-5,7(6H,8H)-dione 138 upon refluxing in acetic acid in the presence of HCl. Treatment of the ester 136 with trichloroacetonitrile in the presence of triethylamine afforded naphthyridine 139.106 NH2 NC CN N CH2CO2Et H2N 136 NH2 NC PhNH2 136 N H2N 137 6-Methyl-2,4-diphenyl-5,6,7,8-tetrahydro-1,6-naphthyridine (140) was synthesised by cyclocondensation of benzylideneaceto- phenone (141) with enamino nitrile 142 in the presence of sodium ethoxide.107 HN BPh2 RNH2, D O 7H2O Me 131 7MeOHHN F3C BPh2 MeOH NR 7Ph2BOMe CHC CH CHNMe2 O 134 O CF3 HN CHNMe2 N CH 72Me2NH NR 133 135 NH2 O NC AcOH, HCl NHO N H2N 138 NH2 NH2 NC CCl3CN N DMF, Et3N N CCl3 H2N CO2Et 139 CN AcOH, HCl 138 CH2CONHPh 307 CN EtONa, EtOH, D MeN 7HCN +PhCH CHCOPh 141 NH2 142 Ph MeN Ph N 140 Of other procedures for the synthesis of 1,6-naphthyridines, noteworthy is three-component cyclisation of 2,3-dioxobutanal phenylhydrazone (143) with diethyl 3-amino-2-cyanopent-2-ene- dioate (144) and ammonium acetate to form substituted 1,6- naphthyridine-5,7(3H,6H)-dione 145. Polyfunctionalised pyri- dine derivative 146 was obtained as the intermediate in this reaction.108 CN H2N Ac PhHNN C + CHO +AcONH4 CO2Et 143 EtO2CCH2144 Me Me CN N CN N AcONH4 PhHNN PhHNN O CO2Et CO2Et 146 NH145 O The synthesis of partially hydrogenated N-substituted tetra- hydro-3-nitro-1,6-naphthyridines 147 involved diene condensa- tion with the inversion of electrophilicity.5-Nitropyrimidine reacted as the diene component with enamines 148. Elimination of HCN from adducts 149 followed by elimination of pyrrolidine from intermediates 150 afforded the products 147.109 R1 N HNR2 N 2 NO2+ R1N NO2 7HCN N N NR22N 148 149 H NO2 NO2 R1N R1N 7HNR22N N NR22 147 150 N.R1=Me, Ac; NR22 = A handy procedure for the one-pot synthesis of the previously unknown 1-substituted 1,2,3,4,5,6,7,8-octahydro-1,6-naphthyri- dines 151, which are convenient building blocks and precursors of Li+7 a NR2 NR2 b R1N R1N 152 NR2 NR2 c Cl NR1 NR1 151 R1=Ph, CH2Ph, (dl)-CH(Me)Ph, (S)-CH(Me)Ph; R2=Me, Bn, COPh; (a) LDA or Et2NLi, THF,730 8C, 30 min; (b) Cl(CH2)3Br,778 8C; (c) D, 4 h.308 biologically active compounds, was developed. The procedure involves metallation of imines 152 with lithium diisopropylamide or diethylamide followed by alkylation of the lithium derivatives with 1,3-bromochloropropane and intramolecular cyclisation of the resulting chloroalkylamines.110 2-Benzyl-1,6-dimethyl-3,10-di(methoxycarbonyl)-8-phenyl- decahydro-1,6-naphthyridin-4-one was synthesised and its struc- ture was confirmed by X-ray diffraction analysis.111 At the end of the section dealing with the synthesis of 1,6- naphthyridines, let us consider special procedures for the prepa- ration of their annelated analogues.Thus a general procedure involves cyclisation of carbonyl derivatives of 4-chloropyridine with aromatic and heterocyclic amines under the action of acids. The reactions of 3-benzoyl-4-chloropyridine with substituted ani- lines afforded amine 153, which underwent cyclisation under the action of polyphosphoric acid to form 1,6-naphthyridine deriva- tives 154 112 (see also the synthesis of the compounds 91a,b 76). R4 COPh R3 N + Cl R2 H2N R1 R4 COPh R3 N PPA 110 8C R2 NH R1 153 R1, R2=H, OH, OMe, Cl; R1±R2=CH=CH7CH=CH; R3=H, OMe, NO2; R4=H, OMe.2-Chloronicotinoyl chloride (155) was used as the starting compound in the synthesis of heteryl-annelated 1,6-naphthyri- dines. Its reactions with 4,5-dihydro-1H-imidazolines (156a) or tetrahydropyrimidines (156b) gave rise to the corresponding N-acyl derivatives 157a,b, which underwent cyclisation under the action of potassium tert-butoxide to yield imidazo- (158a) or pyrimido[1,2-g] [1,6]naphthyridinones (158b), respectively. The compounds 158a,b exhibit antiallergic and antiinflammatory activities.113, 114 COCl (CH2)n HN + N Cl N ArCH2 156a,b 155 a R1 NH2 O O R2 159 b R1=H, Br, Cl; R2=H, OMe, Br; (a) 155, Et3N, PhH, D, 8 h; (b) (160), AcOH; (c) Py or Et3N± PdCl2 .DMF R4 R3 N N R2 R1 154 Pri2NH or Et3N, CH2Cl2 0 8C, 14 ± 16 h NHCO R1 Cl O O R2 161 N Cl N R1 O O 162 R2 V P Litvinov, S V Roman, V D Dyachenko (H2C)n CON N Cl CH2Ar N 157a,b n = 1 (a), 2 (b). The reactions of 3-aminocoumarins 159 with 2-chloronicoti- noyl chloride (155) or 2-chloro-3-formylquinoline (160) afforded N-acyl derivatives 161 or Schiff bases 162, which were dehydro- chlorinated under the action of bases with simultaneous cyclisa- tion to give benzopyrano[3,4-h ] [1,6]naphthyridines 163 or 164, respectively (Scheme 2).115, 116 Flash pyrolysis of aminopyridine derivatives 165a ± c gave rise to intermediates 166a ± c containing the heterodiene system.Intra- molecular cyclisation of the latter involving the substituent R produced annelated naphthyridines 167 ± 169.117 NHCO2But CH(OH)R 600 8C N 165a ± c R =(CH2)3CH=CH2 R= N R= S N N 169 (66%) R=(CH2)3CH=CH2 (a), Ph (b), 2-thienyl (c). Dihydrobenzo[h ] [1,6]naphthyridines 170 and 171 were syn- thesised by Eu(fod)3-catalysed intramolecular Diels ± Alder cycli- sation of the diene system of oxazoles 172 and 173, respectively, with the double bond of the side-chain substituent.118 R1 N R2 c N O HN O 163 R1 c R2 O O CHO Cl N O (CH2)n N ButOK N HN Ar 158a,b NH R CH N 166a ± c N NH 167 (64%) N 168 (100%) S Scheme 2 O NN164309 Pyridopyridines O NHCOCH CHCO2Et O HNEtO2C Eu(fod)3 N 1,2-Cl2C6H4 , D, 16 h N 170 172 COCF3 COCF3 NCH2CH CHCO2Me O N MeO2C Eu(fod)3 N 1,2-Cl2C6H4 , D, 18 h N viz., by oxidation of the dihydro derivative 182 to form pyridine- dicarboxylic ester 183 followed by reduction of the nitro group with simultaneous cyclisation or by storage of a solution of the compound 182 in chloroform in sunlight to give nitroso com- pound 184, which undergoes cyclisation under the action of concentrated HCl.120 Treatment of a solution of pyrido[1,2,3-de] [1,4]benzooxazine 185 with a 60% NaOH solution in DMF followed by the addition of methyl isocyanate afforded the tetracyclic 1,6-naphthyridine derivative 186.The reactions of the latter with substituted pyrrolidines or piperazines gave rise to amino derivatives 187.The compounds 186 and 187 and their salts possess bactericidal properties.121 171 173 OH O O Eu(fod)3=Eu[ButC(O)CH2C(O)(CF2)2CF3]3. F F CO2Me NMe c a, b O N F Me F N O O Me Me186 185 OH O F 1,3-Dipolar cycloaddition of 3-phenylpropynal (174a) or phenyl phenylethynyl ketone (174b) to the N-aminopyridinium salt 175 afforded a mixture of isomeric pyrazolo[1,5-a]pyridines 176a,b and 177a,b. Debenzylation of the compounds 176a,b followed by successive treatment with formaldehyde and sodium cyanoborohydride gave rise to naphthyridines 178a,b. Oxidation of the latter with m-chloroperbenzoic acid yielded N-oxides 179a,b.119 NMe CH2NHCO2Bn O N R2N a O +PhC CC(O)R 7O3SMes Me N+ 174a,b NH2 175 CH2NHCO2Bn C(O)R C(O)R 187 NR2 are substituted pyrrolidinyls or piperazinyls; (a) 60% NaOH, DMF; (b) MeNCO; (c) R2NH, DMSO.+ N N Ph N Ph N H2C 177a,b NHCO2Bn 176a,b O Me Me N R N R b, c, d e Amido and thioamido derivatives of a,b-unsaturated nitriles were used in the synthesis of annelated 1,6-naphthyridines. Thus substituted benzothiazolo[3,2-g] [1,6]naphthyridin-5-ones (188 and 189) were prepared by refluxing 3-hydroxy-1H-pyrido- [2,1-b]benzothiazol-1-one (190) with arylmethylidenecyano(thio)- acetamides 191 and 192, respectively, in ethanol in the presence of triethylamine.122 Ph Ph 176a,b OH S CN Ar N N N N Et3N + 178a,b N 179a,b EtOH, D C(O)NH2 191 O 190 R = H (a), Ph (b); Mes=2,4,6-Me3C6H2 ; (a) K2CO3, MeCN; (b) 30% HBr, AcOH; (c) HCHO; (d ) NaBH3CN; (e) 3-ClC6H4CO2H.HN O S N CN Dihydrobenzo[h ][1,6]naphthyridines 180 and 181 can be prepared from 4,6-dimethyl-2-(2-nitrophenyl)-1,2-dihydropyri- dine-3,5-dicarboxylic esters (182) according to two procedures, Ar O 188 Me Ar=4-MeOC6H4, 4-ClC6H4 . CO2R RO2C SH N S CN Et3N Ar 190+ N Me C6H4NO2-2 EtOH, D CHAr C(S)NH2 NH 192 NH O 189 182 O Me Me Ar=Ph, 4-MeOC6H4. RO2C CO2R RO2C Zn, NH4Cl NOH MnO2 Me N EtOH, H2O C6H4NO2-2 Me N Thermal isomerisation of 3-(2,2-dicyanovinyl)-4-aminopyri- dines 193 in DMSO afforded a mixture of annelated 1,6-naph- thyridines 194 and azepines 195.123 180 183 X X O Me Me X N N DMSO RO2C RO2C N NOH CN CO2R NO HCl CN hn N +N 140 8C, 1 ± 2.5 h N CHCl3 N Me CN Me N CN CN 194 195 NC 184 193 X=CH2, NPh, O, S.Cl 181 R=Me, Et.310 Of other procedures for the synthesis of annelated 1,6-naph- thyridines, noteworthy is three-component condensation of the malononitrile dimer, salicylaldehyde and acetone giving rise to 1,6-naphthyridines 196.124 CHO NC NH2 AcONH4 + +MeCOMe EtOH NC CH2CN OH O N Me N NH2 CN 196 Considerable recent attention has been focused on procedures for modifications of 1,6-naphthyridine derivatives with the aim of searching for new biologically active compounds. Thus conden- sation of 2-amino-3-bromo-1,6-naphthyridines 74 with potassium O-ethyl xanthate in N-methylpyrrolidone afforded thiazolo- [4,5-b ] [1,6]naphthyridine-2(3H)-thiones 197 whose subsequent methylation and hydrolysis of the methyl group in the presence of sodium methoxide afforded thiazolonaphthyridones 198 inhib- iting adenosine-30,50-cyclophosphate phosphodiesterase (cAMP PDE III).64 S HN NH2 S Br N N MeI EtOC(S)SK K2CO3, DMF R N R N 197 74 O SMe HN N S S N N MeONa DMF R N R N 198 R=Me, Et, Prn, Bui, Ph, 4-NO2C6H4.The structural fragment of hydrogenated pyrido[3,2-ij ]-1,6- naphthyridin-6-one 199 is present in the alkaloid matrine (soph- ocarpidine) found in dried roots of several plants belonging to the genus Sophora. The compound 199 was synthesised from 1,6-naphthyridin- 5(6H)-one (200) in six steps.125 The key stage of the scheme consists in the Heck reaction of 8-iodo-5-methoxy-1,6-naphthyr- idine (201) with ethyl acrylate.Reduction of the double bond in the resulting derivative 202 and cyclisation of compound 203 under the action of sodium methoxide complete the synthesis. Cl O O c b a N HN HN N N N 200 I I OMe OMe d e N N N N I 201 CH CHCO2Et 202 V P Litvinov, S V Roman, V D Dyachenko OMe OMe c N N N NH CH2CH2CO2Et O 199 203 (a) I2, 4MNaOH, 80 8C, 4 h; (b) POCl3; (c) MeONa, MeOH; (d) CH2=CHCO2Et, Pd(OAc)2, Et3N, MeCN; (e) H2, PtO2, MeOH. Condensation of 7-amino-8-cyano-1,2,3,4-tetrahydro-1,6- naphthyridine (204) with dimethylformamide diethyl acetal afforded azomethine 205. Its treatment with ammonia gave rise to 1-amino-7,8,9,10-tetrahydropyrimido[4,5-h ] [1,6]naphthyri- dine (206).126 N Me2NCH(OEt)2 H2N NH CN204 N N NH3 (liq) N N Me2NCH NH NH N CN205 NH2 206 Diazotisation of the amine of the naphthyridine series 207 with a solution of sodium nitrite in 10% H2SO4 afforded naph- thyridinone 208, which was used as the starting compound in the synthesis of the antagonist of benzodiazepine receptors 209.127 ± 129 CO2Et BnN BnN CO2Et H2NNHCO2Me NaNO2 O NH2 H2SO4, H2O NH N 207 208 NCO2Pri CONHNHCO2Me BnN NCO2Pri PPh3, Et3N, THF O NH N N OMe O BnN O HN 209 The compounds 210, including 1,6-naphthyridine derivatives, as well as octahydronaphthyridines 211 are of interest from the pharmacological standpoint.130 ± 134 Thus the compounds 210 possess low toxicity and are used as components of drugs for prophylaxis and treatment of infectious diseases caused by various pathogenic bacteria.130 ± 133 Octahydronaphthyridines 211 are used in the treatment of depression or in cases of abstinence.134 COC6H4F R6N O R3 COR2 N X R1 Y R4 R5 R4 R3 R2 NH NR1 211 210 R1=H, Et, (CH2)2OH, (CH2)2Cl, CH2=CH, cyclo-Alk; R2=OH, OAlk; R3=H, Hal, NH2; R4=H, R1=H, Alk(C1±C6), CF3, CN, NO2; R2, R3=H, Alk(C1±C4); R2±R3=(CH2)n, n=2±4; R4=H, Me, Et; R5=OPh, OBn, NO2; R6=H, Alk(C1±C6) Alk, Hal; X, Y=N, CHPyridopyridines The 1,6-naphthyridine fragment is present in the cephem molecule (212), which is active with respect to both gram-positive and gram-negative bacteria.135 OMe N HN S N Me + O N S N H2N CH2 O COO7 N 212 Recently, 1,4-dihydro-4-oxo-1,6-naphthyridine derivatives have also been covered by patents as antibacterial agents.136 1,4-Dihydro-5-isopropoxy-2-methyl-4-(2-trifluoromethylphen- yl)-1,6-naphthyridine-3-carboxylic esters are efficient compo- nents of cytokinin inhibitors.137 1,6-Naphthyridine-2-carboxyl- amides were covered by patents as drugs used in the therapy and prophylaxis of cytomegalovirus infection.138 1,2,3,4,5,6,7,8-Octa- hydro-1,6-naphthyridine derivatives were covered by patents as anticonvulsive agents 139 and tachykinin NR3-receptor antago- nists.140 Oxadiazolyl-substituted 1,6-naphthyridin-2-ones were covered by patents as benzodiazepine receptor antagonists.141 IV.1,7-Naphthyridines Among procedures for the synthesis of 1,7-naphthyridines, four- component condensation of 1-benzyl-3-hydroxy-5-piperidone (213) with 2,3-dichloro-6-fluorobenzaldehyde, acetoacetic esters and ammonium acetate giving rise to 1,4,5,6,7,8-hexahydro-1,7- naphthyridin-5-one-3-carboxylic esters 214 attracts attention due to its simplicity. The compounds 214 and their N-debenzylation products 215 exhibit hypotensive activity.142, 143 Cl O CO2(CH2)2ORNH4OAc, MeOH CH2 + + Cl F D, 6 h BnN CHO O Me OH 213 Cl Cl O F CO2(CH2)2OR H2, Pd/C MeOH, H2O, HCl BnN Me HN 214 Cl Cl F CO2(CH2)2OR HN Me HN 215 R=Ph, 2,3-Cl2C6H3.3-Aroylpyridine-2-carboxylic acids 216 were used as the starting compounds for the preparation of 7-methyl-1,7-naph- thyridin-8(7H)-one derivatives 217, which are new strong neuro- kinin NK1-receptor antagonists.144 CH2CO2Bn C(O)NMe N COOH N c a, b COC6H4R1 COC6H4R1 216 311 O O N N a, d, e, f NMe NMe d COOH R2 C6H4R1 C6H4R1 O N NMe R3 C(O)NCH2 C6H4R1 Me217 R1=H, 4-Me, 4-F; R2=CO2CH2Ph, CN, CHO, CONH2; R3=3,5-(CF3)2, 2-MeO, 2-Cl, 2,5-Cl2; (a) SOCl2, DMF, THF; (b) BnCO2CH2NHMe. HCl, Et3N, CH2Cl2; (c) 1,8-diazabicyclo[5.4.0]undec-7-ene, PhMe; (d) H2, Pd/C, MeOH; (e) R3C6H4CH2NH2, Et3N, CH2Cl2; ( f ) MeI, NaH, DMF. The reactions of N-substituted pyridine-2,3-dicarboxylimides 218 with sodium alkoxides afforded a mixture of 5-hydroxy-1,7- naphthyridin-8(7H)-ones 219 and 8-hydroxy-1,6-naphthyridin- 5(6H)-ones 220.145 O OH O R1 R2ONa NH + NCH2R1 NH R1 N N N O 220 OH 218 O 219 R1=CO2Me, CO2Et, CO2Pri, COPh; R2=Me, Et.Treatment of 1,8-bis(bromomethyl)isoquinolinium bromide (221) prepared from 1,8-dimethylisoquinoline with secondary amines yielded N,N-dialkylbenzo[de] [1,7]naphthyridinium bro- mides 222.146 Br7 NBS, HBr N R2NH + hn NH N MeOH +N Br7 Me Me CH2Br BrH2C R R 221 222 R=Me, Et, Prn, Pri. Condensation of the lithium salt, which was prepared by metallation of Schiff bases 223 with lithium diisopropylamide, with ethyl 3-acetyl-2-halogenonicotinates, 3-cyano-2-halogeno- pyridines or 3-acetyl-2-halogenopyridines 224 gave rise to 6,8- diaryl-1,7-naphthyridines 225.147 R2 C6H4R1-4 a, b 4-R1C6H4CH2N CHC6H4R1-4 N 223 N225 C6H4R1-4 R1=H, Me, OMe; R2=OH, NH2, Me; (a) Pri2NLi; (b) ( N 224), R3=CO2Et, CN, COMe; R4=Cl, Br. R3 R4 The imino derivative of tetrahydro-1,7-naphthyridinedione 226 was synthesised by the reaction of acid chloride 227 with aniline in ether.148 O NPh COCH(Cl) NPh PhNH2 NPh Et2O N COCl N O 226 227V P Litvinov, S V Roman, V D Dyachenko 312 N N3 NN CH2CO2Et EtONa N N CO2Et +O EtOH CH2CO2Et R1 CO2Et 234a,b R2 233 235a,b R1=CN(234a), CO2Et (234b); R2=NH2 (235a), OH (235b).With the aim of searching for new antimalarial drugs among annelated 1,7-naphthyridines, compounds 228 were synthesised.Cyclisation of amino acid 229, which was prepared from 2-iodo- nicotinic acid, under the action of POCl3 in polyphosphoric acid afforded 10-hydroxy-2-methoxybenzo[b ] [1,7]naphthyridine 230. The subsequent reactions with POCl3 and then with aminophe- nols complete the synthesis. The preliminary assay of the bio- logical activities of the compounds 228 demonstrated that these compounds are inferior to the well-known antimalarial agent pyronaridine.149 CO2H CO2H Reduction of 7-nitrobenzo[f ] [1,7]naphthyridine 236 with iron in 80% acetic acid gave rise to amine 237, whose condensation with aromatic aldehydes afforded benzoimidazo[4,3-fg] [1,7]- naphthyridines (238) exhibiting antibacterial and antifungal activ- a b ities.152, 153 I N N NHC6H4OMe-4 R 229 CHO N N Cl OH Fe, AcOH OMe OMe c AcOH d N N N N N 237 236 NH2 NO2 N 230 R1 N OH NH N R2 OMe N N R N 238 228 R1, R2 =H, CH2NEt2 ; R=H, Me, NO2, Cl.(a)H2N OMe; (b) POCl3, PPA; (c) POCl3; R1 (d) H2N OH. R2 The rearrangement of amine of the 1,2-dihydrofuro[3,2-f ] - [1,7]naphthyridine series produced naphthyridine derivative 239 containing the spirocyclopropane fragment.154 The reactions of analogous 1,2-dihydrofuro[2,3-h ] [1,7]naphthyridine derivatives containing the hydroxyalkylamino group with POCl3 afforded imidazo[2,1-f ]- (240a) or pyrimidino[2,1-f ] [1,7]naphthyridinones (240b) bearing the spirocyclopropane fragment.155 O O D, HCl NH N N N O NH2 239 Me O N Me O N POCl3 Treatment of (S)-N-[N-(4-methylpent-3-enyl)pyrrolin-2- ylmethylidene]-o-toluidine (231) with Lewis acids (FeCl3, SnCl4, AlCl3, MeAlCl2, Me2 AlCl, EtAlCl2, Et2 AlCl or BF3 .OEt2) or Brùnsted acids (CF3COOH or TsOH) led to formal Diels ± Alder heterocyclisation giving rise to 1,7-naphthyridines 232a,b in yields of>80%. The isomer ratio depends substantially on the nature of the catalyst used. In the case of monodentate Lewis acids of the MeAlCl2 or EtAlCl2 type, the cis-isomer 232a was obtained as the major product (99%), whereas the reactions involving bidentate Lewis acids of the SnCl4 type or Brùnsted acids afforded predom- inantly the trans-isomer 232b (87% ± 99%).150 N N CHCl3 Cat N (CH2)n NH(CH2)nOH Me2C CH(CH2)2 CH2Cl2, 42 h N 240a,b n = 1 (a), 2 (b).231 NC6H4Me-2 The reactions of 5,8-dihalogeno-1,7-naphthyridines with N N H H H H potassium amide in liquid ammonia were demonstrated to give mixtures of aminohalogeno, monohalogeno and monoamino derivatives of 1,7-naphthyridines.156 + NH NH Me Me H Me H Me Me Me232b 232a The reactivity of 4,6-benzo[h] [1,7]naphthyridine in the Diels ± Alder reactions with maleic acid and dimethyl acetylenedicarbox- ylate was examined. In the former case, the corresponding quaternary salt was obtained, whereas the latter reaction afforded adduct 241.157 6-Ethoxycarbonyl- (242a) and phenacylmethylides (242b) were used as 1,3-dipolar reagents in the reactions with methacrylic acid, methyl methacrylate, butyl vinyl ether, methyl vinyl ketone, maleic anhydride and dimethyl acetylenedicarbox- 1,2,3-Triazolo[1,5-a ] [1,7]naphthyridines 235a,b were syn- thesised by the reactions of diethyl acetone-1,3-dicarboxylate (233) with ortho-substituted azides of the pyridine series 234a,b.151 ylate.157, 158Pyridopyridines N rise to benzo[c] [2,6]naphthyridines 251a ± c.The latter were used in the synthesis of 2,6-naphthyridine derivatives 252 ± 254.173 H CO2Me N + N R N CO2Me MeO2C 7CHR CO2Me 242a,b 241 R=CO2Et (a), PhCO (b). The reactions of 6-acetamido-8-bromo-1,7-naphthyridine (243) with substituted piperazines in the presence ofNaH afforded 6-acetamido-8-(4-R-piperazin-1-yl)-1,7-naphthyridines 244 pos- sessing antiinflammatory, antiarrhythmic, cardiotonic, vasodila- tor, broncholytic, diuretic and antichlolinergic activities.159 ± 162 AcHN N AcHN a N N N N Cl Br 243 NR244 R=H, (CH2)2OH, CHO, Ar; (a)N NR, EtO(CH2)2OH, NaH.252a Hydrochloride of the derivative of 1,7-naphthyridine-3-car- boxylic acid 245, which lowers high blood pressure, was prepared from compound 246 and oxirane 247 upon refluxing in MeOH followed by saturation of the reaction mixture with HCl.163 R = H (a), Cl (b), Me (c). O C6H4Me-2 CH2OAr HCl CO2Me+ O HN 247 Treatment of toluidides of halogen-substituted acetoacetic acids with sodium hydride in dioxane led to intermolecular cyclisation to form completely hydrogenated 2,6-naphthyridine- tetrone 255.174 Me NH 246 O C6H4Me-2 XH2C 2 CO2Me OH .HCl N Me ArOCH2 NH 245 X=Br, I. 1,7-Naphthyridone derivatives 248 were covered by patents as antibacterial agents.164 ± 166 Naphthyridinium derivatives 249 were covered by a patent as antidiabetic drugs.167 R O O HO OHOH R3 OH R2 N+ N N N H11C5 248 R1 R4 C5H11 249 R1=H, Et, Bn; R2=OH, OAlk; R=H, HOCH2CH(OH). R3, R4=Hal, Me2N, OAlk. Indolo[1,7-bc] [2,6]naphthyridine 256 is a strong and selective antagonist of serotonin receptors of the 5-HT2C/2B subtypes possessing relatively weak affinity for 5-HT2A subtype recep- tors.175 ± 177 The compound 256 was synthesised starting from indoline (257) and isonicotinoyl chloride.175 Oxidation of the compound 256 with MnO2 afforded indolonaphthyridine 258.Yet another approach to the compound 256 involves oxidation of the double bond in compound 259 with ozone or osmium tetroxide followed by treatment with methylamine.178 The indo- lonaphthyridines 256 and 258 were covered by patents as drugs for the treatment of appetite disturbance, obsessive states (phobia and depressions) and other diseases.177, 178 The data on different biological activities exhibited by a series of other 1,7-naphthyridine derivatives have also been reported in recent years.157, 168 ± 172 V. 2,6-Naphthyridines 257 Judging from the number of papers published over the last 15 years, the chemistry of 2,6-naphthyridines attracted much less attention than that of 1,5-, 1,6- or 1,7-naphthyridines. Of the synthetic procedures developed over this period, noteworthy is oxidative cyclisation of 4-carbamoyl-3-(2-hydroxyethyl)quino- lines 250a ± c under the action of CrO3 in glacial AcOH giving CONH2 R (CH2)2OH CrO3 AcOH Cl N 250a ± c N Cl R Cl N 252a ± c H2, Pd/C N 72 HCl Cl N MeOH, EtONa NaH HN C6H4Me-4 O O O O C6H4Me-4 N N 4-MeC6H4 O O N a, b + NH COCl 313 O N POCl3 Cl N 251a ± c N N 253 N MeO OMe N 254 255 Me N+ I7 c O N314 Me N d O N CO2Et N f, g O N i, j 256 N 259 (a) Et3N; (b) MeI; (c) NaBH4; (d ) ClCO2Et; (e) hn; ( f ) AlH3 , THF; (g) (7)-di-p-tolyltartaric acid; (h ) MnO2; (i) O3 (OsO4); ( j ) MeNH2.The reactions of 1,6-dialkyl-1,6-diazacyclodeca-3,8-diynes 260 with different organic solvents (cyclohexa-1,4-diene, 9,10- dihydroanthracene or cyclooctane) as scavengers or with such reagents as HCl or MeOH proceed with high regioselectivity according to the AdE2 mechanism to give 1,2,3,5,6,7-hexahydro- 2,6-naphthyridines 261 in quantitative yields.The reaction of the diyne 260 with H2O in the presence of H2SO4 afforded naphthyr- idone 262 in 15% yield.179, 180 C C NR1 R1N C C 260 H2O, H2SO4 260 N R1 R1=Me, Et, Pri; R2=Cl, OMe. Cyclisation of anilide of 3-(2-hydroxyethyl)pyridine-2-car- boxylic acid occurred under the action of diethyl azodicarboxylate and Ph3P. Subsequent treatment of the resulting 2,6-naphthyr- idine 263 with phenacyl bromide and then with sodium borohy- dride afforded octahydro-2,6-naphthyridine 264 covered by a patent as a medicine for prophylaxis and treatment of schizophre- nia and depressions.181 CONHPh N (CH2)OH a N(a) EtO2CN=NCO2Et, Ph3P, THF; (b) BnBr, PhMe; (c) NaBH4, MeOH.It was reported that 2-methoxy-12-methoxycarbonyl-2,6,8,9- tetrahydro-1H-indolo[7a,1a] [2,6]naphthyridine isolated from seeds of Erythrina melanacantha Yarms can be used in the treat- ment of different forms of hypertonia and as an inhibitor of thrombocyte agglutination.182 CO2Et N e O N Me Me N N h N O N 256 258 R2 R1 N HCl (MeOH) N R1 261 O N R1 262 BnN b, c NPh NPh O O 264 263 V P Litvinov, S V Roman, V D Dyachenko VI. 2,7-Naphthyridines Various approaches to the synthesis of 2,7-naphthyridines are available.Thus dianilide of 3-oxoglutaric acid 265 reacted with malononitrile in the presence of sodium acetate to give pyridone derivative 266, whereas the reaction of the compound 265 with ethyl cyanoacetate in pyridine afforded tetrahydropyridinedione derivative 267. The compounds 266 and 267 underwent cylcisa- tion upon heating with triethylamine in DMF to form substituted 2,7-naphthyridine-3,6-diones 268 and 269, respectively.183 CH2CONHPh AcONa O + CH2 (CN)2 AcOH CH2CONHPh 265 CH2CONHPh NH NH2 NC Et3N NPh PhN DMF, D O N H2N O O 268 (60%) Ph 266 (83%) CH2CONHPh NC Et3N Py 265+NCCH2CO2Et DMF, D O O NPh 267 (47%) NH2 O NPh PhN O O 269 (37%) Dieckmann cyclisation of bis(2-methoxycarbonylethyl)alkyl- amines under the action of sodium methoxide gave rise to sodium salts of 4-hydroxy-3-methoxycarbonyltetrahydropyridines 270.Condensation of the latter with cyanoacetamide afforded 4-cyano-1-hydroxy-5,6,7,8-tetrahydro-3-oxo-2,7-naphthyridines whose hydrolysis gave rise to 4-carbamoyl derivatives 271 exhib- iting antiarrhytmic activity.184 ONa CO2Me (CH2)2CO2Me MeONa NCCH2CONH2 RN (CH2)2CO2Me RN 270 CN CONH2 O O concentrated H2SO4 NH RN NH RN OH OH 271 R=Et, Bn. One of procedures for the preparation of alkyl-substituted 2,7- naphthyridines involves acylation of alcohols, ketones or alkenes in the presence of AlCl3 followed by treatment of the reaction mixture with ammonia.However, these reactions are not neces- sarily regioselective. Thus, the reactions of acetyl chloride with tert-butyl alcohol or isobutylene in the presence of AlCl3 followed by treatment with NH4OH gave rise to 1,3,6,8-tetramethyl-2,7- naphthyridine along with other reaction products.185, 186 1-Alkyl-3,6,8-trimethyl-2,7-naphthyridines 273 were regiose- lectively prepared by the reactions of pyran 272 with an excess of fatty acid chlorides in the presence of AlCl3 followed by treatment with liquid ammonia.187Pyridopyridines Ac Me Me N N 1) AlCl3, 35 8C RCOCl + 2) NH3 (liq) Me R Me Me O272 273 R=Me, Et, Prn, Pri. Acylation of ketones 274 or enones 275 in the presence of AlCl3 proceeded nonselectively.Subsequent treatment with liquid ammonia afforded a mixture of 1,3,6,8-tetraalkyl-2,7-naphthyr- idines 276 and 277 along with symmetrically substituted naph- thyridines 278 and picolines.188 Me R1 Me R2 R1 X O 1) R2COCl, AlCl3 274 + N N 2) NH3 (liq) Me R1 R2 R2276 Me O 275 R2 R2 R2 R2 + + N N N N R2 R2 R2 R1 278 277 R1, R2=Me, Et, Pri; X=OH, OMe. Yet another reaction used in the synthesis of 2,7-naphthyr- idine derivatives involves intramolecular cyclisation of enamino nitriles under the action of acid catalysts. Thus compound 279 underwent cyclisation to form 1-hydroxy-2,7-naphthyridine (280), which was used in the synthesis of sedative agent 281.189 NMe2 a c b N N N N N CN Cl OH 280 279 N N SO2 N(CH2)4N N N HN d O 281 N N NH, ButOK, ButOH; (a) 30% HBr, AcOH; (b) POCl3; (c) HN SO2 (d) N(CH2)4Br, K2CO3, PhCl.O Analogous cyclisation of enamino nitrile 282 in polyphos- phoric acid afforded benzo[c] [2,7]naphthyridine 283, which was converted into perlolinium chloride (284) [herbaceous alkaloid perloline (285)].190 The mass spectra of the compound 285 were measured.191 NH NMe2 CN a O b R R N 283 N 282 NH NH c d O O O R N NR O 315 NH NH O e O + Cl7 OH NR 284 NR 285 R=3,4-(MeO)2C6H3; (a) PPA; (b) 3-ClC6H4CO3H; (c) hn, MeOH; (d) [MeO(CH2)2O]2AlH2Na; (e) HCl. Yet another approach to the synthesis of 2,7-naphthyridines involves recyclisation of pyridinothiopyranothiones and pyridi- nopyranones.Thus recyclisation of thiopyranothiones 286 upon heating with morpholine, piperidine or piperazine in anhydrous ethanol proceeded through the nucleophilic attack of the amine on the thiocarbonyl group, the cleavage of the C(1)7S bond of the thiopyran ring and the rearrangement to form finally 2,7-naph- thyridine-3(2H)-thiones 287. It was demonstrated that the com- pounds 287 in the crystalline state exist in the equilibrium with the thiol form 288. Their methylation with Me2SO4 or MeI in aqueous or ethanolic solutions of KOH produced thiomethyl derivatives 289 exhibiting antibacterial activity.192 CN R1 HN NH2 D + R1R2N S Y 286 S CN R1 CN S R1 NH2 R1R2N NH OH7 R1R2N S 7H2S N N HS Y Y 287 CN CN R1 R1 SH SMe R1R2N N R1R2N N Me2SO4 (or MeI) KOH N N Y Y 288 289 R1=H, Me; R2=Me, CH2CH=CH2; Y =O, CH2, NH.Condensation of substituted pyrano[4,3-b ]pyridin-5-one 290 with dimethylformamide dimethyl acetal in xylene yielded enam- ine 291, which underwent recyclisation to 2-alkyl-5-cyano-8- phenacyl-2,7-naphthyridin-1(2H)-ones 292 upon treatment with primary amines.193 It was demonstrated that the change in the sequence of the addition of the reagents to the pyranone 292 (first, treatment with benzylamine and then condensation with di- methylformamide dimethyl acetal) resulted in 8-benzoyl-6-ben- zyl-3-cyano-4-(2-dimethylaminovinyl)-2-phenyl-1,6-naphthyridin- 5(6H)-one (293). NMe2 O O Me NC NC Me2NCH(OEt)2 RNH2 O O xylene Py Ph N Ph Ph N Ph 291 290316 CN Ph NR N PhC(O)CH2 O 292 R=Pr, Bn.NC EtOH 290+PhCH2NH2 Ph NMe2 O NC NCH2Ph N Ph COPh 293 The reactions of 3-(phenylcarbamoyl)coumarin (294) with ketones in the presence of ammonium acetate in ethanol at 20 8C or in the absence of a solvent at 170 8C yielded benzo[c] [2,7]- naphthyridines 295.194CONHPh Br O O294 R1 Br NR1 295 R1=H, Me; R2=Me, Et; R1±R2=(CH2)4. Fused tetrahydro-2,7-naphthyridine-1,6-diones (296 ± 298) were also synthesised by reductive cyclisation of difuropyridine 299 and furopyridine 300.195 HCl O NO2 OO O N 299 EtOH, H2O, AcOH O NO2 O Zn CO2Me O EtOH, H2O, O NH4Cl Me N 300 V P Litvinov, S V Roman, V D Dyachenko Annelated 2,7-naphthyridines were prepared according to a number of other approaches.Thus the reaction of hexahydro-3- oxoisoquinoline-4-carboxamide (301) with dimethylformamide diethyl acetal afforded a mixture of 4,5-trimethyl-1,2,7,8-tetrahy- dro-2,7-naphthyridine-1,8-dione (302) with another O-alkylation product 303.196 CONH2 O O Me +Me2NCH(OEt)2 NH NHCH2Ph Me2NCH(OEt)2 PhH 301 N CH2C(O)Ph CONH2 OEt + N NH HN 303 (10%) O O 302 (69%) 3-(2,3-Dihydrothiazol-2-ylidene)tetrahydropyridinedinone (304) underwent cyclocondensation upon treatment with benzal- dehyde to form thiazolo[2,3-a] [2,7]naphthyridine (305).197 O O Ar NC Ar NC N N PhCHO AcONH4 +R1CH2COR2 O O Me S Ph N S HN R2 PhNHN PhNHN NPh CN CN 305 304 O O Enamine 330 underwent cyclisation to give 2,7-naphthyridine 307 under the action of ammonium chloride.The compound 307 was converted into benzo[ f ]pyrido[20,30:3,4]pyrrolo[2,1-a]- [2,7]naphthyridine 308 on treatment with H2SO4.198 Me2N O O O O O NH H2SO4 NH4Cl N N N N Cl Cl 307 306 N NH O N O O N CH2OH N 296 O(CH2)2OH 308 NH O Zn O Photocyclisation (irradiation with a low-pressure mercury lamp, the power was 10 W, MeOH, N2 atmosphere, 5 h) of 4-methylidene-3-nicotinoyltetrahydro-1,3-benzothiazine 309 gave rise to a mixture of isomeric 6H,8H-naphthyridino-1,3- benzothiazin-8-ones 310 and 311.187 OMe Me OMe N 297 hn H2C OH N S N N O POCl3 297 O 309 O CHCl3 Me N 298317 Pyridopyridines OMe OMe NR NR OMe OMe Me Me EtNCO N Py N N Y + NH2 O NH S N S N N 320 Y321 311 310 O O R=Me, Ac; Y =O, NMe.An ingenious method for the preparation of annelated tetra- methylthio-substituted 2,7-naphthyridine 322 is based on cyclisa- tion of 2,6-bis[bis(methylthio)methylidene]cyclohexylidene- malononitrile (323) under the action of gaseous HBr.202 Intramolecular cyclodehydrobromination of substituted N-[(4-bromo-3-pyridyl)methyl]anilines 312 followed by dehydro- genation of intermediates 313 afforded benzonaphthyridines 314.199 Br R3 SMe MeS SMe MeS R2 R4 HBr HN N R1 Pri2NLi MnO2 MeS SMe N N THF CHCl3 CN NC R1 R2 R4 323 SMe MeS N NH R3 312 322 313 R3 R2 R4 Dehydrogenation of 5,6-dihydro derivatives of 2,7-naphthyr- idines 324, which were prepared by the addition of organolithium compounds to 4-substituted benzo[c] [2,7]naphthyridines (325), afforded naphthyridines 326.203 R1 N N N N 314 THF MnO2 R1 R1 R1, R2, R3, R4=H, Me, OMe.+ R2Li R2 NH PhMe or CHCl3 N325 324 N R1 R2 N Treatment of isoquinoline derivative 315 with ethyl 4-chloro- methylnicotinate 316 in the presence of ButOK gave rise to isoquinoline derivative 317, which underwent cyclisation to form isoquinolino[2,1-b ] [2,7]naphthyridin-8-one (318).200 The latter was not reduced with NaBH4 or NaBH3CN in AcOH or Ac2O, though this compound in AcOH was converted into derivatives 319 upon hydrogenation with hydrogen over Pt.326 Cl R1=F, Cl, CONPri2; R2=Bu, Ar, Het. CO2Et ButOK + N THF N COPh 316 H NC 315 Substituted tetrahydrothieno[2,3-c] [2,7]naphthyridines 327 were prepared by the reaction of naphthyridinethione 328 with derivatives of chloroacetic acid in ethanol in the presence of sodium ethoxide (the Thorpe ± Ziegler reaction).Their neurotro- phic properties were investigated.204 Me Me CN O N N Me EtO2C COPh S H2, Pt MeN EtONa CN NH2 + RCH2Cl AcOH MeMeN NH EtOH N N 317 318 Mor N R Mor S 328 327 O N Mor= N O; R=CN, CONH2 . NR 319 R=H, Ac. UV irradiation of N-aroyleneamide 329 in methanol or aqueous acetonitrile afforded a mixture of regioisomeric spirocy- cloadducts 330 and 331. The solvent was not involved in the reaction performed in cyclohexane, and R=H in the compounds 330 and 331.205 Refluxing of substituted heterylamines 320 with a threefold excess of ethyl isocyanate in pyridine gave rise to partially hydro- genated fused 2,7-naphthyridines 321.201 R N R CH + N NBn N NBn NBn hn MeOH (MeCN), H2O O O O 330 329 331 R=OMe, OH.318 In a number of studies, the structural 2,7-naphthyridine frag- ment was found to be a component of alkaloids alangimaridine (332),206 isoalamarine (333),207 alamaridine (334) and epi-alamar- MeO O N HO N 332 CH2 R1 Me N R2 H N 334 O N HN N 336 O N HN N N R1 R2 NH338 OAc OAc RO O O 340 O N N R1 N OH N 343 R1=O, S, NOAlk(Ar), R2=Alk N R1N HN 345 R2 R1, R2=O, S, NOAlk(Ar). O N NH MeO N Me OH 347 HO O N MeO N HO Me 333 O N NH N 335 Me N H H NH N 337HN R N O N O 339N N N R3 SR1 342 341 R4 R2 N N 344 OR2 R1 N R2 N R3 O 346 Cl N EtO2C Cl Me N 348 NH 349 V P Litvinov, S V Roman, V D Dyachenko idine,208 nauclefine (335),209, 210 naulafine (336),211 normalindine (337),212 eudistones 338,213 cystodytins 339 ± 341,214 kuanoni- amines 342,215 meridine (343) possessing fungistatic activity,216 alkaloids 344 and 345 exhibiting high antifungal and anticancer activities 217 and other biological compounds, such as sampangine (346) possessing antibacterial activity,218, 219 the biosynthetic precursor of camptothecin 347 exhibiting anticancer activity,220 an analogue of olivecin 348 221 and benzo derivatives 349 suitable for the treatment of asthma and lowering of blood pressure.222 The data on antispasmodic activity of 2,7-naphthyridine deriva- tives were also reported.223* * * Analysis of the data on the synthesis, properties and biological activities of five isomeric pyridopyridines (1,5-, 1,6-, 1,7-, 2,6- and 2,7-naphthyridines) published over the last 15 years is indicative of a lively interest in different aspects of the chemistry of these heterocyclic systems.This interest is generally associated with the practical utility of their derivatives possessing a very broad spectrum of biological activities. This review has been written with the financial support of the Russian Foundation for Basic Research (Project No.99-03- 32965). References NHO Me 1. V P Litvinov, S V Roman, V D Dyachenko Usp. Khim. 69 218 (2000) [Russ. Chem. Rev. 69 201 (2000)] 2. A Couture, E Deniau, P Grandclaudon, C Simion Synthesis 1227 (1993) 3. Fr. Appl. 2 548 667; Chem. Abstr. 103 71 302 (1985) 4. Eur. P. 130 878; Chem. Abstr. 102 220 856 (1985) 5. H Irikawa, Y Okumara Bull. Chem. Soc. Jpn. 60 3797 (1987) 6. Eur. P. 327 426; Chem. Abstr. 114 23 970 (1991) 7. G Viti, D Giannotti, R Nannicini, G Balacco, V Pestellini Tetrahedron Lett. 35 5939 (1994) 8. US P. 4 652 567 (1987) 9. US P. 4 742 061 (1988) 10. US P. 4 7421 71 (1986) 11. WO PCT 9 818 795; Chem. Abstr. 128 321 632 (1998) 12.J Malm, A B HoÈ rfeldt, S Gronowitz Heterocycles 37 331 (1994) 13. G B Barlin, S J Ireland Aust. J. Chem. 41 1727 (1988) 14. G B Barlin, C Jiravinyu Aust. J. Chem. 43 1175 (1990) 15. G B Barlin, C Jiravinyu Aust. J. Chem. 44 151 (1991) 16. G B Barlin, W-L Tan Aust. J. Chem. 38 905 (1985) 17. G B Barlin, W-L Tan Aust. J. Chem. 39 51 (1986) 18. G B Barlin, S J Ireland, T M T Nguyen, B Kotecka, K H Rieckmann Aust. J. Chem. 47 1143 (1994) 19. G B Barlin, S J Ireland, Ch Jiravinyu, Yan Jing-He Aust. J. Chem. 44 677 (1991) 20. G B Barlin, W-L Tan Aust. J. Chem. 38 459 (1985) 21. G B Barlin, W-L Tan Aust. J. Chem. 37 2469 (1984) 22. G B Barlin, S J Ireland, C Jiravinyu, T M T Nguyen, B Kotecka, K H Rieckmann Aust. J. Chem. 46 1695 (1993) 23.US P. 5 240 916; Ref. Zh. Khim. 13O85P (1995) 24. US P. 5 110 347; Ref. Zh. Khim. 3O274P (1994) 25. M Wozniak,M Tomula Liebigs Ann. Chem. 471 (1993) 26. R N Castle, S Pakray, G E Martin J. Heterocycl. Chem. 28 1997 (1991) 27. A Achenbach, B DuÈ thorn, R Waibel Liebigs Ann. Chem. 1159 (1992) 28. FR. P. 2 590 990; Ref. Zh. Khim. 15O118P (1988) 29. A A Adesomoju,M V Lakshmikantham,M P Cava Heterocycles 32 1461 (1991) 30. Czech P. 216 622; Ref. Zh. Khim. 8O156P (1985) 31. Swiss P. 646 166; Ref. Zh. Khim. 14O167P (1985) 32. Fr. P. 2 562 894; Ref. Zh. Khim. 15O146P (1986) 33. BRD P. 3 409 185; Ref. Zh. Khim. 19O102P (1986) 34. Jpn. P. 60-248 688; Ref. Zh. Khim. 23O166P (1986) 35. L Czibula, A Nemes, G Visky,M Farkas, Z Szombathelyi, E Karpati, P Sohar, M Kessel, J Kreidl Liebigs Ann. Chem.221 (1993)Pyridopyridines 36. BRD P. 279 887; Ref. Zh. Khim. 8O135P (1991) 37. Jpn. P. 59-93 080; Ref. Zh. Khim. 8O136P (1985) 38. WO PCT 99 29318; Chem. Abstr. 131 44 810 (1999) 39. BRD P. 4 105 386; Chem. Abstr. 118 82 835 (1993) 40. W Sliwa, B Minanowska Pr. Nauk.-Wyzsza Szk. Pedagog. Czestochowa, Chem. 1 177 (1997); Chem. Abstr. 129 175 215 (1998) 41. US P. 4 517 190 (1983) 42. US P. 4 532 247 (1983) 43. US P. 4 567 186 (1985) 44. US P. 4 559 347 (1984) 45. US P. 4 604 399 (1984) 46. US P. 4 650 806 (1985) 47. B Singh, G Y Lesher J. Heterocycl. Chem. 27 2085 (1990) 48. B Singh Synthesis 279 (1992) 49. US P. 4 634 772 (1986) 50. US P. 4 657 915 (1986) 51.US P. 4 697 021 (1986) 52. US P. 4 716 170 (1986) 53. M Balogh, I Hermecz, G Naray-Szabo, K Simon, Z Meszaros J. Chem. Soc., Perkin Trans. 1 753 (1986) 54. M Balogh, I Hermecz, K Simon, L Pusztay J. Heterocycl. Chem. 26 1755 (1989) 55. G Toth, A Kovacs,M Balogh, I Hermecz J. Heterocycl. Chem. 28 497 (1991) 56. J Kleinschroth, K Mannhardt, J Hartenstein, G Satzinger Synthesis 859 (1986) 57. BRD P. 3 327 650; Chem. Abstr. 21O167P (1985) 58. BRD P. 3 431 303; Ref. Zh. Khim. 23O131P (1986) 59. BRD P. 3 502 790; Ref. Zh. Khim. 7O112P (1987) 60. BRD P. 3 605 743; Ref. Zh. Khim. 6O93P (1988) 61. BRD P. 3 602 655; Ref. Zh. Khim. 10O97P (1988) 62. Jpn. P. 477 488; Ref. Zh. Khim. 19O40P (1995) 63. BRD P. 3 609 785; Ref. Zh. Khim. 23O60P (1988) 64.B Singh, E R Bakon, G Y Lesher, Sh Robinson, P O Pennock, D C Bode, E D Pagani, R G Bentley,M J Connel, L T Hamel, P J Silver J. Med. Chem. 38 2546 (1995) 65. I C Ivanov, E V Stoyanov, P S Denkova, V S Dimitrov Liebigs Ann. Chem. 1777 (1997) 66. N Z Tugusheva, L V Ershov, V G Granik, G Ya Shvarts, R D Syubaev,M D Mashkovskii Khim.-Farm. Zh. 20 830 (1986) a 67. N Z Yalysheva, N P Solov'eva, V V Chistyakov, Yu N Sheinker, V G Granik Khim. Geterotsikl. Soedin. 1118 (1986) b 68. A W Erian, S M Sherif, A-Z A Alassar, Y M Elkholy Tetrahedron 50 1877 (1994) 69. K-H Nietsch, R TroschuÈ tz, H J Roth Arch. Pharm. 318 175 (1985) 70. US P. 4 748 246 SShA; Chem. Abstr. 109 110 425 (1988) 71. US P. 4 751 305 (1987) 72. US P. 4 808 612; Chem. Abstr.110 231 609 (1989) 73. A Shiozawa, Y-i Ichikawa, C Komuro, S Kurashige, H Miyazaki, H Yamanaka, T Sakamoto Chem. Pharm. Bull. 32 2522 (1984) 74. J B Sainani, A C Shah, V P Arya Indian J. Chem. B34 17 (1995) 75. BRD P. 3 502 831; Ref. Zh. Khim. 7O113P (1987) 76. C Rivalle, C Huel, E Bisagni J. Heterocycl. Chem. 26 577 (1989) 77. US P. 5 391 554 (1992) 78. V I Sigova,M E Konshin Zh. Obshch. Khim. 54 2083 (1984) c 79. S V Ukhov,M E Konshin Khim. Geterotsikl. Soedin. 238 (1989) b 80. S V Ukhov,M E Konshin Khim. Geterotsikl. Soedin. 92 (1992) b 81. A Nemazany, N Haider J. Heterocycl. Chem. 34 397 (1997) 82. A Nemazany, N Haider Electron. Conf. Heterocycl. Chem. 1996; Chem. Abstr. 127 205 495 (1997) 83. RDClark,DB Repke,AT Kilpatrick, CMBrown,ACMacKinnon, R U Clauge,M Spedding J.Med. Chem. 32 2034 (1989) 84. R D Clark, D B Repke, J Berger, J T Nelson, A T Klipatrick, C M Brown, A C MacKinnon, R U Clauge,M Spedding J. Med. Chem. 34 705 (1991) 85. US P. 4 791 108 (1987) 86. Eur. P. 288 196; Chem. Abstr. 110 173 210 (1989) 87. US P. 4 956 365 (1988) 88. US P. 4 960 891 (1989) 89. US P. 5 229 387 (1991) 90. US P. 5 231 181 (1991) 91. J C Rohloff, N H Dyson, J O Gardner, T V Alfredson, M L Sparacino, J Robinson III J. Org. Chem. 58 1935 (1993) 319 92. I Hutton (Ed.) Pharmacoproject Vol. 13 (Richmond: PJB Publications, 1992) p. 648a 93. R D Clark, A D Michel, R L Whiting, in Progress in Medical Chemistry Vol. 23 (EdsG P Ellis,G B West) (Amsterdam: Elsevier, 1986) p.1 94. P Balczewski, M K J Mallon, J D Street, J A Joule J. Chem. Soc., Perkin Trans. 1 3193 (1990) 95. Jpn. P. 59-48 482; Ref. Zh. Khim. 1O207P (1986) 96. N Nishiwaki,M Komatsu, Y Ohshiro Synthesis 41 (1991) 97. V K Jahalani, L P Ghalsasi, S P Acharya, R N Usgaonkar Indian J. Chem. B28 173 (1989) 98. P J Victory, J Teixido, J I Borrell, N Busquests Heterocycles 36 1 (1993) 99. J N Reed, J Rotchford, D Strickland Tetrahedron Lett. 29 5725 (1988) 100. P-L Compagnon, F Gasquez, T Kimny Bull. Soc. Chim. Belg. 95 49 (1986) 101. V N Nesterov, V D Dyachenko, Yu A Sharanin, Yu T Struchkov Izv. Akad. Nauk, Ser. Khim. 437 (1996) d 102. V D Dyachenko, S V Roman, V P Litvinov Izv. Akad. Nauk, Ser. Khim. 121 (2000) d 103. S V Roman, V D Dyachenko, V P Litvinov, in Organicheskii Sintez i Kombinatornaya Khimiya (Tez.Dokl. Mezhdunar. Konf.), Zvenigorod, 1999 [Organic Synthesis and Combinatorial Chemistry (Abstracts of Reports of the International Conference), Zvenigorod, 1999] P-58 104. S V Roman, V D Dyachenko, V P Litvinov Khim. Geterotsikl. Soedin. 1435 (1999) b 105. L S Vasil'ev, F E Surzhikov, O G Azarevich, V S Bogdanov, V A Dorokhov Izv. Akad. Nauk, Ser. Khim. 1510 (1994) d 106. R M Mohareb, S M Fahmy Z. Naturforsch., B Chem. Sci. 40 1537 (1985) 107. J M Robinson, L W Brent, C Chau, K A Floyd, S L Gilham, T L McMahan, D J Magda, T J Motycka,M J Pack, A L Roberts, L A Seally, S L Simpson, R R Smith, K N Zalesny J. Org. Chem. 57 7352 (1992) 108. I A El-Saaka J.Chem. Res. (S) 434 (1996) 109. A T M Marcelis, H C van der Plas Tetrahedron 45 2693 (1989) 110. E L Gaidarova, A A Borisenko, T I Chumakov, A V Mel'nikov, I S Orlov, G V Grishina Tetrahedron Lett. 39 7767 (1998) 111. T Siener, U Mueller, M Jansen, U Holzgrabe Pharmazie 53 442 (1998) 112. R N Radinov, S D Simova,M A Haimova Izv. Khim. Bulg. Akad. Nauk 22 144 (1989); Ref. Zh. Khim. 24 Zh 432 (1989) 113. R Friary, J H Schwerdt, V Seidl, F J Villani J. Heterocycl. Chem. 27 189 (1990) 114. US P. 5 070 086 (1990) 115. K R Prasad, M Darbarwar Synth. Commun. 22 2479 (1992) 116. K R Prasad, M Darbarwar Org. Prep. Proced. Int. 27 547 (1995) 117. C W G Fishwick, R C Storr, P W Manley J. Chem. Soc., Chem. Commun. 1304 (1984) 118. J I Levin Tetrahedron Lett.30 2355 (1989) 119. Y Miki, O Tomii, M Utsunomiya, S Takemura,M Ikeda Chem. Pharm. Bull. 34 3588 (1986) 120. K GoÈ rlitzer, C Heinrici Arch. Pharm. 321 477 (1988) 121. Jpn. P. 491 090; Ref. Zh. Khim. 2O91P (1995) 122. S M Hussain, A M El-Reedy, T M Abdin Gazz. Chim. Ital. 118 623 (1988) 123. V Ojea, I Muinelo, M C Figueroa,M Ruiz, J M Quintela Synlett 622 (1995) 124. J Svetlik, F Turecek, V Hanus J. Chem. Soc., Perkin Trans. 1 2053 (1988) 125. T Sakamoto, N Miura, Y Kondo, H Yamanaka Chem. Pharm. Bull. 34 2018 (1986) 126. N I Smetskaya, A M Zhidkova, N A Mukhina, V G Granik Khim. Geterotsikl. Soedin. 1287 (1984) b 127. US P. 5 367 078 (1993) 128. US P. 5 424 433 (1994) 129. US P. 5 424 434 (1994) 130. Jpn. P. 58-7626; Ref.Zh. Khim. 2O123P (1985) 131. S Mishio, T Hirose, A Minamida, J-i Matsumoto, S Minami Chem. Pharm. Bull. 33 4402 (1985) 132. US P. 5 180 719 (1991) 133. US P. 5 281 703 (1993)320 134. WO PCT 94-12173; Ref. Zh. Khim. 3O76P (1996) 135. Jpn. P. 59-219 292; Ref. Zh. Khim. 1O184P (1986) 136. A Belicova,M Seman, V Milata, D I Ilavsky, L Ebringer Folia Microbiol. (Prague) 42 193 (1997); Chem. Abstr. 127 63 024 (1997) 137. Jpn. P. 10 182 462; Chem. Abstr. 129 153 228 (1998) 138. WO PCT 97-34 894; Chem. Abstr. 127 318 884 (1997) 139. WO PCT 98-54 184; Chem. Abstr. 130 38 370 (1999) 140. WO PCT 99-00 388; Chem. Abstr. 130 110 251 (1999) 141. WO PCT 99-03 857; Chem. Abstr. 130 125 079 (1999) 142. US P. 4 596 873 (1985) 143. US P. 4 618 678 (1985) 144.H Natsugari, Y Ikeura, Y Kiyota, Y Ishichi, T Ishimaru, O Saga, H Shirafuji, T Tanaka, I Kamo, T Doi,M Otsuka J. Med. Chem. 38 3106 (1995) 145. M Blanco,M G Lorenzo, I Perillo, C B Schapira J. Heterocycl. Chem. 33 361 (1996) 146. A Niimi, Y Nagao, Y Abe, T Misono Chem. Express 6 45 (1991) 147. A Couture, P Grandclaudon, C Simion, P Woisel J. Chem. Soc., Perkin Trans. 1 2643 (1995) 148. L Capuano,W Hell, C Wamprecht Liebigs Ann. Chem. 132 (1986) 149. P-E Zhu, Ch Chen Pharm. Ind. 19 55 (1989); Ref. Zh. Khim. 5O120 (1990) 150. F Linkert, S Laschat,M Knickmeier Liebigs Ann. Chem. 985 (1995) 151. R K Smalley,M Teguiche Synthesis 654 (1990) 152. M Paluszewski, W Sliwa Aust. J. Chem. 46 1115 (1993) 153. Pol. P. 167 956; Ref.Zh. Khim. 20O46P (1996) 154. T Hirota, T Matsushita,K Sasaki, S Kashino Heterocycles 41 2565 (1995) 155. K Sasaki, A S S Rouf, T Hirota J. Heterocycl. Chem. 33 49 (1996) 156. M Wozniak Pol. J. Chem. 62 305 (1988) 157. B Bachowska, T Zujewska Pol. J. Chem. 70 1324 (1996) 158. G Matusiak Aust. J. Chem. 52 149 (1999) 159. Jpn. P. 62-30 780; Ref. Zh. Khim. 2O83P (1988) 160. US P. 4 690 924 (1986) 161. Jpn. P. 63-48 277; Ref. Zh. Khim. 15O88P (1989) 162. US P. 4 866 176; Ref. Zh. Khim. 17O69P (1990) 163. US P. 4 551 534 (1984) 164. S Radl, P Hradil Collect. CÏzech. Chem. Commun. 56 2420 (1991) 165. P Hradil Cesk. Farm. 41 55 (1992) 166. P Hradil Cesk. Farm. 41 194 (1992) 167. Eur. P. 563 797; Chem. Abstr. 120 106 978 (1994) 168.WO PCT 97-03 074; Chem. Abstr. 126 199 561 (1997) 169. WO PCT 97-48 368; Chem. Abstr. 128 88 786 (1998) 170. WO PCT 98-18 796; Chem. Abstr. 128 321 633 (1998) 171. Jpn. P. 10 109 989; Chem. Abstr. 129 16 149 (1998) 172. Y Ikeura, T Ishimaru, T Doi,M Kawada, A Fujishima, H Natsugari Chem. Commun. 2141 (1998) 173. P Rajamanickam, P Shanmugam Synthesis 541 (1985) 174. N S Ibrahim, K U Sadek, S I Aziz, M H Elnagdi Z. Naturforsch., B Chem. Sci. 40 129 (1985) 175. J Nozulak, H O Kalkam, P Floersheim, D Hoyer, P Scoeffter, H R Buerki J. Med. Chem. 38 28 (1995) 176. H MuÈ hle, J Nozulak, J Cercus, E KuÈ sters, U Buetler, G Penn, W Zaugg Chimia 50 209 (1996) 177. BRD P. 473 550; Chem. Abstr. 117 111 583 (1992) 178. BRD P. 4 027 018; Ref. Zh.Khim. 2O76P (1993) 179. R Gleiter, J Ritter Angew. Chem., Int. Ed. Engl. 33 2470 (1994) 180. J Ritter, R Gleiter Liebigs Ann. Chem. 1179 (1997) 181. Br. P. 2 299 582; Ref. Zh. Khim. 12O64P (1997) 182. BRD P. 3 336 194; Ref. Zh. Khim. 10O13P (1986) 183. A M El-Reedy,M K A Ibrahim, S M Hussain, H H Moharram J. Prakt. Chem. 331 745 (1989) 184. E G Paronikyan, S N Sirakanyan, A S Noravyan, T O Asatryan, K Zh Markaryan, R A Aleksanyan Khim.-Farm. Zh. 30 (6) 13 (1996) a 185. C H Erre, C M A Roussel Bull. Soc. Chim. Fr., Pt. 2 449 (1984) 186. F A Selimov, O G Rutman, U M Dzhemilev Izv. Akad. Nauk SSSR, Ser. Khim. 688 (1986) d 187. C Roussel, A Merier,M Gartier J. Org. Chem. 52 2935 (1987) 188. C H Erre, C M A Roussel Bull. Soc. Chim. Fr., Pt. 2 454 (1984) 189. US P. 4 859 671 (1988) 190. A B Ridley, W C Taylor Aust. J. Chem. 40 631 (1987) 191. G Dannhardt, K K Mayer, L Steindl Chem. Ztg. 109 139 (1985) V P Litvinov, S V Roman, V D Dyachenko 192. E G Paronikyan, G V Mirzoyan, A S Noravyan, D A Avakimyan, Yu Z Ter-Zakharyan Khim.-Farm. Zh. 27 (11) 29 (1993) a 193. V A Dorokhov, A Yu Yagodkin, V S Bogdanov, S V Baranin Izv. Akad. Nauk, Ser. Khim. 1079 (1997) d 194. M Y El-Kady, G H Sayed, A M El-Gendy, M El-Sherif Egypt. J. Chem. 28 63 (1985) 195. K GoÈ rlitzer, U Bartke, E Schmidt Arch. Pharm. 324 105 (1991) 196. L T Guss, L V Ershov, V G Granik Khim. Geterotsikl. Soedin. 1696 (1987) b 197. R M Mohareb, H F Zohdi, S M Sherif, W W Wardkhan Tetrahedron 50 5807 (1994) 198. F Bracher J. Heterocycl. Chem. 30 157 (1993) 199. C F Nutailis, S R Marsh J. Heterocycl. Chem. 29 971 (1992) 200. E Reimann, H Bened Arch. Pharm. 327 539 (1994) 201. G Winters, A Sala, A de Paoli, V Ferri Synthesis 1052 (1984) 202. GDR P. 279246; Ref. Zh. Khim. 18O75P (1991) 203. C Cochennec, P Rocca, F Marsais, A Godart, G Queguiner J. Chem. Soc., Perkin Trans. 1 979 (1995) 204. E G Paronikyan, G V Mirzoyan, A S Noravyan, E M Arzanunts, R S Sukasyan, I S Sarkisyan, IMNazaryan, I A Dzhagatspanyan Khim.-Farm. Zh. 31 (10) 34 (1997) a 205. J-C Gramain,Y Troin,D Vallee-Goyet Tetrahedron 47 7301 (1991) 206. Jahangir,M A Brook, D B MacLean, H L Holland Can. J. Chem. 65 2362 (1987) 207. U S Chowdhury Tetrahedron 46 7893 (1990) 208. E Reimann, H Renz Monatsh. Chem. 125 1397 (1994) 209. Jahangir, M A Brook, D B MacLean, H L Holland Tetrahedron 43 5761 (1987) 210. R Vohra, D B MacLean Heterocycles 39 445 (1994) 211. D B Repke Jr,M A Brook, R D Clark, J T Nelson Heterocycles 27 2289 (1988) 212. A W Rey,W A Szarek,D B MacLean Heterocycles 32 1143 (1991) 213. H-Y He, D J Faulkner J. Org. Chem. 56 5369 (1991) 214. M A Ciufolini, N E Byrne J. Am. Chem. Soc. 113 8016 (1991) 215. G P Gunawardana, F E Koehn, A Y Lee, J Clardy, H-Y He, D J Faulkner J. Org. Chem. 57 1523 (1992) 216. P J McCarthy, T P Pitts, G P Gunawardana,M Kelly-Borges, S A Pomponi J. Natur. Prod. 55 1664 (1992); Ref. Zh. Khim. 13 E 80 (1993) 217. US P. 5 182 287 (1991) 218. US P. 5 128 344 (1990) 219. US P. 5 227 383 (1991) 220. B K Carte, C DeBrosse, D Eggleston,M Hemling, M Mentzer, B Poehland, N Troupe, J W Westley, S M Hecht Tetrahedron 46 2747 (1990) 221. L Mathew J. Indian Inst. Sci. 72 345 (1992); Ref. Zh. Khim. 19 Zh 209 (1993) 222. US P. 4 478 834; Ref. Zh. Khim. 13O168P (1985) 223. E G Paronikyan, S N Sirakanyan, R G Paronikyan, A S Noravyan, in Khimiya Fiziologicheski Aktivnykh Soedinenii (Tez. Dokl. Vsesoyuz. Konf.), Chernogolovka, 1989 [The Chemistry of Physiologically Active Compounds (Abstracts of Reports of the All-Union Conference), Chernogolovka, 1989] p. 193 a�Chem. Pharm. J. (Engl. Transl.) b�Chem. Heterocycl. Compd. (Engl. Transl.) c�Russ. J. Gen. Chem. (Engl. Transl.) d�Russ. Chem. Bull. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (4) 321 ± 332 (2001) Mechanism of thermal decomposition of silanes A A Onischuk, V N Panfilov Contents I. Introduction II. Thermal decomposition of monosilane III. Homogeneous pyrolysis of disilane IV. Heterogeneous decomposition of silanes V. Aerosol formation in the pyrolysis of silanes VI. The composition of the solid product VII. Conclusion Abstract. of decomposition thermal of mechanism the on Data Data on the mechanism of thermal decomposition of silanes are generalised. The main stages of the process are silanes are generalised. The main stages of the process are distinguished on concentrated is Attention characterised. and distinguished and characterised. Attention is concentrated on the the elementary from silicon of deposition reactions, homogeneous elementary homogeneous reactions, deposition of silicon from the the gas of models Various aerosols.of formation and phase gas phase and formation of aerosols. Various models of formation formation of and theoretical of results The considered. are aerosols of aerosols are considered. The results of theoretical and exper- exper- imental on Data discussed. are process this of investigations imental investigations of this process are discussed. Data on the the chemical pyrolysis silane of product solid the of composition chemical composition of the solid product of silane pyrolysis (wall (wall deposit that noted is It presented. are particles) aerosol and deposit and aerosol particles) are presented. It is noted that the the hydrogen limits.wide over varies product solid the in content hydrogen content in the solid product varies over wide limits. The The bibliography references 211 includes bibliography includes 211 references. I. Introduction Thermal decomposition (pyrolysis) of silanes is used in micro- electronics to obtain polycrystalline silicon and amorphous silicon layers, which explains the researchers' interest in this process. Dozens of investigations devoted to pyrolysis of silane are published annually, however, it should be emphasised that this process has been studied to a much lesser extent than pyrolysis of their closest structural analogues, viz. alkanes. For example, it is known that the homogeneous pyrolysis (cracking) of alkanes follows a non-branched chain mechanism.Rice's quantitative scheme suggested back in the 1930s (Ref. 1) allows one to calculate the composition of cracking products for any individual hydrocarbon. The results of the calculations obtained by this scheme fit the experimental data. The study of the influence of heterogeneous factors on the pyrolysis of hydrocarbons 2 has led to a concept of the involvement of surface in both the stages of chain initiation and termination and the stages of chain propaga- tion.By now, the mechanism of silane pyrolysis (unlike that of hydrocarbon pyrolysis) has not been studied adequately. Due to A A Onischuk, V N Panfilov Institute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences, ul.Institutskaya 3, 630090 Novosibirsk, Russian Federation. Fax (7-383) 234 23 50. Tel. (7-383) 233 13 22. E-mail: onischuk@ns.kinetics.nsc.ru (A A Onischuk) Tel. (7-383) 233 23 81. E-mail: panfilov@ns.kinetics.nsc.ru (V N Panfilov) Received 10 May 2000 Uspekhi Khimii 70 (4) 368 ± 381 (2001); translated by A V Serov #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n04ABEH000603 321 322 325 326 327 329 330 the lower Si7H and Si7Si bond strengths in silanes in compar- ison to those of the C7C and C7H bonds in hydrocarbons, pyrolysis of silanes occurs at much lower temperatures and high- molecular-mass silane derivatives are not the final products, but rather intermediates. The final products of the thermal decom- position of silanes are gaseous hydrogen and the solid phase, viz., silicon (1) n Si+(n+1)H2 .SinH2n+2 It should be noted that the reaction generally cannot go to completion due to the experimental conditions. So the solid product is usually hydrogenated (containing chemically bound hydrogen) rather than pure silicon. The lower the reaction temper- ature the higher the hydrogen content in the solid product. Besides, under real conditions, stable gaseous substances (Si2H6, Si3H8, Si4H10, etc.) are obtained in addition to hydrogen. Silane pyrolysis is a complex process including numerous homogeneous and heterogeneous elementary steps. At lower pressure (P<1 kPa), heterogeneous stages prevail 3 ±7 and the solid product deposits on the reactor walls, while at high pressures homogeneous elementary reactions dominate and the solid prod- uct is obtained mainly in the form of aerosol particles.8, 9 Usually, the thermal decomposition of silanes is studied in a static reactor (600<T<750 K),10 ± 19 in a flow reactor (800<T< 1000 K) 10, 14, 20 ± 29 and in shock tubes (900<T<2000 K).30 ± 35 The solid product is most often amorphous hydrogenated silicon.The content of hydrogen in the solid product expressed as the [H]/[Si] ratio varies from *2 (Ref. 15) 0.5 (Ref. 27) and 0.15 (Ref. 8) at low temperature and a short reaction period to<0.01 (Ref. 36) at high temperature and a long reaction time. Under definite conditions, the intermediate solid product (hydrogenated silicon) formed in the course of pyrolysis influences the kinetics and mechanism of this process.In particular, such influence results in acceleration of the reaction when silicon with a highly developed surface (owing to the precipitation of aerosol particles) is obtained.37, 38 As the composition of the solid product changes during the reaction, its reactivity changes, i.e., a solid phase of variable composition is formed, which affects the kinetics and mechanism of the process.3 This is the key feature of the pyrolysis of silanes. In the present review, elementary reactions, the processes of silicon precipitation on the reactor walls, the formation of aerosols and elementary reactions occurring in the solid product in the pyrolysis are considered.It should be mentioned that certain aspects of the thermal decomposition of silanes have been322 discussed in detail. Thus homogeneous elementary stages of the process have been described thoroughly in a number of publica- tions.39 ± 44 In the reviews,45, 46 the data on chemical vapour deposition (CVD) are systematised and some peculiarities of this process are explained. The results of studies on aerosol formation in the silane pyrolysis have been analysed.8, 9, 47 In the present review, we avoid repeating the above-mentioned publications; in particular, elementary homogeneous reactions and CVD proc- esses are not considered in detail. The attention is concentrated on those aspects which, in our opinion, either require an additional discussion (for example, the contribution of radical reactions) or provide more complete information on the pyrolysis of silanes.II. Thermal decomposition of monosilane 1. Primary decomposition of the monosilane molecule The reactions of the thermal decomposition of monosilane have been described most thoroughly. There are many viewpoints on whether the process of primary decomposition of SiH4 into SiH2 and H2 under different conditions is homogeneous or heteroge- neous. Thus according to an investigation of the silane pyrolysis in a static reactor (P=5 ± 30 kPa, T=650 ± 700 K) where the initial reaction rate did not depend on the surface to the volume ratio (S/V), the authors came to the conclusion that the primary reaction proceeds homogeneously.15 This process was studied in static and flow reactors 10 (under lower pressure) and a conclusion as to the heterogeneity of the primary pyrolysis reaction in the whole range of pressures from 1 to 104 Pa was made assuming that the rate of the initial decomposition of silane at P=10 Pa is proportional to the S/Vratio.This conclusion 10 was the subject of criticism. In particular, O'Neal and Ring 48 on the basis of results obtained earlier 15 reinterpreted the data given in the paper cited 10. Taking into consideration that the heterogeneous primary decomposition prevails at pressures below 100 Pa, they proved convincingly that under high pressure (>1 kPa) the primary decomposition proceeds as a homogeneous process.Studies on monosilane pyrolysis by laser heating (without influence of the reactor walls) confirm this conclusion. The activation energy of the laser-induced monosilane pyrolysis at 2.7 kPa (Ref. 49) was measured to be 22129 kJ mol71, which corresponds to the results obtained earlier in shock tubes 30 and in a static reactor.15 Investigations into the silane pyrolysis under continuous wave (c.w.) CO2 laser radiation and a total pressure of the mixture equal to 46104 Pa allowed representation of the effective rate constant for the silane pyrolysis as 1011.50.6exp[72198(kJ mol71)/RT] (Ref. 50); the value of the activation energy obtained in this study agrees with those given elsewhere.15, 30, 49 At present, it is believed that at P>1 kPa the primary monosilane decomposition occurs as a homogeneous reaction and at lower pressures it is heterogeneous.5, 51 2.The role of radical reactions in the monosilane pyrolysis A detailed study of the mechanism of the homogeneous mono- silane pyrolysis in a static reactor was first carried out by Purnell and Walsh.15 The kinetics of the pyrolysis was characterised in the initial (conversion below 3%) and middle (conversion of 10%± 20%) stages of the process. In the initial stage, the reaction had the order *3/2 with respect to monosilane, and the stoichi- ometry was described by the equation 2.39 SiH4=Si2H6+0.13 Si3H8+1.26H2 . The total pressure in the reactor did not change with time and the solid product was not deposited.In the intermediate stage, the reaction was first order with respect to monosilane, the total pressure in the reactor increased and the formation of the solid product was observed. Two alternative mechanisms were suggested which explained the experimental result, viz., the molecular mechanism (2) SiH2+H2 , SiH4 A A Onischuk, V N Panfilov (3) Si2H6 , SiH2+SiH4 (4) Si3H8 SiH2+Si2H6 and the radical mechanism .SiH3+H., SiH4 H2+.SiH3 , H.+SiH4 . Si2H6+H., SiH3+SiH4 (5) (6) (7) (8) Si2H6 . 2 .SiH3 It was noted that the reaction order 3/2 can be explained both by the radical and the molecular mechanism (it is suggested that the process in the latter proceeds in the region of the dependence of the rate constant on pressure, viz., in the `fall-off' regime).Later, the calculations of the reaction constant of the unimolecular decomposition of monosilane (reaction 2) by the Rice ± Ramsper- ger ± Kassel ± Marcus (RRKM) method 30 confirmed the reaction order 3/2 observed by Purnell and Walsh. Nevertheless, the discussion on the possible role of radical reactions in the thermal decomposition of monosilane continues to be topical for the last two decades. Cleavage of the Si7H bond rupture in the SiH4 molecule (reaction 5) is the key reaction in the radical mechanism. The Si7H bond energy was measured both by the experimental and theoretical methods. For example, interpreting the experimental data on the kinetics of the reaction between I2 and SiH4 in the gas phase, Doncaster and Walsh 52 estimated the H3Si7H bond energy as 3785.1 kJ mol71.The energy of this bond found in the investigation of the reaction of SiH3 with HI (HBr) was 3842.0 kJ mol71 (298 K),53 and in the experiments on SiH4 photo-ionisation at 150 K it was found to be equal to 3727 kJ mol71 (Ref. 54). It should be mentioned that the values of the H3Si7H bond energy found experimentally agree with the theoretical values. Thus Sax and Kalcher 55 have calculated the enthalpies of for- mation and Gibbs energies of silicon hydrides containing up to three silicon atoms. According to these calculations, the dissoci- ation energy of the H3Si7H bond is 383.7 kJ mol71. The H3Si7H bond energy obtained in the determination of atom- isation energies for a series ofAHn hydrides (whereAis an element of the I7VII groups of the second and third periods of the Periodic Table) is 384 kJ mol71.56 According to the calculation carried out by Ho et al.,57 the value of the energy is 384 kJ mol71, later,58 it was established as 382 kJ mol71.The average value of theH3Si7Hbond energy obtained in the above-mentioned studies is 38113 kJ mol71. This does not agree with the experimental activation energy (233 kJ mol71) of monosilane decomposition in the initial stage of the process.15 (Besides, for the rate of the monosilane chain decay to fit the experimentally observed rates, the chain length should exceed 106, but as it follows from the experimental results obtained in shock tubes,30 the average number of collisions of the monosilane molecule with the molecules of products does not exceed 50 under these conditions.) However, the activation energy value equal to 233 kJ mol71 agrees with the calculated data on the activation energy of the reaction (2).59, 60 ± 64 At present, it is acknowledged that the primary stage of the monosilane decomposition is the dissociation of the SiH4 mole- cule into SiH2 and H2 and not H3Si7H bond cleavage; but the significance of the contribution of the radical reactions to the next pyrolysis stages still remains an open question.Numerous pub- lications deal with the role of SiH3 in the process of the thermal monosilane decomposition. Thus the pyrolysis of an equimolar mixture of SiH4 and SiD4 was studied in flow and static reactors.14 The formation of mixed silanes and disilanes and the differences in the results obtained in static and flow reactors could be rational- ised by the mechanism implying the involvement of SiH3.Thermal decomposition of monosilane in the presence of acetylene in a flow reactor at 690 ± 710 K was studied by HaasMechanism of thermal decomposition of silanes and Ring.24 Based on the analysis of the reaction products, it was suggested that the silyl radicals were the major products of the initial decomposition. In an investigation into the decomposition of monosilane in a single-pulse shock tube at 1200 ± 1300 K, it was concluded 33 that the observed decomposition rate of monosilane cannot be explained by the radical mechanism; therefore, the primary act of the monosilane decomposition is the unimolecular elimination of H2.The rate constants for the H2 elimination from SiH4 were calculated by the RRKM method. Comparison of the calculated and experimental 15 values showed that they differ more than forty-fold, which led to the conclusion that the chain mechanism is predominant under the static reactor conditions. The thermal decomposition of SiH4 in a static reactor in the presence of ethylene was carried out in order to determine whether the primary act of the monosilane decomposition occurs accord- ing to the molecular mechanism (i.e., as the result of H2 elimi- nation from the SiH4 molecule).18 The analysis of the pyrolysis products suggested that the SiH3 radicals formed in the slow heterogeneous process of the Si7H bond cleavage participate, though to a minor extent, in the pyrolysis of silane. This suggestion was confirmed by performing a pyrolysis of a SiH4 ± SiD4 mixture and detecting a rather high content of HD in the hydrogen fraction. The initial acceleration of the silane pyrolysis in static and flow reactors in the range of pressures from 1 to 104 Pa was found,10 which could be explained by the liberation of radicals from the surface into the gas phase or by the hydrogen atom elimination into the gas phase followed by chain reaction [processes (6) and (7)].The gas-phase pyrolysis of monosilane in the mixture of Ar and SF6 under c.w.CO2-laser radiation was investigated.50 The total pressure of the mixture was 46104 Pa, while the partial pressure of monosilane was 102 ± 103 Pa. The influence of propene additives as an inhibitor of radical reactions was studied exper- imentally. It was established that the effective rate constant for the silane pyrolysis decreases on the addition of propene. Presumably, in the presence of an inhibitor the reaction proceeds in accordance with the molecular mechanism and is characterised by an activa- tion energy of 281.5 kJ mol71. In the absence of an inhibitor, the radical stages become sifnificant and the effective activation energy of this process decreases to 217 kJ mol71. A series of experiments on the monosilane decomposition in the presence of alkenes and acetylene in a static reactor was carried out in order to find out whether silyl radicals take part in the pyrolysis of monosilane.11 The analysis of the kinetics of the pyrolysis of the silane ± acetylene mixture gave additional infor- mation on the nature of the monosilane thermal decomposition.According to the first-order kinetics obtained for this mixture, it was determined that silyl radicals do not participate in the decomposition under conditions examined. A possible scheme of the radical mechanism of monosilane decomposition in the presence of alkenes, according to which lower alkanes should be formed, was suggested. The absence of lower alkanes among the experimentally detected products led to the conclusion that the silyl radicals do not contribute considerably to the reaction.The results of the investigation on the pyrolysis of a SiH4 ± SiD4 mixture under static conditions and under reduced pressures, in particular, the kinetic curves of the H2, HD and D2 formation, showed 12 that HDconcentration in the initial stage of the process increased more slowly than in the following stages. The analysis of the formation of isotopic hydrogen molecules shows that the molecular mechanism of the monosilane decomposition is pre- dominant. The sequence of the formation of mixed deuteriosilanes SiH3D SiH4 SiH2D2 , SiHD3 SiD4 was determined as being due to the exchange reactions involving the SiH3 radical, e.g. 323 (9) SiH3D+.SiD3 ..SiH3+SiD4 The concentrations of Si2H6, Si2H2D4 and Si2D6 in the initial stage of the pyrolysis exceed those of other deuteriodisilanes. This fact proves the predominance of the disilane formation upon the reaction SiH2 with SiH4. Thus, the features of formation of isotopic hydrogen molecules and deuteriodisilanes points to the significance of the contribution made by the molecular mecha- nism. In studies on the heterogeneous decomposition of monosilane at 0.5 ± 4 Pa, the liberation of hydrogen atoms and SiH3 radicals into the gas phase was observed.61 The heated surface was placed directly close to the ionisation source in such a way that practically all the radicals evolved from the surface into the gas phase could reach the inlet of the ionisation chamber virtually without collisions with the gas molecules. At T>1800 K, almost all the hydrogen obtained upon the decomposition SiH4 is liberated asH atoms, while a part of hydrogen is liberated as H2 molecules at T<1800 K.The elimination of hydrogen in the temperature range 1300 ± 1800 K is characterised by the activation energy 176 kJ mol71, while at T<1300 K the activation energy is 46 kJ mol71. The effectiveness of the formation of SiH3 radicals in the heterogeneous silane decomposition is 0.014 ± 0.05. The liberation of SiH3 radicals from the surface takes place as the result of the exchange of the hydrogen atom between the SiH4 molecule and the free bond localised on the surface. Direct detection of SiH3 radicals in the thermal decomposition of monosilane in a flow reactor was investigated by the method of laser magnetic resonance.28 A mixture of monosilane and argon (the total pressure was 60 ± 500 Pa, the partial pressure of mono- silane was 2.3 ± 18 Pa) was used.The concentration of SiH3 radicals changed from 1010 to 1012 cm73 as the temperature of the reactor was raised from 830 to 1250 K (10% of SiH4 at the total pressure of 400 Pa). It was suggested that SiH3 radicals are formed in the heterogeneous processes. The possibility of the chemical induction of the monosilane pyrolysis was shown.19 This was carried out in a cylindrical metal reactor. A mixture of SiH4 and O2 with the oxygen percentage below 2% was introduced at room temperature and the reaction was initiated by spark ignition or by chemical ignition.The degree of conversion of monosilane was up to 60%. At definite concen- tration ratio of SiH4 and O2 and pressures, the mixture can ignite spontaneously even at room temperature, which is the chemical method of ignition. If the necessary [SiH4] : [O2] ratio is established in a small volume of a reactor, self-ignition occurs. Then the burning front spreads over the whole volume, which results in the significant burnout of monosilane. The decomposition of silane presumably has branched-chain character under these conditions. Oxygen is required only in the stage of chain initiation. The atoms and radicals formed in this stage take part in the chain reactions of the monosilane decomposition, which results in significant con- sumption of monosilane.The average chain length is*10. Studies of the monosilane pyrolysis under static conditions at P=0.2 ± 5 kPa show that an increase in the specific surface (S/V) at T=669 K gives rise to an increase in the decomposition rate, while at T=779 K it results in its reduction.4 The dependences obtained were interpreted in the terms of the radical chain mechanism and it was suggested that the rate of the heterogeneous decomposition of monosilane increases with the increase in the ratio S/V at low temperatures, while at high temperatures the rate of the heterogeneous chain termination increases. It was supposed that the chain process plays a considerable role in the pyrolysis of monosilane [reactions (6) and (7)].62 The activation energies obtained by ab initio quantum-chemical cal- culations were found to be 36 and 133 kJ mol71 for these reactions, respectively.Hydrogen atom formation during the monosilane pyrolysis in a shock tube was studied.35 The reaction was followed by monitoring light absorption by the hydrogen atoms at the wave- length l=121.6 nm. An Ar/SiH4 mixture was used (the total324 pressure was 1 ± 1.5 atm, the relative concentration of monosilane in the mixture was*1075). The time dependence of the hydrogen atom concentration was obtained. The features of this dependence are as follows: the change in the concentration at low temperatures is S-shaped; the maximum concentration increases as the temper- ature increases to *1600 K and decreases as the temperature increases from 1600 to 2300 K.Furthermore, at T51800 K the two-stage formation of hydrogen atoms is evident, as the charac- teristic time of the accumulation of hydrogen exceeds significantly the decomposition time of SiH4. These features confirm the complex character of the processes of the hydrogen atom for- mation .Si(H)Si+H., .SiH+H., .Si3H+H., (10) (11) (12) (13) H.+H.+M. Si+SiH2 Si+H2 Si+Si(H2)Si H2+M Thus, the monosilane pyrolysis results in significant amount of hydrogen (5% ± 10% of the initial silane concentration) even under homogeneous conditions and at high dilution. The hydro- gen atoms are formed not upon the decomposition of SiH4 or SiH2, but rather in the secondary bimolecular reactions.It was concluded that the hydrogen might take part in the chain process which includes the reaction (6) and (14) .SiH+H2+M, .SiH3+M (15) .Si2H4+H.. SiH4+SiH The rate constants for the radical reactions should be deter- mined to evaluate the importance of these reactions in the thermal decomposition of monosilane. Unfortunately, the information on these reactions in the literature is poor. The rate constants 63 for the reactions of SiH3 with Si2H6, of SiH3 with SiD4 and of H with SiH4 (T=300 K) are 4(74)610715, 4(42)610714, 4(2.50.5)610713 cm3 s71, respectively.63 According to the calculations of the activation energy, the rate constant for the reaction (7) should be 1610710exp[756.1(kJ mol71)/RT] cm3 s71.64 The calculated activation energy for the reaction (7) is 133 kJ mol71.62 For the reaction 65 (16) .Si2H5+H2 .SiH3+SiH4 the rate constant is k16=2.94610712 exp¡18:4 ÖkJ mol¡1Ü .RT The values of the rate constant for reaction (5) calculated based on the variation theory of the transition state for the temperature range 200 ± 1000 K (Ref. 66) agree with the experimental values. The above-mentioned data show that the role of the radical reactions in the monosilane pyrolysis is not clear. Nevertheless, the following conclusions can be made. First, under the conditions of the homogeneous pyrolysis, the primary reaction of the silane decomposition is the dissociation of SiH4 to H2 and SiH2 [reaction (2)].Second, the conditions under which radical reac- tions make considerable contribution exist probably even at relatively high pressures (where homogeneous reactions are pre- dominant). In static and flow reactors, the radicals are apparently formed in heterogeneous reactions.67 The later statement can be confirmed by the following data. Studies of monosilane pyrolysis in a flow reactor for a mixture of 5% SiH4 and 95% Ar (the total pressure was 6.6 kPa) gave the constants of the decomposition of monosilane at 843 and 933 K, viz., 0.25 and 3.6 s71, respectively. Let us estimate the possible contribution of the radical reactions to this process. The concen- tration of SiH3 radicals formed in the monosilane pyrolysis in a flow reactor for a mixture of 10% SiH4 and 90% Ar (the total pressure was 400 Pa) was measured by the method of laser magnetic resonance and was found to be 161010 and A A Onischuk, V N Panfilov 561010 cm73 at 843 and 933 K, respectively.28 The signal for SiH3 was proportional to the total pressure squared (the content of the silane in the initial mixture was constant) and to the partial pressure of the silane (the total pressure was fixed).Extrapolating the results obtained 28 for the mixture of 5% SiH4 and 95% Ar (the total pressure was 6.6 kPa), we obtain [SiH3]&161012 and 561012 cm73 at T=843 and 933 K, respectively. The ratio of the (Ref. 65) reaction rate to [SiH4] can be calculated using the rate constant for the reaction (16).At 843 and 933 K, this ratio is 0.2 and 1.37 s71, respectively. Thus the estimated rate of the reac- tion (16) is close to the rate of the monosilane decomposition measured experimentally. Unfortunately, the development of the expression for the rate constant for the reaction (16) has not been described in detail.65 Thus, the estimate given may suggest that the contribution of the radical reactions to the process of pyrolysis can be significant. 3. The mechanism of homogeneous pyrolysis of monosilane At present, it is believed (see e.g. Ref. 39) that the main contribu- tion to the process of the homogeneous monosilane pyrolysis (at least, in the early stages) is made by the following reactions SiH2+H2 , SiH4 Si2H6 , SiH2+SiH4 (17) Si3H8 , SiH2+Si2H6 Si4H10 . SiH2+Si3H8 Silylene (SiH2) formed in the reaction of primary monosilane decomposition reacts with another molecule of monosilane to give disilane.Silylene reacts consecutively with disilane to give trisi- lane, with trisilane to give tetrasilane, etc. The formation of polysilanes is accompanied by their dissociation and the forma- tion of substituted silylenes (18) H3SiSiH+H2 , Si2H6 (19) SiH3SiH2SiH+H2 . Si3H8 The reactions of substituted silylenes with monosilane result in the formation of trisilane, tetrasilane, etc. (20) Si3H8 , SiH4+H3SiSiH (21) Si4H10 . SiH4+SiH3SiH2SiH In addition, the isomerisation reactions (22) Si2H4 , H3SiSiH (23) Si3H6 , SiH3SiH2SiH unimolecular elimination of hydrogen, e.g., (24) Si2H2+H2 , Si2H4 (25) Si2+H2 , Si2H2 (26) Si3H4+H2 , Si3H6 etc, take place.The intermediate product Si2H2 is represented by two isomers, Si(H2)Si being more stable.39, 68 The Si3H4 species also has two isomers.47 Transition from the initial stage of the pyrolysis to intermedi- ate one is accompanied by the acceleration of the process. Devyatykh and coworkers 16 were the first to investigate this phenomenon. This transition was described in more detail by Purnell and Walsh,15 and the first mechanism of the initial acceleration was suggested in their monograph (Ref. 69). The reactions are autocatalytic processes with SiH3SiH as an inter- mediate product. SiH3SiH+H2 , Si2H6 SiH2+SiH4 (27) SiH2+Si2H6 .Si3H8 SiH3SiH+SiH4Mechanism of thermal decomposition of silanes The results of simulation 39 have shown that the initial accel- eration observed under the conditions of a static reactor 15 agree with the mechanism involving reactions (27). It should be noted that the initial acceleration in the process of the monosilane pyrolysis in a flow reactor was also described based on the mechanism of the reactions (27).27, 67 An alternative heterogeneous mechanism of the initial accel- eration of the silane pyrolysis under the conditions of a static reactor 15 was suggested.13, 70 Modification of the surface during the reaction was supposed to be the reason for the initial accel- eration. As the result, SiH2 is liberated from the surface into the gas phase accelerating the homogeneous process.This mechanism was criticised,39 as the pyrolysis rate and the product ratio did not depend on the S/V ratio.15 Furthermore, the homogeneous mechanism of the initial acceleration was supported by the results obtained in the silane pyrolysis where the influence of the reactor walls was eliminated (shock tubes 30 and laser pyrolysis 49). k=Aexp , ¡ E RT At present, the rate constants for numerous homogeneous elementary reactions occurring in the monosilane pyrolysis have been established. In particular, for the numerical simulation of the monosilane pyrolysis, one can use the parameters of the Arrhenius equation where A is a pre-exponential factor, E is the activation energy.The parameters A and E for the initial decomposition of the monosilane molecule [reaction (2)] in the pyrolysis of undiluted silane for the range of pressures are given in Table 1. Table 1. Parameters of the Arrhenius equation for the reaction of primary monosilane decomposition in the range of temperatures 800 ± 1500 K.60 A /s71 P /Pa 1.56107 1.56108 1.56109 1.561010 1.561011 1.861012 1.761013 3.461015 0.1 1.3 13 1.36102 1.36103 1.36104 1.36105 ?Table 2. Parameters of the Arrhenius equation (A /cm3 s71 or s71; E /kJ mol71) for the reactions occurring in the silane pyrolysis. Reaction SiH2+H2?SiH4 SiH4+SiH2?Si2H6 Si2H6?SiH2+SiH4 Si2H6+SiH2?Si3H8 Si3H8?SiH2+Si2H6 Si3H8+SiH2?Si4H10 Si4H10?Si3H8+SiH2 Si2H6?H3SiSiH+H2 H3SiSiH+H2?Si2H6 Si3H8?SiH3SiH2SiH+H2 SiH3SiH2SiH+H2?Si3H8 SiH4+H3SiSiH?Si3H8 Si3H8?SiH4+H3SiSiH SiH4+SiH3SiH2SiH?Si4H10 Si4H10?SiH4+SiH3SiH2SiH H3SiSiH?Si2H4 Si2H4?H3SiSiH SiH3SiH2SiH?SiH3SiHSiH2 SiH3SiHSiH2?SiH3SiH2SiH E /kJ mol71 214.4 214.4 214.4 214.4 214.8 218.6 223.6 250.4 A E Ref.40 40 43 40 42 40 70 70 70 70 70 70 42 70 70 72 72 70, 71 70, 71 72.3 73.0 213 71.9 222 72.0 223.6 2379.2 2299.2 710.9 209 710.9 210.3 6.5 52.7 6.5 52.7 9.8610713 1.3610710 1.361015 3.1610710 4.961015 3.7610710 6.3161015 7.961015 2.8610710 1.861015 2.83610710 6.6610711 9.361014 6.6610711 4.461015 6.26109 1.361010 6.26109 1.361010 325 In order to apply the data of Table 1 to the mixture of monosilane and argon, the parameters of the Arrhenius equation for several other elementary reactions of the silane pyrolysis are given in Table 2.III. Homogeneous pyrolysis of disilane Possible elementary reactions occurring in the thermal decom- position of disilane are described in the theoretical studies (Refs 72 ± 75), while experimental investigations into the pyrolysis of disilane and higher silanes are presented in other papers (Refs 32, 42 ± 44, 76, 77). Chromatographic analysis of the pyrol- ysis products of disilane carried out under static conditions at P=36103 ± 104 Pa and T=560 ± 610 Kshows that the products are monosilane, trisilane, two tetrasilane isomers and hydrogen.76 The parameters of the Arrhenius equation for the reactions of formation of monosilane, trisilane and hydrogen were deter- mined.The mechanism of decomposition of disilane is following: the initial stage is the Si2H6 dissociation into SiH2 and SiH4, SiH2 reacts further with Si2H6 to form Si3H8 and with Si3H8 to form Si4H10, etc. The experimental data suggest that hydrogen is a secondary reaction product. The kinetics of the disilane decomposition was studied in a single-pulse shock tube in the temperature range 850 ± 1000 Kand at P=(3.1 ± 3.6)6105 Pa.44 It was considered that two primary reactions of the disilane decomposition, viz., the forward reac- tion (18) and the reverse reaction (3) occur simultaneously.The parameters of the Arrhenius equation for these reactions were found (see Table 2). Thermolysis of disilane under static conditions was experi- mentally studied by Martin, Ring and O'Neal.43 The dependences of rate constants for this process on temperature and pressure were obtained. The calculated parameters of the Arrhenius equation differ substantially from those found by Bowrey and Purnell.76 The experimental investigation of the thermolysis of disilane, as well as tri- and tetrasilane under the conditions of a static reactor was continued later.42 The parameters of the Arrhenius equation for the primary decomposition reactions of these silanes were found. Studies of the kinetics of the disilane thermal decomposition in a shock tube at 1000 ± 3100 K have shown that disilane generally decomposes into SiH2 and SiH4.32 The process of the thermolysis of disilane at T=300 ± 1000 K and P=100 ± 1000 Pa was studied by resonance enhanced multi- photon ionisation (REMPI) and multiphoton ionisation (MPI) methods.78 The formation of silicon atoms during the pyrolysis was detected by REMPI method.Thus, according to the above-mentioned data, the reactions SiH2+SiH4 , Si2H6 Si3H8 , Si2H6+SiH2 Si4H10 etc. Si3H8+SiH2 are predominant in the primary stage of the disilane pyrolysis. Furthermore, the formation of silylsilylene H3SiSiH and hydrogen as a result of dissociation of disilane, viz., the forward reaction (18), takes place. Silylsilylene is in equilibrium with its isomer disilene Si2H4 [equilibrium (22)]. Silylsilylene reacts actively with disilane (and other silanes), for example, (28) Si4H10 .H3SiSiH+Si2H6 Thus, pyrolysis of disilane can proceed with an acceleration of the process due to the additional consumption of Si2H6 as the result of its reaction withH3SiSiH and other substituted silylenes.326 IV. Heterogeneous decomposition of silanes Silicon deposition on the surface as the result of the heterogeneous silane decomposition has been described in detail in several reviews (see e.g. 45, 46). Hence, the processes of CVD are considered only in brief in the present review. The publications on the heterogeneous pyrolysis can arbitrarily be divided into two groups: the first group includes studies on the rate of formation of silicon films depending on macroscopic conditions (such as temperature, pressure and composition of a gas mixture); the second group deals with investigations into elementary heteroge- neous reactions on the well-characterised ideal surface of single- crystal silicon using modern methods of analysis of the surface. 1.Growth of silicon films upon heterogeneous decomposition of silanes Studies on the rate of formation of silicon films by the CVD method are aimed at the search for the optimum deposition conditions. Generally, this is associated with technological demands for microelectronics. Experiments were mostly carried out at reduced pressures to avoid the formation of aerosol particles, the deposition of which on the surface of growing layers impairs the product quality.In majority of investigations into CVD, silane 5, 60, 79 ± 93 and disilane 5, 60, 80, 81, 89, 94 ± 99 served as precursors of silicon films. The conditions of deposition in these investigations were so chosen, that the contribution of the homogeneous reactions was negligible. Thus, using the data from these publications, one can determine the reaction sticking a log es 74 75 123456 78910 11 12 13 76 77 b 74 75 76 c 73 74 75 7671 0 1 2 3 logps (Pa) Figure 1. Dependence of the reaction sticking coefficient of monosilane on its partial pressure over the surface of the growing silicon layer at 813 (a), 863 (b) and 950 K (c).Straight lines are plotted using the Eqn (30), the experimental points (1) ± (13) are taken from Refs 92, 5, 81, 87, 86, 84, 88, 96, 95, 91, 85, 93 and 82, respectively. A A Onischuk, V N Panfilov coefficient of monosilane (es) and disilane (eds). For instance, es is calculated using the equation (29) es=Wr Zm , whereWis the rate of the film growth (cm s71); r is the density of the growing silicon film (r&2 g cm73); Z is the collision fre- quency of the monosilane molecules with the unit surface area; m is the mass of the silicon atom. The value of eds is calculated by the same manner. Summing up the results given in the above-cited papers, we derived the equations for the reaction sticking coefficient of monosilane and disilane for a wide range of temperatures and partial pressures of these substances (750<T<1000 K, 261074<ps<2 kPa, 161074<pds<1 kPa) 27 (30) sÜ¡0:59, es=103.0exp ¡ 148 ÖkJ mol¡1Ü Öp RT (31) 119 ÖkJ mol¡1Ü RT ÖpdsÜ¡0:70, eds=101.6exp ¡ where ps and pds are partial pressures of silane and disilane, respectively.The dependences of the reaction sticking coefficient of mono- and disilanes on their partial pressures are shown in Figs 1 ang 2. log eds 72 74 II 1234567 76 I 78 72 71 0 1 2 logpds (Pa) Figure 2. Dependence of the reaction sticking coefficient of disilane on its partial pressure over the surface of the growing silicon layer at 738 (I) and 863 K (II).Straight lines are plotted using the Eqn (31), the experimental points (1) ± (7) are taken from Refs 5, 96, 79, 99, 95, 91 and 94, respectively. The reaction sticking coefficient of trisilane is approximately the same as those of disilane (for the same range of partial pressures).89 It is noteworthy that the transport of the molecules from the gas phase to the reactor walls does not limit the process of film growth in the range of temperatures mentioned.46 Using the Eqns (29) ± (31), the growth rate of the silicon film in the hetero- geneous decomposition of silane and disilane can be estimated for a wide range of temperatures and pressures. 2. Dissociative adsorption of silanes on the ideal surface of single-crystal silicon Information on heterogeneous elementary reactions which occur in the pyrolysis of silane on the surface of a solid product formed, viz., amorphous hydrogenated silicon is rather scarce.However, the mechanism of silicon growth on the well-characterised ideal Si(111) and Si(100) surfaces is discussed in numerous studies. Modern methods of the analysis of surface such as thermodesorp- tion,98 low-energy electron diffraction,98, 100, 101 secondary-ion mass spectrometry,102 scanning tunnelling microscopy,103 ± 108 reflection high-energy electron diffraction 109 ± 111 are used in these investigations.Mechanism of thermal decomposition of silanes It was found that two neighbouring dangling bonds partic- ipate in elementary reactions of heterogeneous decomposition of mono- and disilane.104, 107 The elementary reaction of the hetero- geneous monosilane decomposition can be represented by the reaction 104, 108, 112 ± 114 (32) H+SiH3 .SiH4+2_ Hereinafter, the symbol _ denotes an active centre of the surface (dangling bond), H and SiH3 mean that H and SiH3 are bound to the surface (mono- and trihydride groups, respectively). (33) SiH3+SiH3 . Si2H6+2_ (34) Si2H5+H. Si2H6+2_ The elementary reaction of the heterogeneous decomposition of disilane occurs as follows 105, 115, 116: Alternatively, the following version is possible (Refs 115, 117, 118): Trihydride groups are unstable at high temperatures and decompose in the reaction presented below 105, 113, 116, 119 ± 121: (35) SiH2+H, SiH3+2_ where SiH2 is the group bound to two surface silicon atoms.V. Aerosol formation in the pyrolysis of silanes Three characteristic stages of aerosol formation in the pyrolysis of silanes can be distinguished: (1) nucleation, i.e., the formation of nuclei which grow much faster than they decompose; (2) growth of single particles in the processes of `condensation' and coagulation; (3) the formation of agglomerates consisting of individual primary particles. The stage of nucleation is the least studied. 1. Experimental investigation into aerosol formation Aerosol formation can take place in the process of thermal decomposition of silane under certain conditions. The regimes of an epitaxial reactor which favour aerosol formation were studied by Eversteijn.122 Mixtures of silane with hydrogen under atmos- phere pressure were used.Aerosol formation occurred above certain critical concentration of silane, and this lower limit of silane concentration was registered at different temperatures. The critical silane concentration depended on the temperature of the reactor and changed from 0.248% to 0.002% as temperature increased from 1050 to 1410 K. The process of aerosol formation in an epitaxial reactor was also studied by Murthy et al.123 Mixtures of silane with argon under atmosphere pressure were used. If the concentration of silane at the reactor entrance exceeded 0.2%, `smoke' was observed at the exit from the reactor at 1370 K. The onset of gas-phase nucleation was determined from `smoke' formation.The critical concentration of silane decreased with the increase in temperature. Investigations into aerosol formation in flow reactors were carried out.9, 26, 124 ± 129 Critical concentrations of silane in the decomposition of its mixtures with inert gases and hydrogen at low temperatures were determined by Slootman and Parent.9 These experiments were carried out in a vertical flow reactor under atmosphere pressure. It was found that the critical concen- tration of silane for all carrier gases is inversely proportional to temperature. Homogeneous nucleation in inert gases began at relatively low temperatures. In the experiments with mixtures of silane with hydrogen, aerosol formation was observed at much higher temperatures.It is necessary to consider chemical reactions inhibiting the formation of aerosol particles to rationalise the increase in temperatures. It was suggested that the large excess of hydrogen influences kinetics of gas-phase processes, viz., favours reverse reactions with silylene, which results in the decrease in the concentration of aerosol particle precursors. Qian et al.124 carried out experiments in a low-pressure flow reactor using pure silane. The residence time of silane in a hot zone 327 was *1 s. The beginning of the homogeneous nucleation was established by appearance of a weak smoke-forming flash. The critical concentration of silane (above which the nucleation was observed) increased *1.26-fold on the increase in temperature from 970 to 1020 K.The mechanism of deposition of polycrystalline silicon from silane was studied in a low-pressure flow reactor.125 Mixtures of silane with nitrogen were used. The formation of aerosol particles was determined from laser light scattering. If the pressure of the silane exceeded the critical value (the initial content of silane in the mixture being the same), the intensity of the scattered light increased sharply. Thus for the initial mole fraction of silane equal to 0.3, the flow rate 400 cm3 min71 and temperature 1000 K, an abrupt increase in the intensity of the scattered light was observed at*85 Pa. Electron microscopic studies of particles formed at 130 Pa suggest that these particles are single crystals with characteristic sizes from 5 to 10 nm.Studies of the growth of silicon particles formed upon pyrol- ysis of silane in a multistage flow reactor followed by condensa- tion, showed that the formation of the primary silicon particles as the result of homogeneous nucleation took place in the first stage.126, 127 In the second stage, the primary particles served as the `condensation nuclei' for the products of heterogeneous decomposition of silane and the intermediate products formed in the homogeneous pyrolysis. It was noted that the regime of spontaneous homogeneous nucleation can be realised provided such parameters as flow rate, silane concentration and temper- ature profile are properly selected. The morphology of the particles and their crystal structure were studied by electron microscopy.It was found that the particles are non-agglomerated spheres with the size of*0.1 mm. The major part of the particles has crystal structures. A method of synthesis of spherical particles homogeneous in size in a reactor with five heating zones was described.128 The particle morphology was determined by electron transmission microscopy method and their crystal structures were determined by X-ray and electron diffraction. Particles obtained in a flow reactor were shown to be clusters resulting from the homogeneous decomposition of silane, nucle- ation and coagulation.26 The sizes of the clusters varied from 1 to 10 mm. X-Ray diffraction analysis revealed that the particles have crystalline structures. The results of an experimental study of aerosol formation under static conditions were given in Refs 4 and 17.The formation of aerosols upon thermal decomposition of silane under pressure of 0.3 ± 4.5 Pa and 750 ± 830 K was investigated.4 The aerosol particle formation was followed by monitoring changes in optical transmission at wavelengths of 515 and 755 nm. It was suggested that the precursors of aerosol are silicon atoms. The critical character of the dependence of aerosol formation on the silane pressure was rationalised as being due to competition of the homogeneous nucleation and silicon deposition on the reactor walls.17 The investigation of aerosol formation in the pyrolysis of mono- and disilane diluted with argon was carried out in a shock tube.8 The aerosol particle formation was followed by monitoring the absorption of laser radiation at two different wavelengths.Information on the sizes, compositions and the number concen- trations of particles was obtained. The investigation of the particles by such methods as electron diffraction, electron trans- mission microscopy and secondary-ion mass spectroscopy shows that the particles are spherical with their diameters ranging from 10 to 40 nm. These particles are weakly agglomerated and contain 15 at.% of hydrogen. Anumber of investigations 130 ± 135 was devoted to the study of aerosol formation upon heating of silane by laser radiation. The formation of particles in mixtures of silane with argon under the action of a CO2-laser was investigated by Flint and co- workers.130, 131 The sizes and number concentrations of the particles were determined from the results of the measurement of328 scattered and passed radiation.It was concluded that silicon particles are formed as amorphous silicon with subsequent crystallisation on passing into hotter reaction zones. The silicon powder obtained upon laser heating of silane was analysed.133, 134 Physicochemical parameters of the powder were determined, using a number of methods: the sizes of the particles and crystallites were determined by electron transmission micro- scopy and scanning electron microscopy, while the density was determined by the method of helium pycnometry. The crystal structures and sizes of the crystallites were determined by X-ray diffraction (considering the width of X-ray diffraction lines using the Debye ± Sherrer formula), electron diffraction and dark-field electron microscopy.It was found that both separate spherical silicon particles and chain agglomerates are formed in the laser pyrolysis. The sizes of the primary particles incorporated in these agglomerates are in the range 10 ± 100 nm. The particles are built of separate crystallites with the size*15 nm. The morphology of aerosol silicon particles formed upon decomposition of silane is of particular interest. Coagulation of these particles results in agglomerates of irregular struc- tures.128, 136 Ag,137, 138 of Analogous agglomerates TiO2,136, 139 ± 144 SiO2 141, 144 ± 147 and soot 148 ± 150 are formed in natural and industrial processes.151 These structures can be characterised by the fractal dimension Df<3, which is the exponent in the ratio Df , (36) M! Ra which relates the mass of the agglomerateMto its radiusR, a is the average radius of the primary particles. The silicon agglomerates formed upon pyrolysis of monosilane at temperatures 800 ± 1720 K had the fractal dimension Df=1.5 ± 1.7 (Refs 29, 136).2. The mechanism of aerosol formation Much attention has been paid to the mechanism of aerosol formation in the thermal decomposition of silanes. A mechanism of the formation of particles from silicon atoms was suggested based on investigation of aerosol formation upon thermal decom- position of monosilane under static conditions in the range of temperatures 723 ± 1070 K and pressures 66 ± 1335 kPa.17 Studies of thermal decomposition of silane at 770 ± 1070 K have led to a conclusion that the particles could be formed from Si and Si2 atoms.21, 22 A model of homogeneous polymerisation in the gas phase for the conditions of low-pressure reactors was sug- gested.125 According to this model, the particles are formed from SiH2.A model of particle formation from disilane was considered by White et al.13 and Ring and O'Neal.70 An assumption was made that trisilane could be a precursor of the solid product in the processes of the deposition on the reactor walls and homogeneous polymerisation.39 Decomposition of pentasilane was considered to be the initial stage of aerosol formation.152 A model of aerosol formation in low-pressure reactors which allows identification of the conditions of homogeneous nucleation and those under which heterogeneous reactions predominate was suggested by Sladek.153 This model is based on the competition of nucleation processes and diffusion of aerosol precursors to the reactor walls.A model of aerosol formation upon silane pyrolysis was given.154 The calculations for the temperature range 900 ± 2400 K were made based on the classical nucleation theory and assumption that aerosols are formed from silicon atoms. The critical nucleus contained several silicon atoms at 900 ± 2000 K. Yet another theoretical investigation of silicon nucleation in silane decomposition was made based on the classical nucleation theory.155 Simulation of the aerosol formation process in the silicon layer deposition from mono- and disilane at atmospheric pressure showed that the basis of the mechanism is the growth of aerosol particles as the result of the consecutive addition of silylenes.77 A A Onischuk, V N Panfilov A kinetic model of aerosol formation in thermal decomposi- tion of silane was described.156 The essence of this model is that a nucleus of an (SiHm)n particle is formed in the initial stage as the result of homogeneous reactions; its further growth occurs due to its reaction with silane.A mechanism of the formation of silicon hydride clusters in the silane pyrolysis was discussed.47 This model contains detailed chemical information on the relative stabilities and reactivities of different possible silicon hydride clusters.The model allows one to calculate the rate of the particle nucleation. Thermodynamic properties of the silicon hydride clusters were estimated. The mechanism includes reversible and irreversible formation of silicon hydrides: the former reactions involve particles containing up to 10 Si atoms, while the latter reactions involve particles containing 11 ± 20 Si atoms. The key intermediates and reactions leading to the formation of the particles were determined, the dependence of their formation on the conditions of the reaction was studied. A detailed kinetic model describing pyrolysis of the initial molecules and the homogeneous nucleation of the silicon particles under shock tube conditions was considered.8 The kinetic scheme contains 117 elementary reactions and includes 42 types of molecules (from fully hydrogenated SinH2n+2 to unsaturated SinH2n, Si2 H2 and non-hydrogenated silicon Sin clusters).The model describes the process of coagulation of aerosol particles and their growth due to the CVD process. Numerical solutions of discrete-sectional equations of the aerosol dynamics, which take account of the interaction of separate clusters for various reactor types were described.128, 157 Simplified equations of reaction and coagulation which permit one to predict rapidly the aerosol evolution were derived.The influence of the initial vapour concentration, reaction time, seeds and temperature profile on the properties of the highly dispersed particles formed in silane pyrolysis was estimated. Numerical stimulation of aerosol formation under the hetero- geneous-homogeneous thermal decomposition of silane was made.67 The model allows one to calculate the following time- dependent parameters: the degree of silane decomposition, di- and trisilane concentration, the concentration and dispersity of aero- sol particles formed, the hydrogen content and the concentration of active centers (in the form of dangling bonds) in particles. These data agree reasonably with the experimental results. The results of the above-mentioned studies show that the modern concepts on the process of aerosol formation are contra- dictory and its mechanism is not sufficiently studied.A reason for the disagreement as to the mechanism of aerosol formation is the insufficiency of experimental data. As a rule, the experiment gives only some parameters of the process such as the sizes of the particles,8, 128 or temperature and the concentration of silane required for the onset of homogeneous nucleation.4, 9, 17, 122 ± 126 Nevertheless, based on the available literature data general features of aerosol formation in the silane pyrolysis can be formulated. The initial stage of this process is the formation of SimHn clusters. The thermochemical calculations 47 show that for small sizes of clusters (m410) the rate of their decomposition is relatively high, which leads to a decrease in the number of silicon atoms in the clusters (37) SimH2n+SilH2k .Sim+lH2(n+k) In other words, the concentration of this kind of clusters depends on a quasi-equilibrium between the formation reactions of these clusters due to the agglomeration of smaller clusters (38) Sim+lH2(n+k) , SimH2n+SilH2k and the decay of lager clusters [see reaction (37)], on one side, and involvement of these clusters in analogous reactions of agglomer- ation and decay, on the other side. The rate of cluster decom- position is relatively small for larger cluster sizes (m>10); since the dissociation of a large cluster requires that larger numbers of the Si7Si bonds were cleaved than in the dissociation of a smallerMechanism of thermal decomposition of silanes cluster.Thus, the clusters with m>10 may be regarded as particles and the rate of the formation of clusters with m>10, as the rate of nucleation.47 Further, the rate of the particle growth is determined largely by coagulation.67 VI. The composition of the solid product The detailed investigations on the chemical composition of aerosol particles formed in thermal decomposition of silane are virtually not documented and the data on the solid product of the reaction which deposits on the reactor walls during the silane pyrolysis are very scarce. Therefore, the experimental data con- cerning this aspect should be analysed in more detail than in the previous sections.The solid product (a deposit on the reactor walls and aerosol particles) of thermal decomposition of silane represents polycrys- talline silicon 8 (high temperatures and/or long reaction time) or amorphous hydrogenated silicon 3 (mild temperatures and/or short reaction periods). The content of hydrogen in the solid pyrolysis product vary over a wide range (see Section I). The bound hydrogen present in amorphous hydrogenated silicon has been actively studied over the last three decades. Detailed information on the vibration modes of mono- (SiH) and polyhydride [SiH2, (SiH2)n, SiH3] groups present in amor- phous hydrogenated silicon was obtained (see, e.g., Refs 158 ± 162). IR spectra contain three characteristic absorption bands in the range of frequencies 550 ± 750, 800 ± 950 and 1900 ± 2300 cm71 (Ref.158). Two peaks at *2000 and 2100 cm71 can be found in the range of frequencies 1900 ± 2300 cm71. A doublet with maxima at 845 and 880 cm71 or a single peak with maximum at 880 cm71 is most often observed in the range 800 ± 950 cm71. The data allow one to compare different vibration modes with absorption bands observed in the IR spectra of silanes (Table 3). Table 3. Vibration frequencies of the Si7H bond observed in the IR- spectra of silanes. Frequencies /cm71 Vibration type Group SiH SiH2 2000 630 2090 880 630 SiH3 2120 905, 860 2090 ± 2140 (SiH2)n 890 845 630 stretching bending stretching scissors oscillation stretching deformation stretching scissors oscillation the same The absorption peak at *2100 cm71 has to be discussed separately.It was supposed earlier 163, 164 that the stretching vibrations of the monohydride groups manifest themselves only at *2000 cm71. However, it is well known that the stretching vibrations of the monohydride groups localised on the surface of single-crystal (Refs 116, 165 and references cited therein) and amorphous silicon 166, 167 have the frequencies at *2100 cm71. Furthermore, lines at 2140,168 2100,169 2085,170, 171 2068 cm71 (Ref. 172) were found in the IR spectra of the amorphous hydro- genated silicon, while absorption at 800 ± 950 cm71 (correspond- ing to the polyhydride groups, see Table 3) was absent.Heating of amorphous hydrogenated silicon resulted in an increase in the absorption intensity at 2100 cm71 without any absorption at 800 ± 950 cm71.173, 174 These results suggest that two types of monohydride groups, which have absorption lines at 2000 and 2100 cm71, are possible in amorphous hydrogenated silicon. Using aerosol particles of amorphous hydrogenated silicon as an 329 example, it was shown that the SiH groups with the vibration frequency*2000 cm71 do not interact with atmospheric oxygen, unlike the groups with the frequency*2100 cm71 (Refs 3, 175). The interaction of the latter is characterised by a decrease in the line intensity at 2100 cm71 and appearance of absorption in the region 2170 ± 2300 cm71. The absorption in this region is gener- ally ascribed to monohydride groups where the hydrogen and oxygen atoms are bound to the same silicon atom (Si7O7 Si7H).176 ± 182 The hydrogen evolution method is a most informative method used for the investigation of amorphous hydrogenated sili- con.183 ± 194 This method allows one to obtain quantitative infor- mation on the content of hydrogen in groups of different structures.The essence of this method can be described in the following way. A sample of amorphous hydrogenated silicon is heated at a constant rate (20 ± 40 K min71), which results in the evolution of hydrogen. The amount of the hydrogen evolved is determined by the increasing pressure or by mass spectroscopy. The sample is heated to the T*920 ± 950 K, where crystallisation takes place and all the hydrogen releases to the gas phase.The results of the hydrogen evolution are usually represented as the dependence of the rate of evolution on temperature. Two temper- ature regions, viz., 570 ± 770 and 770 ± 920 Kcan be distinguished. In the former region one 185 ± 191, 193, 194 or two 183, 186, 188, 189, 191 peaks are observed. The hydrogen evolution in this temperature region is described by the first-order process and is not diffusion- limited.185, 190 It is supposed that the low-temperature peak in this region is associated with elimination of hydrogen from the polyhydride groups,183, 187 ± 189, 194 while the high-temperature peak is determined by elimination of hydrogen from the mono- hydride groups.183, 188 In the region 770 ± 920 K, a peak the location of which depends on the rate of heating is observed.Hydrogen evolution in this case is described by the kinetics with an order exceeding the first order.185 ± 187, 189, 192 This peak is associ- ated with the elimination of hydrogen from the monohydride groups.183, 185 ± 189 In relation to the interpretation of the peaks in the evolution spectra of hydrogen, it is worthwhile to note the investigations into the thermal desorption of hydrogen from the (111) and (100) faces of silicon single crystal 113, 195 ± 202 and from the surface of porous crystalline sample.202 It was found 198, 202 that the desorption of hydrogen from di- and trihydride groups occurs at 640 ± 715 K, while from monohydride groups, at 720 ± 820 K.Detailed comparison of the IR spectra and the hydrogen evolution spectra for the aerosol particles of amorphous hydro- genated silicon formed upon thermal decomposition of mono- silane in a flow reactor was described.36, 175 The evolution spectra of the samples of aerosol particles withdrawn from the flow contain two relatively sharp peaks with the maxima at 670 and 760 Kand a broad peak with the maximum at 780 K. It was found that the sharp peaks are determined by the evolution of hydrogen from poly- and monohydride groups with the frequency of stretching vibrations 2000 cm71, while the broad peak is deter- mined by the evolution of hydrogen from monohydride groups with the frequency of stretching vibrations*2100 cm71.The investigations into amorphous hydrogenated silicon by the proton magnetic resonance method (Refs 203 ± 205) show that the spectra represent superposition of two lines, viz., a broad (20 ± 30 kHz) one and a narrow (3 ± 4 kHz) one. The hydrogen giving the broad line is called the `clustered' phase and hydrogen giving the narrow line is called the `dilute' phase. Hydrogen atoms of the both phases are bound to silicon atoms by covalent bonds. The difference between these two phases is that hydrogen atoms of the `clustered' phase are located close to each other, while the hydrogen atoms of the `dilute' phase are isolated. For the aerosol particles of the amorphous hydrogenated silicon obtained upon thermal decomposition of monosilane, the data of IR spectroscopy were compared to those of the hydrogen evolution, on one hand, and the data fromNMRspectroscopy, on the other hand.3 It was concluded that the hydrogen of the330 `clustered' phase is a constituent of mono- (vibration frequency is 2000 cm71) and polyhydride groups.The hydrogen of the `dilute' phase is a constituent of monohydride groups with the vibration frequency 2100 m71. Experimental results suggest that the SiH groups of the `clustered' phase are localised inside the amorphous silicon network 206 as `hydrogen complexes', viz., extended defects stabilised by hydrogen.207 ± 211 The monohydride groups of the `dilute' phase are localised on the surfaces of interconnected microvoids and microchannels.VII. Conclusion Thermal decomposition of silanes is a complex process including different stages. Arbitrarily, the main components of the mecha- nism of this process are as follows: 1. Homogeneous elementary reactions. 2. Heterogeneous reactions occuring on the surface of the silicon deposited on the reactor walls. 3. The formation of precursors of aerosol particles (molecular clusters SimHn, m=10 ± 20) which exist in the equilibrium with gaseous intermediate products. 4. Homogeneous nucleation, i.e., the formation of the smallest aerosol particles, the rates of forward growth reactions for which exceed substantially those of the reverse reactions. 5. Heterogeneous deposition of molecules and smaller clusters on the surface (`condensation').6. Coagulation and coalescence of particles. In the initial stages where the rate of coalescence exceeds that of the coagu- lation, more or less spherical particles are formed. In the following stages where the coagulation becomes predominant, particle agglomerates (the so-called primary particles) are formed. 7. Agglomeration of primary particles, i.e., the transformation of several neighbouring particles into one primary particle. 8. Chemical and structural changes in aerosol particles. To explain the pyrolysis mechanism in general each of the above-mentioned stages has to be investigated in detail because of their close relations and mutual influence on each other. Only thorough study with the help of modern physicochemical methods can give the detailed representation of this process.References 1. F O Rice J. Am. Chem. Soc. 55 3035 (1933) 2. N A Vasil'eva, R A Buyanov, K I Zamaraev Khim. Fiz. 3 688 (1984) a 3. A A Onischuk, V P Strunin, R I Samoilova, A V Nosov, M A Ushakova, V N Panfilov J. Aerosol Sci. 28 1425 (1997) 4. R G Aivazyan, V V Azatyan, L F Satunkina Kinet. Katal. 37 492 (1996) b 5. B A Scott, R D Estes, J M Jasinski J. Chem. Phys. 89 2544 (1988) 6. A G Gasparyan,G B Rostomyan,R A Badalyan,G A Arutyunyan Kinet. Katal. 29 935 (1988) b 7. M S Sukhov Izv. Akad. Nauk, Neorg. Mater. 18 1077 (1982) c 8. M Frenklach, L Ting,H Wang, J M Rabinowitz Isr. J. Chem. 36 293 (1996) 9. F Slootman, J-C Parent J. Aerosol Sci. 25 15 (1994) 10.R Robertson, D Hils, A Gallagher Chem. Phys. Lett. 103 397 (1984) 11. J W Erwin,M A Ring, H E O'Neal Int. J. Chem. Kinet. 17 1067 (1985) 12. N K Serdyuk, V P Strunin, E N Chesnokov, V N Panfilov Kinet. Katal. 26 790 (1985) b 13. R T White, R L Espino-Rios, D S Rogers,M A Ring, H E O'Neal Int. J. Chem. Kinet. 17 1029 (1985) 14. M A Ring, M J Puentes, H E O'Neal J. Am. Chem. Soc. 92 4845 (1970) 15. J H Purnell, R Walsh Proc. R. Soc. London, A Math. Phys. Sci. 293 543 (1966) 16. G G Devyatykh, V M Kedyarkin, A D Zorin Zh. Neorg. Khim. 10 1528 (1965) d 17. S V Maklakov, R G Aivazyan, V V Azatyan, P V Zhirkov, L F Satunkina Mendeleev Commun. 120 (1995) A A Onischuk, V N Panfilov 18. P Neudorfl, A Jodhan, O P Strausz J. Phys. Chem. 84 338 (1980) 19.R G Aivazyan, V V Azatyan, V I Kalachev, V I Rubtsov Fiz. Goreniya Vzryva 30 85(1994) 20. G Cochet, H Mellottee, R Delbourgo J. Chim. Phys. 71 1363 (1974) 21. W G Breiland,M E Coltrin, P Ho J. Appl. Phys. 59 3267 (1986) 22. W G Breiland, P Ho,M E Coltrin J. Appl. Phys. 60 1505 (1986) 23. S K Iya, R N Flagella, F S DiPaolo J. Electrochem. Soc. 129 1531 (1982) 24. C H Haas, M A Ring Inorg. Chem. 14 2253 (1975) 25. B A Scott, R D Estes Appl. Phys. Lett. 55 1005 (1989) 26. G Hsu, R Hogle, N Rohatgi, A Morrison J. Electrochem. Soc. 131 660 (1984) 27. A A Onischuk, V P Strunin,M A Ushakova, V N Panfilov Int. J. Chem. Kinet. 30 99 (1998) 28. L N Krasnoperov, V V Nosov, A V Baklanov, V N Panfilov Khim. Fiz. 7 528 (1988) a 29.A A Onischuk, A I Levykin, V P Strunin, K K Sabelfeld, V N Panfilov J. Aerosol Sci. 31 1263 (2000) 30. C G Newman, H E O'Neal, M A Ring, F Leska, N Shipley Int. J. Chem. Kinet. 11 1167 (1979) 31. V N Votintsev, I S Zaslonko, V S Mikheev, V N Smirnov Kinet. Katal. 27 972 (1986) a 32. H-J Mick,M W Markus, P Roth, V N Smirnov Ber. Bunsenges. Phys. Chem. 99 880 (1995) 33. C G Newman, M A Ring, H E O'Neal J. Am. Chem. Soc. 100 5945 (1978) 34. M Koshi, S Kato, H Matsui J. Phys. Chem. 95 1223 (1991) 35. H-J Mick, P Roth, V N Smirnov, I S Zaslonko Kinet. Katal. 35 485 (1994) a 36. A A Onischuk, V P Strunin,M A Ushakova, V N Panfilov Phys. Status Solidi B 186 43 (1994) 37. A A Onischuk, V P Strunin,M A Ushakova, V N Panfilov Khim. Fiz. 11 1136 (1992) a 38.A A Onischuk, V P Strunin,M A Ushakova, V N Panfilov Khim. Fiz. 13 (5) 129 (1994) a 39. R Becerra, R Walsh J. Phys. Chem. 96 10856 (1992) 40. J M Jasinski, R Becerra, R Walsh Chem. Rev. 95 1203 (1995) 41. I Safarik, V Sandhu, E M Lown,O P Starausz, T N Bell Res. Chem. Intermed. 14 105 (1990) 42. J G Martin, H E O'Neal, M A Ring Int. J. Chem. Kinet. 22 613 (1990) 43. J G Martin,M A Ring, H E O'Neal Int. J. Chem. Kinet. 19 715 (1987) 44. J Dzarnoski, S F Rickborn, H E O'Neal, MA Ring Organometallics 1 1217 (1982) 45. J M Jasinski, S M Gates Acc. Chem. Res. 24 9 (1991) 46. J M Jasinski, B S Meyerson, B A Scott Ann. Rev. Chem. 38 109 (1987) 47. M T Swihart, S L Girshick J. Phys. Chem., B 103 64 (1999) 48. H E O'Neal, M A Ring Chem.Phys. Lett. 107 442 (1984) 49. J M Jasinski, R D Estes Chem. Phys. Lett. 117 495 (1985) 50. V V Nosov, S M Repinskii, F N Dul'tsev Kinet. Katal. 25 530 (1984) a 51. R Robertson, A Gallagher J. Chem. Phys. 85 3623 (1986) 52. A Doncaster, R Walsh Int. J. Chem. Kinet. 13 503 (1981) 53. J A Seetula, Y Feng, D Gutman, P W Seakins,M J Pilling J. Phys. Chem. 95 1658 (1991) 54. J Berkowitz, J P Greene, H Cho, B RusÆ cÆ ic' J. Chem. Phys. 86 1235 (1987) 55. A F Sax, J Kalcher J. Phys. Chem. 95 1768 (1991) 56. J A Pople, B T Luke, M J Frisch, J S Binkley J. Phys. Chem. 89 2198 (1985) 57. P Ho,M E Coltrin, J S Binkley, C F Melius J. Phys. Chem. 89 4647 (1985) 58. P Ho, C F Melius J. Phys. Chem. 94 5120 (1990) 59. M S Gordon, T N Truong, E K Bonderson J.Am. Chem. Soc. 108 1421 (1986) 60. B S Meyerson, J M Jasinski J. Appl. Phys. 61 785 (1987) 61. J Doyle, R Robertson, G H Lin,M Z He, A Gallagher J. Appl. Phys. 64 3215 (1988) 62. A Tachibana, K Yamaguchi, S Kawauchi, Y Kurosaki, T Yamabe J. Am. Chem. Soc. 114 7504 (1992) 63. S K Loh, J M Jasinski J. Chem. Phys. 95 4914 (1991)Mechanism of thermal decomposition of silanes 64. M S Gordon, D R Gano, J S Binkley, M J Frish J. Am. Chem. Soc. 108 2191 (1986) 65. M E Coltrin, R J Kee, J A Miller J. Electrochem. Soc. 133 1206 (1986) 66. J Espinosa-Garsta, J Sanson, J C Corchado J. Chem. Phys. 109 466 (1998) 67. A A Onischuk, A I Levykin, V P Strunin, M A Ushakova, R I Samoilova, K K Sabelfeld, V N Panfilov J.Aerosol Sci. 31 879 (2000) 68. G Raabe, J Michl Chem. Rev. 85 419 (1985) 69. J H Purnell, R Walsh Chem. Phys. Lett. 110 330 (1984) 70. M A Ring, H E O'Neal J. Phys. Chem. 96 10848 (1992) 71. M E Coltrin, R J Kee, G H Evans J. Electrochem. Soc. 136 819 (1989) 72. P Ho,M E Coltrin, J S Binkley, C F Melius J. Phys. Chem. 90 3399 (1986) 73. P M Agrawal, D L Thompson, L M Raff J. Chem. Phys. 92 1069 (1990) 74. H K Moffat, K F Jensen, R W Carr J. Phys. Chem. 96 7695 (1992) 75. H K Moffat, K F Jensen, R W Carr J. Phys. Chem. 96 7683 (1992) 76. M Bowrey, J H Purnell Proc. R. Soc. Lond. A, Math Phys. Sci. 321 341 (1971) 77. C J Giunta, R J McCurdy, J D Chapple-Sokol, R G Gordon J. Appl. Phys. 67 1062 (1990) 78. J E Johannes, J G Ekerdt J.Electrochem. Soc. 141 2135 (1994) 79. R J Bogaert, B N Baron,M T Klein, T W F Russell, R E Rocheleau J. Electrochem. Soc. 136 2960 (1989) 80. P A Breddels, H Kanoh, O Sugiura,M Matsumura Jpn. J. Appl. Phys. 29 L1750 (1990) 81. W A P Claassen, J Bloem, W G J N Valkenburg, C H J van den Brekel J. Cryst. Growth 57 259 (1982) 82. J H Comfort, R Reif J. Electrochem. Soc. 136 2386 (1989) 83. T J Donahue, R Reif J. Electrochem. Soc. 133 1691 (1986) 84. D W Foster, A J Learn, T I Kamins J. Vac. Sci. Technol. B 4 1182 (1986) 85. G Harbeke, L Krausbauer, E F Steigmeier, A E Widmer, H F Kappert, G Neugebauer J. Electrochem. Soc. 131 675 (1984) 86. M K Hatalis, D W Greve J. Appl. Phys. 63 2260 (1988) 87. F Hottier, R Cadoret J. Cryst. Growth 52 199 (1981) 88.T I Kamins, M M Mandurah, K C Saraswat J. Electrochem. Soc. 125 927 (1978) 89. H Kanoh, O Sugiura,M Matsumura Jpn. J. Appl. Phys. 32 2613 (1993) 90. H Kanoh, O Sugiura, P A Breddels,M Matsumura IEEE Electron Device Lett. 11 258 (1990) 91. B S Meyerson, B A Scott, R Tsui Chemtronics 1 150 (1986) 92. A M Beers, J Bloem Appl. Phys. Lett. 41 153 (1982) 93. W L M Weerts, M H J M de Croon, G B Marin J. Electrochem. Soc. 145 1318 (1998) 94. Y Mishima, M Hirose, Y Osaka, K Nagamine, Y Ashida, N Kitagawa, K Isogaya Jpn. J. Appl. Phys. 22 L46 (1983) 95. R J Buss, P Ho,W G Breiland,M E Coltrin J. Appl. Phys. 63 2808 (1988) 96. D Meakin, K Papadopoulou, S Friligkos, J Stoemenos, P Migliorato, N A Economou J. Vac. Sci. Technol. B 5 1547 (1987) 97.R S Rosler Solid State Technol. 20 (4) 63 (1977) 98. Y M Wu, R M Nix Surf. Sci. 306 59 (1994) 99. M Sadamoto, J H Comfort, R Reif J. Electron. Mater. 19 1395 (1990) 100. C C Cheng, J T Yates Jr Phys. Rev. B, Condens. Matter 43 4041 (1991) 101. R Jones Physica B 170 181 (1991) 102. S M Gates, C M Chiang Chem. Phys. Lett. 184 448 (1991) 103. J J Boland Phys. Rev. B, Condens. Matter 44 1383 (1991) 104. U Memmert, A Berko, R J Behm Surf. Sci. 323 L441 (1995) 105. Y Wang,M J Bronikowski, R J Hamers Surf. Sci. 311 64 (1994) 106. D-S Lin, E S Hirschorn, T-C Chiang, R Tsu, D Lubben, J E Greene Phys. Rev. B, Condens. Matter 45 3494 (1992) 107. D Albertini, F Thibaudau, L Masson, F Salvan Surf. Sci. 350 L216 (1996) 108. M Fehrenbacher, J Spitzmuller,M Pitter, H Rauscher, R J Behm Jpn.J. Appl. Phys. 36 3804 (1997) 109. M Suemitsu, H Nakazava, N Miyamoto Surf. Sci. 357 ± 358 555 (1996) 331 110. H Hirayama, T Tatsumi, N Aizaki Appl. Phys. Lett. 52 1484 (1988) 111. K Werner, S Butzke, S Radelaar, P Balk J. Cryst. Growth 136 338 (1994) 112. M Suemitsu, K-J Kim, H Nakazawa, N Miyamoto Appl. Surf. Sci. 107 81 (1996) 113. S M Gates, C M Greenlief, D B Beach, P A Holbert J. Chem. Phys. 92 3144 (1990) 114. Z Jing, J L Whitten Phys. Rev. B, Condens. Matter 44 1741 (1991) 115. L-Q Xia,M E Jones, N Maity, J R Engstrom J. Chem. Phys. 103 1691 (1995) 116. K J Uram, U J Jansson Surf. Sci. 249 105 (1991) 117. M Suemitsu, H Nakazawa, T Morita, N Miyamoto Jpn. J. Appl.Phys. 36 L625 (1997) 118. J Braun, H Rauscher, R J Behm Surf. Sci. 416 226 (1998) 119. Y Suda, D Lubben, T Motooka, J E Greene J. Vac. Sci. Technol. A, Vac. Surf. Films 8 61 (1990) 120. S M Gates, S K Kulkarni Appl. Phys. Lett. 58 2963 (1991) 121. D Lubben,R Tsu, T R Bramblett, J E Greene J. Vac. Sci. Technol. A, Vac. Surf. Films 9 3003 (1991) 122. F C Eversteijn Philos. Res. Rep. 26 134 (1971) 123. J U M S Murthy, N Miyamoto, M Shimbo, J Nishizava J. Cryst. Growth 33 1 (1976) 124. Z M Qian, H Michiel, A van Ammel, J Nijs, R Mertens J. Electrochem. Soc. 135 2378 (1988) 125. C H J van den Brekel, L J M Bollen J. Cryst. Growth 54 310 (1981) 126. J J Wu, R Flagan J. Appl. Phys. 61 1365 (1987) 127. M K Alam, R C Flagan Aerosol Sci. Technol.5 237 (1986) 128. J J Wu, H V Nguyen, R C Flagan Langmuir 3 266 (1987) 129. T Kojima, K Usui, T Furusawa J. Chem. Eng. Jpn. 22 683 (1989) 130. J H Flint, R A Marra, J S Haggerty Aerosol Sci. Technol. 5 249 (1986) 131. J H Flint, J S Haggerty Aerosol Sci. Technol. 13 72 (1990) 132. M Ehbrecht, F Huisken Phys. Rev. B, Condens. Matter 59 2975 (1999) 133. W R Cannon, S C Danforth, J S Haggerty, R A Marra J. Am. Ceram. Soc. 65 330 (1982) 134. J S Haggerty,W R Cannon, in Laser-Induced Chemical Processes (Ed. J I Steinfeld) (New York; London: Plenum, 1981) p. 183 135. R Fantoni, E Borsella, S Piccirillo, C A Nannetti, R Ceccato, S Enzo Appl. Surf. Sci. 43 308 (1989) 136. S N Rogak, R C Flagan, H V Nguyen Aerosol Sci. Technol. 18 25 (1993) 137.A Schmidt-Ott J. Aerosol Sci. 19 553 (1988) 138. A P Weber, U Baltensperger, H W GaÈ ggeler, A Schmidt-Ott J. Aerosol Sci. 27 915 (1996) 139. H D Jang, S K Friedlander Aerosol Sci. Technol. 29 81 (1998) 140. S Morooka, T Yasutake, A Kobata, K Ikemizu, Y Kato Int. Chem. Eng. 29 119 (1989) 141. S H Ehrmann, S K Friedlander,M Zachariah J. Aerosol Sci. 29 687 (1998) 142. S N Rogak, U Baltensperger, R C Flagan Aerosol Sci. Technol. 14 447 (1991) 143. A Kobata,K Kusakabe, S Morooka Am. Inst. Chem. Eng. J. 37 347 (1991) 144. K Okuyama, Y Kousaka, N Tohge, S Yamamoto, J J Wu, R C Flagan, J H Seinfeld Am. Inst. Chem. Eng. J. 32 2010 (1986) 145. A J Hurd, W L Flower J. Colloid Interface Sci. 122 178 (1988) 146. J E Martin, D W Schaefer, A J Hurd Phys.Rev. A, Gen. Phys. 33 3540 (1986) 147. G D Ulrich, J W Riehl J. Colloid Interface Sci. 87 257 (1982) 148. R J Samson,G W Mulholland, J W Gentry Langmuir 3 272 (1987) 149. S di Stasio J. Aerosol Sci. 30 325 (1999) 150. H X Zhang, C M Sorensen, E R Ramer, B J Olivier, J F Merklin Langmuir 4 867 (1988) 151. A A Onischuk, V P Strunin,M A Ushakova, V N Panfilov J. Aerosol Sci. 28 207 (1997) 152. A Yuuki, Y Matsui, K Tachibana Jpn. J. Appl. Phys. 26 747 (1987) 153. K J Sladek J. Electrochem. Soc. 118 654 (1971) 154. C S Herrick, D W Woodruff J. Electrochem. Soc. 131 2417 (1984) 155. F E Kruis, J Schoonman, B Scarlett, J. Aerosol Sci. 25 1291 (1994) 156. V P Pimenov, V I Chernyak, A I Gorbunov Khim. Fiz. 7 1520 (1988) a 157.J J Wu, H V Nguyen, R C Flagan, K Okuyama, Y Kousaka Am. Inst. Chem. Eng. J. 34 1249 (1988)332 158. G Lucovsky,W B Pollard, in The Physics of Hydrogenated Amorphous Silicon (II) (Eds J D Joannopoulous, G Lucovsky) (Berlin; Heidelberg; New York; Tokio: Springer, 1984) p. 86 159. M Cardona Phys. Status Solidi B 118 463 (1983) 160. G Lucovsky J. Non-Cryst. Solids 76 173 (1985) 161. G Lucovsky J. Non-Cryst. Solids 141 241 (1992) 162. W B Pollard, G Lucovsky Phys. Rev. B, Condens. Matter 26 3172 (1982) 163. M H Brodsky,M Cardona, J J Cuomo Phys. Rev. B, Condens. Matter 16 3556 (1977) 164. G Lucovsky, R J Nemanich, J C Knights Phys. Rev. B, Condens. Matter 19 2064 (1979) 165. U Jansson, K J Uram J. Chem. Phys. 91 7978 (1989) 166. Y Toyoshima, K Arai, A Matsuda, K Tanaka J.Non-Cryst. Solids 137/138 765 (1991) 167. N Maley, I Szafranek, L Mandrell,M Katiyar, J R Abelson, J A Thornton J. Non-Cryst. Solids 114 163 (1989) 168. E C Freeman, W Paul Phys. Rev. B, Condens. Matter 18 4288 (1978) 169. D Masson, L Paquin, S Poulin-Dandurand, E Sacher, A Yelon J. Non-Cryst. Solids 66 93 (1984) 170. A T M Wilbers, G J Meeusen,M Haverlag, G M W Kroesen, D C Schram, H Kersten Thin Solid Films 204 59 (1991) 171. D Dellafera, R Labusch, H H Roscher Philos. Mag. B, Phys. Condens. Matter. Electron. Opt. Magn. Prop. 45 607 (1982) 172. M J van den Boogaard, B G C van den Ree, H Meiling, R E I Schropp,W F van der Veg Physica B 170 281 (1991) 173. H Shanks, C J Fang, L Ley,M Cardona, F J Demond, S Kalbitzer Phys. Status Solidi B 100 43 (1980) 174. C J Fang, K J Gruntz, L Ley,M Cardona, F J Demond, G Muller, S Kalbitzer J. Non-Cryst. Solids 35/36 255 (1980) 175. A A Onischuk, V P Strunin,M A Ushakova, R I Samoilova, V N Panfilov Phys. Status Solidi B 193 25 (1996) 176. G Lucovsky, J Yang, S S Chao, J E Tyler,W Czubatyj Phys. Rev. B, Condens. Matter 28 3225 (1983) 177. G Lucovsky, J Yang, S S Chao, J E Tyler, W Czubatyj J. Non-Cryst. Solids 59/60 609 (1983) 178. G Lucovsky, S S Chao, J Yang, J E Tyler,W Czubatyj J. Vac. Sci. Technol. A, Vac. Surf. Films 2 353 (1984) 179. J S Knights, R A Street, G Lucovsky J. Non-Cryst. Solids 35/36 279 (1980) 180. P John, I M Odeh, M J K Thomas, M J Tricker, J I B Wilson, Phys. Status Solidi B 105 499 (1981) 181. G Lucovsky,W B Pollard J. Vac. Sci. Technol. A, Vac. Surf. Films 1 313 (1983) 182. G Lucovsky Solid State Commun. 29 517 (1979) 183. M Pinarbasi,N Maley,A Myers, J R Abelson Thin Solid Films 171 217 (1989) 184. S Furukawa, N Matsumoto, T Toriyama, N Yabumoto J. Appl. Phys. 58 4658 (1985) 185. W Beyer, H Wagner J. Non-Cryst. Solids 59/60 161 (1983) 186. S R Das, S Charbonneau, D F Williame, J B Webb, J R MacDonald, D R Polk, S Zukotynski, J Perz Can. J. Phys. 63 852 (1985) 187. J A McMillan, E M Peterson J. Appl. Phys. 50 5238 (1979) 188. D K Biegelsen, R A Street, C C Tsai, J C Knights Phys. Rev. B, Condens. Matter 20 4839 (1979) 189. M H Brodsky Thin Solid Films 50 57 (1978) 190. S Oguz,M A Paesler Phys. Rev. B, Condens. Matter 22 6213 (1980) 191. F Kozlowski, V Petrova-Koch, A Kux,WStadler, A Fleischmann, H Sigmund J. Non-Cryst. Solids 137/138 91 (1991) 192. W Beyer, H Wagner J. Appl. Phys. 53 8745 (1982) 193. H Fritzsche,M Tanielian, C C Tsai, J Gaszi J. Appl. Phys. 50 3366 (1979) 194. P John, I M Odeh, M J K Thomas, M J Tricker, F Riddoch, J I B Wilson Philos. Mag. B, Phys. Condens. Matter. Electron. Opt. Magn. Prop. 42 671 (1980) 195. K Sinniah,MG Sherman, L B Lewis,WH Weinberg, J T Yates Jr, K C Janda J. Chem. Phys. 92 5700 (1990) 196. S M Gates Surf. Sci. 195 307 (1988) 197. S M Gates, C M Greenlief, D B Beach J. Chem. Phys. 93 7493 (1990) 198. C M Greenlief, S M Gates, P A Holbert Chem. Phys. Lett. 159 202 (1989) A A Onischuk, V N Panfilov 199. S M Gates, R R Kunz, C M Greenlief Surf. Sci. 207 364 (1989) 200. B G Koehler, C H Mak, D A Arthur, P A Coon, S M George J. Chem. Phys. 89 1709 (1988) 201. M L Wise, B G Koehler, P Gupta, P A Coon, S M George Surf. Sci. 258 166 (1991) 202. P Gupta, V L Colvin, S M George Phys. Rev. B, Condens. Matter 37 8234 (1988) 203. M A Petrich Magn. Reson. Rev. 16 183 (1993) 204. J B Boyce, S E Ready Physica B 170 305 (1991) 205. H Eckert Progr. Nucl. Magn. Reson. Spectrosc. 24 159 (1992) 206. A A Onischuk, R I Samoilova, V P Strunin, E N Chesnokov, R N Musin, V S Bashurova, A G Maryasov, V N Panfilov Appl. Magn. Reson. 15 59 (1998) 207. W B Jackson, S B Zhang Physica B 170 197 (1991) 208. N M Johnson J. Non-Cryst. Solids 137/138 11 (1991) 209. F Buda, G L Chiarotti, R Car, M Parinello Physica B 170 98 (1991) 210. A H Mahan, D L Williamson, B P Nelson, R S Crandall Phys. Rev. B, Condens. Matter 40 12024 (1989). 211. M A Robertson, S K Estreicher Phys. Rev. B, Condens. Matter 49 17040 (1994) a�Chem. Phys. Rep. (Engl. Transl.) b�Kinet. Catal. (Engl. Transl.) c�Inorg. Mater. (Engl. Transl.) d�Russ. J. Inorg. Chem. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (4) 333 ± 355 (2001) Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors O N Zefirova, N S Zefirov Contents I. Introduction II. Classification of serotonin receptors III. Ligands of serotonin 5-HT1A subtype receptors IV. Ligands of serotonin 5-HT1B and 5-HT1D subtype receptors V. Ligands of serotonin receptors of other 5-HT1 subtypes VI. Ligands of serotonin 5-HT2 subtype receptors VII. Ligands of serotonin 5-HT3 subtype receptors VIII. Ligands of serotonin 5-HT4 subtype receptors IX. Ligands of serotonin receptors of other subtypes Abstract. active compounds organic of structures the on data The The data on the structures of organic compounds active with respect to serotonin (5-hydroxytryptamine) receptors are with respect to serotonin (5-hydroxytryptamine) receptors are systematised.Various aspects of their design are considered. The systematised. Various aspects of their design are considered. The bibliography includes 296 references bibliography includes 296 references. I. Introduction The problems connected with the design of physiologically active compounds are discussed in the literature in terms of medicinal chemistry;{ therefore, they are usually not very well known to a broad circle of organic chemists. The present review is aimed at the systematisation of the published data on the design of physiolog- ically active compounds interacting with serotonin receptors and the analysis of effects of their chemical structures on their activities.It is known that the physiological role of serotonin (5-hydr- oxytryptamine) (1, 5-HT) in the human brain consists in the regulation of various psychoemotional reactions. In addition, serotonin controls thermal regulation, sensory perception (partic- ularly, pain sensitivity), etc. Recent studies have shown that disturbances in serotonin metabolism and/or functional activities of serotonin regulators are related to the pathogenesis of schizo- phrenia, depressive and anxiety states, alcoholism, drug abuse, etc. (CH2)2NH2 HO NH 1 These findings have stimulated the interest that emerged over the past decade in the design of chemical compounds involved in the neurotransmission mediated by serotonin.O N Zefirova, N S Zefirov Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 30 26. Tel. (7-095) 939 48 78. E-mail: olgaz@org.chem.msu.ru (O N Zefirova) Tel. (7-095) 939 16 20. E-mail: zefirov@org.chem.msu.ru, zefirov@org.chem.msu.su (N S Zefirov) Received 22 January 2001 Uspekhi Khimii 70 (4) 382 ± 407 (2001); translated by R L Birnova #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n04ABEH000654 333 333 334 338 341 342 346 349 350 We hope that this review, which generalises structural approaches to the design of compounds interacting with serotonin receptors, will inspire synthetic chemists to make a radically new contribution to this branch of medicinal chemistry.II. Classification of serotonin receptors Serotonin (1) is a biogenic amine which manifests an extremely broad range of biological activities with respect to the central nervous system and other body organs and tissues. In the human central nervous system, it acts as a neurotransmitter, i.e., a chemical compound which transmits information from one nerve cell to another.3 In the course of transmission, neuro- transmitter molecules are released into the space separating membranes of the interacting cells where they bind to the corresponding receptors on the surfaces of target cells. It is this event that represents perception of information. The specificity of neurotransmitter ± receptor interactions is determined by the structures of both the receptor and the neurotransmitter.Serotonin receptors represent proteins incorporated into cell plasmatic membranes, which transform molecular signals from serotonin and its analogues into specific cell responses. The capability of chemical compounds of binding to their specific receptors underlies their effects on serotoninergic transmission. The effects of the majority of compounds presently used for the treatment of mental disorders, anxiety, schizophrenia, obesity, migraine, etc., are thought to be based on their involvement in serotoninergic transmission. These compounds act as agonists (activators) or antagonists (blockers) of the corresponding recep- tors. It should be noted that the interactions of chemical com- pounds with receptors are considered from the standpoint of both affinity (i.e., binding ability) and intrinsic activity (i.e., the ability to produce a biological response by means of structural or conformational changes in the receptor).(Affinity is numerically expressed by the reverse value of the equilibrium dissociation constant of the ligand ± receptor complex. Apparently, antago- nists possess no intrinsic activities.) The existence of partial { The subject of medicinal chemistry is the search for, and the structural design 1, 2 of, physiologically active compounds (drugs), the elucidation of chemical structure ± activity relationships and, finally, the solution of the opposite structure ± activity problem, viz., the design of compounds with predetermined properties.334 agonists, i.e., those which cannot produce maximum activation of their specific receptors and, as a consequence, maximum bio- logical effect irrespective of their concentrations, is also possible.The discovery of an immense variety of serotonin receptors of different subtypes in the past decade has brought about significant changes in their classification and nomenclature. Here, we shall adhere to the classification proposed by a special IUPAC com- mission according to which the 14 presently known subtypes of serotonin receptors are further divided into seven groups (5HT1 ± 5HT7) depending on their amino acid sequences and the similarity of their signal transmission mechanisms.4 ±6 Activation of serotonin receptors triggers the mechanism which involves a cascade of membrane proteins interacting sequentially with one another.In a definite step, signal transmission involves secondary messengers, i.e., molecules or ions endowed with the ability of inducing conformational changes in the proteins involved in specific cellular processes. 5HT3 receptors which form ionic channels and transmit signals directly from nervous cells by ionic currents are the only exception. The discovery of various subtypes of serotonin receptors led to the conclusion that the majority of presently known drugs, influencing serotoninergic transmission, act simultaneously on several subtypes of 5-HT receptors. Therefore, the design of selective ligands aimed at more efficient treatment with minimum side effects has been and still remains the problem of paramount importance, to say nothing of the fact that the design of selective ligands may culminate in the elaboration of radically new treat- ment schedules.III. Ligands of serotonin 5-HT1A subtype receptors High concentrations of 5-HT1A subtype serotonin receptors were found in the brain cortex where they are thought to play a crucial role in the processes associated with emotions. It has presently been demonstrated that these receptors are involved in the patho- genesis of various mental diseases, whereas ligands (especially agonists) interacting with these receptors were recommended as candidates for the treatment of various anxiety states, phobias, depressions, Alzheimer's disease (senile dementia), as analgesics, etc.7± 9 It is of note that 5-HT1A receptors display a high degree of similarity to other receptor systems of the organism, such as adrenergic and dopaminergic systems, which significantly com- plicates the search for their selective ligands.The simplest structural modifications of the serotonin mole- cule, viz., the introduction of substituents into the benzene ring or restriction of the conformational mobility of the side chain, afforded several high-affinity 5-HT1A agonists, e.g., amines 2 and 3. However, like serotonin itself, these compounds are not selective and interact with different subtypes of serotonin recep- tors (predominantly, with 5-HT1B, 5-HT1D and 5-HT2C).10, 11 NH (CH2)2NH2 H2NC(O) MeO NH 2 NH 3 8-Hydroxy-2-dipropylaminotetralin (4a) was among the first most active and selective agonists of 5-HT1A receptors,12, 13 the (+)-(R)-enantiomer being fully agonist and manifesting the high- est affinity.The main structure ± activity regularities for com- pounds of the type 4 are as follows. The presence of the hydroxy group in position 8 is essential for the binding and provides 500- fold higher selectivity with respect to serotonin receptors in comparison with dopamine receptors (the shift of the OH group into positions 5 or 7 sharply decreases the selectivity), while the hydroxy group in position 8 may be replaced by OMe, Ac or other groups (but not the carboxy group) without any loss in the affinity and selectivity, although the agonistic activity sometimes O N Zefirova, N S Zefirov decreases noticeably, as in substitution by Ac.14 The presence of two propyl substituents at the nitrogen atom is an optimum for affinity, selectivity and agonistic activity, although some varia- tions in substituents, e.g., replacement of one Prn group by CH2CH=CHI (compound 4b),15 enhance significantly the bind- ing of the ligand to the receptor.Prn HO R1 N R3 R2 R1=R2=H: R3=Prn (a), CH2CH=CHI (b); R1=Me, R2=H, R3=Prn (c); R1=H, R2=Me, R3=Prn (d). 4a ± d The introduction of small substituents (e.g., Me) into posi- tions 1 (compound 4c) 16 or 3 (compound 4d) 17 of the alicyclic fragment of the molecule results in compounds manifesting differ- ent activities. The tetralin (1S,2R)-isomer 4c displays nearly the same affinity as compound 4a, but is much inferior to the latter with regard to intrinsic activity and is a weak partial agonist.The derivatives containing a methyl group in position 3 bind less effectively to 5-HT1A receptors, although some of them, e.g., the (2S,3S)-isomer of the tetralin 4d, manifest high intrinsic activ- ities.18 The transition to tricyclic structures afforded several potent 5-HT1A agonists (e.g., the trans-isomer 5),19 whereas the synthesis of rigid octahydrobenzo[g]quinolines yielded weak 5-HT1A ago- nists, among which the amine 6 was the most active compound.17 Notice that these cyclic analogues were used for the preparation of pharmacophore models of compounds interacting with 5-HT1A receptors.{ Prn N NROH OH 6 5 And, finally, it should be noted that the introduction of a fluorine atom into position 5 of compound 4a, which gives compound 7, changes quite unexpectedly the binding pattern.Thus the (R)-isomer of the tetralin 7, is a full agonist and is somewhat less active than compound 4a, while the (S)-isomer of compound 7 is a strong antagonist of 5-HT1A receptors and its affinity is nearly an order of magnitude lower than that of the (R)-enantiomer.20 ± 22 The amine 4a was used as a prototype for the design of yet another structural class of 5-HT1A agonists, viz., trans-(arylcyclo- propyl)amines 8.16, 23 The analysis of the structure ± activity relationships in this series of compounds revealed that the introduction of electron-withdrawing substituents into the phenyl ring decreases the affinity for 5-HT1A receptors, whereas electron donor or aromatic substituents (with the exception of bulky ones), e.g., 2- or 3-thienyl substituents, enhance the affinity.Unlike the aminotetralin 4a, compounds of this series are highly stereo- selective, only (1R,2S)-isomers manifesting high activities. OH NPrn H 2 R NPrn2 H 8 F 7 Yet another class of 5-HT1A agonists, viz., (R)-apomorphine analogues structurally close to aminotetralins 4 are known. { The pharmacophore entails specific spatial arrangement of the binding groups essential for the optimum interaction with the receptor and causing (or blocking) a specific biological reaction.It is a purely structural concept which is neither a genuine molecule nor a genuine association of functional groups.Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors (R)-Apomorphine is a non-selective agonist of dopamine recep- tors. Structure ± activity studies of its derivatives resulted in compounds which do not interact with dopamine receptors but are selective towards serotonin receptors. 11-Hydroxy-10-meth- ylaporphine (9) with a methyl group in position 10 (instead of the hydroxy group in aporphine) is the most potent and selective compound. It is of note that only the (R)-enantiomer manifests agonistic activity, whereas the variations of substituents in posi- tions 10, 11 and at the nitrogen atom reduce this activity in comparison with the amine 9.These findings suggest that a hydrogen bond is formed between the hydroxy group in position 11 and serine-168 in the receptor protein and that a `pocket' for the methyl group in position 10 exists in the binding site of the 5-HT1A receptor.24, 25 2 OH 11 Me 10 8 NMe 9 Compounds 10 ± 12 provide other examples of cyclic systems that interact with 5-HT1A receptors. Although these compounds are not selective, being active with respect to other receptors, they have been used for the construction of computer models of the ligand-binding site of 5-HT1A receptors.26 ± 28 NPrn2 O HN N HN N HN 12 O 11 10 NPrn NEt Other structural variations in compound 4a included the introduction of heteroatoms into the bicyclic fragment.For example, the 5-methoxy derivative 13 manifests the same high reactivity as aminotetralin (4a) and, which is more important, is more selective towards the 5-HT1A receptor.29 The main struc- ture ± affinity relationships for this class of compounds can be summarised as follows. Tertiary amines manifest higher affinity for the receptor than secondary and primary amines; substituents at the nitrogen atom can be rather bulky and the presence of MeO, CO2Me and CONH2 groups in position 5 provides both tight binding and the selectivity towards 5-HT1A receptors.30 OMe NPrn2 O 13 Buspirone derivatives constitute a large and very important structural class of 5-HT1A receptor ligands.31 ± 33 Buspirone (14a) manifests high affinity for 5-HT1A receptors and is endowed with agonistic activity.This is the only known compound specific for this subtype of receptor that has been recommended for clinical application as an antidepressant drug. Structural modifications of the imide and piperazine fragments and variations in the lengths of the alkyl chains were carried out aimed at a search for more active and selective buspirone analogues. These studies showed that the imide fragment of compound 14a can be replaced by bioisosteric groups,} e.g., like those in gepirone (14b), ipsapirone (14c), tandospirone (14d), etc.34 ± 36 These groups should be lipo- philic,37 ± 39 although the role of steric effects may also be significant.40 Interestingly, some compounds, e.g., substituted }A bioisoster is a compound prepared by the replacement of an atom or a group of atoms (e.g., a functional group) by other structural fragments with `preserved' biological activity.The bioisosteric equivalence rules were established experimentally and are widely used in medicinal chemistry. 335 piperazine (15),41 devoid of the cyclic amine fragment in the side chain, also manifest agonistic activities towards 5-HT1A receptors. Heterocyclic substituents at the second nitrogen atom of the piperazine rings of buspirone and its analogues can be replaced by the phenyl substituent either non-substituted or containing an electron-withdrawing substituent in the o- or m-position to ensure higher affinity, e.g., compound 16.42, 43 N N N(CH2)4R N 14a ± d O O O O Me (d).(c), N (b), N (a), N R=N Me S O O O O O O MeO NC6H13-n 3-ClC6H4N N N(CH2)4N 16 15 O The optimum number of methylene groups in the linker needed for manifestation of maximum affinities of buspirone analogues for 5-HT1A receptors is four, although some com- pounds with three and even two (less frequently) methylene groups also manifest noticeable affinities, e.g., compound 17.44 ± 46 Moreover, high affinity can persist upon substitution of heteroatoms for some methylene groups in the chain as is the case with ether 18 47, 48 or amide 19.49 Variations in the chain length and structure strongly affect its selectivity towards serotonin receptors in comparison with adrenoceptors and dopaminergic receptors.50, 51 Thus the selectivities of the bicyclohydantoin derivatives 20a,b for 5-HT1A can be improved by reducing the length of the intermediate alkyl chain to one methylene group (n=1), although such modification causes a significant decrease in their affinities.52 O N 2-MeOC6H4 N7(CH2)27NO 17 OO N N7(CH2)3O 18 N N7(CH2)27NHC(O)Ph 19 O X NC6H4R N7(CH2)n7N N 20a,b O n=1±4;X=CH2 (a), (CH2)2 (b).Computer simulation studies aimed at rationalisation of such an increase in the selectivities of the bicyclohydantoin derivatives 20a,b revealed 53 that compounds with n=1 and n=4 bind differently to the receptor.The fact that the derivatives with n=1 manifest agonistic activities towards 5-HT1A receptors (albeit rather weakly) suggests 53 that the protein molecule has a `non-pharmacophoric' pocket (i.e., the site docking `non-essen- tial' fragments of the ligand molecule rather than the pharmaco- phore group), which docks the imide group of the ligand336 containing one methylene bridge. Its selectivity is attributed to the fact that adrenoceptors have no such pocket. It is noteworthy that the majority of buspirone analogues mentioned above and buspirone itself, despite their high affinities for 5-HT1A receptors, do not manifest any high intrinsic activities and are thus partial agonists. Moreover, some of them, e.g., compounds 17 and 18 or the succinimide analogue 21 54, 55 are sometimes classified as antagonists due to their very low intrinsic activities. ON(CH2)4N N C6H4OMe-2 21 O Studies aimed at the design of 5-HT1A agonists containing simultaneously the groups constituting the essential fragments of the two main classes of 5-HT1A ligands, viz., aminotetralin (4a) and buspirone (14a) analogues, have become very popular in the past decade.An imide or a bioisosteric group, such as substituents R in buspirone (14a), gepirone (14b), ipsapirone (14c), etc., were attached to the amino group of the benzopyran fragment of compound 13 through an alkyl chain, resulting in compounds 22. Studies of structure ± affinity relationships of these com- pounds revealed that compounds 22a ± c containing imide or sulfanylamide substituents manifested the highest activities and selectivities, whereas the preferred length of their side chains was four methylene groups, although sulfonylamide derivatives are also active even if their chains contain two methylene groups.56 OMe N(Prn)(CH2)3ZR O 22a ± c O O (a), N (b); Z=CH2: R=NO O Z=CH2NHSO2, R = C6H4Me-4 (c).High activities and selectivities of analogues of compounds 22 with limited mobilities of their side chains due to incorporation of the alkylamino groups into ring structures were predicted using the pharmacophore model of 5-HT1A receptor agonists 17 and QSAR.} It should be noted that in many cases the restriction of conformational mobilities of functional groups is a classical approach in medicinal chemistry, since this allows stabilisation of the conformation which is the most suitable for the interaction with the receptors.The spiropyrrolidine- and spiropiperidineben- zopyrans 23 and 24 synthesised manifest high agonistic activities and selectivities with respect to 5-HT1A receptors, though not exceeding those of compounds 22a ± c. In our opinion, it is expedient to carry out a detailed analysis of other types of restriction of conformational mobility in the side chains of such compounds. It is also noteworthy that compounds 23 and 24 are partial agonists of 5-HT1A receptors and manifested the proper- ties of antidepressant and anxiolytic drugs in vivo.60 OMe O N N (CH2)4 O O 23 }QSAR is the abbreviation for `quantitative structure ± activity relation- ships'.This branch of medicinal chemistry is aimed at establishing the relationships between the properties (activities) and structural parameters of compounds (see, e.g., Refs 2, 57 ± 59). O N Zefirova, N S Zefirov OMe O N N (CH2)4 O O 24 Benzodioxane derivatives which constitute another structural class of effective 5-HT1A receptor ligands are being studied intensively. Compound 25 is one of the most potent and selective partial agonists.61 ± 63 O O O CH2NH(CH2)NO 25 O HO CH2NH(CH2)3OAr O 26 Ar=C6H4R, Het. In such structures, as in any other buspirone (14a) analogues, the bridging chain may be elongated and modified, as in the case of the high-affinity partial agonists 26, although such modification sometimes causes significant reduction in their selectivities.64 Therefore, the search for structural modifications of benzodiox- anes which might afford compounds endowed with the ability to distinguish among the binding sites of adrenoceptors and dop- aminergic 5-HT1A receptors was continued.It was found 65 that benzopyran (27), the unsaturated analogue of benzodioxanes, manifests high affinity for 5-HT1A receptors and partial agonistic activity; the presence of a double bond in its ring and the absence of the oxygen atom in position 1 strongly depress its binding to adrenoceptors, which makes this compound more selective with respect to serotonin receptors.The acyclic benzodioxane analogue (compound 28) is, contrariwise, less selective towards 5-HT1A receptors than compounds 26, although they retain their high affinities for these receptors. Ph O CH2NH(CH2)2OC6H3(OMe)2-2,6 27 OBn O(CH2)2NH(CH2)2OC6H3(OMe)2-2,6 28 Benzodioxane derivatives with substituents in the benzene rings, e.g., flesinoxan (29), also manifest high selectivities and agonistic activities.66 Studies of structure ± affinity relationships revealed that the main contribution to the binding of compound 29 to the receptor is made by the lipophilic substituent at the N(4) atom (the 4-fluorophenyl fragment may be replaced by the phenyl, thiophene or cyclohexane fragment without any loss in activity, but not by polar five- or six-membered heterocyclic substituents).Also, the amide group hardly takes part in the binding to the receptor but, rather, acts as a bridge and may therefore be replaced by other fragments. Structural modifications may also involve the CH2OHO O N N(CH2)2NHC(O)C6H4F-4 29Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors aryl substituent at N(1); these affect not only the affinities of the analogues but also their selectivities in comparison with dopamine receptors.66 An intensive search for 5-HT1A receptor ligands characterised by high intrinsic activities is now in progress, since the majority of the known ligands are partial agonists. Full 5-HT1A agonists were found in a new class of compounds, viz., 6-substituted pyridylme- thylamine derivatives 30.These ligands were selected using a special model of 5-HT1A receptor activation according to which the intrinsic activities of 5-HT1A ligands depend on their abilities to stabilise the positive charge generated on the receptor protein upon its binding to the agonist. The pyridine ring was chosen as a group endowed with this ability. Therefore, the pyridine deriva- tives of the type 30 were constructed in order to obtain full agonists of 5-HT1A receptors. NC(O)C6H4Cl2-3,4 N R1 CH2NHCH2 R2 30a,b N , R2 = H (b). R1= N, R2 = H (a); R1= O The results of QSAR studies suggest that the abilities of compounds to activate the corresponding receptors strongly depend on the nature and position of substituents in the pyridine ring.Thus a substituent (preferrably an electron-donor substitu- ent) should be present in position 6. Compounds 30a,b were finally found to be full agonists of 5-HT1A receptors and possess the highest affinities and selectivities.67 Moreover, the agonistic activities and selectivities of these compounds and some of their pharmacological characteristics can be improved by introducing the fluorine atom (R2=F).68 As shown above, structure ± activity studies of 5-HT1A recep- tor ligands are mainly directed at the search for selective com- pounds. However, special mention should be made of other relevant tasks, such as the search for compounds acting simulta- neously as dopaminergic receptor antagonists and serotonin 5- HT1A agonists. These compounds hold much promise in the therapy of schizophrenia and other mental disorders; some of them, e.g., the piperazine derivative 31, are undergoing clinical trial.69 Certain structurally related compounds, e.g., ziprasidone (32), manifest antagonistic activities towards 5-HT2 receptors.70 S N N(CH2)4NHCOC6H4NH2-2 N 31 O S NH N N(CH2)2 N Cl 32 Since this review is aimed at the structural analysis of serotonin receptor ligands, their affinities, intrinsic activities and selectivities, we shall not dwell on the search for ligands for the related types of receptors, although in these studies 5-HT-selective ligands can also be found.Thus the search for ligands of dopaminergic receptors revealed that compound 33 possessed affinity and selectivity exclusively with respect to 5-HT1A recep- tors.71 N N NH N N(CH2)2 N N 33 337 To complete the consideration of problems related to the design of 5-HT1A agonists, it should be noted that although such ligands are rather numerous, all of them belong to a limited set of structural classes of compounds.Moreover, only some of these ligands interact with 5-HT1A receptors in a selective manner and very few of them are full agonists. In this context, a search for agonists among novel classes of compounds is a very important task, since it helps elucidation of the structure ± activity relation- ships for ligands and sheds more light on the structural peculiar- ities and functional roles of the corresponding receptors.Antagonists of 5-HT1A receptors have been studied in lesser detail, although they could in principle be used for the treatment of various diseases of the central nervous system. Some presently known b-adrenoceptor blockers, e.g., pindolol (34),72, 73 cyano- pindolol (35),74, 75 etc., manifest antagonistic activities towards 5-HT1A receptors, the (7)-(S)-enantiomers being more active. These compounds are also active with respect to other subtypes of serotonin receptors. But Pri O O HN HN OH OH NC HN HN 35 34 Numerous studies aimed at a search for selective antagonists of 5-HT1A receptors have led to compounds 7 and 36.76, 77 It was also found that many compounds previously assigned to 5-HT1A antagonists, e.g., the piperazine derivative 37,78 spiroxatrine 38,79, 80 etc., are in fact not genuine antagonists but manifest partial agonistic and low intrinsic activities.{ It should be noted that the main approach to the design of compounds 36 ± 38 consists in the introduction of bulky lipophilic substituents into the structures of 5-HT1A receptor agonists.This approach is widely used in medicinal chemistry for the construction of antagonists, since the presence of additional hydrophobic inter- actions between the ligand and the receptor in the vicinity of the binding site increases the strength of binding to such an extent that the receptor appears to be completely blocked by the compound. N N(CH2)2NC(O) N 2-MeOC6H4 36 N NCH2CHC(O)NHBut 2-MeOC6H4 Ph 37 Ph N O CH2N NH O 38 O Until recently, there was no evidence in the literature concern- ing the differences in the structural requirements for agonists or antagonists of 5-HT1A receptors.18 Special mention should be made of investigations into effects of substituents on the intrinsic activities of chromanes of the type 39 which, like benzodioxanes 26, represent high-affinity 5-HT1A ligands.81, 82 Modifications of the benzene ring in their side chains revealed that the p-methoxy group plays a crucial role in both affinity and antagonistic activity of compound 39a, whereas other substituents diminish both the { The term `silent antagonist' was specially introduced to distinguish among genuine antagonists, i.e., those devoid of agonistic activity, and partial agonists.338 binding and the blocking properties.Modifications of the chro- mane ring showed that the introduction of F, Cl andOMe (but not OH and Me) into position 6 results in strong antagonists (com- pounds 39b,c,d), whereas the introduction of analogous substitu- ents into positions 5 and 7 yields partial agonists. The differences in the antagonistic activities of this series of compounds cannot be explained by different distribution of their electron densities, since the introduction of both electron-donor (e.g., OMe) and electron- withdrawing groups (e.g., Cl) affords full antagonists, whereas compound 39e containing the hydroxy group manifests the properties of a partial agonist.7 O(CH2)2NH(CH2)4C6H4OMe-4 6 R 5 O 39a ± e R = H (a), 6-F (b), 6-Cl (c), 6-OMe (d), 6-OH (e). In all probability, the antagonistic properties of the majority of structural classes of 5-HT1A ligands depend on many factors and not necessarily depend on those factors which determine the strength of binding of the ligand to the receptor, i.e., affinity. This circumstance and the fact that even minor structural modifica- tions of 5-HT1A receptor antagonists often cause significant reduction in their affinities, selectivities and antagonistic activities call for further investigations using computer simulation of the receptor structure and ligand docking (in this context, docking is a molecular simulation technique aimed at a search for the optimum matching of the ligand to the binding site).83 On the whole, the search for selective antagonists of 5-HT1A receptors is a compli- cated but very urgent task, since no preparations of this class have been recommended for the clinical use thus far.IV. Ligands of serotonin 5-HT1B and 5-HT1D subtype receptors Serotonin 5-HT1B and 5-HT1D subtype receptors are rather densely localised in the human brain. They contain extremely similar amino acid sequences despite being encoded by two different genes. These two subtypes can pharmacologically be separated using several non-selective 5-HT2 antagonists (e.g., ketanserin, see below);53, 84 ± 86 however, their functional roles in the organism have not been established yet because only a very small number of selective ligands to these receptors are known. It is known that blocking of 5-HT1B receptors results in coronary vasodilation,87 therefore, their selective antagonists can be used for the treatment of various cardiovascular disorders.They are also recommended for use as effective fast-acting antidepressant drugs.88, 89 Agonists of 5-HT1D receptors are putative antimi- graine drugs, since their activation causes the brain vasodilation, diminishes pulsation, etc. In recent years, 5-HT1B and 5-HT1D subtype ligands have attracted considerable attention in connec- tion with the design of an efficient antimigraine drug sumatriptan (40), which appeared to be a 5-HT1B/1D receptor agonist.90 ± 93 Sumatriptan has very few side effects, but causes coronary vaso- dilation, which is controlled by 5-HT1B receptors (see above).This, too, calls for the design of selective agonists of 5-HT1B and 5-HT1D receptors. NH (CH2)2NMe2 HNO MeNHSO2CH2 NH 41 40 NH A fairly large group of 5-HT1B and 5-HT1D receptor agonists includes serotonin or tryptamine derivatives, such as amines 2 and 3, which represent non-selective agonists of 5-HT1B/1D and 5-HT1A receptors. Interestingly, the amine 41,94 ± 96 which is a close analogue of compound 3, is much more active and selective O N Zefirova, N S Zefirov towards the 5-HT1B receptor. Close serotonin analogues include sumatriptan (40) itself and its numerous derivatives, such as naratriptan (42),97, 98 rizatriptan (43),97, 99, 100 zolmitriptan (44),101, 102 eletriptan,103 avitriptan,104 almotriptan 97, 105 and fro- votriptan (45).85, 106 The majority of these compounds are candi- dates for antimigraine drugs and have passed clinical trials or are in the final stages.Their structures vary largely by changing substituents in the indole fragment or by restricting conforma- tional mobility of the side chain containing an amino group. NMe X N MeNHSO2 N N NH 43 NH 42 NHMe H2NC(O) X HN O O NH 44 45 NH X=(CH2)2NMe2 . Compounds 42 ± 45 possess high affinities for 5-HT1D recep- tors. However, they are non-selective and only 1.5 ± 3-fold less active with respect to 5-HT1B receptors. Comparison of the structures obtained suggest that the basic amine, the indole frag- ment and the substituent in position 5 involved in the formation of a hydrogen bond as an acceptor or a donor are the key groups needed for their efficient binding to the receptor.QSAR studies aimed at improving the activities and selectivities of serotonin and tryptamine derivatives towards 5-HT1D receptors included varia- tions of substituents at the C(5) atom. It was found 43 that in the case of alkoxy substituents the optimum length of the alkyl chain needed for the binding to 5-HT1B/1D receptors is 8 or 9 carbon atoms; with shorter chains, the selectivity is lost, while with longer chains, the affinity is decreased. Thus 5-nonyloxytryptamine (46) represents a high- affinity agonist, which is more selective towards 5-HT1B than towards 5-HT1D receptors and is practically inactive towards 5-HT1A receptors. It should be noted, however, that no com- pounds capable of distinguishing 5-HT1B and 5-HT1D receptors with a high degree of selectivity could thus far be obtained by changing the lengths and degree of ramification of the alkoxy substituents.43 (CH2)2NH2 C9H19O NH 46 The variations in the substituents in position 5 revealed that the latter can be rather bulky as in the case of compounds 47,88, 107 48 108 ± 110 and 49,111 ± 112 which are more selective towards 5-HT1D than towards 5-HT1B receptors.This is suggests the existence of a deep pocket in the binding site of 5-HT1D/1B receptor proteins adjacent to C(5) of the serotonin molecule.According to some data, it is in the size of this domain that 5-HT1D/1B receptors differ from the 5-HT1A receptor,113 although some of compounds mentioned, as expected, manifest low affinities for the latter. (CH2)2NH2 NC(O)CH2O N NH 47 MeSO2HNPhysiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors N O (CH2)2NH2 4-MeSO2NHC6H4CH2 N NH 48 (CH2)2NMe2 4-MeOC6H4CH2NHC(O) NH 49 It is noteworthy that the introduction of a triflate group (compounds 50), which is bioisosteric to the carboxy or the methoxy group,114 increases the selectivity towards 5-HT1D receptors 10 ± 15-fold in comparison with 5-HT1B receptors.115 (CH2)2NH2 (CH2)2NR2 But CF3SO3 NH NH 50 51 R=Me, Et.It is of note that until recently the ability of the substituent in position 5 to be involved in the formation of a hydrogen bond was thought to be a necessary prerequisite for high affinity, since virtually all presently known 5-HT1D receptor agonists have at least one heteroatom (N, O or S) in this position. However, the recently discovered 5-alkyltryptamine analogues devoid of hetero- atoms in position 5 possess high affinities for 5-HT1D receptors. This affinity increases as the size of the lipophilic group increases, reaching a maximum in the tert-butyl derivative 51.85 The result- ing compounds are not highly selective, viz., their maximum selectivities towards 5-HT1D receptors are only four times greater than towards 5-HT1B receptors.The fact that the formation of a hydrogen bond involving the substituent at the C(5) atom of serotonin is not critical for the manifestation of its high affinity for 5-HT1D receptors points to the hydrophobicity of the spacious pocket in the vicinity of C(5). This finding allows one to expand significantly the range of possible substituents at this atom. The variations in substituents in position 3 are much less documented. However, there are recent data to show that the selectivity towards 5-HT1B and 5-HT1D receptors can be enhanced by introduction of structurally more complex substituents in this position. Thus it was shown that the tryptamine analogue 52a containing the pyrrolidine ring possesses an order of magnitude higher affinity for 5-HT1D receptors than for 5-HT1B receptors.On the other hand, the analogous N,N-dimethylamino derivative 52b is non-selective.116, 117 N (CH2)2R N N NH 52a,b R=N (a), NMe2 (b). It is thought that the region in the receptor which binds the fragment at the C(3) atom contains an aspartic acid residue.117 The steric requirements for the binding to this domain in both types of receptors can be determined by modifying this frag- ment.117 Studies of series of compounds 53 and 54 revealed that 5-HT1D receptors can bind bulky groups. It was therefore assumed that this domain contains a spacious binding pocket. At the same time, compounds containing smaller substituents bind more effectively to 5-HT1B receptors.For example, compounds 53 or 54 which are full agonists of 5-HT1D receptors manifest 100- and 140-fold higher selectivities towards them. Hence, the selec- tivity towards each of these two receptor subtypes can be provided by a correct choice of a substituent in the structural domain responsible for the binding to the aspartate residue. It is of note that the presence of a bulky substituent at C(3) can in principle 339 increase the affinities for 5-HT1A receptors (see above). However, the above-mentioned series includes compounds (e.g., 54, R1=CH2Ph, R2=H) which are selective with respect to both 5-HT1B and 5-HT1A receptors.117 N (CH2)2N N N R NH 53 (CH2)2N H CH2NR1R2 O NH NH O 54 Many (3-piperazinopropyl)indoles, e.g., compound 55a, also manifest high (up to 200-fold) selectivities towards 5-HT1D vs.5-HT1B receptors.118 The selectivity increases 300- to 500-fold upon introduction of the hydroxy group or the fluorine atom into the propyl chain (compound 55b).119 The distal nitrogen atom of piperazine does not seem to be essential for the binding to this receptor, since the corresponding piperidine analogues, e.g., compound 55c, also manifest high affinities.120 The distal nitrogen atom can also be displaced into the exocyclic position without any loss of its affinity,120 as in the case of the 4-aminopiperidine derivative 55d.116 The selectivities of the ligands 55a ± d towards 5-HT1D receptors suggest that these are less sensitive to changes in the position of the nitrogen atom than 5-HT1B receptors despite close homology of these two proteins.This circumstance can further be used in the structural design. N XR2 CH2CH(R1)CH2N N N NH 55a ± d R1=H,X=N,R2=(CH2)2C6H4F-3 (a); R1=OH, F; X=N, R2=Ph, C6H4F-3 (b); R1=H, X=CH, R2=Bn (c); R1=H, X=CH, R2=N(Me)Bn (d). Among other full and partial agonists of 5-HT1B and 5-HT1D receptors structurally different from tryptamine, special mention should be made of arylpiperazines, e.g., compounds 56 121, 122 and 57,121, 123 which display very low selectivities towards 5-HT1B subtype receptors. As expected, these compounds possess high N NH 3-F3CC6H4N N N F3C 56 NMe 57 affinities with respect to the 5-HT1A receptor. More recently synthesised compounds of this series include alniditane (58) 124 as well as compounds 59 125 and 60.126 NH O CH2NH(CH2)3HN NPrn2N 58 NH MeNHSO2CH2 Cl N S 59 60 The adrenoceptor agonist oxymethazoline (61) manifests agonistic activity towards both 5-HT1B and 5-HT1D receptors340 and is among the most promising lead compounds { for the design of 5-HT1D/1B agonists.127 OH Me But N NH 61 Me However, an analysis of structural requirements for the bind- ing of benzylimidazolines of the general formula 62 to these receptors revealed that far from all substituents in the aromatic ring of compound 61 are necessary for the manifestation of high affinity.Thus the hydroxy group in position 3 of compounds 62 can be removed without any noticeable decrease in their affinities for 5-HT1D subtype receptors.However, such modification decreases 50-fold the affinity for 5-HT1B subtype receptors. More- over, the contribution of the bulky substituent at the C(4) atom to the binding to 5-HT1D receptors is greater. Therefore, some structural modifications of compounds 62 (cf., e.g., compounds 62a,b) lead to a 100-fold increase in the selectivities towards 5-HT1D receptors. Compounds 62a,b are agonists, their affinities for adrenergic receptors are lower than that of compound 61. It should be noted that the transition to amidines 63 decreases both their affinities and selectivities towards 5-HT1D subtype recep- tors.113 N NH R R NH2 RN2 63 62a,b R1=4-But (a), 6-Me (b).Thus, the overwhelming majority of agonists of 5-HT1B and 5-HT1D receptors represent tryptophan analogues; therefore, these types of new lead compounds could be useful in the search for the next generation of agonists able to distinguish between these two subtypes of receptors. The main efforts of chemists engaged in the search for antagonists are directed at the synthesis of compounds more selective towards 5-HT1B/1D subtype receptors in comparison with other serotonin receptors. As mentioned above, some aryl- oxyalkylamines, such as the well-known b-adrenoceptor antago- nist propranolol (64), are characterised also by low affinities for 5-HT1A and 5-HT1B receptors and usually represent antagonists or very weak partial agonists of these receptors.OCH2CH(OH)CH2NHPri 64 Studies of binding of their analogues to 5-HT1D/1B recep- tors 128 revealed that the presence of the hydroxy group in the side chain is not necessary, whereas the substitution of the CH2 group for the oxygen atom in the ether group results in the loss of affinity. The affinity for 5-HT1B subtype receptors can be increased by reducing the length of the side chain to two CH2 groups or by using a primary or a secondary amine with smaller alkyl groups, e.g., compound 65. Replacement of the naphthalene fragment by a substituted phenyl ring is also possible, but in this case the affinity will decrease by an order of magnitude as in compounds 66 and 67.129 In our opinion, the structures of these lead compounds are not very suitable for the design of antago- nists, since the compounds obtained interact with adrenoceptors, { The `lead compound' is one of the basic concepts in medicinal chemistry which denotes a structural prototype for the future drug.The lead compound manifests definite physiological activity and its structure serves as the basis for the design of a new drug.2 O N Zefirova, N S Zefirov but are virtually non-selective towards 5-HT1B subtype receptors in comparison with 5-HT1D subtype receptors. Moreover, some high-affinity compounds, e.g., compound 65, lose their antago- nistic properties and behave as agonists towards the correspond- ing receptors. O(CH2)2NMePrn O(CH2)2NHMe 66 65 N OCH2CH(OH)CH2NHPri 67 It is of note that until recently virtually no selective 5-HT1B/1D antagonists existed and studies into receptor binding were carried out with other compounds, e.g., methiothepin (68), which inter- acts with other serotonin, dopaminergic and adrenergic receptors.S SMe N 68 NMe Benzanilide (69) is the first example of a selective 5-HT1D/1B antagonist, although it does not discriminate between the sub- types. This finding has played a crucial role in the characterisation of 5-HT1D/1B receptors. However, subsequent studies using cloned human receptors revealed that compound 69 is not a full antag- onist, it rather acts as a partial agonist.130, 131 Therefore, the construction of selective antagonists required further investiga- tions. NMe N Me N O OMe CN(H) N O Me 69 Sumatriptan (40), zolmitriptan (44) and other tryptamine analogues were used for the design of 5-HT1B/1D antagonists based on the corresponding selective agonists.QSAR studies of the factors affecting both the activity and selectivity as well as simulation using pharmacophore models of 5-HT1D/1B recep- tors 132 revealed that certain geometrical and conformational restrictions can result in the displacement of the indole nucleus from the site (which is arbitrarily termed as the aromatic binding site) specific, e.g., for zolmitriptan (44). It was demonstrated that the retention of high affinity can be accompanied by a significant decrease in intrinsic activity. Later, it was shown that the antagonistic effect is manifested if only the pyrrole moiety of tryptamine (rather than the whole indole nucleus) is displaced.In other words, it was necessary to modify the structure of the agonist in such a fashion that the interaction of the pyrrole double bond with the p-electrons of the receptor in their specific domain would be blocked and the agonistic response would correspond- ingly be prevented. To achieve this, it was proposed to design an electron-deficient aromatic system by introducing electron-with- drawing substituents into position 2 of zolmitriptan (44). The presence of a substituent in position 2 not only makes the indole system electron-deficient, but also restricts the conformational mobility of the side chain. The alkyl chain in position 5 was also varied in order to confer greater mobility on this substituent.132Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors This approach was successful, viz., the resulting tryptamine derivatives of the type 70 contain heterocyclic substituents com- prising the carbonyl group R2 (this group is critical for selectivity as compared with that of 5-HT2A receptors) 133 and represent highly active full antagonists of 5-HT1B receptors.These are also selective towards other serotonin receptors including sometimes compounds 70a,b 5-HT1D subtype receptors. Substitution of the amide group for the ester group in position 2 afforded less active and less selective antagonists; high affinities and selectivities were restored only after the introduction of bulky substituents to the nitrogen atom, e.g., compound 71.It is believed 134 that this substituent can occupy the aromatic binding site in the receptor by pushing the indole nucleus out from this site. (CH2)2NMe2 R2(CH2)2 CO2R1 NH 70a,b O O NH (b). (a), N R1=Et: R2=N NHMe Me O OO HN (CH2)2NMe2 Me N (CH2)2 Me C(O)NHBn O NH 71 Yet another example is the design of 5-HT1D/1B receptor antagonists from the corresponding agonists.88 Assuming that the tryptamine fragment in compound 47 is responsible for the intrinsic activity (by analogy with serotonin), while the arylpiper- azine moiety is mainly responsible for the binding to the receptor, it was suggested to replace the tryptamine fragment in compound 47 by a non-selective ligand manifesting antagonistic activity towards 5-HT1D/1B receptors (in this particular case, the 1-naph- thylpiperazine fragment) which would result in their selective antagonists. This was the case, e.g., for compounds 72.Studies of these compounds showed that they are indeed selective towards 5-HT1D/1B receptors and many of them are full antagonists. However, the antagonistic activity depends on the aryl substituent of the arylpiperazine side chain and is maximum for Ar=2-MeC6H4 or 2,3-Me2C6H3. This suggests that the binding of this fragment also contributes to the intrinsic activity of the ligand.88 Me NN ArN NC(O)CH2O 72 Recently, a series of antagonists based on the above-men- tioned partial agonist 69, which was successfully used as the lead compound in the design of compounds selectively blocking 5-HT1D and 5-HT1B subtype receptors, have been synthesised.These include primarily benzamide (73),112, 135 the structural analogue of compound 69, which is a strong antagonist much more selective towards 5-HT1B subtype receptors. Variations in the substituent R of its methylated analogues 74a ± d revealed that all these compounds bind to the receptor to approximately the same degree and in the same manner (like compound 73) and manifest low affinities for 5-HT1A and 5-HT1D subtype receptors; their agonistic activities depend critically on the conformation of their side chains. Compound 74a and the (Z)-alkenyl derivative 74b act as strong antagonists resembling compound 73, whereas 341 the alkynyl derivative 74d and the (E)-alkenyl derivative 74c manifest pronounced agonistic activities.89 (CH2)3NMe2 O OH N NH 73 R O OMe N NH 74a ± d R=(CH2)3NMe2 (a), (Z)-CH=CHCH2NMe2 (b), (E)-CH=CHCH2NMe2 (c), C CCH2NMe2 (d).It should be noted that substitution of the flexible aminoalkyl side chain for the piperazine ring in compound 69 (compounds 75a,b) decreases their intrinsic activities with respect to 5-HT1B receptors.89, 136 These data demonstrate the significant role of the spatial orientation of the aminoalkyl fragment in the design of 5-HT1B antagonists, although it is still impossible to describe this notion in terms of conformation, since some conformationally rigid analogues of compounds 69 can also act as full antagonists. Thus compound 76 is a selective 5-HT1B antagonist; its design was based on the use of a computer model of 5-HT1B, 5-HT1D, 5-HT1E, 5-HT2A and 5-HT2C receptors.86, 137 Me Me N O OMe O N X(CH2)2NMe2 NH75a,b X = O (a), CH2 (b).Me N Me Me N O O N N O 76 In conclusion, we shall refer to the structure of compound 77, the recently discovered selective antagonist of 5-HT1D recep- tors.138 This compound possesses 60-fold higher affinity with respect to 5-HT1D subtype receptors than to 5-HT1B subtype receptors and is virtually inactive with respect to other serotonin receptors. NC6H4Cl-3 Ph2CHCH(OH)CH2N 77 It is necessary to stress again that only a few compounds are known which interact with each of the above-mentioned subtypes with high selectivities, which is due to the structural similarity of 5-HT1B and 5-HT1D receptors.However, the present-day struc- tural requirements for the design of such ligands are becoming more and more understandable and their number will be increas- ing. V. Ligands of serotonin receptors of other 5-HT1 subtypes The 5-HT1E and 5-HT1F subtype receptors have been studied to a much lesser degree than those of other 5-HT1 subtypes. It is known that they are present in large concentrations in brain and display high structural similarity as regards their amino acid sequences. The non-selective 5-HT1 agonist, viz., the amine 2, manifests ca.1000-fold lower affinity for 5-HT1E subtype recep-342 tors than for the 5-HT1B subtype receptors.139, 140 This is explained by the presence of the lysine ± glutamic acid pair in the binding site of the 5-HT1E receptor protein instead of the less polar isoleucine ± serine (5-HT1B) pair or the valine ± serine pair (5-HT1D). Presumably, this results in the formation of a salt bridge between lysine and the aspartate residue which plays the key role in the ligand binding. Compound 78a is the agonist of 5-HT1E/1F receptors but does not discriminate between them and possesses low affinity. No selective or highly active ligands of 5-HT1E receptors have been identified so far and the function of these receptors still remains to be established.NMe R NH78a,b R=OH (a), 4-FC6H4CONH2 (b). Recently, studies of the functions and ligands for 5-HT1F receptors have attracted the attention of many research groups due to the detection of high affinity of sumatriptan (40) for 5-HT1F receptors (which was about an order of magnitude lower than for 5-HT1D receptors). It was assumed that some (probably all) effects of sumatriptan in the alleviation of migraine symptoms are due to its binding to these receptors. The recently discovered indole derivative 78b is their first and the only selective ago- nist;141 ± 143 no antagonists of 5-HT1F receptors have been identi- fied yet. VI. Ligands of serotonin 5-HT2 subtype receptors The family of serotonin 5-HT2 receptors is classified into three subtypes, viz., 5-HT2A, 5-HT2B and 5-HT2C.The first two subtypes occur widely in peripheral tissues where they regulate muscle contraction; therefore, their ligands are used in the therapy of cardiovascular diseases. As for 5-HT2C subtype receptors, they are localised exclusively in the central nervous system; their ligands produce psychotropic effects, e.g., their antagonists are used as anxiolytic and sedative drugs. A series of 5-HT2 receptor agonists were obtained by the simplest modifications of the serotonin structure. Thus a-methyl- 5-hydroxytryptamine (79) 48, 144, 145 is a full high-affinity agonist of 5-HT2B subtype receptors which manifests 10- and 100-fold lower affinities for 5-HT2C and 5-HT2A receptors, respectively.Similar selectivity profiles hold following the introduction of the thienylmethoxy substituent into position 5 (compound 80).146 ± 148 NH2 NH2 CH2CH CH2CH HO CH2O S Me Me NH NH 80 79 5-Methoxytryptamine also manifests higher affinity for 5-HT2B receptors than for 5-HT2A and 5-HT2C receptors (25- and 400-fold, respectively), although it is non-selective towards other serotonin receptors. The introduction of a bulky bromo- benzyl substituent into the amino group of 5-methoxytryptamine results in the ligand 81 which possesses high affinity and selectivity with respect to 5-HT2A receptors,149 but manifests the properties of a partial agonist.(CH2)2NHCH2C6H4Br-4 MeO NH 81 Bioisosteric substitution of thienepyrrole fragments for the tryptamine indole nucleus was undertaken in an attempt to design O N Zefirova, N S Zefirov 5-HT2 agonists.However, the resulting compounds 82 and 83 displayed somewhat lower activities towards 5-HT2 receptors than the corresponding indole analogues (compound 82 is nearly 4 times more selective with respect to 5-HT2C receptors than to 5-HT2A and 5-HT2B receptors). Moreover, this transformation increased significantly the affinity for 5-HT1A receptors.150 On the whole, these data corroborate the fact that different subtypes of serotonin receptors are highly sensitive to minor changes in the electronic characteristics of the aromatic systems. (CH2)2NMe2 (CH2)2NMe2 S S 83 82 NH NH Studies of restriction of conformational mobility in the amino- ethyl side chain of serotonin have led to the N-methylpyrrolidine derivative 84 (R=5-OMe) and its analogues which manifested very high agonistic activities, being more selective towards 5-HT2A subtype receptors.The activities of these compounds strongly depend on their stereochemistry, viz., (R)-enantiomers manifest ca. 30-fold higher affinities than (S )-enantiomers.151, 152 R NH2 R Me N NH NH84 85 It is of note that conformationally rigid serotonin analogues containing cyclopropane fragments (85) in their side chains are somewhat more selective towards 5-HT2C receptors, the fluoro derivative (85, R=5-F) being the most active compound.153, 154 Noteworthy, the electron-withdrawing fluorine atom in the ben- zene ring not only decreases the affinity of this compound for 5-HT1A receptors (see above) but also seems to be responsible for higher selectivity towards 5-HT2C receptors in comparison with 5-HT2A receptors.A similar dependence was demonstrated for other fluorotryptamines.155 The agonists manifesting higher selectivities towards 5-HT2C subtype receptors were identified among halogeno derivatives of other structural classes, e.g., of piperazine. The diamine 86, a partial agonist of 5-HT2C receptors, which possesses 10-fold lesser affinity for 5-HT2A receptors, is the most popular among them.48, 156 ± 159 Structural variations in its molecule revealed that substitution of the pyrazine ring for the benzene ring increased significantly its selectivity and intrinsic activity; thus, the affinity of compound 87,160, 161 a full agonist of 5-HT2C subtype receptors, is 25 times higher than for 5-HT2A subtype receptors.N NH N NH 3-ClC6H4N N 87 86 Cl Agonists of 5-HT2C receptors have recently been identified among indole derivatives manifesting high selectivities towards 5-HT2A subtype receptors. Earlier, it was found 162 that isotrypt- amines (88) with the 2-aminopropyl side chain attached to the nitrogen atom are bioisosteric to the corresponding tryptamines as regards their binding to serotonin receptors. Affinity studies of a series of isotryptamines 88 and their analogues, viz., indenopyr- roles 89, revealed 155 that many of them, e.g., compounds 88a,b and 89a,b, are high-affinity full agonists of 5-HT2C receptors which are 20- to 100-fold more selective towards 5-HT2A recep- tors.It is noteworthy that the selectivities of S-isomers are higher than those of the corresponding R-isomers.155 It should be noted that tryptamine structural isomers of some of these derivatives manifest nearly identical affinities but much lower selectivities with respect to 5-HT2C receptors.Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors R2 R2 R1 R2 N R1 N 88a,b CH2CH(Me)NH2 89a,b CH2CH(Me)NH2 R1=7-OMe: R2 = H (a), Me (b). R1=5-F: R2 = H (a), 6-Cl (b). and Phenylisopropylamines phenylethylamines 90a 90b 163 ± 165 which constitute a large group of high-affinity 5-HT2A/2C receptors have been the subjects of in-depth studies into the structure ± activity relationships aimed at establishing the topologies of the binding sites of this particular subtype of serotonin receptors (see, e.g., Refs 166 ± 169).These studies demonstrated that high affinities for 5-HT2A/2C receptors require that certain structural requirements are fulfilled, such as the presence of a two-carbon chain between the amino group and the phenyl ring, the presence of methoxy groups in positions 2 and 5 of the aromatic ring and the presence of the hydrophobic substituent X in position 4. Assuming that the methoxy groups in compounds 90a,b bind to two serine residues in the receptor protein, Nichols et al.166, 170 synthesised the cyclic derivatives 91 ± 94 in order to study the geometry of this binding.It was found that compound 91 having a `syn-conformation' of the tetrahydrofuran ring relative to the aminoalkyl chain is practically inactive towards 5-HT2A/2C receptors, whereas the activity of the analogous compound 92 with the `anti-orientation' of these frag- ments is as active as that of the acyclic analogue 90a. OMe OMe CH2CHNH2 CH2CHNH2 Me R X O 91 90a,b OMe R=Me (a), H (b); X=Alk, Hal, AlkS, F3C etc. OMe CH2CHNH2 Me Br 92 O Substitution of the oxygen-containing heterocycle for the methoxy group in position 2 produces an opposite effect, viz., the `anti-analogues' of the type 93 compounds are inactive, whereas compound 94 is a potent agonist.As expected, structur- ally rigid compounds, e.g., compound 95 with two dihydrofuran fragments mimicking effectively active binding conformations of the methoxy groups in compounds 90a,b, appeared to be high- affinity agonists of 5-HT2A/2C receptors. In this case, benzodi- hydrofuran-containing compounds represent effective conforma- tionally restricted bioisosteres of aromatic methoxy groups.169 O O NH2 CH2CHNH2 Me I X OMe 93 94 OMe Substitution of furan fragments (see, e.g., compound 96) 169 for dihydrofuran fragments made it possible to enhance also the agonistic activity with respect to 5-HT2A/2C receptors by an order of magnitude. Compounds of the type 96 manifest the highest affinities among other known ligands.Unfortunately, none of 343 these compounds discriminate between 5-HT2A and 5-HT2C subtype receptors; the affinities of these compounds for 5-HT2B receptors are severalfold lower because of similarities of the binding domains of the corresponding receptors. O O CH2CHNH2 CH2CHNH2 Me Me Br Br O O 96 95 The attempts to lock the geometry of the aminopropyl side chain in compounds 90a,b showed that its incorporation into the aminotetralin or the aminoindan rings results in the loss of activity towards 5-HT2A/2C receptors, whereas its introduction into the side chain of the cyclopropane ring, as in trans-2,5-dimethoxy- 4-methylphenylcyclopropylamine, enhances the activity.154, 171 As can be seen from the above survey, the structural classes of 5-HT2 agonists are not very diverse and their design is poorly documented in the literature.This can probably be attributed to the strong hallucinogenic effect of these agonists,172 which makes them unsuitable as therapeutics. Antagonists of 5-HT2 subtype receptors can be used as drugs in treatment of ischemia and other vascular disorders.173 So far, several structural classes of 5-HT2 antagonists have been identified. The most prominent position is occupied by ergolines and related compounds which manifest the properties of partial agonists or antagonists with respect to the correspond- ing receptors 174 [the former include the classical hallucinogenic agent, viz., lysergic acid diethylamide (LSD, 97)].174 ± 176 Numer- ous antagonists of 5-HT2 receptors have been synthesised based on ergolines and related compounds; the main strategy consisted in the introduction of various substituents [most frequently, at C(8)] assuming that this might enhance the antagonistic activity due to their additional binding to the receptor. This resulted in potent 5-HT2 antagonists, such as compounds 98,177, 178 99a ± c, etc.176, 179 ± 182 H O NEt2H CH2NHC(O)OBn 8 H NMe H NMe H MeN HN 98 97 COR H OMe (a), HN (b), R=O NMe H HO (c).HN PriN 99a ± c Subsequent computer simulation of the binding of ergolines to 5-HT2 receptors 174 showed that the presence of an aromatic ring in the vicinity of phenylalanine-340 of the receptor is an essential structural feature which ensures high-affinity binding owing to p- or hydrophobic interactions. This prompted synthesis of a series of ergolines and related compounds of the type 100a,b devoid of the acyl fragment at C(8).Indeed, the resulting com- pounds manifested high affinities for 5-HT2 antagonists and high selectivities as regards other subtypes of serotonin receptors. Thus, the ergoline structure of 5-HT2 antagonists can in principle be reduced to a simple tetracyclic indoloquinoline system without any detriment to the affinity and antagonistic activity.182344 Me Me NH NMe H H N RN 101 100a,b R=H (a), Pri (b). Other structural modifications of ergoline analogues included variations of substituents at the nitrogen atom of the indole moiety. It was shown that the introduction of small alkyl substituents, including branched ones, into this position is possi- ble.} In some cases, the antagonistic activity is enhanced due to the presence of a double bond in positions 9 and 10 present also in lysergic acid.182 The literature data on the role of substituents in position 4 of the cyclohexane ring are controversial: in some cases, the methoxy group at C(4) of the cyclohexane ring significantly enhances the antagonistic activity,187 but has no effect in other cases.182 The majority of ergoline derivatives are active with respect to serotonin 5-HT1, adrenergic and other receptors, but are non- selective towards 5-HT2 subtype receptors.However, recent systematic screening of ergoline-related compounds disclosed the naphthyridine (101),188 ± 190 a highly selective antagonist of 5-HT2B/2C receptors which does not interact with 5-HT1 receptors. This compound will probably serve as a prototype for the design of novel selective ligands.Distant analogues of the long-known non-selective ligand 102,191, 192 such as ketanserin (103) 193, 194 and its derivatives which manifest certain selectivities towards 5-HT2A receptors, constitute another large structural class of 5-HT2 anta- gonists.191, 192 Ketanserin (103) is used in the clinical practice as an antihypertensive drug. O COC6H4F-4 S(CH2)3NMe2 N(CH2)2N NHC(O)CH O CHPh 103 102 HN Studies of the structure ± activity relationships in ketanserin derivatives have revealed the following features.The carbonyl group in position 2 of quinazoline can be replaced by the thiocarbonyl group, and a phenolic hydroxy group can be introduced into position 6. The two-carbon side chain can be replaced by the (CH2)4, but not by the (CH2)3 chain, which decreases the activity 100-fold.195 Piperidine ring-opening also decreases the affinity. Although the 4-fluorobenzoyl group is present in the majority of ketanserin analogues, it can be replaced by the benzylidene [as in the case of ritanserin (104)] or a similar group. However, the benzylic carbon atom should be sp2-hybri- dised [cf. the carbonyl group in ketanserin (103) or the double bond in ritanserin (104) and risperidone (105)]. Although a certain role in the affinity for 5-HT2A receptors is played by the quinazo- line-2,4-dione fragment,195 this may be replaced by the pyridopyr- imidine or thiazolopyrimidine fragment 196 or structurally simplified, as in compound 106.197 The latter modification is more appropriate because it affords 120-fold enhancement of } It should be noted that ergolines non-substituted at N(1) manifest higher affinities for human 5-HT2 receptors than for the corresponding rat receptors,183, 184 although the only structural difference between these receptors is associated with the amino acid 242 residue (serine in humans and alanine in rats).185, 186 Presumably, serine and the NH group of indole in non-substituted ergolines form a hydrogen bond, whereas hydrophobic interactions occur between alanine and the alkyl group of the substituted ergolines.selectivity towards 5-HT1C receptors for which ketanserin mani- fests high affinity. It is of note that some ketone 106 analogues having a shortened bridging chain, e.g., compound 107,198 are also potent selective antagonists of 5-HT2 receptors. However, none of the above-mentioned compounds distinguish among different subtypes of 5-HT2 receptors. O (CH2)2N N S Me N 104 O (CH2)2N N Me N 105 N(CH2)4Ph 4-FC6H4CO 106 Variations in the quinazolidinedione fragment of ketanserin analogues have led to a large family of 5-HT2 antagonists, viz., triazaspirodecanones; their classical representative is spiperone (108).197, 199 This compound manifests rather high affinity for dopaminergic and 5-HT1A receptors, and is also one of the few ligands binding more selectively (ca.1000-fold) to 5-HT2A subtype receptors than to 5-HT2C subtype receptors. Ph N 4-FC6H4CO(CH2)3N 108 O Molecular `dissection' of this compound into its two basic fragments revealed that 1,3,8-triazaspirodecanone (109) is abso- lutely inactive towards 5-HT1 and 5-HT2 receptors, while the piperidine derivative 110 (R=H) manifests moderate affinity for 5-HT2A receptors, which is 150-fold lower than that of spiperone (108), but is inactive towards 5-HT1 and 5-HT2C receptors. The introduction of the anilino group into position 4 of piperidine (110, R=NHPh) increases its affinity for 5-HT2A receptors; however, the resulting compound manifests 150-fold lower selec- tivity towards 5-HT2C receptors than spiperone (108).200 These findings are suggestive of the important role of the imidazolidi- none fragment in the reactivity and selectivity of spiperone (108).This conclusion is corroborated by the fact that the spiperone derivatives 111 containing geminal amino groups and an amide group instead of an imidazolidine ring are less selective towards 5-HT2A receptors. Moreover, the affinities of some spiperone analogues decrease dramatically after the removal of the lactam carbonyl group, which points to its significance for the binding to the receptor. O NH 4-FC6H4CO(CH2)4N HN N 109 Ph NHR 4-FC6H4CO(CH2)3N CONH2 111 Other modifications of the spiperone (108) structure showed that substitution of the cyclohexyl fragment for the phenyl substituent at the N(1) atom increases its affinity twofold, whereas the substitution of alkyl groups decreases it.However, the O N Zefirova, N S Zefirov C6H4F-4 C6H4F-4 N OF Ph(CH2)2N CHPh OH 107 NH R 110Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors selectivity towards 5-HT2A subtype receptors increases in com- parison with 5-HT2C, 5-HT1A and dopaminergic receptors.201 Contrariwise, analogous substituents at the N(3) atom signifi- cantly decrease the selectivity.201, 202 It was shown also that the carbonyl group in the spiperone (108) molecule may be replaced by the ether group (compound 112) with the preservation of activity and a small increase in the selectivity towards 5-HT2A receptors in comparison with 5-HT2C receptors.Such analogues can further be used as new structural prototypes in the search for affinity ligands to 5-HT2A receptors.200, 201 O HN N(CH2)3OC6H4F-4 112 NPh The sulfonylamide derivative 113 assayed recently among spiperone analogues (108) is attractive, because unlike compound 108 it manifested antagonistic activity and 100-fold higher selec- tivity towards 5-HT2C receptors in comparison with 5-HT2A and 5-HT2B receptors. Moreover, this compound virtually did not bind to other serotonin receptors.203, 204 O CO(CH2)4N 4-F3CC6H4SO2NH NH MeO OMe HN O 113 Progress has also been achieved in the design of selective antagonists of 5-HT2B receptors. It was found that the indole derivative 114 manifests 150-fold higher affinity for 5-HT2B receptors than for 5-HT2A receptors, but is non- selective with respect to 5-HT2C receptors.205 The selectivity to the latter was virtually unenhanced upon partial restriction of conformational mobility of the linking chain (compound 115); however, this approach makes it possible to increase the affinity for 5-HT2B receptors and to decrease the affinity for 5-HT2A receptors.A highly selective antagonist of 5-HT2B receptors was obtained by substitution of the methylisothiazole fragment for the pyrimidine fragment (114?116), but the activity of the resulting compound is not very high.206, 207 O O NH HN HN N N N Me Me 115 114 O Me HNHN N S Me 116 More active and selective (with respect to 5-HT2B receptors) compounds were prepared based on the alkaloid yohimbine (117), which exhibits low affinity for this type of receptor itself.Eluci- dation of the key elements essential for the antagonistic activity of this alkaloid towards 5-HT2B receptors was carried out by sequential simplification of its structure, resulting in high-affinity derivatives 118 and 119. The affinities of these compounds and some of their analogues for 5-HT2B receptors were 100 ± 200-fold higher than for 5-HT2A and 5-HT2C receptors; these were practi- cally inactive with respect to adrenergic receptors (in contrast to the original alkaloid 117).173 Structure ± activity studies revealed an unusual and interesting fact, namely, it was found that the introduction of the methoxy substituent into position C(6) of compound 118 [which corresponds to position C(5) of the serotonin molecule] decreases the affinity, whereas the introduc- tion of two methyl substituents into positions C(7) and C(8) increases it to a unique extent.173 This suggests that the direct structural analogies of serotonin and the corresponding fragments of compounds 118 and 119 do not reflect their actual orientation during their binding to the receptor and demands further inves- tigation.H N N H H OH H MeO2C 117 NH NH 119 A new structural class of 5-HT2 antagonists based on 5-HT2A agonists of the type 90a has recently been found.Compound 90a (X= CH2CH2CH2Ph) containing a phenylpropyl substituent in position 4 has become the first representative of this class. Structure ± activity studies of some of its analogues established a surprising fact, viz., that the presence of methoxy groups in the ring is not a necessary prerequisite for their high affinities. However, this feature holds only for antagonists (for agonists, see above). This indicates that the phenylpropyl substituent can change the binding profile to the receptor.165, 208 Further develop- ments in this field can be expected in the nearest future. And finally, there exists a vast class of compounds manifesting nearly identical antagonistic activities towards serotonin 5-HT2 and dopamine receptors which can be used as efficient anti- schizophrenic drugs.The neuroleptic drug clozapine (120a) is a classical representative of these compounds.165, 209, 210 This pos- sesses somewhat higher selectivity towards serotonin receptors of the 5-HT2A subtype, whereas its desmethyl derivative 120b manifests higher affinity for 5-HT2C receptors.211, 212 A large number of their presently known analogues comprise a lipophilic tricyclic system,213 which may contain additional heteroatoms (e.g., compound 121) or be devoid of them (e.g., compound 122).214, 215 Tetracyclic analogues such as mianserin (123) are also known.216, 217 A detailed survey of the work devoted to the design of compounds similar to compounds 120 ± 123 or mixed- type antipsychotic drugs belonging to other structural classes is beyond the scope of the present review, which considers only compounds selective towards serotonin receptors.NR N N Cl HN 120a,b R=Me (a), H (b). Me N 122 345 5 NH 6 OMe 7 8 HN OMe 118 NMe N N Me S HN 121 MeN N NH 123346 It should be stressed that the selectivities of antagonists towards serotonin 5-HT2 receptors as opposed to dopaminergic (D2) receptors are difficult to achieve. This problem was the central issue in special studies aimed at comparative structural analysis of binding sites of both types of receptors (see, e.g., Ref. 218). Although the structural requirements for the affinities of both receptors are very close, they have certain steric differ- ences, which makes it possible to predict the selectivities of various compounds to 5-HT2 receptors.To conclude this chapter, it should be emphasised that the search for 5-HT2 ligands employed mainly simple screening procedures, without recourse to computer simulation or pharma- cophore hypotheses. Nevertheless, the design of such ligands entails efficient structural approaches. Moreover, the use of pharmacophore and receptor models developed in the last few years and modern QSAR methods 174, 218, 219 leaves hope that progress in this area will finally be achieved. VII. Ligands of serotonin 5-HT3 subtype receptors Receptors of the 5-HT3 subtype are the only serotonin receptors that form ion channels penetrable for monovalent cations and calcium ions.The ligands interacting with this type of receptor have been in the focus of attention in the past decade, primarily in connection with the design of antagonists, some of which proved to be efficient antiemetic drugs. Antagonists of 5-HT3 receptors can also be used for the treatment of schizophrenia and other mental disorders. The putative therapeutic role of 5-HT3 agonists is based on their ability to control the release of the neuromediator acetylcholine in the brain, which makes these compounds promis- ing for the treatment of neurodegenerative diseases associated with cholinergic transmission disturbances. In addition, 5-HT3 agonists can be used as antidepressant and analgesic drugs.Since the mechanism of signal transmission by 5-HT3 recep- tors is different from those of other serotonin receptors, its ligands often possess unique structures and manifest high selectivity. Thus quaternary amines of the type N,N,N-trimethyl-5-hydroxytrypt- amine do not bind to any other subtypes of serotonin receptors. Agonists of 5-HT3 receptors have been studied less thoroughly than their antagonists and are less numerous. The most well- known of them is 5-hydroxy-2-methyltryptamine (124) which possesses lower affinity than serotonin but much higher selectivity towards 5-HT3 subtype receptors and preserves simultaneously agonistic activity.145, 220, 221 (CH2)2NH2 HO Me NH 124 The other two structural groups of 5-HT3 subtype agonists include arylpiperazines and arylbiguanides; their classical repre- sentatives are quipazine (125a) 222, 223 and 1-phenylbiguanide (126, R=H).224 These compounds had been neglected for a long time, apparently because of the high selectivity but low affinity of phenylbiguanide for 5-HT3 receptors.At the same time, arylpiperazines of the general formula 127 display higher affinities for 5-HT2 receptors but are less selective and are active with respect to a vast variety of other serotonin receptors (see above). The number of publications devoted to the design of arylpiperazine and arylbiguanide analogues has increased mark- edly over the past decade; studies in this field were designed to increase the affinities of arylbiguanides and the selectivities of arylpiperazines. O N Zefirova, N S Zefirov NR HN HN N N NH2 R NH NH 126 125a,b R=H(a), Me (b).R NH N 127 For the first trend, the key hypothesis was that both classes of compounds occupy the same aromatic binding site on 5-HT3 receptors; therefore, the structure ± activity dependences for aryl- piperazines 127 can be extrapolated to arylbiguanides. The main structure ± activity regularity for arylpiperazines 127 is that the chlorine atom in the m-position of the benzene ring is the only substituent (R) studied which produces a significant increase in affinity. This cannot be replaced by any other electron-with- drawing substituent (e.g., CF3). Substitution of the naphthyl substituent (compound 128a) for the phenyl substituent, the introduction of the nitrogen atom into the benzene or the naphthalene fragment (compound 128b) and the conversion of the NH group of piperazine into an exocyclic amino group (compound 129) also affords high-affinity compounds.225 ± 227 It should be noted that the majority of these compounds manifest pronounced antagonistic activities and behave as partial ago- nists.} NH2 Y Cl N N NH N 129 128a,b Y=CH (a), N (b).In contrast, all the arylbiguanide derivatives under study are full agonists of 5-HT3 receptors and manifest high selectivities towards them. As in the case of piperazine compounds, the introduction of the chlorine atom into the m-position affords the high-affinity compound 126 (R=3-Cl), and 2-naphthylbiguanide binds 100 times more effectively than 1-phenylbiguanide.229 ± 233 These data unambiguously demonstrate the validity of the hypothesis about the similarity of binding profiles (or binding sites) of arylpiperazines 127 and arylbiguanides 126. It is of note that the affinities of 3-chlorophenyl- (130) and 2-naphthylgua- nides (131) which also act as 5-HT3 receptor agonists manifest the highest (by one or two orders of magnitude) affinities among other structurally related arylguanides 233 and can thus be regarded as representatives of a new structural class of serotonin receptor agonists.HN NH2 Cl HN NH2 NH NH 130 131 Investigations into the increase in the selectivities of arylpiper- azines towards 5-HT3 receptors have led to the synthesis of a vast variety of effective ligands endowed with a broad range of intrinsic activities, viz., from full agonists to full antagonists.In the majority of studies, quipazine (125a) was used as the structural lead; its simplest modifications showed that the N-methyl ana- logue 125b 225 manifests nearly identical affinity for 5-HT3 recep- } Some arylpiperazine derivatives manifest complex activity profiles with respect to 5-HT3 subtype receptors. Thus the intrinsic activity of quipazine (125a) depends on the type of the cells comprising these receptors. It was shown that quipazine behaves as an agonist for some of these receptors and an antagonist for other receptors.228347 Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors tors but does not bind to 5-HT1B receptors (according to more recent data, the N-methyl substituent is the most suitable sub- stituent for many quipazine analogues 234).Earlier, it was shown that the 5-HT1B receptor ligand 57 which contains a pyrrole ring and is inactive towards 5-HT3 receptors acquires high affinity for these receptors upon removal of the trifluoromethyl group. Therefore, a vast majority of studies aimed at modification of the quipazine structure include the introduction of a condensed pyrrole ring into the quinoline system.234 ± 236 Among compounds of the general formula 132, the derivatives manifesting the highest affinities for 5-HT3 receptors possess the properties of partial (R1=H, 6-, 7- or 9-F; R2=H, Alk, CH2=CHCH2) and full (R1=9-Me, R2=Me; R1=7-OH, R2=Me) agonists, the latter being the strongest 5-HT3 agonists presently known.237 N N N R1 NR2 132 bond between the protonated terminal nitrogen atom of piper- azine and the carboxylate anion of the amino acid residue in 5-HT3 receptors (there is conformational restriction for the bulk of the substituent at this nitrogen atom), the formation of a hydrogen bond between the nitrogen atom in the heterocycle and the corresponding hydrogen bond donor in the receptor and, finally, the specific interaction between the aromatic ring and the corresponding amino acid residue of the receptor (there is a size restriction of substituents in the region of aromatic ring fusion).Presumably, the first two interactions make a weighty contribu- tion to affinity.240 Moreover, favourable van der Waals interac- tions which are supposed 238 to be critical for the modulation of intrinsic activities with respect to 5-HT3 receptors (cf. structures 137 ± 139) can be triggered in that domain of the receptor molecule which corresponds to the quinoline fragment. However, the structural determinants responsible for the intrinsic activities of quipazine derivatives are in fact more complicated, since the changes in their activities can be provoked even by any apparently insignificant changes in the heterocyclic backbone of quipazine. Thus compounds 140a,b which contain neither bulky, nor lip- ophilic substituents in their nuclei behave nevertheless as 5-HT3 antagonists.240, 241 The introduction of lipophilic groups into the quipazine NR N N molecule yields the derivatives 133 ± 135, which behave as antag- onists of the corresponding receptors or partial agonists endowed with low intrinsic activities, e.g., compound 136.228 CN N 140a,b R=H (a), CH2=CHCH2 (b).Me N N N N NMe NMe 134 133 QSAR studies aimed at the search for effective 5-HT3 agonists and antagonists based on quipazine give an insight into the molecular fragments essential for the strong binding of the ligands to the receptors and allow the prediction of novel high-affinity compounds. However, elaboration of structural requirements for their agonistic and antagonistic activities is hardly possible with- out detailed computer simulation studies.O (CH2)n N N N N As indicated above, the main progress in the design of 5-HT3 ligands is associated with the synthesis of its selective antagonists some of which have already found application in clinical practice. The majority of 5-HT3 antagonists represent aryl- or hetarylamide derivatives. We shall briefly summarise the main concepts and results of structure ± activity studies of 5-HT3 antagonists with special emphasis on the latest data. NMe NR 136 135 n=0 ±2,R=H, Alk. QSAR studies of quipazine derivatives containing various heteroaromatic fragments revealed several compounds (e.g., com- pounds 137 ± 139) manifesting high activities, high selectivities and different intrinsic activities with respect to 5-HT3 receptors. Thus compound 137 is the partial agonist, compound 138 is the antagonist, while compound 139 is the agonist of the correspond- ing receptors.238 The first potent and selective 5-HT3 antagonists prepared were based on the non-selective dopamine receptor ligand meto- clopramide (141), which had long been used as an antiemetic drug.Studies from the early 1980's showed that metoclopramide and some cocaine (142) analogues manifest relatively weak antago- nistic activities towards one of the serotonin subtype receptors 242 further identified as 5-HT3.243 This activity is responsible for the antiemetic effect of metoclopramide.Further modifications of metoclopramide structure as well as of the structures of analog- uous and lead compounds consisted in conformational restriction of the side chain, modification of the amide fragment and substitution of an oxygen-containing ring for the o-alkoxy group. NMe O MeO2C Cl N N N N NH(CH2)2NEt2 NMe NMe 138 137 OMe H2N 141 142 PhCO2 N N NMe 139 The introduction of the quinuclidine residue into the side chain of metoclopramide (141) afforded zacopride (143), which is a potent selective antagonist of 5-HT3 receptors.244, 245 Modifica- tions including the introduction of a tropane fragment and variations of substituents in the ring give compound 144a which is also a potent selective antagonist of this subtype of receptors.246 Its analogue 144b containing two methyl groups possesses vir- tually the same activity.It should be noted that side chain This finding was used in the design of a binding model of quipazine derivatives (both agonists and antagonists) to 5-HT3 receptors.238, 239 According to this model, the high-affinity inter- action with the receptor involves the formation of a hydrogen348 modifications included the variations in its `mobile' fragment. Thus phenylurea derivatives containing an alkoxy group in the o-position (e.g., compound 145) manifested high antagonistic activities.247 NMe N O R CO2 Cl NH OMe H2N R 143 144a,b R=Cl (a), Me (b). NMe 2-MeOC6H4NHC(O)HN 145 The benzene ring of these compounds was subjected to structural modification.Thus substitution of the indole ring for the disubstituted benzene ring in compounds 144a,b which affords tropisetron (146) 144 significantly increases the affinity. A similar effect is observed upon the replacement of the aryl substituent in the urea derivative 145 by hetaryl substituents (compounds 147 and 148).248, 249 Tropisetron (146) is a selective 5-HT3 ligand used in clinical practice as an antiemetic drug, like its more active structural analogue granisetron (149) containing a 9-azabicyclo[3.3.1]no- nane residue in the side chain.250 NMe NMe O CO2 N NH NH 147 146 Me Me NMe O N NH O NH 148 Modification of the `mobile' fragment of the side chain resulted in the ketoamide (150), a high-affinity and selective 5-HT3 ligand although manifesting the properties of a partial agonist.251 O NMe C C N(H) C(O)N(H) N O NMe N NMe 150 149 It is of note that these modifications were carried out together with the search for effective 5-HT3 ligands among indole deriva- tives. Thus compounds 151 and 152 containing an imidazole fragment in their side chains were synthesised 252 and ondansetron (153),253 the conformationally rigid analogue of compound 152, a potent selective antagonist of 5-HT3 receptors. Ondansetron (153) is the most well-known 5-HT3 antagonist widely used as an antiemetic drug in combined therapy of tumour diseases.254Me N N N C(O)(CH2)2 C(O)(CH2)2 NH Me 152 151 NMe NMe O N Zefirova, N S Zefirov Me N CH2N O 153 NMe A modification of the amide group of metoclopramide (141) included the formation of the benzotriazine fragment (compound 154).Since this modification increases the activity in comparison with the `open' analogues of the type 155, it is assumed that the active conformation of benzamides in the 5-HT3 receptor should involve a planar arrangement of the carbonyl group.255 Similar results were obtained for a series of ondansetron analogues (153) in which ketones fused to the indole fragment were 4 times as active as the non-cyclic analogue 152.255 NMe NMe O O Cl Cl N NH OMe N N H2N H2N 155 154 Yet another version of cyclisation is the incorporation of an amide group into a new ring in the side chain. The idea of this transformation stems from bioisosterism of benzamide and phe- nylimidazolidin-2-one fragments.Comparison of the structures of metoclopramide (141) and the dopamine receptor antagonist zetidoline (156) prompted an idea about the introduction of azabicyclic or other equivalent structural fragments of 5-HT3 antagonists into the phenylimidazolidinonalkyl group. Indeed, several high-affinity antagonists of the 5-HT3 receptor were identified among a series of compounds of the general formula 157, and a compound with R1=R3=R6=R7=H, R2=R4=Cl, R8=Me is characterised by unusually high affin- ity.256 It should be noted also that the indole derivative 158 containing an oxadiazole residue mimicking the amide group instead of the bicyclic substituent is a specific antagonist of 5-HT3 receptors.257 O NR5 R1 O N R2 Me N(CH2)2N 3-ClC6H4N Me R3 156 157 R4 N O N NMe, , R5= CH2NMe2 N NR6 158 .NMe NR7 R8 Substitution of a fused oxygen heterocycle for the o-alkoxy group is also a modification of the metoclopramide (141) struc- ture. This modification resulted in several high-affinity selective 5-HT3 antagonists, e.g., the quinuclidine derivative 159 258, 259 and zatosetron 160.260 In all these cases, the presence of the chlorine atom is a necessary prerequisite for the high activities of these compounds. NMe O HNNH O N O O Me O Cl Me NMe 159 Cl 160Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors Many pharmacophore models have been suggested for 5-HT3 antagonists. The differences between them originate from differ- ent models and different 5-HT3 ligands used in their development.Critical analysis of five pharmacophore models 247, 248, 251, 261 ± 263 disclosed the following essential components required for mani- festation of high affinity, vis., the presence of an aromatic ring, the carbonyl group directly attached to the aromatic ring and a basic amine. The aromatic ring can represent a structural element of benzoate, benzamide, indole or imidazole fragments. As men- tioned above, the carbonyl group can be replaced by an equivalent bioisosteric group, e.g., the 1,2,4-oxadiazole or the imidazolidin- 2-one fragment.The presence and accessibility of the amine nitrogen plays a significant role in certain models,261 ± 263 although the details of this binding are still unknown.264 It should be stressed, however, that some antagonists (e.g., compound 157) do not fit into these structural models. The lack of correlation brought forth the necessity for modification of the binding profile of the receptor, viz., two additional hydrophobic binding sites were identified in the domain surrounding the aromatic ring.257, 265 Pharmacophore models have proved to be a useful tool for the rational design of 5-HT3 antagonists. Thus these models were used to predict high affinities of 3-aminopyridazines 161 and 162 containing a piperazine fragment for 5-HT3 recep- tors.265 Ph NMe N NMe N N N N N 161 162 Cl N NMe N O 163 Me It should finally be noted that the number of publications devoted to the design of 5-HT3 antagonists has presently lessened, presumably due to progress already achieved in this area.Never- theless, the search for novel structural classes of 5-HT3 antago- nists is a vital problem in the endeavour to go far beyond the scope of tryptamine, metoclopramide and arylpiperazine derivatives. In novel compounds, such as compound 163 containing a seven- membered ring, modifications involve the same fragments. Com- pound 163 is a high-affinity partial agonist of 5-HT3 receptors with low intrinsic activity.266, 267 VIII. Ligands of serotonin 5-HT4 subtype receptors Serotonin 5-HT4 subtype receptors were discovered in the late 1980's and cloned in 1995;268 their molecular structures and some functional characteristics have not been described until very recently.269 Various therapeutic applications for their agonists and antagonists, e.g., in the treatment of intestinal disorders, arrhythmias and diseases associated with the loss of the cognitive ability, are based on the localisation of this receptor in gastro- intestinal tract, atrium and urinary bladder tissues as well as in the central nervous system.Some non-selective antagonists of serotonin receptors, mostly of 5-HT3 receptors, e.g., metoclopramide (141), possess partial agonistic activities towards 5-HT4 receptors. Therefore, metoclo- pramide (141) and its ester analogue 164 270 manifesting the properties of a weak antagonist and 30-fold selectivity towards 5-HT4 receptors than to 5-HT3 receptors were used as the lead compounds in the design of more selective 5-HT4 ligands.Cl CO2(CH2)2NEt2 OMe H2N 164 349 The general scheme of the design of these ligands resembles in many features that for 5-HT3 antagonists and includes, in the first place, different versions of restriction of conformational mobility and modification of the amine side chain. The agonist 165 271 and the antagonist 166 (a quaternised butyl analogue of renzapride (which is a partial agonist of the 5-HT3 receptor) were among the first compounds manifesting somewhat higher selectivity towards 5-HT4 receptors in comparison with 5-HT3 receptors.This modification is based on a significant decrease in the activities of quaternary salts of 5-HT3 antagonists with substituents bulkier than the methyl group at the nitrogen atom.272, 273 Special mention should also be made of the original macrocyclic benza- mide (167) which represents a partial agonist twice as selective towards 5-HT4 receptors as towards 5-HT3 receptors, which does not interact with other receptors.274 O Cl N HN OMe H2N 165 O O + Cl NBun Cl HN HN O H2N N OMe H2N O 166 167 The close structural analogues of compound 164, viz., the piperidine derivatives 168 and 169a,b, are potent selective partial agonists of 5-HT4 receptors (their intrinsic activities are 50%± 60%).275 ± 277 The classical approach to decrease the intrin- sic activity of compound 169b by introducing bulky lipophilic substituents proved to be successful.Thus the derivative 169c 275, 278 manifests high antagonistic activity and 100-fold higher selectivity towards 5-HT4 receptors than towards other serotonin receptors. It is of note that the changes in the structural activity of compound 168 from the agonistic to the antagonistic one is achieved by less significant structural modification, viz., through the introduction of two methyl groups into positions 3 and 5 of the piperidine ring. For the derivative 169a, the decrease in the intrinsic activity is achieved through structural simplifica- tion, whereas the analogue 170 containing an additional carbon atom in the side chain but devoid of the lipophilic butyl sub- stituent is a full 5-HT4 antagonist manifesting high affinity and noticeable selectivity.279, 280 Cl CO2(CH2)2N OMe H2N 168 Cl C(O)(CH2)2 NR2 R1=Me: R2=Bu (a), (CH2)2NHSO2Me (b); R1=CH2C6H3(OMe)2-3,5, R2=(CH2)2NHSO2Me (c).H2N OR1 169a ± c Cl CO2(CH2)3N OMe H2N 170 These and some other examples of this kind, e.g., the fact that metoclopramide (141) is a partial agonist of 5-HT4 receptors and its ester analogue 164 is a 5-HT4 antagonist, suggest that even minor structural changes may provoke dramatic changes in the intrinsic activities of 5-HT4 ligands. A hypothesis of the existence of two binding sites in the 5-HT4 receptor was developed to rationalise this fact.277350 Substitution of an oxygen-containing ring for the o-alkoxy group in the piperidine-substituted metoclopramide derivatives (141) was efficient in the design of both 5-HT3 and 5-HT4 antagonists. Thus chlorine-substituted benzodioxanes 171a,b 281 and the iodo derivative 171c 282, 283 proved to be selective 5-HT4 antagonists with affinities exceeding that of compound 164 1000- fold and even more.O XCH2 OO R1 171a ± c NH2 Substitution of the indole fragment for the benzene rings in many metoclopramide derivatives mentioned above was also effective. Compounds 172 and 173 284, 285 and their tricyclic analogue 174 285 proved to be high-affinity selective 5-HT4 antagonists. The attempts to modify the indole fragment revealed that benzoimidazoles 175 containing the piperazine fragment in their side chains manifest high affinities for 5-HT4 receptors. Interestingly, compound 175 (where R=cyclo-C3H5) behaved as a potent antagonist, whereas that with R=Pri was a partial agonist manifesting very high intrinsic activity.286 This is yet another example of reversal of the pharmacological profile of structurally related compounds, which provides additional evi- dence in favour of the hypothesis postulating the existence of two binding sites in the 5-HT4 receptor.CO2(CH2)2N NH 172 CONHCH2 O N 174 O NH(CH2)2N N O NR 175 Conformational analysis of serotonin and zacopride (143) which also manifests agonistic activity towards its receptor disclosed very potent and selective 5-HT4 agonists among indole derivatives close to serotonin and made it possible to propose a model of a binding site for 5-HT4 agonists corresponding to the serotonin analogue 176a.The latter appeared to be a high-affinity full agonist of 5-HT4 receptors towards which it was more selective than to other serotonin and dopaminergic recep- tors.287, 288 The affinity of this compound can be increased through the introduction of lipophilic substituents into the side chain. Thus the derivatives 176b,c are full agonists with the affinities exceeding that of serotonin severalfold. Compound 176d binds to the receptor 300 times more effectively than serotonin but manifests the properties of a partial agonist.These compounds are among the most potent and selective 5-HT4 agonists presently available.287, 288 Some 5-HT4 ligands were found among the derivatives of quipazine (125a), which also represents a 5-HT3 ligand. The majority of compounds of this kind, e.g., compound 177, manifest partial agonistic activity towards 5-HT4 receptors and simulta- neously act as 5-HT3 antagonists 289 which makes them promising candidates for gastrokinetic agents. NR2 R1=Cl, R2=Bun, X=O(a); R1=Cl, R2=(CH2)3C6H3(OMe)2-3,4, X=CH2 (b); R1=I, R2=Bun, X=O(c). NX CO2CH2 NMe 173 X=(CH2)2NHSO2Me. NBun NMe O N Zefirova, N S Zefirov NR HN NH2 N HO NH 176a ± d R = H (a), C5H11 (b), (CH2)2Ph (c), (CH2)2C6H3Cl2-3,4 (d). O N N N O S 177 It should be stressed in conclusion that despite the construc- tion of a fairly large number of potent and sufficiently selective 5-HT4 ligands, they manifest predominantly antagonistic activ- ities.Full selective agonists are fewer in number, but are ther- apeutically more valuable. Moreover, very many 5-HT4 ligands do not possess high pharmacodynamic characteristics needed for their clinical application. It is for this reason that the main attention has presently been focussed on those modifications of the ligand structure which might improve their pharmacodynamic characteristics. To this end, amide, carbamate or ketone groups are introduced into the side chains of compounds 170 instead of ester groups.290, 291 Among compounds with improved pharma- codynamic characteristics, compounds 171b and 174 and the recently prepared compound 178,290, 292 a long-acting, potent and selective 5-HT4 antagonist, deserve special mentioning.NHCO O NH(CH2)2N PriN N 178 IX. Ligands of serotonin receptors of other subtypes Among other subtypes of serotonin receptors, 5-HT5 (further classified into 5-HT5A and 5-HT5B), 5-HT6 and 5-HT7 receptors are the least known. Highly selective ligands to 5-HT5 and 5-HT6 receptors have not been identified yet; therefore, their functional and, perhaps, therapeutic values have been studied relatively little. It is suggested that 5-HT7 receptors play a role in the pathophysi- ology of certain sleep disturbances, depressive states and schizo- phrenia; also their stimulation causes vasodilation.Studies of the effects of many classical ligands of serotonin receptors on these subtypes revealed that 5-carboxamidotrypt- amine (2) and LSD (97) manifested high, whereas compound 4a manifested moderate, affinities for 5-HT5A and 5-HT7 recep- tors.293 Many non-selective 5-HT2 antagonists, such as methio- thepin (68), methergoline (98), ritanserin (104), etc., bind to 5-HT6 and 5-HT7 receptors manifesting high or moderate affinities. Very strong binding is observed in the interaction of some mixed-action antipsychotic drugs, e.g., clozapine (120a) and pimozide (179), with these receptors.294 (4-FC6H4)2CH(CH2)3N N NH O 179 It is not until very recently that a series of studies devoted to the design of selective ligands of some of the above-mentioned receptor groups have been published.295, 296 Thus it was found that 5-HT6 receptors match small alkyl substituents in position 2 of the indole ring of serotonin.The ethyl derivative 180 which representsPhysiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors a full 5-HT6 agonist manifests the highest affinity and binds to the receptor 5 times more strongly than serotonin. Compound 180 manifests only moderate affinity for some other types of serotonin receptors, being the most selective 5-HT6 agonist known so far. Substitution of the phenyl group for the ethyl substituent does not influence the affinity but decreases dramatically the intrinsic activity.These data suggest that 2-substituted tryptamines can be used in the design of both selective agonists and antagonists of 5-HT6 receptors.296(CH2)2NMe2 MeO Et NH 180 A selective ligand of 5-HT7 receptors was prepared based on sulfamide 181, which has been found as a result of wide screening and manifesting moderate affinity and weak selectivity towards 5-HT7 receptors; its R,R-configuration was found to be important for binding. Modifications of compound 181 revealed that the methyl group in the piperidine ring may be shifted to position 4, whereas the naphthalene ring may be replaced by other aromatic substituents. Thus sulfamide 182 appeared to be two orders of magnitude more selective towards 5-HT7 receptors than to other serotonin receptors.Compound 182 is the first example of an effective selective ligand endowed with 5-HT7 antagonistic activ- ity, although its affinity for these receptors is not very high.295 Me N (CH2)2N O2S Me Me 181 Me Me N (CH2)2N O2S Me 182 Me * * * A vast variety of compounds interacting with serotonin receptors with high affinity have been prepared by the methods employed in medicinal chemistry. Selectivity is the major problem in the design of new ligands, since the family of serotonin receptors is charac- terised by a large set of structurally similar subclasses. Although this problem can successfully be overcome for some of 5-HT subtypes, for other 5-HT receptors, particularly for 5-HT1E, 5-HT1F, 5-HT5A, 5-HT5B, 5HT6 and 5-HT7, the design of selective ligands presents a challenge.Its successful solution will help better understanding of the functional roles of receptors in the organism. Obviously, progress in the search for ligands to serotonin recep- tors will depend on further advances in X-ray diffraction analysis, which will provide a basis for refining of computer models and docking studies of ligands pertaining to different structural classes. References 1. O N Zefirova, N S Zefirov Vestn. Mosk. Univ., Ser. 2, Khim. 41 43 (2000) a 2. O N Zefirova, N S Zefirov Vestn. Mosk. Univ., Ser. 2, Khim. 41 103 (2000) a 3. O N Zefirova, N S Zefirov Zh. Org. Khim. 36 1273 (2000) b 4. D Hoyer, G R Martin Behav. Brain Res. 73 263 (1996) 351 5.D Hoyer, G Martin Neuropharmacology 36 419 (1997) 6. L Uphouse Neurosci. Biobehav. Rev. 21 679 (1997) 7. P Blier, C de Montigny Trends Pharmacol. Sci. 15 220 (1994) 8. A Fletcher, I A Cliffe, C T Dourish Trends Pharmacol. Sci. 14 41 (1993) 9. M Hamon Trends Pharmacol. Sci. 15 36 (1994) 10. M S Beer, J A Stanton, Y Bevan, N S Chauhan, D N Middlemiss Eur. J. Pharmacol. 213 193 (1992) 11. E Zifa, G Fillion Pharmacol. Rev. 44 401 (1992) 12. D N Middlemiss, J R Fozard Eur. J. Pharmacol. 90 151 (1983) 13. D R Helton, W E Colbert J. Pharm. Pharmacol. 46 902 (1994) 14. H Yu, Y Liu, U Hacksell, T Lewander Eur. J. Pharmacol. 231 69 (1993) 15. Z-P Zhuang, M-P Kung, H F Kung J. Med. Chem. 36 3161 (1993) 16. L-E Arvidsson,AMJohansson,UHacksell, J L Nilsson,KSvensson, S Hjorth, T Magnusson, A Carlsson, B Andersson, H WikstroÈ m J.Med. Chem. 30 2105 (1987) 17. C Mellin, J Vallgarda, D L Nelson, L BjoÈ rk, H Yu, N-E Anden, I CsoÈ regh, L-E Arvidsson, U Hacksell J. Med. Chem. 34 497 (1991) 18. L J Cornfield, G Lambert, L E Arvidsson, C Mellin, J Vallgarda, U Hacksell, D L Nelson Mol. Pharmacol. 39 780 (1991) 19. H WikstroÈ m, B Andersson, T Elebring, J Jacyno, N L Allinger, K Svensson, A Carlsson, S Sundell J. Med. Chem. 30 1567 (1987) 20. B B HoÈ oÈ k, L Cortizo, A M Johansson, A Westlind-Danielsson, N Mohell, U Hacksell J. Med. Chem. 39 4036 (1996) 21. L BjoÈ rk, L J Cornfield, D L Nelson, S E Hillver, N E Anden, T Lewander, U Hacksell J. Pharmacol.Exp. Ther. 258 58 (1991) 22. S-E Hillver, L BjoÈ rk, Y-L Li, B Svensson, S Ross, N-E Anden, U Hacksell J. Med. Chem. 33 1541 (1990) 23. J Vallgarda, U Appelberg, L E Arvidsson, S Hjorth, B E Svensson, U Hacksell J. Med. Chem. 39 1485 (1996) 24. M H Hedberg, T Linnanen, J M Jansen, G Nordvall, S Hjorth, L Unelius, A M Johansson J. Med. Chem. 39 3503 (1996) 25. M H Hedberg, J M Jansen, G Nordvall, S Hjorth, L Unelius, A M Johansson J. Med. Chem. 39 3491 (1996) 26. G S Baxter, D A Craig, D E Clarke Naunyn Schmiedebergs Arch. Pharmacol. 343 439 (1991) 27. R D Titus, E C Kornfeld, N D Jones, J A Clemens, E B Smalstig, R W Fuller, R A Hahn,M D Hynes, N R Mason, D T Wong, M M Foreman J. Med. Chem. 26 1112 (1983) 28. R F Heier, L A Dolak, J N Duncan, D K Hyslop,M F Lipton, I J Martin,M A Mauragis,M F Piercey, N F Nichols, P J K D Schreur, M W Smith, M W Moon J.Med. Chem. 40 639 (1997) 29. J M Cossery, H Gozlan, U Spampinato, C Perdicakis, G Guillaumet, L Pichat, M Hamon Eur. J. Pharmacol. 140 143 (1987) 30. E Hammarberg, G Nordvall, R Leideborg, M Nylof, S Hanson, L Johansson, S-O Thorberg, B R Tolf, E Jerning, G T Svantesson, N Mohell, C Ahlgren, A Westlind-Danielsson, I CsoÈ regh, R Johansson J. Med. Chem. 43 2837 (2000) 31. K L Goa, A Ward Drugs 32 114 (1986) 32. A Gobert, J-M Rivet, L Cistarelli, C Melon,M J Millan Neuroscience 93 1251 (1999) 33. B J Cao, R J Rodgers Psychopharmacology (Berlin) 139 185 (1998) 34. D P Taylor, S L Moon Neuropeptides 19 (Suppl.) 15 (1991) 35.K Ishizumi, A Kojima, F Antoku Chem. Pharm. Bull. 39 2288 (1991) 36. I van Wijngaarden,M T M Tulp,W Soudijn Eur. J. Pharmacol. 188 301 (1990) 37. M Abou-Gharbia, U R Patel, M B Webb, J A Moyer, T H Andree, E A Muth J. Med. Chem. 31 1382 (1988) 38. R K Raghupathi, L Rydelek-Fitzgerald,M Teitler, R A Glennon J. Med. Chem. 34 2633 (1991) 39. B J van Steen, I van Wijngaarden,M T M Tulp, W Soudijn J. Med. Chem. 36 2751 (1993) 40. B J van Steen, I van Wijngaarden,M T M Tulp, W Soudijn J. Med. Chem. 37 2761 (1994) 41. J L Mokrosz, M Pietrasiewicz, B Duszynska,M T Cegla J. Med. Chem. 35 2369 (1992) 42. R A Glennon, S -S Hong,M Dukat,M Teitler, K Davis J. Med. Chem. 37 2828 (1994) 43. R A Glennon, S S Hong, M Bondarev, H Law,M Dukat, S Rakhit, P Power, E Fan, D Kinneau, R Kamboj, M Teitler, K Herrick-Davis, C Smith J.Med. Chem. 39 314 (1996)352 44. J M Greuel, T Glaser Eur. J. Pharmacol. 211 211 (1992) 45. T Sharp, L I Backus, S Hjorth, S R Bramwell, D G Grahame- Smith Eur. J. Pharmacol. 176 331 (1990) 46. H T Yang,M Endoh Br. J. Pharmacol. 122 1541 (1997) 47. T Matsuda, Y H Seong, H Aono, T Kanda, A Baba, K Saito, A Tobe, H Iwata Eur. J. Pharmacol. 170 75 (1989) 48. G Baxter,G Kennett, F Blaney, T Blackburn Trends Pharmacol. Sci. 16 105 (1995) 49. Y El Ahmad, E Laurent, P Maillet, A Talab, J F Teste, R Dokhan, G Tran, R Ollivier J. Med. Chem. 40 952 (1997) 50. A Orjales, L Alonso-Cires, L Labeaga,R Corcostegui J. Med. Chem. 38 1273 (1995) 51. J-L Peglion, H Canton, K Bervoets, V Audinot,M Brocco, A Gobert, S Le Marouille-Girardon,M J Millan J.Med. Chem. 38 4044 (1995) 52. M L Lopez-Rodriguez,M L Rosado, B Benhamu, M J Morcillo, A M Sanz, L Orensanz, M E Beneitez, J A Fuentes, J Manzanares J. Med. Chem. 39 4439 (1996) 53. M L Lopez-Rodriguez,M J Morcillo, T K Rovat, E Fernandez, B Vicente, A M Sanz,M Hernandez, L Orensanz J. Med. Chem. 42 36 (1999) 54. G Griebel, R J Rodgers, G Perrault, D J Sanger Psychopharmacology (Berlin) 144 121 (1999) 55. G Griebel, R J Rodgers, G Perrault, D J Sanger Neuropharmacology 39 1848 (2000) 56. T Podona, B Guardiola-Lemaitre, D -H Caignard, G Adam, B Pfeiffer, P Renard, G Guillaumet J. Med. Chem. 37 1779 (1994) 57. R T Kroemer, E Koutsilieri, P Hecht, K R Liedl, P Riederer, J Kornhuber J.Med. Chem. 41 393 (1998) 58. H Kubinyi QSAR: Hansch Analysis and Related Approaches. (Eds R Mannhold, P Krogsgaard-Larsen, H Timmerman) (Weinheim: VCH, 1993) 59. H Kubinyi, in The Quantitative Analysis of Structure-Activity Relationships Vol. 1 (Ed.M E Wolff) (New York: Wiley, 1995) p. 497 60. C Comoy, C Marot, T Podona, M-L Baudin, L Morin-Allory, G Guillaumet, B Pfeiffer, D-H Caignard, P Renard,M -C Rettori, G Adam, B Guardiola-Lemaitre J. Med. Chem. 39 4285 (1996) 61. S Buisson-Defferier, M van den Buuse Eur. J. Pharmacol. 223 133 (1992) 62. S E Gartside, P J Cowen, S Hjorth Eur. J. Pharmacol. 191 391 (1990) 63. P van den Hooff , M Galvan Eur. J. Pharmacol. 196 291 (1991) 64. R E Mewshaw, J Kavanagh, G Stack, K L Marquis, X Shi, M Z Kagan,M B Webb, A H Katz, A Park, Y H Kang,M Abou- Gharbia, R Scerni, T Wasik, L Cortes-Burgos, T Spangler, J A Brennan,M Piesla, H Mazandarani,M I Cockett, R Ochalski, J Coupet, T H Andree J.Med. Chem. 40 4235 (1997) 65. W Quaglia,M Pigini, A Piergentili, M Giannella, G Marucci, E Poggesi, A Leonardi, C Melchiorre J. Med. Chem. 42 2961 (1999) 66. W Kuipers, C G Kruse, I van Wijngaarden, P J Standaar, M T M Tulp, N Veldman, A L Spek, A P Ijzerman J. Med. Chem. 67. B Vacher, B Bonnaud, P Funes, N Jubault,W Koek, M-B Assie, 68. B Vacher, B Bonnaud, P Funes, N Jubault,W Koek, M-B Assie, 40 300 (1997) C Cosi J. Med. Chem. 41 5070 (1998) C Cosi, M Kleven J. Med. Chem. 42 1648 (1999) 69. A M Birch, P A Bradley, J C Gill, F Kerrigan, P L Needham J.Med. Chem. 42 3342 (1999) 70. H R Howard, J A Lowe III, T F Seeger, P A Seymour, S H Zorn, P R Maloney, F E Ewing,M E Newman, A W Schmidt, J S Furman, G L Robinson, E Jackson, C Johnson, J Morrone J. Med. Chem. 39 143 (1996) 71. D Wustrow, T Belliotti, S Glase, S R Kesten,D Johnson,N Colbry, R Rubin,A Blackburn,HAkunne, A Corbin,MDDavis, L Georgic, S Whetzel, K Zoski, T Heffner, T Pugsley, L Wise J. Med. Chem. 41 760 (1998) 72. E M Clifford, S E Gartside, V Umbers, P J Cowen,M Hajos, T Sharp Br. J. Pharmacol. 124 206 (1998) 73. R Corradetti, N Laaris, N Hanoun, A M Laporte, E Le Poul, M Hamon, L Lanfumey Br. J. Pharmacol. 123 449 (1998) 74. H Giles, S J Lansdell,M L Bolofo, H L Wilson, G R Martin Br.J. Pharmacol. 117 1119 (1996) 75. A Hoey, C Jackson, G Pegg, M Sillence Br. J. Pharmacol. 117 712 (1996) O N Zefirova, N S Zefirov 76. A Fletcher, E A Forster, D J Bill, G Brown, I A Cliffe, J E Hartley, D E Jones, A McLenachan, K J Stanhope, D J Critchley, K J Childs, V C Middlefell, L Lanfumey, R Corradetti, A M Laporte, H Gozlan, M Hamon, C T Dourish Behav. Brain Res. 73 337 (1996) 77. E A Forster, I A Cliffe, D J Bill, G M Dover, D Jones, Y Reilly, A Fletcher Eur. J. Pharmacol. 281 81 (1995) 78. I A Cliffe, C I Brightwell, A Fletcher, E A Forster, H L Mansell, Y Reilly, C Routledge, A C White J. Med. Chem. 36 1509 (1993) 79. D B Bylund, H S Blaxall, L J Iversen,MG Caron, R J Lefkowitz, J W Lomasney Mol. Pharmacol. 42 1 (1992) 80.D L Nelson, E W Taylor Eur. J. Pharmacol. 124 207 (1986) 81. T Yasunaga, R Naito, T Kontani, S Tsukamoto, T Nomura, T Yamaguchi, T Mase J. Med. Chem. 40 1252 (1997) 82. T Yasunaga, T Kimura, R Naito, T Kontani, F Wanibuchi, H Yamashita, T Nomura, S Tsukamoto, T Yamaguchi, T Mase J. Med. Chem. 41 2765 (1998) 83. M J Mokrosz, B Duszynska, A J Bojarski, J L Mokrosz Bioorg. Med. Chem. 3 533 (1995) 84. JM Zgombick, L E Schechter, S A Kucharewicz, R L Weinshank, T A Branchek Eur. J. Pharmacol. 291 9 (1995) 85. Y-C Xu, J M Schaus, C Walker, J Krushinski, N Adham, J M Zgombick, S X Liang, D T Kohlman, J E Audia J. Med. Chem. 42 526 (1999) 86. L M Gaster, F E Blaney, S Davies, D M Duckworth, P Ham, S Jenkins, A J Jennings, G F Joiner, F D King, K R Mulholland, P A Wyman, J J Hagan, J Hatcher, B J Jones, D N Middlemiss, G W Price,G Riley,C Roberts, C Routledge, J Selkirk, P D Slade J.Med. Chem. 41 1218 (1998) 87. J Longmore, C M Boulanger, B Desta, R G Hill, W N Schofield, A A Taylor Br. J. Clin. Pharmacol. 42 431 (1996) 88. C Jorand-Lebrun, P J Pauwels, C Palmier, C Moret, P Chopin, M Perez,M Marien, S Halazy J. Med. Chem. 40 3974 (1997) 89. M Lamothe, P J Pauwels, K Belliard, P Schambel, S Halazy J. Med. Chem. 40 3542 (1997) 90. M D Ferrari, P R Saxena Trends Pharmacol. Sci. 14 129 (1993) 91. K L Dechant, S P Clissold Drugs 43 776 (1992) 92. C M Perry, A Markham Drugs 55 889 (1998) 93. K W Weitzel, M L Thomas, R E Small, J V Goode Pharmacotherapy 19 957 (1999) 94.R N Iyer, C W Bradberry J. Pharmacol. Exp. Ther. 277 40 (1996) 95. M D Lee, V J Aloyo, S J Fluharty, K J Simansky Psychopharmacology (Berlin) 136 304 (1998) 96. A C Trillat, I Malagie, K Scearce, D Pons,M C Anmella, C Jacquot, R Hen, A M Gardier J. Neurochem. 69 2019 (1997) 97. M J Gawel Can. J. Clin. Pharmacol. 6 20A (1999) 98. R Salonen Int. J. Clin. Pract. 53 552 (1999) 99. L J Street, R Baker,W B Davey, A R Guiblin, R A Jelley, A J Reeve, H Routledge, F Sternfeld, A P Watt,M S Beer, D N Middlemiss, A J Noble, J A Stantor, K Scholey, R J Hargreaves, B Sohal,M I Graham, V G Matassa J. Med. Chem. 38 1799 (1995) 100. D Deleu, Y Hanssens J. Clin. Pharmacol. 40 687 (2000) 101. C M Spencer, N S Gunasekara, C Hills Drugs 58 347 (1999) 102.R M Gallagher, G Dennish, E L Spierings, R Chitra Headache 40 119 (2000) 103. A Al Shekhlee, R Reed Neurology 55 735 (2000) 104. P R Saxena, P De Vries, W Wang, J P Heiligers, v A Maassen, W A Bax, F D Yocca Naunyn Schmiedebergs Arch. Pharmacol. 355 295 (1997) 105. H C Diener, H Kaube, V Limmroth J. Neurol. 246 515 (1999) 106. A A Parsons, R Valocik, P Koster, P Raval, R Gagnon, N Tilford, G Feuerstein J. Cardiovasc. Pharmacol. 32 995 (1998) 107. M Perez, C Fourrier, I Sigogneau, P J Pauwels, C Palmier, G W John, J-P Valentin, S Halazy J. Med. Chem. 38 3602 (1995) 108. M S Beer, J A Stanton, Y Bevan, A Heald, A J Reeve, L J Street, V G Matassa, R J Hargreaves, D N Middlemiss Br. J. Pharmacol. 110 1196 (1993) 109. A Heald, J A Stanton, S A Osborne, D N Middlemiss, M S Beer Eur.J. Pharmacol. 264 213 (1994) 110. L J Street, R Baker, J L Castro, M S Chambers, A R Guiblin, S C Hobbs, V G Matassa, A J Reeve, MS Beer, D N Middlemiss, A J Noble, J A Stanton, K Scholey, R J Hargreaves J. Med. Chem. 36 1529 (1993)Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors 111. S J Starkey,M Skingle Neuropharmacology 33 393 (1994) 112. J W Clitherow, D I Scopes,M Skingle, C C Jordan,W Feniuk, I B Campbell, M C Carter, E W Collington, H E Connor, G A Higgins, D Beattie, H A Kelly,W C Mitchell, A W Oxford, A H Wadsworth,M B Tyers J. Med. Chem. 37 2253 (1994) 113. H Law, MDukat,MTeitler, D K H Lee, L Mazzocco, R Kamboj, V Rampersad, T Prisinzano, R A Glennon J.Med. Chem. 41 2243 (1998) 114. C Sonesson, T Barf, J Nilsson, D Dijkstra, A Carlsson, K Svensson, M W Smith, I J Martin, J N Duncan, L J King, H WikstroÈ m J. Med. Chem. 38 1319 (1995) 115. T A Barf, P de Boer, H WikstroÈ m, S J Peroutka, K Svensson, M D Ennis, N B Ghazal, J C McGuire, M W Smith J. Med. Chem. 39 4717 (1996) 116. J L Castro, L J Street, A R Guiblin, R A Jelley,M G Russell, F Sternfeld,M S Beer, J A Stanton, V G Matassa J. Med. Chem. 40 3497 (1997) 117. F Sternfeld, A R Guiblin, R A Jelley, V G Matassa, A J Reeve, P A Hunt,M S Beer, A Heald, J A Stanton, B Sohal, A P Watt, L J Street J. Med. Chem. 42 677 (1999) 118. M S Chambers, L J Street, S Goodacre, S C Hobbs, P Hunt, R A Jelley, V G Matassa, A J Reeve, F Sternfeld,M S Beer, J A Stanton, D Rathbone, A P Watt, A M MacLeod J.Med. Chem. 42 691 (1999) 119. M B van Niel, I Collins,M S Beer,H B Broughton, S K F Cheng, S C Goodacre, A Heald, K L Locker, AMMacLeod, D Morrison, C R Moyes, D O'Connor, A Pike, M Rowley, M G N Russell, B Sohal, J A Stanton, S Thomas,H Verrier, A P Watt, J L Castro J. Med. Chem. 42 2087 (1999) 120. J L Castro, I Collins,M G N Russell, A P Watt, B Sohal, D Rathbone,M S Beer, J A Stanton J. Med. Chem. 41 2667 (1998) 121. P Schoeffter, D Hoyer Naunyn Schmiedebergs Arch. Pharmacol. 339 675 (1989) 122. F Chaouloff, S Aguerre, P Mormede Neuropharmacology 37 1159 (1998) 123. R F Neale, S L Fallon, W C Boyar, J W Wasley, L L Martin, G A Stone, B S Glaeser, C M Sinton,M Williams Eur.J. Pharmacol. 136 1 (1987) 124. J E Leysen, W Gommeren, L Heylen, W H Luyten, I van de Weyer, P Vanhoenacker, G Haegeman, A Schotte, P van Gompel, R Wouters, A S Lesage Mol. Pharmacol. 50 1567 (1996) 125. E Schlicker, U Werner, M Hamon, H Gozlan, B Nickel, I Szelenyi, M Gothert Br. J. Pharmacol. 105 732 (1992) 126. M A Petty, J Elands,M P Johnson, M D Linnik, E Hamel, M A Moskowitz, W S Lee, D R McCarty, M Hibert, B M Baron Eur. J. Pharmacol. 336 127 (1997) 127. P Schoeffter, D Hoyer Eur. J. Pharmacol. 196 213 (1991) 128. A M Ismaiel, M Dukat, H Law, R Kamboj, E Fan, D K H Lee, L Mazzocco, D Buekschkens, M Teitler, M E Pierson, R A Glennon J. Med. Chem. 40 4415 (1997) 129. P C Waldmeier,M Williams, P A Baumann, S Bischoff,M A Sills, R F Neale Naunyn Schmiedebergs Arch. Pharmacol.337 609 (1988) 130. P J Pauwels, F C Colpaert Neuropharmacology 34 235 (1995) 131. J M Watson,M J Burton,G W Price, B J Jones,D N Middlemiss Eur. J. Pharmacol. 314 365 (1996) 132. R C Glen, G R Martin, A P Hill, R M Hyde, P M Woollard, J A Salmon, J Buckingham, A D Robertson J. Med. Chem. 38 3566 (1995) 133. G P Moloney, A D Robertson, G R Martin, S MacLennan, N Mathews, S Dodsworth, P Y Sang, C Knight, R Glen J. Med. Chem. 40 2347 (1997) 134. G P Moloney, G R Martin, N Mathews, A Milne, H Hobbs, S Dodsworth, P Y Sang, C Knight,M Williams, M Maxwell, R C Glen J. Med. Chem. 42 2504 (1999) 135. M R MacLean, R A Clayton, A G Templeton, I Morecroft Br. J. Pharmacol.119 277 (1996) 136. C Roberts, G W Price, L Gaster, B J Jones, D N Middlemiss, C Routledge Neuropharmacology 36 549 (1997) 137. I T Forbes, S Dabbs, D M Duckworth, P Ham, G E Jones, F D King, D V Saunders, F E Blaney, C B Naylor, G S Baxter, T P Blackburn, G A Kennett,M D Wood J.Med. Chem. 39 4966 (1996) 353 138. E Schlicker, K Fink, G J Molderings, G W Price, M Duckworth, L Gaster, D N Middlemiss, J Zentner, J Likungu, M Gothert Naunyn Schmiedebergs Arch. Pharmacol. 356 321 (1997) 139. S Leonhardt,K Herrick-Davis,M Titeler J. Neurochem. 53 465 (1989) 140. E M Parker, D G Izzarelli, L Lewis-Higgins, D Palmer, R A Shapiro J. Neurochem. 67 2096 (1996) 141. S Shepheard, L Edvinsson, M Cumberbatch, D Williamson, G Mason, J Webb, S Boyce, R Hill, R Hargreaves Cephalalgia 19 851 (1999) 142.D S Dupuis, F C Colpaert, P J Pauwels Br. J. Pharmacol. 124 283 (1998) 143. KWJohnson, JMSchaus,MMDurkin, J E Audia, SWKaldor, ME Flaugh,NAdham, JMZgombick,ML Cohen, T A Branchek, 144. B P Richardson, G Engel, P Donatsch, P A Stadler Nature 145. H Wilson,W J Coffman,M L Cohen J. Pharmacol. Exp. Ther. 252 146. M S Duxon, G A Kennett, S Lightowler, T P Blackburn, L A Phebus Neuroreport 8 2237 (1997) (London) 316 126 (1985) 683 (1990) K C Fone Neuropharmacology 36 601 (1997) 147. G A Kennett, K Ainsworth, B Trail, T P Blackburn Neuropharmacology 36 233 (1997) 148. G A Kennett, B Trail, F Bright Neuropharmacology 37 1603 (1998) 149. R A Glennon, M Dukat,M El-Bermawy, H Law, J De los Angeles, M Teitler, A King, K Herrick-Davis J.Med. Chem. 37 1929 (1994) 150. J B Blair, D Marona-Lewicka, A Kanthasamy, V L Lucaites, D L Nelson, D E Nichols J. Med. Chem. 42 1106 (1999) 151. M Gerasimov, D Marona-Lewicka, D M Kurrasch-Orbaugh, A M Qandil, D E Nichols J. Med. Chem. 42 4257 (1999) 152. J E Macor, J Blake, C B Fox, C Johnson, B K Koe, L A Lebel, J M Morrone, K Ryan, A W Schmidt, D W Schulz, S H Zorn J. Med. Chem. 35 4503 (1992) 153. S Vangveravong, D E Nichols J. Org. Chem. 60 3409 (1994) 154. S Vangveravong, A Kanthasamy, V L Lucaites, D L Nelson, D E Nichols J. Med. Chem. 41 4995 (1998) 155. M BoÈ s, F Jenck, J R Martin, J L Moreau, A J Sleight, J Wichmann, U Widmer J. Med. Chem. 40 2762 (1997) 156. G Curzon, G A Kennett Trends Pharmacol.Sci. 11 181 (1990) 157. K C Fone, R H Austin, I A Topham, G A Kennett, T Punhani Br. J. Pharmacol. 123 1707 (1998) 158. I Lucki,H R Ward, A Frazer J. Pharmacol. Exp. Ther. 249 155 (1989) 159. Y Sugimoto, J Yamada, T Yoshikawa, K Horisaka Eur. J. Pharmacol. 307 75 (1996) 160. P J Conn, E Sanders-Bush J. Pharmacol. Exp. Ther. 242 552 (1987) 161. S K Hemrick-Luecke, R W Fuller Eur. J. Pharmacol. 311 207 (1996) 162. R A Glennon, J M Jacyno, R Young, J D McKenney, D Nelson J. Med. Chem. 27 41 (1984) 163. R A Glennon, R Raghupathi, P Bartyzel, M Teitler, S Leonhardt J. Med. Chem. 35 734 (1992) 164. D E Nichols, S Frescas,D Marona-Lewicka,X Huang, B L Roth, G A Gudelsky, J F Nash J. Med. Chem. 37 4346 (1994) 165. M R Seggel, M Y Yousif, R A Lyon,M Titeler, B L Roth, E A Suba, R A Glennon J.Med. Chem. 33 1032 (1990) 166. A P Monte, D Marona-Lewicka, M A Parker, D B Wainscott, D L Nelson, D E Nichols J. Med. Chem. 39 2953 (1996) 167. A P Monte, S R Waldman, D Marona-Lewicka, D B Wainscott, D L Nelson, E Sanders-Bush, D E Nichols J. Med. Chem. 40 2997 (1997) 168. A P Monte, D Marona-Lewicka, M M Lewis, R B Mailman, D B Wainscott, D L Nelson, D E Nichols J. Med. Chem. 41 2134 169. M A Parker, D Marona-Lewicka, V L Lucaites, D L Nelson, 170. D E Nichols, S E Snyder, R Oberlender,M P Johnson, X Huang (1998) D E Nichols J. Med. Chem. 41 5148 (1998) J. Med. Chem. 34 276 (1991) 171. M P Johnson, C A Mathis, A T Shulgin, A J Hoffman, D E Nichols Pharmacol.Biochem. Behav. 35 211 (1990) 172. R A Glennon, M Titeler, J D McKenney Life Sci. 35 2505 (1984) 173. J E Audia, D A Evrard, G R Murdoch, J J Droste, J S Nissen, K W Schenck, P Fludzinski, V L Lucaites, D L Nelson, M L Cohen J. Med. Chem. 39 2773 (1996)354 174. M S Choudhary, N Sachs, A Uluer, R A Glennon, R B Westkaemper, B L Roth Mol. Pharmacol. 47 450 (1995) 175. R A Glennon Neuropsychopharmacology 3 509 (1990) 176. M L Cohen, C J Parli, R W Fuller J. Pharmacol. Exp. Ther. 251 1006 (1989) 177. C S Aulakh, J L Hill, D L Murphy J. Pharmacol. Exp. Ther. 263 588 (1992) 178. G Bagdy, A F Sved, D L Murphy, K Szemeredi Eur. J. Pharmacol. 210 285 (1992) 179. M L Cohen, R W Fuller, K D Kurz, C J Parli, N R Mason, D B Meyers, J K Smallwood, R E Toomey J.Pharmacol. Exp. 180. M L Cohen, D W Robertson, W E Bloomquist, H C Wilson Ther. 244 106 (1988) J. Pharmacol. Exp. Ther. 261 202 (1992) 181. J W Misner, W L Garbrecht, G Marzoni, K R Whitten, M L Cohen J. Med. Chem. 33 652 (1990) 182. H H Pertz, H -C Milhahn, E Eich J. Med. Chem. 42 659 (1999) 183. M P Johnson, J E Audia, J S Nissen, D L Nelson Eur. J. Pharmacol. 239 111 (1993) 184. D L Nelson, V L Lucaites, J E Audia, J S Nissen, D B Wainscott J. Pharmacol. Exp. Ther. 265 1272 (1993) 185. M P Johnson, R J Loncharich, M Baez, D L Nelson Mol. Pharmacol. 45 277 (1994) 186. H T Kao,N Adham,M A Olsen,R L Weinshank, T A Branchek, P R Hartig FEBS Lett. 307 324 (1992) 187. W L Garbrecht, G Marzoni, K R Whitten,M L Cohen J.Med. Chem. 31 444 (1988) 188. H O Kalkman, V Neumann, J Nozulak,M D Tricklebank Eur. J. Pharmacol. 343 201 (1998) 189. P O Mora, C F Netto, F G Graeff Pharmacol. Biochem. Behav. 58 1051 (1997) 190. J Nozulak, H O Kalkman, P Floersheim, D Hoyer, P Schoeffter, H R Buerki J. Med. Chem. 38 28 (1995) 191. J E Leysen, F Awouters, L Kennis, P M Laduron, J Vandenberk, P A Janssen Life Sci. 28 1015 (1981) 192. P A Pierce, J Y Kim, S J Peroutka Naunyn Schmiedebergs Arch. Pharmacol. 346 4 (1992) 193. Z Razzaque, J Longmore, R G Hill Eur. J. Pharmacol. 283 199 (1995) 194. J M Van Nueten, P A Janssen, J Van Beek, R Xhonneux, T J Verbeuren, P M Vanhoutte J. Pharmacol. Exp. Ther. 218 217 (1981) 195. Y Watanabe, H Usui, S Kobayashi, H Yoshiwara, T Shibano, T Tanaka, Y Morishima,M Yasuoka,M Kanao J.Med. Chem. 35 196. F Darchen, D Scherman, P M Laduron, J P Henry Mol. 197. J L Herndon, A Ismaiel, S P Ingher,M Teitler, R A Glennon 189 (1992) Pharmacol. 33 672 (1988) J. Med. Chem. 35 4903 (1992) 198. C J Schmidt, V L Taylor, G M Abbate, T R Nieduzak J. Pharmacol. Exp. Ther. 256 230 (1991) 199. P Seeman, H H van Tol Trends Pharmacol. Sci. 15 264 (1994) 200. A M Ismaiel, J De los Angeles, M Teitler, S Ingher, R A Glennon J. Med. Chem. 36 2519 (1993) 201. K A Metwally,M Dukat, C T Egan, C Smith, A DuPre, C B Gauthier, K Herrick-Davis, M Teitler, R A Glennon J. Med. Chem. 41 5084 (1998) 202. R H Mach, J R Jackson, R R Luedtke, K J Ivins, P B Molinoff, R L Ehrenkaufer J. Med. Chem.35 423 (1992) 203. D W Bonhaus, K K Weinhardt,M Taylor, A DeSouza, PMMcNeeley, K Szczepanski, D J Fontana, J Trinh, C L Rocha, M W Dawson, L A Flippin, R M Eglen Neuropharmacology 36 621 (1997) 204. D W Bonhaus, C L Rocha,M W Dawson, R M Eglen Ann. N.Y. Acad. Sci. 861 269 (1998) 205. I T Forbes, G A Kennett, A Gadre, P Ham, C J Hayward, R T Martin,M Thompson,M D Wood, G S Baxter, A Glen, O E Murphy, B A Stewart, T P Blackburn J. Med. Chem. 36 1104 (1993) 206. I T Forbes, P Ham, D H Booth, R T Martin,M Thompson, G S Baxter, T P Blackburn, A Glen, G A Kennett,M D Wood J. Med. Chem. 38 2524 (1995) 207. I T Forbes, G E Jones, O E Murphy, V Holland, G S Baxter J. Med. Chem. 38 855 (1995) O N Zefirova, N S Zefirov 208. C S Dowd, K Herrick-Davis, C Egan, A DuPre, C Smith, M Teitler, R A Glennon J.Med. Chem. 43 3074 (2000) 209. A Y Deutch, B Moghaddam, R B Innis, J H Krystal, G K Aghajanian, B S Bunney, D S Charney Schizophr. Res. 4 121 (1991) 210. H Y Meltzer Psychopharmacology (Berlin) 99 (Suppl.) S18 (1989) 211. S Fox, J Brotchie Eur. J. Pharmacol. 301 27 (1996) 212. M Kuoppamaki, E Syvalahti, J Hietala Eur. J. Pharmacol. 245 179 (1993) 213. G Campiani, V Nacci, S Bechelli, SMCiani, A Garofalo, I Fiorini, H Wikstrlm, P de Boer, Y Liao, P G Tepper, A Cagnotto, T Mennini J. Med. Chem. 41 3763 (1998) 214. D Hoyer, D E Clarke, J R Fozard, P R Hartig, G R Martin, E J Mylecharane, P R Saxena, P P Humphrey Pharmacol. Rev. 46 157 (1994) 215. A Moran, C Velasco, T Salvador,M L Martin, L San Roman Br.J. Pharmacol. 113 1358 (1994) 216. B E Leonard Adv. Biochem. Psychopharmacol. 31 301 (1982) 217. R A Newton, J M Elliott J. Neurochem. 69 1031 (1997) 218. K Andersen, T Liljefors, K Gundertofte, J Perregaard, K P BoÈ gesoÈ J. Med. Chem. 37 950 (1994) 219. T Borowski,M Krol, E Broclawik, T C Baranowski, L Strekowski,M J Mokrosz J. Med. Chem. 43 1901 (2000) 220. A M Ismaiel, M Titeler, K J Miller, T S Smith, R A Glennon J. Med. Chem. 33 755 (1990) 221. D A Craig, R M Eglen, L K Walsh, L A Perkins, R L Whiting, D E Clarke Naunyn Schmiedebergs Arch. Pharmacol. 342 9 (1990) 222. S J Peroutka, A Hamik Eur. J. Pharmacol. 148 297 (1988) 223. N A Sharif, E H Wong, D N Loury, E Stefanich, A D Michel, R M Eglen, R L Whiting Br.J. Pharmacol. 102 919 (1991) 224. S J Ireland,M B Tyers Br. J. Pharmacol. 90 229 (1987) 225. R A Glennon, A M Ismaiel, B G McCarthy, S J Peroutka Eur. J. Pharmacol. 168 387 (1989) 226. A Bachy,M Heaulme, A Giudice, J C Michaud, I A Lefevre, J Souilhac, L Manara, M B Emerit, H Gozlan, M Hamon Eur. J. Pharmacol. 237 299 (1993) 227. E Edwards, E Hampton, C R Ashby, J Zhang, R Y Wang Brain Res. 733 21 (1996) 228. M Anzini, A Cappelli, S Vomero,G Giorgi, T Langer,M Hamon, N Merahi, B M Emerit, A Cagnotto, M Skorupska, T Mennini, 229. G J Kilpatrick, A Butler, J Burridge, A W Oxford Eur. J. 230. M I Sepulveda, S C Lummis, I L Martin Br. J. Pharmacol. 104 536 J C Pinto J. Med. Chem. 38 2692 (1995) Pharmacol. 182 193 (1990) (1991) 231.F G Boess, M I Sepulveda, S C Lummis, I L Martin Neuropharmacology 31 561 (1992) 232. M I Niemeyer, S C Lummis Br. J. Pharmacol. 123 661 (1998) 233. M Dukat, A A Abdel-Rahman, A M Ismaiel, S Ingher,M Teitler, L Gyermek, R A Glennon J. Med. Chem. 39 4017 (1996) 234. GCampiani, A Cappelli, V Nacci,MAnzini, S Vomero,MHamon, A Cagnotto, C Fracasso, C Uboldi, S Caccia, S Consolo, T Mennini J. Med. Chem. 40 3670 (1997) 235. H Prunier, S Rault, J-C Lancelot, M Robba, P Renard, P Delagrange, B Pfeiffer, D-H Caignard, R Misslin, B Guardiola-Lemaitre, M Hamon J. Med. Chem. 40 1808 (1997) 236. S Rault, J-C Lancelot, H Prunier, M Robba, P Renard, P Delagrange, B Pfeiffer, D-H Caignard, B Guardiola-Lemaitre, M Hamon J. Med. Chem. 39 2068 (1996) 237.G Campiani, E Morelli, S Gemma, V Nacci, S Butini, M Hamon, E Novellino, G Greco, A Cagnotto,M Goegan, L Cervo, F D Valle, C Fracasso, S Caccia, T Mennini J. Med. Chem. 42 4362 (1999) 238. A Cappelli, M Anzini, S Vomero, L Mennuni, F Makovec, E Doucet,M Hamon, G Bruni, M R Romeo, M C Menziani, P G de Benedetti, T Langer J. Med. Chem. 41 728 (1998) 239. A Cappelli, A Donati, M Anzini, S Vomero, P G de Benedetti, M C Menziani, T Langer Bioorg. Med. Chem. 4 1255 (1996) 240. A Cappelli, M Anzini, S Vomero, L Canullo, L Mennuni, F Makovec, E Doucet, M Hamon,M C Menziani, P G de Benedetti, G Bruni,M R Romeo, G Giorgi, A Donati J. Med. Chem. 42 1556 (1999) 241. A Monge, J A Palop, J C Del Castillo, J M Caldero, J Roca, G Romero, J Del Rio, B Lasheras J.Med. Chem. 36 2745 (1993)Physiologically active compounds interacting with serotonin (5-hydroxytryptamine) receptors 242. J R Fozard, A T Mobarok Ali Eur. J. Pharmacol. 49 109 (1978) 243. V Derkach, A Surprenant, R A North Nature (London) 339 706 (1989) 244. W L Smith, R S Alphin, C B Jackson, L F Sancilio J. Pharm. Pharmacol. 41 101 (1989) 245. W W Smith, L F Sancilio, J B Owera-Atepo, R J Naylor, L Lambert J. Pharm. Pharmacol. 40 301 (1988) 246. J R Fozard Naunyn Schmiedebergs Arch. Pharmacol. 326 36 (1984) 247. J Bermudez, S Dabbs, F D King J. Med. Chem. 33 1932 (1990) 248. J Bermudez, S Dabbs, K A Joiner, F D King J. Med. Chem. 33 1929 (1990) 249. M Turconi,M Nicola, M G Quintero, L Maiocchi, R Micheletti, E Giraldo, A Donetti J.Med. Chem. 33 2101 (1990) 250. J Bermudez, C S Fake, G F Joiner, K A Joiner, F D King, W D Miner, G J Sanger J. Med. Chem. 33 1924 (1990) 251. R D Clark, J M Muchowski, K K Weinhardt, M P Dillon, C-H Lee, K R Bley, D W Bonhaus, E H F Wong, R M Eglen Bioorg. Med. Chem. Lett. 5 1853 (1995) 252. G J Kilpatrick, K T Bunce,M B Tyers Med. Res. Rev. 10 441 (1990) 253. A Butler, J M Hill, S J Ireland, C C Jordan,M B Tyers Br. J. Pharmacol. 94 397 (1988) 254. M A Liberman, S Howe, M Lane Am. J. Surg. 179 60 (2000) 255. F D King (Ed.) Medicinal Chemistry: Principles and Practice (Cambridge: Royal Society of Chemistry, 1994) 256. F Heidempergher, A Pillan, V Pinciroli, F Vaghi, C Arrigoni, G Bolis, C Caccia, L Dho, R McArthur, M Varasi J. Med. Chem. 40 3369 (1997) 257. C J Swain, R Baker, C Kneen, J Moseley, J Saunders, E M Seward, G Stevenson, M Beer, J Stanton, K Watling J. Med. Chem. 34 140 (1991) 258. T Fukuda,M Setoguchi, K Inaba, H Shoji, T Tahara Eur. J. Pharmacol. 196 299 (1991) 259. K Haga, K Inaba, H Shoji, Y Morimoto, T Fukuda,M Setoguchi Jpn. J. Pharmacol. 63 377 (1993) 260. D W Robertson,W B Lacefield,W Bloomquist, W Pfeifer, R L Simon, M L Cohen J. Med. Chem. 35 310 (1992) 261. T Rosen, A A Nagel, J P Rizzi, J L Ives, J B Daffeh, A H Ganong, K Guarino, J Heym, S McLean, J T Nowakowski, A W Schmidt, T F Seeger,C J Sick, L A Vincent J. Med. Chem. 33 2715 (1990) 262. A W Schmidt, S J Peroutka Mol. Pharmacol. 36 505 (1989) 263. C J Swain, R Baker, C Kneen, R Herbert, J Moseley, J Saunders, E M Seward, G I Stevenson, M Beer, J Stanton, K Watling, R G Ball J. Med. Chem. 35 1019 (1992) 264. R D Clark, A B Miller, J Berger, D B Repke, K K Weinhardt, B A Kowalczyk, R M Eglen, D W Bonhaus, C H Lee, A D Michel J. Med. Chem. 36 2645 (1993) 265. Y Rival, R Hoffmann, B Didier, V Rybaltchenko, J -J Bourguignon, C G Wermuth J. Med. Chem. 41 311 (1998) 266. Y Sato,M Yamada, S Yoshida, T Soneda,M Ishikawa, T Nizato, K Suzuki, F Konno J. Med. Chem. 41 3015 (1998) 267. M Yamada, Y Sato, K Kobayashi, F Konno, T Soneda, T Watanabe Chem. Pharm. Bull. 46 445 (1998) 268. C Gerald, N Adham, H T Kao,M A Olsen, T M Laz, L E Schechter, J A Bard, P J Vaysse, P R Hartig, T A Branchek EMBO J. 14 2806 (1995) 269. O Blondel,M Gastineau, Y Dahmoune, M Langlois, R Fischmeister J. Neurochem. 70 2252 (1998) 270. K H Buchheit, R Gamse, H J Pfannkuche Naunyn Schmiedebergs Arch. Pharmacol. 345 387 (1992) 271. D L Flynn, D L Zabrowski, D P Becker, R Nosal, C I Villamil, G W Gullikson, C Moummi, D-C Yang J. Med. Chem. 35 1486 274. M Langlois,D Yang, B Bremont, S Shen Bioorg. Med. Chem. Lett. (1992) 272. T P Blackburn, G S Baxter, G A Kennet, F D King, D C Piper, G I Sanger, D R Thomas, N Upton,M D Wood Psychopharmacology 110 357 (1993) 273. F D King, M S Hadley, K T Joiner, R T Martin, G J Sanger, D M Smith, G E Smith, P Smith, D H Turner, E A Watts J. Med. Chem. 36 683 (1993) 5 795 (1995) 275. R M Eglen, D W Bonhaus, L G Johnson, E Leung, R D Clark Br. J. Pharmacol. 115 1387 (1995) 355 276. S Elz, A Keller Arch. Pharm. (Weinheim) 328 585 (1995) 277. D Yang, J-L Soulier, S Sicsic,M Mathe-Allainmat, B Bremont, T Croci, R Cardamone, G Aureggi,M Langlois J. Med. Chem. 40 608 (1997) 278. S S Hegde, D W Bonhaus, L G Johnson, E Leung, R D Clark, R M Eglen Br. J. Pharmacol. 115 1087 (1995) 279. R M Eglen, E H Wong, A Dumuis, J Bockaert Trends Pharmacol. Sci. 16 391 (1995) 280. R M Eglen, K Bley, D W Bonhaus, R D Clark, S S Hegde, L G Johnson, E Leung, E H F Wong Br. J. Pharmacol. 110 119 282. A J Kaumann, L M Gaster, F D King, A M Brown Naunyn (1993) 281. R D Clark, A Jahangir, L A Flippin, J A Langston, E Leung, D W Bonhaus, E H Wong, L G Jonhnson, R M Eglen Bioorg. Med. Chem. Lett. 5 2119 (1995) Schmiedebergs Arch. Pharmacol. 349 546 (1994) 283. L M Gaster, A J Jennings, G F Joiner, F D King, K R Mulholland, S K Rahman, S Starr, P A Wyman, K A Wardle, E S Ellis, K J Sanger J. Med. Chem. 36 4121 (1993) 284. J D Gale, C J Grossman, J W Whitehead, A W Oxford, K T Bunce, P P Humphrey Br. J. Pharmacol. 111 332 (1994) 285. L M Gaster, G F Joiner, F D King, P A Wyman, J M Sutton, S Bingham, E S Ellis, G J Sanger, K A Wardle J. Med. Chem. 38 4760 (1995) 286. I Tapia, L Alonso-Cires, P L Lopez-Tudanca, R Mosquera, L Labeaga, A Innerarity, A Orjales J. Med. Chem. 42 2870 (1999) 287. K-H Buchheit, R Gamse, R Giger, D Hoyer, F Klein, E Kllppner, H-J Pfannkuche, H Mattes J. Med. Chem. 38 2326 (1995) 288. K-H Buchheit, R Gamse, R Giger, D Hoyer, F Klein, E KloÈ ppner, H-J Pfannkuche, H Mattes J. Med. Chem. 38 2331 (1995) 289. A Monge,M del Carmen Pena, J A Palop, J M Caldero, J Roca, E Garcia, G Romero, J del Rio, B Lasheras J. Med. Chem. 37 1320 (1994) 290. J M Schaus, D C Thompson, W E Bloomquist, A D Susemichel, D O Calligaro, M L Cohen J. Med. Chem. 41 1943 (1998) 291. J-L Soulier, D Yang, B Bremont, T Croci, U Guzzi,M Langlois J. Med. Chem. 40 1755 (1997) 292. M-L Cohen,W Bloomquist, J M Schaus, D C Thompson, A D Susemichel, D O Calligaro, I Cohen J. Pharmacol. Exp. Ther. 277 97 (1996) 293. M Wood,M Chaubey, P Atkinson, D R Thomas Eur. J. Pharmacol. 396 1 (2000) 294. B L Roth, S C Craigo,M S Choudhary,A Uluer, F J Monsma Jr, Y Shen, H Y Meltzer, D R Sibley J. Pharmacol. Exp. Ther. 268 1403 (1994) 295. I T Forbes, S Dabbs, D M Duckworth, A J Jennings, F D King, P J Lovell, A M Brown, L Collin, J J Hagan, D N Middlemiss, G J Riley, D R Thomas, N Upton J. Med. Chem. 41 655 (1998) 296. R A Glennon, M Lee, J B Rangisetty,M Dukat, B L Roth, J E Savage,A McBride, L Rauser, S Hufeisen,D K H Lee J. Med. Chem. 43 1011 (2000) a�Moscow Univ. Bull. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC