摘要:

Russian Chemical Reviews 69 (7) 551 ± 571 (2000) Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds MG Zuev Contents I. Introduction II. Phase relations inM2O3±V2O5±Ta2O5 systems (M=REE) III. Phase relations in the systemsM2O3±V2O5±Nb2O5 (M=REE) and Y2O3 (La2O3, Sc2O3)±Nb2O5±Ta2O5 IV. Phase relations in systems containing B, Al and Ga V. Phase formation regularities in ternary systems VI. Vibrational spectra of ternary oxides Abstract. considered are region subsolidus the in relations Phase Phase relations in the subsolidus region are considered for including systems ternary 54 for 54 ternary systems including M2O37V2O57R2O5, Y c 2O3(La2O3, S , Sc2 O3)7Nb2O57Ta2O5, M2O37B2O37R2O5, M rare- a s i M l 2O37Ga2O37R2O5, A , Al2O37V2O57R2O5 (where (where M is a rare- earth determining criteria The Nb).Ta, = R and element earth element and R=Ta, Nb). The criteria determining the the formation include to found were compounds oxide ternary of formation of ternary oxide compounds were found to include the the cation ions the of radii the between ratios the and sizes cation sizes and the ratios between the radii of the ions constitut- constitut- ing The compounds. the of part anionic the ing the anionic part of the compounds. The crystal-chemical crystal-chemical characteristics LaRB for presented are characteristics are presented for MR MR2VO9, LaRB2O7, La The La Nd), Pr, La, = n 4VBO10, L , Ln7VBO17 (Ln (Ln=La, Pr, Nd), La3Ga5.5R0.5O14.The vibrational spectra of polycrystals of ternary compounds and vibrational spectra of polycrystals of ternary compounds and solid detail. in described are them on based solutions solid solutions based on them are described in detail. The The bibliography references 154 includes bibliography includes 154 references. I. Introduction One of the primary tasks of chemistry is the development and creation of new materials. Simple and complex oxides have found very wide use in many branches of industry. Synthesis of new materials based on complex oxides and the improvement of existing technologies for the production of these materials are based both on the study of the phase diagrams of oxide systems and on fundamental investigations into the `composition ± struc- ture ± properties' relationships.Owing to the vast diversity of their properties, complex oxides of V, Ta, Nb and elements of the Group III of the Periodic Table are of special interest. For example, oxide crystals of Group III and VB elements are used as photo-, cathodo- and X-ray active luminophores. A topical task is the search for medical and lumi- nescent materials that possess new spectroscopic parameters which might be expected among compounds of transition elements. For instance, efficient tantalum-based X-ray active lumino- phores have acquired ever growing importance in recent years.1 ±3 They possess higher specific density [(7.5 ± 9.7)6103 kg m73] than other X-ray active luminophores used in common practice and a number of other advantages allowing the production of intensifying X-ray screens with higher resolution and speed.This MG Zuev Institute of Solid State Chemistry, Urals Branch of the Russian Academy of Sciences, ul. Pervomaiskaya 91, 620219 Ekaterinburg, Russian Federation. Fax (7-343) 274 44 95. Tel (7-343) 249 34 92. E-mail: zuev@ihim.uran.ru Received 28 December 1999 Uspekhi Khimii 69 (7) 603 ± 623 (2000); translated by S S Veselyi #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n07ABEH000385 551 552 553 554 556 558 makes it possible to decrease significantly the radiation dose that affects the patients during X-ray examinations, a feature espe- cially important for X-ray diagnostics of children.Rare-earth element (REE) tantalates are crystals which effi- ciently absorb X-ray radiation with a quantum energy from 10 to 150 keV and are widely used in medicine. Complex tantalum- containing REE oxides are used as novel X-ray contrast agents (RCA) that are precipitated on body tissues and absorb X-ray radiation. This enhances contrast, which allows one to perform more accurate diagnostics at lower irradiation doses absorbed by the patient.4 ±8 The physicochemical properties of these com- pounds have not been studied sufficiently yet. Studies of the phase diagrams of complex oxides and their solid solutions used as RCA will aid in filling this gap. 4 Compounds of the general formula MR2VO9 exist in ternary systems M2O37V2O57R2O5 (M=REE; R=Nb, Ta).9 Based on these compounds, a new type of complex-oxide phosphor crystal has been developed. Synthesis of these compounds is accompanied by spontaneous formation of optical centres VO47 owing to which the crystals of MR2VO9 possess narrow-band luminescence, predominantly in the red region of the spectrum.These phosphor crystals have an unusual temperature dependence of irradiation: the intensity of red luminescence of the centres increases on heating, whereas that of the support is quenched.10 This effect contradicts the law of thermal luminescence quenching and has not been rationalised yet. A physicochemical substantiation of the properties of com- plex oxides listed above requires primarily a detailed study of the regularities of their formation by considering phase relations in ternary systems.The types of phase diagram often vary for REE- containing systems:11 purely eutectic systems exist in equilibrium with chemical compounds. In addition, one or several compounds can exist only for some of the elements rather than for the entire REE series. As a result, isostoichiometric series are formed which are grouped in complete or incomplete morphotropic and iso- structural series. The change in the type of phase diagrams of ternary systems depending on the nature of the REE is shown below. The purpose of this review is to generalise data on the phase relations in ternary { oxide systems of Group III and VB elements { The phase composition of restriction-type binary systems, viz., M2O37R2O5 (M=B, Al, La, Y, Sc, Ln; R=Ta, Nb, V), M2O37B2O3 (M=La, Y, Sc, Ln,), La2O37Ga2O3, V2O57R2O5 (R=Ta, Nb), Ta2O57Nb2O5, as well as the chemistry of the compounds formed from them have been described in detail elsewhere.12 ± 82552 in a subsolidus region, to reveal the regularities of the formation of ternary oxide systems and to analyse their vibration spectra.The vibration spectra not only provide information on the structure of the crystal lattices of ternary compounds but also make it possible to predict the luminescent properties of phosphor crystals. Based on the analysis of vibration spectra with the use of the concept of elementary lattice excitations (phonons), one can make conclu- sions on the filling of the energy gaps between the excited states of optical centres and describe the processes responsible for the excitation energy transfer in phosphor crystals.II. Phase relations inM2O37V2O57Ta2O5 systems (M=REE) Figure 1 { shows phase relations in tantalum-containing M2O37V2O57Ta2O5 systems in a subsolidus region up to 1200 8C.83 ± 98 For each REE, ternary oxides with the composition MTa2VO9 (henceforth referred to as vanadatotantalates) were found in these systems. When a mixture of the starting compounds is heated, thermodynamic equilibrium is established faster the higher the temperature. Under isothermal sintering conditions at temperatures >1100 8C, the M2O37V2O5 systems are equili- brated in 6 ± 8 h, while the M2O37Ta2O5 systems require 30 ± 70 h (for comparison: the equilibrium in the M2O37Ta2O5 system at 900 8C is established in >100 h).It takes more than 10 h to equilibrate the V2O57Ta2O5 system.18, 44 Equilibration of the ternary systems M2O37V2O57Ta2O5 takes approximately the same time as that of the M2O37Ta2O5 systems.9 Vanadato- tantalates decompose above 1200 8C to give the corresponding REE metatantalates and an unknown phase.86 In a subsolidus region, all compounds of the ternary system are stable. The difference in triangulation depends only on the existence of certain binary compounds in the M2O37V2O5 (up to 50 mol.% M2O3) andM2O37Ta2O5 systems (50 mol.%± 100 mol.%Ta2O5). According to the data of laser radiation second harmonics generation (SHG) measurements, ternary oxides of scandium and cerium subgroup elements can be assigned centrosymmetrical crystals.9 X-Ray diffraction data 9 suggest that the structures of MTa2VO9 (M=La, Y, Pr, Nd, Sm, Eu) are close to those of the corresponding tantalates MTa7O19 which possess hexagonal symmetry.For this reason, solid solutions can be formed in the MTa2VO97MTa7O19 system. Table 1 lists the unit cell parame- { Hereinafter, isothermal sections are shown in the phase diagrams. V2O5 RVO5 MVO4 6 MR2VO9 45 3 VR9O25 12 R2O5 M2O3 8 9 10 7 MRO4 Figure 1. Phase relations in the M2O3±V2O5±R2O5 systems (M=REE, R=Ta, Nb). (1) M12V2O23 (M=Pr, Nd); (2) M5VO10 (M=Y, Sm, Eu, Er); (3) M1.62V0.38O3.38 (M=La, Nd, Eu, Gd, Tb), (4) M8V2O17 (M=Y, La, Ce ± Eu, Dy ± Yb); (5)M3VO7 (M=La, Nd, Eu); (6) La1.42V0.58O3.58; (7) M3RO7 (R=Ta, M=Ce; R=Nb, M=Sc, Ce); (8) MR3 O9 (R=Ta, M=Sc; R=Nb, M=La ± Sm); (9) MR5 O14 (R=Ta, M=La, Ce; R=Nb, M=La ± Nd, Er); (10) MR7 O19 (R=Ta, M=Sc, Yb, Lu; R=Nb,M=La, Ce).MG Zuev Table 1. Unit cell parameters of hexagonal crystals of MTa2VO9 . V /nm 3 c a /nm a a /nm M 1.9573 0.9932 1.9831 1.9864 1.9723 1.9812 1.9831 0.6231 0.6152 0.6124 0.6143 0.6192 0.6121 0.6189 La Ce Pr Nd Sm Eu Yb 0.658 0.653 0.643 0.648 0.654 0.643 0.655 a The error of determination is 0.0005 nm. b The a-YTa2VO9 phase. ters for MTa2VO9 crystals. It should be noted that YTa2VO9 has two crystal modifications conventionally designated as a and b.The a modification is structurally similar to LaTa2VO9, while the b modification is similar to ScTa2VO9. The structures of vanadatotantalates of scandium and yttrium subgroup elements differ from those of compounds of cerium subgroup elements. The structure of MTa2VO9 crystals has not been determined yet. One can see from Table 1 that the parameters a and c and the unit cell volumes of vanadatotantalates do not depend linearly on the lanthanide ionic radii, as opposed to simple tantalates 99 and mixed-cation compounds.100 It is believed 99 that the decrease in the unit cell parameters of tantalates with the decrease in the REE ionic radii can be explained by a decrease in the degree of covalence of the M7O bond, which changes essentially along the series from scandium to yttrium, lanthanum and from cerium to lutetium.Evidently, other factors are important in the case of vanadatotantalates. It is well known that REE compounds display periodicity of structural and structure-dependent properties. These properties change non-monotonically for several reasons; the major of these include the energetic and spatial deepening of orbitals, the alter- ation of the type of exchange interactions between unpaired electrons, spin-orbital stabilisation, the nepheloauxetic effect and the stabilising effect of the crystal field.11 Table 1 shows that praseodymium and europium vanadato- tantalates have the smallest parameters a and unit cell volumes. Hence, it can be assumed that an additional decrease in the energy of the 4 f orbitals occurs in PrTa2VO9 and EuTa2VO9 due to spin- orbital interaction.This affects the contribution of the f orbitals to the formation of chemical bonds and probably the shape of the coordination polyhedra of these compounds. It is noteworthy that there is a small `gadolinium' inflexion of a values for LnTa7O19 crystals (Ln=Ce7Lu).101 No sharp changes in the X-ray diffraction parameters are observed in the series LnTa2VO9 (Ln=Gd7Lu). These crystals are isostructural, which is corroborated by the data from vibra- tional spectra (see Section VI). Some crystal-chemical features of vanadatotantalates were established using NMR spectroscopy.83 In particular, the type of coordination polyhedra formed by oxygen around the vanadium ions was determined.The spectra were assigned using NMR data for known compounds.102 ± 106 The spectra for a sample of YTa2VO9 were recorded at frequencies of 105.2 and 78.86 MHz without and with magic-angle spinning (MAS). The spectrum consisted of a single line. The magic-angle spinning enabled fairly accurate determination of the isotropic shift. The MAS spectra recorded at two operating frequencies of the spectrometer display quadrupole splitting of the central transition. Figure 2 (curve 1) shows a 51V NMR spectrum for YTa2VO9 measured at 78.86 MHz without consideration of the quadrupole correction diso=76652. The small anisotropy of the chemical shift and the value of diso indicate that the vanadium atom is located in an almost regular tetrahedral environment.According to data reported elsewhere,107 the insignificant distortion of the tetrahe-Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds 1234 7550 7600 7650 7700 ppm Figure 2. 51V NMR spectra of YTa2VO9 (1), YNb2VO9 (2), ScNb2VO9 (3), La4VBO10 (4). dron can suggest that one of its oxygen atoms simultaneously belongs to an adjacent VO4 tetrahedron. The NMR spectrum of the vanadatotantalate GdTa2VO9 could not be recorded, prob- ably due to quadrupole effects resulting in line broadening. Thus, it is possible to distinguish a full isostoichiometric series in the LnTa2VO9 class of compounds (where Ln indicates lantha- nides) existing at normal temperature and pressure.However, this is not a full isostructural series: for Ln=Gd, there is a morpho- tropic transition from one structural type (LnTa7O19) for Ln=Y, La ±Eu to another one for Ln=Gd ± Lu. III. Phase relations in the systemsM2O3±V2O5 ± Nb2O5 (M=REE) and Y2O3 (La2O3, Sc2O3) ± Nb2O5±Ta2O5 Figure 1 shows phase relations in the M2O37V2O57Nb2O5 systems in the subsolidus region up to 1100 8C (for M=Sc, up to 1000 8C).83, 89, 94, 108 ± 114 Compounds with the composition MNb2VO9 (henceforth referred to as vanadatoniobates) were found in ternary systems for every M. Under isothermal annealing conditions at temperatures >900 8C, the time required to establish equilibria in the M2O3 ± Nb2O5 and V2O5±Nb2O5 systems is the same as in the corre- sponding tantalum-containing systems.Ternary M2O3±V2O5 ± Nb2O5 systems come to equilibrium in 30 ± 70 h.9 Vanadatonio- bates decompose above 1150 8C to give ortho- (Sc, Y) and metaniobates (La, Ln) and an unknown phase.86 In the subsolidus region, all compounds of the ternary systems are stable. The structure of vanadatoniobates has not been determined yet. According to SHG measurements, Sc, Y and La vanadato- niobates are centrosymmetrical crystals. It follows from compar- ison of the X-ray diffraction characteristics of MTa2VO9 and MNb2VO9 (M=Sc, Y, La) that ScTa2O9 and ScNb2O9 are isostructural; the structures of vanadatoniobates YNb2VO9 and LaNb2VO9 differ from each other and from those of vanadato- tantalates YTa2VO9 and LaTa2VO9, respectively.The structure of vanadatoniobates formed by cerium sub- group lanthanides differs from that of the corresponding vanada- totantalates. Within the series LnNb2VO9 (Ln=Ce ± Eu), the structures of PrNb2VO9 and NdNb2VO9 are similar and differ from the structures of CeNb2VO9, SmNb2VO9 and EuNb2VO9, which have similar lattice symmetry. Probably, as in vanadato- tantalates of cerium subgroup elements, the additional decrease in the energy of the f orbitals participating in the chemical bonds in vanadatoniobates of the elements of this subgroup affects the shape of the coordination polyhedra of the crystals. The structure of vanadatoniobates of yttrium subgroup ele- ments has some specific features. In the X-ray diffraction patterns of compounds in the series Gd± Lu, the interplanar distance (d) between the most intense reflexions decreases from 0.359 [relative 553 intensity (I/I0) 69%] to 0.354 nm (relative intensity 68%) in the case of GdNb2VO9 and to 0.350 (relative intensity 100%) in the case of EuNb2VO9.For TmNb2VO9, YbNb2VO9 and LuNb2VO9, only one reflexion with 100% intensity and d=0.351 nm is observed. Thus, the vanadatoniobates LnNb2VO9, where Ln=Gd, Tb, Dy, Ho, Er, are isostructural, and the structures differ from those of compounds with Ln=Tm, Yb, Lu. This behaviour is probably determined by the Ln3+ ionic radius and the degree of covalency of the Ln7O bond. It should be noted that the structures of the cerium- and yttrium-subgroup compounds are different.The type of coordination polyhedron of the vanadatoniobate lattice was determined using NMR spectro- scopy (see Fig. 2, curves 2,3).83 In polycrystals of YNb2VO9, the V5+ ions occupy two non-equivalent crystallographic positions. The isotropic shifts diso1 and diso2 are 76652 and 76042, respectively, without the quadrupole correction. The small aniso- tropy and the magnitudes of the isotropic shifts suggest that the vanadium atom is located in an almost regular tetrahedral environment and can have one common atom with the adjacent VO4 tetrahedron. It is of note that compounds YNb2VO9 and YTa2VO9 are not isostructural but have almost similar tetrahedral environments with diso=7665.Two lines are observed in the 51V NMR spectrum of ScNb2VO9. The use of the MAS technique made it possible to determine the isotropic shifts for these lines:75815 (line I) and 76032 (line II). The presence of two lines in the spectrum suggests that the lattice contains two non-equivalent vanadium ions. The shape of line I implies an axial anisotropy of the chemical shift tensor equal to d\=7350, whereas line II is almost symmetrical. The magnitude of anisotropy, its type and the magnitude of the isotropic shift make it possible to assume that line I belongs to the vanadium ions in a distorted octahedral environment, while line II belongs to the vanadium ions in a slightly distorted tetrahedral environment. The tetrahedral envi- ronments in ScNb2VO9 and in YNb2VO9 (diso=7604) are almost the same.In the vanadatoniobate ScNb2VO9, the VO4 tetrahedra are linked to each other through one common oxygen atom. The reason for distortion of the VO6 octahedra is that each VO6 octahedron has two common oxygen atoms with the adjacent octahedron.107 Based on the X-ray diffraction characteristics obtained for LnNb2VO9 (Ln=REE), it can be stated that this isostoichiomet- ric series of compounds is not a full isostructural series. Five different structure types exist in the series of LnNb2VO4 vanada- toniobates: the first one for Ln=La, the second one for Ln=Pr and Nd, the third one for Ln=Ce, Sm and Eu, the fourth one for Ln=Gd, Tb, Dy, Ho, Er and the fifth one for Ln=Tm, Yb and Lu.Thus, a morphotropy phenomenon is observed. The Y2O37Nb2O57Ta2O5 system.115 Figure 3 presents phase relations in the Y2O3±Nb2O5±Ta2O5 system in a subsoli- Nb2O51 YNbO4 2 Y3NbO7 Ta2O5 Y2O3 YTa7O19 Y3TaO7 YTaO4 YTa3O9 Figure 3. Phase relations in the Y2O3±Nb2O5±Ta2O5 system. (1) a-Nb2(17x)Ta2xO5; (2) b-Ta2(17x)Nb2xO5.MG Zuev 554 Nb2O5 a-Nb2(17x)Ta2xO5 ScNbO4 b-Ta2(17x)Nb2xO5 Sc5.5Nb1.5O12 Sc6NbO11.5 Ta2O5 Sc2O3 Sc6TaO11.5 ScTaO4 Sc3TaO7 Figure 5. Phase relations in the Sc2O3±Nb2O5±Ta2O5 system. IV. Phase relations in systems containing B, Al and Ga The M2O37B2O37Ta2O5 systems.83, 117 ± 119 Figure 6 a,b presents phase relations in the Sc2O3±B2O3±Ta2O5 and Y2O3 ± B2O3±Ta2O5 systems in a subsolidus region up to 1100 8C.No ternary compounds were found in these systems. Figure 7 demonstrates phase relations in the La2O3±B2O3 ± Ta2O5 system in a subsolidus region up to 1000 8C. A ternary compound with the composition LaTaB2O7, which exists in dus region up to 1350 8C. Solid solutions in this system were studied by optical-immersion analysis. A continuous series of solid solutions with the compositions Y3NbxTa17xO7 with a probable structure of the pyrochlore type was found in the quasibinary section Y3TaO7±Y3NbO7, and solid solutions of two types were found in the quasibinary section YTaO4 ± YNbO4. The first type has the composition YNbxTa17xO4 and is based on the fergusonite M0-modification in the concentration range 04x40.15, while the second type has the same composi- tion and is based on the fergusonite M-modification in the concentration range 0.34x41.A mixture of solid solutions of the first and second types is formed in the intermediate region, viz., 0.15<x<0.3. Above 1400 8C, a continuous series of solid solutions with M-fergusonite structure exists in this section. The La2O37Nb2O57Ta2O5 system.116 In order to determine the location of the quasibinary sections and to perform the triangulation of the La2O3±Nb2O5±Ta2O5 system, physicochem- ical analysis was used to study the phase composition of 30 points located in various hypothetical sections and regions of the system. Figure 4 presents phase relations in the La2O3±Nb2O5±Ta2O5 system in a subsolidus region up to 1250 8C.Various solid substitution solutions were found in the region containing more than 50 mol.%La2O3. The hatched regions 1 and 2 contain solid solutions based on LaNbO4 and La3NbO7, respectively. In addition, a limited series of solid solutions of the general formula LaTa17xNbxO4 (04x40.15) based on LaTaO4 was found in the LaNbO4 ± LaTaO4 section. A continuous series of solid solutions of the general formula La3Nb17yTayO7 (04y41) exists in the La3NbO7±La3TaO7 section. b B2O3 a B2O3 Nb2O5 a-Nb2(17x)Ta2xO5 Y(BO2)3 LaNb7O19 LaNb5O14 LaNb3O9 TaBO4 TaBO4 ScBO3 YBO3 LaNbO4 b-Ta2(17x)Nb2xO5 1 La3NbO7 Ta2O5 YTaO4 Ta2O5 Y2O3 ScTaO4 Sc2O3 2 Sc3TaO7 Sc6TaO11.5 YTa3O9 YTa7O19 Y3TaO7 Ta2O5 La2O3 La3TaO7 LaTaO4 LaTa7O19 LaTa3O9 Figure 6.Phase relations in the Sc2O3±B2O3±Ta2O5 (a) and Y2O3 ± B2O3±Ta2O5 (b) systems. LaTa5O14 B2O3 Figure 4. Phase relations in the La2O3±Nb2O5±Ta2O5 system. The hatched zones indicate the areas of existence of solid solutions based on LaNbO4 (1) and La3NbO7 (2). La(BO2)3 LaTaB2O7 TaBO4 LaBO3 La3BO6 The Sc2O37Nb2O57Ta2O5 system.} The phase composition of the system was studied up to 1400 8C; no ternary compounds were found (Fig. 5). A continuous series of solid substitution solutions of the general formula ScTa17xNbxO4 with the wolf- ramite structure exists in the quasibinary section ScNbO4 ± ScTaO4. Ta2O5 La2O3 La3TaO7 LaTaO4 LaTa3O9 LaTa7O19 LaTa5O14 } The system was studied by E V Arkhipova (Institute of Solid State Chemistry of the Urals Branch of the Russian Academy of Sciences).Figure 7. Phase relations in the La2O3±B2O3±Ta2O5 system.Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds equilibrium with LaBO3, La(BO2)3, B2O3, LaTa3O9 and LaTaO4, was found in the quasibinary section La(BO2)3 ± LaTa3O9. Lan- thanum boratotantalate decomposes at*1200 8C in accordance with the peritectic reaction LaTaB2O7=LaTaO4+B2O3. In the M2O3±B2O3 systems, thermodynamic equilibrium at temperatures above 1000 8C is established in several tens of hours.66 The time required to reach equilibrium in ternary systems is even longer, viz.,*200 h.9 All phases of the systems described above are stable in a subsolidus region. According to SHG laser irradiation measurements, lantha- num boratotantalate has centrosymmetrical crystals.X-Ray dif- fraction studies suggest that it has orthorhombic symmetry. The unit cell parameters a, b and c for LaTaB2O7 are 1.28692, 1.17562 and 1.13692 nm, respectively; the unit cell volume is 1.72004 nm3; the number of formula units in the lattice (z) is 14; the picnometric density is 6.166103 kg m73. The coordination of boron in these crystals was studied by NMRspectroscopy.120 Figure 8 shows the 11B NMRspectrum of the boratotantalate LaTaB2O7 (relative to aqueous H3BO3, magic-angle spinning). The spectrum is a superposition of two doublet components (d77.96 and 721.59) and a narrow central peak (d716.53).The assignment was made using the known literature data (see, e.g., Ref. 121). It was found that the doublet belongs to the boron nuclei located in a trigonal oxygen environ- ment (the quadrupole coupling constant is 2.5 MHz), while the narrow line belongs to the boron nuclei in a tetrahedral environ- ment. It should be noted that the central component in the spectra is shifted by 7 ppm, and the doublet is shifted by 14 ppm with respect to diamagnetic compounds. This is probably caused by paramagnetic admixtures in the samples. The ternaryM2O37B2O3±Ta2O5 systems with the following lanthanides were studied: cerium, praseodymium, neodymium, samarium and europium.122, 123 No ternary compounds were found in these systems.Figure 9 a,b presents phase relations in 0 50 750 ppm Figure 8. 11B NMR spectrum of boratotantalate LaTaB2O7. b B2O3 a B2O3 Ce(BO2)3 Ln(BO2)3 BTaO4 CeBO3 BTaO4 LnBO3 Ln2O3 Ta2O5 Ce2O3 Ta2O5 LnTaO4 CeTaO4 CeTa3O9 CeTa7O19 Ln3TaO7 LnTa7O19 77.9612 716.5282 721.5945 CeTa5O14 LnTa3O9 Figure 9. Phase relations in the Ce2O3±B2O3±Ta2O5 (a) and Ln2O3 ± B2O3±Ta2O5 (Ln=Pr, Nd, Sm, Eu) (b) systems. 555 these systems in a subsolidus region up to 1000 8C. It is seen that borates are in equilibrium with lanthanide tantalates. The M2O37B2O37Nb2O5 systems.83, 118, 119 These systems, where M=REE, are similar to the corresponding tantalum- containing ternary systems in many respects. The equilibrium in these systems is also established in *200 h.9 No ternary com- pounds were found in the Sc2O3±B2O3±Nb2O5 and Y2O3 ± B2O3±Nb2O5 systems, the phase relations in which in a subsolidus region up to 1000 8C are presented in Fig.10 a,b. b B2O3 a B2O3 Y(BO2)3 ScBO3 YBO3 BNb3O9 BNb3O9 Nb2O5 Y2O3 Nb2O5 Sc2O3 ScNbO4 YNbO4 Sc5.5Nb1.5O12 Y3NbO7 Sc6NbO11.5 (a) and Figure 10. Phase relations in the Sc2O3±B2O3±Nb2O5 Y2O3±B2O3±Nb2O5 (b) systems. More complex phase relations are observed in the La2O3 ± B2O3±Nb2O5 system.117, 119 The ternary compound LaNbB2O7 is formed in this system; this compound exists in equilibrium with LaNbO4, LaNb3O9, LaBO3, La(BO2)3 and B2O3. The boratonio- bate LaNbB2O7 is isostructural to the boratotantalate LaTaB2O7. It decomposes peritectically above 1200 8C to give LaNbO4 and B2O3.Figure 11 shows phase relations in this system in a subsolidus region up to 1000 8C. The unit cell parameters a, b and c of LaNbB2O7 are 1.2771(5), 1.1722(5) and 1.1223(5) nm; the unit cell volume is 1.679231 nm3. B2O3 La(BO2)3 LaNbB2O7 LaBO3 BNb3O9 La3BO6 Nb2O5 La2O3 La3NbO7 LaNbO4 LaNb3O9 LaNb7O19 LaNb5O14 Figure 11. Phase relations in the La2O3±B2O3±Nb2O5 system. The M2O37B2O37V2O5 systems. The equilibrium in these systems is established in more than 800 h,9 which is much longer than in theM2O3±B2O3±R2O5 systems (R=Ta, Nb). Figure 12 demonstrates phase relations in the Y2O3±B2O3 ± V2O5 system in a subsolidus region up to 1000 8C. No ternary compounds were found in this system.124 Figure 13 presents phase relations in the La2O3±B2O3±V2O5 system in a subsolidus region up to 900 8C.125 It was found that lanthanum orthovanadate is in equilibrium with all compounds of the binary systems La2O3±B2O3 and B2O3±V2O5 except La3BO6.Aternary compound with the composition La4VBO10 is formed in this system in the region with high La2O3 content; this compound556 B2O3 Y(BO2)3 YBO3 BVO4 V2O5 Y2O3 YVO4 Y5VO10Y8V2O17 Figure 12. Phase relations in the Y2O3±B2O3±V2O5 system. B2O3 La(BO2)3 LaBO3 BVO4 La3BO6 La2O3 V2O5 LaVO4 La4VBO10 Figure 13. Phase relations in the La2O3±B2O3±V2O5 system. is in equilibrium with La2O3, La3BO6, LaBO3 and LaVO4. The melting point of La4VBO10 is unknown. According to SHG measurements, this compound has centrosymmetrical crystals.Figure 2 (curve 4) shows the 51V NMR spectrum of La4VBO10 polycrystals. The spectrum is a single line with diso=7674. The magnitude of the quadrupole correction indi- cates that vanadium is located in a nearly regular tetrahedral environment.126 As in vanadatoniobates, the VO4 tetrahedra can have one common oxygen atom with the adjacent tetrahedra. The signal of the boron nuclei is very weak, probably due to quadru- pole effects. Hence, LaTaB2O7, LaNbB2O7 and La4VBO10 are merely the first members of the corresponding isostoichiometric series under ordinary conditions. An attempt to obtain compounds of scan- dium, yttrium and lanthanides with compositions indicated above failed.83 Gokhman et al.127 synthesised ternary oxides with the com- position Ln7V2BO17 (Ln=La, Pr, Nd).The samples were obtained in air by multi-step sintering of the starting oxides Ln2O3, V2O5 and H3BO3. The temperature of the final sintering step was 1240 8C; the overall sintering time was 30 h. The melting points of the compounds are unknown. Ternary oxides for Ln=Sm, Eu, Gd and Tb could not be obtained. Table 2 lists the unit cell parameters for the crystals of Ln7V2BO17, which possess monoclinic symmetry. Table 2. Unit cell parameters of Ln7V2BO17 crystals. c/nm b/nm a /nm V /nm3 Ln b /deg 1.501(2) 1.425(7) 1.392(2) 99.6(1) 99.7(3) 99.4(1) 1.227(1) 1.205(3) 1.205(1) 1.786(1) 1.747(7) 1.723(2) 0.695(1) 0.687(1) 0.680(1) La Pr Nd MG Zuev b V2O5 a V2O5 NbVO5 AlVO4 TaVO5 AlVO4 VNb9O25 VTa9O25 Al2O3 Nb2O5 Al2O3 AlTaO4 Ta2O5 AlNbO4 AlNb11O29 Figure 14.Phase relations in the Al2O3±V2O5±Ta2O5 (a) and Al2O3 ± V2O5±Nb2O5 (b) systems. The Al2O37V2O57R2O5 systems. Figure 14 a shows phase relations in the Al2O3±V2O5±Ta2O5 system in a subsolidus region up to 1350 8C, and Fig. 14 b shows phase relations in the Al2O3±V2O5±Nb2O5 system in a subsolidus region up to 1100 8C. The times required to establish equilibria in these systems are the same as those for M2O3±V2O5±R2O5 (M=REE, R=Ta, Nb). No ternary compounds were found in these systems.9, 128 ± 130 It was found that the AlTaO4 phase has two crystal modifica- tions in the section AlTaO4±V2O5 (see Fig.14 a). In the vicinity of the section point where the concentrations of the components, i.e., Al2O3, V2O5 and Ta2O5, are 38 mol .%, 24 mol .% and 38 mol .%, respectively, the AlTaO4 crystals are isostructural to alumotantite (its formula is Al0.98Ta0.99Nb0.02O4) with ortho- rhombic symmetry.131 In the other points of this section studied, the AlTaO4 crystals have a structure different from that of alumotantite.132 It should also be noted that for all points of this section containing less than 50 mol.%V2O5 (except for the point specified above), the diffractograms of freshly prepared AlTaO4 samples recorded at room temperature contain additional lines. These lines disappear after storage of the samples in air for a month. These reflexions probably originate from structural order- ing of AlTaO4 crystals with involvement of lattice defects.The AlNbO4 structure is described in the database of X-ray spectra.133 TheM2O37Ga2O37Nb2O5 (Ta2O5) systems (M=La, Pr).42 The isothermal sections in these systems have not been studied. It was found that the La2O3±Ga2O3±Nb2O5 system contains a ternary compound with the composition La3Ga5.5Nb0.5O14 hav- ing the structure of Ca3Ga2Ge4O14 (a=0.8218 nm, c=0.5122 nm; the densities obtained from X-ray estimation and from experimental measurement are 5.9346103 and 5.9036103 kg m73, respectively). The compound has a narrow homogeneity range and is stable up to 1450 8C. compound (a=0.8151 nm, La3Ga5.5Ta0.5O14 The c=0.5105 nm) is formed in the La2O3±Ga2O3±Ta2O5 system, and the compounds Pr3Ga5.5Nb(Ta)0.5O14 exist in the Pr2O3 ± Ga2O3±Nb2O5 (Ta2O5) systems. Starting with M=Nd, no similar compounds isostructural to Ca3Ga2Ge4O14 are formed.V. Phase formation regularities in ternary systems It follows from the experimental phase relation diagrams that vanadatoniobates MNb2VO9 exist in equilibrium preferentially with compounds which have the structural type of fergusonite, perovskite, zircon, wolframite (ScNbO4) and with compounds a-LaNb5O14 and MNb7O19 (M=La, Ce). Vanadatotantalates MTa2VO9 coexist with such compounds as fergusonite, perov- skite, wolframite (ScTaO4), LaTaO4, MTa5O14 (M=La, Ce) and with compounds of hexagonal symmetry such as MTa7O19. All vanadatotantalates and -niobates exist in equilibrium with tetra- gonal phases such as VR9O25 and RVO5 (R=Ta, Nb).Lantha- num boratotantalate and -niobate LaRB2O7 (R=Ta, Nb) coexistPhase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds Table 3. Ionic sizes (Ref. 134) of Group III and VB elements and ratios of ionic radii. Ion (2) a Ion (1) a r1/r2 Radius (r2) /nm Radius (r1) /nm 0.01 B3+ (3) 0.0355 V5+ (4) 0.0355 0.048 0.064 0.064 0.048 0.064 0.064 V5+ (4) Nb 5+ (4) Nb 5+ (6) Ta 5+ (6) Nb5+ (4) Nb 5+ (6) Ta 5+ (6) *0.3 *0.2 *0.16 *0.16 *0.74 *0.55 *0.55 Ga3+ 0.047 ± 0.062 b a The numbers in parentheses indicate the coordination numbers. b Values for different coordination numbers are given.with such compounds as perovskite, aragonite, fergusonite, LaRO4 and monoclinic La(BO2)3, whereas lanthanum boratova- nadate La4VBO10 coexists with such compounds as monazite, aragonite and monoclinic La3BO6. All ternary compounds described above exist in equilibrium with simple oxides R2O5 (R=Ta, Nb, V), B2O3, La2O3. Solid solutions in the Y2O3±Nb2O5±Ta2O5 system coexist with such compounds as fergusonite, pyrochlore, perovskite, Y2O3, R2O5 (R=Ta, Nb); those in the La2O3±Nb2O5±Ta2O5 system coexist with such compounds as weberite, perovskite, LaTaO4 and La2O3, and those in the Sc2O3±Nb2O5±Ta2O5 system coexist with such oxides as Sc3TaO7 and R2O5 (R=Nb, Ta).Thus, it can be stated that ternary phases formed in systems of Group III and VB oxides (except for La2O3±B2O3±V2O5) exist in equilibrium with compounds that have the fergusonite and perovskite structural types and with compounds characteristic of each individual system.It is well known that no theory formulating the phase formation criteria in complex systems exists at present. The phase formation in quasi-binary vanadate systems was analysed by estimation of cation radii and with the use of the spatial- energetical P parameter.44 Let us consider the regularities in the formation of mixed crystals for Group III and VB elements depending on the ion sizes and on the ionic radii ratios in ternary compounds (Table 3). If the ratio of the ionic radius of boron to that of tantalum or niobium is *0.2, no ternary compounds of lanthanides, scan- dium and yttrium are formed in theM2O3±B2O3±Ta2O5 (Nb2O5) systems (M=Sc, Y, Ce ± Lu).The existence of the starting Ternary oxide systems of Group III and VB elements With double phases of constant composition Al2O3±V2O5±Nb2O5(Ta2O5), M2O3±B2O3±Ta2O5 (M=Sc, Y, Ce, Pr, Nd, Sm, Eu), Sc2O3±B2O3±Nb2O5 Y2O3±B2O3±Nb2O5(V2O5) of constant composition M2O3±V2O5±Ta2O5 (M=Sc, Y, La, Ce ± Lu), M2O3±V2O5±Nb2O5 (M=Sc, Y, La, Ce ± Lu), La2O3±B2O3±Ta2O5(Nb2O5, V2O5), Ln2O3±B2O3±V2O5 (Ln=Pr, Nd), La2O3±Ga2O3±Nb2O5(Ta2O5), Pr2O3±Ga2O3±Nb2O5(Ta2O5) 557 Table 4. Ratios of ionic radii in ternary compounds. M Compounds Ion (1) Ion (2) r1/r2 B3+ B3+ B3+ 0.55 ± 0.74 0.2 0.3 0.3 0.97 Sc, Y, La ±Lu V5+ La La La, Pr, Nd La, Pr R5+ R5+ V5+ V5+ R5+ Ga 3+ MR2VO9 (R=Nb, Ta) MRB2O7 (R=Nb, Ta) M4VBO10 M7V2BO17 M3Ga5.5R0.5O14 (R=Nb, Ta) members of the series with M=La, viz., LaTaB2O7 and LaNbB2O7, can be explained by the fact that the radius of the La3+ion is larger than those of scandium and lanthanides.In the Al2O3±V2O5±Ta2O5 (Nb2O5) systems, ternary compounds with hypothetical composition AlxTa(Nb)yVzOk are not formed, either. Probably, the cation radii should also be taken into account in addition to the ionic radii ratios in the anionic parts of the compounds. The ion radius of Al3+ is 0.0390 to 0.0535 nm depending on the coordination number; it is smaller than the ion radii of scandium, yttrium, lanthanum and lanthanides.This is probably the reason why ternary compounds are not formed in the Al2O3±V2O5±Ta2O5 (Nb2O5) systems [unlike in the M2O3 ± V2O5±Ta2O5 (Nb2O5) systems (M=Sc, Y, La ± Lu) where the MR2VO9 compounds exist]. In the M2O3±B2O3±V2O5 systems (M=La, Pr, Nd) at temperatures >900 8C, the M7V2BO17 compounds exist (see Ref. 127), whereas only one compound, namely, M4VBO10 with M=La, was found at temperatures <900 8C. Hence, one can probably expect the formation of compounds with molecular composition M4VBO10 for M=Pr and Nd. Thus, the dimension-related criterion for the formation of ternary compounds in oxide systems includes two conditions: (1) the cation radius should be larger than 0.0535 nm; (2) the radii ratios for the ions forming the anionic parts should be greater than 0.2 but smaller than 1.It follows from the data in Table 4 that if the ionic radii ratio is r1/r2&0.2, ternary compounds exist only in the case of M=La, whereas if this ratio is 0.55 ± 0.74, these exist in the case ofM=Sc, Y, La ± Lu. However, if the radius of theM ion is smaller than that of the Sc ion, ternary compounds are not formed ion(1)=V5+; for r1/r2=0.55 ± 0.74 [where ion(2)=Ta5+, Nb5+]. In the M2O3±Nb2O5±Ta2O5 systems, no ternary compounds with hypothetical composition MxNbyTazOr are formed either if r1/r2=1 andM=Sc, Y, La. Thus, the following classification of ternary oxide systems of Group III and VB elements can be made: (1) systems with an equilibrium between the original components and binary com- pounds; (2) systems where ternary phases are formed.In turn, Scheme 1 With ternary phases of variable composition (with limited and conti- nuous series of solid solutions) La2O3±Nb2O5±Ta2O5 with planar homogeneity ranges Y2O3±Nb2O5±Ta2O5, Sc2O3±Nb2O5±Ta2O5 with linear homogeneity ranges558 ternary phases can have a constant composition (chemical com- pounds) or a variable composition (solid solutions). The arbitrary character of this classification is due to the fact that constant composition phases always contain very narrow homogeneity regions which can be neglected. Scheme 1 shows the classification of ternary oxide systems; this classification can apparently be used to predict the existence of phase equilibria, e.g., in the Ln2O3 ± Nb2O5±Ta2O5 and Ln02O3±Nb2O5±Ta2O5 systems, where Ln and Ln0 are lanthanides belonging to the yttrium and cerium subgroups, respectively.VI. Vibrational spectra of ternary oxides Using NMR spectroscopy in MR2VO9 (R=Ta, Nb) and La4VBO10 crystals, the VO4 and VO6 anionic groups were revealed; they are generally linked through one or two oxygen atoms with the adjacent VO4 or VO6 groups, respectively. The LaTaB2O7 crystals were shown to contain BO3 triangles. Thus, one of the goals in the analysis of the vibrational spectra of ternary oxides is to determine whether the anionic part of a compound is a united complex anion or it consists of anionic groups.1. MTa2VO9 crystals The IR spectra of the MTa2VO9 crystals, where M indicates cerium subgroup elements, contain narrow intense bands in the range of 430 ± 437 cm71, whereas the remaining bands in the spectrum are broad and have medium intensity. The band positions in the range of 600 ± 900 cm71 depend on the nature of the M3+ ion: the vibration frequencies are shifted to the long- wave region with a decrease in the ionic radii in the series Ce ± Eu (Fig. 15). It is well known that a decrease in the ionic radius increases the covalence of bond between a lanthanide ion and the environment and enhances the effect of the crystal field,11 hence these bands can originate from the stretching vibrations of the Ta7O bonds involving the M7O bonds.The shape of the IR spectra of MTa2VO9 crystals, where M indicates yttrium sub- group elements, depends on the nature of the metal and changes non-monotonically (Fig. 16).88 The most significant differences between the spectra of MTa2VO9, where M indicates yttrium and cerium subgroup elements, are observed in the region of 200 ± 500 cm71, which probably corresponds to the deformation vibrations of the lattice polyhedra. The IR spectra of LaTa2VO9 (Fig. 17, spectrum 1) and ScTa2VO9 (Fig. 17, spectrum 3) are similar to the spectra of MTa2VO9, where M indicates yttrium and cerium subgroup elements, respectively. Replacement of 50% of yttrium in YTa2VO9 by europium stabilises the a-phase, and the IR spec- trum of the solid solution Y0.5Eu0.5Ta2VO9 (Fig.17, spectrum 2) Absorption 5 4321 600 800 1000 n /cm71 Figure 15. IR spectra of MTa2VO9, whereM=Ce (1), Pr (2), Nd (3), Sm (4), Eu (5). 2 8 7 65 3 1 4 800 1000 1200 Figure 16. IR spectra of MTa2VO9, whereM=Gd (1), Tb (2), Dy (3), Ho (4), Er (5), Tm (6), Yb (7), Lu (8). 321 600 800 1000 Figure 17. IR spectra of LaTa2VO9 (1), Y0.5Eu0.5Ta2VO9 (2) and ScTa2VO9 (3). resembles that of LaTa2VO9. Tables 5 ± 8 list the frequencies of vibrational spectrum maxima for compounds of the general formula MTa2VO9 (M=Sc, Y, La ± Lu). The bands in the regions of 865 ± 900, 790 ± 835, 455 ± 485 and 370 ± 385 cm71 can probably be assigned to the n1, n3, n4 and n2 vibrations of the VO4 tetrahedra, respectively, while the bands in the regions of 910 ± 930, 580 ± 710 and 430 ± 455 cm71 can be assigned to the n1, n3 and n5 vibrations of the TaO6 octahedra, respectively.Only the most intense lines in the Raman spectra of LaTa2VO9 and TbTa2VO9 were recorded.} The Raman spectra of the vanadato- tantalates CeTa2VO9 and ScTa2VO9 could not be recorded, probably due to strong absorption of radiation. The Raman spectra of MTa2VO9, whereM=Pr, Nd, Sm, Eu, Gd, Dy, Tm, Yb, Lu, in the region of 780 ± 1000 cm71 are similar (Fig. 18 ± 20): each of them contains an intense line at 855 ± 900 cm71 and two lines around 780 ± 835 cm71. The differ- ences between the spectra of the compounds containing cerium and yttrium subgroup elements are mostly observed in the range below 780 cm71.The Raman spectra of the vanadatotantalates HoTa2VO9 and ErTa2VO9 differ from those described above (see Fig. 19, spectra 2 and 3). The spectrum of HoTa2VO9 was recorded using argon laser excitation (excitation with a helium ± neon laser results in a weak vibrational spectrum at the back- ground level). It contains a group of narrow high-intensity lines at } All Raman spectra were recorded with excitation by a helium ± neon laser (l=632.8 nm) unless stated otherwise. Absorption Absorption MG Zuev n /cm71 n /cm71Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds Table 5. Frequencies (cm71) of maxima in vibrational spectra of MTa2VO9 (M=La, Ce, Pr, Nd).Intensity 7 7 7 7 7 7 623 7 7 7 7 610 870 865 790780 795785 760 805 Ce (IR) 750 633 590 470 460 475 La Pr IR Raman IR 960 7 800 7870 845 7 7 7 7 7 7 7 7 7 957 7 7 885 8737 7 7 7802 7 7 7 790 780 7940 884 7 7 7 7 77 7 8027 7 stretching vibrations of VO4, TaO6, VO6 77 785 760 7 7 7 626 965 885 860 77826 8087 7 7 7 7837 7 7 7 677 7 7 7 7 6207 7 618 585 592 7 530 469 525 475 77 520 467 7 7 7 7 431 437 440 435 380 910, 930, 940(sh.), 985, 1000 and 1045 cm71. These lines may be due to resonance phenomena caused by diffusion of laser light near the 5F4 state of the Ho3+ ion. The spectrum of ErTa2VO9 815 805 785 465 679 675 7 543 21 Dn /cm71 Figure 18.Raman spectra of MTa2VO9, where M=Pr (1), Nd (2), Sm (3), Eu (4), Gd (5). 990 855 875 585 Raman IR Raman 7 1150 990 77 7990 7890 870 8557820 7795 7683 750 7650 7 590 570 7 527 472 7437 deformation vibrations of VO4, 535 7470 450 7 7 TaO6, VO6 370 7350 530 7460 7420 370 7350 7360 77270 n(M7O) 305 7245 305 290 250 225 225, 215, 180, 155, 110, 75, 70 2 31 Dn /cm71 Figure 19. Raman spectra of MTa2VO9, where M=Dy (1), Ho (2), Er (3). 370 370 385 375 305 305 305 245 250 260 225 160 70 559 Assignment Nd Intensity955 855 880 825 805 805 translation and rotation of VO4, VO6 785 600 670 580 550 490 480 455 510 430 390 350 380 340 225 280 170 100 150MG Zuev 560 Table 6.Frequencies (cm71) of maxima in vibrational spectra of MTa2VO9 (M=Sm, Eu, Gd, Tb). Assignment Tb Gd Eu Sm Raman IR Raman IR Raman IR Raman IR 990 7 7 7 7 7940 884 7 895 7 7900 7870 990 77865 845 805 790 7960 887 77810 7 7825 7815 805 stretching vibrations of VO4, TaO6, VO6 750 675 7 7 735(sh) 680 690 683 895 77810 777735 681 7 625 627 7 7 7 1080 7 7 7 7 7 7 7 7 7 7875 7 7 7 840 802 7 7 7 7 780 7 7 7 7 76827 7 7 7 626 7 7 7 610 560 810 777 770 7670 7 635 7 7 7 7 7 7 7 590 590 595 590 580 590 527 7 535 475 550 7 7 490 485 deformation vibrations of VO4, TaO6, VO6 550 7 7 7 490 7 7 7 400(sh) 7364 510 470 445 7415 370 7 530 469 7433 398 7360 7437 398 7362 7385 7 7385 7 336 312 n(M7O) 322 7264 7305 260 322 306 288 264 7310 7245 210, 150 345 315 7250 215, 160, 130, 110, 75 translation and rotation of VO4, VO6 Table 7.Frequencies (cm71) of maxima in vibrational spectra of MTa2VO9 (M=Dy, Ho, Er, Tm). Assignment Tm Er Ho Dy Raman IR Raman IR Raman IR Raman IR 7 7 7 7 7 7 7 1010 7 7 7 7 7 7 8857 7 7895 7 7 7 7 stretching vibrations of VO4, TaO6, VO6 7820 7 840 792 7 7 8927 890 870 7 78307 7 845 805 785 812 7 7 7 670 7 7 7 7 985 7 7 7 7 955 7 930 900 77 7 7 890 7 855 7 7 7 835 7815 7 7 7 7 7 7 7 7 7 7 685 6307 7 7 7 7 7 950(sh) 7 7 7 7 7 7880 7 7825 7805 7 7735 7 7675 7600 78247 7 7 760 745 7 730 685 625 605 670 630 580 680 7592 695 640 590 690 640 590 7 7 7 5307 510 5207 7 480(sh) 7 560(sh) 7 7 7 7 7 485 476 deformation vibrations of VO4, TaO6, VO6 554 7 7 7 7 494 480 7 7440 7 7 7 7 430 7 7 7 550 505 490 7 7 7 465 7 7 455 7 7 7 7 3757 7 7 362 777385 7 430 400(sh) 7365561 Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region.Ternary oxide compounds Table 7 (continued). Assignment Tm Er Ho Dy Raman IR Raman IR Raman IR Raman IR 276 n(M7O) 330 77 7 7 7 325 308 250 240 315 275 255 245 325 310 260 236 7 7280 2507210, 170 220 225, 205, 150 220 224 translation and rotation of VO4, VO6 220, 185, 175, 165, 115, 100, 80, 65 Table 8. Frequencies (cm71) of maxima in vibrational spectra of MTa2VO9 (M=Yb, Lu, Sc, Y). Assignment Sc (IR) Lu Yb Y Raman IR Raman IR Raman IR 1090 1050 910 7 7 7 930 890 7 7900 stretching vibrations of VO4, TaO6, VO6 7 7 7 7 7 7 7 7 7 1045 7 7 1025 7 7 7 7 77910 895 7 7 885 7 7 7 830 810 895 7838 7800 815 810 7815 7 730 740 730(sh) 7 7 680 685 7 777 7 7 1020 7 7 7 1010 7 7 7 7 990(sh) 7 7 7 7 7 960(sh) 7 7 7 895 890 7 7 7 7 840 835 820 7 7 7 790 775 7 7 7 765 710 7 7 7 665 600 7927 7 7 730(sh) 7 7 7 692 640 590 670 640 610 680(sh) 630 7 7 7 625 560(sh) 479 77 535 490 7 7 485 7 7 7 574 7 7 7 490 455 7 deformation vibrations of VO4, TaO6, VO6 74807 7 7 7 7 7 7 7 430 77350 450 420 375 7 7 450 7 7 415 370 7 385 7 7360 296 268 n(M7O) 315 265 255 230 77255 7 7275 255 245 230 215, 175 translation and rotation of VO4, VO6 200, 175, 95, 90, 75 210, 170, 150, 100, 65 M=Sc, Y, La ± Lu] displays quite a few maxima (see Table 8), which may be due to vibrations of VO4 groups linked through one oxygen atom.The tantalates MTa7O19 (M=La, Ce) with the structures similar to that of YTa2VO9 are characterised by the presence of distorted TaO6 octahedra. Probably, the structure of MTa2VO9 also contains distorted TaO6 octahedra. The presence of vana- dium-based polyhedra in the lattice of vanadatotantalates pri- marily results in variation of the frequencies originating from internal vibrations of the TaO6 octahedra (n and d for Ta7O and Ta7O7Ta) and in elongation of the bonds in the octahedra contains a group of intense lines in the region of 340 ± 455 cm71. Probably, resonance phenomena involving virtual states due to diffusion of laser light near the 4F9/2 state of the Er3+ ion also occur in this case.Similar phenomena were reported for the 7F0 state of the Eu3+ ion and the 4I9/2 state of the Nd3+ ion in EuTa7O19 and LnTaO4 crystals (Ln=Nd, Eu).135 The empirical assignment of frequencies in the vibrational spectra of YTa2VO9 is based on the fact that its lattice contains almost regular VO4 tetrahedra.136, 137 The region of the stretching vibrations of these tetrahedra [730 ± 900 cm71, cf. the assignment of the stretching vibration frequencies of VO4 tetrahedra 138, 139 for the [VO4]37 ion and for the Na3M(VO4)2 compounds, where562 12 1025 900 890 890 885 3 4 Figure 20. Raman spectra of MTa2VO9, where M=Tm (1), Yb (2), Lu (3), and the Raman spectrum of vanadatotantalate YTa2VO9 (4).accompanied by a decrease in the vibration frequencies of the Ta7O terminal bonds. Some frequencies in the spectra of YTa2VO9 were assigned to vibrations of distorted TaO6 octahedra (see Table 8 and the frequency assignment in Refs 140 and 141). The spectra of orthovanadates MVO4 and the assignment of the vibration frequencies of VO4 tetrahedra have been documented.142 For example, the frequencies for YVO4 are as follows: n1=888, n2=376, n3=836.8, n4=487 cm71. Taking into account the line shapes in the Raman spectrum of the vanadatotantalate YTa2VO9, some of the lines were assigned to the vibration frequencies of the VO4 tetrahedron: n1=885, n2=375, n3=830, n4=485 cm71. It is well known that for some tetrahe- dra, such as VO3¡ 4 , ReO34 ¡, FeO¡4 , RuO24¡, FeO24¡,143 and for MTa2VO9 the frequency n3 is smaller than n1.2. MNb2VO9 crystals The structures of MNb2VO9 crystals have not been determined. Tables 9 ± 12 list the frequencies of the maxima in the vibration spectra of the vanadatoniobates MNb2VO9 (M=Sc, Y, La ± Lu).9 The bands at 865 ± 890, 795 ± 840, 470 ± 490 and 370 ± 385 cm71 can probably be assigned to the n1, n3, n4 and n2 vibrations of the VO4 tetrahedra, respectively. The bands in the IR spectra of these compounds differ in shapes and intensities. For vanadatoniobates of cerium subgroup lanthanides, comparatively intense medium-width bands at 200 ± 600 cm71 and one broad band at 600 ± 1000 cm71 are observed (Fig. 21). For instance, the NdNb2VO9 spectrum has some specific features: the band at 828 cm71 is particularly intense.In the spectra of MNb2VO9, where M=Ce, Pr, Sm, Eu, this band is less intense and is shifted from 828 cm71 towards Intensity 810 815 815 835 835 830 820 810 710 670 680 670 665 485 480 455 485 415 385 375 385 250 370 280 Dn /cm71 265 255 255 210 180 170 175 100 95 85 100 90 65 75 1234 5 600 800 1000 Figure 21. IR spectra of MNb2VO9, where M=Ce (1), Pr (2), Nd (3), Sm (4), Eu (5). higher or lower frequencies. Such frequency shifts are less pro- nounced in the deformation vibration region. The IR spectra of yttrium subgroup compounds are similar: the frequency region characteristic of deformation vibrations displays comparatively intense medium-width bands; the stretching vibration bands merge into one broad band (Figs 22, 23).It should be noted that the IR spectra of LaNb2VO9 (Fig. 24, spectrum 1) and vanada- toniobates of yttrium subgroup elements are similar to the spectra of vanadatoniobates of cerium subgroup elements. The spectra of YNb2VO9 and ScNb2VO9 (Fig. 24, spectra 2, 3) mainly differ in the intensity of the absorption bands. It is noteworthy that for compounds with M=Gd, Tb, Dy, Ho the broad band in the region of deformation vibrations splits into two narrow bands at 43 2 1 600 800 1000 Figure 22. IR spectra of MNb2VO9, where M=Gd (1), Tb (2), Dy (3), Ho (4). 1 2 4 3 900 500 700 Figure 23. IR spectra of MNb2VO9, where M=Er (1), Tm (2), Yb (3), Lu (4).Absorption Absorption Absorption MG Zuev 400 n /cm71 400 n /cm71 300 n /cm71563 Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds Table 9. Frequencies (cm71) of maxima in vibrational spectra of MNb2VO9 (M=La, Ce, Pr, Nd). Assignment Nd Ce (IR) La Pr Raman IR Raman IR Raman IR 7 7 1070 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 990 7 7 7 8757 7 855 7 77 stretching vibrations of VO4, TaO6, VO6 7 71000 7 7910 78507 7 7 7 7 7 7 7 7 7 690 77 7 7 760 690 675 630 ± 620 7 885 865 7850 816 8007 7 7 7 7757 7670 7 7610 8287 7 7 7 7 7 770 6847 7 625 ± 620 7 7 7 7 7 820 8007 7 795 7 7 785 77 7 7 670 620 ± 615 7 7 610 7 560 510 7 7 510 7 7 508 7 7 7 7 7 7 7 7 470 7 7 7 7 450 7 7 480 7 7 7 450 450 deformation vibrations of NbOn, VO4, VO6 7 7 7 7 7 495 77 7 7 7 430 420 420 7 365 7 77 7 7 7 7 7 7 7 7 7 370 365 7 7 7 4207 7 3607 310 7370 7340 286 276 280 n(M7O) 77 7255 7270 185 210, 150, 115 translation and rotation of VO4, VO6 220, 175, 135, 125, 100 Table 10.Frequencies (cm71) of maxima in vibrational spectra of MNb2VO9 (M=Sm, Eu, Gd, Tb). Assignment Tb Gd Eu (IR) Sm Raman IR Raman IR Raman IR 7 7 7 7 1000 7 7 7 7 7 7 7 7 880 880 870 890 7 7 7 990 935 885 8707 7 7 7 7 7 830 7 7 820 stretching vibrations of NbOn, VO4, VO6 7840 840 7 7 7 810 8107 7 810 7 7 7 7777860 77 7 7 7 7 7803 7 7 770 690 ± 670 7 7800 760 690 620 7 7 770 670 620 685 ± 670 7 685 610 670 590 680 620 510 505 7 7 7 7 7 7 7 7 550 510 ± 480 7490 470 7 77450 475 460 deformation vibrations of NbOn, VO4, VO6 480 455 7415 77420 7 77420 7 7 360 506 ± 467 7 7 7 7 7 7450 7 7 7 420 4127 7 7 7 385 7 7 7 415 7365 7 73707 7 360 352 350 7 7 7564 Table 10 (continued).Sm Raman IR 2707 7 7 290 270 7 7255 210, 175, 150 Table 11. Frequencies (cm71) of maxima in vibrational spectra of MNb2VO9 (M=Dy, Ho, Er, Tm).Dy Raman IR 7 1050 7 7 7 7 7 7 7 880 7 7 7 7 840 7 7 815 810 805 7 7 7 7 7 7 7 7 7 7 7 620 7 7 7 7 7 7 7 7 7 7 495 7 475 7 7 480 455 7 7 7 415 7 7 7 7 7 7 370 370 7 7 7 7 7 7 300 290 7 270 7 265 270 270 7 7 260 220, 150, 115 260 ± 270 cm71. This can be due to distortion of the polyhedra formed by theM3+ ions (M=Gd, Tb, Dy, Ho). The positions of certain bands depend on the nature of the M3+ cation: the decrease in the ionic radii in the Ce3+±Lu3+ series is accompa- nied by a shift of the band at 420 cm71 to 412 cm71 (Ho3+ is an exception). This may originate from an increase in the covalency of the bond of the lanthanide with the environment and enhance- ment of the crystal field effect as the M3+ ionic radius decreases.The Raman spectra could be recorded only for the compounds with M=La, Pr, Nd, Sm, Gd, Tb, Dy, Ho, Er; only the most intense lines were recorded for NdNb2VO9. In the spectra of lanthanum, praseodymium, samarium and gadolinium vanada- MG Zuev Assignment Tb Gd Eu (IR) Raman IR Raman IR 290 260 268 n(M7O) 7270 7 77260 175, 110 170 210, 170, 115 translation and rotation of VO4, VO6 Assignment Tm (IR) Er Ho Raman IR Raman IR7 7 7 7 7 7 7 940 860 7 77 7 7 7 stretching vibrations of NbOn, VO4, VO6 7 7 7 7 77810 7 7 810 780 77 7 7685 690 7 77 7 7 7 7 907 7 7 890 7 7 7 7 865 7 7 7 7 850 8407 7 7 7 805 775 7 7 7 7 740 7 7 725 6807 7 7 7 777 690 7 650 7 7 620 7 7 7 605 595 7 7 590 7 7 540 505 7 500 485 7 7 deformation vibrations of NbOn, VO4, VO6 7545 7 7 74887455 7414 7440 7 7370 3707 7 7 550 545 7 7 7 510 500 480 7 7 7 7 455 74157 7 7 7 7 77 7 7 4557 430 4207 405 7 390 7 7 7 7 365 7 7 7 7 356 n(M7O) 7 7280 7 175, 125, 90 230, 175 translation and rotation of VO4, VO6 toniobates, the most intense narrow band is shifted from 850 to 880 cm71 (Fig.25). Similarly, the decrease in ionic radius in the series La3+, Pr3+, Sm3+, Gd3+ results in displacement of the relatively broad band at 670 cm71 to 685 cm71. The pair of lines at 650 and 690 cm71 in the spectrum of HoNb2VO9 has the highest intensity (Fig.26, spectrum 2). This is probably due to the resonance enhancement of the Raman spectrum upon helium ± neon laser excitation of the holmium ion near the 5F5 state. The enhancement of the line at 940 cm71 occurs due to resonance effects, viz., laser light scattering near the 4F9/2 state of the erbium ion in ErNb2VO9. The bands at 850 ± 880 and 670 ± 690 cm71 in the spectra of vanadatoniobates, like those in the case of vanada-Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds Table 12. Frequencies (cm71) of maxima in vibrational spectra of MNb2VO9 (M=Yb, Lu, Sc, Y). Assignment Sc Lu Yb Y990 810 7 7 990 895 7 900 stretching 850 850 850 vibrations of 810 7 810 NbOm, VO4, 7 7 765 VO6, V7O7V 730 7 730 685 7 685 7 7 7 deformation 665, 575 vibrations of NbOm 7 7 615 555 7 550 488 7490 470 475 deformation vibrations of VO4, VO6 1 23 1000 800 600 n /cm71 610 670 610 683 670 600 610 370 Dn /cm71 Intensity 455 412 352 455 415 350 285 280 1000 990 n(M7O) Figure 24.IR spectra of MNb2VO9, whereM=La (1), Y (2), Sc (3). totantalates, originate from the stretching vibrations of the Nb7O bonds involving theM7O bonds. NMR data were used for empirical assignment of certain frequencies in the vibrational spectra of MNb2VO9. In the vanadatoniobate YNb2VO9, vanadium occupies two non-equiv- alent regular tetrahedral positions; the vanadium environment in one of them coincides with the tetrahedral environment of 4 32 1 Figure 25. Raman spectra of MNb2VO9, where M=La (1), Pr (2), Sm (3), Gd (4).Absorption 835 910 885 880 830 820 850 810 670 370 Assignment 255 260 175 4 Figure 26. Raman spectra of MNb2VO9, where M=Tb (1), Ho (2), Dy (3), Er (4). vanadium in the vanadatotantalate YTa2VO9. In ScNb2VO9 crystals, vanadium occupies two non-equivalent crystallographic positions: one position corresponds to a distorted octahedral environment while the other corresponds to a slightly distorted tetrahedral environment; the type of the VO4 tetrahedra in ScNb2VO9 is similar to one of the types of tetrahedron in YNb2VO9. The YNb2VO9 and YTa2VO9 compounds have differ- ent structures.On the other hand, YNb2VO9 and ScNb2VO9 have niobium-formed polyhedra of the same symmetries. The IR and Raman spectra of the vanadatoniobates MNb2VO9, whereM=Sc, Y, La, are virtually similar and mostly differ in the band intensities. Therefore, the structures of MNb2VO9 are likely to contain the same set of polyhedra, namely VO4, VO6 and NbOm; the structural difference between these compounds is that the polyhedra in the lattices of MNb2VO9 crystals are connected in different ways. The existence of vana- dium polyhedra in YNb2VO9 and ScNb2VO9 was established reliably; probable assignment of bands in the IR spectra of these compounds was made on this basis (see Table 12).138, 140, 144, 145 270 210 Intensity stretching vibration ofNbOm stretching vibration of VO4 deformation vibrations of NbOm 135125 115 110 220 115 100 n4(VO4) 12 The majority of lines in the Raman spectra of MNb2VO9 are broadened.This implies that the lattice atoms occupy non- equivalent crystallographic positions and form several types of similar polyhedra with different symmetries. The lines in the Raman spectra of MTa2VO9 crystals are less broadened, hence the MTa2VO9 structure has fewer non-equivalent atomic posi- tions. As in the case of vanadatotantalates, certain frequencies in the Raman spectra of vanadatoniobates were assigned based on the intensities, shapes and positions of particular lines. Comparison of these data with those for MVO4 orthovanadates (see Refs 146, 147) resulted in the assignment given in Tables 9 ± 12 (n3<n1, as in the case of MTa2VO9). The changes in the n1 and n3 frequencies in the Raman spectra of MTa2VO9 and MNb2VO9 polycrystals were analysed (see Ref.9). The spectral lines are sufficiently narrow, hence it is quite reasonable to construct various functional dependences for them. Figure 27 presents the dependences of the n1 and n3 frequencies in the Raman spectra of LnTa2VO9 and LnNb2VO9 on the atomic number (z) and the ionic radius (rLn) of Ln3+ (the ionic radii were taken from Ref. 134). The n1 frequency is given only for LnTa2VO9 crystals. Figure 27 a clearly shows the scatter of experimental points near the straight line drawn using the least- Dn /cm71 940 880 890 815 805 690 680 620 670 650 565 3 620 595 540 550 510 440 405 390 270 230 90566 a n1 /cm71 900 880 860 840 67 63 59 b n1 /cm71 890 870 8500.8 0.9 Figure 27.Dependence of vibration frequencies n1 and n3 of VO4 tetrahe- dra on the atomic number in LnTa2VO9 (a) and the Ln3+ ionic radius in LnTa2VO9 (1, 2) and LnNb2VO9 (3) (b). The experimental points refer to the frequencies n1 (1, 3) and n3 (2); the calculated curves are drawn for the frequencies n1 (solid line) and n3 (dashed line). Table 13. Frequencies (cm71) of maxima in vibrational spectra of La17xSmxTaB2O7 . x a 0 0155, 179, 215, 267 773217 7 352 370 390 423 490 300 315 7350 375 386 400 492 370 777 525(sh) 540 560 7536 564 7 570 7 7 7 595 580 590 620 615 7 7 665 675 7700 614 667 677 77 7749 775 77920 958 7 1045 7765 802 895 923 960 1029 71085 777 7 7 7 7 7 7 1170 11877 7 MG Zuev z n3 /cm71 830 123 810 790 1.0 rLn /A squares method.The n1 frequency for the LnTa2VO9 compounds increases as the lanthanide atomic number changes in the series Pr ± Lu. This phenomenon can be explained by the lanthanide contraction. The scatter of n1 values around the straight line is different for different lanthanides. The greatest deviations are observed for Tb, Nd and Lu. It is well known that non-uniform spreading of f orbitals, which depends on z and originates from spin-orbital coupling, is superimposed on the lanthanide contrac- tion.11 This effect might explain the scatter of n1 for the com- pounds under consideration.Figure 27 b shows the same kind of scatter of experimental data and an increase in n1 and n3 with a decrease in the Ln ionic radius; in addition, the straight lines have sharp breaks around rLn=0.094 nm (the ionic radius of Gd3+). The dependences of other properties in the series of f elements are also characterised by similar breaks.11 As shown by Muck et al.,148 the n1(VO4) frequency in the IR spectra of Y(PO4)x(VO4)17x mixed crystals increases as x changes from 0 to 0.7. This can be due to the effect on the bond vibrations in VO4 tetrahedra exerted, for example, by other VO4 groups.3. LaTaB2O7, La4VBO10, Ln7V2BO17 crystals and solid solutions based on them Table 13 lists the frequencies of maxima in the vibrational spectra of solid solutions based on LaTaB2O7. The bands in the regions of 718 ± 720 and 610 ± 620 cm71 can probably be assigned to the n2 and n4 vibrations of the BO3 triangles. Frequencies up to Assignment 0.1 0.05 0.01 0.001 300 311 325 300 310 320 n(La7O), n(O7B7O), n(O7Ta7O) 300 310 320 352 370 390 427 493 300 310 321 352 372 390 425 490 372 390 425 493 525 542 562 525 540 557 525 540 560 525 540 557 580 577 581 7 615 615 667 677 610 665 673 deformation vibrations of BO3; 6757 7 7 7 stretching vibrations of TaOn 718 720 720 7 stretching vibrations of TaOn 7 7 7 7 775 802 890 7960 1030 765 802 890 920 960 1027 764 802 890 920 960 1026 768 805 890 900 965 1030 7 7 7 1050 1090 1085 1085 1085 7 1170 7 1160 1170 71165 1165 stretching vibrations of BO3 1230 7 7 1230Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds Table 13 (continued).x a 0.01 0.001 0 0 1240 1235 71270 12357 7 7 7 stretching vibrations of BO3 1272 1300 1275 1295 1270 1293 71250 771316 1412 a Raman spectroscopy data. Table 14. Frequencies (cm71) of maxima in vibrational spectra of La4(17x)Ln4xVBO10 (Ln=Eu, Sm).x Assignment 0.01(Eu) 0.05(Eu) 0.05(Sm) 0.01(Sm) a 0 175 translation and rotation of VO4 n(L7O), n(O7B7O) 425 445 485 510 7 7 7440 480 510 n4(BO3) n2(BO3) 445 486 515 7593 613 715 748 594 612 715 740 275, 395 777505 7 7 7 592 612 717 745 760 775 795 829 7777777830 stretching vibrations of VO4, V7O7V 775 795 830 780 800 832 440 482 515 575 594 612 718 746 7 7 7775 800 828 a Raman spectroscopy data. 3 1000 ± 1100 cm71 correspond to the fundamental lattice vibra- tions of the simple rare-earth element tantalates (see Ref. 141). The vibrations of the B7O bonds occur in the region up to 1400 cm71 (see Ref.149). Thus, the frequencies in the region of 1170 ± 1412 cm71 in mixed-anion crystals probably correspond to vibrations of the B7O bonds in BO3. The bands in the region of 155 ± 1100 cm71 belong to the vibrations of groups containing tantalum, boron and lanthanum. The bands at 1170 ± 1270 cm71 probably arise from n3 antisymmetric stretching vibrations of isolated BO3 triangles. It is well known that the isolated BO3¡ anion is characterised by vibrations in the region of 1100 ± 1300 (degenerate antisym- metric stretching vibration) and 700 ± 900 cm71 (nonplanar degenerate deformation vibration). The reduction of boron sym- metry in the case of LaBO3 or the L-modification of SmBO3 causes the frequency of 940 cm71 to appear, which corresponds to symmetric stretching vibrations.150, 151 Association of anions in rings or chains distorts their structures, which leads to broadening and splitting of bands, overlap of regions of different vibrations and often to a high-frequency shift of the stretching vibration bands.The character and span of frequencies of this group of vibrations in the IR spectra of La17xSmxTaB2O7 is not incon- sistent with association of borate anions in rings and chains. Comparison of the spectra presented here with the results obtained for REE tantalates provides a reason to assume that the frequencies 600 ± 830 cm71 include the stretching vibrations of TaOn polyhedra in addition to the deformation vibrations of the BO3 groups, whereas vibrations at 200 ± 500 cm71 can be due to various deformation modes of O7B7O and O7Ta7O.The 567 Assignment 0.1 0.05 1240 7 1275 1300 x Assignment 0.01(Eu) 0.05(Eu) 0.05(Sm) 0.01(Sm) a 0 stretching vibrations of VO4, V7O7V 860 883 864 885 77 865 890 906 927 945 928 942 928 941 7 7 7 7 850 860 885 7 7 7 905 925 941 7 7 7 975 1000 1025 1080 1120 1135 1145 1180 1245 1275 1000 1025 1080 1120 1135 1145 1180 1242 1275 1000 1025 1080 1120 1135 1145 1180 1250 1280 71025 1080 1120 1135 1145 1180 1245 1276 Raman spectrum of the boratotantalate LaTaB2O7 contains a dominant line with n=370 cm71 (Fig.28 a). The presence of an intense line in the region of 370 ± 430 cm71 of the Raman spectrum is characteristic of REE tantalates and comes from the deformation vibration of O7Ta7O. The IR spectra of La17xSmxTaB2O7 solid solutions lack certain frequencies that are present in the spectra of LaTaB2O7 (Fig. 28 b); in addition, a series of frequencies appears (see Table 13). This means that replacement of lanthanum ions with samarium ions results in deformation of lattice polyhedra in solid solutions. This probably changes the selection rule of the vibra- tions. The same is true for the vibrational spectra of La17xEuxTaB2O7 solid solutions (see Ref. 152). Table 14 presents the frequencies of maxima in the vibrational spectra of La4(17x)Ln4xVBO10 solid solutions (Ln=Sm, Eu).According to NMR data, the lattices of the crystals indicated above contain almost regular VO4 tetrahedra which have one common oxygen atom with the adjacent VO4 tetrahedra. The IR spectra of the pure compound and its solid solutions are characterised by the presence of three intense absorption bands at 400 ± 600, 650 ± 950 and 1150 ± 1350 cm71 (see Ref. 153). The spectra of solid solutions lack some frequencies and contain some new frequencies in comparison with the spectrum of La4VBO10 (Fig. 29 a). The bands in the region of 480 ± 486 cm71 can be assigned to the n4 vibrations of the VO4 tetrahedra, while those in the regions of 1275 ± 1280, 941 ± 945, 715 ± 748 and 592 ± 621 cm71, to the n3, n1, n2 and n4 vibrations of the BO3 triangles, respectively.The Raman spectrum of the (La0.09Sm0.01)4VBO10 solid solution contains five high-intensity lines (Fig. 29 b) and two568 1400 Figure 28. Vibrational spectra of LaTaB2O7. (a) Raman spectrum; (b) IR spectrum. 400 Figure 29. IR spectrum of La4VBO10 (a) and Raman spectrum of (La0.99Sm0.01)4VBO10 (b). medium-intensity lines. Using the data for the vibration frequen- cies of BO3 groups (see Refs 138, 150), a series of IR bands were assigned to the vibrations of BO3 borate triangles (see Table 14). It follows from the assignment presented in Table 14 that La4VBO10 is a mixed-anion compound. Probably, the La7O Intensity Intensity Absorption Absorption 920 958 175 a Dn /cm71 800 1045 275 1080 395 158 179 215 1250 267 505 1316 321 b 1000 1200 1412 370 a 800 600 1000 n /cm71 bDn /cm71 600 n /cm71 536 564 830 620 850905 724 700 749 775 MG Zuev bond in the complex borates in question is essentially ionic, as shown for LaMgB5O10, LaBO3, etc.154 The IR spectra of Ln7V2BO17 contain bands ascribed to deformed BO3 groups;127 in particular, at 1172, 1215, 1310 and 1350 cm71 for Ln=Nd.* * * Thus, the possibility of formation of ternary oxide compounds in ternary oxide systems of Group III and VB elements depends both on the ratio of the ionic radii of elements in the anionic part of the compound and on the cation radius.The vibrations of complex anionic groups in ternary oxide crystals can be interpreted as vibrations of quasi-isolated polyhedra, such as VO4, VO6 and ROn, and BO3 triangles.The chemical bonds of solid solutions with complex anions involve the f orbitals, as is evident from the lattice dynamics of these objects. As a rule, ternary compounds decompose to binary ones on heating. This is also characteristic of ternary compounds of other Groups of the Periodic Table. For example, titanoniobates and -tantalates, LnTiNb(Ta)O6, undergo peritectic decomposition to give rutile and the corresponding niobates and tantalates, LnNb(Ta)O4.13 The decomposition prod- ucts of compounds with mixed cations, e.g., Ln2 Ca3(BO3)4 or LnAl3(BO3)4, contain double borates, viz., LnBO3 and AlBO3.66 The data presented in this review can serve as a basis for studies of more complex multicomponent oxide systems, e.g., M2O37V2O57Nb2O57Ta2O5, M2O37M02O37V2O57Nb2O57Ta2O5, etc.The use of spectro- metric methods allows one to understand the regularities of the formation of phases and their structures in the systems under consideration. References 1. L H Brixner Mater. Chem. Phys. 16 253 (1987) 2. Radiology 157 54A (1985) 3. W J Schipper,M F Hoogendorp, G Blasse J. Alloys Compd. 202 283 (1993) 4. P V Sergeev, N K Sviridov, N L Shimanovskii Rentgenokontrastnye Sredstva (X-Ray Contrast Media) (Moscow: Meditsina,1993) 5. V M Kartashov,M G Zuev, A F Dmitriev, A A Fotiev, in Novye Meditsinskie Tekhnologii v Reshenii Problem Mediko-Sotsial'noi Reabilitatsii. Informatsionnye Materialy (New Medical Technologies for the Solution of Medical and Social Rehabilitation Problems. Information Materials) (Ed.L N Kochutina) (Ekaterinburg : Urals Branch of the Russian Academy of Sciences, 1995) p. 93 6. M G Zuev, E Yu Zhuravleva, L P Larionov, V A Sokolov, V M Kartashov, A V Ziya, M A Elistratova, G N Mozhokina, A D Kunigan,M A Bal'bert Rentgenokontrastnye Sredstva � Slozhnye Oksidy Elementov III i V Grupp (Ortotantalat Ittriya), Ekaterinburg, 1997 [Complex Oxides of Group III and V Elements (Yttrium Orthotantalate) as X-Ray Contrast Media, Ekaterinburg, 1997]; article deposited at the VINITI, No. 2868-V97 (Moscow, 1997) 7. M A Bal'bert, Candidate Thesis in Medical Sciences, Institute of Chemistry of Solid, Urals Branch of the Russian Academy of Sciences, Chelyabinsk, 1998 8.Russ. P. 2 097 361; Byull. Izobret. (33) 351 (1997) 9. M G Zuev, L P Larionov Soedineniya RZE s Prostymi i Slozhnymi Anionami Perekhodnykh Metallov V Gruppy. Sintez. Sostav. Stroenie. Svoistva (REE Compounds with Simple and Complex Anions of Transition Metals of Group V. Synthesis. Composition. Structure. Properties) (Ed. V G Bamburov) (Ekaterinburg: Urals Branch of the Russian Academy of Sciences, 1999) 10. M G Zuev Pis'ma Zh. Teor. Fiz. 16 69 (1990) 11. G A Bandurkin, B F Dzhurinskii, I V Tananaev Osobennosti Kris- tallokhimii Soedinenii Redkozemel'nykh Elementov (Khimiya Redkikh Elementov) [The Features of Crystal Chemistry of Rare-Earth Elements (The Chemistry of Rare Element Compounds)] (Ed.V I Pakhomov) (Moscow: Nauka, 1984) 12. N S Afonskii, M Neiman Izv. Akad. Nauk SSSR, Neorg. Mater. 3 1280 (1967) aPhase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds 13. P A Arsen'ev, V B Glushkova, A A Evdokimov, E K Keler, V V Kravchenko, M V Kravchinskaya, A K Kuznetsov, Kh M Kurbanov, A V Potemkin, P A Tikhonov, M N Tseitlin Soedineniya Redkozemel'nykh Elementov. Tsirkonaty, Gafnaty, Niobaty, Tantalaty, Antimonaty (Khimiya Redkikh Elementov) [The Compounds of Rare-Earth Elements. Zirconates, Hafnates, Niobates, Tantalates and Antimonates (The Chemistry of Rare Elements)] (Eds V P Orlovskii, N N Chudinova) (Moscow: Nauka, 1985) 14. P N Iyer, A J Smith Acta Crystallogr., Sect.B 23 740 (1967) 15. A Chre tien, D Bodiot C. R. Hebd. Seances Acad. Sci. 263 882 (1966) 16. Y Yokogawa,MYoshimura, S Somiya Mat. Res. Bull. 22 1449 (1987) 17. T A Kurova, V B Aleksandrov Dokl. Akad. Nauk SSSR 201 10951971) b 18. K I Portnoi,M I Timofeeva Kislorodnye Soedineniya Redkozemel'- nykh Elementov (Spravochnoe Izdanie) [Oxygen Compounds of Rare- Earth Elements (Information Edition)] (Moscow: Metallurgiya, 1986) 19. B Langenbach-Kuttert, J Sturm, R Gruehn Z. Anorg. Allg. Chem. 548 33 (1987) 20. U Schaffrath, R Gruehn Z. Anorg. Allg. Chem. 573 107 (1989) 21. B Langenbach-Kuttert, J Sturm, R Gruehn Z. Anorg. Allg. Chem. 543 117 (1986) 22. H J Rossell J. Solid State Chem. 27 115 (1979) 23. T Tanaka, N Ishizawa,M Yoshimura, F Maruno, H Oganagi J.Solid State Chem. 114 79 (1995) 24. B V Aleksandrov, T A Kurova, D G Ugryumov, in Issledovaniya v Oblasti Prikladnoi Mineralogii i Kristallokhimii (Investigations in Applied Mineralogy and Crystal Chemistry) (Eds D A Mineev, Yu A Pyatenko, V G Feklichev) (Moscow: IMGRE, 1973) p. 69 25. V K Trunov, L N Lykova, N S Afonskii Vestn. Mosk. Univ., Ser. 2, Khim. (1) 55 (1968) c 26. A I Zaslavskii, R A Zvinchuk Dokl. Akad. Nauk SSSR 90 781 (1953) b 27. H P Rooksby, E A D White J. Am. Ceram. Soc. 47 94 (1964) 28. A Santoro,M Marezio, R S Roth J. Solid State Chem. 35 167 (1980) 29. U Schaffrath, R Gruehn Z. Anorg. Allg. Chem. 565 67 (1988) 30. B M Gatehouse J. Solid State Chem. 27 209 (1979) 31.A W S Johnson, B M Gatehouse Acta Crystallogr., Sect. B 36 523 (1980) 32. U Schaffrath, G Steinmann, R Gruehn Z. Anorg. Allg. Chem. 565 54 (1988) 33. S N Putilin, E A Krylov, N F Men'shenina, A A Evdokimov Zh. Neorg. Khim. 30 650 (1985) d 34. J G Allpress, H J Rossell J. Solid State Chem. 27 105 (1979) 35. L H Brixner, H-y Chen J. Electrochem. Soc. 130 2435 (1983) 36. A V Voloshin Tantaloniobaty. Sistematika. Kristallokhimiya. Evolyutsiya. Mineraloobrazovanie v Granitnykh Pegmatitakh (Tantaloniobates. Systematics. Crystal Chemistry. Evolution. Mineral Formation in Granite Pegmatites) (Nauka: St. Petersburg, 1993) 37. E N Isupova, E P Savchenko, E K Keler Izv. Akad. Nauk SSSR, Neorg. Mater. 11 1407 (1975) a 38. A Kahn-Harari, L Mazerolles, D Michel, F Robert J.Solid State Chem. 116 103 (1991) 39. R Hofmann, R Gruehn Z. Anorg. Allg. Chem. 602 105 (1991) 40. E M Levin J. Res. Natl. Bur. Stand. 70A 11 (1966) 41. S S Kiparisov, R A Belyaev, V V Bondarenko, V P Vyskubov Izv. Akad. Nauk SSSR, Neorg. Mater. 9 68 (1973) a 42. G G Khodzhabagyan, B V Mill' Zh. Neorg. Khim. 32 444 (1987) d 43. K Kitayama, D Zoshima, T Katsura Bull. Chem. Soc. Jpn. 56 689 (1983) 44. A A Fotiev, B V Slobodin,M A Khodos Vanadaty. Sostav, Sintez, Struktura, Svoistva (Vanadates. Composition, Synthesis, Structure, Properties) (Moscow: Nauka, 1988) 45. I A Bondar', N V Vinogradova, L N Dem'yanets, Zh A Ezhova, V V Ilyukhin, V Yu Kara-Ushanov, L N Komissarova, E V Lazarevskii, V N Litvin, B P Mel'nikov, D A Murashov, V P Orlovskii, K K Palkin, M A Petrova, I A Rozanov, N N Chudinova, A A Fotiev Soedineniya Redkozemel'nykh Elementov.Silikaty, Germanaty, Fosfaty, Arsenaty, Vanadaty (Compounds of Rare-Earth Elements. Silicates. Germanates, Phosphates, Arsenates and Vanadates) (Eds V P Orlovskii, N N Chudinova) (Moscow: Nauka 1983) 46. V A Fotiev, G V Bazuev Zh. Neorg. Khim. 29 1337 (1984) d 569 47. E M Levin J. Am. Ceram. Soc. 50 381 (1967) 48. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No. 44-391 49. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No.44-392 50. K Kitayama, T Katsura Bull. Chem. Soc. Jpn. 58 948 (1985) 51. H Brusset, F Madaule-Aubry, B Blanck, J P Glaziou, J P Laude Can. J. Chem. 49 3700 ( 1971) 52. H Brusset, R Mahe, J P Laude Bull. Soc. Chim. Fr., Pt. 2 495 (1973) 53. V G Remizov, A K Molodkin, V M Skorikov, Yu E Bogatov Zh. Neorg. Khim. 21 1323 (1976) d 54. K Kitayama, C Mizokuchi, T Katsura Bull. Chem. Soc. Jpn. 56 695 (1983) 55. K Kitayama, T Katsura Bull. Chem. Soc. Jpn. 50 889 (1977) 56. K Kitayama, H Sou, T Katsura Bull. Chem. Soc. Jpn. 56 3415 (1983) 57. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No. 37-79 58. Powder Diffraction File.Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No. 36-142 59. K Kitayama, T Katsura Bull. Chem. Soc. Jpn. 55 1820 (1982) 60. K Kitayama, T Katsura Bull. Chem. Soc. Jpn. 57 1222 (1984) 61. K Kitayama, T Sugihara, T Katsura Bull. Chem. Soc. Jpn. 52 458 (1979) 62. K Kitayama, T Katsura Bull. Chem. Soc. Jpn. 56 1084 (1983) 63. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No. 25-1415 64. O N Ustalova, Candidate Thesis in Chemical Sciences, N S Kurnakov Institute of Organic and Inorganic Chemistry, Academy of Sciences USSR, Moscow, 1980 65. P Beekenkamp Verres Refractaires 18 39 (1964); Chem.Abstr. 61 5332b (1964) 66. P A Arsen'ev, L M Kovba, Kh S Bagdasarov, B F Dzhurinskii, A V Potemkin, B I Pokrovskii, F M Spiridonov, V A Antonov, V V Ilyukhin Soedineniya Redkozemel'nykh Elementov. Sistemy s Oksidami Elementov I ± III Grupp (Khimiya Redkikh Elementov) [Compounds of Rare-Eare Elements. The Systems with Oxides of Group I ± III Elements (The Chemistry of Rare Elements)] (Eds LMKovba, P A Arsen'ev) (Moscow: Nauka, 1983) 67. J J Brown, F A Hummel Trans. Br. Ceram. Soc. 64 419 (1965) 68. J L Waring, R S Roth J. Res. Natl. Bur. Stand. 69A 119 (1965) 69. J L Waring, R S Roth Acta Crystallogr., Sect. B 17 455 (1964) 70. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No.18-1447 71. N A Toropov, V P Barzakovskii, V V Lapin, N N Kurtseva Diagrammy Sostoyaniya Silikatnykh Sistem (State Diagrams of Silicate Systems) (Leningrad: Nauka, 1969) 72. Ya G Goroshchenko Khimiya Niobiya i Tantala (The Chemistry of Niobium and Tantalum) (Kiev: Naukova Dumka, 1965) 73. O Yamaguchi, Y Mukaida, H Shigeta Z. Anorg. Allg. Chem. 574 235 (1989) 74. N A Godina, E P Savchenko, T I Panova, E N Isupova, E K Keler, in Vysokotemperaturnaya Khimiya Silikatov i Okislov (High- Temperature Chemistry of Silicates and Oxides) (Ed. E K Keler) (Leningrad: Nauka, 1972) p. 176 75. E N Isupova, N A Godina, E K Keler Izv. Akad. Nauk SSSR, Neorg. Mater. 6 1465 (1970) a 76.Diagrammy Sostoyaniya Sistem Tugoplavkikh Oksidov. Dvoinye Sistemy (Spravochnoe Izdanie) [State Diagrams of Systems of High- Melting Oxides. Binary Systems (Information Edition)] (Leningrad: Nauka, 1986) No. 5, Pt. 2 77. A Burdese,M L Borlera Ric. Sci. Rend., Ser. A 3 1023 (1963); Chem. Abstr. 60 8722a (1964) 78. R Norin Acta Chem. Scand. 23 1210 (1969) 79. G K Layden J. Am. Ceram. Soc. 46 506 (1963) 80. B F Pedersen Acta Chem. Scand. 16 421 (1962) 81. E K MuÈ ller, B J Nicholson J. Am. Ceram. Soc. 45 250 (1962) 82. O Yamaguchi, T Vegaki, Y Miyata, K Shimizu J. Am. Ceram. Soc. 70 198 (1987)570 83. M G Zuev, Doctoral Thesis in Chemical Sciences, Institute of Chemistry of Solid, Urals Branch of the Russian Academy of Sciences, Ekaterinburg, 1996 84.M G Zuev, in Khimiya i Tekhnologiya Redkikh, Tsvetnykh Metallov i Solei (Tez. Dokl. Vsesoyuz. Konf.), Frunze, 1986 [The Chemistry and Technology of Rare, Nonferrous Metals and Salts (Abstracts of Reports of All-Union Conference), Frunze, 1986] p. 184 85. M G Zuev, A A Fotiev Zh. Neorg. Khim. 31 2425 (1986) d 86. E V Arkhipova,M G Zuev, L A Perelyaeva, L V Kristallov, in Oksidy. Fiziko-Khimicheskie Svoistva (Tez. Dokl. V Vseros. Nauchn. Konf.), Ekaterinburg, 2000 [Oxides. Physicochemical Properties (Abstracts of Reports of the Vth All-Russian Research Conference), Ekaterinburg, 2000] p. 50 87. M G Zuev, E V Arkhipova Zh. Neorg. Khim. 44 1186 (1999) d 88. M G Zuev, A A Fotiev Izv. Akad. Nauk SSSR, Neorg. Mater. 23 168 (1987) a 89.M G Zuev, in Khimiya Tverdogo Tela (Tez. Dokl. Vseros. Konf.), Ekaterinburg 1996 [The Chemistry of Solid (Abstracts of Reports of the All-Russian Conference) (Ed. G P Shveikin), Ekaterinburg, 1996 ] Vol. 2, p. 46 90. M G Zuev, L V Zolotukhina Zh. Neorg. Khim. 35 2185 (1990) d 91. M G Zuev, A A Fotiev Zh. Neorg. Khim. 32 535 (1987) d 92. M G Zuev, A A Fotiev Izv. Akad. Nauk SSSR, Neorg. Mater. 23 2029 (1987) a 93. M G Zuev, in Khimiya Tverdogo Tela. Novye Neorganicheskie Materialy i ikh Fiziko-Khimicheskie Svoistva (The Chemistry of Solids. New Inorganic Materials and Their Physicochemical Properties) (Ed. G P Shveikin) (Ekaterinburg: Urals Branch of the Russian Academy of Sciences, 1997) Vol. 1, p. 77 94. M G Zuev, A A Fotiev, in International Conference on Coordina- tion Chemistry (Abstracts of Reports), Nanjing, 1987 E2 799 95.S V Smirnov,M G Zuev, V A Perelyaev Magnitnaya Vospriimchi- vost' Tantalatov ± Vanadatov Ryada Redkozemel'nykh Elementov, Sverdlovsk, 1987 (Magnetic Susceptibility of Tantalates ± Vanadates of Some Rare-Earth Elements, Sverdlovsk, 1987); article deposited at the NIITEKhIM, No. 1219-khp87 (Cherkassy, 1987) 96. M G Zuev, A A Fotiev Izv. Akad. Nauk SSSR, Neorg. Mater. 24 695 (1988) a 97. A G Malykh,M G Zuev, N P Efryushina, V P Dotsenko, F A Rozhdestvenskii, G G Kasimov, in Termicheskii Analiz i Fazovye Ravnovesiya (Mezhvuzovskii sb.) [Thermal Analysis and Phase Equilibria (Intercollegiate Collection)] (Perm': Perm' State University, 1988) p.98 98. M G Zuev, in Khimiya Tverdogo Tela (Tez. Dokl. Vseros. Konf.), Ekaterinburg 1996 [The Chemistry of Solid (Abstracts of Reports of the All-Russian Conference) (Ed. G P Shveikin), Ekaterinburg, 1996 ] Vol. 1, p. 95 99. A M Sych, A M Golub Usp. Khim. 46 417 (1977) [Russ. Chem. Rev. 46 210 (1977)] 100. P P Mel'nikov, L N Komissarova Koord. Khim. 12 1299 (1986) e 101. M G Zuev, E D Politova, S Yu Stefanovich Zh. Neorg. Khim. 36 1540 (1991) d 102. L Pettersson, B Hedman, A-M Nenner, I Andersson Acta Chem. Scand. Ser. A 39 499 (1985) 103. O W Howarth, J R Hunt J. Chem. Soc., Dalton Trans. 1388 (1979) 104. R N Pletnev, Doctoral Thesis in Chemical Sciences, Institute of Inorganic Chemistry, Siberian Branch, Academy of Sciences of the USSR, Novosibirsk, 1980 105.V M Mastikhin, O B Lapina, I L Mudrakovskii Yadernyi Magnitnyi Rezonans v Geterogennom Katalize (Nuclear Magnetic Resonance in Heterogeneous Catalysis) (Novosibirsk: `Nauka' Siberian Branch, Russian Academy of Sciences, 1992) 106. R N Pletnev, V A Gubanov, A A Fotiev YaMR v Oksidnykh Soedineniyakh Vanadiya (Nuclear Magnetic Resonance in Oxide Compounds of Vanadium) (Moscow: Nauka, 1979) 107. S P Gabuda, R N Pletnev Primenenie YaMR v Khimii Tverdogo Tela (Applications of Nuclear Magnetic Resonance in the Chemistry of Solids) (Ekaterinburg: Izd. `Ekaterinburg', 1996) 108. M G Zuev, A A Fotiev Zh. Neorg. Khim. 33 2417 (1988) d 109. M G Zuev, A A Fotiev, L S Leont'eva, G G Kasimov Zh. Neorg. Khim. 33 2976 (1988) d 110.M G Zuev, A A Fotiev Izv. Acad. Nauk SSSR, Neorg. Mater. 24 2044 (1988) a 111. M G Zuev, A A Fotiev Zh. Neorg. Khim. 34 738 (1989) d MG Zuev 112. M G Zuev, A A Fotiev Zh. Neorg. Khim. 34 2355 (1989) a 113. M G Zuev, in Khimiya Tverdogo Tela (Tez. Dokl. Vseros. Konf.), Ekaterinburg 1996 [The Chemistry of Solids (Abstracts of Reports of the All-Russian Conference) (Ed. G P Shveikin), Ekaterinburg, 1996 ] Vol. 2, p. 45 114. M G Zuev, I A Dmitriev, E V Arkhipova, in Fiziko-Khimicheskie Problemy Sozdaniya Keramiki Spetsial'nogo i Obshchego Nazna- cheniya na Osnove Sinteticheskikh i Prirodnykh Materialov (Tez. Dokl. Vseros. Konf.), Syktyvkar, 1997 [Physicochemical Problems of the Development of Ceramics for Special and General Purposes Based on Synthetic and Natural Materials (Abstracts of Reports of the All-Russian Conference), Syktyvkar, 1997] p.20 115. A G Yasnikov,M G Zuev, A A Fotiev Zh. Neorg. Khim. 35 2927 (1990) d 116. E V Arkhipova,M G Zuev, S A Bekker Zh. Neorg. Khim. 44 156 (1999) d 117. A G Malykh,MG Zuev, A A Fotiev, V P Dotsenko, N P Efryushina, in Khimiya Tverdogo Tela (Mezhvuzovskii Sb.) [The Chemistry of Solids (Intercollegiate Collection)] (Ed. Yu V Egorov) (Sverdlovsk: Urals Polytechnics Institute, 1988) p. 20 118. A G Malykh, V P Dotsenko, N P Efryushina, MG Zuev, A A Fotiev Zh. Neorg. Khim. 35 809 (1990) d 119. M G Zuev Zh. Neorg. Khim. 43 1229 (1998) d 120. V P Dotsenko, M G Zuev, A G Malykh, R N Pletnev, N P Efryushina, V S Startsev, G A Doronina Zh.Neorg. Khim. 36 2153 (1991) d 121. S Schramm, E Oldfield J. Chem. Soc., Chem. Commun. 980 (1982) 122. A G Malykh,M G Zuev Zh. Neorg. Khim. 37 891 (1992) d 123. A G Malykh,M G Zuev, V P Dotsenko, N P Efryushina Zh. Neorg. Khim. 37 2340 (1992) d 124. M G Zuev, V P Dotsenko, N P Efryushina Zh. Neorg. Khim. 37 2097 (1992) d 125. A G Malykh,M G Zuev, A A Fotiev, V P Dotsenko, N P Efryushina Zh. Neorg. Khim. 35 3168 (1990) d 126. V P Dotsenko, M G Zuev, N P Efryushina Neorg. Mater. 29 1541 (1993) a 127. L Z Gokhman, B F Dzhurinskii, V A Krut'ko,M G Komova Zh. Neorg. Khim. 41 1589 (1996) d 128. M G Zuev Zh. Neorg. Khim. 39 512 (1994) d 129. M G Zuev Zh. Neorg. Khim. 40 1573 (1995) d 130. M G Zuev, in Khimiya, Tekhnologiya i Primenenie Vanadievykh Soedinenii (Tez. Dokl. VII Vseros. Soveshch.), Chusovoi, 1996 [The Chemistry, Technology and Applications of Vanadium Compounds (Abstracts of Reports of the VIIth All-Russian Meeting), Chusovoi, 1996] p. 92 131. A V Voloshin, Yu P Menshikov, Ya A Pakhomovskii Zap. Vsesoyuz. Miner. O-va 110 338 (1981) 132. A Burdese,M L Borlera Ric. Sci Rend., Ser. A 7 454 (1967); Chem. Abstr. 63 3711h (1965) 133. Powder Diffraction File. Alphabetical Indexes. Inorganic Phases. Sets 1-47 (International Centre for Diffraction Data: Newton Square, PA: 1997) Card No. 41-347 134. R D Shannon Acta Crystallogr., Sect. A 32 751 (1976) 135. M G Zuev Zh. Prikl. Spektrosk. 49 998 (1988) f 136. M G Zuev, L A Perelyaeva, in Khimiya Tverdogo Tela (Mezhvuzovskii Sb.) [The Chemistry of Solids (Intercollegiate Collection)] (Ed. Yu V Egorov) (Ekaterinburg: Urals State Technological University ± Urals Polytechnic Institute, 1993) p. 40 137. M G Zuev, L A Perelyaeva Zh. Neorg. Khim. 36 2136 (1991) d 138. K Nakamoto Infrared and Raman Spectra of Inorganic and Coordination Compounds (New York: Wiley, 1986) 139. L V Kristallov, M Ya Khodos, L A Perelyaeva Zh. Neorg. Khim. 34 2358 (1989) d 140. N V Gundobin, Candidate Thesis in Chemical Sciences, Moscow Institute Fine Chemical Technology, Moscow, 1974 141. F A Rozhdestvenskii, M G Zuev, A A Fotiev Tantalaty Trekhvalentnykh metallov (Tantalates of Trivalent Metals) (Moscow: Nauka, 1986) 142. E J Baran, M E Escobar, L L Fournier, R R Filgueira Z. Anorg. Allg. Chem. 472 193 (1981) 143. K I Petrov, M E Poloznikova, Kh T Sharipov, V V Fomichev Kolebatel'nye Spektry Molibdatov i Vol'framatov (Vibrational Spectra of Molybdates and Tungstates) (Tashkent: FAN, 1990)571 Phase relations in ternary oxide systems of Group III and VB elements in a subsolidus region. Ternary oxide compounds 144. N V Gundobin, K I Petrov, S S Plotkin Zh. Neorg. Khim. 22 2973 (1977) d 145. N V Gundobin, K I Petrov, V V Kravchenko, S S Plotkin, B N Borisov Zh. Neorg. Khim. 19 894 (1974) d 146. E J Baran, P J Aymonino Z. Anorg. Allg. Chem. 383 220 (1971) 147. A Chaves, S P S Porto Solid State Commun. 10 1075 (1972) 148. A Muck, O Smrckova, B Hajek Collect. Czech. Chem. Commun. 47 1176 (1982) 149. L A Vostricova, K G Ione, V M Mastikhin, A V Petrova React. Kinet. Catal. Lett. 26 291 (1984) 150. J P Lapershes, P Tarte Spectrochim. Acta 22 1201 (1966) 151. Yu N Il'in, Yu M Golovin, in Khimiya i Tekhnologiya Neorgani- cheskikh Soedinenii Bora (Tez. Dokl. VI Vsesoyuz. Soveshch.), Riga, 1987 [The Chemistry and Technology of Inorganic Compounds of Boron (Abstracts of Reports of the VIth All-Union Meeting), Riga, 152. V P Dotsenko, N P Efryushina, M G Zuev Phys. Status Solidi A 153. V P Dotsenko, N P Efryushina,M G Zuev Mater. Chem. Phys. 38 154. V P Dotsenko, M G Zuev, S V Ermakova, E Z Seifulina Zh. Fiz. 1987] p. 80 144 K 103 (1994) 363 (1994) Khim. 63 2106 (1989) g a�Inorg. Matter. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) c�Moscow Univ. Bull. (Engl. Transl.) d�Russ. J. Inorg. Chem. (Engl. Transl.. J. Coord. Chem. (Engl. Transl.) f�Russ. J. Appl. Spectrosc. (Engl. Transl.) g�Russ. J. Phys. Chem. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (7) 573 ± 589 (2000) Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes N Ya Grigorieva, P G Tsiklauri Contents I. Introduction II. Synthesis of the alarm pheromones of the mold mite Tyrophagus putrescentiae, the astigmatid mite Glycyphagus domesticus and the ant Oecophylla longinoda III. Syntheses of ant trail pheromones IV. Syntheses of insect sex pheromones Abstract. construc- the for methods regiocontrolled and Stereo- Stereo- and regiocontrolled methods for the construc- tion regiospecific the for and bond C=C a of tion of a (Z)-trisubstituted )-trisubstituted C=C bond and for the regiospecific introduction total in exemplified are fragment chiral a of introduction of a chiral fragment are exemplified in total syntheses syntheses of insect pheromones belonging to ( The alkenes.of insect pheromones belonging to (Z)-trisubstituted )-trisubstituted alkenes. The bibliography includes 113 references bibliography includes 113 references. I. Introduction In the last two decades, it was found that the biological activities of a series of important natural compounds, such as low-molec- ular-weight bioregulators, viz., polyprenols, dolichols, antibiotics of a strobilurin family and pheromones of different species of insects, are associated with the presence of a (Z)-trisubstituted alkene fragment in these molecules. In this connection, the problem of the regio- and stereocontrolled synthesis of (Z)-tri- substituted alkenes is of substantial methodological and practical interest.A rather small group of insect pheromones containing this structural fragment represents a `proving ground' for the elucidation of the synthetic potential of methods developed for the formation of (Z)-trisubstitutedC=Cbonds. In the total syntheses of these pheromones, all known procedures for the stereoselective construction of (Z)-trisubstituted alkenes were employed. These methods can find use in the preparation of other biologically active compounds. In addition, the approaches developed for the formation of chiral centres are also of obvious interest. In the present review, methods for the regio- and stereo- controlled construction of (Z)-trisubstituted C=C bonds and the regiospecific introduction of a chiral centre, which are of importance for organic synthesis, are exemplified in total synthe- ses of pheromones of the above-mentioned group.We believe that the solution of this problem calls for system- atisation of the available data on the syntheses of the target structures, which are ranked in the order of increasing difficulty. Attention will be focussed on procedures for the construction of (Z)-trisubstituted alkene fragments and methods for the forma- tion of chiral centres. N Ya Grigorieva, P G Tsiklauri N D Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky prosp. 47, 117913 Moscow, Russian Federation. Fax (7-095) 135 53 28. Tel. (7-095) 938 36 96. E-mail: ves@carc.ioc.ac.ru Received 28 March 2000 Uspekhi Khimii 69 (7) 624 ± 641 (2000); translated by T N Safonova #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n07ABEH000588 573 573 574 578 II.Synthesis of the alarm pheromones of the mold mite Tyrophagus putrescentiae, the astigmatid mite Glycyphagus domesticus and the ant Oecophylla longinoda In the late 1970s, Kuwahara et al.1, 2 demonstrated that neryl formate (1) is the alarm pheromone of one of the major pests of cereal crops, viz., the mold mite Tyrophagus putrescentiae. Later,3 this conclusion was confirmed by bioassays of synthetic neryl formate (1) prepared from nerol (2). MnO2 CHO 3 OH HCO2COMe 2 Py OC(O)H 1 Later, Kuwahara and Koshii 4 have found that neral (3), which can also be readily prepared from nerol (2), is the alarm pheromone of the astigmatid mite Glycyphagus domesticus.a,b-Unsaturated aldehyde 4, which is a hexanal dimerisation product, was isolated by Baker and coworkers in 1975.5 This is the alarm pheromone of the African weaver ant Oecophylla longinoda. Although, in strict accordance with the IUPAC nomenclature, aldehyde 4 is the (E)-isomer, the carbon substituents at the C=C bond are in cis positions with respect to each other. Disubstituted a,b-unsaturated aldehydes of this type are precursors of (Z)-meth- ylalkenes. Hence, we believe that it is reasonable to consider the synthesis of the aldehyde 4 in the present review. In 1966, Basavaian and Hyma 6 synthesised the aldehyde 4 with the use of the Baylis ± Hillman reaction 7 for the regio- and stereoselective construction of its trisubstituted alkene fragment.Hydroxy ester 5, which was prepared by the condensation of hexanal with methyl acrylate under conditions proposed by Baylis and Hillman,7 was converted into acetate 6. The reaction of the latter with PrMgBr afforded ester 7 as the only product. Reduc- tion of the ester group and oxidation of the resulting alcohol 8574 with pyridinium chlorochromate gave rise to the aldehyde 4 (the overall yield was 20.5%). HO AcO CO2Me CO2Me a b C5H11 C5H11 5 6 CHO CH2OH CO2Me c d Bun C5H11 Bun Bun C5H11 C5H11 (E)-4 8 7 (a) AcCl, Py; (b) PrnMgBr; (c) Bui2AlH; (d ) PCC (pyridinium chlorochromate), CH2Cl2.Asimpler yet highly stereoselective procedure for the synthesis of the aldehyde 4 8 involves condensation of N-tert-butylhexan- imine with hexanal yielding b-hydroxy imine 9. Acid scission of the latter afforded thermodynamically more favourable 9 alde- hyde (E)-4 in 70% yield (the stereoselectivity was *98%). The aldehyde (E)-4 was also identified by Baker et al.10 as a component of the alarm pheromone of the beetle Amblypelta nitida. a b, c C5H11CHO C5H11CH NBut OH ButN CHCHCHC5H11 d (E )-4 Bun 9 (a) ButNH2; (b) LDA (lithium diisopropylamide); (c) C5H11CHO; (d) H3O+. III. Syntheses of ant trail pheromones 1. Trail pheromones of the fire ant Solenopsis invicta As early as 1959, Wilson 11 demonstrated that fire ants produce a special trail pheromone, which allows members of the community to find routes to sources of food, new areas, etc.Van der Meer and coworkers 12 isolated this pheromone, established that it represents a mixture of at least five closely related compounds and separated these compounds by prepara- tive GLC. Based on the results of studies of the physicochemical properties of individual components of this mixture, two com- pounds were assigned to the a-farnesene series and one compound was assigned to the allofarnesene series. The configurations of the pheromone components were established by comparing the spec- tral data (UV, 1H NMR and mass spectra) of natural and authentic samples of a- and allofarnesenes synthesised in an unequivocal way.Thus dehydration of (Z)-nerolidol 10 in a POCl3 ± Py system at 160 8C afforded a mixture of isomeric a-farnesenes, (Z,Z)- and (E,Z)-11, and (Z)-b-farnesene, (Z)-12. Dehydration of the isomer (E)-10 under the same conditions gave rise to a mixture of (E,E)- and (Z,E)-a-farnesenes (11) and (E)-b- farnesene [(E)-12].12 + OH (Z,Z)-11 (Z)-10 + + (Z)-12 (E,Z)-11 N Ya Grigorieva, P G Tsiklauri + OH (Z,Z)-11 (Z)-10 + + (Z)-12 (E,Z)-11 Based on comparison of the physicochemical characteristics of farnesenes 11 and the components of the fire ant trail pher- omone, it was reliably established that the major component of this pheromone has the structure (Z,E)-11 and one of the minor components has the structure (E,E)-11. It should be noted that a-farnesene (Z,E)-11 was identified as a component of trail pheromones also of other ant species.13 With the aim of establishing the stereochemistry of the components of the fire ant trail pheromone belonging to the allofarnesene series, all eight possible isomeric allofarnesenes were synthesised using the Wittig reaction.14 It appeared that the condensation of ylides 13 and 14 generated from geranyl- and nerylphosphonium bromides, respectively, with tiglic aldehyde (15) proceeded stereoselectively to form allofarnesenes (E,E,E)-16 and (E,E,Z)-16, respectively.To the contrary, the condensation of triphenylphosphonium ethylide 17 with isomeric pseudoionones (E,E)-18 or (E,Z)-18 was non-stereoselective. The reaction of the isomer (E,E)-18 afforded a mixture of (E,E,E)- and (Z,E,E)-16, whereas the isomer (E,Z)-18 gave a mixture of (E,E,Z)- and (Z,E,Z)-16.The condensation of the (Z,E)-pseudoionone, (Z,E)-18, with the ylide 17 yielded a mixture of (E,Z,E)-16 and (Z,Z,E)-16, whereas the reaction of the O PPh3+ 15 13 O+MeCH PPh3 17 (E,E )-18 + (E,E,E )-16 (Z,E,E )-16 15 PPh3 14 (E,E,Z)-16 17 O (E,Z)-18 + (E,E,Z)-16 (Z,E,Z)-16 (E,E,E)-16Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes 17 O (Z,E)-18 + (E,Z,E)-1617 O (Z,Z)-18 + (E,Z,Z)-16 (Z,Z)-pseudoionone [(Z,Z)-18] with the ylide 17 afforded a mixture of (E,Z,Z)-16 and (Z,Z,Z)-16. Comparison of the spec- tral properties (UV, 1H NMR and mass spectra) of the individual allofarnesenes isolated by preparative GLC with those of the third component of the fire ant trail pheromone demonstrated that the latter is the stereoisomer (Z,Z,Z)-16.Bioassays of the allofarne- sene isomers demonstrated that the compounds (E,Z,E)-, (Z,Z,E)- and (E,Z,Z)-16 with the (Z)-configuration of the C(4)=C(5) bond exhibit the trail pheromone activity. The fourth and fifth components of the trail pheromone of S.invicta were identified 15 as (Z,E)- and (E,E)-homofarnesenes 19 by comparing the physicochemical characteristics of these pher- omone components with those of the compounds (Z,E)-, (E,E)-, (E,Z)- and (Z,Z)-19, which have been specially synthesised from nerol or geraniol. OH SPh PhS a, b d O f CHO (Z,E)-19 a, b, c, d, e, f OH (E,Z)-19 (a) PBr3, Py; (b) MeCH(SPh)2; (c) HgCl2; (d) H2C=C(Me)MgBr; (e) PCC, CH2Cl2; (f) Ph3P=CH2. 2.Trail pheromone of the ant Leptogenys diminuta In 1991, Attygalle et al.16 isolated the trail pheromone of the ants Leptogenys diminuta. Based on the results of physicochemical studies, the authors assigned the structure of (Z)-isogeraniol (20) to this pheromone. To decide between the (Z)- and (E)-config- urations of compound 20, the retention times (GLC) of the (Z,Z,E )-16 (Z,Z,Z)-16 cOH e + (E,E )-19 + (Z,Z)-19 575 natural pheromone and the samples of (Z)- and (E)-20 synthesised from citral enol acetates were compared. OH OAc a b CHO + OH OAc+ (Z)-20 (E)-20 (a) Ac2O, NaOAc; (b) NaBH4.3. Trail pheromone of Pharaoh's ant Monomorium pharaonis (faranal) In 1977, Ritter et al.17 isolated the trail pheromone of Pharaoh's ant Monomorium pharaonis, which is a pest of food and a danger- ous carrier of wound infections. Based on the results of phys- icochemical studies, the structure of 3,4,7,11-tetramethyldeca- (6E,10Z)-dienal was assigned to this pheromone named faranal. However, the absolute configurations of the C(3) and C(4) centres in the molecule remained an open question up to 1980 when Kobayashi and coworkers 18 synthesised isomeric aldehydes 21 and demonstrated that only the compound (3S,4R)-21 exhibited biological activity and, hence, the natural faranal had this config- uration.The key step of the synthesis involved condensation of (2E,6Z)-homogeraniol pyrophosphate 22 with (3E)- or (3Z)-3- methylpent-3-enyl pyrophosphate catalysed by farnesyl-pyro- phosphate synthetase leading to chiral pyrophosphates (4S)- and (4R)-23, which was in accord with the results obtained previ- ously.19, 20 These compounds were converted into aldehydes (4S)- and (4R)-24, respectively. Selective reduction of the C(2)=C(3) double bond in the aldehydes (4S)-24 and (4R)-24 afforded a mixture of the diastereomeric aldehydes (3R,4S)- and (3S,4S)-21 Et a OPP 22 Et b, c OPP (4S)-23 Et R1 R2 d, e CHO CHO (4S)-24 (3R,4S)-21 (R1=Me, R2=H); (3S,4S)-21 (R1=H, R2=Me)d, e b, c f CHO 22 OPP (4R)-23 (4R)-24 Et R1 R2 CHO (3R,4R)-21 (R1=Me, R2=H); (3S,4R)-21 (R1=H, R2=Me) , farnesyl-pyrophosphate synthetase; PP is P2O6H3; (a) (CH2)2OPP (b) alkaline phosphatase; (c) MnO2; (d) Et3SiH, (Ph3P)3RhCl; (e) H3O+; , farnesyl-pyrophosphate synthetase ( f ) (CH2)2OPP576 and a mixture of the diastereomers (3R,4R)- and (3S,4R)-21, respectively. The individual compounds were isolated by HPLC.Homogeraniol 25 required for the preparation of the pyro- phosphate 22 was synthesised 21 using a reaction sequence in which the key step of chain elongation (a) was non-stereoselective and, hence, it was necessary to separate isomers by rectification or chromatography. Et Et a d, e, f b, c MeCOEt CO2Me Br Et Et a b CO2Me O Et OH 25 (a) (EtO)2P(O)CH2CO2Me, MeONa; (b) LiAlH4; (c) PBr3, Py; (d ) Mg, MeCOCH2CO2Me; (e) NaOH; ( f) H3O+.Generally, the synthesis of farnal [(3S,4R)-21] involves cou- pling of two building blocks. One of these blocks contains either only (Z)-trisubstituted alkene fragments or simultaneously (Z)- and (E)-trisubstituted fragments, while the second block contains the chiral portion of the molecule. The construction of the blocks of both types and their coupling to form the target faranal molecule [(3S,4R)-21] are considered below. Mori and Ueda 22, 23 used an allylic alcohol 26 as the (Z)-tri- substituted alkene synthon. Its synthesis involved isomerisation of 3-methylpent-1-en-4-yn-3-ol (27) under the action of lithium acetylide and reduction of the triple bond in the alcohol (Z)-28 isolated from its mixture with (E)-28 (the ratio 85 : 15).HC HC a C C CH2OH HC C + 27 OH CH2OH (E)-28 (Z)-28 Et CH2OH b, c (Z)-28 26 (a) HC CLi, (b) NH2NH2 .H2O; (c) H2O2, EtOH. Baker et al.24 performed a highly stereoselective synthesis of the alkene synthon 29 based on the cis-addition of organocopper compounds to terminal alkynes (this reaction had been discovered by Corey and Katzenellenbogen 25). The reagent 30 was added to the starting propyne to give (Z)-homoallylic alcohol 31. The alcohol 31 was converted into enyne 32 via tosylate. Successive treatment of compound 32 with a cuprate complex 3326 and iodine gave the target iodide 29 in 33.7% overall yield with a stereoselectivity exceeding 96%.Et Cu(Me2S) . MgBr2 a b MeC CHEt Et CH e, f c, d OH C 32 31 Et I 29 ; (c) TsCl, Py; (a) EtCu(Me2S) . MgBr2 (30); (b) PrC CLi, O (d ) HC CLi; (e) MeCu(Me2S) . MgBr2 (33); ( f ) I2. An analogous scheme was used in the synthesis of faranal performed by Mori and coworkers.27 They succeeded in increas- N Ya Grigorieva, P G Tsiklauri ing substantially the yield in the step of conversion of the enyne 32 into the iodide 29 by applying Wipf's modification 28 of carbo- alumination. The synthon 34 was prepared employing the cuprate proce- dure for the construction of the (Z)-alkene fragment of faranal.29 Et Et Cu(Me2S) . MgBr2 a b MeC CH O 34 (a) 30; (b) BuC CLi, H2C CHCOMe. An alternative procedure for the construction of the (Z)-tri- substituted alkene fragment was used in the synthesis of bromide 35, which is a key intermediate in the synthesis of faranal.30, 31 Epoxide 36, which can readily be prepared from benzyl ether of geraniol, was converted into allylic alcohol 37 in high yield. Sharpless's epoxidation of the latter, scission of the epoxy group in the resulting epoxide 38 with lithium dimethyl cuprate and, finally, elimination of the hydroxy groups from diol 39 according to a procedure proposed by Tanaka et al.32 afforded benzyl ether 40.The latter was converted into the bromide 35 according to standard procedures in a yield of 24% with respect to geraniol. a b OBn 36 O O c d 38 OH 37 OH Et Et f, g e OBn HO 40 OH 39 Et Br 35 (a) m-Chloroperbenzoic acid (MCPBA); (b) (PriO)3Al, PhMe; (c) ButO2H, VO(acac)2; (d) Me2CuLi; (e) HC(OMe)2NMe2, Ac2O; ( f ) Li/NH3; (g) PBr3, Py.Conversions of the above-described alkene synthon blocks 26, 29, 34 and 35 into faranal [(3S,4R)-21] involved their coupling with blocks containing the 3,4-dimethylbutanal fragment. In many cases, an adduct of butadiene with maleic anhydride 41, which was smoothly converted into cis-dimethylcyclohexene (42), was used as the starting compound for the preparation of these blocks. The ring-opening in the compound 42 and the trans- formation of dicarboxylic acid 43 giving rise to the target synthons have been performed in two ways. The path I 29 involved Path I O c a, b, a CO2H d, a O HO2C 43 42 41 O CO2Et e, f, g, h HOCH2 + CO2H Ph3PCH2 44 I7 (a) LiAlH4; (b) MeSO2Cl, Py; (c) KMnO4, Bu4NHSO4, H2O± PhH; (d ) EtOH, H2SO4, H2O; (e) TsCl, Py; ( f ) LiI, MeCN; (g) NaOH; (h) Ph3P, PhH.Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes oxidation of dimethylcyclohexene 42 with potassium permanga- nate under conditions of phase transfer catalysis followed by the transformation of the carboxy groups in the dicarboxylic acid 43 to form phosphonium salt 44. The path II 24 involved ozonation of dimethylcyclohexene 42 to form the acid 43.Its pyrolysis afforded ketone 45 whose Baeyer ± Villiger oxidation gave rise to lactone 46 (the total yield was 42.8% with respect to the adduct 41).Treatment of the lactone 46 with anhydrous HBr in EtOH afforded bromo ester 47, which was converted into iodide 48 according to standard procedures. The iodide 48 is a synthon for the construction of ()-faranal.{ Path II O e c a, b d 43 42 O O46 45 CH2OTHP CO2Et f, g, h Br I 48 47 THP is tetrahydropyran-2-yl; (a) O3; (b) CrO3; (c) Ba(OH)2, D; (d ) MCPBA; (e) HBr, EtOH; ( f ) LiAlH4; (g) 3,4-dihydro-2H-pyran (DHP), TsOH; (h) LiI, MeCN. The target racemic faranal [()-21] was synthesised 29 by condensation of the ylide generated from the phosphonium salt 44 with the ketone 34. The resulting 54 : 46 mixture of isomeric acids (6E)- and (6Z)-49 was converted into a mixture of the racemic faranal [()-21] and its (6Z)-isomer through the corre- sponding ethyl esters and alcohols.The isomer ()-21 was isolated from the latter mixture by preparative GLC in 9.1% total yield. Et a, b c, d, e 44 CO2H ()-21 49 (a) NaH, DMSO; (b) 34; (c) CH2N2; (d ) LiAlH4; (e) PCC, CH2Cl2. To reach the same goal, Baker et al.24 used condensation of a lithium derivative 50, which was prepared from the iodide 29, with the iodide 48. This condensation is highly stereoselective. The resulting tetrahydropyranyl ether 51 was converted into the ()-faranal [()-21] in a total yield of 18% with respect to the iodide 48. Et a b 29 Li 50 Et c, d CH2OTHP ()-21 51 (a) ButLi, Et2O,790 8C; (b) 48, Et2O±THF (1 : 3); (c) H3O+; (d ) PCC, CH2Cl2. Mori and Ueda 22, 23 were the first to synthesise natural (3S,4R)-(+)-21.The key step of this scheme involved asymmetric resolution of hydroxy acid ()-52, which was prepared from the lactone ()-46, using treatment with (S)-( ± )-a-phenylethylamine. Acidification of the resulting salt 53 afforded the lactone (+)-46 whose (3R,4R)-configuration was confirmed by conversion into (R)-citronellic acid. { Recently, Tolstikov and coworkers 33 have reported an efficient synthesis of methoxymethyl analogues of the compound 48. a b HO ()-46 CO2H 52 c + HO CO¡2 NH3 H 53 Ph e EtO2C O OH (3R,4R)-54 (5R,6R)-55 OTHP BrCH2 (5R,6R)-56 Et l 26 h, i, j, k SO2Ph 57 Et PhSO2 (2R,3R)-58 Et (2R,3R)-59q, g CN Et (a) NaOH; (b) (S)-( ± )-a-phenylethylamine; (c) HCl; (d) Bui2AlH; (e) (EtO)2P(O)CH(Me)CO2Et, NaOEt; ( f ) DHP, TsOH; (g) LiAlH4; (h) BuLi; (i ) TsCl; ( j ) LiBr, Et2O±HMPA (hexamethylphosphoric triamide); (k) PhSO2Na, DMF; (l ) BuLi, THF±HMPA, (5R,6R)-56; (m) Li/NH3; (n) H3O+; (o) TsCl, Py; (p) NaCN, DMSO; (q) NaOH, EtOH; (r) PCC, CH2Cl2. Reduction of (3R,4R)-46 into chiral lactol 54 followed by the Horner ± Emmons reaction of the latter with triethyl 2-methyl-2- phosphonoacetate afforded hydroxy ester (5R,6R)-55, which was converted into bromide (5R,6R)-56 using a sequence of standard reactions.The chiral bromide (5R,6R)-56 was condensed with sulfone 57 prepared from the alcohol 26. Desulfation of the resulting sulfone 58 followed by the replacement of the protective group afforded tosylate (2R,3R)-59, which was the starting product in the final steps of the synthesis of faranal [(3S,4R)- (+)-21] identical with the natural specimen. a b O 61 42 O g O I 60 577 d O O (3R,4R)-46 OH f, g, h, i, j m, n, o OTHPp OTs r CH2OH (3S,4R)-(+)-21 OH c, d, e, f 63578 Et O h, i (3S,4R)-21 O 64 (a) MCPBA; (b) (62); (c) O3, CH2Cl2; (d ) NaBH4; N CH2 NHEt Li (50); (e) MeCOMe, TsOH; ( f ) I2, Ph3P, imidazole; (g) (h) H3O+; (i ) NaIO4.Later, Mori and Murata 27 developed a shorter synthetic route to faranal [(3S,4R)-(+)-21]. The key chiral synthon for this compound, viz., iodide 60, was prepared in five steps starting from dimethylcyclohexene 42. The asymmetric resolution of meso-epoxide 61 using the lithium derivative of (S)-pyrrolidin-2- ylmethyl)pyrrolidine (62) afforded allylic alcohol 63, which was purified by recrystallisation of the corresponding 3,5-dinitroben- zoate to ee *100%.Ozonation of the compound 63, reduction and protection of the hydroxy groups at C(1) and C(2) in the resulting triol followed by substitution of the hydroxy group at C(6) gave rise to the iodide 60. Condensation of 60 with 6-ethyl-2- methylhepta-1,5-dienyllithium (50) resulted in enantiomerically pure acetal 64. Hydrolysis of 64 followed by periodate oxidation of the resulting diol gave rise to (3S,4R)-21 (in a total yield of 12.5% with respect to propyne), which was used for the prepara- tion of the lithium derivative 50.The chiral fragment of faranal was constructed also with the use of lactone (3S)-65 formed from methyl hydrogen (3R)-3- methylglutarate [(3R)-66].30,31 Enzymic hydrolysis of the dimethyl ester 67 using pig liver esterase (PLE) afforded the ester 66 with an optical purity of 85%. This ester was purified by recrystallisation of 1-cinchonidinium salt to ee 100%. Alkylation of the enolate prepared from lactone 65 with the bromide 35 gave rise to lactone (2R,3S)-68 containing *6% of the (2S,3S)-isomer. Transesterification of the lactone 68 followed by protection of the hydroxy group in the resulting hydroxy ester afforded ester 69 whose methoxycarbonyl group was converted into the methyl group in three standard steps. The resulting tetrahydropyranyl ether 70 was converted into (3S,4R)-21 in a total yield of *2% with respect to geraniol, which was the starting compound for the preparation of the bromide 35.c a b CO2Me CO2Me MeO2C HO2C O O (3S)-65 67 (3R)-66 Et d, e 68 O O Et f, g, f OTHP 69 CO2Me Et OTHP h, i (3S,4R)-21 70 (a) PLE; (b) Na/NH3, EtOH; (c) Et2NLi, 35; (d ) MeOH, Et3N; (e) DHP, Py . TsOH; ( f ) LiAlH4; (g) MeSO2Cl, Py; (h) H3O+; (i ) PDC (pyridinium dichromate), CH2Cl2. N Ya Grigorieva, P G Tsiklauri IV. Syntheses of insect sex pheromones 1. Synthesis of sex pheromones of scales a. Sex pheromone of the California red scale Roelofs et al.34, 35 isolated the female-produced sex pheromone of the California red scale Aonidiella auranti, which is the major pest of citrus crops distributed throughout the tropical and subtropical regions.This pheromone was found to consist of two components. Based on the data of physicochemical studies, it was proposed that these components have structures 71 and 72. OAc OAc 72 71 Due to the deficiency of the isolated natural substance, the absolute configurations of these compounds were established only by comparing the physicochemical characteristics of the phero- mone components with those of all possible isomers synthesised starting from (S)- and (R)-carvone. It was found that only the (3S,6R)-acetate 71 and the (3Z,6R)-acetate 72 exhibit the phero- mone activity. The retrosynthetic analysis of the molecule 72 shows that the simplest approach to its construction involves the Wittig reaction between synthon blocks 73 and 74.This approach was used in the first synthesis of the compound 72.36 O OR + Ph3P 72 74 73 The aldehyde (3R)-73 was prepared by oxidative scission of (S)-carvone (75) via epoxide 76 and diol 77. Ethoxy lactone 78 prepared from the diol 77 was converted into compound 79 and then into bromide 80 in three standard steps. Treatment of the bromide 80 with vinyllithium followed by hydrolysis afforded the aldehyde 73. O OHOH O O O c a b 77 76 75 OEt O O e, f, g d CH(OEt)2 EtO2C 79 78 Br j, k h, i CHO CH(OEt)2 80 (3R)-73 OAc (3Z,6R)-72 (a) H2O2, NaOH; (b) HClO4, H2O, THF; (c) Pb(OAc)4, EtOH; (d ) HC(OEt)3, TsOH, EtOH; (e) LiAlH4; ( f ) TsCl, Py; (g) NaBr, HMPA; (h) CH2=CHLi; (i) H3O+; ( j) Ph3P+CH(Me)(CH2)2OH Br7 (81), 2 equiv.of BuLi; (k) Ac2O, Py. The reaction of the aldehyde 73 with ylide obtained from phosphonium salt 81 gave rise to a 52 : 48 mixture of isomeric alcohols. Separation of a mixture of (3Z)- and (3E)-acetates 72Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes obtained from these alcohols by preparative GLC afforded compound (3Z,6R)-72 in 2.53% total yield.{ The acetate (3Z,6R)-72 was also synthesised 37 starting from (R)-limonene 82 according to a similar strategy, but using a more efficient procedure. Ozonolysis of limonene 82 yielded oxo acetal 83, which was converted into enyne acetal 85 through hydroxy acetal phosphate 84 according to Negishi.38 Hydrolysis of the latter yielded aldehyde, which was subjected to hydrogenation to form the diene aldehyde (3R)-73.Condensation of the latter with ylide generated by the reaction of (Me3Si)2NLi with the phospho- nium salt 81 afforded a 2 : 1 mixture of isomeric alcohols, which were separated by chromatography. Acetylation of the (3Z)- isomer produced (3Z,6R)-72 in 37% total yield. O a b CH(OMe)2 82 83 CH C d, e CH(OMe)2 85 OAc (3Z,6R)-72 (a) O3, MeOH,770 8C; (b) LDA, ClP(O)(OEt)2; (c) ; (d) H3O+; (e) H2, Lindlar catalyst; ( f ) 81, (Me3Si)2NLi; (g) Ac2O, Py. A substantial drawback of the above-considered syntheses of the acetate 72 is the non-stereoselectivity of the key step of the construction of the (Z)-trisubstituted C=C bond by the Wittig reaction.Caine and Crews 39 have overcome this drawback using another synthetic scheme. It appeared that when treated with the ylide, which was generated from the salt 81 under the action of BuLi in a mixture of THF and HMPA, b-hydroxy a-phenylthio ketone 86 prepared from the oxirane 76 40 underwent the retro- aldol reaction followed by the Wittig olefination to form alcohol (3Z)-87. This compound did not contain an admixture of the (3E)- isomer but was contaminated with desmethyl analogue 88 (*15%). The resulting mixture was converted into a mixture of the corresponding acetoxy sulfoxides from which sulfoxide 89 was isolated by chromatography. Thermal desulfinylation of the sulfoxide 89 afforded trienone 90, which was converted into a mixture of diacetates 91.Hydrogenolysis of the acetates 91 yielded a 89 : 11 mixture of the pheromone (3Z,6R)-72 and its regioisomer 92. Separation of this mixture was a very tedious procedure. Thus, in spite of the stereoselectivity of the construction of the (Z)-tri- substituted C=C bond, this scheme has no advantages over the scheme proposed by Becker and Sahali 37 as regards both the number of steps and the yield of the target compound 72. Still and Mitra 41 developed a more efficient procedure for the stereoselective construction of the (Z)-trisubstituted alkene frag- ment in the compound 72. It was found that the reactions of BuLi with allylic trialkylstannylmethyl ethers of the type 93 were { Simultaneously, the isomer (3E,6R)-72 was isolated; the isomers (3Z,6S)-72 and (3E,6S)-72 were prepared analogously starting from (R)-carvone.OP(O)(OEt)2 c CH(OMe)2 84 f, g CHO (3R)-73 NH 579 O SPhOH O O a b 76 86 OH SPh OHc, d + Me H O 88 87 OAc SOPh e O 89 OAc f, c O 90 OAc g AcO 91 OAc (6R,3Z)-72+Me 92 (a) PhSH, Et3N, MeCN; (b) 81, 2 equiv. of BuLi, THF, HMPA; (c) Ac2O, Py; (d ) MCPBA; (e) 135 8C, 0.15 mmHg; ( f ) LiAlH4; (g) HCO2NH4, PdCl2(PPh3)2, dioxane. accompanied by a [2,3]-sigmatropic rearrangement yielding (Z)-trisubstituted homoallyl alcohols. OH HO Bu3SnCH2O a b R R R 93 R=Alk; (a) KH, Bu3SnCH2I; (b) BuLi. This finding has been used in the synthesis of the racemic acetate 72.The key intermediate, viz., alcohol 94, was prepared in four steps starting from ethyl 3-methylcrotonate. The transforma- CO2Et CH2Br a, b, c d e f, g OH OCH2SnBu3 95 94 OAc ()-72 (a) LDA, CH2=CH(CH2)2Br; (b) LiAlH4; (c) N-bromosuccinimide (NBS), PPh3; (d ) Mg, H2C=C(Me)CHO; (e) KH, Bu3SnCH2I; ( f ) BuLi; (g) Ac2O, Py.580 tion of the skeleton of the alcohol 94 through organotin com- pound 95 followed by acetylation afforded ()-72 in 32% total yield with the stereoselectivity exceeding 95%. The acetate (3Z,6R)-72 was also synthesised 42, 43 using the Still rearrangement. The required alcohol (5R)-94 was prepared 42 from (+)-camphor. An efficient enantiospecific procedure devel- oped previously 44 ± 46 for the cleavage of 9,10-dibromocamphor 96 yielding bromoacid 97 has been employed.BrBr Br O O c a, b d H HO2CCH2 96 97 O Br Br e, f g, h H H HO(CH2)2 HO(CH2)2 98 OSiMe2But MeO2C d, i, j, k 99 OH i 100 l CHO OH (5R)-94 (3R)-73 (a) Br2, ClSO3H; (b) Zn, AcOH; (c) KOH, DMSO, H2O; (d ) LiAlH4; (e) O3, MeOH; ( f) Me2S; (g) ButMe2SiCl, 4-Me2NC5H4N; (h) NaOMe, MeOH; (i ) PDC, CH2Cl2; (j) H2C=PPh3; (k) Bu4NF; (l) H2C=C(Me)MgBr. Substituted cyclopentanone 98, which was synthesised from compound 97 according to standard procedures, is (like com- pound 96) a b-bromo ketone and underwent smooth ring-opening following protection of the hydroxy group to form acyclic ester 99. The latter was transformed into dienol 100 and then into the aldehyde (3R)-73 according to standard procedures.The reaction of the aldehyde (3R)-73 with isopropenylmagnesium bromide afforded the alcohol (5R)-94 in 8% total yield. Since the con- version of the alcohol ()-94 to the acetate ()-72 had been already performed by Still and Mitra,41 the above-described procedure for the preparation of (5R)-94 can be formally consid- ered as the first total stereoselective synthesis of (3Z,6R)-72. Oppolzer and Stevenson 43 synthesised the acetate (3Z,6R)-72 also with the use of (+)-camphor, but this was used as a chiral matrix rather than as a synthon. This approach was based on the known fact 47 that alkenyl cuprates stabilised by Bu3P add with high stereoselectivity to enoates of the type 101 prepared from camphor.In this case, the induced chirality is completely retained after removal of the matrix. O O 7 steps O O R H 101 ButCH2 N Ya Grigorieva, P G Tsiklauri O d a, b, c O H2C=C(Me)Li O R 102 ButCH2 e, f, g h, i, j CO2H OH 103 (5R)-94 OAc (3Z,6R)-72 R=(CH2)2CH=CH2; (a) Bu3P, CuI,770 8C; (b) BF3 . Et2O; (c) 101; (d ) NaOH, EtOH, H2O; (e) LiAlH4; ( f ) (COCl)2, DMSO, Et3N; (g) H2C=C(Me)Li; (h) KH, Bu3SnCH2I; (i ) BuLi; ( j) Ac2O, Py. Thus enoate 101 [prepared from (+)-camphor in seven steps] added the complex of diisopropenyl cuprate with Bu3P to form ester 102 with an enantioselectivity of *95%. Saponification of the ester 102 afforded chiral acid 103. This acid was converted into the alcohol (5R)-94 according to standard procedures.The latter was transformed into (3Z,6R)-72 using the Still rearrangement in 20% total yield. The synthesis of the triene 72 still attracts interest. This pheromone serves as a test substance for examining every new method proposed for the construction of (Z)-trisubstituted C=C bonds. Thus the compound 72 was chosen 48 as one of the subjects for the comparison of the reactivities of allylic organoiron nucleo- philes with those of allylsilanes and allylstannanes. The synthesis of ()-72 involved the conversion of the commercially available diene 104 into a mixture (*1 : 1) of organoiron compounds (2Z)- and (2E)-106 via allylic chloride 105. The second synthon required for the construction of the triene 72, viz., iodide 107 (a *1 : 1 mixture of the isomers), was prepared from 4-methyl-5,6-dihydro- 2H-pyran 108 according to standard procedures.Cl a b 105 104 c Fp ()-72 106 d, e ICH2 OAc 107 O 108 Fp=C5H5Fe(CO)2; (a) Ca(OCl)2; (b) NaFp; (c) 107, MeNO2; (d ) AcCl, K[PtCl3(C2H4)]; (e) NaI, Me2CO. In spite of the geometric inhomogeneity of the iodide 107, its condensation with organoiron derivatives 106 at *20 8C afforded ()-72 (the yield was 42%) containing <1% of an admixture of the (3E)-isomer. The selectivity of the reaction was attributed to the high reactivity of the (3Z)-iodide 107 due to the anchimeric assistance to ionisation provided by the acetoxy group. This scheme is attractive because it involves a small number of steps and does not require the introduction and removal of protective groups.Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes The syntheses of the aldehyde (3R)-73 and the alcohol (3R)- 100,49 which are the key intermediates in the synthesis of (3Z,6R)- 72, were carried out with the use of a previously discovered 1,4- addition of complex organocopper compounds of the type 109 to acetals 110 prepared from a,b-enals and (R,R)-butane-2,3-diol. Dienol ethers 111 prepared by the highly regio- and diastereose- lective reaction were subjected to hydrolysis to form aldehydes 112 in quantitative yields.O R1 R2 R4 . BF3 . LiHal . Bu3P+ Cu R3 109 110 O R4 R1 R1 R2 HO R2 R4 R3 O R3 CHO 112 111 This scheme was tried out in the synthesis of the aldehyde (3R)-73 starting from propargylaldehyde diethyl acetal (the yield was 50%, ee 85%).The aldehyde (3R)-73 was reduced to the alcohol (3R)-100 in quantitative yield. OEt a b HC CCH(OEt)2 OEt O c, d O O e AcO OH f CHO (3R)-100 (3R)-73 (a) [H2C=CH(CH2)2]2CuLi; (b) (2R,3R)-HOCH(Me)CH(Me)OH, TsOH; (c) H2C=C(Me)Cu . LiBr . BF3 . PBu3; (d) Ac2O, Py; (e) HCO2H; ( f ) NaBH4. The aldehyde (3Z,6R)-72 was also synthesised 50 using a highly stereoselective procedure developed by one of the authors 51 ± 53 of the present review for the construction of (Z)-tri- substituted alkenes based on the facts that (E)-isomers of a,b- disubstituted acroleins are thermodynamically favourable and that they are stereospecifically converted into (Z)-trisubstituted alkenes.According to this methodology, Baudouy and Prince 50 chose the aldehyde (3R)-73 and a-(triethylsilyl)imine 113 as synthons for the construction of (3Z,6R)-72. The aldehyde (3R)-73 was pre- pared from limonene according to a scheme reported previously.37 However, Baudouy and Prince 50 developed a direct procedure for the conversion of the hydroxy acetal phosphate 84 into diene acetal 114, which made it possible to reduce the scheme by two steps. The aldehyde (3R)-73 was formed upon hydrolysis of the acetal 114 in quantitative yield. The second synthon, viz., silylated imine 113, was prepared in three standard steps starting from the monobenzyl ether of butane-1,4-diol.e a, b, c f d 84 83 82 CHO CH(OMe)2 114 (3R)-73 OHC OBn g HO OBn ButN113 SiEt3 115 116 (a) O3, MeOH; (b) (H2N)2CS; (c) HC(OMe)3, CeCl3 .6H2O; (d ) LDA, ClP(O)(OEt)2, THF; (e) Li/NH3, THF; ( f ) HClO4, THF, H2O; (g) (COCl)2, DMSO, Et3N; (h) ButNH2, 4 A molecular sieves; (i ) LDA, Et3SiCl; ( j) BusLi, (3R)-73; (k) H3O+; (l) C5H5N. HCl, CH2Cl2; (m) LiAlH4, Et2O; (n) C5H5N. SO3, THF; (o) LiAlH4, THF; (p) Ac2O, HClO4. Condensation of the imine 113 deprotonated under the action of BusLi with the aldehyde (3R)-73 afforded acrolein 115. The mixture formed initially was kept in a CH2Cl2 solution containing pyridinium chloride for 4 h after which the content of the (Z)-isomer was no higher than 1%.The carbonyl group in compound 115 was reduced to the methyl group in three standard steps. The replacement of the protective benzyl group in the resulting ether 116 by the acetate group gave rise to (3Z,6R)-72 in 30% total yield. The construction of the pheromone ()-72 54 based on a method for the generation of allyllithium compounds by reductive lithiation of readily available allyl phenyl sulfides with aromatic radical anions 55 ± 57 has proved the utility of this procedure. The regiocontrolled addition of unsymmetrical alkyllithium deriva- tives to carbonyl compounds (which had been developed by the authors previously) employed in this synthesis is of particular interest. Thus treatment of the allyllithium intermediate with Ti(OPr)4 allows one to perform the subsequent reaction with an aldehyde at the more substituted position of the allylic system, whereas treatment with CeCl3, on the contrary, makes it possible to perform this reaction at the less substituted position.7 a b SPh SPh 117 d, e 119 HO g OH 121 (a) BusLi; (b) H2C=CH(CH2)2Br; (c) lithium 4,40-di(tert-butyl)bi- phenylide (LDBB),778 8C, Ti(OPr)4, CH2O; (d ) CBr4, Ph3P; (e) 117; ( f ) LDBB,778 8C, CeCl3, CH2O; (g) Ac2O, Py. Thus alkylation of the anion 117 generated from isobutenyl phenyl sulfide under the action of BusLi with but-3-enyl bromide occurred exclusively at the a position to form sulfide 118. Reductive lithiation of the latter with lithium 4,40-di(tert-butyl)- 581 OBn h, i j, k, l OBnm OBn n, o O OH OBn p (3Z,6R)-72 c SPh 118 f PhS 120 ()-72582 biphenylide followed by in situ treatment of the reaction product with Ti(OPr)4 and the reaction of the resulting organotitanium compound with formaldehyde afforded alcohol 119, which was converted into allyl sulfide 120.The sulfide 120 was subjected to reductive lithiation. The successive treatment of the resulting alkyllithium intermediate with CeCl3 and formaldehyde gave rise to alcohol (3Z)-121 from which the acetate ()-72 was obtained in 22.4% total yield. In 1995, Serebryakov and coworkers 58 synthesised the acetate ()-72 using the highly selective 1,4-cis-hydrogenation of con- jugated dienes in the presence of carbonyl chromium complexes 59 for the construction of the (Z)-trisubstituted C=C bond. This method had been employed previously 60, 61 in the synthesis of natural compounds.The triene 122 required for this synthesis was prepared in five steps from ethyl acetoacetate. CO2Et CHO CO2Et e a b, c, d O O O O 123 CO2Et CO2Et c, g f O O O O 125 122 OH h, i ()-72 126 O (a) EtONa, EtOH, H2C=CH(CH2)2Br; (b) HO(CH2)2OH, TsOH; (c) LiAlH4; (d) PCC,CH2Cl2; (e) (EtO)2P(O)CH2CH(Me)=CHCO2Et (124), KOH, PhH, 18-crown-6 (18-C-6); ( f) H2, (PhCO2Me)Cr(CO)3, Me2CO, 120 8C; (g) H3O+; (h) Ph3P=CH2; (i) Ac2O, Na2CO3, 100 8C. Alkylation of ethyl acetoacetate with but-3-enyl bromide, the conversion of the resulting oxo ester (after protection of the carbonyl group) into hydroxy ketal and oxidation of the latter afforded aldehyde 123. Condensation of 123 with triethyl 3-methyl-4-phosphonocrotonate (124) under conditions of phase transfer catalysis according to a procedure developed previ- ously 62 gave rise to a 2 : 1 mixture of the trienes (2E)- and (2Z)-122. Hydrogenation of the conjugated diene system in the presence of a chromium complex afforded (3Z)-ester 125, which virtually did not contain an admixture of the (3E)-isomer.The ester 125 was converted into hydroxy ketone 126 by standard procedures. The reaction of the ketone 126 with methylidenetri- phenylphosphorane followed by acetylation of the resulting alcohol completed the synthesis of ()-72 (the total yield was 4.3%).In the synthesis of the compound 72, Cooke and Burman 63 did not develop a new approach to the construction of the (Z)-trisubstituted C=C bond but used the available (Z)-alkene fragment in chlorohydrin 127 (Z>98%), which had been described previously 64 and which was converted into bromide 128 in two steps. OTHP OH a, b Br Cl 128 127 O c CO2But d Cl ButO2CCH PPh3 7 + 131 130 PPh3Cl N Ya Grigorieva, P G Tsiklauri OTHP O g, h e, f CO2But 129 PPh3 CO2But O 130 PPh3 OH i, j ()-72 O 132 (a) DHP, HCl; (b) NaBr, N-methylpyrrolidone; (c) Cl(CH2)2COCl; (d ) MeONa; (e) H2C=CHCH2Li; ( f ) 128; (g) AcOH, 90 8C; (h) NaOH; (i) Ac2O, Py; ( j) Ph3P=CH2. In the study cited (Ref. 63), an ingenious method for the assembly of the molecule ()-72 was used.Phosphorane 129 was used as the key compound in this synthetic scheme. The com- pound 129 was generated from phosphonium salt 130, which was prepared by acylation of tert-butoxycarbonylmethylidenetriphe- nylphosphorane 131 with b-chloropropionyl chloride. Successive treatment of the phosphorane 129 with prop-2-enyllithium and the bromide 128 afforded the phosphorane 130 containing the major structural fragments of the pheromone ()-72. The phos- phorane 130 was converted into hydroxy ketone 132 by a procedure developed previously.65 Acetylation of the latter fol- lowed by the reaction of the resulting acetate with methylidene- triphenylphosphorane gave rise to ()-72 in 8.5% total yield.b. Sex pheromone of the white peach scale Subsequent to the pheromone of the California red scale 72, a structurally similar female-produced sex pheromone of the white peach scale Pseudaulascaspis pentagona, viz., propionate 133, was isolated.66 OCOEt 133 The structure of compound 133 was established by physico- chemical methods and the absolute configuration of the C(6) centre was determined based on the results of bioassays of both enantiomers, which have been synthesised starting from (+)- and (7)-limonenes according to a scheme analogous to that used in the synthesis of the acetate 72 (see Section IV.1.a). Only the (3Z,6R)-propionate 133 appeared to be the active attractant.67 Serebryakov and coworkers 58 performed a highly stereo- selective synthesis of the compound 133 according to a procedure analogous to that developed by them for the pheromone 72.c. Sex pheromone of the yellow scale Gieselmann et al.68 isolated the female-produced sex pheromone of the yellow scale Aonidiella citrina, a species morphologically very similar to A.auranti, and established its structure. Although the pheromone of the yellow scale (3R,5E)-134 is formally the (5E)-isomer, we consider this compound in the present review because the branches of its major hydrocarbon chain, like those in the pheromones of the California red scale and the white peach scale, are in the cis positions with respect to each other. Because of this, (3R,5E)-134 was synthesised according to procedures analo- gous to those used for the preparation of pheromones containing the (Z)-trisubstituted C=C bond.Based on the results of phys- icochemical studies,68 the structure (5E)-134 was assigned to this pheromone [with the accuracy of the absolute configuration of the C(3) centre]. This structure was confirmed by two independent syntheses. One of these syntheses 69 involved the Horner ± Em- mons reaction of aldehyde 135 with an anion generated from phosphonate 136.Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes a b OMOM OMOM HO O 135 c, d e OMOM AcO EtO2C 138 137 f, d OMOM 139 Pri Pri OAc + OAc (5E)-134 (5Z)-134 MOM is methoxymethyl; (a) PCC, AcONa, CH2Cl2; (b) (EtO)2P(O)CH(Pri)CO2Et (136), EtONa; (c) Bui2AlH; (d) Ac2O, Py; (e) (Me2C=CH)2CuLi; ( f) Cl3CCO2H. The resulting mixture (4 : 1) of the esters (Z)- and (E)-137 was converted into a mixture of acetates 138 through the correspond- ing alcohols. Treatment of the compounds 138 with diisobutenyl- lithium cuprate afforded a mixture of isomers 139, which was converted into a mixture of acetates (5E)-134 and (5Z)-134 in two standard steps also in a ratio of 1 : 4.Bioassays of the isomers separated by preparative GLC demonstrated that only the com- pound (5E)-134 exhibited the attractant activity. Suguro et al.70 reached the same conclusion when performing the stereoselective synthesis of the isomeric acetates 134 from the common precursor, viz., ester (5E)-140. This ester was prepared with a stereoselectivity of 92% by the in situ Claisen ± Johnson [3,3]-sigmatropic rearrangement of orthoester 141 (Scheme 1).Treatment of the ester 140 with MeMgI afforded alcohol 142, which was dehydrated to form a mixture of regioisomeric trienes 143. Hydroboration of this mixture under standard conditions gave rise to a mixture of the corresponding alcohols from which OC(Me)(OEt)2 c a, b Br EtO2C (5Z)-140 Pri 141 + Pri Pri D8-143 D9-143 l j, k (5Z)-140 THPO THPO Pri o, p MeO2C HO 145 Pri Pri (5E)-140 (a) Mg, H2C=CH(Pri)CHO; (b) MeC(OEt)3; (c) EtCO2H, 135 8C; (d ) MeMgI; (e) POCl3, Py; ( f ) 9-borabicyclo[3.3.1]nonane (9-BBN); (g) NaOH; (h) H2O2; (i) Ac2O, Py; ( j ) LiAlH4; (k) DHP, TsOH; (l ) MCPBA; (m) Ph2PLi, MeI; (n) H3O+; (o) PDC, DMF; (p) CH2N2.583 the (5Z)-isomer was isolated by chromatography. Acetylation of the latter afforded the acetate (5Z)-134 in*1% total yield. The ester (5E)-140 required for the preparation of the acetate (5E)-134 was synthesised via epoxide 144 by the Vedejs method.71 Successive treatment of the epoxide 144 with lithium diphenyl- phosphide and methyl iodide afforded the corresponding tetrahy- dropyranyl derivative, which was converted into alcohol 145. Oxidation of the hydroxy group in the alcohol 145 and methyl- ation of the resulting carboxylic acid gave rise to the ester (5E)- 140, which was converted into (5E)-134 in 0.5% total yield by an analogous procedure. Mori and coworkers 72 established the absolute configuration of the C(3) centre in the acetate 134 and performed the synthesis of (3S,5E)-134 exhibiting the high attractant activity.For this purpose, chiral monoester (R)-146, which was synthesised from (R)-citronellic acid, was used as the starting compound. The Hunsdiecker reaction of the monoester (R)-146 afforded bromide 147, which was transformed into acid 149 via selenide 148. The acid 149 was smoothly converted into aldehyde 150. The reaction of the aldehyde 150 with an organomagnesium compound pre- pared from bromide 151 gave rise to alcohol 152, which was subjected to the Still rearrangement. The alcohol 153 containing the trisubstituted alkene fragment was obtained with a stereo- selectivity of>95%. The reaction of an aldehyde prepared from 153 with isopropylidenetriphenylphosphorane afforded triene 154. Hydroboration ± oxidation of the terminal double bond in the triene 154 followed by acetylation of the resulting alcohol 155 completed the synthesis of (3S,5E)-134 (the total yield was 0.2%).Later, Mori and Kuwahara 73 performed a more efficient synthesis of (3S,5E)-134. According to this scheme, the trisubsti- tuted alkene fragment was also constructed by the Still rearrange- ment. The key intermediate, viz., benzyloxy alcohol 156, was prepared by the reaction of the organomagnesium derivative of the bromide 151 with aldehyde 157. The latter was synthesised by the seven-step transformation of methyl (R)-citronellate [(R)-158]. The conversion of the benzyloxy alcohol 156 according to Still giving rise to benzyloxy alcohol 159, oxidation of the latter and treatment of the resulting aldehyde with isopropylidenetriphenyl- phosphorane afforded benzyl ether 160.The target acetate (3S,5E)-134 was prepared from 160 in 7% total yield. Scheme 1 HO e d 142 Pri Pri f, g, h, i OAc Pri (5Z)-134 O n m O7 + THPO PPh2Me Pri 144 Pri d, e, f, g, h, i (5E)-134N Ya Grigorieva, P G Tsiklauri 584 CO2Me c, d CO2Me a, b Br HO2C 147 146 e 2-NO2C6H4 CO2H Se 148 g f CHO CH2OH CO2H 150 149 Julia and coworkers 75 constructed the chiral fragment in the acetate (5E)-134 by the reaction of an anion generated from sulfone 161 with (3R)-methylvalerolactone (162). When treated with tert-butyldimethylchlorosilane, the resulting mixture of hydroxy oxo sulfone 163 and hemiketal 164 was quantitatively converted into the silyl derivative of hydroxy oxo sulfone 163.Reduction of the carbonyl group in 163 followed by acetylation of a mixture of the resulting alcohols afforded a mixture of acetates 165. Elimination of AcOH from the acetates 165 resulted in sulfone 166 containing an admixture (<2%) of the (5Z)-isomer. OH Br j, k h, i l, m a SO2Ph 161 Pri Pri 152 151 Pri O b, c, d o, p, q g, n OH+ O 163 SO2Ph HO 164 PhSO2 HOH2C 154 Pri Pri 153 OAc e OSiMe2But 165 SO2Ph r (3S,5E )-134 OH 155 Pri f, g, h OSiMe2But 166 SO2Ph (a) Ag2O; (b) Br2; (c) 2-NO2C6H4SeCN, NaBH4; (d ) NaOH, D; (e) H2O2; ( f ) LiAlH4; (g) PDC, CH2Cl2; (h) Br2; (i ) 2,5-Me2C6H3OH, NaH; ( j ) Mg; (k) 150, THF; (l ) KH, Bu3SnCH2I; (m) BuLi; n) Ph3P=CMe2; (o) 9-BBN; (p) NaOH; (q) H2O2; (r) Ac2O, Py.CO2Me c, d CO2Me a, b + OAc OAc OH (R)-158 167 (3S,5R)-134 Pri e, f CH2OMOM (a) BuLi, (162); (b) LDA, Me2ButSiCl; (c) NaBH4; OH O O g, h, i CH2OMOM HOCH2 (d) Ac2O, dimethylaminopyridine (DMAP), Py; (e) NaOH; ( f) PriMgBr, FeCl3; (g) Bu4NF; (h) Ac2O, Py. OH j k, i, l OBn O 156 157 OBn m The reaction of the sulfone 166 with PriMgBr in the presence of FeCl3 afforded, like other analogous reactions studied previ- ously,78 a 1 : 1 mixture of products of replacement and elimination of the sulfonyl group. After removal of the silyl protection, acetylation and chromatographic separation of the resulting mixture of the acetates (3S,5E)-134 and 167, (3S,5E)-134 was isolated in 17.2% total yield.OBn HOCH2 159 n, o (3S,5E)-134 OBn 160 Millar 76 constructed the trisubstituted alkene fragment in the compound (5E)-134 based on the above-mentioned highly stereo- selective addition of cuprates to alkynes. Thus successive treat- ment of 3-methylbut-1-yne with (PhMe2Si)2CuLi (Ref. 79) and 3-methylbut-2-enyl bromide gave rise to vinylsilane containing 16% of the SN20-addition product. This mixture was converted into the corresponding iodides. Flash chromatography of the latter afforded iodide 168 in 39% total yield. This iodide was transformed into the target acetate ()-(5E)-134 by cross-cou- pling with an organozinc reagent synthesised from iodide 169.The reaction product was converted into the target acetate ()-(5E)- 134 (the total yield was*16%). (a) (PhSe)2, H2O2; (b) But2O2; (c) LiAlH4; (d ) NaH, MeOCH2Cl (MOMCl); (e) O3, NaHCO3, MeOH; ( f ) LiAlH4; (g) NaH, BnCl; (h) TsOH, MeOH; (i ) PDC, CH2Cl2; ( j) H2C=C(Pri)MgBr; (k) KH, Bun3 SnCH2I; (l ) BuLi; (m) Ph3P=C(Me)2; (n) Li/NH3; (o) Ac2O, Py. Pri Pri a b, c PriC CH I Cu Simultaneously, Mori and Kuwahara 73 prepared (3R,5E)-134 in 10% total yield. Bioassays demonstrated that only the (3S)- enantiomer exhibited the attractant activity.73, 74 SiMe2Ph 168 O e, f d Cl CO2Et O Although Mori and coworkers 70, 72, 73 succeeded in perform- ing stereoselective synthesis of (3S,5E)-134, the synthetic schemes involved many steps and, as a result, the target compound was prepared in low yield.Later, more efficient syntheses of the acetate (5E)-134 were performed.75 ± 77Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes Cl g, h I 169 ()-(5E )-134 (a) 2 equiv. of PhMe2SiLi, CuCN, THF; (b) BrCH2CH=CMe2; (c) I2; (d ) HCl, EtOH; (e) LiAlH4; ( f ) I2 . Ph3P, imidazole; (g) Zn, THF; (h) 168, (Ph3P)4Pd; (i ) NaOAc, HMPA. The pheromone (3S,5E)-134 was also synthesised 77 starting from (S)-citronellyl pivalate (170). The highly stereoselective construction of the trisubstituted alkene fragment took advantage of thermodynamic predominance of the (E)-isomers of a,b- disubstituted acroleins,53 i.e., it is based on the approach used previously 50 in the synthesis of the pheromone of the California red scale 72.The first synthon, viz., aldehyde 171, was prepared from the ester 170 by oxidation of its phenylselenenyl derivative with di-tert-butyl peroxide. c a, b OR OR OH 170 172 d e, f HO O174 j, k OR 175 O n, o OH R=COBut; (a) (PhSe)2, H2O2; (b) But2O2; (c) O3,778 8C; (d ) EtOCH=CH2, H3PO4, 150 8C; (e) ButNH2, 4 A molecular sieves; ( f ) LDA, Et3SiCl; (g) BusLi, 171; (h) CF3CO2H, H2O; (i) C5H5N.HCl; ( j ) MeMgI; (k) (COCl)2, DMSO, Et3N; (l) Ph3P=CH2; (m) Bui2AlH; (n) H2, (Ph3P)3RhCl; (o) Ac2O, Et3N. Subsequent ozonolysis of the resulting tertiary alcohol 172 afforded the aldehyde 171.The second synthon, viz., silylated imine 173, was prepared by condensation of 2-methylbut-3-en-2- ol with ethyl vinyl ether followed by successive treatment of the resulting aldehyde 174 with ButNH2 and Et3SiCl. Condensation of the imine 173 with the aldehyde 171 under conditions used previously 50 in the synthesis of the pheromone 72 yielded acrolein 175, which virtually did not contain the (Z)-isomer. The carbonyl group in the aldehyde 175 was converted into the isopropyl fragment in six steps. The target acetate (3S,5E)-134 was prepared in 25% total yield. This is the best result obtained in the synthesis of the sex pheromone of the yellow scale. d. Sex pheromone of the San Jose scale Almost simultaneously with the pheromone of the yellow scale, the female-produced sex pheromone of the San Jose scale Quad- raspidiotus perniciosus, which is a widespread polyphagous pest of fruit and decorative trees, was isolated.80 Two major components Pri i Cl OR O 171 g, h, i Et3Si NBut 173 l, m OR O OAc (3S,5E)-134 585 of this pheromone, viz., esters 176 and (Z)-177, were identified by physicochemical methods.More recently,81 the ester (E)-177 was also found. It was demonstrated that the pheromone contains these components in a ratio of 47.55 : 47.75 : 4.70.81 Each of these components exhibits approximately the same biological activity. OCOEt CH2OCOEt 176 (Z)-177 CH2OCOEt (E)-177 The ester (Z)-177 was synthesised for the first time 82 with the use of the above-mentioned highly stereoselective addition of dialkyl cuprates to alkynes in the key step of the construction of the (Z)-trisubstituted alkene fragment.a b, c, d MgBr Br 178 e, f, g OCOEt CO2H 179 (Z)-177 (a) Mg; (b) CuBr .Me2S; (c) HC CMe; (d) CO2, (EtO)3P, HMPA; (e) CH2N2; ( f) Bui2AlH; (g) (EtCO)2O, Py. The reaction of cuprate obtained from the magnesium deriv- ative of bromide 178 with propyne proceeded as cis-addition. Treatment of the reaction mixture with CO2 afforded (2Z)-acid 179. The latter was converted into the target pheromone (Z)-177 in 29% total yield using standard procedures. Alderdice et al.83 synthesised the ester (Z)-177 using the stereoselective conversion of enol phosphates of b-oxo esters into (Z)-trisubstituted a,b-unsaturated esters.84 7 7 a b MeCOCH2CO2Me CH2COCHCO2Me 180 O OP(O)(OEt)2 d c CO2Me 182 (E )-183 CO2Me e, f CH2OCOEt CO2Me (Z)-177 (Z)-184 OP(O)(OEt)2 g d, e, f (E)-177 182 CO2Me (Z)-183 (a) NaH, BuLi; (b) H2C=CH(Me)CH2CH2OTs (181); (c) (EtO)2P(O)Cl, Et3N; (d ) MeMgCl, MeCu; (e) Bui2AlH; ( f ) (EtCO)2O, Py; (g) (EtO)2P(O)Cl, NaH.Thus dianion 180 generated from methyl acetoacetate under- went smooth alkylation under the action of tosylate 181 to form b-oxo acid ester 182. The latter was converted into enol phosphate (E)-183. Treatment of (E)-183 with MeMgCl in the presence of a catalytic amount of MeCu afforded ester (Z)-184, which was transformed into the target pheromone (Z)-177 through the corresponding alcohol in *40% total yield.Deprotonation of the oxo ester 182 with NaH instead of Et3N afforded the (Z)-enol phosphate 183. The latter was converted into the component of the pheromone of the San Jose scale, viz., the propionate (E)-177, in three steps. Nerol is a convenient starting compound for the synthesis of the ester (Z)-177. However, isomerisation of the C(6)=C(7) double bond appeared to be a difficult problem. Only after several586 unsuccessful attempts,85 an efficient reagent for this isomerisa- tion, viz., tert-butyl hypochlorite, was found.86, 87 The propionate (Z)-177 was synthesised according to this scheme in three steps in 40% total yield. a b CH2OCOEt CH2OH c (Z)-177 CH2OCOEt Cl (a) (EtCO)2O, Py; (b) ButOCl, SiO2; (c) NaBH4, LiI.Moiseenkov et al.88 synthesised the propionate (Z)-177 start- ing from isoprenoid (Z)-hydroxy sulfonamide 185.89 The syn- thetic scheme involved alkylation of the sulfonamide 185 with 3-methylbut-3-enyl iodide 186 to form sulfonamide 187 and reductive desulfation of the latter in the presence of dibenzo-18- crown-6 (DB-18-C-6). The reaction proceeded virtually without migration of the C=C bond to form alcohol (Z)-188, which was converted into the propionate (Z)-177 in 51% total yield. CH2OH c CH2OH a, b O SO2N O SO2N 187 185 d (Z)-177 CH2OH (Z)-188 (a) BuLi, THF; (b) H2C=C(Me)(CH2)2I (186); (c) Na/NH3, C6H14, DB-18-C-6; (d ) (EtCO)2O, Py. A mixture of propionates (Z)- and (E)-177 was synthesised by Chinese chemists.90 ± 92 The synthesis involved the Wittig reaction of ketone 189 with ylide generated from phosphonium salt 190 followed by the conversion of the resulting mixture of alcohols 188 into a mixture of propionates (Z)- and (E)-177.The key inter- mediate, viz., the ketone 189, was prepared in two ways. The first route 90, 91 involved oxidation of alcohol 191 formed in the reaction of diisopentenyl cuprate with propylene oxide. The second route 92 involved the conversion of heptane-2,6-dione monoketal into the ketone 189 via ketal 192. c a, b CuLi Br 2 OH O d 189 191 O O CH2 g e, f h O O 189 192 i (Z)-177+(E)-177 CH2OH 188Me ; (d ) PCC, CH2Cl2; e) BnO(CH2)3OHon a poly- (a) Li; (b) CuI; (c) O meric support; ( f) Ph3P=CH2; (g) H3O+; (h) (Ph3P+CH2CH2OH)Cl7 (190), B7; (i ) (EtCO)2O, Py.N Ya Grigorieva, P G Tsiklauri 2. Synthesis of the male-produced sex pheromone of the southern green stinkbug Nezara viridula The southern green stinkbug Nezara viridula is a pest of cotton, citrus, cereal and vegetable crops, which is widely distributed in tropical and subtropical regions. The first data on the structure of the male-produced sex pheromone of N.viridula were published in 1971.93 However, its major component was isolated only in 1987.94 Based on the results of physicochemical studies, the structure of epoxybisabolene (193) was assigned to this compo- nent. With the aim of establishing the stereochemistry of the natural pheromone, eight of its isomers were synthesised 94 start- ing from (R)- and (S)-limonenes using the modified Julia proce- dure 75 in the step of construction of the trisubstituted alkene fragment.95 The approach used in the cited study 94 is exemplified in the syntheses of the (Z)- and (E)-isomers of (1S,2R,4S)-193 considered below.The necessary key intermediate, viz., epoxy ketone 194, was prepared in the individual state with the use of a reaction sequence involving epoxidation of (S)-limonene and conversion of the resulting mixture of a- and b-epoxides into a mixture of amino alcohols 195 and 196. The latter compounds were converted into the corresponding tosylates and separated by chromatography. OH O NMe2 OH NMe2 c, d, e a, b f, g + 196 195 (1S,2R,4S)-197 O O O 1 2 h, i, j + 3 4 10 20 O (1S,2R,4S)-194 (Z,1S,2R,4S)-193 (E,1S,2R,4S)-193 O g, h, i c, d, e f 195 (1S,2R,4R)-197 O (1S,2S,4R)-194 O O + (E,1S,2S,4R)-193 (Z,1S,2S,4R)-193 (a) MCPBA; (b) Me2NH, H2O; (c) TsCl, Py; (d ) MeI; (e) KOH; ( f) O3, CH2Cl2; (g) Ph3P; (h) Me2C=CH(CH2)2SO2Ph (198), BuLi; (i ) PhCOCl, Py; ( j ) Na/Hg, THF, MeOH.Amino tosylate corresponding to the amino alcohol 196 was transformed into the quaternary ammonium salt and then into individual epoxide (1S,2R,4S)-197. Ozonation of the double bond in (1S,2R,4S)-197 followed by the reaction of the resulting epoxy ketone (1S,2R,4S)-194 with deprotonated sulfone 198 afforded b-hydroxy sulfone. The latter was converted into benzoate.Subequent reductive elimination gave rise to a 1 : 1 mixture of the (Z)- and (E)-isomers of epoxybisabolene (1S,2R,4S)-193. The individual isomers were isolated by HPLC. The (Z)- and (E)-iso-Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes mers of epoxybisabolene (1S,2S,4R)-193 were prepared analo- gously from epoxy ketone (1S,3S,4R)-194. Four other isomeric epoxybisabolenes were prepared starting from (R)-limonene. Based on comparison of the physicochemical characteristics of the isomers synthesised with those of the male-produced phero- mone of stinkbugs of the Brazilian population, it was reliably established that the pheromone is epoxybisabolene (Z,1S,3R,4S)- 193. Bioassyas demonstrated that only this isomer was attractive to N.viridula females. Later, it was shown 96 ± 99 that other isomers of epoxybisabo- lene 193 with the (Z)-configuration of the C(5)=C(6) double bond as well as (Z)-bisabolene are sex pheromones of stinkbugs from other regions.The ratio between the components in the phero- mone mixture varies depending on the habitat. In this connection, synthesis of racemic epoxybisabolene (Z)-193 starting from race- mic limonene (patented by Mori et al.100) is of interest. The individual epoxybisabolene (Z)-193 was isolated from a mixture with the (E)-isomer by chromatography. O O O a b + O (Z)-193 (E)-193 (a) MCPBA; (b) Ph3P=CHCH2CH=CMe2. The major component of all pheromone mixtures of the southern green stinkbug, viz., epoxybisabolene (Z,1S,2R,4S)- 193, was synthesised 101 from commercially available perillyl alcohol 198.The key step of this scheme involved the Horner ± Emmons reaction of the ketone 194 with phosphine oxide 199. Epoxidation of perillyl alcohol 198 according to Sharpless afforded hydroxy epoxide 200. The hydroxymethyl group of 200 was reduced to the methyl group through the corresponding mesylate. The resulting epoxide 197 was converted into the corresponding epoxy ketone 194 (the total yield was 47%). The reaction of the latter with an anion generated from the phosphine oxide 199 gave rise to a mixture of erythro- and threo-phosphine oxides 201. The former was isolated by chromatography from a mixture with (E)-epoxybisabolene, which was formed due to decomposition of the unstable threo-201, and was converted into the epoxybisabolene (Z,1S,2R,4S)-193 (the total yield was 18%).OH O O OH O e a d b, c O 194 198 200 (1S,2R,4S)-197 O O f (Z,1S,2R,4S)-193 + HO H P(O)Ph2 HO Ph2(O)P erythro-201 Hthreo-201 (a) (D)-(7)-tartaric acid, Ti(OPri)4, ButO2H; (b) MeSO2Cl, Py; (c) LiAlH4; (d ) KMnO4, DB-18-C-6, CH2Cl2; (e) Me2C=CH(CH2)2P(O)Ph2 (199), BuLi; ( f ) NaH, DMF. Efficient syntheses of the isomers (Z,1S,2R,4S)-, (Z,1S,2S,4R)-, (Z,1R,2S,4R)- and (Z,1R,2R,4S)-193 were carried 587 out102 starting from the commercially available mixture of a- and b-epoxides of (S)- and (R)-limonenes. The (Z)-trisubstituted alkene fragment was constructed using the highly stereoselective cis-addition of dialkyllithium cuprates to acetylenecarboxylic esters.The scheme of the synthesis is exemplified below in the preparation of the isomer (Z,1S,2R,4S)-193. O O O e a, b, c d C C O CH 194 CCO2Me 202 O O f, g h (Z,1S,2R,4S)-193 CO2Me Br 203 (a) LDA; (b) (EtO)2P(O)Cl; (c) 2.5 equiv. of LDA; (d ) BuLi, ClCO2Me; (e) Me2CuLi; ( f) Bui2AlH; (g) CBr4, PPh3; (h) ButLi, Me2C=CHBr. Substituted acetylenecarboxylic ester 202 was prepared from the epoxy ketone (1S,2R,4S)-194 in four steps. The addition of dimethyllithium cuprate to the ester 202 afforded ester (2Z)-203. The latter was transformed into (Z,1S,2R,4S)-193 in 24% total yield according to standard procedures.Analogously, the isomer (Z,1S,2S,4R)-193 was prepared from the ketone (1S,2S,4R)-194 and the isomers (Z,1R,2R,4R)- and (Z,1R,2R,4S)-193 were pre- pared from a- and b-epoxides of (R)-limonene, respectively. (Z,1S,2S,4R)-Epoxybisabolene was also synthesised 103 start- ing from acid 204, which was prepared by the asymmetric Diels ± Alder reaction.104 Bromolactonisation of the acid 204 gave rise to a mixture of bromo lactones 205 and 206. Treatment of this mixture with 4-methyl-1-phenylthiopent-3-enyllithium (207) afforded phenylthio ketone 208 in high yield. The reaction of the compound 208 with MeLi gave rise to a mixture of threo- and erythro-hydroxy sulfides (209) separated by flash chromatog- raphy. Treatment of threo-209 with P2I4 in the presence of Et3N produced the target epoxybisabolene 193 in a total yield of 18% with respect to the acid 204.O Br Br c a b O O O + O 206 205 O CO2H SPh 208 204 O O d + H HO H HO threo-209 SPh SPh erythro-209 (Z,1S,2S,4R)-193 (a) NBS, Na2CO3; (b) Me2C=CHCH2CH(Li)SPh (207); (c) MeLi; (d) P2I4, Et3N, CH2Cl2. 3. Synthesis of the sex pheromones of the dry bean beetles Callosobruchus analis and C.maculatus The female-produced sex pheromone of the dry bean beetle Callosobruchus analis was isolated in 1991.105 Based on the data588 of GLC-mass spectrometry, the structure of 3-methylheptenoic acid containing a double bond at position 2 or 3 was assigned to this pheromone. Comparison of the physicochemical properties of the pheromone isolated with those of authentic 3-methylheptenoic acids containing a double bond at different positions (which were synthesised by the Reformatsky, Wittig or Horner ± Emmons reactions) demonstrated that the pheromone of C.analis females is 3-methylhept-(2Z)-enoic acid (210).Bioassays of isomeric 3-methylheptenoic acids demonstrated that 3-methylhept-(3Z)- enoic acid is also the attractant for males of the dry bean beetle C.analis. This acid was identified as the major component of the sex pheromone of the dry bean beetle C.maculatus.105, 106 A highly stereoselective synthesis of 3-methylhept-(2Z)-enoic acid (210) was carried out according to two procedure. The first procedure107 was based on the fact that the (E)-isomers of a,b- disubstituted acroleins are thermodynamically favourable.In this scheme, the key intermediate was the (E)-isomer of substituted acrolein 211, which was formed with a stereoselectivity of >95% upon condensation of the trimethylsilyl derivative of N-tert-butylhexanimine (212) with glycolaldehyde benzyl ether (213).107 Stereospecific reduction of the carbonyl group in the aldehyde 211 to the hydroxymethyl group and then to the methyl group via sulfate afforded benzyl ether 214. Deprotection fol- lowed by oxidation of the resulting alcohol 215 gave rise to the acid 210 (the total yield was 12%). b, c a NBut BunCH2CH BunCH2CHO OHC f b, d, e BunCH(SiMe3)CH NBut Bun CH2OBn 212 211 HOCH2 i g, h Bun Bun CH2OBn CH2OBn 214 j, k Bun Bun CO2H CH2OH 210 215 (a) ButNH2; (b) LDA; (c) Me3SiCl; (d ) OHCCH2OBn (213); (e) H3O+; ( f ) NaBH4; (g) Py.SO3; (h) LiAlH4; (i ) Li/NH3; (j ) MnO2; (k) Ag2O.The second procedure 108 was based on our version of the highly stereoselective Peterson olefination 109, 110 of methyl ketones. The method involves the introduction of a bulky phenyl- thio substituent at the C(3) atom. This substituent can be readily removed from the reaction product. CO2R O a Prn CO2R+ Prn SPh Prn SPh (E)-217a,b SPh (Z)-217a,b 216 b c (Z)-217b Prn Bun CH2OH CH2OH SPh 218 215 R=Me (a), But (b); (a) Me3SiCH2CO2R, LDA; (b) [AlH3]; (c) Na/NH3, C6H12, DB-18-C-6. Olefination of a-phenylthio ketone 216, which was prepared according to Warren's method,111 with alkyl trimethylsilylace- tates afforded a mixture of 4-phenylthio esters (2Z)- and (2E)- 217a,b (the ratios of the isomers were 92 : 8 and 85 : 15 for 217a and 217b, respectively). Reduction of the (Z)-ester 217b isolated by chromatography with AlH3 in situ gave rise to 4-phenylthio alcohol (2Z)-218 desulfation of which afforded the alcohol 215 in 10.3% total yield.It should be noted that the yield of the alcohol 215 prepared according to the first procedure reached 26%. N Ya Grigorieva, P G Tsiklauri 3-Methylhept-(3Z)-enoic acid 219 was synthesised by Sere- bryakov and coworkers.112, 113 The key step of this scheme involved 1,4-cis-hydrogenation of conjugated dienes proposed by Frankel et al.59 a EtCHO+(EtO)2P(O)CH2C(Me) CHCO2Et c b CO2Et Et Prn CO2Et 221 220 Prn CH2CO2H 219 (a) KOH, 18-C-6, PhH; (b) H2, Cr(CO)6, C6H14; (c) KOH, HCl.Condensation of propanal with triethyl 3-methyl-4-phospho- nocrotonate under conditions of phase transfer catalysis afforded a mixture of esters (2E,4E)- and (2Z,4E)-220. Hydrogenation of this mixture over Cr(CO)6 gave rise to ester (3Z)-221 with 99% purity. Saponification of the ester 221 under mild conditions produced the acid 219 in 11.4% total yield. * * * In conclusion, note that the synthesis of insect pheromones is a good `proving ground' for new methods of organic chemistry. All the aforesaid refer in full measure both to stereoselective proce- dures for the construction of (Z)-trisubstituted C=C bonds and to procedures for the construction of chiral synthons required for the preparation of pheromones of the group under consideration. References 1.Y Kuwahara, S Ishii, H Fukami Experientia 31 1115 (1975) 2. Y Kuwahara, H Fukami, S Ishii, K Matsumoto, Y Wada Jpn. J. Sanit. Zool. 30 309 (1979); Chem. Abstr. 92 194 982 (1980) 3. Y Kuwahara, L Sakuma Agric. Biol. Chem. 461855 (1982) 4. Y Kuwahara, T Koshii,M Okamoto, K Matsumoto, T Sazuki Eisei Dobatsu 42 29 (1991); Chem. Abstr. 115 203 722 (1991) 5. J W S Bradshaw, R Baker, P E Howse Nature (London) 258 230 (1975) 6. D Basavaian, R S Hyma Tetrahedron 52 1253 (1996) 7. BRD P. 2 155 113; Chem. Abstr. 77 34 174 (1972) 8. N Ya Grigorieva, I M Avrutov, O A Pinsker, O N Yudina, A I Lutsenko, A M Moiseenkov Izv.Akad. Nauk SSSR, Ser. Khim. 1824 (1985) a 9. N Ya Grigorieva, E P Prokof'ev, A V Semenovskii Dokl. Akad. Nauk SSSR 245 366 (1979) b 10. J T Baker, J D Blake, J K MacLeod, D A Ironside, I C Johnson Aust. J. Chem. 25 393 (1972) 11. E O Wilson Science 129 643 (1959) 12. R K Van der Meer, F D Williams, C S Lofgren Tetrahedron Lett. 22 1651 (1981) 13. A B Attygalle, E D Morgan J. Chem. Soc., Perkin Trans. 1 949 (1982) 14. H J Williams, M R Strand, S B Vinson Tetrahedron 37 2763 (1981) 15. F M Alvarez, R K van der Meer, C S Lofgren Tetrahedron 43 2897 (1987) 16. A B Attygalle, S Steghaus-Kovac, V U Ahmad, U Maschwitz, O Vostrovsky, H J Bestmann Naturwissenschaften 78 90 (1991) 17. F J Ritter, I E M Bruggemann-Rotgans, P E J Verwiel, C J Persoons, E Talman Tetrahedron Lett.2617 (1977) 18. M Kobayashi, T Koyama, K Ogura, S Seto, F J Ritter, I E M Bruggemann-Rotgans J. Am. Chem. Soc. 102 6602 (1980) 19. T Koyama, K Ogura, S Seto J. Am. Chem. Soc. 99 1999 (1977) 20. T Koyama, A Saito, K Ogura, S Seto J. Am. Chem. Soc. 102 3614 (1980) 21. K H Dahm, B M Trost, H Roller J. Am. Chem. Soc. 89 5292 (1967) 22. K Mori, H Ueda Tetrahedron Lett. 22 461 (1981) 23. K Mori, H Ueda Tetrahedron 38 1227 (1982) 24. R Baker, D C Billington, N Ekanayake J. Chem. Soc., Chem. Commun. 1234 (1981)Synthesis of insect pheromones belonging to the group of (Z)-trisubstituted alkenes 25. E J Corey, J A Katzenellenbogen J. Am. Chem. Soc. 91 1851 (1969) 26. A Marfat, P R McGuirk, P Helquist J.Org. Chem. 44 3888 (1979) 27. K Mori, N Murata Liebigs Ann. Chem. 2089 (1995) 28. P Wipf, S Lim Angew. Chem., Int. Ed. Engl. 32 1068 (1993) 29. D W Knight, B Ojhara Tetrahedron Lett. 22 5101 (1981) 30. L Poppe, L Novak, P Kolonits,A Bata,C Szantay Tetrahedron Lett. 27 5769 (1986) 31. L Poppe, L Novak, P Kolonits, A Bata, C Szantay Tetrahedron 44 1477 (1988) 32. S Tanaka, S Yamamoto, H Nozaki, K B Sharpless, R C Michaelson, J D Cutting J. Am. Chem. Soc. 96 5254 (1974) 33. A N Kasatkin, T Yu Romanova, I P Podlipchuk, G A Tolstikov Khim. Prir. Soedin. 459 (1993) c 34. WL Roelofs,MJ Gieselmann, AMCarde, H Tashiro, D S Moreno, C A Henrick, R J Anderson Nature (London) 267 698 (1971) 35. WL Roelofs,MJ Gieselmann, AMCarde, H Tashiro, D S Moreno, C A Henrick, R J Anderson J.Chem. Ecol. 4 211 (1978) 36. M J Gieselmann, C A Henrick, R J Anderson, D S Moreno, W L Roelofs J. Insect Physiol. 26 179 (1980) 37. D Becker, Y Sahali Tetrahedron 44 4541 (1988) 38. E Negishi, A O King, W L I Klimo J. Org. Chem. 45 2526 (1980) 39. D Caine, E Crews Tetrahedron Lett. 25 5359 (1984) 40. D Caine, E Crews, J Salvino Tetrahedron Lett. 24 2083 (1983) 41. W S Still, A Mitra J. Am. Chem. Soc. 100 1927 (1978) 42. J H Hutchinson, T Money Can. J. Chem. 63 3182 (1985) 43. W Oppolzer, T Stevenson Tetrahedron Lett. 27 1139 (1986) 44. WMDadson,MLam, T Money, S E Piper Can. J. Chem. 61 343 (1983) 45. J H Hutchinson, T Money, S E Piper J. Chem. Soc., Chem. Commun. 455 (1984) 46.J H Hutchinson, T Money Tetrahedron Lett. 26 1819 (1985) 47. W Oppolzer, P Dudfied, T Stevenson, T Godel Helv. Chim. Acta 68 212 (1985) 48. J Celebuski, M Rosenblum Tetrahedron 41 5741 (1985) 49. P Mangeney, A Alexakis, J F Normant Tetrahedron Lett. 28 2363 (1987) 50. R Baudouy, Ph Prince Tetrahedron 45 2067 (1989) 51. N Ya Grigorieva, O A Pinsker Usp. Khim. 63 177 (1994) [Russ. Chem. Rev. 63 169 (1994)] 52. N Ya Grigorieva, I M Avrutov, A V Semenovsky Tetrahedron Lett. 24 5531 (1983) 53. N Ya Grigorieva, V V Veselovskii, A M Moiseenkov Khim. Farm. Zh. 21 845 (1987) d 54. D McCullough, M Bhupathy, E Piccolino, T Cohen Tetrahedron 47 9727 (1991) 55. T Cohen, B S Guo Tetrahedron 42 2803 (1986) 56. T Cohen,M Bhupathy Acc.Chem. Res. 22 182 (1989) 57. B S Guo,W Doubleday, T Cohen J. Am. Chem. Soc. 109 4710 (1987) 58. A A Vasil'ev, A L Vlasyuk, G V Kryshtal', E P Serebryakov Izv. Akad. Nauk, Ser. Khim. 2026 (1995) a 59. E N Frankel, E Selke, C A Glass J. Am. Chem. Soc. 90 2446 (1968) 60. M Sodeoka, M Shibasaki Synthesis 643 (1993) 61. A A Vasil'ev, E P Serebryakov Mendeleev Commun. 4 (1994) 62. G V Kryshtal', E P Serebryakov, L M Suslova, L A Yanovskaya, Izv. Akad. Nauk SSSR, Ser. Khim. 2377 (1988) a 63. M P Cooke, D L Burman J. Org. Chem. 47 4955 (1982) 64. E J Corey, C V Kim,M Takeda Tetrahedron Lett. 4339 (1972) 65. M P Cooke, R Goswami J. Am. Chem. Soc. 99 642 (1977) 66. R R Heath, J R McLaughlin, J H Tumlinson, T R Ashley, R E Doolittle J. Chem. Ecol.5 941 (1979) 67. R R Heath, R E Doolittle, P E Sonnet, J H Tumlinson J. Org. Chem. 45 2910 (1980) 68. M J Gieselmann, D S Moreno, J Fargerlund, H Tashiro, W L Roelofs J. Chem. Ecol. 5 27 (1979) 69. R J Anderson, C A Henrick J. Chem. Ecol. 5 773 (1979) 70. T Suguro,W L Roelofs, K Mori Agric. Biol. Chem. 45 2509 (1981) 71. E Vedejs, P L Fuchs J. Am. Chem. Soc. 95 822 (1973) 72. S Masuda, S Kuwahara, T Suguro, K Mori Agric. Biol. Chem. 45 2515 (1981) 73. K Mori, S Kuwahara Tetrahedron 38 521 (1982) 74. W L Roelofs, M J Gieselmann, K Mori, D S Moreno Naturwissenschaften 69 348 (1982) 75. E Alvarez, T Cuvigny, H du Penhoat,MJulia Tetrahedron 44 119 (1988) 589 76. J G Millar Tetrahedron Lett. 30 4913 (1989) 77. R Baudouy,M R Sancho Tetrahedron 47 10015 (1991) 78. E Alvarez, T Cuvigny, H du Penhoat,M Julia Tetrahedron 44 111 (1988) 79. I Fleming,W Trevor, T W Newton F Roessler J. Chem. Soc., Perkin Trans. 1 2527 (1981) 80. M J Gieselmann, R E Rice, R A Jones, W L Roelofs J. Chem. Ecol. 5 891 (1979) 81. R J Anderson,M J Gieselmann, H R Chinn, K G Adans, C A Henrick, R E Rice, W L Roelofs J. Chem. Ecol. 7 695 (1981) 82. R J Anderson, H R Chinn, K Gill, C A Henrick J. Chem. Ecol. 5 919 (1979) 83. M Alderdice, C Spino, L Weiler Tetrahedron Lett. 25 1643 (1984) 84. F W Sum, L Weiler Can. J. Chem. 57 1475 (1979) 85. L Novak, L Poppe,A Kis-Tamas, C Szantay Acta Chim. Acad. Sci. Hung. 118 17 (1985) 86. C Szantay, L Novak, L Poppe, B Majors, A Kis-Tamas, F Jurak, I Vjvary Hung Teljes HU 32985 (1984); Chem. Abstr. 102 166 530 (1985) 87. L Novak, L Poppe, C Szantay Synthesis 939 (1985) 88. A M Moiseenkov, R I Ishchenko, V V Veselovskii, V N Odino- kov, E V Polunin, B G Kovalev, B A Cheskis, G A Tolstikov Khim. Prir. Soedin. 422 (1989) c 89. E V Polunin, A M Moiseenkov, A V Semenovskii Izv. Akad. Nauk SSSR, Ser. Khim. 1354 (1981) a 90. Zh Xu, S Wang, J Jiang, Y Pan Huaxue Tongbao 8 29 (1989); Chem. Abstr. 112 197 897 (1990) 91. S Wang, Zh Xu Huaxue Shigi 14 83 (1990); Chem. Abstr. 117 111 310 (1992) 92. Zh Xu, S Wang, J Chen, J Jiang, Y Pan Yingyong Huaxue 7 83 (1990); Chem. Abstr. 113 77 947 (1990) 93. W C Mitchell, R F L Mau J. Econ. Entomol. 64 856 (1971) 94. R Baker,M Borges, N G Cooke, R H Herbert J. Chem. Soc., Chem. Commun. 414 (1987) 95. P Kociensci, B Lythgoe, D A Roberts J. Chem. Soc., Perkin Trans. 1 829 (1978) 96. J R Aldrich, W R Lusby, B E Marron, K C Nicolaou, M P Hoffmann, L T Wilson Naturwissenshaften 76 173 (1989) 97. J R Aldrich, H Numata,M Borges, F Bin, G K Waite, W R Lusby Z. Naturforsch., C Biosci. 48 73 (1993); Chem. Abstr. 118 230 485 (1993) 98. P Brezot, Ch Malosse,M Renou C.R. Acad. Sci, Ser. 3 316 671 (1993); Chem. Abstr. 119 177 917 (1993) 99. P Brezot, Ch Malosse, K Mori,M Renou J. Chem. Ecol. 20 3133 (1994) 100. Jpn. P. 0 570 449; Chem. Abstr. 119 95 220 (1993) 101. L H B Baptistella, A M Aleixo Liebigs Ann. Chem. 785 (1994) 102. B E Marron, K C Nicolaou Synthesis 537 (1989) 103. S Kuwahara, D Itoh,W S Leal, O Kodama Tetrahedron Lett. 39 1183 (1998) 104. T Poll, A Sobczak, H Hartmann, G Helmchen Tetrahedron Lett. 26 3095 (1985) 105. A Cork, D R Hall, W M Blanney, M S J Simmonds Tetrahedron Lett. 32 129 (1991) 106. Sh Shu,W L Koepnick, G Abata, A Crock, S B Ramaswamy J. Stored Prod. Res. 32 21 (1996); Chem. Abstr. 125 163 761 (1996) 107. O A Pinsker, P G Tsiklauri, N Ya Grigorieva Izv. Akad. Nauk, Ser. Khim. 1385 (1999) a 108. N Ya Grigorieva, P G Tsiklauri, O A Pinsker Izv. Akad. Nauk, Ser. Khim. 1389 (1999) a 109. N Ya Grigorieva, O A Pinsker, A M Moiseenkov Mendeleev Commun. 129 (1994) 110. N Ya Grigorieva, O A Pinsker, A V Buevich, A M Moiseenkov Izv. AN. Ser. khim. 509 (1995) 111. P Brownbridge, S Warren J. Chem. Soc., Perkin Trans. 1 2125 (1976) 112. A A Vasil'ev, G V Kryshtal, E P Serebryakov Mendeleev Commun. 41 (1995) 113. A A Vasil'ev, A L Vlasjuk, G D Gamalevich, E P Serebryakov Bioorg. Med. Chem. 4 389 (1996) a�Russ. Chem. Bull. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) c�Chem. Nat. Compd. (Engl. Transl.) d�Pharm.-Chem. J. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (7) 591 ± 608 (2000) Diacetylene: a candidate for industrially important reactions I A Maretina, B A Trofimov Contents I. Introduction II. The problem of diacetylene utilisation III. Reactions with amines IV. Reactions with alcohols V. Reactions with sulfur-containing nucleophiles VI. Reactions with ketoximes and amidoximes VII. Reactions with hydrazines and semicarbazide VIII. Reactions with guanidine IX. 1-Dialkylaminobut-1-en-3-ynes as building blocks for the synthesis of heterocyclic compounds X. Reactions with carbonyl compounds XI. Conclusion Abstract. synthe- important industrially to devoted is review The The review is devoted to industrially important synthe- ses from product side a diacetylene, on based ses based on diacetylene, a side product from electrocracking, electrocracking, oxidative synthesis plasmochemical and methane of pyrolysis oxidative pyrolysis of methane and plasmochemical synthesis of of acetylene.with derivatives its and diacetylene of Reactions acetylene. Reactions of diacetylene and its derivatives with mono- mono- and chemisorption the in occurring reagents dinucleophilic and dinucleophilic reagents occurring in the chemisorption of of diacetylene stable in resulting and gases industrial from diacetylene from industrial gases and resulting in stable reactive reactive compounds intermediates valuable as used be can which compounds which can be used as valuable intermediates for for organic synthesis and pharmaceutical industry are described. organic synthesis and pharmaceutical industry are described.Some in reactions diacetylene of features specific Some specific features of diacetylene reactions in superbasic superbasic media references 266 includes bibliography The discussed. are media are discussed. The bibliography includes 266 references. I. Introduction Preparation of chemical products on the basis of acetylene is one of the most promising trends in basic industrial organic synthesis. The interest in acetylene-related raw materials, which subsided considerably in 1955 ± 1975, has now rekindled.1 The reorienta- tion towards coal and gas as valuable raw materials of the future and the development of new technologies for acetylene production from coal, e.g. using plasmochemical reactions, played a crucial role in the revision of the place of acetylene in organic synthesis.2 The use of acetylene and its derivatives in small-tonnage produc- tion of sophisticated and expensive organic compounds holds especially great promise.1, 2 The classical studies by A E Favorsky and W Reppe have formed the basis for modern technologies of the chemistry of acetylene, viz., vinylation, ethynylation and carbonylation reac- tions.Avast body of accumulated data on reactions of acetylene is documented in numerous monographs and reviews.1 ±40 On the other hand, some technological aspects of the chemistry of acetylene were obviously underestimated.This is particularly true of diacetylene (butadiyne) formed in the manufacture of I A Maretina, B A Trofimov Irkutsk Institute of Chemistry, Siberian Branch of the Russian Academy of Sciences, ul.Favorskogo 1, 664033 Irkutsk, Russian Federation. Fax (7-395) 239 60 46.Tel. (7-395) 246 14 11. E-mail: admin@irioch.irk.ru (B A Trofimov) Received 29 November 1999 Uspekhi Khimii 69 (7) 642 ± 660 (2000); translated by R L Birnova #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n07ABEH000564 591 591 592 595 598 599 600 601 601 602 605 acetylene. As regards the potential utility of side products derived from acetylene and accompanying compounds (methylacetylene, allene, diacetylene, higher alkynes, phenylacetylene, etc.), di- acetylene occupies the second place after the methylacetylene ± allene fraction.A significant contribution to the chemistry of diacetylene and its commercial application was made by Russian scientists belong- ing to the M F Shostakovskii 3, 8, 11 ± 32, 34 and A A Petrov 36 ± 40 scientific schools. The aim of the present review is to analyse and systematise the published data devoted to industrial syntheses based on diacetylene and to evaluate the prospects for the development of this branch of chemistry. II. The problem of diacetylene utilisation Despite the fact that many compounds of the diacetylene series (1,3-diynes) had been known since the XIX century, the develop- ment of the chemistry of these compounds began only after 1950 3 when they were discovered in nature 4, 5 as well as among the products of electrocracking and oxidative pyrolysis of methane.6 Under industrial conditions, these two reactions yield, in addition to acetylene, some of its homologues and derivatives among which diacetylene constituting 5% of the whole manufac- ture of acetylene presents the greatest practical interest.1 Utilisation of the accompanying diacetylene presents a serious problem for chemical technology. All industrial methods used for the isolation of acetylene are based on the sorption ± desorption principle.Higher alkynes and diacetylene are extracted with the help of selective solvents, e.g., anhydrous ammonia, cooled methanol, N-methylpyrrolidone, DMF.3, 6 The first solution to this problem was proposed and imple- mented at the oldest petrochemical plant `Chemische Werke Huls A.G.' 41 Preparation of 1-methoxybut-1-en-3-yne as a 50% sol- ution in methanol from accompanying diacetylene has been described.42 However, this method is not industrially acceptable because of extremely high diacetylene concentrations in the reactor.Labour protection laws currently in force in the Russian Federation prohibit the use of industrial technologies based on the use of material (gas or liquid) flows containing more than 10% of diacetylene.6 Therefore, problems of diacetylene extraction from gases produced by pyrolysis of hydrocarbon raw materials have592 received recently considerable attention, especially with regard to the safety of acetylene extraction due to the ability of diacetylene to undergo polymerisation and decomposition with explosion.It should be noted in this connection that operations with diacetylene in manufacturing plants and in the laboratory belong to the category of particularly hazardous ones. It is therefore expedient to give due consideration to the vast experience accu- mulated in this area. The formation of a liquid phase in gas flows containing diacetylene is impermissible. The schemes of chemical conversion of diacetylene into practically valuable compounds should employ exclusively alkaline media, because the polymers formed under these conditions as side products are not hazardous. Manipulations with condensates or products prepared based on industrial diacetylene-containing gases differ substantially from those with pure diacetylene, since the polymerisation processes are more pronounced owing to the presence of higher 1,3-diynes.It is therefore preferable to prepare compounds which can easily be removed by distillation. It is rather difficult to isolate high-boiling products from such mixtures by conventional methods. The technologies implemented in Russia were developed with due regard for safety for two main types of the processes, viz., those in which liquid ammonia 36, 43, 44 and N-methylpyrroli- done 37 were used as selective solvents. A possible approach to diacetylene utilisation is based on its ability to react with alco- hols.36, 37, 41, 42 The products of this reaction can be used in the synthesis of some aromatic and heterocyclic compounds (e.g., monomers, dyes and drugs) as well in the production of plasti- cisers, herbicides, rubber vulcanisation accelerants, intermediates for electronic industry, broad-spectrum metal corrosion inhib- itors, etc.36, 37 Other conversions of diacetylene can be used in the synthesis of otherwise inaccessible heterocyclic com- pounds.38 ± 40, 45 ± 47 All this makes diacetylene and its heterofunc- tional derivatives promising candidates for building blocks in the targeted synthesis of heterocyclic compounds.11 ± 17, 38 ± 40, 47 Diacetylene most easily enters into reactions of nucleophilic addition with amines, alcohols and thiols, and high reactivities of the adducts make them promising reagents for practical purposes.In reactions with nucleophilic reagents, diacetylene behaves as a typical alkyne activated by an acceptor group,47 i.e., in this case the addition occurs at the terminal carbon atom of the acetylene bond, eventually resulting in the formation of 1-heteroalk-1-en-3- ynes. III. Reactions with amines Tne addition of primary amines to diacetylene occurs readily at 30 ± 40 8C in the absence of a catalyst. Usually, monoadducts of the type 1 cannot be isolated,48 1-tert-butylaminobut-1-en-3-yne (1f ) being the only exception.49 N-Alkyl(b-aminovinyl methyl)- ketimines or their tautomers, e.g., N-alkyl-3-alkylaminocrotonal- dimines 2, are the reaction products, which are formed in good yields. RNH2 RNH2 HC C C CH NHR 1a ± f R R N N N H Me Me N H 2a ± f R R R=Me (a), Et (b), Prn (c), Pri (d), Bun (e), But (f).According to the published data,48, 49 the adducts 2 have s-Z- conformation which is stabilised by an intramolecular hydrogen bond. The reaction of diacetylene with secondary amines gives 1-dialkylaminobut-1-en-3-ynes (3).3, 47 R2NH HC C C CH 3a ± e R=Me (a), Et (b), Prn (c), Bun (d); R2N= N The stereochemistry of the reaction of diacetylene with diethylamine in various solvents and at different temperatures resulting in the formation of the Z- and E-isomers 3b has been studied.50 The E :Z ratio increases with temperature. The Z-iso- mer prevails in acetonitrile, diethylamine and THF at 0 ± 20 8C, whereas the E-isomer is predominantly formed in ethanol at 20 8C.The Z?E transition is observed only in the photolysis. The kinetics of addition of secondary amines to diacetylene in dioxane at 20 ± 80 8C corresponds to that of second-order reac- tions.51 The formation of compounds 3b,d was monitored by UV-spectroscopy. The low values of activation entropy (DS==755 e.u.) and enthalpy (DH==33.4 kJ mol71) for the amine 3b prompted a conclusion 47 that the addition occurs via intermediate products with considerable charge separation (bipolar ions). The final products result from intramolecular (E-isomers) or intermolecular (Z-isomers) proton transfer. The charge separation in the transition state favours both specific orientation of the amine and dioxane molecules and low values of activation entropy and enthalpy.47 1-Dialkylaminobut-1-en-3-ynes 3 with primary alkyl substitu- ents react with primary amines to give potentially tautomeric aminocrotonaldimines 4 due to the exchange of the amino groups.Reactions of compounds 3 with arylamines yield C-(2-dialkyl- aminovinyl)-C-methyl-N-(aryl)azomethines 5.40, 48 Depending on conditions, the hydration of amines 3 gives either 4-dialkylami- nobut-3-en-2-ones 6 3, 46, 52, 53 or 3-dialkylaminobut-2-enals 7 (Scheme 1).46 R2N Me R12 N NR2 4a ± dR2NH2 R12 N3a ± dH2O Me R12 N O 6a ± dR3OH CH2(CN)2 Me R12 N OR3 R3O8a ± d R1=Me (a), Et (b), Prn (c), Bun (d). The addition of alcohols to enamino ketones 6 which is catalysed by bases gives the vinylogues of aminoacetals 8 40 which easily react with CH-acids, e.g., malonodinitrile, to give the diene 9 (see Scheme 1).40 Stepwise C-alkylation and hydro- genation of the amines 3 results in N-monosubstituted 1-amino- alkanes which represent valuable building blocks for the introduction of the 4-aminobutyl group.54 The reaction of diacetylene with an excess of diethylamine in dilute aqueous solutions of acetonitrile, ethanol or THF results in the diadduct, viz., 1,3-bis(diethylamino)buta-1,3-diene (10); upon subsequent concentration of the reaction mixture, the latter is hydrated by a residual amount of water to be converted first into the hemiaminal 11 and then into the aldehyde 7b following splitting-off of diethylamine.55 I A Maretina, B A Trofimov NR2 O(e).Scheme 1 Me NR12 Me R12 N ArNH2 N Ar 5a ± d O Me 7a ± d NR12 Me R12 N CN NC 9a ± dDiacetylene: a candidate for industrially important reactions Et2NH Et2NH H2O HC C C CH NEt2 3b NEt2 NEt2 10 Me Me NEt2 Et2N ±Et2NH O Et2N HO 7b 11 The reaction of the monoadduct 3b with diethylamine is accompanied by the appearance of a band at 325 nm in the UV spectrum of the reaction mixture; its intensity increases with a simultaneous decrease in the intensity of the band at 275 nm corresponding to the original amine 3b. After completion of the reaction, the UV spectrum contains only the former band corre- sponding to the product 10.55 The absorption band of the original enyne 3b in the IR spectrum is shifted towards the low-frequency region; its intensity decreases simultaneously with the appearance of a band around 1580 cm71, which corresponds to the product 10 having Z-con- figuration.55 The intensity of this band increases as the reaction proceeds further.The first diethylamine molecule adds to diac- etylene in acetonitrile much faster than the second one. Thus, the first step of the conversion of diacetylene into the bisadduct 10 is the rate-limiting step. The enhancing effect of the solvent is attributed to incomplete removal of water and hence to electrophilic catalysis by water molecules which attack the termi- nal atom of the ethynyl group: HO H 10. NR2 NR2 H This transition state is in accord with the observed regioselec- tivity of the reaction and Z-configuration of the diene formed as well as with the high negative entropy and low activation energy of the reaction.The proton transfer and the addition of the nucleo- phile are synchronous processes and are accompanied by the migration of the positive charge to the a-carbon atom as might be expected. Otherwise, the allene 12 would be formed.55 R2NH + 7H+ NR2 NR2 NR2 + + NR2 12 R2N In aqueous solutions, the role of a nucleophile is played by water, which leads to the formation of the enamino ketone 6b. The calculated rate constants for this reaction in aqueous THF are much lower than the rates of addition of the second diethylamine molecule to the amine 3b, the enthalpy of activation is also lower.This attests to a concerted process and the involvement of the nucleophile in the limiting stage of both reactions, since water is a much weaker nucleophile than diethylamine.55 1,2-Diaminoethane reacts with diacetylene to give a mixture of tautomers of 5-methyl-2,3-dihydro-1,4-diazepine (13).56 Me Me HC C C CH H2N(CH2)2NH2 N NH HN N 13 It was expected that the reaction of diacetylene and its homologues with 1,3-diaminopropane would give 1,2,3,4-tetrahy- dro-1,5-diazocine.57 However, the reaction product represents a mixture of D1-tetrahydropyrimidine (14), 2-methyl-D1-tetra- hydropyrimidine (15), 2-methyl- (16) and 2,2-dimethylhexahydro- pyrimidines (17). H2N(CH2)3NH2 HC C C CH + NH N NH N Me 15 H14 This can be rationalised as the addition of two diamine molecules to one diyne molecule, resulting in the diadducts A and B which undergo further cyclisation and fragmentation in two directions (Scheme 2).(CH2)3NH2 H2N(CH2)3 H N N Me A H N N (CH2)3 H2N(CH2)3 NH Me 14 H2N(CH2)3 NH Me C 16 These results are similar to those obtained for the reaction of acetylacetone with 1,3-diaminopropane 57 and formally corre- spond to the scission of b-diketone derivatives. As noted above, reactions of diacetylene with solutions of primary and secondary amines in aqueous ethanol produce enamino ketones 6. It was found, however, that the physical constants of enamino ketones 6 thus prepared do not coincide with those prepared by amination of alkoxy ketones.58 Later, this discrepancy was explained 59 by the formation of enamino alde- hydes 7 along with enamino ketones 6 in reactions of diacetylene with solutions of amines in aqueous ethanol.Depending on the size of the alkyl substituents at the nitrogen atom and the concentration of the aqueous solution of alkali (3% ± 5%), the content of the aldehyde in the reaction mixture changes and may exceed 40% with respect to the ketone.59 Thus if 3-dimethyl- aminobut-2-enal (7a) is formed only in trace amounts, the yields of the products 7b and 7d amount to 35% and 43%, respectively.59 It is believed 46 that the formation of enamino ketones 6 from amines 3 can also occur via aldehydes 7. With an increase in temperature, the ketone : aldehyde ratio increases.46 Thermal isomerisation occurs quantitatively, probably, via the oxetene cycle.The aldehyde 7b is fully isomerised into the ketone 6b in 4 h at 150 8C. Et Et N N EtO Me Me 7b O N Me Et 6b The addition of water to amines 3a,b occurs easily in the presence of catalytic amounts of an alkali exclusively at position 3 and results in alkyl 2-dimethylaminovinyl ketones 18.60 593 + + NH HN NH HN H Me Me Me 17 16 Scheme 2 (CH2)3NH2 H2N(CH2)3 H N N Me B H N N (CH2)3NH2 (H2C)3 HNMe 15 (CH2)3NH2 HN D 17 Me Et O Et N O Et Et594 NMe2 HO7, H2O R NMe2 O18a,b 3a,b R R=Me (a), Et (b). These conversions were used to develop a procedure for chemisorption of diacetylene by aqueous solutions of primary and secondary amines aimed at separation of diacetylene in the manufacture of acetylene by oxidative pyrolysis of methane.59 This approach has been approved at the Yerevan Industrial Corporation `Nairite' with both a real abgas { and a diacetylene- enriched condensate.Their compositions are shown in Table 1.59 The presence of methane, nitrogen, ethylene and some other components in the abgas can be explained by the fact that diacetylene-containing gases are diluted with a cracking gas from which acetylene had previously been removed by absorption down to concentrations of HC:C7C:CH (<10 vol.%) which ensure their safety upon transportation.59 Chemisorption of diacetylene is performed by a 25% aqueous solution of dimethyl- amine.The number of absorption plates and the design of the absorption apparatus ensure the absorption of 92%± 94% of diacetylene at its average concentration of 6%± 7%. The addition of dimethylamine to diacetylene occurs at 25 ± 30 8C. Since this reaction is exothermic, the temperature of the reaction mass increases slowly to 50 8C.{ The amine 3a formed in the first step was isolated from the reaction mixture in its individual state and characterised.59 Table 1. The composition of abgases and their condensates. Content (vol.%) Component in condensate in abgas 777 35 ± 40 20 ± 25 7.0 ± 11.0 6.0 ± 8.0 1.0 ± 2.0 1.0 ± 1.5 1.0 ± 1.5 0.5 ± 1.0 0.5 ± 1.8 36 ± 52 75.5 ± 7.5 5.0 ± 7.0 3.5 ± 5.5 4.0 ± 4.5 Hydrogen Methane Nitrogen Diacetylene Ethylene Methylacetylene Benzene Allene Vinylacetylene Then, the reaction mass is loaded into a reactor where the triple bond is hydrated at 60 ± 65 8C. The ketones 6 formed are rather stable and convenient C4-synthons for the synthesis of heterocyclic compounds.38 ± 40 1-Dialkylaminobut-3-yn-2-ones (19), which are promising reagents for the synthesis of heterocyclic compounds, are easily prepared by bromination and dehydrobromination of enamino ketones 6.62 ± 67 Br + O NR2 Br2 Me R2N Br7 6a ± c Me OH Br O Me R2N R2N C C 7HBr Me 19a ± c O R=Me (a), Et (b), Prn (c).{ Abgas is a technical term, which designates an absorption gas. { Among a vast array of products present in diacetylene as admixtures, vinylacetylene ranks second in reactivity after diacetylene.However, even this compound reacts with aqueous solutions of secondary amines at a detectable rate only at 100 8C in an autoclave.61 I A Maretina, B A Trofimov The structures and reactivities of enamino ketones 6, 18 and ynamino ketones 19 were described in numerous publica- tions.38, 39, 52, 62 The effects of substituents on the distribution of electron density in the molecules of en- and ynaminocarbonyl compounds were studied by 13C-NMR spectroscopy. These studies were aimed at the determination of spin-spin coupling constants (JCC) 13C713C at the double and triple bonds.62 The JCC values were found to be significantly smaller than those calculated according to the additive scheme.This is due to delocalisation of the multiple bonds by direct polar conjugation, through the double (triple) bond involving the lone pair on the nitrogen atom and the p-system of the carbonyl fragment. R R + N O N O7 R R Me Me 6 O O7 R + C C C N R2N C Me Me R 19 The contribution of the bipolar resonance form was estimated to be *30% for enamino ketones 6 and exceeded 50% for ynamino ketones 19.62 Comparison of chemical shifts in the 13C NMR spectra of carbon atoms at the triple bond of sub- stituted ynamines with the corresponding values for the C atoms at the double bond of enamines led to a conclusion about cross- polarisation of the orthogonal p-system of the triple bond in ynamines containing acceptor groups.62 This effect is manifested in the abnormally low shielding exerted by the dialkylamino group on the carbon atom of the multiple bond nearest to the nitrogen atom.Data on the structures of ynamino carbonyl compounds suggest that the conjugation of the donor and acceptor groups through the multiple bond (the so- called `push-pull effect') strongly influences the distribution of electron density in the molecules and, as a consequence, their reactivities. As a result, these compounds acquire specific proper- ties which differ radically from those of both ynamines and ketones containing a triple bond.40 The behaviour of ynamino ketones 19 in cyclisation reactions differs from that of enamino ketones 6 and 18.38, 39, 52 Whereas enamino carbonyl compounds typically `deliver' three-carbon fragments in cyclisation,40 ynamino ketones mostly react like simple ynamines, i.e., with a twofold attack by bifunctional reagents on the carbon atoms of the triple bond (1,1-attack).52 HO Me X HX(CH2)2NH2 N 20a ± c NH2 Me O O N YH R2N C C 19a ± c Me Y 21a,b NH2 NR2 N NH2 N Me 22a ± c 19, 22: R=Me (a), Et (b), Prn (c); 20: X=NH (a), O (b), S (c); 21: Y=O(a), S (b).Thus the reaction of ynamino ketones 19 with ethylenedi- amine, ethanolamine and 2-mercaptoethanol is accompanied by the elimination of the dialkylamino group and the formation ofDiacetylene: a candidate for industrially important reactions 1-(4,5-dihydroimidazol-2-yl)prop-1-en-2-ol (20a), 1-(4,5-dihydro- oxazol-2-yl)prop-1-en-2-ol (20b) and 1-(4,5-dihydrothiazol-2- yl)prop-1-en-2-ol (20c), respectively, containing a fully enolised carbonyl fragment.68, 69 The use of derivatives 20 as the Xa-inhi- bition factor 70 and anticonvulsants, e.g., in the treatment of industrial electronic shock traumas, was described.71 The ynamino ketones 19 react with aromatic a,b-bifunctional reagents, e.g., 2-aminophenol and 2-aminobenzenethiol, like ordinary ynamines 72 ± 76 to produce 1-(benzooxazol-2-yl)propan- 2-one (21a) and 1-(benzothiazol-2-yl)propan-2-one (21b), respec- tively.77, 78 Benzothiazole 21b has been used in the synthesis of compounds for modulation of information signals 79 and phar- maceutical preparations against acidophilia, bronchial asthma, allergies,80 diabetes and nephropathies.81 Benzothiazoles of the type 21b manifest herbicidal activities 82 and are used in the production of liquid-crystal compositions.83 The cyclisation of ynamino ketones 19 with o-phenylenedi- amine results in 2-dialkylamino-4-methyl-3H-1,5-benzodiaze- pines 22 69 in up to 83% yields.Photosensitive materials based on benzodiazepines 22 were patented.84 Compounds of this class are used in the therapy of cholecystitis, gastritis 85 and meningitis 86 as well as squalene synthase inhibitors,87 atypical neuroleptics 88 and fungicides against Penicillum notatum.89 Thus, benzodiazepines of the type 22, which have a wide range of practical applications, can also be synthesised from ynamino ketones.The reaction of diacetylene with primary amines in the presence of copper(I) salts occurs as 1,4-addition and results in the formation of pyrrole or 1-alkylpyrroles.3 This reaction is carried out at 140 ± 160 8C both in the absence of the solvent and in methanol, ethanol or DMF. In the latter case, the resulting pyrroles have a high degree of purity and are formed in higher yields. The reaction of ammonia and primary alkyl- and aryl- amines with disubstituted symmetrical 1,3-diynes in the presence of copper(I) chloride affords 1,2,5-trisubstituted pyrroles 23.3 R1 Cu2Cl2 R2 N R1C C C CR1+R2NH2 23 R1 R1=H, Me, Et, Ph; R2=H, Me, Et, Bun, Ph. According to Shostakovsky and Bogdanova,3 the role of a catalyst in this reaction consists in the formation of a non-polar p-complex with one of the triple bonds where the electron density is evenly distributed between both carbon atoms, which facilitates the interaction of the nucleophilic nitrogen with the fourth carbon atom of the conjugated diyne system.The smooth course of this reaction made it possible to develop a general scheme for the synthesis of substituted pyrroles.3 The practical significance of this class of compounds is documented in numerous monographs and reviews.13 ± 16, 21, 29, 90 ± 94 In the absence of a catalyst, the reaction of diphenylbutadiyne with benzylamine affords 2,3,6-tri- phenylpyridine (24) in 50% to 70% yield.95 Ph Ph Ph N24 PhCH2NH2 PhC C C CPh Ph Cu2Cl2 N CH2Ph+24 Ph The specific feature of this reaction is that in the absence of a catalyst only pyridine 24 is formed; however, in the presence of Cu2Cl2 it proceeds simultaneously in two directions. This phe- 595 nomenon is interpreted by the ability of copper(I) chloride to coordinate both at the nitrogen atom and at the triple bond.95 IV.Reactions with alcohols The synthesis of 1-alkoxybut-1-en-3-ynes 25 by the reaction of diacetylene with alcohols in the presence of a catalyst was described;3 however, the physical constants obtained by different authors differ.96 This may be due to the difference in the ratios of E- and Z-isomers in the specimens prepared under different conditions or due to their contamination with products of further transformations of the ethers 25.HO7 HC C C CH+ROH OR 25a ± e R=Me (a), Et (b), Prn (c), Bun (d), n-C5H11 (e). The addition of methanol, ethanol, propan-1-ol, butan-1-ol, 2-methylpropan-2-ol, pentan-1-ol and hexan-1-ol to diacetylene has been studied under identical conditions at the alcohol : di- acetylene ratio of 3 : 1 in the presence of catalytic amounts of KOH.96 The content of E- and Z-isomers was determined on the basis of 1H NMR spectroscopy and GLC data. The ethers 25a,b appeared to be pure Z-isomers; their homo- logues 25c,d contained approximately equal amounts of the E-isomer (12.8% ± 13.3%), whereas compounds 25e,f represented equimolar mixtures of E- and Z-isomers.96 It is probable that the rule of trans-nucleophilic addition is observed only for lower alcohols.However, if the reaction mixture contains compounds capable of forming complexes with diacetylene, the above rule may not be observed as is the case even with the reaction of diacetylene with methanol, which gives up to 15% of the E-isomer with dioxane as a solvent.96 Kinetic studies of reactions of diacetylene with lower aliphatic alcohols in the presence of KOH were carried out using UV spectroscopy and GLC.97, 98 The reaction of diacetylene with methanol (5 moles/mole of HC:C7C:CH) in the presence of KOH (0.22 moles/mole HC:C7C:CH) is described by an equation for a pseudo-unimolecular reaction with the activation enthalpy, DH=, of 87.8 kJ mol71 and the activation entropy, DS=, of 719 e.u.The experimental results suggest 47 that the addition of the alkoxide ion to the terminal carbon atom of diacetylene followed by rapid protonation of the carbanionic intermediate 26 formed is the rate-limiting step of this reaction. sp2-Hybridisation of the carbon atom orbitals at the double bond and spatial hindrances created by bulky alkoxy and ethynyl groups direct the attack of the proton donor to the trans-position relative to the alkoxide ion. H 7 HC C C CH+ROH RO7 RO H OR 26 UV spectroscopy and GLC were used to study 98 the kinetics of the reaction of diacetylene with methanol and ethanol under conditions close to the industrial ones,37 in the presence of potassium hydroxide and sodium hydroxide at 50 ± 73 8C and at a 1,3-diyne concentration of 2.5 mass%± 10 mass %.Such a narrow temperature range can be explained by the fact that at lower temperatures diacetylene is predominantly polymerised due to the slow rate of the main reaction. At temperatures above 73 8C, this reaction proceeds further.98 Since the rate of formation of the ether 25b is proportional to the concentration of diacetylene in the solution, the reaction should be carried out at the maximally permissible (up to 10 mass%) concentration of HC:C7C:CH. The dependence of the rate constant for the formation of the ether 25b on KOH concentration is more complex. With an increase in KOH concentration from 1% to 5%, the rate constant increases abruptly. Within the concentra-596 tion range ofKOHfrom 5 mass%to 15 mass %, the reaction rate is not changed (kaver=2.7 ± 2.8 l mol71 s71, 60 8C), but this drops drastically after further increase in the catalyst concentra- tion.98 This is ascribed to the simultaneous effect of several factors.At low concentrations of the catalyst, the increase in the reaction rate is mainly due to an increase in the concentration of the alkoxide ions responsible for this reaction. At 5 mass%± 15 mass% of KOH, the concentration of alkoxide ions increases to such an extent that its fluctuations within the indicated range do not influence the reaction rate. At KOH concentrations above 15 mass %, the rate of side reactions increases drastically. From these data, it was concluded that the optimum concentration of KOH is 5 mass %.98 When diacetylene was passed through a solution of mono- ethanolamine in benzene at 0 8C, a crystalline product was formed which represents 2-[2-(2-hydroxyethyl)iminopropyl]oxazolidine (27) as can be judged from the 1H and 13C NMR spectral data.NH O HC C C CH+2HO(CH2)2NH2 N(CH2)2OH Me 27 This adduct of diacetylene with ethanolamine is used for stabilisation of biologically active substances in grass fodders.99 The binding of 1,2- and 1,3-diols to diacetylene was studied in the presence of catalytic amounts of alkali and resulted in a mixture of cyclic acetals of but-3-ynal, buta-2,3-dienal (28) and but-2-ynal (29); their ratio changed depending on experimental conditions.100 R1 R2 R1 O HO7 + (CH2)n HC C C CH+HOCH(CH2)nCHOH O R2 R1 R1 O O Me C C (CH2)n (CH2)n + + O O R2 R2 29 28 n=0: R1=R2=H; n=1: R1=R2=H, Me, R1=Me, R2=H.Bifunctional derivatives of the aromatic series, such as o-, m- and p-aminophenols, add to diacetylene in the presence of alkalis 2ROH ROH HC C C CH RO25a,b 3ROH ROH ROH, H2O RO Me RO H2O RO H2O C C Me RO OR OR OR 31a,b 33a,b OR O 32a,b ROH 2ROH Me 33a,b RO OR OR 34a,b R=Me (a), Et (b). I A Maretina, B A Trofimov like ordinary alcohols, i.e., through the hydroxy group, to give 1-aryloxybut-1-en-3-ynes (30).101 H2N H2N HO7 O HC C C CH+ HO 30 Alcohols (glycols, amino alcohols) and phenols add to diac- etylene under much milder conditions than to acetylene and vinylacetylene.37, 98 The vinylation of lower alcohols with acety- lene requires the use of superbasic media 19, 20, 23, 28, 102 or heating up to 120 ± 140 8C, and their addition to vinylacetylene occurs at temperatures around 100 8C, while the optimum temperature for the reaction of diacetylene with methanol in the presence of KOH (5 mass %) is as low as 72 ± 73 8C.37, 98 This circumstance is extremely important because the substantial difference between the rates of the reaction of methylacetylene, vinylacetylene and diacetylene with lower aliphatic alcohols allows selective chem- isorption of diacetylene from gas mixtures produced by acetylene- manufacturing plants 36, 37 which contain these particular hydro- carbons.Moreover, extraction of diacetylene from industrial gases allows implementation of technologies for industrial syn- theses of isopropenylmethyl ether, acetone and acetylacetone based on the accompanying methylacetylene ± allene fraction.1 Numerous transformations of the ether 25a have been described under conditions of alkaline and acid catalysis (Scheme 3).103 Thus the addition of the second methanol molecule to the ether 25a in the presence of an alkali results in 1,1-di- methoxybut-3-yne (31a).Its analogue, viz., 1-methoxy-1-phen- ethoxybut-2-yne, is used in perfumery.104 1,1-Dimethoxybutan-3-one (32a) is formed in 80% yield as a result of hydration of the triple bond of acetal 31a under conditions of acid and mercury salt catalysis (H2SO4 , HgSO4) in boiling aqueous methanol (see Scheme 3).The addition of an alcohol and water to the ether 25a in the presence of an acid catalyst also results in the acetal 32a in 85% yield. 1,1,3,3-Tetramethoxybutane (33a) and a small amount of 1,1,3-trimethoxybut-2-ene (34a) are formed under identical con- ditions but in the presence of dry methanol (3 equiv.). Upon mild hydrolysis at room temperature, this mixture is converted into the acetal 32a; its complete hydrolysis yields 1,3,5-triacetylbenzene, which was first observed by Claisen.105 Thus, the acetals 32a,b appear to be the most convenient and stable intermediates prepared from diacetylene.106 Thermal or catalytic elimination of methanol from the acetal 32a gives 4-methoxybut-3-en-2-one (35a);106 a mixture of com- pounds 32a and 35a (7 : 3) is formed in the presence of catalytic amounts of acids or alkalis.Its homologue 32b (R=Et) yielded Scheme 3 H2O Me Me RO O 35a,b 7ROHDiacetylene: a candidate for industrially important reactions an equilibrium mixture of 32b and 35b in a 7.5 : 2.5 ratio upon storage under conditions close to those employed in industry.98 The structure of the ketone 35a was studied in detail using UV, IR and NMR spectroscopy.107 Analysis of thermochemical and 1H NMR spectral data revealed 108 that this compound has E-configuration.108 A kinetic analysis of acid-catalysed hydrolysis of the ethers 25a,b by polarography 109, 110 provided support for the initial rapid hydration of the triple bond with subsequent slow proto- nation of ketones 35a,b, which are further decomposed to 3-oxo- butanal (36); the latter undergoes self-cyclisation into 1,3,5- triacetylbenzene (37).In water, the rate of this reaction is an order higher than in 25% aqueous DMF. H+ H+, H2O H2O OR Me OR OR + 25a,b O 35a,b Me O O Me Me Me 36 O 37 O O R=Me (a), Et (b). Evidence for primary hydration of the ethers 25a,b was also obtained in kinetic studies of formation of 1,1-dialkyloxybutan-3- ones by GLC.111 ± 113 The reaction of 1-methoxypent-1-en-3-yne (38a), 1-meth- oxyhex-1-en-3-yne (38b) and 1-methoxyhept-1-en-3-yne (38c) with methanol and water under conditions of acid catalysis gave 1,1-dimethoxyalkan-3-ones 39a ± c, respectively.114 MeO H2O, MeOH, H+ R OMe O MeO 38a ± c R 39a ± c R=Me (a), Et (b), Prn (c).Derivatives of the aldehyde 36 containing functional groups in the methylene group have been synthesised. The synthesis of 2-bromo-3-oxobutanal (40) by direct bromination of the acetal 32 was described.115 However, reproduction of the experiments and structural analysis of the reaction products by GLC revealed the formation of three compounds, which suggests that bromina- tion of the original acetal occurs in all the three possible direc- tions.115 The bromosubstituted ketone 40 was synthesised in *60% yield by bromination of enamino ketones 6a,b in an aqueous medium. This reaction is accompanied by hydrolysis of 4-dialkylamino-3,4-dibromobutan-2-one 41 formed with elimina- tion of the secondary amine and dehydrohalogenation.115 Br Me Me Br2, H2O R2N R2N H2O + 7[R2NH2Br7] O Br 41a ± d O 6a,b,e,f Br Me O O 40 R=Me (6a, 41a), Et (6b, 41b); R2N=(CH2)4N (6e, 41c), (CH2)5N (6f, 41d).Intensive development of the chemistry of b-functionalised vinyl ketones has begun following synthesis of b-chlorovinyl methyl ketone by addition of acetyl chloride to acetylene.116 The synthetic potentialities of b-chlorovinyl ketones have been reviewed.116 ± 118 The transfer of the acetylvinyl residue was named by A N Nesmeyanov and N K Kochetkov as `ketoviny- lation'.119 The synthetic potential of this reaction was further expanded owing to the involvement of other b-substituted vinyl ketones RCOCH=CHY (Y=OR, NR2 , SR, NO2, N3 , 597 SO2R);120 the activities and the selectivities of their reaction centres C(1) and C(3) can be altered by using different solvents.121 A pilot plant for the synthesis of the acetal 32b based on dilute diacetylene-containing industrial gases accompanying manufac- ture of acetylene by electrocracking of methane (see Scheme 3) was installed at the Scientific-and-Industrial Association `Azot' (Saratov) in 1974.The ether 25b was obtained as a by-prod- uct.37, 122 This version of diacetylene utilisation was implemented on an industrial scale with DMF and N-methylpyrrolidone as selective solvents. A pilot-plant unit for isolation of diacetylene formed as a side product in the manufacture of acetylene by oxidative pyrolysis of methane and preabsorption with ammo- nium was installed at the Scientific-and-Industrial Association `Azot' (Novomoskovsk) in 1980.36, 43, 44, 123 ± 127 The technological process 36, 123 involved isolation of diacetylene from the bottoms of an ammonium preabsorption unit followed by the synthesis of the ether 25a.Its hydration and methoxylation in the presence of catalytic amounts of H2SO4 yield the acetal 32a, which is further used in the synthesis of 4-methylpyrimidine-2-thiol (42) by the reaction with thiourea in the presence of HCl and its oxidation product, viz., bis(4-methylpyrimidin-2-yl) disulfide (43).36, 123 Me Me MeO N (H2N)2C S, HCl SH N42 OMe O 32aMe Me N N N N S S 43 The pyrimidinethiol and disulfide thus prepared accelerate the vulcanisation of rubbers having a unique spectrum of action.127, 128 Their addition to rubber mixtures increases the resistance of the compositions to prevulcanisation by 150%.The disulfide 43 is a novel accelerant having a retarded initial period of action. The use of such accelerants prevents early cross-linking in processing of rubbers.128 ± 130 Pyrimidinethiol 42 was used in the preparation of antitumour drugs 131 and novel nematocides.132 It is also used as a lustre-enhancing coating in the production of electrolytic copper foils.36, 133 The addition of small (0.003 ± 0.03 g litre71) concentrations of this compound to elec- trolytes for the production of copper foil increases its yield by 60% and improves its physico-mechanical properties.The use of pyrimidinethiol 42 in the production of photographic materials for thermal development is widely known.134 Thus, the interest in the practical use of diacetylene is determined by industrial demands, on the one hand, and by broad potentialities of the syntheses on the basis of 1,3-bifunc- tional compounds, on the other hand. The latter are the nearest derivatives of diacetylene and offer promising approaches to the synthesis of various classes of organic compounds. The list of products which can be prepared from diacetylene is given below. Price per kg /USDa Compound 201.75 1716 1600 1290 236 660 2566 5200 2117.5 770 1,1-Dimethoxybutan-3-one (E)-4-Methoxybut-3-en-2-one 1,3,3-Trimethoxybutane 4-Diethylaminobutan-2-one Thiophene 5-Methylisoxazole 3-Methylpyrazole 4-Methylpyrimidine 2-Amino-4-methylpyrimidine 4-Methylpyrimidine-2-thiol hydrochloride aThe prices of 1997 are taken from Ref.135.598 V. Reactions with sulfur-containing nucleophiles 1. Reactions with hydrogen sulfide and sulfide ions Systematic studies of reactions of acetylene and its derivatives with hydrogen sulfide and sulfide ions 12, 17, 20, 23, 26 ± 32, 136 ± 138 have been extended to diacetylene.12 The reaction of diacetylene with hydrated sodium sulfide occurs through the formation of the ethynylvinylthiolate anion (44) and results in both di(2-ethynylvinyl) sulfide (45) and thiophene (46),12, 136 ± 138 which is the product of cyclisation of the anion 44 or the corresponding thiol.Na2S, H2O HC C C CH S7 7NaOH 44 HC C C CH S 45 H2O S 46 Earlier, an attempt was made 139 to synthesise thiophene from diacetylene by reaction of hydrated sodium sulfide with diacety- lene in aqueous ethanol at pH 8 ± 10; however, in this case the yield of the target product did not exceed 20%. When this reaction was carried out in a superbasic medium (KOH± DMSO), thiophene was obtained in high yields (up to 94%) and with high selectiv- ity.12, 136 The sulfide 45 is not formed under these conditions. If the reaction is carried out in dimethyl sulfoxide, the addition of an alkali is unnecessary, since it is liberated in this reaction, but the yield of thiophene does not exceed 55%.The effect of reaction conditions on the yield of thiophene is demonstrated in Table 2. Table 2. The effect of conditions of the reaction of diacetylene with sodium sulfide on the yield of thiophene (solvent, DMSO). Concentration of reagents/mol Time /min T/8C Yield of thiophene (%) C4H2 Na2S .9H2O 44.1 93.9 54.9 40.0 0.108 0.083 0.061 0.14 0.054 0.038 a 0.120 0.072 90 75 90 75 20 55 70 100 a In the presence of 0.08 mol of KOH. It should be emphasised that the addition of alkali is only necessary to trigger the synthesis, since the alkali liberated accel- erates further the reaction. High selectivity of this reaction is worthy of note, viz., the purity of raw thiophene distilled from the reaction mixture without additional purification reaches 99.9%.Since this reaction represents nucleophilic addition of the sulfide ion to diacetylene with subsequent cyclisation of the intermediate product,140 its rate should be higher in the solvents which do not suppress the activity of anions owing to their solvation. By virtue of their high dielectric permittivity, aprotic dipolar solvents (e.g., DMSO, hexamethylphosphoric triamide) enhance electrolytic dissociation of sodium sulfide and sharply increase the concentration of the weakly solvated sulfide ion.140 N-Methylpyrrolidone was also used as a solvent, and the reaction was carried out under identical conditions (55 8C, equimolar amounts of KOH and sodium sulfide).At 70 8C, the reaction of diacetylene with sodium sulfide carried out in an aqueous medium in the presence of an alkali gives thiophene in 52.5% yield.12 The conditions of the reaction of diacetylene with the sulfide ion giving exclusively the sulfide 45 (yields up to 89.5%) have been found.140 In this reaction, liquid ammonia is used as a solvent and I A Maretina, B A Trofimov sulfide ions are generated from ammonium sulfide formed in situ from ammonia and hydrogen sulfide: NH3 (liquid) HC C C CH+(NH4)2S S 45 The product 45 has Z,Z-configuration, which points to high trans-stereoselectivity of this reaction.140 Among reactions of diacetylene with sulfur-containing com- pounds, synthesis of thiophene undoubtedly presents the greatest interest from the practical point of view.This stimulated the development of a procedure for continuous synthesis of thiophene from diacetylene and sodium sulfide in aqueous-alkaline media12 with continuous removal of thiophene and replenishment of the reaction mixture with the starting reagents. Circulation of diac- etylene lasted for 6 h at 70 8C; the yield of thiophene of 99% purity exceeded 70%. The results were reproducible. An increase in temperature caused a dramatic decrease in the yield of thio- phene. It should be noted that polymers formed from diacetylene in these media are not explosive. However, the synthesis performed in DMSO is more efficient, since it results in thiophene of high purity (99.9%) and in 94% yield.136 The obvious advantages of this method include its simplicity, one-step procedure and waste- lessness as well as high yields and high purity of the final product.Presumably, diacetylene present in waste gases from acetylene- manufacturing plants can also be used in large-scale productions, since the activities of the accompanying products (e.g., methyl- acetylene, vinylacetylene) in the reaction with sulfide ions is much lower.12 2. Reactions with thiols The addition of thiols to diacetylenes has been carried out under diverse conditions.3 In the presence of catalytic amounts of an alkali, thiols are easily added to one of the triple bonds of diacetylene to form 1-organylthiobut-1-en-3-ynes 47.3 This reaction is neither accelerated by free-radical initiators nor is decelerated by antioxidants, which suggests its heterolytic mechanism.3 Like many other ionic reactions of nucleophilic reagents with alkynes, it occurs as trans-addition and results in the Z-isomers 47. However, this reaction is complicated by the addition of thiol to the sulfide 47 and the formation of 1,4-bis(organylthio)buta-1,3-dienes 48, which significantly dimin- ishes the yield of the primary adduct 47.RSH, HO7 SR RSH HC C C CH RS SR 47 48 R=Alk, Ar, Het. The in situ generation of the thiolate anion from S-alkyliso- thiouronium salts 141 or thioacetates 142 under the action of alkalis prevents competitive free-radical addition of the free thiol thus reacting to sulfides 47 in high yields.The reaction of diacetylene with thioacetates is carried out in the presence of KOH (1.2 moles per mole of the starting acetate) with methanol as a solvent. This reaction gives sulfides 47 in up to 80% yields.141, 142 Me SR MeOH RSK HC C C CH SR KOH O 47a ± d R=Et (a), Prn (b), Pri (c), Bun (d). The IR spectra of sulfides 47a ± d contain characteristic absorption bands at 3290 (:CH), 2100 (C:C) and 1560 cm71 (C=C7S). In 1H NMR spectra, the spin-spin coupling constant of vicinal protons (JHH) is 10 Hz, i.e., it is characteristic of the cis-arrangement of vinylic protons. The use of liquid ammonia or of aqueous ammonia for nucleophilic thiylation of 1,3-diynes opens up new prospects forDiacetylene: a candidate for industrially important reactions the synthesis of unsaturated sulfides and practical utilisation of diacetylene in those technological processes where liquid ammo- nia is used as a selective solvent.36 The reaction of diacetylene with thiols was studied 143 both in liquid ammonia and in aqueous ammonia at concentrations of ammonia varying from 25% to 70% within the temperature range of 733 to +20 8C.The aim of these studies was to reach the maximum use of the high solubilising capacity of ammonia with respect to diacetylene, which is determined by its moderate dielectric permittivity and polarity as well as by its relatively high basicity and ability to form hydrogen bonds and ammonium salts with thiols, which exist in ammoniacal solutions as contact ionic pairs and free ions.These factors altogether enhance the reactivity of thiolate ions, which are weakly solvated by ammonia. In fact, the reaction of diacetylene with thiols in liquid ammonia occurs under mild conditions (733 8C) and is completed immediately after mixing of the reactants. Irrespective of the diacetylene : thiol molar ratio (1.0 : 0.5, 1 : 1, 1.0 : 1.2, 1 : 2), the selectivity of the reaction is not changed and the reaction results exclusively in sulfides 47 in 84% to 98% yields. This suggests that the secondary reaction leading to disulfides 48 is completely excluded in this case.143 RS HC C C CH RSH, NH3 SR SR 48 47a,d ± j R=Et (a), Bun (d), n-C6H13 (e), (CH2)2OH (f), CH2CH(OH)CH2OH (g), CH2Si(OMe)3 (h), (CH2)2Si(OMe)3 (i), CH2Si(OEt)3 (j).In this reaction, ammonia activates both the thiolate ion and diacetylene due to the formation of hydrogen bonds. The violation of the stereospecificity of the addition of thiols to diacetylene in liquid ammonia is attributed to the formation of diacetylene ± ammonium complexes.143 The adducts 47a,d ± j represent mix- tures of E- and Z-isomers; the spin-spin coupling constants for the protons of E- and Z-ethylene fragments in the 1H NMR spectra have characteristic values (15 and 10 Hz, respectively). The proportion of the Z-isomer increases with an increase in the bulk of the radical R, reaching a maximum (30%) in the case of compounds 47h ± j, i.e., at the highest branching in the immediate vicinity of the sulfur atom.This suggests a significant contribution of spatial effects to the stereochemistry of this reaction, which seems to be predominantly kinetic, for the ratio of E- and Z-isomers is not changed after thermal treatment of the reaction products. High reactivity of diacetylene with respect to thiols in liquid ammonia 143 permits, in principle, its quantitative isolation from gas mixtures in the form of sulfides 47. Mild conditions of this reaction allow the involvement of thiols containing various func- tional groups. 3. The reaction with dithiocarbamate ions Reactions of diacetylene with carbon disulfide and secondary amines or its direct reaction with ammonium dithiocarbamate in protic or aprotic solvents represent an efficient utilisation of diacetylene as a valuable product.144 This reaction proceeds very easily in the absence of a catalyst when gaseous diacetylene is passed through a solution of a mixture of a secondary amine and carbon disulfide in MeOH, EtOH, Et3N, THF, DMSO, acetone, benzene or ether or through a solution of dithiocarbamate at 20 ± 50 8C and results in s-(Z)-(but-1-en-3-ynyl) N,N-dialkyldi- thiocarbamates 49 and 1,4-bis-(N,N-dialkylthiocarbamoylthio)- buta-1,3-dienes 50 in up to 94% yields.S CS2, R1R2NH CS2, R3R4NH HC C C CH S NR1R2 49a ± f 599 S S S S R1R2N NR3R4 50a,b 49: R1=R2=Et (a), Prn (b); R1±R2=(CH2)5 (c), (CH2)2O(CH2)2 (d); R1=R2 = H (e), R1=H,R2=Me (f); 50: R1=R2=R3=R4=Et (a); R1=R2=Et, R3±R4=(CH2)5 (b). Mild conditions of this reaction are determined by high nucleophilicities of dithiocarbamate ions.The reaction products are Z-isomers.144 However, the reaction of diacetylene with ammonium N,N-diethyldithiocarbamate is not stopped at the addition of one mole of dithiocarbamic acid and is complicated by further conversions resulting in the formation of bis(dithio- carbamate) 50. Dithiocarbamates are used as nematocides and flotation agents 145, 146 as well as in the production of photo- sensitive materials. The information about the use of alkynes in the synthesis of such esters is scarce.147 VI. Reactions with ketoximes and amidoximes It might be expected that the extension of the reaction of ketoximes with alkynes, which results in pyrroles (the Trofimov reaction),13, 15, 16, 22, 29, 90, 93 to diacetylene 13 would open a route to the synthesis of ethynylpyrroles of the type 51 which could be regarded as promising monomers and reagents in fine organic synthesis. However, the route resulting in O-adducts, e.g., O-(but- 1-en-3-ynyl)ketoximes 52, proved to be preferable due to the high reactivity of triple bonds of diacetylene with respect to nucleo- philic reagents.13, 148, 149 R1 KOH±DMSO±H2O N HC C C CH + CH2R2 HO R2 R1 NH 51a,b R1 CH2R2 O N 52a,b R1=Me, R2 = H (a); R1±R2=(CH2)4 (b).In aqueous DMSO, diacetylene exothermally adds to ketox- imes in the presence of catalytic amounts of alkalis resulting in vinylketoximes 52.148, 149 It was of interest to find out whether more drastic conditions, which are normally used in the synthesis of pyrroles from ketoximes and acetylene, would result in cyclisa- tion of vinylketoximes 52 through a [3,3]-sigmatropic shift char- acteristic of simple O-vinylketoximes.13 However, under different reaction conditions diacetylene reacted with ketoximes to give either vinylketoximes 52 or black-coloured insoluble polymers formed upon more profound transformations of diacetylene itself and its adducts.13 Ethynylpyrroles 51 were not found.The best yields of the adducts 52 (36% ± 41%) were obtained in the presence of *10%± 30% of water in DMSO and at a concen- tration of KOH 1%± 2%. A decrease in the reagent's concen- tration increased the yield of the adduct, since polymerisation is inhibited under these conditions.An increase in the alkali concen- tration has a negative effect on the reaction yields. In neither case were bisadducts formed. The reaction of ketoximes with diacety- lene is stereospecific and occurs as trans-addition. According to the 1H NMRspectroscopic data, the adducts formed have Z-con- figuration (JHH=6.5 ± 6.7 Hz), which corresponds to the nucleo- philic addition to the triple bond with concerted attachment of a600 proton from the reaction medium. Vinylketoximes 52 are rather unstable mobile fluids, these do not withstand heating up to 70 ± 90 8C, being rapidly and quantitatively converted into a black insoluble powder even in solution. Thermolysis products largely contained polymers formed with involvement of the triple bonds.Pyrroles were not detected, either. The reaction of diacetylene with amidoximes in the presence of KOH in aqueous DMSO gives O-adducts 53. Their rearrange- ment into the corresponding ethynylimidazoles was not observed.150 R KOH±DMSO±H2O N NH2 HC C C CH +HOR NH2 O N 53 R=Me, Ph. VII. Reactions with hydrazines and semicarbazide In the syntheses of heterocyclic systems, 1,3- and 1,4-dicarbonyl compounds are widely used.151 The role of such C3- and C4-synthons can be played by 1,3-diynes.12, 13, 16, 17, 38 ± 40 Three independent groups of investigators reported simulta- neously on the synthesis of pyrazole derivatives by addition of hydrazines to diacetylene and its mono- and disubstituted homo- logues.38, 39, 49, 56, 152 ± 155 Apparently, hydrazine reacts with di- acetylene as a primary amine to give an intermediate 54; further intramolecular attack of the second amino group at the C(3) atom results in the formation of the pyrazole ring 55.48 NH2NH2 HC C C CH HNNH2 54 HN N Me Me N 55 NH Z-Configuration of the adduct favours the cyclisation. The condensation of methyldiacetylene with hydrazine hydrate gives 3,5-dimethylpyrazole together with 3(5)-ethylpyrazole,155 which confirms the feasibility of another type of attack at the conjugated diyne system.In anhydrous media, the reactions of diacetylene with monosubstituted hydrazines result in 1,3-dialkylpyrazoles 56, whereas 1,5-dialkylpyrazoles 57 are formed in the presence of water.The reaction of diacetylene with methylhydrazine in an ethanolic solution gives a mixture of pyrazoles 56a and 57a (4 : 1), which corresponds to the predominant attack of hydrazine at the carbon atom bearing the hydrogen atom.156 Me HC C C CH RNHNH2 N R + N R Me N 56a,b N 57a,b R=Me (a), Et (b). Pyrazole 55 has become accessible due to the reaction of diacetylene with hydrazine.152, 154, 155 Direct synthesis of pyrazoles seems to be one of possible clues to the practical solution of the problem of diacetylene utilisation;155 the resulting pyrazoles with various substituents are used as starting compounds in the syn- thesis of C- and N-vinylpyrazoles.155 Syntheses of vinylpyrazoles and their homo- and copolymer- isation acquired further development.157 ± 159 Syntheses of bio- logical additives to polymeric materials based on pyrazoles I A Maretina, B A Trofimov prepared from diacetylene were also described.157, 158 The syn- thesis of C-vinylpyrazoles from pyrazole 55 was carried out.159 Vinyl derivatives of some azoles, including pyrazole, were used in various branches of industry, such as production of plastics and synthetic fibres, radio engineering and medicine.159 Some novel herbicides,160 ± 162 pesticides,163,164 antidiabetic drugs,165 immu- nostimulators 166 and perfume compositions 167 prepared from pyrazoles 55 ± 57 have been covered by patents.Synthesis of polychelating agents of the pyrazole series designed for solid- ification of epoxide resins was described.157 The possibilities of extractive isolation of noble metals (Pd, Au, Ag) with pyrazole 55 hydrochlorides,168, 169 of preparation of complex-forming ion- exchange resins,170, 171 absorbents for concentration of platinum metals,172 nitration catalysts 173 and disproportionation catalysts in the synthesis of monoalkyl- and monoaryl-halogenosilanes 174 were demonstrated.The reaction of diacetylene with semicarbazide, which mani- fests much weaker basic properties than hydrazines, yields a mixture of 3- and 5-alkylpyrazolecarboxamides (58, 59).175 The amides 58, which are irreversibly isomerised into the adducts 59 under these conditions, are the main reaction products.175 Other 1,3-diynes produce exclusively amides of the type 59.175 CH2R H2NNHCONH2 N CONH2 RC C C CH N 58a ± d H+ NH N CONH2 N N RH2C RH2C 59a ± d 60a ± d R = H (a), Me (b), Et (c), Prn (d).The presence of a pyrazole ring was confirmed 175 by acid hydrolysis of the adducts 59, which was accompanied by decar- boxylation and formation of pyrazoles 60. Synthesis of 1-guanyl- 3-methylpyrazole (yield 86%) from pyrazole 55 was described.176 Pyrazole-1-carboxylic acid derivatives are used as pesticides 177 and insecticides.178 Functionalised pyridines were prepared based on diacetylene and compounds containing a weakly nucleophilic nitrogen atom and an active CH-component (e.g., acetylacetoneimine and ethyl b-aminocrotonate).176 ± 178 The reaction products are 3-acetyl-2,4- dimethylpyridine (61) and ethyl 2,4-dimethylnicotinate (62), respectively.O Me R Me Na R HC C C CH+ NH2 O N Me 61, 62 R=Me (61), OEt (62). Since the nucleophilicity of the nitrogen atom in the starting enamino ester and the enamino ketone is lowered, the reaction with diacetylene occurs in the presence of metallic sodium. The formation of one of two possible isomers in both reactions was confirmed by 1H NMR and IR spectroscopy as well as by chromatography. Moreover, pyridine 62 was obtained independ- ently by oxidation of ethyl 2,4-dimethyl-1,4-dihydronicotinate synthesised from crotonaldehyde and ethyl b-aminocrotonate.178 The IR spectra of compounds 61 and 62 contain bands corre- sponding to stretching vibrations of the pyridine ring (1560, 1580 cm71), ester (1076 cm71) and carbonyl (1721 cm71) groups.176, 178 The protons of the pyridine ring in the 1H NMR spectrum of compound 62 resonate as two doublets in the low- field region at d 8.24 and 6.84 (J=5.2 Hz).The former belongs to the proton at the C(6) atom, which is especially stronglyDiacetylene: a candidate for industrially important reactions deshielded by the nitrogen atom of the pyridine ring, while the latter corresponds to the proton at the C(5) atom. Pyridines prepared from diacetylene are candidates for the synthesis of nicotinic acid. However, the above-described reaction is inefficient because of low yields of the target products. A promising procedure for large-scale synthesis of nicotinic acid was developed 37, 130 on the basis of 4-aminobut-3-en-2-one (63) obtained by transamination of the ether 25b.58 Aminobutenone 63 cyclises spontaneously into 5-acetyl-2-methylpyridine (64),179 which is further oxidised to pyridine-2,5-dicarboxylic (isocincho- meronic) acid (65), which in turn is decarboxylated into nicotinic acid (66) at 200 8C.37 Me O NH3 O [O] Me OEt 7NH3 N Me 25b 64 H2N 63 O O OH OH D HO N N O 65 66 VIII.Reactions with guanidine The reaction of diacetylene and monosubstituted 1,3-diynes with guanidine in the presence of equimolar amounts of metallic sodium yields 4-alkyl-2-aminopyrimidines 67.180, 181 CH2R N RC C C CH HN C(NH2)2 NH2 N 67a ± d R = H (a), Me (b), Et (c), Prn (d).The structure of cyclisation products was established by a comparison of physical constants of isolated compounds and their picrates with those of known compounds obtained by independ- ent synthesis.180, 181 Amines of the pyrimidine series are widely used in large-scale organic synthesis, particularly, in the pharmaceutical industry.182 Some of them are intermediates in the synthesis of sulfonylamide drugs, e.g, sulfomerazine 2-(p-aminophenylsulfonylamido)-4- methylpyrimidine (68),182 which can be prepared in high yields from the acetal 32a or the ether 25a.130 NH2 OMe + H2N SO2NH 25a NH Me N SO2NH H2N N 68 Sulfomerazine (68) is less toxic and more easily accessible (one-step procedure, high yields) than the well-known drug sulfodimezine [2-(p-aminophenylsulfonylamido)-4,6-dimethyl- pyrimidine] prepared from 2-amino-4,6-dimethylpyrimidine by a six-step procedure.Sulfomerazine is manufactured by the phar- maceutical company `Polfa' and is used in the treatment of streptococcal and pneumococcal infections. Some derivatives of the pyrimidine 67a, which include protein tyrosine kinase inhib- itors,183 vitronectin receptor antagonists,184 potent psychotropic drugs,185 novel fungicides186 and perfume compositions,187, 188 were patented. 601 IX. 1-Dialkylaminobut-1-en-3-ynes as building blocks for the synthesis of heterocyclic compounds Some cyclisation reactions occur exclusively with the amines 3. Their reaction with 1,2-ethanedithiol and 1,3-propanedithiol result in 5-dialkylamino-7-methyl-5H-2,3-dihydro-1,4-dithie- pines 69a ± c and 2-dialkylamino-4-methyl-2,6,7,8-tetrahydro- 1,5-dithiocines 70a ± c, respectively.189, 190 R2N NR2 +HS(CH2)nSH 3a,b,e S(CH2)n S Me 69a ± c, 70a ± c O (3e, 69c, 70c); R=Me (3a, 69a, 70a), Et (3b, 69b, 70b); R2N= N n = 2 (69a ± c), 3 (70a ± c).The condensation of 2-aminobenzenethiol with the amines 3a,b,e is performed at room temperature in the absence of a catalyst and gives 2-dialkylamino-4-methyl-2,5-dihydro-1,5-ben- zothiazepines 71a ± c in up to 90% yields.190, 191 NR2 S NH2 NR2 + 3a,b,e SH Me NH 71a ± c O (3e, 71c). R=Me (3a, 71a), Et (3b, 71b); R2N= N High yields of the reaction products and the absence of routes to their direct synthesis from diacetylene makes this method for the preparation of thiazepines promising.1,4-Dithiepine derivatives are biologically active and can be used as tranquillisers.192, 193 1,5-Benzothiazepines manifest anti- arrhythmic activity 194 and are used as regulators of the cardio- vascular function.195, 196 Compounds 72 are tranquillisers,197 and 1,5-benzothiazepines 73 possess antifungal activity.198 Ph S S N N OPh S 72 CONR(CH2)nNMe2 73 n=2, 3. The salient advantage of 1-dialkylaminobut-1-en-3-ynes over diacetylene and its other derivatives as building blocks becomes especially obvious in the synthesis of isoxazoles. The latter cannot be obtained by the reaction of diacetylene with hydroxylamine hydrochloride.47 1-Alkoxy- and 1-alkylthiobutenynes give a mix- ture of 3- and 5-methylisoxazoles.37, 38, 47 On the other hand, 1-dialkylaminobut-1-en-3-ynes 3 give exclusively 5-methyl- isoxazole (74) in reaction with hydroxylamine (Scheme 4).37, 47 The latter yields cyanoacetone (75) upon alkaline treatment.199 Cyanoacetone reacts with phenylhydrazine to give the cyclisation product, viz., 5-amino-3-methyl-1-phenylpyrazole (76).37 The latter is subjected to Wilsmeyer ± Haack formylation to give 5-amino-3-methyl-1-phenylpyrazole-4-carbaldehyde (77),200 which is used in the construction of fused pyrazolopyridines 78 and 79 and pyrazolopyrimidines 80 (see Scheme 4).201, 202 Novel long-acting sulfonylamide drugs of the type 81 can be prepared from the isoxazole 74.202 Aminopyrazole 76 is used in the synthesis of pharmacologically active compounds (e.g., anti- depressants, anticonvulsants, analgesics, etc.),203 biologically active reagents, some key synthons,204 diazo dyes for colour offset,205 heterocyclic anionic monoazo dyes 206 and compositions for colour photography.207 ± 209602 Scheme 4 NR12Me PhNHNH2 HO7 NH2OH.HCl N Me CN O O 75 74 3a,b CHO Me Me Me O HC NH2 N N N N NH2 2HC ONH2 NH2 (POCl3) N N N N80 Ph 77 76 Ph Ph NC O R2 O PhSO2X OEt Me R3 Me Me R2 NC N NH N N N N N O S O N R3 O Ph NH Ph Ph Ph 79 81 78 R2=R3=Me, Et. X. Reactions with carbonyl compounds General methods for the synthesis of diacetylene alcohols and glycols include the reaction of sodium derivatives of 1,3-diynes with carbonyl compounds in liquid ammonia, condensation of mono- and bis(bromomagnesium) derivatives of 1,3-diynes with aldehydes and ketones in THF, the Favorsky condensation of 1,3-diynes with carbonyl compounds in the presence of KOH powder in ether or THF, oxidative condensation of a-acetylenic alcohols in the presence of copper salts and the Hodkevich ± Kadyo condensation of monoacetylenic alcohols.3 This review provides a detailed description of reactions of diacetylene with carbonyl compounds which are applicable to industrial diacetylene and result in the formation of valuable products.Since some technological schemes used in acetylene production include selective extraction of diacetylene with liquid ammonia,36 its extraction with ammoniacal and aqueous ammo- niacal solutions also seems to be efficient.A one-step procedure was developed for the synthesis of bis-tertiary diacetylenic glycols by the reaction of diacetylene with ketones in the presence of a catalytic system (ammonia ± water ± alkali).210 The use of a small excess of the ketone allows preparation of glycols 82 in nearly quantitative yields without the formation of the diacetylenic alcohols 83. R2 R1R2C O R1 HC C C CH R1R2C O C C C CH OH 83a ± e R2 R2 R1 C C C C R1 OH OH 82a ± e 82, 83: R1=R2=Me (a); R1=Me, R2=Et (b); R1±R2=(CH2)5 (c); R1=Me, R2=Bun (d); R1=Me, R2=n-C6H13 (e). A simple and efficient one-pot synthesis of tertiary alkyl- thioenyne alcohols 84 from diacetylene, ketones and thiols in liquid ammonia has been developed.211 ± 214 HC C C CH R2 R2 R3SH R1R2C O C C R1 C C C CH R1 OH OH 83a ± c,f R3S 84a ± f I A Maretina, B A Trofimov Thioalcohol 84 R3 in the thiol R3SH Diacetylenic alcohols 83 83a (R1=R2=Me) Et Prn Bun Et 84a 84b 84c 84d 84e 84f 83b (R1=Me, R2=Et) 83c (R1±R2=(CH2)5) Et 83f (R1±R2=(CH2)4) Et The alcohols 84a ± f are characterised by a broad spectrum of biological activities and exhibit interest as monomers and intermediates for fine organic synthesis and for the pharmaceut- ical industry.215 Condensation of diacetylene with ketones in the presence of catalytic amounts of an alkali (0.01 ± 0.1 moles per mole of HC:C7C:CH) in liquid ammonia with subsequent addition of a thiol was studied with the aim of developing a reliable procedure for the synthesis of alkylthioenyne alcohols 84 and glycols 85.213 This reaction is carried out with a 6%±10% diacetylene solution in liquid NH3 at the temperature of liquid ammonia and at the diacetylene : ketone : thiol molar ratio of 1.0 : 0.5 : 1.2.Alkali metal cations catalyse the condensation of diacetylene with ketones and simultaneously accelerate the reac- tion of diacetylenic alcohols 82 with thiols resulting in compounds 84 (yields 83% ± 96%).213 An excess of diacetylene (relative to the ketone) directs the reaction towards the formation of diacetylenic alcohols. Alkylthioenyne glycols 85 are formed in 89% to 95% yields at the diacetylene : ketone : thiol molar ratio of 1.0 : 2.0 : 1.3.213, 216 NaOH HC C C CH+R1R2C O R3SH 84 83 R2 R3SH R2 82 R1 C C R1 R3S OH85 OH R1=Me: R2=Me, Et; R1±R2=(CH2)4; R3=Et, Prn, Pri, Bun, Bui.The sulfides 84 can also be synthesised by the condensation of enynyl sulfides 47 with acetone under conditions of the Favorsky reaction.217 Since the presence of the hydroxy functional group often excludes the possibility of selective attack at other reaction centres, Volkova et al.211 have converted alcohols 84a,b into ethers (86b), esters (87a,b), b-cyanoethyl (88a,b), trimethylsilyl (89a,b) and glycidyl (90a,b) derivatives and acetals (91a) (Scheme 5). Scheme 5 Me Me ZR Me Me ZR OSiMe3 89a,b O(CH2)2CN 88a,b Me Me Me ZR Me ZR Me ZR Me OCOMe 87a,b OH 84a,b OMe 86b Me Me ZR Me ZR Me OEt O O 91a Me 90a,b O ZR=C:C7CH=CH7SR, where R=Et (a), Prn (b).Diacetylene: a candidate for industrially important reactions The functionalisation of the thioethers 47a ± c was described.142 These sulfides, contrary to the published data,3 enter into the Mannich reaction with secondary amines on heating in dioxane in the presence of copper(I) chloride to give 1-alkylthio- 5-dialkylaminopent-1-en-3-ynes 92a ± c.Cu+ C C R22 NCH2 +H2C O+R22 NH SR1 47a,b 92a ± c R1S 47: R1=Et (a), Prn (b); 92: R1=R2=Et (a); R1=Prn, R2=Et (b); R1=Et; R2=(CH2)2OH (c). The reaction of 2-methyl-6-phenylhexa-3,5-diyn-2-ol with alcohols and ketones 218 was used to obtain the corresponding ethynylvinyl ethers and dioxolanes which contain an enyne frag- ment in the side chain and manifest pharmacological activities.219 An unusual reaction of tertiary alkylthioenyne alcohols 84 with CO2 was carried out under a pressure of 70 ± 73 atm at 70 ± 75 8C in the presence of catalytic amounts of copper(I) salts and triethylamine without solvents.220, 221 The products of this reaction are 5-(3-alkylthioprop-2-enylidene)-4,4-dimethyl-1,3- dioxolan-2-ones 93a,b.220 Me Me CO2 C C Me Me CuX RS OH O O RS 84c,g O 93a,b R=Bun (84c, 93a), Me (84g, 93b); X =Cl, OAc.Stable 1,2-bis(1,3-dioxolan-2-on-5-ylidene)ethanes 94a ± d are formed from diacetylenic glycols 82a ± d in more than 90% yields under analogous conditions.221 R1 R1R2 R2 R2 R2 CO2 O O O O R1 C C C C R1 CuX, Et3N OH OH O O 82a,b,f,g 94a ± d X=Cl, OAc.Product 94 Glycol 82 R2 R1 Me Me Prn 94a 94b 94c 94d 82a 82b 82f 82g Me Et Prn (CH2)4 The reaction of the tertiary diacetylenic alcohol 83a or the bis- tertiary diacetylenic glycol 82a with hydrogen sulfide 222 in the presence of alkaline catalysts result in dihydrofuranthione 95. S Me H2S Me C C C CH Me Me Me OH 83a O95 Condensation of diacetylene with carbonyl compounds was carried out under conditions close to the industrial ones.223 ± 225 2,7-Dimethylocta-3,5-diyne-2,7-diol (82a) was obtained in 72% yield (acetone, 25% aqueous NH3 , 0.01 ± 0.04 mole of KOH or NaOH per mole of acetone, 10-fold excess of diacetylene).224 A decrease in the catalyst's concentration to 0.01 mole decreases the yield of glycol to 11% and simultaneously increases the yield of 2-methylhexa-3,5-diyn-2-ol (83a) from 9% to 61%.An increase in the ammonia concentration in the reaction mixture directs the reaction exclusively towards the formation of glycol 82a; its yield reaches 94%. Butylmethyl- and hexylmethyl ketones react with diacetylene in the presence of an alkali (0.3 mole per mole of ketone) to form diacetylenic alcohols 83 (yields*60%).224 It was 603 shown that the anion-exchange resin AB-17-8P (OH) catalyses the condensation of diacetylene with carbonyl compounds.226 Tetra- methylammonium hydroxide also catalyses the reaction of di- acetylene with formaldehyde.227 The reaction with formaldehyde occurs smoothly at 20 ± 40 8C and yields, in addition to hexa-2,4- diyne-1,6-diol (96), 1,3-dioxolanes 97 and 98 which are the products of subsequent reaction of formaldehyde with the diol 96 and 2,4-pentadiynol.227 H H Me4NOH H+ C C C C H HC C C CH+H2C O OH OH H 96 H OH + + O O O O 98 97 Alkylation of alcohols 83a ± d (dialkyl sulfate, 5 ± 15 8C, 2 h) was used to prepare ethers of tertiary diacetylenic alcohols 99 in up to 80% yields.228 R2 R2 (R3O)2SO2 C C C CH R1 C C C CH R1 OR3 99a ± d OH 83a,d,e Ether 99 Alcohol 83 R3 R2 R1 99a 99b 99c 99d 83a 83a 83e 83d Me Bun Me Et Me Me n-C6H13 Bun Me Me Me Me b-Diketones and acetylenic ketones were synthesised from ethers 99.229 In the presence of HgSO4 (1 h, 65 ± 70 8C), the hydration of the terminal triple bond of the ethers 99 gives ketones 100 in up to 93% yields.229 b-Diketones 101 are formed with an increase in temperature to 80 8C and with larger amounts of HgSO4 in 8 ± 10 h as a result of total hydration of both triple bonds (yields 60%± 80%).229 Me O C C R1 Me Me OR2 100 Hg2+ C C C CH R1 O O Me OR2 99 R1 Me 101 OR2 A number of papers are devoted 230 ± 237 to the synthesis of potentially physiologically active compounds based on acetylenic and diacetylenic alcohols and glycols as well as to the analysis of biological activities of reaction products and their applications in agriculture.It was shown that diacetylene and diynes containing terminal triple bonds add to the carbonyl group of 1,2,3-trimethyl- piperidin-4-one under conditions of the Favorsky reaction to yield a mixture of isomeric alcohols and glycols containing predominantly epimers 102, the products of axial addition of diynes to the C=O bond.230 ± 237 Me Me Me Me Me N N N HC C C CH Me C C C C Me OH HO Me Me 102 O A correlation was found between the biological activity and chirality of asymmetrical diacetylenic glycols.230 Acpinol, which represents a product of exhaustive hydrochlorination of a mixture of isomers of the glycol 102, was recommended for use in agriculture.236 ± 240 A method of manufacture of acpinol employs diacetylene present in the abgases from acetylene-manufacturing604 plants.The purity of the product depends on the composition of the abgas.230 The cyclohydration of alkoxyvinylacetylenic alcohols 103 ± 105 obtained from ethoxybutenyne (25b) and butoxy- butenyne (25d) 241, 242 under conditions of the Kucherov reaction or in diluteH2SO4 at room temperature results in dihydropyran-4- ones 106 ± 108.243, 244 KOH R1R2C O+ OR3 25b,d O R2 Hg2+ R1 R2 OH O R3O R1106 ± 108 103 ± 105 R1=R2=Me (103, 106); R1±R2=(CH2)5 (104, 107); R1±R2=(CH2)2NMe(CH2)2 (105, 108); R3=Et (b), Bun (d). The synthesis of 2-alkyl-2,3-dihydro-g-pyrones 109 from the ether 25a and aldehydes was described.245 Condensation of lithium methoxybutenynide with various aldehydes (from MeCHO to iso-C5H11CHO) in THF at 778 8C yields secondary alcohols which undergo hydration at the triple bond with sub- sequent cyclodehydration to give dihydropyrones 109a ± e.245 MeO MeO HO THF 30% HClO4 C C +RCHO Li C C R R O O OH O R 109a ± e R=Me (a), Et (b), Prn (c), Pri (d), iso-C5H11 (e).The yield of the reaction products increases with an increase in the length of the alkyl radical. Thus pyrones 109a and 109e are formed in 40% and 80% yields, respectively. The reaction of the lithium derivative of the ether 25a with lactones affords spiro- ketals 111.246 MeO O MeO C C BunLi O HC C 25a O OH 110O 30% HClO4 CH2Cl2 O 111 O OMe OMe O MeOH OMe OH 112 For example, the hydroxy ketone 110 was obtained from d-valerolactone in THF at 778 8C in 98% yield.Its conversion into the spiroketal 111 can be carried out in two ways,246 viz., by direct conversion under the action of 30% HClO4 inCH2Cl2 (yield 58%) or via an intermediate trimethoxy ketone 112 (yield 86%). These dihydropyrones are used as fungicides.247 A general method for the synthesis of functionalised spiro- ketals from lactones and (Z)-lithium derivatives of the ether 25a was proposed.248 I A Maretina, B A Trofimov Some novel antiinflammatory drugs and a pharmaceutically stable inhibitor of LTA4-hydrolase based on the E-isomer of the ether 25a were patented.249 These compounds are used as con- stituents of composite fuels for stabilisation of combustible mixtures.250 Ethynylation with diacetylene was used in the synthesis of additives to perfume compositions, e.g., 2,6,11,15-tetramethyl- hexadecane-6,11-diol 251 or 6,11-dihydroxycrocetane and croce- tane.252 A commercial synthesis of squalane by ethynylation of geranylacetone with a diacetylene ±N-methylpyrrolidone com- plex (1 : 1) in the presence of KOH with subsequent reduction over Raney nickel was described.253 The alkoxy- and alkylthioenyne alcohols prepared from diacetylene can be used in the synthesis of various polyfunctional unsaturated compounds.254 Hydrolytic scission of vinyl ethers 103a,b by dilute acids results in acetylenic hydroxy aldehydes 113, which are readily isomerised into oxoaldehydes 114 in up to 75% yields.254 The reaction of vinyl ethers 103a,b with ethyl vinyl ether was used to obtain acetals 115 (yield 80%).The oxidation of vinyl ethers 103a,b with chromic acid in acetone gives a mixture of aldehydes 113 and 114 in a total yield of 61% and the oxo acid 116 in 10% yield. RO Me C C Me OH 103a,b O O Me Me H2O, H+ C C Me O Me OH 114 113RO Me OEt C C Me OCH(OEt)Me 115 Me O H2CrO4 113+114+ CO2H Me 116 R=Me (a), Et (b). Reactions of tertiary alcohols 103a ± d with acetone, acryloni- trile, acetic anhydride, epichlorohydrin and glycols (Scheme 6) were studied.255 OR1 R1O Scheme 6 O Me Me Me H2C H+ C C Me C C O O Me OAc 119a,d 120 Me Me2 C O Ac2O Me117a,d R1O R1O R1O Me Me CH2=CHCN C C Me C C Me OH103a ± d O(CH2)2CN 118a,d 25a ± d R2 R3 Cl O HOCH(CH2)nCHOH 7R1OH R1O R2 Me O O Me (CH2)n C C Me O R3 Me O O 122a ± c, 123 121a ± c R1=Me (a), Et (b), Prn (c), Bun (d); R2, R3=H, Me; n=0 (122), 1 (123).Diacetylene: a candidate for industrially important reactions The reactions of the vinyl ether 103a,d with acetone result in 5-(3- alkoxyallylidene)-2,2,4,4-tetramethyl-1,3-dioxolanes 117a,d.255 Cyanoethylation of these ethers with acrylonitrile occurs in the presence of an 40% aqueous solution of KOH or MeONa in tert- butyl alcohol and results in 1-alkoxy-5-(2-cyanoethoxy)-5-methyl- hex-1-en-3-ynes 118a,d in up to 50% yields.Acetylation of the alcohols 103a,d with acetic anhydride in pyridine at 70 ± 90 8C results in 5-acetoxy-1-alkoxy-5-methylhex-1-en-3-ynes (119a,d) in 21% to 49% yields.255 During distillation, acetates undergo deacetylation and dehydration to give the enynal 120.255 The reaction of epichlorohydrin with ethers 103a ± c in an alkaline medium afforded glycidyl ethers 121a ± c in 60%± 65% yields (see Scheme 6).256 The resulting products are candidates for epoxy- containing monomers.215 The reaction of bisprimary and primary-secondary a- and b-glycols with ethers 103a,b,d in the presence of the Newland catalyst (HgO ± BF3) results in 2-(4-methyl-2-oxopent-3-enyl)- 1,3-dioxolanes 122 in up to 64% yields (see Scheme 6).257 Pre- sumably, this reaction begins with isomerisation of the alkoxy- enyne alcohol into an unstable alkoxydivinyl ketone. The latter adds a glycol molecule to give an asymmetrical hydroxy acetal, which easily loses the alcohol molecule and cyclises into 1,3- dioxolanes 122 or 1,3-dioxanes 123.257 Moderate neuroactivity of 1,3-dioxolanes and 1,3-dioxanes prepared on the basis of alkoxy- enyne tertiary alcohols was established.258 It is worth noting that the most reactive compounds of this series manifest both general depressant activity and a stimulating effect on the central adreno- and dopaminergic systems of the brain, which makes them similar to antidepressants.258 Base-catalysed thermolysis (the retro-Favorsky reaction) of 8-methoxyoct-7-ene-3,5-diyn-2-ol (124) was studied.259 In organic media, this is accompanied by different reactions the direction of which depends on the nature of the solvent.259 In benzene or THF, the thermolysis at the boiling temperature of the reaction mixture results in the formation of 1-methoxyhex-1-ene-3,5-diyne (125) in 85% yield.259 MeO MeO Me KOH, D Me HC C C C 7Me2C O C C C C OH 125 124 KOH, Me2C O Me Me CHOMe CHC CCHO CHC CHC Me Me O O O O Me 127 Me Me Me 126 The formation of 2,2,4,4-tetramethyl-5-(5-methoxypent-4-en- 2-ynylidene)-1,3-dioxolane (126) isolated from the reaction mix- ture as a side product can be attributed to the reaction of the original alcohol with acetone in the presence of KOH. In air, the product 126 is oxidised to 2,2,4,4-tetramethyl-5-(3-formylpropy- nylidene)-1,3-dioxolane (127).259 Vinyl ethers 103 react with nitrogen bases (hydrazine, meth- ylhydrazine, hydroxylamine) in acid and alkaline media with the formation of heterocyclic compounds of the pyrazole and iso- xazole series.260 Some natural thiophene derivatives (e.g., 7-thienylheptatri- enal) were synthesised from (E)-methoxybutenyne and thio- phene-2-carbaldehyde.261 MeO O MeO S Pd/H+ S HO 605 O OMe Ph3P CHCO2Me S S O S XI.Conclusion The data accumulated thus far demonstrate the possibility of utilisation of diacetylene in industrial syntheses of valuable products, reagents, starting compounds for pharmaceutical indus- try, materials for new technologies, etc. The processes aimed at acetylene production based on pyro- lysis of hydrocarbon raw materials will long be used as a source for large-tonnage synthesis of diacetylene the utilisation of which is still a challenge for chemists and also common sense, since combustion of hydrocarbon gases is most senseless in this case.As D I Mendeleev used to say, `one can fire a furnace with banknotes'.262 It has now become quite clear that diacetylene can be used for small-capacity, profitable manufacture of thiophene, vitamins A and PP, geraniol and phytol derivatives, pyrazoles, pyrimidines and pyridines. Studies of diacetylene polymerisation reactions present special interest. The first example of polyacetylene syn- thesis from a diacetylenic monomer has been described recently.263 The diacetylene present in industrial gases enters into reac- tions of nucleophilic addition selectively.In this review, it is these transformations which encompass only some of possible applica- tions of diacetylene and its derivatives that are analysed in detail. Thus copolymers of diacetylenic glycols and their derivatives are used as a basis in the preparation of stereoregulatory membranes mimicking cell liposomes 14 and thermochromic polymeric layers 264, 265 for laser image recording and observations over conformational transitions in polymers at ambient temperatures. The role of compounds produced from diacetylene will be increasing with the development of new technologies. The syn- thesis of reagents based on diacetylene can be the first step towards its commercial utilisation.This, in turn, will give an impetus to the development of new procedures for the synthesis of a vast variety of useful compounds from inexpensive industrial raw materials using reactions discussed in the present review.266 References 1. O N Temkin, G K Shestakov, Yu A Treger Atsetilen. Khimiya. Mekhanizmy Reaktsii. Tekhnologiya (Acetylene. Chemistry. Reaction Mechanisms. Technology) (Moscow: Khimiya, 1991) 2. R Tedeschi Acetylene Based Chemicals from Coal and Other Natural Resources (New York: Marcel Dekker, 1982) 3. M F Shostakovskii, A V Bogdanova Khimiya Diatsetilena (The Chemistry of Diacetylene) (Moscow: Nauka, 1971) 4. F Bohlmann, T Burkhardt, C Zdero Naturally Occurring Acetylene (New York; London: Marcel Dekker, 1973) 5.V F Kucherov, M V Mavrov, A R Derzhinskii Prirodnye Poli- atsetilenovye Soedineniya (Natural Polyacetylene Compounds) 6. V N Antonov, A S Lapidus Proizvodstvo Atsetilena (Production of (Moscow: Nauka, 1972) Acetylene) (Moscow: Khimiya, 1970) 7. HN Viehe (Ed.) Chemistry of Acetylenes (New York: Marcel Dekker, 1969)606 8. M F Shostakovskii, B A Trofimov, B V Prokop'ev, in Irkutskii Nauchnyi Tsentr SO AN SSSR, Irkutsk, 1967 (Irkutsk Research Centre of Siberian Branch of the Academy of Sciences of the USSR, Irkutsk, 1967) p. 267 9. S Patai (Ed.) The Chemistry of Functional Group. The Chemistry of the Carbon7Carbon Triple Bond (London: Wiley, 1978) 10. L Brandsma Preparative Acetylenic Chemistry (Amsterdam: Elsevier, 1988) 11.B A Trofimov Geteroatomnye Proizvodnye Atsetilena. Novye Poli- funktsional'nye Monomery, Reagenty i Poluprodukty (Heteroatomic Derivatives of Acetylene. New Polyfunctional Monomers, Reagents and Intermediate Products) (Moscow: Nauka, 1981) 12. B A Trofimov, S V Amosova Divinilsul'fid i ego Proizvodnye (Divinyl Sulfide and Its Derivatives) (Novosibirsk: Nauka, 1983) 13. B A Trofimov, A I Mikhaleva N-Vinilpirroly (N-Vinylpyrroles) (Novosibirsk: Nauka, 1984) 14. B A Trofimov, in Problemy Khimii i Metallurgii Tsentral'nogo Kazakhstana (The Problems of Chemistry and Metallurgy of Central Kazakhstan) (Alma-Ata: Nauka, 1985) p. 26 15. B A Trofimov, in Azotsoderzhashchie Geterotsikly (Nitrogen- Containing Heterocycles) (Novosibirsk: Nauka, 1987) p.7 16. B A Trofimov, in The Chemistry of Heterocyclic Compounds Vol. 48 (Ed. R A Jones) (New York: Wiley 1992) p. 131 17. B A Trofimov, B A Shainyan, in The Chemistry of Functional Groups. Supplement S: The Chemistry of Sulfur-Containing Functional Groups (Eds S Patai, Z Rappoport) (New York: Wiley, 1993) p. 659 18. M F Shostakovskii, B A Trofimov, A S Atavin , V I Lavrov Usp. Khim. 37 2070 (1968) [Russ. Chem. Rev. 37 907 (1968)] 19. B A Trofimov Usp. Khim. 50 248 (1981) [Russ. Chem. Rev. 50 138 (1981)] 20. B A Trofimov Zh. Org. Khim. 22 1991 (1986) a 21. L N Sobenina, A I Mikhaleva, B A Trofimov Usp. Khim. 58 275 (1989) [Russ. Chem. Rev. 58 163 (1989)] 22. B A Trofimov Usp. Khim. 58 1703 (1989) [Russ.Chem. Rev. 58 967 (1989)] 23. B A Trofimov Zh. Org. Khim. 31 1368 (1995) a 24. L N Sobenina, A I Mikhaleva, B A Trofimov Khim. Geterotsikl. Soedin. 291 (1989) b 25. B A Trofimov, A G Mal'kina, Yu M Skvortsov Zh. Org. Khim. 29 1268 (1993) a 26. B A Trofimov Sulfur Rep. 3 83 (1983) 27. B A Trofimov, S V Amosova Sulfur Rep. 4 323 (1984) 28. B A Trofimov Z. Chem. 26 41 (1986) 29. B A Trofimov, A I Mikhaleva Heterocycles 37 1193 (1994) 30. N A Nedolya, B A Trofimov Sulfur Rep. 15 237 (1994) 31. N A Nedolya, B A Trofimov Sulfur Rep. 15 339 (1994) 32. BATrofimov Phosphorus Sulfur Silicon Relat. Elem. 95 ± 96 145 (1994) 33. I L Kotlyarevskii, L G Karpitskaya Khimiya Atsetilena (Chemistry of Acetylene) (Tomsk: Tomsk State University, 1981) 34.I N Azerbaev Khimiya Proizvodnykh Atsetilena (Chemistry of Derivatives of Acetylene) (Alma-Ata: Nauka, 1976) 35. D Yusupov, K Akhmerov Kataliticheskaya Geterotsiklizatsiya Atsetilenovykh Soedinenii (Catalytic Heterocyclisation of Acetylene Compounds) (Tashkent: Fan, 1985) 36. I A Maretina Zh. Prikl. Khim. 73 528 (2000) c 37. I A Maretina Zh. Prikl. Khim. 69 353 (1996) c 38. I A Maretina, A E Tsil'ko, Yu A Zaichenko Usp. Khim. 50 1252 (1981) [Russ. Chem. Rev. 50 657 (1981)] 39. IAMaretina Usp. Khim. 60 103 (1991) [Russ. Chem. Rev. 60 57 (1991)] 40. V Shrot Khim. Geterotsikl. Soedin. 1443 (1985) b 41. BRD P. 871 006; Chem. Abstr. 52 19 982 (1958) 42. M Schroeder Chem. Ing. Tech. 46 395 (1974) 43. USSR P. 785 292; Byull.Izobret. (45) 95 (1980) 44. USSR P. 570 580; Byull. Izobret. (32) 64 (1977) 45. M Takami, Y Nihagava, Y Omure Heterocycles 16 146 (1981) 46. A Preiss,W Schroth Z. Chem. 17 286 (1977) 47. I A Maretina, Doctoral Thesis in Chemical Sciences, Lensovet Leningrad Technological Institute, Leningrad, 1975 48. W Schroth, J Peschel, A Zschunke Z. Chem. 9 110 (1969) 49. W Schroth, J Peschel, A Zschunke Z. Chem. 9 108 (1969) 50. A N Volkov, A N Vavilova, L V Sokolyanskaya, M V Sigalov Izv. Akad. Nauk SSSR, Ser. Khim. 687 (1982) d 51. A A Ul'yanov, I A Maretina. Zh. Org. Khim. 7 2052 (1971) a I A Maretina, B A Trofimov 52. I A Maretina, G V Asratyan, I G Ostroumov Khim. Geterotsikl. Soedin. 867 (1992) b 53. A N Volkov, L V Sokolyanskaya, R Ya Mushii Zh.Org. Khim. 9 1599 (1973) a 54. W Schroth Z. Chem. 22 347 (1982) 55. B A Trofimov, A N Vavilova Zh. Org. Khim. 22 475 (1986) a 56. W Schroth, J Peschel, A Zschunke Z. Chem. 9 143 (1969) 57. WSchroth, A Preiss,MBethin, W-D Bartsch Z. Chem. 17 216 (1977) 58. N K Markova, Yu A Zaichenko, A E Tsil'ko, I A Maretina Zh. Org. Khim. 20 962 (1984) a 59. I G Ostroumov, Doctoral Thesis in Chemical Sciences, Saratov State University, Saratov, 1999 60. N A Pogorzhel'skaya, I A Maretina, A A Petrov Zh. Org. Khim. 5 223 (1969) a 61. V A Engelhardt. J. Am. Chem. Soc. 78 107 (1956) 62. L B Krivdin, A G Proidakov, I G Ostroumov, I A Maretina, V V Shcherbakov, G A Kalabin Zh. Org. Khim. 25 698 (1989) a 63. I G Ostroumov, A E Tsil'ko, I A Maretina, A A Petrov Zh. Org.Khim. 23 1635 (1987) a 64. I G Ostroumov, E Yu Khakunova, A E Tsil'ko, I A Maretina, A A Petrov Zh. Org. Khim. 26 2508 (1990) a 65. L N Sukhova, I G Ostroumov, I A Maretina Khim. Geterotsikl. Soedin. 1409 (1992) b 66. L N Sukhova, I G Ostroumov, I A Maretina Zh. Org. Khim. 30 51 (1994) a 67. Ya F Freimanis, in Khimiya Enaminoketonov, Enaminoefirov, Enaminotionov (The Chemistry of Enamino ketones, Enaminoesters and Enaminothions) (Riga: Zinatne, 1974) p. 107 68. USSR P. 1 330 133; Byull. Izobret. (30) 108 (1987) 69. I G Ostroumov, A E Tsil'ko, I A Maretina, A A Petrov Zh. Org. Khim. 24 1165 (1988) a 70. WO 9828269 PCT; Chem. Abstr. 129 109 090 (1998) 71. BRD P. 19 653 355; Chem. Abstr. 129 81 731 (1998) 72.USSR P. 910 626; Byull. Izobret. (9) 85 (1982) 73. S E Tolchinskii,M V Kormer, I A Maretina,A A Petrov Zh. Org. Khim. 23 2097 (1987) a 74. E A Sokolova, I A Maretina, A A Petrov Zh. Org. Khim. 20 1800 (1984) a 75. USSR P. 998 464; Byull. Izobret. (7) 157 (1983) 76. USSR P. 1 081 163; Byull. Izobret. (11) 81 (1984) 77. USSR P. 1 362 732; Byull. Izobret. (48) 71 (1987) 78. I G Ostroumov,A E Tsil'ko, I A Maretina Zh. Org. Khim. 24 2321 (1988) a 79. WO 9 827 092 PCT; Chem. Abstr. 129 95 500 (1998) 80. WO 9 801 429 PCT; Chem. Abstr.128102 090 (1998) 81. WO 9 422 845 PCT; Chem. Abst. 122 239 686 (1995) 82. WO 9 427 983 PCT; Chem. Abstr. 122 239 689 (1995) 83. Eur. P. 625 513; Chem. Abstr. 122 238 687 (1995) 84. BRD P. 3 207 001; Chem. Abstr.98 9999 (1983) 85. WO 9 424 151 PCT; Chem. Abstr. 122 239 728 (1995) 86. WO 9 749 690 PCT; Chem. Abstr. 128 102 106 (1998) 87. WO 9 748 701 PCT; Chem. Abstr. 128 102 107 (1998) 88. WO 9 629 316 PCT; Chem. Abstr. 126 8143 (1997) 89. V M Gurav, S V Agorkar. J. Indian Chem., Sect. B, Org. Chem. Incl. Med. Chem. 37B 161 (1998); Chem. Abstr. 129 95 479 (1998) 90. B A Trofimov, A I Mikhaleva Khim. Geterotsikl. Soedin. 1299 (1980) b 91. B A Trofimov, A I Mikhaleva, L V Morozova Usp. Khim. 54 1034 (1985) [Russ. Chem. Rev. 54 609 (1985)] 92. B A Trofimov, A I Mikhaleva Zh. Org. Khim. 32 1127 (1996) a 93. B A Trofimov Adv. Heterocycl. Chem. 51 177 (1990) 94. B A Trofimov Sulfur Rep. 12 207 (1992) 95. A J Chalk Tetrahedron Lett. 3487 (1972) 96.Yu A Zaichenko, L N Reshetova, I A Maretina Zh. Org. Khim. 7 2461 (1971) a 97. A A Ul'yanov, I A Maretina Zh. Org. Khim. 7 1560 (1971) a 98. L N Reshetova, Candidate Thesis in Chemical Sciences, Lensovet Leningrad Technological Institute, Leningrad, 1974 99. USSR P. 1 095 907; Byull. Izobret. (21) 5 (1984) 100. A N Volkov, A N Khudyakova, V V Keiko, G N Mazneva Zh. Org. Khim. 10 953 (1974) a 101. A N Volkov, L V Sokolyanskaya, E O Tsetlina, R Ya Mushii Zh. Org. Khim. 9 64 (1973)Diacetylene: a candidate for industrially important reactions 102. B A Trofimov, S V Amosova, A I Mikhaleva, N K Gusarova, E P Vyalykh, in Fundamental'nye Issledovaniya. Khimicheskie Nauki (Fundamental Research. Chemical Sciences) (Novosibirsk: Nauka, Siberian Branch of Academy of Sciences of the USSR, 1977) p.174 103. W Franke, R Kraft, D Tietjen, H Weber Berichte 86 793 (1953) 104. US P. 4 588 521; Chem. Abstr. 105 102 359 (1986) 105. L Claisen, N Stylos Berichte 21 1144 (1888) 106. W Franke, R Kraft Angew. Chem. 67 395 (1955) 107. J Dabrowsky, M Tencer Tetrahedron 32 587 (1973) 108. E Taskinen, H Kalva Finn. Chem. Lett., 208 (1975) 109. A N Khudyakova, A N Volkov, B A Trofimov, in Reaktsionnaya Sposobnost' Organicheskikh Soedinenii (Reactivity of Organic Compounds) (Tartu: Tartu State University, 1973) Vol. 10, p. 987 110. A N Khudyakova, A N Volkov, B A Trofimov Zh. Org. Khim. 12 1384 (1976) a 111. A E Tsil'ko, I A Maretina Zh. Org. Khim. 10 929 (1974) a 112. A E Tsil'ko, I A Maretina Zh.Org. Khim. 10 447 (1974) a 113. A E Tsil'ko, L N Reshetova, I A Maretina Zh. Org. Khim. 9 655 (1973) a 114. A N Belyaeva, I A Maretina, E A Lisitsyn Zh. Org. Khim. 6 1532 (1970) a 115. Yu A Zaichenko, M V Kormer, A I Vil'davskaya, K B Rall', I A Maretina Zh. Org. Khim., 19 923 (1983) a 116. W Franke, J Kuther. Fette ± Seifen ± Anstrichmittel 82 82 (1980) 117. W Franke Fette ± Seifen ± Anstrichmittel, 82 370 (1980) 118. R Gripp, in Methoden der Organischen Chemie (Houben Weyl) Vol. VII/2e (Stuttgart: Georg Time Verlag, 1977) p. 2481 119. M I Rybinskaya, A N Nesmeyanov, N K Kochetkov Usp. Khim. 38 961 (1969) [Russ. Chem. Rev. 38 433 (1969)] 120. H Kaufmann,WStamm. Fette ± Seifen ± Anstrichmittel 59 946 (1977) 121. S Hmnig, G Wehner Chem.Ber. 113 302 (1980) 122. N V Korol'kov, L N Reshetova, A E Tsil'ko, A N Belyaeva, I A Maretina, A A Petrov, in Tez. Dokl. V Vsesoyuz. Konf. po Khimii Atsetilena (Abstracts of Reports of the Vth All-Union Conference on Chemistry of Acetylene) (Tbilisi: Metsniereba, 1975) p. 477 123. V N Stepanova, Candidate Thesis in Chemical Sciences, Lensovet Leningrad Technological Institute, Leningrad, 1980 124. V N Stepanova, G K Babalov, A A Zuev, G I Kudirov, L P Romashkina, A A Petrov, A E Tsil'ko Khim. Promst (3) 145 (1984) 125. V M Antonenkov, V N Stepanova Azotn. Promst (2) 10 (1977) 126. V N Stepanova, A E Tsil'ko, I A Maretina, S A Grigor'eva Zh. Prikl. Khim. 57 1801 (1984) c 127. V N Stepanova, A E Tsil'ko, I A Maretina Azot. Promst (2) 63 (1977) 128.E N Polivoda, V M Kharchevnikov, A E Tsil'ko Kauchuk Rezina (10) 31 (1986) 129. USSR P. 768 788; Byull. Iobret. (37) 87 (1980) 130. I A Maretina, in Issledovanie v Oblasti Khimii i Tekhnologii Pere- rabotki Goryuchikh Iskopaemykh (Research in the Field of Chemistry and Technology of Processing of Combustible Minerals) (Leningrad: Lensovet Leningrad Technological Institute, 1979) p. 40 131. Jpn. P. 06 306 044; Chem. Abstr. 122 291 217 (1995) 132. Br. P. 2 281 295; Chem. Abstr. 122 265 397 (1995) 133. WO 9 828 300 PCT; Chem. Abstr. 129 95 506 (1998) 134. Jpn. P. 5 915 013; Ref. Zh. Khim. 24 N 257P (1984) 135. Catalogue Handbook of Fine Chemicals (Aldrich) Aldrich Chemical Company Milwaukee, WI, 1997, p. 84; 928; 942;1025; 1071; 1280; 1478; 3097 136.BRD P. 2 818 580; Chem. Abstr. 92 94 228 (1980) 137. M G Voronkov, B A Trofimov, V V Kryuchkov, Yu M Skvortsov, S V Amosova, A N Volkov, A G Mal'kina, R Ya Mushii Khim. Geterotsikl. Soedin. 1694 (1981) b 138. V V Kryuchkov, Candidate Thesis in Chemical Sciences, Irkutsk State University, Irkutsk, 1976 139. K E Schulte, J Reisch, L Hlrner Chem. Ber. 95 1943 (1962) 140. A N Volkov, R N Kudyakova, B A Trofimov Zh. Org. Khim. 15 1554 (1979) a 141. USSR P. 417 418; Byul. Izobret. (8) 68 (1974) 142. R N Kudyakova, A N Volkov Zh. Org. Khim. 13 934 (1977) a 607 143. A N Volkov, K A Volkova, E P Levanova, B A Trofimov Izv. Akad. Nauk SSSR, Ser. Khim. 831 (1981) b 144. L V Sokolyanskaya, A N Vavilova, A N Volkov, B A Trofimov Zh.Org. Khim. 17 2269 (1981) a 145. N N Mel'nikov Khimiya Pestitsidov (The Chemistry of Pesticides) (Moscow: Khimiya, 1968) 146. US P. 3 787 471; Chem. Abstr. 80 82 142 (1974) 147. USSR P. 556 137; Byul. Izobret. (16) 66 (1977) 148. L V Sokolyanskaya, A N Volkov, B A Trofimov Zh. Org. Khim. 12 905 (1976) a 149. B A Trofimov, L V Sokolyanskaya, A N Volkov, A I Mikhaleva Izv. Akad. Nauk SSSR, Ser. Khim. 2803 (1980) d 150. A N Volkov, L V Sokolyanskaya, B A Trofimov Izv. Akad. Nauk SSSR, Ser. Khim. 1430 (1976) d 151. A Quilico The Chemistry of Heterocyclyc Compounds Vol. 17 (Eds A Weisberger, E Teilor) (New York: Interscience, 1962) p. 7 152. USSR P. 237 905; Byul. Izobret. (9) 28 (1969) 153. W W Paudler, A G Zeiler J.Org. Chem. 34 999 (1969) 154. E G Darbinyan, S G Matsoyan Arm. Khim. Zh. 23 610 (1970) 155. E G Darbinyan, Doctoral Thesis in Chemical Sciences, Institute of Organic Chemistry, Academy of Sciences of Arm. SSR, Erevan, 1981 156. A N Belyaeva, I A Maretina Zh. Org. Khim. 9 832 (1973) a 157. USSR P. 432 144; Byul. Izobret. (22) 71 (1974) 158. USSR P. 241 447; Byul. Izobret. (14) 24 (1969) 159. USSR P. 602 512; Byul. Izobret. (14) 86 (1978) 160. Jpn. P. 0 625 177; Chem. Abstr. 121 83 326 (1994) 161. WO 9 631 507 PCT; Chem. Abstr. 126 7984 (1997) 162. WO 9 829 392 PCT; Chem. Abstr. 129 95 490 (1998) 163. WO 9 828 279 PCT; Chem. Abstr. 129 95 489 (1998) 164. WO 9 828 277 PCT; Chem. Abstr. 129 109 087 (1998) 165. WO 9 825 907 PCT; Chem.Abst. 129 81 725 (1998) 166. Jpn. P. 0 702 786; Chem. Abstr. 122 239 697 (1995) 167. BRD P. 4 314 317; Chem. Abstr. 122 38 518 (1995) 168. V A Pronin, M V Usol'tseva, Z N Shastina, A N Volkov, V N Seraya Zh. Neorg. Khim. 19 800 (1976) e 169. V A Pronin, M V Usol'tseva, Z N Shastina, A N Volkov, L V Sokolyanskaya Zh. Anal. Khim. 31 1767 (1976) f 170. WO 9 511 319 PCT; Chem. Abstr. 123 14 301 (1995) 171. Russ. P. 2 011 657; Byull. Izobret. (8) 69 (1994) 172. N I Shcherbinina, G V Myasoedova, S S Kolobov, T V Druzhinina, A S Aleksandriiskii, L A Nazar'ina Zh. Anal. Khim. 50 795 (1995) f 173. BRD P. 4 237 688; Chem. Abstr. 121 194 346 (1994) 174. Eur. P. 610 809; Chem. Abstr. 121 256 035 (1994) 175. Yu A Zaichenko, I A Maretina, A A Petrov Zh.Org. Khim. 8 2605 (1972) a 176. Yu A Zaichenko, I A Maretina, A A Petrov Zh. Org. Khim. 8 1328 (1972) a 177. Yu A Zaichenko, I A Maretina, A A Petrov Zh. Org. Khim. 11 708 (1975) a 178. Yu A Zaichenko, Candidate Thesis in Chemical Sciences, Lensovet Leningrad Technological Institute, Leningrad, 1975 179. USSR P. 1 122 652; Byull. Izobret. (41) 86 (1984) 180. M A Kirillova, I A Maretina, A A Petrov Zh. Org. Khim. 6 1347 (1970) a 181. MA Kirillova, I A Maretina, A A Petrov Zh. Org. Khim. 7 14 (1971) a 182. M D Mashkovskii, in Lekarstvennye Sredstva (Medicinals) (Khar'kov: Torsing, 1997) p. 361 183. WO 9 828 281 PCT; Chem. Abstr. 129 95 505 (1998) 184. BRD P. 19 653 646; Chem. Abstr. 129 95 606 (1998) 185. WO 9 631 488 PCT; Chem.Abstr. 126 8135 (1997) 186. WO 9 749 696 PCT; Chem. Abstr. 128 102 097 (1998) 187. Eur. P. 663 204; Chem. Abstr. 123 179 083 (1995) 188. BRD P. 4 400 757; Chem. Abstr. 123 179 084 (1995) 189. USSR P. 1 109 401; Byull. Izobret. (31) 67 (1984) 190. E A Sokolova, I A Maretina, A A Petrov Zh. Org. Khim. 24 88 (1988) a 191. USSR P. 1 130 565; Byull. Izobret. (47) 81 (1984) 192. BRD P. 2 707 072; Chem. Abstr. 87 184 560 (1977) 193. US P. 4 109 006; Chem. Abstr. 90 87 477 (1979) 194. L H Werner, S Ricca, A Rossi, G deStevens J. Med. Chem. 10 575 (1967)608 195. W Ried, R Teubner Liebigs Ann. Chem. 741 (1978) 196. T Nagao,M Sato, H Nakajima, A Kiyomoto Chem. Pharm. Bull. 21 92 (1973) 197. US P. 3 748 321; Chem. Abstr. 79 92 303 (1973) 198.US P. 3 911 126; Chem. Abstr. 84 155 700 (1976) 199. B Gonzalez, A Gonzalez, F Pulido Synth. Commun. 25 1005 (1995) 200. A Simay,K Takacs,K Horvath, P Dvortsak Acta Chim. Acad. Sci. Hung. 105 127 (1980) 201. J Hiufel, E Breitmaier Angew. Chem. 86 671 (1974) 202. M Elnagdy Heterocycles 20 2437 (1984) 203. Eur. P. 190 457; Chem. Abstr. 105 226 578 (1986) 204. G Diadone,M Bajarch, D Rafa Synth. Commun. 25 1441 (1995) 205. Eur. P. 591 103; Chem. Abstr. 121 233 064 (1994) 206. BRD P. 3 524 295; Chem. Abstr. 105 228 491 (1986) 207. US P. 4 562 248; Chem. Abstr. 105 80 662 (1986) 208. Jpn. P. 05 289 232 Yaponiya; Chem. Abstr. 121 95 825 (1994) 209. GDR P. 234 014; Chem. Abstr. 105 226652 (1986) 210. E B Oleinikova,A N Volkov, Yu M Skvortsov Zh.Prikl. Khim. 50 225 (1977) c 211. K A Volkova, E P Levanova, A N Nikol'skaya, A N Volkov, B A Trofimov Zh. Org. Khim. 16 1382 (1980) a 212. USSR P. 771 090; Byull. Izobret. (38) 135 (1980) 213. K A Volkova, E P Levanova, A N Volkov, A N Nikol'skaya, B A Trofimov Zh. Prikl. Khim. 53 1683 (1980) c 214. A N Volkov, K A Volkova, A N Nikol'skaya, E P Levanova, B A Trofimov Zh. Org. Khim. 17 83 (1981) a 215. T T Manakova, A N Volkov, T A Usmanova, L V Morozova, Yu M Skvortsov, B A Trofimov Zh. Prikl. Khim. 50 2547 (1977) c 216. USSR P. 791 741; Byull. Izobret. (48) 94 (1980) 217. A N Nikol'skaya, E P Levanova, A N Volkov, R N Kudyakova, K A Volkova Izv. Akad. Nauk SSSR, Ser. Khim. 1403 (1977) d 218. A N Volkov, I I Danda, Yu M Skvortsov Izv.Sib. Otd. Akad. Nauk, Ser. Khim. Nauk (2) 173 (1975) 219. WO 9 630 359 PCT; Chem. Abstr. 125 328 723 (1996) 220. B A Trofimov, K A Volkova, E P Levanova, A N Volkov, A N Nikol'skaya Zh. Org. Khim. 15 1319 (1979) a 221. K A Volkova, A N Nikol'skaya, E P Levanova, A N Volkov, B A Trofimov Khim. Geterotsikl. Soedin. 1617 (1979) b 222. A N Volkov, Yu M Skvortsov, Yu V Kind,M G Voronkov Zh. Org. Khim. 10 174 (1974) b 223. S I Shergina, A S Zanina, I E Sokolov, I L Kotlyarevskii, in Tez. Dokl. V Vsesoyuz. Konf. po Khimii Atsetilena (Abstracts of Reports of the Vth All-Union Conference on the Chemistry of Acetylene) (Tbilisi: Metsniereba, 1975) p. 545 224. S I Shergina, A S Zanina, I E Sokolov, I L Kotlyarevskii Izv. Akad. Nauk SSSR, Ser.Khim. 1447 (1977) d 225. Jpn. P. 7 746 009; Chem. Abstr. 87 101 930 (1977) 226. A S Zanina, I L Kotlyarevskii, S I Shergina, I E Sokolov, L I Shishkina Izv. Akad. Nauk SSSR, Ser. Khim. 690 (1972) d 227. S I Shergina, A S Zanina, I E Sokolov, I L Kotlyarevskii Izv. Akad. Nauk SSSR, Ser. Khim. 1876 (1973) d 228. A S Zanina, I L Kotlyarevskii, S I Shergina, I E Sokolov, A A Kirchanov Izv. Akad. Nauk SSSR, Ser. Khim. 1636 (1977) d 229. A S Zanina, S I Shergina, G N Khabibulina, I E Sokolov, A A Kirchanov, R N Myasoedova, I L Kotlyarevskii Izv. Akad. Nauk SSSR, Ser. Khim. 2781 (1977) d 230. K B Erzhanov, Doctoral Thesis in Chemical Sciences, Institute of Chemical Sciences, Academy of Sciences, Kaz. SSR, Alma-Ata, 1991 231. I N Azerbaev, K B Erzhanov, T S Sadykov, A N Nasonova Izv. Akad. Nauk Kaz. SSR, Ser. Khim. (6) 40 (1971) 232. USSR P. 602 881; Byull. Izobret. (6) 52 (1976) 233. I N Azerbaev, K B Erzhanov, T S Sadykov, in Khimiya Atsetilena i Tekhnologiya Karbida Kal'tsiya (The Chemistry of Acetylene and Technology of Calcium Carbide) (Alma-Ata: Nauka, 1972) p. 70 234. USSR P. 259 074; Byull. Izobret. (2) 32 (1970) 235. USSR P. 276 043; Byull. Izobret. (23) 35 (1970) 236. I N Azerbaev, K B Erzhanov, T S Sadykov Izv. Akad. Nauk Kaz. SSR, Ser. Khim. 66 (1974) 237. A A Lekerov, K B Erzhanov, L G Potapov Khim. Sel'sk. Khoz. 21 46 (1983) 238. USSR P. 340 383; Byull. Izobret. (18) 8 (1972) 239. USSR P. 425 607; Byull. Izobret. (16) 13 (1974) 240. USSR P. 483 972; Byull. Izobret. (34) 8 (1975) I A Maretina, B A Trofimov 241. I N Azerbaev, K B Erzhakov, T S Sadykov, R Ya Mushii, in Tez. Dokl. V Vsesoyuz. Konf. po Khimii Atsetilena (Abstracts of Reports of the Vth All-Union Conference on the Chemistry of Acetylene) (Tbilisi: Metsniereba, 1975) p. 537 242. N I Azerbaev, K B Erzhanov, T S Sadykov Izv. Akad. Nauk Kaz. SSR, Ser. Khim. 66 (1976) 243. K B Erzhanov, M A Musin, T S Sadykov, in Prirodnye i Sinteti- cheskie Biologicheski Aktivnye Veshchestva (Natural and Synthetic Biologically Active Substances) (Alma-Ata: Nauka, 1980) p. 40 244. T S Sadykov, K B Erzhanov, S S Kolkhoseva Izv. Akad. Nauk Kaz. SSR, Ser. Khim. 68 (1988) 245. M T Crimmins, D M Bankaitis Tetrahedron Lett. 5303 (1983) 246. M T Crimmins, D M Bankaitis Tetrahedron Lett. 4551 (1983) 247. Jpn. P. 80 102 504; Chem. Abstr. 93 199 231 (1980) 248. M M Crimmins, R O'Mahony J. Org. Chem. 55 5894 (1990) 249. Eur. P. 360 246; Chem. Abstr. 113 152 228 (1990) 250. Jpn. P. 06 192 665; Chem. Abstr. 121 208 941 (1994) 251. Jpn. P. 7 783 704; Chem. Abstr. 87 200 777 (1977) 252. Jpn. P. 76 125 010; Chem. Abstr. 86 190 276 (1977) 253. N Takashi, N Yoishi, I Kazuo, F Yoshiji Bull. Chem. Soc. Jpn. 56 2805 (1983) 254. A N Nikol'skaya, A V Bogdanova, T L Kovtonogova, A N Volkov Izv. Akad. Nauk SSSR, Ser. Khim. 397 (1971) d 255. A N Volkov, A N Nikol'skaya, E P Levanova, K A Volkova Zh. Org. Khim. 9 1594 (1973) a 256. A N Nikol'skaya, K A Volkova, E P Levanova, A N Volkov, M V Sigalov Izv. Akad. Nauk SSSR, Ser. Khim. 2792 (1975) d 257. A N Volkov, E P Levanova, K A Volkova, M V Sigalov, A N Nikol'skaya Zh. Org. Khim. 11 516 (1975) a 258. A N Nikol'skaya, A N Volkov, E P Levanova, K A Volkova, S E Metkalova,D A Sarkisyan, E F Lavretskaya Khim.-Farm. Zh. 12 61 (1978) g 259. A N Volkov, A N Nikol'skaya, E P Levanova, V A Pestunovich, E O Tsetlina, V V Keiko. Zh. Org. Khim. 10 1160 (1974) a 260. E P Levanova, A N Volkov, A N Nikol'skaya, K A Volkova, E O Tsetlina Zh. Org. Khim. 10 1835 (1974) a 261. F Bohlmann, C Hmhn Chem. Ber. 110 1183 (1977) 262. D I Mendeleev, in Sochineniya (The Collected Works) (Leningrad; Moscow: Academy of Sciences of the USSR, 1949) Vol. 10, p. 463 263. A Sarkar, S Okada, H Matsuda, H Nakanishi Chem. Lett. 1073 (1998) 264. H Koch, H Ringsdorf Makromol. Chem. 182 255 (1981) 265. US P. 4 215 208; Chem Abstr. 94 74 762 (1981) 266. A A Petrov, I A Maretina, A E Tsil'ko, G V Asratyan, R O Matevosyan, in Tez. Dokl. Vsesoyuz. Soveshch. po Perspekti- vam Rasshireniya Assortimenta Reaktivov dlya Obespecheniya Vedushchikh Otraslei Narodnogo Khozyaistva i Nauchnykh Issledo- vanii, Yaroslavl', 1987 (Abstracts of Reports of the All-Union Meeting on the Perspectives of Extending the Assortment of Chemicals for Provision of Leading Branches of National Economy and Scientific Investigations, Yaroslavl', 1987) p. 26 a�Russ. J. Org. Chem. (Engl. Transl.) b�Chem. Heterocycl. Compd. (Engl. Transl.) c�Russ. J. Appl. Chem. (Engl. Transl.) d�Russ. Chem. Bull. (Engl. Transl.) e�Russ. J. Inorg. Chem. (Engl. Transl.) f�Russ. J. Anal. Chem. (Engl. Transl.) g�Chem. Pharm. J. (

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (7) 609 ± 621 (2000) Direct fluorination of fullerenes { O V Boltalina, N A Galeva Contents I. Introduction II. Synthesis of fluorofullerenes III. Methods of study and physical properties of fluorofullerenes IV. Chemical properties of fluorofullerenes V. Thermodynamic studies of the fluorofullerene C60F48 VI. Conclusion Abstract. fluorocarbon of class new a of synthesis the on Data Data on the synthesis of a new class of fluorocarbon compound, their of reactions by fluorofullerenes, compound, fluorofullerenes, by reactions of C60, C70 , their mixtures are fluorine molecular with fullerenes higher and mixtures and higher fullerenes with molecular fluorine are sur- sur- veyed. properties thermodynamic and physicochemical The veyed.The physicochemical and thermodynamic properties are are presented formed compound only the is which pure for presented for pure C60F48 , which is the only compound formed selectively in direct fluorination of C fluorofullerene for and selectively in direct fluorination of C60 , and for fluorofullerene mixtures. The bibliography includes 70 references mixtures. The bibliography includes 70 references. I. Introduction Very few naturally occurring compounds of fluorine exist, these are primarily mineral inorganic compounds. The unique proper- ties inherent in fluorine and its compounds 1, 2 have served as the basis for the development of new modern technologies and a large number of fluorine-containing materials both in organic and inorganic chemistry.3, 4 The interest in the reaction of molecular fluorine with the new form of carbon, viz., fullerenes, is undoubtedly due to the widely known useful properties of the products of direct fluorination { of graphite, such as high thermal stability (up to 400 8C) of graphite fluorides (CF)n and (C2F)n, the possibility of using fluorine- intercalated graphites CxF with high degrees of fluorination as cathodic materials for lithium batteries, as well as the good antifriction properties of these compounds.5, 6 On the one hand, the polyhedral clusters of carbon, fullerenes, can be regarded as inorganic compounds, as this is an allotropic modification of carbon.On the other hand, fullerenes can be placed among organic compounds, as their behaviour in chemical transformations is similar to that of unsaturated hydrocarbons. Formally, C60 and C70 fullerenes have 30 and 35 double bonds, respectively. Owing to delocalisation of the p-electron density, fullerenes undergo both addition to the double bonds and electrophilic addition with the intermediate formation of charge transfer complexes and, eventually, various derivatives. The C60 fullerene can add up to 24 chlorine or bromine atoms or methyl groups.As O V Boltalina, N A Galeva Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 12 40. Tel. (7-095) 939 53 73. E-mail: OVB@capital.ru (O V Boltalina) Received 6 March 2000 Uspekhi Khimii 69 (7) 661 ± 674 (2000); translated by S S Veselyi #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n07ABEH000579 609 610 613 619 619 620 the size of the functional groups being added decreases, their number increases; for example, hydrogenation of C60 can result in C60H36.What is the maximum number of addends in the case of fluorine? On the one hand, the van der Waals radius of fluorine (1.35 A) is greater than that of hydrogen (1.2 A), and on the other hand, fluorine forms stronger bonds with carbon. The authors of theoretical work 7±9 published before experiments on hydrogena- tion and halogenation of fullerenes were carried out had predicted that the completely fluorinated fullerene C60F60 would be stable.However, in this case, the steric interaction of the adjacent fluorine atoms would result in elongation of the carbon ± carbon bonds in the fullerene framework to 1.67 A,10 as well as in a decrease in theC7F bond energy by 15% in comparison with that of the C7F bond in CF4.11 The authors of yet another theoretical work 12 noted that the formation of theC60F60 molecule is possible provided that steric strain is relieved by distortion of the carbon framework with retention of the icosahedral symmetry. It was anticipated that, owing to its unique spherical symmetry, per- fluorofullerene would possess good lubricating properties, sim- ilarly to graphite fluorides. It did not take the experimenters too long to answer these questions, and as early as in 1991, the first study was published 13 dealing with the synthesis of highly fluorinated fullerene by the reaction of C60 with molecular fluorine.It was shown that addition of 36 (as in the reaction with hydrogen) or more fluorine atoms can occur. Thus, a new line of research, i.e., studies on the fluorination of fullerenes, appeared in fluorine chemistry. This review is aimed at generalisation of the literature data available since 1991 and the own author's results concerning fluorination of fullerenes with molecular fluorine. We intention- ally limit ourselves to direct fluorination reactions. First, the reaction of F2 with fullerenes drew the highest attention of researchers in 1991 ± 96; these were the peak years in the activity of scientific groups in the USA, Great Britain, France and Japan, who were experienced in direct fluorination and possessed the corresponding experimental equipment.Second, direct fluorina- tion reactions are well-developed technological processes that have been implemented in large-tonnage manufacture of various fluorine-containing products. If areas of fluorofullerene applica- tion demanding their production on a relatively large scale are { To the cherished memory of V F Bagryantsev. { The term `direct fluorination' designates the reaction involving molec- ular fluorine, as it is generally accepted in the literature (see e.g. Ref. 4).610 found, it is the reactions with molecular fluorine that can prove to be the most efficient and economically sound.We leave the studies on the alternative methods for fullerene fluorination, including the reactions with fluorides of transition metals, noble gases, etc., outside the scope of this review; these studies have been covered in recent reviews.14, 15 The authors of almost all pioneering publications of the early 1990s used samples obtained by direct fluorination ofC60 orC60/C70 mixtures to study the physical and chemical properties of fluorofullerenes. On the other hand, despite the abundance of original works published in this period, neither detailed analysis of the results obtained nor coordination of the data obtained by different authors regarding the composition and properties of fluorination products have been carried out to date.II. Synthesis of fluorofullerenes 1. Direct fluorination of C60 and C60/C70 The first synthetic study 13 on fullerene fluorination was carried out soon after the development of the procedure for their synthesis in macroscopic quantities. Selig et al.13 studied the fluorination of fullerenes using a Sartorius magnetic balance, which allowed them to detect the changes in the sample weight to within 0.1 mg. The samples were kept in a fluorine atmosphere for at least 20 h to attain a constant weight. The largest weight gain occurred within the first several minutes, after which the absorption of fluorine markedly retarded. The electron impact (EI) mass spectrum (70 eV) of the yellowish-brown reaction product with the compo- sition F :C60=38 (the composition was determined by measure- ment of the weight gain and by titrimetry) contained peaks corresponding to the C60F30 ± 44 molecules, the most intense peak being due to the C60F36 molecule.It should be noted that Selig et al.,13 who used an IR cell for the fluorination, observed accumu- lation of CF4 during the reaction (nC7F=1283 cm71). It was shown 16 that fluorination of fullerenes with molecular fluorine is accompanied by side reactions to give perfluoroalkanes CnF2n+2 (n=1 ± 4). The mass spectrum of a gas phase sample taken directly from the reactor zone during the experiment (Fig. 1) showed, in addition to the Fá2 ions, the presence of CnFák ions (n=1±4, k=1 ± 9). In 1991, a paper by Holloway et al.17 was published.Direct fluorination of C60 fullerene at 70 8C was monitored for several days. Three to four days after the beginning of fluorination, the black C60 powder underwent an abrupt colour change to dark- brown; this colour remained apparently unchanged for another three to four days and then abruptly changed to light-brown. Subsequent fluorination resulted in gradual decolourising of the sample to white eventually. The stepwise change in the colour of the powder during the reaction was considered as evidence for consecutive addition of fluorine atoms to the fullerene molecule. It was assumed that the two intermediates (dark- and light-brown) were stable, partially fluorinated fullerene derivatives, C60F6 and CFá3 615 Ná2C2Fá Fá25 H2O+ Oá2 C3Fá5 7 7 C3Fá C4FáC4Fá9 Figure 1.Mass spectra of a gas phase sample taken from the reactor during the reaction of fullerenes with molecular fluorine. O V Boltalina, N A Galeva C60F42, while the white powder was C60F60. The formation of C60F60 was proven by NMR spectroscopy [the 19F NMR spec- trum contained one intense band with a chemical shift (d) of 7150.5 ppm]. On the other hand, analysis of the mass spectra of the white powder showed peaks corresponding to species with compositions ranging from C60F6 to C60F42; however, species with higher fluorine contents (> 42) were not found. Tuinman et al.18 performed the fluorination of fullerenes in various temperature and time modes using reaction vessels made of different materials (Pyrex, quartz, steel).For example, the following procedure was described: through a tube with an inserted capillary containing the fullerene, helium was first passed at 160 8C at a flow rate of 50 ml min71 for two hours in order to remove traces of moisture and then fluorine was passed at a flow rate of 1 ml min71 together with helium. When the temperature was increased, no visible changes in the sample occurred until helium feeding to the reactor was stopped (at 210 8C); as this was done, the volume of the powder increased and its colour changed from black to yellow. Then the temperature was increased to 275 8C and fluorine was passed for additional 4 h (the temper- ature was not further increased because of the vigorous reaction of fluorine with quartz).According to mass spectroscopic data, the main products, viz., white powders, were mixtures of fluorofuller- enes C60F38 ± 46 (Fig. 2); however, no perfluorinated fullerene C60F60 was detected, the maximum degree of fluorination being 48 (the C60F48 molecule). It was noted that molecules containing more than 48 fluorine atoms were not formed upon longer treat- ment of fullerenes with fluorine. C60F44 C60F42 C60F46 Positive ions C60F40 C60F38 Figure 2. Electron impact mass spectra (70 eV) of a sample synthesised by Tuinman et al.18 Kniaz et al.19 monitored the course of C60 fluorination using a Sartorius balance. The most intense peak in the mass spectrum of the air-stable yellowish-white reaction product corresponded to the C60F40 molecule.In addition, a 1075 times less intense peak corresponding to the C60F60 molecule was observed. In order to improve the quality, the resulting product was sublimed at 300 8C; this gave a yellow sublimate and a white non-volatile residue which included black grains of amorphous carbon. Unlike the spectra of the starting sample, the mass spectra of the sublimate and the residue did not display peaks corresponding to the highly fluorinated C60Fx molecules (x>48). The absence of these peaks was explained by the liberation of molecular fluorine from C60Fx (x>48) during sublimation. Okino et al.20 ± 23 determined the composition of fluorofuller- enes by X-ray photoelectron spectroscopy (XPS) in thin films obtained by deposition of C60 on nickel plates which were then treated with fluorine under the same conditions as the C60 powders (Table 1).The XPS data are in good agreement with mass spectroscopic data. For example, a sample with the compo- sition F:C60=45 (XPS data) contained C60F44 and C60F46 as the major components. It should be noted that the X-ray photo- electron spectra of the samples with low fluorine content (F :C60<20) contained peaks belonging to the C60 fullerene and C60Fx fluorofullerene (x&40) phases. It was assumed that even short-term fluorination converts a fraction of fullerene molecules to derivatives with high fluorine content, whereas the data of chemical analysis and XPS of the C60Fx samples, where x<20,Direct fluorination of fullerenes Table 1.Conditions used for the synthesis of fluorofullerenes and their characteristics. Fluorination conditions C60 sample p /atm time /h T /8C Film Sample weight See a 200 150 70 70 200 200 48 ± 72 100 4 24 *288 *3364 *672 30 48 1 7 111 7 C60F9 77 7 7 225 b 7 7 7 7 Note. The following designations are used: CA is chemical analysis; MS is mass spectroscopy; XPS is X-ray photoelectron spectroscopy. a Sample weight 30 mg; bmm Hg. showed only the presence of unreacted C60 together with highly fluorinated molecules. Hamwi et al.24, 25 fluorinated C60 using a setup made of Monel and steel. A fullerene powder was placed in an aluminium boat; the reaction was carried out at 20 ± 300 8C, either under static conditions (p=1 ± 5 atm) or in a stream of fluorine (the flow rate was 2 ml min71) for 10 ± 336 h.Thermogravimetric analysis showed that fluorofullerenes started to decompose at 180 8C. For this reason, the reaction at temperatures above 180 8C was stopped at the instant when the product weight started to decrease. This resulted in a series of differently coloured powders (from black to pale yellow). The maximum weight gain recorded for one of the samples corresponded to the stoichiometry F :C60=58.5. On the other hand, without mass spectroscopic analysis, it is difficult to make reliable conclusions on the actual presence of molecules with such a high fluorine content. Short-term reactions of the same compounds gave fluoroful- lerenes with low fluorine stoichiometric factors (e.g., C60F9), as determined from weight gain.According to IR spectroscopic and X-ray diffraction data, these samples contained only the non- fluorinated fullerene, while the 19F NMR spectrum exhibited one broad peak, which proved the presence of fluorine in the prod- ucts.25 It was found 24, 25 that C60/C70 fullerene mixtures are more reactive than C60 fullerene. For example, the reaction of C60/C70 with fluorine at room temperature for 4 h resulted in an increase in the sample weight; however, the weight of pure C60 samples did not change upon longer treatment (48 h) under the same con- ditions. Fluorination of C60/C70 at 300 8C for 4 h afforded a pale Table 2.Experimental data on the fluorination of fullerenes with molecular fluorine and composition of the products obtained.16 Mode Tmax /8C o (%) a Ratio F :C60 in the products according to data Starting compound CA see b C60 (see d) 43 22 94 70 461 441 42 4 C60 C60Fx e C60Fx e C60Fx e C60 C60 481 491 C60 (see d) C60 (see d) C60 16 19 7 7 7 7 7 3.2 7 7 7 7 7 6.9 0 7 7 14 7 7 1.5 7 7 7 0 7 7 7 0 7 7 04.7 275 315 315 315 315 315 315 340 340 350 350 10 38 37.4 31.1 6.1 34 7 7 481 462 56 88 88 80 51 84 dynamic quasi-static "static dynamic """""" C60 491 7 7 48.00.5 a The reaction yield (o) was calculated using the formula: o=(mexp/mtheor)6100%, where mexp and mtheor are the C60F48 masses; b the calculation is based on the weight gain; c the measurement accuracy is 0.02; d F2 diluted with 90 vol.% Ar was used; e the fluorination of fluorofullerenes C60Fx was carried out successively in quasi-static, static and dynamic modes.611 Ref. Reaction product according to data XPS MS CA C60F45 C60F42 ± 46 7 7 C60F40 7 7 C60F20 C60F22 C60F45 C60F44 ± 48 C60F45 20 20 20 20 7 7 21 7 21 22 7 23 yellow product with the ratio F :C60=44, whereas for pure C60, it took 8 h to attain the same degree of fluorination. Bagryantsev et al.16 fluorinated a C60/C70 fullerene mixture and found that if the reaction temperature is raised from 70 to 315 8C, the content of fluorofullerenes with higher degrees of fluorination gradually increases.The C60Fx (x=38, 40) and C60F48 fluorofullerenes were the main components among the products obtained at 70 and 315 8C, respectively. The trend towards higher degrees of fluorination at elevated temperatures was also observed in the case of C60 fullerene (Table 2). For example, the C60F46 fluorofullerene with a small admixture of C60F44 and C60F48 was the main reaction product at 275 8C. Fluorination of C60 at 315 8C under dynamic conditions resulted in a mixture of fluorofullerenes, in which C60F48 was the major product, C60F46 was a minor product, whereas the content of C60F44 was insignificant. Figure 3 shows a diagram demonstrat- ing the effect of fluorination time of C60 at 315 8C on the C60F48 content in the resulting samples.Repeated fluorination increased the C60F48 : C60F46 ratio but also resulted in an appreciable loss of the original fluorofullerene due to thermal destruction (partial burning and sublimation); according to mass spectroscopic data, the content of the main product (C60F48) was as high as 95%. Fluorination at higher temperatures (up to 355 8C) for 2 ± 3 h resulted in *90% of C60F48; the reaction yield was >80% in most cases. This procedure was beneficial regarding the reaction time and the consumption of reactants. A study of the effect of the method of feeding fluorine to the reactor showed that the reaction under dynamic conditions in the absence of flashes mostly gave two components, viz., C60F48 and Composition according to mass spectroscopy data NMR XPS C60F46 C60F44 C60F48 F:C60 (see c) 7 46.00.5 77 45.77 45.45 47.23 46.40 47.92 46.53 47.37 47.74 47.79 47.87 47.52 4.3 13 63 51 95 49 68 85 88 91 81 77 50 32 333.8 33 29 13 106.3 12612 o (%) 100 80 60 40 200 1 3 5 7 t /h Figure 3.Effect of time (t) ofC60 fluorination with molecular fluorine at 315 8C on the C60F48 content (o) in the samples synthesised. C60F46. The product obtained under quasi-static conditions (Fig. 4) contained C60Fx fluorofullerenes, where x=40 ± 48, and an admixture of the C60F44OandC60F46Ooxyfluorides (empirical formula C60F45.45O0.05).It is evident from Fig. 4 that the absorp- tion of fluorine at low temperatures (up to 110 8C) virtually does not occur (section I). In section II, starting from 110 8C, notice- able fluorine absorption at a specified, constant rate of heating was observed. The rate of fluorine absorption in the temperature range from 160 to 31 8 C (section III) was almost constant. When the temperature was maintained constant (section IV, 315 8C), noticeable fluorine absorption ceased. T /8C 350 W/mol 0.16 300 0.12 250 200 0.08 150 100 0.04 IV III II I 50 0 0 t /min 100 200 500 400 300 Figure 4. Changes in the temperature and in the fluorine amount absorbed (W) during the fluorination ofC60 under quasi-static conditions. Matsuo et al.26 fluorinated C60 using various methods: with molecular fluorine, with a mixture of fluorine and hydrogen fluoride, with molecular fluorine under UV irradiation, and by passing a stream of fluorine through a CFCl3 solution of fullerene (Table 3).The last-mentioned technique, which is the only reported example of liquid-phase fluorination where a low degree of fluorination is attained, might be a promising method for the Table 3. Conditions of the synthesis of fluorofullerenes from C60 and their characteristics.26 Time /days Fluorinating agent T /8C 20 ± 90 20 ± 85 20 ± 70 20 ± 65 F2 F2 ±HF F2 under UV irradiation (254 nm) F2 (see b) 11.5 ± 27 12 ± 35 11 ± 14 0.2 ± 14 a The calculation is based on the weight gain; b the stream of fluorine was passed through a CFCl3 solution of C60; O V Boltalina, N A Galeva synthesis of fluorofullerenes with low fluorine contents. However, the methods which were used to analyse the products do not allow unambiguous conclusions to be made about the actual molecular composition of the fluorides, in particular, about the formation of products with low fluorine content.In general, the molecular formulae established for the fluorofullerenes were in agreement with the data reported by other investigators. For example, the largest degree of fluorination was attained when UV irradiation was used (see also the data in Ref. 27). Chilingarov et al.28 studied C60 fluorination under conditions of a mass spectroscopic experiment where molecular fluorine was fed from a cylinder into an effusion cell.The temperature (500 ± 800 K) and the reaction time, as well as the fluorine pressure (1074 ±1075 atm) were varied. In the F2 (gas) ±C60 (so- lid) system, only molecules with an even number of fluorine atoms, C60F2n (2<2n<44), were formed. The composition of the gas phase was largely temperature-dependent: an increase in temperature shifted the distribution of ions in the mass spectrum towards smaller masses. For example, the following major prod- ucts were obtained: C60F38 at 510 K, C60F8 at 760 K, and C60F2 and C60F4 at 800 K. Longer reaction times at 745 and 800 K increased the content of C60F18 and C60F22±C60F46, respectively. A search for conditions for selective synthesis of a certain fluorofullerene in the gas phase was performed.For example, the maximum yield of C60F18 (50%) was attained if the reaction was carried out for 20 h at 720 K and at a fluorine pressure of 261074 atm in the presence of NiF2. In the presence of MnF2 (10 h, 615 K, pF2=261074 atm), up to 90% of C60F36 was formed. No study of solid-phase fluorination products, which could be obtained, e.g., by condensation from the gas phase, was carried out.28 Therefore, it is impossible to estimate the applic- ability of the above approach for the selective synthesis of macro- scopic quantities of fluorofullerenes. Thus, the studies on direct fluorination of fullerenes described in this Section showed that (1) the reaction with molecular fluorine occurs readily and the products of fullerene fluorination can form even at room temper- ature; (2) as a rule, mixtures of fluorofullerenes C60Fn with a wide range of compositions are formed (94n454, see Ref.25); the samples with fluorine content n<36 are mixtures of highly fluorinated products (364n448) and unreacted C60 (the higher the content of the latter in the mixture, the smaller the stoichio- metric coefficient for the fluorine atom); (3) fluorofullerene molecules from C60F2 (see Ref. 28) to C60F102 were detected in the gas phase.27 The studies were carried out in the temperature range of 20 ± 300 8C (on increasing the temperature above 300 8C, either no noticeable weight gain in the sample occurred or fluorine reacted vigorously with the reactor material); C60F44 and C60F46 were the main fluorination products at 275 ± 300 8C.16 2.Fluorination of fullerene bromo and chloro derivatives In order to obtain fullerene derivatives with low fluorine content, an attempt was made to fluorinate bromo and chloro fullerene derivatives such as C60Br24, C60Br8, C60Br6 and C60Cl24 with Ratio F :C60 in the products according to data Product colour see a XPS CA 7 7457 7yellowish-white 7 7 41 42 ± 45 47 ± 52 27 ± 33 40 7>45 7Direct fluorination of fullerenes molecular fluorine.29 The idea was to perform the reaction in two stages: in the first step, fluorine had to add to the carbon atoms unoccupied by bromine or chlorine, while in the second stage, due to the lower strength of the C7Cl and C7Br bonds with respect to the C7F bond, abstraction of Br2 and Cl2 on heating was expected giving rise to fluorofullerenes with a low fluorine con- tent.However, the reaction of C60Br24 with molecular fluorine resulted in complete displacement of bromine even without addi- tional heating to give fluorofullerenes with variable composition up to C60F44 with predominant formation of C60F36. Fluorination of C60Br8 gave C60F36; the resulting product contained mostly C60F18 as well as mono- and dioxides of fluorofullerenes. As in the reaction of C60 fullerene with gaseous fluorine, fluorine com- pounds with an even number of fluorine atoms in the molecule were formed.3. Synthesis of pure fluorofullerenes by direct fluorination The use of the reaction of C60 with molecular fluorine for the synthesis of fluorofullerenes with a particular number of fluorine atoms was first reported by Gakh et al. 30 Fluorination was carried out in two stages. A stream of molecular fluorine (u= 1.3 ml min71) was passed over a mixture of C60 with NaF at 250 8C for 20 h, then the product was extracted with CFCl3 and dried; a new portion of NaF was added, and the mixture was again treated with a stream of F2 (275 8C, 30 h). After extraction and drying, the composition of the sample was determined by mass spectroscopic analysis using field desorption for ionisation. The intensity of the ion corresponding to theC60F48 molecule was 69% of the overall intensity of the ions corresponding to the C60F2n molecules (2n448).A study of the 19F NMR spectrum of a sample with a sufficiently high content of C60F48 allowed the researchers to suggest the molecular structure of C60F48 (see Section III.5). Considerable progress 16, 31, 32 in the synthesis of pure C60F48 (with >95% of C60F48 in the samples) was attained in studies of the dependence of the composition of the products of C60 fluorination with molecular fluorine on the temperature (70 ± 355 8C) and on the reaction conditions. At each intermediate step before the next increase in the temperature, the molecular composition was determined by high-temperature mass spectro- scopy with electron impact ionisation and evaporation of samples from the effusion cell.The fluorination products had the form of white powders stable in air for at least three years. The molecular structure of the C60F48 fluorofullerene obtained by this method was identical to that reported by Gakh et al.30 4. Fluorination of C70 and higher fullerenes The fluorination of C60/C70 mixtures resulted in a series of products C70Fn, where n=48 (see Ref. 24) and 54 (see Ref. 18) for the components with the maximum dergee of fluorination. The reaction of pure C70 with fluorine resulted in mixtures of fluorides containing mainly C70F40 (see Ref. 13), C70F38 (see Ref. 16) and C70F54 (see Ref. 31). In these mixtures, C70F46 (see Ref. 13), C70F48 (see Ref. 16) and C70F56 (see Ref. 31), respectively, had the maximum degrees of fluorination.Experiments carried out successively at 250, 280 and 350 8C showed an increase in the degree of fluorination with increase in temperature. However, even at 350 8C, the fluorination product was a mixture of compounds from C70F38 to C70F56 (Fig. 5).16 The only reference to the fluorination of higher fullerenes with molecular fluorine can be found in the paper by Bagryantsev et al.31 The mass spectrum of the product of fluorination of a fullerene mixture containing small amounts of C76 and C84 showed the presence of C76F54 and C84Fn (n=56 ± 62). The reaction was carried out under the same conditions as the fluorination of C70 to give C70F56.31 Thus, the maximum number of fluorine atoms that can be added toC60 andC70 (48 and 56, respectively) corresponds to 80% 613 Intensity (%) 1828 100 1714 1752 1866 80 1790 60 1638 1676 40 1562 1600 1904 200 m/z 1800 1600 1700 1500 Figure 5.Electron impact mass spectrum of a C70Fn sample (n=38 ± 56) synthesised at 350 8C. The maximum degree of fluorination corresponds to C70F56 (1904 amu) and the most intense peak corresponds to 1828 amu, i.e., C70F52. occupation of the reaction sites in fullerene molecules. More extensive fluorination on treatment of fullerenes with molecular fluorine at temperatures of up to 350 8C is probably hampered by the inaccessibility of the remaining doubleC=Cbonds because of steric hindrance caused by the size of the fluorine atoms and the rigidity of the spherical carbon skeleton.Pure fluorides of the C70 fullerene will probably be hardly accessible due to the lower symmetry of C70 in comparison with C60. In the case of heavy fullerenes, the prospects for the synthesis of pure compounds are also complicated by the existence of geometrical isomers. III. Methods of study and physical properties of fluorofullerenes 1. IR spectroscopy The IR spectroscopic data for fluorofullerenes are reported in many papers. As the compositions of fluorofullerene samples are known, one can follow a correlation between the vibration frequency of the C7F bonds and the fluorine content in the samples. For example, the broad unresolved band at 1148 ± 1165 cm71 typical of fluorofullerene mixtures C60Fx (364x448) 13, 19 shifts towards 1170 ± 1176 cm71 for the pure C60F48 fluorofullerene.16 A similar shift of bands with increase in the fluorine content was also observed in the case of fluorinated graphite: 1110 ± 1120 cm71 for fluorine-intercalated graphite with the composition C2F (see Ref.26) and 1220 cm71 for graphite fluorides (CF)n, (C2F)n.33, 34 The main distinction of the IR spectra of pure C60F48 from those of fluorofullerene mixtures is that fine structure in the region of C7F vibrations and a group of intense bands in the region of 600 ± 800 cm71 are displayed (see Refs 26, 32). One can see in Fig. 6 that the most intense peaks at 650 and 723 cm71 are characteristic only of the C60F48 fluorofullerene but are absent from the IR spectrum of a C60F46 sample.Evidently, the bands in this spectrum region can serve as `fingerprints' of pureC60F48. The IR spectra for certain C60F48 isomers were calculated in a theoretical study,35 but no detailed assignment of the normal vibration modes has been carried out to date. IR spectroscopic data for some C60Fx fluorofullerene samples are presented in Table 4.614 T (%) 50 1 40 30 20 70 2 60 50 40 301500 1300 1100 900 Figure 6. IR spectra of fluorofullerene samples: C60F46 (1) and C60F48 (2).32 Table 4. IR spectroscopic data for some samples of C60Fx fluorofuller- enes. x Vibration frequencies /cm71 1164, 1133 1148 1150, 750, 510 1165 1168 1150, 750, 600 1170, 1141, 755, 736, 723, 650, 604 36 (see a) 64x442 384x448 304x444 47 27 ± 52 48 a The fluorofullerene was synthesised in the reaction with manganese trifluoride.2. Electronic structure and electrochemical properties The electronic structure of fluorinated fullerenes differs from the electronic structure of the original fullerenes. The high electro- negativity of fluorine decreases both the LUMO and the HOMO energies, thus the electron affinity and the first ionisation potential of C60Fx and C70Fx increase in comparison with those of C60 and C70. Obviously, the changes in the electronic potentials occurring due to transformation of fullerenes to fluoro derivatives affect their electrochemical and photoelectrochemical behaviour. Data on the electronic structures of the C60F48 and C60 molecules obtained by XPS, from the near fine structure and from the near-edge X-ray absorption fine structure spectra (NEXAFS) are presented in Fig.7. It was found 36 that the work functions for C60Fx are 5.6, 5.2 and 5.4 eV for x=30, 36 and 48, respectively. This is considerably higher than the work functions for organic compounds (usually, 4.1 ± 5.0 eV) and for C60 fullerene (4.7 eV 39), which points to strong electron-withdrawing properties of fluorofullerenes. The addition of fluorine to fullerene shifts the bands in the absorption spectra from 1.8 eV for C60 fullerene to 6.4 eV for C60Fx (424x448).40 Many researchers have studied fluorofullerenes by XPS,25, 26, 41 ± 43 but the data of different authors are in poor agreement due to the use of different types of X-ray radiation, different calibrations of the instruments and different substrate materials.In certain cases, surface charging and a shift of the absolute energy scale were observed, which complicated consid- erably the interpretation of the spectra and made it difficult to obtain reliable data on the composition of the fluorofullerene samples and on the chemical shifts. According to some data of 34 17 18 13 24 26 31 723 cm71 700650 Ref. gas Evac=0 2.65 LUMO4.95 0.95EF 4.3 1.4 HOMO <7.6 C1s 289.0 Figure 7. Energy diagram for the C60 and C60F48 molecules in the gas and solid phases 36 (the HOMO value was changed in accordance with the data in Refs 37 and 38).multicomponent photoelectron C1s spectra,22, 25, 41, 42 destruction processes take place on a sample surface exposed to hard X-radiation, which is accompanied by a flow of high-energy electrons. In addition, it was found 26 that the relative intensity of peaks in the photoelectron spectra depended on the duration of the exposure of samples to X-radiation. This was explained by the destructive effect of the electron flux. For example, according to the data reported by Matsuo et al.26 the C60 :F ratio was 42 after irradiation of a fluorofullerene sample for 33 s, while an increase in the irradiation time to 750 s decreased the content of fluorine to C60 :F=38. The use of electron-free synchrotron radiation (40 ± 500 eV) made it possible to prevent the destruction of the fluorofullerene surface layer and to eliminate the charge accumulation on the sample surface.The photoelectron spectra obtained under these conditions 43 displayed peaks at 287.2 and 284.9 eV, which were interpreted as the energies of the 1s electrons of the carbon atoms (Eb) involved in the formation of the C7F and C7C bonds, respectively (Fig. 8 a). Similar two-component C1s spectra were also obtained in other studies.26, 36, 45, 46 The two lines in the F1s Pulse s71 1000 800 600 400 200 0 282 600 400 2000 Figure 8. X-Ray photoelectron spectra ofC60F48: (a) spectrumof carbon 1s electrons; (b) spectrum of fluorine 1s electrons excited by X-ray radiation of AlKa1,2 with Teflon as the standard.44 O V Boltalina, N A Galeva C60 C60F48 solid solid 3.6 1.4 0.95 5.0 1.9 2.6 1.1 EF 5.4 6.2 8.4 2.2 284.1 284.7 286.2 284.1 287.7 289.1 2.6 291.6 288.4 293.8 a C7F C7C (CF2)n 286 290 294 Eb /eV b C7F (CF2)n 690 685 Eb /eV gas 4.05 LUMO 7.9 12.0 HOMO C1sDirect fluorination of fullerenes photoelectron spectrum of C60F48 fluorofullerene (Fig. 8 b) at 685.9 and 689.5 eV represent two states of fluorine, namely, those in fluorofullerene and in Teflon, respectively.In certain studies, the XPS method was used to determine the compositions of the samples obtained (see Tables 1, 2). However, the accuracy of these estimates is insufficient, 2 fluorine atoms per fullerene molecule.x It is well known that, depending on the fluorine content in fluorine-intercalated graphite CxF, the nature of the C7F bond changes from ionic to partially ionic (or partially covalent).5 The F1s photoelectron spectra of CxF compounds with low fluorine content display one peak at 684.5 eV; this corroborates the existence of only one type of fluorine atoms in the compound, which are in the same electronic state as those in metal fluorides (for instance, the peak at 685.6 eV represents LiF 22). It is believed 5 that intercalated fluorine atoms distributed between carbon layers undergo ionisation and interact with the Cdá polycation. The increase in x in CxF to 10 is accompanied by some delocalisation of the electron density, hence the C7F bond cannot be considered as purely ionic.The F1s spectrum displays a peak at 685.7 ± 685.8 eV. Subsequent increase in the fluorine content in CxF produces a second peak in the C1s spectrum, which is explained by a covalent contribution to the energy of the fluorine ± carbon bond. Thus, the CxF compounds with the composition 24x44 contain fluorine ± carbon bonds of two types: ionic and partially ionic (or partially covalent). The observed chemical shifts in the photoelectron spectra are listed in Table 5. Matsuo et al.26 compared the photoelectron spectra of fluo- rofullerene, fluorine-intercalated graphite C2F and graphite fluo- rides (C2F)n and (CF)n (Fig. 9). It is important that the high- energy peak for C60F43 fluorofullerene (291.1 eV) is located approximately between the corresponding peaks of C2F and graphite fluorides, whence the authors conclude that the degree of covalency of the C7F bond in fluorofullerene is intermediate between those of the C7F bonds in C2F intercalates and in graphite fluorides.The differences between the electron binding energies (DEb) in carbon atoms with different chemical environments were com- pared;26 the difference is manifested in the C1s photoelectron spectra as two peaks. It was concluded that the greater DEb, the Table 5. X-Ray photoelectron spectroscopy data for fluorine-intercalated graphite, graphite fluorides and fluorofullerenes. Compound a Eb(C1s) /eV Eb(F1s) /eV C7C C7F CxF (20<x) CxF (4<x<10) CxF (24x44) C2F C4F (C2F)n (CF)n C60 C60 C60F48 C60Fx (x=44, 46, 48) C60F47 C60F43 C60Fx (x=31, 33) C60Fx C60F48 C60F48 C60Fx (x=40±48) 284.6 284.1 284.1 ± 284.6 287.5 283.4 290.8 7284.6 284.7 286.2 284.9 288.5, 290.7, 293.3 288.9 288.0 284.6 291.5 285.6 286.8 a The sample composition is given in accordance with original work.684.5 685.7 ± 685.8 686.0 ± 687.1 687 7690 689.6 7777777688 7685.9 7 615 1234284 288 292 296 Eb /eV Figure 9. X-Ray photoelectron spectra of carbon 1s electrons in fluo- rine-intercalated graphite C2F (1), C60F43 fluorofullerene (2) and graphite fluorides (C2F)n (3) and (CF)n (4).26 more ionic and less covalent the C7F bond.For example, DEb for C60F33 and C60F31 samples is smaller than that for fluorine- intercalated graphite C2F and larger than that for graphite fluoride (C2F)n, which suggests that the C7F bond in fluoroful- lerene has an intermediate character between covalent and semi- ionic. Comparison of the DEb values in the C1s spectra in C60F48 fullerene (1.8 eV) and in the spectra of fluorine-intercalated graphite (2.6 eV) and graphite fluorides (1.9 eV) showed that the C7F bond in C60F48 is covalent. The average DEb values were 2.2 ± 2.4 eV (see Table 5). The electrochemical behaviour of fluorofullerenes was first studied by cyclic voltammetry. Holloway et al.17 studied the electrochemical reduction of C60F48 in a CH2Cl2 solution and detected the reaction products using a mass spectrometer.It was shown that the reduction of C60F48 occurs reversibly and that the reduction potential of C60F48 is 1.38 V higher than that of C60. This agrees completely with the fact that the electron affinity of fluorinated fullerenes is about 1.5 eV higher than that of full- erenes. Based on the mass spectra obtained using the electro-spray method, it was concluded 47 that the electrochemical reduction mechanism includes a series of consecutive reactions of electron transfer and abstraction of fluorine atoms. A study of the electrochemical properties of fluorofullerenes in solid electrolytes based on polyphosphazine derivatives doped with Li+ was carried out.48 The cathodic mixture contained a fluorofullerene, graphite and the electrolyte in the mass ratio Ref.DEb(C1s) /eV 7 7 7 72.7 ± 3.4 2.6 3.2 1.9 7 286.8 ± 288.0 290.1 286.6 292.7 290.4 2.2 2.3 288.4 287.2 555 26 26 26 33 7 7 22 7 7 37 37 43 7 7 26 26 26 42 45 46 46 2.2 2.4 4.9 2.0 1.8 2.5 291.1 290.4 289.5 293.5 287.4 289.3616 1 : 1 : 2. The voltammograms of C60Fx and C70Fx fluorofullerenes in solid-phase lithium batteries showed peaks at potentials of 3.1 and 3.4 V relative to the E(Li+/Li) potential, which corresponded to irreversible reduction of C60Fx and C70Fx. After two voltam- metric cycles, XPS showed a decrease in the relative intensity of the signal at 289 eV in the C1s spectrum for the carbon atoms involved in the C7F bonds and the appearance of a new signal at 685 eV in the F1s spectrum.This is evidence for irreversible reduction of C7F bonds during the experiment and the forma- tion of LiF. Similar experiments with fluorinated graphite (CF)n showed that its reduction potential is 0.5 V more negative than that of fluorofullerene. This implies that the C7F bond in the C60Fx molecule is reduced more readily than the C7F bond in (CF)n. It was assumed that reduction of fluorofullerenes under galvanostatic conditions is homogeneous, unlike the heterogene- ous reduction of graphite fluoride (CF)n. The electrochemical behavior of fluorofullerenes the synthesis of which is described in Section II and Table 3 was studied as a function of the elementary composition (from C60F15 to C60F47).49 It was shown that the reduction potentials of fluoro- fullerenes approach the reduction potential of C60 fullerene as the fluorine content decreases.3. Crystal structure The structure of C60 fullerene at room temperature is described by a face-centred cubic (FCC) lattice with the parameter a=14.17 A; the distance between the neighbouring C60 mole- cules is 10.02 A, and the density is 1.72 g cm72 (which corre- sponds to 1.4461021 molecules cm72).10 Data of numerous X-ray diffraction studies (see, e.g., Refs 19, 20, 50 ± 53) indicate that at room temperature, the products of fullerene fluorination have the same FCC lattice (Table 6) as the matrix compound C60 but with an increased spacing.The increase in the distance between two neighbouring molecules following fluorination manifests itself as a shift of diffraction lines to smaller reflection angles: the first maximum is at 2Y=98,13 which is much smaller than for crystalline C60. Some papers reported the absence of crystallinity, especially for samples obtained at rather low temperatures.20, 24, 25 The diffraction pattern usually con- tained two diffuse peaks indicating the absence of long-range order, characteristic of crystals. Table 6. Lattice parameters, the number of molecules in a unit cell and the density of fluorofullerenes at room temperature. Lattice type Compound C60 C60F45 (see b) C60F40 (see b) C60F20 (see b) C60F22 (see b) C60F54 (see c) C60F53 (see b) C60F48 C60F40 (see d) C60F46 (see d) C60F48 (see d) FCC FCC FCC FCC FCC FCC FCC FCC BCT-40% FCC-60% FCC BCT FCT BCT FCC orthorhombic C60F48 (see d) C60F46 (see d) C60F48 (see d) Note.The following designations are used: FCC is face-centred cubic lattice, BCT is body-centred tetragonal lattice, and FCT is face-centred tetragonal lattice; a dcalc=M/(VNA), whereMis molecular mass, V is the cell volume, NA is the Avogadro number; b composition was determined by photoelectron X-ray spectroscopy; c composition was determined from the weight gain; d composition corresponds to the most intense peak in the mass spectrum; e lattice parameter b is equal to 37.97(8) A.Figure 10. X-Ray diffraction patterns of samples with increasing contents of C60F48 (1± 3), C60F46 fluorofullerene (4) and C60 fullerene (5).46 The a parameters for fluorofullerene samples obtained by different research groups and described by a FCC lattice range from 16.98 (see Ref. 20) to 17.6 A,24 which approximately corre- sponds to an increase in the molecular volume by a factor of 1.8 relative to that of C60.50 Due to the replacement of the short double bonds by longer single bonds, the radius of the carbon cage in a fluorofullerene molecule increases by 1% (from 3.53 to 3.56 A) 47 with respect to that of C60, or by 10% according to other data.21 Representing a fluorofullerene molecule by two concentric spheres which corre- spond to the positions of the carbon and fluorine atoms (the C7F bond length is 1.36 A) and considering that fluorination results in an increase in the radius of the carbon cage by 10%, Okino et al.21 calculated the X-ray spectra for C60F45 and C70F32 samples.Kniaz et al.19, 51 studied a sublimed fluorofullerene powder; the most intense peak in its mass spectrum was that of C60F40. It was shown that it contained 60% of a FCC phase and 40% of a hexagonal phase (see Table 6) and had a higher degree of crystallinity than unsublimed fluorofullerene. Lattice parameters /A a c 14.17 17.07 16.98 17.3 17.3 17.600.15 17.21 19.28 17.91(1) 17.9 17.83(2) 17.27 11.85 16.677 17.158(3) 11.852(8) 16.75 11.859(3) 17.24(1) 12.06(3) e 12.21(4) 10 20 z Vcell/z /A3 711 1250 1224 1294 1294 1363 1274 1288 1354 1160 1263 1258 1256 1254 1281 1398 4444444424424241 O V Boltalina, N A Galeva 12 345 2Y Ref.dcalc a /g cm73 10 20 20 20 20 24 25 47 19, 51 19, 51 46 46 46 52 52 53 1.681 2.093 2.009 1.412 1.412 2.128 2.252 2.105 772.096 2.155 2.096 2.162 2.116 1.939Direct fluorination of fullerenes It was found 46 that at room temperature, C60F48 has a body- centred tetragonal cell with the parameters a=11.852(8) and c=17.91(1) A. It was assumed previously 53 that the C60F48 structure can be described by an orthorhombic cell (see Table 6).Comparison of X-ray diffraction data for fluorofullerene samples with different molecular compositions (Fig. 10) shows that, where a mixture of fluorofullerenes with averaged composition is con- cerned, the crystal structure corresponds to an FCC lattice similar to that of C60, whereas the C60F48 fullerene has a body-centred tetragonal lattice. 4. NMR spectroscopy 19F NMR of solutions. The NMR method was first employed to study the products of fullerene fluorination back in 1991.17 The 19F NMR spectra of solutions of fluorofullerenes in acetone or tetrahydrofuran exhibit singlet peaks with chemical shifts of 7150.5 and 7152.7 ppm. It was suggested 17 that either one or both peaks were due to the C60F60 molecule in which all fluorine atoms are equivalent.However, the mass spectra of these samples did not show the presence of C60F60. Some 19F NMRspectra had a signal at7188 ppm, due to the hydrogen fluoride resulting from the reactions of fluorides with the residual water in acetone.17 It has been found 54 that the integral intensity of the 19F NMR signal of the fluoride ion in THF depends on the amount of water in THF. Kniaz et al.19 have employed the 19F NMR method to study acetone solutions of the product of fluorination of C60, the product of its sublimation at 300 8C and the non-volatile residue after sublimation. The 19F NMR spectra thus obtained were identical to those reported in the study cited above.17 Analysis of the initial fluorofullerene by mass spectroscopy showed that the intensity of the peak due to the C60F60 molecule was 105 times lower than that of the most intense peak, which corresponded to the C60F40 molecule. After sublimation, the mass spectra of neither the sublimate nor the residue exhibited signals for C60Fn, where n550.The researchers explained this outcome by assuming that less stable, higher fluorofullerenes were destroyed during sublimation to give F2 or CF4. It was found that the NMR signal intensity passes through a maximum (corresponding to the composition C60F36) as fullerene gets saturated with fluorine. Hence, the signal at7149 ppm could belong to yet another highly symmetrical molecule, C60F36. The 19F NMR spectrum of a THF solution of fluorofullerene (prepared 17 by direct fluorination of C60) studied by Boltalina et al.55 looked like those reported previously.17, 19, 54 Unlike earlier studies, in this study, the solution was analysed by mass spectro- scopy.The spectra contained intense peaks in the region of small masses (which were not identified) but no signals corresponding to fluorofullerenes. This outcome may be due to the chemical interaction of fluorofullerene with the solvent; this would account for the fact that the data obtained by NMR spectroscopy for solutions of fluorofullerenes in THF or acetone do not agree with the data obtained by other methods which do not require dissolution of samples. The line occurring at*7150 ppm in the 19F NMR spectrum of a THF or acetone solution of fluorofullerenes can apparently be attributed to the product of decomposition of fluorofullerenes accompanied not only by the loss of fluorine but also by destruction of the carbon skeleton. The composition of the products of reactions between THF or acetone and fluorofuller- enes C60Fx has not yet been identified.To eliminate the influence of solvents, NMR studies of solid samples should be carried out. 19F NMR of solid samples. Static 19F NMR spectra of solid C60Fn fluorofullerenes have been reported.24, 25 The spectra con- sist of one broad line with d 7153 ppm and the half-width Dn1/2=60 ± 120 ppm (Fig. 11). The d values observed in the NMR spectra of a black sample (the mass spectrum of which exhibited ions corresponding to C60, C70 and low-fluorine fluo- rofullerenes) and of a pale-yellow sample consisting mainly of fluorofullerenes were identical. TheNMR spectra differed only in 617 1 2 0 7400 ppm Figure 11.19F NMR spectrum (relative to CF3CO2H) of solid samples of C60Fn fluorofullerenes at 20 (1) and 300 8C (2).35, 36 the Dn1/2 value, which was 115 ppm for the pale-yellow sample and 140 ppm for the black one. Fluorination of C60 with molec- ular fluorine for two weeks at room temperature gave a light- brown fluorofullerene, which was then sublimed at 150 8C. The 19F NMR spectrum of this sample exhibited a narrow peak at 7155 ppm with Dn1/2=8 ppm. It was suggested 25 that this peak width corresponded to highly symmetrical molecules such as C60F60 or C60F36.The spectra of some fluorofullerene samples sublimed at 110 8C and 150 8C had an additional signal with d=7133 ppm; this was attributed to the presence of oxyfluor- ides C60F6072yOy. Unfortunately, the researchers did not inves- tigate the composition of the samples; therefore, the reliability of the conclusions drawn cannot be judged. 13C NMR. A solid sample having C60F46 as the major com- ponent (according to the data of mass spectroscopy) was studied by 13C NMR.41 The technique of 13C719F polarisation exchange showed that the sample was a mixture of compounds differing from one another in molecular motion parameters. For one of the components characterised by fast molecular re-orientations (sim- ilar to those observed in plastic crystals), the ratio of the sp2 : sp3 resonance lines was found to be 14 : 46, i.e.it complied with the stoichiometry C60F46, determined from the mass spectroscopy data. The resonance lines of the sp2-hybridised carbon atoms occurred in the region of 140.0 ± 142.5 ppm, while those of sp3 carbon were manifested at 86 ± 90 ppm. However, in addition to the lines assigned to C60F46, the 13C NMR spectrum exhibited a broad band at 118 ppm having Dn1/2=40 ppm and an intensity of 2010% of the total peak intensity. The dynamics of cross- polarisation of this line indicated that the other component of the sample comprises molecules immobile in the solid phase. 5. Molecular structure of C60F48 Experimental data on the structure of C60F48 were first obtained by Gakh et al.30 Mass-spectroscopic analysis of the specimen they studied (the procedure of the synthesis is described in Section II.3) showed that the intensity of the ion derived from the C60F48 molecule amounted to 69% of the total intensity of ions derived from C60F2n (2n448). The 19F NMR spectrum of a solution of this sample in the CFCl3 ± CDCl3 mixture exhibited eight main lines with equal intensities (Fig.12). It was found 30 that this spectrum can be attributed to 14 isomers of C60F48 including optical isomers which contain eight groups of equivalent fluorine atoms. Using 19F (COSY) homonuclear correlation spectroscopy, a two-dimensional spectrum of C60F48 was recorded,30 in which the cross-peaks were due to the spin ± spin scalar interactions of the fluorine nuclei in the system.Based on this spectrum, the molecular structure of the fluorofullerene shown in Fig. 13 a was proposed. The use of homonuclear two-dimensional 19F±19F COSY spectroscopy for C60F48 31 revealed additional cross- peaks, which were consistent with the proposed structure.30 Note that Boltalina et al.31 fluorinated a fullerene extract (C60 :C70=3 : 1) and, according to mass spectrometry, the fluo- rination product contained up to 30% of C70Fx (x=44 ± 56) and traces of fluorinated higher fullerenes (C76Fy and C84Fz). How- ever, these impurities barely influenced theNMRspectral pattern.618 7125 Figure 12. 19F NMR spectrum of a solution of C60F48 CFCl3 ± CDCl3 system.30 h d A B c h d f b B c Figure 13.Schlegel diagrams for optical isomers of the C60F48 molecule with D3 symmetry (a) 30 and with S6 symmetry (b).35, 36 The atoms equivalent in symmetry are denoted by the same characters, (a ± f) for F atoms and A and B for C atoms. 7155 7145 7135 acB b A g a f g a e b e A f B e d c d f h h a A c c b h B g B d A b f g g a e b a e A f dh bc B b A g a g a b e A f e d Bc d f h h a A c c b h B B g d A b f g a e a e A f dh 7165 h d f e a g b B ch d f ee a A g g B b c O V Boltalina, N A Galeva ppm in the Semi-empirical quantum-mechanical calculations for the rel- ative stability of nine isomers of C60F48, namely, five chiral pairs with D3 symmetry and four meso forms with S6 symmetry, have shown 35 that two of these isomers, namely, those presented in Fig.13, are more stable than any other isomer by at least 175 kJ mol71. In addition, the difference between the energies of formation of these isomers is negligibly small and amounts to 2 kJ mol71. In view of the size of the molecular system in question and the degree of uncertainty inherent in semi-empirical methods, these isomers can be regarded isoenergetic. The double bonds in them are located only at the edges shared by pentagons and hexagons in the carbon cage (see Fig. 13). The presence of a double bond at an edge shared by two hexagons increases the energy of formation of C60F48 by 174.7 kJ mol71.Conjugation of double bonds is a destabilising factor, which increases the DfH value by 500 kJ mol71 with respect to that of the most stable isomer. Since the shape of fluorofullerene molecules is nearly spher- ical, it would be natural to expect molecular mobility in the crystalline phase similar to that observed for C60 fullerene.10 Figure 14 presents the variation of the static 19F NMR spectra of fluorofullerenes vs. temperature. These data (Fig. 14 a) indicate that, as the temperature decreases to 320 K and below, the velocity of re-orientation of the C60F48 molecules in the fluoro- fullerene crystal lattice smoothly decreases, and at temperatures below 260 K, the rotational re-orientation of molecules stops on the NMR time scale, the spectrum exhibiting one broad line with Dn1/2=40 000 Hz. At temperatures above 320 K, the C60F48 molecules rotate isotropically, the velocity of rotation being sufficient for observing the fine structure in the signals of the two groups of fluorine atoms in fluorofullerene molecules.56 a b 1 1 2 2 3 3 Dn1/2 4 4 0 710072007300 ppm 0 710072007300 Figure 14.Variation of the static 19F NMR spectrum of crystalline C60F48 (a) and C60F46 (b) fluorofullerenes as a function of temperature /K: (1) 200, (2) 260, (3) 300, (4) 340. The high-resolution 19F NMR spectrum of a polycrystalline sample of C60F48 consists of eight lines with equal intensities, which correspond to eight groups of magnetically non-equivalent fluorine atoms in the C60F48 molecule.It should be noted that the chemical shifts of these lines 56 virtually coincide with the chemical shifts observed in the 19F NMR spectrum recorded for a solution of C60F48 in CDCl3.30 Thus, the crystal lattice does not exert significant influence, compared to the influence of the solvent, on the 19F chemical shifts of C60F48. The four groups of low-intensity lines (< 10% of the total intensity of peaks) observed in the spectra of both solutions and solid samples can apparently be assigned to C60F2n impurities (2n<48) or to other C60F48 isomers. The high-resolution 13C NMR spectra of a C60F48 sample exhibit two equally intense lines in the region of signals corre- sponding to sp2 carbon atoms, pointing to the presence in the C60F48 molecule of two magnetically non-equivalent groups of carbon atoms not attached to fluorine atoms.56 In accordance with the 19F NMR spectrum, the region of signals of the sp3Direct fluorination of fullerenes ppm 90 140 100 150 Figure 15.13C NMR{1H, 19F} spectrum of a solution of C60F48 in CDCl3 (recorded by Prof. T Ono, Japan). carbon nuclei should contain eight peaks corresponding to eight magnetically nonequivalent groups of carbon atoms attached to fluorine atoms. However, the spectral pattern was fairly complex because each of the eight expected lines was split due to the scalar 13C±19F spin ± spin interaction. The complex multiplet was resolved by using equipment for C7F decoupling (Fig.15). It is worth noting that 13C NMR data obtained by other investiga- tors 41 only pointed out the presence of non-resolved broad bands in the regions of sp2 and sp3 carbon signals. Thus, the compound C60F48 has been studied comprehen- sively byNMRspectroscopy; it was found that the isomeric purity of this compound prepared by direct fluorination of C60 reaches 90%, which is a startling result if one recalls that tens of thousands geometrical isomers of C60F48 can theoretically exist. To elucidate the molecular structure more precisely (to choose between the most probable structures, D3 and S6), experimental and theoret- ical data of IR and Raman spectroscopy and crystallography should be resorted to. Additional experiments are needed includ- ing the synthesis of C60F48 with an endohedral 3He atom and subsequent analysis of the chemical shifts in the 3He NMR spectra.IV. Chemical properties of fluorofullerenes Fluorocarbon polymers such as Teflon are known to be excep- tionally chemically inert, being able to react slowly only with liquid sodium metal. This is due to the great strength of the C7F bond (460 ± 540 kJ mol71), the non-polar character of this bond and the rigidity of carbon chains covered by a shell of fluorine atoms. A similar type of behaviour could have been expected of fluorofullerenes. However, fluorofullerenes were found to exhibit oxidative properties. A typical feature of fluorinated fullerene derivatives is high reactivity. Strong repulsion of fluorine atoms from one another is supposed to be the driving force of reactions.The C7F bond length in fluorofullerenes is 1.49 A;23 for compar- ison, this bond length in aliphatic fluorocarbons is 1.38 A.4 Correspondingly, the energy of the C7F bond in fluorofullerenes is 150 ± 250 kJ mol71 lower than that in aliphatic fluorocarbons. According to both kinetic and thermodynamic factors, fluoroful- lerenes have lost the intertness of the C7F bond, typical of fluorocarbons; fluorine can be transferred to electron-saturated compounds and electrons can be captured from some reagents such as benzene. The chemical properties of fluorofullerenes have been poorly studied yet. This is due to several reasons, namely, inaccessibility of macroscopic quantities of the compounds and the difficulty of the analysis of the reaction products (i.e.the formation of mixtures of compounds with closely similar properties). It has been reported 57 that highly fluorinated fluorofullerenes enter into a number of oxidative reactions (with sodium iodide or isopropanol) and into fluorination reactions (with aromatic com- pounds in the presence of Lewis acids). Mixing a solution of a 619 C60F44±C60F46 fluorofullerene mixture with 0.1 MNaI in acetone results in immediate evolution of iodine; after keeping the mixture for several hours at room temperature, a black precipitate is formed. Analysis of the reaction solution by mass spectroscopy shows the presence of fluorofullerenes with a fluorine content lower than that in the initial sample.I2+C60Fm+NaF C60Fn+NaI n=44±46, m=36 ± 40. It was shown that fluorofullerenes react with bromides in a similar way. In the reaction with tetrabutylammonium bromide in chloroform, immediate evolution of bromine was observed. The reaction with isopropanol proceeding on prolonged heating (85 8C) afforded 1.6 moles of acetone per mole of fluo- rofullerene. Me2CO+C60Fm+HF C60Fn+Me2CH(OH) n=44±46, m=36 ± 40. The reaction of fluorofullerenes with benzene in the presence of BF3 . Et2O was studied. Mass-spectroscopic analysis of the reaction mixture, the colour of which slowly changed from yellow to brown, indicated the formation of fluorobenzene and the presence of fluorofullerenes with lower mass numbers.It was noted that degradation of fluorofullerenes C60Fx is accelerated as x increases. Some reactions result in the formation of a black resin. PhF+C60Fm+HF. C60Fn+PhH n=44±46, m=36 ± 42. It should be noted that even pyridine, which is well known to be inert towards electrophilic substitution, does react slowly with higher fluorofullerenes at elevated temperatures giving rise to 2-fluoropyridine (the yield reaches 10%). Unfortunately, it cannot be understood from the results obtained what is the mechanism of these reactions, whether it is ionic (fluorine transfer) or radical (electron transfer). On the one hand, the predominant formation of biphenyl in the reaction of fluorofullerene with phenyllithium supports the radical mecha- nism, while, on the other hand, fluorination of pyridine is a typical ionic process (by analogy with the well known reactions of pyridine with other electrophilic fluorinating agents).In one of the first publications dealing with fullerene fluori- nation,58 it was noted that fluorination products tend to enter into a nucleophilic substitution reaction, namely, the OH7 anion reacts with a double bond in fluorofullerene, then HF is elimi- nated and fluorofullerene oxide is produced. HO7 OH F F F O C C C C C C C C C C C C 7HF 7F7 As noted above, detailed investigation of the chemical proper- ties of fluorofullerenes is possible only provided that individual compounds are available and their molecular structures and the degrees of purity are known.In the last two years, certain progress has been attained along this line. In particular, the isolation of pure C60F18 and of its oxo derivative C60F18O are worthy of note.59, 60 An example of chemical transformation of individual fluorofullerenes is the reaction of C60F18 with benzene in the presence of FeCl3, known as the Friedel ± Crafts reaction, which resulted in the synthesis of C60F15Ph3, called `triumphene'.61 V. Thermodynamic studies of the fluorofullerene C60F48 The thermodynamic properties of C60F48 known to date are summarised in Table 7. The low-temperature heat capacity of C60F48 in the temper- ature range from 8 to 356 K was measured by adiabatic calorim- etry.62, 65 These data were used to calculate the numerical values of620 Table 7.Thermodynamic characteristics of C60F48. Value Characteristic Ref. Cp,m(298.15) /J K71 mol71 62 62 62 62 62 1365.22.7 192.480.83 1137.25.7 146.62.9 Hm(298.15)7Hm(0) /kJ mol71 S m(298.15)7S m(0) /J K71 mol71 Gm(298.15)7Hm(0) /kJ mol71 DfS m(298.15) /J K71 mol71 DsHm(476) /kJ mol71 74071.65.8 1097 63 VI. Conclusion DfHm(cryst.) /kJ mol71 64 64 DfHm(gas) /kJ mol71 77563166 77454166 the thermodynamic functions H(T) ¡¾H(0), S (T) ¡¾ S (0) and G(T) ¡¾H(0). The curve for the variation of the specific heat capacity of C60F48 as a function of temperature, presented in Fig. 16 displays a phase transition in the temperature range of 315 ¡¾ 345 K (Ttr=329.60.4 K and DtrH=7.00:7 kJ mol71).According to NMR and X-ray diffraction data, this transition is due to the transformation of the ordered body-centred tetragonal phase into a face-centred cubic plastic phase. cp /J K71 g71 1.0 0.6 0.2 0 300 T /K 200 100 Figure 16. Specific heat capacity of C60F48 vs. temperature.56 The partial pressures of C60F48 have been measured in the temperature range of 395 ¡¾ 557 K by high-temperature mass spectroscopy using PbI2 and C60F36 as internal standards.63 The researchers suggested calculating the partial pressures (in Pa) using the equation lnapOC60F48Ua a ¢§13 146 a 24:34. T The enthalpy of sublimation ofC60F48 (see Table 7) calculated using the second law of thermodynamics is lower by 66 kJ mol71 than the enthalpy of sublimation of C60 (1753 kJ mol71), which is due to the smaller energy of intermolecular interactions in the fluorofullerene.The energy of combustion of C60F48 for burning in a rotating bomb calorimeter (724 638163 kJ mol71) and the standard molar enthalpy of formation of C60F48 in the crystalline state have been determined 64 (see Table 7). In 1992, the DfHm(gas) values for C60F48 were estimated by semi-empirical MNDO and AM1 methods.66 The values found for C60F48 isomers formed upon the addition of fluorine atoms to C60 with and without rupture of the C7C bonds in the carbon cage were77395 and75508 kJ mol71, respectively. The former value is close to the enthalpy of formation found experimentally; however, in view of the molecular structure of fluorofullerene C60F48 now established (see Section IV), which rules out the presence of =CF2 groups in the molecule, this appears to be an accidental coincidence.The energy of the C7F bond in C60F48 calculated from experimental data 64 proved to be 36% lower than that in CF4 (Fig. 17). The reason is elongation of the C7F bonds in C60F48 with respect to those in tetrafluoromethane, difluoroethane or difluorofullerene due to steric factors. O V Boltalina, N A Galeva C60F48 C60F2 CF4 C2H4F2 250 400 300 350 450 500 EC7F /kJ mol71 Figure 17. Energy of the C7F bond in some fluorocarbons. In conclusion, we would like to mention some lines of investiga- tion into direct fluorination of fullerenes which appeared recently and are, in our opinion, the most interesting and promising.In the near future, the search for the routes of the selective synthesis of fluorofullerenes, especially those with low fluorine contents (less than 36 fluorine atoms per fullerene molecule) would be contin- ued. As shown above, no convincing evidence supporting the assumption that low-fluorine compounds can form in substantial amounts in the reaction of F2 with fullerene can be found in the literature. Indirect data indicate that these compounds could be prepared by liquid-phase fluorination; however, exhaustive spec- troscopic investigation of the reaction products is still to be performed. Recent studies of Chilingarov et al.28, 67 in which fullerene mixed with transition metal difluorides is treated with molecular fluorine at elevated temperatures deserve attention.The use of transition metal difluorides as catalysts restricts substantially both the number of the products formed and the degree of saturation of fullerene molecules with fluorine. Nevertheless, it is still unclear whether this method would prove more efficient than the known solid-phase reactions involving trifluorides of metals of the first transition row. Development of methods for purification and separation of fluorination products would remain among the most topical tasks. Attempts of using various methods for the separation of fluoro- fullerene mixtures resulting from direct fluorination have been reported in the literature.In particular, a technique used fairly often for this purpose is sublimation. However, these attempts have not yet been a success because the properties of compounds with close contents of fluorine (36 to 48 fluorine atoms per fullerene molecule) are close. Nevertheless, a series of recent publications dealing with chromatographic separation of several fluorofullerenes, their oxidised forms and even fluorofullerene isomers prepared by reactions with metal fluorides demonstrated high efficiency of this method as applied to the given class of compound.68, 69 The most impressive example is isolation of>40 individual compounds from a C70Fx sample prepared by the reaction of C70 with MnF3. Among these 40 compounds, seven isomers of C70F36 alone were present.70 This review was partially financially supported by the Russian Foundation for Basic Research (Project No. 99-15-96044), the INTAS (Project No.97-30027), and the Science and Engineering Program `Fullerenes and Atomic Clusters' (the `Sphere' project). Reverences 1. G Olah, R Chambers, G K S Parkash (Eds) Synthetic Fluorine Chemistry (New York: Wiley-Interscience, 1992) 2. MHome-Grant (Ed.) Fluorine Chemistry. A Comprehensive Treatment (New York: Wiley, 1995) 3. J F Liebman, A Greenberg,WR Dolbier (Eds) Fluorine-Containing Molecules. Structure, Reactivity, Synthesis, Applications (Weinheim: VCH, 1988) 4. N Ishikawa (Ed.) Soedineniya Ftora. Sintez i Primenenie (Fluorine Compounds. Synthesis and Applications) (Translated into Russian; Moscow: Mir, 1990) 5.T Nakajima, in Fluorine ¡¾ Carbon and Fluorine ¡¾ Carbon Materials. Chemistry, Physics, Applications (Ed. T Nakajima) (New York: Marsel Dekker, 1995) p. 1Direct fluorination of fullerenes 6. N Ishikawa (Ed.) Novoe v Tekhnologii Soedinenii Ftora (New in Technology of Fluorine Compounds) (Translated into Russian; Moscow: Mir, 1984) 7. J Cioslowski Chem. Phys. Lett. 181 68 (1991) 8. D Bacowies, W Thiel Chem. Phys. Lett. 193 236 (1992) 9. G E Scuseria Chem. Phys. Lett. 176 423 (1991) 10. M S Dresselhaus, G Dresselhaus, P C Eklund, in Science of Fullerenes and Carbon Nanotubes (New York: Academic Press, 1995) p. 171 11. B I Dunlap, D W Brenner, J W Mintmire, R C Mowrey, C T White J.Phys. Chem. 95 5763 (1991) 12. P W Fowler, H W Kroto, R Taylor, D R M Walton J. Chem. Soc., Faraday Trans. 2 87 2685 (1991) 13. H Selig, C Lifshitz, T Peres, J E Fischer, A R McGhie, W J Romanov, J P McCauley Jr, A B Smith III J. Am. Chem. Soc. 113 5475 (1991) 14. O V Boltalina J. Fluorine Chem. 101 273 (2000) 15. R Teilor Izv. Akad. Nauk, Ser. Khim. 852 (1998) a 16. V F Bagryantsev, A S Zapol'skii, O V Boltalina, N A Galeva, L N Sidorov Zh. Neorg. Khim. 45 1121 (2000) b 17. J H Holloway, E G Hope, R Taylor, G J Langley, A G Avent, T J Dennis, J P Hare, H W Kroto, D R M Walton J. Chem. Soc., Chem. Commun. 966 (1991) 18. A A Tuinman, P Mukherjee, J L Adcock, R L Hettich, R N Compton J. Phys. Chem. 96 7584 (1992) 19. K Kniaz, J E Fischer, H Selig, G B M Vaughan,W J Romanov, D M Cox, S K Chowdhury, J P McCauley, R M Strongin, A B Smith III J.Am. Chem. Soc. 115 6060 (1993) 20. F Okino, H Touhara, K Seki, R Mitsumoto, K Shigematsu, Y Achiba Fullerene Sci. Technol. 1 425 (1993) 21. F Okino,H Fujimoto,R Ishikawa,H Touhara Trans. Mat. Res. Soc. Jpn. 14 1205 (1994) 22. F Okino, S Yajima, S Suganuma, R Mitsumoto, K Seki, H Touhara Synth. Met. 70 1447 (1995) 23. F Okino, S Kawasaki, Y Fukushima, M Kimura, T Nakajima, H Touhara Fullerene Sci. Technol. 4 873 (1996) 24. A Hamwi, C Fabre, P Chaurand, S Della-Negra, C Ciot, D Djurado, J Dupuis, A Rassat Fullerene Sci. Technol. 1 499 (1993) 25. A Hamwi, C Latouche, V Marchand, J Dupuis, R Benoit J. Phys. Chem. Solids 57 991 (1996) 26.Y Matsuo, T Nakajima, S Kasamatsu J. Fluorine Chem. 78 7 (1996) 27. A A Tuinman, A A Gakh, J L Adcock J. Am. Chem. Soc. 115 5885 (1993) 28. N S Chilingarov, A V Nikitin, J V Rau, L N Sidorov Fullerene Sci. Technol. 6 577 (1998) 29. A J Adamson, J H Holloway, E G Hope, R Taylor Fullerene Sci. Technol., 5 629 (1997) 30. A A Gakh, A A Tuinman, J L Adcock, R A Sachleben, R N Compton J. Am. Chem. Soc. 116 819 (1994) 31. O V Boltalina, L N Sidorov, V F Bagryantsev, V A Seredenko, A S Zapol'skii, J M Street, R Taylor J .Chem. Soc., Perkin Trans. 2 2275 (1996) 32. V F Bagryantsev, A S Zapol'skii, O V Boltalina, N A Galeva, L N Sidorov Dokl. Akad. Nauk 357 487 (1997) c 33. Y Kita, N Watanabe, Y Fujii J. Am. Chem. Soc. 101 3832 (1979) 34.O V Boltalina, A Ya Borschevskii, L N Sidorov, J M Street, R Taylor J. Chem. Soc., Chem. Commun. 529 (1996) 35. S J Austin, P W Fowler, J P B Sandall, F Zerbetto J. Chem. Soc., Perkin Trans. 2 155 (1996) 36. R Mitsumoto, T Araki, E Ito, Y Ouchi, K Seki, K Kikuchi, Y Achiba, H Kurosaki, T Sonoda, H Kobayashi, O V Boltalina, V K Pavlovich, L N Sidorov, Y Hattori, N Liu, S Yajima, S Kawasaki, F Okino, H Touhara J. Phys. Chem. 102 552 (1998) 37. H Steger, U Mische, W Kamke, A Ding,M Fieber-Erdmann, T Drewello Chem. Phys. Lett. 39 (1997) 38. Yu V Vasil'ev, O V Boltalina, L N Sidorov, R F Tuktarov, V A Mazunov Int. J. Mass Spectrom. Ion Proces. 173 113 (1998) 39. G Gensterblum, K Heversi, B Y Han, LMYu, J Pireaux, P A Thiry, R Caudano, A A Lucas, D Bernaerts, S Amelincx, G van Tendello, G Bendele, T Buslaps, R L Johnson,M Foss, R Feidenhanst, G Lelay Phys.Rev. B, Condens. Matter 50 11 981 (1994) 40. K Kanno Mater. Sci. Forum 239197 (1997) 621 41. DMCox, S D Cameron, A A Tuinman, A A Gakh, J L Adcock, R N Compton, EWHagaman, K Kniaz, J E Fischer, RMStrongin, M A Cichy, A B Smith III J. Am. Chem. Soc. 116 1115 (1994) 42. H Touhara, F Okino, S Yajima, S Suganuma, K Seki, R Mitsumoto, H Fujimoto Trans. Mat. Res. Soc. Jpn. 14 1173 (1994) 43. P J Benning, T R Ohno, J H Weaver, P Mukherjee, J L Adcock, R N Compton, B I Dunlap Phys. Rev. B, Condens. Matter 47 1589 (1993) 44. N A Galeva, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1998 45. A Hamwi, V Marchand Fullerene Sci. Technol. 4 835 (1996) 46. S Kawasaki, T Aketa, F Okino, H Touhara, O V Boltalina, I Gol'dt, S I Troyanov, R Taylor J. Phys. Chem. 103 1223 (1999) 47. F Zhou, G J Van Berkel, B T Donovan J. Am. Chem. Soc. 116 5485 (1994) 48. N Liu, H Touhara, F Okino, S Kawasaki, Y Nakacho J. Electrochem. Soc. 143 2267 (1996) 49. Y Matsuo, T Nakajima Electrochim. Acta 47 15 (1996) 50. V P Glazkov, V P Zhukov, V A Somenkov, S Sh Shil'shtein Poverkhnost. Rentg. Sinkhrotron. Neitron. Issled. (7) 45 (1997) 51. H Selig, K Kniaz, G B M Vaughan, J E Fischer, A B Smith Macromol. Symp. 82 89 (1994) 52. N A Galeva, A I Druzhinina, I V Gol'dt, FM Spiridonov, R M Varushchenko, O V Boltalina, L N Sidorov, V I Privalov, V Agafonov, H Allouchi, in The 15th International Conference on Chemical Thermodynamics (Abstracts of Reports), Porto, 1998 PW1-4 53. O V Boltalina, I Gol'dt, T Zatsepin, FM Spiridonov, S I Troyanov, L N Sidorov, R Taylor In Molecular Nanostructures (Ed. H Kuzmany) (World Scientific Publ., 1997) p. 114 54. R Taylor, J H Holloway, E G Hope, A G Avent, G J Langley, T J Dennis, J P Hare, H W Kroto, D R M Walton J. Chem. Soc., Chem. Commun. 665 (1992) 55. O V Boltalina, A K Abdul-Sada, R Taylor J. Chem. Soc., Perkin Trans. 2 981 (1995) 56. V I Privalov, O V Boltalina, N A Galeva, R Teilor Dokl. Akad. Nauk 360 499 (1998) c 57. A A Gakh, A A Tuinman, J L Adcock, R N Compton Tetrahedron Lett. 34 7167 (1993) 58. R Taylor, G J Langley, J H Holloway, E G Hope, A K Brisdon, H W Kroto, D R M Walton J. Chem. Soc., Perkin Trans. 2 181 59. O V Boltalina, V Yu Markov, R Taylor,M P Waugh J. Chem. Soc., (1995) Chem. Commun. 2549 (1996) 60. A G Avent, O V Boltalina, A Yu Lukonin, R Taylor Chem. Commun. 436 (1998) 61. O V Boltalina, J Street, R Taylor Chem. Commun. 1827 (1998) 62. A I Druzhinina, N A Galeva, R M Yarushchenko, O V Boltalina, L N Sidorov, in Fullerenes, Recent Advances in the Chemistry and Physics of Fullerenes and RelatedMaterials (Eds KMKadish, R S Ruoff) (Pennington: Electrochemical Society, 1998) Vol. 6, p. 1189 63. O V Boltalina, V Yu Markov, A Ya Borschevskii, N A Galeva, L N Sidorov, G Gigli, G Balducci J. Phys. Chem. 103 3828 (1999) 64. T S Papina, V P Kolesov, V A Lukyanova, O V Boltalina, N A Galeva, L N Sidorov J. Chem. Thermodyn. 31 1328 (1999) 65. A I Druzhinina, N A Galeva, R M Varushchenko, O V Boltalina, L N Sidorov J. Chem. Thermodyn. 31 1469 (1999) 66. N Matsuzava, T Fukunaga, D A Dixon J. Phys. Chem. 96 10747 (1992) 69. O V Boltalina, J H Holloway, E G Hope, J M Street, R Taylor 67. N S Chilingarov Dokl. Akad. Nauk 356 774 (1999) c 68. O V Boltalina, J Street, R Taylor J. Chem. Soc., Perkin Trans. 2 650 (1998) J. Chem. Soc., Perkin Trans. 2 1845 (1998) 70. R Taylor, A K Abdul-Sada, O V Boltalina, J Street J. Chem. Soc., Perkin Trans. 2 1013 (2000) a�Russ. Chem. Bull. (Engl. Transl.) b�Russ. J. Inorg. Chem. (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC

摘要:

Russian Chemical Reviews 69 (7) 623 ± 638 (2000) The formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys V V Chesnokov, R A Buyanov Contents I. Introduction II. The carbide cycle mechanism of carbon formation III. Regularities in the formation of carbon of different morphological types on iron subgroup metals IV. Conclusion Abstract. upon formed carbon filamentous of structure The The structure of filamentous carbon formed upon catalytic subgroup iron on hydrocarbons of decomposition catalytic decomposition of hydrocarbons on iron subgroup metals metals and deposition the of regularities The considered. is alloys their and their alloys is considered. The regularities of the deposition of of carbon cycle carbide The generalised.are metals these on carbon on these metals are generalised. The carbide cycle mech- mech- anism of carbon formation is considered in detail. The growth anism of carbon formation is considered in detail. The growth models of modifications morphological some of models of some morphological modifications of filamentous filamentous carbon references 151 includes bibliography The discussed. are carbon are discussed. The bibliography includes 151 references. I. Introduction Deactivation of catalysts may be determined by a number of factors,1 one of which is carbonisation (coking) of catalysts in the processing of hydrocarbons. The development of an efficient method for combating this phenomenon would allow substantial extension of the lifetime of catalysts; therefore, the interest of investigators in the mechanisms of catalytic formation of carbon has not decreased for many decades.This interest is also due to the fact that in a number of cases, the carbonised mineral phases (metals, oxides, and others) and carbon itself can be regarded as carbon-mineral and carbon composite materials possessing spe- cial properties. The formation of filamentous carbon upon decomposition of hydrocarbons on iron subgroup metals and their alloys is one of the most exotic (from the viewpoint of its mechanism) and practi- cally important processes. The properties of carbon graphite deposits on catalysts are mainly determined by their morpholog- ical and crystallographic features. The main task in the develop- ment of technologies for the preparation of a broad spectrum of such composites should be the elaboration of a scientifically substantiated procedure of process control.To the end, it is necessary to know the mechanisms underlying all its stages. II. The carbide cycle mechanism of carbon formation 1. Decomposition of hydrocarbons on metallic iron Various suggestions on the possible role of carbides as intermedi- ates in the formation of carbon deposits were put forward in early V V Chesnokov, R A Buyanov G K Boreskov Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences, Prosp. Akad. Lavrent'eva 5, 630090 Novosibirsk, Russian Federation. Fax (7-383) 234 37 66. Tel. (7-383) 234 45 53.E-mail: buyanov@catalysis.nsk.su (R A Buyanov) Received 28 February 2000 Uspekhi Khimii 69 (7) 675 ± 692 (2000); translated by V D Gorokhov #2000 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2000v069n07ABEH000540 623 623 626 636 studies generalised by Storch et al.2 in their monograph nearly five decades ago. It has been noted, in particular, that the hypothesis according to which carbides are intermediates in the formation of free carbon is attractive but stands in need of evidence. Kagan et al.3 considered carbonisation of an iron catalyst to be the result of two independent reactions 2CO Fe2C + CO2 . C + CO2 and 2 Fe + 2CO However, a later study proved 4 that the important parts of the process of carbon formation in the decomposition of hydro- carbons on iron are the steps of carbide formation and decom- position.The proposed carbide mechanism of carbon formation could be confirmed by the establishment of a common limiting step and a common intermediate in the formation of a carbide and carbon. To this end, the activation energies of carbon and iron carbide formation 5, 6 were measured in the experiments on carbonisation of iron in the butadiene decomposition. The formation of the carbide in the temperature range 573 ± 723 K is virtually unac- companied by its decomposition and liberation of carbon. The formation of iron carbide Fe3C is a topochemical reaction, i.e., it occurs through formation and growth of nuclei of the carbide phase.An integral kinetic curve of the reaction is S-shaped; therefore, an approximate formula 7 is used to determine the specific rate of carbidisation W=2Wmax pS , where Wmax is the maximum observed rate of carbidisation, S is the specific surface of a sample. The activation energy was determined from theWmax values at different temperatures. In the range of temperatures 808 ± 874 K, the activation energy of carbide decomposition (197 kJ mol71) is substantially higher than that of its formation (88 kJ mol71). For this reason, the ratio between the rates of these reactions changes with temperature. Above 1023 K, the rate of iron carbide decomposition exceeds the rate of its formation, and the carbide phase is not formed. Investigations into the kinetics of carbon formation on iron showed that the activation energy of carbon formation at 1023 ± 1027 K amounts to 96 kJ mol71 and becomes close to that of carbide formation. Obviously, the common limiting step of both reactions is the formation of an intermediate, surface carbide-like compound of the type [Fe7C].This makes it possible to represent the mechanism of formation of carbon deposits on iron with the following scheme: C4H6 Fe + C . Fe [Fe7C]624 In our studies, this mechanism has been called the carbide cycle mechanism.4± 6,8± 17 The reactions proceed in an open flow system from which the gaseous decomposition products are removed; therefore, the thermodynamic restrictions here are of minor significance.The carbide cycle mechanism includes two basic steps:5 the `chemical' step in which the catalytic decomposition of hydro- carbons occurs on the surface of a metal particle with the formation of carbon atoms whose concentration increases to definite limiting values, and the `physical' step in which the crystallisation centres (nuclei) of the graphite phase are formed on definite faces of a metal particle. In this step, migration (diffusion in the bulk of the metal particle) of carbon atoms towards these centres and growth of a definite variety of graphite particles, predominantly in the form of filaments begin. Despite the fact that the formation of carbon comprises two steps, the mechanism, on the whole, was called the mechanism of the carbide cycle since the carbon atoms involved in the graphite formation `originate' from intermediate carbide compounds.The character of carbon deposits is determined by the proper- ties of metal catalyst particles, the nature of a hydrocarbon, as well as by the distinctive features of crystallisation centres and mass- transfer of carbon atoms. The mechanism of the carbide cycle considered above in the example of the model system Fe+C4H6 can also be extended to the decomposition of other hydrocarbons. Naturally, the activa- tion energies of formation of carbon deposits from different hydrocarbons differ substantially: e.g., for methane, butane, propylene, isobutylene and butadiene they are 6, 18 209, 184, 146, 138 and 96 kJ mol71, respectively.The relationship between the activation energies (Ea) and enthalpies (DH) of the iron carbide formation from different hydrocarbons (Fig. 1) is described by the BroÈ nsted ± Polanyi equation. Ea/kJ mol71 C4H10 200 CH4 C3H6 150 C4H8 C4H6 60 DH/kJ mol71 720 0 20 Figure 1. Dependence of the activation energy of carbon formation on the enthalpy of iron carbide formation in the decomposition of various hydrocarbons. The formation of carbon from aromatic hydrocarbons can also occur by the carbide cycle mechanism. This was demon- strated in the example of benzene decomposition on metallic iron.19 At temperatures above 773 K, the catalyst in the steady state is in the form of metal, whereas below this temperature the metallic iron passes to the carbide.The formation of carbon from benzene on metallic iron in the temperature range 773 ± 998 Kcan be described by the scheme C6H6 [Fe7C6H5] (and [Fe7C6H4]) Fe + C . [Fe7C] At temperatures above *898 K, the limiting step is the dissociative adsorption of benzene accompanied by the formation of C6H5 or C6H4 species,20 ± 22 while below this temperature the reaction is limited by the scission of the benzene ring with the formation of intermediate carbide-like compounds. V V Chesnokov, R A Buyanov 2. The role of topochemical processes in the mass-transfer of carbon in the carbide cycle mechanism Studies of the carbon phase formation on metallic iron suggest the elucidation of not only the mechanism of decomposition of hydrocarbons with liberation of carbon atoms, but also of the mode of generation and formation of a new phase from these atoms.The theory of solid-phase reactions 23 and the theory of crystallisation 24 are based on the assumption that the initial formation of nuclei of a new phase is often the limiting step of these processes. Under the conditions of topochemical process, the nuclei are formed on the active centres the role of which can be played by surface defects, associates, clusters, sites of the exit of dislocations onto the surface of crystals, etc. In the decomposition of hydrocarbons accompanied by the formation of intermediate carbide-like compounds, carbon atoms are liberated which diffuse into the bulk of the metal and form there a solid solution.13 The question arises of how the nuclei of the new phase (graphite) are formed and how does the transport of carbon to these nuclei occur.In accordance with the carbide cycle mechanism,4 ± 6, 10 ± 18, 25 the subsurface layer of the metal becomes supersaturated (by over two orders of magnitude) with the carbon liberated in the decomposition of intermediate carbide compounds. Indeed, the decomposition of intermediate carbide compounds forms a mix- ture of carbon and the metal in which the C: Fe ratio appears to correspond to the composition Fe3C (*6 mass%± 7 mass% of carbon), whereas the saturated solution of carbon in iron contains not more that 0.025 mass% of carbon.26 The concentration of carbon is averaged over the bulk of the metal, and this results in the appearance of a graphite nucleus at one of the sites of the metal particle surface.The growth of the graphite phase occurs by virtue of excess carbon in the supersaturated solution. High gradients of carbon concentration appear between the surface site where decomposition of hydrocarbons takes place and the site where the graphite phase is formed. The presence of such a gradient ensures the diffusive transport of carbon atoms through the bulk of a metal particle. The appearance of a concentration gradient in a solution of carbon in the metal is explained by the specific features of carbon formation through the carbide cycle mechanism and by the difficulties of graphite nuclei formation. The above considerations make it possible to present a general scheme of all the steps of the graphite phase formation.The activation energies of deposition of carbon formed upon decom- position of acetylene on a-iron (Ec) and diffusion of carbon through the a-iron phase (Ed) are close,27 their values are within the range 67 ± 80 kJ mol71. This is only possible provided the activation energy of decomposition of hydrocarbons by the carbide cycle mechanism (E 0) is smaller than Ed. In this case, the growth of the carbon phase is limited by the step of diffusive transfer of carbon, which predetermines the apparent equality Ec'Ed. However, the activation energy of the decomposition of a number of more stable hydrocarbons by the carbide cycle mech- anism may be rather high and E0>Ed; in this case it is not the carbon diffusion but rather the formation of intermediate com- pound [Fe7C], i.e., decomposition of hydrocarbons, that becomes the limiting step.Then the values of E0 and Ec coincide. Thus, the diffusion of carbon in the metal from the site of its liberation towards the nucleus or the graphite phase on which it condenses is an important step in the catalytic process of carbon deposition on metals.13 Deposition of carbon on iron subgroup metals occurs at increased temperatures when carbides of these metals are not formed (for Fe, Ni and Co, these temperatures amount to 1023, 673 and 573 K, respectively). Relevant experiments have shown 28 that the formation of graphite nuclei is a more energy-consuming process than the carbide phase formation.This may be explained by the fact thatThe formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys carbon is formed by the carbide cycle mechanisms through an intermediate surface carbide compound which is transformed into carbide after saturation with carbon from its solid solution in the metal.13 Under definite conditions, the deposition of carbon can begin with the formation of a thermodynamically unstable meta- stable carbide phase.29 ± 32 The carbide formation accompanied by the liberation of elementary carbon was observed, in particular, in CO disproportionation on iron.7 In this case, the accumulation of carbon in the carbide phase may be regarded as only an inter- mediate step in the formation of graphite nuclei. Owing to this step, the concentration of carbon in the metal bulk may be considerably higher than merely in the saturated solution of carbon in the metal. 3.The formation of carbon on iron carbide As noted above, under definite conditions a metal catalyst under- goes some changes under the action of the reaction medium and can exist in the form of a metastable carbide in the steady state. For this reason, irrespective of whether the metal or the carbide is the active origin of the process of carbon formation, one can say that carbon will be formed by the carbide cycle mechanism. Accumulation of carbide-like compounds leads first to the appearance of the Fe3C phase nuclei and then to the complete phase transformation of the metal into the carbide. The carbide mechanism of the formation of carbon deposits on Fe3C is preserved in any case; however, the features of its implementation are determined by the chemical nature of the active phase.The chemical nature of the active phase is primarily man- ifested in the step of decomposition of an intermediate carbide- like compound. The limiting step of the formation of carbon deposits on the metal is the formation of a carbide-like compound, whereas its decomposition and aggregation of the liberated carbon atoms occur rather quickly. The limiting step of carbon formation on the carbide is the liberation of carbon atoms from the carbide, the carbide passing to a transient state with a disturbed stoichiometry.The disturbed stoichiometry of the sur- face carbide is recovered upon hydrocarbon decomposition. As the bulk of a catalyst contains a considerable amount of carbon, the diffusive transfer of carbon atoms through the carbide bulk virtually does not take place, while their surface migration is negligible. In this case, the carbon phase is formed as islets or a continuous film shielding the catalyst's surface. The rate of carbide decomposition increases with temperature. Direct measurements of the initial rate of Fe3C decomposition were performed using a high-temperature X-ray chamber.6, 12 It was established that a-Fe2O3 undergoes various trans- formations under the action of the reaction medium (buta-1,3- diene diluted with helium at the molar ratio 1 : 10) Fe3C , Fe3O4 a-Fe a-Fe2O3 which occur over 10 min; after this the composition of the catalyst corresponds to iron carbide and remains unchanged with time.The contents of a-Fe and Fe3C in a sample were estimated from the heights of the corresponding peaks (110) and (210) on the diffractograms. In an inert medium, iron carbide decomposes into carbon and a solid solution of carbon in a-Fe (ferrite). The initial rate of Fe3C decomposition coincided with the rate of carbon formation in the presence of the carbide phase.6 This fact confirms that the formation of carbon deposits is limited by the step of carbon liberation from iron carbide.As a result, iron carbide is depleted in carbon to give a non-stoichiometric carbide with which a hydrocarbon interacts. The carbon atoms formed during decomposition of hydrocarbon by the carbide mechanism are again involved in finishing the `build-up' of the carbide. X-Ray structural studies showed that the interplanar distances in the carbide correspond to the standard values. Consequently, in the course of reaction the system contains stoichiometric carbide Fe3C, while the slow step is the liberation of carbon from it. A detailed scheme of the catalytic process of formation of carbon deposits by the carbide cycle mechanism has the following form: C4H6 Fe [Fe7C] where [Fe7C] is the carbide-like intermediate compound, [Fe3C17x] n is the intermediate surface carbide with disturbed stoichiometry, nx=1.Depending on the conditions, either the carbide phase or the metal phase can be stable; although, naturally, only one phase exists in the steady state. The reaction route 1 corresponds to the formation of carbon from hydrocarbons on a metal without formation of the carbide phase. Reaction 2 describes the phase chemical transformations occurring under the action of the reaction medium. Reaction 3 is related to the liberation of carbon and the subsurface non-stoichiometric carbide which can be saturated with carbon under the action of gaseous hydrocarbons. 4. Specific features of the carbide cycle mechanism in the formation of carbon on nickel The above-cited scheme of the carbide cycle describes the proc- esses occurring on nickel, cobalt and a number of other metals.Naturally, individual features of each metal catalyst are mani- fested in individual steps of the formation of carbon deposits, though this takes place within the limits of the regularities described. Let us consider, for example, the features of formation of carbon deposits from butadiene on nickel.6, 15 Using a high-temperature X-ray chamber, it was shown that in the hydrocarbon medium, nickel carbide is formed at temper- atures below 623 K. In this case, deposition of carbon proceeds very slowly and does not change the kinetics of carbide formation, which was therefore studied in the temperature range 548 ± 623 K, while the kinetics of carbon formation was studied at 773 ± 923 K.The former reaction is of first order with respect to butadiene and is characterised by an activation energy of 96 kJ mol71, while the latter is zero order and has an activation energy of 134 kJ mol71. These distinctions indicate that the reactions of carbide formation and carbon depositions on nickel have different limit- Table 1. Coefficients of carbon diffusion through iron subgroup metals and their temperature dependences. Diffusion coefficient T/K Ref. D/cm2 s71 Iron 773 873 41 43 26 41 43 26 261078 161078 261078 161077 561078 861077 Cobalt 3610711 44 5.74610710 44 773 873 Nickel 773 873 41 42 41 42 8610711 2.5610711 8610710 2.5610710 a Activation energy Ea is expressed in kJ mol71.625 1 Fe + C3 2 Fe3C [Fe3C17x]n + nxC C4H6 Temperature dependence a Ref. D=D0 exp(7Ea/RT) D=3.961073 exp(780.3/RT) 43 D=861073 exp(783/RT) 26 D=0.21 exp(7147/RT) 44 D=0.1 exp(7138/RT) 41 D=2.48 exp(7168/RT) 42626 ing steps. The value of Ea for the second reaction is 134 kJ mol71, which corresponds to the activation energy of carbon diffusion through this metal.27, 33 ± 39 Consequently, it may be presumed that the limiting step of carbon formation on metallic nickel is its diffusion through nickel to the nuclei and the graphite phase which is being formed. Thus, the limiting steps of the processes of formation of carbon deposits on iron and nickel are different. This is explained by the fact that the coefficient of carbon diffusion in a-Fe is nearly three orders of magnitude larger than that in nickel (Table 1).40 ± 44 At temperatures below 623 K, the formation of Ni3C carbide occurs considerably faster that its decomposition; in this temper- ature range, carbon is formed on nickel carbide.Substitution of benzene or hexane for butadiene does not affect the rate of carbon formation. This may indicate that the formation of carbon from the above-mentioned hydrocarbons on nickel is limited by the same step, i.e., the diffusion of carbon atoms through the bulk of the metal. 5. Regularities of carbon formation from methane on nickel The results of studies of the nature of the active phase of Ni/SiO2 catalysts in the reforming of methane by carbon dioxide con- firmed the important role of the surface layers of nickel carbide in both implementation of the major transformation and formation of carbon.45 The reaction proceeds on the surface layer which is formed in the beginning of the process and which is close in its composition to nickel carbide.The adsorbed forms of carbon resulting from the activation of methane on the nickel surface are in equilibrium with methane in the gas phase. They can be irreversibly transformed to CO by reacting with adsorbed oxygen (in different forms), which results from the activation of carbon dioxide. A portion of carbon in the amount equivalent to its content in the subsurface carbide layer is rapidly dissolved in the bulk of nickel.This carbon diffuses into the bulk of the metal and is involved in the formation of carbon deposits. Owing to high thermodynamic stability of methane, its behaviour in the process of nickel carbonisation is different from that of other hydrocarbons. Chemisorption and subsequent decomposition of methane on the surface of nickel single crystals were studied by high-resolution spectroscopy of characteristic electron energy losses and by Auger spectroscopy.46 ± 55 It was established that at lower temperatures, methane is chemisorbed in the dissociative mode with formation of the methyl radical and a hydrogen atom. On temperature increase from 373 to 600 K, the radical CH3 undergoes stepwise dehydrogenation C. CH CH2 CH3 Schouten et al.56 carried out experiments at 600 K and have proposed a similar scheme of the formation of surface nickel carbide: CHx Ni3C .CH4 Kinetic studies of methane decomposition were performed at 723 ± 823 K in a catalytic reactor with a microbalance.57 ± 59 Analysis of the results obtained was based on the model of Grabke 60, 61 according to which the decomposition of methane on a-Fe and g-Fe surfaces occurs as a result of successive splitting of hydrogen atoms, while the limiting step is the decomposition of the methyl radical. On the other hand, Alstrup and Tavares 57, 59 showed that the limiting step could be both dissociative adsorp- tion of methane and dehydrogenation of adsorbed methyl groups, whereas Snoeck et al.62 state that this process is limited by the step of detachment of the first hydrogen atom from the methane molecule adsorbed on the nickel surface.Our kinetic studies of methane decomposition on alumi- nium ± nickel catalysts 63 showed that the activation energy of carbon formation on nickel decreases as the temperature increases. In the temperature range 798 ± 873 K, the activation energy is 75 kJ mol71, which is close to the literature data [cf. V V Chesnokov, R A Buyanov 88 kJ mol71 (Ref. 58), 90 kJ mol71 (Ref. 59), 96 kJ mol71 (Ref. 64) and 97 kJ mol71 (Ref. 65)]. Apparently, the limiting step is dehydrogenation of the adsorbed methyl radical with formation of surface nickel carbide, which is confirmed by the study of the kinetics of nickel carbide formation from methane on the (110) face of a nickel single crystal.56 The activation energy of the surface nickel carbide formation is 8712 kJ mol71 (see Ref.56). In the temperature range 923 ± 1023 K, the activation energy of carbon formation from methane decreases to 26 kJ mol71. Similar values of the activation energy of carbon formation from methane (0 ± 40 kJ mol71) were obtained for the nickel-contain- ing catalysts,66, 67 as well as for the (111) and (100) faces of nickel.50, 51, 68 The value 26 kJ mol71 coincides with the exper- imental Ea, which characterises the dissociative adsorption of methane on the face (100) of the nickel surface with formation of the methyl radical and a hydrogen atom.68 This makes it possible to conclude that the limiting step of methane decomposition on nickel catalysts at 923 ± 1023 K is the dissociative adsorption of methane with formation of a hydrogen atom and methyl radical.III. Regularities in the formation of carbon of different morphological types on iron subgroup metals 1. Varieties of carbon deposits Carbon formed on the iron subgroup metals or their alloys with other metals is produced in the form of deposits possessing various crystallographic characteristics, which depend on the specific features of catalysts and conditions of the process. Using these dependences, one can purposefully control the technologies of preparation of carbon deposits of various morphologies.69 Car- bon can be deposited as filaments of different configuration, nanotubes, plates and other forms.The most interesting varieties are characteristic of the forms of filaments. Figure 2 shows three basal structures of carbon fila- ments.a b c Figure 2. The structures of graphite filaments with different positions of basal planes; (a) coaxial-cylindrical; (b) coaxial-conical; (c) flat-parallel (in the form of stacks of plates of graphite planes). Carbon nanotubes have first been revealed as a concomitant product in the preparation of fullerenes by vaporisation of graph- ite in an electric arc.70 It was shown later that nanotubes could also be obtained by catalytic methods.71 ± 73 Nanotubes differ from carbon filaments primarily in their smaller sizes. Carbon nano- tubes consist of 1 ± 50 cylindrical graphite-like layers coaxially inserted into each other;70, 74, 75 the diameters of cylindrical cavities are 0.7 ± 6 nm, whereas the lengths of the tubes reach several micrometers.In a chemical respect, the nanotubes obtained by graphite vaporisation differ from the known forms of carbon in their increased resistance to oxidation.76 2. Regularities of the formation of filamentous carbon The growth of carbon filaments is catalysed by finely dispersed metal particles. The limits of sizes were established outside which the metal particles are coated with carbon and loose their abilitiesThe formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys to perform catalytic functions.77 ± 80 The lower limit is *30 A, while the upper limit depends on the coefficient of carbon diffusion in the bulk of the metal particle.For nickel and cobalt, these are *1000 ± 1500 A. The carbon diffusion coefficient in iron is 2 ± 3 orders of magnitude larger than that in nickel and cobalt; therefore, the upper limiting size of iron particles is at least one order of magnitude larger.79 The growth of carbon filaments on finely dispersed particles of the iron subgroup metals and their alloys with some other metals is a physical step of the carbide cycle mechanism.5, 6, 9 ± 18, 77 ± 79 This process is to some extent unique and is of great interest for fundamental science. The growth of carbon filaments was directly observed in the gas chamber of an electron microscope.33, 35 The following regu- larities were revealed: (1) a particle of a catalyst is present at the growing ends of each filament; (2) the diameter of filaments is controlled by the size of the catalytic particle at the growing end, which is fragmented in most cases, and this initiates growth of finer filaments; (3) the diameters of filaments range from 100 to 1500 A, while their lengths can reach 86104 A; (4) the grown filaments have various forms (loops, helices, interwoven mass); (5) the dependence of the rate of filament growth on time is S-shaped; (6) the time for the carbon filament formation depends on the reaction temperature: at 923 K, the growth of some filaments lasts for 10 min, while at 1243 K their growth stops after 30 s.The catalytic particle at the end of the filament is pear- shaped.33, 35 An electron-transparent channel often passes along the entire filament length to its top. Reasonably, the filament has an external layer which is more resistant to oxidation than the internal one. The thickness of the external layer makes up*10% of the outer diameter. In order to understand the regularities of formation of carbon filaments, it is necessary to answer at least the following questions. (1) Why carbon atoms are formed predominantly on one crystal face of a metal particle and crystallise into graphite structures on other faces? (2) What are the driving forces for the migration of carbon atoms to the sites of graphite phase formation and what are the peculiarities of the mass-transfer of carbon atoms? (3) What are the conditions for the appearance of the graphite phase crystallisation centres and specific features of these centres? (4) How do the crystallographic and crystallochemical fea- tures of metal particles influence the properties of graphite structures formed on them? (5) What possibilities exist for the targeted formation of graphitic phases of different morphological types? There are two explanations of the reasons for the transfer of carbon atoms from the front part of catalytic particles where they are formed to the rear side where the filament grows.Some investigators presume that this transfer occurs owing to the diffusion of carbon atoms over the surface of the catalytic particle.81 However, the more popular is another viewpoint according to which the growth of filamentous carbon occurs as a result of diffusion of carbon atoms in the bulk of the metal particle.33 ± 35, 39, 82 ± 91 The mechanism of filamentous carbon growth proposed by Baker et al.33, 35 is presented schematically in Fig.3. After formation of small metal particles on a support (Fig. 3 a), depo- sition of amorphous carbon takes place at the site of particle contact with the support (Fig. 3 b). The formation of amorphous carbon is explained by the gas-phase polymerisation of acetylene. At the same time, acetylene is decomposed on a metal particle surface, and the carbon atoms are dissolved in the bulk of the metal.Decomposition of acetylene and solubilisation of carbon are accompanied by heat release, and this creates temperature gradient along the particle. Owing to this gradient, carbon moves 627 b a c d Cessation of filament growth Excess carbon C C C2H2 C Metal Support Figure 3. Schematic representation of the mechanism of the filamentous carbon growth from acetylene on a metal (see the text for comments). to a colder part of the particle in contact with the support. The deposited carbon raises the metal particle and removes it from the support. The initial filament growth causes curvature of the metal particle which becomes pear-shaped, and this in turn leads to the formation of a channel (Fig.3 c). The increase in the filament growth rate is explained by the insulation of the particle from the support and the corresponding increase in the temperature gradient. The excess carbon formed on the outer surface of the particle migrates from this surface and builds the external wall of the filament. It is presumed that the external wall of the filament has a different structure from that of its core. The filament growth stops when the `hot' side of the particle is covered by carbon (Fig. 3 d ). In the authors' opinion,35 their hypothesis is confirmed by the fact that the activation energies of the growth of carbon filaments on a-Fe, Co, Cr and Ni are close to the activation energies of carbon diffusion in the bulk of the corresponding metal.However, the values of the growth rates of carbon filaments on different metals 33, 35 are contradictory. If the formation of filaments is governed only by the diffusion of carbon atoms in the correspond- ing metal, then under the same experimental conditions (1010 K, acetylene as the carbon source, catalytic particles of the same sizes), the growth rates should be proportional to the diffusion coefficient of carbon atoms in the corresponding metal. As noted above, the coefficient of carbon diffusion in metallic iron is 2 ± 3 orders of magnitude larger than that in cobalt or nickel. Con- sequently, the rate of growth of carbon filaments on iron should be proportionally higher. However, in practice it proved to be lower on iron than on nickel or cobalt.33, 35 In this context, it is reasonable to suggest that the limiting step on iron is the interaction of acetylene with the surface of the metal.The driving force of diffusion was presumed 33, 35 to be the temperature gradient arising between the sites where the decom- position of hydrocarbons occurs and the sites where the graphite phase grows. The higher the metal temperature the larger the solubility of carbon in it; for this reason, concentration gradient also appears. The concept according to which the growth of filamentous carbon is brought about by carbon diffusion in a crystal has been accepted by most investigators, whereas the appearance of mass gradient determined by the temperature gradient has become the subject of lively discussion.81, 82, 84 ± 87, 92 The arguments of the opponents were based on the fact that the hydrocarbons, the decomposition of which is an endothermic process (e.g., methane), also provide the growth of carbon filaments.The suggestion according to which the deposition of carbon is explained by the presence of admixtures of other hydrocarbons in methane was disproved by the fact that with high-purity methane (99.99 vol.%) the degree of its decomposition reached 30%.87 In order to assess the magnitude of the temperature gradient, Tibbetts et al.93 have simulated the process of the growth of the carbon filaments from acetylene on g-Fe. To simplify calculations, they suggested that the filament has no hollow channel and its628 diameter is equal to the diameter of the metal crystal initiating the filament growth.The thermal balance of the growing filament end was calculated on the assumption that all the heat released in the decomposition of hydrocarbons is absorbed by the metal particle. Calculations showed that the temperature gradient is smaller than 261073 K, and such a small magnitude cannot ensure the really observed growth rates of carbon filaments. Nonetheless, unsuc- cessful attempts were undertaken to prove a substantial influence of the temperature gradient.94 ¡¾ 96 An alternative explanation of the nature of the driving force of diffusion is based on the appearance of concentration gradient. In the model proposed by Rostrup-Nielsen and Trimm,97 it was suggested that the solubility of carbon at the hydrocarbon ¡¾ metal boundary is higher than that at the metal ¡¾ graphite boundary.However, the data on solubility used in this model and based on the results reported by Wada et al.88 are questionable. The hypothesis of concentration gradient was also supported by the authors of a series of studies,98 ¡¾ 100 who investigated the growth of filamentous carbon upon disproportionation of CO on Co/Fe alloys. It was presumed that the activity of carbon at the gas-metal interface is determined by the carbon activity in the gas phase. The activity of carbon at the metal ¡¾ filament interface was assumed to be equal to the activity of graphite. On the basis of this simple assumption, a kinetic equation was deduced according to which the rate of carbon deposition depends linearly on the carbon activity in the gas phase, However, the explanation of the reasons for the appearance of the concentration gradient 97 ¡¾ 100 appears to be supported with insufficient argumentation.Based on the carbide cycle mechanism, a model of the filamentous carbon growth was constructed which does not contradict most experimental results.5, 6, 15 We believe that the mass-transfer of carbon atoms occurs owing to their diffusion in the bulk of metal particles from the formation site to the crystallisation centres. Depending on the ratios of the rates of formation and diffusion of carbon atoms, the limiting steps of the process may be different.Let us consider the growth of carbon filaments under the conditions where the limiting step is the diffusion of carbon atoms in the metal particle. In a saturated solution, the solid and dissolved phases possessing equal chemical potentials are in equilibrium. In our case, however, carbon on the front and rear sides of the metal particle is in different states which are characterised by different chemical potentials. The pool of carbon on the front side of the particle is continually replenished due to the decomposition of hydrocarbons. On the rear side, carbon is formed as the graphite phase. The concentrations of saturation (solubility) of carbon at these sites are substantially different; therefore, a concentration gradient arises, which ensures continuous dissolution of carbon on the front side and its diffusive mass-transfer to the rear side of the metal particle.This gradient is rather high. The concentration of carbon atoms on the front side of the nickel particle formed by decomposition of intermediate carbide-like compounds is close to the concentration of carbon in Ni3C carbide, i.e., it reaches 25 at. %. This value is confirmed by the calculation of the linear rate of carbon filament growth.15 The linear rate of the carbon filament growth (V) is deter- mined by the amount of carbon Q which has diffused through an area section unit of a metal particle in unit time. Q a DOc1 ¢§ c2U and V a DOc1 ¢§ c2U , L Ld where D is the coefficient of carbon diffusion in the metal, c1 is the carbon concentration in the surface carbide-like compound, c2 is the carbon concentration in the saturated solution of carbon in nickel on the rear side of the metal particle, L is the diameter of the metal particle at the end of the filament, d is the density of graphite.These formulae were used to calculate the rates of the growth of carbon filaments 600 A in diameter on metallic nickel at 873 K. V V Chesnokov, R A Buyanov Decomposition of an intermediate carbide-like compound of the Ni3C type gives the concentration c1=0.514 g cm73; the values of c2=3.8610 g cm73, D=3.89610710 cm2 s71, d= 2 g cm73 were taken from the monograph of Byuanov.5 The growth rate of carbon filaments found (1650 A s71) is close to the values measured experimentally 33, 34 (800 ¡¾ 1600 A s71).An analogous explanation of the reasons for the diffusion of carbon atoms in the metal particle is given in Refs 101 ¡¾ 107. At present, the concept of the formation of carbon filaments by carbide cycle mechanisms is commonly accepted for the iron subgroup metals and their alloys.108 3. The effect of the orientation of faces of metal single crystals on the growth of carbon filaments Detailed crystallographic studies of the role of the orientation of catalytic particles in the preparation of filamentous carbon were carried out by Audier et al.98 ¡¾ 100, 109 ¡¾ 115 using mainly the disproportionation reaction of CO on Fe7Co and Fe7Ni alloys. It turned out that on all alloys with the body-centred cubic (BCC) structure, the carbon filaments are formed; the metal particles which initiate their growth are oriented in such a manner that the direction [100] of single crystals is parallel to the filament axis.The (100) face of the crystal is free from carbon. The alloys with the face-centred cubic (FCC) structure also catalysed the growth of filaments; however, they had a different orientation relative to the carbon filament. The direction [110] of crystals was always parallel to the filament axis. Electron microscopy studies of the metal particle ¡¾ filamen- tous carbon interface showed 98 ¡¾ 100, 109 ¡¾ 116 that the liberation of carbon and generation of graphite layers occur on the vicinal faces of the catalyst particle.Near the metal ¡¾ carbon interface, the graphite planes are arranged essentially parallel to the metal surface. The experiments were performed under the following conditions: the gas phase was compressed (75% CO and 25% CO2), the pressure was 105 Pa and the temperature range was 723 ¡¾ 923 K. However, an investigation of the structure and texture of the filamentous carbon obtained from benzene on iron and from buta-1,3-diene on nickel showed 77 that the particles of metallic iron are oriented differently relative to the filament axis. The iron particles which initiated the growth of filaments were single crystals in which it was the direction [110] that was parallel to the filament axis. Analogous results were obtained later by Rodriguez et al.117 In crystals of metallic nickel the direction [100] was parallel to the filament axis, while the planes (100) and (110) were free from carbon.Similar orientation and disposition of faces were established by Yang and Chen.94 Such an orientation of metal crystals and the presence of low-index faces in them are not occasional. Different faces possess different activities in the decomposition of hydrocarbons; the diffusion rate of carbon atoms through single crystals of metals can also be anisotropic.77 Thus, different faces and sites of the surface of metal crystals where the mechanism of the carbide cycle of decomposition of hydrocarbons is valid are not equivalent and they exhibit different properties in the general process of formation of carbon deposits.The distinction between the adsorption and catalytic properties of different faces and surface sites is confirmed by numerous data. For example, in the studies of methane decomposition on nickel in the temperature range 300 ¡¾ 773 K it was found 56, 118, 119 that this reaction proceeded only on the faces (100) and (110) [but not on the (111) face]. At temperatures up to 673 K, surface carbide structures are formed on the surface of these faces and diffusion of the carbon formed in the bulk of nickel crystal becomes noticeable above 573 K. No chemisorption of methane or formation of carbon on the (111) face could be detected. Decomposition of ethylene on the nickel faces (111), (110) and 5(111)6(110) is yet another example.120 On the stepped faces 5(111)6(110), the cleavage of bonds occurs in the lower temper- ature range than on the smooth faces (111) and (110).The formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys Presumably,77 ± 79 the generation of the graphite phase is easier on the face (111) than on the faces (100) and (110) because of high degree of its geometric correspondence to the structure of the basal graphite plane.This conclusion is in good agreement with the results reported by Eizenberg and Blakely 121 who studied the isolation of carbon from a supersaturated Ni7C solution on the faces (111), (311), (110), (210) and (110) and the vicinal high-index surfaces of nickel.In the course of the monolayer filling, the vicinal surfaces of nickel are spontaneously restructured (faceted) into stepped faces formed by the terraces (111) and the mono- atomic steps (110) or (310) depending on the index of the initial face. 4. Steps of the growth of carbon filaments from methane on nickel-containing catalysts Four individual sites (stages) which differ in the features of the carbon filament growth can be distinguished on the kinetic curves describing the decomposition of methane on the catalyst contain- ing 85 mass% of Ni/Al2O3 122 at 823 K. Analogous stages were observed in the formation of carbon from acetylene.27 Crystals of nickel of various shapes are formed after the reduction of an aluminium ± nickel catalyst. In conformity with the carbide cycle mechanism, the carbon atoms liberated as a result of methane decomposition in the induction period of filament growth diffuse into the bulk of the metal and form there a supersaturated solution of carbon in nickel relative to the graphite phase.As soon as the critical level of supersaturation is reached and a crystalline graphite nucleus appears at some site, condensation of the carbon atoms from the supersaturated solution occurs on the nucleus formed. In this case, the concen- tration of dissolved carbon in the zone adjoining the graphite phase decreases to its level in the saturated solution. The (111) faces of nickel are the most favourable for the formation of graphite nuclei, since the symmetry and parameters of the flat networks (111) of nickel (Fig.4 a) and (002) of graphite coincide. The graphite filaments formed upon carbonisation of Ni/MgO or Ni/Al2O3 catalysts have a coaxial conical structure where the graphite (002) layers are inserted one into another as funnels (see Fig. 2 b). The correspondence of the structures of the face surface of nickel particles and the graphite filament can serve as a virtually classical example. It was shown 63 that the frontal side of a nickel crystal is formed by the (100) and (110) faces, while the faces (111) form its rear part. After the appearance of the graphite crystallisation centres, the growth of graphite filaments begins, which is accompanied by the reconstruction of metal particles.During this period, the rate of carbon formation comes to stationary level. The shape of the metal crystals depends substantially on the rate of liberation of carbon atoms, their mass-transfer and specific features of the graphite structure.122 The more intense the process the more labile is the structure of the metal particle. Boellaard et al.105 noted that the formation of carbon fila- ments with the coaxial-conical structure should result in slippage of the carbon layers relative each other. Due to irregular supply of carbon atoms to the back, rear side of a cone-like metal particle, the carbon layers loose the form of ideal cones, and the body of filaments acquires an imperfect graphite structure with numerous c b a (100) Ni Ni ±Cu Ni (111) (111) a (111) (111) C C C Figure 4.Schematic representation of a mechanism of the generation and growth of carbon filaments on Ni7Cu and Ni catalysts (see the text for comments). 629 defects, in particular with a considerable number of edge disloca- tions. However, in our opinion, as the new graphite layers grow, the nickel particle should be lifted a little, while the (111) face of nickel and the basal graphite face should slide relative each other. Each new graphite plane slides to the central part of the carbon filament. Owing to high mobility of nickel atoms at the Ni7C interface, the surface metal atoms dislocate towards the filament axis, and this explains considerable transformation of the metal particle during the accelerated growth of the carbon layer observed on kinetic curves.In this case, the nickel particle (Fig. 4 b) changes its shape (Fig. 4 c). According to the crystallo- graphic laws, the angle a between the faces (111) should be approximately equal to 71 8. However, at 823 K the transfer of nickel atoms out to the `tail' of the crystal occurs rather strongly, which decreases the angle a to 50 ± 60 8. It may be suggested that the relief of the above-mentioned faces is determined by the atomic steps which change the slope of the plane. The nickel particles which initiate the growth of filamentous carbon from methane on Ni/Al2O3, are well formed (see Fig. 4 c). The face (100) is oriented towards the filament growth, while the faces (111) in the `tail' part of the nickel crystal form a pyramid with its top facing the side opposite to the filament growth direction.63 It is clear that after the induction period the appearance of new crystallisation centres at some sites of the metal surface becomes virtually impossible.As a result, the site where the graphite growth occurs becomes the back (rear) side of the particle, while all the other faces, which are not closed by graphite, begin to play the role of the front side on which methane is decomposed by the carbide cycle mechanism. At relatively low temperatures of 748 ± 798 K, such a form of nickel crystal remains invariable for a long time, which ensures stable growth of filamentous carbon. This stage is associated with the steady-state period of the filamentous carbon growth.The velocity of filamentous carbon growth from methane at 823 K was estimated on the basis of the rate of carbon formation. Rather simple calculations 122 showed that the mean filament growth velocity is 6 A s71, while the carbon concentration on the front side of the nickel particle is only 2 ± 3 times higher than that in the saturated solution in nickel at 823 K. It should be remembered that these results refer to the decomposition of the most stable hydrocarbon, viz., methane. Therefore, in this case the limiting step is the decomposition of CH4 molecules rather than the diffusive transfer of carbon atoms in the bulk of nickel particles. If the growth of filaments is limited by the diffusion of carbon in the bulk of the metal, then, as it was shown above, the gradient of carbon concentration becomes extremely large.This gradient predetermines the magnitude of diffusive flux of carbon atoms in the metal particle and has a strong effect on the phase state of the metal. Transition metals (Fe, Co, Ni) have small atomic radii. In particular, it is 1.24 A for nickel, and for this reason, the carbon atom with a radius of 0.77 A is too large to occupy the octahedral voids of the most dense FCC-packing of nickel. Thus, diffusion of carbon through the nickel particle results in distortion of the crystal lattice of the latter. In some cases, the catalytic particle can become highly labile owing to a high carbon concentration, intense motion of carbon atoms through the metal bulk and the release of heat upon their condensation in the form of graphite phase.Centres of carbon crystallisation on the metal particle are formed if a high energy barrier, which is 170 kJ mol71 for nickel 123 and 220 ± 300 kJ mol71 for iron, is overcome.17 This is possible owing to a considerable supersaturation of carbon solutions in the metal which is especially large in the early stages of hydrocarbon decomposition. Studies using a high-temperature X-ray chamber showed 124 that in the course of induction period the concentration of carbon atoms on nickel exceeds 2 at.%. Based on a thermodynamic analysis, Parmon 125 concluded that owing to a strong supersaturation of iron or nickel with630 carbon, which can reach >10 mass %, melting of the metal should begin at *920 K.In our opinion, this relates primarily to the induction period of the filamentous carbon growth because after the formation of graphite nuclei the maintenance of such a high carbon concentration in the metal is hardly probable. The formation of carbon filaments is often associated with the high mobility of metal ± carbon particles. In this connection, one should mention the VLS (vapour ± liquid ± solid) mechanism proposed for the description of the growth of silicon filaments on drops of a liquid metal.126 The problem of the possibility of melting of metal particles has not been totally solved yet, and the temperature boundary where the metal particles catalysing the growth of filamentous carbon or nanotubes can pass to the liquid state, has not been established yet.Some authors state 127, 128 that the finely dispersed catalytic particles of iron are capable of melting even at temperatures below 1173 K. They believe that the particles of liquid iron are `respon- sible' for the efficient growth of carbon filaments. However, according to other data (see, e.g., Refs 129 and 130), the `lique- faction' of iron particles occurs at*1400 K. Relevant calculations confirm that the growth of carbon filaments from methane is initiated by edged particles oriented in a definite manner relative to the filament axis; the growth of carbon filaments from alkenes, dienes and acetylene is initiated by the pear-shaped particles, the form of which is less expressed. In the so-called `extrusion' method, which is less common than the above-considered (classical) method for the formation of carbon filaments, the crystal initiating the carbon deposition remains in contact with the support. This process occurs in the growth of carbon filaments on Pt/Fe alloys 36 and on pure metals.35, 131, 132 5.Specific features of the formation of carbon filaments on metallic iron In a number of cases, deposits of carbon on iron catalysts have the form of filaments.77 A general view of such filaments is shown in Fig. 2 a. Figure 5 shows a micrograph of one such filament. As a rule, the carbon filament contains inclusions both at end of the filament and in its hollow channel, which, according to micro- 100A Figure 5. A micrograph obtained by the method of high-resolution electron microscopy (HREM) of a carbon filament formed on metallic iron from an atmosphere of benzene diluted with argon in the molar ratio C6H6 :Ar=1 : 15 at 923 K.77 V V Chesnokov, R A Buyanov diffraction data, represent the metal phase.Reflexes on electro- nograms indicate that this phase has a body-centred cubic lattice of a-Fe. The characteristic feature is that the crystallites which are constituents of the filament structure are oriented in a definite manner relative to its axis. The crystallographic direction of a-Fe [110] is parallel to the axis. The metallic inclusion at the end of the filament has usually the drop-like form, which is narrowed in its lower part and is weakly edged. The inclusions have a block structure.This follows from the existence of boundaries between the blocks as seen on electron micrographs and from the periodic contrasting of the areas of overlap of individual blocks. The carbon of the filament is graphitised, and this is clearly seen in micrographs obtained by the high-resolution electron microscopy method. The distance between the planes is 3.50.2 A, which is close to the interplanar distance d 002 for graphite. Zaikovskii et al.77 showed that the graphite layers are concentrically arranged around the filament axis and the inter- channel metal particle. The distribution of carbon around the metal particle at the end of the filament is irregular.The smallest thickness of the carbon layer (only a few monolayers) is observed on the side of the metal particle where the decomposition of a hydrocarbon occurs. As the distance from the top of the metal particle increases, the thickness of the carbon layer increases. The orientation of graphite layers at different sites of the metal surface is also different. In contrast to the upper part of the metal particle, in its lower part the graphite layers form a certain angle with the surface and are in contact with the metal by their terminal sides (Fig. 2 a). In the hollow channel of the filament there are very thin (2 ± 10 monolayers) partitions composed of carbon layers. Note in conclusion that the mechanism of generation and growth of carbon filaments on a-Fe is extremely complex.When considering these processes, one should take into account the following characteristic features of carbonisation of metallic iron: (1) high diffusion coefficient of carbon in the bulk of the metal ensuring its rapid supply to different surface sites; (2) incomplete epitaxial correspondence between the parame- ters of a-Fe face (111) and the basal graphite plane (002), which impedes the formation of nuclei of the graphite phase. (3) the tendency of metallic iron to form carbides. 6. Regularities of the formation of carbon deposits in the `octopus' form On nickel ± copper alloys, the formation of unusual morpholog- ical `octopus' carbon was observed:63, 106, 133 ± 135 one particle of nickel ± copper alloy initiates the growth of several filaments in different directions (Fig.4 a). In contrast to nickel, metallic copper has low activity in the processes of carbon formation. It was used as an additive to Ni/MgO and Ni/Al2O3, because it has the same face-centred lattice like nickel and forms alloys with nickel in a wide range of concentrations. After reduction by hydrogen of the catalyst containing 10 mass% of MgO, 12 mass% of Cu and 78 mass% of Ni, the X-ray patterns reveal the predominance of the lines of MgO and a nickel ± copper alloy of cubic structure 63 with the unit cell parameter a=3.53 A; this parameter corresponds to an alloy of approximate composition Ni0.9Cu0.1. The mean size of the alloy particles found from the half-width of the line (311) is 210 A.Kinetic studies of the formation of carbon from methane on the catalysts Ni/MgO and Ni7Cu/MgO showed that the rate of this process decreases as the content of copper in the sample increases.63 Addition of 12 mass% of copper decreases this rate 2 ± 3-fold. As a result, the aggregation of carbon atoms proceeds in a more ordered manner, and one observes a rather clear-cut trend towards the formation of graphite on the (111) face of the nickel ± copper alloy (see Fig. 4 a). The lattice parameter of the alloy increases with the increase in the copper content, and the crystallo- graphic correspondence of the (111) faces of the alloy and the faces (002) of graphite becomes more and more complete.The formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys Electron microscopy analysis showed 63, 79, 133 that carbon- isation of the Ni7Cu/MgO catalyst upon decomposition of methane for 1 min at 873 K leads to the formation of carbon filaments, the ends of which contain metal particles.This metal particle usually has a well edged form, while its diameter exceeds the diameter of the carbon filament. The data from electronic microdiffraction (EMD) suggest that the (111) face of the FCC lattice of the metal is the most developed. It is perpendicular to the direction of the filament growth and is in contact with the (002) planes of graphite. The lateral (100) faces of the alloy are free, they have no contact with carbon and are the centres of methane decomposition.The form presented in Fig. 4 a corresponds to the initial stage of the growth of filamentous carbon on particles of the Ni7Cu alloy. Carbon can be formed simultaneously on two opposite faces (111) of the flattened alloy particle; this can give birth to two carbon filaments in the form of rings containing a metal particle inside. Sometimes, carbon is deposited simultaneously on three (111) faces of a nickel ± copper alloy and the growth of filaments begins in three directions (Fig. 6). According to the data from EMD, the structure of such filaments is formed by the graphite layer (002), arranged perpendicularly to the filament growth direction.63, 79, 133 500A Figure 6.A micrograph showing the filamentous carbon growth in three directions on the Ni7Cu alloy particles.63 The sizes and shapes of faces which determine the equilibrium habits are known to be related to the surface energies of these faces. The difference in the habits of catalytically active particles in nickel and nickel ± copper catalysts is explained by different degrees of proximity of the parameters of hexagonal lattices of the nickel and nickel ± copper alloys to the parameter of the graphite lattice. The link of carbon with a particle of nick- el ± copper alloy over the face (111) is stronger; consequently, the surface energy in this case is lower than on the nickel particle, and the size of (111) face is increased.After carbonisation with methane for 2 min at 873 K the particles loose their clearly expressed edging.63 The increase in the carbonisation time to 5 min induces changes in the structure of metal particles. It becomes of micro- block-type, the size of the blocks being 50 ± 200 A. The defective- ness of the alloy surface deteriorates the ordering of the graphite layers formed and in some places the filament is separated into layers (Fig. 7). This favours the formation of bundles of filaments growing from one particle. Very often, thin branches appear from the major bundle or from the block-polycrystalline metal particle. At the ends of such branchings, there are also metal particles.63, 79 Carbon deposits, formed on the Ni7Cu/MgO catalyst upon decomposition of CH4 for 3 h at 875 K present a porous mass consisting of filaments intertwoven in a complex manner. In 631 600A Figure 7.Morphology of the carbon formed from methane on the block particle of Ni7Cu alloy after a 5-min relaxation at 873 K. addition to large filamentous formations, numerous thin fila- ments 50 ± 100 A in diameter are present. Taking this into account, it may be concluded that large particles of the alloy break into individual blocks which catalyse the growth of thin filaments. Thus, one of the stages in the formation of carbon deposits is the diffusion of carbon atoms in the metal particle determined by different concentrations of these atoms near those faces where a hydrocarbon is decomposed and graphite is condensed.The directed diffusion of carbon atoms at elevated temperatures leads to the displacement of the matrix atoms (Cu and Ni). This effect is the stronger the higher the ability of components to form chemical bonds with carbon. Stratification and fragmentation of a single crystal of the initial alloy in the process of the carbon filament growth may be explained by bulk diffusion of atoms. Consequently, as the carbon accumulates, the morphological structure of the carbon mass begins to resemble a branched tree.63 7. The reasons for deactivation of a catalytic particle in the growth of a carbon filament The phenomenon of deactivation of catalysts in their carbon- isation by hydrocarbons has been investigated.5, 136 Deactivation is caused by destruction and grinding of catalysts, removal of some components from them and by the mechanical capture of individual microparticles by the growing carbon filaments.This phenomenon was called carbon erosion. Electron microscopy studies of carbon filaments which ceased to grow showed that the front side of nickel particles is covered by a carbon film. Treatment of such filaments with hydrogen or oxygen leads to the recovery of their activity and continuation of growth.27 In our view, the cessation of the filamentous carbon growth is explained by the peculiarities of the process itself. As noted above, the diffusion of carbon atoms in the nickel particle is determined by the existence of a concentration gradient.The concentration of carbon on the faces (100) and (110) on the frontal side of the crystal is larger than its equilibrium concentration at the nick- el ± graphite interface. Supersaturation of carbon solution in the metal is the driving force of structural transformations on the surface of nickel. The larger the supersaturation the faster the deactivation. Wesner et al.137 considered surface autodiffusion of carbon atoms on the (100) face of a nickel single crystal which is induced by catalytic hydrogenation of CO with the formation of methane. On this face, sinusoidal grooves 1 mm deep were made at a distance of 7 mm. After the catalytic hydrogenation of CO for 6 h, the shape of the section profile changed substantially.Facet- ing of the groove was spreading from the ridge down to the632 `valleys'. Taking this into account, it may be suggested that the nickel atoms dislocate from the `valleys' in the direction of ridge peaks with simultaneous creation of cutting. The height of ridges increases with time. There are grounds to believe that in the process of carbon deposition reconstruction occurs of the surface of the front side of the metal particle, on which the steps corresponding to the (111) faces appear. Thus, conditions are created for the formation of new graphite nuclei on the front side of the nickel crystal. Therefore, the front side of the nickel crystal becomes coated with a thin carbon film with time despite the efflux of carbon atoms inside the crystal.This is the primary mechanism of deactivation of metal particles which catalyse the growth of carbon filaments. One manifestation of the carbon erosion is the destruction of metal particles initiating the growth of carbon filaments with the formation of smaller and smaller fragments which can eventually be totally `overgrown' with carbon and become isolated. A spectacular example may be the above-described growth of filamentous carbon on the Ni7Cu alloy particles in the Ni ± Cu/MgO catalyst. The initially formed filaments 500 ± 1000 A in diameter are branched with time. This is due to the fact that the diffusion flux of carbon atoms in the bulk of alloy particles favours their fragmentation and crushing into subcrys- tals, a part of nickel being removed with carbon.As the initial particle is fragmented, a deposition of carbon begins not only on major filaments, but also on the surfaces of a contact between individual subcrystals. As a consequence, the initial particles undergo the ever-increasing splitting, while each new smaller particle becomes the centre of the growth of new filaments of a smaller diameter. Gradual extinction of the process of carbon deposition is explained by the fact that the metal particles of some minimal size prove to be covered by a graphite envelope and are switched off the process. 8. The effect of carbonisation temperature on the regularities of growth of carbon filaments in the decomposition of methane on nickel and nickel ± copper catalysts The increase in the reaction temperature influences the peculiar- ities of the filamentous carbon growth on both nickel and nickel ± copper catalysts.Figure 8 shows the kinetic curves of carbon formation from methane at different temperatures on an Ni7Cu/MgO catalyst (analogous kinetic curves were also obtained for an Ni/MgO catalyst).122 The rate of carbon forma- tion increases with temperature. However, at 973 K the higher rate is maintained only for a few minutes, after which it drops abruptly. Figure 9 shows an electron micrograph of a filamentous carbon obtained from methane at 1023 K on an Ni/Al2O3 catalyst. It is seen that the nickel particles appear to stretch along the filament length. As a result, the channel inside the carbon filament is filled with metallic nickel.Analogous morpho- logy is also characteristic of the filamentous carbon obtained from G(mass %) 2 1 2000 1500 1000 500 3 0 20 t /min 40 Figure 8. Kinetic curves of the carbon formation from methane on an Ni7Cu/MgO catalyst at 823 (1), 873 (2) and 973 K, respectively. G is the carbon increment. V V Chesnokov, R A Buyanov 500A Figure 9. Amicrograph of the filamentous carbon formed from methane on an Ni/Al2O3 catalyst at 1023 K. methane on an Ni7Cu/MgO catalyst at 973 K.A question arises: what is the reason for the internal cavity formation in the graphite filament and what forces `drag' the metal particle along this cavity? The second mechanism of deactivation of metal particles which catalyse the growth of carbon filaments consists in the following.The increase in temperature and the higher rate of the filamentous carbon growth lead to an increase in the viscous fluidity of the metal particle, which facilitates sliding of the graphite planes formed over the surface of the metal 122 and intensifies the efflux of nickel atoms in the direction of the diffusion flow of carbon atoms, i.e., into the `tail' part of the metal particle. As a result, the particle is elongated, while the angle a (Fig. 10 a) decreases considerably; the catalytic particle is trans- formed (Fig. 10 b). In the limit, the particle is so elongated that it acquires the shape of a needle (Fig. 11). Since in this case, the (111) faces become virtually parallel, the graphite deposited on them does not push the nickel crystal ahead but rather brings about lateral compression, which makes it still more elongated.Because of a large length of the nickel needle, the accumulation of carbon and supersaturation of the tail part in it proceed slowly. After reaching the critical supersaturation of the tail part of the nickel crystal, rapid condensation of carbon atoms begins with the formation of graphite layers which create carbon partitions inside the filaments. This in turn leads to irregular forward movement of the metal needle by pulsation. This results in the periodic appearance of carbon partitions which are seen in Fig. 11. However, the situation changes considerably if carbonisation is carried out at 973 K in an atmosphere of methane diluted with hydrogen in the molar ratio 1 : 1.The morphology of the fila- mentous carbon formed on both Ni/Al2O3 and Ni7Cu/MgO c b a Ni Ni Ni a C C C Figure 10. Transformation of catalytic particles upon the reaction tem- perature increase (see the text for comments).The formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys 500A Figure 11. Amicrograph of the filamentous carbon formed from methane on the Ni7Cu/MgO catalyst at 973 K. becomes different. A hollow channel appears inside the carbon filament. (Fig. 10 c). We have already mentioned that as the intensity of the process increases, the probability of restructuring of the metal becomes higher until some faces disappear and the other faces develop.In this case, upon addition of hydrogen to methane, the methanation of carbon atoms on the front side becomes possible. In other words, the reaction takes place which is opposite to methane decomposition, the carbon concentration on the front side decreases and the entire process is slowed down. It is logical to suggest that in this labile system reconstruction of the structure shown in Fig. 10 a into the structure shown in Fig. 10 b occurs. On the rear, a less active face (100) is formed in the side outgrowth; this is surrounded by more active sites (111).111, 113 The carbon is predominantly deposited at the sites marked with the arrows in Fig.10 c, and this determines the speed of the motion of the entire particle. Thus, the reason for the appearance, along the filament axis, of a hollow channel the diameter of which is commensurable with the size of the `passive' face (100) on the rear side of the particle is the stretching of the catalytic particle. Baker and Harris 27 explained the formation of a hollow channel by the slower supply of carbon atoms because of the increase in its diffusion route. However, this is rather the consequence of the catalytic particle stretching. The graphite layers which arise slowly on this face are sealed off periodically, making partitions in the hollow of the channel. At a more rapid influx of carbon atoms, the hollow channel should have filled with carbon.9. The effect of hydrogen on the formation of carbon filaments The knowledge of general regularities and specific features of the carbide cycle mechanism allows prediction of the influence of hydrogen on the deposition of carbon. Hydrogen can hydrogenate metal (M) carbides with the formation of lower hydrocarbons, mainly, methane: M + C M + CnHm H2 [M7C] M + CH4 Owing to this property, hydrogen also affects the formation of carbon; in the case of nickel, this effect is stronger than in the case of cobalt or iron. Upon dilution of buta-1,3-diene with argon alone, the carbon- isation of nickel proceeds rapidly at the beginning of the reaction, but then the rate of carbon deposition drops abruptly (Fig.12). Addition of hydrogen decreases the initial rate of carbon for- mation. The larger the partial pressure of hydrogen in the reaction mixture the stronger is this decrease. At a partial hydrogen pressure pH2 of 17.5 kPa, the rate becomes constant and does not depend on the carbon content on the catalyst. At higher pressures, an induction period appears the duration of which increases with 633 G(mass %) 2 1 8 3 64 4 2 5 0 2 4 6 8 t /min Figure 12. Effect of hydrogen on carbonisation of nickel by buta-1,3- diene at 698 K. The partial pressure of hydrogen/kPa: (1) 26.2; (2) 17.5; (3) 8.8; (4) 3.5; (5) 0. pH2 . If a hydrocarbon is diluted with hydrogen 300 times, the induction period will become so long that carbon will not be formed over 2 h.138 An electron microscopy study showed 123 that carbon deposits formed upon dilution of a hydrocarbon with argon or hydrogen to pH2=17.5 kPa, present mostly a film which screens the nickel surface.At a higher hydrogen pressure, carbon is deposited in the form of filaments which do not block the surface. When carbonisation occurs in the medium of a pure or argon- diluted hydrocarbon, favourable conditions are provided for the appearance of the graphite nuclei, since in this case the carbon atoms do not have enough time to escape from the place of hydrocarbon decomposition. During subsequent growth of the nuclei, these atoms form a regular layer blocking the nickel surface. Upon partial or complete replacement of an inert diluent by hydrogen, hydrogenation of an intermediate carbide-like com- pound begins to occur in addition to the hydrocarbon decom- position.At the same time, hydrogenation of carbon with atomic hydrogen formed as a result of the dissociation ofH2 molecules on the nickel surface becomes possible. As the partial pressure of hydrogen is increased, the hydrogenation rate increases. This decelerates the formation of the carbon phase nuclei on the surface accessible to hydrogen and leads to the appearance of an induc- tion period. The activation energy of the graphite nuclei formation is 170 kJ mol71 during the deposition of the first 1%± 2% of carbon on nickel. In a hydrocarbon ± hydrogen medium, such nuclei are formed on the surface to which the access of reagents is hampered since hydrogen hydrogenates both the carbon atoms formed and the graphite nuclei.Owing to diffusion, a portion of carbon atoms is transferred to the interblock boundaries of nickel microcrystal- lites where new graphite nuclei are formed on defects. Deposition of carbon on the interblock boundaries of nickel crystallites favours their dispersion and increases the number of metal particles which are actively involved in the interaction with the reaction medium. After the induction period and with the accumulation of graphite nuclei, the activation energy of carbon formation decreases and becomes constant (140 kJ mol71) after deposition of 10% of carbon. The transfer of carbon to the graphite phase nuclei (diffusion of carbon atoms in the metal bulk) becomes the limiting step of carbonisation.The value 140 kJ mol71 is close to the reference value of the activation energy of the diffusion of carbon atoms in nickel. Fragmentation of polycrystalline nickel into particles coated with carbon on the side inaccessible to the gas medium creates prerequisites for the formation of filamentous carbon. In this case, the carbon constituting the growing filament is not hydrogenated because hydrogen is virtually incapable of penetrating to the surface of contact of the metal particle with the carbon filament634 phase (on the rear side), though a constant influx of carbon atoms takes place. The frontal (active) part of the metal particle, which catalyses the carbon filament growth, is not screened by carbon for the above-mentioned reasons.Thus, the addition of hydrogen prevents the deposition of carbon in the form of a film and creates favourable conditions for the formation of carbon filaments. Nolan et al.139 gave an interesting explanation of the reasons for the formation of carbon deposits of different morphological types. They believe that carbon is accumulated as open forms, e.g., carbon filaments, in the presence of hydrogen. The edge faces of graphite with unsaturated bonds come out, onto the external surface of the filaments, and hydrogen saturates these bonds. In the absence of hydrogen, carbon should be deposited predom- inantly in such forms as nanotubes or carbon films which do not have any unsaturated bonds on the edge faces.Thus the major products of decomposition of pure CO on highly dispersed nickel particles are carbon films and nanotubes, whereas the hydrogen- diluted CO forms carbon filaments with the coaxial-conical disposition of the graphite planes in the filament body. It is stated that the slope of the graphite cones depends on the hydrogen concentration in the reaction medium. However, in our opinion, Nolan et al.139 give too much attention to the compensation of the energy of the free valences of the edge faces due to the formation of C7H bonds. It is unclear why CO2 cannot react with carbon atoms and compensate thereby the energy of free valences. Furthermore, it is not taken into account that the neighbouring graphite layers coming onto the external surface of filaments can interact with each other thus closing the free valences.140 Coming back to the role of hydrogen, let us note that it influences primarily the rate of the filamentous carbon growth, and this, as was shown above, changes the morphology of filaments.10. The formation of carbon filaments at temperatures above 1273 K The fibres used for the preparation of carbon composite materials are known to be obtained from viscose, polyacrylonitrile, as well as from petroleum or coal pitch. However, these materials have a substantial disadvantage � they are costly. Therefore, the pros- pect of their replacement by filaments prepared by catalytic methods seems rather attractive.A method for the preparation of carbon fibres from the gas phase [vapour-grown carbon fibres (VGCF)] is being actively developed in the USA, Japan and China.141 ± 145 To this end, use is made of iron-containing catalysts deposited on ceramic supports or of injection of a benzene solution of ferrocene into a reactor. The sources of carbon are benzene or methane diluted with hydrogen as well as the gaseous waste from the production of coke or from the converter produc- tion of steel (CO7CO27H2). A useful effect is given by the addition of small amounts of sulfur-containing gases (e.g., H2S) to the gas flow of reagents. The reaction is carried out at temper- atures 1273 ± 1423 K. The process starts with the catalytic growth of primary fibres, then they become thicker due to the deposition of pyrocarbon. The fibres prepared by theVGCFmethod have the following characteristics: length, a few millimetres; diameter, 2 ± 3 mm; density,*2 g cm73.Structural studies showed that the use of iron particles at temperatures above 1173 K can yield long filaments uniform in size, which consist of densely packed cylindrical basal graphite planes and always have a hollow core. Baird et al.81 have proposed a scheme of the growth of a tubular filamentous carbon which presumes diffusion of small carbon clusters over the surface of a metal particle. This model was further developed in a study of Oberlin et al.,146 which is one of the most detailed investigations of the structure of filamentous carbon formed from a benzene ± hydrogen mixture at 1373 K.The reaction products were tubular filaments of cylindrical shape with a hollow channel inside them. The diameter of the channel ranged from 20 to 500 A and was not always constant along the filament length. In their cross-sections, the filaments had a `dendritic' V V Chesnokov, R A Buyanov structure: apparently, the graphite layers formed concentric rings. In the core of the filament, the graphite structure was more ordered than in the outer layers. Based on the results obtained, the authors have postulated a two-step mechanism of the filamentous carbon formation. In the first step, the filament core is formed which consists of long parallel carbon layers twisted around the hollow central channel.In the second stage, the filamentous carbon becomes thicker owing to pyrolytic deposi- tions. The ends of filaments contain the catalyst's particles. The interplanar distances measured by the method of electron micro- diffraction make it possible to presume that the catalytic particles are either iron carbide (Fe3C) or a-Fe. However, no final conclusions have been made. Tibbetts 147 attempted to establish the reason for the tubular structure of carbon filaments. As the basal face possesses lower specific surface energy than the edge faces, the free energy of the growing filament should be a minimum when the filament consists of curved basal planes forming cylinders. However, as the diameters of cylinders decrease, the degree of curvature of the planes increases and, naturally, the energy required for such curving also increases.Thermodynamic considerations indicate that the formation of a hollow core is more advantageous energetically than the deposition of carbon in the form of strongly deformed cylindrical planes of small diameter. The model pro- posed made it possible to predict accurately the minimum inner diameter of a carbon filament and to substantiate theoretically the tubular structure of the filamentous carbon. Deactivation of catalysts at high mperatures is associated with the blockage of the surface of catalytic particles; however, there are different suggestions about the mechanism of this phenomenon. Some investigators believe 93, 147, 148 that deactiva- tion is due to the adsorption, on the frontal side of the metal particle, of highly condensed pyrolysis products, which are formed in the gas phase and are precursors of soot particles.The equation obtained on the basis of this hypothesis provides a satisfactory description of the histogram of the distribution of filaments with respect to their length which are formed upon methane decomposition on g-Fe at 1320 K. According to Gadelle,128 the catalyst is active in the liquid state, and even small changes in the surface composition induce its transition to a less active non-melting form which is easily blocked by carbon. 11. Growth of carbon filaments under the action of catalytic particles present inside the filament It was shown that disproportionation of CO on the Fe/Ni and Fe/Co alloys yields the so-called bifilaments.100 It was estab- lished 111, 112 that the growth of bifilaments is catalysed by biconical microcrystals.In their structures, each of the bicones is analogous to a `simple cone' which initiates the growth of classical filaments. It was established that the directions [100] of the alloy crystal with the body-centred cubic lattice and [110] of the alloy with the face-centred cubic lattice are parallel to the filament axis. In the alloys with the face-centred cubic lattice, the twinning of biconical crystals was often observed. No simple mechanism has yet been proposed for the explanation of the growth of these unusual filaments.12. Regularities and the mechanism of the growth of helical carbon forms Helical filaments or bifilaments first described by Baker et al.37 were formed on an Fe/Sn alloy from acetylene at temperatures above 1073 K. One catalytic particle of the alloy initiates simulta- neous growth of two filaments in the opposite directions. The two corkscrew-like filaments grow with approximately the same velocities, and in the course of the process, identical fragments are reproduced symmetrically in either filament. Unusual helical forms of filamentous carbon were also observed in the carbonisation of an Ni7Cu/MgO catalyst which was carried out in the medium of buta-1,3-diene diluted with argon and hydrogen in the molar ratio C4H6:Ar :H2=2:40 : 75The formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys 1000A Figure 13.A micrograph of the mirror-symmetrical carbon filaments growing from one particle of an Ni7Cu alloy. Due to the increase in the carbon content on the Ni7Cu/MgO catalyst, its mass increased 28-fold (the same refers to Figs 14, 15). at 723 K.149 According to the X-ray diffraction data, after the reduction with hydrogen, the Ni7Cu alloy of cubic structure is predominant in the catalyst and the MgO phase is also present. An electron micrograph of the mirror-symmetrical carbon filaments (Fig. 13) shows that the filaments are twisted into helices and originate from one metal particle. The thickness of these filaments ranges from 300 to 1000 A, while their lengths reach 1 mm.Obtained by the HREM method, micrographs of metal particles on which carbon grows indicate (Fig. 14) that they possess a nearly symmetrical oval shape with the symmetry planes in the middle part and represent multiple twins. On the frontal side of the surface, in the middle of the oval metal particle there is a 100A Figure 14. A micrograph obtained by the HREM method of an oval particle of an Ni7Cu alloy. The arrows indicate the positions of the twinning boundaries. 635 25A Figure 15. A micrograph obtained by the HREM method of the Ni7Cu/ MgO catalyst. microparticle 10 ± 15 A thick and 30 ± 50 A wide with the struc- ture of the hexagonal nickel. It is on these microparticles that the `chemical' step of the carbide cycle mechanism is realised.In the crystalline particles of the Ni7Cu alloy, there is a series of twinning planes (111) which are parallel to the mirror symmetry plane. The twinning planes separate the blocks of twins with the face-centred cubic structure. The interplanar distances in each block of twins are: d111=2.03 A and d002=1.7 A. In a micro- graph obtained by the HREM method, one can easily notice that the twinning planes are grouped under a microphase particle on the frontal side of the crystal (Fig. 15), while they are rarely found in other places. In the catalysts on which the deposition of carbon reaches 2000% ± 5000%, the widths of twin blocks in a metal usually exceed 20 A.At the beginning of the reaction the frequency of the appear- ance of twins in the metal is much higher than that at its end. In a micrograph of the Ni7Cu/MgO catalyst subjected to carbon- isation for 10 min (Fig. 16), one can see the carbonisation boundaries with a period of *4 A. This distance is sufficient for the placement of only three densely packed layers of atoms of the face-centred cubic lattice of a metal with the interplanar distance d111=2.03 A. The hexagonal nickel carbide Ni3C (see Ref. 150) and metallic nickel 151 have virtually the same crystal lattices (hexagonal close packing). Carbon atoms in carbide occupy 1/3 of the octahedral cavities in the packing of metal atoms. According to the mechanism of formation of symmetrical carbon filaments proposed by Zaikovskii et al.149, in the induction 25A Figure 16.A micrograph of the Ni7Cu/MgO catalyst containing 75 mass.% of carbon. In the middle part of the Ni7Cu alloy particle one can see the boundaries of twinning after each 4 A. The sequence of densely packed layers corresponds to the alternation of the hexagonal and cubic structures of the alloy.636 period of the filamentous carbon growth, decomposition of buta- 1,3-diene yields a particle of Ni3C carbide microphase metastable at 723 K, while the Ni7Cu alloy itself is supersaturated with carbon. When critical supersaturation is attained and at some site of the alloy surface a crystalline graphite nucleus is formed, then a rapid `discharge' of carbon atoms onto it occurs.The graphite phase nuclei are preferentially formed on the (111) face of the Ni7Cu alloy since the symmetry and interatomic distances in flat networks (002) of graphite and (111) of the alloy oriented towards the graphite filaments coincide. Owing to a strong supersaturation of the frontal zone of the metal particle with carbon, an intense diffusion flux of carbon to the places of its crystallisation in the form of graphite phase arises. Rapid diffusion of carbon atoms leads to an enhanced generation of defects in the metal. The metal, while remaining solid, passes to the viscous-fluid state. As a result, the particle of the Ni7Cu alloy acquires spherical shape which is the most advantageous from the energetic viewpoint. However, the diffusive fluxes of carbon atoms also induce dislocation of metal atoms, and the spherical particle is somewhat stretched (Fig.17). There arises a regime of self-organisation of the symmetrical form of metal particles. As a result, the metal particles acquire an oval form with the symmetry plane in their centre part. In this case, each of the symmetrical halves is oriented in such a manner that the [111] directions of the alloy coincide with the directions of the growth of carbon filaments. C4H6 Ni3C Zone of alternation of layers with hexagonal and cubic packing An Ni7Cu alloy with the cubic lattice C C Figure 17. The mechanism of the growth of a helical carbon filament obtained from buta-1,3-diene on the Ni7Cu crystals. A particle of the hexagonal Ni3C carbide microphase on the frontal side is a nucleus for the growth of a hexagonal structure piercing the catalytic particle of the Ni7Cu alloy.Under the reaction conditions, the catalytic particle consists of alternating layers of cubic and hexagonal Ni7Cu alloys. The passage of carbon atoms through the interlayer of the hexagonal metal may be represented as continuously alternating steps of formation and decomposition of the carbide. However, the saturation of hexa- gonal interlayers with carbon should decrease as they go farther and farther from the microphase of the hexagonal carbide. Consequently, in the course of reaction the layers of the hexagonal metal include most probably a non-stoichiometric carbide of variable composition.During the reaction, the twinning planes are dislocated from the central zone of the Ni7Cu alloy to the crystal periphery. The driving forces of the dislocation of the interblock boundaries and, respectively, of restructuring is the diffusion of carbon atoms and the lesser stability of the alloy with the hexagonal structure than that of the alloy with the cubic structure. Owing to an intense directed flux of carbon atoms from the central part of the crystal to its periphery (see Fig. 17), defects arise in the layers adjoining the twinning planes which induce restructuring of the layers. In other words, the plane of twinning begins to dislocate in the direction of the movement of carbon atoms to the site of carbon crystallisation into the graphite phase.Thus, the hexagonal planar inclusions representing a non-stoi- chiometric carbide of variable composition `disperse' from the middle of the oval metal ± carbon particles. The removal of these V V Chesnokov, R A Buyanov interlayers from the central zone is probably compensated by their repeated regeneration by a microparticle on the surface of a large oval particle in such a way that the process is continuous. Graphite filaments grow symmetrically and virtually over the entire surface of the Ni7Cu alloy particle. The route of the diffusion of carbon atoms from the place of their entry into the alloy particle through the nickel carbide microphase to different sites of the graphite phase growth is not the same (see Fig.17). Because of this, the velocities of the graphite phase growth become also different and twisting of filaments occurs to the side opposite to the place where the carbide microparticle is situated. IV. Conclusion The above-discussed observations and conclusions are helpful in understanding the factors that influence the process of carbon filament formation and determine their properties. These proper- ties include primarily the spatial order of the construction of the basal graphite planes in filaments, the diameters and lengths of filaments, the presence or absence of the inner hollow along the axis of filaments, their morphologies, the existence and degree of dispersion of other mineral components in the graphite structure of filaments, etc.Adsorption, catalytic, chemical, magnetic, mechanical and many other properties of filaments depend on these characteristics. Note that the methods for controlling the synthesis of carbon materials are still in their infancy. For this reason, we will touch upon this issue only in brief and in most general terms. The most substantial effect on the properties of filamentous carbon materials is the spatial distribution of the basal planes of graphite. The distribution of graphite layers in a filament deter- mines such properties of the carbon material as its mechanical strength, adsorption capacity, chemisorption and catalytic char- acteristics, intercalation capacity and others. As mentioned above, depending on the properties of the catalytic metal particle (head) and conditions of the implementa- tion of the process, it is possible to synthesise graphite filaments of three major `basal' structural types: with the coaxial-cylindrical, coaxial-conical and stack structures (see Fig.2). The properties of metal particles which are of interest to us can be regulated by varying the method of their preparation and by adding other metals. The diameter of a filament is determined by the size of the metal particle and differs little from it. The length of a filament depends on the time of its growth, which in turn is determined by the time during which the metal particle can efficiently play the role of the filament growth centre without undergoing deactiva- tion or breaking into subcrystals. The regime of the process has a substantial effect on properties of carbon materials not only directly, but also indirectly, through deactivation of metal catalyst particles.A decrease in the temper- ature of the process, dilution of a hydrocarbon with hydrogen, use of a less coke-producing hydrocarbon will make it possible to increase the yield of carbon formations per unit mass of a metal and to improve the graphite structure. However, in this case, naturally, the rate of carbon formation decreases. This review was prepared with a financial support of the Russian Foundation for Basic Research (Projects Nos 99-03- 32420, 00-03-32431) and the Grant for Leading Scientific Schools of Russian Federation (No.00-15-97440). References 1. R A Buyanov Kinet. Katal. 28 157 (1987) a 2. H Storch, N Golumbic, R Anderson The Fischer ± Tropsch and Related Syntheses (New York: Wilay, 1951) 3. Yu B Kagan, A N Bashkirov, S V Kamzolkina, A Ya Rozovskii Zh. Fiz. Khim. 23 2706 (1959) bThe formation of carbon filaments upon decomposition of hydrocarbons catalysed by iron subgroup metals and their alloys 4. R A Buyanov, V V Chesnokov, A D Afanas'ev, V S Babenko Kinet. Katal. 18 1021 (1977) a 5. R A Buyanov Zakoksovanie Katalizatorov (Coking of Catalysts) (Novosibirsk: Nauka, 1983) 6. V V Chesnokov, Candidate Thesis in Chemical Sciences, Institute of Catalysis, Siberian Branch of the Academy of Sciences of the USSR, 7. A Ya Rozovskii Katalizator i Reaktsionnaya Sreda (Catalyst and Novosibirsk, 1982 Reaction Medium) (Moscow: Nauka, 1988) 8.V V Chesnokov, R A Buyanov, A D Afanas'ev Kinet. Katal. 20 477 (1979) a 9. R A Buyanov Kinet. Katal. 21 237 (1980) a 10. R A Buyanov,M A Tanatarov, in Nauchnye Osnovy Tekhnologii Katalizatorov (Scientific Foundations of the Technology of Catalysts) (Novosibirsk: Institute of Catalysis, Siberian Branch of the Academy of Sciences of the USSR, 1981) No. 13, p. 67 11. R A Buyanov,M A Tanatarov, in Sbornik Vsesoyuz. Shkoly po Katalizatoram. Lektsii (Collection of All-Union School on Catalysts. Lectures) (Novosibirsk: Institute of Catalysis, Siberian Branch of the Academy of Sciences of the USSR, 1980) Part 3, p. 4 12. V V Chesnokov, R A Buyanov, A D Afanas'ev, L M Plyasova Izv.Sib. Otd. Akad. Nauk SSSR. Ser. Khim. (5) 82 (1980) 13. R A Buyanov, V V Chesnokov, A D Afanas'ev Izv. Sib. Otd. Akad. Nauk SSSR. Ser. Khim. (4) 28 (1981) 14. R A Buyanov, V V Chesnokov, A D Afanas'ev Kinet. Katal. 20 207 (1979) a 15. V V Chesnokov, R A Buyanov, A D Afanas'ev Izv. Sib. Otd. Akad. Nauk SSSR. Ser. Khim. (2) 60 (1982) 16. V V Chesnokov, R A Buyanov, A D Afanas'ev, V S Babenko, in Kataliticheskaya Konversiya Uglevodorodov (Catalytic Conversion of Hydrocarbons) (Kiev: Naukova Dumka, 1978) No. 3, p. 38 17. V V Chesnokov, R A Buyanov, in Proceedings of the Second Indo-Soviet Seminar on Catalysis, Hyderabad, 1986 p. 22 18. V V Chesnokov,R A Buyanov,A D Afanas'ev Kinet. Katal.24 1251 (1983) a 19. V V Chesnokov, R A Buyanov Kinet. Katal. 28 403 (1987) a 20. I Stensgaard, L Ruan, E Laegsgaard, F Besenbacher Surf. Sci. 337 190 (1995) 21. WHuber, MWeinelt, P Zebisch, H-P Steinruck Surf. Sci. 253 72 (1991) 22. D R Huntley, S L Jordan, F A Grimm J. Phys. Chem. 96 1409 (1992) 23. Yu D Tret'yakov Tverdofaznye Reaktsii (Solid-Phase Reactions) (Moscow: Khimiya, 1978) 24. E V Khamskii Kristallizatsiya iz Rastvorov (Crystalisation from Solutions) (Leningrad: Nauka, 1967) 25. V V Chesnokov, Doctoral Thesis in Chemical Sciences, Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, 1999 26. A K Damask, K D Khoman Diffuziya v Metallakh s Ob'emno- tsentrirovannoi Reshetkoi (Diffusion in Metals with Body-Centred Lattice) (Moscow: Metallurgiya, 1969) 27.R T K Baker, P S Harris, in Chemistry and Physics of Carbon (Eds P L Walker Jr, P A Thrower) (New York: Marsel Deker, 1978) p. 83 28. Y Miura, T Uchijima, S Makishima J. Chem. Soc. Jpn., Ind. Chem. Soc. 71 86 (1968) 29. V V Veselov, P S Pilipenko Kinet. Katal. 11 1590 (1970) a 30. P S Pilipenko, V V Veselov Poroshk. Metall. 9 150 (1975) 31. V V Veselov, Doctoral Thesis in Chemical Sciences, L V Pisarzhevskii Institute of Physical Chemistry, Academy of Sciences of the Ukrainian SSR, Kiev, 1975 32. V V Veselov, P S Pilipenko, in Kataliticheskaya Konversiya Uglevodorodov (Catalytic Conversion of Hydrocarbons) (Kiev: Naukova Dumka, 1975) No. 2, p. 35 33. R T K Baker,M A Barber, P S Harris, F S Feates, R J Waite J.Catal. 26 51 (1972) 34. S A Bernardo, L S Lobo J. Catal. 37 267 (1975) 35. RTKBaker, P S Harris,RB Thomas,RJ Waite J. Catal. 30 86 (1973) 36. R T K Baker, R J Waite J. Catal. 37 101 (1975) 37. R T K Baker, P S Harris, S Terry Nature (London) 253 37 (1975) 38. R T K Baker, G R Gadsby, R B Thomas, R J Waite Carbon 13 211 (1975) 39. R T K Baker, P S Harris, H J Henderson, R B Thomas Carbon 13 17 (1975) 637 40. S D Gertsriken, I Ya Dekhtyar Diffuziya v Metallakh i Splavakh v Tverdoi Faze (Diffusion in Metals and Alloys in the Solid Phase) (Moscow: Fizmatgiz, 1960) 41. P L Gruzin, Yu A Polikanov, T B Fedorov Fiz. Met. Metalloved 4 94 (1975) 42. W Seith, T Heumann Diffusion in Metallen (Berlin: Springer, 1955) 43.J D Fast Interaction of Metals and Gases Vol. 2 (London: Macmillan, 1971) 44. I I Kovenskii Fiz. Met. Metalloved. 16 613 (1963) 45. VKKh Krol', P Delizhyur,KMirodatos Kinet. Katal. 37 698 (1996) a 46. M B Lee, Q Y Yang, S L Tang, S T Ceyer J. Chem. Phys. 85 1693 (1986) 47. M B Lee, Q Y Yang, S T Ceyer J. Chem. Phys. 87 2724 (1987) 48. J D Beckerle,Q Y Yang,A D Johnson, S T Ceyer J. Chem. Phys. 86 7236 (1987) 49. S T Ceyer, J D Beckerle, MB Lee, S L Tang, Q Y Tang,MA Hines J. Vac. Sci. Technol. A, Vac. Surf. Films 5 501 (1987) 50. A V Hamza, R J Madix Surf. Sci. 179 25 (1987) 51. T P Beebe, D W Goodman, B D Kay, J T Yates Jr J. Chem. Phys. 87 2305 (1987) 52. A G Sault, D W Goodman J. Chem. Phys. 88 7232 (1988) 53.X Jiang, D W Goodman Catal. Lett. 4 173 (1990) 54. I Chorkendorff, I Alstrup, S Ullmann Surf. Sci. 227 291 (1990) 55. L Hanley, Z Xu, J T Yates Surf. Sci. Lett. 248 L265 (1991) 56. F C Schouten, E W Kaleveld, G A Bootsma Surf. Sci. 63 460 (1977) 57. I Alstrup,M T Tavares J. Catal. 135 155 (1992) 58. M C Demicheli, E N Pouzi, A O Ferretti, A A Yermain J. Chem. Eng. 46 129 (1991) 59. I Alstrup,M T Tavares J. Catal. 139 524 (1993) 60. H J Grabke Ber. Bunsenges. Phys. Chem. 69 409 (1965) 61. H J Grabke Metall. Trans. 1 2972 (1970) 62. J-W Snoeck, G F Froment, M Fowles J. Catal. 169 250 (1997) 63. V V Chesnokov, V I Zaikovskii, R A Buyanov, V V Molchanov, L M Plyasova Kinet. Katal. 35 146 (1994) a 64. E R Gilliland, P Harriott Ind.Chem. Eng. 46 10 (1954) 65. G G Kuvshinov, Yu I Mogilnykh, D G Kuvshinov Catal. Today 42 357 (1998) 66. P S Pilipenko, V V Veselov, in Kataliticheskaya Konversiya Uglevodorodov (Catalytic Conversion of Hydrocarbons) (Kiev: Naukova Dumka, 1978) No. 3, p. 33 67. E G M Kuijperty, J W Jansen,A J van Dillen, J W Gent J. Catal. 72 75 (1981) 68. R A Campbell, J S Lenz, D W Goodman Catal. Lett. 17 39 (1993) 69. V B Fenelonov Poristyi Uglerod (Porous Carbon) (Novosibirsk: Institute of Catalysis of Russian Academy of Sciences, 1995) 70. S Iijima Nature (London) 354 56 (1991) 71. M Jose-Vacaman, M Miki-Yoshida, I Rendon, J G Santiesteban Appl. Phys. Lett. 62 657 (1993) 72. P E Nolan, A H Cutler, D C Lynch, in Engineering, Construction, and Operations in Space IV (Eds RGGalloway, S Lokaj) (New York: American Society of Civil Engineers, 1994) Vol.2, p. 1199 73. P Chen, H-B Zhang, G-D Lin, Q Hong, K R Tsai Carbon 35 1495 (1997) 74. T W Ebbesen, P M Ajayen Nature (London) 358 220 (1992) 75. C McKee Appl. Catal. A 98 N2 (1993) 76. L S K Pang, J D Saxby, S P Chattield J. Phys. Chem. 97 6941 (1993) 77. V I Zaikovskii, V V Chesnokov, R A Buyanov Appl. Catal. 38 41 (1988) 78. R A Buyanov, V V Chesnokov Zh. Prikl. Khim. 70 978 (1997) c 79. R A Buyanov, V V Chesnokov Khim. Inter. Ustoich. Razv. 3 177 (1995) 80. N A Prokudina, V V Chesnokov, N A Zaitseva, L M Plyasova, V I Zaikovskii, A V Golovin, V Yu Gavrilov Sib. Khim. Zh. (3) 122 (1991) 81. T Baird, J R Fryer, B Grant Carbon 12 591 (1974) 82.J R Rostrup-Nielsen Catalysis, Science and Technology (Eds J R Anderson, MBoudart) (Berlin: Springer, 1984) Vol. 5, p. 1 83. F J Derbyshire, D L Trimm Carbon 13 189 (1975) 84. S D Robertson Carbon 8 365 (1970) 85. E L Evans, J M Thomas, P A Thrower, P L Walker Carbon 11 441 (1973) 86. C W Keep, R T K Baker, J A France J. Catal. 47 232 (1977) 87. J R Rostrup-Nielsen J. Catal. 27 343 (1972)638 88. T Wada, H Wada, J F Elliott, J Chipman Metall. Trans. 2 2199 (1971) 89. R T K Baker, J J Chludzinski Jr J. Catal. 64 464 (1980) 90. L S Lobo, D L Trimm J. Catal. 29 15 (1973) 91. J L Figueiredo, D L Trimm J. Appl. Chem. Biotechnol. 28 611 (1978) 92. S D Robertson Nature (London) 221 1044 (1969) 93. G G Tibbetts, M G Devour, E J Rodda Carbon 25 367 (1987) 94.R T Yang , J P Chen J. Catal. 115 52 (1989) 95. R T Yang , K L Yang J. Catal. 90 194 (1984) 96. R T Yang , K L Yang J. Catal. 93 182 (1985) 97. J R Rostrup-Nielsen, D L Trimm J. Catal. 48 155 (1977) 98. M Audier, M Coulon, L Bonnetain Carbon 21 93 (1983) 99. M Audier, M Coulon, L Bonnetain Carbon 21 105 (1983) 100. M Audier, M Coulon, L Bonnetain Carbon 21 99 (1983) 101. A Sacco Jr, P Thacker, T N Chang, A T S Chiang J. Catal. 85 224 102. A Sacco Jr, P Thacker, in Proceedings of the 16th Biennial 103. P K DeBokx, A J H M Kock, E Boellaard, W Klop, J W Geus 104. A J H M Kock, P K DeBokx, E Boellaard, W Klop, J W Geus (1984) Conference on Carbon, San Diego, CA 1983 p. 525 J. Catal. 96 454 (1985) J. Catal. 96 468 (1985) 105. E Boellaard, P K DeBokx, A J H M Kock, J W Geus J. Catal. 96 481 (1985) 106. I Alstrup J. Catal. 109 241 (1988) 107. E C Bianchini, C R F Lund J. Catal. 117 455 (1989) 108. G G Kuvshinov, S G Zavarukhin, Yu I Mogil'nykh, D G Kuvshinov Khim. Promst 5 48 (1998) 109. M Audier, M Coulon Carbon 23 317 (1985) 110. M Audier, A Oberlin, M Oberlin, M Coulon, L Bonnetain Carbon 19 217 (1981) 111. M Audier, A Oberlin, M Coulon J. Cryst. Growth 55 549 (1981) 112. M Audier, A Oberlin, M Coulon J. Cryst. Growth 57 524 (1982) 113. M Audier, M Coulon, A Oberlin Carbon 18 73 (1980) 114. M Audier, J P Simon, P Guyot Acta Metall. 34 1982 (1986) 115. M Audier, P Bowen, W Jones J. Cryst. Growth 63 125 (1983) 116. L Bonnetain, P Gadelle,M Auder, in Carbon Fibres, Filaments and Composites (NATO ASI, Ser. C, Math. Phys. Sci.) (Eds J L Figueiredo, C A Bernardo, R T K Baker, K J Huttinger) (Dordrecht: Kluwer Academic, 1990) Vol. 177, p. 507 117. N M Rodriguez, A Chambers, R T K Baker Langmuir 11 3862 (1995) 118. F C Schouten, O L J Gijzeman, G A Bootsma Surf. Sci. 87 1 (1979) 119. F C Schouten, O L J Gijzeman, G A Bootsma Bull. Soc. Chim. Belg. 88 541 (1979) 120. G A Somorjai Surf. Sci. 242 481 (1991) 121. M Eizenberg, J M Blakely Surf. Sci. 82 228 (1979) 122. R A Buyanov, V V Chesnokov Khim. Inter. Ustoich. Razv. 5 619 (1997) d 123. V V Chesnokov, R A Buyanov, A D Afanas'ev Kinet. Katal. 25 1152 (1984) a 124. L B Avdeeva, O V Goncharova, D I Kochubey, V I Zaikovskii, L M Plyasova, B N Novgorodov, Sh K Shaikhutdinov Appl. Catal. A 141 117 (1996) 125. V N Parmon Catal. Lett. 42 195 (1996) 126. R S Wagner,W C Ellis Trans. AIME 233 1053 (1965) 127. F Benissad, P Gadelle, M Coulon, L Bonnetain Carbon 26 425 (1988) 128. P Gadelle, in Carbon Fibres, Filaments and Composites (NATO ASI, Ser. C, Math. Phys. Sci.) (Eds J L Figueiredo, C A Bernardo, R T K Baker, K J Huttinger) (Dordrecht: Kluwer Academic, 1990) Vol. 177, p. 95 129. G G Tibbetts, E J Rodda Mater Res. Soc. Symp. Proc. 111 49 (1988) 130. G G Tibbetts, M P Balogh, in The 23rd Biennial Conference on Carbon (Expanded Abstracts and Program) Pennstate Campus, 1997 Vol. 2, p. 288 131. H P Boehm Carbon 11 583 (1973) 132. M P Manning, R C Reid Ind. Eng. Chem., Proc. Des. Dev. 16 358 (1977) 133. V V Chesnokov, V I Zaikovskii, R A Buyanov, V V Molchanov, L M Plyasova Catal. Today 24 265 (1995) V V Chesnokov, R A Buyanov 134. C A Bernardo, I Alstrup, J R Rostrup-Nielsen J. Catal. 96 517 (1985) 135. P C M van Stiphout, D E Stoble, F Th V D Scheur, J W Geus Appl. Catal. 40 219 (1988) 136. R A Buyanov, A D Afanas'ev, V V Chesnokov Kinet. Katal. 19 1072 (1978) a 137. D A Wesner, F P Coenen, H P Bonzel, in The Structure of Surfaces II (Eds J F van der Veen,MAVan Hove) (Berlin: Springer, 1988) Vol. 11, p. 612 138. R A Buyanov, A D Afanas'ev, V V Chesnokov, V S Babenko, in Geterogennyi Kataliz (Tr. IV Mezhdunarod. Simp.) [Heterogeneous Catalysis (Proceedings of the IVth International Symposium)] (Varna: Bulgarian Academy of Sciences, Centre on Chemistry, 1979) Part 1, p. 355 139. P E Nolan, M J Schabel, D C Lynch, A H Cutler Carbon 33 79 (1995) 140. Sh K Shaikhutdinov, V I Zaikovskii, L B Avdeeva Appl. Catal. A. 148 123 (1996) 141. G G Tibbetts J. Cryst. Growth 73 431 (1985) 142. M Ishioka, T Okada, K Matsubara,M Endo Carbon 30 859 (1992) 143. M Ishioka, T Okada, K Matsubara,M Endo Carbon 30 865 (1992) 144. M Ishioka, T Okada, K Matsubara Carbon 30 975 (1992) 145. T Kato, K Haruta, K Kusakabe, S Morooka Carbon 30 989 (1992) 146. A Oberlin, M Endo, T Kayama J. Cryst. Growth 32 335 (1976) 147. G G Tibbetts J. Cryst. Growth 66 632 (1984). 148. G G Tibbetts, in Carbon Fibres, Filaments and Composites (NATO ASI, Ser. C, Math. Phys. Sci.) (Eds J L Figueiredo, C A Bernardo, R T K Baker, K J Huttinger) (Dordrecht: Kluwer Academic, 1990) Vol. 177, p. 73 149. V I Zaikovskii, V V Chesnokov, R A Buyanov Kinet. Katal. 40 612 (1999) a 150. S Nagakura J. Phys. Soc. Jpn. 12 482 (1957) 151. G Carturan, G Cocco, S Enzo, R Gauzerla,M Jenarda Mater. Lett. 7 47 (1988) a�Kinet. Catal. (Engl. Transl.) b�Russ. J. Phys. Chem. (Engl. Transl.) c�Russ. J. Appl. Chem. (Engl. Transl.) d�Chem. Sastainable Dev. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2000

数据来源: RSC