摘要:

Russian Chemical Reviews 68 (5) 343 ± 344 (1999) To the 275th anniversary of the Russian Academy of Sciences The Russian Academy of Sciences was founded as an `assem- bly of educated people' by a decree of Peter the Great dated January 28, 1724. The three main goals clearly defined in the decree were (1) production of the scientific knowledge itself, (2) training of scientists and (3) education of young people in the fundamentals of sciences. It is remarkable that none of these goals has lost its urgency nowadays and the modern Charter of the RAS could easily begin with the decree of Peter the Great. Moreover, Peter the Great thought the activity of the Academy of Sciences in the education sphere to be one of the most important. That is why in 1726, a secular University (a higher-education establishment), later called Academic, which was the first in Russia, was founded by the Academy.Unfortunately, it stopped existing in 1766, despite the efforts of M V Lomonosov, who was the vice-chancellor of the University from 1758 to 1765. Thus, the efforts undertaken now for integration of the fundamental science and higher education are directed to the realisation of the idea put forward by the founder of the Academy of Sciences, of course, in the present-day conditions. The beginning of the establishment of the chemical science in Russia is associated with the name ofMV Lomonosov. Chemical laboratories with analytical, technological and pharmaceutical profiles had also existed in Russia before Lomonosov, but he created the first research chemical laboratory.And what is more, his laboratory had been the prototype of the research institutes of the Academy, which appeared later. This laboratory existed till 1793 in its original form. Then hard times began, when academi- cians in chemistry had to carry out their research work either in their home laboratories or in the chemists' shops or higher- education establishments. It was not until the beginning of the XIX century that a room in the main building of the Academy on the Neva quay was allocated for a chemical laboratory. K S Kirchhoff, G I Gess and Yu F Fritzsche carried out their experiments there. After a conflagration in 1859, a new laboratory was built by the initiative of Yu F Fritzsche andN N Zinin in the Eighth Line of the Vasil'evskii island; this building still exists.A M Butlerov, N N Beketov, F F Beilstein, P I Walden, N S Kurnakov and V N Ipatieff also worked in this well-known laboratory. An Academy printing office and type-foundry were founded during the Peter the Second's reign. The Academy started spread- ing its scientific production in publications. In the course of time, the Russian Academy of Sciences started publishing periodicals and monographic editions in various fields of science. In the early stages, virtually no reviews were published in scientific journals because there was no need for them. However, the increase in the amount of scientific information and the appearance of new fields of science, mostly in border areas, caused the necessity of summarising and analysis of the scientific infor- mation.In 1932, the `Uspekhi Khimii' journal was founded by a decree of the USSR People's Commissariat on Education in order to `publish the major reviews in theoretical and experimental chem- istry and chemical technology, which are of high generalising and summarising importance and are the result of creative and critical #1999 Russian Academy of Sciences and Turpion Ltd thought uniting the latest scientific achievements into a harmo- nious single whole.' The early stage of activity of the `Uspekhi Khimii' journal is associated with the names of its founders � N P Gorbunov, A N Bakh, A M Berkengeim, B M Berkengeim, S I Vol'fko- vich, N D Zelinsky, N S Kurnakov and A N Frumkin.{ The main aim of the journal in those years was to acquaint people with the achievements of the fundamental chemical science and the practical use of them in the national economy.`Uspekhi Khimii' efficiently responded to the needs of the chemical industry in the solution of fuel and energy problems, development of the effective chemical and technological processes, chemisation of agriculture and forestry and development of the mineral-stock base. In particular, an article by A E Fersman entitled `New problems in the investigation of mineral raw materials' was published in the first issue of the journal. In the initial stage, `Uspekhi Khimii' paid considerable attention to the reviewing of monographs and information material.Reports on scientific sessions and general meetings in the USSR Academy of Sciences, on international conferences and congresses and Mendeleev Congresses on General and Applied Chemistry were published. From the methodological point of view, the most significant achievement of `Uspekhi Khimii' in its starting period was elaboration of an independent and consistent strategy of reporting on the present-day chemical achievements. This strategy, in particular, was to orient the journal towards the quantitative aspects of various fields of chemistry and elucidation of the physical processes underlying chemical transformations. Yet another important feature of `Uspekhi Khimii' was the fact that since its foundation, the journal has published not only reviews in chemistry but also those in related branches of science.Papers on X-ray diffraction analysis (W Bragg), Raman spectro- scopy (L I Mandel'shtam and G S Landsberg), geochemistry (A P Vinogradov), internal rotation in molecules and rotational isomerism (M V Vol'kenshtein) and the problems of chemical topology (V Prelog) were published in the journal in different years. Being aware of the fact that the progress of chemical science is largely determined by the advances in modern theoretical and experimental physics, the journal published in these years many purely physical reviews of Russian and foreign researchers such as P Debye, E V Shpolsky, F Joliot-Curie, I Joliot-Curie, O Hahn, N Bohr, E Rutherford, E SchroÈ dinger and others.The develop- ment of quantum mechanics in these years served as a powerful stimulus for the progress in the classical fields of physical chemistry and for the creation of a new one referred to as chemical physics. The pre-war decades have been marked by great achievements in the development of chemical physics. The journal showed a great efficiency at publishing the fundamental works on this {A comprehensive analysis of the activity of the `Uspekhi Khimii' journal from 1932 to 1982 can be found in the paper by V A Nikanorov published in issue 11, 1982.344 subject, which were performed not only in Russia but also abroad. N N Semenov, D A Frank-Kamenetsky, S Hinshelwood and V N Kondratiev were among its authors.For a long period, along with the reviews specially written for it, the journal also published translations of articles from foreign review journals. Later, this was no longer necessary and the journal became written mainly by Russian scientists, although some outstanding works of foreign authors still appeared. The scope of publications in `Uspekhi Khimii' at that period was extremely broad. It included organic chemistry (the reviews by N Ya Dem'yanov and M Tiffeneau on the molecular rear- rangements of cyclic hydrocarbon systems discovered by them), coordination chemistry (works by I I Chernyaev and A A Grin- berg), the chemistry of organometallic compounds (A N Nes- meyanov and K A Kocheshkov), geochemistry (A P Vino- gradov), and the polymer chemistry (H Staudinger and W Carothers).It is typical that in different years, the journal has published reviews by almost all the winners of the Nobel Prize in chemistry. The 1930s were characterised by intense development of the chemistry of natural compounds. Suffice it to say that during fifteen pre-war years, about ten Nobel Prizes were awarded in this field. The majority of the laureates appeared in the pages of `Uspekhi Khimii' with their reviews: A Windaus (sterols), P Kar- rer (carotenoids and flavins, vitamins A and B2), R Kuhn (car- otenoi and vitamins of group B), A Butenandt (hormones) L RuzÏ icÏ ka (polymethylenes and higher terpenoids), and H Wie- land (cholic acids and oxidative dehydrogenation in the series of natural products).During the post-war years, the journal published reviews by R Robinson (study of plant anthocyanines), R B Woodward (vitamin B12 and the development of the chemistry of natural substances), A Todd (nucleotides and coenzymes), D Crowfoot- Hodgkin (X-ray diffraction analysis of protein crystals) and I N Nazarov (the chemistry of vinylethynylcarbinols). In the pre-war and post-war years, the journal published a series of important reviews by N D Zelinsky, A A Balandin, B A Kazansky, N I Kobozev and A V Frost devoted to various aspects of catalysis and catalytic transformations. During the war and the first post-war years, the size of scientific publications in `Uspekhi Khimii' diminished but the information and bibliographic material was exceptionally diversi- fied.In the 1950s and the 1960s, vigorous development of chemical science and industry started.The size of the journal in that period increased and the activity of the Editorial Board, which directed the journal to new lines of research in chemistry and promising related branches of science, was enhanced. One new branch of science was the chemistry of organoele- ment compounds. Beginning with the paper by A N Nesmeyanov entitled `The organoelement compounds and the Periodic system,' which appeared just after the war, the journal has regularly published reviews on various aspects of the chemistry of organo- element compounds. The chemistry of polymers was yet another priority of the journal in the 1960s; various aspects of this branch of chemistry have been considered in reviews by V A Kargin, V V Korshak and other scientists.The 1960s were marked by one more event important for the journal. Since 1960, it started to be translated into English under the title `Russian Chemical Reviews'; this significantly broadened the area in which the journal was distributed and gave the possibility for the foreign readers to get acquainted more exten- sively with the achievements of Russian scientists. When observing the present-day period of activity of `Uspekhi Khimii,' it is necessary to mention the truly avalanche-like growth of the size of primary information accompanied by an increase in its difficulty and weight. In this connection, review journals become even more important.In addition, a modern review To the 275th anniversary of the Russian Academy of Sciences journal is not only to inform the reader on the new scientific achievements but also to orient him correctly and neatly. An important quality characteristic of review scientific liter- ature is the so-called `data compression index,' which is the ratio of the size of primary information to the size of the review. As early as 1936, this coefficient for the `Uspekhi Khimii' journal was equal to 20 and by 1976, it reached 50, which is considered by specialists to be the optimum value. On further increase in this index, the information may become distorted. Yet another quantitative parameter for estimating the quality and topicality of scientific publications is the `citation index'.It is not by chance that the highest citation index among all the chemical periodicals belongs to `Chemical Reviews'. Regarding this index, the `Uspekhi Khimii' journal occupies a honourable place among other journals with similar subject matter all over the world and has the highest citation index among all the chemical periodicals in Russia. The citation index sharply increased after the English version of the journal started being published by the Russian Academy of Sciences together with the Royal Chemical Society of the UK in 1993. The English version of the journal is an exact translation of the Russian one and appears almost simultaneously with it.The whole size of the journal is approximately 1200 pages per year, which makes up 12 issues. Each issue contains 4 to 6 reviews. From 1996 the reviewing of the papers has become stricter (two Russian referees and one British referee), which contributes to the quality of publications. Since 1997, in addition to the hard copy of the journal, an electronic version has also been produced. Annual archive CD- ROM's are also produced. Undoubtedly, all this is favourable for the popularity and availability of the journal. At present, the journal publishes reviews on the following fields of chemistry: molecular structure and quantum chemistry, chemistry of coordination compounds, analytical chemistry, chemical physics, physical chemistry (including catalysis), organic and organometallic chemistry, chemistry of macromolecules, biochemistry and bioorganic chemistry and materials science. The majority of papers published in the last years have been submitted by scientists working in the RAS or in the Academies of Sciences of the CIS countries. The geographic area in which the authors of the journal work is quite large. Moscow together with some nearby scientific centres (Chernogolovka, Pushchino) comes first in the number of articles published, St. Petersburg occupies the second place. The `Uspekhi Khimii' journal makes a significant contribu- tion to the development of the chemical science in Russia and to popularisation of the achievements of the Russian science abroad. It properly represents the Russian Academy of Sciences, which celebrates its 275th anniversary.

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (5) 345 ± 363 (1999) Molecular ferromagnets V I Ovcharenko, R Z Sagdeev Contents I. Introduction II. Systematisation principles III. M2M0 systems IV. M3M0 systems V. M4M0, M5M0 and M7M0 systems VI. The problem of organic ferromagnets VII. Conclusion Abstract. The state-of-the-art, achievements, problems and pro- spects of an actively developing area of supramolecular chemistry, viz., the design of molecular ferromagnets, are considered. The bibliography includes 343 references. I. Introduction This review deals with a rapidly developing area of modern chemistry, viz., the design of molecular ferromagnets. Like any other area of natural sciences which is still in its infancy, it is not yet liable to generalisations, as many works are pioneering.Various approaches are used to solve the diverse problems arising in these studies, which require a high level of integration of the efforts of specialists in different fields of knowledge. Nevertheless, certain approaches to the design of molecular ferromagnets have now taken shape.1±18 It is obviously impossible for a single review to consider every aspect of the design of molecular ferromagnets in detail. There- fore, in this paper we dwell on the basic classes of the chemical compounds which are characterised by established synthetic strategies and the principal ways for the construction of highly dimensional multispin systems. When the development of methods for the synthesis of highly dimensional systems is concerned, one has, to some extent, to part with the classical chemical thinking, which generally deals with the design, modification or transformation of an individual molecule.No doubt, a molecule taken as a subunit is also important. However, the arrangement of the molecules is no less important in the design of molecular ferromagnets. Even more important is their selective arrangement in a highly dimensional assembly. Arrangement of individual molecules in a highly effective assem- bly bound by exchange interactions is a key problem in the design of molecular ferromagnets. Before we start a detailed discussion of this problem it is necessary to define the notion of a `molecular ferromagnet'. V I Ovcharenko, R Z Sagdeev International Tomography Centre, Siberian Branch, Russian Academy of Sciences, ul.Institutskaya 3a, 630090 Novosibirsk, Russian Federation. Fax (7-383) 233 13 99. Tel. (7-383) 233 19 45. E-mail: ovchar@tomo.nsc.ru (V I Ovcharenko). Tel. (7-383) 233 14 48. E-mail: root@tomo.nsc.ru (R Z Sagdeev) Received 2 February 1999 Uspekhi Khimii 68 (5) 381 ± 400 (1999); translated by S S Veselyi #1999 Russian Academy of Sciences and Turpion Ltd UDC 538.22+541.67+541.49+547.78+548.736 345 346 347 350 354 357 358 The term `molecular ferromagnet' may seem to a certain degree paradoxical. The classical ferromagnet is a macroobject, for example, a piece of iron, nickel, or cobalt or an alloy of these metals, but in no way a separate molecule.By its very nature, a magnetic phase transition into the ferromagnetic state is a cooperative phenomenon inherent in a macroobject as a phase. Nevertheless, in recent years the term `molecular ferromagnet' has been widely used in scientific publications dealing with the design of multispin systems.{ Let us define the essence of this term more precisely. The studies in the field of molecular ferromagnets deal with the synthesis of organic, organometallic and, currently, mostly coordination compounds the solid phases of which are formed from separate molecules or ions that existed initially in a solution and contained paramagnetic centres. In these solid phases, as in classical ferro- and ferrimagnets or weak ferromagnets, magnetic phase transition into the ferro-, ferrimagnetic or weak ferromag- netic state is detected below a critical temperature. These days, researchers can plan beforehand, with a certain degree of success, the synthesis of compounds with a desired solid state structure that will be formed from molecular precursors in solution.These molecules should form layered or framework polymers when crystallised from solution. As a matter of fact, this field of studies which implies the design of highly dimensional structures is the subject of supramolecular chemistry. The formation of layered or framework structures in the solid phase of molecular ferromagnets is a prerequisite for the magnetic phase transition into the ferromagnetic state. This condition is necessary but insufficient in the chemical design of molecular ferromagnets.For this transition to occur, the paramagnetic centres in polymers should be linked with such an atomic chain serving as an exchange channel 13 that the polymer has the ability to undergo magnetisation in a magnetic field. The following consideration is important in this case: the higher the efficiency of the exchange channels in exchange interactions between the unpaired electrons of the paramagnetic centres the higher critical temperature (Tc) that can be achieved. It is within the limits set by these requirements that regular studies of the effect of conditions of directed chemical synthesis on the magnetic parameters (the { Reviews and monographs on `molecular ferromagnets' have been published, the titles of which, like the present paper, sometimes contain this word combination (or the equivalent expressions: `molecular mag- netics', `molecular magnets', `molecule-based ferromagnets') as the key words.1 ±14346 Curie temperature or the Neel temperature, spontaneous magnet- isation, etc.) of the final magnetically active product are being performed, i.e., the effects of the composition, structure of the starting molecules, conditions of the synthesis and crystallisation on the magnetic properties of the resulting solid phases are being studied.Magnetic-structural relationships understood as relation- ships between the observed magnetic properties of a compound and its chemical and structural characteristics are used in the search for subjects with the required properties.These relation- ships are either calculated or determined experimentally. Thus, the word `ferromagnet' in the term `molecular ferro- magnet' is used in the ordinary sense, while the definition `molecular' reflects the methodology of the approach: the starting individual molecules and/or molecular ions should have a struc- ture which would enable them to form a spatial structure favourable for a magnetic phase transition during formation of a solid phase. The problem can be solved by various chemical methods combined by the term `chemical design of molecular ferromagnets'. The development of principles for the synthesis of molecular ferromagnets and the determination of their inherent magnetic- structural relationships are fundamental problems.There are few natural magnetic materials. Most of them are transition metals and their oxides; the number of magnetic materials was mainly increased by virtue of metal alloys and binary oxides. However, studies on the design of molecular ferromagnets make it possible to expand considerably the range of magnetically active com- pounds and magnetic materials. In addition, the design of molecular magnets and conventional, mild synthetic approaches provide a possibility to create new magnetic materials using technologies alternative to the traditional power-consuming met- allurgical technologies for the production of magnetically active oxide materials. The number of studies in the field of the design of molecular ferromagnets increases rapidly.Owing to the aforementioned high integration requirements concerned with the problem, experts in the field of inorganic, organometallic or organic chemistry and in materials science, chemical physics and physics of magnetism are involved in collaboration. The results of these studies are extremely `scattered' in scientific publications; therefore, biennial international conferences on molecule-based magnets play an important uniting role (see, e.g., Refs 15 and 16). The proceedings of these conferences are published in the journal Molecular Crystals and Liquid Crystals (see, e.g., Refs 17 and 18). However, these are far from being the only international conferences with scientific sections devoted to molecular ferromagnets (see, e.g., Refs 19 and 20).In this review, we made an attempt at the generalisation and systematisation of the main approaches and strategies existing at present in the field of the synthesis of molecular ferromagnets. As a rule, the design of particular compounds triggers an entire complex of studies dealing with the analysis of structure, mag- netic-structural relationships and mechanisms of exchange inter- actions between the unpaired electrons in paramagnetic centres. In all of the molecular ferromagnets discussed in the present review the structure of a molecule or a bridging ligand involves at least one carbon atom. The presence of one carbon atom, or better several carbon atoms, reflects the common desire of the research- ers to use as soon as possible all the potential of synthetic organic chemistry in the given area.Here we do not consider the classical magnetically active systems in which the metal atoms in the solid phase are separated by the individual oxygen, carbon, sulfur, halogen or boron atoms. II. Systematisation principles For a systematic generalisation of chemical compounds involved in the design of molecular ferromagnets, we offer a specific V I Ovcharenko, R Z Sagdeev `language'. It is therefore reasonable to discuss the principles underlying its construction. In this work we speak of metal- containing molecular ferromagnets as being two or three-dimen- sional systems (as regards their structures);{ hence the common features of all of the systems under consideration are the presence of a paramagnetic metal ion in a compound and the presence of a bridging group of atoms linking the metal ions to form a layer or a framework.Therefore, the description of a particular system can be given by the key formula MnM0, where M and M0 are paramagnetic metal ions (the same, different or existing in differ- ent oxidation states) and n is the smallest number of atoms in a bridging unit connecting theMandM0 ions. Thus, we shall refer cyanide complexes of metals containing, e.g., such fragments as V3+7C:N7V3+, Cr3+7C:N7Ni2+ and Fe2+7C:N7Fe3+ to the M2M0 class, since the bridging cyanide is a diatomic unit. Carboxylate, oxalate, bipyrimidine, azide or dicyanamide polynuclear complexes containing triatomic bridges as the shortest ones belong to the M3M0 class.R N N O O C Mn O Mn O Ni Ni Cr Cr N N O O Cu N N N N Cu N Cu C N C Cu Cu According to the above approach, the complexes with 3- and 2-imidazoline derivatives are regarded as M4M0 systems. Ni Co R N N R N N O O Ni Co Heterospin polynuclear metal complexes with nitronyl nitroxyl radicals or 2-imidazoline 3-oxide paramagnetic deriva- tives, which play the bridging role, belong to the M5M0 systems. Ni O N R NO Ni More complicated situations are also possible where one bridging ligand can form exchange channels with different num- bers of atoms between the metal ions or the bridging function is played by several ligands creating the exchange channels with different numbers of bridging atoms.We designate such systems asMnn0M0, if the ligands are the same, andMn(n0)M0 if the ligands are different. Fragments of an M31M0 system with one ligand and M3(3)M0 and M3(1)M0 systems with different bridging ligands are shown below as examples. { It should be noted that structural dimensionality, i.e., molecular (zero- dimensional), chain (one-dimensional), layered (two-dimensional) and framework (three-dimensional) compounds, should not be confused with magnetic dimensionality, i.e., isolated exchange clusters (zero-dimen- sional), exchange chains (one-dimensional), etc.; in many cases, the structural and magnetic dimensionalities do not correlate.Molecular ferromagnets NNN M N N N N M M NNN O O M M N O O N N M NNN O O O O M M M O O O O NNN The systematisation suggested here is general and is not limited to specific metal ions or bridging ligands.III. M2M0 systems The molecular ferromagnets of the M2M0 class primarily include coordination compounds with the bridging cyanide ligands. The ability of homo- and heterometallic hexacyanides to undergo ferro- and ferrimagnetic ordering at low temperatures has been known for a long time (see, e.g., Refs 21 and 22). However, the problem was recently considered in a different way23 ± 31 and an essentially new contribution to the design of high-temperature molecular ferromagnets of theM2M0 type based on metal cyanide complexes was made.The first synthetic coordination compound, viz., Prussian blue, whose framework structure was established by Buser et al.,32 may be considered as an ancestor of this group of molecular ferromagnets. The highly symmetrical octahedral coordination anion [M(CN)6]n7 serves as a starting block for the construction of two- and three-dimensional structures.33 If an additionalM0 ion is introduced, the system can undergo polymerisation along two or three mutually perpendicular directions. LM0 L N L L L C C C M N M0 M0 N C L L L L N M0 L M0 M0 N N C C C C N M M0 M0 N C C N N M0 M0 N N M N N O O N M O OM 347 Prussian blue itself, FeIII 4 [FeII(CN)6]3 .xH2O(x=14 ± 16), has a framework structure (cubic face-centred structure) and under- goes a magnetic phase transition at 5.6 K.34 The low temperature of the magnetic ordering is due to the fact that the FeII ion is in the low-spin d 6 state. As a result, the exchange interactions between the paramagnetic centres (the FeIII ions) located in the FeIII7N:C7FeII7C:N7FeIII fragments at a distance of 10.6 A Ê from each other are rather weak.26 Thus, it seemed reasonable to assume that the magnetic ordering temperature will be higher if all of the metal ions are paramagnetic. Verdaguer et al.23 ± 31 were the first to perform a directed synthesis of Prussian blue analogues in which the iron ions were replaced by other metal ions in variable ratios.Table 1 presents some data for hexacya- nometallates of the general formula M3[M0(CN)6]2 from which it is possible to follow the regularities of changes in the magnetic ordering temperature (we do not specify its type, i.e., ferromag- netic, ferrimagnetic or antiferromagnetic). Table 1. Magnetic ordering temperatures for hexacyanometallates M3[M0(CN)6]2. M M Tc /K Tc /K M3[Mn(CN)6]2 M3[Cr(CN)6]2 37 37 MnII NiII 315 240 66 VII CrII CuII MnII 60 M3[Fe(CN)6]2 53 23 16 NiII CoII FeII 209 23 15 CuII MnII NiII CoII In a series of complexes containing the [Cr(CN)6]37 anions, the values of Tc are always higher than those for the analogues with the [Mn(CN)6]37 or [Fe(CN)6]37 anions; the highest Tc values are observed for V3[Cr(CN)6]2 and Cr3[Cr(CN)6]2.The latter is commonly depicted as Cr5(CN)12. Thus, there is an obvious trend of the increase in the critical temperature if the compounds involve VII, CrII and CrIII containing a significant number of unpaired electrons in the t2g orbitals. Hexacyanome- tallates (Et4N)0.5Mn1.25[V(CN)5](H2O)2 (Tc=230 K), Cs2Mn[V(CN)6] (Tc=125 K),30 Cs0.75[Cr2.125(CN)6] (Tc= 190 K)25 characterised by high Tc also contain the VII or CrII ions, which is not quite convenient from the practical viewpoint as they are unstable in this oxidation state under ordinary condi- tions. Nevertheless, the compound VII 0:42VIII 0:58[CrIII(CN)6]0.86 . . 2.8H2O with Tc=315 K obtained in an inert atmosphere (it is sensitive to the atmospheric oxygen) retains its magnetic proper- ties for several weeks when dispersed in kerosene or mixed with a glue.27 Based on this compound, a thermomagnetic switch and a device for the absorption of solar energy were designed and these were shown to function properly.24, 35 The character of exchange interactions in theM7N:C7M0 clusters can be predicted with high reliability under the assump- tion that these interactions between the unpaired electrons on the eg orbitals of one metal and the unpaired electrons on the t2g orbitals, orthogonal to the former, of another metal should be ferromagnetic, while the interactions between the unpaired elec- trons on the t2g orbitals of both metals should be antiferromag- netic.23 ± 25, 36, 37 By calculating the number of possible channels of exchange interaction according to the ferromagnetic (F) and antiferromagnetic (AF) mechanisms, it is possible to predict the character of ordering in the solid phase of a specific hexacyano- metallate.Below the critical temperature, ferromagnetic ordering348 of the spins of the paramagnetic centres predominates if F>AF, and antiferromagnetic ordering predominates if AF>F. According to this scheme, in all of the hexacyanometallates with Tc>100 K discussed above, where the t2g7t2g interactions predominate and the exchange interactions between the para- magnetic centres of the neighbouring metal ions are antiferro- magnetic, the magnetic moments of the sublattices of one metal ion are partially compensated by the magnetic moments of the sublattices of the other metal ion below the critical temperature. In systems of the type Ni3[Cr(CN)6]2 and Cu3[Cr(CN)6]2, the unpaired electrons lie on orbitals of different symmetries (eg and t2g), and the exchange interactions between the paramagnetic centres in these systems are ferromagnetic, which is favourable for the synthesis of compounds with high spontaneous magnet- isation.23 The results obtained 23 ¡¾ 31 stimulated the study 38, 39 of a series x MnII 1¢§x)1.5[CrIII(CN)6] .8H2O. The solids with this of solids (NiII composition were obtained in two ways: (1) by mechanically mixing pure Ni1.5[Cr(CN)6] . 8H2O and Mn1.5[Cr(CN)6] . 7.5H2O; in the former, the unpaired electrons of the neighbouring paramagnetic centres (NiII, CrIII) are coupled by ferromagnetic exchange interactions [the exchange interaction integral (J) is positive]; in the latter, the spins of the neighbouring centres(MnII,CrIII) mostly interact antiferromagnetically (J<0); (2) by mixing at the molecular level, i.e., in the synthesis of x MnII 1¢§x)1.5[CrIII(CN)6] .8H2O; the value of x was set by the (NiII initial ratio of the reagents NiCl2, MnCl2 and K3Cr(CN)6. As expected, the characteristics of the mixture obtained by the former method are additive values proportional to the fractions x and 17x of the mixed compounds; for example, the magnet- isation of the mechanical mixture grows linearly as x increases from 0 to 1.The variation of the characteristics of the phase obtained by the latter method (magnetisation, the Weiss temper- ature, coercive field) with x is essentially different due to the competition between the ferro- (J>0) and antiferromagnetic (J<0) interactions. As a result, e.g., the spontaneous saturation magnetisation (Js) quickly decreases from 4.38 mB to almost zero when x increases from 0 to 0.43; when x increases from 0.43 to 1, Js quickly increases from 0 to 5.57 mB according to the equation 32 a 3ax ¢§ 2:5O1 ¢§ xUa 2 Js a 2mB . The coercive field is a maximum (680 G) at x ' 0:43 where the parallel spins of CrIII and NiII are almost completely compensated by the antiparallel spins ofMnII. For this reason, the experimental temperature dependences of magnetisation for the (NiII 0:38MnII 0:62)1.5[CrIII(CN)6] .8H2O phase coincided with the the- oretical Neel curves: when measurements were carried out in an external field of 1000 G (stronger than the coercive field), the Js(T ) dependences contained two maxima, while in an external field of 10 G (weaker than the coercive field) one maximum was observed and the curve intersected the abscissa axis (i.e. the magnetisation sign changed) at 40 K. (It should be noted that at x=0 or 1, the magnetisation of the specimens below the critical temperature increased monotonically as the temperature was decreased; these measurements were carried out in a field of 1000 G.) As a result, the magnetisation of a specimen at temper- atures below 40 K was positive in an external magnetic field of H>Hc and negative in a field ofH<Hc.Thus, it was shown38, 39 that by varying the composition of a compound by chemical methods, it was possible to obtain phases in which the direction of the magnetic poles of the species was changed under the effect of an external magnetic field. The critical temperatures of Prussian blue analogues capable of magnetic ordering can also be changed by electrochemical 40 or optical stimulation.41 In the latter case, consecutive irradiation of, e.g., K0.2Co1.4[Fe(CN)6] . 6.9H2O with red (l=660 nm) and blue (l=450 nm) light at a sufficiently low temperature (5 K) made it possible to increase and then to decrease the magnetisation of the V I Ovcharenko, R Z Sagdeev x CrII 1¢§x)1.5[- specimen; this is of undoubted interest for those engaged in the development of magnetic-optical memory blocks.Pioneer stud- ies 40, 41 showed that the targeted selection of photosensitive hexacyanometallates can be carried out by molecular design. Japanese researchers implemented light-controlled inversion of magnetic poles on specimens of the mixed ferri-ferromagnet (Fe0.4Mn0.6)1.5[Cr(CN)6] . 7.5H2O42 and developed a technology for growing thin coloured magnetic films of (FeII CrIII(CN)6] . zH2O on electrode surfaces.43 A principle for the synthesis of hexacyanometallates with high critical temperatures was suggested.25, 30 It was found in studies of specific hexacyanometallate series that the synthesis of molecular ferromagnets (structural analogues of Prussian blue) should be based on ions of the first transition period metals (VII, VIII, CrII), despite their small stability under ordinary conditions.Compar- ison of the magnetic properties of Cs2MnII[VII(CN)6] (see Ref. 30) with those of CsMnII[CrIII(CN)6] .H2O (see Refs 44 and 45) and MnII[MnIV(CN)6] . xH2O, which are isostructural to the former, is indicative in this respect (see also Table 1).46 All the three compounds contain MnII ions in the high-spin d5 configuration in a weak ligand field (N6-octahedral coordination environment) and metal ions in the d 3 configuration (VII, CrIII and MnIV) located in a strong ligand field (C6-octahedral coordination environment).The basic difference between the compounds of this period is that the energy of the t2g orbitals for d 3 ions decreases with an increase in the element atomic number, i.e., on going from VII to CrIII and then to MnIV. For example, the magnetic ordering temperatures of 125, 89.8 and 48.7 K for Cs2MnII[VII(CN)6], CsMnII[CrIII(CN)6] .H2O and MnII[MnIV(CN)6] . 1.14H2O, respectively, show that the introduction of a d 3 transition metal ion with higher energies of the t2g orbitals results in compounds with higher temperatures of the magnetic phase transition into the ferromagnetic state. In other words, the more effective the dative bonding of the bridging cyanide ligands due to bonding of their p* orbitals the higher the energy of interaction between the spins of the neighbouring paramagnetic centres.In addition, the increase in the energy of exchange interaction between the unpaired electrons of the neighbouring paramagnetic centres is known to correlate with the increase in the magnetic ordering temperature.47 Unfortunately, it has to be noted that the technique for the synthesis of hexacyanometallates as single crystals of high quality, which is important for a detailed study of magnetic-structural relationships, has not been developed yet. As a rule, conclusions on the structure of compounds are based only on the analysis of powder diffractograms. In addition, all Prussian blue analogues with a cubic symmetry of the crystal structure do not manifest magnetic anisotropy.48 These facts stimulated the synthesis of other cyanide complexes.Of the studies in this field, it is necessary to note the synthesis of the first paramagnetic tetrahedral cyano- metallate [Ph3P=N=PPh3]2[MnII(CN)4] the structure of which has been established completely.49 Based on this compound, MnII[MnII(CN)4] having a interpenetrating diamond-like struc- ture was obtained.50 By itself, MnII[MnII(CN)4] is ordered anti- ferromagnetically, but the authors have a good reason to believe that representatives of the future family M0[MnII(CN)4] (M06aMnII), unlike MnII[MnII(CN)4], should have uncompen- sated spins and should behave as ferrimagnets below Tc. Single crystals of Mn2(H2O)5[Mo(CN)7] . 4H2O (a-phase) and Mn2(H2O)5[Mo(CN)7] .4.75H2O (b-phase) were obtained by slow mutual diffusion of deoxygenated solutions of K4[MoIII(CN)7] . 2H2O and Mn(NO3)2 . 6H2O in a nitrogen atmosphere.48 Both phases are monoclinic and have the same temperature of magnetic phase transition into the ferromagnetic state (51 K). The phases are unstable in air, but in special experi- ments it was possible to detect strong magnetic anisotropy on single crystals of the a-phase. Thus, the transition from Prussian blue analogues with cubic structure to less symmetrical compounds has expanded the range of magnetically active polymeric cyanometallates with strongMolecular ferromagnets anisotropy of magnetic properties. Compounds of this type are still rather rare, but their number increases gradually.A two- dimensional heptacyanomolybdate-based polymer, viz., KMn1.5(H2O)3[Mo(CN)7] . 3H2O grown as single crystals,51 and octacyanometallates [M(CN)8]3,47, where M=NbIV, MoV or WV, were reported at the Sixth Conference on molecular ferro- magnets. The latter also undergo magnetic phase transition at rather high temperatures, viz., at *40 K for Cu3[W(CN)8]2 . . 3.4H2O and at*50 K for Ni3[W(CN)8]2 . nH2O.52 A problem in the design of molecular ferromagnets based on cyanometallates is that the cyano group, which is a bridging ligand, cannot be functionalised. As a matter of fact, this limits the possibilities of synthetic organic chemistry in the design of molecular magnets to the variation of the metal ions.The situation can be improved if some of the metal ions have previously been `dressed in a coordination coat' in such a way as to reserve coordination sites for the approach of bridging cyanide groups. Certainly, the number of bonds between such metal ions and the bridging cyanide ligands is decreased, but the coordination environment acquires the possibility to affect the electronic structure of M0 and the steric structure (and hence the magnetic characteristics) of the compound as a whole. Scheme 1 shows the simplest versions of the formation of three- or two-dimensional polymeric structures withMandM0 ions alternating along differ- ent directions, viz., the framework motif (a),53 the square (b)54, 55 and the honeycomb (c) lattices.56 a M0 M0 N N C C N C N M M0 M0 C C C N N M0 M0 c M M0 M0 M M M0 M0 M M M0 M0 MM0 M M0 M M0 is a chelate matrix with a connectedness of 2; cyanide fragment with a connectedness of 3; In fact, the situation can turn out to be much more complex.In a hybrid version of the design of molecular ferromagnets based on a hexacyanometallate of one metal and a chelate of another,54, 57, 58 the reaction of K3[Fe(CN)6] with trans- [NiCl2(en)2] (where en is ethylenediamine) gave the complex Scheme 1 b M0 N N C C N C N M M0 M0 C C C N N M0 M M0 M0 MM0 M M0 M0 MM0 M M0 M a is is a hexa- hexa- 349 [Ni(en)2]3[Fe(CN)6]2 . 2H2O, whose structure was characterised. It was found that in the solid phase, only three of the six cyanide ligands in the Fe(CN)6 fragments participate in the formation of the bridging bonds with the NiII ions of the Ni(en)2 fragments; 2/3 of the Ni(en)2N2 fragments contain a cis-coordinated NiII ion, while 1/3 contain a trans-coordinated NiII ion.As a result, in a crystal polymeric bands are extended along one of the crystallo- graphic axes, while polymeric layers are formed along the other two directions. The interaction between the unpaired electrons in the NiII7N:C7FeIII fragments, as in the usual hexacyanofer- rate Ni3[Fe(CN)6]2 (see Table 1), is of ferromagnetic character, and at temperatures below 18.6 K the complex [Ni(en)2]3[- Fe(CN)6]2 . 2H2O undergoes a magnetic phase transition into the ferromagnetic state.To date, a rather wide range of molecular ferromagnets of this group with various organic ligands have been studied. These include PPh4[Ni(pn)2][M(CN)6] .H2O, [Ni(pn)2]2[Fe(CN)6]X and [Ni(1,1-dmen)2]2[Fe(CN)6]X, whereM=Fe, Cr and Co, pn is 1,2- diaminopropane,59 1,1-dmen is 1,1-dimethylethylenediamine and X=ClO¡4 , BF¡4 , PF¡6 , CF3SO¡3 , BzO7, I7, N¡3 , NCS7,NO¡3 (see Ref. 60) and [Ni(tren)]3[Fe(CN)6]2 . 6H2O, where tren is the tetradentate tris(2-aminoethyl)amine.61 Using a classical macro- cyclic ligand, 1,4,8,11-tetraazacyclotetradecane (cyclam), the fol- lowing [Ni(cyclam)]3[Cr(CN)6]2 . 5H2O,56 compounds: [Ni(cyclam)]3[Fe(CN)6]2 . 6H2O and [Fe(cyclam)][Fe(CN)6]2 . . 6H2O, which are polymeric in the solid phase, have been synthesised and studied.62 The efficiency of complexes of trivalent manganese with various polydentate Schiff's bases in the forma- tion of heterometallic complexes with hexacyanometallates [Mn(CN)6]37 and [Fe(CN)6]37 has been shown.63 ± 67 The critical temperatures range from 2 to 16.2 K for the group of compounds studied.The [Ni(bpm)2]3[Fe(CN)6]2 . 7H2O complex, where bpm is bis(1-pyrazolyl)methane, with a rather high critical temperature has been synthesised and studied.68 The solid phase of this compound is formed from pentanuclear heterometallic complexes [Ni(bpm)2]3[Fe(CN)6]2 linked into a diamond-like lattice by a system of hydrogen bonds involving molecules of crystallisation water.Although water molecules are generally poor conductors of exchange interactions,69 ± 72 high Tc (23 K) has been found for the [Ni(bpm)2]3[Fe(CN)6]2 . 7H2O phase. Dehydration of the com- pound results in the destruction of the structure and, as a consequence, in the loss of the magnetic ability. This situation is rather typical, i.e., the destruction of a system of hydrogen bonds forming a two- or three-dimensional structure of a compound results in the loss of the ability to undergo magnetic phase transition into the ferro- or ferrimagnetic state.73, 74 However, it should not be considered that disintegration of a highly dimen- sional structure into separate polymetallic exchange clusters always results in zero-dimensional magnetic systems.In recent years, studies in a specific area of the design of molecular ferromagnets, namely, large exchange clusters, is being developed intensely.14, 75 ± 78 Compounds of this type are nano- scale polynuclear molecular particles with a high degree of aggregation, such as Mn12O12(CH3COO)16(H2O)4 . 2HOAc, Co8O4(PhCOO)12(DMF)4, NBu4[Mn8O6Cl6 . (PhCOO)7(H2O)2] and Fe11O6(OH)6(PhCOO)15. Large polynuclear complexes arouse interest from the biochemical viewpoint as models of active polymetallic centres in enzymes.79 On the other hand, despite the molecular nature of these compounds, the same effects as for the classical bulk magnets, including the original hysteresis loops that appear due to tunnelling effects between multiple spin conditions of different multiplicities, have been observed for them.One can find more details on the magnetic properties of these systems elsewhere, e.g., in Refs 80 ± 88. Almost all of them contain polynuclear complexes with carboxylate, oxo and/or hydroxo bridging ligands. A new possibility in the design of heterometallic polynuclear molecular complexes with the use of hexacyanometallates dis- cussed above was shown.24, 89 The reaction of K3[Cr(CN)6] with350 [NiL(H2O)](ClO4)2, where L is tetraethylenepentamine, which blocks five coordination sites in the coordination sphere of NiII, resulted in {Cr[(CN)NiL]6}(ClO4)9; its solid phase is formed by heptanuclear molecular exchange clusters CrIIINiII 6 with ferro- magnetic exchange interactions between the CrIII and NiII ions.N N N N Ni N N N NN Ni N N N C N N C N N N N Ni C Ni Cr N C N N N N N C C N N N N N N N Ni N N Ni N N N One cannot rule out that new dendritic molecular magnets will be obtained from polynuclear compounds of this type.90, 91 The examples presented above show that the systematic development of methods for the synthesis of polymeric cyanide complexes with paramagnetic metal ions has made it possible to overcome, to a considerable degree, the above-mentioned `draw- backs' inherent in the nature of Prussian blue analogues and, in addition, to contribute significantly to the supramolecular chem- istry of metal cyanide complexes.33, 92 ± 100 IV.M3M0 systems 1. Oxalates The oxalate complexes of metals constitute the most representa- tive group of molecular ferromagnets of the M3M0 class.101 ± 127 Okawa et al.101 ± 103 were the first to pay special attention to polymeric heterometallic oxalates.They chose these compounds taking into account the available data on polynuclear complexes with ferromagnetic exchange interactions between the unpaired electrons of the paramagnetic centres located on orbitals of different symmetries.37, 104 ± 108 It was thus desirable to obtain two- or three-dimensional heterospin structures `pierced' by ferromagnetic pair exchange interactions (Fig. 1). It is absolutely unnecessary for a magnetic phase transition that all intra- and intermolecular exchange interactions be exclusively ferromag- netic; however, it should be noted that if only ferromagnetic exchange channels are present in the structure of a compound, higher values of magnetisation can be reached than in the case of b a 6 4 40 3 7 5 30 8 2 9 1 2 0 10 0 10 Figure 1.Scheme of the formation of a layered structure (a) and a fragment of framework structure with identical chiral centres (b); (a) `honeycomb' motif; the light and dark circles indicate the alternating chiral (L and D) fragments of M(ox)3 (ox is oxalate); the squares indicate the cavities that can be occupied by spacers (lacking counterions or solvent molecules);102 (b) `fused decagons' motif.111, 117 Table 2. Critical temperatures for oxalate complexes. Compound Tc /K Structure dimensionality 6667 10 12 14 7728 43 49 2786 777777777 2D 2D 2D 2D 2D 2D 2D 2D 2D 2D 2D 2D 2D 3D 3D 3D 3D 3D 3D 3D 3D 3D 3D 3D 3D <3.5 3D 2D NBu4[MnIICrIII(ox)3] NPr4[MnIICrIII(ox)3] PPh4[MnIICrIII(ox)3] NBu4[CuIICrIII(ox)3] NBu4[CoIICrIII(ox)3] NBu4[FeIICrIII(ox)3] NBu4[NiIICrIII(ox)3] NBu4[CrIICrIII(ox)3] PPh4[CrIICrIII(ox)3] NBu4[NiIIFeIII(ox)3] NBu4[FeIIFeIII(ox)3] NBu4[MnIIFeIII(ox)3] N(C5H11)4[MnIIFeIII(ox)3] [CoIII(bipy)3][CoIICoII(ox)3]ClO4 [FeII(bipy)3][CoIICoII(ox)3] [NiII(bipy)3][CoIICoII(ox)3] [FeII(bipy)3][FeIIFeII(ox)3] [CrIII(bipy)3][NaICrIII(ox)3]ClO4 [CrIII(bipy)3][MnIIMnII(ox)3]ClO4 [CrIII(bipy)3][MnIIMnII(ox)3]BF4 [CoIII(bipy)3][NaICrIII(ox)3]PF6 [NiII(phen)3][NaICoIII(dto)3] .C3H6O 7 [NiII(bipy)3][[MnIIMnII(ox)3] [FeII(bipy)3][NaIFeIII(ox)3] [FeII(bipy)3][LiICrIII(ox)3] [FeII(bipy)3][LiIFeIII(ox)3] NBu4[FeIIRuIII(ox)3] 13 Note.The following abbreviations are used: bipy is bipyridyl, phen is phenanthroline, and dto is dithiooxalate. antiferromagnetic interactions between unpaired electrons. The directed design of heterospin systems has made it possible to obtain 101±103 solid of NBu4[MCr(ox)3] (M=CuII, MnII, FeII, CoII, NiII) with magnetic ordering at 6 ± 14 K (Table 2). A three-dimensional framework structure was originally postulated for NBu4[MCr(ox)3], although it was shown later109 ± 113 that these solids have layered honeycomb structures.One has to note the studies of NBu4[MnCr(ox)3] single crys- tal,109, 110 where strong anisotropy of the magnetic properties has been revealed. If measurements were made along the crystallo- graphic axis c, which is perpendicular to the polymeric layers and coincides with the direction of the magnetic moments of MnII and CrIII, the dependence of magnetisation on the magnetic field strength M(H) was saturated in fields of*0.05 T. If the external magnetic field was oriented perpendicularly, i.e., inside the nets of the polymeric layers, the M(H) curve was far from saturation in the field of 104 G=1 T. Later, a study of magnetic-structural relationships for NPrn4 [MnIICrIII(ox)3] and NBun4 [MnIIFeIII(ox)3] single crystals and a study of the magnetic structure in a neutron diffraction experiment with [P(C6D5)4][MnIICrIII(ox)3] 114 powder confirmed that the magnetic properties of this type of subjects have strong anisotropy.From the conceptual viewpoint, a series of studies 111, 117, 118 into the role of templates promoting the formation of either two- dimensional layered polymeric structures with a honeycomb motif (see Fig. 1 a) or three-dimensional frameworks with a motif of fused decagons (see Fig. 1 b) is important. The following princi- ples for the construction of two- and three-dimensional polymeric oxalate structures were revealed: assembling a structure from various alternating distorted octahedral chiral (L and D) M(ox)3 V I Ovcharenko, R Z Sagdeev Ref.102, 109 114 112 101 102 102 102 116 116 115 115 123 126 122 122 122 111 120 120 120 120 120 118 118 118 118 127Molecular ferromagnets fragments results in polymeric layers with metal ions arranged in one plane, whereas assembling a polymeric structure from the same fragments of the same chirality gives a polymeric framework structure. As a whole, the two-dimensional honeycomb structure is achiral, and the distance between the layers is determined by the size of the cations located in the space between the layers. The three-dimensional structure formed from fused decagons is always chiral. It was found that the [XR4]+ cations, where X=N or P and n n n R=Ph, Pr, Bu or C5H11, favour the separation of two-dimen- sional honeycomb-like layered structures containing [MIIMIII(ox)3]n¡ n , where MII=V, Cr, Mn, Fe, Co, Ni, Cu, Zn; MIII=V, Cr, Fe, into the solid phase, while tris-chelated metals cations, e.g., [M(bipy)3]2+/3+ (bipy is bipyridyl), cause the cross- linking of the polymeric anionic oxalate framework of [MII2(ox)3]2n¡, [MIMIII(ox)3]2n¡ or [MIIMIII(ox)3]n¡ into a three- dimensional framework of decagons.117, 120 Another motif found in the solid phase of lanthanide-containing oxalates [LnIII- CrIII(ox)3] .nH2O is composed of polymeric chains the shape of which resembles `rope ladders'.119 It was also found that frame- work structures similar to oxalates can also be obtained with the use of dithiooxalates (dto, see Table 2) as the bridging ligands.120 The progress of studies into the design of layered and frame- work molecular ferromagnets based on oxalate complexes has also stimulated interest in these compounds as a new type of microporous materials suitable for the separation of racemic mixtures and for directed asymmetric synthesis, as well as for catalysis.It has been shown for the layered polymeric micro- porous compound Na2(NMe3Ph)5[Cr(ox)3]2Cl . 5H2O that partial removal of the template from the compound by cation exchange at room temperature and even its complete removal by thermal treatment of the specimen accompanied by removal of water of crystallisation did not result in the loss of crystallinity.121 This behaviour is similar to that of zeolite materials, therefore the phases under consideration can actually prove to be efficient for the separation of chiral subjects.Compounds with high negative magnetisation of approxi- x MnII 1¡x)1.5. mately the same magnitude as that of solid solutions of classical ferrospinels and the cyanide phases (NiII .[CrIII(CN)6] . 8H2O discussed above, were described in Refs 124 and 125. The authors studied the specially selected series of layered polymeric oxalate complexes AI[MIIFeIII(ox)3], where AI=NPrn4 , NBun4 , N(n-C5H11)4, PBun4 , PPh4, NBun3 (C6H5CH2), Ph3PNPPh3, AsPh4; MII=Mn, Fe, in which the interlayer distance can be adjusted by varying the cation AI. Compounds with MII=Fe behave as ferrimagnets with the ordering temper- ature of 33 ± 48 K.For five of these compounds, transition from positive to negative magnetisation has been recorded at*30 K in weak magnetic fields (*100 G). Such transitions are most char- acteristic of compounds with the highest critical temperatures. Moreover, characteristic changes were found in the series N(n- CnH2n+1)4[FeIIFeIII(ox)3], where n=3 ± 5. It was established that Tc increases with the interlayer distance; at low temperatures, magnetisation changes from positive for n=3 to negative for n=4 and 5.125 Single crystals of the layered polymeric heterometallic oxalate N(n-C5H11)4[MnIIFeIII(ox)3] could be grown and the anisotropy of its magnetic properties was studied.109, 110, 114, 125 Studies into the anisotropy of the magnetic properties of single-crystal molec- ular ferromagnets are still rather rare, therefore they always arouse vivid interest.It was found 126 that N(n-C5H11)4. .[MnII- FeIII(ox)3] with Tc=27 K, as NBu4[MnCr(ox)3] discussed above, has strong magnetisation anisotropy. The uncompensated mag- netic moment directed perpendicularly to the plane of the poly- meric layers (i.e., in the same direction as in NBu4[MnCr(ox)3]) is rather small, viz., 8.781075 mB per magnetically active atom, which can be explained by the small angle of rotation of the 351 magnetic sublattices (*0.001 8) and the small difference between the g-factors of the d5 metal ions.126 Table 2 lists the characteristics of a large group of polymeric metal oxalates, which allows some generalisations to be made.However, the overall picture for the understanding of the mag- netic properties of compounds belonging to this class is still far from completion.127, 128 Nevertheless, methods for directed syn- thesis of layered polymeric or framework polymeric metal oxa- lates have been developed to date, and it has become clear what the resulting chirality of the subjects would be and what spatial structure they would have. In addition, it is possible to adjust the distances between the layers in a series of layered compounds by varying the size of the cation AI located in the interlayer space.74, 121 Experiments show that the highest magnetic ordering temper- atures (*40 ± 50 K) are reached if a compound contains a combination of metal ions having high spins (high-spin MnII, FeII, FeIII), in accordance with the correlation Tc!JS1S2 (where S1 and S2 are the spins of the first and second paramagnetic centres, respectively). When metal ions are combined, it is possible to predict the temperature range in which the critical temperature of the expected compound will fall. Thus, the introduction of diamagnetic ions (for example, lithium or sodium ions) in polymeric oxalate systems leads to an increase in the distances between the paramagnetic centres, a reduction of the energy of exchange interaction between them and, hence, a sharp decrease in the critical temperature. On the other hand, the fact that the critical temperatures of the known framework oxalates (3D structures) are much lower than those of layered polymeric compounds (2D structures) requires a quantum-chemical explan- ation. The role of relativistic effects is not quite clear; the necessity of taking them into account in the description of the magnetic properties of heterometallic polymeric oxalates containing para- magnetic 4d- or 5d-metal ions was pointed out in Ref.127. Such properties appear due to a significant spin-orbital interaction in these elements. It may seem that the design of oxalate systems is limited only to the replacement of metal ions and template cations. This is hardly true. Certainly, oxalate polymers are the basic systems for the synthesis of the M3M0 class ferromagnets. On the other hand, current studies encourage a wider use of various diamagnetic and paramagnetic ligands which are structural analogues of the oxalate anion and hence favour the development of the synthesis of new molecular magnets.For example, the aforementioned dithiooxalate,120, 129, 130 imidazoles,131 imidazoleacetate 132 and biimidazole 133 may be regarded as such ligands. 7 7 N N S O M M M M N N S O 7 7 In addition, investigations into the design of polynuclear mixed-ligand compounds, in which not all but only one or two ligands are oxalates,134, 135 are certainly of interest. For example, the structure of complexes MII 2 (dpm)(ox)2 . xH2O, where M=Mn, Cu; x=6, 5, respectively, and dpm is bipyrimidine (Fig. 2),135, 136 is topologically identical with the honeycomb structure of layered polymeric oxalates (see Fig.1 a). The way of self-organisation of three-dimensional frameworks upon forma- tion of mixed-ligand complexes with an oxalate anion and bipyridyl has been studied.137 Thus, addressing mixed-ligand complexes is also promising in the design of new molecular ferromagnets.352 N N M N N M M N M N N N N N M M MM N N N N M N N M M N M N N N N M N M M N M N N N M N N M M N M N N N N N M M N N Figure 2. Scheme of a layered polymeric structure (`honeycomb' motif) ofMII 2 (dpm)(ox)2 . xH2O135, 136 and MII 2 (dpm)(N3)4 .1387140 The double lines indicate the bridging oxalate or azide anions connecting the metal ions into chains, which are, in turn, linked by the bridging bipyrimidine ligands into layers.2. Azides The above approach to the synthesis of layered polymeric com- pounds based on mixed-ligand complexes of metal oxalates with bipyrimidine, but with the use of azide ligands instead of oxalate ones, has been developed further.138 ± 140 The authors succeeded in synthesising and studying a series of layered polymeric complexes with compositionM2(N3)4(dpm), whereM=CoII, FeII andMnII. The structural motif of polymeric honeycomb-like layers in M2(N3)4(dpm) is exactly the same as that inMII 2 (dpm)(ox)2 . xH2O (see Fig. 2) with only one difference: the role of a bridge is played by two bridging m-(1,1)-azide ligands instead of the oxalate anion, i.e., according to the systematics accepted, the M2(N3)4(dpm) complexes correspond to the M3(1)M0 class.The authors believed that the introduction of the azide anion into polymeric structures would be expedient for obtaining polymers with exchange inter- actions alternating in sign (ferromagnetic7antiferromagnetic). As a rule, strong antiferromagnetic exchange interaction (7200 to 7400 cm71) predominates in exchange clusters linked by a bridging oxalate anion or bipyrimidine.13, 141, 142 However, it is known that in binuclear exchange clusters linked by azide follow- ing the m-(1,3) pattern, the exchange interaction is antiferromag- netic,143 while in those linked following the m-(1,1) pattern, it is ferromagnetic.144, 145 ± 162 M N N N M M N N N N N N N N N M m-(1,1) m-(1,3) Experimental studies of the magnetic properties of the M2(N3)4(dpm) complexes have shown that competition between ferro- and antiferromagnetic exchange interaction channels does exist in them, and at low temperatures (<12 K) these compounds behave as metamagnets with high critical magnetic fields (>2 T).138 It should be noted that the ferromagnetic character of exchange in m-(1,1) type binuclear azide exchange clusters was confirmed by quantum-chemical studies of the [M(m-1,1-N3)2M] and [M(m-1,1-N3)3M] fragments, whereM=CuII, NiII, MnII (see Ref.146). For example, for the copper complex, the results of V I Ovcharenko, R Z Sagdeev calculation for the distribution of spin density qualitatively agree with the experimental results of a polarised neutron diffraction study.The spin density distribution obtained in the calculated model is a superposition of the electron delocalisation and spin polarisation effects: delocalisation of unpaired electrons from the dx2¡y2 orbitals to the donor atom of the ligand decreases in the series Cu>Ni>Mn, while spin polarisation in the same series increases because of the increase in the number of unpaired electrons.146 It was also found that the exchange is hardly affected by deviations of the uncoordinated azide ligand from the plane of the M(m-N)(m-N)M fragment, which opens the possibilities for a more flexible and versatile design of highly dimensional structures with the m-(1,1) bridging azide. In the synthesis of a layered polymer Ni(N3)2L, where L is 2,2- dimethylpropane-1,3-diamine,163, 164 advantage was taken of the diversity of coordination capabilities of the N¡3 ligand known for polynuclear complexes with azide ligands.147 ± 157 The bidentate ligand was shown not to participate in the formation of bridging bonds.The layered polymeric structure is formed by only the azide anions, some of them serving as m-(1,1) bridges and others as m-(1,3) bridges. The magnetic ordering temperature for Ni(N3)2L was found to be rather high (Tc=55 K), although the residual magnetization at 4.2 Kis very low (661073 mB). Below the critical temperature, this compound behaves as a weak ferromagnet, but the reason why the weak ferromagnetism appeared (antisymmet- ric exchange 165 ± 168 or g-factor anisotropy of the nickel ion 169) remains obscure.Similar magnetic behaviour has been recorded for a series of layered polymeric coordination compounds Mn(N3)2L2, where L stands for monodentate ligands, e.g., pyridine, 3-acetylpyridine, 4-acetylpyridine, 4-cyanopyridine or ethylisonicotine.170 ± 172 In polymers with exclusively m-(1,3) bridging function of the azide ligand, antiferromagnetic exchange interactions between the manganese ions occur, whereas in polymers with alternating m-(1,1) and m-(1,3) bridging azides, alternation of ferromagnetic and antiferromagnetic channels of exchange interaction is observed. At low temperatures (<16 ± 40 K) the ESR signal from polycrystalline specimens broadens and compounds behave as weak ferromagnets with small magnetic moments.172 3.Dicyanamides The first reported molecular ferromagnets 173 ± 178 were based on polymeric dicyanamide derivatives of transition metals (ML2)?, where L=N(CN)¡2 ; M=Co, Ni. In solid (ML2)?, each ligand has tridentate bridging coordination (Fig. 3 a). Each N(CN)¡2 bridging ligand is bound to metal ions through two terminal nitrogen atoms of the nitrile groups and the central amide nitrogen atom. In turn, each metal ion is bound to six different N(CN)¡2 ligands, four of which are coordinated by the nitrile nitrogen atoms and two are coordinated by the amide nitrogen atoms. This results in a framework structure similar to that of rutile. These compounds undergo magnetic phase tran- sition into the ferromagnetic state at 9 (Co) and 20 K (Ni).An important result of these studies is that no ordering effects were detected down to 2 K for the isostructural compound (CuL2)? characterised by the pronounced Jahn7Teller effect, which is inherent in CuII. (CuL2)? behaves as a usual paramagnet over the entire measurement range. In our opinion, this behaviour is due to a considerable elongation of the bonds between the copper ions and the amide nitrogen atoms in the structure of (CuL2)?,173 i.e. the magnetic structure consists of the chains [7Cu7(N:C7N7C:N)27Cu7(N:C7N7C:N)27]? with CuII ions that weakly interact through five bridging atoms (Fig. 3). As regards the magnetism, (CuL2)?belongs to theM5M0 class rather than toM3M0 as is the case for (CoL2)? and (NiL2)?. Band chains of this type with weakly interacting manganese ions, [7Mn7(N:C7N7C:N)27Mn7(N:C7N7C:N)27]?, have been reported.179 They were detected for solid MnL2(CH3OH)2, which confirms the assumption stated above.Molecular ferromagnetsa b X N N C C N ; = =MII.Figure 3. Scheme of decomposition of a three-dimensional framework structure of (ML2)? (a) into separate chains (b) upon increase in the metal ± amide nitrogen atom distance. Experiments with the use of neutron diffraction on (ML2)? powders (M=Co, Ni, Mn) and measurements of specific heat capacity of these compounds in the region of magnetic phase transition were carried out.176 ± 178 It was possible to determine correctly the Curie temperatures for the ferromagnets, viz., cobalt (8.7 K) and nickel (19.7 K), the Neel temperature for an anti- ferromagnet (MnL2)? (16 K), to determine the magnetic struc- ture of the compounds and to establish that (CoL2)? exists as two polymorphic modifications, one of which (a) can undergo ferro- magnetic ordering below the critical temperature. In the studies of the magnetic properties of (ML2)? (M = Fe, Co, Ni) and alloys [Ni0.5Co0.5L2]?Ni(OH)27xLx , some unusual properties of these compounds were reported.180, 181 First, an increase in the pressure to 17 kbar makes it possible to increase the critical temperature of the phase by *10%; this is an important feature of molecular ferromagnets that rather commonly have a `loose' structure.Second, anomalously high coercive fields are characteristic of the molecular ferromagnets studied: 710, 7975 and 17 800 ê, respec- tively, for the nickel, cobalt and iron complexes.182 ± 184 The latter value is not only the highest in the series of molecular ferromag- nets but also exceeds the coercive fields for Sm7Co and Re7Nd7B alloys (the coercive field for classical magnetically hard materials reaches 103± 104 ê).182 Thus, the examples presented above show that the organic dicyanamide anion can be efficiently used in the design of molecular ferromagnets. 4. Dioxamates An original approach to the design of polymeric heterospin systems based on mixed-ligand complexes with polydentate ligands was developed.185 ± 195 The construction of these systems is a multistage directed design of molecular ferromagnets starting from the synthesis of a specific polyfunctional ligand (various dioxamates) and allowing one to obtain a solid heterospin layered polymer with a strictly determined alternation of paramagnetic centres. In the first stage, a complex of CuII with a tetradentate ligand is formed, and then cross-linking of copper-containing fragments into a chain occurs through the formation of chelate complexes of MnII with the oxygen atoms of the oxamate groups.The MnII ion cannot compete with CuII for the donor nitrogen atoms, as its thermodynamic constants of complex formation with 353 27 X O O N N O O Mn2+ Cu2+ HN NH Cu O O O O O O OH OH X X O O N N N O O N Mn Cu Cu Mn O O O O O O O O X=H, OH.nitrogen-containing ligands are much smaller than those for copper. As a result, a chain with regular alternation of CuII and MnII is formed. Mutual arrangement of the paramagnetic centres belonging to different chains is an essential factor. It was found185 ± 187 that in the case of N,N0-propane-1,3-diyldioxamate (X=H), the CuII ions in solid MnCuL. 5H2O are arranged above CuII ions, while the MnII ions are above MnII ions; the magnetic moments of ions belonging to different chains are oriented oppositely (due to the predominant antiferromagnetic exchange interactions) and com- pensate each other (Fig. 4 a). The magnetic moments in the copper and manganese sublattices generally tend to compensate each other.As a result, the compound is ordered antiferromagnetically at low temperature (Tc=2.2 K). b aCu Cu Cu Cu Cu Cu Mn Mn Mn Mn Mn Mn Cu Cu Cu Cu Cu Mn Mn Mn Figure 4. Antiferromagnetic spin ordering of neighbouring chains with resulting zero (a) and non-zero (b) magnetic moment. In an attempt at changing the mutual arrangement of the metal ions, the authors synthesised the MnCuL0 . 3H2O complex, where L0 is N,N0-(2-hydroxypropane-1,3-diyl)dioxamate (X= OH). In the solid phase of this compound, the chains (...Cu...Mn...)? are shifted relative to each other in such a way that the OH groups of one chain form hydrogen bonds with the water molecules coordinated by the MnII ions of the other chain.As a result, a staggered arrangement of the CuII and MnII ions appears (Fig. 4 b). The magnetic moments of ions from different chains tend to be oriented oppositely. However, in this case each Cu7Mn pair forms a difference magnetic moment, and on the whole, uncompensated magnetic moment of the manganese and copper sublattices appears, as inside these sublattices all spins are oriented in the same way. As the temperature is decreased, the MnCuL0 . 3H2O structure is ordered ferromagnetically (Tc= 4.6 K). A series of molecular ferromagnets with the Curie temper- atures of 5 ± 22 K were obtained using this approach.188 ± 193 The antiferromagnetic character of exchange interaction between the neighbouring CuII and MnII ions in these compounds was confirmed by polarised neutron diffraction.194 A method for the synthesis of the multispin compound R2Mn2(CuL)3(DMSO)2 .2H2O, where R is the 2-imidazoline 1-oxyl 3-oxide-containing radical cation; L is the o-phenylene-354 dioxamate anion, was developed.196, 197 The structure of the compound consists of a system of almost perpendicular (73 8) interlocked networks analogous to those of layered polymeric oxalates (see Fig. 1 a), the only difference being that the vertices of each hexagon are occupied by the MnII ions, while the midpoints are CuII ions. Mn Cu Cu Mn Mn R Cu Cu CuMn R Mn Mn Cu Cu Cu MnCu Mn MnCu Cu Mn The tetradentate ligands L play the role of bridges connecting MnII and CuII.Each of the oxygen atoms of the nitronyl nitroxide is coordinated by CuII ions from different networks, which results in the formation of additional polymeric chains in the crystal; the compound is ordered as a ferrimagnet below 22.5 K. Using this approach, the authors obtained a series of molecular ferromagnets (R-Rad)2M2(CuL)3, where R-Rad is the 2-(4-N-alkylpyridinium)- 4,4,5,5-tetramethyl-2-imidazoline 1-oxyl-3-oxide cation, M = MnII, NiII, CoII, with critical temperatures in the range of 22 ± 37 K.198 V. M4M 0, M5M 0 and M7M 0 systems In the Introduction, we discussed the importance of involvement of synthetic organic chemistry in the design of molecular ferro- magnets. However, in the classes of compounds considered above, it is not always possible to attach additional functional groups to the bridging ligand atoms.Besides, simple elongation of the hydrocarbon chain will result in a sharp decrease in the energy of exchange interactions between the unpaired electrons of the paramagnetic centres. To resolve the situation, one can use stable nitroxides; the methods for their synthesis and modification are well developed (see, e.g., Ref. 11). The bridging heterospin struc- tures are shown below. Depending on the nature of the group X, these systems can contain a minimum of four atoms in the bridge. Fragments of bridging structures representing derivatives of stable 2-imidazoline 199 ± 214 and 3-imidazoline 215 ± 227 nitroxides and acyclic polynitroxyls linked through benzene or heterocyclic rings are also shown.12, 228 ± 231 M X R3 R1 C R2 N R4 O M A M ON R1 N R2 R3 R4 R5 O M D Mn Cu Mn M M R3 X R2 N R4 C R1 R2 R1 N R3 N R5 O O M C B MR3 R4 R2 R6 R1 N N R5 O O E M M V I Ovcharenko, R Z Sagdeev R3 R3 R4 R4 R2 R2 R6 R1 N N N R5 R5 O O O M M M F Substituents Rn in nitroxides may be varied over a broad range.The presence of one or several N7O . groups (preferrably, with the greatest possible delocalisation of spin density over the bridging organic fragment) makes it possible to maintain efficient exchange interactions between the unpaired electrons of the metal ions and the nitroxyl radicals. A group of chain polymers based on mixed-ligand complexes of MnII and NiII hexafluoroacetylacetonates with nitronyl nitro- xides (structure D) and capable of ferrimagnetic ordering at T<8 K has been reported for the first time in Refs 232 ± 235.The solid phases formed by polymeric chains of [Mn(hfacac)2L]?, where hfacac is hexafluoroacetylacetonate and L represents poly- nitroxyls, are characterised by similar ordering.236, 237 Studies of these subjects led to the conclusion that, despite the strong intrachain exchange (the integral of exchange interaction is greater than 250 cm71),234 one cannot hope to obtain ferromag- nets with high Curie temperatures from these compounds due to the very low energies of intermolecular exchange interactions. However, the design of molecular magnets makes it possible to pass from chain structures to structures with higher dimension- ality.In particular, synthesis of a series of triradicals with structures of the F type has been developed. High conformational mobility of acyclic nitroxyl fragments linked by single bond to an aromatic ring enabled the realisation of a framework structure in the complex [Mn(hfacac)2]3L2 (L is a triradical; R2, R4, R5=H; R1, R3, R5=But), which is ordered as a ferrimagnet at 46 K.238 ± 240 Each coordination matrix [Mn(hfacac)2] is bound to two ligands, whereas each ligand is surrounded by three different Mn(hfacac)2 units. Nevertheless, it should not be thought that a framework polymer is always formed at the stoichiometric ratio M(hfacac)2 : - nitroxyl=3 : 2.We found that this composition can also be observed for specific trinuclear molecules with radicals 241 which are formed particularly often in the synthesis of mixed-ligand complexes Cu(hfacac)2 with nitroxides of (Fig. 5).204, 209, 219, 221, 242 Ph N MeO [Cu(hfacac)2]3L2 Cu(hfacac)2+ N MeO O If the compounds have the same composition, one has to pay special attention to the possibility of formation of polymorphic modifications, as their magnetic properties can differ consider- ably.199, 222, 232, 243, 244 OCu N Figure 5. Molecular structure of the complex [Cu(hfacac)2]3L2.Molecular ferromagnets Ph Ph N N O N O M0 M0 O N a-[Mn(hfacac)2L]? M0 M0 O N O N O N Ph N Ph O M0 O Ph N Ph N O N N M0 O O M0 b-[Mn(hfacac)2L]6 Scheme 2 shows that trans-coordination of the bridging 2-imidazoline nitronyl nitroxyl ligands in Mn(hfacac)2 results in the formation of polymeric chains (a) in the solid which undergoes three-dimensional magnetic ordering below 6 K.232 cis-Coordi- nation of radicals results in hexanuclear separate molecules (b) due to closure of a 36-membered macrocycle formed by the manganese atoms and the O7N7C7N7O fragments of the paramagnetic ligand.199 The ground state of hexanuclear molecules is that with S=12 in the temperature range from 5 to 300 K; the b-modification behaves as an ordinary paramagnet.The bischelate of divalent nickel with acetonyl 3-imidazoline nitroxide derivative NiL2 (Fig.6) also exists as two polymorphic modifications (a and b).222 Both modifications have layered polymeric structures. The goffered layered structure of the solid is formed owing to coordination of the nitroxyl oxygen atoms of the neighbouring bischelates by the nickel ion. The coordinated oxygen atoms of the nitroxyl groups are trans-coordinated in the a-modification and cis-coordinated in the b-modification. In solid a-NiL2, the antiferromagnetic exchange interactions between the unpaired electrons of the paramagnetic centres in the 7NiII7O7 .N fragments predominate, and mef N7O . approaches zero as the temperature is lowered. In b-NiL2, the antiferromagnetic (J^7115 cm71) and ferromagnetic (J^10 cm71) exchange compete, which results in antiferromag- netic ordering at Tc=14 K (Fig.7). Taking into account the stereochemical non-rigidity of bis- chelates of the NiL2 type, it is possible, by selecting the crystal- lisation conditions, to obtain modifications having either a molecular structure incapable of cooperative magnetic ordering or a layered polymeric structure that displays a tendency towards cooperative magnetic ordering.243, 245 Perhaps, the only factor that limits the possibilities of the design of molecular ferromagnets from complexes with nitroxyl radicals is that the N7O . group, which results from the oxidation of the corresponding hydroxylamine, is a weak donor of electron density. However, coordination of the nitroxyl group or any other way of its inclusion in the network of polymeric exchange channels are necessary for cooperative magnetic ordering.220 Therefore, Scheme 2 O O N N Ph OM0 O Ph N N O a b ON F3C O N Ni N O CF3 NO cd Figure 6.Fragments of layered polymeric structures of a- and b-NiL2; (a) NiL2 complex; (b) coordination of the Ni atom in b-NiL2; (c) layered polymeric structure of a-NiL2; (d ) layered polymeric structure of b-NiL2. metal hexafluoroacetylacetonates have mostly been used as acceptor matrices in early studies of heterospin systems based on complexes of transition metals with stable nitroxides.7, 11 It was shown246 that fluorinated carboxylates of metals can be used instead of hexafluoroacetylacetonate matrices; heterospin com- pounds capable of magnetic ordering at 21 ± 24 K were obtained from mixed-ligand complexes of manganese bis(pentafluoro- benzoate) with 2-imidazoline nitroxide.It is possible in principle to get rid of the `ballast' diamagnetic anionic ligands, such as hexafluoroacetylacetonate or carboxy- late, if the organic radical is originally constructed in such a way that it can play both the roles of a bridge and an anion. We succeeded in synthesising a vast group of molecular ferromagnets of this type with layered polymeric or framework structure (Table 3). Scheme 3 illustrates the formation of polymeric layers in heterospin bischelates based on complexes of transition metals with derivatives of the stable 3-imidazoline nitroxide.355 Ni ON356 mef /mB 4 2 32 1 10 100 103 wn 80 40 10 0 Figure 7. Temperature dependences of the effective magnetic moment for a- (1) and b-NiL2 (2) (a) and the specific magnetic susceptibility (wn) for b-NiL2 (b). R1 O R2 N NO R1= CF3, COOMe, COOEt; R2=Me, H, Cl, Br;M =Cu, Ni, Co. High-quality single crystals of intracomplex chelates can be obtained from various organic solvents; therefore, it was possible to determine the crystal and molecular structures for all of the heterospin compounds listed in Table 3. A distinctive feature of the composition of the compounds studied is the presence of non- crystallisation molecules of alcohols in the crystal. In the solid phase, the metal ion coordinates the oxygen atoms of the OH groups belonging to two ROH fragments, then the ROH7ML27HOR groups are linked by hydrogen bonds ROH_O7 .N with two neighbouring ROH7ML27HOR units.As a result, the bridging alcohol molecules `cross-link' the a 400 T/ K 300 200 b 2 TN T /K 20 Scheme 3 O H R3 N O R1 N O R2 M R2 O N R1 O N R3 H O ON R1 N O R2 M R2 O N R1 NO Table 3. Magnetic ordering temperatures for complexes with 3-imidazo- line nitroxyl radicals. Compound Tc /K R1=CF3, R2=H 6.5 75.3 6.2 6.0 5.9 4.8 6.3 6.8 4.0 7.1 8.1 4.0 7 NiL2(MeOH)2 NiL2(H2O)2 NiL2(EtOH)2 NiL2(C3H5OH)2 NiL2(PrnOH)2 NiL2(BunOH)2 NiL2(BuiOH)2 NiL2(n-C5H11OH)2 NiL2[HO(CH2)4OH] NiL2[HO(CH2)5OH] CoL2(MeOH)2 CoL2(EtOH)2 CoL2[HO(CH2)4OH] CuL2 R1=CF3, R2=Cl 5.0 7.1 5.0 NiL2(MeOH)2 CoL2(MeOH)2 CuL2 R1=CF3, R2=Me 5.8 NiL2(MeOH)2 R1=COOEt , R2=H 4.2 CuL2 R1=COOMe, R2=H 5.0 CuL2 R1=COOEt , R2=Br 4.3 CuL2 ML2 fragments to form polymeric layers (monohydric alcohols) or a three-dimensional framework (dihydric alcohols), i.e.the two- or three-dimensional structure is formed by hydrogen bonds.247, 248, 250 ± 253 It should be noted that a separate metal ion in ML2(ROH)2 is surrounded by a large number of `organic' atoms (C, H, N, O). For example, there are 90 C, H, N and O atoms altogether per each NiII ion in NiL2(n-C5H11OH)2; in spite of this, the compound is capable of cooperative magnetic ordering at low temperature.Removal of the alcohols and transition to layered polymeric structure in solid ML2 leads to the loss of its ability to undergo ferromagnetic ordering.222, 248 At first glance, this seems para- doxical, as the transition from ML2 to ML2(ROH)2 is accompa- nied by an abrupt increase in the fraction of the organic component in the compound, and hence, a decrease in the number of paramagnetic centres per unit volume of the solid phase of the complex. Nevertheless, it is such changes that determine the ability to undergo magnetic phase transition typical of weak ferrimag- nets.247 It was found by comparison of the magnetic properties of the isostructural complexes NiL2(ROH)2 and NiL2(H2O)2 (see Ref. 72) that the introduction of water molecules in the exchange N7O .channel _HOH7MII7HOH_O7 .N results in a sharp decrease in the exchange interaction energy and, hence, the loss of ability to undergo magnetic phase transition, despite the structural identity of the exchange channels in NiL2(H2O)2 and NiL2(ROH)2. The high kinetic stability of complexes with alcohols, especially dihydric ones, was noted. For example, no changes in the composition and properties of NiL2(MeOH)2 and V I Ovcharenko, R Z Sagdeev Ref. Structure dimensionality 247, 250 72 247, 250 248 248 249 249 248 247 247 251 11, 248 249 254, 255 2D 2D 2D 2D 2D 2D 2D 2D 3D 3D 2D 2D 3D molecular 252 252 253 2D 2D 2D 252 2D 253 2D 253 2D 253 2DMolecular ferromagnets NiL2[HO(CH2)4OH] were recorded upon their storage under normal conditions for 10 years.Large single crystals (with sizes from several millimetres to two centimetres) of many compounds of this series are also storage-stable. The possibility of obtaining complexes of transition metals with nitroxyl radicals in the form of sufficiently large single crystals makes them convenient subjects for studying the anisotropy of the magnetic proper- ties 233 ± 235, 238, 245, 247 and the spin density distribution using polarised neutron diffraction.256 ± 259 For example, Fig. 8 shows the results of a study of the magnetisation anisotropy for NiL2(MeOH)2 and NiL2(EtOH)2 and demonstrates the relation between the direction of easy magnetisation axis for single-crystal NiL2(EtOH)2 as a macrosubject and the direction of the second- order crystallographic axis of the crystal microstructure.The coincidence of these directions is a prerequisite (within the frame- work of the classical magnetism concept) for single crystals belonging to the space group P21/c.168 It was established in systematic studies of transition metal complexes with stable 3-imidazoline nitroxides that the structural dimensionality of a solid can be controlled by varying the substituents R1 and R2 in the side chain of the ligand. If R1 is not an electron density acceptor (alkyl, aryl), the solid phase of the complex has a molecular structure with a square or tetrahedral environment of the metal ion.If R1 is a strong electron density acceptor (for example, CF3, COOEt), the central metal ion a 0y b 1073M /G cm3 mol71 y 0.8 0.6 x 0.4 0.2 z 0 4 2 6 1073 H /G Figure 8. Scheme demonstrating the coincidence of the directions of the twofold symmetry axis (arrow inside the cell) and the direction of the easy magnetisation axis in the single crystal (arrow inside the crystal) for NiL2(EtOH)2 (a) and dependences of m(H) showing the magnetisation anisotropy of single crystals of NiL2(MeOH)2 (b) and NiL2(EtOH)2 (c).247 OCFNNi z c yxz 4 2 6 1073 H /G 357 coordinating the donor atoms of the neighbouring molecules also becomes a strong acceptor. The coordination number of the central ion is completed to 6 and a layered polymeric structure is formed.72, 260 In the case where the acceptor ability of a metal ion attained only due to the introduction of an electrophilic substitu- ent R1 is insufficient to implement a two-dimensional solid state structure, the introduction of additional acceptor groups, for example, halogen atoms, as the substituent R2 is effective.253 In the examples of heterospin complexes discussed above, the 3-imidazoline nitroxyl radical plays the role of a tridentate bridge.Recently, heterospin layered polymers with deprotonated nitronyl nitroxyl radicals (NIT) containing an imidazole (Im) or benzoi- midazole (BzIm) substituent at position 2 of the paramagnetic heterocycle were described.261, 262 Mn Mn 7O 7O N7 N7 N+C N+C N N N N O O Mn Mn Such substituents perform the tetradentate bridging function (M53M0 systems).Complexes with composition [Mn2(NITIm)3]. .ClO4, [Mn2(NITBzIm)3]ClO4 and [Mn2(NITIm)3]BF4 ordered as weak ferromagnets at 1.4, 40 and 3.6 K, respectively, deserve special attention.262 The layered polymeric structure of these complexes with honeycomb motif is the same as that of the layered polymeric oxalate (see Fig. 1 a) or azide complexes (see Fig. 2). This fact agrees completely with the general requirement for the arrange- ment of a layered polymeric structure with a metal : bridging radical ratio of 3 : 2 with alternating chiral (L and D) fragments formed by the metal ions. This indicates that molecular ferromag- nets of this structural type have a common architecture.A systematic study of magnetic-structural relationships only for compounds of this structural type will make it possible to obtain a series of molecular ferromagnets with different magnetic char- acteristics (critical temperatures, spontaneous magnetisation, coercive fields). Like for oxalate molecular ferromagnets,125, 263 the critical temperature of heterospin layered polymers was found to depend on the nature of uncoordinated anions (ClO¡4 , BF¡4 );261, 262 vary- ing the latter serves as an additional means for affecting the magnetic parameters of the solid phase. In addition, the interlayer distances in the imidazole complex [Mn2(NITIm)3]ClO4 increase as the temperature is decreased, which decreases the critical temperature of this complex in comparison with the benzoimida- zole complex [Mn2(NITBzIm)3]ClO4.VI. The problem of organic ferromagnets The general problem of the creation of molecular ferromagnets was intensely fed 15 ± 20 years ago by theoretical studies 264 ± 269 and experiments 270 ± 275 aimed at the synthesis of the so-called high-temperature, purely organic ferromagnets, i.e., compounds the solid phase of which has no metal-containing components at all. Experimental results on the synthesis of purely organic ferromagnets have not been confirmed in many aspects and are subject to criticism.276 ± 281 In the review 1 where the results of studies on organic ferromagnets were generalised and analysed in detail, it was noted that a most complex problem in the creation of such compounds is that the intermolecular exchange interactions in solid organic paramagnets are rather weak because of the large distances between the paramagnetic centres.In fact, the Curie temperatures for the purely organic ferromagnets known to date do not exceed 1 ± 2 K.282 ± 303358 7 O O N O + N+C N N N O7 O O N N Cl S N N Perhaps, the only exception is the value of 16.1 Krecorded for the adduct of C60 fullerene with tetrakis(dimethylamino)ethylene (TDAE) with composition C60(TDAE)0.86 (see Ref. 304). How- ever, the nature of the ferromagnetism of C60(TDAE)0.86 remains unclear, although there was no doubt that the adduct did not contain ferromagnetic admixtures.A series of molecular organometallic ferromagnets based on salts of sandwich radical cations of decamethylferrocenium type with radical anions of such p-acceptors as tetracyanoethylene or tetracyanoquinodimethane have been synthesised and studied.3 ± 6, 305 ± 316 N N R 7 R + + C C N or M N N M C C N N N R R The solid phases of these compounds involve molecular packing with alternating paramagnetic donors and acceptors (...D+.A7....D+.A7....). The reasons for the bulk magnetic ordering in them became the subject of extensive theoretical studies.1, 6, 317 ± 319 The compounds have critical temperatures from 2.3 to 8.8 K and are stable only in an inert atmosphere. A similar series of donor-acceptor systems capable of mag- netic ordering at 3 ± 13 K was obtained from derivatives of porphyrin or phthalocyanine macrocyclic complexes with organic radical anions.320 ± 325 Judging by the temperature of bulk mag- netic ordering, compounds of metals with the tetracyanoethylene radical anion (TCNE) with composition M(TCNE)2(CH2Cl2)y [M=MnII (Tc=75 K), FeII (Tc=75 K)] 326, 327 and especially the vanadium compound V(TCNE)x(CH2Cl2)y, where x^2 and y*0.5, are prominent among these subjects.The latter com- pound behaves as a ferromagnet up to the temperature of thermal decomposition (>350 K).6, 328, 329 Unfortunately, the compound is extremely unstable in contact with air, and its structure could not be elucidated. Nevertheless, it is substantiated that a phase which is satu- rated with an organic component and contains solvate molecules of dichloromethane (tetrahydrofuran or acetonitrile) 329 is capable of magnetic ordering at high temperatures.This means that, in principle, it is not forbidden to obtain purely organic ferromag- nets with high critical temperatures in which the exchange interactions between the paramagnetic centres are sufficiently strong. Recently, we found in a study of the first stable vinyl- nitroxyl (L) 330 CN O Ph N L O that antiferromagnetic exchange chains with an exchange energy of J/k^100 K are formed in this solid. Due to such a high antiferromagnetic exchange energy, the magnetic moment of the V I Ovcharenko, R Z Sagdeev radical L does not reach the theoretical value (1.73 mB) even at room temperature and equals 1.49 mB, i.e.high exchange inter- action energies can also occur in compounds with a molecular structure. On the other hand, as already noted, mutual orientation of magnetic orbitals in the solid phase of L is such that the exchange interactions between the paramagnetic centres are anti- ferromagnetic, and the magnetic moment approaches zero as the temperature is lowered. However, examples of organic radicals are known in the solid phases of which the intermolecular exchange interactions are ferromagnetic 297, 298, 331 ± 333 and even ferromagnetic chains are formed.334 Besides, antiferromagnetic exchange is not an insurmountable obstacle to cooperative order- ing, because a ferrimagnet can be obtained 267 by using heterospin bi- or polyradicals335 or by cocrystallisation of mono- and bi- or polyradicals.267 It is possible that a purely organic ferromagnet, maybe even one with a molecular structure and with a sufficiently high critical temperature (exceeding the boiling point of liquid helium) will be synthesised.VII. Conclusion This review analyses the main classes of chemical compounds for which the basic methods for designing high-dimensionality multi- spin systems which tend to cooperative magnetic ordering have been discussed. The development of any method for designing molecular ferromagnets is undoubtedly a laborious and lengthy process. About ten years ago, the determination of the first approaches to the design of various classes of molecular ferro- magnets had only started.To date, vast and valuable experimental material has been accumulated: critical temperatures above room temperature (315 ± 330 K) have already been achieved for molec- ular ferromagnets;336, 337 compounds with anomalously high coercive fields have been synthesised;180, 181 a number of com- pounds have been obtained as single crystals stable under ordinary conditions.8, 247 The prospects of the practical application of molecular ferromagnets in various electronic and magnetic devi- ces as well as for data recording, magnetic visualisation and protection against low-frequency magnetic fields are under dis- cussion.3, 6, 24 An original photomagnetic device with a cyanide molecular ferromagnet as a working element was shown at the XXXIIIrd International Conference on Coordination Chemis- try.35 The prospects of progress in the area of molecular ferromag- nets are difficult to predict. However, it is doubtless that these compounds are promising, as they represent light, transparent and mainly dielectric materials obtained under mild conditions, with specific properties commonly not typical of the classical magnetic materials.338 We did not dwell upon details of analysis of the magnetic properties of molecular ferromagnets, as this has been done exhaustively in several monographs,13, 339 ± 342 and a detailed discussion of the diversity of magnetic properties of molecular ferromagnets might be an independent topic of a separate review.Let us just note that the authors often introduce a large number of exchange parameters (J1, J2, J3 etc.) in the spin-Hamiltonian when analysing the experimental temperature dependences w(T ) or m(T ) for complex multispin systems with various channels of exchange interactions. If this approach is used, it should be taken into account that the introduction of additional parameters improves the fitting of the experimental curves, but the errors of the model in the determination of the parameters can increase to such an extent (by several orders) that these additional parameters would be utterly uncertain.343 The reliability of the exchange parameters obtained from the experiment is extremely important for the subsequent quantum-chemical analysis of the mechanisms of exchange interaction between the paramagnetic centres.Molecular ferromagnets References 1.A L Buchachenko Usp. Khim. 59 529 (1990) [Russ. Chem. Rev. 59 307 (1990) 2. A L Buchachenko Mol. Cryst. Liq. Cryst. 176 307 (1989) 3. J S Miller Adv. Mater. 6 322 (1994) 4. J S Miller, A J Epstein,W M Reiff Science 240 40 (1988) 5. J S Miller, A J Epstein,W M Reiff Acc. Chem. Res. 21 114 (1988) 6. J S Miller, A J Epstein Angew. Chem., Int. Ed. Engl. 33 385 (1994) 7. A Caneschi, D Gatteschi, R Sessoli, P Rey Acc. Chem. Res. 22 392 (1989) 8. A Caneschi, D Gatteschi, P Rey Progr. Inorg. Chem. 39 331 (1991) 9. A F Williams, C Floriani, A E Merbach (Eds) Perspectives in Coordination Chemistry (Weinheim: VCH, 1992) 10.A Bencini, D Gatteschi EPR of Exchange Coupled Systems (Berlin: Springer, 1990) 11. L B Volodarsky, V A Reznikov, V I Ovcharenko Synthetic Chemistry of Stable Nitroxides (Boca Raton, FL: CRC Press, 1994) 12. H Iwamura, K Inoue, T Hayamizu Pure Appl. Chem. 68 243 (1996) 13. O Kahn Molecular Magnetism (New York: VCH, 1993) 14. V V Pavlishchuk Teor. Eksp. Khim. 33 341 (1997) a 15. The Vth International Conference on Molecule-Based Magnets (Abstracts of Reports), Osaka, 1996 16. The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 17. Mol. Cryst. Liq. Cryst. (Proceedings of the Symposium on the Chemistry and Physics of Molecular Based Magnetic Materials) 232 ± 233 (1992) 18.Mol. Cryst. Liq. Cryst. (Proceedings of the 5th International Conference on Molecule-Based Magnets) 305 ± 306 (1997) 19. Synth. Met. 85 (1997) 20. The XXXIIIth International Conference on Coordination Chemistry (Abstracts of Reports), Florence, 1998 21. R M Bozorth, H J Williams, D E Walsh Phys. Rev. 103 572 (1956) 22. A N Holden, B T Matthias, P W Anderson, H W Lewis Phys. Rev. 102 1463 (1956) 23. V Gadet, T Mallah, I Castro, M Verdaguer J. Am. Chem. Soc. 114 9213 (1992) 24. M Verdaguer, T Mallah, S Ferlay, V Gadet, C Helary, C Auberger, R Ouahgs, J Vaissermann, P Veillet, in Current Trends in Coordination Chemistry (Eds G Ondrejovich, A Sirota) (Smolenice: Slovak Technical University Press, 1995) p.291 25. T Mallah, S Thie baut, M Verdaguer, P Veillet Science 262 1554 (1993) 26. M Verdaguer Science 272 698 (1996) 27. S Ferlay, T Mallah, R Ouahe s, P Veillet,M Verdaguer Nature (London) 378 701 (1995) 28. O Kahn Nature (London) 378 667 (1995) 29. T Mallah, C Auberger,M Verdaguer, P Veillet J. Chem. Soc., Chem. Commun. 61 (1995) 30. W R Entley, G S Girolami Science 268 397 (1995) 31. W R Entley, G S Girolami Inorg. Chem. 33 5165 (1994) 32. H J Buser, D Schwarzenbach, W Petter, A Ludi Inorg. Chem. 16 2704 (1977) 33. K R Dunbar, R A Heintz Progr. Inorg. Chem. 45 283 (1997) 34. F Herren, P Fischer, A Ludi,W HaÈ lg Inorg. Chem. 19 956 (1980) 35. M Verdaguer, A Bleuzen, V Marvaud,M Seuleiman, J Vaissermann, C Desplanches, A Scuiller, R Garde, G Gelly, C Lomenech, I Rosenman, P Veillet, in The XXXIIIth International Conference on Coordination Chemistry.(Abstracts of Reports), Flor- ence, 1998 p. 133 36. A P Ginsberg Inorg. Chim. Acta Rev. 5 45 (1971) 37. Y Pei, Y Journaux, O Kahn Inorg. Chem. 28 100 (1989) 38. S Ohkoshi, O Sato, T Iyoda, A Fujishima, K Hashimoto Inorg. Chem. 36 268 (1997) 39. S Ohkoshi, T Iyoda, A Fujishima, K Hashimoto Phys. Rev. B, Con- dens. Matter 56 11642 (1997) 40. O Sato, T Iyoda, A Fujishima, K Hashimoto Science 271 49 (1996) 41. O Sato, T Iyoda, A Fujishima, K Hashimoto Science 272 704 (1996) 42. S Ohkoshi, S Yorozu, O Sato, T Iyoda, A Fujishima, K Hashimoto Appl. Phys. Lett. 30 1040 (1997) 43. S Ohkoshi, A Fujishima, K Hashimoto J.Am. Chem. Soc. 120 5349 (1998) 359 44. W-D Griebler, D Babel Z. Naturforsch., B Chem. Sci. 37 832 (1982) 45. D Babel Comments Inorg. Chem. 5 285 (1986) 46. R Klenze, B Kanellakopulos, G Trageser, H H Eysel J. Chem. Phys. 72 5819 (1980) 47. L Ne el Ann. Phys. 3 137 (1948) 48. J Larionova, J Sanchiz, S Gohlen, L Ouahab, O Kahn J. Chem. Soc., Chem. Commun. 953 (1998) 49. W E Buschmann, A M Arif, J S Miller Angew. Chem., Int. Ed. Engl. 37 781 (1998) 50. J L Manson,W E Buschmann, J S Miller Angew. Chem., Int. Ed. Engl. 37 783 (1998) 51. J Larionova, R Cle rac, O Kahn, S Gohlen, L Ouahab, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 237 52. R Garde, C Desplanches, A Bleuzen,M Verdaguer, P Veillet, in The VIth International Conference on Molecule-Based Magnets.(Abstracts of Reports), Seignosse, 1998 p. 249 53. N Fukita, M Ohba, H Okawa, K Matsuda, H Iwamura Inorg. Chem. 37 842 (1998) 54. H Miyasaka, H Okawa, A Miyazaki, T Enoki Inorg. Chem. 37 4878 (1998) 55. H Miyasaka, N Matsumoto, N Re, E Gallo, C Floriani Inorg. Chem. 36 670 (1997) 56. S Ferlay, T Mallah, J Vaissermann, F Bartolome , P Veillet, M Verdaguer J. Chem. Soc., Chem. Commun. 2481 (1996) 57. M Ohba, N Maruono, H Okawa, T Enoki, J-M Latour J. Am. Chem. Soc. 116 11566 (1994) 58. M Ohba, H Okawa, T Ito, A Ohto J. Chem. Soc., Chem. Commun. 1545 (1995) 59. M Ohba, N Usuki, N Fukita, H Okawa Inorg. Chem. 37 3349 (1998) 60. M Ohba, H Okawa, N Fukita, Y Hashimoto J.Am. Chem. Soc. 119 1011 (1997) 61. M S Fallah, E Rentschler, A Caneschi, R Sessoli, D Gatteschi Angew. Chem., Int. Ed. Engl. 35 1947 (1996) 62. E Colacio, J M Dominguez-Vera, M Ghazi, R KivekaÈ s,M Klinga, J M Moreno J. Chem. Soc., Chem. Commun. 1071 (1998) 63. H Miyasaka, N Matsumoto, H Okawa, N Re, E Gallo, C Floriani Angew. Chem., Int. Ed. Engl. 34 1446 (1995) 64. H Miyasaka, N Matsumoto, H Okawa, N Re, E Gallo, C Floriani J. Am. Chem. Soc. 118 981 (1996) 65. N Re, E Gallo, C Floriani, H Miyasaka, N Matsumoto Inorg. Chem. 35 6004 (1996) 66. H Miyasaka, H Ieda, N Matsumoto, N Re, R Crescenzi, C Floriani Inorg. Chem. 37 255 (1998) 67. N Re, E Gallo, C Floriani, H Miyasaka, N Matsumoto Inorg. Chem. 35 5964 (1996) 68. K van Langenberg, S R Batten, K J Berry, D C R Hockless, B Moubaraki, K S Murray Inorg.Chem. 36 5006 (1997) 69. F A Cotton, T R Felthouse Inorg. Chem. 21 2667 (1982) 70. D Luneau, P Rey, J Laugier, E Belorizky, A Cogne Inorg. Chem. 31 3578 (1992) 71. M Dickman, R J Doedens Inorg. Chem. 22 1591 (1983) 72. V I Ovcharenko, K E Vostrikova, A V Podoplelov, R Z Sagdeev, G V Romanenko, V N Ikorskii Polyhedron 13 2781 (1994) 73. Y Pei, O Kahn, K Nakatani, E Codjovi, C Mathoniere, J Sletten J. Am. Chem. Soc. 113 6558 (1991) 74. O A Bogdanova, E I Zhilyaeva, R N Lyubovskaya, N S Ovanesyan, S I Pirumova, O S Roshchupkina Izv. Akad. Nauk, Ser. Khim. 2327 (1996) b 75. R Sessoli, D Gatteschi, A Caneschi,M A Novak Nature (London) 365 141 (1993) 76.G L Abbati, A Cornia, A C Fabretti, A Caneschi, D Gattechi Inorg. Chem. 37 3759 (1998) 77. D Gatteschi Adv. Mater. 6 635 (1994) 78. M Hennion, L Pardi, L Mirebeau, E Suard, R Sessoli, D Gattechi Phys. Rev. B, Condens. Matter 56 8819 (1997) 79. K R Reddy, M V Rajasekharan, S Padhye, F Dahan, J-P Tuchagues Inorg. Chem. 33 428 (1994) 80. S M J Aubin, Z Sun, I A Guzei, A L Rheingold, G Christou, D N Hendrickson J. Chem. Soc., Chem. Commun. 2239 (1997) 81. H J Eppley, H L Tsai, N deVries, K Folting, G Christou, D N Hendrickson J. Am. Chem. Soc. 117 301 (1995)360 82. Z Sun, D Ruiz, E Rumberger, C D Incarvito, K Folting, A L Rheingold, G Christou, D N Hendrickson Inorg. Chem. 37 4758 (1998) 83. D Ruiz, Z Sun, B Albela, K Folting, J Ribas, G Christou, D N Hendrickson Angew.Chem., Int. Ed. Engl. 37 300 (1998) 84. J Tejada, J Maciejewski, R Ziolo J. Appl. Phys. 79 6031 (1996) 85. D A Garanin, E M Chudnovsky Phys. Rev. B, Condens. Matter 56 11 102 (1997) 86. F Luis, J Bartolome, J F Fernandez Phys. Rev. B, Condens. Matter 57 505 (1998) 87. K Takeda, K Awaga Phys. Rev. B, Condens. Matter 56 14 560 (1997) 88. A M Gomes, M A Novak, R Sessoli, A Caneschi, D Gatteschi Phys. Rev. B, Condens. Matter 57 5021 (1998) 89. S Ferlay, T Mallah, R Ouahes, P Veillet,M Verdauger Inorg. Chem. 38 229 (1999) 90. M Trollsss, J L Hedrick J. Am. Chem. Soc. 120 4644 (1998) 91. K Fegy, K E Vostrikova, D Luneau, P Rey Mol. Cryst. Liq. Cryst. 305 69 (1997) 92. B F Hoskins, R Robson J.Am. Chem. Soc. 112 1546 (1990) 93. L R MacGillivray, S Subramanian,M J Zaworotko J. Chem. Soc., Chem. Commun. 1325 (1994) 94. B F Abrahams,M J Hardie, B F Hoskins, R Robson, E E Sutherland J. Chem. Soc., Chem. Commun. 1049 (1994) 95. T Soma, H Yuge, T Iwamoto Angew. Chem., Int. Ed. Engl. 33 1665 (1994) 96. J Lu,W T A Harrison, A J Jacobson Angew. Chem., Int. Ed. Engl. 34 2557 (1995) 97. S Subramanian, M J Zaworotko Angew. Chem., Int. Ed. Engl. 34 2127 (1995) 98. L Carlucci, G Ciani, D M Proserpio, A Sironi Angew. Chem., Int. Ed. Engl. 34 1895 (1995) 99. P Schwarz, E Siebel, R D Fisher, D C Apperley, N A Devis, R K Harris Angew. Chem., Int. Ed. Engl. 34 1197 (1995) 100. B F Hoskins, R Robson, N V Y Scarlett J. Chem. Soc., Chem. Commun.2025 (1994) 101. Z J Zhong,N Matsumoto,H Okawa, S Kida Chem. Lett. 87 (1990) 102. H Tamaki, Z J Zhong, N Matsumoto, S Kida, M Koikawa, N Achiva, Y Hashimoto, H Okawa J. Am. Chem. Soc. 114 6974 (1992) 103. H Okawa, N Matsumoto, H Tamaki, M Ohba Mol. Cryst. Liq. Cryst. 233 257 (1993) 104. O Kahn, J Galy, Y Journaux, J Jaud, I Morgenstern Badarau J. Am. Chem. Soc. 104 2165 (1982) 105. Y Journaux, O Kahn, J Zarembowitch, J Galy, J Jaud J. Am. Chem. Soc. 105 7585 (1983) 106. N Matsumoto, K Inoue, H Okawa, S Kida Chem. Lett. 1251 (1989) 107. Z J Zhong, N Matsumoto, H Okawa, S Kida Inorg. Chem. 30 436 (1991) 108. G P Gupta, G Lang, C A Koch, B Wang,W R Scheidt, C A Reed Inorg. Chem. 29 4234 (1990) 109. L O Atovmyan, G V Shilov, R N Lyubovskaya, E I Zhilyaeva, N S Ovanesyan, Yu G Morozov, S I Pirumova, I G Gusakovskaya Pis'ma Zh.Eksp. Teor. Fiz. 58 818 (1993) 110. N S Ovanesyan, G V Shilov, L O Atovmyan, R N Lyubovskaya, A A Pyalling Mol. Cryst. Liq. Cryst. 273 175 (1995) 111. S Decurtins, H W Schmalle, P Schneuwly, H R Oswald Inorg. Chem. 32 1888 (1993) 112. S Decurtins, H W Schmalle, H R Oswald, A Linden, J Ensling, P GuÈ tlich, A Hauser Inorg. Chim. Acta 216 65 (1994) 113. R P Farrell, T W Hambley, P A Lay Inorg. Chem. 34 757 (1995) 114. R Pellaux, H W Schmalle, R Huber, P Fischer, T Hauss, B Ouladdiaf, S Decurtins Inorg. Chem. 32 2301 (1997) 115. H Tamaki, M Mitsumi, K Nakamura, N Matsumoto, S Kida, H Okawa, S Iijima Chem. Lett. 1975 (1992) 116. C J Nuttall, C Bellitto, P Day J.Chem. Soc., Chem. Commun. 1513 (1995) 117. S Decurtins, H W Schmalle, R Pellaux, P Fischer, A Hauser Mol. Cryst. Liq. Cryst. 305 227 (1997) 118. S Decurtins, H W Schmalle, P Schneuwly, J Ensling, P GuÈ tlich J. Am. Chem. Soc. 116 9521 (1994) 119. S Decurtins,M Gross, H W Schmalle, S Ferlay Inorg. Chem. 37 2443 (1998) V I Ovcharenko, R Z Sagdeev 120. S Decurtins, H W Schmalle, R Pellaux, P Schneuwly, A Hauser Inorg. Chem. 35 1451 (1996) 121. L O Atovmyan, G V Shilov, R N Lyubovskaya, E I Zhilyaeva, N S Ovanesyan, O A Bogdanova, S I Pirumova Koord. Khim. 23 683 (1997) c 122. M Hernandez-Molina, F Lloret, C Ruiz-Perez, M Julve Inorg. Chem. 37 4131 (1998) 123. W M Reiff, J Kreisz, L Meda, R U Kirss Mol. Cryst.Liq. Cryst. 273 181 (1995) 124. C Mathonie re, S G Carling, D Yusheng, P Day J. Chem. Soc., Chem. Commun. 1551 (1994) 125. C Mathonie re, C J Nuttall, S G Carling, P Day Inorg. Chem. 35 1201 (1996) 126. S G Carling, C Mathonie re, P Day, K M A Malik, S J Coles, M B Hursthouse J. Chem. Soc., Dalton Trans. 1839 (1996) 127. J Larionova, B Mombelli, J Sanchiz, O Kahn Inorg. Chem. 37 679 (1998) 128. P Day Mol. Cryst. Liq. Cryst. 305 533 (1997) 129. A Gleizes,M Verdaguer J. Am. Chem. Soc. 103 7373 (1981) 130. A Gleizes,M Verdaguer J. Am. Chem. Soc. 106 3727 (1984) 131. S J Rettig, A Storr, D A Summers, R C Thompson, J Trotter J. Am. Chem. Soc. 119 8675 (1997) 132. M A M Lorente, J-P Tuchagues, V Pe troule as, J-M Savariault, R Poinsot,M Drillon Inorg.Chem. 30 3587 (1991) 133. M A Martinez-Lorente, F Dahan, Y Sanakis, V Petrouleas, A Bousseksou, J-P Tuchagues Inorg. Chem. 34 5346 (1995) 134. R Vicente, A Escuer, X Solans,M Font-Bardta J. Chem. Soc., Dalton Trans. 1835 (1996) 135. G De Munno, R Ruiz, F Lloret, J Faus, R Sessoli, M Julve Inorg. Chem. 34 408 (1995) 136. G De Munno, M Julve, F Nicolo , F Lloret, J Faus, R Ruiz, E Sinn Angew. Chem., Int. Ed. Engl. 32 613 (1993) 137. F D Rochon, R Melanson,M Andruh Inorg. Chem. 35 6086 (1996) 138. G De Munno, T Poerio, G Viau,M Julve, F Lloret, Y Journaux, E Rivie re J. Chem. Soc., Chem. Commun. 2587 (1996) 139. G De Munno, M Julve, G Viau, F Lloret, J Faus, D Viterbo Angew. Chem., Int. Ed. Engl. 35 1807 (1996) 140.R Corte s, L Lezama, J L Pizarro, M I Arriortua, T Rojo Angew. Chem., Int. Ed. Engl. 35 1810 (1996) 141. M Julve,M Verdaguer, A Gleizes,M Philoche-Levisalles, O Kahn Inorg. Chem. 23 3808 (1984) 142. M Julve,M Verdaguer, G De Munno, J A Real, G Bruno Inorg. Chem. 32 795 (1993) 143. A Escuer, R Vicente, J Ribas, M S El-Fallah, X Solans, M Font-Bardta Inorg. Chem. 33 1842 (1994) 144. P Chaudhuri, T Weyhermmller, E Bill, K Wieghardt Inorg. Chim. Acta 252 195 (1996) 145. J Ribas, M Monfort, C Diaz, C Bastos, X Solans Inorg. Chem. 33 484 (1994) 146. E Ruiz, J Cano, S Alvarez, P Alemany J. Am. Chem. Soc. 120 11122 (1998) 147. F Wagner, M T Mocella,M J D'Aniello Jr , A H-J Wang, E K Barefield J. Am. Chem. Soc. 96 2625 (1974) 148. C G Pierpont, D N Hendrickson, D M Duggan, F Wagner, E K Barefield Inorg.Chem. 14 604 (1975) 149. P Chaudhuri,M Guttman, D Ventur, K Wieghardt, B Nuber, J Weiss J. Chem. Soc., Chem. Commun. 1678 (1985) 150. J Ribas, M Monfort, C Diaz, C Bastos, X Solans Inorg. Chem. 32 3557 (1993) 151. M I Arriortua, A R Cortes, L Lezam, T Rojo, X Solans, M Font-Bardia Inorg. Chim. Acta 174 263 (1990) 152. R Vicente, A Escuer, J Ribas, M S El Fallah, X Solans, M Font-Bardia Inorg. Chem. 32 1920 (1993) 153. J Ribas,M Monfort,R Costa,X Solans Inorg. Chem. 32 695 (1993) 154. A Escuer, R Vicente, J Ribas, M S El Fallah, X Solans Inorg. Chem. 32 1033 (1993) 155. R Vicente, A Escuer, J Ribas, X Solans Inorg. Chem. 31 1726 (1992) 156. L K Thompson, S S Tandon, M E Manuel Inorg.Chem. 34 2356 (1995) 157. K R Reddy, M V Rajasekharan, J-P Tuchagues Inorg. Chem. 37 5978 (1998) 158. S Sikorav, I Bkouche-Waksman, O Kahn Inorg. Chem. 23 490 (1984)Molecular ferromagnets 159. R Corte s, J I Ruiz de Larramendi, L Lezama, T Rojo, K Urtiaga, M I Arriortua J. Chem. Soc., Dalton Trans. 2723 (1992) 160. A Escuer, R Vicente, J Ribas, X Solans Inorg. Chem. 34 1793 (1995) 161. M G B Drew, C J Harding, J Nelson Inorg. Chim. Acta 246 73 (1996) 162. R Corte s, J L Pizarro, L Lezama,M I Arriortua, T Rojo Inorg. Chem. 33 2697 (1994) 163. M Monfort, J Ribas, X Solans J. Chem. Soc., Chem. Commun. 350 (1993) 164. J Ribas, M Monfort, X Solans,M Drillon Inorg. Chem. 33 742 (1994) 165. I Dzyaloshinsky J. Phys.Chem. Solids 4 241 (1958) 166. I F Silvera, J H M Thorneley, M Tinkman Phys. Rev. A, Gen. Phys., Ser. 3 136 695 (1964) 167. F Keffer Phys. Rev. 126 896 (1962) 168. E A Turov Fizicheskie Svoistva Magnitnouporyadochennykh Kristallov (Physical Properties of Magnetically Ordered Crystals) (Moscow: Izd. Akad. Nauk SSSR, 1963) 169. T Moriya Phys. Rev. 117 635 (1960) 170. A Escuer, R Vicente,M A S Goher, F A Mautner Inorg. Chem. 34 5707 (1995) 171. A Escuer, R Vicente,M A S Goher, F A Mautner Inorg. Chem. 35 6386 (1996) 172. A Escuer, R Vicente,M A S Goher, F A Mautner Inorg. Chem. 36 3440 (1997) 173. S R Batten, P Jensen, B Moubaraki, K S Murray, R Robson J. Chem. Soc., Chem. Commun. 439 (1998) 174. K S Murray, K van Langenberg, B Moubaraki, K J Berry, P Jensen, D J Price, S R Batten, R Robson, D C R Hockless, in XXXIIIth International Conference on Coordination Chemistry (Abstracts of Reports), Florence, 1998 p.389 175. K S Murray, B Moubaraki, P Jensen, D J Price, S R Batten, R Robson in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 87 176. C R Kmety, J L Manson, Q Huang, J W Lynn, R W Erwin, J S Miller, A J Epstein, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 88 177. J L Manson,D W Lee, A L Rheingold, J S Miller Inorg. Chem. 37 5966 (1998) 178. J L Manson, C R Kmety, Q-Z Huang, J W Lynn, G M Bendele, S Pagola, P W Stephens, L M Liable-Sands, A L Rheingold, A J Epstein, J S Miller Chem.Mater. 10 2552 (1988) 179. C J Kepert, M Kurmoo, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 214 180. Y Iwasa,M Kurmoo, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 212 181. M Kurmoo, J P Kapler, W Grange, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 213 182. M Kurmoo, in The VIth International Conference on Molecule- Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 60 183. M Malinowska, C Meny, P Panissod,M Kurmoo, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p.215 184. T Jestidt, M Kurmoo, S J Blundell, B W Lovett, F L Pratt, R I Bewley, K H Chow, W Hayes, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 216 185. Y Pei,M Verdaguer, O Kahn, J Sletten, J-P Renard J. Am. Chem. Soc. 108 7428 (1986) 186. Y Pei,M Verdaguer, O Kahn, J Sletten, J-P Renard Inorg. Chem. 26 138 (1987) 187. O Kahn, Y Pei,M Verdaguer, J P Renard, J Sletten J. Am. Chem. Soc. 110 782 (1988) 188. F Lloret, K Nakatani, Y Journaux, O Kahn, Y Pei, J P Renard J. Chem. Soc., Chem. Commun. 642 (1988) 189. K Nakatani, J Y Carriat, Y Journaux, O Kahn, F Lloret, J P Renard, Y Pei, J Sletten, M Verdaguer J. Am. Chem. Soc. 111 5739 (1989) 361 190. H O Stumpf, Y Pei, O Kahn, J Sletten, J P Renard J. Am.Chem. Soc. 115 6738 (1993) 191. H O Stumpf, Y Pei, L Ouahab, F Le Berre, E Codjovi, O Kahn Inorg. Chem. 32 5687 (1993) 192. F Lloret, M Julve, R Ruiz, Y Journaux, K Nakatani, O Kahn, J Sletten Inorg. Chem. 32 27 (1993) 193. O Kahn,H Stumpf, Y Pei, J Sletten Mol. Cryst. Liq. Cryst. 233 231 (1993) 194. V Baron, B Gillon, O Kahn, C Mathonizhre,M Bonnet, J X Boucherle Mol. Cryst. Liq. Cryst. 233 247 (1993) 195. M Fettouhi, L Ouahab, A Boukhari, O Cador, C Mathonizhre, O Kahn Inorg. Chem. 35 4932 (1996) 196. H O Stumpf, L Ouahab, Y Pei, D Grandjean, O Kahn Science 261 447 (1993) 197. H O Stumpf, L Ouahab, Y Pei, P Bergerat, O Kahn J. Am. Chem. Soc. 116 3866 (1994) 198. M G F Vaz, H O Stumpf, L Ouahab, O Cador, C Mathonizhre, O Kahn, in The VIth International Conference on Molecule-Based 199.A Caneschi, D Gatteschi, J Laugier, P Rey, R Sessoli, C Zanchini 200. P Rey, J Laugier, A Caneschi, D Gatteschi Mol. Cryst. Liq. Cryst. 201. C Benelli, A Caneschi, D Gatteschi, M C Melandri, P Rey Inorg. 202. A Caneschi, D Gatteschi, M C Melandri, P Rey, R Sessoli Inorg. 203. C Benelli, A Caneschi, D Gatteschi, L Pardi, P Rey Inorg. Chem. 204. D Luneau, P Rey, J Laugier, P Fries, A Caneschi, D Gatteschi, 205. A Caneschi, D Gatteschi, R Sessoli, P Rey Inorg. Chim. Acta 184 Magnets (Abstracts of Reports), Seignosse, 1998 p. 180 J. Am. Chem. Soc. 110 2795 (1988) 176 337 (1989) Chim. Acta 172 137 (1990) Chem. 29 4228 (1990) 29 4223 (1990) R Sessoli J.Am. Chem. Soc. 113 1245 (1991) 67 (1991) 206. C Benelli, A Caneschi, D Gatteschi, L Pardi Inorg. Chem. 31 741 (1992) 207. A Caneschi, D Gatteschi, R Sessoli Inorg. Chem. 32 4612 (1993) 208. A Caneschi, P Chiesi, L David, F Ferraro, D Gatteschi, R Sessoli Inorg. Chem. 32 1445 (1993) 209. A Caneschi, L David, F Ferraro,D Gatteschi, A C Fabretti Inorg. Chim. Acta 217 7 (1994) 210. F L de Panthou, D Luneau, R Musin, L OÈ hrstroÈ m, A Grand, P Turek, P Rey Inorg. Chem. 35 3484 (1996) 211. N C Schiudt, F F Biani, A Caneschi, D Gatteschi Inorg. Chim. Acta 248 139 (1996) 212. G GoÈ rlitz, T Hayamizu, T Itoh, K Matsuda, H Iwamura Inorg. Chem. 37 2083 (1998) 213. M Tanaka,K Matsuda, T Itoh,H Iwamura Angew. Chem., Int. Ed.Engl. 37 810 (1998) 214. F L de Panthou, E Belorizky, R Calemczuk, D Luneau, C Marcenat, E Ressouche, P Turek, P Rey J. Am. Chem. Soc. 117 215. V I Ovcharenko, A B Gel'man, V N Ikorskii Zh. Strukt. Khim. 30 11247 (1995) (5) 142 (1989) d 216. N V Pervukhina, N V Podberezskaya, V I Ovcharenko, S V Larionov Zh. Strukt. Khim. 32 (1) 123 (1991) d 217. A B Burdukov, V I Ovcharenko, V N Ikorski, N V Pervukhina, N V Podberezskaya, I A Grigor'ev, S V Larionov, L B Volodarskii Inorg. Chem. 30 972 (1991) 218. V I Ovcharenko, V N Ikorskii, K E Vostrikova, A B Burdukov, G V Romanenko, N V Pervukhina, N V Podberezskaya Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk 100 (1990) 219. A B Burdukov, V I Ovcharenko, N V Pervukhina, V N Ikorskii, N V Podberezskaya Zh.Neorg. Khim. 36 2058 (1991) e 220. V I Ovcharenko, V N Ikorskii, G V Romanenko, V A Reznikov, L B Volodarskii Inorg. Chim. Acta 187 67 (1991) 221. A B Burdukov, V I Ovcharenko, N V Pervukhina, V N Ikorski, L Greci, P Carloni, E Damiani, P Stipa, V A Reznikov Polyhedron 12 1705 (1993) 222. V I Ovcharenko, G V Romanenko, V N Ikorskii, R N Musin, R Z Sagdeev Inorg. Chem. 33 3370 (1994) 223. A B Burdukov, V I Ovcharenko, R Z Sagdeev, A V Podoplelov, N V Pervukhina, V N Ikorskii, I A Kirilyuk, I A Grigor'ev Polyhedron 14 501 (1995)362 224. V Ovcharenko, A Burdukov, R Musin Mol. Cryst. Liq. Cryst. 273 89 (1995) 225. A B Burdukov, V I Ovcharenko, R Z Sagdeev, N V Pervukhina, V N Ikorskii, I A Kirilyuk, I A Grigor'ev Polyhedron 15 4211 (1996) 226.V Ovcharenko, I Korobkov, S Fokin, A Burdukov, G Romanenko, V Ikorskii, N Pervukhina, D Mazhukin, V Reznikov Mol. Cryst. Liq. Cryst. 305 311 (1997) 227. V I Ovcharenko, S V Fokin, V A Reznikov, V N Ikorskii, G V Romanenko, R Z Sagdeev Inorg. Chem. 37 2104 (1998) 228. K Inoue, H Iwamura Angew. Chem., Int. Ed. Engl. 34 927 (1995) 229. K Inoue, T Hayamizu, H Iwamura Chem. Lett. 745 (1995) 230. K Inoue, H Iwamura J. Am. Chem. Soc. 116 3173 (1994) 231. Y Ishimaru,M Kitano, H Kumada, N Koga, H Iwamura Inorg. 232. A Caneschi, D Gatteschi, P Rey, R Sessoli Inorg. Chem. 27 1756 233. A Caneschi, D Gatteschi, J P Renard, P Rey, R Sessoli Inorg. 234. A Caneschi, D Gatteschi, J P Renard, P Rey, R Sessoli Inorg. Chem. 37 2273 (1998) (1988) Chem.28 1976 (1989) Chem. 28 3314 (1989) 235. A Caneschi, D Gatteschi, R Sessoli, P Rey Mol. Cryst. Liq. Cryst. 176 329 (1989) 236. K Inoue, H Iwamura J. Chem. Soc., Chem. Commun. 2273 (1994) 237. K Inoue, H Iwamura Synth. Met. 71 1793 (1995) 238. K Inoue, T Hayamizu, H Iwamura, D Hashizumi, Y Ohashi J. Am. Chem. Soc. 118 1803 (1996) 239. K Inoue, T Hayamizu, H Iwamura Mol. Cryst. Liq. Cryst. 273 67 (1995) 240. K Inoue, H Iwamura Adv. Mater. 8(1) 73 (1996) 241. V I Ovcharenko, V N Ikorskii, N V Podberezskaya, N V Pervukhina, S V Larionov Zh. Neorg. Khim. 32 1403 (1987) e 242. A Caneschi, D Gatteschi, R Sessoli, S K Hoffman, J Laugier, P Rey Inorg. Chem. 27 2390 (1988) 243. V N Ikorskii, V I Ovcharenko Zh. Neorg.Khim 35 2093 (1990) e 244. V N Ikorskii, V I Ovcharenko, K E Vostrikova, N V Pervukhina, N V Podberezskaya Zh. Neorg. Khim. 37 1177 (1992) e 245. Y G Shvedenkov, V N Ikorskii, G V Romanenko, V I Ovcharenko, E Y Fursova, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), 246. A Caneschi, D Gatteschi, J P Renard, P Rey, R Sessoli J. Am. 247. V N Ikorskii, V I Ovcharenko, Y G Shvedenkov, G V Romanenko, Seignosse, 1998 p. 158 Chem. Soc. 111 785 (1989) S V Fokin, R Z Sagdeev Inorg. Chem. 37 4360 (1998) 248. V I Ovcharenko, K E Vostrikova, V N Ikorskii, G V Romanenko, N V Podberezskaya, S V Larionov, R Z Sagdeev Sintez, Struktura i Magnitnye Svoistva Nizkotemper- aturnykh Ferromagnetikov na Osnove Kompleksov Ni(II) i Co(II) s Imidazolinovym Nitroksil'nym Radikalom (Preprint 89-2) [Synthesis, Structure and Magnetic Properties of Low-Temperature Ferromagnetics Based on Ni(II) and Co(II) Complexes with Imi- dazoline Nitroxide (Preprint 89-2)] (Novosibirsk: Institute of Inor- ganic Chemistry, Siberian Branch of Academy of Sciences of the USSR, Institute Chemical Kinetics and Combustion, Siberian Branch of Academy of Sciences of the USSR, 1989) 249.G V Romanenko, V N Ikorskii, S V Fokin, V I Ovcharenko Zh. Strukt. Khim. 38 930 (1997) d 250. V I Ovcharenko, K E Vostrikova, G V Romanenko, V N Ikorskii, N V Podberezskaya, S V Larionov Dokl. Akad. Nauk SSSR 306 115 (1989) f 251. V I Ovcharenko, K E Vostrikova, V N Ikorskii, S V Larionov, R Z Sagdeev Dokl.Akad. Nauk SSSR 306 660 (1989) f 252. V N Ikorskii, G V Romanenko,M K Sygurova, V I Ovcharenko, F L Pontu, P Re, V A Reznikov, A V Podoplelov, R Z Sagdeev Zh. Strukt. Khim. 35 (4) 76 (1994) d 253. V I Ovcharenko, V N Ikorskii, G V Romanenko, R Z Sagdeev, F L de Panthou, P Rey Synth. Met. 85 1639 (1997) 254. V I Ovcharenko, S V Larionov, V K Mokhosoeva, L B Volodarskii Zh. Neorg. Khim. 28 151 (1983) e 255. N V Pervukhina, V N Ikorskii, V I Ovcharenko, N V Podberezskaya, P S Nikitin, A B Gel'man, S V Larionov, V V Bakakin Zh. Strukt. Khim. 27 (4) 61 (1986) d V I Ovcharenko, R Z Sagdeev 256. E Ressouche, J-X Boucherle, B Gillon, P Rey, J Schweizer J. Am. Chem. Soc. 115 3610 (1993) 257. E Ressouche, A Zheludev, J X Boucherle, B Gillon, P Rey, J Schweizer Mol. Cryst.Liq. Cryst. 232 13 (1993) 258. M Bonnet, J Laugier, V I Ovcharenko, Y Pontillon, E Ressouche, P Rey, P Schleger, J Schweizer Mol. Cryst. Liq. Cryst. 305 401 261. K Fegy, N Sanz, D Luneau, E Belorizky, P Rey Inorg. Chem. 37 (1997) 259. I Ovcharenko, A Grand, N Sanz, V Ovcharenko, Y Pontillon, E Ressouche, P Rey, J Schweizer, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 178 260. K E Vostrikova, V I Ovcharenko, G V Romanenko, V N Ikorskii, N V Podberezskaya, V A Reznikov, L B Volodarskii Zh. Neorg. Khim. 37 1755 (1992) e 4518 (1998) 262. K Fegy, D Luneau, T Ohm, C Paulsen, P Rey Angew. Chem., Int. Ed. Engl. 37 1270 (1998) 271. Y V Korshak, T V Medvedeva, A A Ovchinnikov, V N Spector 263.P Day J. Chem. Soc., Dalton Trans. 701 (1997) 264. A A Ovchinnikov Dokl. Akad. Nauk SSSR 236 928 (1977) f 265. A A Ovchinnikov Theor. Chim. Acta 47 297 (1978) 266. A L Buchachenko Khim. Fiz. 3 3 (1984) g 267. A L Buchachenko Dokl. Akad. Nauk SSSR 244 1146 (1979) f 268. R Breslow Pure Appl. Chem. 54 927 (1982) 269. T J LePage, R Breslow J. Am. Chem. Soc. 109 6412 (1987) 270. Yu V Korshak, A A Ovchinnikov, AM Shapiro, T V Medvedeva, V N Spektor Pis'ma Zh. Eksp. Teor. Fiz, 43 309 (1986) Nature (London) 326 370 (1987) 272. Y Cao, P Wang, Z Hu, S Li, L Zhang, J Zhao Synth. Met. 27 B625 (1988) 273. J B Torrance, S Oostra, A Nazzal Synth. Met. 19 709 (1987) 274. A A Ovchinnikov, U Abdurakhmanov, A A Kuznetsov, M A Magrupov, V N Spektor Dokl.Akad. Nauk SSSR 302 885 (1988) f 275. S Galaj, A LeMe haute , F Tsobnang, D Cottevieille, A Le austic, R Cle ment, V Cagan, M Guyot, J C Fayet, B Villeret, L Noirez, A Pe richaud J. Magn. Magn. Mater. 140 ± 144 1445 (1995) 276. A L Buchachenko, R P Shibaeva, L P Rozenberg, A A Dadali Khim. Fiz. 6 773 (1987) g 277. J S Miller, D T Glatzhofer, J C Calabrese, A J Epstein J. Chem. Soc., Chem. Commun. 322 (1988) 278. D W Wiley, J C Calabrese, J S Miller J. Chem. Soc., Chem. Commun. 1523 (1989) 279. R Laversanne, A Chakraborty, D T Glatzhofer, A J Epstein, J S Miller Synth. Met. B353 (1988) 280. J S Miller, D T Glatzhofer, R Laversanne, T B Brill, M D Timken, C J O'Connor, J H Zhang, J C Calabrese, A J Epstein, S Chittipeddi, P Vaca Chem.Mater. 2 60 (1990) 281. T A Aleksandrova, A M Vasserman, A Yu D'yakonov, A M Shapiro,A L Buchachenko Dokl. Akad. Nauk SSSR 302 1399 (1988) f 282. M Tamura, Y Nakazawa, D Shiomi, K Nozawa, Y Hosokoshi, M Ishikawa,M Takahashi,M Kinoshita Chem. Phys. Lett. 186 401 (1991) 283. Y Nakazawa,M Tamura, N Shirakawa, D Shiomi, M Takahashi, M Kinoshita, M Ishikawa Phys. Rev. B, Condens. Matter 46 8906 (1992) 284. R Chiarelli, M A Novak, A Rassat, J L Tholence Nature (London) 363 147 (1993) 285. P Day Nature (London) 363 113 (1993) 286. A Rajca Chem. Rev. 94 871 (1994) 287. T Nogami, K Tomioka, T Ishida, H Yoshikawa,M Yasui, F Iwasaki, H Iwamura, N Takeda, M Ishikawa Chem. Lett.29 (1994) 288. T Ishida, H Tsuboi, T Nogami, H Yoshikawa,M Yasui, F Iwasaki, H Iwamura, N Takeda, M Ishikawa Chem. Lett. 919 (1994) 289. Y Hosokoshi, M Tamura, H Sawa, R Kato, M Kinoshita J. Mater. Chem. 5 41 (1995) 290. K Takeda, K Konishi,M Hitaka, T Kawae, T Tamura, M Kinoshita J. Magn. Magn. Mater. 140 ± 144 1451 (1995) 291. K Mukai, K Konishi, K Nedachi, K Takeda J. Magn. Magn. Mater. 140 ± 144 1449 (1995)Molecular ferromagnets 292. M Kinoshita, P Turek,M Tamura, K Nozawa, D Shiomi, Y Nakazawa,M Ishikawa, M Takahashi, K Awaga, T Inabe, Y Maruyama Chem. Lett. 1225 (1991) 293. E Hernendez,M Mas, E Molins, C Rovira, J Veciana Angew. Chem., Int. Ed. Engl. 32 882 (1993) 294. Y Hosokoshi, M Tamura, D Shiomi, N Iwasawa, K Nozawa, M Kinoshita, H A Katori, T Goto Physica B 201 497 (1994) 295.M Kinoshita Mol. Cryst. Liq. Cryst. 232 1 (1993) 296. T C Kobayashi,M Takiguchi, C U Hong,K Amaya, A Kajiwara, A Harada, M Kamachi J. Magn. Magn. Mater. 140 ± 144 1447 (1995) 297. F M Romero, R Ziessel, M Drillon, J L Tholence, C Paulsen, N Kyritsakas, J Fisher Adv. Mater. 8 826 (1996) 298. Y Pei, O Kahn,M A Aebersold, L Ouahab, F Le Berre, L Pardi, J L Tholence Adv. Mater. 6 681 (1994) 299. J Cirujeda, E Hernendez-Gasio, C Rovira, J L Stranger, P Turek, J Veciana J. Mater. Chem. 5 243 (1995) 300. A Zheludev, R Chiarelli, B Delley, B Gillon, A Rassat, E Ressouche, J Schweizer J. Magn. Magn. Mater. 140 ± 144 1439 (1995) 301. A Zheludev, E Ressouche, J Schweizer, P Turek,M Wan,H Wang J. Magn. Magn. Mater. 140 ± 144 1441 (1995) 302. J Cirujeda,M Mas, E Molins, F L de Panthou, J Laugier, J G Park, C Paulsen, P Rey, C Rovira, J Veciana J. Chem. Soc., Chem. Commun. 709 (1995) 303. D Shiomi, M Tamura, H A Katori, T Goto, A Hayashi, Y Ueda, H Sawa, R Kato,M Kinoshita J. Mater. Chem. 4 915 (1994) 304. P-M Allemand, K C Khemani, A Koch, F Wudl, K Holczer, S Donovan, G GruÈ ner, J D Thompson Science 253 301 (1991) 305. J S Miller, J C Calabrese, H Rommelmann, S R Chittipeddi, J H Zhang,W M Reiff, A J Epstein J. Am. Chem. Soc. 109 769 (1987) 306. J S Miller, A J Epstein Mol. Cryst. Liq. Cryst. 233 133 (1993) 307. J S Miller, J C Calabrese, R L Harlow, D A Dixon, J H Zhang, W M Reiff, S Chittipeddi,M A Selover, A J Epstein J. Am. Chem. Soc. 112 5496 (1990) 308. J S Miller, J H Zhang, W M Reiff J. Am. Chem. Soc. 109 4584 (1987) 309. J S Miller, D T Glatzhofer, D M O'Hare, W M Reiff, A Chakraborty, A J Epstein Inorg. Chem. 28 2930 (1989) 310. J S Miller, J C Calabrese, A J Epstein Inorg. Chem. 28 4230 (1989) 311. J S Miller, D M O'Hare, A Chakraborty, A J Epstein J. Am. Chem. Soc. 111 7853 (1989) 312. J S Miller, A J Epstein Mol. Cryst. Liq. Cryst. 176 347 (1989) 313. J S Miller, J C Calabrese, A J Epstein, R W Bigelow, J H Zhang, W M Reiff J. Chem. Soc., Chem. Commun. 1026 (1986) 314. J S Miller Adv. Mater. 2 98 (1990) 315. W E Broderick, J A Thompson, E P Day, B M Hoffman Science 249 401 (1990) 316. D M Eichhorn, D C Skee,W E Broderick, B M Hoffman Inorg. Chem. 32 491 (1993) 317. C Kollmar, O Kahn J. Am. Chem. Soc. 113 7987 (1991) 318. C Kollmar,M Couty, O Kahn J. Am. Chem. Soc. 113 7994 (1991) 319. C Kollmar, O Kahn J. Chem. Phys. 98 453 (1993) 320. J S Miller, J C Calabrese, R S McLean, A J Epstein Adv. Mater. 4 498 (1992) 321. E J Brandon, K-I Sugiura, A M Arif, L Liable-Sands, A L Rheingold, J S Miller Mol. Cryst. Liq. Cryst. 305 269 (1997) 322. E J Brandon, D K Rittenberg, A M Arif, J S Miller Inorg. Chem. 37 3376 (1998) 323. E J Brandon, R D Rogers, B M Burkhart, J S Miller Chem. Eur. J. 4 1938 (1998) 324. S P Sellers, B J Korte, J P Fitzgerald,W M Reiff, G T Yee J. Am. Chem. Soc. 120 4662 (1998) 325. H Winter, E Dormann, R Gompper, R Janner, S Kothrade, B Wagner, H Naarmann J. Magn. Magn. Mater. 140 ± 144 1443 327. J Zhang, J Ensling, V Ksenofontov, P GuÈ tlich, A J Epstein, (1995) 326. A J Epstein, C M Wynn, C R Kmety, A Y Mishurov, J Zhang, J S Miller, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 89 J S Miller Angew. Chem, Int. Ed. Engl. 37 657 (1998) 328. J M Manriquez, G T Yee, R S McLean, A J Epstein, J S Miller Science 252 1415 (1991) 363 329. A J Epstein, J S Miller Mol. Cryst. Liq. Cryst. 233 171 (1993) 330. V A Reznikov, I V Ovcharenko, N V Pervukhina, V N Ikorskii, A Grand, V I Ovcharenko J. Chem. Soc., Chem. Commun. 539 (1999) 331. J Cirujeda, L E Ochando, J M Amigd, C Rovira, J Rius, J Veciana Angew. Chem., Int. Ed. Engl. 34 55 (1995) 332. R Chiarelli, S Gambarelli,A Rassat Mol. Cryst. Liq. Cryst. 306 257 (1997) 333. T Akita, K Kobayashi Mol. Cryst. Liq. Cryst. 305 455 (1997) 334. A Lang, Y Pei, L Ouahab, O Kahn Adv. Mater. 8 60 (1996) 335. P M Lahti, B Esat, R Walton J. Am. Chem. Soc. 120 5122 (1998) 336. G S Girolami, W R Entley, C R Treadway, S M Holmes, in The VIth International Conference on Molecule-Based Magnets (Abstracts of Reports), Seignosse, 1998 p. 63 337. E Dujardin, S Ferlay, X Phan, C Desplanches, C Cartier dit Moulin, P Sainctavit, F Baudelet, E Dartyge, P Veillet, M Verdaguer J. Am. Chem. Soc. 120 11347 (1998) 338. A L Buchachenko Usp. Khim. 68 99 (1999) [Russ. Chem. Rev. 68 85 (1999) 339. R L Carlin Magnetochemistry (Berlin: Springer, 1986) 340. V T Kalinnikov, Yu V Rakitin Vvedenie v Magnetokhimiyu. Metod Staticheskoi Magnitnoi Vospriimchivosti (Introduction in Magneto- chemistry. The Method of Static Magnetic Susceptibility (Moscow: Nauka, 1980) 341. Yu V Rakitin, V T Kalinnikov Sovremennaya Magnetokhimiya (Modern Magnetochemistry) (St.-Petersburg: Nauka, 1994) 342. A A Ovchinnikov, V N Spector The World Polymer Encyclopedia CRC Press, Boca Raton, FL, 1996 343. I V Ovcharenko, Yu G Shvedenkov, R N Musin, V N Ikorskii Zh. Strukt. Khim. 40 (1) 36 (1999) d a�Teor. Exp. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Transl.) c�Russ. J. Coord. Chem. (Engl. Transl.) d�Russ. J. Struct. Chem. (Engl. Transl.) e�Russ. J. Inorg. Chem. (Engl. Transl.) f�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) g�Russ. J. Chem. Phys. (

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (5) 365 ± 379 (1999) The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements YuMKiselev, Yu D Tretiyakov Contents I. Introduction II. A concept of oxidation state stabilisation III. Rare-earth elements IV. Transition elements V. Matrix oxidation state stabilisation VI. Valence possibilities of transition elements VII. Regularities of changes in the extreme oxidation states of transition elements Abstract. The general principles of the concept of oxidation state stabilisation are formulated. Problems associated with the prepa- ration and provision of the highest valent forms of transition elements are considered. The empirical data concerning the syn- thesis of new compounds of rare-earth elements and d elements in unusually high oxidation states are analysed.The possibility of occurrence of the oxidation states +9 and +10 for some elements (for example, for iridium and platinum in tetraoxo ions) are discussed. Approaches to the realisation of these states are outlined and it is demonstrated that solid phases or matrices containing alkali metal cations are the most promising systems for the stabilisation of these high oxidation states. Selected thermo- dynamic features typical of metal halides and oxides and the regularities of the changes in the extreme oxidation states of d elements are considered. The bibliography includes 266 references. I. Introduction According to traditional concepts, chemical compounds contain elements in definite oxidation states (OS).These concepts form the basis of, for example, authoritative handbooks on inorganic and coordination chemistry.1, 2 A hypothesis for the stabilisation of the oxidation states has been offered rather long ago. The theoretical and experimental aspects of this hypothesis were discussed in the early 1960s (see, for example, Refs 3 ± 7). More recently, the experimental aspect of the problem of the existence and OS stabilisation was repeatedly considered in connection with the discovery of noble gas compounds,8 studies of chemical properties of transactinides 9, 10 and the preparation of com- pounds of NpVII, PuVII,10 AmVII,11 ± 14 CmVI,15 RnVI,16 CfV,17 CmV,15,18 AuV,19 AgV,20 EsIV,18 CuIV,21 TmIV 22 and lanthanides and actinides in the lowest oxidation states.23 ± 29 Worthy of mention are investigations that have demonstrated the similarity Yu MKiselev, Yu D Tretiyakov Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation.Fax (7-095) 932 88 46. Tel. (7-095) 939 22 52. E-mail: kiselev@coord.chem.msu.ru (Yu MKiselev) Tel. (7-095) 939 20 74 (Yu D Tretiyakov) Received 1 December 1998 Uspekhi Khimii 68 (5) 401 ± 415 (1999); translated by T N Safonova #1999 Russian Academy of Sciences and Turpion Ltd UDC 541.49 365 365 369 371 372 373 374 of the properties of heavy actinides and lanthanides, the discovery of the specificity of the so-called tetrad effect in lanthanide compounds,23, 29, 30 the preparation of FeVIII (see.Refs 31 and 32) and the detection of CoVI in a cesium ferrate matrix.33, 34 Studies of the properties of elements with atomic numbers 104, 105 and 106, viz., rutherfordium, dubnium and seaborgium,35, 36 and the possibility of generation of HgIV (see Refs 37 and 38) are also worthy of note. Recently, new aspects of the problem under consideration became apparent to researchers in connection with the develop- ment of novel techniques for the stabilisation of oxidation states. With this in mind, Yatsimirskii 39, 40 and Parry 41 examined the OS stabilisation upon coordination based on the original data. Previously, we have considered different aspects of this problem.In particular, we have discussed the stability of the highest OS of lanthanides in fluoride systems,29, 42 the effect of outer-sphere cations on the stability of coordination compounds,43 the ways of attaining high OS in the oxidative fluorination reactions,44 procedures for the identification of the valent forms of transition elements 45 and some problems associated with the concept of valence,46 in particular, some aspects of the problem of OS stabilisation in the solid phase47 and the possibility of generation of high valent forms (5+8) of the heaviest transition elements (Ir, Pt, Au and Hg).48 The validity of the principles reported in Ref. 48 was confirmed by direct quantum-chemical calcula- tions.49 In the present review, we consider some problems of OS stabilisation in relation to new evidence on their stability and touch on the problem of the extreme valent possibilities of transition elements.The concept of `oxidation state stabilisation' is useful in this discussion because it provides ideas and mecha- nisms which allow researchers not only to perform experimental studies but also to interpret the results obtained. Let us formulate the general principles of this concept. II. A concept of oxidation state stabilisation It is generally assumed that chemical compounds contain elements in certain OS, which are stabilised by various forces. The character of chemical processes which result in a change in the oxidation state depends on the properties of the ion to be stabilised and the corresponding coordination environment.7, 50, 51 In the studies of OS stabilisation, the possibility of formation of compounds of a particular element in the particular OS and their stability are determined and the conditions for the synthesis of these com-366 pounds are searched for, i.e., techniques and procedures for the preparation of compounds with the predetermined property, viz., OS, are developed and factors that affect the stability of OS are revealed (for example, phenomenologically).Quite apparently, it is impossible to reveal the stabilisation factors without considering the problem of stabilisation of coor- dination polyhedra and crystalline phases. The strength of the corresponding chemical bonds and the stability of OS can be enhanced by stabilising coordination polyhedra.52 ¡À 56 Within the framework of the concept under consideration, it is assumed that the OS stabilisation implies an increase in the lifetime of the oxidised (reduced) form in an oxidant ¡À reductant dynamic system compared to the standard.This system is charac- terised by certain thermodynamic and kinetic parameters (for example, by the Gibbs potential, vapour pressure, the activation energy, the reaction order, etc.). To stabilise OS, one can vary these parameters or change the `thermodynamics' and/or `kinetics' for those that are more favourable for the increase in the lifetime of the oxidised (reduced) form. Let us consider the following transformation: PrIII (1) aq Praq IV +e¡¦aq.The PrIV aq aqua ion exists in aqueous media for *200 ns,57, 58 whereupon this ion is transformed into the PrIII aq aqua ion. When phosphotungstic acid is introduced into this dynamic system, a PrIV heteropolycompound [containing an ion of the probable composition PrIVW12O57 40 (see Ref. 59)] is formed and the lifetime of the `oxidised' form containing PrIV is increased to 10 ¡À 15 min. Consequently, a change in the coordination environment about the PrIV ion is accompanied by a change in its stability. We do not consider the kinetic characteristics of redox processes in the absence and in the presence of heteropolyacids. However, it is evident that a change in the nearest environment of the central ion in the complex is reflected in the kinetic parameters.The problem of stabilisation has two aspects. The first aspect is associated with the ways of attaining the desired oxidation state. The second aspect concerns the stability of this state. It is the provision of the stability of the OS over a period of time which is sufficient for its formation, identification and utilisation of the properties that receives primary attention. The stability of chemical compounds with respect to decom- position into elements is determined by the total energy of the ligand ¡À central atom system: (2) Etot=PEi+E0+E 00+Ec, where Ei is the bond energy of the ith molecular orbital of the isolated complex formed upon interaction of the electron system of the central atom with the ligand, E0 is the energy of stabilisation due to loss of degeneracy, E 00 is the energy of repulsion between the ligands and Ec is the energy of stabilisation due to establish- ment of the long-range order (crystal lattice).51 Analysis of Eqn (2) demonstrates that the stability of com- pounds is governed by a combination of properties of the central atom and the ligand environment as well as by the crystal structure.Many researchers have reached this conclusion in an empirical consideration of the problem (see, for example, Refs 4 and 5). The factors that affect the stability of OS are diverse in nature and generally act simultaneously and competitively.5 However, these factors can be, though rather arbitrarily, divided into thermodynamic (the first group) and kinetic (the second group) factors. The factors of the first group determine the possibility of formation of a particular ligand environment about the central atom (the formation of the short-range or long-range order).The factors of the second group affect the rate of formation of the desired environment. Taking into account the aforesaid, the following ways of stabilising OS can be distinguished. 1. Stabilisation through coordination. This stabilisation is realised in the chemical process owing to the formation of the nearest environment of the central atom in the complex, which is YuMKiselev, Yu D Tretiyakov more stable than those in free ions in the gaseous state or in solvated ions in solution. In this case, the ligands, the geometry, etc.are selected.7 2. Stabilisation by the crystal lattice, which is realised due to an increase in the coordination saturation of the central atom in the course of formation of crystals compared to the coordination saturation of ions in the gaseous state or in solution.7, 47 3. Kinetic stabilisation due to minimisation of kinetic barriers that appear in the course of formation of the thermodynamically more favourable nearest environment of the central atom.7 A particular technique for the stabilisation should be chosen taking into account the properties of both the reagents and the expected reaction products containing the ion (atom) to be stabilised. Thus, the stability of compounds in which the coordi- nation polyhedron about the `highly oxidised central atom' is formed from the most stable F7 and O27 ions, which are the strongest, difficultly oxidisable and difficultly deformable, are generally a maximum.Probably, only these ligands in the nearest coordination sphere of the central atom are suitable for stabilisa- tion of the high and the highest unstable OS. Therefore, com- pounds with oxide and fluoride coordination polyhedra are generally considered in studies on the stabilisation of the highest OS.Direct and inverse problems can be distinguished in the analysis of problems associated with the stability of valent forms. The problem is said to be direct if physicochemical character- istics of an oxidant ¡À reductant system, for example, redox poten- tials, are determined experimentally or calculated.The problem is said to be inverse if the conclusion about the stability of the oxidation state is made from a change in the selected parameters (properties) which can be related to OS in some series of compounds. These problems can in principle be solved because rather simple correlations exist between the energy characteristics of the ions to which particular valent forms are assigned. Actually, the quantum-mechanical parameters which charac- terise the electron density distribution and which are calculated taking into account only the electrostatic interactions between nuclei and electrons (the first-order perturbation theory) directly correlate with the oxidation states of the elements. This can be exemplified by the linear equation for the chemical shifts in the Mo�� ssbauer spectra (Fig.1).32, 60, 61 d �� a �¢ bx, where d is the chemical shift, a=1.834, b=70.456 and x is OS. There are rather simple correlations between the enthalpies of formation of halides MHaln and the enthalpy increments ofMn+ ions calculated according to the additive schemes.62 These corre- lations can be interpreted in a different way (see, for example, Ref. 63) because of the ambiguity of the approaches used by chemists or, more precisely, due to the fact that none of approaches can be extended indiscriminately to all compounds. d /mm s71 1.0 0.0 71.0+1 +2 +3 +4 +5 +6 +7 +8 OS Figure 1. Dependence of the chemical shift relative to sodium nitroprus- side on OS of the metal for iron derivatives.32, 60, 61The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements In the general case, the direct and inverse problems cannot be solved.This is associated with many factors but the main reason is that Mn aq aqua ions that exist in solutions are known only for a limited number of elements (generally, in the lowest OS). As the atomic number of the element is increased, the covalence of the M¡Ó L bond and the role of relativistic effects, including spin- orbital interactions, are increased, which is accompanied by the enhancement of the characteristic features of the ions more complex than aqua ions (for example, of actinyl ions 64).An analogous situation is observed as the OS is increased. The definition of the notion `oxidation state' is the key in the analysis of the problems associated with the concept under consideration. This problem, which is directly related to the `problem of valence,' is intricate. In particular, the problem of the relationship between the valence number, the effective charge and OS is still topical, although it has been discussed many times. This aspect of the chemical structure theory turned to be so contradictory that it has been suggested,65 without going into terminological details, to consider the valence only as a number, to retain its classical content and to give up attempting to treat all modern discoveries in the field of chemical bonds in the context of this notion.However, more recently (see, for example, Refs 66 and 67) attempts to calculate the valence numbers have been undertaken. No consideration is given to these and other calcu- lation procedures in the present discussion, but it should be noted that none of the available procedures is versatile. The measure of the valence or OS can be most readily determined by the number of gram-equivalents of the oxidising or reducing agent per gram-atom of the element (metal or metal- loid) which is either consumed or acquired when the element is transformed to the compound or vice versa. It is advantageous to take a one-equivalent oxidation or reduction process as the standard process, the choice of a particular one-equivalent con- version being of no importance.The above-described determina- tion of the valence number { is convenient because it is based on experimental grounds and is invariant with respect to the model of the electronic structure of the compound. We emphasise that the positive or negative value of OS does not imply that the corresponding charge is located on the atom. The determination of the valence on the basis of the number of gram-equivalents means the thermodynamic relations from which neither the spatial arrangement of the atoms nor characteristics of the bonds (for example, the bond order, which cannot be determined spectroscopically) can be directly obtained. The gram-equivalent is a complex parameter, which covers all inter- actions in the compound, and it cannot be represented either as a charge or as a number of bonds.The valence (or OS) is a classification notion used in crystal chemistry or for systematising educational data on inorganic chemistry. For example, Kossel's views were consistently used in the monograph 69 for tetrahedral systems. Chemical compounds of the element in the same oxidation state can be characterised by different electron density distributions, coordination polyhedra and the bond character. This is nong because specific electronic properties of atoms, which can be judged from the results of physical methods, are nothing more than a consequence of the manifestation of a particular valence. The above-considered determination of the valence number does not contradict the notions 71 of the electronic state fluctuation, processes of elec- tronic ordering and, consequently, fractional OS for transition elements.These are the properties of the element which exists in a { In this approach, the valence numbers coincide with the so-called Jorgensen's formal OS (or the Russian term `degrees of oxidation'). The application of the latter is improper because of the absence of charges on atoms. Crystal-chemists use `formal charges' on atoms in calculations 68, 69 and reject the reality of these charges by introducing the notion `effective charges.' The valence balance or valence strengths would be more properly used (this approach was used, for example, in Ref. 70). 367 particular OS or possesses a definite valence. These properties should be specially studied and systematised. The results of calculations of the enthalpies of formation of inorganic halides and oxides according to additive schemes provide certain support for the above approach to the determi- nation of the valence.62, 72 ¡Ó 75 Calculations of the enthalpies (energies) of inorganic halides and oxides are based on the classical structure theory, which has been conceptually refined by Tatevskii.76, 77 According to the known approximation of this theory, the macrocharacteristic, such as the self-energy, can be represented as a sum of increments (Ej) that characterise contributions of the so-called effective jth atoms with the valence number (n), j (n): (3) j n .The application of this equation to inorganic subjects is DU=PE limited. Equation (3) cannot be directly used even in the seemingly simple case, viz., in the case of compounds of various metals of the same stoichiometry, because one has to solve a system of k equations of the type (4) DUAxBy=xEAn +yEB [EA=F(n), EB=const] in k+1 unknowns.One more equation is required to obtain a solution. We found an equation of relationship,62, 73, 78 which is an empirical correlation between the enthalpy of formation of com- pounds MXn and the enthalpy of the following type: i DsHM, Mn : 1 PI I where Ii are stepwise ionisation potentials and DsHM is the enthalpy of sublimation of the metal. The equation takes the form of the following additive four- membered equation: (5) DfMXn =HMn +nHX+nHMn£¾X+yHX£¾X, where HM(n) is the enthalpy corresponding to the contribution of the central atom, HX, HMn7X and HX7X are the enthalpies corresponding to the contributions of the ligand X and of the M(n)7X and X7X bonds, respectively.Examples of selected dependences of the enthalpies of formation of chlorides and fluorides on the first term of Eqn (5) are given on Fig. 2 and in Table 1. DfHoMHal(gas) / kcal mol71 1 2W 80 Ta Zr 40 V Ti Zr V 0 La Be Sc Al Ti Ga 740 Al Li La Sc Be Ga 780 Li 100 0 DHM(1) /kcal mol71 Figure 2. Dependence of the enthalpy of formation of gaseous mono- halides on the partial enthalpy HM(1); (1) chlorides; (2) fluorides.368 Table 1. Coefficients of the linear dependence DfHo=A+BHM(n) for gaseous metal halides. Halide Metal MF MF2 MF3 MCl MCl2 MCl3 Na, K, Rb, Cs, Ag, Tl 7112.7 Li, Be, Al, Sc, Ti, V, Ga, Zr, Ta, La ± Lu 7117.4 Ca, Sr, Ba, Cr, Mn, Fe, Co, Ni, Cu, In 7126.3 Zn, Cd, Hg, Sn, Pb, Zr, Hf, Y 726.0 Mg, Ba, Ra, Zn, Cd, Hg, Yb Be, Cr, Mn, Fe, Co, Ni, Ga, Mo, Tc, Ru, 7193.6 Rh, Pd, Ag, In, Sn, Ta, W, Re, Os, Ir, Pt Sn, Pb, Zr, Hf Sr, Sc, La ±Lu 7217.9 Cu, Zn, Pb, Ru, Rh, Pd, Os, Ir, Pt 7141.8 Al, Ti, V, Ga, Y, Nb, Mo, Tc, La, Ta, W 7340.1 Al, Ga, In, Tl, Mn Al, Ti, Zr 7345.5 Sc, Y, La ± Lu 7302.7 Na, K, Rb, Cs, Ag, Tl, Mg, Ca, Sr, Ba 787.1 Li, Al, Be, Sc, Ti, V, Ga, Zr, Ta, La ± Lu 784.9 Cr, Mn, Fe, Co, Ni, Cu, In, Sn 7100.3 Zn, Cd, Hg, Sn, Pb, Zr, Hf, Y 5.0 Mg, Ba, Ra, Zn, Cd, Yb Cr, Mn, Fe, Co, Ni, Mo, Tc, Ru, Rh, 7129.5 Pd, Ag, In, Re, Os, Ir, Pt, Au, Tm Sn, Pb, Hf, Hg Be, Al, Sc, Ti, V, Ga, Y, Ca, Sr, Ba, Zr, Nb, 7144.4 Mo, Tc, Ru, Ag, La, Hf, Ta, W, Re, Ce ± Lu Cu, Zn, Rh, Pd, Os, Ir, Pt, Pb 777.3 Al, Ga, In, Tl Cr, Mn, Fe, Ir Sc, Y, La ±Lu Note.The characteristic groups, i.e., the sets of compounds of elements of the corresponding vertical groups in the short version of the Periodic system, are given in boldface type. DfH8=A+BHMÖnÜ . A similar situation is also observed for other halides and oxides. The so-called characteristic sets of compounds can be distin- guished in these linear dependences. The Group I elements are characteristic for monohalides (see Fig. 2), the Group II elements are characteristic for dihalides, etc. In other words, the character- istic sets of compounds of elements with the ith valence are comparable with the lists of the elements in the vertical groups of the short version of the Periodic system.Taking into account the additive equation (5) and the linear dependences of the enthalpies of formation of gaseous halides on the enthalpies HM(n), it can be concluded that grouping of compounds is dictated by the argument, while the character of the change in the function is determined by the remaining terms of the additive sum in Eqn (5). The relationship between the enthalpy increments is such that the first terms of the sum (5) have the maximum values. This makes it possible to construct the hierarchy which defines the energetics of compounds and which is based on the classical structure theory.76, 77 This theory assumes the application of the model of pairwise interactions.A certain energy increment can be related to each pairwise interaction (the procedure is described in the mono- graph 76). The values of the increments depend on the following characteristics: chemical individuality, sort, type, kind and vari- ety. On the whole, these characteristics comprise a `tree' of chemical states. The first characteristic is determined by the position of the element in the Periodic system. The second characteristic is associated with the valence. The third character- istic is connected with the geometry of the nearest environment of A B1.636 0.970 1.479 0 115.0 71.930 0.328 789.9 70.0999 0.252 0.252 0.486 744.1 70.280 0.0841 01.636 0.970 1.479 0 105.1 71.930 0.328 726.0 70.0999 0.252 0.252 57.8 70.280 0.239 0 7233.6 7170.2 YuMKiselev, Yu D Tretiyakov the central atom. The last two mentioned characteristics reflect finer details of the structure, for example, geometrical or optical isomerism.The characteristics of the sort (valence), which determine, in particular, clustering of halides (or oxides) in the dependences analogous to those shown in Fig. 2, can be assigned to the parameter HM(n). As a consequence, the valence is determined through the gram-equivalent, i.e., through the thermodynamic relations. 7 The oxidation states can be determined by direct (chemical) and indirect (physical) methods. Oxidimetric and electrochemical techniques were referred to as the chemical methods in Ref.45. Procedures for the analysis based on one-equivalent transitions of the types I7?(1/2)I2 and CeIII?CeIV as well as on processes described by a known number of equivalents (MnO74 ?MnII and Cr2O27?CrIII) are very popular. Of the above-mentioned pro- cedures, iodometric techniques are most widely used because of their simplicity, availability and versatility. For example, the applications of the iodometric analysis to fluoride-containing oxidants, which include compounds of d and f transition elements, and to platinum compounds were consid- ered.79,80 Fluorides of LnIV, CuII, CuIII, AgII, AgIII, AgV, AuV, NiIII, NiIV, MnIII, MnIV, CoIII, FeIII, FeIV, ClIII, BrIII, BrV, XeII, XeIV, XeVI, etc. were analysed. The iodometric procedure was used in the analysis of alkaline solutions of iron derivatives in the highest oxidation states (FeIV, FeVI and FeVIII).81 In this case, not only the OS of iron but also the concentration of the metal in solution was determined.The determination of the final state after titration is the key to oxidimetric (in particular, iodometric) procedures. Thus experi- ments with platinum compounds demonstrated 80 that Pt2.3+ was the final state in the titration of iodide complexes, which does not prevent the analysis. Evidently, the main drawback of oxidimetric procedures is the absence of selectivity. This calls for studies aimed at revealing reasons for redox transformations. For example, the analysis of alkaline solutions of iron in the highest oxidation state was accompanied by additional studies of solutions by MoÈ ssbauer spectroscopy, which have demonstrated that the iodometric data correspond to the valence transitions of iron.81 Much the same is true for electrochemical methods, viz., for polarography and coulometry. Cyclic voltammetry in combina- tion with coulometry is a convenient procedure for the determi- nation of gram-equivalents in metal complexes.82 For example, two reversible one-electron waves corresponding to the IrVI?Ir- V?IrIV transformations are observed in the cyclic voltammo- grams of reduction of the electrochemically oxidised hydroxo complexes of IrIV.83, 84 Electrochemical methods can be applied not only to solutions but also to melts.For example, these procedures were used for revealing the nature of the processes in salt melts and for determining conditions for stabilisation of the highest transition-element OS in these melts.85 Chemical methods for the OS determination are effective only in combination with physical methods.The most commonly used methods are magnetochemistry, ESR, electronic, MoÈ ssbauer and X-ray spectroscopy, X-ray photoelectron spectroscopy, X-ray diffraction analysis and methods of vibrational spectroscopy.45 To answer the question about the element OS, it is necessary to know, first, the correlations between OS and the parameter observed and, second, the reasons for deviations from these correlations if these deviations are observed or expected. For example, the number of outer electrons withdrawn from the electronic configuration of the free atom often corresponds to the element OS.In addition, the number of gram-equivalents of the oxidising (reducing) agent consumed in the oxidimetric titra- tion corresponds to the number of d (or f) electrons of the metal ion (for a series of important regularities, see Refs 86 ± 88). Below are considered procedures for the stabilisation of the highest OS of particular elements.The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements III. Rare-earth elements The known OS 89, 90 for rare-earth elements (REE, Ln) are listed in Table 2. For rare-earth elements, OS +3 is generally realised in a rich variety of their compounds; OS +2 belongs to the lowest OS and is not considered in the present paper because many studies were devoted to this problem and the results have been reviewed.23 Table 2.Known oxidation states of rare-earth elements.89, 90 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu OS +2 +3 ? ? + + ? + + ? ? + + + + + ? + + + + + + + + + + + + + + + +4 +++ + ++ + Note. Hereinafter, the following notations are used: +, this OS exists; ?, indirect evidence is available and reliable proof is absent. 1. Complexes of LnIV with organic ligands The problem of preparation of rare-earth complexes in high OS with organic ligands arose in the 20th century.91 ± 93 These com- pounds are often characterised by high stability constants and are very promising, for example, for the separation of rare-earth elements and actinides.94 ± 96 In the 1970s, it was established that intramolecular redox processes cannot be completely eliminated even in CeIV compounds (see, for example, Ref. 97), and the number of studies aimed at this particular application became fewer. However, it was reported 98 that the influence of these processes in rare-earth complexes with diphthalocyanines was a minimum, and it was concluded that these subjects contain LnIV ions throughout the entire lanthanide family.Nevertheless, a number of facts do not allow one to accept this interpretation of the data made in the work cited. In studies of transition metal complexes in the highest OS with carboxyl-containing chelating agents, hydroxy carboxylic acids, polydentate acyclic ligands, crown ethers and b-diketones, it was demonstrated that redox processes occur in all cases, though at different rates.97 In other words, intramolecular redox interactions can only be partially suppressed (but not completely excluded) by choosing an appro- priate ligand.Phthalocyanine can be readily reduced due to the presence of a large number of multiple bonds. Even if rare-earth complexes under consideration possess high stability, one cannot expect that intramolecular redox processes would be completely suppressed taking into account extremely high values of the redox potentials of the LnIII/LnIV pairs (for example, the values of E were estimated as equal to 2.83.0 V for Ln=Tb and Pr 99, 100). Apparently, additional supporting evidence for the existence of LnIV in diphthalocyanines is required.101 2.Complexes of LnIV with inorganic ligands CeIV=CeIII = The data on the behaviour of LnIV (Ln=Ce, Pr, Nd, Tb and Dy) in aqueous media are limited mainly by those for CeIV com- pounds. The CeIV ion is a rather strong oxidant (E 1.66 V99, 100). These ions are formed in the reactions with the participation of oxidants such as bromate (see, for example, Ref. 102) and are stabilised by various inorganic ligands with different efficiency. Interestingly, CeIV compounds containing chloride ions in the nearest coordination sphere exist although Cl7 ions are known as reductants. For example, a large series of hexachlorocerates(IV) M2 ICeIVCl6 (MI=NRá4 , PRá4 , AsRá4 , H3O+, Rb+ and Cs+) were prepared and their structures and stabilities were studied.103 ± 112 It was demonstrated 113 that the anionic [CeCl27 6 ] complex exists in aqueous media for a rather long period of time and an excess of chloride ions even somewhat stabilises this complex.The cesium salt (air-stable below 300 8C) is a convenient gravimetric form in the chemical analysis and separation of alkali metal cations.105 369 Data on the existence of TbIV and PrIV in aqueous media are scarce. Note that the information on the preparation of PrIV and TbIV acido complexes (nitrates, sulfates and chlorides) in aqueous solution was disproved.114 As mentioned above, reliable data on the PrIV ions were obtained by pulse radiolysis:57, 58 the short-lived absorption was assigned to the PrIV aqua ion.The PrIV ion is unstable due to the larger value of the PrIV/PrIII pair redox potential. The preparation of aqueous solutions of PrIV phospho- tungstate was reported.115 ± 117 The possibilities of studying this deep coloured complex are limited since this complex exists in a solution for only several minutes even at low concentrations. As a consequence, detailed information on this complex is absent. Tetravalent terbium (E TbIV=TbIII '2.8 V) is also stabilised in a salt of phosphotungstic acid.59, 117, 118 Its solutions can be stored at room temperature for several months. Phosphotungstic acid can be very efficiently used for stabilising strong oxidants, such as actinide ions, and the complexation leads to a decrease in the E value by *1.0 V.118, 119 The evidence on the existence of PrIV and TbIV ions in these compounds suggests that the above conclusion is true for lanthanides as well. 4 8 While on the subject of aqueous media, PrIV and TbIV hydroxides should be mentioned.Treatment of alkaline suspen- sions of TbIII hydroxide with such reagents as O3, S2O27 and IO7 is accompanied by the formation of TbIV hydroxide.120 Treatment of binary and complex PrIV and TbIV fluorides with alkaline solutions also resulted in the isolation of hydroxides.121 Note that the yield of LnIV hydroxide decreases as the stability of LnIV ions decreases. Thus when the corresponding fluoride complex was treated with cold alkali, TbIVO2 .nH2O was obtained in almost quantitative yield, the yield in the case of PrIV was substantially lower and DyIV hydroxide was formed only in trace amounts.121 To summarise, the stabilisation of LnIV in solutions is insuffi- ciently efficient due to a limited number of inorganic ligands which are able to decrease the redox potential of the system to a reasonable value. The stability of LnIV is substantially higher in the solid state than in solutions, which is manifested not only in the higher thermal stability of the solid compounds compared to that of the complexes in solutions but also in the broad assortment of the corresponding tetravalent ions 89, 122, 123 that occur in solid phases.Solid fluoride complexes of CeIV, TbIV, PrIV, NdIV, DyIV, TmIV and SmIV (the data on the last two mentioned complexes were reported in Refs 22 and 29) and binary CeIV, TbIV and PrIV fluorides (the synthesis of pure PrF4 was described for the first time in Refs 124 and 125) have been prepared to date. Binary oxides PrO27x and TbO27x (x<0.5) and complex compounds were detected in oxygen-containing systems.126, 127 A large num- ber of LnIV ions (PrIV, TbIV, NdIV and DyIV) are stabilised in BaCeO3,128 ± 130 which serves as a matrix. Stabilisation was performed under an atmosphere of oxygen at 1200 ± 1400 8C.128, 129, 131 Therefore, the stability of a metal ± ligand system in the solid phase (i.e., in the presence of long-range order) is higher than that of this system in solution.The coordination effects that occur in the solid state diminish the intermolecular metal ± ligand distances in the nearest coordination sphere about the central atom of the complex compared to the internuclear distances in the binary compounds,132 which also favours the stabilisation of the LnIV ions. The currently available evidence allows one to consider the regularities of stabilisation of the highest OS of lanthanides. Analysis of the stability of LnIV compounds based on the published data 29, 43, 133 ± 137 allowed conclusions 29, 137 that the efficiency of LnIV stabilisation can be judged from only two groups of data. The first group includes the data on redox potentials or ionisation potentials of lanthanide atoms.However, this information does not allow reliable conclusions about the stability of LnIV because ELnIV=LnIII are calculated by empirical methods and only the data on CeIV derivatives provide trust- worthy basis for these calculations. The experimental E values370 for other LnIV/LnIII pairs cannot be obtained due to the high oxidising capacity of LnIV, which does not allow one to construct an equilibrium electrochemical cell. The solution of the direct problem of the stability presents difficulties because the initial data are scarce. The series of the relative stabilities of oxidation states 122, 123 obtained on the above basis reflect only roughly the actual situation. Therefore, an alternative method which consists in the solution of the above-mentioned inverse problem is used for obtaining data on the stability of LnIV.The thermodynamic characteristics of LnIV fluorides with composition Cs3LnF7 were reported.135 The thermal stabilities of the same fluorides were also studied by methods of non- isothermal kinetics. The thermodynamic values were determined according to procedures of differential thermal analysis,138 which have been developed and tested in special experiments, with the use of a specially designed sealed nickel cell whose sensitivity is close to that of the cell of a differential scanning calorime- ter.139 ± 141 These procedures make it possible to obtain data on the enthalpies of fluorination reactions. The most reliable values of the enthalpies and Gibbs energies of the reactions (6) M3LnIVF7 M3LnIIIF6+12 F2 for all rare-earth elements (Table 3) can be obtained on the basis of the available data 135 ± 137 and results of calculations of the Born ± Haber cycle.135, 142 ± 144 Table 3.Results of calculations of the enthalpies of process (6) (kJ mol71) according to the Born ± Haber cycle.135 Ln Ln 7DrH8 DU 7DrH8 DU 3809 3827 3852 3904 3986 3998 Tb Dy Ho Er Tm Yb 331 164 90 31 768 7184 3608 3656 3730 3753 3774 3793 Ce Pr Nd Sm Eu Gd 239 95 22 58 136 47 It was confirmed 135, 142 ± 144 that in the LnF4 series the cerium compound is the most stable one, this is followed by the terbium and praseodymium derivatives in order of decreasing stability.The thermal stability of the M3LnF7 complexes decreases in the following order: Ce>Tb>Pr>Dy.137, 145, 146 The thermodynamic stability of the complexes with respect to elimination of fluorine according to reaction (6) decreases on going from Ce to Sm: Ce>Tb>Pr>Dy>Tm>Nd>Sm. The stability of the highest OS (+4) of lanthanides decreases in the same order. The regularity which is observed in the lanthanide series and is referred to as the `tetrad effect' has been refined.29, 137 In the series of the rare-earth elements, `nodal points' exist in which the probability of changes in the composition, structure, the type of phase diagram and some other parameters is a maximum. The lanthanide family is divided into four segments (from whence the name `tetrad' follows).This effect is most pronounced in the middle of the 4f series, which corresponds to the `gadolinium break.'29, 147, 148 This effect is most clearly manifested in the atomic properties (ionisation potentials, electron affinity, Gibbs energy, enthalpies of gaseous ions, etc.). The tetrad effect was attributed to the specific change in the electrostatic term of the exchange interaction of the f electrons in the Ln series.145 The thermodynamic data (DH and DG) for LnIV fluorides reproduce the tetrad effect.29, 135 However, the distinguishing feature is that this effect is of different importance for the middle elements of each half of the series.This effect is more pronounced in the middle of the second half of the series. As a consequence, the YuMKiselev, Yu D Tretiyakov stability of TmIV is anomalously high compared to that of SmIV due to the effect of the condensed state. Actually,137, 149 the difference between the lattice energies of LnIIIF3 and LnIVF4 or ofM3LnIVF7 andM3LnIIIF6 (DU) and the differences between the corresponding enthalpies of formation (DH) are related to the fourth ionisation potentials of the lanthanide atoms (I4) by the following equation: DH8+DU+const=I4. The DU values in the lanthanide series change `smoothly' (see Table 3) so that these values for the second half of the 4f series are substantially larger than those for the first one.Since the depend- ence of I4 on the atomic number of the element is nonlinear and is governed by the tetrad effect, the DH(DG) values should be changed analogously. This implies that the enthalpies (Gibbs energies) of formations of TmIV compounds should be larger in magnitude than the corresponding values for the SmIV com- pounds. 3. The effect of complexation A large number of tetravalent rare-earth elements are stabilised in cesium fluoro complexes. The higher stability of the complexes compared to the corresponding binary compounds has been discussed many times (see, for example, Refs 29 and 42). In the solid phase, the stabilisation occurs due to the positive charge of the outer-sphere cation, which compensates for the negative charges on the complex anions.43 The absence of the outer-sphere cations in the structure of a compound leads to a distortion of the coordination polyhedron of the central atom as a result of which the overall atomic system is destabilised.For example, the structure of AuF5 is built of octahedra linked to each other and twisted into a helix.150 In AuV complex fluorides, there is no need for such twist and the anion has an octahedral shape.151 4. Valence possibilities of lanthanides Efforts were made to extend the series of lanthanides which could be transformed into the highest OS and to reveal the possibilities of attaining OS higher than +4 by experimental meth- ods.121, 124, 137, 140, 141 In the solution of the first problem, reactions were found which are characterised by minimal kinetic barriers to processes of formation and the maximum possibility of the LnIII?LnIV transformation.44 The reaction that proceeds according to the following general equation: Cs3LnF7+3.5 Xe+3 Cl2.Cs3LnCl6+3.5 XeF2 is most suitable. This reaction was successfully used for the preparation of Cs3TmF7; TmIV was identified by physicochemical methods. Experimental evidence (the data from differential thermal analy- sis) for the existence of SmIV was obtained.44 The solution of the second problem was considered using praseodymium, which exhibits the most promise for preparing the highest valent forms, as an example.137 Various high- and low- temperature fluorination reactions of binary and complex chlor- ides, fluorides, nitrates and sulfates under diverse reaction con- ditions (with variations of the temperature and the duration of the reaction, in the presence or in the absence of a solvent, etc.) were studied.The most powerful oxidants, including fluorine, krypton difluoride, etc., were used. It appeared that in all cases fluorina- tion afforded only PrIV compounds. Taking into account this fact and based on the theoretical grounds, it was concluded 29, 137 that +4 is the extreme OS for 4f elements due to strong localisation of 4f states in the atomic core of lanthanides and the presence of serious energy hindrances to the subsequent ionisation. These 4f elements differ sharply from 5f elements. In conclusion, is should be noted that OS +4 for the rare- earth elements are also stabilised in other solid systems, in particular, in perovskites BaTbO3, SrCeO3,152 Ba17ySryPrO3, BaCeyPr17yO3 and SryBa17yPrO3,153 oxide superconductorsThe problem of oxidation state stabilisation and some regularities of a Periodic system of the elements Pb2Sr2Ce17xCaxCu3O8154 and Pb2Sr2RCu3O8 (R=Ce, Pr, Tb or Am155) and manganites Y17xPrxSr2Cu2.85Re0.15O7+d,156 Pr17x(Ca,Sr)xMnO3,157 Pr0.6Ca0.4MnO3158 and Pr0.6(Ca,Sr)0.4..MnO3.159 IV. Transition elements Of all transition metals, the Group VIII elements, silver, gold and mercury are of most interest because researchers faced serious problems when preparing their highest valent forms. 1. Highest valent forms of iron Until recently, it was assumed that the maximum valence of iron is no higher than +6 and this fact was even substantiated theoret- ically.Quantum-chemical calculations for Fe, Ru and Os tetr- oxides 160 demonstrated that FeO4 is characterised by high populations of the atomic states, which excludes the formation of strong Fe7O bonds. However, this fact does not necessarily indicate that this compound, though unstable, cannot be pre- pared. For iron, an OS of +5 was observed only in one compound, viz., in La2LiFeO6, which was prepared at high pressure and was well studied.161 The seemingly well-known ferrates(VI) continue to attract the attention of researchers. For example, the mechanism of decomposition of ferrate(VI) was considered in Ref. 162. At the same time, FeIV compounds remain poorly studied.The processes that occur in alkaline media upon anodic dissolution of metallic iron under conditions of high current density were studied in detail.31, 32, 163 ± 178 The conditions for formation of ferrate(VI) were found,31, 32, 168 ± 173 qualitative notions of the mechanism of this process were developed and explanations were offered for the decomposition of concentrated ferrate(VI) solutions to form iron(III) hydroxide. Abnormal prop- erties of ferrate solutions were observed. The detailed studies of these properties allowed conclusions to be made about the existence of ferrates(IV) in alkaline media (previously, it was believed that these compounds are unstable in aqueous solutions) as well as about the occurrence of FeO4.31, 32 These compounds were identified by electronic spectroscopy of solutions, MoÈ ssba- uer spectroscopy and some other methods.The justified assump- tions were made that ferrate(VI) may undergo disproportionation and thermal stimulation of the latter afforded FeVIII and FeIV derivatives. Iron tetroxide is very unstable. It can be isolated in a relatively pure form by extraction from solutions of ferrates in CCl4. The extracts were distilled at 30 8C to give products in yields no higher than 25% ± 30%. These properties were used in the preparation of highly pure iron.179 ± 181 Solutions of FeVIII undergo autocatalytic decomposition to FeIII hydroxide and can exist only when the concentration of iron is lower than 0.01 mol litre71.The MoÈ ss- bauer spectra (77 K) of aqueous ± alkaline systems containing FeVIII have a singlet line belonging (judging from the value of the chemical shift d) to the highest valent form of iron.32 Ferrate(VI) in an aqueous ± alkaline medium is so stable that its solutions under normal conditions can be stored for at least two years. This compound is formed not only in the course of the above-mentioned disproportionation but also under prolonged action of oxygen on alkaline solutions of FeIII hydroxo com- plexes.178 ± 181 The data on FeO4 essentially supplement the known regular- ities of the Periodic system and, in addition, make it possible to establish a qualitative correlation between the stability of the compounds and the occupancies of the orbitals determined by quantum-chemical methods. 2.Copper, silver and gold The data on OS of copper, silver and gold are given in Table 4. For this group of elements, OS5+3 are unstable, although, strictly speaking, even OS of +2 is unstable. For example, one of the manifestations of the effect of variable valence, viz., the fluctua- 371 Table 4. Oxidation states of copper, silver and gold. Au Ag Cu OS +(see a) +(see a, b) +(see c, d) +(see c, d) +(see a) ? (see b) +(see a) ? (see b) +(see d) +(see a) +(see b, d) +(see c, d) ? (see b) +(see d) +1 +2 +3 +4 +5 a Known in many compounds. b Undergoes disproportionation. c For oxides. d For fluorides. tion of the charge state, is observed for CuII ions.Apparently, this results in such a value of the redox potential E CuII=Cu that the CuII ion can be considered as an oxidant. The problem of the existence of the highest valent forms for 1=2AgIII 1=2F6 Cu, Ag and Au is still not completely solved. This is particularly true for gold. The first AuV compound, viz., the fluoro complex with the bulky Xe2Fá11 cation, was prepared only in 1972.182 Since that time, some other AuV compounds have been synthesised. In particular, AuF5 and fluoro complexes with cations of alkali and alkaline-earth metals were prepared. However, no compounds with ligands other than fluoride were found.183 The questions of whether AuIV can be isolated and whether OS higher than +5 can be realised are still open.The same is true for silver compounds. Thus, only the AgV derivative with composition Cs2AgV is known.184 considered problems These many have been times.150, 151, 185 ± 197 New binary (AgF3) and complex fluorides of CuIII, AgI, AgII, AgIII, AuIII, AuV and AgIII,V with various outer- sphere cations were synthesised and studied. Many of the syn- thetic reactions of these compounds have been examined and the factors that affect the yields of the products and, consequently, the stabilisation of the corresponding OS have been revealed.197 Finally, data which can be interpreted in the context of dispro- portionation of AuV were obtained in studies of thermolysis products of AuV fluoro complexes.150 In fact, there is no direct evidence for the presence of a disproportionation product, viz., of a fluoride more enriched with fluorine than AuF5, in the gaseous phase.The data on the preparation of CuIV as Cs2CuF6 and CuIV oxide were reported in Refs 198 and 199, respectively. The latter experiments were carried out under high oxygen pressure. The oxide contains CuIII and CuIV stabilised in the perovskite lattice with composition La17xSrxCuO3 (04x40.25). The formation of mixed-valence states is typical of Cu, Ag and Au both in the fluoride and oxide systems. In oxide superconduc- tors and related compounds, these states of copper are of great importance. In superconductors, the OS of copper varies from +1 to +2.33. In this case, the p-orbitals of oxygen atoms make significant contributions to the corresponding molecular orbitals.The mixed-valence states 200, 201 are responsible, for example,202 for a `charge-stabilised' distortion of NaBa2Cu3O6. Of the recent works in this field, let us note Refs 203 ± 205. The unusual properties of gold oxides in oxide films were described.206, 207 In this connection, recall the effects associated with the dynamics of the electron density in the solid phase.71 The unit cell can contain metal ions in different OS that occupy crystallographically non- equivalent positions, metal ions in different OS that occupy equivalent positions and, finally, metal ions in the so-called fluctuating state where fast oscillations (which are equivalent to electron transitions between different OS) of electron density occur at each lattice centre.The existence of the mixed-valence state is an indication of valence instability. This problem as well as the factors favourable for disproportionation were considered.48 The existence of copper in the mixed-valence state (CuIII, IV) in oxides is indicative of a decrease in the stability of OS +4 of copper compared to that observed in the fluoride due, apparently,372 to the presence of the essential p-component of the metal ± ligand bond in the oxide. Taking into account the existence of copper in the CuIII/CuIV state, it is believed that +4is the extreme oxidation state for copper. 3. Heavy platinum elements Difficulties emerge in the realisation of the highest OS for heavy 5d transitions elements (Table 5).This is particularly true for Ir, Pt, Au and Hg. It can be seen that the maximum oxidation state attained for platinum metals is equal to +6, which is, apparently, far from the extreme value. The main and most reliable informa- tion on the properties of the strongly oxidised ions of the above- mentioned elements concern predominantly their fluorides, although scattered data on oxide systems have been recently reported. Thus attempts were made to synthesise iridium compounds in the highest OS under high oxygen pressure (it is known 211, 212 that high pressure favours attainment of unusual OS of transition elements).208 ± 210 The properties of a new series of oxides with the distorted K2NiF4 structure of composition A2MIrVIO6 (A=Ba or Sr;M=Sr, Ca,Mgor Zn) containing iridium in OS of+6were reported.208, 209 The structural and magnetic properties of the Sr2RhIVO4 and Sr2IrIVO4 compounds, which are analogues of La2CuO4, were studied.213 X-Ray and neutron diffraction studies demonstrated the distortion of the ideal structure of the K2NiF4 type. Oxoplatinates with composition Ln2MPtIVO6 (Ln=La, Pr, Nd, Sm, Eu or Gd; M=Mg, Co, Ni or Zn) were prepared.214 Note also the electrochemical studies of hydroxo complexes of iridium in OS of +4, +5 and +683, 84 and studies aimed at preparing platinum and iridium compounds in the highest OS in alkaline solutions.215 ± 219 Solutions of hydroxo complexes of platinum and iridium were treated with ozone and the resulting compounds were studied by various methods, including ESR, electronic absorption spectroscopy, cyclic voltammetry and chem- ical analysis.It was established that the platinum hydroxo com- plexes afforded the previously unknown platinum superoxo complexes [Pt(OH)5(O72 )]27, [(OH)5Pt(m-O72 )Pt(OH)5]37 and [(OH)4Pt(m-O72 )(m-OH)Pt(OH)4]27. In the case of iridium, hydroxo complexes of iridium in OS of up to +6 inclusive were detected in solution along with the peroxo and superoxo com- plexes of iridium of the [(OH)5IrIV(m-O72 )IrIII(OH)5]57 {or [(OH)5IrIII(m-O72 )IrIII(OH)5]57} type.218 This suggests that OS of +6 for iridium can be attained even under rather mild conditions (in solution), which indicates that treatment of solu- Table 5.Known oxidation states of heavy transition elements. Os Re OS W +(see a) +(see a) +(see a) +1 +(see c) +(see b) +(see b) +(see d) +(see a) +(see c) +2 +(see f, g) +(see d) +(see a, b) +(see d) +(see c) +(see g) +3 +(see c, d) +(see b, e) +(see c) +(see c) +(see c) +4 +(see c) +(see c) +(see c) +(see d) +(see e. f) +(see f) +5 +(see f, g) E (see d) +(see d) +(see e, f) +(see e, f) +(see i) +6 ? (see i) E (see d) +(see d) +7 ? (see i, j) +8 E (see e, h) +9 ? (see k) +10 7(see j) Note. The following notations are used: E, the highest OS, which is, apparently, the extreme one; `minus sign,' this OS is absent. a Known, for example, in carbonyls and complex phosphines. bKnown, for example, in halides and cyanides. c Known in many compounds.dKnown in oxides and halides. e For oxides. f For fluorides. g Disproportionation occurs. h For hydroxo complexes. iA very unstable higher fluoride with an unknown composition formed upon heating of fluoro AuV complexes. j See the discussion in the text. k The formation of this OS in a compound formed upon b-decay of 191OsO4 was reported. YuMKiselev, Yu D Tretiyakov tions of hydroxo complexes with oxidants under more drastic conditions has considerable promise. V. Matrix oxidation state stabilisation Recently, the so-called method of matrix OS stabilisation has been developed. This method is based on the fact that the tracer atoms incorporated into the crystalline matrix are interchangable with the metal atoms of the base upon heating.Thermal treatment is apparently one of a few procedures used in this case because the mechanism of transfer of the atoms is obviously of a diffusion character and the diffusion is enhanced at higher temperatures. The efficiency of the exchange reaction is associated with the possibility of the change in the size of tracer atoms or of a change in the electronic configuration (and, consequently, in OS) of the above-mentioned atoms, i.e., with the lability of the correspond- ing electron cloud. The crystal field of the matrix `forces' the doping ions to behave as the electron cloud in the crystal lattice, which is in some cases accompanied by charge transfer from one metal atom to another,220 i.e., by a change in OS.The geometrical criteria (ionic radii, local symmetry, etc.) are of great importance because they reflect the effect of the thermo- dynamics of the mixed matrix ± tracer system. The stability of OS should be determined to a large extent by the isomorphism rules because the geometry of the vacancy into which a transition metal ion is inserted plays a great role in the stabilisation of the latter. The effect of the matrix on the properties of the doping ions can be illustrated by a change in the colour of CrIII ions placed at the matrix nodes with different local symmetry. In corundum (a-Al2O3), beryl (Al2Be3Si6O18) and kyanite (aluminosilicate, Al2SiO5), the CrIII ions are red, green and blue, respectively.221 Shifts of the absorption bands in energy depend on various factors, including the coordination number, and the absorption bands corresponding to the charge transfer depend on the valence transitions or on the stability of OS.For example, a change in OS of cerium in the same matrix, viz., in ZrO2 (fianites), leads to a change in the colour from bright red-claret (CeIV) to yellow- orange (CeIII).222 Many researchers use the matrix stabilisation of the highest OS in spite of the fact that the lattice contributions to the total energy are insignificant. This method was used for preparing laser crystals doped with transition metal ions (see, for example, Refs 223 and 224). However, note that it is difficult to obtain metal ions in high OS in these subjects because of their specificity.Hg Au Pt Ir +(see c) +(see c) +(see j) 7(see j)The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements In matrices which are more suitable for stabilisation (for example, in BaMIVO3, where M=Ce, Th or Zr), PrIV, TbIV, NdIV and DyIV were obtained.128 ± 130 Interestingly, it is possible to obtain NdIV in all the above-mentioned matrices, while DyIV was obtained only in BaCeIVO3. The MnIII,224 CrIV and CrV as well as FeVI ions 225, 226 are stabilised in matrices of garnets, chromates and sulfates (the systems were prepared from solution) and TbIV ions are stabilised in (SrMgF4 : TbIV)-phosphors (prepared in the solid phase).227 In studies of matrix oxide systems involving oxoferrates and oxoruthenates,33, 34, 228 ± 236 the conditions for doping these matri- ces with Fen+ and Con+ ions (OS were established by MoÈ ssbauer spectroscopy with the use of 57Fe and 57Co isotopes) were found.Of the most interesting results, note the stabilisation of CoVI ions in the ferrate Cs2FeVIO4 (see Refs 33 and 34) and the stabilisation of FeVI ions in the mixed-valence compound with composition Cs3(RuVI 1=2RuVII 1=2O4).232, 235 Attempts to attain the FeVII state are of particular interest. 4 This OS can be realised for Fe. Thus it was concluded 165 that the disproportionation of FeO27 ions (in alkaline media) occurs according to the following scheme: FeVIII+FeIV. 2FeVI Since a two-equivalent exchange occurs, the intermediate one- electron conversion with the participation of FeVII could be expected in accordance with Scheffer's principle.It was attempted to solve the problem of stabilisation of this OS by placing 57Fe in the Cs3(RuVI 1=2RuVII 1=2O4) matrix. In this ruthenate, the iron ion is stabilised primarily in the FeVI form. However, an unstable satellite was also observed in the slope of the intense line in the MoÈ ssbauer spectrum. The chemical shift value of this satellite corresponds to that expected for FeVII.232, 236 Unfortunately, there is no more convincing evidence for the presence of FeVII in this system. According to the data of MoÈ ssbauer spectroscopy (4 K),229, 237 ± 241 the FeVII state can also be stabilised in the Na7Fe7O system at a molar ratio [Na] : [Fe]>50 : 1 (at lower ratios, only FeIV compounds are formed 242).Unlike other researchers, the authors of the publica- tions 229, 237 ± 242 applied oxidation processes using sodium and cesium peroxides under rather mild conditions. The latter are suitable for the realisation of high OS if the strongest oxidants (Na2O2 and CsO2) as well as an efficient stabilising agent (in particular, Cs+ cations) are present in the reaction mixture.43 In addition, peroxides as such can stabilise the highest OS of, for example, iron and cobalt on condition that their contents in the reaction mixtures are low.229, 241 Highly reactive reagents (pre- cursors) which are formed, for example, upon ultrasonic treat- ment show promise.It is by this procedure that specimens of highly disperse Fe2O3 were prepared.243 The procedures for the preparation of the corresponding precursors have been summar- ised.244 VI. Valence possibilities of transition elements 4 It was hypothesised that the heaviest 5d elements can attain high OS,48 which can be realised for Ir, Pt, Au and Hg in the tetrahedral oxo ions MOn 4 . Estimates of the energies of molecular orbitals which were obtained by the scattered wave Xa method for a series of tetraoxo anions MVIIO7 of the third-row elements 245, 246 indicate that rather stable compounds can be formed. Quantum- chemical and thermodynamic calculations were performed for the tetrahedral MOq4 ions isoelectronic with OsO4 (M=Ir, Pt, Au or Hg, q=+1, +2, +3 or +4, respectively).49 Quantum-chemical calculations were carried out with the use of quasirelativistic and relativistic pseudopotentials (cf.Ref. 247), which are believed to be very reliable. Ab initio calculations for the IrIXOá AuVIIO7 minima. For the PtXO2á and AuIXOá 4 4 , IrVIIO74 , PtVIIIO04 and 4 ions demonstrated 49 that the potential surfaces have 4 cations, minima are also 373 observed but they are apparently too shallow to expect the formation of more or less stable compounds. An analogous conclusion can be drawn on the basis of the results of thermody- namic calculations. The main outcome of the study 49 was the conclusion that transition elements in very high OS (which have not yet been realised), up to +10 inclusive, can theoretically occur.4 4 It was demonstrated 49 that the probability of the existence of HgVIIIO04 is distinct from zero because calculations gave a reason- able value for the internuclear Hg7O distance of 2.12 A (HgO2á ions and all the more so, HgO4á ions are unstable). Probably, mercury compounds in OS of lower than +8 are rather more stable. For mercury and its analogues (Zn and Cd), +2 is still considered to be the highest OS. Note that the short-lived complex of trivalent mercury with composition [Hg(cyclam)](BF4)2 was reported,248 although these data require additional proof. It was suggested that an unstable intermediate containing HgIII is formed upon fluorination of organomercurials with xenon difluoride.249, 250 Based on the results of X-ray photoelectron spectroscopy of mercury copper ± oxide superconductors with composition HgBa2(Ca0.86Sr0.14)2Cu3O8+d, it was concluded that these superconductors contain HgIII.251, 252 On the basis of the general regularities of the Periodic system, Jorgensen also believed that the possibility exists of attaining high OS for mercury.37 He reasoned that the existence of I(5d) of *14 eV for mercury(II) fluoride may favour oxidation of HgII to HgIII or HgIV, particularly taking into account that HgF2 crystallises in the fluorite structure with a low lattice energy.The realisation of the 5d 8 configuration and consequently of HgIV in the square-planar HgF4 molecule is more probable than the 5d9 configuration and HgIII.In this connection, the publication 38 in which the molecular and electronic structure and the stability of HgF4, ZnF4 and CdF4 were compared is of fundamental importance. The possibility of stabilisation of HgF4 upon anionic complexation was considered. Calculations were performed with the use of the same pseudopo- tentials as those used previously 247 for HgF2. It was shown that the fluoride HgF4 may well exist in the gaseous state.38 To the contrary, the existence of CdF4 and all the more ZnF4 is highly improbable. In addition, it follows that complexation does not enhance stabilisation of HgIV with respect to HgII in anions.38 The conclusion that complexation affects only slightly the stability of HgIV is unexpected because this is considered as one of the ways of stabilising the highest OS of f and d transition elements.39, 40 Since HgIV should exhibit the properties of d transition elements, the statement that the formation of the HgFx7 4áx anionic complexes has a favourable effect on the stability of HgIV should be fulfilled.The introduction of an equivalent amount of cations into the system containing anions decreases substantially the total energy of the system. In this case, the mutual cation ± anion influence should occur and stabilisation should be realised through the electrostatic forces. In particular, this is consistent with the results of calcu- lations of the electronic structures of theMVF76 ions (M=Hf, Ta, W, Re, Os, Ir, Pt, Au or Hg) 253 by the spin-polarisation discrete variation Xa method.The change in electron affinity in the HfF76 ?HgF76 series as well as the experimental data on the stability of the complex fluorides of OsV, IrV, PtV and AuV (compared to the data on the corresponding MF5) indicate that the effect of complexation does exist. Thus the temperature of decomposition of complex fluoroaurates(V) is 100 8C higher than that of AuF5.45, 150, 192 Hence, the conclusion that coordination has no effect on the stability of higher mercury fluorides is questionable although it was not disproved (HgF4 or other HgIV fluorides have not been obtained yet despite attempts, as can be judged from the published data,38 aimed at their synthesis).Weemphasise that in the above-mentioned theoretical studies, only the presence of the minimum on the potential surface was noted, while the depth of this minimum can be determined only374 experimentally. Thus derivatives of CsIII7CsIX and krypton oxides { have not yet been obtained, although calculations predicted that they can exist.255 Apparently, this is also true for HgF4. VII. Regularities of changes in the extreme oxidation states of transition elements The highest OS of transition elements (according to the exper- imental data and results of quantum-chemical calculation) 89 are shown in Scheme 1. The extreme OS (E) are separated from the higher OS by a thick line, which assumes that the latter cannot be attained. It can be seen that the values of the extreme OS in the series of transition metals increase, reach a maximum and then decrease. This character of the change in the maximum OS is one of the main regularities of the Periodic system.For some OS, the experimental data are lacking but there is reason to hope that such information will be obtained. In relation to the aforesaid, it is necessary to make some comments. For the first-row transition metals, the maximum OS is observed for iron (+8). The existence of MnVIII in tetraoxide MnO04 , which is formed upon photodissociation of theMnO74 ion, is suggested on the basis of spectroscopic data.256 Probably, this conclusion is erroneous. For cobalt, +6 is the highest OS.33, 34 This state was identified by emission MoÈ ssbauer spectroscopy.The detection of OS by this method may indicate that it is highly unstable. In view of this fact, it is believed that +6 is the extreme OS for cobalt. For Ni, +4 is the highest OS attained. It is assumed that NiVI can be realised (handbooks even contain the E value for the NiVI/NiII pair). In our opinion, the existence of NiVI is still an open question. The preparation of NiV is more probable. More- over, the identification of the OS in this case is a very complicated problem. The stability of NiVI and NiV derivatives should be rather low and probably the only suitable technique for observing OS is MoÈ ssbauer spectroscopy which can detect the chain of transformations 57Ni?57Co?57Fe. The detection of particular OS in studies by emission MoÈ ss- bauer spectroscopy [provided that radiative transitions (known as post-effects) are absent] is very revealing.In fact, for the A?B process where compound B has OS which can be observed by MoÈ ssbauer spectroscopy, aA*1 and aB=1078 (a is the activity). The latter value can readily be obtained on the assumption that the activity of the source was *1 mCi (57Co). Then keq=aB/aA *1078 and DrG=RTlnkeq is about 50 kJ mol71, i.e., it is chemical calculations, this OS can be attained.49 c The possibility of formation of derivatives upon fluorination of organomercurials at low temperature was reported.249, 250 impossible to obtain macroscopic amounts of a compound con- taining the desired OS. For the second-row transition metals (see Scheme 1), the maximum OS is attained for Ru(+8).Compounds containing RhVI and PdV were described. On the basis of the published data,246 both states can hardly be assigned to the limiting states. In this connection, the question of whether higher OS are possible remains open. In the third-row transition elements, the maximum OS for iridium is +948, 49 (apparently, this is the extreme OS), while OS +10 can be attained for platinum.49 The probability of occurrence of the nona- and decavalent states for the elements is an open question. According to Abegg's rule, each element may have a set of OS consisting of no more than 8 values. It is known 37 that this rule is fulfilled for the main-group elements and for the post-transition elements.Only four exceptions to this rule are known, namely, Mn, Ru, Re and Os, for which ten OS can be realised.37 Assuming that OS of +10 and +9 can be attained for platinum and iridium, respectively, and thus postulating sets of 10 OS for these elements, we extend the series of the above `exceptions.' Without consider- { It was reported that a krypton compound with the orthotellurate ligand was presumably obtained though under cryogenic conditions.254 This compound exploded even under these conditions. More detailed data in support of the individuality of this compound and of the presence of the Kr7O bond are lacking. YuMKiselev, Yu D Tretiyakov Scheme 1 Me- Oxidation state of the metal tal 2+ 3+ 4+ 5+ 6+ 7+ 8+ 9+ 10+ E Sc E Ti E V E Cr E Mn E Fe E Co E Ni H ?? Cu E Zn E E E ?? H E Y E Zr E Nb E Mo Tc Ru Rh Pd H Ag E Cd see b La E Hf E Ta E W E Re E Os E Ir ?? ?(see a) H ?? Pt see b H ?? ?? Au see b see b H Hg H see c see b The following notations are used: H, the high OS is known, but it is unclear whether it is the extreme value; ??, the data are lacking, but there is the probability that this state can be attained.a The data were not confirmed. b According to the results of quantum- ing the nature of this phenomenon (which is, of course, governed not only by the electrostatics of metal ± ligand interactions), we emphasise that the above `exception' for transition elements is apparently the rule.Evidently, the presence of a particular set of OS for the element as well as the maximum valence are determined primarily by the principle of the arrangement of the Periodic system (the so-called `Aufbau' principle; this follows from the ideology of the publication 37) which is based on thermodynamic factors, primarily on the energies of the detachment of electrons from the free atom. This can be judged, for example, from the values of the ionisation potentials 257, 258 of atoms } realising that it is the differences between the ionisation potentials, which can characterise the tendency for OS stabilisation, rather than the absolute values of these potentials that are of importance. Analysis of the ionisation potentials demonstrated that it is necessary to overcome an energy barrier the value of which is } An analogous procedure was used in the analysis of the regularities of the changes in the ionisation potentials of atoms of the heavy transition elements.48 Speculative conclusions obtained on this basis were subse- quently confirmed by direct quantum-chemical calculations.49The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements reflected in the stability of the OS, on going from one OS to another.These barriers can be overcome in chemical (or in physical) processes. The latter is illustrated in the observation of a high OS for the metal atoms upon b-decay of 191Os,259 Am and Bk in oxides,15, 16 which provides short-lived IrIXOá4 , CmVIO22á and CfVOá2 ions, respectively.In other words, the existence of OS +9 or +10 for some elements does not contradict the principles of construction of the Periodic system. It was of interest to verify the regularities of the changes in the extreme OS and the observations of OS higher than the extreme values. On the one hand, the fundamental regularity as declared above seemingly does not allow for the formation of these OS. On the other hand, these OS are not theoretically forbidden and only energy barriers occur, which can be overcome in some cases, as demonstrated above.} Presently, there is no point in going through all the details because the experimental data are lacking. However, attempts can, in principle, be undertaken to obtain this information, for example, in experiments on bombardment of solid and liquid targets with heavy highly charged ions.For example, calcium targets were bombarded with Ar12+ ions,262, 263 helium was treated with C6+, N7+, O8+, Ar16+, I16+ and Xe30+ ions264 and the recombination of Kr34+ ions on hydrogen molecules was considered.265 Evidently, the appearance of an almost bare nucleus in the matrix should cause a cascade of Auger electrons. These secondary processes are known and are being studied. For example, the manifestations of the secondary processes in the Ka spectra which were observed in experiments on bombardment of sulfur targets with Ar5+ and Xe8+ ions were analysed.266 Actually, the problem of OS stabilisation was not solved in the papers cited because ions and targets evidently were not optimum (in this context, the above-discussed studies on the matrix OS stabilisation are of importance).The questions are how long these superhigh OS live and whether these states can be detected by the available physical methods. To summarise, the present review considers the state-of-the- art of the problem of the existence and stabilisation of the highest and extreme valent forms of transition elements. In the discussion of these problems, many authors (see, for example, Refs 23 and 29) attempted to reveal the relationship between the electronic structures of the atoms and the properties of the compounds. Evidently, this relationship can be also revealed in the present review. However, d transition elements (particularly, heavy ele- ments) are characterised by a high degree of covalence, which makes these correlations poorly informative.Hence, we restrict ourselves to the discussion of the chemical evidence and the statement of empirical regularities. The advantage of the empirical method is its independence from the electronic structure and the inherent nature of the subjects under consideration. This is of particular importance when the data are scarce as in the case of elements in the highest unstable OS which are few in number. At the same time, the discussion of the general situation on the basis of the available reliable quantum-chemical and experimental data allows us to offer the hypothesis that superhigh and short-lived OS exist and to suggest the procedure for its verification. The review has been written with the financial support of the Program `Russian Universities' and the Russian Foundation for Basic Research (Projects Nos 96-03-33285 and 99-03-32488) } This question for main-group elements was considered by Sidorov,260 who has suggested that BeF3 and AlF4 can be obtained. However, more recent studies demonstrated 261 that attempts to obtain the above com- pounds and, consequently, OS higher than the extreme values (+2 for Be and +3 for Al) in chemical processes failed.375 References 1. F A Cotton,G Wilkinson Advances Inorganic Chemistry (New York: Wiley, 1988) 2. N N Greenwood, A Earnshaw Chemistry of the Elements (Oxford: Pergamon Press, 1984) 3.J C Bailar, H J Emeleus, R S Nyholm, A F Trotman-Dickenson (Eds) Comprehensive Inorganic Chemistry Vol. 1 ± 5 (Oxford: Pergamon Press, 1973) (Eds) Comprehensive Inorganic Chemistry Vol. 1 ± 5 (Oxford: Pergamon Press, 1973) 4. J Kleinberg Unfamiliar Oxidation States and Their Stabilisation (Lawrence: University Kansas, 1950) 5. B E Douglas J. Chem. Educ. 29 119 (1952) 6. R S Nyholm Proc. Roy. Soc. (London) 273 (1961) 7. R Keller, R Perry, in The Chemistry of the Coordination Compounds (Eds J C Bailar, D H Busch) (New York: Reinold Publ., 1956) Part 4 8. N Bartlett, F O Sladky, in Comprehensive Inorganic Chemistry Vol. 1 (Eds J C Bailar, H J Emeleus, R Nyholm, A F Trotman- Dickenson) (Oxford: Pergamon Press, 1973) p.213 9. NMEdelstein (Ed.) Actinides in Perspective (Oxford: Pergamon Press, 1982) 10. E K Hulet Radiochim. Acta 32 23 (1983) 11. N N Krot, A D Gel'man Dokl. Akad. Nauk SSSR 177 124 (1967) a 12. N N Krot, V P Shilov, V B Nikolaevskii, A K Pikaev, A D Gel'man, V I Spitsyn Dokl. Akad. Nauk SSSR 217 589 (1974) a 13. N N Krot, A D Gel'man, M P Mefod'eva, V P Shilov, V F Peretrukhin, V I Spitsin Semivalentnoe Sostoyanie Neptuniya, Plutoniya, Ameritsiya (Heptavalent State of Neptunium, Plutonium, Americium) (Moscow: Nauka, 1977) 14. V F Peretrukhin, C H Delegard, in Proceedings of the 5th Interna- tional Conference on Radioactive Waste Management of Environ- mental Remediation. ICEM'95 Vol. 1 (Ed.S Slate) (New York: American Society Mechanic Engineering, 1995) p. 425 15. V F Peretrukhin, E A Erin, V I Dzyubenko, V V Kopytov, V G Polyukhov, V Ya Vasil'ev, G A Timofeev, A G Rykov, N N Krot, V I Spitsyn Dokl. Akad. Nauk SSSR 242 1359 (1978) a 16. V V Avrorin, R N Krasikova, V D Nefedov,M A Toropova Radiokhimiya 27 511 (1985) b 17. V N Kosyakov, E A Erin, V M Vityutnev, V V Kopytov, A G Rykov Radiokhimiya 24 551 (1982) b 18. G Bouissieres, B Jouniaux, Y Legoux, J Merinis, F David, K Samhoun Radiochem. Radioanal. Lett. 45 121 (1980) 19. K Leary, N Bartlett Rev. Chim. Miner. 13 82 (1976) 20. P Sorbe, J Grannec, J Portier, P Hagenmuller C. R. Hebd. Seances Acad. Sci., Ser. C 284 231 (1977) 21. R Hoppe Isr. J. Chem. 17 48 (1978) 22. Yu M Kiselev, S A Goryachenkov, L I Martynenko, V I Spitsyn Dokl.Akad. Nauk SSSR 278 126 (1984) a 23. N B Mikheev, L N Auerman, I A Rumer, A N Kamenskaya, MZ Kazakevich Usp. Khim 61 1805 (1992) [Russ. Chem. Rev. 61 990 (1992)] V I Spitsyn Izv. Akad. Nauk SSSR, Ser. Khim. 818 (1982) c R A D'yachkova Dokl. Akad. Nauk SSSR 216 1154 (1981) c Radiokhimiya 22 316 (1980) b V G Korshunov Radiokhimiya 28 82 (1986) b (1976) b Aktinidov (Electronic Structure of Actinides) (Moscow: Nauka, 1986) 24. N B Mikheev, L N Auerman, I A Rumer, R A D'yachkova, 25. V I Spitsyn, N B Mikheev, L N Auerman, I A Rumer, 26. R A D'yachkova, L N Auerman, N B Mikheev, V I Spitsyn 27. N B Mikheev, V I Spitsyn, G V Ionova, L N Auerman, 28. N B Mikheev, V I Spitsyn, G V Ionova Radiokhimiya 18 550 29.G V Ionova, V G Pershina, V I Spitsyn Elektronnoe Stroenie 30. Yu M Kiselev Zh. Neorg. Khim. 39 1266 (1994) d 31. Yu M Kiselev, N S Kopelev, V I Spitsyn, L I Martynenko Dokl. 32. Yu D Perfil'ev, N S Kopelev, Yu M Kiselev, V I Spitsyn Dokl. 33. N S Kopelev, L A Kulikov, Yu D Perfil'ev, Yu M Kiselev Zh. 34. N S Kopelev, L A Kulikov, Yu D Perfiliev Hyperéne Interact. 90 Akad. Nauk SSSR 292 628 (1987) a Akad. Nauk SSSR 296 1406 (1987) a Neorg. Khim. 40 838 (1995) d 377 (1994)376 35. G V Ionova, V G Pershina, G A Gerasimova Zh. Neorg. Khim. 41 1190 (1996) d 36. I I Zvara Ross. Khim. Zh. 40 (6) 53 (1996) e 37. C K Jùrgensen Z. Anorg. Allg. Chem. 540/541 91 (1986) 38. M Kaupp,M Dolg, H Stoll, H G von Schnering Inorg. Chem.33 2122 (1994) 39. K B Yatsimirskii Bull. Pol. Acad. Sci., Chem. 42 299 (1994) 40. K B Yatsimirskii Proc. Am. Chem. Soc., Symp. Ser. 565 207 (1994) 41. R W Parry Proc. Am. Chem. Soc., Symp. Ser. 565 320 (1994) 42. V I Spitsyn, Yu M Kiselev, L I Martynenko Zh. Neorg. Khim. 31 2764 (1986) d 43. Yu M Kiselev Zh. Neorg. Khim. 40 817 (1995) d 44. Yu M Kiselev Koord. Khim. 23 (2) 73 (1997) f 45. Yu M Kiselev Zh. Neorg. Khim. 43 2010 (1998) d 46. Yu M Kiselev Priroda (11) 86 (1996) 47. Yu M Kiselev Zh. Neorg. Khim. 43 759 (1998) d 48. Yu M Kiselev Zh. Neorg. Khim. 41 1067 (1996) d 49. A I Dement'ev,M V Kuznetsov, Yu M Kiselev Zh. Neorg. Khim. 42 1167 (1997) d 50. L Pauling The Nature of the Chemical Bond and the Structure of Molecules and Crystals.An Introduction to Modern Structural Chem- istry (New York: Cornell Univ. Press, 1947) 51. I B Bersuker Elektronnoe Stroenie i Svoistva Koordinatsionnykh Soe- dinenii (Electronic Structure and Properties of Coordination Com- pounds) (Leningrad: Khimiya, 1986) 52. A J Downs, in New Pathways in Inorganic Chemistry (Eds E A V Ebsworth, A G Maddock, A G Sharp) (Cambridge: University Press, 1968) 53. C K Jùrgensen Oxidation Numbers and Oxidation States (Berlin: Springer, 1969) 54. P Hagenmuller Proc. Indian Acad. Sci. 92 1 (1983) 55. P Hagenmuller, in Studies in Inorganic Chemistry Vol. 3 (Ed. R Metselaar) (Amsterdam: Elsevier, 1983) p. 49 56. P Hagenmuller Z. Chem. 23 1 (1983) 57. M Faraggi, A Feder J. Chem. Phys.56 3294 (1972) 58. A K Pikaev, G K Sibirskaya, V I Spitsyn Dokl. Akad. Nauk SSSR 209 1154 (1976) a 59. A S Saprykin, V P Shilov, V I Spitsyn, N N Krot Dokl. Akad. Nauk SSSR 226 853 (1976) a 60. F Menil J Phys. Chem. Solids 46 763 (1985) 61. N S Kopelev, inMoÈssbauer Spectroscopy Sophisticated Oxides (Eds A Vertes, Z Homonnay) (Budapest: Akademia Kiado, 1997) p. 305 62. Yu M Kiselev, A I Popov Additivnyi Raschet Ental'pii Obrazovaniya Gazoobraznykh Ftoridov i Khloridov Metallov, Moskva, 1984 (Additive Calculation of Enthalpy of Formation of Gaseous Fluorides and Chlorides of Metals, Moscow, 1984); article deposited at the VINITI No. 766 (1984) 63. R S Drago Physical Methods in Chemistry (Philadelphija: WB Saunders, 1977) 64. G V Ionova, G A Gerasimova Usp. Khim.61 1656 (1992) [Russ. Chem. Rev. 61 907 (1992)] 65. E M Shustorovich Khimicheskaya Svyaz' (Chemical Bond) (Moscow: Nauka, 1973) 66. G P Shveikin, V A Gubanov, A L Ivanovskii, R N Pletnev, E Z Kurmaev, A V Dmitriev Dokl. Akad. Nauk SSSR 288 1383 (1986) a 67. V I Nefedov, V I Vovna Elektronnaya Struktura Khimicheskikh Soedinenii (Electronic Structure of Chemical Compounds) (Moscow: Nauka, 1987) p. 307 68. M A Porai-Koshits Prakticheskii Kurs Rentgenostrukturnogo Analiza (Practical Course of X-Ray Analysis) (Moscow: Moscow State University, 1960) Vol. 2 69. E Parthe Elements of Inorganic Structural Chemistry (Wien: KSP, 1990) 70. V A Efremov, Doctoral Thesis in Chemical Sciences, Moscow State University, Moscow, 1993 71.D I Khomskii Usp. Fiz. Nauk 129 443 (1979) g 72. Yu M Kiselev, V Ya Leonidov, in Tez. Dokl. Xth Simpoz. po Khimii Neorganicheskikh Ftoridov (Abstracts of Reports of the Tenth Sym- posium on the Chemistry of Inorganic Fluorides) (Moscow: Moscow State University, 1998) p. 79 YuMKiselev, Yu D Tretiyakov 73. Yu M Kiselev Termodinamika Stabilizatsii Sostoyanii Okisleniya Metallov v Tverdykh Telakh. Analiz Ental'pii Obrazovaniya Binarnykh Galogenidov, Oksidov i Nekotorykh Kompleksnykh Galogenidov, Moscow, 1986 (Thermodynamics of Stabilisation of Oxidation States of Metals in Solids. Analysis of Enthalpy of Formation of Binary Halides, Oxides and Some Complex Halides, Moscow, 1986); article deposited at the VINITI No. 8903-2 (1986) 74.Yu M Kiselev, N S Kopelev Additivnyi Raschet Ental'pii Obrazova- niya Oksidov Metallov, Moscow, 1986 (Additive Calculation of Enthalpy of Formation of Metal Oxides); article deposited at the VINITI No. 8433 (1986) 75. Yu M Kiselev, A I Popov, N S Kopelev, V I Spitsyn, in Tez. Dokl. XI Vsesoyuz. Konf. po Kalorimetrii i Khimicheskoi Termodinamike (Abstracts of Reports of the XIth All-Union Conference on Calorimetry and Chemical Thermodynamics) (Novosibirsk: Nauka, 1986) p. 42 76. V M Tatevskii Klassicheskaya Teoriya Stroeniya Molekul i Kvantovaya Mekhanika (Classical Theory of Structure of Molecules and Quantum Mechanics) (Moscow: Khimiya, 1973) 77. V M Tatevskii Stroenie i Fiziko-Khimicheskie Svoistva Molekul i Veshchestv (The Structure and Physicochemical Properties of Molecules and Substances) (Moscow: Moscow State University, 1993) 78. A I Popov, N S Kopelev, Yu M Kiselev, V I Spitsyn Dokl.Akad. Nauk SSSR 298 1416 (1988) a 79. A I Popov, N E Fadeeva, Yu M Kiselev Zh. Anal. Khim. 43 465 (1988) 80. Yu M Kiselev, D A Pankratov Vestn. Mosk. Univ., Ser. 2, Khim. 36 (1) 51 (1995) i 81. Yu M Kiselev, N S Kopelev, N A Zav'yalova, P E Kazin Zh. Neorg. Khim. 34 2199 (1989) d 82. G A Mabbott J. Chem. Educ. 60 697 (1983) 83. I N Kiseleva, N A Ezerskaya, in Tez. Dokl. XIV Vsesoyuz. Chernyaevskogo Soveshchaniya po Khimii, Analizu i Tekhnologii Platinovykh Metallov (Abstracts of Reports of the XIVth Chernyaev All-Union Meeting on the Chemistry, Analysis and Technology of Platinum Metals) (Novosibirsk: Nauka, 1989) Vol.1, p. 144 84. N A Ezerskaya, I N Kiseleva, I F Golovaneva, A P Klyagina Koord. Khim. 16 1127 (1990) f 85. S A Kuznetsov, P T Stangrit Zh. Prikl. Khim. 66 2702 (1993) j 86. V I Nefedov Stroenie Molekul i Khimicheskaya Svyaz' (Itogi Nauki i Tekhniki) [The Structure of Molecules and Chemical Bond (Advances Science and Engineering Series)] (Moscow: Izd. VINITI, 1973) Vol. 1 87. V I Nefedov Stroenie Molekul i Khimicheskaya Svyaz' (Itogi Nauki i Tekhniki) [The Structure of Molecules and Chemical Bond (Advances Science and Engineering Series)] (Moscow: Izd. VINITI, 1975) Vol. 3 88. V I Nefedov Rentgenoelektronnaya Spektroskopiya Khimicheskikh Soedinenii (X-Ray Spectroscopy of Chemical Compounds) (Moscow: Khimiya, 1984) 89.Yu M Kiselev Vestn. Mosk. Univ., Ser. 2, Khim. 38 (1) 3 (1998) i 90. N Ya Turova Spravochnye Tablitsy po Neorganicheskoi Khimii (Information Tables on Inorganic Chemistry) (Moscow: All-Russian Chemical College, Russian Academy of Sciences, 1997) 91. V V Serebrennikov Khimiya Redkozemel'nykh Elementov (Skandii, Ittrii, Lantanidy) [The Chemistry of Rare-Earth Elements (Scandium, Yttrium, Lanthanides)] (Tomsk: Tomsk State University, 1959) Vol. 1 92. K A Gschneidner (Ed.) Handbook on the Physics and Chemistry of Rare-Earth Vol. 1 (New York: Interscience, 1978) 93. Gmelin Handbook of Inorganic Chemistry (Berlin: Springer, 1983) Pt. D6B C7B 94. Yu M Kiselev, V I Spitsyn, N I Pechurova, in Tez. Dokl. Vsesoyuz.Konf. po Primeniyu Ekstraktsionnykh i Sorbtsionnykh Metodov Raz- deleniya Lantanidov i Aktinidov (Abstracts of Reports of the All- Union Conference on Application of Extraction and Sorption Meth- ods of Separation of Lanthanides and Actinides) (Moscow: Nauka, 1984) p. 58 95. Q Su, Z Wu, C Gao, in Value Adding Solvent Extraction Vol. 1 (Eds D C Shallcross, R Paimin, LMPrvcic) (Melbourne: University Melbourne, 1996) p. 605 96. T H Con, in Value Adding Solvent Extraction Vol. 1 (Eds D C Shallcross, R Paimin, LMPrvcic) (Melbourne: University Melbourne, 1996) p. 565The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements 97. V I Spitsyn, L I Martynenko (Eds) Koordinatsionnaya Khimiya Redkozemel'nykh Elementov (Coordination Chemistry of Rare- Earth Elements) (Moscow: Moscow State University, 1979) p.132 98. L G Tomilova, K M Dyumaev, O P Tkachenko Izv. Akad. Nauk, Ser. Khim. 425 (1995) c 99. L J Nugent, R D Baybarz, J L Burnett, R L Ryan J. Inorg. Nucl. Chem. 33 2503 (1971) 100. L J Nugent MTP Int. Rev. Sci., Ser. 2 2 195 (1975) 101. Yu M Kiselev, in Proceedings of the XXIst Rare-Earth Research 102. M F Barakat,M M Abdel-Hamid J. Radioanal. Nucl. Chem. 211 103. Yu M Kiselev, A Brandt, L I Martynenko, V I Spitsyn Dokl. 104. A Brandt, Yu M Kiselev, L I Martynenko Z. Anorg. Allg. Chem. 105. A Brandt, Yu M Kiselev, L I Martynenko, V I Spitsyn Dokl. 106. A Brandt, Yu M Kiselev, L I Martynenko Zh. Neorg. Khim. 26 927 107. A Brandt, Yu M Kiselev, L I Martynenko Zh.Neorg. Khim. 26 Conference (Abstracts of Reports), Duluth, 1996 p. 44 461 (1996) Akad. Nauk SSSR 246 879 (1979) a 474 233 (1981) Akad. Nauk SSSR 257 1189 (1981) a (1981) d 2950 (1981) d 108. A Brandt, Yu M Kiselev, L I Martynenko Zh. Neorg. Khim. 28 2806 (1983)d 113. Yu M Kiselev, I Yu Filatov, L I Martynenko, V I Spitsyn Izv. 109. Yu M Kiselev, A I Popov Zh. Neorg. Khim. 28 348 (1983) d 110. V M Agre, Yu M Kiselev Zh. Neorg. Khim. 28 597 (1983) d 111. Yu M Kiselev, A Brandt, G V Ionova, V G Pershina, in Tez. Dokl. XIV Vsesoyuz. Chugaevskogo Soveshch. po Khimii Koordinat- sionnykh Soedinenii, Ivanovo, 1981 (Abstracts of Reports of the XIVth Chugaev All-Union Meeting on the Chemistry of Coordi- nation Compounds, Ivanovo, 1981) (Moscow: Nauka, 1981) p.687 112. Yu M Kiselev, V I Spitsyn, L I Martynenko, in Tez. Dokl. VI Vsesoyuz. Konf. po Khimii i Tekhnologii Redkikh Shchelochnykh Metallov, Ashkhabad, 1984 (Abstracts of Reports of the VIth All- Union Conference on the Chemistry and Technology of Rare Alkali Metals, Ashkhabad, 1984 (Moscow: Nauka, 1984) p. 119 Akad. Nauk SSSR, Ser. Khim. 7 (1985) c 114. I G Bogdanovich, N I Pechurova, V I Spitsyn, L I Martynenko Zh. Neorg. Khim. 15 2145 (1970) d 115. B F Myasoedov, I A Lebedev Radiokhimiya 19 87 (1977) b 116. A S Saprykin, V I Spitsyn, N N Krot Dokl. Akad. Nauk SSSR 228 649 (1976) a 117. V N Kosyakov, G A Simakin, A A Baranov, G A Timofeev Radiokhimiya 19 360 (1977) b 118. V M Nikolaev, G A Timofeev, V Ya Vasil'ev, I I Kapshukov, E A Karelin, N S Kosulin Issledovaniya v Khimii i Khimicheskoi Tekhnologii Transuranovykh Elementov (Preprint NIIAR, No.5) [Investigation of the Chemistry and Chemical Technology of Transuranium Elements (Preprint of Research Institute of Nuclear Reactor, No. 5)] (Moscow: Tsniiatominform, 1986) p. 686 119. T Isobe, Yo Okaue Kidorui 28 240 (1996); Chem. Abstr. 126 69 323 (1997) 120. I G Bogdanovich, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1971 121. Yu M Kiselev, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1976 122. L B Asprey, B B Cunningham, in Progress in Inorganic Chemistry Vol. 2 (Ed. F A Cotton) (New York; London: Interscience, 1961) p.267 123. B B Cunningham, in Rare Earth Researches (Proceedings of the XVIIth International Congress on Pure and Applied Chemistry) (Ed. F Kleber) (New York: McMillan, 1961) p. 127 124. V I Spitsyn, Yu M Kiselev, L I Martynenko, V N Prusakov, V B Sokolov Dokl. Akad. Nauk SSSR 219 621 (1974) a 125. Yu M Kiselev, V B Sokolov Zh. Neorg. Khim. 29 857 (1984) d 126. R G Haire, L-R Eyring, in Handbook on the Physics and Chemistry of Rare-Earth Vol. 18 (EdsKAGschneidner, LREyring, G R Choppin, G H Lander) (New York: Elsevier, 1994) 127. R G Haire J. Alloys Comp. 223 185 (1995) 128. G Brauer, H Kristen Z. Anorg. Allg. Chem. 456 41 (1979) 129. G Brauer, H Kristen Z. Anorg. Allg. Chem. 462 35 (1980) 377 130. A J Jacobson, B C Tofield, B E F Fender, in Proceedings of the Xth Rare-Earth Research Conference (Abstracts of Reports), Arizona, 1973 p.134 131. V B Gromov, N Yu Konstantinov, A R Kaul, A S Strelkov, Yu M Kiselev, in Proceedings of the IInd International Symposium on Solid State Chemistry (Abstracts of Reports), Praga, 1989 p. 121 132. V I Spizin, L I Martynenko, J M Kiselev Z. Anorg. Allg. Chem. 495 39 (1982) 133. V V Nikulin, S A Goryachenkov, M V Korobov, Yu M Kiselev, L N Sidorov Zh. Neorg. Khim. 30 2530 (1985) d 134. A I Popov, M N Brekhovskikh, V A Fedorov, Yu M Kiselev Zh. Neorg. Khim. 34 1021 (1989) d 135. Yu M Kiselev, N E Fadeeva,M V Korobov, V I Spitsyn Izv. Akad. Nauk SSSR, Ser. Khim. 1012 (1989) c 136. Yu M Kiselev, S A Goryachenkov, N E Fadeeva, L I Martynenko, in VII Vsesoyuz. Simpozium po Khimii Neorgani- 137.YuMKiselev, Doctoral Thesis in Chemical Sciences, Moscow State cheskikh Ftoridov (Tez. Dokl.) [The VIIth All-Union Symposium on the Chemistry of Inorganic Fluorides (Abstracts of Reports) (Moscow: Nauka, 1984)] p. 178 University, Moscow, 1992 138. J Paulik, F Paulik, in Comprehensive Analytical Chemistry Vol. 12 (Amsterdam: Elsevier, 1981) 151. A I Popov, M A Val'kovskii, Yu M Kiselev, N A Chumaevskii 139. Yu M Kiselev, S A Goryachenkov Zh. Neorg. Khim. 28 16 (1983) d 140. Yu M Kiselev, S A Goryachenkov, L I Martynenko Zh. Neorg. Khim. 28 1153 (1983) d 141. Yu M Kiselev, S A Goryachenkov, L I Martynenko Zh. Neorg. Khim. 29 69 (1984) d 142. Yu M Kiselev Radiokhimiya 25 463 (1983) b 143.Yu M Kiselev Energii Reshetki i Termodinamicheskie Kharakteris- tiki Oksidov i Ftoridov Redkozemel'nykh Elementov i Aktinidov, Moscow, 1985 (Lattice Energies and Thermodynamic Character- istics of Oxides and Fluorides of Rare-Earth Elements and Actinides, Moscow, 1985); Article deposited at the VINITI, No. 4416 (1985) 144. Yu M Kiselev, N E Fadeeva, in VIII Vsesoyuz. Simpozium po Khimii Neorganicheskikh Ftoridov (Tez. Dokl.) [The VIIIth All- Union Symposium on the Chemistry of Inorganic Fluorides (Abstracts of Reports)] (Moscow: Nauka, 1986) p. 184 145. Yu M Kiselev, S A Goryachenkov, L I Martynenko, V I Spitsin Dokl. Akad. Nauk SSSR 278 122 (1984) a 146. USSR P. 1 312 073; Byull. Izobr. (19) 100 (1987) 147.G V Ionova, V G Vokhmin, V I Spitsyn Zakonomernosti Izmene- niya Svoistv Lantanidov i Aktinidov (Regularities of the Variation of the Properties of Lanthanides and Actinides) (Moscow: Nauka, 1990) 148. S Bratsch, J J Lagowski J. Phys. Chem. 89 3310 (1985) 149. Yu M Kiselev, in Proceedings of the 2nd International Conference on f-Elements (ICFE-2) (Abstracts of Reports), Helsinki, 1994 p. 185 150. Yu M Kiselev, A I Popov, A V Goryunov, N A Chumaevskii, L N Savinova, V B Sokolov, S N Spirin Zh. Neorg. Khim. 35 611 (1990) d Zh. Neorg. Khim. 37 1902 (1990) d 152. J Dumschat, R Lubbers, I Felner, G Lucazeau, G Wortmann J. Phys. IV 7 1019 (1997) 153. Yu Hinatsu J. Solid State Chem. 119 405 (1995) 154. U Staub, L Soderholm, S Skanthakumar,M R Antonio J.Phys. IV 7 1077 (1997) 155. L Soderholm, S Skanthakumar, U Staub,M R Antonio, C W Williams J. Alloys Compd. 250 623 (1997) 156. M Murugesan, P Selvam, M Sharon, L C Gupta, R Pinto, M S R Rao, C S Gopinath, S Subramanian J. Appl. Phys. 80 6884 (1996) 157. A Maignan, C Martin, B Raveau Z. Phys. B, Condens. Matter 102 19 (1997) 158. S Das, P Mandal Z. Phys. B, Condens. Matter 104 7 (1997) 159. F Damay, A Maignan, C Martin, B Raveau J. Appl. Phys. 82 1485 (1997) 160. I A Topol', V P Vovna, M V Kazachek Teor. Eksp. Khim. 23 456 (1987) k 161. B Buffat, G Demazeau,M Pouchard, L Fournes, J-M Dance, P Hagenmuller C. R. Hebd. Seances Acad. Sci., Ser. C 292 509 (1981)378 162. K Bouzek, I Rousar J. Appl. Electrochem. 27 679 (1997) 163.Yu M Kiselev, V I Spitsyn, in Tez. Dokl. VIII Vsesoyuz. Chugaev- skogo Soveshch. po Khimii Koordinatsionnykh Soedinenii (Abstracts of Reports of the VIIIth Chugaev All-Union Meeting on the Chemistry of Coordination Compounds) (Moscow: Nauka, 1987) Part 1, p. 105 164. Yu M Kiselev, G V Ionova, A A Kiseleva, N S Kopelev, A P Bobylev, A B Yatskevich, V I Spitsyn Dokl. Akad. Nauk SSSR 293 1407 (1987) a 165. Yu M Kiselev, N S Kopelev, V I Spitsyn Dokl. Akad. Nauk SSSR 295 882 (1987) a 166. N S Kopelev, Yu D Perfil'ev, Yu M Kiselev, in Yaderno-Spek- troskopicheskoe Issledovanie Sverkhtonkikh Vzaimodeistvii (Tez. Dokl. II Vsesoyuz. Konf.) [Nuclear Spectroscopic Investigation of Hyperfine Coupling (Abstracts of Reports of the Second All-Union Conference)] (Moscow: Moscow State University, 1987) p.134 167. Yu M Kiselev, N S Kopelev, Yu D Perfil'ev Zh. Neorg. Khim. 32 2982 (1987) d 168. Yu M Kiselev, N S Kopelev, A P Bobylev Zh. Neorg. Khim. 34 2654 (1989) d 169. N S Kopelev, Yu D Perfil'ev, Yu M Kiselev Zh. Neorg. Khim. 35 1793 (1990) d 170. N S Kopelev, Yu D Perfiliev, Yu M Kiselev J. Radioanal. Nucl. Chem. 157 401 (1992) 171. N S Kopelev, Yu D Perfiliev, Yu M Kiselev J. Radioanal. Nucl. Chem. 162 239 (1992) 172. N S Kopelev, Yu M Kiselev, V I Spitsyn, in Tez. Dokl. XVI Vse- soyuz. Chugaevskogo Soveshch. po Khimii Koordinatsionnykh Soe- dinenii (Abstracts of Reports of the XVIth Chugaev All-Union Meeting on the Chemistry of Coordination Compounds) (Moscow: Nauka, 1987) Part 1, p.327 173. N S Kopelev, Yu M Kiselev, in Proceedings of the XXVIIIth International Conference on Coordination Chemistry, Gera, 1990 Vol. 2, p. 70 174. USSR P. 1 303 553; Byull. Izobret. (14) 108 (1987) 175. USSR P. 1 318 530; Byull. Izobret. (23) 76 (1987) 176. USSR P. 14 147 83; Byull. Izobret. (29) 97 (1988) 177. USSR P. 1 567 655; Byull. Izobret. (20) 112 (1990) 178. USSR P. 1 759 959; Byull. Izobret. (33) (1992) 179. N S Kopelev, Yu M Kiselev, V I Spitsyn, in Tez. Dokl. VIII Vse- soyuz. Konf. po Metodam Polucheniya i Analiza Vysokochistykh Soedinenii (Abstracts of Reports of the VIIIth All-Union Confer- ence on the Methods of Synthesis and Analysis of High-Purity Compounds) (Moscow: Nauka, 1988) Part 2, p. 155 180.A A Kamnev, B B Ezhov, N S Kopelev, Yu M Kiselev, Yu D Perfiliev, in The XXVIth Colloquium Spectroscopic International, Soéa, 1989 (Ed. D L Tsalev) Vol. VIII, p. 77 181. USSR P. 1 604 863; Byull. Izobret. (42) 132 (1990) 182. K Leary, N Bartlett J. Chem. Soc., Chem. Commun. 903 (1972) 183. N Bartlett, K Leary Rev. Chim. Miner. 13 82 (1976) 184. P Sorbe, J Grannec, J Portier, P Hagenmuller J. Fluor. Chem. 11 243 (1978) 185. A I Popov, Yu M Kiselev, V F Sukhoverkhov, V I Spitsyn Dokl. Akad. Nauk SSSR 296 615 (1987) a 186. A I Popov, Yu M Kiselev Zh. Neorg. Khim. 32 2276 (1987) d 187. A I Popov, Yu M Kiselev Zh. Neorg. Khim. 33 965 (1988) d 188. A I Popov, Yu M Kiselev, A A Timakov, K P Bukharin Zh. Neorg. Khim. 33 1252 (1988) d 189. Yu M Kiselev, A I Popov, K P Bukharin, A A Timakov, M V Korobov Zh.Neorg. Khim. 33 3205 (1988) d 190. A I Popov, Yu M Kiselev Zh. Neorg. Khim. 33 2289 (1988) d 191. A I Popov, Yu M Kiselev Zh. Neorg. Khim. 33 2294 (1988) d 192. Yu M Kiselev, A I Popov, V B Sokolov, S N Spirin Zh. Neorg. Khim. 34 434 (1989) d 193. Yu M Kiselev, A I Popov,M V Korobov, L N Savinova, V B Sokolov, S N Spirin Zh. Neorg. Khim. 34 2240 (1989) d 194. USSR P. 1 281 518; Byull. Izobret. (1) 75 (1987) 195. USSR P. 1 281 519; Byull. Izobret. (1) 75 (1987) 196. USSR P. 1 281 517; Byull. Izobret. (1) 75 (1987) 197. A I Popov, Yu M Kiselev, in Tez. Dokl. XVI Vsesoyuz. Chugaev- skogo Soveshch. po Khimii Koordinatsionnykh Soedinenii (Abstracts of Reports of the XVIth Chugaev All-Union Meeting on the Chemistry of Coordination Compounds) (Moscow: Nauka, 1987) Part 1, p. 342 YuMKiselev, Yu D Tretiyakov 198.R Hoppe Angew. Chem. 93 641 (1981) 199. G Demazeau, S Darracq, J H Choy High Pressure Res. 12 323 (1994) 200. W E Pickett Rev. Mod. Phys. 61 433 (1989) 201. C M Varma Physica C 263 39 (1996) 202. J K Burdett, S A Gramsch, B T Schilf Z. Anorg. Allg. Chem. 621 1508 (1995) 203. E Goodilin,M Limonov, A Panfilov, N Khasanova, A Oka, S Tajima, Y Shiohara Physica C 300 250 (1998) 204. E Goodilin,M Kambara, T Umeda,Y Shiohara Physica C289 251 (1997) 205. E Goodilin,A Oka, J G Wen,Y Shiohara,M Kambara, T Umeda Physica C 299 279 (1998) 206. L M Dall'Antonia, G Tremiliosi-Filho, G Jerkiewicz Proc. Electrochem.Soc. 96 (18) 292 (1996) 207. A Krozer, M Rodahl J. Vac. Sci. Technol., A, Vac. Surf. Films 15 1704 (1997) 208. G Demazeau, D-Y Jung Europ. J. Solid State, Inorg. Chem 32 383 (1995) 209. G Demazeau, D-Y Jung, L Fournes, A Largeteau, J P Sanchez, in High Pressure Sciences Technology (Proceedings of the International Conference, 1995) (Ed. WA Trzeciakowski) (Singapoore, 1996) p. 269 210. G Demazeau J. Korean Chem. Soc. 42 135 (1998) 211. P Hagenmuller (Ed.) Rapport d'Activite Scientiéque (Laboratorie de Chimie du Solide, CNRS, 1977) 212. V A Mukhanov, in Kristally, Rost, Svoistva, Real'naya Struktura, Primenenie (Tr. III Mezhdunar. Konf.), Aleksandrov, 1997 [Crys- tals, Growth, Properties, Real Structure, Application (Proceedings of the Third International Conference), Aleksandrov, 1997] p.171 213. M K Subramanian, M Crawford, R L Harlow, T Ami, J A Fernandez-Baca, Z R Wang,D C Johnston Physica C 235/240 743 (1994) 214. K Ouchetto, F Archaimbault, J Choisnet, M Et-Tabirou Mater. Chem. Phys. 51 117 (1997) 215. Yu M Kiselev, D A Pankratov, N A Ezerskaya, I N Kiseleva, L A Shundrin, M P Popovich Zh. Neorg. Khim. 39 1340 (1994) d 216. Yu M Kiselev, D A Pankratov, L A Shundrin, I N Kiseleva Zh. Neorg. Khim. 41 2069 (1996) d 217. D A Pankratov, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1998 218. D A Pankratov, Yu M Kiselev, in Tez. Dokl. XVI Mendeleevskogo S'ezda po Obshchei i Prikladnoi Khimii (Abstracts of Reports of the 222. V P Golenko, V A Vanyshev, E V Polyanskii, E G Yarotskaya, XVIth Mendeleev Congress on General and Applied Chemistry) (Moscow: Nauka, 1998) Vol.1, p. 239 219. D A Pankratov, P N Komozin,Yu M Kiselev, I B Podnebesnaya, V D Dolzhenko, in Problemy Sol'vatatsii i Kompleksoobrazovaniya v Rastvorakh (Tez. Dokl. VII Mezhdunar. Konf.) [The Problems of Solvation and Complexing in Solutions (Abstracts of Reports of the Seventh International Conference)] (Moscow: Nauka, 1998) p. 302 220. G Blasse Struct. Bond. 76 154 (1991) 221. D T Sviridov, R K Sviridova, Yu F Smirnov Opticheskie Spektry Ionov Perekhodnykh Metallov v Kristallakh (Optical Properties of Transition Metal Ions in Crystals) (Moscow: Nauka, 1976) V G Yarotskii Razvedka Okhrana Nedr (3) 19 (1995) 223.D Reinen, U Kesper, M Atanasov, J Roos Inorg. Chem. 34 184 (1995) 224. S Kueck, S Hartung, S Hurling, K Petermann, G Huber Phys. Rev. B, Condens. Matter. 57 2203 (1998) 225. M Herren, H U Guedel Inorg. Chem. 31 3683 (1992) 226. T C Brunold, H U Guedel, S Kueck, G Huber J. Lumin. 65 293 (1996) 227. A B P Lever Inorganic Electronic Spectroscopy (Amsterdam: Elsevier, 1984) 228. L N Kholodkovskaya, Yu D Perfil'ev, Yu M Kiselev Zh. Neorg. Khim. 42 517 (1997) d 229. S K Dedushenko, L N Kholodkovskaya, Yu D Perfiliev, Yu M Kiselev, A A Saprykyn J. Alloys Comp. 262 ë 263 78 (1997) 230. N S Kopelev, in Yaderno-Spektroskopicheskoe Issledovanie Sverkhtonkikh Vzaimodeistvii (Tez. Dokl. VI Vsesoyuz. Konf.) [Nuclear Spectroscopic Investigation of Hyperfine Coupling (Abstracts of Reports of the Sixth All-Union Conference)] (Moscow: Moscow State University, 1992) p.149The problem of oxidation state stabilisation and some regularities of a Periodic system of the elements 231. Yu D Perfil'ev, L A Kulikov, N S Kopelev, Yu M Kiselev, in Tez. Dokl. I Ross. Konf. po Radiokhimii, Dubna, 1994 (Abstracts of Reports of the First Russian Conference on Radiation Chemistry, Dubna, 1994) (Moscow: Russian Scientific Centre of Cosmic Investigation, 1994) p. 74 232. Yu D Perfiliev, L N Kholodkovskaya, L A Kulikov, Yu M Kiselev, in Proceedings International Conference on the Applications of theMoÈssbauer Effect (ICAME'95) (Abstracts of Reports), Pisa, 1995 p. 04-C16 233. Yu M Kiselev, Yu D Perfil'ev, in Tez.Dokl. XVII Chugaevskogo Soveshch. po Khimii Koordinatsionnykh Soedinenii (Abstracts of Reports of the XVIIth Chugaev Meeting on the Chemistry of Coordination Compounds) (Moscow: Nauka, 1997) p. 63 234. Yu D Perfil'ev, L A Kulikov, L N Kholodkovskaya, Yu M Kiselev, in Tez. Dokl. XVII Chugaevskogo Soveshch. po Khimii Koordinatsionnykh Soedinenii (Abstracts of Reports of the XVIIth Chugaev Meeting on the Chemistry of Coordination Com- pounds) (Moscow: Nauka, 1997) p. 99 235. L N Kholodkovskaya, Yu D Perfiliev, L A Kulikov, in Proceedings International Conference on Solid State Chemistry'96 (Abstracts of Reports), Bratislava, 1996 p. 130 236. Yu D Perfiliev, L N Kholodkovskaya, Yu M Kiselev, L A Kulikov Ital. Phys. Soc. 50 517 (1996) 237. S K Dedushenko, Yu D Perfiliev, L N Kholodkovskaya, Yu M Kiselev, A A Saprykin, in Proceedings of the 12th Interna- tional Conference of Solid Compounds of Transition Elements (Abstracts of Reports), Sent-Malo, 1997 p. O-23 238. S K Dedushenko, Yu D Perfiliev, Yu M Kiselev, V A Andrianov, M G Kozin, in Proceedings of the 14th International Symposium on Quadrupole Interactions (Abstracts of Reports), Pisa, 1997 p. 99 239. S K Dedushenko, Yu D Perfiliev, Yu M Kiselev, V A Andrianov, M G Kozin, in Proceedings of the International Conference on the Applications of theMoÈssbauer Effect (ICAME'97)(Abstracts of Reports), Rio de Janeiro, 1997 MO. T4. P 03 240. Yu D Perfil'ev, S K Dedushenko, L A Kulikov, Yu M Kiselev, in Tez. Dokl. XVI Mendeleevskogo S'ezda po Obshchei i Prikladnoi Khimii (Abstracts of Reports of the XVIth Mendeleev Congress on General and Applied Chemistry) (Moscow: Nauka, 1998) Vol. 1, p. 245 241. S K Dedushenko, Yu D Perfiliev, L A Kulikov, L A Ponomarenko, Yu M Kiselev, in Proceedings of the XIth International Conference on Hyperéne Interactions, Durban, 1998 p. We19 242. L A Ponomarenko, S K Dedushenko, Yu M Kiselev, N A Chumaevsky Mendeleev Commun. 169 (1998) 243. V M Kuznetsov, A N Baranov, N N Oleinikov, E V Kisterev, O V Abramov, Yu D Tretiyakov Dokl. Akad. Nauk 358 204 (1998) a 244. Yu D Tretiyakov, N N Oleinikov, A A Vertegel Zh. Neorg. Khim. 41 932 (1996) d 245. G L Gutsev, A I Boldyrev Zh. Strukt. Khim. 26 (1) 22 (1985) l 246. G L Gutsev, A I Boldyrev Chem. Phys. Let. 108 255 (1984) 247. P Schwerdtfeger, P D W Boyd, S Brienne, A K Burrell Inorg. Chem. 31 3481 (1991) 248. R L Deming, A L Allred, A R Dahl, A W Herlinger, M O Kestner J. Am. Chem. Soc. 98 4132 (1976) 249. T A Magdesieva, K P Butin, Yu M Kiselev, O A Reutov J. Organomet. Chem. 235 127 (1982) 250. K P Butin, Yu M Kiselev, T A Magdesieva, O A Reutov Izv. Akad. Nauk SSSR, Ser. Khim. 716 (1982) c 251. C S Gopinath, N H Hur, S Subramanian Phys. Rev. B, Condens. Matter. 52 R9879 (1995) 252. C S Gopinath, S Subramanian Physica C 232 222 (1994) 253. G L Gutsev, A I Boldyrev Koord. Khim. 10 1455 (1984) f 254. B B Chaivanov, V B Sokolov, S N Spirin Research in the Field of Inorganic Fluorine Chemistry (Moscow: Atominform, 1989) 255. V I Spitsyn, G V Ionova, A A Kiseleva Dokl. Akad. Nauk SSSR 267 676 (1982) a 256. T I Maksimova, H Vogelzang, H Stolz, W von der Osten Fiz. Tv. Tela 39 1567 (1997) 257. C E Moore National Standard Reference Data. Series 34 (Washington, DC: National Bureau of Standards, 1970) 258. T A Carlson,C V Nestor, J N Wasserman Atomic Data 2 63 (1970) 259. P Rother, F Wagner, U Zahn Radiochim. Acta 11 203 (1969) 260. L N Sidorov Koord. Khim. 3 1128 (1977) f 379 261. MVKorobov, Doctoral Thesis in Chemical Sciences, MoscowState University, Moscow, 1990 262. P Kuerpick, W-D Sepp, B Fricke Phys. Rev. A, At. Mol. Opt. Phys. 51 3693 (1995) 263. Y D Wang, C D Lin, N Toshima, Z Chen Phys. Rev. A, At. Mol. Opt. Phys. 52 2852 (1995) 264. W Wu, C L Cocke, J P Giese, F Melchert, M L Raphaelian, M Stoeckli Phys. Rev. Lett. 75 1054 (1995) 265. K E Zaharakis, R R Haar, O Woitke, M Zhu, J A Tanis, N R Badnell Phys. Rev. A, At. Mol. Opt. Phys. 52 2910 (1995) 266. S M Blokhin, V V Kolesnikov Zh. Tekhn. Fiz. 65 (5) 183 (1995) Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) Radiochemistry (Engl. Transl.) Russ. Chem. Bull. (Engl. Transl.) Russ. J. Inorg. Chem. (Engl. Transl.) Mendeleev Chem. J. (Engl. Transl.) Russ. J. Coord. Chem. (Engl. Transl.) Physics-Uspekhi (Engl. Transl.) Russ. J. Anal. Chem. (Engl. Transl.) Moscow Univ. Bull. (Engl. Transl.) Russ. J. Appl. Chem. (Engl. Transl.) Theor. Exp. Chem. (Engl. Transl.) J. Struct. Chem. (Engl. Transl.)

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (5) 381 ± 392 (1999) Synthetic diamond in electrochemistry Yu V Pleskov Contents I. Introduction II. Preparation of diamond thin-film electrodes III. Crystal structure and basic electrophysical characteristics of diamond films IV. Synthetic diamond films as electrodes: resistance to corrosion and background current ± potential curves V. Impedance spectroscopy and semiconductor properties of diamond electrodes VI. Kinetics of the electrode reactions VII. Photoelectrochemistry of diamond VIII. Conclusion Abstract. The results of studies on the electrochemistry of dia- mond carried out during the last decade are reviewed. Methods for the preparation, the crystalline structure and the main electro physical properties of diamond thin films are considered.Depend- ing on the doping conditions, the diamond behaves as a superwide-gap semiconductor or as a semimetal. It is shown that the `metal-like' diamond is corrosion-resistant and can be used advantageously as an electrode in the electrosynthesis (in partic- ular, for the electroreduction of compounds that are difficult to reduce) and electroanalysis. Kinetic characteristics of some redox reactions and the impedance parameters for diamond electrodes are presented. The results of comparative studies of the electrodes made of diamond single crystals, polycrystalline diamond and amorphous diamond-like carbon, which reveal the effect of the crystalline structure (e.g., the influence of intercrystallite bounda- ries) on the electrochemical properties of diamond, are presented.The bibliography includes 99 references. I. Introduction Diamond is a material endowed with unique properties: it is a perfect dielectric characterised by extremely high atomic density and hardness. Historical records of intensive electrophysical, physicochemical and optical studies of diamond is only several- decade long.1, 2 Its practical uses have diversified and embrace nowadays such fields as materials technology,3 microelectronics 4 and other. Depending on the doping level, diamond behaves as a super- wide-gap semiconductor or as a semimetal. Owing to its excep- tional chemical stability, semiconductor diamond is undoubtedly promising as an electrode material. However, in contrast with other carbon materials of long-standing use (graphite, glassy carbon and other), electrochemical studies of diamond were initiated approximately a decade ago.Until recently, such inves- tigations have been seriously hampered by two circumstances: firstly, diamond remained an exotic, difficultly accessible mate- Yu V Pleskov A N Frumkin Institute of Electrochemistry, Russian Academy of Sciences, Leninsky prosp. 31, 117071 Moscow, Russian Federation. Fax (7-095) 952 08 46. E-mail: pleskov@electrochem.msk.ru Received 18 November 1998 Uspekhi Khimii 68 (5) 416 ± 429 (1999); translated by V D Gorokhov #1999 Russian Academy of Sciences and Turpion Ltd UDC 541.148 381 382 382 383 384 387 390 391 rial; secondly, by its nature, diamond is an insulator and hence is unsuitable material for electrodes.The situation has radically changed in connection with the achievements in the technology of manufacture of diamond thin films from the gas phase under subatmospheric pressure. Highly efficient methods have been developed for the growing of poly- crystalline diamond films on non-diamond substrates; such films are relatively inexpensive when manufactured on a large scale. Doping with boron, as an acceptor admixture, allowed one to synthesise semiconductor films with good conductivity, while high doping levels led to quasimetallic films.5 The first studies on diamond electrodes have opened a new branch of electrochemistry of semiconductor materials 6 known now as electrochemistry of diamond.Formally, the first publication on diamond electrochemistry was that describing attempts to make an electrode from a dielectric diamond crystal.7 In order to make electroconducting at least the thin outer layer of the specimen, the authors implanted Ar+ions into the diamond crystal. Argon is not a doping impurity for the diamond and does not change its conductivity, but as the result of such an ion implantation the external diamond layer becomes amorphous (e.g., see Refs 8, 9). A noticeable conductiv- ity arises because of the destruction of the diamond crystalline lattice and the appearance of a large number of carbon atoms with the sp2-hybridisation of their orbitals.10 For this reason, strictly speaking, the study of Iwaki et al.7 dealt with the electrochemistry of non-diamond (amorphous) carbon and not with the diamond proper.In fact, the electrochemistry of diamond was started with the study reported in Ref. 11. In this work, the I ±V characteristic was derived for the first time and the differential capacitance at the diamond/electrolyte solution interface was measured; photosen- sitivity of diamond electrodes was also revealed and related to the semiconductor properties of diamond. A few years later, these investigations of Russian researchers were followed by studies carried out by electrochemists in Japan, Israel, France, USA and other countries. For the last three years the number of publica- tions devoted to diamond electrodes has increased dramatically.This review generalises the most important results from studies of electrochemical properties of synthetic semiconductor diamond and application of electrochemical methods for the determination of characteristics of diamond films and outlines perspectives of the development of this new branch of electro- chemistry.382 II. Preparation of diamond thin-film electrodes Since the work reported in Ref. 11 until now the subjects of virtually all studies have been diamond films prepared by the method of chemical vapour deposition (CVD). The studies by Derjaguin et al.12 carried out in the 1970s were an important stage in the development of this method. The CVD method relies on the deposition under subatmospheric pressure from an activated gas phase consisting of a volatile organic substance and hydrogen.3, 13 The gas phase activation is required in order to create sufficiently high concentration of active carbon-containing species.These particles, by colliding against the surface of a heated substrate, disintegrate and release carbon atoms from which the crystalline lattice of diamond is formed. In the temperature and pressure ranges used, the diamond is metastable; the stable form is another allotropic carbon modification known as graphite. The gas phase activation should lead to the appearance of a reagent suppressing the deposition of non-diamond forms, viz., graphite and amor- phous carbon. Atomic hydrogen is such a reagent; in its atmos- phere graphite `burns out', while diamond remains unaltered.Current practices rely basically on the use of thermal and electric (sometimes, chemical and photochemical) methods of gas activation. In the case of thermal activation, a reactor for vapour deposition of diamond films is used (Fig. 1). Above the substrate, on which diamond is supposed to grow, an activator is suspended in the form of wire made of a high-melting metal (W, Ta). This wire is heated to incandescence (*2000 8C) at which temperature the H2 molecules dissociate at a sufficiently high rate. The gas phase is a mixture of carbon-containing gases (methane, acetone or methanol vapour, etc., usually at concentrations of up to 5%) with hydrogen. Their contacts with the activator surface yield, along with the hydrogen atoms, excited carbon-containing mole- cules and radicals that are transferred towards the substrate surface, where deposition of diamond film takes place.6 17 2 3 54 9 8 Figure 1. Scheme of a reactor for the chemical vapour deposition of diamond films under thermal activation;3 (1) reactor body; (2) substrate; (3) substrate heater; (4) thermocouple; (5) heated tungsten wire; (6) gas input; (7) diffuser; (8) pipe-line to a vacuum pump; (9) pipe-line to a manometer. Electrical activation is induced using various forms of electric (arc, glow) discharge with the use of constant and alternating current (high frequency, ultra-high frequency, pulse mode). Each version of the electric method has its own advantages and disadvantages.For example, the electric arc discharge permits high deposition rates ranging from tens to hundreds of micro- meters per hour. Under the ultra-high-frequency activation, the plasma ensures homogeneous conditions for the deposition over a large surface area of the substrate. Yu V Pleskov Depending on the aim of particular studies, the role of the substrate is played by metals (W, Mo), silicon (wafeas with a mirror-polished surface that are frequently used in microelec- tronics), glassy carbon, etc. Crystallisation is primed by diamond microcrystals seeded on the support surface. Later, the silicon substrate can be easily etched to obtain the so-called free-standing films. If the substrate is a diamond single crystal (natural or synthetic), then epitaxial growth takes place: the growing film repeats the crystalline structure of the substrate and forms a single crystal (such films are called homoepitaxial).Usually, the film thickness is a few micrometers, but in some cases it may reach several millimetres. The area covered with a diamond film reaches 10610 cm and more, though the electrochemical studies usually use electrodes*161 cm in size. As noted above, in order to make a diamond film electro- conducting, one should dope it (usually in the process of its growth) with boron. For this purpose, the gas phase is supple- mented with a volatile boron compound, e.g., trimethyl borate, in the concentration ranging from 10 to 10 000 ppm. Alternatively, a solid source of boron is placed near the activator and the substrate.In the latter case, the boron is transferred into the gas phase as a result of the activation. It has been established empirically that the content of boron in the growing film is approximately proportional to its content in the gas phase.14 The specific resistance of diamond depends on the content of boron in the film: it can vary, e.g., from 104 Ohm cm (at a content of boron atoms in diamond of*1018 cm73) to tenth and even thousandth fractions of Ohm cm (at a content of boron atoms in diamond *1021 cm73). As its resistivity decreases, the dielectric diamond acquires successively the properties of a semiconductor, a degen- erate semiconductor and then of a semimetal.The energy of boron ionisation in diamond is great compared with the thermal energy (kT). At room temperature, the activation energy is 0.37 eV and decreases appreciably at high boron concentrations.15 Hence, the concentration of the majority charge carriers (holes) is 2 ± 3 orders of magnitude lower than that of the boron atoms (acceptors). It should be pointed out that the methods of chemical and physical vapour deposition are widely used for the preparation of not only crystalline (diamond) films but also films of amorphous hydrogenated carbon a-C:H 10 in which there is no long-range ordering of the C7C bonds (hence, there is no crystalline structure), but the short-range ordering exists. Of a wide diversity of amorphous carbon films, one can conditionally distinguish graphite- and diamond-like films.Hybridisation of the carbon atom orbitals is predominantly of the sp2-type in the former and of the sp3-type in the latter. The graphite-like films are characterised by a small width (<1 eV) of the gap and have high electric conductivity. The diamond-like films have a wider (2 ± 4 eV) gap and are appreciably less conductive; certain properties of these films are similar to those of diamond (and this explains their name). In addition to thin films, individual diamond crystals grown at high temperature and high pressure can also be used as electrodes. Such crystals are obtained from a carbon solution in molten metals (e.g., Ni7Fe7Mn) in the pressure and temperature ranges corresponding to the region of thermodynamic stability of dia- mond; diamond crystals are doped with boron during their growth.III. Crystal structure and basic electrophysical characteristics of diamond films 1. Crystal structure Polycrystalline films consist of crystallites the size of which increases with the film growth approximately proportionally to its thickness and reaches a few micrometers. Facets of crystallites usually form a characteristic faceted morphology of the film surface.Synthetic diamond in electrochemistry The perfection of the crystalline structure of deposited films is monitored by recording their Raman spectra. A typical spectrum of a diamond film is shown in Fig. 2. The crystalline diamond is characterised by a narrow peak at 1332 cm71 due to the first- order scattering on the crystal lattice. Two blurred bands at *1350 and *1550 cm71 correspond to the non-diamond car- bon form with the predominant sp2-hybridisation.When compar- ing the diamond peak and the maximum determined by the sp2- carbon in the spectrum, one should bear in mind that the magnitude of Raman signal on the ordinate (Fig. 2) is 50 times more sensitive to the non-diamond carbon form than to the crystalline diamond.16 In the best specimens employed as electro- des, the admixture of the non-diamond form did not exceed 1% (as confirmed by independent electrochemical measurements). According to Hashimoto et al.,17 the admixture content decreases in the film substrate surface direction.Nevertheless, modern physical techniques (Raman and Auger spectroscopy, atomic- force microscopy) do not provide sufficiently exhaustive charac- terisation of non-diamond carbon inclusions in polycrystalline diamond films. It is by electrochemical measurements that these inclusions are revealed.18 1073 n /pulse s71 1.2 0.8 0.4 1620 1440 1260 1080 Raman shift /cm71 Figure 2. Raman spectrum of a polycrystalline diamond film. In polycrystalline diamond films, the non-diamond form is represented primarily by a disordered carbon of intercrystallite boundaries, the outcrops of which on the diamond surface were observed using high-resolution electron microscopy.19 The thick- ness of the intercrystallite boundaries in diamond films reaches a few nanometers. In addition, the crystal lattice of films has various defects.In some instances, a thin (about several nanometers) non- diamond layer coating the diamond film was revealed. This layer is often formed in the latest (poorly controlled) stages of film growth after cessation of the gaseous phase activation. To remove it, samples are heated in air to 520 8C or boiled in concentrated HClO4, dichromate-sulfuric acid mixture or solutions of other strong oxidants. Such a treatment of the non-diamond carbon form results in its oxidation to CO2. 2. Electroconductivity of films Specific resistance of diamond films is determined not only by the concentration but also by the mobility of free charge carriers.Usually, the mobility of the holes in polycrystalline films is not high; it is lower than that in natural single crystals (usually, of the magnitude of 1 cm2 V71 s71),20 ± 22 although in some cases it reaches tens and hundreds of cm2 V71 s71 (see, e.g., Refs 23 and 24). The mobility of the holes in single-crystal (homoepitaxial) films is 1 ± 2 orders of magnitude higher than that in polycrystal- line films grown under the same conditions.22 In films of amor- phous diamond-like carbon, the mobility is several orders of magnitude lower than that in crystalline diamond.20 The character of conductivity of doped polycrystalline films and its relation to the crystal structure were studied by the method of impedance spectroscopy.25 The results obtained have made it 383 possible to describe a conducting polycrystalline film using a simple model of a heterogeneous medium. According to this model, the medium consists of regularly distributed well-conduct- ing crystallites of the same size separated by poorly conducting intercrystallite boundaries, which form a continuous matrix.Undoubtedly, such a model simplifies the structure of a real film in which crystallites are distributed over its thickness with regard to their size, while individual acicular crystallites penetrate through the entire film thickness. Nevertheless, this model is applicable in the studies of moderately doped diamond films. The theory of complex conductivity of a heterogeneous medium 26 developed for the model described has made it possible to determine the bulk proportion of the boundary phase in the film (*5%, which in order of magnitude correlates with the above- mentioned results of Raman spectroscopy) and to estimate con- ductivities of crystallites { and intercrystallite boundaries as 1074 and 1077 Ohm71 cm71, respectively.Similar conclusions about the character of conductivity of polycrystalline films were made by Sokolina et al.27 Subsequent studies devoted to thorough investigation of the frequency dependence of the impedance of polycrystalline films 28 ± 30 have shown that at high temperatures the electric conductivity due to the motion of free holes in the valence band should be supple- mented with a component determined by the jump-over of holes between localised traps (possibly, associated with the intercrys- tallite boundaries).IV. Synthetic diamond films as electrodes: resistance to corrosion and background current ± potential curves In the assessment of novel electrode materials used in electrolysis, electroanalysis and manufacture of electrochemical sensors and other devices, the main attention is paid to the magnitude of background current and the width of the potential range in which this current remains sufficiently low and does not interfere with measurements as well as to the corrosion resistance and stability of materials. The diamond electrode meets all of these criteria. Let us consider the current-potential curves in solutions of indifferent electrolytes as the initial characteristic of the electrode behaviour of diamond.31 ± 33 It is these curves that form the `background' against which the kinetic, impedance, photoelectro- chemical and electroanalytical characteristics of diamond electro- des are manifested.An advantage of diamond electrodes is a wide potential range of weak background currents, which is sometimes called the range of ideal polarisability. It is obvious in comparison with other materials, which are widely used as electrodes (platinum, glassy carbon, graphite, etc.). Figure 3 shows characteristic polarisation curves recorded using a solution of an indifferent electrolyte (0.5 M H2SO4) in the potentiodynamic mode of potential cycling on the polycrystal- line diamond thin-film electrodes and the electrodes from plati- num and highly oriented pyrographite which are traditionally applied in electrochemistry.According to Martin et al.,32 all the diamond films used as electrodes may be divided into `good' and `bad' films, depending on the width of the ideal polarisability range. For the `good' films, this range reaches *3 V at the current density j of several mA cm72 and at least about 1.5 V for the j values making up a few decimal fractions of 1 mA cm72 (Fig. 3a). For the `bad' films, the range of ideal polarisability is appreciably narrower (Fig. 3b); there is a noticeable hysteresis, which naturally restricts their { In the studies reported in Refs 11 and 25, the conductivity of diamond films was associated not with the boron incorporation but rather with the presence of unidentified admixtures and/or defects due to specific con- ditions of deposition (in particular, an elevated temperature of the substrate).384 b a j /mA cm72 j /mA cm72 250 725 d c 250 1 E /V 2 E /V 71 72 725 Figure 3.Polarisation curves of the electrodes made of polycrystalline diamond (a, b), platinum (c) and highly oriented pyrographite (basal plane) (d) in 0.5 M H2SO4;33 (a) `good' film; (b) `bad' film; potential was measured relative to the normal hydrogen electrode (NHE). applicability as electrodes. Martin et al.32 did not give any accurate definition of `good' films but indicated that they were obtained at a low content (*1%) of a carbon-containing gas (methane) in the reaction system (the `bad' films are formed at a methane content of*4%).Comparison of the background currents for diamond and glassy carbon has shown that this current is 10 times as low on diamond.34 As regards its resistance to corrosion in aggressive media and under extreme potentials, the diamond demonstrates it in full measure. Multiple cycles of potential variation on the diamond electrode from the potential of hydrogen evolution on the cathode to the potential of oxygen evolution on the anode in the aggressive solution 1 M HNO3+0.1 M HF had no effect on either the surface morphology of diamond (according to electron micro- scopy data) or the ratio of diamond and non-diamond carbon forms on the electrode surface (according to Raman spectra).Under the same conditions, the electrodes from glassy carbon and pyrographite underwent strong degradation.35 A similar study was conducted in the solution 1 M HNO3+0.1 M NaCl under the potentials of the chlorine evolution on the anode.36 But in this case too, X-ray photoelectron spectroscopy showed that neither the morphology of the diamond surface nor the content of admixtures on it were altered. On the other hand, prolonged `cycling' of the diamond electrode potential in concentrated (15%) KOH solution can lead to corrosive degradation of the diamond surface. The extent and character of this degradation depend on the conditions of diamond growth, although this dependence has not been clearly established.The intercrystallite boundaries composed of the non- diamond carbon undergo the strongest degradation.37 In some cases (probably, in the presence of latent flaws in the film), prolonged polarisation of the diamond electrode in solution leads to the appearance in it of pinholes through which the electrolyte penetrates to the substrate.38 ± 40 On the electrode surface, some inactive sites may appear 41 as the result of local detachment of diamond films from the substrate or possibly because of a non-uniform doping of diamond with boron. The well-prepared films prove to be not only resistant to corrosion themselves but also to protect the metal substrate.39, 42, 43 If a diamond film was obtained in a hydrogen atmosphere, its surface is covered by a hydrogen monolayer and displays hydro- phobic properties.This type of surface is also obtained after treatment of diamond in a hydrogen plasma.44 On the contrary, an oxidised diamond surface contains the hydroxy groups and hence is hydrophilic.32, 45 These two types of surface can also be obtained by electrochemical treatment consisting, respectively, in the cathodic or anodic polarisation in an aqueous electrolyte solu- tion.46 Yu V Pleskov In conclusion, let us consider properties of the electrodes made of diamond-like carbon. Here, the data are scarce and are some- what contradictory. For example, Howe 47 has reported that even thin films (50-nm thick) are well resistant to corrosion, while Sakharova et al.48, 49 have pointed out that such films proved to be permeable to the electrolyte, which by penetrating to the metal substrate resulted in its corrosion and led eventually to the film peeling.Films from diamond-like carbon of submicron thickness were found to be more resistant. Their background current ± po- tential curves resemble analogous curves for the diamond elec- trode (Fig. 3a,b) but have a narrower range of ideal polarisability.50 V. Impedance spectroscopy and semiconductor properties of diamond electrodes 1. Linear impedance. Equivalent circuits Measurement of frequency characteristics of electrochemical systems makes it possible to derive information on their micro- scopic structure, as well as on the nature and kinetics of the charge transfer processes that occur in them.Linear impedance can be measured if a harmonic electrical signal with the frequency f is applied to the cell and then the real (ReZ) and imaginary (ImZ) components of the impedance Z are determined by analysing the spectrum of the system response with an appropriate device. In studies of the impedance of diamond electrodes, the great- est attention is paid to the following peculiarities: the semicon- ductor character of diamond (for moderately doped films) and a special type of the frequency dependence of the impedance characteristic of the overwhelming majority of films. In its general form, the impedance of a thin-film electrode may be represented by the three equivalent circuits shown in Fig.4. The circuit depicted in Fig. 4 a includes two RC-chains, one of which belongs to the diamond film itself and the other relates to the diamond/electrolyte solution interface. In the first chain, Rs is the ohmic resistance of the diamond film and Cg is the geometrical capacitance of the film; in the second chain, Csc is the differential capacitance of the space charge region at the diamond/solution interface and RF is the differential resistance to the process of charge transfer at this interface (the so-called Faraday resistance). The geometrical capacitance of most films is low (of the order of 1079 F cm72); as regards the order of magnitude, the corre- sponding impedance is larger than the film resistance Rs (except for very lightly doped films), and hence the latter may be neglected.In this case, we obtain a simple equivalent circuit with three elements (Fig. 4 b) which is known as the Randles ± Ershler circuit.51 It is essential that all the elements of the above circuits are independent of signal frequency. The circuit represented by Fig. 4 c includes a frequency-dependent, constant phase-shift element (CPE) (see below). a Csc Cg RF Rs b Csc Rs RF c CPE Rs RF Figure 4. Equivalent circuits of the impedance (see text).Synthetic diamond in electrochemistry /mF72 cm4 1 C2sc 4.0 2.00 E /V 1.0 0.5 Figure 5. The Mott ¡¾ Schottky plot for a boron-doped polycrystalline electrode with frequency-independent capacitance in 0.5 M H2SO4.52 In certain cases, the impedance spectra of diamond electrodes are well described by the equivalent circuit (Fig.4b) with a virtually frequency-independent (in the range of 1 ¡¾ 105 Hz) capacitance.52 Figure 5 presents the Mott ¡¾ Schottky plot, which expresses the dependence of the squared reciprocal capacitance Csc on the polycrystalline diamond electrode potential E in 0.5 M H2SO4. Physically, it reflects the dependence of the space charge layer thickness in a semiconductor on potential.6 The space charge arises because the near-surface region of semiconductors is depleted in mobile charge carriers (the so-called depletion layer). Extrapolation of the linear segment until its intersection with the abscissa provides the value of flat-band potential of the electrode Efb, while the slope of this line can be used for calculating the concentration of non-compensated acceptors in a semiconductor by the formula ¢§1 N , (1) sc U dOC¢§2 dE a a ¢§ 2 ee0e where e is the electron charge, e and e0 are the dielectric constants of diamond and vacuum, respectively.Formula (1) was derived under assumptions that, first, the potential drop during polar- isation takes place in the space charge region (and, e.g., not in the Helmholtz layer at the interface); second, there is no ionisation of `deep' acceptors; third, the measured capacitance is that of the space charge regionCsc (and does not include, e.g., the capacitance of surface states, adsorption capacitance, etc.). For the electrode under discussion (see Fig.5), Na ^ 3.161018 cm73. The depletion layer is usually formed in a doped wide- bandgap semiconductor which is in contact with another phase (metal, electrolyte solution or vacuum). The potential barrier associated with the depletion layer formation (in particular, upon contact with the conducting phase) is called the Schottky barrier. The contact of polycrystalline diamond films with certain metals (Au, Pt, Pd) manifests a rectifying effect (see, e.g., Refs 53 and 54), while its capacitance characteristics resemble those of the diamond/indifferent electrolyte contact. However, specimens with a frequency-independent capaci- tance represent an exception rather than the rule. Usually, the capacitance of diamond films depends on frequency.Figure 6 features a characteristic electrode impedance spectrum presenta- tion on a complex plane.55 At high frequencies (1 ¡¾ 100 kHz), the complex-plane plot forms a straight line that does not pass through the origin of coordinates (see Fig. 6a). A similar impe- dance feature was observed in a series of studies (see, e.g., Refs 56 ¡¾ 59). At low frequencies, the7ImZ vs. ReZ plot is curved due to the presence of the Faraday resistance in the equivalent circuit (Fig. 6). Under the anodic or cathodic polarisation, the curvature becomes more appreciable due to the decrease in RF, and the shape of the low-frequency segment of the curve approaches a semicircle. 385 b a 7ImZ /Ohm cm2 7ImZ /Ohm cm2 0.02 1200 15 0.04 15 10 800 30 5 400 200 0 0 0.07 0.15 2 8 2 4 ReZ /Ohm cm2 800 400 0 1200 ReZ /Ohm cm2 Figure 6.High-frequency (a) and low-frequency (b) segments of the complex-plane plot of the impedance of a single-crystal electrode in 0.5 M H2SO4 under a steady-state potential.35 Numerals at the curves indicate the signal frequencies /kHz. Such a characteristic type of the frequency dependence suggests that the equivalent circuit of the electrode includes the so-called constant phase element (CPE) with its impedance equal to (2) ZCPE=s71(io)7a , where the frequency-independent parameter s has the dimension- ality Fa Ohma71 cm72, the index a determines the character of the frequency dependence, i is the imaginary unity, o=2pf is the angular frequency of the alternating current.Analysis of the impedance spectra is based on the equivalent circuit (Fig. 4c) derived from the Randles ¡¾ Ershler circuit by replacing the fre- quency-independent capacitance with CPE. The elements of the equivalent circuit are usually calculated by the least-squares method with minimisation of the standard deviation of the logarithm of the impedance modulus measured at different frequencies from the logarithm of the impedance modulus calcu- lated for the equivalent circuit using the corresponding computer programme.60 Typical s and a values, film resistance Rs, calculated diamond resistivity r and the Faraday resistance RF are listed in Table 1 for the polycrystalline and single-crystal (thin-film and compact) diamond electrodes and the electrode made of amorphous diamond-like carbon.50, 61 Comparison of the data presented in Table 1.Parameters a, Rs (Ohm cm2), s (mFa Ohma71 cm72), r (Ohm cm) and RF (Ohm cm2) found in the studies of the impedance of polycrystalline and single-crystal diamond films in 0.5 M H2SO4 solution. Electrode a Rs s r RF Sub- strate Film thickness /mm 2.4 W 0.92 1.5 1 ¡¾ 1.5 0.76104 104 ¡¾107 Polycrys- talline film 0.1 0.96 32 2.7 10 106 ¡¾108 Single- crystal film syn- thetic diamond single crystal 30 106 7 7 0.92 2.8 2.5 Single crystal a Amor- 0.07 0.86105 106 0.9 7 0.05 glassy carbon phous diamond- like film aA single crystal grown under high pressure and high temperature.386 Table 1 suggests that in both the qualitative respect (the equivalent circuit form and the presence of CPE in it) and the quantitative respect (typical a and s values) the polycrystalline and single- crystal electrodes are virtually the same: a varies within a narrow range 0.9 ± 0.96, s ^ 1 ± 3 mFa71 Ohma71 cm72, RF&104 ± 108 Ohm cm2 (in an indifferent electrolyte solution under a sta- tionary potential).Assuming that the measured integral impe- dance parameters a and s characterise the double electrical layer on diamond electrodes, it may be concluded that the structure of this layer is approximately the same in the polycrystalline and single-crystal specimens. Furthermore, on the whole, the structure of the double electrical layer is identical in the diamond and wide- bandgap diamond-like carbon electrodes.Rather high values of the Faraday resistance RF reflect the above-mentioned low background currents and the negligibly small corrosion of diamond in aqueous solutions. Larger values of s in comparison with those in other semiconductor electrodes are indicative of a relatively high doping level of the diamond films studied (the acceptor concentration Na51018 cm73). Note that according to different authors, the flat-band poten- tial (Efb) values determined by the extrapolation of the Mott ± - Schottky plots are in the range from 0.5 V (Ref. 52) to 1.2 V (Ref. 55) [relative to the saturated calomel electrode (SCE)]. Such a spread of data can in part be explained by different degrees of diamond surface oxidation (resulting from pretreatment of sam- ples).It is known 6 that the anodic and cathodic oxidation of numerous semiconductors shifts their flat-band potentials toward positive and negative regions, respectively. This is largely due to an appreciable extent to the dipolar potential drop associated with the oxygen monolayer adsorbed on the electrode surface and determined by the polarity of the C7O bonds. Note in conclusion that in the CPE-containing equivalent circuit of the electrode, no separate frequency-independent capacitance was revealed that could be attributed to the space charge region in diamond.62 This means that it is the CPE [with its dependence on potential which is characteristic of semiconductor electrodes (see Section V.2 for more detail)] that describes the properties of the space charge region.62 Recently, the diamond crystals grown under high pressure and high temperature have been used for the first time as electrodes.Their impedance characteristics proved to be close to those of thin-film electrodes (see Table 1).63, 64 2. On the nature of the frequency dependence of the differential capacitance of diamond electrodes Naturally, the dependence of capacitance on frequency that follows from the impedance complex-plane plot (see Fig. 6) determines the slope of the straight line in the Mott ± Schottky plot.65 The researchers investigating the frequency dependence of the differential capacitance of diamond electrodes must answer two major questions, viz., what processes lead to the appearance of this dependence and what is the most convenient manner of its formal presentation.The nature of the frequency dependence of the slope of the Mott ± Schottky plot for the semiconductor electrodes has been subject to discussion for over three decades (see, e.g., Refs 6 and 66). Possible reasons for its appearance are believed to be the dependence of dielectric relaxation on frequency in the space charge region,67 roughness of the electrode surface,66 slow ionisa- tion of deep donors (acceptors) in the space charge region of semiconductors 68 and the influence of the surface states. Evstefeeva et al.69 investigated the dependence of capacitance on frequency in its application to the diamond electrode assuming that the electrode surface roughness should not necessarily lead to the appearance of CPE in its impedance.70 The hypothesis 59 of slow ionisation of atoms with a deep-lying energy level (e.g., boron) in the space charge region of diamond crystals has been assumed for a tentative explanation of the frequency dispersion of the Mott ± Schottky plots of diamond electrodes.This explanation is supported by the fact that the magnitude of s varies regularly with changes in the thickness of the space charge layer (in conformity with Schottky's theory), whereas the magnitude of a is virtually independent of both the potential and the solution dilution, i.e., the factors influencing the electrode surface state.Therefore, it has been concluded that the observed dependence of capacitance on frequency is not associated with the surface states but is determined by the process occurring in the space-charge region of diamond.{ Let us consider different forms of the Mott ± Schottky plots when they are frequency-dependent using the diamond electrode as an example. As noted above, this situation is typical of the electrochemistry of various semiconductor electrodes. Figures 7 and 8 illustrate three ways of presentation of such plots. The first pattern is a `sheaf' of C72 ± E straight lines plotted from the capacitance values measured directly at a given frequency (a series of frequencies) and potential (see Fig. 7). The second one is the plot C¡2 calc ± E (see Fig.8, curve 2). When plotting this curve, the real equivalent electrode circuit (see Fig. 4c) was replaced by a simpler circuit (see Fig. 4b) including the frequency-independent capacitance instead of CPE. In its essence, the Ccalc is the capacitance Csc averaged over the entire range of the alternating current frequencies studied. As this step uses the body of exper- imental data for the whole frequency range, the value of Ccalc obtained may be regarded formally as the frequency-independent one. Finally, the third pattern (for the cases where the equivalent circuit contains CPE) is the s72 ±E plot (see Fig. 8, curve 1). Note that s?Ccalc at a?1. 10713 1 C2 /F72 cm4 1 3.0 2 2.0 345 6 1.0 7 0 0.4 0 Figure 7.The Mott ± Schottky plot for a boron-doped single-crystal electrode with a frequency-dependent capacitance in 0.5 M H2SO4. Frequency/Hz: (1) 21 544; (2) 10 000; (4) 4642; (5) 2154; (6) 1000; (7) 215; (3) the curve is calculated for Ccalc;69 the potential values were measured relative to the Ag/AgCl electrode. When comparing the above-listed patterns, one should pay attention to the fact that the plot s72 vs. E is frequency- independent, since s is the frequency-independent parameter by definition [see formula (2)]. It may be considered as the most general case and be used for comparing results of different experiments, since in this case the contribution from the intrinsic frequency-dependent element is ruled out.However, taking into { It is unclear until now why the differential capacitance of some diamond electrodes is independent of frequency. Yu V Pleskov E /VSynthetic diamond in electrochemistry (s/sE=0.4V)72, (Ccalc/CcalcE=0.4V)72 3 1.8 2 1 1.4 1.0 0.6 0.2 0.2 70.2 Figure 8. Dependence of (s/sE=0.4 V)72 (1), (Ccalc/CcalcE=0.4 V)72 (2) and a (3) on potential for a single-crystal electrode (calculated using the data of Fig. 7).69 Potentials were measured relative to the Ag/AgCl electrode. account the dimensionality of s, correct comparison of the calculated data with the published results of direct capacitance measurement on diamond electrodes seems questionable. The use of the C¡2 calc ± E dependence removes the dimensiona- lity problems, although a somewhat formal character of the Ccalc parameter is a disadvantage of this form of the Mott ± Schottky plot presentation.The sheaf of C72 ± E straight lines may be interpreted as manifestation of a slow process (e.g., the above-mentioned ionisation of the deep acceptor) occurring in the space charge region of a semiconductor. Unfortunately, the nature of the dependence of differential capacitance on frequency is unknown; therefore, one should consider as approximate the literature data on the concentration of acceptors in diamond calculated on the basis of Schottky's theory for an arbitrary frequency of the measured signal. Nonetheless, it appears more convenient to use the dependence s72 ± E (especially for the cases when the expo- nent a is close to 1) or C¡2 calc ± E for qualitative comparison of electrodes made of the same semiconductor material.VI. Kinetics of the electrode reactions 1. Kinetics of outer-sphere reactions. Metal-like diamond electrodes It is common practice to divide electrochemical reactions into outer-sphere (occurring without cleavage of interatomic bonds in the reacting molecule) and inner-sphere ones (occurring with the cleavage or formation of such bonds). An example of outer-sphere reactions is the electron transfer in the Fe(CN)3¡=4¡ system, while the inner-sphere reactions may be exemplified by the oxidation and reduction of water molecules with the evolution ofO2 andH2, respectively. Let us start discussion of the kinetics of reactions on the diamond electrode with the former type of reactions.A number of outer-sphere reactions were investigated by the method of potentiodynamic curves with a linear potential sweep. This method proved to be convenient for both qualitative and quantitative characterisation of the electrode behaviour. The theory of this method has been considered in textbooks.71, 72 6 Typical cyclic voltammograms recorded for four redox systems are presented in Fig. 9. The outer-sphere reactions occur in solutions of Fe(CN)3¡=4¡, Ru(NH3)2á=3á and IrCl2¡=3¡, while the inner-sphere raction occurs in the quinone/hydroquinone system. A characteristic shape of all curves with peaks of anodic and cathodic currents demonstrates that in all cases the stage of charge transfer at the interface is rather fast; the reaction is hampered due to the insufficiently rapid reagent mass transfer in solution toward the electrode surface. Similar curves were a 1.0 0.8 0.6 0.4 0.2 0 E, B 0.6 6 6 6 387 b a I /mA I /mA 4 4 0 0 74 74 400 E /mV 200 0 E /mV 0 7200 d c I /mA I /mA 4 4 0 0 74 74 0 700 500 400 E /mV 900 E /mV 7400 6 6 6 Figure 9.Cyclic voltammograms in 1074 M solutions of Fe(CN)3¡=4¡ (a), Ru(NH3)2á=3á (b), IrCl2¡=3¡ (c) against the background of 0.1 M KCl and quinone/hydroquinone (d) against the background of 0.1 M HClO4.73 Potentials were measured relative to the SCE. obtained in a series of studies of the ferrocyanide/ferricyanide 74 and quinone/hydroquinone systems,75 Ce3+ ions (see Ref.76), etc. Analysis of these curves makes it possible to determine the following characteristics of the diffusion and interphase charge transfer: diffusion coefficient, transfer coefficients of the cathodic (a) and anodic (b) reactions and the rate constant (k0). Quantita- tively, the larger the difference between the potentials of the cathodic and anodic peaks the less reversible the electrode reaction. According to the theory of electrode kinetics,71, 72 for a reversible one-electron reaction at room temperature, DEp=56 mV irrespective of the reagent concentration and the rate of potential sweep. For the systems, the voltammograms of which are shown in Fig. 9a ± c, the values of DEp are 88, 72 and 65 mV, respectively.73 Consequently, on a relatively strongly doped polycrystalline electrode, the processes under consideration occur in the mode of a quasi-reversible reaction.On the contrary, the electrode reaction is markedly irreversible in the quinone/ hydroquinone system: DEp=560 mV. These qualitative esti- mates correlate with the values of rate constants determined for the above-listed redox systems.77 Reagent Fe(CN)3¡=4¡ Ru(NH3)2á=3á IrCl2¡=3¡quinone/hydroquinone 6 6 6 461073 <1076 361073 k0 /cm s71 961074 The rate constants strongly depend on both the level of diamond doping with boron (see below) and the prior preparation of the electrode surface; these data are presented here for the purpose of illustration.On the whole, based on the studies of a series of systems [including methyl viologen; FeL3(ClO4)2, where L is o-phenanthroline; 1,10-dimethoxyferrocene, etc.], the follow- ing qualitative inferences have been made:78, 79 the outer-sphere reactions are closer to reversible reactions that the inner-sphere ones; the reactions in the systems with a positive equilibrium potential are closer to reversible reactions than those in the systems with a negative potential; } on the heavily doped metal- like diamond the reaction rate is higher than on the moderately doped electrodes characterised by semiconductor properties. But even on the metal-like diamond electrodes the reaction rate is appreciably lower than that on the metal electrodes with high electrocatalytic activity, e.g., platinum.Consequently, upgrading } This regularity is invalid for the Ce3+/4+ system (see Ref. 75) which is discussed in more detail in the next section.388 of electrocatalytic properties of diamond electrodes intended for practical applications is a topical problem. First experiments have shown 80, 81 that the deposition of microquantities of platinum onto the diamond electrode surface reduces the overvoltage of the cathode reactions. 2. Semiconductor properties of diamond and the electrode kinetics 6 The systems Fe(CN)3¡=4¡, quinone/hydroquinone } and Ce3+/4+ have been investigated quantitatively.75, 80, 82 The dependences of the potentials of the anodic and cathodic current peaks Ep in the potentiodynamic curves on the logarithm of the rate of potential sweep u were determined experimentally; the data characterising the quinone/hydroquinone system are shown in Fig.10.75 The transfer coefficients of the anode and cathode reactions were calculated from the slope of the straight line (Fig. 10 c) using the theory of the method of potentiodynamic curves.72 a b ja /mA cm72 70.2 0 0.2 E /V 0 1 5 2 3 71 43472 2 1 3210 5 73 0.6 0.8 E /V jc /mA cm72 c Ep,a /V Ep,c /V 0.2 0.8 0 0.6 70.2 71 72 log u (V s71) Figure 10. Potentiodynamic curves recorded during the linear potential sweep in the course of hydroquinone (0.01 M) oxidation (a) and quinone (0.01 M) reduction (b) in 0.5 M H2SO4 on a polycrystalline diamond electrode, and dependences of the potentials of the cathodic (Ep,c) and anodic (Ep,a) current peaks on the logarithmof the sweep rate (c).75 The rate of potential sweep /V s71: (1) 0.005; (2) 0.01; (3) 0.02; (4) 0.05; (5) 0.1. Potentials were measured relative to the Ag/AgCl electrode.Table 2 presents the measured a and b values typical of the 6 single-crystal and polycrystalline thin-filmelectrodes and a single crystal (grown under high pressure and high temperature) in Ce3+ and Ce4+ solutions. For the Ce3+/4+ system(as for the two other systems, viz., Fe(CN)3¡=4¡ and quinone/hydroquinone), the transfer coefficients are less than 0.5, especially for the cathodic reaction, and the suma+b proves to be less than unity.Accord- ing to the theory of the electrochemical kinetics on semiconduc- tors,6 in the case of an `ideal' p-type semiconductor electrode, the transfer coefficients for the reactions occurring with the partic- ipation of holes should be as follows: a=0, b=1; a+b=1. In practice, however, the ideal case is extremely rare even for single- crystal semiconductor materials (silicon, gallium arsenide, etc.) that are prepared using much more advanced technologies than the technology for the preparation of polycrystalline CVD- diamond. The reason for the deviation from `the ideal situation' }Note that the reactions occurring in the quinone/hydroquinone system are inner-sphere ones. Yu V Pleskov Table 2.Kinetic characteristics of diamondelectrodes ina 0.01 MCe3+ (or Ce4+)+0.5 M H2SO4 solution.Electrode a b k0 /cm s71 Polycrystalline film 55 0.16 0.46 661079 Single-crystal film 55 0.2 0.43 6.561077 Single crystal grown 0.33 0.27 4.461076 under high pressure and high temperature 63 6 appears to lie in the fact that during the electrode polarisation, part of the interphase potential drop is lost in the Helmholtz layer (e.g., due to a high density of the surface states). As a result, the transfer coefficients a and b have the values intermediate between those characteristic of semiconductors (0 or 1) and metals (*0.5). Extrapolation of the impedance complex-plane plot until its intersection with the ReZ axis (Fig 6a) gives the diamond film resistance Rs from which the diamond resistivity r can be calculated. The dependences of RF on r for two redox systems are presented inFig.11. It is seen that, first, the reaction rate in the Fe(CN)3¡=4¡ system is approximately three orders of magnitude higher than that in the quinone/hydroquinone system, which is consistent with the above-cited rate constants; second, RF is observed to be directly proportional to r over a wide range of diamond resistivity. The existence of such a dependence for diamond electrodes was first reported by Modestov et al.80 It reflects the fundamental law of the kinetics of reactions on semiconductor electrodes: the rate of an electrochemical reaction is proportional to the concentration of free carriers on the electrode surface. Indeed, the concentration of holes in the diamond bulk p0 is inversely proportional to its resistivity r.{ The concentration of holes on the electrode surface ps is related to p0 by the Boltzmann equation 6 ps=p0exp ¡efsc , kT where fsc is the potential drop in the space charge layer.If the chemical interaction of diamond with the solution and hence the potential distribution over the diamond/solution inter- face is assumed to be independent of the level of diamond doping withboron, thenat the givenelectrode potential whichis set by the redox system the value of efsc is the same for all electrodes. Consequently, the reaction rate (*1/RF) should be proportional logRF (Ohm cm2) 1 2 65432 4 5 6 7 logr (Ohm cm) 6 Figure 11.Dependence of the Faraday resistance under equilibrium potential in the Fe(CN)3¡=4¡ (1) and quinone/hydroquinone (2) systems on the filmresistivity.83 { In the discussion of the relation between the concentration of holes in specimens and their resistivity, it is assumed that the mobility of holes is independent of the boron concentration in specimens.Synthetic diamond in electrochemistry to the diamond conductivity, which is in fact observed in experi- ments. 3. Crystal structure and the electrode kinetics: comparison of electrodes from single-crystal and polycrystalline diamond and amorphous carbon Polycrystalline diamond films are heterogeneous systems consist- ing of diamond crystallites and disordered carbon of intercrystal- lite boundaries (see Section III.1). In this context, the following question arises: to what extent do the intercrystallite boundaries influence the electrochemical properties of polycrystalline dia- mond electrodes? In order to answer this question, one should compare the electrochemical behaviour of crystalline diamond with that of diamond-like carbon, which may be arbitrarily regarded as a model for the material of intercrystallite boundaries in the polycrystalline diamond. Such a comparison shows that the kinetic characteristics (a and b) of single-crystal and polycrystalline diamond electrodes are similar. In solutions of redox systems, both types of diamond electrodes possess electrocatalytic activity (though not very great, which is characteristic of semiconductor electrodes).6 The properties of amorphous carbon vary within very wide limits. In their electrode properties, the `graphite-like' thin-film electrodes resemble conventional carbonaceous materials (e.g., glassy carbon); redox reactions occur on them at high rates. On the contrary, in the case of `diamond-like' electrodes there are no peaks of oxidation and reduction currents in such systems as Ce3+/4+, Fe(CN)3¡=4¡ and quinone/hydroquinone; the currents measured do not differ from the background current. The fact that the diamond-like carbon proved to be electrochemically inactive suggests that the intercrystallite boundaries make no practical contribution to the electrode behaviour of polycrystalline dia- mond; it is totally determined by diamond crystallites.This conclusion relies on the results of comparison of the impedance (Table 1) and kinetic (Table 2) characteristics of polycrystalline and single-crystal diamond films and seems to be valid at least in the range of moderate electrode polarisation.55 On the other hand, at high anode potentials (+1.7 V relative to the NHE, just before the beginning of oxygen evolution) a small peak is observed in the polarisation curve of a polycrystalline electrode in an indifferent electrolyte solution (similar to that shown in Fig. 3a).32 Its height is as small as *1 mA cm72; therefore, it can hardly be seen in Fig. 3a (and other similar graphs). As to the quantity of electricity passed corresponding to this current peak, it may be concluded that within one such 'passage' of the curve, *0.05% of the monolayer substance of the electrode surface takes part in the reaction.This current peak is better expressed for the diamond films with a relatively high content of non-diamond carbon, but it is not observed on single- crystal electrodes. According to Martin et al.,32 this peak is due to the oxidation of the non-diamond carbon of the intercrystallite boundaries, which is less resistant to oxidation than the crystalline diamond. No other peculiarities of the electrode behaviour of diamond, which could be attributed to the intercrystallite boun- daries, have been revealed. Note in conclusion that the wide-bandgap diamond-like carbon a-C:H which is usually electrochemically inactive acquires an electrode activity after the incorporation of a sufficiently large quantity (*10%) of platinum into the bulk of film (in the course of its deposition).In the shape of their potentiodynamic curves and the values of kinetic parameters, the a-C:H/Pt electrodes thus obtained resemble qualitatively the polycrystalline (boron-doped) diamond electrodes. For example, in the redox system Ce3+/4+ no current peaks in the curves are observed at the Pt level ranging from traces to 1%, but they become well expressed at 12% of platinum; coefficients a and b being equal to 0.13 and 0.18, respectively. The redox reactions are irreversible; they occur under a mixed diffusion-kinetic control and have a moderate overvoltage.61 389 The results obtained point to the possibility of creation of fairly active electrodes from corrosion-resistant carbonaceous materials, viz., synthetic diamond and diamond-like carbon, by incorporating into their bulk rather small quantities of metal catalysts such as platinum.Platinum can be distributed in a carbon material in the form of individual atoms or clusters depending on the mode of its incorporation.82 An essential feature is that platinum is not a doping admixture in a-C:H in the sense of this term as used in the physics of semiconductors. Impedance measurements have shown that the Pt admixture does not virtually change the material resistivity. Therefore, it may be suggested that the effect of platinum is of a purely catalytic nature: its atoms on the electrode surface play the role of active sites on which the adsorption and the electrode reaction proceed at high rates.50, 61 4.Kinetics of inner-sphere reactions. Diamond electrodes in preparative electrolysis and electroanalysis The evolution of oxygen from water on the anode and hydrogen on the cathode are regarded as the inner-sphere processes. Poten- tials of these processes restrict the region of `ideal polarisability' in aqueous solutions of electrolytes (see Fig. 3a,b). Overvoltage of these reactions increases with higher quality of electrode films. For the anodic evolution of oxygen, the exchange current is 5610711 A cm72, transfer coefficient b=0.25; for the hydrogen evolution on the cathode the exchange current is 10710 A cm72 and a=0.30 (see Ref.32). The region of `ideal polarisability' may be restricted by decomposition potentials of not only the solvent but also dis- solved electrolyte, e.g., halides. The reactions of anodic oxidation of Cl7, Br7 and I7 ions with the evolution of the corresponding halogens proceed irreversibly and are accompanied by a much higher (by 1 V for Cl7) overvoltage than on the platinum or graphite electrodes. Most probably, this is associated with low adsorbability of intermediate products (chlorine, bromine and iodine atoms) on the diamond surface, which thus proves to be passive.78 (The outer-sphere reactions discussed in Section VI.1 occur without adsorption of reagents or intermediate products on the electrode; for this reason, they are insensitive to the poor adsorption capacity of the diamond surface.) In addition to the above-mentioned processes, the inner- sphere reactions also include oxidations of numerous organic compounds, in which the interphase charge transfer is accompa- nied by transformations of organic molecules.Polycrystalline diamond electrodes have been used to investigate the anodic oxidation of chloropromazine,41 ascorbic acid,84 anthraquinone- 2,6-disulfonate 85 and several amines.86 One of the most efficient and promising applications of the diamond electrodes is their use for the electroreduction of com- pounds that are difficult to reduce. For example, the reduction of nitrate and nitrite ions to ammonium has been investigated.87 ± 89 In an alkaline solution ofNO73 , the following reaction takes place on a heavily doped electrode (with Na up to 1021 cm73) under anodic polarisation: NH4OH+9OH7 .NO¡3 +7H2O+8e At a potential of 75 V, the current density reaches 300 mA cm72 but no corrosion of diamond is observed even after a few days of continuous electrolysis. The most striking feature of this process is that the anodic current efficiency exceeds several-fold the theoretical limit calcu- lated by the equation for the reaction! Attempts were undertaken to rationalise this fact by the occurrence of the conjugated oxidation of silicon (the diamond film was prepared by deposition on a silicon substrate), e.g., through pinholes in the film.However, special experiments on a free-standing diamond film (see Section II) have shown that the current efficiency, though being somewhat decreased, still remained higher than its theoretical level. The reason of this effect is still unclear.390 The electrodeposition of metals (Li, Ag, Cu, Pb, Pt, Hg) leading to the formation of a new phase stands somewhat apart. Data on such processes are scanty and rather contradictory. Lead and mercury are deposited as islets a few micrometers in diameter predominantly over the intercrystallite boundaries.81 According to Li et al.,18, 90 lithium is deposited in the same mode with concomitant intercalation into the non-diamond carbon. Obser- vation of the lithium intercalation ± deintercalation may be used as a technique for revealing non-diamond inclusions on the surface of diamond films.Finally, there is indirect evidence that the electrodeposition of copper, the first stages of which follow the `underpotential deposition' mechanism, also takes place on the intercrystallite boundaries.8 Electroanalytical methods have been developed on the basis of certain electrochemical reactions on diamond electrodes. Judging from the results of relevant studies,8, 81 the diamond electrode is well-suited for use in the commonly accepted anodic-stripping analysis of solutions for low concentrations of metal ions, which consists in the cathodic deposition of small quantities of metals on an inert working electrode with their subsequent anodic dissolu- tion.A flow-type thin-layer cell with a diamond electrode has been devised 85, 91, 92 for the determination of ethylamine, ethylenedi- amine and other amines, as well as azide ion N73 .Such determi- nations are carried out by injecting a microportion of the solution under study into the electrolyte flow with subsequent recording of the current jump induced by the oxidation of the compound under analysis (Fig. 12). The electrode function is linear in the concen- tration range of 1072± 1075 M; the detection limit, e.g., of azide ion, is 8 nmol l71 (see Ref. 92). I /mA0.2 mA 2 min t /min Figure 12. Chronoamperogram in the determination of ethylamine con- centration by injecting an aliquot (20 ml of 0.25 mM solution) into a flow- type cell with 0.1 M carbonate buffer; solution flow rate, 0.5 ml min71 (see Ref.91). VII. Photoelectrochemistry of diamond Like any other semiconductor, the diamond demonstrates a photoelectric effect due to photogeneration of free charge carriers upon illumination. An increase in the concentration of charge carriers affects the current across the diamond contact with another conducting phase or the interface charge. In the potentio- static regime, a photocurrent is recorded and a photopotential is recorded in the galvanostatic and coulombostatic regimes. Starting from the first study,11 the photoelectrochemistry of diamond has been investigated by illuminating electrodes with light with quanta energy hn smaller than the bandgap (Eg ^ 5.5 eV); this energy is insufficient for the transition of excited electrons from the valence band to the conduction band. Free carriers were apparently formed due to ionisation of some unidentified photosensitive sites in the bandgap of diamond, viz., admixtures or structural defects (the so-called extrinsic or impur- ity photosensitivity).Curves of photocurrent in the diamond electrode elicited by band-to-band electron transitions (the so- a jph /mA cm72 1 0 740 2 780 7120 b jph /mA cm72 4 1 0 2 74 0 70.5 7871.0 Figure 13. Dependence of dark current (1) and photocurrent (2) upon illumination of a polycrystalline diamond electrode in 0.1 M KH2PO4 under intrinsic (a) and impurity-related (b) photoexcitation.The quanta energies were 6.4 eV (a) and 5 eV (b). Density of illumination power was 80 mW cm72 (Ref. 93); the potential values were measured relative to the SCE. called intrinsic photosensitivity 93) have been derived quite recently using an excimer laser with a wavelength of 193 nm (which corresponds to quantum energy hn=6.4 eV) as the high- energy light source. Figure 13 illustrates well the differences in the photocurrent density jph in the intrinsic and impurity-related photoexcitations. In the impurity-related excitation the quantum yield of photocurrent was *0.01%,94 whereas in the intrinsic excitation it reached 34%.93 The absolute magnitude of photocurrent is increased with increase in cathodic polarisation.This effect may be understood by examining Fig. 14 showing the square of photocurrent as a function of the potential.11 This plot is equivalent to the Mott ± - Schottky plot (see Fig. 5) because, like the latter, it reflects the variation of the width of the space charge layer in the semi- conductor diamond as a function of the potential. Upon illumi- nation of this semiconductor, the light-generated charge carriers with opposite signs are separated by an electric field in the space charge region. The minority carriers drift in the electric field Ip2h/mA2 0.3 0.2 0.1 0 70.5 71 Figure 14. Dependence of the squared photocurrent on the diamond electrode potential in 1 M KCl;11 the potential values were measured relative to the SCE.Yu V Pleskov 1.0 E /V 0.5 0.5 E /VSynthetic diamond in electrochemistry toward the interface, where they are `captured' by the electro- chemical reaction or they charge (in the coulombostatic regime) the electrode, whereas the majority carriers migrate into the bulk of diamond and then, through the ohmic contact, to the external circuit of the electrochemical cell. Therefore, the thicker the space charge layer from which the minority carriers are `harvested' the stronger the photocurrent. By extrapolating the straight line in Fig. 14 until its intersection with the potential axis, one can determine the flat-band potential in the same way as when using the Mott ± Schottky plot; both methods give consistent values for the same electrode (see Section V.1). In a series of studies 11, 93, 95 the photocurrent was measured in the steady-state mode upon illumination of the electrode with light of constant intensity. Later,94, 96, 97 pulsed illumination was used; the photopotential and dynamics of its decay were measured after the application of a light pulse. The current of photogenerated minority charge carriers charges the electrode capacitance and thereby determines the potential shift.The analysis conducted with the aid of a somewhat more complex version of the Randles ± Ershler circuit (see Fig. 4b) { afforded a satisfactory description of the dynamics of photopotential decay. A typical spectrum of the cathodic photocurrent Iph of a diamond electrode is shown in Fig.15a. Processing of such spectra in the coordinates (Iphhn)1/2 vs. hn (see Fig. 15b) provided values of threshold energies of indirect electron transitions that determine the photogeneration of charge carriers (for more details, see Ref. 6). In some cases, the spectra include two regions and give, respectively, two values of threshold energy. The problem of the elucidation of the physical nature of the energy levels in the bandgap of diamond which participate in such transitions is still far from its solution. Some transitions with a threshold energy close to 4 eV have been identified with the photoexcitation of electrons of the valence band on the energy levels of N+ ions in the bandgap of diamond localised *1.7 eV below the conduction band.94 b a Iph /nA (Iphhv)1/2 (rel. units) 105 20 70 10 35 0 500 350 0200 2.0 3.0 4.0 hv /eV 1.0 l /nm Figure 15.Spectrum of the cathodic photocurrent of a boron-doped polycrystalline diamond electrode in 0.5 M H2SO4 under a potential of 0.05 V (relative to NHE) (a) and its presentation in the (Iphhn)1/2 vs. hn coordinates, which is convenient for the spectrum processing for the threshold energy determination.94 The photocurrent ± potential curves and the photocurrent spectra were also recorded for the films of amorphous diamond- like carbon.48, 49, 98 Processing of the spectra similar to that described above has made it possible to determine the width of the mobility gap in the disordered structure of this material.Upon illumination, the diamond-like carbon displays properties char- acteristic of intrinsic semiconductors. Photoelectrochemical measurements make it possible to derive certain characteristics of the semiconductor diamond. Thus comparison of the spectra of light absorption coefficient and open-circuit photopotential allowed determination of the { The Randles ± Ershler circuit was supplemented with a branch which included a capacitance and a resistance connected in series and modelled the process of charging the surface states. 391 diffusion length (2 ± 4 mm) of the minority carriers in polycrystal- line diamond films.99 VIII. Conclusion About ten years have passed since the first studies in the field of diamond electrochemistry. This decade has witnessed the develop- ment of a somewhat schematic but rather comprehensive system of concepts of diamond as a novel electrode material.Investiga- tions of diamond electrodes have afforded better understanding of the relationship between the crystal structure of solids and their electrode behaviour. It is most likely that further studies on diamond electro- chemistry will develop both extensively (a wider range of studied reactions, search for new applications of diamond electrodes in fundamental and applied electrochemistry) and intensively (a deeper insight into the relationship between the semiconductor and structural properties of carbon allotropes and their electro- chemical behaviour). In the near future, the corrosion-resistant diamond film electrodes may be expected to replace (at least in part) the conventional dimension-stable electrodes based on the platinum group metals, which are used nowadays for the electro- chemical synthesis and analysis.Preparation of this review was supported by the Russian Foundation for Basic Research (Project No. 98-03-32057). References 1. V S Vavilov, A A Gippius, E A Konorova Elektronnye i Opticheskie Protsessy v Almaze (Electronic and Optical Processes in Diamond) (Moscow: Nauka, 1985) 2. J Field (Ed.) The Properties of Natural and Synthetic Diamond (London; New York: Academic Press, 1992) 3. K E Spear J. Am. Ceram. Soc. 72 171 (1989) 4. V S Vavilov Usp. Fiz. Nauk 167 17 (1997) a 5. V S Vavilov Usp. Fiz. Nauk 164 287 (1994) a 6.Yu Ya Gurevich, Yu V Pleskov Fotoelektrokhimiya Poluprovod- nikov (Photoelectrochemistry of Semiconductors) (Moscow: Nauka, 7. M Iwaki, S Sato, K Takahashi, H Sakairi Nucl. Instrum. Methods, 8. F Bouamrane, A Tadjeddine, R Tenne, J E Butler, R Kalish, 1983) Pt II 209/210 1129 (1983) C Levy-Clement J. Phys. Chem. B 102 134 (1998) 9. B Miller, R Kalish, L C Feldman, A Katz, N Moriya, K Short, A E White J. Electrochem. Soc. 141 L41 (1994) 12. D V Fedoseev, V P Varnin, B V Derjaguin Usp. Khim. 53 753 (1984) 10. J Robertson Adv. Phys. 35 317 (1986) 11. Yu V Pleskov, A Ya Sakharova,M D Krotova, L L Bouilov, B V Spitsyn J. Electroanal. Chem. 228 19 (1987) [Russ. Chem. Rev. 53 435 (1984)] 13. V P Varnin, in Tekhnika Sredstv Svyazi (Tekhnologiya Proizvodstva i Oborudovanie) [Technology of Means of Communication (Production Technology and Equipment)] (Moscow: TsOONTIEKOS, 1989) No.1, p. 44 14. R Locher, J Wagner, F Fuchs,M Maier, P Gonon, P Koidl Diam. Relat. Mater. 4 678 (1995) 15. M Werner, R Job, A Zaitsev, W R Fahrner, W Seibert, C Johnston, P R Chalker Phys. Status Solidi A 154 385 (1996) 21. Yu V Pleskov, A R Tameev, V P Varnin, I G Teremetskaya 16. P Bou, L Vandenbulcke J. Electrochem. Soc. 138 2991 (1991) 17. K Hashimoto, Y Muraki, R Baba, A Fujishima J. Electroanal. Chem. 397 339 (1995) 18. L-F Li, D A Totir, N Vinokur, B Miller, G Chottiner, E A Evans, J C Angus, D A Scherson J. Electrochem. Soc. 145 L85 (1998) 19. P J Fallon, L M Brown Diam. Relat. Mater.2 1004 (1993) 20. Yu Pleskov, A Tameev, V Varnin, I Teremetskaya, A Baranov Fiz. Tekhn. Poluprovod. 31 1142 (1997) J. Solid State Chem. 3 25 (1998) 22. D M Malta, J A von Windheim, H A Wynands, B A Fox J. Appl. Phys. 77 1536 (1995)392 23. M Deguchi, M Kitabatake, T Hirao Thin Solid Films 281/282 267 (1996) 24. L S Pan, S Han, D R Kania,M A Plano, M I Landstrass Diam. Relat. Mater. 2 820 (1993) 25. A Ya Sakharova, A E Sevast'yanov, Yu V Pleskov, G L Teplitskaya, V V Surikov, A A Voloshin Elektrokhimiya 27 263 (1991) b 26. N G Bukun, E A Ukshe, E I Moskvina, A E Ukshe Elektrokhimiya 21 269 (1985) b 27. G A Sokolina, A A Botev, L L Builov, S V Bantsekov, O I Lazareva, A F Belyanin Fiz. Tekhn. Poluprovod. 24 175 (1990) 28.S Nath, J I B Wilson Diam. Relat. Mater. 5 65 (1996) 29. Y Muto, T Sugino,K Kobashi, J Shirafuji Jpn. J. Appl. Phys. 2 31 L4 (1992) 30. M M Bataineh, D K Reinhard Diam. Relat. Mater. 6 1689 (1997) 31. T Yano, D A Tryk,K Hashimoto, A Fujishima J. Electrochem. Soc. 145 1870 (1998) 32. H B Martin, A Argoitia, U Landau, A B Anderson, J C Angus J. Electrochem. Soc. 143 L133 (1996) 33. GMSwain, A B Anderson, J C Angus Mater. Res. Bull. 23 56 (1998) 34. J W Strojek,M C Granger, G M Swain, T Dallas, M W Holtz Anal. Chem. 68 2031 (1996) 35. G M Swain J. Electrochem. Soc. 141 3382 (1994) 36. Q Chen,M Granger, T E Lister, G M Swain J. Electrochem. Soc. 144 3806 (1997) 37. R DeClements, G M Swain J. Electrochem. Soc. 144 856 (1997) 38. P U Natishan, A Morrish Mater.Res. 8 269 (1989) 39. I Garsia, A Conde, J J de Danborenea, A J Vazquez Thin Solid Films 310 217 (1997) 40. R Ramesham, M F Rose Corros. Sci. 39 2019 (1997) 41. C H Goeting, F Jones, J S Foord, J C Eklund, F Marken, R G Compton, P R Chalker, C Johnston J. Electroanal. Chem. 442 207 (1997) 42. R Ramesham, M F Rose Diam. Relat. Mater. 6 17 (1997) 43. R Ramesham, M F Rose Thin Solid Films 300 144 (1997) 44. R Ramesham, M F Rose Thin Solid Films 315 222 (1998) 45. E P Smirnov, O G Taushkanova, V B Aleskovskii Dokl. Akad. Nauk SSSR 290 901 (1986) c 46. P E Pehrsson,M Marchywka, J P Long, J E Butler, in Applications of Diamond Films and Related Materials (Proceedings of the 3rd International Conference), Gaitherburg, MD, 1995 p.267 47. A T Howe J. Electrochem. Soc. 134 2470 (1987) 48. A Ya Sakharova, Yu V Pleskov, F Di Quarto, S Piazza, C Sunseri, S S Gerasimovich, V V Sleptsov J. Electroanal. Chem. 398 13 (1995) 49. A Ya Sakharova, Yu V Pleskov, F Di Quarto, S Piazza, C Sunseri, S S Gerasimovich, V V Sleptsov Elektrokhimiya 32 1298 (1996) b 50. M D Krotova, Yu E Evstefeeva, Yu V Pleskov, V V Elkin, A M Baranov Elektrokhimiya 34 1039 (1998) b 51. Z B Stojnov, B M Grafov, B Savova-Stojnova, V V Elkin Elektrokhimicheskii Impedans (Electrochemical Impedance) (Moscow: Nauka, 1991) 52. A Ya Sakharova, Yu V Pleskov, F Di Quarto, S Piazza, C Sunseri, I G Teremetskaya, V P Varnin J. Electrochem. Soc. 142 2704 (1995) 53. W P Kang, Y Gurbuz, J L Davidson, D V Kerns J.Electrochem. Soc. 141 2231 (1994) 54. J A von Windheim, V Venkatesan, D M Malta, K Das Diam. Relat. Mater. 2 841 (1993) 55. Yu E Evstefeeva, M D Krotova, Yu V Pleskov, V V Elkin, V P Varnin, I G Teremetskaya Elektrokhimiya 34 1171 (1998) b 56. A Sakharova, L Nyikos, Yu Pleskov Electrochim. Acta 37 973 (1992) 57. Yu V Pleskov, V Ya Mishuk, M A Abaturov, V V Elkin, M D Krotova, V P Varnin, I G Teremetskaya J. Electroanal. Chem. 396 227 (1995) 58. R Ramesham Thin Solid Films 322 158 (1998) 59. YuVPleskov,VYa Mishuk,MAAbaturov,VVElkin,MDKrotova, V P Varnin, I G Teremetskaya Elektrokhimiya 33 67 (1997) b 60. B A Boukamp Solid State Ion. 20 31 (1986) 61. Yu V Pleskov, Yu E Evstefeeva, M D Krotova, V V Elkin, A M Baranov, A P Dement'ev Diam.Relat. Mater. 8 64 (1999) 62. Yu V Pleskov, V V Elkin, M A Abaturov,M D Krotova, V Ya Mishuk, V P Varnin, I G Teremetskaya J. Electroanal. Chem. 413 105 (1996) 63. Yu E Evstefeeva, M D Krotova, Yu V Pleskov, V A Laptev Elektrokhimiya 35 138 (1999) b Yu V Pleskov 64. Yu V Pleskov, Yu E Evstefeeva, M D Krotova, V A Laptev Electrochim. Acta (1999) (in the press) 65. M A Abaturov, V V Elkin,M D Krotova, V Ya Mishuk, A Ya Sakharova, Yu V Pleskov Elektrokhimiya 31 1214 (1995) b 66. G Oskam, D Vanmaekelbergh, J J Kelly J. Electroanal. Chem. 315 65 (1991) 67. E C Dutoit, R L van Meirhaeghe, F Cardon, W P Gomes Ber. Bunsenges. Phys. Chem. 79 1206 (1975) 68. J F McCann, S P S Badwal J. Electrochem. Soc. 129 551 (1982) 69. Yu E Evstefeeva, M D Krotova, Yu V Pleskov, V M Mazin, V V Elkin, V Ya Mishuk, V P Varnin, I G Teremetskaya Elektrokhimiya 34 1493 (1998) b 70. T Paikossy, in The 1997 Joint Meeting of ECS (192nd) and ISE (48th) (Abstracts of Reports), Paris, 1997 p. 972 71. P Delahay New Instrumental Methods in Electrochemistry (New York: Interscience, 1954) 72. Z Galus Teoretyczne Podstavy Electroanalizy Chemicznej (Warszawa: Panstwowe Wydawnictwo Naukowo, 1971) 73. J Xu,M C Granger, Q Chen, J W Strojek, T E Lister, G M Swain Anal. Chem. 69 591A (1997) 74. G M Swain, R Ramesham Anal. Chem. 65 345 (1993) 75. Yu E Evstefeeva, Yu V Pleskov, V P Varnin, I G Teremetskaya Elektrokhimiya 34 234 (1998) b 76. Y Maeda, K Sato, R Ramaraj, D A Tryk,A Fujishima Electrochim. Acta (1999) (in the press) 77. S Alehashem, F Chambers, J W Strojek, G M Swain, R Ramesham Anal. Chem. 67 2812 (1995) 78. N Vinokur, B Miller, Y Avyigal, R Kalish J. Electrochem. Soc. 143 L238 (1996) 79. R Ramesham, M F Rose J. Mater. Sci. Lett. 16 1693 (1997) 80. A D Modestov, Yu V Pleskov, V P Varnin, I G Teremetskaya Elektrokhimiya 33 60 (1997) b 81. M Awada, J W Strojek, G M Swain J. Electrochem. Soc. 142 L42 (1995) 82. K I Schiffmann Mikrochim. Acta 125 107 (1997) 83. A D Modestov, Yu E Evstefeeva, Yu V Pleskov, V M Mazin, V P Varnin, I G Teremetskaya J. Electroanal. Chem. 431 211 (1997) 84. R DeClements, G M Swain, T Dallas, M W Holtz, R Herrik III, J L Stickney Langmuir 12 6578 (1996) 85. J Xu, Q Chen, G M Swain Anal. Chem. 70 3146 (1998) 86. M D Koppang, J Blau, G M Swain Anal. Chem. (1999) (in the press) 87. R Tenne, K Patel, K Hashimoto, A Fujishima J. Electroanal. Chem. 347 409 (1993) 88. C Reuben, E Galun, H Cohen, R Tenne, R Kalish, Y Muraki, K Hashimoto, A Fujishima, J M Butler, C Levy-Clement J. Electroanal. Chem. 396 233 (1995) 89. F Bouamrane, A Tadjeddine, J E Butler, R Tenne, C Levy-Clement J. Electroanal. Chem. 405 95 (1996) 90. L-F Li, D Totir, B Miller, G Chottiner, A Argoitia, J C Angus, D Scherson J. Am. Chem. Soc. 119 7875 (1997) 91. S Jolley,M Koppang, T Jackson, G M Swain Anal. Chem. 69 4099 (1997) 92. J Xu, G M Swain Anal. Chem. 70 1502 (1998) 93. L Boonma, T Yano, D A Tryk, K Hashimoto, A Fujishima J. Electrochem. Soc. 144 L142 (1997) 94. A Ya Sakharova, Yu V Pleskov, F Di Quarto, S Piazza, C Sunseri, I G Teremetskaya, V P Varnin Elektrokhimiya 31 188 (1995) b 95. Yu V Pleskov, V P Varnin, I G Teremetskaya, A V Churikov J. Electrochem. Soc. 144 175 (1997) 96. Yu V Pleskov J. Chem. Vapor Depos. 5 126 (1996) 97. Yu V Pleskov, A Ya Sakharova, A V Churikov, V P Varnin, I G Teremetskaya Elektrokhimiya 32 1164 (1996) b 98. A Pinkowski, U Stimming, J Winter J. Electrochem. Soc. 136 1849 (1989) 99. Yu V Pleskov, A Ya Sakharova, E V Kasatkin, V A Shepelin J. Electroanal. Chem. 344 401 (1993) a�Uspekhi-Physics (Engl. Transl.) b�Russ. J. Electrochem. (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (5) 393 ± 414 (1999) Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions V A Barkhash,MP Polovinka Contents I. Introduction II. Conformational control of reactions III. The effect of polyfunctionality of the original substrates IV. Factors affecting the pathway of cationic rearrangements V. Selection rules VI. Computational methods for predicting the most probable pathways of multistep cationic rearrangements VII. A system of `blocks' for aliphatic models VIII. Synthetic aspects of terpene reactions in superacidic media IX. Biogenetic aspects of terpene reactions in superacidic media X. Conclusion Abstract. The main factors affecting the pathways of multistep cationic molecular rearrangements of natural terpenes and their analogues are considered.Methods for predicting the most probable pathways of cationic rearrangements are proposed. The bibliography includes 85 references. I. Introduction Carbocationic reactions play a prominent role in molecular rearrangements of organic compounds. The regularities of these reactions are most efficiently studied by modelling methods based on `direct visualisation' of ion conversions in hypernon-nucleo- philic superacidic media at superlow temperatures. Specially synthesised simple organic compounds are normally used as subjects for the investigation. The use of this approach in a particular field of organic chemistry in which cationic molecular rearrangements play a central role (e.g., terpene chemistry) seems to be highly efficient. The study of regularities of carbocationic reactions of terpenoids and their analogues pursues three main goals: 1.Elucidation of regularities of molecular rearrangements of complex polycyclic cations. 2. Broadening of the range of available natural compounds suitable for the preparation of novel substances. 3. Chemical proof of validity of biogenetic schemes. II. Conformational control of reactions Polyfunctionality and conformational mobility are the most characteristic features of terpenes. Studies of conformational mobility and its effect on the reactivity of terpenes at low temper- atures have generated an entirely new field of research, viz., the V A Barkhash, MP Polovinka Novosibirsk Institute of Organic Chemistry, Siberian Branch of the Russian Academy of Sciences, 630090 Novosibirsk, prosp.Akad. Lavrent'eva 9, Russian Federation. Fax (7-383) 234 47 52. Tel. (7-383) 234 38 70. E-mail: root@orchem.nsk.su (V A Barkhash) Received 26 June 1998 Uspekhi Khimii 68 (5) 430 ± 453 (1999); translated by R L Birnova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.931 393 393 398 399 406 406 409 411 412 414 chemistry of conformationally controlled carbocationic reactions. In synthetic organic chemistry, conformational control is usually restricted to deamination of aliphatic amines and related com- pounds, but it plays a prominent role in biogenetic reactions of carbocationic cyclisation. The discovery of conformationally controlled reactions of stable ions derived from terpenes at low temperatures has made it possible to fill the gap between the biogenetic and chemical processes and provided a way for experimental realisation of biogenetic reactions.Conformational control can be manifested both in initial steps of reactions and in subsequent transformations of the intermedi- ate carbocations. The former can be exemplified in rearrange- ments of a sesquiterpene diene, viz., caryophyllene (1). According to 13C NMR spectral data, which are consistent with the molec- ular mechanics (MM2) calculations, compound 1 may exist in four stable conformations, viz., ba, aa, ab and bb.1, 2 7120 8C Con- former bb-1 ab-1 ba-1 aa-1 12.4 14.2 17.3 DHf (MM2), 12.6 /kcal mol71 0 3 58 39 Content (MM2) (%) The experimental barrier to the transition between the pairs ba>aa and ab>bb (16.25 kcal mol71) cannot be virtually overcome at temperatures employed for the ion generation (HSO3F±SO2FCl, from 7120 to 7100 8C).Indeed, the ratio of reaction pathways of two conformers, viz., aa and bb, the trans- annular cyclisation of which gives radically different intermediate carbocations and different reaction products, correlates with the ratio of the above conformers. aa-1 ba-1 7120 8C + 2394 + 730 8C bb-1 ab-1 7120 8C 3 X + 730 8C + 9 Heating of the diene 1 in HSO3F±SO2FCl to730 8C and its subsequent neutralisation yield a mixture of enantiomeric dienes 6 and 7 in which the diene 6 with a clovane backbone [judging from optical rotation dispersion (ORD) curves)] is predominant.The reaction proceeds via the intermediate allylic ions 4 and 5. The ratio of dienes 6 and 7 corresponds to the ratio of the conformers ba-1 and bb-1 in SO2FCl at7120 8C. The stable ions derived from conformers aa and bb can be generated separately from clovene 2 and 1-substituted caryolanes 3 via ions 8 and 9. The conformational control can be effected after the carboca- tion has been formed and be manifested in different ways. The conformationally controlled molecular rearrangements of 4b,5b- (10) and 4a,5a-epoxides (11) of isocaryophyllene in superacids occur enantioselectively and result in the formation of the enantiomeric carbocations 12 and 13 from an epimeric pair of epoxides.3 H H HSO3F, SO2FCl 7130 8C O 10 H H HSO3F, SO2FCl 7130 8C O 11 Presumably, such enantioselective rearrangements take place when the opening of the epimeric monoepoxides generated from an optically active and conformationally flexible cyclic polyene occurs in acidic media with the participation of s- and p-bonds.The direction of skeletal shifts is determined by the spatial structure of the epoxide. The optically active diene ultimately gives two enantiomeric products without recourse to any optically active reagents, i.e., `partial asymmetric synthesis' takes place. The reaction which is controlled by the conformation of the intermediate carbocation may proceed in a structurally selective manner, as is the case with the low-temperature cyclisation of trans-5,6- (14) and cis-5,6-j-ionones (15) in a superacidic medium.It was reported previously 4, 5 that the structural direction of cyclisation of isoprenoids containing a conjugated 5,6-double bond does not depend on its configuration. This was explained by decreased nucleophilicity of the double bond as a result of conjugation and, ultimately, a decrease in the cyclisation rate. +4 6 +8 7 5 + *H,*Me HO+ 7H2H2O+ 12+ *H,*Me 7H2 +OH H2O+ 13 V A Barkhash,MP Polovinka More recent studies have shown 6 that protonation of the ketones 14 and 15 in HSO3F±SO2FCl at 7130 8C yields con-formers 16 and 17 of the same dication.Quenching of an acidic solution of the conformer 16 leads to elimination of the axial proton from the tertiary carbon atom linked with the side chain to give b-ionone (18). The conformer 17 eliminates the axial proton from the secondary carbon atom of the ring (in conformity with the ionic elimination rules) and is converted into a-ionone (19). H O + 5 H+ k3 6 7130 8C + O 18 14 k2 HO16 k1 O + H+ k4 H 7130 8C + 19 OH O 15 17 k3, k4>k1, k2 Thus, in this case the conformational stability of intermediate monocyclic carbocations at low temperatures ensures conforma- tional control and, as a result, the structural selectivity of the rearrangement. A more complex situation is observed in the protonation of cis- (20) and trans-j-irones (21).7 Protonation of the ketone 20 in HSO3F±SO2FCl (7130 8C) gives the dication 22 in which the methyl group at C(6) is in the trans-position relative to the side chain at C(2).The elimination of the axial proton from the secondary carbon atom in the dication 22 is accompanied by its conversion into the `apparent' ion 22; `quenching' of the latter gives trans-a-irone (24). Heating of an acidic solution of the ion 23 to720 8C results in pentadienyl ions 25 and 26; their `quenching' yields b-irone (27). Protonation of the latter at 7100 8C gives a mixture of ions 25 and 26. It is noteworthy that the ion 23 is not generated in the temperature range from 7100 to 710 8C.Structurally selective and stereo- selective transformation of the ketone 20 in HSO3F±SO2FCl is probably due to the lack of conformational transitions at 7130 8C (Scheme 1). Scheme 1 HO Mee 6 + 2 + H+ 7130 8C H + O OH 22 20 23 O 7H+ 24 HO OH O + 720 8C + H+,7100 8C 25 26 27Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions The transformation of the trans-isomer 21 in HSO3F± SO2FCl occurs differently. At 7130 8C, it gives a mixture of dications 22 and 28. The latter exists as a mixture with the dication 29 due to conformational transitions of the chair>chair type. `Quenching' of the reaction mixture gives ketones 24 and 27 and cis-a-irone (30). 24 22 H+ H 7130 8C Mee + + O MeaH 21 + + HO OH 29 28 O O 30 27 Generation of an ion, which exists in two different conforma- tions that are more rapidly stabilised by intramolecular interac- tions than by interconversions, from different conformational isomers is yet another typical example of conformational control.However, in contrast with the above examples, this process ensures not only structural selectivity but also stereospecificity of cyclisation reactions. Thus the cyclisation of trans- (31) and cis- geranylacetone (32) in 100% sulfuric or fluorosulfonic acids is stereospecific and yields trans- (33) and cis-chromenes (34), respectively.8, 9 It has been suggested 10 ± 12 that the stereospecific- ity of cyclisation of dienones 31 and 32 is due to the formation of stable conformers 35 and 36.However, it has been found 6 that monocyclic ions 35 and 36 were not detected in the NMR spectra after dissolution of dienones 31 and 32 in HSO3F±SO2FCl even at 7120 8C because of their rapid cyclisation into carboxonium ions 37 and 38. It should be noted that the methylcyclohexyl cation is a rather stable species.13 H O33 7H+ H a k3 H + + O O O 31a 37 35 k2 k1 H k4 a + H + O O 36 32a O38 7H+ H O 34 (a) HSO3F, SO2FCl,7130 8C. 395 1H and 13C NMR spectral data point to the formation of carboxonium ions 37 and 38, which have not been considered previously at all as intermediates in the cyclisation of dienones 31 and 32 into chromenes 33 and 34.10 ± 12 Presumably, the stereo- specificity of these reactions is determined not by the conforma- tional stability of monocyclic ions 35 and 36 as such, but rather by the ratio of rate constants of their conformational transitions as well as by the ratio of rate constants for the formation of bicyclic ions (k3, k4 k1, k2).6 In this case, we are dealing with conforma- tional control, since the ratio of the reaction products is approx- imately the same as the conformer ratio.14 The formation of conformationally isomeric monocyclic ions 35 and 36 from dienones 31 and 32 seems to be due to the fact that they react in the most stable conformations 31a and 32a.10 This conclusion is confirmed by the strain energies (E) of the conformations of dienones 31 and 32 calculated by the molecular mechanics method using parametrisation described in Ref.15. It was found 6 that the majority of geranylacetone atoms in conformations 31a and 32a fall into the attraction field of non-valent (van-der-Waals) inter- actions. Therefore, the energies of these conformations are by 6 ± 8 kcal mol71 lower than those of non-valent interactions in other conformations, e.g., 31b and 32b. O O 32b 31b 32b 225.5 31b 226.7 32a 219.5 Conformation E /kcal mol71 31a 218.6 The calculations of the barriers DH# to the conformational ion transitions 35>36 (MMX) and the opening of heterocycles 37 and 38 (PM3)16 corroborate the conclusion 6 that k3, k4 k1, k2. The stepwise mechanism of generation of ions 37 and 38 proposed by Gavrilyuk et al.6 agrees well with the transition state energies of cyclisation of cis- and trans-undeca-6,10-dien-2-yl cations into decalin derivatives calculated by the MINDO/3 method.These calculations showed that the activation energies of synchronous processes are by 23 ± 24 kcal mol71 higher than those of stepwise processes.17 In addition, experimental data suggest that biogenetic cyclisation of isoprenoids occurs in a stepwise manner.18 The formation of conformational isomers of cyclic ions as the manifestation of conformational control in the cyclisation of 2,6- dienes stereoisomeric with respect to the 4,5-double bond having a carbonyl or a hydroxy oxygen atom at C(1) is of general character 19 and is observed in the cyclisation of stereoisomeric 2,6,9,9-tetramethyl-2,6-decadien-10-ols. 4 HSO3F, SO2FCl + 1 7100 8C 5 OH OH + HSO3F, SO2FCl 4 1 7100 8C OH 5 OH In the above examples, the interaction of the cationic centre in the conformers of monocyclic carbocations containing suffi- ciently nucleophilic carbonyl or hydroxy groups occurs faster than conformational transitions.In the cyclisation of trans, trans- (39) and cis,trans-farnesols (40), which are isomeric at the 6,7- double bond, the configuration of this double bond does not influence the steric direction of the reaction. Both isomers yield a396 product with trans-fused rings. Apparently, this is due to decreased nucleophilicity of the 2,3-double bond caused by the 7I-effect of the hydroxy group.20 6 3 7 2 1.HSO3F, SO2FCl,7110 8C 2. MeOH 39 CH2OH 67 40 CH2OH Conformational control is also effected in the stage of formation of intermediate carbocations where the same ion exists in different conformations which give rise to different (but not isomeric) ions.21 Thus protonation of methyl 2,3-trans- (41) and cis-geranate (42) in HSO3F±SO2FCl at 7120 8C gives dications 43 and 44, which are conformers but are not practically inter- converted at temperatures below 770 8C. In the dication 43, the 1,2-hydride shift prevails, so that the cationic centres become more remote. Subsequent addition of the fluorosulfonyloxy group results in the ion 45. In the dication 44, the axial proton is eliminated exclusively from the secondary carbon atom to give the ion 46.It is interesting that `quenching' of acidic solutions of ions 45 and 46 yields the same compound, viz., methyl a-cyclo- geranate (47). Thus, the studies of intermediate carbocations generated at low temperatures under `long-life' conditions allowed Gavrilyuk et al.21 to conclude that it is the configuration of the 2,3-double bond that determines the structure of primary cyclisation products (viz., intermediate carbocations) in the cycli- sation of acyclic isoprenoids containing an ester group conjugated with the double bond. It has been shown 22 that conformational control can influence the ratio of intra- and intermolecular reaction products in the stage of generation of an intermediate carbocation. Thus stable a 50 8C HSO3F 49 48 770 8C b H 70 8C + 51 H CH2OH H H + + A H + BH H + 7H+ 54 55 H H b a O + OMe + OH MeO 41 43 a + 7H+ O + OH H MeO MeO 42 44 (a) HSO3F, SO2FCl,7120 8C; (b) OSO2F7,*H; (c) MeOH.diastereomeric allylic ions 49 ± 51 generated from (7)-a-murolene (48) differ only in the configuration of the asymmetric centre at C(6) or C(10).23 H H 1 9 2 8 10 3 + 7 5 6 4 H 48 49 This, in turn, affects the conformation of the corresponding ion and, ultimately, the dihedral angle between the vacant p-orbital of the cationic centre and the C(10) ± C(9) s-bond. This angle determines the probability of intramolecular rearrange- ments in these ions with a `turnover' of a six-membered ring relative to another.Thus one of the epimers (ion 49) undergoes intramolecular ring `turnover' at 50 8C to give the diene 52 (Scheme 2), while the other epimer (cation 50) is converted into the ion 49 at770 8C. In the case of the ion 51, only intermolecular rearrangement into the ion 54 is realised even upon heating to 70 8C; its `quenching' gives compound 55 (see Scheme 2). H MeOH 52 H *H MeOH + + V A Barkhash,MP Polovinka OH+OMe O OSO2F45 c OMe OH 47 +OMe 46 H H + + 51 50 Scheme 2 H 53Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions Intramolecular rearrangement of the ion 49 may occurs in two pathways, a or b. Which of them is realised in each concrete case can be established by comparing the ORD curves of dienes 55 and 56 mimicking compounds 52 and 53, respectively. H 56 The experimental results 21 suggest that the rearrangement follows pathway a.This is probably due to the fact that of the two secondary spirocations A and B only the former can be stabilised by homoallylic conjugation. In the case of pathway b, such an efficient interaction is impossible due to the perpendicular posi- tion of the p-orbitals of the cationic centre and the double bond. It should be noted that superacidic media favour profound molecular rearrangements that cannot be achieved in `ordinary' media even under most drastic conditions.Since solutions of stable ion salts can be heated up to 100 8C and above, direct observation of `cation ± cation' rearrangements with an activation barrier of *30 kcal mol71 and higher can be observed. Molec- ular cationic rearrangements which occur in superacidic media with an activation barrier of *20 kcal mol71 cannot be repro- duced under conditions of an ordinary acid catalysis (cf. Ref. 24). The possibility of intermediate formation of the spirocyclic carbocation 58 in the previously discovered 22 rearrangement of allylic cations of the type 49 with the cadalene skeleton has been demonstrated with a sesquiterpenoid hydrocarbon b-alaskene (57) having a spiro[4.5]decane backbone as an example.25 HSO3F, SO2FCl 1,3*H + 7120 8C 57 + *H *C C +58 H MeOH + HSO3F 20 8C H H MeOH + + Steric factors have a profound influence on the ratios of intra- and inter-molecular transformations of stable carbocations.They can also affect the regio- and stereoselectivity of these rearrange- ments as can be exemplified in the isomerisation of cis- (59) and trans-caranes (60) in HSO3F±SO2FCl (7120 8C) resulting in the formation of apparent ions 61 and 62, respectively.26 3 4 2 HSO3F, SO2FCl 1 5 7120 8C 6 7 +63 59 HSO3F, SO2FCl 7120 8C 60 64 The opening of the cyclopropane rings in epimers 59 and 60 is accompanied by seeming inversion of configuration. This is due to the fact that in the case of long-living cations the conversions of initially formed ions 63 and 64 into thermodynamically more stable cations 61 and 62 occur through hydride shifts as can be judged from the molecular mechanics calculations.27 The calcu- lations of the geometry of the ground states of compounds 59 and 60 made with the use of the parameters described in Ref.28 suggest certain elongation of the C(1) ± C(7) bond relative to the C(6) ± C(7) bond in the cis-isomer 59 and of the C(6) ± C(7) bond relative to the C(1) ± C(7) bond in the trans-isomer 60. These data are in good agreement with the results of conversions of com- pounds 59 and 60 in superacids. III. The effect of polyfunctionality of the original substrates Yet another essential feature of terpenes as starting compounds in the generation of stable ions is their polyfunctionality.Thus terpenes may serve first as potential precursors of both mono- and dications with identical skeletons. The pathways of molecular rearrangements of mono- and dications often differ, which largely increases the diversity of terpene transformations. Thus it has been shown 23 that molecular rearrangements of mono- (65) and dications (66) generated from the same natural diene 48 give rise to diastereomeric allylic ions 49 and 67. H H + a H H 48 65a H b H+ 48 + + H 66 65b (a) HSO3F, SO2FCl,770 8C; (b) HSO3F, SO2FCl,7120 8C. Second, a polyfunctional terpenoid molecule contains groups that can be both converted into an electrophile and simultane- ously act as an internal nucleophile. The direction of a reaction often depends on the structural peculiarities of the original terpenoid.Minor structural changes in such strongly internally interrelated multicentre systems can result in dramatic changes in relative reactivities of different groups, in relative stabilities of intermediate carbocations and, ultimately, in the cyclisation path- way. The results of cyclisation of analogous acyclic terpenoids, viz., citral (68), citronellal (69) and methyl geranate 41, were compared under identical conditions (HSO3F±SO2FCl, 7110 8C).29 397 +61 + + 62 H +49 H H + 750 8C + H 67398 In the citral molecule (68), the proton is added to the most nucleophilic 6,7-double bond. The oxygen atom of the carbonyl group, which attacks the cationic centre to give rise to cyclic allylic ions, acts as a nucleophile; quenching of allylic ions with alcohols gives 7-alkoxy-substituted 6,7-dihydrocitrals in good yields.OR 7 O + H+ CHO O 6 CHO ROH + + 68In contrast, in the aldehyde 69 the proton is added to the oxygen atom of the carbonyl group. The carbon atom in the latter attacks the 6,7-double bond to generate an allylic ion. Its `quenching' gives 7-methoxymenthene. + OMe + OH CHO MeOH H+ 7H2O 69 Finally, similarly to the aldehyde 68, protonation of the ester 41 results in the addition of a proton to the 6,7-double bond. However, in this case the role of a nucleophile is played not by the oxygen atom of the carbonyl group but by the 2,3-double bond, eventually resulting in the ester 47 after `quenching'.COOMe COOMe COOMe MeOH H+ + 41 47 It has been shown 30 that isomeric bicyclic labdane diterpe- noids 70 and 72, which differ in the position of the allylic hydroxy group in the side chain undergo radically different conversions under identical conditions (HSO3F±SO2FCl,7100 8C). Thus the protonation of the terpene 70 involves the exocyclic double bond to yield a cyclic carbocation; the role of the internal nucleophile is played by the double bond in the side chain. This structurally selective and stereoselective rearrangement gives rise to the bifunctionalised isoagatane 71. CH2OH CH2OH + 70 OH CH2OH 71 The generation of the cationic centre in the tertiary alcohol 72 is of by the solvolytic type and is accompanied by elimination of the allylic hydroxy group from the side chain.In this case, the exocyclic double bond plays the role of the internal nucleophile, eventually resulting in the tetracyclic hydrocarbon 73. V A Barkhash,MP Polovinka OH + 72 73 If a cationic centre is formed in a terpenoid molecule, it significantly affects the appearance of the second cationic centre in the same molecule (if this does take place) and its further behaviour. Thus the opening of a three-membered ring in carane (59) occurs through proton attack at C(1) of the cyclopropane ring.26 HSO3F, SO2FCl *H, *H 7100 8C + + 59 61 In a-trans-3,4-epoxycarane (74), the initial opening of the epoxide ring (HSO3F±SO2FCl,7100 8C) is followed by b-cyclo- propylmethyl rearrangement and results ultimately in the forma- tion of the ketone 75.31 O OH OH + *H,*C7C + 74 + O OH *H 7H+ 75 IV.Factors affecting the pathway of cationic rearrangements The behaviour of stable p-conjugated carbocations is largely determined by charge delocalisation and the relative capacity to migration of substituents in 1,2-shifts.32 If ions are generated from terpenes or their analogues, the crucial role will be played by other factors, e.g., the nature of the acidic medium, the mode of generation of the cationic centre (solvolysis or electrophilic addition to the double bond) and the temperatures at which cations are generated and the acidic solutions are defrosted.The 1,2-shifts toward the cationic site are controlled by the competitive (rather than relative) migration capacity of the adjacent groups. 1. The nature of the acidic medium Elucidation of the relationship between the chemical behaviour of stable carbocations generated in superacidic media and trans- formations of these original substrates under conditions of acidic catalysis is an important problem. Its solution will allow the use of regularities established in the studies of stable carbocations in prediction of the most probable pathways of multistep carboca- tionic rearrangements under conditions of conventional acidic catalysis. The solution of this problem is also important for designing organic syntheses in general and syntheses of terpene compounds in particular.However, the publications devoted to this problem are scarce and the conclusions made therein areCationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions rather controversial. Thus Olah, who was a pioneer in the studies of the behaviour of cations generated in superacidic media, considers the chemistry of stable ions as an entirely independent field and conducts even applied investigations under extremely `hypernon-nucleophilic' conditions.33 In 1975, Sorensen et al.34 generalised the data on carbocationic rearrangements of several monoterpenes in media of different acidities. According to these authors, the increase in the DG= barrier to `ion-to-ion' rearrange- ments is of the same order both in superacids and under conditions of acidic catalysis, while the stability of alkenes parallels that of the corresponding cations.It was thus concluded that the path- ways of carbocationic rearrangements in acidic catalysis can be predicted from the results of direct observation of their rearrange- ments in superacids. No exceptions were reported.34 In our opinion, this can be due to the limited choice of experimental objects investigated in this study 34 (as a rule, compounds with a norbornane skeleton were used). We have found several substan- tial limitations of Sorensen's statements. Thus in a more recent study, Brown 35 has shown that the chemical behaviour of stable ions and the pathway of solvolysis are in rather poor agreement with each other.In 1983, Arnett and Hofelich 36 established a correlation between the rate constants of solvolysis of certain chlorides in ethanol and enthalpies of ionisation of these chlorides in superacids and thus confirmed the formation of intermediate short-living carbocations in solvolysis that are analogous to stable ions generated in superacids. However, further multistep conver- sions of the initially formed ions have not been investigated.36 We believe that the data obtained in the studies of molecular rearrangements in superacids can be used for predicting the most probable pathways of their transformations in ordinary acids with allowance made for possible changes in the nature of the medium, e.g., the site of generation of the cationic centre, the nature of the internal nucleophile, the `qualitative' changes in the ratio of rates of intra- or inter-molecular rearrangements, the non-correspond- ence of changes in the relative stabilities of alkenes and the cations derived from them, etc.This must be taken into consideration altogether in designing isoprenoid syntheses. We have studied the effect of the nature of medium on the pathway of terpenoid rearrangements. Thus in the case of trans, trans-farnesol (39), the site of formation of the cationic centre has been shown to depend on the nature of the medium.37 H + HSO3F CH2OH CH2OH 76 39 CH2OH + HCOOH OH 77 Protonation of the double bond takes place in superacids endowed with high protonating capacity and results in drimenol (76).Formic acid, characterised by high ionising capacity, favours solvolysis with the formation of an allylic ion. Its subsequent cyclisation gives bisabolol (77). The medium acidity affects the behaviour of labdanoids 70 and 72 (see above) in different ways. Cyclisation of compound 72 containing a tertiary allylic group in the side chain in both superacids and ordinary acids gives the tetracyclic product 73. Compound 70 containing a primary allylic group generates the tricyclic diol 71 in superacids and the tetracycle 73 in ordinary acids.30 399 A study of transformations of nerolidol (78) in various acidic media has shown37 that carbocylisation reactions involving differ- ent p-bonds can take place even in the presence of a good leaving group (e.g., a tertiary allylic hydroxy group) and in the case where a cationic centre at the same carbon atom is formed in different media. For example, conformational control favours the appro- aching of the chain termini in superacids at low temperatures to give the secondary ion 79. In weak acids and at room temperature, this reaction proceeds with the formation of a more stable tertiary bisabolyl ion 80.a + + 79 + H OH b 78 +OH2 80 (a) HSO3F, SO2FCl,7100 8C; (b) [H+], 20 8C. It has been shown 38 that cyclisation of chrysanthemyl alcohol (81) occurs differently in different media, since the medium determines the mode of reaction of a particular internal nucleo- phile with the cationic centre.This, in turn, significantly affects the ratio between carbo- and heterocyclisation reactions. Heterocyc- lisation in ordinary acids and carbocyclisation in superacids give cations 82 and 83, which have identical stability (DHof =114.9 and 114.6 kcal mol71, respectively, calculated by the MINDO/3 method). In other words, the reaction pathway is not determined by thermodynamic factors but is rather due to a lesser contribu- tion of heterocyclisation because of the protonation of the heteroatom in superacids. In the study by Baig et al.,39 the effect of the medium was neglected and the structure 84 was ascribed to the ion generated from the alcohol 81 in a superacid.As a matter of fact, this ion has the structure 8238 (Scheme 3, see Ref. 40). Scheme 3 + HSO3F CH2OH 82 + H+ H+ CH2OH CH2OH 81 +H O 83 HSO3F + O 84 The direction of a reaction can be affected not only by the nature of the medium but also by changes in the superacid:substrate ratio. Thus cyclisation of the alcohol 78 in the presence of a large (20 : 1) excess of a superacid results predominantly in carbocyclisation with the formation of the ion 79, whereas heterocyclisation resulting in the cation 85 occurs at the HSO3F : 78 ratio=5 : 1.37400 OH+ HSO3F, SO2FCl 78 7100 8C H 85 A similar effect of the excess of an acid on the carbo- to heterocyclisation ratio was observed with the diol 86.41 OH OH 86 + a OH b + (a) HSO3F : 86=2.5 : 1; (b) HSO3F : 86=5:1.In the above examples, the ratio of carbo- and heterocyclisa- tion reactions changed with variations in the excess of an acid as the result of translocation of the cationic centre and the internal nucleophile. The effect of the medium can also be manifested in a different way. By varying medium acidity and excess of acid with respect to the substrate, one can generate structurally different mono- or dications from the same substrate. Thus by conducting the cyclisation of the ketone 14 in acids of various strengths, one can generate allylic ions (pathway a) or dications (pathway b) at the same temperature.6 O OH + H+ + 14 OH + a HSO3F 7H+ OSO2F OH + *H, *Me b + (a) HSO3F, SO2FCl; (b) HSO3F: SbF5=1:1, SO2FCl,*H,*Me.Asignificant effect can be produced by varying only the excess of an acid relative to the substrate without changing medium acidity and reaction temperature. Thus by varying the excess of an acid in the cyclisation of the ketone 31, one can generate either the bicyclic carboxonium ion 37 or monocyclic dications with six- or five-membered rings.6 a + O HO+ 31 +OH + c 31 (a) HSO3F:SbF5=1: 1, SO2FCl; (b) HSO3F : 31=3: 1; (c) HSO3F : 31=8:1. Probably, the first example of a change in the ratio of the rates O for intramolecular rearrangements upon transition from a super- acid to ordinary acidic catalysis, which may alter both the stereo- isomeric composition and the structure of reaction products, was described in 1981.42 Thus the difference in the optical activities of isolongifolene (88) specimens ([a]20 D =08 and 783 8) obtained by isomerisation of longifolene (87) in HSO3F and AcOH can be attributed to the change in the ratio of the rates for intramolecular rearrangements upon transition from a superacid to a degenerate (pathway a, racemisation) or non-degenerate acetic acid (pathway b, retention of optical activity).On going to a more nucleophilic solvent (AcOH), the degree of solvation of ion 89 increases, while the `requirement' of the cationic centre formed at C(1) for the s- bond decreases. As a result, the orientation of the vacant p-orbital of the cationic site relative to the s-orbitals of the migrating groups and atoms changes, which affects the rate of their migra- tion (sp2-alignment factor). H H+ + 89 87 a MeOH + 7H+ b + OH + (a) HSO3F; (b) AcOH.OH + + In the studies by Sorensen et al.,34 it was shown that the rearrangements in mono- and polymethylnorbornyl cations formed as intermediate species independently of the medium gave 2-norbornyl ions; their further transformations occurred only by intramolecular pathways. In other cases, the contribution of intermolecular reactions can influence the comparability of rearrangement processes that occur in different media. The change in the ratio of rates of intermolecular rearrange- ments of an isocaryofyllene isomer, viz., the diene (90) (Scheme 4), has been established.43 In superacids (pathway a), the double bond from position 3,4 migrates to position 4,5 faster than from position 8,12 to position 8,9.An opposite ratio is observed in ordinary acids (pathway b). This finding cannot be explained by thermodynamic factors as can be seen from the calculation of enthalpies of formation of the final ions and alkenes (hereinafter DHof and DDHof values are expressed in kcal mol71). The pathways of neoclovene (91) conversions in superacids and ordinary acids differ drastically despite the similar character V A Barkhash,MP Polovinka H b +O 37 Et +OH + 88a 88bCationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions H a H H + + H H H 2 3 1 45 DHf =135.4 9 6 8 7 H H H 90 b H+ + H (a) HSO3F; (b) [H+].of stabilities of the final ions and alkenes.44 Calculations of the enthalpies of formation of ions and alkenes were carried out using the parameters described in Refs 45 and 46. According to NMR spectral data, pathway b is predominant in superacids and yields the ion 94, which is more stable than the ion 93 formed from the ion 92 by pathway a. It is still unclear, however, why in sulfuric acid this reaction gives the alkene 95, which is less stable than the alkene 96 generated by pathway b. a 7H+ + + 95 H+ DHf =719.7 93 DDHf =72.4 92 + 91 DHf =715.6 DDHf =0.0 b 7H+ 96 94 DDHf =713.0 DHf =729.8 (a) H2SO4; (b) HSO3F.In conversions of the intermediate ion 97 in various acidic media, the changes in relative stabilities of carbocations are opposite to those of the corresponding alkenes (NMR data and calculations). In ordinary acids, the direction of these rearrangements is determined by relative stabilities of intermediate alkenes. + a (*Me) HSO3F DHf =8.7 DDHf =711.6 + + b (*C7C) 97 H2SO4 DHf =78.2 DDHf =6.6 As can be seen from DHof and DDHof values, the Sorensen principle of parallelism of relative stabilities of carbocations and the corresponding alkenes 34 is violated in this case. It may be assumed that if several alternative pathways of rearrangements resulting in generation of ions with different skeletons (bridged, 401 Scheme 4 H H 7H+ HDHf =730.2 H H H H + 7H+ H H DHf =149.7 DHf =714.7 open, monocyclic or fused structures) do exist, the trends of changes in relative stabilities of ions and the corresponding alkenes form opposite series.Thus bridged structures can give rise to stable non-classical ions and the corresponding unstable `anti-Bredt' alkenes. Stable `Bredt' alkenes corresponding to unstable classical ions are generated from open, monocyclic and fused compounds. 2. The mode of generation of the cationic centre A difference in the stereochemical results of solvolysis of an acyloxy derivative and of protonation of the corresponding alkene in identical weakly acidic media has been established.47, 48 Accord- ing to Brown,47 this effect is due to the fact that more `loose' carbocations formed upon protonation are trapped by the nucle- ophile prior to complete establishment of an equilibrium between the ions.An alternative hypothesis 48 rests on the formation of isomeric non-classical (solvolysis) and classical ions (protona- tion). However, according to literature data, stable carbocations generated upon protonation of alkenes or upon solvolysis of esters (halides) normally yield identical species irrespective of the reaction type (e.g., see Ref. 49). HSO3F, SO2FCl + OTs In a more recent study,50 models were selected for the first time in such a way that isomeric stable ions could be generated upon alkene protonation and ester solvolysis.Their formation was confirmed by NMR spectroscopy. a F F F + + MeO MeO Me Me Me OMe a + F F F + OCOCF3 MeO MeO Me Me OMe Me (a) HSO3F, SO2FCl,7100 8C. On transition from these model compounds to terpenes, the problem of elucidating the dependence of the main pathway of402 cationic rearrangements on the mode of generation of the cationic centre becomes more complicated. First of all, this is due to polyfunctionality and conformational mobility of terpenes. In the following, we will compare the behaviours of an alkene in the protonation and its monoepoxide in the solvolysis. This is of interest for both physical organic chemistry (studies of solvolysis) and terpene chemistry.The opening of epoxides can initiate biomimetic processes. Two cases should be taken into consideration in comparing the behaviours of an alkene and the corresponding monoepoxide, viz., the coincidence or non-coincidence of the charge localisation sites. However, further rearrangements of the carbocation may occur according to different pathways even in the former case. For example, the conversions of a-humulene (98) and its 6,7-epoxide 99 51 with the identical localisation of the initially formed cationic centre occur by different routes, viz., by concerted, conformation- ally controlled pathway a (epoxide) or by barrier-free conversion of the alkene into the cation 101 (pathway b) and then into the ion 102. The configuration of all the chiral centres in the ion 100 formed by pathway a was confirmed by X-ray analysis of the `quenching' product. The original epoxide 99 seems to have the most stable CT-conformation.Analysis of the 13C NMR spectra of compound 99 at 7120 to 770 8C revealed the absence of conformational transitions at these low temperatures. H H O OH a [O] H + H+ 99 100 + + b H+ 98 102 101 Isocaryophyllene 43 and epimeric epoxides 10 and 11 3 derived from it generate different stable ions in superacidic media at all temperatures. HH H H HSO3F, SO2FCl + + 7100 8C H H 103 O + [O] HSO3F, SO2FCl 7100 8C +OH2 12, 13 10, 11 In isocaryophyllene, the energy barriers between individual conformers analogous to those of caryophyllene 1 are rather low, as a result of which it generates only one ion, 103.The transition to epimeric monoepoxides 10 and 11 is accompanied by stabilisation of individual conformations which further generate enantiomeric ions 12 and 13 (see Section II). V A Barkhash,MP Polovinka Caryophyllene (1) 2 gives rise to a mixture of carbocations corresponding to the original mixture of conformers (HSO3F ± - SO2FCl, 7120 8C). Since diene 1 conformers cannot be isolated in the individual state, the structure of the ion generated from one or another conformation cannot be established experimentally. Epimeric 4,5-monoepoxides 104 and 105, which may be regarded as `fixed' conformers, were used to generate one ion each (106 and 107) corresponding to those derived from the diene 1.3 O HO + HSO3F, SO2FCl 7120 8C 104 [O] 106 O 1 + HSO3F, SO2FCl 7120 8C 105 HO 107 Naturally, the mode of formation of the carbocationic centre is affected by other parameters of cationic rearrangements.Studies with the epoxide 105 have shown 3 that a change in temperature alters the position of the cationic centre. This, in turn, changes the composition of the reaction products formed from the monoepoxide 105 at different temperatures and, corre- spondingly, the pathways of rearrangements in the original diene 1 and its monoepoxide 105 occurring under identical conditions. Obviously, it is the ratio of epoxide 105 conformers that is changed with a change in temperature. As the barriers to the conformational transitions 105a>105b and to the reactions of the primary ions are comparable, a change in temperature results in the change in the rearrangement pathways.OH H+ HSO3F, SO2FCl O + 7130 8C 105a HSO3F, SO2FCl O + 780 8C OH H+ 105b 107 The epoxide transformations described above are concerted processes. However, they can be realised only for such relative arrangement of the epoxide ring that is opened and the part of the molecule where hydride and alkyl shifts toward the cationic site formed occur without difficulty. In other cases, non-concerted epoxide conversions may take place. The latter case has been described 52 where stable ions 110 and 111, differing in config- uration of the carbon atom containing a hydroxy group, were generated in superacids from neoclovene epoxides 108 and 109.The stereochemical result of this reaction suggests that conver- sions of epoxides 108 and 109 into the corresponding alcohols 110 and 111 occur in a non-concerted way.Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions O O [O] + 109 91 108 HSO3F, SO2FCl, 7100 8C HSO3F, SO2FCl, 7100 8C + + + + H2O H2O H H 111 110 The non-coincidence of pathways of rearrangements of an alkene and the derived epoxide may reflect the differences in the behaviour of mono- and dications with identical skeletons. For example, different stable ions were obtained from neoclovene (91) and epimeric epoxides 108 and 109 with identical backbones and identical sites of formation of the cationic centres under the action of a superacid. The epoxide opening generates a dication with charges on the carbon atom and the oxygen atom of the hydroxy group.Spatial separation of the charges is the driving force for the rearrangement of such a dication. It thus follows that a rearrange- ment of a cyclic alkene and the derived epoxide will yield the same ion only in the case where the rearrangement of an ion generated from the alkene is accompanied by translocation of the positive charge away from the site of proton addition. H + HSO3F, SO2FCl + 91 7100 8C H +OH2 + HSO3F, SO2FCl + 108, 109 + 7100 8C H2O H 110, 111 The mode of generation of the cationic centre is also reflected in the structure of carbodications generated from dienes and the corresponding hydrogen chloride addition products.Stable iso- meric carbodications 66 and 113 were generated from a-murolene 48 and the dichloride 112 with identical backbones and identical position of the cationic centres formed under identical conditions (SbF5 ±HSO3F±SO2FCl,7120 8C).53 Cl H H 2HCl Cl H H 48 112 H 1 H H 9 + 2 8 10 5 + 7 + 3 4 6 H H 11 + 113 66 As has been noted earlier, isomeric stable ions were obtained upon changes in the mode of their generation.50 The difference in the reaction routes was ascribed 50 to the formation of non- classical ions upon ester solvolysis and of classical ions upon 403 alkene protonation.However, non-classical ions are formed mainly in reactions of bridged polycyclic systems.54 If the con- version of the dichloride 112 into the dication 113 occurs through a scheme involving a 1,5-hydride shift from C(11) to C(1), the quenching product should possess the structure 114a. If this process involves only 1,2-hydride shifts, the diene formed should have the structure 114b. It has been established 53 that ionisation of the dichloride 112 gives the diene 114b and, correspondingly, translocation of the positive charge from C(1) to C(11) occurs as consecutive 1,2-hydride shifts. H H H H 114b 114a In the dichloride 112, elimination of the chlorine anion from C(7) is followed by translocation of the positive charge to C(11) through 1,2-hydride shifts and only after this process has been completed, the second chlorine anion is split off from C(1) giving rise to the dication 113.Protonation of the diene 48 results in immediate generation of the dication 66 in which translocation of the positive charge through 1,2-hydride shifts from C(1) to C(11) does not occur, since the barriers to transition from a classical tertiary carbocation to a classical secondary isomeric carbocation usually exceed 14 kcal mol71 at 7120 8C (see Ref. 55). Further translocation of the positive charge from C(1) to C(10) is hamp- ered due to the presence of a cationic centre at C(7). Thus, depending on the mode of cation generation, the differ- ence in reaction routes can be caused by: (i) intermediate for- mation of classical and non-classical isomeric ions; (ii) concerted epoxide rearrangement and non-concerted rearrangement upon alkene protonation; (iii) change in the site of formation of the cationic centre; (iv) non-coincidence of pathways of rearrange- ment of mono- and dications having identical backbones and (v) peculiarities of the spatial structure of the original compounds including the ratios between the conformation barriers and barriers to reactions of the primary ions.The effect of spatial structure of original compounds, i.e., configuration and the nature of the leaving group, on the direction of a cationic rearrangement can be well illustrated by the following reactions:56 + HSO3F, SO2FCl 7100 8C OH OH OH + HSO3F, SO2FCl 770 8C COOCF3 COOCF3 F3COOC HSO3F, SO2FCl 760 8C OH OH 3.Cation generation temperature The temperature at which carbocations are generated is a crucial factor which determines the pathways of multistep cationic rearrangements. It is assumed that ions generated at low temper- atures are not converted upon heating into cations that are formed at higher temperatures. In some cases, a change in cation generation temperature influences the position of the cationic404 centre. Thus the diene 114b predominantly generates the ion 115 at780 8C and the ion 67 at750 8C in HSO3F±SO2FCl:53 H + + 780 8C H H 115 H H H + 114b 750 8C H + 67 A change in the conformational composition of the original diene 114b does not account for the temperature-dependent generation of allylic ions, since the 13C NMR spectra of this diene in CD2Cl2 are not changed in the temperature range from 785 to +10 8C.Rather, a change in temperature may affect the direction (the exo- or the endocyclic double bond) of primary protonation. A temperature-dependent change in the ratio of protonation rates of two endocyclic double bonds has been described.23 Thus ions 66 and 49 are generated upon protonation of the diene 48 at 7120 8C in HSO3F±SO2FCl, whereas heating to 750 8C yields ions 67 and 49 in a *2.5 : 1 ratio. If the ions are generated at 770 8C, the ion 49 is predominantly formed (see Section II). Thus, the ratio of diastereomeric ions can be altered by varying the temperature at which these ions are generated.The effect of temperature on the ratio of bi- and tricyclic ions formed has been observed.57 Thus a mixture of bicyclic (49) and tricyclic (116) ions is formed from the diene 48 (HSO3F±SO2FCl, 730 8C) in a*2 : 1 ratio (data from 1H NMR spectroscopy). H H H 48 HSO3F, SO2FCl + + + H H 116 49 H 117 H H + HSO3F, SO2FCl *H 48 49 7120 8C 65a Since the temperature at which ions are generated does not influence the conformational composition of the original diene 48 (the 13C NMR spectra are not changed at temperatures from 790 8C to +10 8C), it may be supposed that its protonation at 7120 to 770 8C will yield the cation 65a in the most stable half- chair ± chair conformation (according to calculations carried out by the MM2 method).1,2-Hydride shifts in the ion 65a result in the allylic ion 49. In the conformation 65c, the reaction centres are at the shortest distance favourable for transannular cyclisation. The conformation of the ion 65a is more favourable for trans- annular cyclisation than that of the ion 65b as can be judged from V A Barkhash,MP Polovinka the formation of the ion 116 (in addition to the ion 49) from g-murolene (117) (HSO3F±SO2FCl, 730 8C). At 7120 8C, the protonation of the diene 117 in HSO3F±SO2FCl gives the ion 49;23 at lower temperatures, it is the monocation 65a that is exclusively formed due to the greater ease of protonation of the exocyclic double bond.g-Cadinene (118), the isomer with a trans- fusion of the rings, does not generate either the ion 116 or any other tricyclic cations in HSO3F±SO2FCl at730 8C. HH 118 Obviously, the DH= value calculated by theMM2method for the conformational transition 65a?65c (10.9 kcal mol71) significantly exceeds the barriers to 1,2-hydride shifts in the ion 65a in the conversion of one tertiary ion into another (<4 kcal - mol71).58 This explains the effect of the conformational compo- sition of the ion 65 on the ratio of ions 49 and 116 at 7120 to 770 8C. The rise in ion generation temperature (but not in defrosting temperature) favours the conformational transitions 65a?65c and increases the contribution of transannular inter- action of the carbocationic centre with the double bond, resulting in the initial generation of ions 119 and 120.Taking into account the difference in DHof values for these ions (22.9 kcal mol71) and the short C(3)7C(9) and C(4)7C(9) distances in the conforma- tion 65c, it may be supposed that the cyclisation of the ion 65c gives a more stable ion 119. The latter is further rearranged into the ion 116 according to the schemes 57 suggested on the basis of estimation of free energies of activation.59 1 +9 >730 8C 65a 116 2 + 4 3 65c 119 According to 1H NMR spectral data, dissolution of a-copaene (120) in HSO3F±SO2FCl at7115 8C and subsequent heating to 730 8C yield a mixture of ions 49 and 116 (*3 : 1). Hydrocarbon 120 is a direct precursor of the ion 119; which rearranges further into the ion 121.The observed rearrangement corroborates the pathway proposed for generation of the ion 116. HSO3F, SO2FCl 119 49+116 + 121 120 An interesting example of the effect of ion generation temper- ature on the pathway of a cationic rearrangement has been described.56H + HSO3F,*H 2 9 7H2O OH OH X 123 122Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions + *Me X HX+ X is fluorosulfonate residue. The hydroxy group at the C(9) atom in the diol 122 is the first to be eliminated under the action of HSO3F. Subsequent 1,2- and 1,3-hydride shifts in the cation formed give the ion 123. The latter is subject either to a 1,2-shift of the methyl group with subsequent elimination of an HX species (its nature is unknown) to give the ion 124 or to elimination of a proton and the HX species 4 to give ultimately the ion 4.The ratio of the pathways of ion 123 conversions is determined by the temperature of its generation from the diol 122. Thus, the ion 124 is predominantly formed at 7100 8C, while the ion 4 is generated at 20 8C. 4. Defrosting temperature Yet another factor which controls cationic rearrangements is the temperature at which an acidic solution is defrosted . First, the reaction mixture obtained following treatment with superacids (HSO3F, HSO3F ± SbF5±SO2FCl in a solvent or with- out it) can be heated from7130 8Cto +100 8C.This temperature range allows one to overcome the barriers to ion-to-ion transitions of up to 30 kcal mol71. Second, these `hypernon-nucleophilic' media favour profound molecular rearrangements which have no analogues in acid-catalysed reactions. Thus cyclisation of dehy- drolinalool (125) in `ordinary' acidic media gives a mixture of compounds 126 and 127.60 H+ OH 125 In superacids, the rearrangements resulting in the formation of stable ions 128 ± 133 are much more diverse.61 +OH 128H + OSO2F 131 According to literature data,62 acid-catalysed cyclisation of isolone (134) gives exclusively florione (135). Ketone 134 gener- ates the ion 136 in HSO3F±SO2FCl at 7110 8C. Heating of the + 7HX 124 + 7HX *H,*Me +4 O + COMe 127 126 H + + OH 129 130 OSO2F + + O O 132 133 405 resulting acidic solution to 710 8C yields a mixture of carboxo- nium ions 137 and 138.63 COMe COMe a 134 135 + O+ H O 710 8C b + 134 + H O 138 137 136 (a) [H+]; (b) HSO3F, SO2FCl,7110 8C.The ketone 134 generates the dication 139 (HSO3F± SbF5 ± - SO2FCl,790 8C); heating to760 8C results in the dication 140. + + OH OH HSO3F, SbF5 760 8C + 134 + SO2FCl 790 8C 140 139 Longifolene 87 is predominantly converted into isolongifolene 88b in acidic media.64 Four carbocations (89 and 141 ± 143) were generated in HSO3F±SO2FCl by varying the defrosting temper- ature;42 these carbocations were successively interconverted with the rise in temperature.a 88b b 780 8C 740 8C + + 87 141 89720 8C + + 143 142 (a) [H+]; (b) HSO3F, SO2FCl,7120 8C. The effect of defrosting temperature on the structure of carbocations formed has been observed.56 It has been shown that ditosylate 144 generates the ion 145 in HSO3F± SbF5 ± SO2FCl at 790 8C. Heating of the mixture to 750 8C leads to its rearrangement into the ion 146, which is further isomerised into the ion 4 with the rise in temperature to720 8C. + 750 8C HSO3F, SbF5 SO2FCl,790 8C OTs OTs 145 144 + + 720 8C 146 4406 V. Selection rules Irrespective of the complex structure of the original substrates, high selectivity of rearrangements of stable ions generated from terpenes is a very important factor which determines their behaviour in superacidic media.This selectivity is provided by a combination of regio- and stereocontrol. Thus studies of diene 48 conversions in HSO3F ± SbF5 ± - SO2FCl have shown 23 that at temperatures from 7120 to 750 8C all ionic isomerisations occur exclusively as 1,2-hydride shifts and deprotonation-protonation, but not as C7C shifts. Although the theoretically possible protonation ± deprotonation reactions and 1,2-hydride shifts in the diene 48 can lead to an extremely complex mixture, this gives rise to a limited number of stable ions that are generated by several relatively short pathways. The selection rules realised in this case can be formulated as follows: 1. Protonation of double bonds and further rearrangements occur predominantly through the formation of both intermediate and final tertiary carbocations; secondary ions are not normally formed.2. In superacidic media, 1,2-hydride shifts occur more easily than deprotonation reactions, if tertiary and allylic ions are formed. 3. Allylic ions in which all three carbon atoms are incorpo- rated in a single ring and which do not involve two carbon atoms of the bond common to both rings are the most stable. 4. Protonation and deprotonation occur regio- and stereo- selectively. This may be due to both inherent steric hindrances for approaching a protonating reagent or a base (the conformational transitions are strongly hampered at 7120 to 750 8C) and stereoelectronic factors in the deprotonation.These peculiarities distinguish the compounds under study from more simple, mono- cyclic or aliphatic models, e.g., those in which protonation is non- selective as a rule. The conversions of the dication 66 into the allylic ion 67 described above may occur through stereoselective addition of protons to the double bonds of intermediate dienes. An alternative scheme implies that under kinetically controlled conditions the protonation occurs non-selectively, whereas under thermo- dynamically controlled conditions the equilibrium between dia- stereomeric ions is practically completely shifted toward the ion 67. This suggestion seems unlikely, since diastereomeric ions 49 ± 51 generated separately are stable and are not interconverted under these conditions.23 H + H+ 780 8C HSO3F, SO2FCl 48 + 7120 8C H H 66 H + + 1.7H+ 2.H+ 750 8C H 67 VI. Computational methods for predicting the most probable pathways of multistep cationic rearrangements The `selection rules' are valid in small groups of related com- pounds 23 and allow one to predict qualitatively the most probable pathways of multistep carbocationic rearrangements. Several approaches to quantitative estimations have been proposed.16 An original computer programme `ICARUS' was used to inves- tigate plausible mechanisms of multistep carbocationic rearrange- V A Barkhash,MP Polovinka ments.65 This programme entails generation of the whole variety of multistep rearrangements of a definite chemical structure and control over realisation of the generated routes.All the cationic rearrangements can be divided into six main groups: (1) 1,2-shift of the s-bond + + (2) addition at the multiple bond +CH2 + (3) b-fragmentation +CH2 +(4) allylic rearrangement +CH2 + (5) b-cyclopropyl ± carbinyl rearrangement + + (6) a-cyclopropyl ± carbinyl rearrangement + + Validation of the pathways proposed by the programme is carried out using an empiric data base on the peculiarities of carbocationic rearrangements. In particular, the ICARUS data base involves the following selection rules: 1. Only the rearrangements occurring as 1,2-shifts of the C7C and C7H bonds fall under investigation; more distant 1,3-, 1-4- shifts, etc., are not analysed.2. Only intramolecular rearrangements (i.e., those not result- ing in degradation of the original structure into two independent species, e.g., proton elimination followed by generation of an alkene) are subject to analysis. 3. Reactions resulting in generation of unstable intermediates (primary cations, structures containing strained small rings, etc.) are excluded from consideration. Also excluded are b-fragment- ation reactions, since they produce high-energy intermediates: the formation of a double bond significantly increases the enthalpy of the reaction. 4. 1,2-Shifts of the s-bonds in allylic ions are excluded from consideration. However, estimation of the probability of such rearrange- ments on the basis of qualitative speculations alone does not always permit one to rule out improbable mechanisms.Quantita- tive estimations require calculation of thermodynamic character- istics of the intermediates by molecular mechanical or quantum mechanical methods. The enthalpy of a reaction can be calculated using a set of special force field parameters of the MM2 pro- gramme and bond increments for carbocations.66 Estimation of the energy barriers demands simultaneous account of thermo-Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions dynamic and orbital factors, since their separate consideration often gives contradictory predictions. The enthalpies of formation of primary and final ions can be used as a thermodynamic factor.The orbital factor is associated with the value of the dihedral angle between the vacant p-orbital and the neighbouring migrating bond. It was assumed 59 that in the first approximation the internal barrier to the 1,2-shift (L) of a degenerate rearrangement is connected with the interorbital angle (j) by equation (1): (1) L=L0/cos j, where L0 is the internal barrier at j=08. For a non-degenerate process, the thermodynamic factor was calculated by the Marcus equation (2) (in isoentropic approxi- mation): (2) DG== 1 á DDH 2 4 or in its linear Polanyi ± Semenov approximation (3): (3) DG==L+bDDH, where DDH is the difference between the DHof values of the generated and rearranged ions; b=0.25 (for 74L<DDH<0), 0.75 (for 0<DDH<4L), DG==0 (for DDH<74L), DDH(for DDH>4L).The internal barriers, L0, for the orbitally optimum situation of a 1,2-shift were estimated according to Sorensen 24 from the values of barriers to 1,2-shifts of the C7C bond and the hydrogen atom in polymethylnorbornyl ions. Degenerate 1,2-shifts can be estimated from equation (4): L0=DG=S cosjS, (4) where DG6àS is the barrier to the shift according to Sorensen and jS is the orbital factor in a norbornyl cation. Since according to Sorensen's data the difference in the enthalpies of formation of secondary and tertiary isomeric nor- bornyl cations (in DDH) is equal to 5.5 kcal mol71, non-degener- ate processes can be estimated from equation (5): (5) L0=(DG=S +5.5 b) cosjS.The above equations allow one to predict the most probable pathways for the rearrangements of 1-chlorocaryolane,59 clovene (2),2 a-copaene (120) 57 and neoclovene (91) 44 in superacidic media. Thus calculations have shown 44 that the ion 92 generated initially from neoclovene (91) is further converted into a mixture of two carbocations 94 and 147 with an energy gain, DDHof ,713.0 and710.7 kcal mol71, respectively. This result was confirmed in experiments on protonation of the alkene 91 in HSO3F±SO2FCl at7120 8C.44 + + HSO3F, SO2FCl 91 94 + 7120 8C 92 147 Thus, the relatively simple technique for estimating the barriers to 1,2-shifts for the migrating ion 59 can efficiently be used for predicting rearrangements in alicyclic carbocations.However, this approach has some drawbacks. For example, the angle j is not an unambiguous characteristic of the transition state strain. The internal barriers, L0, were estimated 59 on the basis of kinetic data for rearrangements of s-delocalised 2-nor- bornyl ions. The choice of such specific models has led to overestimated barriers for open and monocyclic ions. The correlation ratios (6) and (7) for estimating DG6à for 1,2- shifts of the hydrogen atom and the methyl group in aliphatic and alicyclic carbocations which actually determine the barrier to a degenerate rearrangement have been proposed.67 407 DG6àH=15.33710.79qpx+0.232 logKi, (6) r=0.986, s=0.9, n=8, DG6àCH (7) f and 3 =25.74721.48qpx+0.303 logKi, r=0.995, s=0.8, n=7, where qpx is the value of the HuÈ ckel charge on the atom toward which the migrating group is displaced and Ki is the basicity constant of the ith position formed upon migrant detachment. In this approach, the alkene formed upon elimination of the migrating group from the original cation serves as a model of the transition state of a 1,2-shift. The non-degeneracy of this process can be estimated by the Marcus equation (2).An advantage of this approach is the possibility of indirect estimation of the orbital factor and `olefin strain',68 since both these factors are included in the proton affinity value. However, the correlation ratios (6) and (7) are inapplicable to the Wagner ± Meerwein rearrangement; in this case, DG6à can be determined only as described in Ref.59. The use of the above approaches 59, 67 for estimating DHo DG6à has made it possible to propose the most probable pathway for a-cedrene (148) rearrangement in superacids resulting in the ion 149.69 The latter could be obtained experimentally (the figures under the structures refer to the DDHof values; the figures above the arrows indicate the values of the barrier to 1,2-shift in kcal mol71). 13 H+ + 1 3 + + 148 0.0 12.5 2.2 + 4 4 + 9 11 + + 71.9 71.4 8.4 72.4 + + 2 5 3 + 6.6 14.2 16.8 18 5 3 + + + 0.5 14.7 16.5 + + 1 + 2 7 + 4.1 72.3 2.3 72.5 149 The primary site at which the cationic centre is formed in the alkene 148 having a trisubstituted double bond leaves no doubt.The primary proton attack in ledene (150) may involve both the double bond and the cyclopropane ring. Estimation of the proton affinity of the double bond and the three-carbon ring by the molecular mechanics method (MMX programme) revealed that the protonation of the double bond is much more preferable (by 3.5 kcal mol71).69 As for the regio- and stereoselectivity of proton addition to the double bond, it has been found 69 that the protonation of the double bond at the carbon atom containing a408 methyl group is more preferable for both a- and b-attack (by 9.2 and 8.3 kcal mol71, respectively) (MMX). H H + + H H 151 H H 150 + + H 152 The isomeric ions 151 and 152 formed (DHof = 157.4 kcal - mol71) have identical stability.Therefore, the stereodirection of their protonation was established with the use of additional data, viz., analysis of the Dreiding models of the alkene 150, estimation of stability of epimeric epoxides prepared therefrom and estima- tion of steric accessibility of the double bond of the alkene 152 by the method described in Ref. 70. These data altogether suggest that the proton attack will occur largely from the a-side to yield the ion 152.69 The use of these approaches permits one to predict the rearrangement into the ion 153 and then into the ion 154 as the most probable pathway for alkene 150 transformations in `long- living' ions; this conclusion was confirmed by the results of the reaction carried out in HSO3F±SO2FCl at 7110 8C.The calcu- lations predicted a gain in energy for the conversion of the ion 152 into the ion 154 (10.7 kcal mol71) (MMX).+ H+ *H *(C7C) 150 152 153 H H + 7H+ 154 It should be noted that neglect of calculation methods in predicting the most probable pathways of multistep carbocation rearrangements may lead to serious misinterpretations of exper- imental data. Thus analysis of 13C NMR data has led to a conclusion that dissolution of the epoxide 155 in HSO3F±SO2 at 770 8C yields the ion 158 according to the following scheme:71 H+ + + OH OH156 157 O155 + HO CHOH + + O 158 The CH2 group in the ion 157 undergoes a shift, although it is migration of theMe2Cgroup that should occur in this type of ions.To overcome this controversy, an alternative scheme has been proposed where the ion 156 is not converted into the ion 157 but V A Barkhash,MP Polovinka undergoes an unusual for epoxides cleavage of the C7C bond in a three-membered ring.71 This is followed by a hydride shift to the tertiary carbon atom with generation of the ion 159 and then into the ion 158. H + O+ 158 155 156 CHO 159 Later it was shown 72 that this scheme of epoxide ring cleavage contradicts both the experimental and calculated data. According to the calculations performed by the AM1 method, the ion 156 is converted into the ion 157 (but not into the ion 159) without a barrier. Further rearrangements of the ion 157 occur predomi- nantly by pathway a resulting in the ion 160 (DHof =145.1 kcal - mol71) but not by pathway b (the ion 161, DHof =150.3 kcal mol71), which is attributed to the greater stability of the ion 160.a + HO 160 + b OH OH 157 + 161 (a) DG==5.9 kcal mol71; (b) DG==6.5 kcal mol71. In addition, it has been shown 72 that dissolution of the epoxide 155 in HSO3F±SO2FCl at 7110 8C and subsequent `quenching' with methanol yield acetal 162 as the main reaction product. The carboxonium ion 163 generated from the ion 158 by a standard route with a 1,2-shift of the Me2C group is its immediate precursor. + + HOCH HOCH H+, *Me p C+ 160 + + 7H+ OH 157 O O MeOH + H OMe H163 H162 We can cite another example of an error resulting from disregard of calculation methods.As has been noted above, the structure 84 was ascribed to the ion generated from the alcohol 81.39 However, quantum-chemical calculations by the MINDO/3 method (AMPAC programme) revealed that the cation 84 was converted into the ion 164 without a barrier. This suggests that the ion 84 could not be regarded as a stable species. *H + + CH2OH O 164 81 O 84Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions VII. A system of `blocks' for aliphatic models The synthesis of complex organic compounds from renewable plant material, particularly from widely accessible aliphatic monoterpenes by cationoid cyclisation acquires increasing impor- tance (see Section VIII).Therefore, prediction of the most probable pathways of these processes becomes especially urgent. Unfortunately, pathways of structure-dependent cyclisation of acyclic compounds cannot be predicted either by quantum chem- istry or molecular mechanics methods. By virtue of its `non-rigid' geometry, the original compound can exist in different, relatively stable and energetically close conformations. Therefore, its cycli- sation may occur by several different pathways; their choice is hampered by imperfection of the methods used. An alternative approach to the experimental proof of the cyclisation schemes consists in fixation of intermediate cations in the rearrangement of aliphatic monoterpenes and establishment of a correlation between their generation and the presence of definite structural `blocks' in the molecule of the original com- pound.73 Systematic studies of the behaviour of terpenoids with identi- cal carbon skeletons [e.g., citral (68), ionone (14) and methyl geranate (41)] in superacidic media have made it possible to propose three main pathways of primary cyclisation,73, 74 viz., proton attack at the terminal double bond resulting in generation of a tertiary carbocation and its subsequent addition to the carbon (pathway a) or oxygen atom (pathway b); addition of a proton to the oxygen atom and subsequent attack of the cationic centre at the terminal double bond (pathway c, Scheme 5).Scheme 5 O O + a R R + O H+ R R O O + + b R + OH O + OH c H H However, the cyclisation of citral (68) and citronellal (69), i.e., two aldehydes with identical skeletons, occurs by totally different pathways.29 H+ O O H H H+ 68 (pathway b) 69 (pathway c) In some cases, compounds under study contain a common structural block which determines the pathway of cyclisation.This structural block is not necessarily identical with the carbon skeleton commonly used for classification of terpene compounds. For example, cyclisation in HSO3F±SO2FCl of j-damascone (165) and methyl geranate 41 having different skeletons and belonging to different classes of organic compounds occurs by a similar route (pathway a, Scheme 5).21 409 OH O FO2SO 790 8C 7100 8C + 165 OH+ 740 8C +O O OH FO2SO 7100 8C OMe 780 8C OMe + 41 OH+ MeO 740 8C OMe +O Studies with trans- (166) and cis-4,8-dimethylnona-3,7-dien-2- ones (168) have shown that reactions of isomeric acyclic isopre- noids with a conjugated 3,4-double bond can occur in HSO3F ± - SO2FCl in a stereospecific way (pathway b, Scheme 5) both in the stage of generation of intermediate ions and the formation of `quenching' products from acid solutions.Hence, the structural direction and stereochemical result of cyclisation of a,b-unsatu- rated carbonyl compounds of the terpene series depend on the configuration of the 3,4-double bond. O O O MeO + b MeOH 7110 8C 167 166 O O OMe O b MeOH + 7110 8C 168 At first glance, ketones 166 and 168 share a common struc- tural block with compounds 41 and 165 but the pathways of their primary cyclisation are totally different.Comparison of these models allows one to conclude that carbocyclisation requires that the structural fragment, Me2C=CH(CH2)2C(Me)=CHCO, was linked with a substituentXable to stabilise the positive charge by a mesomeric effect (OMe, CH=CHMe); otherwise heterocyc- lisation will take place. Substitution of the hydrogen atom for the methyl group at C(4) in the ketone 166 (? 169) changes substantially the path- ways of cyclisation of compound 166 (Scheme 5, pathway b) and compound 169 (Scheme 5, pathway c) due to secondary conver- sions of the homologous carbocations 167 and 170.O OH + O H+ + 750 8C 7100 8C 167 166 H+ O O O + 7100 8C 750 8C 169 170410 + O Varying the conditions of ion generation, particularly the acidity of the medium and temperature, makes it possible to carry out directed multistep rearrangements, e.g., carbo- or heterocyc- lisation of ketones 166 and 168.73 In this case, heterocyclisation strongly depends on the configuration of the conjugated 3,4- double bond. Cyclisations of the `ionone' type are independent of the spatial structure of the original substrate. 167 OH + + 2H+ O 166 + O OH + + O It is of note that compound 171 is the main product formed from the ketone 166 in dilute sulfuric acid.73 In this case, the main pathways of cyclisation in superacids and `ordinary' acids are different, since in the latter case intramolecular rearrangements prevail, which is not the case with superacids.O H+ 166 7H+ O 171 It was shown 21, 73, 74 that many acyclic terpenoids (com- pounds 14, 15, 41, 165, 166, 168) undergo similar conversions in SbF5 ±HSO3F±SO2FCl according to the following scheme: +RH +RH R 790 8C *H *H + + 172 +RH +RH +RH + *H *Me *H + + R+ +RH + 174 173 R=COCH=CHMe, COMe, CH=CHCOMe, COOMe.Electrostatic repulsion of positive charges in the primary dication 172 is the reason for its rearrangement into the dication 173 (from ketones 14 and 15) or a bridged ion of the type 174 (from the ester 41 and ketones 166 and 168). Erroneous structures were ascribed to the conversion products of geraniol (175) and nerol (176) in superacids.75, 76 We assumed 77 that rearrangements of aliphatic alcohols 175 and 176 proceeded in accordance with the above scheme and gave the ion 177 and the ester 178, since they have a common structural block with compounds 165, 166 and 168.This hypothesis was confirmed experimentally. CH2OH 175 SbF5, HSO3F SO2FCl,7100 8C HO 176The type of rearrangements of alcohols 175 and 176 depends on reaction conditions. Thus the cyclic oxonium ion 179 was obtained in HSO3F±SO2FCl at 7110 8C, while the tetracyclic ion 182 was formed from intermediates 180 and 181 upon heating of the reaction mixture to790 8C. Deprotonation of the ion 182 gives the ether 183.Such a conversion of a monoterpene into a diterpene has no analogues either among cyclisation reactions of acyclic isoprenoids occurring in superacids or among mono- terpene reactions realised in ordinary acidic media.78 CH2OH HSO3F, SO2FCl 7110 8C 175, 176 +CH2 X 180 OH + 182 CH2OH. X=The outcome of rearrangements of stereoisomeric alcohols 175 and 176 is independent of the configuration of the 2,3-double bond,77 while reactions of geranyl (184) and neryl acetates (185) gave different products. In HSO3F±SO2FCl (7120 8C), these compounds are predominantly converted into acyloxonium ions 186 and 187, respectively. `Quenching' of acidic solutions of ions 186 and 187 results in diastereomeric hydroxy acetates 188 and 189.H CH2OAc H+ 184 186 H CH2OAc H+ 185 187 V A Barkhash,MP Polovinka O O H + 177 178 + CH2OH2 790 8C 179 +CH2OH 181 O 7H+ 183 CH2OAc MeOH O + OH O 188 CH2OAc MeOH O + OH O 189Cationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions VIII. Synthetic aspects of terpene reactions in superacidic media One of the goals of studies of behaviour of terpenoids in super- acids is to extend the range of new synthetic substances that can be prepared from accessible natural compounds and their analogues. Structural features of these compounds and the use of superlow temperatures and hypernon-nucleophilic media in the reactions ensure great diversity of molecular rearrangements.In contrast with acid-catalysed processes, the use of superacids allows for: (a) multistep conversions due to hypernon-nucleophilicity of the medium; (b) the possibility to obtain dications, which has no analogues in acid-catalysed reactions; (c) rapid conversion of original substrates into carbocations, which prevents cationic polymerisation due to high acidity; (d) high structural and stereo- chemical selectivity owing to the use of superlow temperature. The preparation of 19 stable carbocations (which represent optical isomers) from an optically active diene, viz., (7)-a- murolene (48), serves as an illustrative example of synthesis of different compounds from terpenes in superacids.53 The direction of reaction was changed by varying the mode of ion generation (protonation of the diene or solvolysis of the corresponding dichloride), the acidity of the reaction system and the temperature at which ions were generated or `defrosted'.Reactions of these ions with nucleophiles gave rise to different dienes, viz., natural terpenes or previously unknown isomers. H H + + H + + H H H + Cl H H + + H + + + + The conversions of compounds 14, 15, 31, 32 and 125 considered in this review demonstrate the possibility of synthesis of various compounds through intermediate generation of di- and monocations having identical skeletons. Conducting reactions in superacidic media provides greater possibilities for varying the structure of reaction products in comparison with ordinary acidic media.72, 79 Thus treatment of the epoxide 74 with proton acids usually gives products with retention of a carane skeleton and few compounds with a bicyclo[3.1.0]hexane skeleton.80 H H + + H H + + + H H + H + H Cl + H H + + + In HSO3F±SO2FCl at 411 7110 8C, this epoxide generates a mixture of compounds 190 ± 195.O OMe O O 190 192 191 HO O MeO OMe OMe O 195 194 193 Compounds 190 ± 195 have not been previously obtained upon acid-catalysed cleavage of the epoxide 74. Their formation can be ascribed to the initial opening of the epoxide ring, subsequent rearrangements with cleavage of the cyclopropane ring and stabilisation of cations by virtue of reactions with internal or external nucleophiles or elimination of a proton.The behaviour of the epoxide 155 in HSO3F±SO2FCl (7110 8C) has been studied.72 The three main reaction products were acetal 162 and esters 196 and 197. OMe OMe O a, b O + + O H 162 155 OMe 196 197 (a) HSO3F, SO2FCl,7100 8C; (b) MeOH. These compounds have also not been isolated earlier upon acid-catalysed cleavage of the epoxide 155. 6-Methyl-3-(3-methylbut-2-enyl)hept-5-en-2-one 198 does not yield monomeric products under acid-catalysed conditions but is resinified. However, it gives a mixture of compounds 199 and 200 (yield 80%) upon treatment with HSO3F±SO2FCl (7110 8C) with subsequent `quenching'.29 OH O O O a, b +MeO MeO 199 198 200 (a) HSO3F, SO2FCl,7110 8C; (b) MeOH.4 As has been noted above, the use of liquid superacidic media allows one to obtain a more complete set of substrate transforma- tion products. If the problem is to prepare a specific compound, it is necessary to choose an appropriate solid catalyst, e.g., a solid superacid. The role of such catalysts may be played by sulfates of transition metal oxides, which, according to literature data, have not been used previously as acid catalysts in terpene reactions. It has been shown 79 that the use of TiO2/SO27 as an acid catalyst in isomerisation of the epoxide 74 allows one to obtain a mixture of reaction products in which the aldehyde 201 and the ketone 202 predominate, which is different from the results obtained with liquid superacids.Earlier, compounds 201 and 202 were consid- ered to be intermediate products in the preparation of the acetal 190 and the ketone 195 in liquid superacids. CHO O O TiO2/SO2¡ 4 + 74 201 202412 The pathways of conversions of conformationally flexible substrates in both liquid and solid superacids can be different. Thus the isomerisation of the epoxide 99 in a liquid superacid (HSO3F±SO2FCl) differs from that on a solid catalyst (Al2O3/ SO27 4 ) and alcohols 203 and 204 are formed, respectively.81 a O OH + b 99 (a) HSO3F,780 8C; (b) Al2O3/SO2¡ 4 , 20 8C. This phenomenon can be explained by the difference in conformational control due to low temperature and epoxide binding on the catalyst surface.IX. Biogenetic aspects of terpene reactions in superacidic media The biogenetic aspect is the last to be considered in this review. A question then arises about the relationship between the processes occurring under natural conditions and in superacidic media. According to Nishizawa et al.,18 biomimetic cyclisation of terpene alkenes is a `stepwise' process which involves intermediate for- mation of cationic species. Their properties are best studied by modelling the behaviour of carbocationic salts under `long-life' conditions. Low temperatures favour regio- and chemoselectivity of the reactions and conformational control, whereas hypernon- nucleophilic media are beneficial for more profound conversions. An example of such a conversion is cyclisation of isomanool (205) in HSO3F±SO2FCl (7110 8C) into tetracyclic compound 73, which occurs in nine steps.30 HO 205 +H+ *H + H *H H OH+ OH 203 OH 7H+ 204 + + *C7C + *C7C + *Me V A Barkhash,MP Polovinka *H + The conditions of acid catalysis are much more distant from those of biogenetic reactions because they occur in accordance with the Curtin ± Hammett hypothesis, i.e., in the absence of conformational control, and the reactions do not involve pro- found rearrangements and are often stopped due to interactions with nucleophiles.Thus, superacidic media favour biomimetic reactions `mimicking' natural processes and provide chemical evidence for practicability of various biogenetic schemes.Thus it has been shown 30 that cyclisation in superacids (HSO3F±SO2FCl, 7100 8C) of labdane diterpenes 206 and 207 with primary allylic oxygen-containing groups in the side chain can be used for stereoselective syntheses of bifunctional com- pounds of the isoagatane series of specific structures. The ease of their cyclisation at low temperatures suggests that bicyclic lab- danes might be biogenetic precursors of isoagatane compounds. Cyclisation of these compounds in ordinary acids yields a very complex mixture of reaction products in which the content of isoagatanes does not exceed 10%. O O 206 OH CH2OAc OAc 207 OHCH2OAc A similar mechanism has earlier been proposed for biogenetic cyclisation by Nishizawa et al.18 + OAc A key role in biogenetic schemes of sesquiterpenes has been ascribed to the diene 48, which is considered as a precursor of tricyclic compounds with cubebane, copabornane and copaane skeletons.However, attempts to synthesise tricyclic products from the diene 48 and the alkene 120 by acid-catalysed reactions have failed. In contrast, tricyclic compounds 208 and 209 were obtained in HSO3F±SO2FCl.57 73 O+O MeOH O+O MeOH + OOCationic molecular rearrangements of natural terpene compounds in superacidic media: reality and predictions H a, b H 48 H c, b H 120 (a) HSO3F, SO2FCl,730 8C; (b) MeOH; (c) HSO3F, SO2FCl,7110 8C. Wenkert 82 has hypothesised that tetracyclic diterpenoids with different skeletons are formed from bicyclic diterpenoids accord- ing to the following scheme: CH2OH 70 + However, the attempts at in vitro synthesis of tetracyclic diterpenes from tricyclic ones have not met with success.83 Thus treatment of pimaradienes with ordinary acids resulted merely in their isomerisation.It was therefore concluded that in contrast with the Wenkert scheme, pimaradienes are not biogenetic precursors of tetracyclic compounds. However, the use of super- acidic media allowed, for the first time, cyclisation of pimara- dienes into the tetracyclic compound 211.84 The failure to effect 13 COOMe +8 a COOMe 210 + b COMe +OH (a) [H+]; (b) HSO3F, SO2FCl,7100 8C. 208 209 + + +CH2 tetracyclic diterpenes.7H+ *H+ COOMe MeOH COOMe 211 413 this cyclisation under conditions of acid catalysis was due to the higher rate of cation deprotonation at C(8) (compound 210) resulting in the formation of a sterically hindered tetra-substituted double bond in comparison with the rate of cyclisation involving the vinyl group. The skeleton of compound 211 is similar to that of stemarine 212 isolated recently from a natural source.85 OH CH2OH 212 It is assumed that the latter is biosynthesised from pimaradienes. The preparation of the same stemarane-type stereoisomers from compounds of the pimaric and isopimaric series that are diaster- eomeric at the C(13) atom 84 provides evidence for the practic- ability of the Wenkert scheme postulating intermediate generation of a non-classical ion.X. Conclusion Thus, the main factors that control multistep multiroute cationic molecular rearrangements in superacidic media at superlow temperatures are the following: the nature of the acidic medium, the mode of generation of the cationic centre (solvolysis or proton addition to the alkene double bond), the temperature at which the cationic centre is generated and the temperature at which the acidic solution is `defrosted'. The most probable pathways of multistep cationic rearrangements can thus be predicted with due regard to the orbital and thermodynamic factors. In practical syntheses of products of multistep cationic rearrangements, preference should be given to solid acidic (particularly super- acidic) catalysts rather than to liquid superacids. References 1.H Shirahama, E Osawa, B Chnabra, T Shimokawa, T Yokono, T Kanaiwa, T Amiya, T Matsumoto Tetrahedron Lett. 22 1527 (1981) 2. V P Gatilova, D V Korchagina, T V Rybalova, Yu V Gatilov, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 25 320 (1989) a 3. G A Nisnevich, D V Korchagina, V I Makal'skii, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 29 524 (1993) a 4. A V Semenovskii, V A Smit, V F Kucherov Dokl. Akad. Nauk SSSR 132 1107 (1960) b 5. M Kurbanov, A V Semenovskii, V A Smit Izv. Akad. Nauk SSSR, Ser. Khim. 390 (1973) c 6. O A Gavrilyuk, D V Korchagina, Yu V Gatilov, V I Mamatyuk, S A Osadchii, E A Fisher, Zh V Dubovenko, V A Barkhash OMe H Zh. Org.Khim. 23 457 (1987) a 7. L E Tatarova, D V Korchagina, V A Barkhash Zh. Org. Khim. 31 83 (1995) a 8. V A Smit, A V Semenovskii, V F Kucherov Izv. Akad. Nauk SSSR, Ser. Khim. 2193 (1960) c 9. G E Muntyan, M Kurbanov, V A Smit, A V Semenovskii, V F Kucherov Izv. Akad. Nauk SSSR, Ser. Khim. 633 (1973) c 10. I G Marsakulov, A V Semenovskii, V A Smit Izv. Akad. Nauk SSSR, Ser. Khim. 384 (1967) c 11. A V Semenovskii, V A Smit, A I Dyadchenko Izv. Akad. Nauk SSSR, Ser. Khim. 743 (1974) c 12. A I Dyadchenko, A V Semenovskii, V A Smit Izv. Akad. Nauk SSSR, Ser. Khim. 2276 (1976) c 13. R P Kirchen, T S Sorensen J. Am. Chem. Soc. 100 1487 (1978) 14. N S Zefirov Tetrahedron 33 2719 (1977) 15. V G Dashevskii Konformatsii Organicheskikh Molekul (Confor- mations of Organic Molecules) (Moscow: Khimiya, 1974) p.432 16. Yu V Gatilov, Doctoral Thesis in Chemical Sciences, Novosibirsk Institute of Organic Chemistry, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, 1997414 17. M J S Dewar, C H Reynolds J. Am. Chem. Soc. 106 1744 (1984) 18. M Nishizawa, H Takenaka, Y Hayashi J. Am. Chem. Soc. 107 522 19. L E Tatarova, D V Korchagina, V A Barkhash Zh. Org. Khim. 29 20. M P Polovinka, N D Ungur, V B Perutskii, D V Korchagina, (1985) 326 (1993) a Yu V Gatilov, I Yu Bagryanskaya, V I Mamatyuk, G E Sal'nikov, P F Vlad, V A Barkhash Zh. Org. Khim. 27 2116 (1991) a 21. O A Gavrilyuk, D V Korchagina, I Yu Bagryanskaya, Yu V Gatilov, A A Kron, V A Barkhash Zh. Org.Khim. 23 2124 (1987) a 22. M P Polovinka, S A Osadchii, D V Korchagina, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 17 1223 (1981) a 23. S A Osadchii, M P Polovinka, D V Korchagina, N A Pankrushina, Zh V Dubovenko, V A Barkhash, V A Koptyug Zh. Org. Khim. 17 1211 (1981) a 24. T S Sorensen Acc. Chem. Res. 9 257 (1976) 25. M P Polovinka, D V Korchagina, V A Khan, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 21 1442 (1985) a 26. L A Popova, I I Bardyshev, D V Korchagina, Zh V Dubovenko, Yu V Gatilov, V A Barkhash Zh. Org. Khim. 18 815 (1982) a 27. E M Engler,M Farcasiu, A Sevin, J M Cense, P v R Schleyer J. Am. Chem. Soc. 95 5769 (1973) 28. E M Engler, J D Andose, P von R Schleyer J. Am. Chem. Soc. 95 8005 (1973) 29. O A Gavrilyuk, D V Korchagina, S A Osadchii, V A Barkhash Zh.Org. Khim. 24 966 (1988) a 30. P F Vlad, N D Ungur, A N Barba, L E Tatarova, Yu V Gatilov, D V Korchagina, I Yu Bagryanskaya, V P Gatilova, E N Shmidt, V A Barkhash Zh. Org. Khim. 22 2519 (1986) a 31. M P Polovinka, O G Vyglazov, D V Korchagina, V A Barkhash Zh. Org. Khim. 27 2623 (1991) a 32. V G Shubin Top. Curr. Chem. 116/117 267 (1984) 33. G A Olah Angew. Chem., Int. Ed. Engl. 34 1393 (1995) 34. R Haseltine, E Huang, K Ranganayakulu, T S Sorensen Can. J. Chem. 53 1056 (1975) 35. H C Brown The Nonclassical Ion Problem (New York: Plenum Press, 1977) 36. E M Arnett, T C Hofelich J. Am. Chem. Soc. 105 2889 (1983) 37. M P Polovinka, D V Korchagina, Y V Gatilov, I Yu Bagrianskaya, V A Barkhash J.Org. Chem. 59 1509 (1994) 38. D V Korchagina, L E Tatarova, Yu V Gatilov, V A Barkhash Zh. Org. Khim. 29 103 (1993) a 39. M A Baig, D V Banthorpe, G Carr, D Whittaker J. Chem. Soc., Perkin Trans. 2 163 (1990) 40. L Crombie, R Houghton, D Woods Tetrahedron Lett. 4553 (1967) 41. P F Vlad, N D Ungur, A N Barba, D V Korchagina, I Yu Bagryanskaya, Yu V Gatilov, V P Gatilova, V A Barkhash Khim. Prir. Soedin. 198 (1988) d 42. M P Polovinka, S A Osadchii, D V Korchagina, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 17 1629 (1981) a 43. TMKhomenko,DVKorchagina, YuVGatilov, IYu Bagryanskaya, A V Tkachev, A I Vyalkov, O B Kun, V L Salenko, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 26 2129 (1990) a 44. TMKhomenko,DVKorchagina, YuVGatilov, IYu Bagryanskaya, T V Rybalova, G E Sal'nikov, V I Mamatyuk, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 27 570 (1991) a 45. P MuÈ ller, J Blanc, J Mareda Helv. Chim. Acta 69 635 (1986) 46. N L Allinger J. Am. Chem. Soc. 99 8127 (1977) 47. H C Brown Tetrahedron 32 179 (1976) 48. A Yu Spivak, V A Barkhash Zh. Org. Khim. 12 1223 (1976) a 49. G A Olah Acc. Chem. Res. 9 41 (1976) 50. G A Nisnevich, A I Vyalkov, G T Kamshii, V I Mamatyuk, V A Barkhash Zh. Org. Khim. 19 2081 (1983) a 51. V P Gatilova, D V Korchagina, Yu V Gatilov, I Yu Bagryanskaya, V A Barkhash Zh. Org. Khim. 27 2301 (1991) a 52. T M Khomenko, D V Korchagina, Yu V Gatilov, I Yu Bagryanskaya, V A Barkhash Zh. Org. Khim. 27 599 (1991) a 53. M P Polovinka, D V Korchagina, S A Osadchii, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 21 2102 (1985) a 54. V A Barkhash Neklassicheskie Karbokationy (The Nonclassical Carbocations) (Novosibirsk: Nauka, 1984) p. 295 55. E M Arnett, N Pienta, C Petro J. Am. Chem. Soc. 102 398 (1980) V A Barkhash,MP Polovinka 56. G A Nisnevich, V I Makal'skii, D V Korchagina, Yu V Gatilov, I Yu Bagryanskaya, V A Barkhash Zh. Org. Khim. 30 161 (1994) a 57. M P Polovinka, V I Mamatyuk, D V Korchagina, G E Sal'nikov, Yu V Gatilov, T V Rybalova, L E Tatarova, S G Molodtsov, Zh V Dubovenko, V A Barkhash Zh. Org. Khim. 27 999 (1991) a 58. V A Koptyug, V G Shubin Zh. Org. Khim. 16 1977 (1980) a 59. V P Gatilova, D V Korchagina, I Yu Bagryanskaya, Yu V Gatilov, Zh V Dubovenko, V A Barkhash, V A Koptyug Zh. Org. Khim. 21 7 (1985) a 60. D Merkel Z. Chem. 9 63 (1969) 61. O A Gavrilyuk, D V Korchagina, S A Osadchii, V A Barkhash Zh. Org. Khim. 24 966 (1988) a 62. US P. 3 076 022; Chem. Abstr. 59 2672 (1963) 63. O A Gavrilyuk, D V Korchagina, L E Tatarova, E A Burdin, I Yu Bagryanskaya, Yu V Gatilov,V A Barkhash Zh. Org. Khim. 23 1723 (1987) a 64. J E McMurry J. Org. Chem. 36 2826 (1971) 65. E V Gordeeva, V V Scherbukhin, N S Zefirov Tetrahedron Comp. Method. 3 429 (1990) 66. P MuÈ ller, J Blanc, J Mareda Chimia 39 234 (1985) 67. Yu V Gatilov, T V Rybalova, V A Barkhash, V A Koptyug Zh. Org. Khim. 27 1129 (1991) a 68. T W Bentey, B Goer, W Kirmse J. Org. Chem. 53 3066 (1988) 69. M P Polovinka, A A Shal'ko, D V Korchagina, G A Zenkovets, Yu V Gatilov, V V Shcherbukhin, V A Barkhash Zh. Org. Khim. (1999) (in the press) a 70. D N J White, M S Bovill J. Chem. Soc., Perkin Trans. 2 225 (1983) 71. G Carr, G Dosanjh, A P Millar, D Whittaker J. Chem. Soc., Perkin Trans. 2 1419 (1994) 72. M P Polovinka, D V Korchagina, Yu V Gatilov, O G Vyglazov, V A Barkhash Zh. Org. Khim. (1999) (in the press) a 73. O A Gavrilyuk, D V Korchagina, V A Barkhash Zh. Org. Khim. 28 111 (1992) a 74. L E Tatarova, D V Korchagina, G V Cherkaev, A A Kron, I S Aul'chenko, V A Barkhash Zh. Org. Khim. 28 1183 (1992) a 75. D V Banthorpe, P A Boullier, W D Fordham J. Chem. Soc., Perkin Trans 1 1637 (1974) 76. GCarr, C Dean,DWhittaker J. Chem. Soc., Perkin Trans. 2 71 (1989) 77. O A Gavrilyuk, D V Korchagina, I Yu Bagryanskaya, Yu V Gatilov, V A Barkhash Zh. Org. Khim. 28 1602 (1992) a 78. W E Erman Chemistry of Monoterpenes Part A (New York: Marcel Dekker, 1985) p. 329 79. M P Polovinka, D V Korchagina, Yu V Gatilov, O G Vyglazov, G A Zenkovets, V A Barkhash Zh. Org. Khim. (1999) (in the press) a 80. Z G Isaeva, G A Bakaleinik Izv. Akad. Nauk SSSR, Ser. Khim. 648 (1997) c 81. T M Khomenko, G A Zenkovets, V A Barkhash Zh. Org. Khim. 33 655 (1997) a 82. E Wenkert Chem. Ind. 282 (1955) 83. P F Vlad Izv. Akad. Nauk Mold. SSR, Ser. Biol. Khim. Nauk 67 (1977) 84. E N Shmidt, Yu V Gatilov, I Yu Bagryanskaya, D V Korchagina, N M Bardina,M P Polovinka, S A Osadchii, S A Shevtsov, V A Barkhash Zh. Org. Khim. 21 793 (1985) a 85. P S Marchand, J F Blount J. Chem. Soc., Chem. Commun. 894 (1975) a�Russ. J. Org. Chem. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) c�Russ. Chem. Bull. (Engl. Transl.) d�Chem. Nat. Compd. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (5) 415 ± 433 (1999) Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands { I D Sadekov, A I Uraev, A D Garnovskii Contents I. Introduction II. Complexes with organic derivatives of monocoordinated tellurium III. Complexes with organic derivatives of dicoordinated tellurium IV. Conclusion Abstract. Published data on the synthesis, reactivity and structure of complexes formed in the reactions of organic derivatives of tellurium with metal carbonyls and cyclopentadienyl carbonyls are described systematically and generalised. The bibliography includes 142 references. I. Introduction The substantial progress achieved in synthetic organotellurium chemistry has stimulated studies dealing with metal complexes with organotellurium ligands.1±11 In recent years, considerable attention has been paid to metal complexes, in particular, com- plexes of metal carbonyls and cyclopentadienyl carbonyls with organochalcogens, including organotellurium, ligands. Study of these complexes has allowed researchers to elucidate the influence of the nature of donor atoms (sulfur, selenium and tellurium) on the reactivity and structural and spectral features of compounds of the same type containing these atoms.Compounds prepared by the reactions of metal carbonyls with phosphine tellurides 12 ± 14 and diphenyl ditelluride 15 and of tetracarbonylmanganese com- plexes with tellurium-containing ligands 16, 17 are convenient pre- cursors of metal tellurides, the synthesis of which can be performed at relatively low temperatures.This review is concerned with the synthesis, structures and reactivity of complexes formed by metal carbonyls and cyclo- pentadienyl carbonyls with organotellurium ligands. Various tellurium-containing clusters which contain either telluride (± Te ±) or ditelluride (± Te ± Te ±) bridges have been considered in earlier reviews; 2 ± 4, 18 therefore, we do not dwell on them here. II. Complexes with organic derivatives of monocoordinated tellurium This Section contains data on the reactions of metal carbonyls and cyclopentadienyl carbonyls with organic derivatives of mono- coordinated tellurium. This type of derivative includes phosphine tellurides R3P=Te and compounds with C=Te bonds.Since organyltellurolate anions RTe7 can also be formally classified as I D Sadekov, A I Uraev, A D Garnovskii Research Institute of Physical and Organic Chemistry of the Rostov State University, prosp. Stachki 194/2, 344090 Rostov-on-Don, Russian Federation. Fax (7-863) 228 56 67. Tel. (7-863) 228 08 94 (I D Sadekov). Tel. (7-863) 228 57 76. E-mail: garn@ipoc.rnd.runnet.ru (ADGarnovskii) Received 22 July 1998 Uspekhi Khimii 68 (5) 454 ± 473 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 546.24.07+546.711+547.8 415 415 422 431 derivatives of monocoordinated tellurium, this Section also presents data concerning complexes with bridging and terminal RTe groups.1. Phosphine tellurides Methods for the synthesis and some reactions of complexes of phosphine tellurides with metal carbonyls and cyclopentadienyl carbonyls have been considered fairly comprehensively in a review.19 Therefore, here we present only recent data concerning the structure, synthesis and transformations of some of these compounds (Table 1). Spectral characteristics of complexes 1a ± c are listed in Table 2. Table 1. Some complexes of phosphine tellurides with metal carbonyls and cyclopentadienyl carbonyls. Initial compounds Ref. Reaction products R3P=Te Metal complexes But3P=Te M(CO)6 (M=Cr, Mo, W) (CO)5M7Te=PBut3 (1a ± c) a 20 (M=Cr (a), Mo (b), W(c)) Et3P=Te Co2(CO)8 Mn2(CO)10 [(Et3P)2(CO)2CoTe]2 (2) 13, 14 (Et3P)4(CO)6Co4Te2 (3) (Et3P)6Co6Te8 (4) b [(Et3P)2(CO)3MnTe]2 (5) 1216, 17 BnTeMn(CO)3(PMe3)2 (6a) Me3P=Te BnMn(CO)5 Et3P=Te BnMn(CO)5 MeMn(CO)5 BnTeMn(CO)3(PEt3)2 (6b) 16, 17 MeTeMn(CO)3(PEt3)2 (6c) c, d 16, 17 MeMn(CO)3(PEt3)2 (7) 23, 24 R3P=Te [CpFe(CO)2THF]+.[CpFe(CO)2TePR3]+BF¡4 .BF¡4 (9) (10) e a As opposed to thiophosphorane complexes of a similar structure, 4 phosphine telluride in the complexes 1a ± c cannot be replaced by Ph3P or CO even at 80 8C.21 bA complex with a similar structure, (Ph2PrP)6Co6Te8, was prepared by the reaction of Ph2PrP with PhTe- SiMe3 and CoCl2.22 cA mixture of 6c and 7 is formed. d The complex 6c can also be prepared by the reaction of BrMn(CO)3(PEt3)2 (8) with MeTeLi.17 e Decomposition of the complex 10 affords the complex [CpFe(CO)2PR3]+BF¡ (11), while the reaction of 10 with I2 gives [CpFe(CO)2I]2 (12).23 { Dedicated to the blessed memory of Doctor of Chemical Sciences Gusein Kurban-Ismailovich Magomedov.416 Table 2.Spectral characteristics of the complexes 1a ± c.20 dC=O d31 d125Tea P Com- pound trans cis 1a 218.5 224.9 7635 53.5 (744)b (7839)b 1b 206.8 213.1 54.7 7749 1c 199.4 200.4 49.9 7770 a With respect to Me2Te. b The chemical shifts for the free ligand are given in parentheses. According to X-ray diffraction analysis,20 complex 1c has an octahedral structure, the P ± Te bond length being equal to 2.439A. Comparison of this value with the lengths of the P=Te bonds in phosphine tellurides, which range from 2.305 to 2.371A,25 ± 31 shows that the P ± Te bond in the complex 1c is mainly a single bond.This finding was explained 20 by assuming that the bipolar ionic structure But3P+7Te=M27(CO)5 makes the predominant contribution to the ground electronic state of the complex 1. The structures of clusters 2 ± 5 have been studied by X-ray diffraction analysis. Here we shall briefly describe only the structures of the compounds 2 14 and 5 12 closest to compounds that constitute the subject of this review. The molecules of the clusters 2 and 5 contain ditelluride bridges, Te ± Te, the tellurium atoms of these bridges being bound to the Co and Mn atoms, respectively. The Te ± Te bonds (2.765 A in the cluster 2 and 2.763A in the compound 5) are somewhat longer than those in diaryl ditellurides [2.697 A in di(4-methylphenyl) ditelluride,32 2.702 A in di(4-chlorophenyl) ditelluride,33 2.710 A in di(4-meth- oxyphenyl) ditelluride 34 and 2.712 A in diphenyl ditelluride 35].The Co ±Te ± Te and Mn± Te ± Te bond angles (101.78 and 105.58, respectively) are also somewhat greater than the C± Te ± Te angle in diaryl ditellurides [for example, in di(4-methoxy- phenyl) ditelluride, this angle is 99.88]. The coordination polyhe- Table 3. Complexes of transition metal carbonyls with monocoordinated tellurium. Initial compounds metal carbonyls (CO)5W=CPhR (13a,b) [R=Ph (a), H (b)] (CO)5W.THF M(CO)5 .MeCN [M=Cr (a), Mo (b), W (c)] ac Spectral characteristics of the compound 16c: d13C=301 (C=Te), d125Te=2858, lmax=825 nm.40 b Diferrocenyl telluroketone does not form complexes with metal carbonyls.39 Spectral characteristics of the compound 17: d13C=285 (C=Te), d125Te=1783, lmax=525 nm.42 d The complex 19a (M=Cr) decomposes on exposure to light to give complex EtN(CH2)2N(Et)Cr(CO)5 (20).nC=O / cm71 J31P¡125Te /Hz 1502 (1600)b 1602 1600 2052, 1927, 1986, 1913, 1937 2055, 1937, 1988, 1912, 1941 2063, 1927, 1982, 1909, 1932 tellurium derivatives (Ph3P=N=PPh3)+(TeCN)7 Me Me C Te(16a ± c) R Me Me [R=(CH2)2 (a), CH=CH (b), o-C6H4 (c) a] (R=o-C6H4) NEt Te (18) NEt I D Sadekov, A I Uraev, A D Garnovskii dron about the Co atom is a distorted trigonal bipyramid in which the apical positions are occupied by the Et3P groups; the environ- ment of the Mn atoms is a distorted octahedron with trans- arranged phosphine ligands.According to X-ray diffraction data,16, 17 the crystal unit cell in the complex BnTeMn(CO)3(PEt3)2 6a contains two crystallo- graphically independent molecules, which differ in the conforma- tion of the ethyl groups at the phosphorus atoms. The bond geometry about the manganese atom is a slightly distorted octahedron with trans-arrangement of the Et3P groups. The Mn±Te bond length (2.702 A) is close to the bond length in the cluster 5 (2.763 A),12 and the C± Te bond length is within the range normal for this type of bond.36 2. Tellurocarbonyl compounds Synthesis and reactions of complexes of metal carbonyls with tellurocarbonyl compounds have been described in detail in a review.19 Here we present only some examples of these complexes for which X-ray diffraction data have been obtained or chemical reactions have been studied (Tables 3, 4).The bond lengths in the complex 17, determined by X-ray diffraction analysis,42 point to the presence of a weak coordina- tion bond W± Te, the length of which (2.808 A) is close to the upper limit of the lengths of single W± Te bonds (2.88 A).46 The C=Te bond length amounts to 1.987 A, which is close to the length of this bond in S=C=Te (1.904 A47). This indicates that the carbon ± tellurium bond in the complex 17 is close to a double bond. The fact that upon heating of the complex 17 in a MeCN solution, the solvent displaces the telluroketone 16c to give the complex (CO)5W.MeCN confirms the conclusion that the coor- dination in this complex is weak.42 The molecular and crystal structures of the complex 19a were studied by X-ray diffraction analysis.43 The geometry of bonds about the chromium atom corresponds to a distorted octahedron.The C± Te bond length (2.12 A) falls into the range of lengths typical of single C± Te bonds,36 while the Cr ±Te ±C bond angle is 96.18. The Cr ± Te bond length (2.765 A) differs only slightly from + EtN EtN +NEt NEt 7 7 Te Te (CO)5Cr (CO)5Cr Ref. Reaction products 37, 38 (CO)5W7Te CPhR (14a,b) (CO)5WTe CPhH (15) W(CO)5 Me Me R C Te.W(CO)5 (17) b, c 40 ± 42 Me MeEt N 43 Te .M(CO)5 (19a ± c) d NEtSynthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands Table 4.Products of the addition of the complexes 14a,b to unsaturated compounds. Ref. Complex Unsaturated Product compound (CO)5W Me Me 39, (CO)5WTe=CPhR Te (14a,b) 44 (R=Ph (a), H (b)) Me Me Ph R21a,b (CO)5W Te 34, (CO)5WTe=CHPh (14b) 44Ar=Ph, 4-MeOC6H4 . Ph H exo-22, endo-22 NEt2 45 (CO)5WTe=C RC:CNEt2 C=CPh2 (CO)5WTe=CPh2 (14a) 23 R those in other complexes containing chromium and tellurium. The C± Te bond in the complex 19a is much longer than the carbon ± tellurium bond in the compound 17 (1.987 A), which is due to the predominant contribution of polar resonance structures to the ground state of the molecule.43 3.Organyltellurolate complexes An important method for the synthesis of the complexes 24, containing metal carbonyl fragments and organyltellurolate (mainly, aryltellurolate) bridging groups is based on reactions of various metal carbonyls 25 with diaryl ditellurides.48 ± 54 Dodecacarbonyltriiron,48 ± 51 dodecacarbonyltriruthenium,49 dodecacarbonyldimanganese 53, 54 and hexacarbonylmolybde- num 52 enter into this reaction. The number of diaryl ditellurides that have been involved in this reaction is quite limited. The use of diphenyl ditelluride,49, 52 ± 54 bis(pentafluorophenyl) ditellur- ide 50, 51 and di(4-methoxyphenyl) ditelluride 48 has been reported.7CO [M(CO)zTeAr]2 24a ± f Mx(CO)y+Ar2Te2 25a ± d Ref. Mx(CO)y [M(CO)zTeAr]2 Fe3(CO)12 (25a) [Fe(CO)3TeC6F5]2 (24a) 50, 51 [Fe(CO)3TePh]2 (24b) 49 [Fe(CO)3Te C6H4OMe-4]2 (24c) 48 [Ru(CO)3TePh]2 (24d) 49 [Mo(CO)4TePh]2 (24e) 52 [Mn(CO)4TePh]2 (24f) 53, 54 Ru3(CO)12 (25b) Mo(CO)6 (25c) Mn2(CO)10 (25d) These reactions occur under relatively mild conditions Whereas in the case of Fe3(CO)12 25a, the complexes 24a ± c are the only reaction products, the reaction of ruthenium carbonyl 25b with diphenyl ditelluride yields, depending on the ratio of the reactants, either a mixture of 24d and [Ru(CO)2(TePh)2]n (n=6, 7, 12) 26 or only the latter compound.49 When a molar excess of diphenyl ditelluride is used, only the complex 26 is formed, whereas with a 25b : Ph2Te2 ratio of 3 : 2, a mixture containing the compound 26 (*90%) and the complex 24d (*10%) is produced.49 Variation of the reaction temperature has a substantial influence on the composition of manganese complexes.Indeed, UV irradiation of a tetrahydrofuran solution of Mn2(CO)10 25d and diphenyl ditelluride at room temperature affords the bis- bridged complex 24f,53 whereas refluxing of the initial solution yields paramagnetic compound 27, containing three bridging PhTe groups.53 417 THF, D 7CO Mn2(CO)10+ Ph2Te2 25d Mn2(CO)6(TePh)3 27 Unlike other carbonyls, Co2(CO)8 reacts with diaryl ditellur- ides in an atmosphere of CO (pressure 10 ± 100 atm) to give telluroesters.55, 56 It was assumed 56 that these reactions yield complexes 28 as intermediates.Ar2Te2+Co2(CO)8 MeCN, 10 ± 100 atm 125 ± 200 8C ArCTeAr [ArTeCo(CO)4] 28 O In the case of dialkyl ditellurides, no telluroesters are formed, the corresponding dialkyl tellurides being the only reaction products.56 Diaryl tellurides can be used instead of diaryl ditellurides as the sources of bridging ArTe ligands in the preparation of complexes of the type 24. Thus the complex 24f is formed when a mixture of manganese carbonyl 25d and diphenyl telluride is heated under rigorous conditions.48 p-xylene, 130 8C, 125 h [Mn(CO)4TePh]2 Mn2(CO)10+ Ph2Te 24f 25d Yet another approach to the synthesis of binuclear manganese complexes of type 24, containing various bridging ligands TeR [R=Ph, Alk, (Me3Si)3Si] is based on the nucleophilic replace- ment of the bromine atom in bromo(pentacarbonyl)manganese by lithium organyltellurolates.17, 53 The reaction products 29 formed initially are quite unstable; they eliminate a CO molecule being thus converted into the binuclear complexes 24 in 10% ±52% yields.17 The attempts to stabilise the mononuclear complexes 29 by introducing the bulky tris(trimethylsilyl)silyl group failed.16 BrMn(CO)5+ RTeLi 7CO 7LiBr [RTeMn(CO)5] 29 D MnTe [Mn(CO)4TeR]2 24f ± k R=Ph (f), Me (g), Et (h), Pri (i), CH2SiMe3 ( j), (Me3Si)3Si (k). The reductive thermolysis of the complexes 24g, h carried out by heating them under a hydrogen atmosphere at 300 and 200 8C, respectively, affords films containing manganese telluride in addition to manganese and tellurium.17 The attempts to convert the complexes 24 into the complexes 6 (cf.Section II.1) by substituting triorganylphosphines for the CO groups were unsuccessful.17 At room temperature, the reaction proceeded very slowly, while at an elevated temperature, it gave a complex mixture of non-identified products. Nucleophilic substitution of chlorine in iridium complex 30 occurring on treatment of this complex with relatively stable lithium 2,4,6-tris(tert-butylphenyl)tellurolate has been used to prepare complex 31 the structure of which was studied by X-ray diffraction analysis.57 (Ph3P)2Ir(CO)Cl + 2,4,6-But3C6H2TeLi 7LiCl 30 (Ph3P)2Ir(CO)TeC6H2But3-2,4,6 31 (80%) Recently the complex 24g was synthesised 58 by virtue of a new tellurium-containing reagent, dimethyl(methyltelluro)tellu- ronium tetrafluoroborate 32 (which was prepared by alkylation of dimethyl ditelluride by trimethyloxonium tetrafluoroborate in acetonitrile).Treatment of a THF solution of the compound 32 with salt 33 gives initially, like the reaction of bromo(penta- carbonyl)manganese with lithium methyltellurolate described418 above,17 the mononuclear complex 29 (R=Me), the existence of which in solution at 0 8C was proved by spectroscopy (IR, 1H NMR). This product rapidly dimerises with abstraction of CO to give the complex 24g. + THF 7(PPN)BF4 ,7Me2Te2 Me2Te TeMe+(PPN)[Mn(CO)5] BF¡ 32 33 4 7CO [(CO)4MnTeMe]2 24g + MeTeMn(CO)5 29 PPN = Ph3P=N=PPh3 .Since the natures of both components can be varied, this reaction can become a convenient method for the synthesis of binuclear complexes of type 24. The prospects for using the salt 32 and its analogues in preparative organic chemistry are not less important. Sulfonium salts with similar structures containing alkyl and aryl radicals at the disulfide bridge 59 ± 61 and also dimethyl(phenylseleno)sulfonium tetrafluoroborate 62 are used in the synthesis of episulfonium salts, S-arylation (alkylation) of alkenes, phenylselenation of aromatic compounds and in other reactions. Manganese complexes of the type 24 have also been obtained using tellurium-containing anionic compounds 34a,b (their syn- thesis is described below).63 Treatment of the salt 34a with triphenylmethyl tetrafluoroborate gave the complex 24f in 34% yield.The same complex was isolated in 30% yield together with anionic complex 35 when salt 36a (M=Te) was made to react with `pentacarbonyl(bromo)manganese.58 Ph3C+BF¡4 7(PPN)BF4 (PPN)[(CO)3Mn(TeR)3Mn(CO)3] 34a,b [(CO)4MnTePh]2 24f R=Ph (a), Me (b). 36a (PPN)[Mn(CO)4(MPh)2]+BrMn(CO)5 36a (PPN)[(CO)3BrMn(MPh)2Mn(CO)4] + 24f 35 M=Te. Now we shall briefly describe two other reactions resulting in the formation of complexes with bridging aryltellurolate groups. Thus the compound 24f has been synthesised by refluxing com- plexes 37, prepared by the reaction of octacarbonyldihaloman- ganeses (for details, see Section III.2) with diphenyl ditelluride, in heptane 64 or benzene .53 D 7MnX2 ,7CO,7Ph2Te2 [(CO)4MnTePh]2 24f Mn2X2(CO)6 .Te2Ph2 37a,b X=Br (a), I (b). Osmium complex 38 (yield 7%) has been obtained, together with non-identified compounds, on treatment of complex 39 with di(4-methoxyphenyl)tellurium dichloride at room temperature.65 Thermolysis of the complex 38 (refluxing in octane for 2 h) affords cluster 40 (yields 6%). The mechanism of transformation of the compound 39 into 38 has not been discussed. (PPN)[HOs3(CO)11]+(4-MeOC6H4)2TeCl2 D 39[HOs3(CO)10TeC6H4OMe-4] 38 Os3Te2(CO)9 40 The molecular and crystal structures of the complexes 24e,52 24f,53, 63 and 24g 58 have been studied by X-ray diffraction analysis (Fig.1). The structure of the molecules 24e ± g is a planar rhombus M2Te2, in which the methyl groups occupy syn-positions;58 the phenyl groups in the complex 24f occupy trans-positions. The most important bond lengths and bond angles in the complexes 24e ± g are presented in Table 5. The 1H, 13C and 125Te NMR spectra of the complex 24g indicate the presence of two different TeMe groups, which is I D Sadekov, A I Uraev, A D Garnovskii R R R MTe COCO CO OC OC OC COCO CO OC OC OCOC OC OC OC R B A Figure 1. Molecular geometry of 24e ± g; A is the syn-isomer, B is the anti-isomer. Table 5. Bond lengths and bond angles in the complexes 24e ± g.58 Bond angles/ deg.Bond lengths/ A Complex M7Te7M Te7M7Te C7Te M7Te 111.2 68.8 2.133 2.756 84.04 95.96 2.129 2.661 [Mo(CO)4TePh]2 24e [Mn(CO)4TePh]2 83.66 95.71 2.183 2.661 24f [Mn(CO)4TeMe]2 24g consistent with the finding that two conformers (A and B, see Fig. 1) exist in solution in a ratio of 4 : 5.17 The signals of the two TeMe groups in this complex coalesce at 361 K, and those for 24j coalesce at 343 K, although the 125TeNMR spectrum of the latter compound exhibits only one 125Te signal. The data of multi- nuclear NMR spectroscopy for solutions of the complexes 24h,i also point to the presence of two isomers. Judging by the fact that 125Te NMR spectra of the complexes 24f,k exhibit singlet 125Te signals, these complexes exist as a single, more stable, isomer due to fast interconversion of the isomers.17 The binuclear manganese complex 27,53 which contains three bridging PhTe groups, has been described above.Recently,63, 66 trinuclear complexes containing either five PhTe bridges (42a, M=Te) or two PhTe bridges and three PhSe bridges (42b, M=Se) have been prepared. The complex 42a was synthesised by the reaction of cobalt perchlorate hexahydrate with diphenyl ditelluride in the presence of the anionic complex 36a; the reaction is accompanied by oxidation of Co2+ to Co3+.63 The mixed selenium-and-tellurium complex was prepared from cobalt per- chlorate hexahydrate, diphenyl diselenide and the anionic com- plexes 36a and 36b.66 Evidently, the compounds 42a,b result from rearrangement of the complexes 41a,b formed intermediately.(PPN)[Mn(CO)4(TePh)2]+(PPN)[Mn(CO)4(MPh)2] + 36a 36a,b +Co(ClO4)2 . 6H2O+Ph2M2 Ph Ph CO Te OC Mn OC Te CO CO M CO Co Mn MPhM CO CO Ph Ph41a,b Ph Ph CO MPh M CO OC Te Mn Mn Co CO M Te OC CO CO CO Ph Ph 42a,b (>90%) M=Te (a), Se (b).Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands According to X-ray diffraction data,63, 66 the complexes 42a,b contain linear chains of atoms Mn±Co ± Mn. The Co(III) ion is bound either to five PhTe bridging ligands and a terminal CO group (42a) or to two PhTe bridging groups, three PhSe bridging groups and a carbonyl group (42b).Thus, the coordination environment of the Co(III) atom in both complexes is a distorted octahedron. One Mn atom in 42a,b is coordinated to two PhTe bridging ligands and four terminal carbonyl groups; the other manganese atom is linked to three carbonyls and three PhTe (42a) or PhSe (42b) bridges. Thus, the coordination polyhedron formed about each manganese is also a distorted octahedron. The Te ± metal bond lengths in 42a are the following: Co ± Te (Mn ± Te) 2.657A;62 in 42b, they are 2.576 and 2.654A, respectively; the Co ± Se bond length is 2.450 A.66 Unlike octacarbonyldihalo-manganese 64 and -rhenium ,67 which react with diphenyl ditelluride with retention of the Te ± Te bond (for details, see Section III.2), oxidative addition of diaryl ditellurides to the carbonyl chloro iridium complex 43 occurs with rupture of the Te ± Te bond and gives complexes 44 containing terminal ArTe ligands.Spectral characteristics of the complexes (shift of the singlet signal in the 31PNMRspectrum by*100 ppm in relation to that in the spectrum of the initial complex 43 and the 125Te chemical shifts) are consistent with the structure 44; how- ever, the exact geometry of these complexes is unknown. During the reaction, a broad EPR signal typical of tellurium-centred radicals was observed; based on this fact, a reaction mechanism was proposed which includes the initial addition of the ditelluride to the Ir(I) complex 43 followed by homolytic cleavage of the Te ± Te bond to give the Ir(III) complexes 44.68 PhMe, D 2 h IrCl(CO)(PPh3)2+Ar2Te2 43 IrCl(TeAr)2(CO)(PPh3)2 44 Ar=4-EtOC6H4 , 4-MeC6H4 , Ph, 4-ClC6H4 .Among other halo carbonyl complexes containing ArTe ligands, rhodium complex 45 should be noted. It was synthesised in a yield of more than 50% by passing CO through a benzene solution of complex 46.69 (Ph2Te)2Rh(CO)TePhCl2 (Ph2Te)2RhTePhCl2+CO 45 46 In addition to the neutral carbonyl and halo carbonyl com- plexes described above, anionic carbonyl complexes with terminal RTe ligands have been reported. The synthesis of these compounds is based on a reaction typical of diorganyl ditellurides, namely, cleavage of the Te ± Te bond by various nucleophiles: organolithium and organomagnesium compounds, alkalis, cyanide anion, etc.70 When anionic metal carbonyls such as [Mn(CO)5]7 (see Refs 58, 63) and [HFe(CO)4]7 (see Refs 71 ± 74) are used as nucleophiles, the reactions with diorganyl ditellurides result in the synthesis of the anionic tellurium- containing complexes 36a, 47a,b,58, 63 and 48,71 ± 74 the yields of the complexes being higher than 90%.The reactions occur under mild conditions (THF, room temperature) at a reactant molar ratio of 1 : 1. Dimethyl ditelluride reacts with [Mn(CO)5]7 similarly to diphenyl ditelluride but the reaction occurs more slowly. The formation of an intermediate complex of the type 36b was judged only by the IR spectra of the reaction mixture; the reaction with dimethyl ditelluride yields the complex 34b as the final product.63 THF, 20 8C 7CO X[Mn(CO)4(TeR)2] 36a, 47a,b X[Mn(CO)5] +R2Te2 33 X=PPN: R=Ph (36a), Me (47a); X=Na-18-C-6, R=Ph (47b).THF, 20 8C (PPN)[HFe(CO)4] + Ph2Te2 7PhTeH (PPN)[PhTeFe(CO)4] 48 419 The complexes 36a, 47a,b, and 48 are crystalline compounds, stable in the absence of air and readily soluble in polar solvents. The molecular and crystal structures of the complexes 47b 63 and 48 72 were studied by X-ray diffraction analysis. The crystal lattice of the former compound contains separate [Na+18-C-6 . 2 THF] cations and cis-[Mn(CO)4(TePh)2]7anions. As was to be expected, the manganese ion has an octahedral configuration of bonds; the average Mn± Te bond length is 2.673 A. The complex 48 72 has a trigonal-bipyramidal configuration of bonds about the Fe atom; the TePh group occupies the axial position.The Fe ± Te bond length is 2.630 A. The Fe ±COax bond (1.83 A) is substantially longer than the Fe ±COeq bond (1.73 A). The transformation of the compounds 36a and 47a on reflux- ing in a THF solution and the transformation of the complex 47b at room temperature afford the complexes 34a,b and 49 in high yields.63 36a, 47a,b D 7CO X[(CO)3Mn(TeR)3Mn(CO)3] 34a,b, 49 X=PPN: R=Ph (34a), Me (34b); X=Na-18-C-6, R=Ph (49). According to X-ray diffraction data,63 the crystals of 34a,b consist of discrete cations and anions with slightly distorted octahedral configurations of bonds about each manganese atom (three bridging RTe ligands and three terminal carbonyl groups).TheMn±Te bond lengths are 2.674 A in 34a and 2.688 A in 34b.63 It is noteworthy that theMn± Te bond in the neutral binuclear triply bridged compound 27 (2.638 A) 53 is on the average some- what shorter than the corresponding bonds in the anionic com- plexes 34a,b. The Mn± Te ±Mn bond angles (average values) are also appreciably different; in the neutral derivative 27, this angle (70.9 8) 53 is much smaller than those in the anionic complexes 34a and 34b (82.8 8 and 81.0 8, respectively).63 The complex 36a is detellurated upon the reaction with iodine 58 or NOPF6 in acetonitrile 63 to give complexes 50 and cis-51. NOPF6 (PPN)[Mn(CO)3(MeCN)3] MeCN 50 I2 , THF (PPN)[Mn(CO)4(TePh)2] 36a 7Ph2Te2 (PPN)[Mn(CO)4I2] cis-51 The reaction of the complex 34a with Ph3C+BF¡4 giving rise to the complex 24f 63 and the reaction of the complex 36a with pentacarbonyl(bromo)manganese giving a mixture of 24f and 35 58 have been described previously.The anionic iron complex 52, containing three terminal PhTe groups, is formed in the reaction of the anionic iron carbonyl derivative with two moles of Ph2Te2 in THF at room temper- ature.71 The reaction rate markedly increases when the amount of Ph2Te2 in the reaction mixture increases. The complex 48 is formed as the intermediate compound. (PPN)[HFe(CO)4] + Ph2Te2 [48] 7CO 7CO fac-(PPN)[Fe(CO)3(TePh)3] 52 It should be noted that the complex 52 is formed in the reaction of (PPN)[HFe(CO)3PPh3] with diphenyl ditelluride instead of the expected complex (PPN)[Fe(CO)2PPh3(TePh)3], i.e.in this case, the triphenylphosphine ligand rather than CO is replaced.71 The crystal structure of the solvate 52 . 0.5THF is formed from separate cations and anions.71 The coordination polyhedron around the iron atom is a distorted octahedron with a Fe ± Te bond length of 2.630 A. The complex 48 is a convenient starting compound for the synthesis of both new and known neutral 72, 73 and anionic420 tellurium-containing metal carbonyl complexes.74 Thus its alky- lation with methyl iodide gives rise to an oily telluride complex 53,72, 73 while the reaction with HBF4 affords the binuclear compound 24b, described above.72 [Fe(CO)3TePh]2 24b HBF4 7CO, 7H2, 7(PPN)BF4 (PPN)[PhTeFe(CO)4] 48 MeI 7(PPN)I PhTeMeFe(CO)4 53 It follows from published data 71, 74 that the complex 48 undergoes oxidative addition to diphenyl ditelluride to give the complex 52.The reaction of the complex 48 with Ph2Se2 follows a similar pattern and gives rise to complex 54.74 Diphenyl disulfide does not react with 48. THF, 20 8C 7CO (PPN)[PhTeFe(CO)4] + Ph2Se2 48(PPN)[Fe(CO)3TePh(SePh)2] 54 (90%) Thus, the ability of diphenyl chalcogenides to enter into oxidative addition reactions with the complex 48 decreases in the sequence Ph2Te2 > Ph2Se2 > Ph2S2 due to the decrease in the nucleophi- licity of the chalcogens. Complex 55, in which the ratio of PhTe to PhSe groups is the opposite to that in the complex 54, was prepared by the reaction of the selenium analogue of the complex 48 with diphenyl ditellur- ide.74 THF, 20 8C (PPN)[PhSeFe(CO)4] + Ph2Te2 7CO (PPN)[Fe(CO)3SePh(TePh)2] 55 (90%) The molecular and crystal structures of the complexes 54 and 55 were studied by X-ray diffraction analysis.74 The reactions of diaryl ditellurides with metal cyclopenta- dienyl carbonyls are mostly accompanied by rupture of the Te ± Te bonds and give complexes containing terminal or bridging ArTe ligands.Depending on the conditions (temperature, solvent, the presence or absence of UV radiation), the reactions of diaryl ditellurides with molybdenum,75 ± 77 chromium 78 and iron 79, 80 cyclopentadienyl carbonyls give rise to complexes of four types.Under relatively mild conditions (room temperature or short-term refluxing in benzene), mononuclear complexes 56 with terminal ArTe ligands are formed. [Z5-C5R5M(CO)n]2+Ar2Te2 Z5-C5R5M(CO)nTeAr+[Z5-C5R5M(CO)n71TeAr]2 57a ± e 56a ± e n Ref. Ar M R Compound 56 abcde 75 78 76 79 80 Ph Ph Ph Ph 4-EtOC6H4 Mo Cr Mo Fe Fe HHMe HH 33322 An X-ray diffraction study showed that the Te ±C and Te ±Cr bond lengths in the complex 56b are 2.117 and 2.763 A, respec- tively; the C± Te ±Cr bond angle is 105.48.78 In some cases, in addition to the complexes 56, binuclear complexes 57 with bridging ArTe groups were also isolated. They become the major products when the reaction is carried out under more drastic conditions (many-hour heating in toluene at 60 8Cor under reflux for chromium and molybdenum complexes or refluxing in benzene for iron compounds).The complexes 57 can also be obtained on heating or exposure to UV radiation of solutions of the compounds 56. The complex 57a is formed in 50% yield when [CpMo(CO)2]2 is made to react with diphenyl ditelluride in toluene at room temperature.77 Refluxing of bis(tricarbonylcyclopentadienyl)molybdenum with diphenyl ditelluride in xylene results in the formation of completely decarbonylated complex 58.75 When bis(tricarbonyl- cyclopentadienyl)chromium reacts with diphenyl ditelluride in toluene at 808C, complex [CpCr(TePh)]2Te 59, containing three telluride bridges, is produced.78 CpCr [CpMo(TePh)2]x 58 It should be noted that the reaction of tricarbonyl(cyclopen- tadienyl)molybdenum with diphenyl diselenide occurs under milder conditions than that with diphenyl ditelluride.75 In this case, the complexes CpMo(CO)3SePh and [CpMo(SePh)2]x were isolated, although judging by the IR spectra, the solution con- tained the selenium analogue of the complex 57.The reaction with diphenyl disulfide gave the completely decarbonylated complex of the type 58 as the only product.75 Thus, the capability of ArE groups (E=S, Te, Te) of stabilising mono- and binuclear com- plexes decreases in the series Te>Se>S.75 The same series of the stability of complexes as a function of the chalcogen nature holds for iron mononuclear complexes 56.79 Several geometric isomers are possible for the complexes 57a,c.However, the 1H NMR spectrum of the complex 57a contains only one set of signals of cyclopentadienyl protons, indicating that only one isomer is present in the solution.75 The binuclear complexes of iron 57d,e can exist as five geometric isomers (A± E, Fig. 2). Judging by the 1H NMR spectra, the complex 57d exists in solution as a mixture of two isomers; the attempts to separate the mixture into components failed.79 Te Cp Cp Fe Fe Ar CO OC Te Ar A (cis-I) Ar Te Cp Cp Fe Fe Ar CO OC Te C (cis-III) Ar Te CO Cp Fe Fe Ar Cp OC Te E (trans-II) Figure 2. Geometric isomers of the complexes 57d,e. The given nomen- clature is proposed in Ref.80. I D Sadekov, A I Uraev, A D Garnovskii Ph Te3 Te1 CrCp Te Ph2 59 Ar Te Cp Cp Fe Fe CO OC Te Ar B (cis-II) Ar Te CO Cp Fe Fe Cp OC Te Ar D (trans-I)Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands In the case of 57e, crystals of two isomers (A and D, see Fig. 2) have been isolated in <5% yields.80 The molecular structures of these isomers were studied by X-ray diffraction analysis.80 Both isomers contain Fe2Te2 rings, which are slightly bent along the axis connecting the iron atoms. The Te ± Fe bond lengths (aver- age) are 2.541 A for isomer A and 2.547 A for isomer D; the Fe ± Te ± Fe angles are 82.888 and 94.168, respectively.The environ- ments of the iron and tellurium atoms in both isomers are approximately the same. Each iron atom is situated at the centre of a distorted tetrahedron and the tellurium atoms are at a vertex of a trigonal pyramid. The difference between the isomers is that identical substituents (cyclopentadienyl, p-ethoxyphenyl, and carbonyl groups) in isomer A (Ar=p-EtOC6H4) occupy cis- positions with respect to each other, whereas in isomer D (Ar=p-EtOC6H4), they are trans-arranged. According to a proposed nomenclature,80 the former compound is defined as the cis-I isomer and the latter one is referred to as trans-I. The molecular structures of binuclear molybdenum (57a) 81 and chromium (57b) 78 complexes have also been studied by X-ray diffraction analysis.The structure of the complexes 57a,b is close to the structure of the corresponding sulfur derivatives. Both compounds incorpo- rate non-planar four-membered M2Te2 rings with Mo± Te dis- tances of 2.822 A and Cr ±Te distances of 2.721 A. The Te ±Mo ± Te and Mo± Te ±Mo bond angles are 70.18 8 and 97.05 8, respec- tively, and Te ±Cr ± Te and Cr ± Te ± Cr, are 70.39 8 and 98.12 8. The Te ± Te distance in the complex 57a is 3.24A, while that in 57b is 3.14 A. Taking into account the fact that the sum of the van der Waals radii of tellurium atoms is 4.4 A,82 it can be considered that there exist weak bonds between the tellurium atoms in these complexes. In the molybdenum complex 57a, the phenyl groups, the cyclopentadienyl rings, and the carbonyl groups occupy trans- positions with respect to each other.81 The bond lengths and bond angles in compound 59 (X-ray diffraction data 78) are listed in Table 6.Table 6. Average bond lengths and angles in the complex 59.78 Size/ deg Bond angle Length/ A Bond 68.47 67.06 67.34 Cr7Te17Cr Cr7Te27Cr Cr7Te37Cr 2.15 2.13 2.609 2.65 2.647 Te27C Te37C Te17Cr Te27Cr Te37Cr The heteronuclear cyclopentadienyl carbonyl complexes 60 and 61 with bridging PhE ligands (E=S, Se, Te) have been synthesised by the reaction of bis(phenylchalcogeno)niobocenes with nitrosyl carbonyl mercury complexes.83 Cp2Nb(EPh)2Fe(NO)CO Cp2Nb(EPh)2+Hg[Fe(CO)3NO] 60 (70% ± 80%) E=S, Se, Te.Niobium-and-cobalt-containing complexes 61 were prepared according to the following scheme: Cp2Nb(EPh)2+Hg[Co(CO)4] Cp2 Nb(EPh)2Co(CO)2 61 E=S, Se, Te. The rate of formation of the complexes 60 and 61 increases in the sequence S<Se<Te, which was attributed 83 to the increase in the nucleophilicity of chalcogens in the same series. The complexes 60 and 61 are more stable than the initial cyclopentadienyl derivatives of niobium. They are stable in DMSO solutions under an inert atmosphere but rapidly decom- pose in air.83 The IR spectra of the complexes 60 exhibit bands at 1610 (NO) and 1840 cm71 (C=O). The complexes 61 are characterised by 421 two bands, 1850 and 1910 cm71. The 1H NMR spectra of the compounds 61 (E=Se, Te) { contain three singlets for the cyclo- pentadienyl protons, which was explained by cis ± trans isomer- ism.The 1H NMR spectra of the complexes 60 exhibit six singlets due to the cyclopentadienyl protons, which correspond to two cis- (60a,b) and one trans-isomer (60c).83 Ph Ph E E CO Cp NO Cp Ph Fe Nb Nb Ph Fe NO Cp CO Cp E E 60b 60a Ph E Cp NO Fe Nb Cp CO EPh 60c Apart from the neutral complexes described above, which contain bridging and terminal RTe groups, several anionic and cationic cyclopentadienyl carbonyl complexes with organyl tel- luride ligands have been synthesised. Thus anionic complex 62 is formed when paramagnetic monomeric complex 63b, which occurs in solution in equilibrium with diamagnetic dimer 63a, is reduced by sodium borohydride.84 [Cp(CO)2MnTePh]2 63a NaBH4 Cp(CO)2MnTePh 63b [CpMn(CO)2TePh]7 +NMe4[{Cp(CO)2Mn}2TePh]7 62 The coordination polyhedron around the tellurium atom in the complex 62 is a pseudo-tetrahedron, the Mn± Te bond length being 2.556 A.84 Anionic complex 64, containing a bridging MeTe ligand, is formed in 85% yield when complex 65 85 is made to react with methyllithium.86, 87 1.MeLi 2. (PPN)Cl [{CpMn(CO)2}3TeMe](PPN) [CpMn(CO)2]3Te 64 65 Oxidation of neutral cyclopentadienyl carbonyl complexes 63a and 66 by silver hexafluorophosphate has given rise to cationic complexes 67a 84 and 67b (yield 56%).88 In the cation 67a, the tellurium atom forms a pyramidal configuration of bonds; the Mn± Te bond length is 2.513 A.84 AgPF6 [Z5-R5C5(CO)2MnTeAr]+PF¡ [Z5-R5C5(CO)2MnTeAr]2 6 63a, 66 67a,b R=H, Ar=Ph (63a, 67a); R=Me, Ar=Mes (66, 67b).Cationic chromium complex 68 has been prepared in a nearly quantitative yield by methylation of the corresponding neutral complex with methyl trifluoromethanesulfonate followed by treatment of the reaction mixture with NH4PF6.86 { The spectrum of the sulfur analogue could not be recorded due to its low solubility.422 1. CF3SO3Me 2. NH4PF6 [CpCr(CO)3]2Te [{CpCr(CO)3}2TeMe]+PF¡¦6 68 III. Complexes with organic derivatives of dicoordinated tellurium This Section surveys data on the synthesis, structure and reactions of carbonyl and cyclopentadienyl carbonyl complexes of transi- tion metals containing as ligands organic derivatives of dicoordi- nated tellurium such as diorganyl tellurides R1TeR2, tellurium- containing heterocycles, diorganyl ditellurides RTe ¡À TeR and organoelement compounds with C¡À Te ¡À E and E ¡À Te ¡ÀE frag- ments (E is a Group IV or V element).For tellurols RTeH, tellurenyl halides RTeHal and tellurocyanates RTeCN, com- plexes of this type are unknown. 1. Complexes of diorganyl tellurides Several types of complexes of metal carbonyls and cyclopenta- dienyl carbonyls with acyclic and cyclic (tellurium-containing heterocycles) diorganyl tellurides have been obtained. Thus heating or UV irradiation of solutions of diorganyl tellurides and hexacarbonyls of Group VI metals [or complexes (CO)5M.THF, which can be readily prepared from them] results in the formation of pentacarbonyl complexes 69a,89 70a,b 90 and 69b.91 THF R2Te+M(CO)6 (CO)5M. TeR2 7CO 69a,b M=Cr, R=2-Me2NCH2C6H4 (a); M=W, R=2,4,4-tris(trimethylsilyl)cyclopenta-2,5-dien-1-yl (b). W(CO)5 N N N W(CO)6 R R R Te Te Te 70a,b W(CO)5 R=Me (a), Ph (b). The structure of the complexes 69a,b was proved by X-ray diffraction analysis.90, 91 However, in another study,92 the product of reaction between chromium hexacarbonyl and dimethyl tel- luride was identified as [Cr(CO)5]2TeMe2 (IR spectroscopy and elemental analysis data). The results obtained by 1HNMRspectroscopy indicate that in solutions of the compounds 70a,b [in C6D6 or (CD3)2CO], an equilibrium between Te- and N-centered complexes exists, as in the case of metal carbonyl complexes of benzothiazoles and benzoselenazoles.93 When examining the formation of benzotellurazole complexes of the type 70, one can follow the influence of the nature of the metal carbonyl on the structure of the final product, which has been repeatedly noted for other ligands.Thus the reaction of 2-methylbenzotellurazole with Fe3(CO)12 in boiling toluene gives rise to a mixture, formed in a low yield (6%¡À13%), of detellurated complexes 71 and 72 and tellurium-containing cluster 73.94 N Me+Fe3(CO)12 Te Me C Fe(CO)3 N N Fe Fe(CO)3+ Fe Me+Te2Fe3(CO)9 73 Fe(CO)3 (CO)3 O (CO)2CO C 72 71 I D Sadekov, A I Uraev, A D Garnovskii The reaction of benzoisotellurazole with Fe3(CO)12 is also accompanied by the destruction of the heterocycle and gives complex 74 and the cluster 73.94 +73 Fe(CO)3 N N Fe Te Te N+Fe3(CO)12 Te Fe(CO)3 74 (15%) Whereas in 69a,b and 70a,b the metal is coordinated only by the tellurium atom, the reaction of tellurophene, possessing heteroaromatic properties, with the chromium acetonitrile com- plex affords a different type of product, namely, p-complex 75 (yield 80%).95 Similar complexes are formed with thiophene and selenophene.95 Bu2O, 50 ¡À 60 8C Cr(CO)3 Cr(CO)3(MeCN)3+ Te 75 Te Recently,96, 97 a new approach to the synthesis of the com- plexes 69c ¡À e has been proposed, which is based on the reaction of metal hexacarbonyls with diaryl telluroxides in THF.Presum- ably,97 the reaction occurs via a four-membered transition state. C O M (CO)5 M(CO)6+Ar2TeO 7CO2 (CO)5M. TeAr2 69c ¡À e Ar2Te O Ref. Yield (%) M Ar Compound 69 96 97 97 50 42 29 Cr Mo W 4-MeOC6H4 4-MeOC6H4 4-MeOC6H4 cde It should be noted that in the presence of pyridine, chromium hexacarbonyl reacts with di(4-methoxyphenyl) telluroxide to give the corresponding telluride and a complex of pyridine with chromium pentacarbonyl.96 CHCl3 Cr(CO)6+(4-MeOC6H4)2TeO+Py 7CO2 (CO)5Cr . Py+(4-MeOC6H4)2Te The reaction of hexacarbonyls of Group VI metals and iron pentacarbonyl with organoelement oxides RnEO (n=2, 3; E=Se, Te, As, Sb) is a fairly general process. It has been reported 96 that the reactions of chromium hexacarbonyl with the oxides obey second-order kinetics and that their relative rates decrease in the sequence TeO>SeO>SbO>AsO.Meanwhile, the least basic compounds among Group V and VI element oxides D triphenylphosphine oxide and diphenyl sulfoxide D do not react with hexacarbonyls of chromium group metals at all. It has been shown 97 that the rates of reactions of di(4-methox- yphenyl) telluroxide with metal hexacarbonyls M(CO)6 decrease in the seriesW>Mo>Cr (the relative rates are 9.4, 4.4 and 1.0, respectively). It follows from the above data that in order to prepare complexes of the type 69, metal carbonyls should be made to react with diorganyl tellurides . Diorganyl tellurides can arise as intermediates in the reactions of metal hexacarbonyls with other organotellurium compounds and then react with excess metal hexacarbonyls, present in the reaction mixture, to give complexes of the type 69.This idea has been confirmed by the synthesis of complexes 76a,b, containing a dimethyl telluride ligand, from an organic compound containing tetracoordinated tellurium (dime- thyltellurium diiodide) and salts 77.98 However, these complexes exist only in solution and rapidly decompose when the solvent is removed in vacuo.98Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands [M(CO)n]2TeMe2 Na[M(CO)n]+Me2TeI2 77 76a,b M=Co, n=4 (a);M=Mn, n=5 (b).The vigorous reaction of octacarbonyldicobalt with dimethyl- tellurium diiodide affords polymeric complex 78.98 However, the mechanism of this reaction is unknown. Co2(CO)8+ Me2TeI2 7CO [Co(CO)5TeMe2]n 78 It is of interest that the photoinduced reaction ofMe2TeI2 with decacarbonyl-dimanganese and -dirhenium affords iodo carbon- yls 79 in 60%¡À 80% yields.98 hn M2(CO)10+ Me2TeI2 7CO IM(CO)4TeMe2 79a,b M=Mn (a), Re (b). The chromium complex 69f has been synthesised using a specific method for the synthesis of diethyl telluride derivatives, namely, alkylation of anionic complex 80 with triethyloxonium tetrafluoroborate.99 1. Na2Te, EtOH, D 2. (PPN)Cl Et3O+BF¡¦4Cr(CO)6 (PPN)[Cr(CO)5TeH] 7(PPN)BF4 80 (CO)5CrTeEt2 69f (>60%) Another example in which alkylation has been used to prepare telluride complexes D the synthesis of the complex 53 72, 73 D is described in Section II.3.On treatment with iodine, the complex 53 is oxidised to give halo carbonyl complex 81a (yield 94%), which was characterised by spectroscopy (1H NMR, 13C NMR, IR, UV).73 I2 (MeTePh)Fe(CO)3I2 81a (MeTePh)Fe(CO)4 53 The diphenyl analogue of the compound 81a D complex 81b Dhas been synthesised from diphenyltellurium diiodide. This complex reacts with (CO)5W.THF to give the diphenyl telluride derivative 69g.73 The mechanism of formation of this product and the structures of other reaction products have not yet been elucidated. THF W(CO)5 .THF Fe(CO)5+Ph2TeI2 7CO Ph2TeFe(CO)3I2 81b (CO)5W.TePh2 69g (62%) The molecular and crystal structures of the complexes 69a ¡À c,g and 70b were studied by X-ray diffraction analysis. The most important bond lengths and bond angles are listed in Table 7. The coordination polyhedron about the metal atom in all the studied complexes is a distorted octahedron; the tellurium atom normally has a pyramidal environment. The secondary intra- molecular coordination N?Te, occurring in the free ligand, is retained in the structure 69a.89 The N¡À Te distance (3.096 A) in 69a coincides almost completely with the average N¡À Te distance in the ligand. In the complex 70b, formed by a bidentate tellurium-contain- ing heterocycle, 2-phenylbenzotellurazole, with tungsten hexacar- bonyl, of the two possible sites of coordination, namely, the `hard' nitrogen atom and the `soft' tellurium atom, it is the latter that is involved in coordination with tungsten, in conformity with the principle of hard and soft acid and bases (see Table 7).100 In all probability, the complex 70b is the first structurally characterised 423 Table 7.Bond lengths and bond angles in the complexes (CO)5M. TeR2 . Bond lengths /A C7Te7C Ref. bond angle /deg Com- M R po- und M7Te C7Te a 89 97.2 2.133 2.667 69a Cr 2-Me2NCH2C6H4 Me3Si 91 98.5 2.139 2.839 69b W Me3Si SiMe3 97 73 95.5 94.8 2.131 2.125 2.684 2.809 69c Cr 4-MeOC6H4 69g W Ph N 90 78.0 2.17 2.809 70b W Ph Te a The average bond lengths and angles are given.complex of a benzochalcogenazole with chalcogen ¡À metal coor- dination. The reactions of binuclear and trinuclear metal carbonyls with diorganyl tellurides give different products, depending on the nature of the substrate. The reaction of Mn2(CO)10 with diphenyl telluride, accompanied by rupture of the C¡À Te bond and yielding the complex 24f with bridging PhTe ligands,48 has been described above (see Section II.3). The reaction of the trinuclear complex Fe3(CO)12 with the same telluride gives complex 82 in a moderate yield (25%).48 When dodecacarbonyltriiron is made to react with telluro- phene in benzene, a mixture of FeTe, ferrole 83 and complex 84 is formed.101 Based on the mass spectra [molecular ion peak with m/z =459 and peaks due to ions with m/z =459728n (n=076)] and the IR spectra (absorption bands for three terminal carbonyl groups in the range 1995 ¡À 2064 cm71), a metallocene structure was ascribed to 84.101 Tetraphenyltelluro- phene reacts with Fe3(CO)12 to give complex 85 in a low yield.102 [Fe(CO)4]3+ Ph2Te Fe(CO)4TePh2 82 Fe(CO)3 [Fe(CO)4]3 FeTe+ Fe(CO)3 + Te Fe(CO)3 Te Fe(CO)3 84 83 Ph Ph Ph Ph [Fe(CO)4]3 Fe(CO)3 Ph Ph Ph Ph Te Te 85 According to the data of McWhinnie et al.,101 when condi- tions of the reaction of tellurophene with Fe3(CO)12 are changed (refluxing in heptane, 2.5 h), this reaction yields ferrole 83 (45%) and FeTe, which appears to arise upon elimination of CO and Fe(CO)5 molecules from ferrole 83.Under the same conditions, dibenzotellurophene is converted into dibenzoferrole in 28% yield, FeTe being formed simulta- neously.101 C7H16 , D, 2.5 h Fe(CO)3+FeTe +Fe3(CO)12 Fe(CO)3 Te The reaction of telluroindan follows a different route.Reflux- ing a mixture of telluroindan with Fe3(CO)12 gives rise to new424 complex 86 (yield 15%) and the cluster 73. The structure of the compound 86 was confirmed by X-ray diffraction analysis.101 C7H16 , D, 3 h Te+Fe3(CO)12 (CO)3Fe +Te2Fe3(CO)9 73 86 The reaction of Co2(CO)8 with diphenyl telluride is accom- panied by rupture of both C± Te bonds, binuclear complex 87 being the final reaction product.48 Heating of a mixture of diethyl telluride, Co2(CO)8 and Fe3(CO)12 affords cluster 88 in 30% yield.103 PhH , D Co2(CO)8+Ph2Te Co2Te(CO)5 87 C6H14 , D Et2Te+Fe3(CO)12+Co2(CO)8 FeCo2(m3-Te)(CO)9 88 Aryl ethynyl telluride 89 reacts with octacarbonyldicobalt in a different way.104 The reaction carried out in a 1 : 1 pentane ± toluene mixture at 25 8C gives complex 90 in 45% yield.Accord- ing to X-ray diffraction data,104 in this complex, as in other alkynyl complexes, the cobalt atom is coordinated to the multi- ple-bond carbon atoms rather than to tellurium.(CO)3 Co C C ArTe SiMe3 7CO ArTeC CSiMe3+Co2(CO)8 89 Co (CO)3 90 Ar=2,4,6-Me3C6H2 . The lengths of the Co ± Co, C± C, Ar ± Te and C± Te bonds in the complex 90 are 2.465, 1.31, 2.311 and 2.066 A, respectively.104 The reaction of nickel carbonyl with diphenyl telluride, unlike the reactions of manganese and cobalt carbonyls, is accompanied by complete dearylation and decarbonylation, tellurium and nickel being the only reaction products.48 It is noteworthy that nickel carbonyl reacts with derivatives of tetracoordinated tellu- rium, ArTeCl3 and Ar2TeCl2, to give (after hydrolysis of the reaction mixture) carboxylic acids.105 The structure of complexes resulting from the reaction of the osmium carbonyl cluster 91 with five-membered aromatic hetero- cycles is determined by the nature of the heteroatom.106, 107 Thus when the complex 91 reacts with furan 107 or thiophene,106 the corresponding hydrido furyl (92a) or hydrido thienyl (92b) com- plex is obtained.The complexes 92a,b have similar spectral characteristics.According to the X-ray diffraction data obtained for 92a,107 the heterocyclic fragments in these compounds act as m,Z1,Z2-vinylic ligands. The 1H NMR spectra imply that the thiophene complex 92b exists in solution at 755 8C as a mixture of exo- and endo-isomers (ratio 4 : 1), which undergo fast inter- conversion at room temperature.106 The reactions of the complex 91 with selenophene and tellurophene in cyclohexane give rise to complexes 93a,b.106 The spectral characteristics of these compounds are close to each other but sharply differ from those of the complexes 92. The X-ray diffraction data obtained for the selenophene complex 93a indi- cate that it has resulted from insertion of the Os atom into the Se ± C rather than C±H bond.106E Os3(CO)10(MeCN)2+ 91 I D Sadekov, A I Uraev, A D Garnovskii E E E=O, S Os(CO)3 (CO)4Os Os(CO)3 (CO)4Os H Os (CO)3 endo-92a,b Os H (CO)3 exo-92a,b (CO)3 Os E=Se, Te (CO)4Os EOs 93a,b (CO)3 E=O(92a, 46%), S (92b, 57%), Se (93a, 20%), Te (93b, 35%).Carbonyl-containing complexes of various types are also formed in the reactions of diorganyl tellurides with alkyl and acyl manganese carbonyls and manganese, rhenium, ruthenium, rhodium and iridium halo carbonyls. The reactions of pentacarbonyl(methyl)manganese 94a with dialkyl tellurides in THF at room temperature occur in the same way as reactions with other Lewis bases, giving rise to (acetyl)te- tracarbonyl(dialkyl telluride)manganeses 95a ± c in yields of more than 80%.108 The reaction with diisopropyl telluride occurs very slowly; the complex 95c has not been isolated in an analytically pure state.When dialkyl tellurides are made to react with benzyl(pentacarbonyl)manganese 94b, the complexes 95d,e are formed together with substantial amounts of decarbonylated products 96. Thus the reaction of 94b with dimethyl telluride affords an unseparable mixture consisting of the initial carbonyl (10%), the complex 95d (80%) and the corresponding decarbony- lated product 96a (10%). The reaction of 94b with diethyl telluride in boiling pentane yields the decarbonylated complex 96b as the only product.108 CO O TeR22 Mn CO R1 C R1Mn(CO)5+ R22 Te D 7CO 94a,b OC CO 95a ± e 2 R1Mn(CO)4TeR2 96a,b 96b 96a 95e 95d 95c 95b 94b 95a 94a Com- pound Bn Me Bn Et Bn Me Bn Et Bn Me Me Pri Me Et Me Me R1 R2 1H and 125Te NMR spectra (singlets in the region of 153.0 to 688.6 ppm) and IR spectra indicate that the complexes 95a ± e exist in solutions as single isomers with cis-arrangement of the acetyl and dialkyl telluride ligands.108 The complexes 95a ± e decompose when distilled in vacuo (0.5 mm Hg, 40 ± 70 8C), the reaction pathway depending on the nature of the substituent at the tellurium atom.108 R=Pri[Mn(CO)4(m-TePri)]2 R=Me, Et MeCOMn(CO)4TeR2 95a ± c R2Te+MeMn(CO)5 The complexes 95f, g have been synthesised by the reaction of pentacarbonyl(propionyl)manganese with dialkyl tellurides.108 EtCOMn(CO)5+ R2Te 7CO EtCOMn(CO)4TeR2 95f,g (>70%) R=Me (f), Et (g).The reductive thermolysis of the complexes 95b,c,g under a hydrogen atmosphere at 200 8C yields films containing manga- nese and tellurium.108Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands The reactions of manganese 48, 108 ± 111 and rhenium 111 ± 113 halo carbonyls with dialkyl and diaryl tellurides have been studied. Dimethyl and diethyl tellurides react with bromo(pentacarbonyl)manganese 97a giving rise to dialkyl tel- luride complexes 79c,d, their yields exceeding 80%.108 The reac- tion of pentacarbonylchloromanganese 97b with dibutyl telluride affords the binuclear complex [(CO)4MnCl]2.48 When the complex 79d is treated with a second mole of diethyl telluride, it is converted into bis(diethyl telluride) complex 98a.108 Compounds of this type are formed as the only products in reactions of diphenyl telluride 48, 109, 110 and phenoxatellurine 48 with manga- nese 48 and rhenium 111 halo carbonyls.The complex 98b also arises in the reaction of pentacarbonylchlororhenium 97c with dibutyl telluride,112 i.e. the rhenium and manganese compounds react in different ways. Evidently, this implies that the rhenium complexes 98 are more stable than the corresponding manganese analogues. Rhenium bis(dialkyl telluride) complexes 98c,d were prepared by the reaction of di-m-chlorobis[tetracarbonyl- rhenium(I)] 99b with diethyl telluride under drastic conditions 113 and by the reaction of complex 100 with dimethyl telluride at room temperature.111 R2Te XM(CO)5 97a XM(CO)4TeR2 79c ± e (CO)4M(m-X)2M(CO)4 99a,b Et2Te 79d BrMn(CO)3(TeEt2)2 98a M=Mn [(CO)4MnCl]2 Bu2Te M=Re ClRe(CO)3(TeBu2)2 ClM(CO)5 97b,c 98b Et2Te 79f CO PhMe Me2Te XRe(CO)3(TeR2)2 98c, d (CO)4Re(m-X)2Re(CO)4 99b Re2X2(CO)6(THF)2 100X R M Compound Mn Re Mn Mn Mn Re Re Re Re Br Cl Cl Br Br Cl Br Cl Cl 97a, 99a 97b, 99b 97c 79c 79d, 98a 79e 79f, 98c 98b 98d Me Et Me Me Br Et The reaction of the binuclear rhenium complex 99b with dimethyl telluride 113 gives the complexes 79e, while the reaction of the similar manganese complex 99a with dimethyl telluride performed under relatively mild conditions affords the complex 79c.When carbon monoxide is passed through a toluene solution of the complex 98c at room temperature, the complex 79f is produced.111 The rhenium complexes 79e,f are much more stable than the manganese analogues. Thus it has been found 108 that the man- ganese complex 79c exists only in solution, the attempted isolation resulting in its complete decomposition. Study of the dimethyl chalcogenide ligand exchange in the rhenium complexes 98 111 shows that the equilibrium in these reactions is shifted towards the formation of complexes contain- ing heavier chalcogen atoms.111 425 ReBr(CO)3(E1Me2)2+E2Me2 ReBr(CO)3(E2Me2)2+E1Me2 E1=S, E2=Se; E1=Se, E2=Te.A new pathway to the iodo carbonyl complexes 79a,b is based on the reaction of decacarbonyl-dimanganese and -dirhenium with dimethyltellurium diiodide.98 The compounds 79a,b were isolated by sublimation in vacuo at 60 8C. The reactions of other dialkyltellurium dihalides with the manganese and rhenium car- bonyls have not been studied. hn IM(CO)4TeMe2 M2(CO)10+ Me2TeI2 79a,b M =Mn (a, 80%), Re (b, 60%). Passing NO through a benzene solution of the complex 98e gives the nitrosyl complex 101, which, however, has not been isolated in an analytically pure state.48 7CO ClMn(CO)3(TePh2)2+NO 98e Mn(NO)3TePh2+Ph2Te+Ph2TeCl2 101 The reaction of the manganese complex 98f with nitrogen- containing ligands results in the replacement of either or both diphenyl telluride ligands, depending on the nature of the reagent used.110 L1 [Mn(CO)3L13 ]Br L2 [Mn2(CO)6L23 ]Br2 BrMn(CO)3(TePh2)2 L3 [Mn2(CO)6L32 ]Br2 98f L4 [Mn(CO)3(TePh2)L4]Br O , , L1= ; HN HN HN L2=HN NH2 ; NH, H2N , NH NH L3= NH NH2 , N N NH N CH .; L4=H2N Both Rh(I)- and Rh(III)-containing complexes have been synthesised from rhodium halo carbonyls. Compounds of the former type (102a ± d) were prepared by the reaction of di-m- chlorobis[dicarbonylrhodium(I)] 103 with diorganyl tellur- ides,69, 114 including tellurium-containing heterocycles,115 at room temperature in pentane, hexane or toluene. According to IR spectroscopy, after the isolation of the complex with tellurox- anthene 102c (nCO 1967 cm71), the mother liquor contains the intermediate dicarbonyl complex 104 (nCO 2080, 2033 cm71), which has resulted from cleavage of the chloride bridges in the complex 103 by the tellurium-containing ligand.115 CO CO R2Te R2Te R2Te Rh Rh 7CO [Rh(CO)2Cl]2+R2Te 103 CO Cl TeR2 Cl102a ± d 104 R=Et (a), Ph (b); NEt (d).(c), R2Te= Me Me Te Te426 The reactions of [Rh(CO)2Cl]2 with tellurium-containing ligands generally resemble its reactions with triorganylphos- phines. However, tellurium-containing ligands, in particular, telluroxanthene and 3,7-dimethyl-N-ethylphenotellurazine, are weaker donors than phosphines. Indeed, Rh(acac)(CO)2 readily reacts with triphenylphosphine at room temperature to give Rh(acac)(CO)PPh3116, 117 but does not react with telluroxanthene even at 80 8C.115 However, telluroxanthene and 3,7-dimethyl-N- ethylphenotellurazine replace the alkene ligand in the rhodium complex 105, giving rise to complexes 106 with trans-arrangement of organic ligands.115 Rh(Oxq)(C8H14)CO+ R2Te Rh(Oxq)(CO)TeR2 7C8H14 105 106 Oxq� , C8H14�cyclooctene.N O7 The weaker donor properties of diorganyl tellurides compared to those of tertiary phosphines account for the fact that telluride ligands in the complexes 102c,d and 106 are readily substituted by triphenyl- and tributylphosphines.115 The rhodium carbonyl complexes 102a,b undergo fast oxida- tive addition of halogens and thiocyanogen to give octahedral Rh(III) complexes 107a ± c,114 107d ± f 69 and 107g,h.115 Rh(CO)(TeR2)2ClX2 Rh(CO)(TeR2)2Cl+X2 107a ± h 102a,b R=Et: X=Cl (a), Br (b), I(c); R =Ph: X =Br (d), I (e), SCN (f); NEt (h); X=I.(g), R2Te= Me Me Te Te The complex 107a was also prepared by the reaction of the diethyl ditelluride complex 102a with hydrogen chloride;114 102a also adds several other reagents to give Rh(III) complexes 108a,b.114 It is noteworthy that the diphenyl telluride analogue 102b, unlike the compound 102a, does not react with methyl iodide.69 HCl 7H2 Rh(CO)(TeEt2)2Cl3 107a XY Rh(CO)(TeEt2)2Cl+HCl 102a Rh(CO)(TeEt2)2ClXY 108a,b XY=MeI (108a), PhSO2Cl (108b). The complex 108a easily reacts with carbon monoxide; the insertion of CO into the Rh ±Me bond gives rise to a rhodium(III) acyl derivative.114 Rh(CO)(TeEt2)2(COMe)ICl Rh(CO)(TeEt2)2MeICl+CO 108a On passing from the planar rhodium(I) complexes 102 to the octahedral rhodium(III) complexes 107 and 108, the nCO value increases by*100 cm71.69, 114, 115 The CO stretching vibration frequencies in the spectra of the complexes 102c (1956, 1946 cm71), 102d (1962 cm71), and the analogue of 102d containing phenothiazine as a ligand (1966 cm71) 115 indicate that the p-dative transfer of electron density from the central atom to a carbonyl group is less significant in the complex with phenothiazine than in the com- plexes with telluroxanthene and tellurazine.A similar effect, although less pronounced, noted for diethyl chalogenide com- plexes 102,114 may be due to the enhancement of the s-donor properties of chalcogen atoms in the series S<Se<Te.The substantial increase in nCO following the transition from the complexes 102c,d to the phenothiazine complex attests that the Rh(I) ± S bond is weaker than Rh(I) ± Te.115 I D Sadekov, A I Uraev, A D Garnovskii The complex 102a not only enters into oxidative addition reactions but also reacts with tetracyanoethylene and fumaroni- trile to give stable 1 : 1 adducts 109 in almost quantitative yields.118 In these complexes, the metal is coordinated to the double bond of the cyanoalkene. Similar complexes with acrylonitrile and cinna- monitrile are not formed.118 CN R PhH, C5H12 R(NC)C C(CN)R Rh(CO)(TeEt2)2Cl+ 102a 109 Rh(CO)(TeEt)2Cl R NC R=H, CN. The rhodium(III) complexes 107 have also been obtained by replacement of one tellurium-containing ligand in complexes 110 by CO, which is attained by passing carbon monoxide through solutions of the complexes 110 in chloroform or toluene.69, 119 Rh(CO)(TeR2)2Cl3 (R2Te)3RhCl3+ CO 7R2Te 107 110 Cl , .R2Te=Ph2Te, Te Te Data on halo carbonyl complexes of iridium with diorganyl tellurides can scarcely be found in the literature. The only representative of this class of compound, complex 111, has been synthesised by the reaction of polymeric iridium chloro carbonyl with diethyl telluride.120 [Ir(CO)3Cl]n+ Et2Te 7CO ClIr(CO)2TeEt2 111 Only few representatives of complexes of diorganyl tellurides with iron 48 and ruthenium 121, 122 halo carbonyls are known.Iron diphenyl telluride complexes 112 were prepared by substituting diphenyl telluride for one carbonyl group in tetracarbonyldiha- loirons. The structure of these products was not specially studied; they were identified by analogy with other complexes.48 TePh2X OC Fe Fe(CO)4X2+Ph2Te 7CO OC X CO 112 X=Br, I. A mixed iron carbonyl nitrosyl complex was synthesised in a similar way.48 Et2O, 20 ± 25 8C Fe(NO)2(CO)TePh2 Fe(NO)2(CO)2+Ph2Te 7CO Complexes 113a ± d, containing two telluride ligands, were prepared by the reactions of polymeric ruthenium halo carbonyls with diorganyl tellurides.121, 122 These reactions also give small amounts of tritelluride complexes 114a ± d.121 They were not isolated in a pure state and their existence in solution was confirmed by IR spectra.PhH, 100 8C [Ru(CO)2X2]n+R2Te Ru(CO)2(TeR2)2X2+Ru(CO)(TeR2)3X2 114a ± d 113a ± d X=Br: R=Bu (a), Ph (b); X=I: R =Bu (c), Ph (d). The mixture of the compounds 113e and 114e, resulting from the reaction of a CO-saturated alcoholic solution of RhCl3 .H2O with Ph2Te, was successfully separated by fractional crystallisa- tion.121 RuCl3 .H2O+CO+Ph2TeSynthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands Ru(CO)2(TePh2)2Cl2+Ru(CO)(TePh2)3Cl2 114e (25%) 113e (40%) The reactions of complexes of the type 113 with halogens lead to the rupture of the Te ±Ru bonds, unlike similar reactions of analogous complexes containing nitrogen and phosphorus ligands, in which less active halogens are replaced by more active ones.121 Ru(CO)2X2+R2TeX2+I2 Ru(CO)2(TeR2)2I2+X2 113c,d R=Bu (c), Ph (d); X=Cl, Br.Of complexes of metal cyclopentadienyl carbonyls with dio- rganyl tellurides, compounds of Group V elements, vana- dium 123, 124 and niobium,125 have been described. These complexes, having the composition CpM(CO)3TeR2 115, are prepared by exposure of a solution of tetracarbonyl(cyclopenta- dienyl)-vanadium or -niobium and a large excess of a diorganyl telluride to UV or visible light. Complexes of the type 115 with other diorganyl chalcogenides have also been synthesised.THF CpM(CO)4+ ER2 7CO CpM(CO)3ER2 115 Ref. M Compound 115 R2E abcdef 124 124 124 123 123 123 VVVVVV Me2S Me2Se Me2Te Ph2S Ph2Se Ph2Te g 125 Nb S h 125 Nb Se i 125 Nb Te When other solvents (toluene, heptane) are used instead of THF, the complexes 115g ± i substantially decompose and their yields diminish.125 The low thermal stability of dimethyl chalco- genide complexes is the obstacle preventing large-scale prepara- tion of these compounds.124 Niobium complexes with tetra- hydrochalcogenophenes are the most stable; they have been isolated as individual crystalline compounds. However, they are still fairly sensitive to air both in the crystalline state and especially in solutions.125 All complexes of the type 115 tend to decompose in solution to give diorganyl chalcogenides and CpM(CO)4.124, 125 The stability of the diorganyl chalcogenide complexes with the same metal increases in the series S<Se<Te.123 ± 125 This stability series was attributed 125 to the increase in the s-donor capacity on passing to heavier chalcogens.This is confirmed by IR spectra, which display a decrease in nCO in the order S>Se>Te both for the vanadium 124 and niobium 125 diorganyl chalcogenide complexes 115. The reaction of formation of the complexes 115a ± c is reversible, the equilibrium constant increasing in the order S<Se<Te (4.561073, 1.861072 and 1071, respectively). The dimethyl sulfide ligand in the complex 115a can be displaced by dimethyl selenide and dimethyl telluride ligands; the equili- brium constants for these reactions are 4.0 and 1.3610 2, respec- tively.124 CpV(CO)3EMe2+Me2S 115b,c CpV(CO)3SMe2+Me2E 115a 427 The reactions of tetracarbonyl(cyclopentadienyl)-vanadium and -niobium with diorganyl chalcogenides can occur by either dissociative (pathway a) or associative (pathway b) mechanism.125 a R2E, fast [CpM(CO)3] 7CO CpM(CO)3ER2 115 CpM(CO)4 115 [CpM(CO)4ER2] fast 7CO b R2E Kinetic measurements 125 indicate that the formation of the niobium complexes 115g ± i occurs according to both mechanisms.The tetrahydrochalcogenophene ligands in the complexes 115g ± i are easily replaced by various phosphine ligands.125 The reaction rate increases in the series Te<Se<S, thus confirming the fact that on passing to heavier chalcogens, the complex stability increases, which has been repeatedly noted above.CpNb(CO)3PR23 CpNb(CO)3ER12 +R23 P 7R12 E 115g ± i R2=Bu, Ph, OPh. In addition to the neutral complexes of diorganyl tellurides with metal cyclopentadienyl carbonyls 115 described above, cationic cyclopentadienyl carbonyl complexes of iron 126, 127 and molybdenum 128 and anionic cyclopentadienyl carbonyl com- plexes of vanadium 129, 130 have been synthesised. When dimethyl chalcogenides react with cationic complexes 116a ± c under mild conditions, they substitute the THF molecule, which results in the formation of cationic complexes 117a ± c, their yields exceeding 90%.126 On UV irradiation of solutions contain- ing the complexes 117a ± c and dimethyl chalcogenides, the latter replace a CO molecule to give cationic monocarbonyl complexes 118a ± c (yields 47% ± 82%).126 The cationic complexes 117a ± c and 118a ± c react with salts containing nucleophilic anions (CN7, I7) to give neutral complexes 119, 120a ± c, in which X is cyano group or iodine.126 CH2Cl2 4 [CpFe(CO)2THF]+BF¡4 +Me2E [CpFe(CO)2EMe2]+BF¡ 117a ± c 116 a CpFe(CO)2X 119 b a 4 [CpFe(CO)EMe2]X 120a ± c [CpFe(CO)(EMe2)2]+BF¡ 118a ± c E = S (a), Se (b), Te (c); X=CN,M=K; X=I, M=Li; (a) MX, CH2Cl2,*20 8C, Ar; (b) Me2E, CH2Cl2 , hn.Exposure of solutions of the complexes 117a,b and dimethyl chalcogenides containing a heavier chalcogen to UV radiation leads to the displacement of a CO molecule and the formation of compounds 121, which incorporate two different dimethyl chal- cogenide ligands.127 [CpFe(CO)2E1Me2]+BF¡4 +E2Me2 hn 7CO 117a,b 4 [CpFe(CO)E1Me2(E2Me2)]+BF¡ 121 E1=S: E2=Se, Te; E1=Se, E2=Te.Treatment of the complex 117c with organoelement com- pounds of Group V elements results in the replacement of one or more CO ligands giving rise to compounds 122 ± 124.127428 MR¡¦34 7CO [CpFe(CO)TeMe2(MR3)]+BF¡¦ 122 2PR234 [CpFeTeMe2(PR23 )2]+BF¡¦472CO [CpFe(CO)2TeMe2]+BF¡¦ 117c 123 + CpFeTeMe2 BF¡¦ PPh2 Ph2P 4 X 7CO (CH2)n 124 MR13 =PMe3 , P(NMe2)3 , PPh3 , AsPh3 , SbPh3; R2=Me, OMe, OPh; X=Ph2P(CH2)nPPh2 .The molecular and crystal structures of the complex 117c have been studied by X-ray diffraction analysis.131 In this compound, the iron atom is linked to two terminal carbonyl groups, one TeMe2 ligand and one cyclopentadienyl ring, which form a pseudo-octahedral configuration. No tight contacts between the [CpFe(CO)2TeMe2]+ cations and the BF¡¦4 anions were found in the crystals. The Fe ¡À Te distance is 2.533A. The dimethyl telluride ligand occupies the anti-position in relation to the cyclopenta- dienyl ring, which minimises steric contacts. The Te ¡ÀC bond length (2.107 A) and the C¡À Te ¡ÀC bond angle (93.18) fall within the ranges of these values normal for organic tellurides. The extended Hu�� ckel quantum-chemical calculations 131 for the [CpFe(CO)2EMe2]+ cations confirmed the conclusion, based on the study of reactions of the cations 117, that the stability of the Fe ¡À E bond increases in the series S<Se Te.A different approach has been used to prepare molybdenum cyclopentadienyl carbonyl complexes 125.128 These monocarbo- nyl complexes were synthesised in 39%¡À 91% yields by treatment of tricarbonyl(cyclopentadienyl)iodomolybdenum with silver tet- rafluoroborate in the presence of the corresponding dimethyl chalcogenides (at a molar ratio of the reactants of 1 : 1 : 1). CH2Cl2 CpMo(CO)3I+AgBF4+EMe2 4 7AgI [CpMo(CO)3EMe2]+BF¡¦ 125 E=S, Se, Te. Dicarbonyl molybdenum-containing cations 126 were obtained in 86%¡À 93% yields by oxidation of the dimer of dicarbonyl(cyclopentadienyl)molybdenum with ferrocenium tet- rafluoroborate in the presence of dimethyl chalcogenides.128 These compounds are formed as mixtures of cis- and trans-isomers the latter being formed in substantially larger amount.128 CH2Cl2 [CpMo(CO)2]2+EMe2+[Cp2Fe]+BF¡¦47Cp2Fe 4 [CpMo(CO)2(EMe2)2]+BF¡¦ 126 E=S, Se, Te.Anionic vanadium pentacarbonyl complexes 127, containing diorganyl chalcogenides as ligands, have been synthesised by ligand exchange, namely, by the replacement of the THF molecule in the anionic complexes of vanadium 128 by diorganyl chalcoge- nide.129, 130 7THF Et4N+[V(CO)5THF]7+R1R2E 128 Et4N+[V(CO)5ER1R2]7 127a ¡À d Ref. R2 R1 Compound 127 E 130 129, 130 129, 130 129, 130 Ph Ph Ph Ph Ph Ph Ph Me abcd SSe Te S Among the synthesised compounds, only the tellurium com- plex 127c is stable at room temperature.130 The 51VNMRspectra I D Sadekov, A I Uraev, A D Garnovskii of the complexes 127a ¡À c demonstrate that the metal atom becomes more shielded with an increase in the size of the chalcogen atom. The d51V values for E=S, Se, Te are 71618, 71658 and71753 ppm, respectively.130 2. Complexes of diorganyl ditellurides It has been shown in Section II that bi- and trinuclear iron, ruthenium and manganese carbonyls and molybdenum hexacar- bonyl react with diaryl ditellurides with cleavage of the Te ¡À Te bond to give complexes containing bridging ArTe ligands.How- ever, when diphenyl ditelluride reacts with chromium and tung- sten hexacarbonyls, the Te ¡À Te bond is retained in the resulting complexes.15, 132 Depending on the ratio of the substrates, this gives either 1 : 1 (129) 132 or 2 : 1 (130) 15 complexes. Complexes of the former type are produced in 40% ¡À60% yields on UV irradiation of tetrahydrofuran solutions of metal hexacarbonyls and diphenyl ditelluride at room temperature.132 The use of adducts of metal carbonyls with THF instead of hexacarbonyls makes it possible to prepare the complexes 129 in approximately the same yield without UV irradiation.132 THF, hn, 20 8C (CO)6M+Ph2Te2 7CO 20 8C (CO)5MTe2Ph2 129a,b (CO)5M.THF +Ph2Te2 7THF M=Cr (a), W (b). The complexes 129a,b are orange crystalline compounds, stable in air and readily soluble in organic solvents.132 The structures of complexes of both types were studied by X-ray diffraction analysis.132 The metal atom is bound only to one tellurium atom; the coordination polyhedron is a slightly distorted octahedron.The Cr ¡ÀTe and W¡À Te bond lengths are 2.679 and 2.810 A, respectively. The complexation has only a slight effect on the Te ¡ÀTe bond length in the ligand; in the complex 129a, the length of this bond is 2.737A, and in 129b, it is 2.718A(the Te ¡À Te bond length in Ph2Te2 is 2.712 A32). The 125Te NMR spectra of the complexes 129 exhibit, in conformity with the structure proposed for them, two singlet signals due to 125Te nuclei; for the chromium complex 129a, they occur at 289 and 250 ppm (with respect to Ph2Te2), while for the tungsten complex, they are observed at 340 and 34 ppm.132 Since the positions and shapes of the 125Te signals do not depend on the temperature, no exchange processes occur in solutions.It remains unknown why the difference between the chemical shifts of the two 125Te nuclei in the tungsten cox 129b (Dd=306 ppm) is much greater than that in the chromium derivative 129a (Dd=39 ppm). Heterometallic complexes 130a ¡À c have been prepared in 40%¡À 55% yields by the reaction of equimolar amounts of 129a ¡À c with (CO)5M2 . THF.15 THF (CO)5M1(Te2Ph2)+(CO)5M2 .THF 129a ¡À c [(CO)5M1][(CO)5M2]Te2Ph2 130a ¡À c M1=M2=Cr (a), W (b); M1=Cr,M2=W(c).The complexes 130a ¡À c can also be synthesised by UV irradiation of a mixture of metal hexacarbonyls and diphenyl ditelluride (taken in a molar ratio of 2 : 1) followed by evaporation of the reaction mixture in vacuo.15 The compounds 130 form dark lilac crystals, readily soluble in aromatic and chlorinated solvents. They are fairly stable in air but gradually decompose in solutions. In coordinating solvents (ace- tonitrile, THF), especially in the light or on exposure to UV radiation, they decompose to give the complexes 129.15Synthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands According to X-ray diffraction data,15 each metal atom in the complex 130a ± c is linked to one Te atom.The coordination polyhedron is a distorted octahedron. The Cr ± Te bond in 130a (2.656 A) and the W± Te bond in 130b (2.779 A) are somewhat shorter than the corresponding bonds in the complexes 129. However, the Te ± Te bond lengths in the complexes 130a ± c (2.812, 2.808 and 2.828 A, respectively) are markedly longer than the corresponding bonds in the complexes 129 and diphenyl ditelluride. The C2Te2 fragment is planar, the M(CO)5 groups occupying trans-positions with respect to the Te ± Te bonds.15 This molecular geometry implies one signal in the 125TeNMR spectrum. For the complex 130b, d125Te is 417. However, for the complexes 130a,c, no signals for the 125Te nuclei were detected; this was attributed 15 to weakening of the Te ± Te bond and its possible dissociation to give paramagnetic monomers.However, no evidence for such a homolytic bond cleavage (which could be provided, e.g., by EPR spectroscopy) is presented in this study. Thermal decomposition of the complexes 129a,b and 130a ± c was studied by differential scanning calorimetry.15 The complexes 129a,b decompose at 185 ± 192 8C to give inorganic ditellurides CrTe2 and WTe2. The binuclear complexes 130a ± c lose the carbonyl and phenyl groups at 180 ± 234 8C to yield, depending on the composition of the initial compound, CrTe, WTe and CrWTe2. Metal halo carbonyls, similarly to metal carbonyls, can react with diorganyl ditellurides both with rupture and with retention of the Te ± Te bond.The reaction of trans-chlorocarbonylbis(triphe- nylphosphine)iridium 43 with di(4-ethoxyphenyl) ditelluride, accompanied by Te ± Te bond rupture and giving the complex 44 with bridging ArTe ligands, was described above (see Section II).68 Meanwhile, the reactions of the dimeric rhenium bromo carbonyl 100 (PhMe, room temperature) or pentacarbonyl(halo)- manganese 97a,d (Et2O, Pri2O, D) 64 and -rhenium 97e,f (Pri2O, D) 67 with Ph2Te2 give the binuclear complexes 37a ± d in which the Te ± Te bond is retained. The dimeric octacarbonyl halo deriva- tives Mn2X2(CO)8 are formed as intermediate compounds in the reactions of manganese complexes.64 An intermediate with the same composition has been postulated for the reaction involving pentacarbonyl(iodo)rhenium.67 Ph2Te2 7THF [M2X2(CO)6(THF)2] 100 Ph2Te2 [M2X2(CO)6Te2Ph2] 37a ± d 7CO XM(CO)5 97a,d ± f M=Mn: X=Br (37a, 97a), I (37b, 97d); M=Re: X=Br (37c, 97e, 100), I (37d, 97f). The complexes 37a ±d are red or red-brown crystalline com- pounds.The manganese complexes 37a, b exist in air in the crystalline state only for several minutes; in solutions, they are fairly unstable thermally and photochemically unlike the rhenium complexes 37c,d. Thermal decomposition of manganese com- plexes affords the complexes 24f with bridging PhTe ligands (see Section II.3). The molecular and crystal structures of the rhenium complex 37c were studied by X-ray diffraction analysis.67 The geometry of bonds about the rhenium atoms is a pseudo-octahedron, each rhenium atom being bound to three carbonyl groups, one tellu- rium atom of the ditelluride fragment and two bridging bromine atoms.The average length of the Te ±Re bond is 2.760A, and the Te ± Te bond length is equal to 2.794A, which is somewhat longer than this bond in diphenyl ditelluride (2.712A).32 The length of the bridging Re ± Br bond is 2.642A; the Te_Br contact (3.674 A) is almost 0.5 A shorter than the sum of the van der Waals radii of the corresponding atoms (3.91A) and attests apparently to a weak interaction between them.82 The fact that the IR spectra of the rhenium bromo carbonyl 37c are similar to those of the iodo carbonyl 37d suggests that they have similar structures.67 The IR spectra of the complexes 37a ±d 429 and their sulfur and selenium analogues show a regular decrease in the nCO value following an increase in the atomic mass of the chalcogen.In the opinion of the researchers cited,67 this regularity is largely due to the increase in the s-donor capacity of chalcogens in the series S<Se<Te. The diphenyl ditelluride ligand in the complexes 37a ±d can be substituted by other ligands under mild conditions. Thus, on dissolution in THF, the diphenyl ditelluride molecule is ejected to be replaced by a THF molecule.64, 67 However, the adduct of bromo(tricarbonyl)manganese with THF, formed from the man- ganese complex 37a, is unstable even at room temperature 64 and decomposes to give a mixture of non-identified products.BrMn(CO)3(THF)2 7Ph2Te2 Mn2Br2(CO)6Te2Ph2+THF 37a Dissolution of the rhenium complex 37c and its S and Se analogues results in the equilibrium Re2Br2(CO)6(THF)2+Ph2E2 Re2Br2(CO)6E2Ph2+2THF E=S, Se, Te. The equilibrium constants diminish in the series S (3.15)>Se (7.561072)>Te (561074).64 The diphenyl ditelluride ligand in the complexes 37a ,b is replaced relatively easily at room temperature by carbon mon- oxide.64 Mn2Br2(CO)8 7Ph2Te2 Mn2Br2(CO)6Te2Ph2+2CO 37a,b For the manganese complexes of diphenyl dichalcogenides Mn2Br2(CO)6E2Ph2 (E=Se, Te), the rate of this reaction for E=Se is higher than that for E=Te, which reflects the increase in the bond strength in this series.64 A 13CNMRstudy of the exchange of diphenyl dichalcogenide ligands in the reaction of rhenium complexes 37 enriched in 13C with diphenyl dichalcogenides showed that the strength of the Re ± chalcogen bond increases on passing from Se to Te.The ligand containing a heavier chalcogen expels the ligand containing a lighter chalcogen from the complex. CDCl3 Re2Br2(13CO)6E12 Ph2+E22 Ph2 Re2Br2(13CO)6E22 Ph2+Ph2E12 E1=S, E2=Se; E1=Se, E2=Te. Complexes of metal cyclopentadienyl carbonyls with diaryl ditellurides are relatively little studied.88, 133 The dark lilac com- plex of diphenyl ditelluride with dicarbonyl(cyclopentadi-enyl)- manganese 131 has been prepared in 70% yield by the reaction of tricarbonyl(cyclopentadienyl)manganese with diphenyl ditellur- ide in heptane on exposure to UV radiation.133 This complex was also obtained by the reaction of diphenyl ditelluride with [CpMn(CO)2]THF at room temperature.133 C7H16 , hn CpMn(CO)3+Ph2Te2 7CO [CpMn(CO)2]Te2Ph2 131 [CpMn(CO)2]THF+Ph2Te2 7THF The complex 131 is fairly sensitive to air.It is readily soluble in THF, CH2Cl2 and in aromatic hydrocarbons and poorly soluble in alkanes. Its solution in THF does not exhibit signals in the EPR spectrum, and a toluene solution produces no signal in the 125Te NMR spectrum.133 The complex [MeCp(CO)2Mn]2(MesTe)2 with a similar struc- ture has been synthesised in 33% yield by the reaction of the adduct [MeCp(CO)2Mn]THF with Mes2Te2.88 It follows from X-ray diffraction data 133 that each CpMn(CO)2 fragment in complex 131 is linked to one tellurium atom.The Mn± Te bond length is 2.486A. The metal atoms are in the trans-positions430 relative to the Te ¡À Te bond, the atoms in the Mn¡ÀTe ¡À Te ¡ÀMn fragment thus formed lying in one plane. The coordination to the metal scarcely influences the Te ¡ÀC bond length (2.128A) but substantially influences the Te ¡À Te bond length, which is equal to 2.884 A in the complex 131 and 2.712 A in diphenyl telluride. The Te ¡À Te ¡ÀC bond angle also considerably changes D it is 88.48 in 131 and 97.48 and 100.38 in free Ph2Te2.32 This was interpreted 133 as being due to the additional dative interaction of the lone electron pairs of manganese with vacant d orbitals of tellurium. The Te ¡ÀTe bond in diorganyl ditellurides is retained when they react with the cationic iron complex 116.The reaction occurs under mild conditions (CH2Cl2, *20 8C) and affords complexes 132a,b in high yields (85% ¡À 98%).134 The sulfur and selenium analogues of the complexes 132 were prepared in a similar way in 89%¡À 98% yields.134 [CpFe(CO)2THF]+BF¡¦4 +RTe TeR 7THF 116 + Cp OC Fe TeR BF¡¦4 OC TeR 132 R=Me (a), Ph (b). When diphenyl ditelluride reacts with two equivalents of the salt 116 or when the complex 132b is treated with one equivalent of 116, complex 133 is formed in 81% yield. Obviously, in this compound, diphenyl ditelluride acts as a bridging ligand, both its Te atoms being linked to iron atoms.134 The sulfur and selenium analogues of 133 were synthesised in a similar way in 84% and 87% yields, respectively.134 It should be noted that the complex 133 and its chalcogen analogues react with MeCN to give mixtures of complexes 132b and 134.[CpFe(CO)2THF]+BF¡¦4 +PhTe TePh 7THF 116 + Ph Cp 116 OC Fe Te BF¡¦47THF Te OC Ph 132b 2+ Cp Cp CO OC Fe Fe 2BF¡¦4Te Te CO OC Ph Ph 133 + Cp MeCN 133 BF¡¦ N CMe Fe 132b+ OC 4 OC 134 Diphenyl dichalcogenide complexes of the type 132 behave differently on heating in (CD3)2CO. In the sulfur complex, the solvent displaces the disulfide ligand, which leads to complex 135.134 + + Cp Ph Cp (CD3)2CO, D O S OC Fe OC Fe BF¡¦ BF¡¦ C(CD3)2 4 4 7Ph2S2 S OC OC 135 Ph I D Sadekov, A I Uraev, A D Garnovskii In the case of the selenium analogue, the data of 13C NMR spectroscopy indicate the occurrence of a dynamic process, which was interpreted as a 1,2-shift of the CpFe(CO)2 fragment.134 + + Se Ph Cp Cp Ph OC Fe OC Fe Se Se BF¡¦ BF¡¦ 4 4 OC OC Se Ph Ph For the tellurium complex 132b, no splitting of the 13C signals of the phenyl groups is observed even at 200 K.131 3.Complexes of metal carbonyls with compounds with tellurium¡À element bonds The formation of complexes from metal carbonyls and com- pounds containing Te ¡À element bonds can, in principle, involve coordination of the metal atom both to tellurium and to the other element. The interaction of metal hexacarbonyls with compounds (Me3E)2Te (136, E=Ge, Sn, Pb) gives rise to complexes 137, their yields being 17%¡À 45%.135, 136 The reactions proceed in THF on exposure to UV radiation.(Me3E)2Te+M(CO)6 THF, hn 7CO 136 (Me3E)2Te M(CO)5 137 M=Cr, Mo, W; E=Ge, Sn, Pb. The complexes 137 are rather sensitive to moisture and atmospheric oxygen. Complexes with similar compositions and structures are formed when Cr, Mo and W hexacarbonyls are made to react with bis-element-substituted sulfides 137 and sele- nides.135 The reactions of Mn(CO)5Br and Re(CO)5Cl with (Me3Sn)2E (E=Se, Te) afford binuclear complexes 138a ¡À d, which contain apparently bridging (Me3Sn)E ligands.138 Unlike the selenium derivatives 138a,b, which are formed in high yields, the yields of their tellurium analogues 138c,d are quite low.Thus the manga- nese complex 138c was isolated in a yield of only 11%, and the formation of the rhenium derivative 138d was judged only from the IR spectra of the reaction mixture.138 SnMe3 E M(CO)4 (CO)4M (CO)5MHal+(Me3Sn)2E 7CO,7Me3SnHal E 138a ¡À d SnMe3 E=Se: M=Mn (a), Re (b); E=Te:M =Mn (c), Re (d). The complexes 138 can be transformed via two routes giving either compounds 139 or 140.138 The thermal transformation of 138 into 139 can be accomplished only for selenium-containing derivatives 138a,b; the tellurium analogue could not be synthes- ised. On treatment with HCl, the complexes 138a ¡À c undergo destannylation to give the compounds 140a ¡À c. [(CO)3MSeM(CO)3] D 7CO 139a,b [(CO)4MESnMe3]2 138a ¡À c [(CO)4MEH]2 HCl 7Me3SnCl 140a ¡À c E=Se: M=Mn (a, 69%), Re (b, 66%); E=Te, M=Mn (c, 57%).The proton signals of the EH groups in the 1H NMR spectra of the complexes 140a ¡À c substantially shift upfield on passing from Se to Te. The chemical shifts of the SeH group in the spectra of 140a,b are 76.20 and 74.13 ppm, respectively, whereas the signal for the TeH group in the spectrum of 140c occurs atSynthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands 712.80 ppm.138 For comparison, the chemical shifts of protons of the TeH groups in stable tellurols (Me3Si)3ETeH are 75.10 (E=C),78.82 (E=Si),79.02 (E=Ge).139 Unlike complexes formed by metal carbonyls with com- pounds containing Te ±E bonds with Group IV elements, in which the tellurium atom acts as the coordination site, in the case of complexes in which E is a Group V element, both tellurium and this element can be involved in coordination. The complexes 141a ± c have been prepared in relatively low yields (14% ± 27%) by reactions of metal carbonyls with com- pounds 142a ± c at room temperature.140, 141 THF M(CO)5 .THF+(CF3)2ETeMe 142a ± c (CO)5M(CF3)2ETeMe 141a ± c M=Cr: E=P (a), As (b);M=Mo, E = P (c).The complexes 141a ,c with phosphorus ligands are relatively stable in the light both in the solid state and in solutions in the absence of atmospheric oxygen. The arsenic analogue 141b decomposes in solution over a period of several days even in the dark and under an inert atmosphere.Based on the spectroscopy data,140, 141 it was concluded that in phosphorus complexes, the metal is coordinated only to the P atom, while the arsenic complex exists as two isomers in which the metal is coordinated to either As or Te. The reaction of the complex 141a with trimethyltin hydride occurs with rupture of the P ± Te bond.142 Cr(CO)5(CF3)2PTeMe+Me3SnH 141a Cr(CO)5(CF3)2PH+MeTeSnMe3 A common drawback of studies describing metal carbonyl complexes with tellurium-containing ligands with Te ± element bonds is that no X-ray diffraction data are presented. IV. Conclusion In the last 10 ± 15 years, substantial progress has been achieved in the synthesis and study of complexes of metal carbonyls and cyclopentadienyl carbonyls with organic derivatives of telluriu- m(II).This includes synthesis of the complexes of metal carbonyls and cyclopentadienyl carbonyls with tellurocarbonyl compounds and triorganylphosphine tellurides, the use of organic derivatives of tricoordinated (diaryl telluroxides) and tetracoordinated (s- telluranes) tellurium for the synthesis of Te(II) complexes and the possibility of thermolytic synthesis of metal tellurides. The reac- tions of bidentate N,Te-containing organic ligands also present interest; in particular, this refers to the synthesis of the first structurally characterised complex of benzochalcogenazoles with metal ± tellirium coordination. It is significant that many types of complexes of organotellurium compounds considered here are more stable than complexes containing structurally related organic compounds of sulfur and selenium.Further research in this field should be directed apparently along the following lines: (1) systematic studies of complexes with broad variation of the electronic and spatial structures of organo- tellurium ligands and metal carbonyls; (2) the search for new preparative methods for the synthesis of complexes of Te(II) organic derivatives using various compounds of tricoordinated (telluronium ylides, tellurimides, telluronium salts) and tetracoor- dinated (organyltellurium trihalides, aryltellurinic acid deriva- tives) tellurium as well as organyltellurenyl halides, organyl tellurocyanates, and telluroesters; (3) elucidation of the relation- ship between the structural and spectral characteristics of com- plexes and the structures of both initial components and elucidation of the influence of the nature of the chalcogen atom on the structure and reactivity of complexes containing similarly constructed organochalcogen ligands; (4) synthesis of complexes with various ambidentate tellurium-containing ligands; (5) prep- 431 aration of complexes with organotellurium ligands of various types containing bulky organic substituents (Me3Si)3E (E=C, Si, Ge) and the use of these complexes as initial compounds for low- temperature synthesis of metal tellurides.The review was financially supported by the Russian Founda- tion for Basic Research (Projects No.96-03-32502a, No. 96-03- 32026a) and by the INTAS (Grant 94-4675). References 1. S G Murray, F R Hartley Chem. Rev. 81 365 (1981) 2. H J Gysling Coord. Chem. Rev. 42 133 (1982) 3. H J Gysling, in Proceedings of the Vth International Conference on the Organic Chemistry of Selenium and Tellurium, Birmingham, 1983 (Eds F J Berry,WR McWhinnie), p. 32 4. H J Gysling, in The Chemistry of Organic Selenium and Tellurium Compounds Vol. 1 (Eds S Patai, Z Rappoport) (New York: Wiley Interscience, 1986) p. 680 5. D K Kershner, F Basolo Coord. Chem. Rev. 79 279 (1987) 6. A K Singh, V Srivastava J. Coord. Chem. 27 237 (1992) 7. E G Hope,W Levason Coord. Chem. Rev. 122 109 (1993) 8. I Haiduc, R B King,M G Newton Chem.Rev. 94 301 (1994) 9. E Hey-Hawkins Chem. Rev. 94 1661 (1994) 10. J Arnold Prog. Inorg. Chem. 43 353 (1995) 11. S V Larionov, S M Zemskova Ross. Khim. Zh. 40 171 (1996) a 12. M L Steigerwald, C E Rice J. Am. Chem. Soc. 110 4228 (1988) 13. M L Steigerwald, T Siegrist, S M Stuczynski Inorg. Chem. 30 2256 (1991) 14. M L Steigerwald, T Siegrist, S M Stuczynski Inorg. Chem. 30 4940 (1991) 15. A A Pasynskii, Yu V Torubaev, I L Eremenko, D Vegini, S E Nefedov, Zh V Dobrokhotova,A I Yanovskii, Yu T Struchkov Zh. Neorg. Khim. 41 2006 (1996) b 16. K McGregor, G B Deacon, R S Dickson, G D Fallon, R S Rowe, B O West J. Chem. Soc., Chem. Commun. 1293 (1990) 17. A P Coleman, R S Dickson, G B Deacon, G D Fallon, M Ke, K McGregor, B O West Polyhedron 13 1277 (1994) 18.L C Roof, J W Kolis Chem. Rev. 93 1037 (1993) 19. I D Sadekov, A A Maksimenko, V L Nivorozhkin Usp. Khim. 67 219 (1998) [Russ. Chem. Rev. 67 193 (1998)] 20. N Kuhn, H Schumann, G Wolmershauser J. Chem. Soc., Chem. Commun. 1595 (1985) 21. E W Ainscough, A M Brodie, A R Furness J. Chem. Soc., Dalton Trans. 2360 (1973) 22. J F Corrigan, S Balter, D Fenske J. Chem. Soc., Dalton Trans. 729 (1996) 23. N Kuhn, H Schumann J. Chem. Soc., Dalton Trans. 541 (1987) 24. N Kuhn, H Schumann J. Organomet. Chem. 304 181 (1986) 25. S Pohl Z. Naturforsch. B, Chem. Sci. 33 610 (1978) 26. S Pohl Z. Naturforsch. B, Chem. Sci. 34 256 (1979) 27. C Romming, A J Iversen, J Songstad Acta Chem. Scand., Ser. A 34 333 (1980) 28. C Romming, K Maartmann-Moe, J Songstad Acta Chem.Scand., Ser. A 38 349 (1984) 29. N Kuhn, H Schumann, G Wolmershauser Z. Naturforsch. B, Chem. Sci. 42 674 (1987) 30. N Kuhn, G Henkel, H Schumann, R Flohlich Z. Naturforsch. B, Chem. Sci. 45 1010 (1990) 31. M Bochman, G C Bwembya, N T Whilton, X Song, M B Hursthouse, S J Coles, A Karaulov J. Chem. Soc., Dalton Trans. 1887 (1995) 32. M R Spirlet, G van den Bosshe, O Dideberg, P Dupont Acta Crystallogr., Sect. B 35 1727 (1975) 33. F H Kruse, R E Marsh, J D McCullough Acta Crystallogr. 10 201 (1957) 34. S Ludlow, A E McCarthy J. Organomet. Chem. 219 169 (1981) 35. G Llabres, O Dideberg, L Dupont Acta Crystallogr., Sect. B. 28 2438 (1972) 36. T H Allen, O Kennard, D G Watson, L Brammer, A G Orpen, R Taylor J.Chem. Soc., Perkin Trans. 2 31 (1987) 37. H Fisher, S Zeuner J. Organomet. Chem. 252 C63 (1983) 38. H Fisher, A FruÈ h, C Troll J. Organomet. Chem. 415 211 (1991) 39. P Denifl, B Bildstein J. Organomet. Chem. 453 53 (1993)432 40. M Minoura, T Kawashima, R Okazaki J. Am. Chem. Soc. 115 7019 (1993) 41. M Minoura, T Kawashima, N Tokitoh, R Okazaki Phosphorus Sul- fur Silicon Relat. Elem. 136 ± 137 549 (1998) 42. M Minoura, T Kawashima, N Tokitoh, R Okazaki J. Chem. Soc., Chem. Commun. 123 (1996) 43. M F Lappert, T R Martin, G M McLaughlin J. Chem. Soc., Chem. Commun. 635 (1980) 44. H Fisher, U Gerbing J. Organomet. Chem. 299 C7 (1986) 45. H Fisher, I Pashalidis J. Organomet. Chem. 348 C1 (1988) 46. D Rabinovich, G Parkin J.Am. Chem. Soc. 115 9822 (1993) 47. G Silvey, W Hardy, C H Townes Phys. Rev. 87 236 (1952) 48. W Hieber, T Kruck Chem. Ber. 95 2027 (1962) 49. E D Schermer, W H Baddley J. Organomet. Chem. 30 67 (1971) 50. E Kostiner,M L N Reddy, D S Urch, A G Massey J. Organomet. Chem. 15 383 (1968) 51. M L N Reddy,M R Wiles, A G Massey Nature (London) 217 740 (1968) 52. S E Nefedov, B I Kolobkov, A A Pasynskii, I L Eremenko, I D Sadekov, A I Yanovskii, Yu T Struchkov Zh. Neorg. Khim. 37 335 (1992) b 53. A A Pasynskii, Yu V Torubaev, I L Eremenko, D Vegini, A I Yanovskii, Yu T Struchkov Koord. Khim. 22 924 (1996) c 54. A A Pasynskii, Yu V Torubaev, I L Eremenko, D Vegini, A I Yanovskii, Yu T Struchkov Koord. Khim. 22 369 (1996) c 55. H Takahashi, K Ohe, S Uemura, N Sugita J.Organomet. Chem. 334 C43 (1987) 56. S Uemura, H Takahashi, K Ohe, N Sugita J. Organomet. Chem. 361 63 (1989) 57. P J Bonasia, J Arnold J. Organomet. Chem. 449 147 (1993) 58. W F Liaw, S J Chiou,W Z Lee, P Gene-Hsiang, S M Peng J. Chin. Chem. Soc. (Taipei) 43 29 (1996) 59. A S Gybin, W A Smith, V S Bogdanov,M Z Krimer, J B Kalyan Tetrahedron Lett. 21 383 (1980) 60. Yu B Kal'yan,M Z Krimer, E G Cherepanova, V S Bogdanov, V A Smit Izv. Akad. Nauk SSSR, Ser. Khim. 378 (1982) d 61. G Capozzi Pure Appl. Chem. 59 989 (1987) 62. P G Gassman, A Miura, T Miura J. Org. Chem. 47 951 (1982) 63. W F Liaw, D S Ou, Y S Li,W Z Lee, C Y Chuang, Y P Lee, G S Lee, S M Peng Inorg. Chem. 34 3747 (1995) 64. J L Atwood, I Bernal, F Calderazzo, L G Canada, R Poli, R D Rogers, C A Veracini, D Vitali Inorg.Chem. 22 1797 (1983) 65. P Mathur, D Chakrabarty J. Organomet. Chem. 373 129 (1989) 66. WF Liaw, C Y Chuang,WZ Lee, G H Lee, C K Lee, SM Peng Inorg. Chem. 35 2530 (1996) 67. F Calderazzo, D Vitali, R Poli, J L Atwood, R D Rogers, J M Cummings J. Chem. Soc., Dalton Trans. 1004 (1981) 68. R T Mehdi, J D Miller J. Chem. Soc., Dalton Trans. 1065 (1984) 69. S A Gardner J. Organomet. Chem. 190 289 (1980) 70. I D Sadekov, V I Minkin Sulfur Rep. 19 285 (1997) 71. WF Liaw,MH Chiang, C J Liu, P J Harn, L K Liu Inorg. Chem. 32 1536 (1993) 72. W F Liaw, S J Chiou,W Z Lee, S M Peng J. Chin. Chem. Soc. (Taipei) 40 361 (1993) 73. W F Liaw,M H Chiang, C H Lai, S J Chiou, G H Lee, S M Peng Inorg.Chem. 33 2493 (1994) 74. W F Liaw, C H Lai,M H Chiang, C K Hsien, G H Lee, S M Peng J. Chin. Chem. Soc. (Taipei) 40 437 (1993) 75. E W Tillay, E D Schermer,W H Baddley Inorg. Chem. 7 1925 (1968) 76. P Jainter J. Organomet. Chem. 233 333 (1982) 77. P Jainter, W Wohlgenannt Inorg. Chim. Acta 101 L43 (1985) 78. L Y Goh,M S Tay, C Wei Organometallics 13 1813 (1994) 79. E D Schermer, W H Baddley J. Organomet. Chem. 27 83 (1971) 80. R E Cobbledick, N S Dance, F W B Einstein, C H W Jones, T Jones Inorg. Chem. 20 4356 (1981) 81. P Jainter, W Wohlgenannt J. Organomet. Chem. 297 281 (1985) 82. A Bondi J. Phys. Chem. 68 441 (1964) 83. M Sato, T Yoshida J. Organomet. Chem. 94 403 (1975) 84. G Huttner, S Schuler, L Zsolnai, M Gottlieb, H Braunwarth, M Minelli J.Organomet. Chem. 299 C4 (1986) 85. M Herberhold, D Reiner, D Neugebauer Angew. Chem. 95 46 (1983) 86. W A Herrmann, J Rohrmann, C Hecht J. Organomet. Chem. 290 53 (1985) I D Sadekov, A I Uraev, A D Garnovskii 87. W A Herrmann, C Hecht, E Herdtweck J. Organomet. Chem. 331 309 (1987) 88. P Lau, G Huttner, L Zsolnai J. Organomet. Chem. 440 41 (1992) 89. R Kaur, H B Singh, R J Butcher Organometallics 14 4755 (1995) 90. A D Garnovskii, A I Uraev, A S Antsyshkina, I D Sadekov, V I Minkin Koord. Khim. 22 512 (1996) a 91. G Thaler, K Wurst, F Sladky Organometallics 15 4639 (1996) 92. G Bremer, R Boese, M Keddo, T Kruck Z. Naturforsch. B, Chem. Sci. 41 981 (1996) 93. J C Weis, W Beck J. Organomet. Chem. 44 325 (1972) 94. K Badyal, W R McWhinnie, T A Hamor, H Chen Organometallics 16 3194 (1997) 95. K OÄ fele, E Dotzauer J. Organomet. Chem. 42 C 87 (1972) 96. J K Shen, Y C Gao, Q Z Shi, A L Rheingold, F Basolo Inorg. Chem. 30 1868 (1991) 97. Y C Gao, X Yana, Q Z Shi Inorg. Chim. Acta 240 661 (1995) 98. P Jainter, W Winder Inorg. Chim. Acta 134 201 (1987) 99. H Hausmann, M Hofler, T Kruck, H W Zimmerman Chem. Ber. 114 975 (1981) 100. A D Garnovskii, O A Osipov, S B Bulgarevich Usp. Khim. 41 648 (1972) [Russ. Chem. Rev. 41 341 (1972)] 101. K Singh,W R McWhinnie, R William, H L Chen,M Sun, T A Hamor J. Chem. Soc., Dalton Trans. 1545 (1996) 102. E H Braye,W HuÈ bel, I Caplier J. Am. Chem. Soc. 83 4406 (1961) 103. C E Strouse, L F Dahl J. Am. Chem. Soc. 93 6032 (1971) 104. H Keller, J C Daran, H Lang J. Organomet. Chem. 482 63 (1994) 105. J Bergman, L Engman J. Organomet. Chem. 175 233 (1979) 106. A J Arce, A J Deeming, Y De Sanctis, R Machado, J Manzur, C Rivas J. Chem. Soc., Chem. Commun. 1568 (1990) 107. D Himmelreich, G MuÈ ller J. Organomet. Chem. 297 341 (1985) 108. A P Coleman,M Ke, R S Dickson, G B Deacon, B O West Polyhedron 13 2301 (1994) 109. W Hieber, K Wollmann Chem. Ber. 95 1522 (1962) 110. W Hieber, F Stanner Chem. Ber. 102 2930 (1969) 111. A Belforte, F Calderazzo, D Vitali, P F Zanazzi Gazz. Chim. Ital. 115 125 (1985) 112. W Hieber,W Opavsky,W Rohm Chem. Ber. 101 2244 (1968) 113. F Faraone, S Sergi, R Pietropaolo J. Organomet. Chem. 24 453 (1970) 114. F Faraone, R Pietropaolo, S Sergi J. Organomet. Chem. 24 797 (1970) 115. A D Garnovskii, G M Abakarov, I D Sadekov, V I Minkin, T G Cherkasova, Yu S Varshavskii Koord. Khim. 10 234 (1984) c 116. F Bonati, G Wilkinson J. Chem. Soc. 3156 (1960) 117. Yu S Varshavskii, T G Cherkasova Zh. Neorg. Khim. 12 1709 (1967) b 118. F Faraone, R Pietropaolo, S Sergi, P Pira J. Organomet. Chem. 24 805 (1970) 119. A Z Al-Rubaie, Y N Al-Obaidi, L Z Yousif Polyhedron 9 1141 (1990) 120. P Piriano, F Faraone, R Pietropaolo Atti. Accad. Peloritana Pericolanti. Cl. Sci., Fis., Mat. Natur. 51 283 (1971); Chem. Abstr. 79 13 011 (1973) 121. W Hieber, P John Chem. Ber. 103 2161 (1970) 122. P John Chem. Ber. 103 2178 (1970) 123. M Hoch, D Rehder J. Organomet. Chem. 288 C25 (1985) 124. A Belforte, F Calderazzo, P F Zanazzi Gazz. Chim. Ital. 115 71 (1985) 125. J W Freeman, F Basolo Organometallics 10 256 (1991) 126. N Kuhn, H Schumann J. Organomet. Chem. 276 55 (1984) 127. N Kuhn, H Schumann, E Zauder J. Organomet. Chem. 327 17 (1987) 128. N Kuhn, H Schumann, E Zauder J. Organomet. Chem. 354 161 (1988) 129. D Rehder, K Ihmels, D Wenke, P Oltmanns Inorg. Chim. Acta 100 L11 (1985) 130. K Ihmels, D Rehder Organometallics 4 1340 (1985) 131. H Schumann, A M Arif, A L Rheingold, C Janiak, R Hoffmann, N Kuhn Inorg. Chem. 30 1618 (1991) 132. A A Pasynskii, Yu V Torubaev, A G Drukovskii, I L Eremenko, D Vegini, E V Krasil'nikova, V I Privalov, A I Yanovskii, Yu T Struchkov Zh. Neorg. Khim. 42 42 (1997) bSynthesis, reactions and structures of complexes of metal carbonyls and cyclopentadienyl carbonyls with organotellurium ligands 133. A A Pasynskii, Yu V Torubaev, I L Eremenko, E V Krasil'nikova, A I Yanovskii, Yu T Struchkov Zh. Neorg. 134. N Kuhn, H Schumann J. Organomet. Chem. 287 345 (1985) 135. H Schumann, R Weis Angew. Chem., Int. Ed. Engl. 9 246 (1970) 136. H Schumann,R Montachemi,H J Kroth,U Frank Chem. Ber. 106 137. H Schumann, O Stelzer, W Gick Angew. Chem. 81 256 (1969) 138. V Kullmer, H Vahrenkump Chem. Ber. 110 228 (1977) 139. P J Bonasia, V Christou, J Arnold J. Am. Chem. Soc. 115 6777 Khim. 42 730 (1997) b 2049 (1973) (1993) 140. J Grobe, D Le Van Z. Naturforsch. B, Chem. Sci. 34 1653 (1979) 141. J Grobe, D Le Van Z. Naturforsch. B, Chem. Sci. 35 694 (1980) 142. J Grobe, D Le Van Z. Naturforsch. B, Chem. Sci. 35 666 (1980) a�Mendeleev Chem. J. (Engl. Transl. ) b�Russ. J. Inorg. Chem. (Engl. Transl.) c�Russ. J. Coord. Chem. (Engl. Transl.) d�Russ. Chem. Bull. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC