摘要:

DALTON PERSPECTIVE J. Chem. Soc., Dalton Trans., 1998, 3133–3148 3133 The kinetics and mechanisms of the crystallisation of microporous materials Robin J. Francis and Dermot O’Hare *† Inorganic Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, UK OX1 3QR. E-mail: dermot.ohare@chemistry.ox.ac.uk Received 25th March 1998, Accepted 18th June 1998 The synthesis of microporous materials is one of the major activities in contemporary solid state chemistry. This perspective brings together the current thinking regarding the synthesis and growth mechanisms of these technologically important materials.Also reviewed are the latest experiments being utilised to probe the complex chemistry which occurs during the synthesis of these materials. 1 Introduction Microporous materials, often referred to as molecular sieves or open-framework materials, are a class of inorganic solids which possess regular pores or voids in the size range 5–20 Å (Fig. 1). Zeolites are the most well known family of such materials. Following the successful synthesis of the first artificial zeolite in 1948 (see below),1 their utility as catalysts for the production of petrochemicals stimulated great interest in the synthesis of other zeolitic materials and this first report was quickly followed by many others. Today, there are hundreds of known microporous materials, and investigations on these compounds remain an extremely active area of research.2 The continued eVort to synthesize new materials in this class, and to gain a greater understanding of their crystallisation mechanisms, is driven by the broad range of useful and unique properties they possess.Microporous materials contain uniformly sized pores in the range 5–20 Å, and can thus display molecular recognition, discriminating and organisational properties with a resolution of less than 1 Å. They are therefore of great interest as materials for a range of molecular recognition applications, Robin J.Francis Dermot O’Hare † D. O’Hare is the Royal Society of Chemistry Sir Edward Frankland Fellow. Robin Francis was born in Solihull, England in 1972. He received his BA from the University of Oxford in 1994, having done undergraduate research in Dr. O’Hare’s group studying the hydrothermal synthesis of open-framework tin sulfides. His D.Phil. research, also performed under the supervision of Dr. O’Hare at Oxford, was concerned with the application of in situ diffraction techniques to elucidate the kinetics and mechanisms of the hydrothermal synthesis of microporous materials.He is currently engaged in postdoctoral research in materials chemistry under the supervision of Professors A. J. Jacobson and S. C. Moss at the University of Houston, Texas. Dermot O’Hare obtained his first degree in Chemistry from the University of Oxford in 1982 and then remained at Oxford for his D.Phil. under the supervision of Malcolm Green, where he worked on carbon–hydrogen bond activation using metal atom chemistry.In 1990 he was appointed to a University Lectureship in Inorganic Chemistry and Septcentenary Tutorial Fellowship at Balliol College in 1990. His current interests span a wide range of inorganic chemistry from synthetic molecular organometallic chemistry through to solid state chemistry. He received the RSC’s Sir Edward Frankland Fellowship in 1996–1997. In 1998 he received the Corday-Morgan Medal and Prize and the Exxon Chemicals European Science and Engineering award.as well as for the more familiar applications such as catalysts, absorption and ion exchange. For these reasons, the synthesis of new zeolitic materials and an understanding of their mode of formation continues to be of paramount importance. The vast majority of microporous materials are constructed from linked TO4 tetrahedra (where T = tetrahedral atom, e.g. Al, Si, P, etc.) in which each oxygen is shared between two adjacent tetrahedra to give frameworks with an O/T ratio of 2.The tetrahedra are linked in such a way as to form regularly sized pores, channels and cages within the materials such that a significant fraction (up to 50% in some cases) of the materials is literally ‘empty space’. Zeolites are hydrated aluminosilicates constructed from linked AlO4 and SiO4 tetrahedra with the general formula Mn1 x/n[(AlO2)x(SiO2)]x2?zH2O.3 Pure silica zeolites contain neutral frameworks, whereas aluminosilicates contain a negatively charged oxide framework (one negative charge per Al31).This negative charge is balanced by extraframework positive ions (Mn1) which reside inside the channels and cages of the zeolite. Aluminophosphates (AlPO4s) constitute another large class of open-framework materials.4,5 Their basic frameworks are built from linked tetrahedral Al31 and P51 units and have the general formula Al2O3?1 ± 0.2P2O5?xR? yH2O, where R is the amine or quaternary ammonium salt used in the original synthesis.Conceptually, a neutral AlPO4 framework can be considered to be derived from a neutral pure silica zeolite by replacement of two Si41 cations with one Al31 and one P51 cation. Isomorphic substitution of framework Al31 and P51 ions by divalent metal cations or silicon produces the MeAPO (metal aluminophosphate) and SAPO (silicon aluminophosphate) family of materials respectively.6 All of the above materials contain only tetrahedrally co-ordinated units.However, recently novel microporous materials have been3134 J. Chem. Soc., Dalton Trans., 1998, 3133–3148 Fig. 1 The free pore diameters and structures of several well known microporous materials (after Davis and Lobo 32). synthesized in which the metal phosphate framework is constructed from units in which the metal atom resides in complexes of higher co-ordination than tetrahedral such as vanadium,7–11 cobalt,12–15 molybdenum,16,17 iron,18 gallium,19–26 and indium 27,28 phosphates.Nowadays, the synthesis and characterisation of new microporous materials is a huge field; over a thousand research papers a year are published on the subject, and new molecular sieve materials are regularly being discovered. This research eVort is not only driven by pure academic interest but by the continued discovery and development of useful applications for these materials due to their unique chemical, physicochemical and catalytic properties.Unfortunately, the processes by which microporous materials form in hydrothermal crystallisations are very complex and poorly understood (see below). Lok et al.29 have described understanding molecular sieve crystallisations as one of the most challenging chemical problems of today. The lack of mechanistic understanding of the formation of molecular sieves has meant that the discovery of new microporous materials, to date, has been a mainly heuristic exercise involving the systematic exploration of many reaction variables, and requiring a fair degree of serendipity for the successful synthesis of new materials.A more complete understanding of the fundamental processes occurring during hydrothermal crystallisations leading to more rational syntheses of new molecular sieve materials would therefore be of great value. In this short article we review the current state of the art relating to the kinetics, growth and mechanisms of formation of these materials, and the experimental techniques that are currently being applied to the study of their formation.This perspective is not intended to be a detailed review of the synthesis and structures of microporous materials, and for much more detailed information the reader is directed towards the many excellent books and reviews that have been written on the subject.3,5,6,30–33 2 Historical background Although zeolites ‡ were first identified as a class of minerals in ‡ Zeolitic materials were first recognised as a new class of compounds by the Swedish mineralogist A.F. Cronstedt in 1756. He observed that when an unidentified silicate material (since identified as stilbite) was heated in a blowpipe flame it fused with marked intumesence (swelling). This result led to all other minerals that showed this property to be called zeolites, which is derived from the Greek words zeo meaning ‘to boil’ and lithos meaning ‘stone’. Since then approximately 40 natural zeolites have been discovered. 1756, attempts artificially to synthesize zeolitic materials did not begin until 1862. Early attempts concentrated on simulating the high temperatures and pressures (T > 200 8C, P > 100 bar; bar = 105 Pa) of geological conditions under which natural zeolites were believed to form. However, it was not until 1948 that the successful synthesis of a zeolitic material without a natural counterpart was reported by Barrer.1 The first large scale synthetic methodologies for the synthesis of zeolites were pioneered by Milton and co-workers at the Union Carbide laboratories in the late 1940s, who developed hydrothermal zeolite syntheses at low temperature (ca. 100 8C) and low pressure (autogenous) using alkali metal aluminosilicate gels. The next major advance in zeolite synthesis occurred in 1961, when Barrer and Denny 34 reported the synthesis of zeolites using organic alkylammonium cations instead of alkali metal cations.The introduction of organic cations allowed the synthesis of zeolites with a much higher Si/Al ratio than is found in natural zeolites. The use of organic cations in zeolite synthesis increased rapidly after the work of Barrer and Denny, and the ready availability of a large range of organic cations allowed many new high silica zeolites to be prepared. A further significant advance in molecular sieve synthesis occurred in 1982, when Wilson et al.4 reported the synthesis of aluminophosphate (AlPO4) molecular sieves.These materials were synthesized under acidic or neutral conditions as opposed to the strongly basic conditions used in the synthesis of zeolites. Following this discovery other related classes of materials such as MeAlPOs and SAPOs were rapidly synthesized in which the addition of other elements into the reaction gels results in their incorporation into the framework.6 Recently, the synthesis of microporous structures has begun to spread across all parts of the Periodic Table, and a number of new classes of microporous materials incorporating a large variety of main group and transition metal elements have been synthesized. 3 Synthesis of microporous materials Molecular sieves are almost exclusively synthesized under hydrothermal conditions at temperatures of between 100 and 250 8C under autogenous pressure, under either strongly basic conditions (for zeolites), or weakly acidic or neutral conditions (for metal phosphates or derivatives).The versatility of the hydrothermal technique derives from the extremely eVective solvating ability of water under these conditions. This allows the dissolution and mixing of the solid reagents to form an inhomogeneous gel in the initial stages of the reaction. AtJ. Chem. Soc., Dalton Trans., 1998, 3133–3148 3135 Fig. 2 Schematic diagram illustrating Ostwald’s law of successive reactions operating during the synthesis of zeolite A and sodalite.Zeolite A is the kinetically favoured product and at short reaction times it is the dominant product of the reaction. At longer reaction times a transformation to the more thermodynamically stable sodalite structure occurs. At very long reaction times conversion into the thermodynamically most stable product, SiO2 and Al2O3, will eventually occur. later times nucleation centres are formed which subsequently grow as the reaction proceeds to form the final crystalline product.The chief diYculty of the hydrothermal technique is the very large number of possible reaction variables, all of which aVect the pathway and kinetics of the reactions in ways that are not generally understood.32 Typical reaction variables include time, temperature, pressure, reactant source and type, pH, the inorganic or organic cation used, ageing time of the gel, reaction cell fill volume, and so on. Since, in general, variation of any one of these parameters can have an eVect on several others, it is often diYcult to evaluate the eVect of varying one parameter in a straightforward way.Owing to this diYculty, the synthesis of new molecular sieves has generally proceeded by systematically exploring what the eVect of changing each variable is on the synthesis. Since this eVectively means exploring a vast n-dimensional reaction space (where n is the number of reaction variables), this process can be very time consuming and not very eYcient. A greater understanding of the processes occurring during hydrothermal syntheses leading to a more rational approach to the synthesis of new molecular sieves would be very desirable.Nevertheless, despite these diYculties, certain guidelines for the eVects of various reaction variables on hydrothermal syntheses can be given. It is generally found that as the temperature of synthesis is raised there is a trend towards forming species with a lower intercrystalline void space and lower water content.For example, zeolites such as A and X which have porosities at the high end of the range for zeolites (up to 50% void space) are generally synthesized at temperatures close to 100 8C, whereas reactions at much higher temperatures (e.g. 350 8C) often yield dense phases. This can easily be rationalised by the exponential increase in the autogenous pressure of water with increasing temperature. Of course, it is evident that changing the initial reaction composition will aVect the nature of the final phase formed.However, it is unfortunately not simply the case that one can tailor the composition of the final product simply by using the desired ratios of the starting materials in the reaction. This is because molecular sieve syntheses are generally inhomogeneous reactions consisting of both liquid and solid components, and changing the qualities of any one component changes the chemical composition of both the solution and the solid phase.Hence, the chemical composition of the solid product does not reflect the overall composition of the mixture. Molecular sieve materials are metastable phases which are thermodynamically unstable with respect to dense oxide phases. It is therefore clear that the formation of zeolitic materials cannot be rationalised on the basis of thermodynamics alone, and kinetics must also play a large part in determining which particular phases are formed.Time is therefore also an important factor governing the products formed in molecular sieve syntheses. The synthesis of zeolitic materials obeys Ostwald’s law of successive reactions. This law states that an initial metastable phase is successively converted into a thermodynamically more stable phase until the most stable phase is produced. Ostwald’s law has been observed in a number of zeolite syntheses (Fig. 2). For example, zeolite A converts into the more stable sodalite after long reaction times.The successful commercial synthesis of the former phase relies on controlling the reaction time to produce optimum yields. In molecular sieve syntheses it is found that the nature of the cation used in the synthesis is a critical factor in determining the composition and structure of the final product formed. In zeolite syntheses the use of alkali metal cations generally results in the synthesis of aluminium rich zeolites; if organic cations are used silicon rich zeolites are formed.Silicon rich zeolites can only be synthesized in the presence of organic cations. (The exception of ZSM-5, which is a high silica zeolite which can be synthesized in the absence of any organic cations over a very narrow range of Na1 and aluminium concentrations.)35 This can be rationalised by a consideration of the much larger size of organic cations compared with alkali metal cations. This greater size means that fewer cations can be contained within the zeolite framework.Hence, for the material to be charge neutral the charge density of the host framework must be lower, i.e. the Si/Al ratio must be higher. The large variety of sizes and shapes of organic cations when compared to spherical alkali metal cations means there is much more scope for the synthesis of new silicon rich zeolites compared with aluminium rich zeolites. The ability to control the steric and electronic nature of the organic cation adds a new dimension to the chemist’s ability to control the interactions occurring during crystallisation, and consequently the structure of the final product. Implicit in this is the idea that the organic molecule is acting as much more than simply a charge balancing cation and is playing a structure directing or ‘templating’ role during the crystallisation of the zeolite.This idea is central to much molecular sieve synthesis and is discussed in much more detail in the next section.The synthesis of metal phosphate molecular sieves and derivatives follows broadly the same principles as for zeolites, with the important diVerence that phosphate molecular sieves are always synthesized under either acidic, neutral or mildly basic conditions as opposed to the highly basic conditions used in the synthesis of zeolites.4–6 The vast majority of phosphate molecular sieves are synthesized at a pH of between 4.0 and 6.5 and temperatures of between 130 and 200 8C.Phosphate based molecular sieves can only be synthesized in the presence of organic amines or alkylammonium ions. Thus, as with the synthesis of high silica zeolites, the organics appear to be acting in a structure directing template role. An interesting and notable exception to this rule is the synthesis of the extra-large pore aluminophosphate VPI-5. In the original synthesis reported by Davis et al.36 a structure directing agent was used. However, subsequently, Duncan et al.37 reported that under certain3136 J.Chem. Soc., Dalton Trans., 1998, 3133–3148 conditions VPI-5 could be synthesized in the absence of any organic molecules. However, there is also some suggestion that the occluded water in as synthesized VPI-5 is ordered in a triple helix within the aluminophosphate framework.38–41 Although the precise details are not fully understood it has been postulated that the water structure in VPI-5 in some way acts in the role usually played by the inorganic.In this report the synthesis of VPI-5 is unique. This kind of eVect has not been observed in any other AlPO4 syntheses. 4 The concept of templating The concept that the organic molecules used in molecular sieve syntheses were not simply acting as charge balancing cations, and were in fact playing an active role in ‘directing’ the synthesis of a particular molecular sieve structure, was first suggested because of the close correlation that was often seen between the size and shape of the template and the size and shape of the cavities formed.For example, in the synthesis of sodalite using tetramethylammonium cations it is found that in the final product the NMe4 1 cations are located at the centre of the sodalite cages, from which the cation is too big to either enter or leave, suggesting that the sodalite framework must have formed around the cations.42 Another well known example is the synthesis of high silica ZSM-5 using the tetrapropylammonium cation as the organic species.In this case it is found that the cation is located at the intersection of the two intersecting channel systems, with the four long alkyl chains lying along the four individual channels.43 These and other observations led to the suggestion that the organic molecules were acting as ‘templates’ and ‘building’ the molecular sieve structure around themselves by directing the condensing oxide tetrahedra into a particular geometry.29,44 However, it is clear that the situation is very much more complex than originally suggested.For example, although it is true that NPr4 1 is an eVective structure directing agent for the synthesis of ZSM-5, ZSM-5 can also be synthesized in the absence of any organic molecules.35 There are also a plethora of examples of the same organic species forming a variety of diVerent molecular sieve structures, or the same structure being formed by a variety of diVerent organic molecules (there are at least 22 diVerent organic molecules that can be used to synthesize ZSM-5 for example).Furthermore, the correlation between template shape and pore shape is often weak. In the synthesis of aluminophosphates the necessity of using organic molecules, and the specific requirement of using a particular organic agent in the synthesis of certain frameworks [e.g. AlPO4-20 (SOD) can only be synthesized using NMe4OH] 4 also suggested a templating eVect. However, again, it is also true that some AlPO4s can be synthesized using a number of templates [for example, AlPO4-5 (AFI) can be synthesized using over 20 diVerent species],4 and the same template can be used to synthesize a number of structures.A further interesting point is that the very-large pore phosphate based molecular sieves (e.g. VPI-5,36 JDF-20,45 AlPO4-8,46 cloverite,21 ULM-523 and ULM- 1625) have all been synthesized using relatively small organic molecules. Although it is obvious that the organic molecules are playing some structure directing role in the synthesis of molecular sieve materials, it seems clear that in most cases they are not acting in a true templating manner; i.e.not directing the formation of a unique zeolitic structure which reflects the geometric and electronic structure of the template. Some authors have sought to draw a distinction between ‘templating’ and ‘structure direction’.47 In this context, ‘templating’ refers to the process described above, in which a unique template leads to the formation of a unique structure which reflects the geometrical and electronic structure of the template, whereas ‘structure direction’ describes a more subtle eVect in which the use of a particular organic moiety leads to a preference for the synthesis of a particular structure via a combination of factors such as pH modification, solubility modification, and electrostatic interactions with the solubilised silica, aluminium and phosphate species in the reaction mixture.Davis and Lobo 32 in their review published in 1992 further extended this idea, and suggested that organic guest molecules can act in three distinct ways: (i) as space filling species, (ii) as structure directing agents, and (iii) true templates. Space filling refers to the situation in which the role of the organic is simply to exclude water from the voids in the zeolite framework, decreasing unfavourable energetic interactions between the solvent water and the growing molecular sieve.It is clear that in the cases where an organic is simply acting as a space filler, the precise nature of the organic is not of great importance. Therefore the converse can also be implied, namely that in those cases where a great many diVerent templates can be used to synthesize the same structure (such as ZSM-5 and AlPO4-5) the primary role of the organic in these cases must be simply as a space filling agent.Actual structure direction by an organic molecule implies that the use of a particular template leads to a unique structure which cannot be synthesized by the use of any other templates. There are several examples of this kind of structure direction in the zeolite field. For example the synthesis of CIT-1 was achieved by using a very specific organic template.48 There have been several studies published on structure direction in the synthesis of zeolites. One of the most detailed was Gies and Marler’s work49 on the structure directing eVects of organic molecules during the crystallisation of porosils.§ By studying the synthesis of these materials using the simplified system SiO2/organic/H2O they were able to avoid the complicating eVects of the mutual interactions that occur between the various components of the more complex zeolite syntheses studied by previous researchers.Gies and Marler found that there was a very high correlation between the size and shape of the organic used and the size and shape of the framework pore produced.They concluded that, since there are no ionic interactions between the guest molecules and the framework, the closeness of the geometrical fit between the host and guest must be due to an optimised arrangement for maximising the van der Waals contacts between host and guest. These results were supported by subsequent solid state NMR measurements by Burkett and Davis which suggested that weak non-covalent interactions between the organic molecules and the silicate species are important (see below).These results, together with work by Weibcke,51 oVer a fairly convincing argument for structure direction by organic molecules in the synthesis of high silica zeolites. However, it is not clear how true this is for other types of molecular sieves. Little work has been done on the mechanism of structure direction in the synthesis of phosphate molecular sieves.However, since the frameworks of AlPO4s are neutral it seems clear that van der Waals interactions between the host framework and the organic guest must be the dominant factor in determining the structure formed. It has also been suggested that the amines also play a role in modifying the gel chemistry in AlPO4 syntheses. It is known that under the low pH conditions used in AlPO4 syntheses the tetrahedral AlO4 species that form the precursors to the crystalline products are unstable with respect to octahedral aluminium species.52,53 It has therefore been suggested that the amine stabilises the AlO4 units by bonding to them and forming a hydrophobic shell which resists nucleophilic attack by the solvent water.54 Examples of templating in the truest sense of the word (i.e. corresponding to Davis’ third definition) are much rarer.One notable example where true templating may be occurring is the formation of the zeolite ZSM-18. This was first synthesized in § Porosils are porous tectosilicates of the general formula xM?SiO2 (where M is the organic guest and x can vary over a very wide range) and include clathrasils and pure silica zeolites such as ZSM-5, -11 and -48.J.Chem. Soc., Dalton Trans., 1998, 3133–3148 3137 1970 using the triquaternary amine (C18H30N3 31) shown in Fig. 3.55 The extremely close registry between the shape of the organic molecule and the shape of the pore system in ZSM-18 suggested that true templating had indeed occurred.Energy minimisation calculations by Davis and Lobo 32 indicated that, in the lowest energy conformation, the template was held in a cage in the zeolite framework that has the same three-fold symmetry as the organic template, and that the organic molecule was not able to rotate in this cage which implies that true templating has taken place. Subsequent work by Stevens et al.,56 who performed Monte Carlo simulated annealing calculations to predict the location and orientation of organic molecules inside zeolite hosts, supported these calculations.Their calculations showed that along the channels in ZSM-18 there is a perfect match between the shape of the organic and the shape of the zeolite pore. These results showed that the shape of the template plays a crucial role in determining the location of the TO4 groups around it, and hence the structure of the final product. It is these strong guest–host interactions that distinguish true templating from structure direction.It should be obvious from the above discussion that the precise role that organic molecules play in molecular sieve synthesis is not entirely understood and the matter is still very much a subject of debate and discussion. Nevertheless, it is clear that the use of specific organics is frequently necessary for the successful synthesis of a particular molecular sieve structure. For this reason many authors (including us) often use the terms ‘template’ and ‘structure directing agent’ interchangeably and synonymously when referring to organic species. 5 Mechanisms of formation of molecular sieves Conceptually, the crystallisation of a zeolitic material can be considered to follow an idealised process involving three states. Initial dissolution of the solid starting reagents by the solvent water to form a randomly distributed array of reaction components, followed by an ordering of some of these components on a microscopic level (i.e.formation of nucleation sites), and finally the growth of these nucleation sites to form the final material in which long range order is observed (i.e. the formation of crystals). The diYculty in trying to understand the precise details of the mechanisms and processes occurring during the formation of a zeolitic material is due to the extreme complexity of hydrothermal crystallisations. The reactions occur in multicomponent systems in which there are a plethora of interactions, chemical reactions, equilibria, and crystal nucleation and growth processes taking place throughout a heterogeneous reaction mixture.Further to complicate matters, many of these processes are interrelated and change with time over the course of the crystallisation. Nevertheless, several authors have proposed mechanisms for Fig. 3 The organic template, ‘tri-quat’, used to synthesize ZSM-18. the synthesis of molecular sieve materials. In particular, there are two postulated mechanistic processes that comprise the two extremes of the range; (i) the solution-mediated transport mechanism, and (ii) the solid hydrogel transformation mechanism. The solution-mediated transport mechanism involves dissolution of the reagents in the solution phase followed by transport of the dissolved silicate/phosphate species via solution-mediated diVusion to the nucleation sites where crystal growth takes place.The solid hydrogel transformation mechanism involves the reorganisation of the solid phase from an initially amorphous state to one with long range order (i.e.the crystallised zeolite). It is clear that in any particular case the true mechanism could lie somewhere between these two extremes, or could proceed via a combination of both. There are many examples of zeolitic crystallisations that appear to proceed via the solution transport mechanism, particularly in the synthesis of aluminium rich zeolites.Perhaps the most convincing examples are provided by Ueda et al.57 who crystallised zeolites Y, S and P from clear solutions and Testa et al.58 who crystallised zeolites ZSM-5 and ZSM-11 from similar clear solutions. In these cases the possibility of any solid–solid transformation appears to be ruled out. A good example of the solid hydrogel transformation mechanism is demonstrated by Xu et al.59 They synthesized ZSM-5 and ZSM-35 by first dehydrating an amorphous aluminosilicate gel at 550 8C, and then treating this mixture with liquid triethylamine and ethylenediamine in the absence of water at 160 8C.They observed no silicate or aluminate species in the liquid phase during the crystallisation of the zeolite indicating that a solid phase transformation must be occurring. However, work by Iton et al.60 and Bodart et al.61 has shown that ZSM-5 can be synthesized via either of the two extreme reaction mechanisms depending on the reaction conditions.Aluminophosphates also appear to crystallise via both mechanisms depending on the reaction conditions. For example, Pang et al.62 have synthesized element substituted AlPO4s from clear solutions, while Davis et al.63 have demonstrated that VPI-5 can be crystallised via a solid hydrogel mechanism (although as discussed previously the crystallisation of VPI-5 is unusual and may not be representative of the crystallisation of aluminophosphates in general). Since diVerent molecular sieves can crystallise via diVerent mechanisms, and the same molecular sieve can crystallise via diVerent mechanisms, or a combination of mechanisms, depending on the reaction conditions, no general conclusions can be drawn as to which of the postulated mechanisms is occurring in any particular system or class of molecular sieves.Also, neither of the proposed mechanisms addresses the question of exactly what are the detailed processes taking place during crystallisation.In particular, the issues that need to be clarified are: what controls which specific molecular sieve structure is formed?; what are the interactions between the various components in the system?; what species are formed in the solid and in solution as the reaction proceeds?; how is the mechanism of structure direction taking place? (i.e. how is the geometry and electronic structure of the template being transmitted to the silicate/phosphate species in solution in such a way as to translate in the structure of the final zeolitic product?); how does nucleation occur, and how does crystal growth take place once nucleation has occurred? In an eVort to gain a greater understanding of the processes occurring during the formation of a molecular sieve interest has focused on the formation and role in the nucleation process of small inorganic clusters present in molecular sieve synthesis gels. Many spectroscopic techniques have been used to try and identify particular silicate/phosphate molecular species which are postulated to condense initially to form nucleation centres which then grow to form the infinite framework structure.Spectroscopic techniques that have been used include NMR, MAS-NMR, IR and Raman. These studies have identified several fundamental structures found in molecular sieve3138 J. Chem. Soc., Dalton Trans., 1998, 3133–3148 materials (such as single and double four membered rings), and yielded some interesting information about the interaction of species present in zeolitic precursor gels.For example, in the synthesis of zeolite A, Dutta and Shieh 64 observed the formation of 4 T-atom membered rings in the amorphous aluminosilicate solid, and observed that the solid reorganises via interaction with Al(OH)4 2 ions in solution to form nuclei of zeolite A. However, it is also clear that in many cases species are observed spectroscopically that are never seen in the zeolitic products.Knight et al.65 in their recent overview of NMR studies of zeolite crystallisations suggest that the silicate species observed by NMR are merely spectator species and do not condense to form the zeolite framework, and furthermore many of the previous assignments of the silicate species may be in error. Therefore, it is not clear exactly what the relationship is between the species observed spectroscopically and the final zeolitic products, and it seems unlikely that the very small species observable by spectroscopy react directly to form the final product.Davis and Lobo 32 have suggested that it seems more likely that these small species are not directly incorporated into the final lattice, but instead form more extended structures with medium range order that are not observable by short range spectroscopic techniques such as Raman/NMR. It is further suggested that it is these extended structures that participate in the formation of the nucleation centres and then, in turn, the long range structure of the material, whether it be by the solution transport or solid hydrogel transformation mechanism, or a combination of both.Therefore emphasis should be placed on understanding the formation of structures and nucleation centres on larger length scales than those probed by spectroscopy. Recently, a number of such longer length scale studies have been published. These experiments are described in detail in section 6.There is evidence that at least in some cases molecular sieve structures are formed via a layer by layer growth mechanism and that extended sheet structures are important building units in their formation. In particular, this is suggested by the formation of layered intergrowth structures, for example ZSM-5/ ZSM-11,66 FAU/EMT,67,68 and SSZ-26/SSZ-33/CIT-1 48,69 intergrowths, in which it is diYcult to account for these structures by anything other than a layer by layer growth mechanism. The fact that the precise layer stacking sequence can be controlled by systematic manipulation of the organic templating agent adds weight to this view.48,67,70 Finally, Vaughan71 has provided evidence that the presence of sodium in an aluminosilicate gel promotes the formation of faujasite sheets.One of the most detailed studies of the mechanism of formation of a molecular sieve was Burkett and Davis’ study of the formation of pure silica ZSM-5 (Si-ZSM-5 or silicalite) using tetrapropylammonium.50,72,73 The formation of ZSM-5 in the presence of NPr4 1 has long been considered to be a classical example of structure direction in the formation of zeolites, based on the close correspondence between the shape of the cation and the intersection of the channels in ZSM-5, and the tight enclathration of the NPr4 1 within the zeolite post synthesis. 43 In the original report of the synthesis of Si-ZSM-5 it was postulated that the mechanism of structure direction by NPr4 1 was via the preorganisation of silicate species around the organic cation prior to zeolite crystallisation.43 Gies and Marler’s work49 suggested that this interaction was primarily via van der Waals contacts.Subsequent studies by, for example, 29Si MAS NMR were hampered by the inappropriate length scales probed by such techniques, as discussed above. By applying the technique of solid state cross polarised magic angle spinning Burkett and Davis 50 could probe in detail, on the appropriate length scales, the interactions occurring before and during the crystallisation.By performing 1H]29Si CP MAS NMR between the protons of the NPr4 1 and the silicon atoms of the zeolite precursors they were able to study the interaction between the organic and inorganic components. They found that short range intermolecular interactions (i.e. of the order of van der Waals interactions) are established in the synthesis gel before the development of long range order indicative of a crystalline solid.Furthermore the NMR data suggested that the NPr4 1 adopt a conformation within the composite organic– inorganic zeolite precursor similar to that which they have in the final product. Burkett and Davis’ work provides the first direct evidence of preorganised organic–inorganic composite structures during the synthesis of Si-ZSM-5, and is consistent with a mechanism of structure direction in which these organic–inorganic composite structures from the precursors to the formation of the ZSM-5 channel intersections.Further, more detailed, work on the NPr4 1/Si-ZSM-5 system allowed them to refine their proposed mechanism.72–74 As shown in Fig. 4 they suggested that the formation of the organic–inorganic composite species is initiated by overlap of the hydrophobic hydration spheres around the NPr4 1 cation 75 and hydrophobically hydrated domains of soluble silicate species. (Hydrophobic hydration is the reorientation of water molecules in the vicinity of a hydrophobic solute species in order to accommodate them whilst still maintaining a fully hydrogen bonded network.) 76 This allows the establishment of favourable van der Waals contacts between the alkyl chains of the NPr4 1 and the hydrophobic silica species, whilst at the same time allowing the release of the water molecules from the ordered hydration spheres around the NPr4 1 and silica species.This process provides both an enthalpic and entropic driving force for the formation of the organic–inorganic species, which provide the precursor units for the formation of the final crystalline product.Furthermore, Burkett and Davis postulate that crystal growth occurs via diVusion of these composite species to the growing crystalline surface in a layer-by-layer growth fashion which is consistent with known layered intergrowth structures such as ZSM-5/ZSM-1166 and SSZ-26/SSZ-33/CIT- 1 48,69 intergrowths.Burkett and Davis’ proposed mechanism which incorporates all these ideas is shown in Fig. 4. Further insight was provided by studying the formation of Si-ZSM-5 and Si-ZSM-48 using hexanediamine as the organic structure directing agent. At lower temperatures (120 8C) ZSM-5 is formed, and interactions are again seen between the inorganic and organic components, suggesting that organic– inorganic composite units form prior to the formation of a long range ordered material.At higher temperatures ZSM-48 is formed and in this case no interactions are seen between the inorganic and organic components, suggesting that structure direction is not occurring in the same way in this case. Burkett and Davis suggest that the higher temperature could disrupt the hydrophobic hydration spheres around the hexanediamine molecules. This result is interesting in the light of the fact that Si-ZSM-48 can be synthesized in the presence of many amines suggesting that the amines are playing a space filling role, whereas Si-ZSM-5 can only be synthesized in the presence of NPr4 1 or hexanediamine suggesting that the amines are playing a specific structure directing role.This raises the question of whether the ability to form a hydrophobic hydration sphere is a pre-requisite for a species to serve as a structure directing agent, whether the formation of a hydrophobic hydration sphere is simply a reflection of other properties that make an organic molecule an eVective structure directing agent, or the correlation is simply coincidental due to the relatively small number of syntheses studied so far (seven in the Burkett/Davis study).It also suggests that the synthesis of aluminosilicate ZSM-5, which is known to be eVected by a large variety of organics, is formed by a diVerent mechanism in which the aluminosilicate gel chemistry is the controlling factor rather than organic– inorganic interactions.The mechanisms of formation of aluminophosphate and related classes of molecular sieves have been less thoroughly studied. It is generally considered that the first stage in the synthesis of aluminophosphate materials is the reaction ofJ. Chem. Soc., Dalton Trans., 1998, 3133–3148 3139 Fig. 4 Schematic diagram of Burkett and Davis’ proposed mechanism for the formation of Si-ZSM-5 (adapted with permission from ref. 73). H H H H H H H H H H H H H H H H H H H H H H Si Si Si Si Si Si Si Si O O O O O O O O O O O O O O O OO O O O O O Si Si Si Si Si Si Si hydrophobic hydration Si overlap of hydrophobic hydration spheres formation of composite species nucleation crystal growth N+ N+ the aluminium containing starting material (usually pseudoboehmite or an aluminium alkoxide) with the phosphoric acid to form an amorphous aluminophosphate layer.77,78 The next stage in the process is less clear, however.Some authors have argued that a direct solid-state transformation occurs to give the final crystalline product.Others have argued that complete dissoluton of the aluminophosphate layer occurs to produce small solution phase building units, similar to those postulated to be involved in zeolite syntheses, which subsequently condense to form the final products.20 However, as with the synthesis of zeolites, although there is some evidence for the existence of aluminophosphate entities in solution, there is little evidence that these species are the direct precursors of the final crystalline product.More recently, Ozin and co-workers 79 have proposed a model for the formation of aluminophosphates in which two- and three-dimensional structures are formed via hydrolysis and condensation of an initial chain structure which forms first in solution. However, whilst there is compelling evidence for the transformation of chain structures to layered structures in some systems studied,80 it is far from clear at this time that this is a general pathway by which aluminophosphate molecular sieves are produced.An alternative approach to the understanding of the formation mechanisms of microporous materials is to use computational techniques and molecular modelling as tools to probe the relationship between the templates used and the framework structures formed.81 Some of the most successful approaches have used molecular mechanics methodologies to study the interactions between particular organic molecules and framework structures.56,82–86 Such studies can lead to a more thorough understanding of how these interactions aVect the eYciency of a particular organic molecule to act as a template for a given host structure, and can therefore be used as a guide for selecting an eVective template for a given target framework structure.The application of these techniques in relation to the synthesis of ZSM-18 has already been referred to.56 More recently, work by Lewis and co-workers 87,88 has made the rational ‘design’ of a target microporous materials a much more realisable goal.They have developed ‘de novo’ molecular design methodology in which potential template molecules are ‘grown’ from an initial seed molecule. Potential templates grown in such a way are then ranked according to the binding energy within a given pore system, which gives a good guide to the likely eVectiveness of a particular organic molecule as a template for that molecular sieve structure.A recent good example of the application of this technique was provided by the Lewis et al.89 synthesis of DAF-5 (a CoAPO with the Chabazite structure). Using a computationally designed 4-piperidinopiperidine molecule as a template they were able to synthesize DAF-5 using short preparation times in the absence of any other microporous phases, a task impossible to achieve using the smaller organic templates known to result in the synthesis of Chabazitic cobalt aluminophosphates.Whilst the experimental studies and computational approaches referred to above have shed light on the processes occurring during hydrothermal syntheses, and revealed details of the mechanisms of crystal nucleation and growth occurring for specific cases, it is clear that, in general, the mechanisms of hydrothermal syntheses are still not well understood. Although progress has been made, the rational a priori ‘design’ of a molecular sieve is in general still diYcult to achieve because of a lack of mechanistic understanding of their synthesis.Given the3140 J. Chem. Soc., Dalton Trans., 1998, 3133–3148 broad range of useful properties that molecular sieves possess a more complete mechanistic understanding of their formation leading to a more rational approach to their synthesis is still a highly sought after goal in materials science. The main diY- culty is the lack of a universal crystallisation mechanism for molecular sieve materials, with the result that each individual synthesis must be studied using techniques such as those described above, which are often laborious, demanding, and sometimes ambiguous. For this reason, increasing use is being made of non-invasive ‘in situ’ studies to probe the course of zeolitic crystallisations which are capable of delivering far more information.Cheetham and Mellot90 have recently reviewed the application of in situ techniques to the study of a wide variety of materials synthesized from sol–gel precursors.Here we review in detail the specific application of in situ techniques to study the hydrothermal crystallisation of molecular sieve materials. 6 In situ measurements of crystallisation The studies described above are all ‘ex situ’ studies. That is, they were performed by periodically removing aliquots of the reaction mixture, quenching the reaction, working up the synthesized products, and finally analysing the products using conventional techniques such as X-ray diVraction (XRD), MAS NMR and SEM. Besides being labour intensive and providing relatively few data points, such techniques inevitably raise the question of whether the reaction is being aVected by the analysis process, and if the species observed in these experiments are really representative of species present in the reaction medium at the time of quenching.In situ real time studies that probe the processes occurring during crystallisation, without the need for quenching, not only allow the continuous monitoring of the reaction (thus vastly increasing the amount of data that can be obtained), but also allow the reactions to be studied under normal reaction conditions.Thus the question of the method of analysis causing unknown structural changes does not arise. The advantages of in situ experiments over ex situ experiments can be summarised as follows. (i) In situ experiments eliminate the need for sample quenching and work-up, during which the sample may undergo significant and indeterminable structural changes.(ii) Since the reaction is monitored continuously the information gained per reaction is vastly higher. In particular, the time resolution that can be routinely achieved in most experiments is much higher than with conventional techniques; which is important for kinetic and mechanistic studies. (iii) In situ studies allow the direct observation of intermediate phases and their subsequent conversion into the final crystalline product.(iv) In situ experiments provide a ready method for easily probing the eVect of changing synthesis parameters such as temperature, pressure, reagents used, and the gel composition, and allow one to monitor the interconversion of phases as the conditions are varied. Unfortunately, the nature of molecular sieve crystallisations means that in situ studies of hydrothermal reactions are a far from routine procedure.The chief diYculty is the necessity of constructing reaction cells that are able to withstand the relatively high temperatures and pressures required for the synthesis of molecular sieve materials, whilst also conforming to the constraints imposed by the environment in which the measurements will be performed, for example a diVractometer or an NMR machine. However, despite these diYculties, a number of in situ studies on the formation of molecular sieves have been published over the last few years.A variety of techniques have been successfully applied, including NMR, IR/ Raman, EXAFS, optical microscopy, light and neutron scattering, X-ray diVraction, neutron diVraction and small angle diffraction. These experiments have shed light on the species involved in nucleation, the mechanisms of nucleation and growth, the kinetics and energetics of growth, the influence of reaction conditions on the course of the reactions, and the observation and identification of intermediate phases. 6.1 In situ spectroscopic and optical techniques The range of spectroscopic and optical techniques that have been used to study molecular sieve synthesis in situ includes NMR,91 IR/Raman,92 EXAFS,93–96 optical and electron microscopy, 97 and light scattering.98 In situ NMR studies are relatively uncommon despite the large number of good NMR nuclei present in molecular sieve materials, e.g. 27Al, 29Si, 31P, 1H, 13C and 15N.This is probably a reflection of the diYculty of designing suitable reaction cells and heating systems for studying crystallisations at high temperatures that are compatible with being contained within an NMR chamber. There have been a number of NMR studies of the unheated gel prior to reaction but, as was discussed above, the relationship of the species observed in these studies and the final zeolitic product have proved diYcult to determine, and hence the value of such studies is limited.An example of a true in situ NMR measurement of a zeolite crystallisation is provided by Shi et al.91 who studied the synthesis of zeolite A at 65 8C, by constructing a rotor which could be sealed and held at relatively high temperatures whilst recording MAS NMR spectra. They studied the time variation of the 27Al and 29Si spectra during the crystallisation. Unfortunately, during the nucleation period both spectra remained unchanged reflecting the fact that the NMR technique is insensitive to the length scales at which changes are occurring during this period of the reaction.However, they were able to observe that during the growth of the zeolite the NMR lines narrowed (see Fig. 5) indicating that NMR can be used to probe the onset of long range ordering within a growing zeolitic structure. Taulelle and co-workers 99,100 have also developed apparatus capable of studying hydrothermal syntheses in situ using NMR spectroscopy. Using specially designed NMR tubes capable of withstanding high temperatures and pressures 99,101 they have been able to extend the range of conditions that can be studied, and have used the technique to observe the formation of solution phase species during the synthesis of zeolites under true hydrothermal conditions.They have demonstrated that it is possible to achieve suYcient time resolution using this experimental technique to perform kinetic studies and observe transient intermediate phases that are formed during the syntheses.There have been very few in situ studies of hydrothermal syntheses by IR or Raman techniques, although there have been numerous in situ IR studies of catalysis by molecular sieves. In part, this again reflects the diYculty of constructing suitable cells, but also the presence of unwanted scattering from typical zeolitic gels that hampers observation. The principal work of interest in this area was performed by Twu et al.92 who studied the synthesis of faujasite zeolites, and the eVects of changing the silica source on the reaction.They were able to identify the species forming in solution, as well as the solid zeolite product as it crystallised from the aluminosilicate gel. The main point of interest was that they were able to detect the crystallisation of the zeolite well before it could be detected in powder X-ray diVraction patterns. The work of Twu et al. was not really in situ as such, since the measurements were performed on gels that had been separated from solution by centrifugation, but they are illustrative of the kind of information that could be obtained by in situ IR and Raman spectroscopies.In situ light scattering studies are a potentially valuable technique for studying hydrothermal syntheses. This is because they provide information about the very small particles (<200 Å) present in molecular sieve gels that are believed to play an important role in the initial nucleation stage of a zeolitic crystallisation, a process about which very little is understood.A recent example of the application of in situ light scattering techniques was a study of the synthesis of NPr4 1-silicaliteJ. Chem. Soc., Dalton Trans., 1998, 3133–3148 3141 reported by Schoeman.98 In contrast to an earlier ex situ study by the same author,102 by using a laser as the light source he was able to study the synthesis in a non-invasive manner.He found that there were two distinct populations of particles present in the reaction gel; subcolloidal silicate particles with a diameter of ca. 33 Å which remained present throughout the crystallisation and remained essentially unchanged in size, and a second population of larger crystallites which were detected at later times in the reaction. These larger particles were found to grow in size as the reaction proceeded, and correspond to growing silicalite crystals. Evidence was obtained that the larger crystals may grow from the subcolloidal particles, and that the subcolloidal particles may possess short range order, i.e.they may be considered to be zeolite nuclei. A similar earlier study was reported by Twomey et al.103 who studied the influence of various synthetic parameters on the silicalite system and were able to distinguish between the nucleation and growth stages of the synthesis. Fig. 5 In situ 27Al MAS NMR spectra of a zeolite A synthesis from a gel at 65 8C showing the narrowing of linewidths as long range order is established (Figure reproduced with permission from ref. 91). Useful information can also be gained about the kinetics and energetics of crystal growth by using optical microscopy. For example, Iwasaki et al.97 described methods for monitoring the growth of zeolites by in situ optical microscopy from both clear solutions and gels, and illustrated these by studying the growth of ZSM-5 and silicalite.The disadvantage of this technique is the limited resolution obtainable using optical techniques. This disadvantage could be overcome by performing electron microscopy, but such experiments are hampered by the diYculty of constructing suitable reaction cells. This obstacle has proved to be insurmountable to date. In situ EXAFS studies are discussed at the end of the following section due to their frequent combination with in situ diVraction studies. 6.2 In situ diVraction and scattering studies Most in situ studies performed over recent years have been diffraction or scattering studies using X-rays or neutrons.This reflects both the increased availability of high flux X-ray and neutron sources which are generally necessary for this work (see below), and also that many of the questions relating to the synthesis of molecular sieve materials, such as the kinetics and mechanisms of nucleation and growth, the existence of intermediate phases, and the eVects of changing reaction variables, are readily amenable to investigation by diVraction and scattering techniques.Barnes and co-workers first described the potential advantages of in situ diVraction studies over conventional studies in a series of seminal papers in the early 1990s.104–107 Since then a number of such studies have been published, which have began to shed light on some of the issues relating to the kinetics and mechanisms of molecular sieve crystallisations. 6.2.1 In situ X-ray diVraction studies. Although some in situ X-ray diVraction studies have been performed using conventional laboratory equipment, the vast majority of studies have made use of high flux X-ray and neutron sources, in particular synchrotron X-ray sources. The extremely high flux at high energies of such sources allows the X-rays to penetrate the thick walls of typical reaction cells without significant attenuation, and enables the collection of high quality diVraction data using very short acquisition times, even when the cells are constructed of, for example, several millimetres of steel.There have been two very distinct ways in which in situ X-ray diVraction experiments have been performed; in the energy dispersive diVraction mode using ‘white’ polychromatic radiation,104–111 and angular dispersive diVraction mode using a monochromatic beam.112–123 In energy dispersive X-ray diVraction (EDXRD) full spectrum polychromatic (‘white’) radiation emanating from the synchrotron source is allowed to impinge on the reaction cell, and an energy discriminating detector is employed which is held at a fixed angle of 2q.Thus, diVraction from diVerent Bragg reflections is separated by energy coordinate, instead of 2q spatial coordinate as in conventional X-ray diVraction. DiVraction occurs from a lozenge-shaped diVraction volume which is defined by the collimation geometry and the scattering angle 2q as shown as an inset in Fig. 6.The principal advantages of the EDXRD technique are twofold. First, by employing the entire spectral range of the radiation produced by the synchrotron source, the total flux used in the experiment is extremely high. This allows the use of very short acquisition times (of the order of seconds in some cases) whilst still obtaining very high quality data. Secondly, the fixed angle geometry of the EDXRD technique simplifies reaction cell design because only very small windows are required for the incident and diVracted beams.This allows the construction of reaction cells with sophisticated environmental control systems in which a large volume of sample can be kept at a controlled temperature and pressure. This is important because it allows3142 J. Chem. Soc., Dalton Trans., 1998, 3133–3148 Fig. 6 Schematic diagram of the experimental apparatus used for in situ EDXRD studies of hydrothermal reactions. The apparatus is designed for use on station 16.4 of the UK SRS at Daresbury Laboratory.To Exhaust System Bleed System Solid State Energy Discriminating Detector Diffracted Beam Bursting Disc Pressure Relief Valve Steel Cell Incident White X-ray Beam Copper Heating Block Stirrer Motor Lozenge 2q Pressure Transducer one to perform reproducible experiments under conditions very similar to those employed in conventional zeolite syntheses. The chief disadvantage to the EDXRD technique is the rather low resolution of the data obtained, which is caused by the poor energy resolution obtainable using presently available solid state detectors, and the unrefinability of the data obtained.This can particularly cause problems if the materials being studied are low symmetry, or if structural refinement of the data is desirable. The first study of a zeolite crystallisation using in situ EDXRD was performed by Barnes and co-workers 106 in 1992. By adapting an environmental cell which they had previously used for studying the hydration of cement,105 they were able to study the crystallisation of zeolite A and sodalite from an amorphous aluminosilicate gel and kaolinite respectively. Their cell could operate in both open and closed modes, and temperature regulation was achieved by circulating either water (for temperatures up to 95 8C), or silicone oil (for temperatures up to 110 8C).The inner lining of the cell was coated to provide an inert lining to contain the corrosive alkaline solutions, and the cell could be spun to try to prevent sample settling.Despite the ground-breaking nature of this work, there were a number of limitations to the cell design. Primarily, these related to the fairly unsophisticated nature of the sample environment control system and sample containment system. Most importantly, the maximum temperature at which syntheses could be studied was fairly low (they did not study any syntheses at temperatures greater than 110 8C).Most zeolitic syntheses are performed at between 130 and 200 8C, with some being carried out at up to 250 8C. In addition the question of sample settling and accurate sampling of the reaction mixture by the synchrotron beam was not really addressed. Over the past few years we have developed a large volume (ca. 30 ml) hydrothermal reaction cell suitable for collecting in situ, time-resolved EDXRD data under autogenous pressures of up to ca. 30 bar and temperatures of up to 250 8C.The high time resolution and high quality of data that can be achieved using this system allow one to follow the structural changes that occur during hydrothermal crystallisations in real time. A schematic illustration of the cell and the experimental set-up is shown in Fig. 6; a full description of the experiment has been published previously.108 We have used this facility to study a number of hydrothermal syntheses. Fig. 7 shows a three-dimensional plot of the crystallisation of the pyridine templated aluminophosphate [Al2P3O12H2(Hpy)] 124 at 180 8C and is illustrative of the type Fig. 7 Three-dimensional plot showing the EDXRD spectra as a function of time during the synthesis of a pyridine templated aluminophosphate. Each spectrum was acquired in 300 s. Also shown inset is a single 300 s EDXRD spectrum of the final observed product.J. Chem. Soc., Dalton Trans., 1998, 3133–3148 3143 Fig. 8 Schematic representation of the proposed mechanism of formation of TMA-SnS-1.(i) Dissolution of the solid reagents to form solution phase ions such as [Sn2S6]42. (ii) Condensation of these ions under the influence of the template to form a disordered layered material. (iii) Ordering between adjacent tin sulfide sheets to give a material with crystallographic registry in all three spatial dimensions. Stages (i) and (ii) are rapid and take place on a timescale of minutes to hours, whereas stage (iii) is much slower and takes place on a timescale of days (Figure reproduced with permission from ref. 10). and quality of data that can be obtained. Each spectrum was collected in 300 s and an individual spectrum is shown inset in Fig. 7. After an induction time of 10 min when only a broad amorphous background is observed, the aluminophosphate product crystallises smoothly from the reaction as shown by the steady growth of the Bragg diVraction peaks. No intermediate crystalline products are seen.Integration of the intensities of each of the observed Bragg reflections using a Guassian fitting routine shows that each reflection grows at the same rate, and obeys second order Avrami kinetics of order 2, i.e. the intensities can be fitted to the expression a = 1 2 exp[2k(t 2 to)]2, where a = fraction of crystallised material, k = rate constant, t = time, and to = time of onset of crystallisation. Of particular note is the unexpected rapidity of the reaction. In the original reported synthesis the reagents are heated for between 36 and 48 h, whereas our experiments indicate that the crystallisation is complete after only ca. 1 h.This unexpected rapidity is a common feature of the systems we have studied to date. Another system we have investigated is the synthesis of the open-framework tin sulfide TMA-SnS-1 [empirical formula (NMe4)2Sn3S7?xH2O] under basic hydrothermal conditions.110 In this case the time-resolved X-ray data revealed highly anisotropic growth of the TMA-SnS-1 product.This is a layered material in which the tin sulfide microporous layers lie in the 001 plane. It was found that the 002 Bragg reflection corresponding to the interlamellar separation appeared first and grew very rapidly in the initial stages of the reaction. Other hk0 and hkl reflections appeared later and grew much more slowly. In particular, the 111 reflection corresponding to interlamellar ordering was still increasing in intensity after 62 h of reaction, well after the 002 reflection had reached maximum intensity.This suggests that ordering along the direction perpendicular to the layers occurs at a much faster rate than the ordering either within the layers or between adjacent layers. Combined with kinetic data on the initial growth of the 002 reflection which show that the rate of crystallisation is dependent on the initial pH of the reaction mixture, these data can be used to provide an overall picture of the mechanism of formation of TMASnS- 1 involving three stages: (1) initial dissolution of the starting materials by OH2 to give solution phase ions such as [Sn2S6]42, which has been suggested as the dominant solution phase ion present in these syntheses,125,126 (2) rapid condensation/ polymerisation of these ions under the influence of the template to form a layered but poorly ordered material in the initial stages of the reaction, and (3) a much slower process in which registration between tin sulfide sheets occurs to give a crystalline product with crystallographic order in all three spatial directions.A schematic diagram of this process is shown in Fig. 8. More recently we have been studying the synthesis of members of the ULM-n family of materials,20 in particular one member, ULM-5, a microporous oxy-fluorinated gallophosphate containing very large 16 T-atom pores.23,127 It is synthesized hydrothermally via the reaction of gallium oxide, a phosphorus source, hydrofluoric acid and the templating agent 1,6-diaminohexane.These studies have highlighted the critical importance of the form of the phosphorus source on the reaction pathway and the kinetics of crystal growth.111 Using phosphoric acid ULM-5 is found to crystallise smoothly from the amorphous starting materials at an extremely rapid rate (halflife ca. 1 min) after a short induction time. No other crystalline materials are observed at any stage, and the reaction is complete in ca. 40 min. Kinetic analysis of the growth of diVraction peaks indicates that the crystallisation of ULM-5 under these conditions is an essentially diVusion controlled process, and that the rate of crystallisation is independent of temperature over the range 130–200 8C.128 In marked contrast, when phosphorus pentaoxide is used as the phosphorus source the crystallisation of ULM-5 proceeds at a much slower rate via a previously unobserved crystalline intermediate phase which subsequently reacts to form ULM-5.111 The course of the reaction at 180 8C is shown as a three-dimensional plot in Fig. 9, together with a plot of the integrated intensity of the 13.1 Å reflection and the ULM-5 final product 002 reflection.The growth and decay of the intermediate Bragg reflections and the product Bragg reflections are highly correlated which, although not conclusive, strongly suggests that the two phases are related and the intermediate phase converts directly into the final product.Recent in situ results have revealed that there are in fact two diVerent intermediate phases, and that in any particular reaction ULM-5 may form via either one or other of the two intermediate phases exclusively, or a mixture of both phases.129 The relative proportion of each intermediate phase formed is found to be critically dependent on the precise quantity of phosphorus pentaoxide, hydrofluoric acid and 1,6-diaminohexane used in the reaction.128 The rate of conversion of the intermediate to product is strongly dependent on the temperature of reaction, with the transformation taking place at a much slower rate as the temperature is lowered.The determination of the composition and structure of the intermediate phases is clearly of great interest with regard to the mechanism of formation of ULM-5, and eVorts are continuing in this regard. Very recent in situ experiments have revealed that under certain conditions the syntheses of ULM-3,3144 J.Chem. Soc., Dalton Trans., 1998, 3133–3148 Fig. 9 Three-dimensional plot showing the evolution of the EDXRD spectra as a function of time during the synthesis of ULM-5 using phosphorus pentaoxide as a starting material. The peak labels correspond to the indices of the Bragg reflections of ULM-5. Each spectrum was acquired in 60 s at a diVraction angle of 2q = 1.228. Also shown is a plot of the intensity of the (100) Bragg reflection of the intermediate (r) and the (002) Bragg reflection (d) of ULM-5 as a function of time at 180 8C.Inset: plot of intensity of the (100) reflection of the intermediate at 180 and 150 8C. -4 and -16 also proceed via crystalline intermediate phases.130 Given the known structural relationships between these phases,20 this raises the fascinating prospect that the syntheses of these materials may proceed via structurally related phases, allowing one to build a coherent picture of their formation mechanisms.These experiments also dramatically illustrate the complexity of hydrothermal syntheses, and how subtle changes in the reaction conditions, such as the use of slightly diVerent starting materials or reactant ratios, can greatly aVect the kinetics and mechanisms of these reactions. It also demonstrates the power of in situ EDXRD techniques quickly and eYciently to gain unique information about these reactions and the factors aVecting them.Rey et al.131,132 and Davies et al.133 have also made use of the same experimental set-up to study hydrothermal crystallisations in situ. To date, they have focused on the synthesis of metal substituted aluminophosphates (MeAlPOs). Rey et al. studied the template-mediated formation of AlPO-5 and its cobalt substituted derivative CoAlPO-5. They found that whilst AlPO-5 crystallised directly from the reaction mixture, in the CoAlPO-5 case competitive formation of a metastable chabazite type phase occurred.131,132 Further work by Davies et al.showed that the competitive formation of chabazite only occurs for reactions with a Co/P ratio of above ca. 0.04. Turning their attention to kinetics, they found that the rate of crystallisation of CoAlPO-5 increased with increasing cobalt content up to the critical Co/P ratio of 0.04, possibly due to the competitive formation of the chabazite phase, and both AlPO-5 and CoAlPO-5 crystallised at an increasing rate with increasing temperature.Relative to AlPO-5 it was found that cobalt and manganese substituted AlPO-5 (Co/P = 0.04 and Mn/P = 0.04) crystallised faster than unsubstituted AlPO-5, whilst silicon substituted AlPO-5 (SAPO-5, Si/Al = 0.04) crystallised at a slower rate. Davies et al. attributed this somewhat surprising result to the fact that metals substitute exclusively for Al31, whereas silicon has been found to substitute for both Al31 and P51 resulting in silicon island formation.133 In contrast to the energy dispersive diVraction experiments reported above, the angular dispersive diVraction technique makes use of monochromated X-ray radiation, and the diVerent Bragg reflections are split by spatial coordinate rather than energy coordinate.The chief advantage of using constant wavelength monochromatic radiation is that the resolution obtained can be much higher, and the data obtained are suitable for structure refinement. The possibility of performing timeresolved Rietveld refinement thus becomes available.134–136 The chief disadvantage of the angular dispersive technique is that the much lower flux of a monochromatic beam does not allow the construction of large volume cells with bulky sample environment control systems.Generally, angular dispersive in situ X-ray diVraction experiments have employed a system in which an external pressure is applying to a capillary containing the sample, and the reaction occurs by heating a small zone of the sample.This inevitably raises questions about whether such a system is operating under true hydrothermal conditions, and about the reproducibility of such experiments. Norby and co-workers 116,119–121 have developed a facility to study hydrothermal reactions in situ using angular dispersive X-ray diVraction for use on beamline X7B at the NSLS at Brookhaven, USA. Samples are contained in 0.7 mm capillaries mounted in a Swagelock fitting with a Vespel ferrule.Pressure is then applied to the capillary from an external source (usually a nitrogen cylinder), and a hot air stream is used to heat a part of the sample approximately 5 mm wide (much smaller than the height of the beam). A schematic diagram of the experimental set-up is shown in Fig. 10. Norby et al.119 have used this experimental set-up to study a number of hydrothermal syntheses. In a study of the synthesis of CoAPO-5 they studied the temperature dependence of the crystallisation rate.By fitting the crystallisation curves by using the Avrami equation, a = 1 2 exp[2k(t 2 to)]n (n = the order of the reaction, all other symbols as above), they were able to show that the reaction obeys first order kinetics, i.e. the nucleation rate has very little influence on the reaction. The formation of magnesium aluminophosphate was also studied and the influence of the templates on the final products formed investigated.115 Another study investigated the hydrothermal conversion of zeolite LTA into zeolite Li-A(BW) and LiAlSiO4?H2O using LiCl, with a view to determining the mechanism of transformation.121 Fig. 11 shows a three-dimensional plot of the evolution of the X-ray pattern as a function of time and shows the decay of the Bragg reflections of zeolite LTA and the corresponding growth of Bragg reflections of zeolite Li-A(BW). It was found that the conversion is a solution mediated process, but that only very small amounts of aluminosilicate are in the solution phase.There was no evidence for the formation of any amorphous phases during the course of the reaction. Using the same experimental apparatus Morris et al.118 studied the formation of single crystals ofJ. Chem. Soc., Dalton Trans., 1998, 3133–3148 3145 Fig. 10 Schematic diagram of the experimental apparatus developed by Norby and co-workers for studying hydrothermal reaction in situ using time resolved angular dispersive X-ray diVraction. Fig. 11 Three-dimensional plot of the evolution of the in situ angular dispersive X-ray powder diVraction profiles with time during the hydrothermal conversion of zeolite LTA into zeolite Li-A(BW). Data recorded at a wavelength of 1.1727(2) Å and a temperature of 200 8C (reproduced with permission from ref. 121). the clathrasil dodecasil-3C under solvothermal conditions. The high time resolution that they were able to achieve using this system allowed the accurate fitting of rate expressions to the experimental data, and thus to study the crystallisation kinetics of dodecasil-3C in detail.Although not strictly examples of hydrothermal crystallisations, in situ angular dispersive techniques have also been used to study the structural changes that occur during the dehydration of molecular sieve materials. A good example is a study of the dehydration of the natural zeolite laumonite performed by Stahl et al.122,137 Powder diVraction profiles were collected at time intervals of 5 min as the temperature was raised in steps of 5 K from 310 to 584 K.Rietveld refinements of each of the profiles proved possible and thus a dynamic picture of the dehydration process and the host structure’s response to the dehydration was obtained. More recently Norby et al.138 have reported a study of the migration of Na1 and Cs1 cations within the cavities of zeolite Cs(Na)-Y during dehydration. Time-resolved Rietveld refinement of the in situ data allowed the cation populations of the various sites within the zeolite framework to be determined with excellent precision throughout the entire dehydration process. 6.2.2 In situ neutron diVraction studies. There have been very few in situ neutron diVraction studies of hydrothermal syntheses. This is despite the fact that neutrons have low absorption cross-sections relative to X-rays and thus they appear suited for studying reactions carried out in special environmental cells.Of course, neutrons also have the advantage that they can be used to probe the behaviour of light atoms in the presence of heavy atoms. Unfortunately, the successful application of in situ neutron diVraction techniques is severely limited by the low fluxes obtainable at neutron sources when compared to synchrotron X-ray sources. Consequently the collection of high quality spectra requires acquisition times that are inappropriately long for in situ time-resolved studies.Therefore, in spite of the potential advantages, there has only been one report of an in situ neutron diVraction study of a hydrothermal synthesis of3146 J. Chem. Soc., Dalton Trans., 1998, 3133–3148 a molecular sieve material (zeolite A).139 Nevertheless, future advances in instrument design may reduce the time needed for the collection of powder neutron data,90 and thus make the collection of time-resolved neutron diVraction data a more technically realisable goal, especially for syntheses that take place over a period of hours to days rather than minutes. 6.2.3 In situ X-ray and neutron scattering studies. Despite the power of in situ diVraction methods as a tool for understanding hydrothermal syntheses, they are, of course, probes of long range order. They are therefore insensitive to the structures of nucleation centres that are formed in the initial stages of the reaction and which are too small to be observed by diVraction. Small angle X-ray and neutron scattering studies (SAXS and SANS) provide information about the size, shape and fractal dimension of aggregates of particles in the size range 1–1000 Å.140 They therefore complement diVraction experiments by probing much shorter length scales, and obtain information about the critical nucleation step of crystallisations, before long-range order has been established.In addition, it is possible to use isotopic substitution to modify the scattering length densities of the various components (inorganic, template and solvent) of the reaction mixture.Of particular importance is the ability to label the organic template with deuterium and thus be able to diVerentiate between purely inorganic particles and those incorporating template molecules. Two groups have made extensive use of in situ SAXS/SANS experiments to study the formation of ZSM-5 and its pure silica analogue silicalite. White and co-workers 60,141–143 investigated the room temperature ageing and crystallisation of both ZSM-5 and silicalite.The SANS experiments revealed that on mixing of NPr4 1 and soluble silicate species the cations are rapidly incorporated into amorphous NPr4 1/silicate structures.60 Such a process is consistent with the formation of composite NPr4 1/ silicate species in the mechanism of silicalite formation proposed by Burkett and Davis.50 Further studies focussed on the changes that occur during the ageing process (the elapsed time from mixing of the reagents before the onset of heating).141 They found that particles in the size range 10–50 Å formed in the first 60 h and their numbers grew until an equilibrium state was reached.Interestingly, in the absence of aluminium no such process occurred suggesting that the aluminium acts as a cross linking agent and drives the formation of loose networks of aluminosilicate species. The SANS measurements showed that these particles did not contain template molecules, and they could not therefore be considered to be crystal nuclei. However, in recent subsequent studies of the high temperature crystallisation of silicalite 142,143 the same group have detected the growth of particles of 50–100 Å in size during the induction period of the crystallisation.The SANS measurements indicated that these particles have a composition close to that of silicalite and contain NPr4 1, which strongly suggests that these are the nuclei for the growth of silicalite crystals.They propose that the crystallisation occurs by a process whereby these nuclei assemble preferentially along the c axis to form 330 Å cylindrically shaped ‘primary crystallites’, which subsequently aggregate to form polycrystalline particles of approximate length 6000 Å. Van Santen and co-workers 144–147 have also performed extensive SAXS/WAXS and SANS experiments on the silicalite/ ZSM-5 system. They also observed the formation of amorphous particles in the 50–70 Å range both prior to and during the formation of crystallites with long range crystallographic order.144,145 However, additionally, they obtained evidence that a structural reorganisation of the gel occurs prior to crystallisation, and proposed a mechanism of formation of silicalite involving additional aggregation and densification steps.146 Their proposed mechanism can be summarised as follows: (1) the formation of small NPr4 1-silicate clusters less than 16 Å in size, (2) aggregation of these clusters into amorphous particles ca. 60–70 Å in size, (3) densification of these 60–70 Å particles, and (4) combination of these to densified aggregates into crystallites containing long range order. Initially a secondary aggregation step prior to the formation of the final crystalline product was proposed,146 but recent studies indicate that this step does not in fact occur.147 It was also shown that the formation of the aggregate particles is dependent on the Si/OH ratio used in the synthesis.No aggregate particles were observed when the reactions were performed under conditions of high alkalinity (Si/OH = 2.4). 6.2.4 In situ EXAFS/XRD studies. Extended X-ray absorption fine structure (EXAFS) is a potentially very powerful method of studying zeolitic syntheses because it enables the determination of the environment around an atom (coordination number, bond lengths and angles) even when that particular element is only present in small quantities and is part of an amorphous or highly disordered phase.Thus, it is highly suited to the study of zeolitic gels prior to crystallisation. The use of a synchrotron as the X-ray source is highly desirable because it provides a tuneable, broad-band source of X-rays over a wide spectral range making it possible to study several diVerent elements within the same reaction system. It is especially advantageous to record a combination of in situ EXAFS and in situ XRD measurements on the same reaction simultaneously, because this allows one to correlate the changes in both the short and long range order which occur as the system undergoes nucleation and crystallisation.Couves et al.96 first demonstrated the power of such a combined EXAFS/ XRD approach with a detailed study of the synthesis of an active Cu/ZnO catalyst from a powdered precursor phase. A disadvantage of this original design was that it was performed in energy dispersive mode which makes collection and analysis of high quality EXAFS spectra diYcult.Clausen et al.93 first developed a combined EXAFS/XRD experimental method in the quick scanning mode (QuEXAFS) which does not have these drawbacks. They used this system to study the calcination and reduction of Cu based methanol catalysts. These studies provided information about the presence of possible intermediate phases during the calcination and reduction processes. More recently Sankar et al.94 also reported the construction of a combined QuEXAFS/XRD cell.They demonstrated the application of the technique with a study of the crystallisation of CoAlPO-5 in which they observed a change in the coordination of the Co21 ion from octahedral to tetrahedral immediately prior to crystallisation.148 7 Conclusion This review brings together the current thinking regarding the kinetics and mechanisms of the formation of microporous materials. It is clear that, despite the complexity of these reactions, great strides are being made towards understanding the processes by which molecular sieves are synthesized.In particular, the range of in situ experiments that are currently being developed are beginning to demonstrate their power for elucidating in detail the processes occurring during hydrothermal syntheses. In situ techniques now span the full range of atomic length scales from short range probes (NMR and EXAFS), to medium range probes (SAXS/SANS), up to probes of long range order (X-ray and neutron diVraction).Continuing experimental developments are likely to bring further insights. In particular, combined multitechnique in situ experiments will also be of great importance since they allow a range of length scales to be probed simultaneously. Thus, although to date no new microporous structure type has been designed and synthesized ab initio, the developments described above suggest that such a breakthrough will be achieved in the near future.J.Chem. Soc., Dalton Trans., 1998, 3133–3148 3147 8 Acknowledgements We would like to thank the EPSRC and the Leverhulme Trust for financial support, and Daresbury Laboratory for access to the Synchrotron Radiation Source (SRS). We also thank the science and technical staV at the SRS for helping in the design, construction and operation of the EDXRD cell. Finally, we would like to thank the many research colleagues and collaborators that have been involved in the EDXRD experiments described above. 9 References 1 R.M. Barrer, J. Chem. Soc., 1948, 127. 2 W. M. Meier, D. H. Olsen and C. Baerlocher, Atlas of Zeolite Structure Types, 4th edn., Elsevier, Amsterdam, 1996. 3 R. M. Barrer, Hydrothermal Chemistry of Zeolites, Academic Press, New York, 1982. 4 S. T. Wilson, B. M. Lok, C. A. Messina, T. R. Cannan and E. M. Flanigen, J. Am. Chem. Soc., 1982, 104, 1146. 5 S. T. Wilson, B. M. Lok, C. A. Messina, T.R. Cannan and E. M. Flanigen, ACS Symp. Ser., 1983, 218, 79. 6 E. M. Flanigen, R. L. Patton and S. T. Wilson, Stud. Surf. Sci. Catal., 1988, 37, 13. 7 M. I. Khan, Y.-S. Lee, C. J. O’Conner, R. C. Haushalter and J. Zubieta, Chem. Mater., 1994, 6, 721. 8 M. I. Khan, Y.-S. Lee, C. J. O’Conner, R. C. Haushalter and J. Zubieta, J. Am. Chem. Soc., 1994, 116, 4525. 9 M. I. Khan, L. M. Meyer, R. C. Haushalter, A. L. Schweitzer, J. Zubieta and J. L. Dye, Chem. Mater., 1996, 8, 43. 10 V. Soghomonian, Q. Chen, R. C. Haushalter, J. Zubieta, C. J. O’Conner and Y.-S. Lee, Chem. Mater., 1993, 5, 1690. 11 V. Soghomonian, Q. Chen, R. C. Haushalter and J. Zubieta, Chem. Mater., 1993, 5, 1595. 12 A. M. Chippindale and R. I. Walton, J. Chem. Soc., Chem. Commun., 1994, 2453. 13 A. M. Chippindale and A. R. Cowley, Zeolites, 1997, 18, 176. 14 J. Chen, R. H. Jones, S. Natarajan, M. B. Hursthouse and J. M. Thomas, Angew. Chem., Int. Ed. Engl., 1994, 33, 639. 15 P. Y. Feng, X.H. Bu and G. D. Stucky, Nature (London), 1997, 388, 735. 16 R. C. Haushalter, K. Strohmaier and F. W. Lai, Science, 1989, 246, 1289. 17 R. C. Haushalter and L. A. Mundi, Chem. Mater., 1992, 4, 31. 18 M. Cavellec, D. Riou and G. Ferey, Acta Crystallogr., Sect. C, 1995, 51, 2242. 19 G. Ferey, T. Loiseau and D. Riou, Mater. Sci. Forum, 1994, 152, 125. 20 G. Ferey, J. Fluorine Chem., 1995, 72, 187. 21 M. Estermann, L. B. McCusker, C. Baerlocher, A. Merrouche and H. Kessler, Nature (London), 1991, 352, 320. 22 T. Loiseau and G. Ferey, Eur. J. Solid State Inorg. Chem., 1993, 30, 369. 23 T. Loiseau and G. Ferey, J. Solid State Chem., 1994, 111, 403. 24 T. Loiseau, R. Retoux, P. Lacorre and G. Ferey, J. Solid State Chem., 1994, 111, 427. 25 T. Loiseau and G. Ferey, J. Mater. Chem., 1996, 6, 1073. 26 T. Loiseau, F. Serpaggi and G. Ferey, Chem. Commun., 1997, 1093. 27 A. M. Chippindale, S. J. Brech, A. R. Cowley and W. M. Simpson, Chem. Mater., 1996, 8, 2259. 28 I. D. Williams, J. Yu, H. Du, J. Chen and W. Pang, Chem. Mater., 1998, 10, 773. 29 B. M. Lok, T. R. Cannan and C. A. Messina, Zeolites, 1983, 3, 282. 30 R. M. Barrer, Zeolites and Clay Minerals as Sorbents and Molecular Sieves, Academic Press, New York, 1978. 31 R. Szostak, Molecular Sieves: Principles of Synthesis and Identification, Van Nostrand Reinhold, New York, 1989. 32 M. E. Davis and R. F. Lobo, Chem. Mater., 1992, 4, 756. 33 D. W. Breck, Zeolite Molecule Sieves: Structure, Chemistry and Use, Wiley, New York, 1974. 34 R. M. Barrer and P. J. Denny, J. Chem. Soc., 1961, 971. 35 F. Y. Dai, M. Suzuki, H. Takahashi and Y. Saito, Stud. Surf. Sci. Catal., 1986, 28, 223. 36 M. E. Davis, C. Saldarriaga, C. Montes, J. Garces and C. Crowder, Nature (London), 1988, 331, 698. 37 B. Duncan, R. Szostak, K. Sorby and J. G. Ulan, Catal. Lett., 1990, 7, 367. 38 L. B. McCusker, C. Baerlocher, E. Jahn and M. Bulow, Zeolites, 1991, 8, 183. 39 M. J. Duer, H. Y. He, W. Kolodziejski and J.Klinowski, J. Phys. Chem., 1994, 98, 1198. 40 S. Prasad and R. Vetrivel, J. Phys. Chem., 1994, 98, 1579. 41 G. Cheetham and M. M. Harding, Zeolites, 1996, 16, 245. 42 C. Baerlocher and W. M. Meier, Helv. Chim. Acta, 1969, 52, 1853. 43 E. M. Flanigen, J. M. Bennett, R. W. Grose, J. P. Cohen, R. L. Patton, R. M. Kirchner and J. V. Smith, Nature (London), 1978, 271, 512. 44 L. D. Rollmann, Adv. Chem. Ser., 1979, 173, 387. 45 R. H. Jones, J. M. Thomas, J. Chen, R.Xu, Q. Huo, S. Li, Z. Ma and A. M. Chippindale, J. Solid State Chem., 1993, 102, 204. 46 R. M. Dessau, J. L. Schlenker and J. B. Higgins, Zeolites, 1990, 10, 522. 47 A. Moini, K. D. Schmitt, E. W. Valyocsik and R. F. Polomski, Zeolites, 1994, 14, 504. 48 R. F. Lobo and M. E. Davis, J. Am. Chem. Soc., 1995, 117, 3766. 49 H. Gies and B. Marler, Zeolites, 1992, 12, 42. 50 S. L. Burkett and M. E. Davis, J. Phys. Chem., 1994, 98, 4647. 51 M. Wiebcke, J. Chem. Soc., Chem. Commun., 1991, 1507. 52 E. Jahn, D. Muller, W. Wieker and J. Richtermendau, Zeolites, 1989, 9, 177. 53 G. Finger, J. Richtermendau, M. Bulow and J. Kornatowski, Zeolites, 1991, 11, 443. 54 B. L. Newalkar, B. V. Kamath, R. V. Jasra and S. G. T. Bhat, Zeolites, 1997, 18, 286. 55 J. Ciric, U.S. Pat., 39 504 496, 1976. 56 A. P. Stevens, A. M. Gorman, C. M. Freeman and P. A. Cox, J. Chem. Soc., Faraday Trans., 1996, 2065. 57 S. Ueda, W. Kageyama and M. Koizumi, Proc. 6th Int. Zeolite Conf., 1984, 905. 58 F. Testa, R. Szostak, R. Chiappetta, R. Aiello, A. Fonseca and J. B. Nagy, Zeolites, 1997, 18, 106. 59 W. Xu, J. Li, W. Li, H. Zhang and B. Liang, Zeolites, 1989, 9, 468. 60 L. E. Iton, F. Trouw, T. O. Brun and J. E. Epperson, Langmuir, 1992, 8, 1045. 61 P. Bodart, J. B. Nagy, Z. Gabelica and E. G. Derouane, J. Chim. Phys., 1986, 83, 777. 62 W. Pang, S. Qiu, Q. Kan, Z. Wu, S. Peng, G. Fan and D. Tian, Stud. Surf. Sci. Catal., 1989, 49, 281. 63 M. E. Davis, C. Montes, P. E.Hathaway and J. M. Garces, Stud. Surf. Sci. Catal., 1989, 49, 199. 64 P. K. Dutta and D. C. Shieh, J. Phys. Chem., 1986, 90, 2331. 65 C. T. G. Knight, R. T. Syvitski and S. D. Kinrade, Stud. Surf. Sci. Catal., 1995, 97, 483. 66 G. R. Millward, S. Ramdas, J. M. Thomas and M. T. Barlow, J. Chem. Soc., Faraday Trans. 1, 1983, 1075. 67 J. P. Arhancet and M. E. Davis, Chem. Mater., 1991, 3, 567. 68 S. L. Burkett and M. E. Davis, Microporous Mater., 1993, 1, 265. 69 R. L. Lobo, M.Pan, I. Chan, H.-X. Li, R. C. Medrud, S. I. Zones, P. A. Crozier and M. E. Davis, Science, 1993, 262, 1543. 70 O. Terasaki, T. Ohsuna, V. Alfredsson, J.-O. Bovin, D. Watanabe, S. W. Carr and M. W. Anderson, Chem. Mater., 1993, 5, 452. 71 D. E. W. Vaughan, Stud. Surf. Sci. Catal., 1991, 65, 275. 72 S. L. Burkett and M. E. Davis, Chem. Mater., 1995, 7, 920. 73 S. L. Burkett and M. E. Davis, Chem. Mater., 1995, 7, 1453. 74 M. E. Davis, Stud. Surf. Sci. Catal., 1995, 97, 35. 75 H.S. Frank and M. W. Evans, J. Chem. Phys., 1945, 13, 50. 76 N. Muller, Acc. Chem. Res., 1990, 23, 23. 77 H. Y. He and J. Klinowski, J. Phys. Chem., 1994, 98, 1192. 78 M. E. Davis and D. Young, Stud. Surf. Sci. Catal., 1991, 60, 53. 79 S. Oliver, A. Kuperman and G. A. Ozin, Angew. Chem., Int. Ed. Engl., 1998, 37, 46. 80 S. Oliver, A. Kuperman, A. Lough and G. A. Ozin, Chem. Mater., 1996, 8, 2391. 81 D. W. Lewis, C. R. A. Catlow and J. M. Thomas, Faraday Discuss. R. Soc. Chem., 1997, 106, 451. 82 R. E. Boyett, A. P. Stevens, M. G. Ford and P. A. Cox, Zeolites, 1996, 17, 508. 83 R. E. Boyett, A. P. Stevens, M. G. Ford and P. A. Cox, Stud. Surf. Sci. Catal., 1997, 105, 117. 84 C. M. Freeman, D. W. Lewis, T. V. Harris, A. K. Cheetham, N. J. Henson, P. A. Cox, A. M. Gorman, S. M. Levine, J. M. Newsam, E. Hernandez and C. R. A. Catlow, ACS Symp. Ser., 1995, 589, 326. 85 T. V. Harris and S. I. Zones, Stud. Surf. Sci. Catal., 1994, 84, 29. 86 D. W. Lewis, C. M.Freeman and C. R. A. Catlow, J. Phys. Chem., 1995, 99, 11 194. 87 D. W. Lewis, D. J. Willock, C. R. A. Catlow, J. M. Thomas and G. J. Hutchings, Nature (London), 1996, 382, 604. 88 D. J. Willock, D. W. Lewis, C. R. A. Catlow, G. J. Hutchings and J. M. Thomas, J. Mol. Catal. A, 1997, 119, 415. 89 D. W. Lewis, G. Sankar, J. K. Wyles, J. M. Thomas, C. R. A. Catlow and D. J. Willcock, Angew. Chem., Int. Ed. Engl., 1997, 36, 2675. 90 A. K. Cheetham and C. F. Mellot, Chem. Mater., 1997, 9, 2269.3148 J.Chem. Soc., Dalton Trans., 1998, 3133–3148 91 J. M. Shi, M. W. Anderson and S. W. Carr, Chem. Mater., 1996, 8, 369. 92 J. Twu, P. K. Dutta and C. T. Kresge, Zeolites, 1991, 11, 672. 93 B. S. Clausen, K. Grabaek, G. SteVensen, P. L. Hansen and H. Topsoe, Catal. Lett., 1993, 20, 23. 94 G. Sankar, P. A. Wright, S. Natarajan, J. M. Thomas, G. N. Greaves, A. J. Dent, B. R. Dobson, C. A. Ramsdale and R. H. Jones, J. Phys. Chem., 1993, 97, 9550. 95 I. J. Shannon, T.Maschmeyer, G. Sankar, J. M. Thomas, R. D. Oldroyd, M. Sheehy, D. Madill, A. M. Waller and R. P. Townsend, Catal. Lett., 1997, 44, 23. 96 J. W. Couves, J. M. Thomas, D. Waller, R. H. Jones, A. J. Dent, G. E. Derbyshire and G. N. Greaves, Nature (London), 1991, 354, 465. 97 A. Iwasaki, M. Hirata, I. Kudo, T. Sano, S. Sugawara, M. Ito and M. Watanabe, Zeolites, 1995, 15, 308. 98 B. J. Schoeman, Zeolites, 1997, 18, 97. 99 C. Gerardin, M. In and F. Taulelle, J. Chim. Phys., 1995, 92, 1877. 100 M. Haouas, C. Gerardin, F. Taulelle, C. Estournes, T. Loiseau and G. Ferey, J. Chim. Phys., 1998, 95, 302. 101 S. D. Kinrade and T. W. Swaddle, J. Magn. Reson., 1988, 77, 569. 102 B. J. Schoeman, J. Sterte and J. E. Otterstedt, Zeolites, 1994, 14, 568. 103 T. A. M. Twomey, M. Mackay, H. P. C. E. Kuipers and R. W. Thompson, Zeolites, 1994, 14, 162. 104 P. Barnes, D. Hausermann and S. E. Tarling, Inst. Phys. Conf. Ser., 1990, 61. 105 P. Barnes, S. M. Clark, D. Hausermann, E. Henderson, C. H. Fentiman, M. N. Muhamad and S. Rashid, Phase Transitions, 1992, 39, 117. 106 J. Munn, P. Barnes, D. Hausermann, S. A. Axon and J. Kilinowski, Phase Transitions, 1992, 39, 129. 107 H. Y. He, P. Barnes, J. Munn, X. Turrillas and J. Klinowski, Chem. Phys. Lett., 1992, 196, 267. 108 J. S. O. Evans, R. J. Francis, D. O’Hare, S. J. Price, S. M. Clarke, J. Flaherty, J. Gordon, A. Nield and C. C. Tang, Rev. Sci. Instrum., 1995, 66, 2442. 109 S. O’Brien, R. J. Francis, S. J. Price, D. O’Hare, S. M. Clark, N. Okazaki and K. Kuroda, J. Chem. Soc., Chem. Commun., 1995, 2423. 110 R. J. Francis, S. J. Price, J. S. O. Evans, S. O’Brien, D. O’Hare and S. M. Clark, Chem. Mater., 1996, 8, 2102. 111 R. J. Francis, S. J. Price, S. O’Brien, A. M. Fogg, D. O’Hare, T. Loiseau and G. Ferey, Chem. Commun., 1997, 521. 112 A. N. Christensen, P. Norby and J. C. Hanson, J. Solid State Chem., 1995, 114, 556. 113 A. N. Christensen, P. Norby and J. C. Hanson, Acta Chem. Scand., 1995, 49, 331. 114 A. N. Christensen, P. Norby, J. C. Hanson and S. Shimada, J. Appl. Crystallogr., 1996, 29, 265. 115 A. N. Christensen, P. Norby and J. C. Hanson, Acta Chem. Scand., 1997, 51, 249. 116 A. Gualtieri, P. Norby, G. Artioli and J. Hanson, Phys. Chem. Miner., 1997, 24, 191. 117 A. Gualtieri, P. Norby, G. Artioli and J. Hanson, Microporous Mater., 1997, 9, 189. 118 R. E. Morris, S. J. Weigel, P. Norby, J. C. Hanson and A. K. Cheetham, J. Synchrotron Rad., 1996, 3, 301. 119 P. Norby, A. N. Christensen and J. C. Hanson, Stud. Surf. Sci. Catal., 1994, 84, 179. 120 P. Norby, Mater. Sci. Forum, 1996, 228, 147. 121 P. Norby, J. Am. Chem. Soc., 1997, 119, 5215. 122 G. Artioli, K. Stahl and J. C. Hanson, Mater. Sci. Forum, 1996, 228, 369. 123 G. Cruciani, G. Artioli, A. Gualtieri, K. Stahl and J. C. Hanson, Am. Miner., 1997, 82, 729. 124 A. M. Chippindale, A. V. Powell, L. M. Bull, R. H. Jones, A. K. Cheetham, J. M. Thomas and R. R. Xu, J. Solid State Chem., 1992, 96, 199. 125 G. A. Ozin, T. Jiang and R. L. Bedard, Adv. Mater., 1994, 6, 860. 126 T. Jiang, A. Lough, G. A. Ozin and R. L. Bedard, J. Mater. Chem., 1998, 8, 733. 127 T. Loiseau, D. Riou, F. Taulelle and G. Ferey, Stud. Surf. Sci. Catal., 1994, 84, 395. 128 R. J. Francis and D. O’Hare, J. Am. Chem. Soc., submitted. 129 D. O’Hare, J. S. O. Evans, R. J. Francis, P. S. Halasyamani, P. Norby and J. Hanson, Microporous Mesoporous Mater., in the press. 130 R. J. Francis, R. I. Walton, T. Loiseau and D. O’Hare, unpublished work. 131 F. Rey, G. Sankar, J. M. Thomas, P. A. Barrett, D. W. Lewis, C. R. A. Catlow, S. M. Clark and G. N. Greaves, Chem. Mater., 1995, 7, 1435. 132 F. Rey, G. Sankar, J. M. Thomas, P. A. Barrett, D. W. Lewis, C. R. A. Catlow, S. M. Clark and G. N. Greaves, Chem. Mater., 1996, 8, 590. 133 A. T. Davies, G. Sankar, C. R. A. Catlow and S. M. Clark, J. Phys. Chem. B, 1997, 101, 10 115. 134 L. Lutterotti, A. Gualtieri and S. Aldrighetti, Mater. Sci. Forum, 1996, 228, 29. 135 A. Gualtieri, P. Norby, J. Hanson and J. Hriljac, J. Appl. Crystallogr., 1996, 29, 707. 136 P. Norby, J. Appl. Crystallogr., 1997, 30, 21. 137 K. Stahl, G. Artioli and J. C. Hanson, Phys. Chem. Miner., 1996, 23, 328. 138 P. Norby, F. I. Poshni, A. F. Gualtieri, J. C. Hanson and C. P. Grey, J. Phys. Chem. B, 1998, 102, 839. 139 E. Polak, J. Munn, P. Barnes, S. E. Tarling and C. Ritter, J. Appl. Crystallogr., 1990, 23, 258. 140 C. G. Windsor, J. Appl. Crystallogr., 1988, 21, 582. 141 J. Dougherty, L. E. Iton and J. W. White, Zeolites, 1995, 15, 640. 142 J. N. Watson, L. E. Iton and J. W. White, Chem. Commun., 1996, 2767. 143 J. N. Watson, L. E. Iton, R. I. Keir, J. C. Thomas, T. L. Dowling and J. W. White, J. Phys. Chem. B, 1997, 101, 10 094. 144 W. H. Dokter, T. P. M. Beelen, H. F. van Garderen, C. P. J. Rummens, R. A. van Santen and J. D. F. Ramsay, Colloids Surf. A, 1994, 85, 89. 145 W. H. Dokter, T. P. M. Bellen, H. F. van Garderen, R. A. van Santen, W. Bras, G. E. Derbyshire and G. R. Mant, J. Appl. Crystallogr., 1994, 27, 901. 146 W. H. Dokter, H. F. van Garderen, T. P. M. Beelen, R. A. van Santen and W. Bras, Angew. Chem., Int. Ed. Engl., 1995, 34, 73. 147 P. E. A. de Moor, T. P. M. Beelen, B. U. Komanschek, O. Diat and R. A. van Santen, J. Phys. Chem. B, 1997, 101, 11 077. 148 G. Sankar, J. M. Thomas, F. Rey and G. N. Greaves, J. Chem. Soc., Chem. Commun., 1995, 2549. Paper 8/02330A

ISSN:1477-9226    

DOI:10.1039/a802330a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1998, 3149–3150 3149 First synthesis of a unique dilead SchiV base complex Pravat Bhattacharyya, Jonathan Parr,* Andrew T. Ross and Alexandra M. Z. Slawin Department of Chemistry, Loughborough University, Loughborough, Leics., UK LE11 3TU Received 3rd August 1998, Accepted 20th August 1998 The reaction of the tripodal Schiff base ligand N[CH2CH2N]] C(H)C6H3(OMe)-3-(OH)-2]3 (H3L) with lead(II) chloride in methanol in the presence of a base led to the bimetallic complex [Pb2L]Cl 1 in good yield; the cation was characterised crystallographically as the perchlorate salt 2, revealing dissimilar environments for the Pb(II) centres with one lead lying within the central cavity of (L)32 and the second bound externally to the three phenolate oxygens of the ligand.Tripodal SchiV base ligands of general formula N[CH2CH2N]] C(H)C6H3(R)OH-2]3 have a well explored coordination chemistry with d- and f-block metals. Complexation of such trivalent metals by triply deprotonated ligands gives either octahedral complexes with [N3O3] donor sets, or for some larger cations additional participation of the apical nitrogen atom gives a seven-coordinate monocapped octahedron with an [N4O3] donor set.1 The p-block metals have, in contrast, received less attention.We have previously demonstrated that polydentate SchiV base ligands support a range of unusual coordination polyhedra at Pb(II):2,3 here we report the preparation of a new bimetallic species [Pb2L]Cl 1, where H3L = tris[2-(3-methoxysalicylideneamino) ethyl]amine, characterised crystallographically as its perchlorate salt 2.The reaction of H3L with an equimolar quantity of lead(II) chloride and triethylamine in refluxing methanol generates exclusively [ Pb2L]Cl 1, isolable in good yield from the reaction as an orange microcrystalline solid.† This contrasts with the corresponding reaction using the unsubstituted tripodal ligand N(CH2CH2N]] CHC6H4OH-2)3 (H3saltren) which gives exclusively [Pb(Hsaltren)].3 Anion metathesis of 1 using Ag[ClO4] or Pb[ClO4]2?3H2O in methanol gives the perchlorate salt [Pb2L][ClO4] 2 (Scheme 1).The 1H NMR spectra of 1 and 2 in d3-acetonitrile at room temperature each have broadened peaks indicating that at this temperature, the molecule is stereochemically non-rigid. They both show two distinct imine resonances in the ratio 2 : 1, suggesting a slow exchange process in which two legs of the tripodal ligand experience a diVerent environment from the third.This situation is highly reminiscent of the fluxional behaviour of [Pb(Hsaltren)], which we have previously ascribed to an interchange involving all three legs of the tripod. Unfortunately, because of the insolubility of 1 and 2 in solvents of suitable melting point, we were unable to obtain VT NMR data over a suYciently wide temperature range to throw any light upon the mechanism of this fluxionality.The solid state structure of 2 has been determined by single crystal X-ray analysis.‡ The two lead(II) atoms in 2 [Fig. 1(a)] exist in dissimilar environments, whereby Pb(1) is bound by all of the nitrogen atoms and phenolate oxygens of L32 and is seven-coordinate. This [PbL]2 unit acts as a tridentate ligand to Pb(2) binding through the phenolate oxygen atoms O(6), O(16) and O(26) to give the dinuclear cation [Pb2L]1, which possesses pseudo-C3v symmetry [Fig. 1(b)]. Complex 2 provides a rare example where all three phenolate oxygen atoms of a tripodal SchiV base ligand bridge two metal centres, a unique instance amongst main group elements. The central [Pb2O3] core of 2 can be considered as a trigonal bipyramid, with Pb(1) and Pb(2) providing the apical vertices. Pb(2) has close contacts with an oxygen atom O(41) from the perchlorate counter ion [Pb(2) ? ? ? O(41) 3.19(1) Å] and the o-methoxy groups of L [Pb(2) ? ? ? OMe 2.69(1)–2.78(1) Å]; the intermetallic distance Pb(1) ? ? ? Pb(2) is 3.558(1) Å.While the o-methoxy groups of L32 have been identified as donor sites in heterodimetallic lanthanide complexes [LnLn9L]- [NO3]3 (Ln,Ln9 = Gd; Ln = Yb, Ln9 = Gd),5 we have previously discounted such bonding interactions in related Pb(II) complexes. 2 The Pb(1)–O and Pb(1)–N distances [2.518(3)–2.595(4) Å and 2.557(5)–2.651(4) Å respectively] are discernibly longer than in [Pb(Hsaltren)] [2.274(5)–2.518(7) Å] in which the lead(II) atom is bound by two phenolate oxygens and two imido nitrogens only;2 additionally the Pb(2)–O lengths [2.339(3)– 2.358(4) Å] are elongated compared with SchiV base complexes of lead(II) in which the central cavity of the ligand is occupied by the phenolic protons [2.27(1)–2.34(1) Å].2 The Pb(1)–N(1) distance of 2.651(4) Å is comparable to the mean Pb(1)–Nimine length of 2.60 Å, which seems to confirm that a genuine bonding interaction exists.The O–Pb(1)–O, N(1)–Pb(1)–N and chelating N–Pb(1)–O angles lie within a range 67.60(12)– 71.82(11)8, the O–Pb(2)–O angles being slightly more open [74.96(13)–78.21(12)8].All of the remaining X–Pb(1)–N angles (where X = phenolate O or N) between diVerent chelate rings fall within a larger range, 103.2(14)–140.62(12)8. The coordination geometry about Pb(1) in 2 is best described as a distorted monocapped octahedron, with the triangular faces defined by the three imine nitrogens and the three Scheme 1 The preparation of compounds 1 and 2.3150 J.Chem. Soc., Dalton Trans., 1998, 3149–3150 phenolate oxygens atoms respectively, with the apical nitrogen atom N(1) capping the imine face. The o-methoxy substituents of L32 increase the electron density at the phenolate oxygen atoms suYciently to allow binding of a second Pb(II) centre; thus Pb(2) closes the ligand pocket around Pb(1), the mismatch in size leading to distortions in the overall geometry of the Fig. 1 (a) Structure of the cation in 2 (hydrogen atoms omitted for clarity), (b) core coordination environment. Selected bond lengths (Å) and angles (8): Pb(1)–O(6) 2.595(4), Pb(1)–O(16) 2.534(4), Pb(1)–O(26) 2.518(3), Pb(1)–N(1) 2.651(4), Pb(1)–N(3) 2.616(5), Pb(1)–N(13) 2.594(4), Pb(1)–N(23) 2.577(5), Pb(2)–O(6) 2.351(4), Pb(2)–O(16) 2.339(3), Pb(2)–O(26) 2.358(4); O(26)–Pb(1)–O(16) 71.82(11), O(26)– Pb(1)–N(23) 71.55(13), O(16)–Pb(1)–N(23) 123.13(14), O(26)–Pb(1)– O(6) 69.29(12), O(16)–Pb(1)–O(6) 67.60(12), N(23)–Pb(1)–O(6) 132.19(13), O(26)–Pb(1)–N(13) 135.96(13), O(16)–Pb(1)–N(13) 70.88(13), N(23)–Pb(1)–N(13) 111.7(2), O(6)–Pb(1)–N(13) 115.40(13), O(26)–Pb(1)–N(3) 117.63(14), O(16)–Pb(1)–N(3) 126.03(14), N(23)– Pb(1)–N(3) 109.2(2), O(6)–Pb(1)–N(3) 68.05(13), N(13)–Pb(1)– N(3) 103.02(14), O(26)–Pb(1)–N(1) 139.57(13), O(16)–Pb(1)–N(1) 140.62(12), N(23)–Pb(1)–N(1) 69.24(14), O(6)–Pb(1)–N(1) 135.70(13), N(13)–Pb(1)–N(1) 69.98(14), N(3)–Pb(1)–N(1) 68.01(14), O(16)– Pb(2)–O(6) 74.96(13), O(16)–Pb(2)–O(26) 78.21(12), O(6)–Pb(2)–O(26) 76.24(13).encapsulated metal atom. While several authors have claimed the lone pair at Pb(II) centres of high coordination number to be stereochemically inactive, its precise steric requirements seem to be sensitive to several factors.6 Although Pb(2) can be described as having a distorted tetrahedral geometry in which the lone pair is in the hemisphere oriented away from Pb(1), the position of the lone pair at Pb(1) cannot be unequivocally defined.Acknowledgements We are grateful to the EPSRC (P. B. and A. T. R.) for financial support, and to the EPSRC National MS Service Centre at Swansea for mass spectral data. Notes and references † Experimental conditions and instrumentation are as described elsewhere, 2,3 H3L was prepared from tris(2-aminoethyl)amine and ovanillin. 4 CAUTION perchlorate salts are potentially explosive and should be handled with care. [Pb2L]Cl 1.A mixture of H3L (2.42 g, 4.4 mmol) and lead(II) chloride (1.22 g, 4.4 mmol) in methanol (40 cm3) was heated at reflux for 30 min. Addition of triethylamine (3 cm3) gave a clear orange solution which was heated for a further 2 h, during which time a precipitate of triethylammonium hydrochloride formed. The solution was allowed to cool and filtered, orange microcrystals of 1 grew from the filtrate over one week. Yield 1.45 g, 66% based on Pb. Found (Calc. for C30H33ClN4O6Pb2): C, 36.6 (36.2); H, 3.4 (3.3); N, 5.6 (5.6)%.IR(KBr disc): n(CN) 1624s cm21. dH(250.1 MHz, CD3CN): 8.28 (s, 2H, CH]] N), 8.22 (s, 1H, CH]] N), 6.96 (m, br, 6H, aryl), 6.57 (m br, 3H aryl), 3.88 (s br, 9H, OCH3), 3.70 (s br, 6H, CH2), 2.85 (s br, 6H, CH2). MS(FAB1): m/z 961 [Pb2L]1. [Pb2L][ClO4] 2. To a stirred solution of 1 (0.17 g, 0.17 mmol) in methanol (15 cm3) was added Ag[ClO4] (0.045 g, 0.22 mmol) in methanol (3 cm3). After 30 min the orange solid was collected by centrifugation, extracted into acetonitrile (10 cm3) and filtered through glass wool–Celite.Removal of the solvent in vacuo gave crude 2, recrystallised from acetonitrile–diethyl ether as yellow microcrystals (49 mg, 28% yield). Found (Calc. for C30H33ClN4O10Pb2): C, 34.1 (34.1); H, 3.2 (3.0); N, 5.3 (5.3)%. IR(KBr disc): n(CN) 1630s, n(ClO4) 1097vs and 623m cm21. dH(250.1 MHz, CD3CN): 8.55 (s, 2H, CH]] N), 8.34 (s, 1H, CH]] N), 7.13 (m 4H, aryl), 6.99 (m br, 2H, aryl), 6.71 (m, 2H, aryl), 6.60 (m br, 1H, aryl), 4.00 (s, 3H, OCH3), 3.90 (s, 6H, OCH3), 3.64 (m, 6H, CH2), 2.88 (m, 6H, CH2).‡ Crystal data and data collection parameters for 2: C30H33ClN4O10Pb2, M = 1059.43, monoclinic, space group P21/c, a = 12.218(1), b = 16.274(10), c = 16.443(1) Å, b = 91.04(1)8, U = 3268.99(3) Å3, Z = 4, m(Mo-Ka) = 10.431 mm21, T = 293 K, R1 = 0.0212 for 4754 unique reflections. CCDC reference number 186/1127. 1 D. F. Cook, D. Cummins and E. D. McKenzie, J. Chem. Soc., Dalton Trans., 1976, 1396; S. K. Chandra, P. Chakraborty and A. Chakravorty, J. Chem. Soc., Dalton Trans., 1993, 863; M, Kanesato, T. Yokoyama, O. Habashi, T. M. Suzuki and M. Shiro, Bull. Chem. Soc. Jpn., 1996, 69, 1297; G. Hunter and N. Kilcullen, J. Chem. Soc., Dalton Trans., 1989, 2115; P. K. Bhasadwaj, A. M. Lee, S. Mandal, B. W. Skelton and A. H. White, Aust. J. Chem., 1994, 47, 1799; A. Smith, S. J. Rettig and C. Orvig, Inorg. Chem., 1988, 27, 3929. 2 J. Parr, A. T. Ross and A. M. Z. Slawin, Polyhedron, 1997, 16, 2765. 3 J. Parr, A. T. Ross and A. M. Z. Slawin, J. Chem. Soc., Dalton Trans., 1996, 1509. 4 S. Liu, L-W. Yang, S. J. Rettig and C. Orvig, Inorg. Chem., 1993, 32, 2773. 5 J.-P. Costes, F. Dahan, A. Dupuis, S. Lagrave and J.-P. Laurent, Inorg. Chem., 1998, 37, 153. 6 L. Shimoni-Levy, J. P. Glusker and C. W. Bock, Inorg. Chem., 1998, 37, 1853. Communication 8/06058D

ISSN:1477-9226    

DOI:10.1039/a806058d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1998, 3151–3153 3151 Coupling Mo2 n1 units via dicarboxylate bridges F. Albert Cotton,*a Chun Lin a and Carlos A. Murillo *ab a Department of Chemistry and Laboratory for Molecular Structure and Bonding, PO Box 300012, Texas A&M University, College Station, TX 77842-3012, USA. E-mail: cotton@tamu; edumurillo@tamu.edu b Department of Chemistry, University of Costa Rica, Ciudad Universitaria, Costa Rica Received 5th August 1998, Accepted 18th August 1998 The compounds Mo2(DArF)3(O2CC2H5) 1, (DArF)3Mo2- (O2CCO2)Mo2(DArF)3 2 and (DArF)3Mo2(O2CC6F4CO2)- Mo2(DArF)3 3 (DArF = N,N9-diarylformamidinate, Ar = p-anisyl) have been selectively prepared in good yield from the reactions of Mo2(DArF)3Cl2 and the corresponding carboxylate salts in the presence of NaHBEt3; their crystal structures, electronic spectra and electrochemistry have been studied.Compounds built upon M2 n1 units display many kinds of interesting properties and chemistry,1 not the least of which is electrochemistry.Clearly, however, considerably more elaborate and interesting electrochemistry would be expected if two (or more) such centers could be connected so that they are electronically coupled. As a long-range goal it would be interesting to study species of the schematic type I. Synthesis and structural characterization of such species is clearly the first task, and it would seem equally clear that the “first of the first” task is to make and characterize the first member (I, n = 0).A promising first step in this direction was reported in 1991 by Chisholm et al.2 who described several species of type I (n = 0) wherein the metal atoms were Mo, W, the non-connective ligands were pivalate anions (Piv) and the connective ligands were dicarboxylates, specifically, oxalate, 1,4-C6F4(CO2)2 22, 1,8-anthracenedicarboxylate and 9,10-dihydroanthracene-1,8- dicarboxylate. They made some very significant observations concerning these eight compounds, but were unable to obtain crystallographic structural information on any of them.These were made by two types of reactions (1 and 2) using quadruplybonded Mo2 41 species as starting materials: 2M2(O2CR)4 1 HO2CR9CO2H toluene r.t. [M2(O2CR)3]2(m-O2C–R9–CO2) 1 2RCO2H (1) 2Mo2(O2CBut)3 1(BF4)2 1 (M1)2bridge22 THF 0 8C [Mo2(O2CBut)3]2(bridge) 1 2MBF4 (2) Unfortunately, serious synthetic limitations were found, i.e. the reversibility of the substitutions when using M2(O2CR)4 compounds, and the formation of higher oligomers that compete with the targeted tetranuclear species.Reaction 1 is an equilibrium process, and thus the separation of the products is non-trivial. We have recently employed compounds of the type II as starting materials 3 in the preparation of compounds in which Mo2(DArF)3 (DArF = N,N9-diarylformamidinate) units are linked by m-H, m-OH and m-O groups.4 A variety of Ar groups may be employed in the formamidinate ligands.5 We have now used a molecule of type II with Ar = p-MeOC6H4 and in this way we have cleanly made compounds of type I (n = 0) that can be characterized structurally.Two of these are described here. The preparations† were carried out by the general reaction (3) and the crystallographic studies ‡ led to the structures 2Mo2(DArF)3Cl2 1 2NaHBEt3 1 (NBun 4)2(O2C]h]CO2) CH2Cl2 (DArF)3Mo2(O2C]h]CO2)- Mo2(DArF)3 1 2NaCl 1 2NBun 4Cl 1 2BEt3 1 H2 (3) shown in Figs. 1 and 2. In 2 where the connective ligand is oxalate the main structural features that are relevant here are: (1) The oxalate connector is in an end-to-end (as opposed to a conceivable side-to-side) posture. (2) The entire Mo2O2CCO2- Mo2 unit is planar. (3) The Mo–Mo distances show that Mo– Mo quadruple bonds are retained. In 3, the plane of the central C6F4 unit is rotated by 30.48 from the two coplanar Mo2O2C units. The electrochemical results (Table 1) obtained on 2 and 3 are very similar to those obtained by Chisholm et al.on the pivalate analogues.2 Acknowledgements We are grateful to the National Science Foundation for support and to Dr Lee M. Daniels for helpful crystallographic advice.3152 J. Chem. Soc., Dalton Trans., 1998, 3151–3153 Table 1 Electrochemical data for compounds 1, 2 and 3 Compounda Mo2(DArF)4 1 Mo2(Piv)4 2 Piv analogue 3 Piv analogue E2� 1 (1/0)/mV b 119 218 850 260 860 332 e 990 DEp/mVc 90 74 78 64 79 133 e 92 E2� 1 (21/1)/mV ——— 472 1140 —— DEp/mV ——— 68 110 —— DE2� 1 /mVd ——— 212 280 41 f 65 a The cyclic votammograms were recorded on a BAS 100 electrochemical analyzer on 0.1 M (Bun)4NPF6 solution (CH2Cl2) with Pt working and auxiliary electrodes and a Ag/AgCl reference electrode, and scan rate of 100 mV s21.All the potential values are referred to Ag/AgCl, and under the present experimental conditions, the E2� 1 (Fc1/Fc) was consistently measured at 440 mV. The values for pivalate analogues were taken from ref. 2, and have been converted to be referred to Ag/AgCl. b E2� 1 = (Epa 1 Epc)/2. c DEp = Epa 2 Epc. d DE2� 1 = E2� 1 (21/1) 2 E2� 1 (1/0). e E2� 1 and DEp values for 3 are average values for both redox couples, i.e., 21/0. f DE2� 1 for 3 was obtained from diVerential pulse votammetry (rate = 2 mV s21, pulse amplitude = 10 mV). Fig. 1 The molecular structure of 2 in 2?2CH2Cl2. Selected bond distances (Å) and angles (8): Mo(1)–Mo(2) 2.0900(7), Mo(1)–O(7) 2.145(3), Mo(2)– O(8) 2.115(3), Mo–N (av) 2.141[7]; Mo(2)–Mo(1)–O(7) 90.9(1), Mo(1)–Mo(2)–O(8) 93.1(1), cis-N–Mo–O (av) 85.5[3], N(3)–Mo(1)–O(7) 176.8(2), N(4)–Mo(2)–O(8) 173.2(2), Mo–Mo–N (av) 92.7[6], trans-N–Mo–N (av) 169[1], cis-N–Mo–N (av) 94.4[4].Fig. 2 A thermal ellipsoid drawing of the core of 3 in 3?2CH2Cl2? C6H14. Selected bond distances (Å) and angles (8): Mo(1)–Mo(2) 2.0902(9), Mo(1)–O(8) 2.124(5), Mo(2)–O(7) 2.162(5), Mo–N (av) 2.133[7]; Mo(1)–Mo(2)–O(7) 90.4(1), Mo(2)–Mo(1)–O(8) 93.1(1), cis-N–Mo–O (av) 86.2[7], N(3)–Mo(1)–O(8) 173.3(2), N(4)–Mo(2)– O(7) 177.4(2), Mo–Mo–N (av) 92.6[4], trans-N–Mo–N (av) 171[2], cis-N–Mo–N (av) 93.6[4].Notes and references † The following procedure describes the preparation of 2 [(DArF)3- Mo2(O2CCO2)Mo2(DArF)3]. A similar method was used for 1 [Mo2- (DArF)3(OCC2H5)] and 3 [(DArF)3Mo2(O2CC6F4CO2)Mo2(DArF)3]. To a mixture of Mo2(DArF)3Cl2 (154 mg, 0.15 mmol) and (Bun 4N)2- (C2O4) (43.0 mg, 0.075 mmol) in 60 mL of CH2Cl2 was added NaHBEt3 (1.0 mmol).The reaction mixture was stirred for 24 h at room temperature, then the volatile materials were removed under vacuum, and the resulting residue was washed with Et2O (10 mL), EtOH (2 × 20 mL), H2O (2 × 10 mL) and EtOH (10 mL). Finally it was extracted with CH2Cl2 (3 × 7 mL). The combined extracts were concentrated to 4 mL. Hexanes (25 mL) were then carefully added without stirring, and the solution was kept for 12 h at room temperature.The bright red blockshaped crystals that formed were collected by filtration and dried in vacuo. Yield: 80.5 mg, 53.6%. Single crystals suitable for X-ray analysis were grown by diVusion of hexanes into a CH2Cl2 solution. 1H NMR d (ppm, in CD2Cl2): 8.52 (s, 2H, NCHN), 8.47 (s, 4H, NCHN), 6.56 (m, 32H, aromatic), 6.45 (d, 8H, aromatic, 3J = 9.0), 6.23 (d, 8H, aromatic, 3J = 8.9 Hz), 3.67 (s, 24H, OCH3), 3.64 (s, 12H, OCH3). IR (KBr, cm21): 1700w, 1646w, 1609m, 1548s, 1504s, 1464m, 1441m, 1296s, 1246s, 1217s,J.Chem. Soc., Dalton Trans., 1998, 3151–3153 3153 1178m, 1107w, 1034s, 938w, 829s, 771m, 590w, 536w. UV/VIS, lmax/nm (e/M21 cm21): 296 (54300), 450 (6210). Compound 1: Yield, 45.3%. 1H NMR d (ppm, in CD2Cl2): 8.46 (s, 2H, NCHN), 8.42 (s, 1H, NCHN), 6.65 (d, 8H, aromatic, 3J = 8.9), 6.51 (d, 8H, aromatic, 3J = 9.0), 6.43 (d, 4H, aromatic, 3J = 9.0), 6.22 (d, 4H, aromatic, 3J = 8.9), 3.70 (s, 12H, OCH3), 3.63 (s, 6H, OCH3), 2.85 (q, 2H, O2CCH2CH3, 3J = 7.6), 1.34 (t, 3H, O2CCHR (KBr, cm21): 1609w, 1540s, 1503s, 1463m, 1292m, 1245s, 1217s, 1176m, 1108w, 1034s, 937w, 829m, 764w, 722w, 650w, 618w, 590w, 536w. UV/VIS, lmax/nm (e/M21 cm21): 290 (sh), 303 (47900), 385 (sh), 430 (sh). Compound 3: Yield, 46.5%. Single crystals suitable for X-ray analysis were grown by diVusion of hexanes into a CH2Cl2 solution. 1H NMR d (ppm, in CD2Cl2): 8.52 (s, 2H, NCHN), 8.51 (s, 4H, NCHN), 6.66 (d, 16H, aromatic, 3J = 9.2), 6.57 (d, 16H, aromatic, 3J = 9.2), 6.46 (d, 8H, aromatic, 3J = 9.0), 6.25 (d, 8H, aromatic, 3J = 8.9 Hz), 3.70 (s, 24H, OCH3), 3.64 (s, 12H, OCH3).IR (KBr, cm21): 1541s, 1501s, 1466m, 1459m, 1438w, 1388w, 1289m, 1245s, 1217s, 1178m, 1107m, 1035s, 993w, 938w, 828m, 765w, 742m, 644w, 589w, 468w. UV/VIS, lmax/nm (e/M21 cm21): 297 (94370), 477 (8230). Elemental analyses were satisfactory for all compounds. ‡ Crystal data for 2?2CH2Cl2: C94H94Cl4Mo4N12O16, M = 2173.37, monoclinic, space group P21/c, a = 13.998(2), b = 17.441(3), c = 20.229(1) Å, b = 97.40(1)8, U = 4898(1) Å3, Z = 2, m(Mo-Ka) = 0.678 mm21.Data were collected at 213(2) K. The structure, refined on F 2, converged for 6356 unique reflections and 596 parameters to give R1 = 0.052 and wR2 = 0.121 and a goodness-of fit = 1.096. Crystal data for 3?2CH2Cl2?C6H14: C106H108Cl4F4Mo4N12O16, M = 2407.60, triclinic, space group P1� , a = 9.791(1), b = 15.756(2), c = 18.090(3) Å, a = 76.65(1)8, b = 79.61(1)8, g = 87.63(1)8, U = 2670.8(6) Å3, Z = 1, m(Mo-Ka) = 0.635 mm21. Data were collected at 213(2) K. The structure, refined on F 2, converged for 6664 unique reflections and 647 parameters to give R1 = 0.068 and wR2 = 0.166 and a goodness-of- fit = 1.036. CCDC reference number 186/1126. See http://www.rsc.org/ suppdata/dat/1998/3151/ for crystallographic files in .cif format. 1 F. A. Cotton and R. A. Walton, Multiple Bonds between Metal Atoms, Oxford University Press, Oxford, 2nd edn., 1993. 2 R. H. Cayton, M. H. Chisholm, J. C. HuVman and E. B. Lobkovsky, J. Am. Chem. Soc., 1991, 113, 8709. 3 F. A. Cotton, G. T. Jordan IV, C. A. Murillo and J. Su, Polyhedron, 1997, 16, 1831. 4 F. A. Cotton, L. M. Daniels, G. T. Jordan IV, C. Lin and C. A. Murillo, Inorg. Chem. Commun., 1998, 1 109; F. A. Cotton, L. M. Daniels, G. T. Jordan IV, C. Lin and C. A. Murillo, J. Am. Chem. Soc., 1998, 120, 3398. 5 C. Lin, J. D. Protasiewicz, E. T. Smith and T. Ren, Inorg. Chem., 1996, 35, 6422. Communication 8/0616

ISSN:1477-9226    

DOI:10.1039/a806161k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3155–3165 3155 Crystal supramolecular motifs. Ladders, layers and labyrinths of Ph4P1 cations engaged in fourfold phenyl embraces Marcia Scudder and Ian Dance * School of Chemistry, University of New South Wales, Sydney 2052, Australia. E-mail: I.Dance@unsw.edu.au Received 8th May 1998, Accepted 23rd July 1998 Tetraphenylphosphonium cations in crystals associate with attractive fourfold phenyl embraces (4PE), in which two phenyl rings from each cation engage in concerted edge-to-face and oVset-face-to-face local interactions.The names of the orthogonal (O4PE) and parallel (P4PE) subclasses refer to the relationship between the relevant C–P–C planes in the two cations. One-dimensional ladder networks with O4PE along the sides and P4PE as rungs, and twodimensional networks with O4PE in both directions, P4PE in both directions, or O4PE in one direction and P4PE in the other have been explored. The layers can be planar or corrugated.Crystalline [Ph4P1][Cl3Te(OCH2CH2O)]2 contains a zeolite-like network of cations involved in 4PE, with the anions in channels. Most of the crystals included in this survey contain mono-negative anions, ranging from Cl3 2 to metal complexes with chelate ligands, and include structurally non-molecular anions such as [{Cu4I5ReS4}22]•. The crystalline compound [Ph4P1]2[C60]2[I2]x contains a high symmetry two-dimensional network of cations in P4PE, which functions as an eVective host lattice completely encapsulating the [C60]2, with phenyl ? ? ? fullerene interactions which are both face-to-face and edge-to-face.Introduction Our investigations of supramolecular motifs are based on the excellence of a molecular crystal as a supramolecular entity,1,2 and the occurrence of crystallographic data for ca. 170,000 crystals (mainly of molecular compounds) in the Cambridge Structural Database.3 One outcome has been the recognition of multiple phenyl embraces as significant supramolecular motifs for compounds containing Ph4P1 cations and compounds with PPh3 ligands.4–6 While individual phenyl groups engage in edgeto- face (ef) and oVset-face-to-face (oV) attractive phenyl ? ? ? phenyl interactions, the multiple phenyl embraces are motifs in which these local attractions are concerted.The name embrace signifies the three attributes of (1) participation of two or more phenyl groups from each partner molecule, (2) geometrical concertedness, and (3) strong attraction.The main categories of multiple phenyl embrace are the sixfold phenyl embrace or 6PE † containing six edge-to-face (ef) attractions involving three phenyl rings from each Ph4P1, first identified by Muller,7 and two types of fourfold phenyl embrace or 4PE.† In the orthogonal fourfold phenyl embrace (O4PE) the two C–P–C planes for the four phenyl rings are approximately orthogonal and engaged in four ef interactions, while in the parallel fourfold phenyl embrace (P4PE) these two C–P–C planes are approximately parallel and the motif comprises one oV and two ef interactions.6 The net attractive energy for the phenyl rings in the 6PE is calculated to be in the range 8–11 kcal mol21 (for a pair of complete Ph4P1 cations the attractive † In our original papers the 6PE was named the sextuple phenyl embrace and abbreviated as SPE, and the 4PE was called the quadruple phenyl embrace, or QPE.Subsequent investigations of multiple phenyl embraces revealed more elaborate motifs involving larger numbers (eight and twelve) of phenyl rings, and so we have revised the abbreviations to include the number of phenyl rings involved.‡ These intermolecular energies are calculated as the sum of the van der Waals and coulombic atom–atom components, dependent on empirical parameters which are under continuing refinement: more recent energy values may diVer from earlier published values.energy is 5–8 kcal mol21):‡ the energy of the 6PE is comparable with that of the stronger hydrogen bonds. The heteroaromatic rings of bipy ligands in [M(bipy)3]z complexes can adopt similar sixfold aryl embraces,8 and crystalline tris(anthracenyl)phosphine forms an inverted 6PE.9 We subsequently described the general occurrence of continuous chains of these embraces, as the zig-zag infinite chain of sixfold phenyl embraces (ZZI6PE, originally ZZISPE†) 10 and the linear infinite chain of translational fourfold phenyl embraces (LIT4PE, originally LITQPE:† translational repetition of Ph4P1 along a pseudo-twofold axis generates O4PE motifs).6,10 Muller 11 had previously classified the columnar structures with fourfold crystal symmetry known in 1980.These one-dimensional extensions of the isolated embraces arise because each phenyl ring can function as both H “donor” and p “acceptor” in ef interactions. In the LIT4PE only two of the “edges” of the Ph4P1 tetrahedron are involved in 4PE,10 and therefore there are additional opportunities for multiple phenyl embraces.Similarly, in the ZZI6PE only two “faces” of the tetrahedron are used.10 Both the LIT4PE and the ZZI6PE motifs could engage in additional embraces and form more highly connected supramolecular motifs, and it is these elaborations which are investigated in this and the following paper. The Cambridge Structural Database (October, 1997) contains data for 1060 crystals containing Ph4P1, and 944 of these contain at least one pair of cations with P ? ? ? P < 9 Å, signifying interpenetration of volumes and possible multiple phenyl embraces.Examination of this large collection of crystal structures reveals that more elaborate networks of multiple phenyl embraces occur. We classify these networks according to: (1) the number of connections at each Ph4P1; (2) the embrace type; and (3) the dimensionality of the network. We have already described the group of compounds containing RPh3P1 cations which crystallise with hexagonal arrays of 6PE (HA6PE†), which are further connected by other embraces.12,13 This paper focusses on one-, two- and three-dimensional networks constructed from 4PE, and the following paper 14 on networks based on 6PE.3156 J.Chem. Soc., Dalton Trans., 1998, 3155–3165 Table 1 Parameters used in calculations of interatomic energies: see eqns. (3), (4), (5) Atom type All C in phenyl ring All H in phenyl ring P A/Å12 kcal mol21 1116550.0 23058.4 6025894.0 B/Å12 kcal mol21 644.5 42.9 2195.6 da i/Å 3.9 3.2 4.2 ea i/kcal mol21 0.093 0.020 0.20 qi 20.10 10.15 10.40 Methodology Data were obtained from the October, 1997 release of the Cambridge Structural Database,3,15 and analysed using the Cambridge Quest3D graphical software and the MSI programs16 InsightII and Catalysis for construction, analysis and portrayal of crystal lattices. Intermolecular energies E were calculated as the sum of interatomic energies,17,18 using the Lennard-Jones 6-12 interatomic potential for attractive and repulsive van der Waals energies EvdW ij [eqn.(3)], and the coulombic components Ecoulombic ij [ eqn. (5)]. The atom partial charges qi were obtained from a density functional (blyp) calculation of the electron density of Ph4P1 followed by optimisation of the charge array to best reproduce the electrostatic potential of Ph4P1. Because Ph4P1 has low polarity and no lone pairs, the coulombic energy need not incorporate atomic multipoles.19 The van der Waals parameters have been refined to reproduce (together with the coulombic component) the best experimental and theoreticalJ.Chem. Soc., Dalton Trans., 1998, 3155–3165 3157 Table 2 Crystal structures which contain isolated regular ladders of Ph4P1 cations. The interactions along the ladder edges are O4PE and the rungs are PQPE. In all crystals the ladders are parallel P ? ? ? P dimensions/Å Angle of Refcode BITXUX BITYAE JETNUR SAPDIW TPCBPT VEJLAX WAKGUK YIMSUI ZUDPUJ Anion [SnCl3]2 [SnBr3]2 [Cl4Re(NSSN)]2 [Cu(SSNS)2]2 [Cl3Pt{C(CH2OH)}2]2 [Cl5W{h2-C(H)C(Ph)}]2?CH2Cl2 [NCl2Os(OCH2CH2O)]2 [I2(B7CH6)]2 [B6H6(CH2CH2)B6H6]22?EtOH Space group P1� P1� P1� P1� P2c P21/n P1� P21/n Along ladder O4PE 7.42 7.44 7.67 7.08 7.47 7.51 8.17 7.84 7.53 Rungs, P4PE 8.20 8.27 8.23 8.99 8.56 8.00 8.61 8.50 8.53 rung to edges/8 78 76 78 79 79 76 88 83 69 information on the intermolecular energy of benzene.The parameters and charges used in eqns. (1) to (5) are presented in E = SEij (1) Eij = EvdW ij 1 Ecoulombic ij (2) EvdW ij = [(Aij × dij)212 2 (Bij × dij)26] = ea ij [(d a ij/dij)12 2 2(d a ij/dij)6] (3) Aij = (Ai × Aj)0.5, Bij = (Bi × Bj)0.5, d a ij = (d a i 1 d a j)/2, ea ij = (ea i × ea j)0.5 (4) Ecoulombic ij = qi qj/e dij (5) Table 1: dij is the interatomic separation; Aij, Bij are the parameters used in the MSI program Discover;16 e a ij and d a ij are respectively the energy depth and the interatomic distance for the attractive well of the vdW potential;5 and the permittivity e was set equal to dij.5 In the Discussion the energies are considered per pair of {Ph4}, and per Ph4P1 cation.The energies marked on the figures are per Ph2 pair. Volumes were computed using a Connolly surface with a probe of zero radius, and van der Waals radii Cl 1.75 and Br 1.85 Å. Results One-dimensional arrays When two strands of LIT4PE are located near each other so that embrace motifs occur between the strands, the result is a ladder.A regular ladder structure is illustrated in Fig. 1(a), for the compound Ph4P1SnCl3 2 as it occurs in the crystals BITXUX (six letter identifiers for crystal structures are the reference codes of the Cambridge Structural Database). The rungs of the ladder are formed from P4PE. Fig. 1(b) shows more of the crystal lattice, and how the ladders are parallel and separated by the pyramidal [SnCl3]2 anions. Table 2 lists other instances (with dimensions) of similar regular ladder structures comprised of O4PE along the edges and P4PE as rungs.The rungs of the ladder are approximately orthogonal to the edges, with the angle ranging from 69 to 888. Note the variation in size and volume of the anions. Another group of compounds has embracing cations in ladders comprised of triangular rather than quadrilateral segments. These compounds and the P ? ? ? P dimensions of the ladders are listed in Table 3, and one example, WEYYUU, Ph4P1[m-Br(NBS)2]2 (NBS = N-bromosuccinimide), is illustrated in Fig. 2. The sides of the ladder are made up of O4PE (7.6 Å) while each of the zig-zag rungs involves an interaction between one ring from one cation and three from the other. The diVerence between this ladder and that shown in Fig. 1 can be easily visualised as being caused by a translation of one edge of the ladder relative to the other, by a distance of about half the Fig. 1 Representations of the crystal packing in Ph4P1SnCl3 2 [BITXUX]. (a) Two cells of the ladder of Ph4P1, with translationally repeated O4PE (7.4 Å) forming the ladder supports (vertical), and P4PE (8.2 Å) as the rungs: P black, C grey, H atoms omitted. P ? ? ? P distances are marked with arrows, while the non-arrowed numbers are the interaction energies (kcal mol21 per Ph2) between phenyl rings. Ph ? ? ? Ph energies smaller than 1.0 kcal mol21 are not marked.There are centres of inversion on and between the rungs (space group P1� ) with the c axis along the ladder. (b) View along the ladders (marked with arrows) showing the separation of the ladders by the pyramidal SnCl3 2 anions: Sn black, Cl speckled.3158 J. Chem. Soc., Dalton Trans., 1998, 3155–3165 Table 3 Crystal structures with zig-zag ladder structures of the type shown in Fig. 2 Refcode FASWEB SELGOF TOPZIH WEYYOO WEYYUU WEZBAE WEZBEI Anion [V(SCH2CH2S)2(OSiMe3)]2 [{Cu4I5ReS4}22]•?CH3CN [{PbI3}2]•?DMF [m-Cl(NBS)2]2 [m-Br(NBS)2]2 [m-Cl(NIS)2]2?CH3CN [m-Br(NIS)2]2?CH3CN Space group P1� P212121 Pna21 P21/n P21/n Pna21 Pna21 Ladder dimensions P ? ? ? P along, across/Å 8.94, 8.75/8.04 7.26, 8.88 7.99, 8.65 7.64, 8.81 7.62, 8.79 7.28, 8.98 7.25, 8.96 Angles of rungs to edges/8 62, 64, 54 66, 48, 66 63, 55, 63 65, 51, 65 65, 51, 65 66, 48, 66 66, 48, 66 Relationship between ladders Parallel Canted Parallel Parallel Parallel Canted Canted NIS = N-iodosuccinimide. O4PE length.Comparison of Figs. 1(b) and 2(b) shows that while there are diVerences within the ladders, both lattice types Fig. 2 Representations of the crystal packing in Ph4P1[m-Br(NBS)2]2 [WEYYUU]: P black, H atoms omitted for clarity. (a) Face view of the ladder: cell translation generates the O4PE along the ladder, while the rungs are generated by a 21 screw axis in space group P21/n. P ? ? ?P distances are marked with arrows, while the non-arrowed numbers are the energies (kcal mol21 per Ph2) between phenyl rings.(b) View along the ladders (marked with arrows) showing how the ladders are separated by layers of [m-Br(NBS)2]2 anions. Br speckled, O striped, N crosshatched. have channels containing the anions between the ladders, and that the ladders are clearly separated by the anions, and confirm that the one-dimensional ladder entity is the primary supramolecular motif in these crystals.Just as ZZI6PE chains can be parallel or canted relative to each other,10 these ladders also vary in their relative orientation, as shown in Table 3. Two-dimensional arrays Next we describe several classes of crystal packing which contain two-dimensional arrays of cations maintained by 4PE interactions. The first class of layer is propagated by O4PE in one direction and P4PE in the orthogonal dimension, with the P atoms exactly or nearly coplanar. This is illustrated in Fig. 3 for the compound Ph4P1Br3 2 [BEPZEB]. This packing structure is the two-dimensional extension of the ladder shown in Fig. 1. The infinite sequences of O4PE in this layer are strong, as shown by the phenyl ? ? ? phenyl interaction energies marked on Fig. 3 for the embraces with a P ? ? ? P separation of 7.6 Å (the interaction energy for the O4PE is 27.8 kcal mol21 per {Ph4}2). The phenyl ? ? ? phenyl energies in the other dimension of the layer are weaker, and the interaction is a poor P4PE.Nevertheless, the integrity of this O4PE/P4PE layer is demonstrated by its general occurrence in a variety of compounds. Table 4 lists the relevant details of compounds with anions as varied as Br3 2, Cl2 (with [Pr(OH2)6Cl2]1 also in the lattice), and [Cu(h2-Cp)2]2, which contain this layer motif of Ph4P1 cations. The anions are located between the layers. Fig. 3 The layer of Ph4P1 cations interacting attractively with O4PE (7.6 Å) and P4PE (8.8 Å) in crystalline Ph4P1Br3 2 [BEPZEB]: P ? ? ?P distances are marked with arrows. In space group P2/c the net directions are b (7.6) and c/2 (8.8 Å): there are centres of inversion at the mid-points of the P4PEs.The P atoms (black) are coplanar. The Br3 2 anions are located between these layers. Non-arrowed numbers are the energies (kcal mol21 per Ph2) between phenyl rings.J. Chem. Soc., Dalton Trans., 1998, 3155–3165 3159 Table 4 Crystal structures with two-dimensional nets of Ph4P1 cations, in which each cation is surrounded by four other cations and the net is propagated by O4PE motifs in one direction and by P4PE motifs in the other.All these nets have the P atoms exactly, or very nearly coplanar Refcode BEPZEB TPHOSI CUPKAZ DALNAF01 HESJIY PPHTCQ YIMSOC YULLIA YULLOG YULYUZ ZEZQOK Anion and other components of the lattice [Br3]2 [BrIBr]2 [(CO)3Cr(h5-formylcp)]2 [Pr(OH2)6Cl2]1, 2H2O, 2Cl2 [CpLiCp]2 [L2]2 [I(B7CH7)]2 [Cr(CO)5(NO2)]2 [W(CO)5(NO2)]2 [(BMe)2B4H5]2 [Cu(h2-Cp)2]2 P ? ? ? P distance for the O4PE/Å 7.60 7.67 7.03 6.71 6.97 7.77 7.47 6.95 7.05 7.26 7.40 P ? ? ? P distance for the P4PE/Å 8.79 8.88 8.83 8.21 8.87 8.23 8.20, 8.93 8.94 8.98 8.30 8.91 L = Tetracyanoquinodimethanide.The second class of layer is a primitive rectangular net of repeating Ph4P1, which are oriented such that each Ph4P1 forms an O4PE with each of its four neighbours. The O4PEs which propagate the layer in one direction are shorter and more attractive than those in the other direction, as shown by the data in Fig. 4 for the compound Ph4P1[AsCl4(THF)2]2 [VUDTOD]. We have identified only two other instances of this motif, listed in Table 5, but agadraw attention to the variety of anion type and shape which can be associated with this net. The third type of two-dimensional net is propagated by P4PE embraces in both directions, as illustrated in Fig. 5 for the com- Fig. 4 Representations of the crystal structure of Ph4P1[AsCl4- (THF)2]2 [VUDTOD].(a) The rectangular net of Ph4P1 cations involved in O4PE attractive interactions in both dimensions. The space group is P2/c: the 8.1 Å interaction is b, the 8.8 Å interaction is a/2. The calculated interaction energies per {Ph4}2 are 25.6 kcal mol21 for the 8.1 Å O4PE, and 22.1 kcal mol21 for the 8.8 Å O4PE: the total intralayer interaction energy per cation is 27.7 kcal mol21 per {Ph4} [22.2 kcal mol21 per Ph4P1]. (b) Side view of two nets of cations sandwiching the [AsCl4(THF)2]2 anions: As white, Cl speckled.pound Ph4P1[SAs(S7)]2 [KIYJIL]. Although the P4PE are longer than in isolated occurrences of this embrace, there are still strongly attractive Ph ? ? ? Ph energies, as shown in Fig. 5(a). The slight distortions from a primitive square net of cations in this compound are due to the presence of the anion, which nestles in the cusps of the cation layer, as shown by the side view of the layers in Fig. 5(b). Other instances of this P4PE/P4PE layer motif are listed in Table 6.We draw attention to the crystal packing in the three isostructural compounds which contain C60 2 as one of two anions, namely [Ph4P1]2[C60]2X2 (X = Cl, Br or I) [YEBDUE, YUXCAV, LAZPAD], illustrated in Fig. 6. Each layer of cations has a four-fold array of P4PE [Fig. 6(a)] in a high symmetry tetragonal (I4/m) lattice. These layers of cations are stacked according to mirror planes between them, such that the lattice contains tetragonal prismatic cavities surrounded by eight cations.The C60 2 and I2 ions occupy these cavities (although the I2 positions are incompletely occupied 20). The I2 is surrounded by eight H atoms (the 4-positions of the phenyl rings), forming weak C–H ? ? ? I hydrogen bonds [Fig. 6(a)]. The encapsulation of the C60 2 by the eight Ph4P1 cations is particularly significant and is described in more detail. Fig. 6(b) shows how the C60 2 nestles between four Ph4P1 cations engaged in four P4PE, and Fig. 6(c) showing the underside of this nest demonstrates how this square of four Ph4P1 in P4PE can almost completely cover the C60 2.Most of the phenyl rings covering the C60 2 are in face(Ph)-to-face(C60) geometry, as is common for C60 derivatives containing phenyl groups,21–25 but four of the phenyl rings from each of the two layers approach the C60 in edge-to-face geometry: these four rings are evident in Fig. 6(b). Fig. 6(d) shows a side view of the C60 site, including cations from the cation layers above and below, and confirms that the C60 is almost totally covered by the eight cations.In this crystal structure we see the two-dimensional array of Ph4P1 cations linked by P4PEs as indented with approximately hemispherical nests, able to hold a molecule such as C60 which engages in aryl ? ? ? aryl attractive interactions. Two such nests are able to enclose the fullerene. This role of the cation array is then comparable with that of host molecules of the calixarene 26,27 and cyclotriveratrylene 28 classes.The host lattice formed by Ph4P1 cations using multiple phenyl embraces as the supramolecular factor is comparable with the hydrogen-bonded quinol networks which also include fullerenes.29,30 At this point we note that the rectangular layer motifs containing Ph4P1, in which each cation is surrounded by four other cations, can be propagated by O4PE in both dimensions, or P4PE in both dimensions, or by O4PE in one dimension and P4PE in the other.This is possible by rotations of the cations in the layers, as is evident by comparison of Figs. 3(a), 4(a) and 5(a).3160 J. Chem. Soc., Dalton Trans., 1998, 3155–3165 Fig. 5 Representations of the crystal packing in Ph4P1[SAs(S7)]2 [KIYJIL]. (a) The two-dimensional net of Ph4P1 cations involved in P4PE motifs in both dimensions. The space group is Pna21, and the net (generated by the n-glide operation) is along the two ab diagonals and is half the length of each. Ph ? ? ? Ph energies smaller than 0.9 kcal mol21 are not marked.(b) Side view of two layers of cations, showing how the [SAs(S7)]2 anions (all atoms black) are positioned on the edges of the layers, not midway between them. The shortest S ? ? ? S distance between anions is 3.88 Å. Table 5 Crystal structures with two-dimensional nets of Ph4P1 cations, in which each cation is surrounded by four other cations and the net is propagated by O4PE motifs in both directions Refcode VUDTOD WEZVOM YAPHUS Anion and other components of the lattice [Cl4As(THF)2]2 [L1 2], Cl2 [VL2 2]2?H2O Planarity of the net Flat Flat Slightly pleated P ? ? ? P distance for the shorter O4PE/Å 8.09 7.46 7.81 P ? ? ? P distance for the longer O4PE/Å 8.82 7.97 8.55 L1 = succinimide, L2 = 2O2CCHMeN(O2)CHMeCO2 2.In the three types of two-dimensional four-connected nets already presented, the layers are exactly or almost planar, and there is only one type of embrace propagating the net in each dimension.There exists another class of two-dimensional four-connected nets which are corrugated and which have two diVerent embraces alternating through the corrugations. The compounds which crystallise with this non-planar layer of four-connected cations are collected in Table 7, and one example, Ph4P1[NCCH2(B6H6)]2 [WEYZUV] is shown in Fig. 7. Table 6 Crystal structures with two-dimensional nets of Ph4P1 cations, in which each cation is surrounded by four other cations and the net is propagated by P4PE motifs in both directions Refcode KIYJIL LAZPAD YEBDUE YUXCAV Anion and other components of the lattice [SAs(S7)]2 [C60]2, I2 [C60]2, Cl2 [C60]2, Br2 Planarity of the net Flat Flat Flat Flat P ? ? ? P distance for each embrace/Å 8.94 8.90 8.89 8.87 In addition to the corrugated layers just described, the fourconnected net can be modified by part-translations of the linear sequences of principal motifs within the net in a way analogous to that described above for generation of the triangular ladder (Fig. 2) from the regular ladder (Fig. 1). This is best illustrated with the example of [Ph4P1]2[BrAg(m-Br)2AgBr]22 [GIDWEV] shown in Fig. 8. Now each cation is surrounded by five other cations, and has remarkably favourable attractive Ph ? ? ? Ph interactions with each of these cations. This could be called a five-connected net of embracing cations. Sections of the pleated layer have similarities to other motifs: the approximately rectangular section with P ? ? ? P distances of 7.5 and 7.9 Å is like the ladder in Fig. 1 while the triangular section is like the ladder in Fig. 2. Other compounds with this structure type are listed in Table 8.J. Chem. Soc., Dalton Trans., 1998, 3155–3165 3161 Fig. 6 Representations of the crystal packing in [Ph4P1]2[C60]2[I2]x [LAZPAD], space group I4/m. (a) View of the fourfold net of P4PE (purple rods) along ab diagonals of the cell, with the C60 2 anions (twofold disorder is not shown) located in cusps between four P4PE, and the I2 ions located on 4/m sites (with partial occupancy).The cations are at 4� sites, and the P4PE network is generated by the fourfold symmetry. The C60 2 and I2 ions are at the same z coordinate, and covered by the next net of cations above. The C–H ? ? ? I2 weak hydrogen bonds (C ? ? ? I 3.67 and H ? ? ? I 2.82 Å) are marked in one quadrant. (b) Space-filling representation of the C60 anions nestled between the phenyl rings of four Ph4P1 cations engaged in four P4PE: note that there are face-to-face and edge-to-face interactions between Ph rings and C60 2.(c) Underneath view of part (b), showing how the Ph4P1 cations completely enclose the C60 2, and showing also the good edge-to-face and oVset-face-to-face interactions of the P4PEs. (d) Side view of the two layers of embracing Ph4P1 cations almost completely enclosing the C60 2: in this image the two layers of cations are drawn on the left and right rather than top and bottom.In parts (b), (c) and (d) the60 has been coloured blue for contrast. In addition to the linear chain, ladder or layer categories already described, there is a small number (less than 10 identi- fied so far in the CSD) of structures with LIT4PE linked to form three-dimensional arrays of Ph4P1. One of them is Ph4P1[Cl3Te(OCH2CH2O)]2 [COETTE], which contains parallel LIT4PE chains which are linked in a hexagonal array to form a porous network akin to a zeolite, as shown in Fig. 9. The channels so formed contain double columns of anions. Each cation takes part in a total of five embraces. The O4PE (along a) has P ? ? ? P 7.5 Å and an interaction energy of 27.2 kcal mol21 per {Ph4}2. The additional interactions are a centrosymmetric P4PE with P ? ? ? P 9.0 Å and an interaction energy of 22.3 kcal mol21 per {Ph4}2, and a second interaction with P ? ? ? P 8.2 Å and an energy of 25.9 kcal mol21 per {Ph4}2.Inter-Ph4P1 energies The calculated inter-cation energies are the sum of the van der Waals and coulombic components, and are expressed in two ways: (a) as the energy due to the embracing phenyl rings, that is3162 J. Chem. Soc., Dalton Trans., 1998, 3155–3165 Table 7 Crystal structures with two-dimensional nets of Ph4P1 cations, in which each cation is surrounded by four other cations and the net is pleated Refcode FARXEB JAGMIN10 WEYZUV YAPHOM YEYXUV YIBLAW YUCSEU ZEHTIP Anion and other components of the lattice [Br4W(m-S)(m-S2)WBr4]22, CH2Cl2, H2S [{(m-Se4)Ag}2]• [NCCH2(B6H6)]2 [VL2]2 [In(SCH2CH2S)2]2 [I(B6H6)]2 [Me(B6H6)]2 Cl3 2 Planarity of the net Pleated Pleated Pleated Pleated Slightly pleated Pleated Pleated Pleated P ? ? ? P distance (Å) and type for the principal propagating embrace 7.53 O4PE 7.08 O4PE 7.50 O4PE 7.92 O4PE 7.36 7.51 O4PE 7.41 O4PE 7.89 O4PE P ? ? ? P distance (Å) and embrace type in the other dimension 8.92 8.38, 8.82 both P4PE 8.34, 8.83 both P4PE 8.56 8.96 O4PE 8.23, 8.95 both P4PE 8.18, 8.73 both P4PE 8.14, 8.53 both P4PE L = 2O2CCH2N(O2)CH2CO2 2. per pair of {Ph4}, to focus on the reason for the embrace; (b) with inclusion of the additional contributions from the partially positive P atoms, and including all of the defined cation– cation interactions in the first sphere around Ph4P1 to give a total energy per single Ph4P1 cation.For the most common Fig. 7 Representations of the crystal structure of Ph4P1[NCCH2( B6H6)]2 [ WEYZUV] which contains corrugated two-dimensional networks with O4PE and P4PE.(a) View towards the network. The 7.5 Å arrowed interaction is the O4PE (the cell a axis), and the 8.3 and 8.8 Å arrowed interactions are P4PEs. Ph ? ? ? Ph energies smaller than 1.0 kcal mol21 are not marked. The calculated energies between cations are 28.3 kcal mol21 for the O4PE, 24.8 kcal mol21 for the 8.3 Å pair, and 22.1 kcal mol21 for the 8.8 Å interaction, all per {Ph4}2.Within the network the total interaction energy per cation is 212.6 kcal mol21 [or 26.0 kcal mol21 per Ph4P1]. (b) View parallel to the layers, showing their corrugation and the interleaving anions: B black. motifs, the sum of the additional contributions due to the P atoms is calculated to be 13.3 kcal mol21 (per {Ph4P1}2) for a 6PE, 13.0 kcal mol21 (per {Ph4P1}2) for an O4PE and 12.8 kcal mol21 (per {Ph4P1}2) for a P4PE.Table 9 summarises the calculated interaction energies for the various networks of embracing Ph4P1 cations. In the ladder structures BITXUX and WEYYUU, the calculated interaction energies per {Ph4} for the components of the ladder are ca. 27 kcal mol21 for the O4PE sides, and ca. 23 kcal mol21 for the P4PE or other interactions as rungs. With inclusion of contributions from the P atoms, each cation in the ladder is attracted to neighbouring cations by 24.9 (BITXUX) or 23.3 (WEYYUU) kcal mol21.For the layered arrays of cations, the inter-cation energies within the layers vary according to the quality of the embraces: when assessed for the Ph rings alone the attractive energy per cation ranges from 24.8 to 213.8 kcal mol21 in the best structure, [Ph4P1]2[BrAg(m-Br)2- Fig. 8 The two-dimensional network of Ph4P1 cations in [Ph4P1]2- [BrAg(m-Br)2AgBr]22 [GIDWEV]. There are centrosymmetric P4PEs with P ? ? ? P 7.5 Å and calculated energy of 28.3 kcal mol21, O4PEs with P ? ? ? P 7.9 Å (cell translation) and calculated energy of 24.8 kcal mol21, while the P ? ? ? P 8.5 Å interactions occurring as a zig-zag generated by a 21 screw axis each involve a pair of phenyl rings with total energy 24.6 kcal mol21.Within the network the total interaction energy per cation is 213.8 kcal mol21 [26.6 kcal mol21 per Ph4P1].J. Chem. Soc., Dalton Trans., 1998, 3155–3165 3163 Table 8 Crystal structures with layers of Ph4P1 cations in which each cation is surrounded by five others, as in Fig. 8 Refcode DEHKAC GIDWEV KEWFEX SEHFUG SIHRAC TUJYEC WEZVUS Anion [ICu(m-I)2CuI]22 [BrAg(m-Br)2AgBr]22 [{(m-I)(m3-I)2(m4-I)Ag3}2]• [F2Cl2Re(NSSN)]2 [(CO)4MoL]2 [SeCl4]22, (CH3CN)2 [m-Cl(NClS)2]2 P ? ? ? P distances (Å) and motifs for principal chains in the layer 7.82 O4PE 7.91 O4PE 8.50 O4PE 7.13 O4PE 7.48 O4PE 7.26 O4PE 7.53 O4PE P ? ? ? P distances (Å) for longer motifs 7.69 P4PE, 8.57 7.48 P4PE, 8.46 7.78, 8.85, 8.87 all P4PE 8.90 P4PE, 8.95 8.56 P4PE, 8.94 8.26 P4PE, 8.92 8.73, 8.98 P4PE L = 2,29-bipyridine-5-sulfonate. AgBr]22 [GIDWEV] (Fig. 8) which has corrugations and displacements in the layers. For this structure the full calculated energy for each cation with its surrounding cations is 26.6 kcal mol21. Only the poorest of the layered motifs, Ph4P1[SAs(S7)]2 [KIYJIL], has a small (0.7 kcal mol21) net repulsion between cations. The three-dimensional network of Ph4P1 using 4PE, [COETTE] (Fig. 9), has a total interaction energy per cation of 26.7 kcal mol21 per Ph4P1. Discussion We have shown instances of Ph4P1 embracing mainly through 4PE in the six diVerent network types shown in Fig. 10, and also a three-dimensional network. A total of 43 diVerent crystal structures and 50 diVerent compounds are included in this classification. The LIT4PE network, Fig. 10(a), was described previously10 (and there are now some 70 structures in the 1997 CSD which adopt this motif) while the others are new.There Fig. 9 The three-dimensional network of Ph4P1 cations in [Ph4P1]- [Cl3Te(OCH2CH2O)]2 [COETTE] viewed along the LIT4PE chains, with the pseudo-hexagonal array of these chains outlined. The independent P ? ? ? P distances and prominent Ph ? ? ? Ph energies are marked. In the O4PE which are perpendicular to the plane of this figure, the individual Ph ? ? ? Ph energies for the four ef interactions are 21.4, 21.4, 21.9 and 21.9 kcal mol21.The anions are located in pairs within the channels in the cation network: Te black, Cl speckled. are two types of ladder, where the sides are constructed from O4PE [Fig. 10(b) and Fig. 10(c)]. The rectangular array of Fig. 10(d) has been demonstrated with orthogonal 4PE in both directions, parallel 4PE in both directions, and with O4PE in one direction and P4PE in the other: these variations arise by rotation of the Ph4P1 cations about real or pseudo-two-fold axes in or perpendicular to the layer plane. In general the O4PEs in these nets are tighter and more attractive than the P4PEs. The O4PE usually results from unit cell translation (and hence very many of the compounds described here have one cell dimension of about 7.5 Å), while the P4PE often surrounds a centre of symmetry.The relative conformations of the two rings on each cation which form the P4PE dictate its quality. The P4PE has two components, the oV interaction and the two ef Fig. 10 Networks of Ph4P1 cations (symbolised as circles) in crystals, representing the one- and two-dimensional concerted supramolecular motifs comprised of fourfold phenyl embraces. (a) is the same as the LIT4PE.The connecting lines represent well-defined local 4PE (or in some instances, embraces involving three phenyl rings) in some crystals. In many crystals the concerted supramolecular motif is considered to be the complete network, which is more significant than the local 4PE.3164 J.Chem. Soc., Dalton Trans., 1998, 3155–3165 Table 9 Summary of the energy components (kcal mol21) for interacting cations in the networks described Other Total Refcode O4PEa P4PEa interactions a For Ph rings b For Ph4P1c Ladder BITXUX WEYYUU 27.8 26.8 23.3 22.6 210.0 29.4 24.9 23.3 Layer BEPZEB VUDTOD KIYJIL WEYZUV GIDWEV 27.8 25.6, 22.1 28.3 24.8 20.5 22.7, 22.1 24.8, 22.1 28.3 24.6 28.8 27.7 24.8 212.6 213.8 22.4 22.2 10.7 26.0 26.6 Labyrinth COETTE 27.2 22.3 25.9 215.1 26.7 a Energy per {Ph4}2 due to the embracing phenyl rings. b Energy, per cation, summed over the Ph ? ? ? Ph interactions for Ph4P1 ions in the first sphere around Ph4P1.c Energy, per Ph4P1, summed over the identified Ph4P1 ? ? ? Ph4P1 interactions to the first sphere around Ph4P1. interactions, but the energy derived from the oV interaction is the one which dominates the total energy of the interaction, and as a result there is wide variability in the energy associated with the P4PE.Ladders can be parallel, or canted: parallel ladders can be coplanar, or not. When they are coplanar, the obvious comparison is of the ladder (e.g. BITXUX, Table 2, Fig. 1) with a layer (e.g. BEPZEB, Table 4, Fig. 3), created simply by moving the ladder edges closer together. When the ladders are not coplanar, they can be envisaged as creating puckered layers when brought closer together. In both WEYYOO and WEYYUU (Table 3) the ladders are slightly non-coplanar, and when brought closer together form the slightly puckered layer of the isomorphous WEZVUS (Table 8).In JETNUR (not pictured) the ladders are distinctly non-coplanar. In the crystal packing arrangements classified in this paper there are segregated regions of cations and anions. This distinction from the conventional desegregated array of ions of opposite charge emphasises again 10,13 that the packing in crystals containing arylated cations, such as Ph4P1, is frequently dominated by attractive interactions between cations.The common occurrence of the identified cation–cation motifs in crystals with diVerent anions indicates that these cation-based supramolecular features could be more significant than cation– anion interactions. The anions in the compounds surveyed in this paper generally do not have arylated surfaces (C60 2 is the exception) and are not able to participate in phenyl embraces. In a subsequent paper we will describe the crystal supramolecularity which occurs when the anion is also able to engage in multiple phenyl embraces, as for example in Ph4P1Ph4B2 and related compounds.31 The size and shape of the anion are clearly factors in determining the crystal packing in a particular structure.The anions in the compounds included in this analysis are generally approximately the same size as the cation, although in some cases are smaller, for example [SnCl3]2 [BITXUX] and Br3 2 [BEPZEB]. Crystals of Ph4P1 with the comparable anions Br3 2, BrIBr2 and Cl3 2 are included in this survey. The first two are isostructural and form planar layer structures created by O4PE in one direction and P4PE in the other, but Ph4P1Cl3 2 contains a highly puckered layer.While the shape of the anions is the same, there is a significant diVerence in their volumes: Br3 2 73 Å3, Cl3 2 60 Å3. Only those crystals in which the anion is slightly larger than the cation form the P4PE/P4PE layer, namely [SAs(S7)]2 [KIYJIL] and the three C60 2X2 compounds.The larger anions appear to distend and weaken the cation array, as is evident by comparison of the energies for KIYJIL with the others in Table 9. In the case of the three C60 2X2 compounds there are compensating attractions between the cations and C60 2, and between the cations and X2. Why then does KIYJIL have a structure similar to the C60 2 compounds? The shape of the anion is not dissimilar to one half of a C60 molecule, and so it fits neatly into the same nest.There are attractive S ? ? ? phenyl interactions which add stability to this arrangement [Fig. 5(b)]. In some of the networks described in this paper there are local phenyl ? ? ? phenyl attractions involving more than two cations [see Figs. 1(a), 3, 7(a)], and therefore the networks are more than the sums of the standard phenyl embraces for cation pairs. We regard these larger networks as the concerted supramolecular motifs demonstrated by these crystals.As for our previous classifications of the ZZI6PE and the LIT4PE as one-dimensional supramolecular motifs based on multiple phenyl embraces, we consider that the networks illustrated in Fig. 10 provide a valuable conceptual foundation of concerted supramolecular motifs for crystal engineering using Ph4P1 cations. In support of this contention we note that the crystal structures described in this paper are free of crystallographic disorder of the cations, indicating the significance of these extended multiple phenyl embraces.In the following paper 14 we describe further two- and threedimensional networks of Ph4P1 cations in crystals, where the 6PE is the dominant pairwise motif. Acknowledgements This research is funded by the Australian Research Council through a project grant and through VisLab computing facilities. References 1 J. D. Dunitz, Pure Appl. Chem., 1991, 63, 177. 2 G. R. Desiraju, The Crystal as a Supramolecular Entity, Perspectives in Supramolecular Chemistry, ed.J. M. Lehn, John Wiley, Chichester, 1996. 3 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 131. 4 I. G. Dance and M. L. Scudder, J. Chem. Soc., Chem. Commun., 1995, 1039. 5 I. G. Dance, in The Crystal as a Supramolecular Entity, ed. G. R. Desiraju, John Wiley, New York, 1996, pp. 137–233. 6 I. Dance and M. Scudder, Chem. Eur. J., 1996, 2, 481. 7 U. Muller, P. Klingelhofer, J. Eicher and R. Bohrer, Z.Kristallogr., 1984, 168, 121. 8 I. Dance and M. Scudder, J. Chem. Soc., Dalton Trans., 1998, 1341. 9 I. Dance and M. Scudder, Polyhedron, 1997, 16, 3545. 10 I. Dance and M. Scudder, J. Chem. Soc., Dalton Trans., 1996, 3755.J. Chem. Soc., Dalton Trans., 1998, 3155–3165 3165 11 U. Muller, Acta Crystallogr., Sect. B, 1980, 36, 1075. 12 C. Hasselgren, P. A. W. Dean, M. L. Scudder, D. C. Craig and I. G. Dance, J. Chem. Soc., Dalton Trans., 1997, 2019. 13 M. Scudder and I. Dance, J. Chem. Soc., Dalton Trans., 1998, 329. 14 M. Scudder and I. Dance, following paper. 15 F. H. Allen, J. E. Davies, J. J. Galloy, O. Johnson, O. Kennard, C. F. Macrae and D. G. Watson, Chem. Inf. Comput. Sci., 1991, 31, 204. 16 MSI, http://www.msi.com, 1998. 17 A. J. Pertsin and A. I. Kitaigorodsky, The atom-atom potential method. Applications to organic molecular solids, Springer Series in Chemical Physics, Springer-Verlag, Berlin, 1987. 18 A. Gavezzotti, Theoretical Aspects and Computer Modelling of the Molecular Solid State, Molecular Solid State Series, John Wiley, Chichester, 1997. 19 D. S. Coombes, S. L. Price, D. J. Willock and M. Leslie, J. Phys. Chem., 1996, 100, 7352. 20 A. Penicaud, A. Perez-Benitez, R. Gleason, E. Munoz and R. Escudero, J. Am. Chem. Soc., 1993, 115, 10392. 21 A. L. Balch, V. J. Catalano and J. W. Lee, Inorg. Chem., 1991, 30, 3980. 22 A. L. Balch, V. J. Catalano, J. W. Lee, M. M. Olmstead and S. R. Parkin, J. Am. Chem. Soc., 1991, 113, 8953. 23 A. L. Balch, V. J. Catalano, J. W. Lee and M. M. Olmstead, J. Am. Chem. Soc., 1992, 114, 5455. 24 A. L. Balch, J. W. Lee and M. M. Olmstead, Angew. Chem., Int. Ed. Engl., 1992, 31, 1356. 25 A. L. Balch, J. W. Lee, B. C. Noll and M. M. Olmstead, J. Chem. Soc., Chem. Commun., 1993, 56. 26 J. L. Atwood, G. A. Koutsantonis and C. L. Raston, Nature (London), 1994, 368, 229. 27 C. L. Raston, J. L. Atwood, P. J. Nichols and I. B. N. Sudria, Chem. Commun., 1996, 2615. 28 J. L. Atwood, M. J. Barnes, M. G. Gardiner and C. L. Raston, Chem. Commun., 1996, 1449. 29 O. Ermer and C. Röbke, J. Am. Chem. Soc., 1993, 115, 10077. 30 O. Ermer, Helv. Chim. Acta, 1991, 74, 1339. 31 B. F. Ali, D. C. Craig, I. G. Dance, A. D. Rae and M. L. Scudder, unpublished work. Paper 8/03463J

ISSN:1477-9226    

DOI:10.1039/a803463j

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3167–3175 3167 Crystal supramolecular motifs. Two- and three-dimensional networks of Ph4P1 cations engaged in sixfold phenyl embraces Marcia Scudder and Ian Dance* School of Chemistry, University of New South Wales, Sydney 2052, Australia. E-mail: I.Dance@unsw.edu.au Received 8th May 1998, Accepted 23rd July 1998 The sixfold phenyl embrace (6PE) which is a widespread supramolecular motif for Ph4P1 cations in crystals can occur as a zig-zag chain motif (ZZI6PE) in which each Ph4P1 engages two 6PE, and these ZZI6PE allow further multiple phenyl embraces at each Ph4P1 cation.The Cambridge Structural Database has been explored for occurrences of the two- and three-dimensional networks of multiple phenyl embraces involving the common cation Ph4P1. A prevalent motif involves parallel ZZI6PE chains linked by fourfold phenyl embraces (the parallel variant, P4PE) to form skewed hexagonal nets, called ZZI6PE/P4PE layers.Another type has ZZI6PE chains running in orthogonal directions and linked by the orthogonal variant of the 4PE: these are the orthogonal ZZI6PE/O4PE three-dimensional nets. An expanded layer involves eight-rings of Ph4P1 built from finite sequences of 6PE. The compounds [Ph4P1]2[Te4]22 and [Ph4P1]4[(S6)Cu(S8)Cu(S6)]42 contain more elaborate three-dimensional nets comprised of 6PE, 4PE and 3PEs. These nets are more highly concerted crystal supramolecular motifs. The nets described are diVerent from those based on 4PE, and the diVerence is related to the charge of the anion. Anions are usually located over or in the cycles of cations in the layered motifs, and in cavities of the three-dimensional motifs.The concerted supramolecular motifs of cations, rather than cation–anion attractions, are believed to be the dominant factors in the crystal supramolecularity. Calculated energies for the motifs are reported. It is concluded that a single Ph4P1 cation can participate in up to four multiple phenyl embraces, and that the maximum attractive interaction energy per Ph4P1 cation is ca. 9 kcal mol21. This is comparable with the energies of hydrogen bonds. Introduction A feature of the supramolecularity of phenylated molecules is the widespread occurrence of multiple phenyl embraces, in which individual intermolecular phenyl ? ? ? phenyl attractive interactions operate in concert with significant net attraction.1–7 These supramolecular motifs have been recognised through analysis of the packing in molecular crystals (supramolecular entities par excellence 8,9) using the voluminous accurate geometrical data in the Cambridge Structural Database.10 The characteristic attributes of multiple phenyl embraces are (1) participation of two or more phenyl groups from each partner molecule, (2) geometrical concertedness, and (3) strong attraction.The most prevalent multiple phenyl embrace is the sixfold phenyl embrace, 6PE, in which three phenyl rings from each molecule are involved in six edge-to-face local phenyl ? ? ? phenyl attractions: the net attractive energy of the 6PE is in the range 8–11 kcal mol21.† The other major type is the fourfold phenyl embrace, 4PE, which has orthogonal and parallel variants, namely the O4PE and P4PE respectively.The 6PE, O4PE and P4PE motifs are illustrated in the preceding paper,7 where we described how Ph4P1 cations in crystals can form one-, two- and three-dimensional networks constructed from 4PE.In this paper we describe two- and three-dimensional networks comprised of Ph4P1 cations linked mainly by 6PE. We have previously identified and described the one-dimensional ziz-zag infinite chain of sixfold phenyl embraces (ZZI6PE) involving Ph4P1. In the ZZI6PE, each 6PE uses three phenyl rings or one face of the Ph4P1 tetrahedron, and the two faces or 6PE are therefore inclined at the tetra- † These intermolecular energies are calculated as the sum of the van der Waals and coulombic atom–atom components, dependent on empirical parameters which are under continuing revision: more recent energy values may diVer from earlier published values.hedral angle at the P atom, giving rise to the zig-zag.11 In the ZZI6PE the required threefold array within each of the two sets of three phenyl rings for each 6PE is allowed by the intramolecular constraints on the conformations of the phenyl rings in Ph4X.12 At each Ph4P1 in the ZZI6PE the two phenyl rings on the interior of the bend are engaged in the two 6PE, but each of the two phenyl rings on the exterior of the bend is engaged in only one 6PE, and is still available for involvement in further multiple phenyl embraces.It is these additional embraces which give rise to the networks described in this paper. Methodology The use of the Cambridge Structural Database, and the calculational procedures for evaluation of attractive energies, are described in the preceding paper.7 van der Waals radii used in the calculations of volumes are Fe 2.0, S 1.8, Br 1.85, I 1.98 and H 1.1 Å.In the Fig. captions the calculated energies of specific multiple phenyl embraces (identified by their P ? ? ?P distances) are quoted in kcal mol21 per {Ph4}2 set, and [in3168 J. Chem. Soc., Dalton Trans., 1998, 3167–3175 Fig. 1 Representation of the multiple phenyl embraces in crystalline [Ph4P1]2[Pb(Se4)2]22 [VIPTAP].The chains of 6PE are denoted by purple rods, and the P4PE motifs which connect them are shown as orange rods, with a minimal set of phenyl rings which comprise the interactions also drawn. There are two crystallographically distinct ZZI6PE chains, with 6PE interactions identified by their P ? ? ? P distances of 6.09, 6.55 Å alternating in one chain and 6.66, 6.90 Å alternating in the other: all P4PE interactions have P ? ? ? P 8.00 Å, and all P atoms lie in the same plane.One anion, located over the approximately hexagonal gap in the cation layer, is shown. The calculated energies of the embraces (identified by their P ? ? ?P distances) are given in kcal mol21 per {Ph4}2 set, and in square brackets per {Ph4P}2 set: 6PE at 6.09 Å, 210.4 [27.4]; 6PE at 6.55 Å 26.8 [23.8]; 6PE at 6.66 Å 28.0 [24.6]; 6PE at 6.90 Å 25.8 [22.7]; P4PE at 8.00 Å 25.9 [22.8] kcal mol21. square brackets] per {Ph4P}2 set. In Table 3 the energies are presented as net attractive energies between one Ph4P1 cation and its neighbouring Ph4P1 cations.Results The hexagonal ZZI6PE/P4PE layer The two exterior phenyl rings just mentioned can readily form a P4PE, as is illustrated in Fig. 1, for the compound [Ph4P1]2[Pb- (Se4)2]22 [VIPTAP]. The ZZI6PE chains are denoted by purple rods, and the P4PE motifs which connect them are shown as orange rods. This motif, called the hexagonal ZZI6PE/P4PE layer, has an approximately planar net of P atoms.The characteristic of the parallel 4PE is that the Cipso–P–Cipso planes on the two molecules are exactly or almost coplanar (see previous paper 7), and consequently the two phenyl rings at each P have inequivalent roles. Therefore the contiguous ZZI6PE chains in the ZZI6PE/P4PE layer do not exactly oppose each other, but are partly displaced along their length in order to optimise the geometry of the P4PE, as is evident in Fig. 1. This collection of cations forms a concerted two-dimensional motif, in which each cation is three-connected, and the array of cations is distorted hexagonal. The calculated attractive energies of the various interactions are provided in the caption to Fig. 1. Table 1 lists the many other instances of this crystal packing type. One distinctive characteristic of crystals which adopt this hexagonal ZZI6PE/P4PE layer motif is their space group, which is almost invariably P1� . It is to be noted that the anions are generally small, and that the anionic metal complexes have monatomic or oligoatomic ligands.Almost all of the anions have double negative charge, with the consequence that one anion is associated with each distorted hexagon of cations, as in Fig. 1. In general (although not in VIPTAP, Fig. 1), the anion is situated at one of the inversion centres between the layers. Solvent molecules occupy ttice holes in a number of structures, and are related by another of the inversion centres. When the charge is single negative, there are two anions associated with each hexagon, related by one of the inversion centres: BUNHEX, COGHEL, PASWIP, BIHDAX, TAGFAI and JEDIAC01 fall into this category.The predominance of space group P1� is entirely in accord with, and a demonstration of, a principle of crystal packing enunciated by Brock and Dunitz.13 They drew attention to the favourability of the centre of inversion as a symmetry element for intermolecular motifs, and the unfavourability of rotation axes and mirror planes as intermolecular symmetry elements.Both the 6PE and the P4PE can accommodate a centre of inversion without distortion: the 6PE requires heterochiral partners, and is disposed towards centrosymmetry. The space group which combines only centres of inversion and translation is P1� . Intramolecular symmetry possible for Ph4P1 is restricted to S4 and subsets D2 and C2, but since these cannot be incorporated in the 6PE or P4PE motifs, the cation array can only support space group P1� .In most of the few examples in Table 1 with higher symmetry space groups, the anions are located on two-fold special positions, and the embraces populate the centres of inversion (except in the cases of FAZPEB and FOBNAL). There are two examples in Table 1 which include mixed cations, JEGBAY and ZIPROF, both of which contain a similar pseudo-hexagonal array of multiply embracing Ph4P1 cations. In [Ph4P1]2 [Bu4N1]2[Fe8S9Cl6]42 [JEGBAY] there is one anion associated with each pseudo-hexagon of cations.However, the anion is 42, and the crystal also contains Bu4N1 cations equal in number to the Ph4P1 cations. As shown in Fig. 2 there is a clear segregation of the Bu4N1 cations fromJ. Chem. Soc., Dalton Trans., 1998, 3167–3175 3169 Table 1 Crystal structures containing the hexagonal ZZI6PE/P4PE layer. All structures have space group P1� unless otherwise stated. The listing groups compounds with chemically similar anions Refcode CUTMEJ VIKJAA JAVYOU KAVHEU KAVHIY GAYWIM BIRHAL10 FUYXEC GAZNOK GAZNUQ JAVKEW COYWES COYWIW SIWKUE FUWSAR TOSYIJ TOSYUV DEMYEZ01 VAXDUT FAZPEB C2/c FOBNAL Pbcn BUNHEX COGHEL CEXRIG VIPTAP KAZBUI10 PERGIC YOLGEL YOLGIP ZAZBIL VOFMOS VOFNAF C2/c RARZEP PASWIP TAGFAI WACXAZ WACXED DURXUJ C2/c BIHDAX JEDIAC01 P21/c POJRAH02 Ccca VECGUF10 VOXWOU ZIPROF ZAQRUE01 JEGBAY C2/c JANHIP10a KOMWUEb PAHPODc Anion [Br2Se(m-Br)2SeBr2]22 [Cl2Se(m-Cl)2SeCl2]22 [Se2W(m-Se)2WSe2]22 [Cl2Zn(m-Cl)2ZnCl2]22 [Cl2Cd(m-Cl)2CdCl2]22 [Br2Mn(m-Br)2MnBr2]22 [OCl2Se(m-Cl)2SeCl2O]22 [WS2Cl4]22?2CH2Cl2 [RuCl6]22?2CH2Cl2 [UCl6]22?2CH2Cl2 [ZrCl6]22?2CH2Cl2 [WOCl5]22?2CH2Cl2 [ReOCl5]22?2CH2Cl2 [NbSCl5]22?2CH2Cl2 [Cl4ReReCl4]22?2CH2Cl2 [NC-Se-B6H5]22 [NC-Se-B12H11]22?CH3CN Br2 Br2?H2O [S(S2)2W(m-S)W(S2)2S]22?CH3CN [O(S2)2W(m-S)W(S2)2O]22?CH3CN [MoBr3(NO)2OH2]2 [RuCl4(NS)OH2]2 [OsCl6]22?DMF [Pb(Se4)2]22 [Pt(Se4)3]22?DMF [Tl2(S4)2]22 [(CO)4Mo(m-S)2Mo(m-S)2Mo(CO)4]22 [(CO)4W(m-S)2W(m-S)2W(CO)4]22 [Ag2Se6Te3]22 [(m-Se3)(m-Se2)Au2]22 [(m-Se4)(m-Se2)Au2]22 [OV{OP(O)(CH3)OP(O)(CH3)O}2]22 [OAs(OH)3]Cl2 [CH3C(]] NH)Cl]Cl2 [Hg(c-C6H8S2)2]22?4H2O [Cd(c-C6H8S2)2]22?4H2O [Pd(SCH2CH2S)2]22?4H2O [Me2SMoCl2(m-Cl)3MoCl2SMe2]2 [SRe(S4)2]2 [{Cu3I4}2]• [OW(CN)6]22?H2O [Cl2Sb(m-S)(m-Cl)2SbCl2]22?CH3CN [Mo6Se2Cl12]42[H3O1]2 ?4CH3OH [SeSb4(Se)4Se]22 [(Cl3Fe4S4)(m-S)(Fe4S4Cl3)]42[Bu4N1]2[Ph4P1]2 [(Se4)2InSe5In(Se4)2]42 [Cl3W(m-SEt)(m-Cl)(m-SEt2)WCl3]22?CH2Cl2 [{Cr(CO)5}2TeTe{Cr(CO)5}2]22?CH2Cl2 P ? ? ? P distances in the ZZ16PE/Å 6.51, 6.69 6.34, 6.54 6.47, 6.89 6.52, 6.71 6.58, 6.63 6.59, 6.60 6.45, 6.49 6.37, 6.77 6.38, 6.78 6.42, 6.83 6.41, 6.81 6.37, 6.76 6.34, 6.73 6.37, 6.81 6.13, 6.61 6.15, 6.27 6.08, 6.16 6.33, 6.40 6.31, 6.39 6.53, 6.53 6.68, 6.68 6.51, 6.58 6.70, 6.80 6.37, 6.58 6.09, 6.55 6.66, 6.90 6.26, 6.89 6.26, 6.89 6.41, 6.44 6.45, 6.60 6.48, 6.70 6.47, 6.66 6.22, 6.29 6.20, 6.29 6.44, 6.44 6.28, 6.54 6.27, 6.43 6.14, 6.62 6.58, 6.65 6.07, 6.58 6.62, 6.64 6.06, 6.62 6.12, 6.86 6.32, 6.34 6.93, 6.93 6.42, 6.42 6.56, 6.95 6.31, 6.42 6.42, 6.68 6.47, 6.71 6.05, 6.72 6.34, 6.88 6.43, 6.99 6.34, 6.34 6.39, 6.90 P ? ? ? P distances in the P4PE/Å 8.36 8.18 8.05 8.14 8.25 8.42 8.39 8.43 8.42 8.55 8.50 8.37 8.41 8.47 8.45 7.99 8.30 7.92 7.91 8.73 8.98 8.55 8.42 8.57 8.00 8.78 8.54 8.88 8.36 8.32 8.13 7.96 8.20 8.05 8.98 8.11 8.41 skewed P4PE 8.42 skewed P4PE 8.34 8.58 8.79 8.70 7.15 8.75 8.75 8.50 7.98 8.86 8.60 8.53 8.96 a Additional chains of alternating 6PE (6.22 Å) and P4PE (8.37 Å).b Additional layers composed of 6PE (5.73) and P4PE (7.97 and 8.10 Å). c Additional layers composed of 6PE (6.46) and P4PE (8.10 and 8.28 Å). the Ph4P1 cations in the hexagonal ZZI6PE/P4PE array. This segregation is a manifestation of the dominance of the attractive energies in the two-dimensional network of multiple phenyl embraces. The orthogonal ZZI6PE/O4PE net In the ZZI6PE the two phenyl groups exterior to the chain at each cation are symmetrically arrayed in the plane of the chain.Two ZZI6PE chains orthogonal to each other could therefore form an O4PE between pairs of these orthogonal phenyl groups, and this motif indeed occurs, albeit usually distorted, in the crystal structures of the compounds listed in Table 2. A good example is [Ph4P1]2[Fe4S4(SH)4]22 [FAGREK], shown in Fig. 3. All except one of the compounds listed in Table 2 adopt space group C2/c where the equivalent ZZI6PE chains align with the ab diagonals of the unit cell.However, these structures are not as ideally orthogonal in the 4PE as hinted above because the two Cipso–P–Cipso planes are not orthogonal. Table 2 contains details of the angles between the normals to the two Cipso–P–Cipso planes involved in the 4PE: these values range from 38 to 688, while in the ideal O4PE this angle is 908. It is evident from Fig. 3 that this apparent distortion is mainly local.The crystal structure of [Ph4P1]2[Fe4S4Br4]22 [DEXXIN10] is very similar to that of the related compound [Ph4P1]2-3170 J. Chem. Soc., Dalton Trans., 1998, 3167–3175 Fig. 2 Representations of the crystal structure of [Ph4P1]2[Bu4N1]2[Fe8S9Cl6]42 [JEGBAY]. (a) View of the hexagonal ZZI6PE/P4PE layer of Ph4P1 cations, and the locations of the anions under the hexagons. (b) Side view of the layered structure, showing the segregated layer of Ph4P1 cations, with the Bu4N1 cations interspersed in the layer of anions. [Fe4S4(SH)4]22 [FAGREK] shown in Fig. 3. In both, the planes of the ZZI6PE sequences (as defined by the P atoms) are not exactly orthogonal, just as the Cipso–P–Cipso planes of the 4PE are not orthogonal (Table 2). However, the homologous compound [Ph4P1]2[Fe4S4I4]22 [KAKFIL] (Fig. 4) adopts the higher symmetry space group I41/a, crystallising with exact orthogonality between the directions of the ZZI6PE chains and between the normals to the P ? ? ?P? ? ? P planes of the ZZI6PE chains.Within the O4PE, the two Cipso–P–Cipso planes are almost orthogonal. The P ? ? ? P vector of the O4PE lies in the planes of both ZZI6PE chains. In this structure there is only one type of 6PE, one type of ZZI6PE, and one type of O4PE. We have sought the reason for the subtle diVerence between [Ph4P1]2[Fe4S4Br4]22 [DEXXIN10] and [Ph4P1]2[Fe4S4I4]22J. Chem. Soc., Dalton Trans., 1998, 3167–3175 3171 Table 2 Crystal structures containing two orthogonal ZZI6PE chains linked by a further 4PE Refcode BECPII10a POPCECa SUPZIMa,b ZUPFIZa FURHEFa GEPFIQa GEPFOWa TADGAGa YUBDUUa NAXXALa DEXXIN10c FAGREKc KAKFIL DAMNUA Anion [NCCu(m-S)2MoS2]22 [NCCu(m-Se)2MoSe2]22 [NCCu(m-Se)2WSe2]22 [NCAu(m-Se)2MoSe2]22 [ClCo(N3)3]22 [MoSe4]22 [WSe4]22 [NiCl4]22 [S3NbSH]22 [CdBr4]22 [Fe4S4Br4]22 [Fe4S4(SH)4]22 [Fe4S4I4]22 space group I41/a [Fe4S4Cl2{(m-S)2CN(Et)2}2]22 P ? ? ? P distances along ZZI6PE chain/Å 6.45, 6.66 6.46, 6.64 6.54, 6.62 6.44, 6.85 6.47, 6.48 6.47, 6.50 6.37, 6.50 6.42, 6.49 6.41, 6.65 6.56, 6.69 6.53, 6.69 6.68 6.54, 6.95 P ? ? ?P distance in the 4PE/Å 7.30 7.31 7.25 7.26 7.38 7.32 7.28 7.35 7.41 7.37 7.48 7.48 8.63 Angle between normals to the two Cipso–P&ndashlanes of the 4PE/8 39.0 38.6 39.2 40.1 42.4 42.5 41.0 41.5 32.9 67.6 66.9 87.2 59.2 a These structures all have unit cells of similar dimensions, 11 × 20 × 20 Å, b ª 928.b No coordinates available, isostructural with POPCEC.c These two structures have unit cells of similar dimensions, 16 × 14 × 24 Å, b ª 1108. Fig. 3 Representation of the crystal structure of [Ph4P1]2[Fe4S4(SH)4]22 [FAGREK], with ZZI6PE chains denoted by the purple rods, and the O4PE motifs as yellow rods. One set of ZZI6PE chains runs top to bottom, the other almost perpendicular to the figure. An asymmetric set of phenyl groups is included, and the anion locations between the ZZI6PE chains are marked.The calculated energies of the embraces (identifi ed by their P ? ? ? P distances) are given in kcal mol21 per {Ph4}2 set, and in square brackets per {Ph4P}2 set: 6PE at 6.53 Å, 29.1 [26.1]; 6PE at 6.69 Å, 29.0 [25.4]; O4PE at 7.48 Å, 25.9 [22.8] kcal mol21. [KAKFIL]. We believe that the network of embracing Ph4P1 around [Fe4S4Br4]22 has distorted in order to contract the dimensions of the cavity containing the anion, and that the larger [Fe4S4I4]22 anion has the correct volume for the ideal orthogonal ZZI6PE/O4PE network. The volumes of the anions in the three crystal lattices discussed here are [Fe4S4(SH)4]22 [FAGREK] 231 Å3, [Fe4S4Br4]22 [DEXXIN10] 238 Å3, and [Fe4S4I4]22 [KAKFIL] 264 Å3, supporting this view.Corrugated sheets with eight-rings of Ph4P1 The compound [Ph4P1]3[CRe7(CO)21]32 [BEJBOH] crystallises with a two-dimensional network of cations engaged in diVerent types of multiple phenyl embraces. The essential features of the crystal packing are shown in Fig. 5. The cation array is comprised of centrosymmetric sequences of five 6PE, linked by P4PE (P ? ? ? P 8.43 Å) and 4PE (P ? ? ? P 7.23 Å) interactions into eight-membered cycles, and the anions lie over these eightrings. Note that the larger eight-ring here, relative to the sixrings in the pseudo-hexagonal cation arrays already described, accommodates the larger size of the anion and its 32 charge: the stoichiometry of cation contributions in an eight-ring is (6 × 1 3 – ) 1 (2 × ��� ) = 31.The sheets of cations are corrugated (see Fig. 5) and the anions nestle in the hollows. More elaborate three-dimensional networks of Ph4P1 Fig. 6 represents the crystal lattice of [Ph4P1]2[Te4]22 [KIZWEV] which contains two diVerent well-developed zig-zag chains of SPE, both running in the same direction, and linked in the3172 J. Chem. Soc., Dalton Trans., 1998, 3167–3175 Fig. 4 Comparison of the crystal structures of [Ph4P1]2[Fe4S4Br4]22 [DEXXIN10] in space group C2/c (left) and [Ph4P1]2[Fe4S4I4]22 [KAKFIL] in space group I41/a (right).Colour coding is as for Fig. 3. The slight distortions of the lattice in DEXXIN10 are obvious. Only one 6PE of the ZZI6PE chain parallel to the viewing direction can be seen. Fig. 5 Simplification of the crystal structure of [Ph4P1]3[CRe7(CO)21]32 [BEJBOH]. Purple rods signify 6PE, orange rods are P4PE, and the white rods are asymmetric 4PE. This view of the P4PE at 8.43 Å illustrates well the oVset-face-to-face and two edge-to-face Ph ? ? ? Ph interactions characteristic of P4PE.Note that the cation involved in the 6PE identified as 6.37, 6.29 Å is folded out of the cation layer, and that this occurs in such a way that the sheet of cations is corrugated left-right across the figure. The calculated energies of the embraces (identified by their P ? ? ?P distances) are given in kcal mol21 per {Ph4}2 set, and in square brackets per {Ph4P}2 set: 6PE at 6.29 Å, 28.9 [25.9]; 6PE at 6.37 Å, 26.7 [23.3]; 6PE at 6.75 Å, 24.8 [21.9]; 4PE at 7.23 Å, 25.2 [22.6]; P4PE at 8.43 Å, 24.1 [21.0] kcal mol21.other two dimensions. One set of ZZI6PE (P ? ? ? P 5.98, 6.34 Å) is linked by P4PE to form the hexagonal ZZI6PE/P4PE layer already described, while the other set of ZZI6PE (P ? ? ? P 6.39, 6.72 Å) does not contain these links. The two sets of ZZI6PE alternate through the crystal lattice, and are linked by 3PE in which a phenyl ring on one cation is directed between the pairJ. Chem.Soc., Dalton Trans., 1998, 3167–3175 3173 Fig. 6 Representation of the labyrinth of multiple phenyl embraces in crystalline [Ph4P1]2[Te4]22 [KIZWEV]. The purple rods signify 6PE, orange P4PE, and grey 3PE (P ? ? ? P 7.95 Å) in which one phenyl ring of one cation is attracted to two phenyl rings on the adjacent cation. The calculated energies of the embraces (identified by their P ? ? ? P distances) are given in kcal mol21 per {Ph4}2 set, and in square brackets per {Ph4P}2 set: 6PE at 5.98 Å, 210.4 [27.2]; 6PE at 6.34 Å, 29.9 [26.2]; 6PE at 6.39 Å, 29.2 [26.1]; 6PE at 6.72 Å, 27.6 [24.4]; P4PE at 7.95 Å, 25.8 [22.6]; 3PE at 8.79 Å, 24.7 [21.9] kcal mol21.of phenyl rings external to an adjacent ZZI6PE and forms attractive edge-to-face interactions with them. As shown in Fig. 6 the abnormally small anions thread between these 3PE. The majority of the crystal lattice is comprised of Ph4P1 cations in attractive embraces. An even more elaborate network of multiple phenyl embraces occurs in [Ph4P1]4[(S6)Cu(S8)Cu(S6)]42 [COVCAR], shown in Fig. 7. ZZI6PE chains are a prominent feature, and they occur as two types, one (type A) with P ? ? ? P separations of 6.29, 6.90 Å, the other (type B) 6.27, 6.97 Å. ZZI6PE chains of each type occur in layers, and form pseudo-hexagonal planar networks, although the connections (P ? ? ? P 8.75, 8.65 Å respectively) between the chains in each network are long and not significant.The layers are viewed almost edge-on in Fig. 7. Between the planar hexagonal layers (A, B) of cations there are puckered hexagonal nets of cations (type C), connected by 6PE and P4PE as fused six-rings in chair conformation. Cations in the planar layers and the puckered layers are further connected by 3PE (P ? ? ? P 8.55 Å for A to C and P ? ? ? P 8.39 Å for B to C). The sequence of layers of cations, evident in Fig. 7, is then –A–(3PE)–C–(3PE)–B–(3PE)–C–(3PE)–A–.Cations in the puckered layers (C) are four-connected, while cations in planar layers (A,B) are eVectively three-connected because the 8.65, 8.75 Å linkages are insignificant. The very flexible anion, in extended conformation, threads through this labyrinth of cations: each CuS6 chelate unit in the anion is located between a planar layer and a puckered layer of cations, in a pseudoadamantanoid cage of cations. Our interpretation of the crystal packing in this compound is that the aggregate of the attractive energies of the embracing cations will be the dominant contribution to the lattice energy, and that intramolecular energies for the anion will have negligible eVect.The calculated energies of the identified motifs are included in the caption to Fig. 7. We note that the network of alternating planar and puckered layers of six-rings, i.e. alternating graphitic and diamondoid layers, which occurs in [Ph4P1]4[(S6)Cu(S8)Cu(S6)]42 is present also in the crystal structure of a quite diVerent copper sulfide compound, the CuS mineral covellite.14 Lastly, we describe briefly a crystal structure which manifests a one-dimensional chain containing 6PEs which is diVerent from the ZZI6PE motif.In Ph4P1[CpMo(Se4)2]2 [SIRGAB] there is a ladder motif which is dominated by repeated 6PEs, but as illustrated in Fig. 8(a) there are other appreciable Ph ? ? ? Ph attractive energies within the ladder.Fig. 8(b) shows the separation of these layers by the [CpMo(Se4)2]2 anions. Discussion We have shown that sequences of 6PE, particularly as the ZZI6PE, are prevalent in crystals containing Ph4P1, and that they further associate through well developed multiple phenyl embraces to form two-dimensional and three-dimensional networks of Ph4P1. The combination of the ZZI6PE and the P4PE is the most common structure type, forming two-dimensional nets of cations arranged in hexagons whicskewed as a requirement of the P4PE.Alternatively, orthogonal ZZI6PE chains can be linked through O4PE. There is substantial chemical diversity in the anions which form these networks, particularly the ZZI6PE/P4PE net, although there is a preponderance of doubly negative anions and of anions containing chalcogenide and polychalcogenide metal complexes. This latter bias reflects laboratory custom, and the knowledge of preparative chemists exploring these areas that the Ph4P1 cation generates good crystals of such anions.We predict that more widespread use of Ph4P1 as a crystallising cation will broaden the scope of the crystals demonstrating the networks of multiple phenyl embraces described here. We note that the majority of the crystals described in the preceding paper,7 with networks based on the 4PE, contain mono-negative anions, while the majority of crystals forming more elaborate networks of 6PE (this paper) contain di-3174 J.Chem. Soc., Dalton Trans., 1998, 3167–3175 Fig. 7 Representation of part of the crystal packing and multiple phenyl embraces in [Ph4P1]4[(S6)Cu(S8)Cu(S6)]42 [COVCAR]. Purple rods signify 6PE, orange rods are P4PE, and grey rods are 3PE. The anion has CuS6 chelate rings linked by an extended S8 chain. The calculated energies of the embraces (identified by their P ? ? ? P distances) are given in kcal mol21 per {Ph4}2 set, and in square brackets per {Ph4P}2 set: 6PE at 6.04 Å, 210.2 [27.4]; 6PE at 6.27 Å, 28.5 [25.4]; 6PE at 6.29 Å, 27.4 [24.7]; 6PE at 6.90 Å, 28.5 [25.2]; 6PE at 6.97 Å, 26.8 [23.1]; P4PE at 8.07 Å, 25.2 [22.4]; 3PE at 8.39 Å, 20.5 [12.1]; 3PE at 8.55 Å, 24.4 [21.4]; 4PE at 8.67 Å, 22.7 [10.3] kcal mol21.Fig. 8 Representations of the crystal packing in Ph4P1[CpMo(Se4)2]2 [SIRGAB]: P, Mo black, C grey, Se speckled, H atoms omitted for clarity. (a) Face view of the ladder: P ? ? ? P distances are marked with arrows, while the non-arrowed numbers are the energies (kcal mol21 per Ph2) between individual phenyl rings.The centrosymmetric 6PE (6.4 Å) and the strong attractive interactions along the edges of the ladder are evident. (b) View along the ladders (marked with arrows) showing how the ladders are separated by the [CpMo(Se4)2]2 anions. negative anions. What is the reason for this pattern? The 4PE layer has one positive charge per cycle and so can conveniently accommodate one X2 anion above (or below) each cycle.The 6PE layer has two positive charges per cycle and so can accommodate one X22 anion per cycle. Alternatively, there can be one X2 on each side of the cycle, and this has been found to occur. Because of the relatively large space between layers, solvent is often accommodated there also. A crystal structure which might be expected is the onedimensional motif comprised of an isolated pair of ZZI6PE linked by P4PE to form a ladder (analogous to those in the preceding paper): this would be a one-dimensional section ofJ.Chem. Soc., Dalton Trans., 1998, 3167–3175 3175 Table 3 Total calculated attractive energies between one Ph4P1 cation and neighbouring Ph4P1 cations in crystals containing networks of 6PE and other multiple phenyl embraces Crystal refcode VIPTAP FAGREK BEJBOH BEJBOH BEJBOH KIZWEV KIZWEV COVCAR COVCAR COVCAR VOFMOS Number of surrounding Ph4P1 cations 3 32333444 43 Types of multiple phenyl embraces (6PE)2(P4PE) (6PE)2(O4PE) (6PE)2 (6PE)2(4PE) (6PE)(4PE)(P4PE) (6PE)2(3PE) (6PE)2(P4PE)(3PE) (6PE)2(P4PE)(3PE) (6PE)(P4PE)(3PE)(4PE) (6PE)2(4PE)(3PE) (6PE)2(P4PE) Total interaction energy per Ph4P1a/kcal mol21 27.0 25.1 27.2 24.6 23.9 24.8 26.2 28.9 24.0 25.5 23.7 24.5 28.5 27.9 a Intermolecular energy contributions from all atoms in Ph4P1 are included.The energy is per Ph4P1: each local intermolecular interaction energy is halved before the summation, so that no energies are doubly counted.the two-dimensional ZZI6PE/P4PE net. However we have not yet found an instance of this: the closest approach to this type is SIRGAB, Fig. 8. The networks described in this paper raise the question about the number of multiple phenyl embraces which can be sustained by a Ph4P1 cation, and the total attractive energy involved. The most common pattern is (6PE)2(4PE) at one Ph4P1, and for this we calculate that each 6PE contributes ca. 7 kcal mol21 of attractive energy per [Ph4P1]2 pair and the 4PE ca. 3 kcal mol21 of attractive energy per [Ph4P1]2 pair (in the following discussion the contributions from all atoms of the Ph4P1 cations are included in the energies quoted). Two of the structures illustrated show one Ph4P1 cation participating in four multiple phenyl embraces: in KIZWEV (Fig. 6) there are are two 6PE, one P4PE, and one 3PE linking to one cation, while in COVCAR (Fig. 7) diVerent Ph4P1 cations participate in one 6PE, two 4PE and one 3PE, or two 6PE, one 4PE and one 3PE.Of greater significance to the crystal packing energy is the total interaction energy between a cation and its neighbouring cations, not subdivided according to type or number of local motifs. For this we calculate the energies given in Table 3. It appears that the maximum attractive interaction energy which a Ph4P1 can achieve with surrounding Ph4P1 cations participating in multiple phenyl embraces is ca. 9 kcal mol21. All of these embraces between Ph4P1 cations are as strong as conventional hydrogen bonds, and the extended supramolecular motifs demonstrated here are dominant lattice-determining factors. Acknowledgements This research is funded by the Australian Research Council through a project grant and through VisLab computing facilities. References 1 I. G. Dance and M. L. Scudder, J. Chem. Soc., Chem. Commun., 1995, 1039. 2 I. G. Dance, in The Crystal as a Supramolecular Entity, ed. G. R. Desiraju, John Wiley, New York, 1996, pp. 137–233. 3 I. Dance and M. Scudder, Chem. Eur. J., 1996, 2, 481. 4 C. Hasselgren, P. A. W. Dean, M. L. Scudder, D. C. Craig and I. G. Dance, J. Chem. Soc., Dalton Trans., 1997, 2019. 5 I. Dance and M. Scudder, New J. Chem., 1998, 481. 6 M. Scudder and I. Dance, J. Chem. Soc., Dalton Trans., 1998, 329. 7 M. Scudder and I. Dance, preceding paper. 8 J. D. Dunitz, Pure Appl. Chem., 1991, 63, 177. 9 G. R. Desiraju, The Crystal as a Supramolecular Entity, Perspectives in Supramolecular Chemistry, ed. J. M. Lehn, John Wiley, Chichester, 1996. 10 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 131. 11 I. Dance and M. Scudder, J. Chem. Soc., Dalton Trans., 1996, 3755. 12 N. A. Ahmed, A. I. Kitaigorodsky and K. V. Mirskaya, Acta Crystallogr. Sect. B, 1971, 27, 867. 13 C. P. Brock and J. D. Dunitz, Chem. Mater., 1994, 6, 1118. 14 H. J. Gotsis, A. C. Barnes and P. Strange, J. Phys. Condens. Matter, 1992, 4, 10461. Paper 8/03464H

ISSN:1477-9226    

DOI:10.1039/a803464h

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3177–3182 3177 Lantern type heterobimetallic complexes. Tetra-Ï-4-methylpyridine- 2-thiolato bridged platinum(II)cobalt(II) and oxidation complexes † Ken’ichi Kitano,a Kazuyuki Tanaka,a Takanori Nishioka,a Akio Ichimura,a Isamu Kinoshita,*a Kiyoshi Isobe a and Shun’ichiro Ooi b a Department of Molecular Materials Science, Graduate School of Science, Osaka City University, Sumiyoshi-ku, Osaka 558-8585, Japan. E-mail: isamu@sci.osaka-cu.ac.jp b Department of Applied Chemistry, Osaka Institute of Technology, Asahi-ku, Osaka 535-8585, Japan Received 11th May 1998, Accepted 4th August 1998 A lantern type bimetallic complex [(CH3CN)Co(4-mpyt)4Pt] 1 (4-Hmpyt = 4-methylpyridine-2-thiol) was obtained by the reaction of [Pt(4-Hmpyt)4]Cl2 with CoCl2 in the presence of base and recrystallization from acetonitrile. The complex was characterized by single-crystal X-ray crystallography and is paramagnetic.It has a tetragonal Co(4-mpyt)4Pt core in which 4-mpyt ligands bridge Pt and Co atoms.The Pt atom has a square-planar co-ordination by four S atoms, whereas the Co atom has a square-pyramidal co-ordination by five N atoms, the N atom of acetonitrile being located at the axial position of Co. The Co ? ? ? Pt distance is 2.573(2) Å. Oxidation of 1 yielded diamagnetic [CoII(4-mpyt)4PtIIIX] type complexes (X = Cl 2a, Br 2b, N3 2c or NCS 2d): 2a and 2b were obtained by the reaction of 1 with iodobenzene dichloride or CeIV in the presence of Br2 respectively, while the reaction with pseudo-halides in air gave 2c and 2d.They have been characterized by variable-temperature 1H NMR and X-ray photoelectron spectroscopies. Complexes 2a–2d exist as the oligomer [{Co(4-mpyt)4PtX}n] where up to four tetragonal [Co(4-mpyt)4PtX] units are linked by Pt–X–Co bridges. Cyclic voltammetry (CV) of 1 in acetonitrile displayed a one-electron oxidation peak at 20.05 V and two rereduction peaks at 20.17 and 20.40 V vs.an Ag–Ag1 electrode at 25 8C. The CVs of 1 and 2a show similar temperature dependences which can be interpreted on the basis of the dissociation–association equilibrium of the CoIIPtIII species in the co-ordinating solvent. Introduction Binuclear complexes with metal–metal bonds have attracted considerable interest for studying metal–metal bonding 1a–q and catalytic activity 2a,b in the simplest manner. Most of those studies have been made for homometal complexes and much less for heteronuclear metal–metal complexes.1a,3a–o A heteronuclear complex is produced by bridging diVerent kinds of metal ions with appropriate ligands, but the synthetic reaction may aVord two diVerent homonuclear complexes at the same time.Sometimes the latter are produced dominantly over the former. This tendency causes the diYculty in the synthesis of heteronuclear metal complexes. The complex [Pt(4-Hmpyt)4]Cl2 (4-Hmpyt = 4-methylpyridine- 2-thiol), in which the Pt atom is co-ordinated by four sulfur atoms in a square-planar configuration and pyridine nitrogens are protonated, acts as a binucleating reagent upon deprotonation of its nitrogen atoms which capture another metal ion to give a [Pt(4-mpyt)4M] type complex.By reaction of [Pt(4-Hmpyt)4]Cl2 with Ni(CH3CO2)2 we have prepared a lantern type binuclear [PtII(4-mpyt)4NiII], where the Pt and Ni atoms are co-ordinated by four S and four N in pyridine, respectively. 4a Such a d8 ? ? ?d8 complex has a potential ability to form a d7–d7 complex.5,6 The d7–d7 complex could also be derived from a d8 ? ? ?d7 complex by one-electron oxidation. In addition to the NiIIPtII complex, an attempt has been made to prepare † Supplementary data available: variable temperature 1H NMR spectra. For direct electronic access see http://www.rsc.org/suppdata/dt/1998/ 3177/, otherwise available from BLDSC (No. SUP 57425, 4 pp.) or the RSC Library. See Instructions for Authors, 1998, Issue 1 (http:// www.rsc.org/dalton).lantern type CoIIPtII and CoIIPtIII complexes. We report here the syntheses and characterizations of the binuclear CoIIPtII and CoIIPtIII complexes. Experimental Materials and measurements 4-Methylpyridine-2-thiol was prepared from 4-methylpyridine by modifying the literature method.7 The complex [Pt(4- Hmpyt)4]Cl2 was prepared as described previously.4a Iodobenzene dichloride was prepared by the literature method.8 Solvents were dried and distilled before use.All other chemicals were purchased from Wako, Nacalai and Aldrich, and used without further purification. Silica gel column chromatography was used for purification of the products. Electronic spectra were recorded on an Hitachi 330 spectrophotometer at room temperature, 1H NMR spectra on JEOLGX400 and -a400 spectrometers at 25 8C, SiMe4 being used as internal standard. Electrochemical measurements were performed with a Bioanalytical Systems Inc.CV-50W Voltammetric Analyzer. Cyclic voltammograms were recorded between 25 and 225 8C by use of a glassy-carbon-disc working electrode and an Ag– Ag1 reference electrode, acetonitrile and 0.1 mol dm23 Bu4NPF6 being used as a solvent and supporting electrolyte respectively.9 The formal potential of the ferrocenium– ferrocene couple was 10.52 V with respect to this reference electrode. Controlled potential coulometry was carried out with a Hokuto 501 potentiostat. The working electrode was reticulated vitreous carbon.Molecular weight measurements were performed with a3178 J. Chem. Soc., Dalton Trans., 1998, 3177–3182 Knauer model 11.00 vapour pressure osmometer using dichloromethane as a solvent. The XPS spectra were recorded on a VG Scientific ESCA LAB MKII spectrometer by use of monochromated Mg-Ka (1253.6 eV) radiation, the standard technique being employed. Preparations [(CH3CN)Co(4-mpyt)4Pt]?CH3CN 1?CH3CN. The mixture of [Pt(4-Hmpyt)4]Cl2 (200 mg, 0.26 mmol), CoCl2?6H2O (105 mg, 0.26 mmol), KHCO3 (105 mg, 1.04 mmol) and toluene (10 cm3) in a sealed tube with a greaseless bulb was heated for 3 h at 120 8C under a nitrogen atmosphere. The solid compounds were dissolved in toluene, the solution changing from green to red-orange.After cooling to room temperature, the product was purified by short silica gel column chromatography under a nitrogen atmosphere using acetonitrile as an eluent.Red crystals of 1?CH3CN were obtained by concentrating the red eluate (135 mg, 62%) (Found: C, 39.98; H, 3.53; N, 9.95. C28H30- CoN6PtS4 requires C, 40.38; H, 3.63; N, 10.09%); meff(room temperature) = 3.20 mB. [Co(4-mpyt)4PtCl]?CHCl3 2a?CHCl3. Iodobenzene dichloride (9 mg, 0.03 mmol) dissolved in chloroform (3 cm3) was added to a chloroform solution (5 cm3) of complex 1 (50 mg, 0.06 mmol) at 240 8C and stirred for 2 min under a nitrogen atmosphere. After addition of chilled n-hexane (10 cm3), the solution was left in a refrigerator (240 8C).Dark brown microcrystals were recrystallized from chloroform to give 2a?CHCl3 (48 mg, 84%) (Found: C, 32.83; H, 2.76; N, 6.21. C25H25Cl4- CoN4PtS4 requires C, 33.16; H, 2.78; N, 6.19%); dH(CD2Cl2) 8.71 (2 H, br, H6), 6.97 (1 H, br, H3), 6.89 (1 H, br, H3), 6.54 (2 H, br, H5) and 2.20 (6 H, s, CH3). [Co(4-mpyt)4PtBr]?3CHCl3 2b?3CHCl3. A methanol solution (3 cm3) of potassium bromide (7 mg, 0.06 mmol) was added to a chloroform solution (5 cm3) of complex 1 (50 mg, 0.06 mmol) at 230 8C and stirred for 2 min under a nitrogen atmosphere.Addition of ammonium cerium(IV) nitrate (35 mg, 0.06 mmol) in methanol (3 cm3) gave rise to a change from orange to greenblack. After stirring for 2 min, chilled n-hexane (10 cm3) was added. Green-black 2b?3CHCl3 was obtained by recrystallization from chloroform (33 mg, 44%) (Found: C, 26.31; H, 2.19; N, 4.83. C27H27BrCl9CoN4PtS4 requires C, 27.28; H, 2.29; N, 4.71%); dH(CD2Cl2) 8.71 (2 H, br, H6), 6.98 (1 H, br, H3), 6.91 (1 H, br, H3), 6.51 (2 H, br, H5) and 2.17 (6 H, s, CH3).[Co(4-mpyt)4Pt(N3)] 2c. A methanol solution (3 cm3) of sodium azide (4.3 mg, 0.06 mmol) was added to a chloroform solution (5 cm3) of complex 1 (50 mg, 0.06 mmol) at 230 8C and stirred for 2 min. After addition of chilled n-hexane (10 cm3), the resulting solution was left in a refrigerator. Brownblack 2c was obtained by recrystallization from chloroform (47 mg, 94%) (Found: C, 35.92; H, 3.03; N, 12.02.C24H24- CoN7PtS4 requires C, 36.31; H, 3.05; N, 12.37%); dH(CD2Cl2) 7.88 (2 H, d, H6, 3JHH = 6.4), 6.95 (2 H, br, H3), 6.61 (2 H, d, H5, 3JHH = 6.4 Hz) and 2.17 (6 H, s, CH3). Complex 2c was also obtained from 2a by the substitution reaction [Co(4-mpyt)4- PtCl] 1 NaN3 æÆ [Co(4-mpyt)4Pt(N3)] 1 NaCl at 230 8C in methanol–chloroform. [Co(4-mpyt)4Pt(SCN)]?2CHCl3 2d?2CHCl3. A methanol solution (3 cm3) of sodium thiocyanate (5 mg, 0.06 mmol) was added to a chloroform solution (5 cm3) of complex 1 (50 mg, 0.06 mmol) at 230 8C with stirring.After chilled n-hexane (10 cm3) was added, the solution was left in a refrigerator overnight. Black crystals of 2d?2CHCl3 (53 mg, 81%) were obtained (Found: C, 35.34; H, 2.92; N, 7.97. C26H26Cl2CoN5PtS5 requires C, 34.94; H, 2.93; N, 7.84%); dH(CD2Cl2) 8.03 (2 H, br, H6), 7.00 (1 H, br, H3), 6.92 (1 H, br, H3), 6.62 (2 H, br, H5) and 2.20 (6 H, s, CH3). [(CH3CN)Ni(4-mpyt)4Pt]?CH3CN 3?CH3CN.The preparation of complex 3 was reported previously.4a The complex [Pt(4-Hmpyt)4]Cl2 (767 mg, 1.0 mmol), Ni(CH3CO2)2?4H2O (249 mg, 1.0 mmol), KHCO3 (400 mg, 4.0 mmol) and 15 g of naphthalene were put in a sealed tube with a greaseless bulb and heated for 1 h at 120 8C under a nitrogen atmosphere. The solution gradually turned from yellow to red-brown during the reaction. After cooling to room temperature, naphthalene was removed by treatment with n-hexane.The orange precipitate was purified by silica gel chromatography using acetonitrile– dichloromethane (1:10, v/v) as eluent. Red crystals of 3?CH3CN (525 mg, 63%) were obtained (Found: C, 40.20; H, 3.59; N, 10.03. Calc. for C28H30N6NiPtS4: C, 40.39; H, 3.63; N, 10.10%); meff(room temperature) = 2.83 mB. Crystallography Crystal data. Compound 1. C28H30CoN6PtS4, M = 832.85, orthorhombic, space group Pbca (no. 61), a = 21.381(4), b = 18.292(4), c = 16.692(5) Å, U = 6527(2) Å3, T = 296 K, Z = 8, Dc = 1.695 Mg m23, F(000) = 3272, l(Mo-Ka) = 0.71069 Å, m(Mo-Ka) = 5.1 mm21, wR(F2) 0.057, R1 = 0.051.Compound 3. C28H30N6NiPtS4, M = 832.62, orthorhombic, space group Pbca (no. 61), a = 21.362(3), b = 18.237(2), c = 16.608(2) Å, U = 6470(2) Å3, T = 296 K, Z = 8, Dc = 1.709 Mg m23, F(000) = 3280, l(Mo-Ka) = 0.71069 Å, m(Mo-Ka) = 5.2 mm21, wR(F2) 0.039, R1 = 0.032. Data collection and reduction. Compound 1. Single crystals of compound 1 were obtained by slow evaporation of its acetonitrile solution.A red prism with dimensions 0.50 × 0.30 × 0.20 mm was mounted in a glass capillary. A total of 10294 intensities were measured to 2qmax 608, all independent. An absorption correction based on empirical y scans was applied, with transmission factors 0.791–0.998. Cell constants were refined from setting angles of 25 reflections in the range 2q 21.9–23.48. For every intensity measurement three standard reflections, monitored every 150, showed no significant variation during data collection.Compound 3. Single crystals of compound 3 were obtained by slow evaporation of dichloromethane into an acetonitrile solution. A red prism with dimensions 0.40 × 0.40 × 0.20 mm was mounted in a glass capillary. A total of 10275 intensities were measured to 2qmax 608, of which 10272 were independent (Rint = 0.510). An absorption correction based on empirical y scans was applied, with transmission factor 0.618–0.992. Cell constants were refined from setting angles of 25 reflections in the range 2q 29.5–30.08.For every intensity measurement three standard reflections, monitored every 150, showed no signifi- cant variation during data collection. Structure solution and refinement. The two structures were solved by the direct method, and the positional and thermal parameters refined by full-matrix least squares on F. Positions of almost all hydrogen atoms were found on the Fourierdi Verence map and the remaining ones were located on the calculated positions (C–H 0.95 Å).The positional and isotropic thermal parameters were included in the least-squares calculation. All calculations were carried out by use of the TEXSAN crystallographic software package.10 CCDC reference number 186/1117. See http://www.rsc.org/suppdata/dt/1998/3177/ for crystallographic files in .cif format. Results and discussion Initially complex 1 was prepared by the reaction of [Pt(4- Hmpyt)4]Cl2 with CoCl2 in the presence of base in naphthalene similar to the preparation of 3 in low yield, presumably owing to the diYculty in removing naphthalene completely from theJ.Chem. Soc., Dalton Trans., 1998, 3177–3182 3179 reaction mixture. The yield was improved significantly using toluene instead of naphthalene as a solvent. Complex 1 is immediately decomposed upon dissolution in chloroform at room temperature but is more stable in acetonitrile than in nonco- ordinating solvents.Presumably ligation of acetonitrile to the Co atom stabilizes 1. An ORTEP drawing of complex 1 (Fig. 1) shows that it is isostructural with the NiIIPtII complex as inferred from the similar cell dimensions of 1?CH3CN to those of 3?CH3CN.4b Table 1 lists selected bond lengths and angles of 1 and 3. Complex 1 has an approximate tetragonal symmetry and the four S and four N atoms are co-ordinated to the heavier and the lighter metal atom respectively. While the former has a square-planar coordination by four S atoms, the latter is co-ordinated to four N atoms at the basal sites and an acetonitrile molecule at the axial site to give a square-pyramidal co-ordination.The CoN4 equatorial co-ordination square is parallel to, but rotated by 28.78 (mean) from, the PtS4 co-ordination squares. The Co atom deviates from the four N co-ordination plane toward the apical MeCN ligand by 0.10 Å. The PtII ? ? ? CoII distance is 0.042 Å less than PtII ? ? ? NiII, and the Co–N (4-mpyt2) and (MeCN) distances are also slightly longer than the corresponding Ni–N distances of 3?CH3CN, reflecting a larger atomic and covalent radius for Co than Ni.Compound 1 is paramagnetic with meff = 3.20 mB (298 K), which suggests no appreciable intermetallic interaction in this complex. The reaction of complex 1 with iodobenzene dichloride in Fig. 1 An ORTEP drawing of [(CH3CN)Co(4-mpyt)4Pt]?CH3CN 1?CH3CN. Table 1 Selected bond lengths (Å) and angles (8) for complexes 1 (M = Co) and 3 (M = Ni) Pt ? ? ?M Pt–S(1) Pt–S(2) Pt–S(3) Pt–S(4) M–N(1) M–N(2) M–N(3) M–N(4) M–N(5) S(1)–Pt–S(2) S(1)–Pt–S(4) S(2)–Pt–S(3) S(3)–Pt–S(4) N(1)–M–N(2) N(1)–M–N(4) N(2)–M–N(3) N(3)–M–N(4) N(1)–M–N(5) N(2)–M–N(5) N(3)–M–N(5) N(4)–M–N(5) M–N(5)–C(26) 1?CH3CN 2.573(2) 2.328(4) 2.329(4) 2.323(5) 2.331(5) 2.16(1) 2.14(1) 2.16(1) 2.16(1) 2.17(2) 90.4(2) 88.9(2) 90.9(2) 89.7(2) 91.6(5) 89.1(5) 89.9(5) 88.9(5) 94.1(6) 93.5(5) 91.3(6) 91.3(5) 162(2) 3?CH3CN 2.531(1) 2.321(2) 2.322(2) 2.324(2) 2.321(2) 2.130(5) 2.104(5) 2.119(5) 2.124(4) 2.119(5) 90.85(6) 89.74(6) 90.58(6) 88.82(6) 89.6(2) 89.4(2) 91.2(2) 89.5(2) 90.5(2) 92.9(2) 93.4(2) 90.7(2) 164.3(5) chloroform gave dark brown diamagnetic 2a, and the bromoanalogue 2b was obtained by the reaction of 1 with (NH4)2Ce(NO3)6 in the presence of KBr.These reactions were carried out under a nitrogen atmosphere and at temperatures below 230 8C, taking precautions against the decomposition of 1 prior to the reaction. However, the reactions of 1 with NaN3 and NaSCN in chloroform–methanol in air at 230 8C aVorded diamagnetic 2c and 2d without strong oxidant, and no appreciable decomposition of 1.Formation of 2a and 2b was also observed in the reaction of 1 with NaX (X = Cl or Br) in air at <230 8C, but not so selective as those of 2c and 2d.12 The 1 : 1 reaction of 2a with NaN3 and NaSCN in chloroform–methanol gives 2c and 2d. Complex 2c reverts back to 2a on reaction with twice as many moles of iodobenzene dichloride in dichloromethane. 13 The reactions of 2a with Bun 4NBr gave 2b, but that of 2b with Bun 4NCl aVorded a mixture of 2a and 2b. In the oxidation of 1 one electron may be removed from either CoII or PtII to give a CoIIPtIII or CoIIIPtII complex. In the former case the d7–d7 bond formation is responsible for the diamagnetic nature of the products. In the latter the strong preference of CoIII for octahedral six-co-ordination to give a monomeric complex may result in a destruction of the {Co(4-mpyt)4Pt} core structure.That 2a–2d are CoIIPtIII complexes is supported by XPS data as shown later as well as the following facts. As described below, 2a shows a cyclic voltammogram similar to that of 1 in pattern as well as peak potentials for redox waves in acetonitrile at 21.0 to 10.5 V vs. Ag1–Ag at room temperature. This electrochemical interconversion process suggests the possibility of the chemical reconversion of 2a into 1.Indeed the reaction of 2a–2d with NaBH4 in thf gives a red-orange compound identified as 1 based on the UV/VIS spectrum. The 1 2 redox interconversions and the substitution reactions among 2a–2d strongly indicate the retention of the {Co(4-mpyt)4Pt} core structure during the reactions (Scheme 1). As the preparation of single crystals of complexes 2a–2d has been unsuccessful, their structures have been investigated by using the results of reactivity, electrochemical measurement, 1H NMR spectra, and osmometric molecular weight measurement.Fig. 2 shows the dynamic behaviour of the 1H NMR spectrum of 2c. At 20 8C all protons exhibit a sharp signal except for H3, the broad signal of which begins to split into two peaks at 25 8C. At much lower temperature, however, the H5, H6 and CH3 signals also become broad. At 280 8C the CH3 signal splits into more than three and each of the H3, H5 and H6 signals into more than four peaks.On the other hand, 2a gives, even at 20 8C, a broad signal for all protons and two peaks for H3. The extent of the broadening is similar to that of 2c at 240 8C. At 260 8C the H6 signal splits into six peaks and the CH3 signal into three. Signal-pattern variations with temperature for 2b and 2d resemble those of 2a than 2c. At 260 8C the CH3 signal Scheme 1 Reactions of (CoPt)41/51 complexes. C6H5ICl2 SCN–/CHCl3 N3 –/CHCl3 [Co(4-mpyt)4PtBr] [(CH3CN)Co(4-mpyt)4Pt] Br-+Ce4+ [Co(4-mpyt)4Pt(SCN)] NaBH4/thf [Co(4-mpyt)4PtCl] Cl– N3 – [Co(4-mpyt)4Pt(N3)] SCN– 1 2a 2c 2b 2d NaBH4/thf NaBH4/thf NaBH4/thf3180 J.Chem. Soc., Dalton Trans., 1998, 3177–3182 of 2d split into more than eight peaks which indicates the complex behaviour of 2d in solution.14 These spectra could not be explained on the basis of a tetragonal [Co(4-mpyt)4PtX] monomer structure simply supposed from the analytical results. The existence of [Co(4-mpyt)4PtX] cores was confirmed by the redox reactions, electrochemistry and XPS measurement.The osmometric molecular weights of 2a and 2d at 2.40 mg per 1 ml CH2Cl2 and 8.38 mg per 1 ml CH2Cl2 were 4060 ± 400 and 3430 ± 400, respectively.15 The result shows the existence of a tetrameric unit of [{Co(4-mpyt)4PtX}4] for 2a and 2d in CH2Cl2 solution. The tetramers of the {Co(4-mpyt)4Pt} core are the key point for understanding the complex 1H NMR behaviour. Molecular weight measurements indicate only the average value, so it is likely there are equilibria involving oligomers, namely the trimer, tetramer and pentamer, their ratio being determined by the concentration, solvent, axial ligand, temperature and so on.Such oligomeric aggregation is consistent with the highly symmetric spectra at the high temperature limit, and the very complex signals at lower temperature. The signals at room temperature tend to cause broadening at higher concentration. The most plausible model for this type of aggregation involves stacking through the Pt–X–Co interaction. At high temperature rapid exchange of the {Co(4-mpyt)4Pt} core between oligomers averages the signals to give a highly symmetrical signal.At lower temperature the exchange between oligomers is slowed as is the rate of rotation around the aggregation axis. Finally, each oligomer as well as the rotamer contributes to the NMR signal, resulting in the very complex pattern. Thus, the methyl singlet for 2d at 20 8C splits in to more than eight peaks at 260 8C.14 There is no evidence for terminal ligation at either platinum or cobalt; the oxidation state of the metals might determine the co-ordination site of the outer axial ligand.Fig. 3 shows a plausible structure of the [{Co(4-mpyt)4PtX}n] oligomer. The Pt–X bonds in the bridging Pt–X–Co segment, similar to those in [XPt(bridge)4PtX] complexes,6 may be significantly longer than those of the in-plane bonds. In the tetramer there are at least four diVerent environments for the protons.In the presence of an excess of halide ion, however, the oligomer may be decomposed into [XCo(4-mpyt)4PtX]2 monomer owing to the ligation of halide ion to CoII. Indeed a CDCl3 solution of complex 2a showed sharp 1H NMR signals even at room temperature (Fig. 4) on addition of a large excess of Bun 4NCl or Bun 4NBr, and each of the spectra was explainable based on the tetragonal [XCo(4-mpyt)4PtX]2 monomer.16 The UV/VIS spectral data are given in Table 2.In view of the transition energy and e value the absorption bands of complex 1 are assignable as d–d transitions localized on the CoN5 core. In contrast to 1, 2a–2d show much more intense absorption in CH2Cl2, which indicates participation of two metal atoms in these electronic transitions. These observations strongly indicates that 2a–2d possess a Pt–Co bond. Fig. 2 The variable temperature 1H NMR spectra of [Co(4-mpyt)4- Pt(N3)] 2c.The oxidation states of the metal atoms were investigated by XPS. The platinum (4f7– 2) binding energy of complex 2a is 0.5 eV higher than the corresponding one of 1 but the cobalt (2p3– 2) binding energy of 2a agrees with that of 1 within the experimental error (Table 3). In the case of the [XPtIII(4-mpyt)4PtIIIX] and [PtII(4-mpyt)4PtII] couple, the platinum (4f7– 2) binding energy of the former is 1.2 eV higher than that of the latter, the usual diVerence for a 11 increment of the oxidation state.The increment of the platinum (4f7– 2) binding energy for 2a–1 is less than that of [XPtIII(4-mpyt)4PtIIIX] and [PtII(4-mpyt)4PtII]. However the significant increase in binding energy of 2a in comparison with 1 indicates the Pt atom in 2a to be in the trivalent state, one odd electron of which is coupled with that of CoII to give diamagnetic 2a. Thus 2a is considered as a CoIIPtIII complex. From analogy to PtIII 2 complexes, a plausible structure for Fig. 3 The tetramer model of [{Co(4-mpyt)4PtX}4] (X = Cl 2a, Br 2b, N3 2c or SCN 2d).Fig. 4 The 1H NMR spectra of [Co(4-mpyt)4PtCl] (a), [Co(4-mpyt)4- PtCl] 1 1000 equivalents Bun 4NCl (b) and [Co(4-mpyt)4PtCl] 1 1000 equivalents Bun 4NBr (c). Asterisks denote signals assignable to [XCo- (4-mpyt)4PtX]2.J. Chem. Soc., Dalton Trans., 1998, 3177–3182 3181 Table 2 The UV/VIS absorption spectral data [lmax/nm (e/dm3 mol21 cm21)] CoII–PtII 1 a 540 (360) 470 (355, sh) CoII–PtIII–Cl 2ab 530 (4000, sh) 435 (8480) 410 (8400, sh) CoII–PtIII–Br 2b b 520 (3350, sh) 430 (6370, sh) 400 (7300) CoII–PtIII–N3 2c b 550 (6700) 445 (12700) CoII–PtIII–SCN 2d b 515 (4980, sh) 425 (11030) 400 (10200, sh) NiII–PtII 3 a 510 (160, sh) 350 (10100, sh) 310 (17100, sh) 282 (24800) a Measured in CH3CN.b In CH2Cl2. 2a–2d is [Co(4-mpyt)4PtX] which has a {Co(4-mpyt)4Pt} core and X bound to a PtIII atom. Fig. 5 shows the cyclic voltammograms of complexes 1 and 2a at 25 8C in acetonitrile.Both change dramatically with temperature (Figs. 6 and 7). At 25 8C 1 exhibits an oxidation peak at 20.05 V, whereas there are two rereduction peaks at 20.17 and 20.40 V on the reverse scan. The controlled-potential coulometry of 1 at 10.20 V indicated the oxidation to be a oneelectron process. With the dependence on temperature, the rereduction peak at 20.17 V tends to diminish and the peak potential of the other rereduction peak at 20.40 V at 25 8C moves negatively to 20.65 V.The CV of 2a (Fig. 7) shows two reduction peaks at 20.30 and 20.60 V at 25 8C. The temperature dependence for the reduction peaks of 2a is quite similar to that of 1. Both peaks diminish and the peak at 20.60 V Fig. 5 Cyclic voltammograms of [(CH3CN)Co(4-mpyt)4Pt] 1 (——) and [Co(4-mpyt)4PtCl] 2a (– – –) in 0.1 mol dm23 Bu4NPF6–CH3CN at a glassy-carbon electrode with a scan rate of 50 mV s21. Fig. 6 Cyclic voltammograms of [(CH3CN)Co(4-mpyt)4Pt] 1 in 0.1 mol dm23 Bu4NPF6–CH3CN at a glassy-carbon electrode with a scan rate of 50 mV s21 at 25 (——), 5 (– – –), 26 (- - -), 217 (—-—-—) and 226 8C (—- -—).Table 3 XPS Data (eV) Compound 1 [(CH3CN)Co(4-mpyt)4Pt] 2a [Co(4-mpyt)4PtCl] [Pt(4-mpyt)4Pt] [IPt(4-mpyt)4PtI] Pt(4f7– 2 ) 72.9 73.4 73.0 74.2 Co(2p3– 2 ) 779.8 779.9 moves negatively to 20.85 V. At 25 8C the reoxidation peak appeared at 20.05 V which is almost the same as the oxidation potential peak potential of 1. The broad signal around 20.40 V increases and shifts to 20.60 V as the temperature decreases and is likely coupled with the reduction peak at 20.85 V.The variable temperature data suggest complementary behaviour for 1 and 2a. The observation that the one-electron oxidized species of 1 shows two reduction peaks indicates that the two reduction processes arise from diVerent species. As illustrated in Scheme 2, the electrode oxidation of 1 generates LCoIIPtIII species (L may be either solvent or PF6 2), parts of which are immediately coupled together to give the (LCoIIPtIII)n oligomer.In the reverse scan the LCoIIPtIII and (LCoIIPtIII)n species stabilized by oligomer formation are reduced at 20.17 and 20.40 V, respectively, to give LCoIIPtII and (LCoIIPtII)n. At lower temperature the equilibrium shift to oligomer takes place which diminishes the monomer reduction peak and the reduction peak of the oligomer moves to negative potential. Even though the oxidation peak at 20.05 V diminishes in peak height significantly at low temperature, the peak position does not change, which is the same as that of 1.This phenomenon indicates the peak is assignable to the oxidation of Co–Pt. Since the stabilized Fig. 7 Cyclic voltammograms of [Co(4-mpyt)4PtCl] 2a in 0.1 mol dm23 Bu4NPF6–CH3CN at a glassy-carbon electrode with a scan rate of 50 mV s21 at 25 (——), 15 (– – –), 4 (- - -), 25 (—-—-—), 212 (—--—) and 225 8C (— — —). Scheme 2 Square electrochemical–chemical–electrochemical–chemical step dimerization mechanism.(LCoIIPtIII)+ (LCoIIPtII) n (LCoIIPtIII) n n+ -0.05 V -e– +e– -e– -0.40 V -e– +e– +e– -e– -0.60 V -0.30 V +e– -0.17 V CoIIPtIIICl (CoIIPtIIICl) n CoIIPtII + Cl- (CoIIPtIICl) n n– + CoIIPtIIICl + + (LCoIIPtIII)+ -0.05 V -0.40 V 1 2a + (LCoIIPtII) (LCoIIPtII) + + + CoIIPtII + Cl- + (CoIIPtII) (CoIIPtIII)3182 J. Chem. Soc., Dalton Trans., 1998, 3177–3182 oligomer of (CoIIPtII)n is reoxidized at around 20.60 V at lower temperature, the decrease in concentration of the monomer CoIIPtII causes a significant change in the peak intensity at 20.05 V for 2a.These phenomena are consistent with the existence of a monomer–oligomer equilibrium for both LCoIIPtII and (LCoIIPtIII)n. Acknowledgements This work was financially supported by Grants-in-Aid from the Ministry of Education, Japan (No. 06640731 and 09874135). References 1 (a) F. A. Cotton and R. A. Walton, Multiple Bonds between Metal Atoms, 2nd edn., Clarendon Press, Oxford, 1993; (b) F.A. Cotton, J. H. Matonic and C. A. Murillo, Inorg. Chim. Acta, 1997, 264, 61; (c) C. Tejel, M. A. Ciriano, J. A. Lopez, F. J. Lahoz and L. A. Oro, Organometallics, 1997, 16, 4718; (d ) E. Zangrando, F. Pichierri, L. Randaccio and B. Lippert, Coord. Chem. Rev., 1996, 156, 275 and refs. therein; (e) T. Wienkotter, M. Sabat, G. Fusch and B. Lippert, Inorg. Chem., 1995, 34, 1022; ( f ) A. Schreiber, O. Krizanovic, E.C. Fusch, B. Lippert, F. Lianza, A. Albinati, S. Hill, D. M. L. Goodgame, H. Stratemeier and M. A. Hitchman, Inorg. Chem., 1994, 33, 6101; ( g) G. Frommer, F. Lianza, A. Albinati and B. Lippert, Inorg. Chem., 1992, 31, 2434; (h) M. Krumm, B. Lippert, L. Randaccio and E. Zangrando, J. Am. Chem. Soc., 1991, 113, 5129; (i) I. Mutikainen, O. Orama, A. Pajunen and B. Lippert, Inorg. Chim. Acta, 1987, 137, 189; ( j) H. Schollhorn, U. Thewalt and B. Lippert, Inorg. Chim. Acta, 1987, 135, 155; (k) W.Micklitz, G. Muller, J. Riede and B. Lippert, J. Chem. Soc., Chem. Commun., 1987, 76; (l) B. Lippert, U. Thewalt, H. Schollhorn, D. M. L. Goodgame and R. W. Rollins, Inorg. Chem., 1984, 23, 2807; (m) D. Neugebauer and B. Lippert, J. Am. Chem. Soc., 1982, 104, 6596; (n) B. Lippert and U. Schubert, Inorg. Chim. Acta, 1981, 56, 15; (o) T. V. O’Halloran, M. M. Robert and S. J. Lippard, Inorg. Chem., 1986, 25, 957; (p) L. S. Hollis and S. J. Lippard, J. Am. Chem. Soc., 1983, 105, 3494; (q) L.S. Hollis and S. J. Lippard, Inorg. Chem., 1983, 22, 2600. 2 (a) J. H. Sinfelt, in Metal–Metal Bonds and Clusters in Chemistry and Catalysis, ed. J. P. Fackler, jun., Plenum, New York, 1989 and refs. therein; (b) L. Hao, J. Xiao, J. J. Vittal and R. J. Puddephatt, Organometallics, 1997, 16, 2165. 3 (a) W. Clegg, M. Capdevila, P. Gonzalez-Duarte and J. Sola, Acta Crystallogr., Sect. B, 1996, 52, 270; (b) A. Erxleben and B. Lippert, J. Chem. Soc., Dalton Trans., 1996, 2329; (c) G.Fusch, E. C. Fusch, A. Erxleben, J. Huttermann, H. J. Scholl and B. Lippert, Inorg. Chim. Acta, 1996, 252, 167; (d) C. Mealli, F. Pichierri, L. Randaccio, E. Zangrando, M. Krumm, D. Holtenrich and B. Lippert, Inorg. Chem., 1995, 34, 3148; (e) M. Krumm, E. Zangrando, L. Randaccio, S. Menzer, A. Danzmann, D. Holthenrich and B. Lippert, Inorg. Chem., 1993, 32, 2183; ( f ) M. Krumm, E. Zangrando, L. Randaccio, S. Menzer and B. Lippert, Inorg. Chem., 1993, 32, 700; (g) M.A. Ciriano, J. J. Perez-Torrente, F. J. Lahoz and L. A. Oro, Inorg. Chem., 1992, 31, 969; (h) G. P. A. Yap and C. M. Jensen, Inorg. Chem., 1992, 31, 4823; (i) J. G. Reynolds, S. C. Sendlinger, A. M. Murray, J. C. HuVman and G. Christou, Angew. Chem., Int. Ed. Engl., 1992, 31, 1253; ( j) J. H. Yamamoto, W. Yoshida and C. M. Jensen, Inorg. Chem., 1991, 30, 1353; (k) M. A. Ciriano, F. Viguri, J. J. Perez-Torrente, F. J. Lahoz, L. A. Oro, A. Tiripicchio and M.Tiripicchio-Camellini, J. Chem. Soc., Dalton Trans., 1989, 25; (l) W. Micklitz, J. Riede, B. Huber, G. Muller and B. Lippert, Inorg. Chem., 1988, 27, 1979; (m) W. Micklitz, G. Muller, B. Huber, J. Riede, F. Rashwan, J. Heinze and B. Lippert, J. Am. Chem. Soc., 1988, 110, 7084; (n) L. A. Oro, M. A. Ciriano, F. Viguri, A. Tiripicchio and M. Tiripicchio-Camellini, New J. Chem., 1986, 10, 75; (o) H. Schollhorn, U. Thewalt and B. Lippert, Inorg. Chim. Acta, 1985, 108, 77. 4 (a) T.Nishioka, I. Kinoshita, K. Kitano and S. Ooi, Chem. Lett., 1992, 883; (b) The previously reported crystal structure of complex 3 was revised and the space group determined as Pbca instead of P212121; 1 and 3 are also isostructural. 5 F. A. Cotton and R. A. Walton, Multiple Bonds between Metal Atoms, 2nd edn., Clarendon Press, Oxford, 1993, ch. 8. 6 K. Umakoshi, I. Kinoshita, A. Ichimura and S. Ooi, Inorg. Chem., 1987, 26, 3551. 7 J. R. Thirtle, J. Am. Chem. Soc., 1946, 68, 342. 8 H. Lucas and E. R. Kennedy, Org. Synth., 1955, 3, 482. 9 Acetonitrile was purified by distilling three times from CaH2; Bu4NPF6 was recrystallized three times from ethanol solution. In each case, working solutions were checked by recording the cyclic voltammogram of the electrolyte solution before addition of the complex. 10 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, TX, 1992. 11 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 12 A solution (3 cm3) of LiCl (2.5 mg, 0.06 mmol) in methanol was added to a solution (5 cm3) of complex 1 (50 mg, 0.06 mmol) in chloroform at 230 8C and stirred for 2 min. Addition of chilled n-hexane (10 cm3) gave a dark brown solid, for which 1H NMR in CDCl3 disclosed that although the reaction aVords 2a a few unidentified compounds are also produced at the same time. 13 A dichloromethane solution (3 cm3) of iodobenzene dichloride (5.2 mg, 0.019 mmol) was added to a dichloromethane solution (5 cm3) of complex 2c (30 mg, 0.038 mmol) at 230 8C with stirring. Addition of chilled n-hexane (10 cm3) aVorded a dark green precipitate, which was found to be an approximately 1 : 1 mixture of 2a and 2c from its 1H NMR spectrum. Iodobenzene dichloride is thought to give rise to the oxidation followed by replacement of the azide ligand to give 2a. 14 The variable temperature 1H NMR spectra of [Co(4-mpyt)4- Pt(SCN)] 2d is provided in the Supplementary data (SUP 57425). 15 The large uncertainties in the molecular weights arises from the low solubility of complexes 2a and 2d. Measured values are provided in SUP 57425 (Figs. S2 and S3). 16 A mixture of Bun 4NX (X = Cl or Br) (2.5 mmol), complex 2a (2 mg, 0.0025 mmol) and CDCl3 (1 cm3) was stirred at 20 8C for 5 min. It was filtered into an NMR sample tube and the 1H NMR spectrum measured. In the case of X = Br the spectrum showed only tetragonal [BrCo(4-mpyt)4PtBr]2 monomer; when X = Cl an approximately 1 : 1 mixture of [ClCo(4-mpyt)4PtCl]2 monomer and 2a was revealed. Paper 8/03489C

ISSN:1477-9226    

DOI:10.1039/a803489c

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3183–3190 3183 Nitrido-ruthenium(VI) and -osmium(VI) complexes containing chelating multianionic (N, O) ligands. Synthesis, crystal structures and reactions with triphenylphosphine Pui-Ming Chan, Wing-Yiu Yu, Chi-Ming Che * and Kung-Kai Cheung * Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong Received 4th June 1998, Accepted 15th July 1998 A series of nitrido-ruthenium(VI) and -osmium(VI) complexes containing chelating di-, tri- and tetra-anionic ligands was synthesized by ligand substitution reaction in methanol under room conditions in the presence of 2,6-dimethylpyridine.All the newly prepared complexes are air-stable diamagnetic solids. The crystal structures of seven complexes have been established by X-ray crystallography. The Ru]] ] N (1.615–1.594 Å) and the Os]] ] N (1.612–1.621 Å) bond distances are rather insensitive to the electron-donating power of the auxiliary ligands.All the nitridoruthenium(VI) complexes react spontaneously with triphenylphosphine, and the intermediate [RuIV(N]] PPh3)- L1(py)Cl] has been isolated and characterized spectroscopically for the reaction with [RuVIN(L1)Cl]. However, for those nitridoruthenium(VI) complexes bearing the tri- (L2)32 and tetra-anionic (L3,4)42 ligands, the phosphiniminatoruthenium( IV) intermediate undergoes further reaction with pyrazole to generate a bis(pyrazole)ruthenium(IV) complex as the product.Introduction The study of metal–ligand multiple bonded (M]] X) complexes has received considerable attention.1 Of our particular interest is the chemistry of high-valent oxo-, nitrido- and imido-ruthenium and -osmium complexes, and their reactivities toward organic substrates.2 While high-valent nitridoosmium( VI) complexes of some amine and/or polypyridyl ligands are known to exhibit interesting photochemistry 3 and electrochemistry,4,5 related studies on the ruthenium(VI) analogues are sparse in the literature.Electrochemical studies revealed that high-valent oxoruthenium(VI) complexes are at least 500 mV more oxidizing than their osmium(VI) counterparts; 2,6 therefore, it is discernible that high-valent nitridoruthenium( VI) complexes could be rather electrophilic and reactive. Indeed, a N]] ] Ru species has been postulated to be the active intermediate in the oxidation of bound ammonia to nitrite.7 Herein we describe synthetic and structural studies of some nitrido-ruthenium(VI) and -osmium(VI) complexes containing chelating multianionic N, O ligands.Some of them have been characterized by X-ray crystallography, and their nitrogen atom transfer reactions to triphenylphosphine have also been examined. Experimental Materials All solvents were purified by the standard methods before use. The compounds [NBun 4][RuVINCl4],8 2,6-bis(2-hydroxy-2,2- diphenylethyl)pyridine (H2L1) 9 and 1,2-bis(2-hydroxybenzamido) benzene 10 (H4L3) were prepared according to the reported procedures. Physical measurements Infrared spectra were recorded as Nujol mulls on a Nicolet model 20 FXC FT-IR spectrometer, fast atom bombardment (FAB) mass spectra on a Finnigan MAT 95 mass spectrometer using 3-nitrobenzyl alcohol as matrix, electrospray (ES) mass spectra on a Finnigan LCQ mass spectrometer and NMR spectra using Bruker DPX 300 and 500 pulsed Fourier transform instruments. Cyclic voltammetric measurements were conducted using a Princeton Applied Research model 273 potentiostat with a glassy carbon electrode and Ag–AgNO3 (0.1 mol dm23 in MeCN) as the reference electrode, and the deaerated dichloromethane solutions in 0.10 M tetra-nbutylammonium hexafluorophosphate containing the sample complexes were scanned under an argon atmosphere.Elemental analyses were performed by Butterworth Co. Ltd., UK. X-Ray crystallography X-Ray diVraction data were collected on either a Rigaku N OH HO Ph Ph Ph Ph H2L1 HN NH N O O HO Cl Cl tBu H3L2 HN NH O O HO Cl Cl OH H4L4 HN NH O O HO OH H4L3 HN NH N N O O H2 L53184 J.Chem. Soc., Dalton Trans., 1998, 3183–3190 AFC7R or MAR diVractometer using graphite-monochromatized Mo-Ka radiation (l = 0.71073 Å) at 301 K. Intensity data were corrected for Lorentz-polarization eVects, and the structures were solved by the Patterson method and expanded by Fourier methods (PATTY).11 Structure refinements were performed by full-matrix least squares using the software package TEXSAN12 on a Silicon Graphics Indy computer.For complexes 3a, 3b and 4a, disorder in the terminal carbon atoms of one of the n-butyl groups of the [NBun 4]1 cation was treated by assigning occupation numbers of ca. 0.5 to the disordered carbon atoms: 0.5 each to C(36) and C(37) in 3a, as well as to C(36) and C(369) in 3b; 0.55 and 0.45 to C(36) and C(369) in 4a, where the thermal parameters of these disordered carbon atoms are comparable. For structures 5a and 5b the NH protons were located in the Fourier diVerence syntheses and their positional parameters included in the least squares refinement.CCDC reference number 186/1097. See http://www.rsc.org/suppdata/dt/1998/3183/ for crystallographic files in .cif format. Ligand syntheses 1-(4-tert-Butylpyridine-2-carboxamido)-4,5-dichloro-2-(2- hydroxybenzamido)benzene (H3L2).10 To a stirred anhydrous 1,4- dioxane (60 cm3) solution of N-(2-amino-4,5-dichlorophenyl)- 4-tert-butylpyridine-2-carboxamide (3.0 g, 8.9 mmol) was slowly added O-acetylsalicyloyl chloride (1.8 g, 9.0 mmol).After the mixture was stirred for 12 h water (ca. 100 cm3) was added cautiously to the solution with stirring, and the resulting white precipitate was collected by filtration and washed with water. The solid was redissolved in 1,4-dioxane (50 cm3), and concentrated hydrochloric acid (15 cm3) added; the mixture was stirred for 12 h.The reaction mixture was treated by dropwise addition of water (ca. 100 cm3) to induce precipitation of the product, and the solid was then collected on a frit and washed with water. The crude product was recrystallized from acetone. Yield: 2.4 g, 59%. IR (Nujol, cm21) 3275 (nOH), 3111 (nNH), 2846, 1667 (nC]] O) and 1645 (nC]] O). 1H NMR (300 MHz, 298 K, CDCl3) d 1.38 (s, 9 H), 6.92 (m, 1 H), 7.00 (dd, 1 H, J = 8.4, 1.0), 7.43 (m, 1 H), 7.5 (dd, 1 H, J = 5.2, 1.9), 7.54 (s, 1 H), 7.75 (dd, 1 H, J = 8.0, 1.3), 8.02 (s, 1 H), 8.31 (d, 1 H, J = 1.6), 8.50 (dd, 1 H, J = 5.2, 0.5 Hz), 10.12 (s, 1 H), 10.32 (s, 1 H) and 12.05 (s, 1 H). 13C NMR (75.47 MHz, 298 K, CDCl3) d 30.49, 35.24, 114.24, 118.67, 119.02, 119.90, 124.37, 124.61, 125.37, 126.38, 127.93, 129.32, 129.73, 130.32, 134.78, 147.93, 148.45, 162.19, 162.70, 164.28 and 168.72. MS m/z 458 (M1) (Found: C, 60.18; H, 4.58; N, 9.23. Calc. for C23H21Cl2N3O3: C, 60.27; H, 4.62; N, 9.17%). 1,2-Dichloro-4,5-bis(2-hydroxybenzamido)benzene (H4L4).A similar procedure was employed as for the preparation of H3L2. Yield: 1.8 g, 76%. IR (Nujol, cm21) 3280 (nOH), 3100 (nNH), 2880, 1650 (nC]] O). 1H NMR [300 MHz, 298 K, (CD3)2CO] d 6.96 (d, 4 H, J = 7.9), 7.47 (dt, 2 H, J = 7.9, 1.4), 8.02 (dd, 2 H, J = 8.4, 1.6 Hz), 8.14 (m, 2 H), 10.15 (s, 2 H) and 11.49 (s, 2 H). 13C NMR [75.47 MHz, 298 K, (CD3)2CO] d 116.61, 118.46, 120.37, 127.74, 129.19, 129.63, 131.76, 135.45, 160.65 and 168.62. MS m/z 417 (M1) (Found: C, 57.49; H, 3.29; N, 6.81.Calc. for C20H14Cl2N2O4: C, 57.57; H, 3.38; N, 6.71%). 2,3-Dimethyl-2,3-bis(pyridine-2-carboxamido)butane (H2L5). Pyridine-2-carboxylic acid (1.06 g, 8.17 mmol) was added to neat SOCl2 (ca. 10 cm3), and the mixture refluxed for 3 h. An excess of SOCl2 was removed by vacuum evaporation. The solid residue was redissolved in dichloromethane (30 cm3), and the solution added dropwise to another dichloromethane solution (30 cm3) of 2,3-diamino-2,3-dimethylbutane (0.5 g, 4.31 mmol) and triethylamine (ca. 3 cm3). The mixture was allowed to stand overnight with stirring. Removal of solvent under vacuum left an oily substance. The crude product was purified on a silica gel column using diethyl ether–light petroleum (bp 40–60 8C) (9 : 2 v/v) as the eluent. The product was obtained as a white solid. Yield: 0.85 g, 60%. IR (Nujol, cm21) 3100 (nNH), 3000, 2890, 1680 (nC]] O). 1H NMR (300 MHz, 298 K, CDCl3) d 1.69 (s, 12 H), 7.39 (t, 2 H, J = 6.1), 7.82 (t, 2 H, J = 7.7), 8.18 (d, 2 H, J = 7.7), 8.57 (d, 2 H, J = 3.8 Hz) and 8.91 (s, 2 H). 13C NMR (75.47 MHz, 298 K, CDCl3) d 22.34, 60.68, 121.80, 124.61, 125.82, 137.19, 147.96, 150.91 and 164.25. MS m/z 326 (M1) (Found: C, 66.31; H, 6.82; N, 17.12. Calc. for C18H22N4O2: C, 66.24; H, 6.79; N, 17.17%). Syntheses of nitrido-ruthenium(VI) and -osmium(VI) complexes [RuVIN(L1)Cl] 1. To a stirred solution of H2L1 (1 g, 2.1 mmol) and 2,6-dimethylpyridine (0.5 cm3) in anhydrous chloroform (10 cm3) was added a methanolic solution (30 cm3) of [NBun 4]- [RuVINCl4] (1 g, 2.0 mmol) and the mixture stirred for 2 h.The purple precipitate formed was collected by filtration, and washed first with chloroform then with diethyl ether. The product was dried on a frit by vacuum suction. Yield: 0.46 g, 37% (Found: C, 63.70; H, 4.25; N, 4.45. Calc. for C33H27- ClN2O2Ru: C, 63.92; H, 4.39; N, 4.52%); IR (Nujol, cm21) 3060, 2922, 1600, 1573 and 1025. 1H NMR (300 MHz, 298 K, CDCl3) d 3.75 (d, 2 H, J = 14.8), 4.55 (d, 2 H, J = 14.8 Hz) and 6.9–8.2 (m, 23 H). FAB MS m/z 585 (M1 2 Cl) and 570 (M1 2 Cl 2 N). [MVIN(L2)] 2 (M 5 Ru a or Os b). To a methanolic solution (30 cm3) of [NBun 4][MVI(N)Cl4] (0.2 g, 0.4 mmol) were added H3L2 (0.13 g, 0.4 mmol) in chloroform (10 cm3) and 2,6- dimethylpyridine (ca. 0.5 cm23). The mixture was stirred for 2 h and a brown precipitate gradually deposited. The brown solid was collected, washed with acetone, and then dried in vacuo.For 2a: yield 0.10 g (56%) (Found: C, 48.36; H, 3.04; N, 9.76. Calc. for C23H18Cl2N4O3Ru: C, 48.43; H, 3.18; N, 9.82%); IR (Nujol, cm21) 2890, 1680 (nC]] O), 1620 (nC]] O) and 1102; 1H NMR [300 MHz, 298 K, (CD3)2SO] d 1.51 (s, 9 H), 6.94 (t, 1 H, J = 7.2), 7.24 (d, 1 H, J = 7.5), 7.42 (t, 1 H, J = 6.7), 8.29 (m, 3 H), 8.69 (s, 1 H) and 9.30 (d, 2 H, J = 5.7 Hz); FAB MS m/z 571 (M1 1 1) and 558 (M1 2 N). For 2b: yield 0.19 g (72%) (Found: C, 41.66; H, 2.65; N, 8.44.Calc. for C23H18Cl2N4O3Os: C, 41.89; H, 2.75; N, 8.49%). IR (Nujol, cm21) 2890, 1690 (nC]] O), 1620 (nC]] O) and 1068; 1H NMR [300 MHz, 298 K, (CD3)2SO] d 1.45 (s, 9 H), 6.93 (m, 1 H), 7.23 (d, 1 H, J = 8.1), 7.43 (m, 1 H), 8.23 (dd, 1 H, J = 8.1, 1.5), 8.29 (d, 2 H, J = 2.6 Hz), 8.66 (s, 1 H), 9.21 (s, 1 H) and 9.27 (m, 1 H); FAB MS m/z 661 (M1 1 1). [NBun 4][MVIN(L3)] 3 (M 5 Ru a or Os b). To a stirred methanolic solution (10 cm3) of [NBun 4][MVINCl4] (0.5 g, 1.0 mmol) was added H4L3 (0.35 g, 1.0 mmol) and 2,6-dimethylpyridine (ca. 0.5 cm3), and the mixture was stirred for 1 d. Solvent evaporation by vacuum left an oily residue, which was chromatographed on an alumina column (activity 90, neutral) using chloroform as the eluent. The orange (Ru) or yellow (Os) band was collected in one portion and concentrated to ca. 2 cm3 by rotary evaporation. Addition of diethyl ether to the red solution led to isolation of the complex as orange or yellow prismshaped crystals.The crystalline solid was collected on a frit and dried in vacuo. For 3a: yield 0.3 g (43%) (Found: C, 61.52; H, 6.76; N, 7.76. Calc. for C36H48N4O4Ru: C, 61.61; H, 6.89; N, 7.98%); IR (Nujol, cm21) 3066, 2850, 1615 (nC]] O), 1580 and 1073; 1H NMR (300 MHz, 298 K, CDCl3) d 0.78 (t, 12 H, J = 7.0), 1.13 (m, 16 H), 2.63 (m, 8 H), 6.91 (m, 2 H), 6.99 (m, 2 H), 7.27 (d, 2 H, J = 6.9), 7.36 (m, 2 H), 8.20 (dd, 2 H, J = 7.9, 1.6 Hz) and 8.89 (m, 2 H); FAB MS (negative) m/z 460 (M2).For 3b: yield 0.54 g (68%) (Found: C, 54.48; H, 6.19; N, 7.22. Calc. for C36H48N4O4Os: C, 54.66; H, 6.12; N, 7.08%); IR (Nujol, cm21) 3064, 2580, 1625 (nC]] O), 1595 and 1108; 1H NMR (300 MHz, 298 K, CDCl3) d 0.80 (t, 12 H, J = 7.0), 1.15J. Chem. Soc., Dalton Trans., 1998, 3183–3190 3185 (m, 16 H), 2.63 (m, 8 H), 6.97 (m, 4 H), 7.36 (m, 4 H), 8.27 (d, 2 H, J = 7.8) and 8.97 (dd, 2 H, J = 6.3, 3.5 Hz); FAB MS (negative) m/z 548 (M2).[NBun 4][MVIN(L4)] 4 (M 5 Ru a or Os b). The complexes were synthesized in a manner similar to that for 3. For 4a: yield 0.14 g, 53% (Found: C, 56.03; H, 6.06; N, 7.32. Calc. for C36H46- Cl2N4O4Ru: C, 56.10; H, 6.02; N, 7.27%); IR (Nujol, cm21) 3020, 2965, 2870, 1600 and 1010; 1H NMR (300 MHz, 298 K, CDCl3) d 0.84 (t, 12 H, J = 7.1), 1.18 (m, 8 H), 1.32 (m, 8 H), 2.80 (m, 8 H), 6.93 (m, 2 H), 7.28 (d, 2 H, J = 9.2), 7.35 (m, 2 H), 8.19 (d, 2 H, J = 7.9 Hz) and 9.21 (s, 2 H); FAB MS (negative) m/z 528 (M2).For 4b: yield 0.66 g (77%) (Found: C, 50.33; H, 5.23; N, 6.67. Calc. for C36H46Cl2N4O4Os: C, 50.28; H, 5.39; N, 6.52%); IR (Nujol, cm21) 3020, 2990, 2850, 1600(nC]] O) and 1110; 1H NMR (300 MHz, 298 K, CDCl3) d 0.85 (t, 12 H, J = 7.1), 1.19 (m, 8 H), 1.32 (m, 8 H), 2.76 (m, 8 H), 6.95 (m, 2 H), 7.37 (m, 4 H), 8.26 (dd, 2 H, J = 8.0, 1.5 Hz) and 9.28 (s, 2 H); FAB MS (negative) m/z 617 (M2). [MVIN(H2L5)Cl3] 5 (M 5 Ru a or Os b]. To a stirred solution of [NBun 4][MVI(N)Cl4] (0.5 g, 1.0 mmol) in methanol (10 cm3) were added H2L5 (0.326 g, 1.0 mmol) in chloroform (5 cm3) and a few drops of 2,6-dimethylpyridine and stirred for 3 h.The pink solid formed was collected, washed with methanol and dried in vacuo. For 5a: yield 0.21 g (38%) (Found: C, 39.36; H, 4.22; N, 12.62. Calc. for C18H22Cl3N5O2Ru: C, 39.46; H, 4.05; N, 12.78%); IR (Nujol, cm21) 3320 (nNH), 2950, 1650 (nC]] O), 1640 (nC]] O) and 1067; 1H NMR (300 MHz, 298 K, CDCl3) d 1.53 (s, 6 H), 1.61 (s, 6 H), 7.53 (t, 1 H, J = 6.0), 7.92 (m, 2 H), 8.20 (d, 1 H, J = 7.9), 8.42 (t, 1 H, J = 7.3), 8.54 (s, 1 H), 8.59 (d, 1 H, J = 4.4), 8.78 (d, 1 H, J = 7.9), 9.27 (d, 1 H, J = 7.1 Hz) and 11.40 (s, 1 H); FAB MS m/z 512 (M1 2 Cl).For 5b: yield 0.35 g (55%) (Found: C, 33.77; H, 3.55; N, 10.83. Calc. for C18H22Cl3- N5O2Os: C, 33.94; H, 3.48; N, 10.99%); IR (Nujol, cm21) 3320 (nNH), 2950, 1660 (nC]] O), 1640 (nC]] O) and 1112; 1H NMR (300 MHz, 298 K, CDCl3) d 1.52 (s, 6 H), 1.62 (s, 6 H), 7.54 (dd, 1 H, J = 7.1, 4.7), 7.87 (t, 1 H, J = 6.5), 7.95 (dt, 1 H, J = 7.8, 1.4), 8.21 (d, 1 H, J = 7.8), 8.31 (t, 1 H, J = 7.8), 8.57 (s, 1 H), 8.60 (d, 1 H, J = 4.4), 8.82 (d, 1 H, J = 8.0), 9.15 (d, 1 H, J = 5.4 Hz) and 11.63 (s, 1 H); FAB MS m/z 637 (M1) and 602 (M1 2 Cl).Reactions with triphenylphosphine [RuIV(NPPh3)L1(py)Cl]. Triphenylphosphine (42 mg, 0.16 mmol) was added under an argon atmosphere to a dichloromethane solution (20 cm3) of complex 1 (100 mg, 0.16 mmol) and pyridine (1 cm3) with stirring.An instantaneous change from yellow-orange to a dark green solution occurred; this was stirred for 1 h and the solvent removed by rotary evaporation. The green residue was then loaded on an alumina column and eluted with chloroform, the green band being collected in one fraction. After solvent evaporation a green solid was obtained. Yield: 90 mg (58%) (Found: C, 69.88; H, 4.75; N, 4.55. Calc.for C56H47ClN3O2PRu: C, 69.95; H, 4.93; N, 4.37%); IR (Nujol, cm21) 2850, 1650 (nC]] O), 1605 and 1160 (nN]] P). 1H NMR (300 MHz, 298 K, CDCl3) d 3.32 (d, 2 H, J = 14.6), 4.90 (d, 2 H, J = 14.6 Hz) and 6.90–7.90 (m, 38 H). 31P NMR (202.48 MHz, 298 K, CDCl3) d 166.41. FAB MS m/z 926 (M1). [NBun 4][RuIVL(Hpz)(pz)] (L 5 L3 or L4). A light orange acetonitrile solution containing the nitridoruthenium complex 3a or 4a (100 mg, 0.13 mmol) and pyrazole (Hpz, 198 mg, 0.26 mmol) was treated with triphenylphosphine (34 mg, 0.13 mmol) under an argon atmosphere; an instantaneous change to dark green occurred.The reaction mixture was stirred for 15 min and then rotary evaporated to dryness to leave a dark green solid. The solid was extracted by diethyl ether (3 × 20 cm3), and the combined organic extracts were evaporated to dryness to leave HN]] PPh3 as a white solid [EI MS 277 (M1); 31P NMR d 130.1]. The dark green residue was recrystallized by slow diVusion of diethyl ether into its acetonitrile solution, and a dark green microcrystalline solid was collected by filtration.For [NBun 4]- [RuIVL3(Hpz)(pz)]: yield 80 mg (76%) (Found: C, 61.11; H, 6.66; N, 11.89. Calc. for C42H55N7O4Ru: C, 61.29; H, 6.74; N, 11.91%); IR (Nujol, cm21) 2925, 1595 (nC]] O) and 1460; ES MS (negative) m/z 581 (M2). For [NBun 4][RuIVL4(Hpz)(pz)]: yield 96 mg (83%) (Found: C, 56.45; H, 5.88; N, 10.78. Calc. for C42H53Cl2N7O4Ru: C, 56.56; H, 5.99; N, 10.99%); IR (Nujol, cm21) 2930, 1600 (nC]] O) and 1460; ES MS (negative) m/z 649 (M2).[RuIVL2(Hpz)(pz)]. A round-bottom flask (25 cm3) was charged with complex 2a (100 mg, 0.17 mmol), pyrazole (198 mg, 0.35 mmol), triphenylphosphine (46 mg, 0.17 mmol) and dichloromethane (10 cm3). The orange-brown suspension was stirred for 0.5 h, and a homogeneous dark brown solution resulted. After complete removal of solvent by rotary evaporation a brown solid was isolated. The solid residue was extracted by diethyl ether (3 × 20 cm3), and the combined organic extracts were evaporated to dryness to leave HN]] PPh3 as a white residue [EI MS m/z = 277 (M1); 31P NMR d 130.1].The remaining brown solid was recrystallized by slow diVusion of diethyl ether into its chloroform solution, and a brown microcrystalline solid was isolated by filtration. Yield: 100 mg, 88% (Found: C, 50.25; H, 3.45; N, 14.32. Calc. for C29H25- Cl2N7O3Ru: C, 50.37; H, 3.64; N, 14.18%); IR (Nujol, cm21) 2850, 1680 (nC]] O), 1630 (nC]] O) and 1460. FAB MS m/z 693 (M1 1 1).Results and discussion Synthesis and spectral characterization Multianionic chelating ligands are strong s donors capable of stabilizing metal ions in high oxidation states. Notable examples are complexes of CuIII, MnV and FeIV which are readily attainable by using tetraanionic amide ligands.13 In this work we have prepared a series of auxiliary (N, O) ligands, H2L1, H3L2 and H4L3,4, with diVerent electronic charges upon deprotonation of the amide and hydroxyl groups.The H2L1 ligand was first reported by Berg and Holm,9 and its high-valent dioxo- and nitrido-osmium(VI),14 oxo-9 and bis(imido)-molybdenum( VI) 15 derivatives have been prepared and structurally characterized. On the contrary, the use of tetradentate trianionic ligands to support high-valent metal centres has received relatively less attention, a noteworthy example being the preparation of an iron(IV) metallocorrole complex by Vogel et al.16 The H3L2 ligand was prepared by the method developed by Kabanos and co-workers.10,17 The presence of a tert-butyl substituent rendered the ligand more soluble in organic solvents.We are interested to examine the eVect of the electrondonating strength (in terms of formal anionic charge) of the auxiliary ligands on the M]] ] N functions, especially on its electrophilicity. We envisaged that the increase in s donation from the ligand would promote the electron density at the metal centre; as a result the electrophilicity of the nitrido ligand would be reduced.1 Several methods are known for the preparation of highvalent terminal nitrido-ruthenium(VI) and -osmium(VI) complexes: (1) oxidative deprotonation of co-ordinated amines,3a (2) thermal decomposition of azidoosmium complexes,8,18 and recently (3) reduction of nitrosyl complexes.19 A synthetic route based on the ligand substitution reactions of [MVINCl4]2 provides the easiest entry to our target complexes, and similar methods had previously been exploited by Shapley and coworkers 20 for the preparation of nitrido-ruthenium(VI) and -osmium(VI) bearing anionic N, S, O or carboxylate ligands. Treatment of [NBun 4][MVINCl4] with HnL1–4 (n=1–4) in methanol–chloroform (5: 1 v/v) with a few drops of 2,6- dimethylpyridine (acting as a base) aVords the desired nitrido-3186 J.Chem. Soc., Dalton Trans., 1998, 3183–3190 metal complexes in moderate to good yields (Scheme 1).However, this method is apparently limited to those ligands bearing OH groups, for example a pyridine amide ligand (H2L5) lacking any OH groups failed to eVect metal encapsulations, instead pyridyl nitrido-ruthenium(VI) and -osmium(VI) adducts [MVIN(H2L5)Cl3] 5a and 5b were produced (Scheme 2). Complexes 1, 2a, 2b and 5a, 5b were isolated as insoluble solids in good purities as revealed by 1H NMR spectroscopy. Analytically pure samples of 3a, 3b and 4a, 4b were obtained after purification by column chromatography on alumina using chloroform as the eluent.Good quality crystals were grown by slow diVusion of diethyl ether into dichloromethane or chloroform solutions. All the nitrido complexes prepared in this work are diamagnetic solids, consistent with the formulation of a singlet (dxy)2 electronic ground state (taking the M]] ] N axis as the z direction). Their infrared spectra do not show any NH and/or OH stretches, suggesting that the chelating ligands are in completely deprotonated forms.However the M]] ] N stretches cannot be unequivocally assigned because of extensive overlap with the Scheme 1 (i) For H2L1 and H3L2, 2,6-dimethylpyridine, MeOH–CHCl3, room temperature (r.t.); (ii) for H4L3 and H4L4, 2,6-dimethylpyridine, MeOH, r.t. [NBun 4] M = Ru, Os N MVI Cl Cl Cl Cl N M N M Cl O N O N N O N But 1 2 N M N N O O N M N N O O Cl Cl 3 4 VI VI VI VI (i) (ii) Scheme 2 M Cl Cl N N NH O HN O H2L5 MeOH, r.t.[NBun 4] M = Ru, Os N MVI Cl Cl Cl Cl 5 Cl N ligand absorptions in the same spectral region. Their UV/VIS spectra are featureless and dominated by intense absorption(s) at l < 250 nm. Complexes 1, 3a, 3b and 4a, 4b show similar 1H NMR spectral patterns to those of their respective unbound ligands, suggesting that the mirror symmetry of the chelating ligands is retained after co-ordination. For 3 and 4 the aromatic protons on the ligands are downfield-shifted relative to the unbound states probably due to inductive withdrawal of electrons by the electrophilic metal centre.All the 1H NMR spectra of the nitridometal complexes show no amido and phenolic proton absorptions, consistent with the infrared results that the ligands are in their deprotonated forms. The 1H NMR spectra of the ruthenium and the analogous osmium complexes revealed similar patterns implying an isostructural relationship. Structural studies of the nitrido-ruthenium and -osmium complexes The crystal structures of some nitrido-ruthenium(VI) (1, 3a, 4a and 5a) and -osmium(VI) (2b, 3b and 5b) complexes were determined by single-crystal X-ray diVraction method.The crystal data, selected bond distances and angles are listed in Tables 1 and 2 respectively. All the nitridometal complexes prepared in this work are five-co-ordinated (except 5a and 5b); this can be ascribed to the strong trans eVect exerted by the nitride (N32) ligand.As revealed by X-ray crystal analysis, complexes 3a and 4a are isostructural, and for illustration a perspective view of 4a is shown in Fig. 1. Both complexes adopt a distorted square pyramidal co-ordination with the terminal nitride ligand in the apical position. The metal centre is slightly elevated by ª0.58 Å from the mean equatorial plane consisting of the four N- and O-donor atoms; a similar structural feature has been found for some other nitridometal complexes.21 The RuVI]N (amide) bond distances for 3a and 4a are in a range of 1.991(6)–2.006(3) Å, significantly shorter than the RuII]N(sp3) {2.144(4) Å in [Ru(NH3)6]I2},22 RuIII]N(sp3) {2.104–2.117 Å in cis-[RuIII- ([14]aneN4)Cl2]Cl} 23 and RuIV]N(sp3) {2.085–2.141(5) Å in trans-[RuIV(tmc)O(MeCN)][PF6]2} 24 ([14]aneN4 = 1,4,8,11- tetraazacyclotetradecane, tmc = 1,4,8,11-tetramethyl-1,4,8,11- tetraazacyclotetradecane) distances involving saturated amine ligands, but the values are close to those [1.987(5)–2.044(5) Å] of [RuIV(chbae)(PPh3)(py)] 25 [chbae = 1,2-bis(3,5-dichloro-2- hydroxybenzamido)ethane tetraanion]. Complexes 5a and 5b are six-co-ordinated and isostructural; the metal atoms adopt a distorted octahedral co-ordination where the pyridine amide ligand (H2L5) binds to the M]] ] N moiety in a bidentate fashion via the pyridyl nitrogen and the carbonyl oxygen atoms.The carbonyl oxygen atom is opposite to the N32 ligand. The Ru]] ] N and Os]] ] N distances are found to be 1.598(3) and 1.612(5) Å respectively.Fig. 1 Perspective view of [RuVIN(L4)]2 4a.J. Chem. Soc., Dalton Trans., 1998, 3183–3190 3187 The structure of complex 1 (Fig. 2) is again isostructural to the osmium(VI) analogue [OsVIN(L1)Cl] reported earlier,14 and the co-ordination polyhedron around the metal atom can be described as distorted trigonal bipyramidal with the N(1), O(1) and O(2) atoms on the trigonal plane. The N(2)]Ru]Cl(1) axis significantly deviates from linearity [167.0(1)8].The respective N(1)]Ru]O(1), O(1)]Ru]O(2) and N(1)]Ru]O(2) bond angles are 117.7(2), 129.2(1) and 112.4(2)8, and the two six-membered chelate rings are in a gauche conformation. The trigonal plane defined by the Ru, N(1), O(1) and O(2) atoms is a near isosceles triangle. The Ru]O(1) and the Ru]O(2) bond distances are 1.902(3) and 1.921(3) Å respectively, comparable to those [2.007(3) and 1.897(3) Å] found in [NPr4][RuVO(phab)] 26 [phab = 1,2-bis(2-hydroxy-2,2-diphenylethanamido)benzene tetraanion].The Ru]] ] N distance of 1.615(4) Å is close to that found in the tetraanionic (L3)42 [1.594(4) Å] and (L4)42 [1.609(6) Å] analogues. On the other hand the alkoxide (O2) and amide (N2) donor atoms of the L3 and L4 tetraanions are regarded as considerably more basic than the chloride ligand; notwithstanding, these Ru]] ] N distances are comparable to those found for [RuVIN(H2L5)Cl3] 5a [Ru]] ] N 1.598(3) Å], [RuVINCl4]2 (Ru]] ] N 1.570 Å) 22 and [RuVINMe4]2 (Ru]] ] N 1.58 Å).27 Although we could not obtain crystals of [RuVIN(L2)] 2a suitable for X-ray crystal analysis, the molecular structure of the osmium(VI) analogue 2b has been determined. As shown in Fig. 3, complex 2b is five-co-ordinated and exhibits distorted square pyramidal co-ordination at the osmium centre; the Os]] ] N bond distance was found to be 1.621(4) Å. Similar insensitivity of the Os]] ]N bond distances to the number of anionic donor atoms is also observed for the isostructural nitridoosmium(VI) analogues [OsVIN(L1)Cl] [1.634(5) Å],22 [NBun 4][OsVIN(L3)] 3b [1.618(7) Å], [OsVIN(H2L5)Cl3] 5b [1.612(5) Å] and [OsVINCl5]22 (1.614 Å).28 Unlike the oxoruthenium( IV) systems, the O]] Ru bond strength is sensitive to the electron-donating power of, as well as the presence of p Fig. 2 Perspective view of [RuVIN(L1)Cl] 1. Fig. 3 Perspective view of [OsVIN(L3)] 2b. unsaturation in, equatorial macrocyclic amine ligands.24 The Os–O(1) (2.010 Å) and Os–N (amide) (2.004 and 1.962 Å) distances are comparable to those for [NBun 4][OsVIN(L3)] 3b [Os– O 1.97 Å, Os–N (amide) 2.010 and 2.011 Å].The Os–N (py) distance (2.088 Å) is also close to the reported value (2.119 Å) for [OsVIN(L1)Cl].14 Nitrogen-atom transfer reactions of nitridoruthenium to PPh3 To examine the possible influence of the electron-donating power of the auxiliary ligands on the electrophilicity of the Ru]] ] N function, triphenylphosphine was used as a nucleophile to eVect a series of nitrogen atom transfer reactions. When [RuVIN(L1)Cl] was treated with a stoichiometric amount of PPh3 in dry dichloromethane–pyridine (20 : 1) an instantaneous change from a yellow orange to dark green solution occurred (Scheme 3).After column chromatography on alumina a green solid was isolated. The IR spectrum of the green compound revealed a strong absorption peak at 1106 cm21 which is conspicuously absent from the spectrum of complex 1.According to related studies this could be assigned as a N]] P stretch.29 The 31P NMR spectrum of the compound showed a singlet at d 166.4 vs. 85% H3PO4, which is close to other reported values for a phosphiniminate co-ordinated to a metal centre.30 With the aid of mass spectroscopy (M1 2 Cl m/z = 926), 1H NMR and elemental analysis, the green solid can be formulated as [RuIV(N]] PPh3)L1(py)Cl]. When the triphenylphosphine reduction was monitored by 31P NMR spectroscopy in a CD2Cl2– C5D5N mixture [RuIV(N]] PPh3)L1(py)Cl] (31P = d 166.4) was formed instantaneously upon mixing the reactants, no absorption (d 23.5) corresponding to free PPh3 being detected indicates that all PPh3 was consumed on the NMR timescale.However, on prolonged standing (14 h) of the reaction mixture the 31P NMR spectrum revealed only a singlet at d 130.1 (Ph3P]] NH) and the complex [RuIIIL1(py)3]Cl was isolated and characterized by spectroscopic as well as X-ray crystallographic means.31 Likewise complexes 2a, 3a and 4a undergo facile reactions with triphenylphosphine, however we were unable to isolate analytically pure ruthenium(IV) phosphiniminato complexes when CH2Cl2–pyridine (20 :1) was employed as the reaction medium (Scheme 4). The 31P NMR spectra recorded for their reactions with PPh3 in CD2Cl2–C5D5N mixture revealed a weak singlet absorption at d 130.1 (Ph3P]] NH) only, and yet no free PPh3 was detected.When the triphenylphosphine reductions were carried out using CH2Cl2 (2a) or MeCN (3a and 4a) as solvent in the presence of an excess of pyrazole (Hpz) (10 equivalents vs.Ru]] ] N) some green crystalline solids were obtained by addition of diethyl ether (>76% isolated yield). The product complexes are paramagnetic solids with meff = 2.9 mB (mB = 9.27 × 10224 J T21) consistent with a ground state electronic configuration with two unpaired electrons (S = 1) Scheme 3 py/CH2Cl2 prolonged reaction structure determined Ru N O N Cl O PPh3 Ru O N O N PPh3 IV VI py Cl Ru O N O III (py)3 Cl3188 J.Chem. Soc., Dalton Trans., 1998, 3183–3190 Table 1 Crystallographic data for the nitrido-ruthenium(VI) and -osmium(VI) complexes Formula M Crystal symmetry Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z DiVractometer Dc/g cm23 No. collected data No. unique data No. data used No. parameters m(Mo-Ka)/cm21 F(000) RR 9 Goodness of fit 1 C33H27ClN2O2Ru 620.11 Monoclinic P21/n (no. 14) 9.609(7) 18.444(7) 16.099(8) 104.56(5) 2761(2) 4 Rigaku AFC7R 1.491 5355 5040 3454 352 6.98 1264 0.041 0.046 1.80 2b C23H18Cl2N4O3Os 659.53 Monoclinic P21/n (no. 14) 14.107(2) 10.225(1) 15.216(9) 91.581(7) 2194.0(4) 4 Rigaku AFC7R 1.997 4051 3884 3223 298 60.87 1272 0.020 0.028 1.39 3a C36H48N4O4Ru 701.87 Orthorhombic Pbca (no. 61) 15.346(3) 19.721(3) 22.744(4) 6883(1) 8 MAR 1.354 53014 7048 3902 405 4.99 2944 0.042 0.056 2.14 4a?CHCl3 C37H47Cl5N4O4Ru 890.14 Triclinic P1� (no. 2) 12.074(2) 13.118(2) 13.648(2) 75.65(2) 84.14(2) 86.46(2) 2081.8(6) 2 MAR 1.420 28929 5737 3695 458 7.39 916 0.056 0.076 1.52 3b C36H48N4O4Os 791 Orthorhombic Pbca (no. 61) 19.816(2) 22.746(4) 15.430(2) 6955(1) 8 Rigaku AFC7R 1.511 6753 6753 3047 405 37.08 3200 0.031 0.041 1.29 5a C18H22Cl3N5O2Ru 546.83 Monoclinic P21/n (no. 14) 11.817(2) 10.165(2) 19.091(3) 100.72(2) 2254.3(8) 4 MAR 1.611 20165 4128 2986 265 10.74 1100 0.032 0.046 2.24 5b C18H22Cl3N5O2Os 635.96 Monoclinic P21/n (no. 14) 11.825(2) 10.276(4) 19.011(1) 100.952(7) 2267.9(8) 4 Rigaku AFC7R 1.862 4464 4247 2944 265 59.96 1228 0.028 0.030 1.33 R = S||Fo| 2 |Fc||/S|Fo|; R9 = [Sw(|Fo| 2 |Fc|)2/SwFo 2]� �� where w = 4Fo 2/s2(Fo 2).J. Chem. Soc., Dalton Trans., 1998, 3183–3190 3189 Table 2 Selected bond distances (Å) and angles (8) for the ruthenium(VI) and osmium(VI) nitrido complexes Complex 1 Ru]N(1) Ru]N(2) Ru]O(1) Ru]O(2) Ru]Cl Complex 3aa Ru]N(1) Ru]N(2) Ru]N(3) Ru]O(1) Ru]O(2) Complex 3ba Os]N(1) Os]N(2) Os]N(3) Os]O(1) Os]O(2) 1.615(4) 2.125(4) 1.902(3) 1.921(3) 2.366(1) 1.594(4) 2.006(3) 1.992(4) 1.956(3) 1.957(3) 1.618(7) 2.011(6) 2.010(6) 1.969(5) 1.977(5) N(1)]Ru]N(2) O(1)]Ru]O(2) N(1)]Ru]O(1) N(1)]Ru]O(2) O(1)]Ru]O(2) N(2)]Ru]Cl N(1)]Ru]N(2) N(1)]Ru]N(3) N(1)]Ru]O(1) N(1)]Ru]O(2) N(2)]Ru]O(2) O(1)]Ru]N(3) N(1)]Os]N(2) N(1)]Os]N(3) N(1)]Os]O(1) N(1)]Os]O(2) N(2)]Os]O(2) O(1)]Os]N(3) 95.3(2) 129.2(1) 117.7(2) 112.4(2) 129.2(1) 167.0(1) 105.8(2) 106.0(2) 108.4(2) 108.0(1) 146.2(1) 145.6(1) 107.0(3) 106.1(3) 109.0(3) 107.4(3) 145.5(2) 144.9(2) Complex 2b Os]N(1) Os]N(2) Os]N(3) Os]N(4) Os]O(1) Complex 4a Ru]N(1) Ru]N(2) Ru]N(3) Ru]O(1) Ru]O(2) 1.621(4) 2.004(3) 1.962(3) 2.088(3) 2.010(3) 1.609(6) 1.991(5) 1.992(6) 1.960(4) 1.955(5) N(1)]Os]N(2) N(1)]Os]N(3) N(1)]Os]N(4) N(1)]Os]O(1) N(2)]Os]N(4) O(1)]Os]N(3) N(1)]Ru]N(2) N(1)]Ru]N(3) N(1)]Ru]O(1) N(1)]Ru]O(2) N(2)]Ru]O(2) O(1)]Ru]N(3) 104.0(2) 107.4(2) 98.2(1) 105.4(2) 154.7(1) 147.2(1) 105.8(3) 106.8(3) 108.3(3) 107.3(3) 146.9(2) 14 Same atom labelling scheme is adopted.indicating that the ruthenium centre is in the 14 oxidation state. According to the mass spectroscopic analyses, molecular ion peaks at m/z = 693 (M1 1 1), 581 (M2) and 649 (M2), these green solids could be formulated as [RuIVL2(Hpz)(pz)], [RuIVL3( Hpz)(pz)]2 and [RuIVL4(Hpz)(pz)]2 respectively. The cyclic voltammogram of [NBu4][RuIVL4(Hpz)(pz)] in 0.1 mol dm23 NBu4PF4 (CH2Cl2) showed three quasi-reversible couples at 10.8, 10.04 and 21.24 V vs.Ag–AgNO3. All the triphenylphosphine reductions of the nitridoruthenium complexes take place spontaneously. When the reaction with [NBun 4][RuVIN(L4)] was monitored by UV/VIS spectroscopy immediately after mixing of the reactants a new species exhibiting an absorption band at 551 nm was generated. This species slowly converted into [NBun 4][RuIVL4(Hpz)(pz)] (over 12 h) manifested by the gradual decay of the 551 nm band, suggesting that the putative RuIV(N]] PPh3) species is unstable toward subsequent ligand substitution.The failure to Scheme 4 N RuVIL N RuIVL PPh3 Hpz RuIVL pz N But N N N N Cl Cl + n PPh3 / Hpz (excess) CH2Cl2 or MeCN n � n = 0; when L = N N O O O O O (L2)3- (L3)4- (L4)4- or n = –1; when L = HN PPh3 observe the ruthenium(IV) intermediate by 31P NMR spectroscopy could be ascribed to the paramagnetic nature of the molecule. Despite previous reports that [OsVIN(terpy)Cl2]Cl will also react with PPh3 to give [OsIV(N]] PPh3)(terpy)Cl2]Cl (terpy = 2,29:69,20-terpyridine),32 the analogous reaction of [NBun 4]- [OsVIN(L4)] was found to be sluggish.This observation is consistent with the findings from related studies on the oxoruthenium( VI) and -osmium(VI) complexes that high-valent ruthenium is a stronger oxidant than its osmium counterpart.2,6 A 31P NMR spectrum recorded after 12 h of reaction revealed that the mixture contained largely unchanged PPh3 (d 23.5).Conclusion A series of nitrido-ruthenium(VI) and -osmium(VI) complexes containing di-, tri- and tetra-anionic ligands was prepared via ligand substitution reactions. The M]] ] N bond distance is invariant to the electron-donating power of the auxiliary ligands. All the nitridoruthenium(VI) complexes react readily with triphenylphosphine, and the intermediate [RuIV(N]] PPh3)(L1)- (py)Cl] can be isolated and characterized spectroscopically for the reaction with [RuVIN(L1)Cl]. However, for those nitridoruthenium complexes bearing the tri- (L2)32 and tetra-anionic (L3,4)42 ligands the phosphiniminatoruthenium(IV) intermediate undergoes further reaction with pyrazole to generate a bis- (pyrazole)ruthenium(IV) complex as the product.Acknowledgements We gratefully acknowledge support from The University of Hong Kong and The Hong Kong Research Grants Council. References 1 W. A. Nugent and J. M. Mayer, Metal–Ligand Multiple Bonds, Wiley, New York, 1988. 2 C.-M. Che, Pure Appl. Chem., 1994, 67, 225. 3 (a) C.-M. Che, T.-C. Lau, H.-W. Lam and C.-K. Poon, J. Chem. Soc., Chem. Commun., 1989, 114; (b) H.-W. Lam, C.-M. Che and K.-Y. Wong, J. Chem Soc., Dalton Trans., 1992, 1411; (c) K.-F. Chin, K.-K. Cheung, H.-K. Yip, T. C.-W. Mak and C.-M. Che, J. Chem. Soc., Dalton Trans., 1995, 657. 4 C.-M. Che, K.-Y. Wong, H.-W. Lam, K.-F. Chin, Z.-Y. Zhou and T. C.-W. Mak, J. Chem. Soc., Dalton Trans., 1993, 857.3190 J.Chem. Soc., Dalton Trans., 1998, 3183–3190 5 D. W. Pipes, M. Bakir, S. E. Vitols, D. J. Hodgson and T. J. Meyer, J. Am. Chem. Soc., 1990, 112, 5507. 6 C.-M. Che and V. W.-W. Yam, Adv. Inorg. Chem., 1992, 39, 233; Adv. Transition Metal Coord. Chem., 1996, 1, 209. 7 W. R. Murphy, jun., K. J. Takeuchi and T. J. Meyer, J. Am. Chem. Soc., 1982, 104, 5817; D. W. Pipes and T. J. Meyer, Inorg. Chem., 1984, 23, 2466; W. R. Murphy, jun., K. Takeuchi, M. H. Barley and T. J.Meyer, Inorg. Chem., 1986, 25, 1041. 8 W. P. GriYth and D. Pawson, J. Chem. Soc., Dalton Trans., 1973, 1315. 9 J. M. Berg and R. H. Holm, J. Am. Chem. Soc., 1985, 107, 917. 10 A. D. Keramidas, A. B. Papaioannou, A. Vlahos, T. A. Kabanos, G. Bonas, A. Makriyannis, C. P. Rapropoulou and A. Terzis, Inorg. Chem., 1996, 35, 357. 11 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smith and C. Smyklla, The DIRDIF program system, Technical Report of the Crystallography Laboratory, University of Nijmegen, 1992. 12 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, TX, 1985 and 1992. 13 T. J. Collins, Acc. Chem. Res., 1994, 27, 279. 14 Z.-Y. Li, W.-Y. Yu, C.-M. Che, C.-K. Poon, R.-J. Wang and T. C.-W. Mak, J. Chem. Soc., Dalton Trans., 1992, 1657. 15 V. C. Gibson, E. L. Marshall, C. Redshaw, W. Clegg and M. R. J. Elsegood, J. Chem. Soc., Dalton Trans., 1996, 4197. 16 E. Vogel, S. Will, A. S. Tilling, L. Neumann, J. Lex, E. Bill, A. X. Trautwein and K. Wieghardt, Angew. Chem., Int. Ed. Engl., 1994, 33, 731; see also R. R. Schrock, Acc. Chem. Res., 1997, 30, 9. 17 T. A. Kabanos, A. D. Keramidas, A. B. Papaioannou and A. Teris, J. Chem. Soc., Chem. Commun., 1993, 643. 18 C. J. Barner, T. J. Collins, B. E. Mapes and B. D. Santarsiero, Inorg. Chem., 1986, 25, 4322. 19 D. Sellmann, M. W. Wemple, W. Donaubauer and F. W. Heinemann, Inorg. Chem., 1997, 36, 1397. 20 J. J. Schwab, E. C. Wilkinson, S. R. Wilson and P. A. Shapley, J. Am. Chem. Soc., 1991, 113, 6124. 21 W. P. GriYth, Coord. Chem. Rev., 1972, 8, 369. 22 H. C. Stynes and J. A. Ibers, Inorg. Chem., 1971, 10, 2304. 23 C.-M. Che, S.-S. Kwong, T.-F. Lai, C.-K. Poon and T. C.-W. Mak, Inorg. Chem., 1985, 24, 1359. 24 C.-M. Che, K.-Y. Wong and C.-K. Poon, Inorg. Chem., 1986, 25, 1809. 25 C.-M. Che, W.-K. Cheng, W.-H. Leung and T. C.-W. Mak, J. Chem. Soc., Chem. Commun., 1987, 418. 26 N. L. P. Fackler, S. S. Zhang and T. V. O’Halloran, J. Am. Chem. Soc., 1996, 118, 481. 27 P. A. Shapley, H. S. Kim and S. R. Wilson, Organometallics, 1988, 7, 928. 28 D. Bright and J. A. Ibers, Inorg. Chem., 1969, 8, 709. 29 B. F. G. Johnson, B. L. Haymore and J. R. Dilworth, in Comprehensive Coordination Chemistry, ed. G. Wilkinson, Pergamon, Oxford, 1987, vol. 2, pp. 122–125. 30 B. M. Schomber, J. W. Ziller and N. M. Doherty, Inorg. Chem., 1991, 30, 4488; A. Aistars, R. J. Doedens and N. M. Doherty, Inorg. Chem., 1994, 33, 4360. 31 P.-M. Chan and C.-M. Che, unpublished work. 32 M. Bakir, P. S. White, A. Dovletoglou and T. J. Meyer, Inorg. Chem., 1991, 30, 2835. Paper 8/0420

ISSN:1477-9226    

DOI:10.1039/a804204g

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3191–3194 3191 Reactions of oxo- and peroxo-molybdenum complexes with B(C6F5)3: crystal structures of cis-[MoO{OB(C6F5)3}(Á2-ONEt2)2] and cis-[MoO{OB(C6F5)3}{Á2-PhN(O)C(O)Ph}2] Linda H. Doerrer,a Jane R. Galsworthy,a Malcolm L. H. Green,a Michael A. Leech b and Matthias Müller b a Inorganic Chemistry Laboratory, South Parks Road, Oxford, UK OX1 3QR b Chemical Crystallography Laboratory, 9 Parks Road, Oxford, UK OX1 3PD Received 27th April 1998, Accepted 28th July 1998 The oxometal complexes [MoO2(h2-O-NR2)2], R = Et or CH2Ph, reacted with the strong Lewis acid B(C6F5)3 at their oxo functionality to give cis-[MoO{OB(C6F5)3}(h2-ONR2)2], (R = Et 1 or CH2Ph 2).Reaction of the peroxo complex [MoO(O2){h2-PhN(O)C(O)Ph}2] with the same Lewis acid led initially to the formation of [MoO(O2){B(C6F5)3}{h2- PhN(O)C(O)Ph}2] 3, which decomposes to form [MoO{OB(C6F5)3}{h2-PhN(O)C(O)Ph}2] 4. Compounds 1 and 4 have been characterised by X-ray crystallography.Introduction Alkyl peroxide transition metal complexes, A, play a central role as reactive intermediates in selective oxidation reactions which employ alkyl hydrogenperoxide as the oxygen source.1 Reaction is postulated as proceeding according to Scheme 1 via formation of an h1-co-ordinated O–OR ligand. Co-ordination of the organic substrate to the metal centre followed by oxygen transfer results in oxidation of the organic molecule.Sheldon and Van Doorn2c proposed that the main function of the metal catalyst in the co-ordinated peroxide complex was to act as a Lewis acid and remove electron density from the peroxidic oxygen. However, to date, the synthesis of well defined, soluble and reactive molybdenum alkyl peroxide complexes remains an unattainable goal. N,N-Dialkylhydroxylamino complexes containing an h2-ONR2 bound ligand, B, are structurally and electronically closely related to metal alkyl peroxide complexes and have been examined as well defined compounds for catalytic activity.3 Unfortunately these species do not exhibit similar reactivities to those of their alkyl peroxide analogues and are ineVective oxidation catalysts.4 The lack of reactivity has been attributed to strong complexation of the dialkylhydroxylamino ligand to the metal centre.In certain catalytic reactions oxometal complexes may be activated by addition of a Lewis acid co-catalyst which is thought to co-ordinate to the M]] O functionality thereby increasing the electrophilicity of the metal centre.5 We have recently reported the synthesis of several oxometal complexes containing a stable, approximately linear, M]] O–B(C6F5)3 moiety.6 As part of our continuing studies of the reactivity of B(C6F5)3 we describe the reactions of this Lewis acid with oxomolybdenum complexes containing ancillary dialkylhydroxylamino or peroxo ligands.Results and discussion Treatment of cis-[MoO2(h2-ONEt2)2] with 1 equivalent of B(C6F5)3 in toluene under ambient conditions yields cis- M O O R M O N R R A B [MoO{OB(C6F5)3}(h2-ONEt2)2] 1, which can be isolated as colourless crystals.The 11B NMR spectrum of compound 1 exhibits a signal at d 2.4, typical of a four-co-ordinate boron species and shifted upfield from that of B(C6F5)3 (d 59). A 1H NMR spectrum reveals two triplet signals, assignable to the methyl groups, and four multiplet resonances, due to the methylene protons, clearly indicating that co-ordination has occurred at a single Mo]] O unit.The methylene protons are also diastereotopic in the parent complex. An IR spectrum of complex 1 exhibits an absorption at 1008 cm21, tentatively ascribed to the n(N–O) stretching vibration, but assignment of the characteristic Mo]] O stretches is hampered by strong absorption of the fluorinated aryl rings in this region. Full characterising data for compound 1 is detailed in Table 1. The solid state structure of compound 1 has been determined by X-ray crystallography; selected bond angles and distances are reported in Table 2.The structure determination of complex 1 (Fig. 1) reveals that a molecule of B(C6F5)3 is bound to one Mo]] O unit. The B–O bond length [1.510(2) Å] is typical of those found within this family of compounds and the Mo]] O–B unit deviates slightly from linearity [170.09(8)8].6 The molybdenum centre displays pseudo-tetrahedral geometry with terminal oxo units occupying two of the vertices and the midpoint of the N–O bond of each h2-ONEt2 ligand occupying the remaining two.The presence of a Mo]] OB(C6F5)3 moiety and a non-co-ordinated Mo]] O unit within compound 1 allows us to assess the electron withdrawing capabilities of the boron Lewis acid. A significant lengthening of the Mo]] O bond of approximately 0.1 Å is observed upon co-ordination to the Lewis acid; Mo–O(1) 1.808(1), Mo–O(2) 1.6823(1) Å. For comparison the Mo]] O bond distances in the parent complex are 1.714(2) and 1.713(2) Å.6 The benzyl compound, [MoO2{h2-ON(CH2Ph)2}2], was synthesized from [MoO2(acac)2] (acac = h2-C5H7O2). The former complex has been previously reported as synthesized from Na2MoO4?2H2O and N,N-dibenzylhydroxylamine.3e Colourless Scheme 1 Proposed mechanism for alkene oxidation using metal alkyl peroxide complexes.M O O R M O R O A alkene +3192 J. Chem. Soc., Dalton Trans., 1998, 3191–3194 Table 1 Analytical and spectroscopic data for compounds 1, 2, 3 and 4 Complex a 1 [MoO{OB(C6F5)3}(h2-ONEt2)2] C, 38.4 (38.2); H, 2.6 (2.5); B, 1.3 (1.3); N, 3.2 (3.4) Spectroscopic data b IR: 3100–2850w, 1646vs, 1517vs, 1469vs, 1386vs, 1373vs, 1284vs, 1108m, 1096vs, 1008s, 978vs, 940m, 911vs, 791m, 772m, 762m, 748m, 729m, 689m, 676m, 668m, 656m, 628m, 588m, 576m 1H: 2.98 (2 H, m, J 7, CH2), 2.74 (2 H, m, J 7, CH2), 2.66 (2 H, m, J 7, CH2), 2.50 (2 H, m, J 7, CH2), 0.74 (6 H, t, J 7, CH3), 0.48 (6 H, t, J 7, CH3) 13C: 148.3 (d, J 240, C6F5), 140.1 (d, J 240, C6F5), 137.5 (d.J 250, C6F5), 51.7 (s, CH2), 50.6 (s, CH2), 10.4 (s, CH3), 9.4 (s, CH3) 11B: 2.4 (br) 2 [MoO{OB(C6F5)3}{h2-ON(CH2- Ph)2}2] C, 51.6 (51.9); H, 2.5 (2.6); B, 1.1 (1.0); N, 2.7 (2.6) IR: 3100–2820w, 1640m, 1512s, 1488m, 1467vs, 1393m, 1375m, 1354m, 1284m, 1096s, 1015m, 977vs, 949vs, 919s, 904s, 853m, 790m, 773m, 767m, 747m, 699m, 683m, 677m, 668m, 662m, 628m, 613m, 599m 1H: 7.41–6.92 (5 H, m, C6H5), 4.24–3.99 (2 H, m, CH2) 13C: 148.1 (d, J 233, C6F5), 140.2 (d, J 230, C6F5), 137.4 (d, J 236, C6F5), 131.8 (s, C6H5), 131.2 (s, C6H5), 130.7 (s, C6H5), 130.2 (s, C6H5), 130.0 (s, C6H5), 128.8 (s, C6H5), 61.2 (s, CH2), 60.1 (s, CH2) 11B: 2.7 (br) 3 [MoO(O2){B(C6F5)3}{h2-PhN- (O)C(O)Ph}2] C, 48.30 (48.89); H, 2.04 (1.85); B, 0.99 (1.02); N, 2.58 (2.59) 1H:c 7.24–6.65 (m, C6H5) 11B: c 3.8 (br) 4 [MoO{OB(C6F5)3}{h2-PhN(O)- C(O)Ph}2]?0.5C5H12 C, 50.66 (50.73); H, 2.42 (2.36); B, 1.09 (1.00); N, 2.48 (2.55) IR: 3120–2840w, 1900–1790w, 1613w, 1596w, 1493vs, 1453vs, 1447vs, 1351s, 1325m, 1237m, 1203m, 1069m, 1026m, 1014m, 920s, 904vs, 845m, 804m, 757s, 746m, 737m, 700s, 620m, 606s, 595m, 515m, 508m 1H: 7.52–7.17 (m, C6H5) 13C: 148.1 (d, J 237, C6F5), 139.8 (d, J 244, C6F5), 137.0 (d, J 247, C6F5), 133.7 (s, C6H5), 133.4 (s, C6H5), 131.3 (s, C6H5), 130.1 (s, C6H5), 129.9 (s, C6H5), 129.6 (s, C6H5), 129.0 (s, C6H5), 128.5 (s, C6H5), 126.9 (s, C6H5), 126.0 (s, C6H5), 118 [s(br), BC] 11B: 3.3 (br) a Analytical data given as found (calculated) in %, IR data (cm21) determined for KBr discs.b The NMR data (CDCl3, 298 K), unless otherwise stated, given as: chemical shift (d) [relative intensity, multiplicity (J in Hz), assignment]. c In C6D6. crystals of [MoO2{h2-ON(CH2Ph)2}2] were obtained, the single crystal structure of which confirmed the cis-oxo geometry of the starting material but the data were not of suYcient quality to be published. Upon reaction of [MoO2{h2-ON(CH2Ph)2}2] with 1 equivalent of B(C6F5)3 cis-[MoO{OB(C6F5)3}{h2-ON(CH2Ph)2}2] 2 was obtained as colourless, diamond shaped crystals. Compound 2 was fully characterised by spectroscopic techniques and elemental analysis (Table 1) and displays similar features to those of its ethyl analogue, 1.A general structural feature of cis-[MoO2(h2-ONR2)2] complexes is a relatively large O]] Mo]] O bond angle,3a,4 compared to related cis-[MoO2L2] complexes.This suggests that such dialkylhydroxylamino- complexes might be sterically capable of Fig. 1 View of the structure of [MoO{OB(C6F5)3}(h2-ONEt2)2] 1. Fluorine atoms omitted for clarity. binding two molecules of B(C6F5)3. However, reaction of [MoO2{h2-ON(CH2Ph)2}2] with 2 equivalents of the Lewis acid yielded only compound 2, even using prolonged reaction times. We then investigated the reaction of B(C6F5)3 with the peroxomolybdenum complex [MoO(O2){h2-PhN(O)C(O)Ph}2]. This compound has been shown eVectively to oxidise primary and secondary alcohols to the corresponding carbonyl compounds and to be capable of stereospecifically epoxidising allylic alcohols.7 Treatment of [MoO(O2){h2-PhN(O)C(O)Ph}2] with 1 equivalent of B(C6F5)3, in hexanes, gave a red-orange precipitate, 3.Elemental analysis data for 3 is consistent with the empirical formula [MoO(O2){B(C6F5)3}{h2-PhN(O)C(O)- Ph}2]. The compound exhibits a signal at d 3.8 in its 11B NMR spectrum indicative of a four-co-ordinate boron atom whilst the 1H NMR spectrum shows several resonances in the phenyl region, shifted from those of the starting complex.These data suggest that the Lewis acid is bound to either the oxo or peroxo functionality (Fig. 2) and that the organic ligands have Fig. 2 Possible structures of [MoO(O2){B(C6F5)3}{h2-PhN(O)C(O)- Ph}2] 3. Mo O O O O O O O N N Ph Ph Ph Ph Mo O O O O O O O N N Ph Ph Ph Ph B(C6F5)3 B(C6F5)3 Table 2 Selected bond distances (Å) and angles (8) for compound 1 Mo(1)–O(1) Mo(1)–O(2) Mo(1)–O(3) Mo(1)–O(4) Mo(1)–N(3) Mo(1)–N(4) O(1)–B(1) O(3)–N(3) O(4)–N(4) 1.808(1) 1.683(1) 1.936(1) 1.938(1) 2.140(1) 2.140(1) 1.510(2) 1.423(2) 1.425(2) O(1)–Mo(1)–O(2) O(1)–Mo(1)–O(3) O(1)–Mo(1)–O(4) O(2)–Mo(1)–O(3) O(2)–Mo(1)–O(4) O(3)–Mo(1)–N(3) O(4)–Mo(1)–N(4) B(1)–O(1)–Mo(1) 122.94(5) 114.32(4) 113.89(5) 107.84(5) 108.02(5) 40.47(4) 40.54(4) 170.09(8)J. Chem.Soc., Dalton Trans., 1998, 3191–3194 3193 retained their integrity. The interaction of a related oxoperoxomolybdenum complex, [MoO(O2)2L2], L = 2-(1-octylpyrazol- 3-yl)pyridine, with Brønsted and Lewis acids has been described.8 The NMR spectroscopic studies (17O and 1H) indicate that trifluoroacetic anhydride selectively attacks the peroxo ligand in the pyrazolylpyridine complex to form a mixture of isomers of the type [MoO(OR)42x(OR9)xL2] [R = C(O)CF3, R9 = OC(O)CF3, x = 0–4].In contrast, electrophilic attack of chlorotrimethylsilane occurs preferentially, but not selectively, at the peroxo ligands.8 These studies suggest that the peroxo ligands may provide the most likely site within [MoO(O2){h2- PhN(O)C(O)Ph}2] for electrophilic attack by B(C6F5)3.However, any stretches due to the Mo]] O and Mo(O2) units in the IR spectrum of 3 are masked by those of the Lewis acid and so, without structural determination or 17O labelling studies, the exact location of the B(C6F5)3 moiety cannot be confidently predicted.In order to determine the bonding mode of the Lewis acid moiety attempts were made to crystallise compound 3 from toluene solutions. However it decomposes under these conditions to give small orange crystals of [MoO{OB(C6F5)3}{h2- PhN(O)C(O)Ph}2] 4. Compound 4 can be independently synthesized in good yield by reaction of [MoO2{h2-PhN(O)- C(O)Ph}2] with B(C6F5)3 and has been fully characterised by IR and NMR spectroscopies, elemental analyses (Table 1) and a crystal structure determination.The NMR spectroscopic data for compound 4 are significantly diVerent from those of 3. The structure of compound 4 is shown in Fig. 3 and signifi- cant bond angles and distances are detailed in Table 3. The structure determination reveals the presence of a cis- MoO{OB(C6F5)3} unit with similar features to those described for compound 1; B(1)–O(1) 1.508(3), Mo(1)–O(1) 1.775(3), Mo(1)–O(2) 1.674(3) Å, and Mo(1)–O(1)–B(1) 169.07(18)8.It has been observed that the oxoperoxo-complex [Mo- O(O2){h2-PhN(O)C(O)Ph}2] will undergo gradual conversion into [MoO2{h2-PhN(O)C(O)Ph}2] and so the formation of 4 from 3 is unsurprising. In conclusion, we have demonstrated that oxo- and perhaps peroxo-functionalities in molybdenum complexes are suf- ficiently nucleophilic to form a dative interaction with B(C6F5)3. In the case of dioxomolybdenum complexes containing ancillary h2-ONR2 ligands, attack of the Lewis acid occurs preferentially at the oxometal unit and no evidence for reaction at the h2-ONR2 ligand is observed.Whilst the analogous reaction of Fig. 3 View of the structure of [MoO{OB(C6F5)3}{h2-PhN(O)C(O)- Ph}2] 4. Fluorine and hydrogen atoms omitted for clarity. B(C6F5)3 with a peroxooxomolybdenum complex may initially occur at the peroxo ligand it is followed by decomposition on standing to yield a B(C6F5)3 substituted dioxomolybdenum species. The Mo]] OB(C6F5)3 unit is reasonably stable to air and two complexes containing this motif have been crystallographically characterised.Experimental Fourier-transform 1H and 11B NMR spectra were recorded on a Bruker WM 300 spectrometer at 300 and 96 MHz respectively, 13C NMR spectra on a Bruker WM 300 spectrometer at 75.5 MHz or Varian Unity 500 spectrometer at 125.7 MHz: 1H and 13C shifts are reported with respect to d 0 for SiMe4, 11B with respect to d 0 for BF3?OEt2; all downfield shifts are positive.Infrared spectra were recorded on either a Mattson ‘Polaris’ Fourier-transform, Perkin-Elmer FT 1710 spectrophotometer, or Perkin-Elmer 457 grating spectrometers. Microanalyses were obtained from the microanalytical department of this department. All reactions were carried out under nitrogen using standard Schlenk techniques. Solvents were dried over suitable reagents and freshly distilled under N2 before use. The compounds HONEt2, HON(CH2Ph)2, [MoO2(acac)2] were used as received (Aldrich); [MoO2(h2-ONEt2)2],4 [MoO2{h2-PhN(O)C(O)Ph}2],7 [MoO(O2){h2-PhN(O)C(O)Ph}2] 7 and B(C6F5)3 9 were prepared as previously described.Preparations cis-[MoO{OB(C6F5)3}(Á2-ONEt2)2] 1. White [MoO2(h2- ONEt2)2] (0.304 g, 1 mmol) was partially dissolved in toluene (20 cm3) and a toluene solution (20 cm3) of B(C6F5)3 (0.512 g, 1 mmol) added. The mixture was stirred for 4 h during which time a yellow solution formed. After removal of solvent in vacuo the residue was washed with pentane and the desired product then extracted with toluene.This solution was concentrated and cooled to 220 8C resulting in the formation of colourless crystals of compound 1. Yield: 0.63 g, 77%. Alternative preparation of cis-[MoO2{Á2-ON(CH2Ph)2}2]. Orange [MoO2(acac)2] (1.22 g, 3.74 mmol) was suspended in CH2Cl2 and HON(CH2Ph)2 (1.22 g, 3.74 mmol) dissolved in CH2Cl2 (50 cm3) added. Ethanol (50 cm3) was added and the reaction mixture stirred for 1 h until an oV-white precipitate had formed.The solvent was removed under vacuum and the residue washed with Et2O to remove any unchanged hydroxylamine. The residue was extracted with CH2Cl2. Concentration and cooling to 220 8C resulted in the formation of colourless crystals. Yield: 0.57 g, 74%. cis-[MoO{OB(C6F5)3}{Á2-ON(CH2Ph)2}2] 2. White [MoO2- {h2-ON(CH2Ph)2}2] (0.552 g, 1 mmol) was suspended in CH2Cl2 (20 cm3) and a CH2Cl2 solution (20 cm3) of B(C6F5)3 (512 mg, 1 mmol) added. Over about 15 min the solid dissolved and a very pale yellow solution formed.After stirring for 1.5 h in total the solvent was removed in vacuo and the residue washed with hexane. The residue was extracted with toluene and the solution filtered oV. Both the hexane and toluene filtrates were separately concentrated and cooled to 220 8C leading to the formation of colourless crystals. Combined yield: 0.92 g, 86%. Table 3 Selected bond distances (Å) and angles (8) for compound 4 Mo(1)–O(1) Mo(1)–O(2) Mo(1)–O(3) Mo(1)–O(4) Mo(1)–O(5) Mo(1)–O(6) O(1)–B(1) 1.775(3) 1.674(3) 1.982(2) 1.979(2) 2.088(3) 2.164(3) 1.508(3) O(1)–Mo(1)–O(2) O(1)–Mo(1)–O(3) O(1)–Mo(1)–O(4) O(2)–Mo(1)–O(3) O(2)–Mo(1)–O(4) O(2)–Mo(1)–O(5) B(1)–O(1)–Mo(1) 103.68(13) 86.22(11) 105.95(11) 106.33(12) 87.47(12) 91.06(13) 169.07(18)3194 J.Chem. Soc., Dalton Trans., 1998, 3191–3194 [MoO(O2){B(C6F5)3}{Á2-PhN(O)C(O)Ph}2] 3. Yellow [Mo- O(O2){h2-PhN(O)C(O)Ph}2] (0.568 g, 1 mmol) was suspended in hexane (20 cm3) and a hexane solution (20 cm3) of B(C6F5)3 (512 mg, 1 mmol) added.There was an immediate change to red-orange and the reaction stirred for 30 min. The pale yellow filtrate was removed and the red-orange precipitate washed with hexane (3 × 10 cm3) and then dried in vacuo. cis-[MoO{OB(C6F5)3}{Á2-PhN(O)C(O)Ph}2] 4. OV-white [MoO2{h2-PhN(O)C(O)Ph}2] (0.552 g, 1 mmol) was suspended in toluene (20 cm3) and a toluene solution (20 cm3) of B(C6F5)3 (512 mg, 1 mmol) added.There was an immediate change to orange and after 1 h all the solid had dissolved. The solvent was removed in vacuo and the residue washed with hexane. The residue was extracted with toluene and the filtrate concentrated and layered with pentane leading to the formation of orange microcrystals. Yield: 0.87 g, 82%. Crystal structure determination of compounds 1 and 4 Crystals of compound 1 were grown from toluene solution at 253 K and of 4 from toluene layered with pentane at 298 K.In each case a crystal from the mother-liquid was immersed in highly viscous perfluoropolyether to exclude oxygen and prevent solvent loss. It was mounted on a glass fibre and plunged into a cold (150 K) nitrogen stream. Crystal data. Compound 1, C26H20BF15MoN2O4?0.5C7H8, M = 816.21 1 46.04, triclinic, space group P1� , a = 10.764(1), b = 12.107(1), c = 12.563(1) Å, a = 86.673(2), b = 85.919(2), g = 86.480(2)8, V = 1627.6 Å3, Z = 2, Dc = 1.76 g cm23, m = 5.138 cm21, colourless, crystal dimensions 0.23 × 0.31 × 0.18 mm.Compound 4, C44H20BF15MoN2O6?0.5C6H12, M = 1094.45, triclinic, space group P1� , a = 10.2840(8), b = 12.4090(8), c = 18.598(2) Å, a = 102.500(5), b = 98.190(4), g = 106.397(4)8, V = 2169.74 Å3, Z = 2, Dc = 1.68 g cm23, m = 4.14 cm21, yellow block, crystal dimensions 0.25 × 0.25 × 0.10 mm. Data collection and processing. The data for compounds 1 and 4 were collected at 150 and 100 K respectively on an Enraf- Nonius DIP2000 image plate diVractometer with graphitemonochromated Mo-Ka radiation (l = 0.71069 Å).For compound 1 19362 reflections were measured (1 < q < 268, 213 < h < 13, 215 < k < 15, 215 < l < 15). 6304 Unique reflections were obtained giving 5950 reflections with I > 3s(I). For compound 4 4580 reflections were measured (2 < q < 258, 0<h<12, 213 < k < 13, 221 < l < 20). 4580 Unique reflections were obtained giving 4201 reflections with I > 3s(I).The images were processed with the DENZO and SCALEPACK programs.10 Corrections for Lorentz-polarisation eVects were performed but not for absorption. Structure solution and refinement. The crystal structures were solved by direct methods and refined by the full-matrix leastsquares method. Compound 1 crystallised with toluene in a 1 : 0.5 ratio. The toluene molecules are disordered at the crystallographic inversion centre with a translation of about 1.4 Å along their molecular twofold axis. All non-hydrogen atoms of 1 were refined with anisotropic displacement parameters. All hydrogen atoms of the molybdenum compound could be located in Fourier-diVerence maps and were refined isotropically.The hydrogen atoms of the disordered toluene were added geometrically and included in the final refinement with fixed positional and thermal parameters. For compound 1, 548 refined parameters and 5960 observations resulted in an observation/refined parameter ratio of 10.9 : 1.Corrections for secondary extinction were applied and refinement completed using a Chebyshev weighting scheme 11 with parameters 1.67, 0.875, 1.28. Refinement on F converged at R = 0.025, R9 = 0.031 and goodness of fit = 1.07. A final Fourier-diVerence synthesis showed minimum and maximum residual electron densities of 20.46 and 0.37 e Å23. Compound 4 was crystallised from toluene solution layered with pentane. One molecule of pentane is incorporated in the unit cell with the central carbon atom, C(101), lying on the centre of inversion such that one half of a pentane molecule is in the asymmetric unit. Hydrogen atoms were generated geometrically and allowed to ride on the corresponding carbon atoms.For 4, 646 refined parameters and 4201 observations resulted in an observation/refined parameter ratio of 6.50 : 1. Corrections for secondary extinction were applied and refinement completed using a Chebyshev weighting scheme11 with parameters 1.30, 0.078, 0.968.Refinement on F converged at R = 0.0465, R9 = 0.0455 and goodness of fit = 1.1318. A final Fourier-diVerence synthesis showed minimum and maximum residual electron densities of 20.62 and 0.75 e Å23. All crystallographic calculations were carried out using the CRYSTALS program package.12 Neutral atom scattering factors were taken from ref. 13. CCDC reference number 186/1105. See http://www.rsc.org/suppdata/dt/1998/3191/ for crystallographic files in .cif format.Acknowledgements We thank the University of Oxford for a Violette and Samuel Glasstone Fellowship (J. R. G.), the Deutsche Gemeinschaft Forschung (M. M.), St. John’s College, Oxford (L. H. D.) and the EPSRC for support of this work. References 1 For reviews concerning transition-metal catalysis of epoxidation reactions see: J. E. Lyons, in Aspects of Homogeneous Catalysis, ed. R. Ugo, Reidel, Dordrecht, Boston, 1977, vol. 3, ch. 1 and refs. therein; R.A. Sheldon and J. K. Kochi, Metal Catalyzed Oxidations of Organic Compounds, Academic Press, New York, 1981. 2 (a) S. Chan-Cheng, J. W. Reed and E. S. Gould, Inorg. Chem., 1973, 12, 337; (b) R. A. Sheldon, Recl. Trav. Chim. Pays-Bas, 1973, 92, 253, 367; (c) R. A. Sheldon and J. A. Van Doorn, J. Catal., 1973, 31, 427; (d) M. N. Sheng and J. G. Zajacek, Adv. Chem. Ser., 1968, 76, 418; (e) M. N. Sheng and J. G. Zajacek, J. Org. Chem., 1970, 35, 1839; ( f ) T. N. Baker, G.J. Mains, M. N. Sheng and J. G. Zajacek, J. Org. Chem., 1973, 38, 1145; ( g) G. R. Howe and R. R. Hiatt, J. Org. Chem., 1971, 36, 2493; (h) J. Sobczak and J. Ziolkowski, Inorg. Chim. Acta, 1976, 19, 15. 3 (a) K. Wieghardt, W. Holzbach, J. Weiss, B. Nuber and B. Prikner, Angew. Chem., Int. Ed. Engl., 1979, 18, 548; (b) P. Jaitner, W. Huber, A. Gieren and H. Betz, Z. Anorg. Allg. Chem., 1986, 538, 53; (c) K. Wieghardt, W. Holzbach, E. Hofer and J. Weiss, Inorg. Chem., 1981, 20, 343; (d) C. Redshaw, G. Wilkinson, B. Hussain-Bates and M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1992, 555; (e) K. Wieghardt, E. Hofer, W. Holzbach, B. Nuber and J. Weiss, Inorg. Chem., 1980, 19, 2927. 4 L. Saussine, H. Mimoun, A. Mitschler and J. Fisher, New J. Chem., 1980, 4, 235. 5 J. Fischer, J. Kress, J. A. Osborn, L. Ricard and M. Wesolek, Polyhedron, 1987, 6, 1839; J. Kress, M. Wesolek, J.-P. Le Ny and J. A. Osborn, J. Chem. Soc., Chem. Commun., 1982, 514. 6 J. R. Galsworthy, M. L. H. Green, M. Müller and K. Prout, J. Chem. Soc., Dalton Trans., 1997, 1308; J. R. Galsworthy, J. C. Green, M. L. H. Green and M. Müller, J. Chem. Soc., Dalton Trans., 1998, 15. 7 H. Tomioka, K. Takai, K. Oshima and H. Nozaki, Tetrahedron Lett., 1980, 21, 4843. 8 W. R. Thiel, Chem. Ber., 1996, 129, 575. 9 A. N. Chernega, A. J. Graham, M. L. H. Green, J. Haggitt, J. Lloyd, C. P. Mehnert, N, Metzler and J. Souter, J. Chem. Soc., Dalton Trans., 1997, 2293. 10 D. Gewirth, The HKL Manual, written with the co-operation of the program authors, Z. Otwinowski and W. Minor, Yale University, 1995. 11 E. Prince, Mathematical Techniques in Crystallography and Material Sciences, Springer, New York, 1982. 12 D. J. Watkin, C. K. Prout, J. R. Carruthers and P. W. Betteridge, CRYSTALS, Issue 10, Chemical Crystallography Laboratory, University of Oxford, 1996. 13 International Tables for Crystallphy, Kluwer, Dordrecht, 1992, vol. C. Paper 8/03126F

ISSN:1477-9226    

DOI:10.1039/a803126f

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3195–3198 3195 Vibrational and crystallographic studies of dioxohalogenomolybdenum(VI) complexes with crown ethers Michael J. Taylor,*a Clifton E. F. Rickard a and Lars A. Kloo b a Department of Chemistry, University of Auckland, Private Bag, 92019, Auckland, New Zealand b Inorganic Chemistry 1, Department of Chemistry, Lund University, PO Box 124, S-221 00 Lund, Sweden Received 1st May 1998, Accepted 28th July 1998 Crown ether adducts [MoO2X2(H2O)2]?2H2O?18-crown-6 (X = Br 1 or Cl 2), obtained by combining solutions of MoVI in HBr or HCl with 1,4,7,10,13,16-hexaoxacyclooctadecane (18-crown-6), were shown by X-ray crystallography to contain MoO2X2(H2O)2 units with cis,trans,cis ligand arrangement.Recrystallisation from methyl alcohol yielded [MoO2Cl2(MeOH)2]?18-crown-6 3 which was also characterised by X-ray analysis. Adding potassium ions to the HCl system gave [K(18-crown-6)][MoO2Cl3(H2O)] 4 which was identified by studying Raman and IR spectra of the series.Introduction The cis-MoO2 unit is widespread in molybdenum(VI) chemistry. For example, it occurs at the active site in the molybdenum- containing oxotransferase enzymes which catalyse oxygen atom transfer to and from biological substrates in the nitrogen, sulfur and carbon cycles. Thus it is particularly important to understand the structural chemistry and ligand exchange behaviour of the MoO2 unit in relatively simple systems.Furthermore, dioxomolybdenum complexes with halogen ligands are on the borderline of stability between the 15 and 16 states of molybdenum, as shown by the fact that the MoO2 species with bromide ligands undergoes photoreduction. Such complexes are therefore potentially useful as model systems for the investigation of this redox process. The nature of molybdenum(VI) species in aqueous solution is not well understood, especially for acidic solutions containing HCl or HBr.An object of the present work was to isolate dioxohalogenomolybdenum(VI) complexes from solution for X-ray and spectroscopic examination, in order to extend our earlier study of molybdenum(VI) solutions and solvent extracts.1 Molybdenum(VI) exists in alkaline solutions as the simple molybdate ion MoO4 22 but on acidification this is converted into polyoxomolybdate complexes which separate as white solids of variable composition. The addition of an excess of hydrochloric or hydrobromic acid causes the white precipitate to dissolve yielding solutions in which halogen-containing molybdenum(VI) anions apparently exist.Previously we obtained 95Mo NMR and Raman evidence indicating the presence of neutral [MoO2X2(H2O)2] complexes in these HCl and HBr solutions, and showed that these species are the predominant form in which molybdenum(VI) undergoes solvent extraction into an ether phase.1 The solid Cs2[MoO2Cl4] is produced when CsCl is added to the HCl solution of MoVI, although the anion concerned, [MoO2Cl4]22, is not detected in the parent solution.2 Molybdenum(VI) is eYciently extracted from its HCl or HBr solutions into an ether or ketone solvent phase, which provides a useful method of separation.3,4 In this investigation crown ethers are used to isolate the species [MoO2X2(H2O)2] (X = Br or Cl) within crystalline solids, enabling their structures to be characterised by vibrational spectroscopy and X-ray crystallography.Introduction of potassium ions to form [K(18-crown-6)]1 allows the novel species [MoO2Cl3(H2O)]2 to be isolated and identified. Results and discussion Solids were isolated from MoVI–HCl and –HBr systems by introducing crown ethers which formed air-stable crystalline adducts with the molybdenum complexes. The ethers 18- crown-6, 15-crown-5 and 12-crown-4 gave similar products, as indicated by the Mo–O and Mo–X bands in the infrared spectra. The compounds involving 18-crown-6 (1,4,7,10,13,16- hexaoxacyclooctadecane) were chosen for particular study.Three preparative methods were employed: (i) diethyl ether extraction of the MoVI–HCl or –HBr solution, followed by addition of crown ether to the separate extract phase, (ii) solvent extraction of the MoVI–HX solution using a solution of the crown ether in Et2O, (iii) direct addition of the crown ether to a solution of the molybdenum complexes in HCl or HBr. The source of MoVI was potassium molybdate or ammonium molybdate.In the case of HCl solutions methods (ii) and (iii) gave a product which incorporated K1 ions. Crystalline samples 1–4 were obtained (Experimental section) and investigated by X-ray diVraction and by their vibrational spectra. X-Ray crystallography Fig. 1 gives the structure of [MoO2Br2(H2O)2]?2H2O?18- crown-6 1, consisting of the uncharged molybdenum complex [MoO2Br2(H2O)2] accompanied by a molecule of 18-crown-6 and two extraneous water molecules.The H2O ligands take part in hydrogen bonds measuring 1.86(2) and 1.79(2) Å to the extra water molecules which form hydrogen bonds of 2.07(2) Å directed towards oxygens of the crown ether. Table 1 contains the principal interatomic distances and angles of the [MoO2Br2(H2O)2] structure which consists of the expected cis-MoO2 group, with the H2O ligands trans to the Mo]] O bonds, and a pair of mutually trans-Br atoms making up an octahedral complex. Bromide is an uncommon ligand for molybdenum(VI) as such systems are prone to photoreduction and few structures are available for comparison.5 (2,29- bipyridyl)dibromodioxomolybdenum(VI) 6 has cis-M]] O bonds of 1.643(17) and 1.826(18) Å, and trans-Mo–Br bonds of 2.461(3) and 2.781(3) Å with Br–Mo–Br angle of 159.7(1)8 which may be compared with the dimensions of the present, less distorted structure.The Mo]] O distance of 1.695 Å in 1 is slightly shorter than the average value of 1.704 Å derived from a large number of dioxomolybdenum structures.73196 J.Chem. Soc., Dalton Trans., 1998, 3195–3198 Crystals of [MoO2Cl2(H2O)2]?2H2O?18-crown-6 2 were also investigated and found to have a kindred structure in which the pattern of hydrogen bonding resembles that of complex 1. This compound contains the octahedral complex [MoO2Cl2(H2O)2] which has been encountered before in [MoO2Cl2(H2O)2]?H2O? NEt4Cl 8 and [MoO2Cl2(H2O)2]?2C5H5NHCl.9 The geometries are similar, and the average dimensions of our earlier study 8 are given in parentheses in the following comparison with the present bond lengths (Å) and angles (8): Mo]] O 1.677 (1.687), Mo–O 2.210 (2.272) and Mo–Cl 2.368 (2.346); O]] Mo]] O 103.5 (103.5), O–Mo–O 75.8 (78.5) and Cl–Mo–Cl 159.9 (157.9).Fig. 1 The structure of [MoO2Br2(H2O)2]?2H2O?18-crown-6 1, showing the atomic labelling. Table 1 Selected bond distances (Å) and angles (8) for [MoO2- Br2(H2O)2]?18-crown-6 1, [MoO2Cl2(H2O)2]?2H2O?18-crown-6 2 and [MoO2Cl2(MeOH)2]?2H2O?18-crown-6 3 Mo]] O Mo]O Mo]X* O]] Mo]] O O]Mo]O X]Mo]X* 1 1.695(2) 2.204(2) 2.5310(3) 103.2(2) 76.1(1) 161.47(2) 2 1.677(3) 2.210(3) 2.3682(9) 103.5(3) 75.8(2) 159.90(5) 3 1.696(3) 2.240(3) 2.3730(10) 102.8(2) 76.9(1) 160.81(6) * X = Br in complex 1 or Cl in 2 and 3.Previously the distances to the pair of H2O ligands diVered significantly, but in the present instance these bonds are equal. Recrystallisation of complex 2 from methanol results in the replacement of both the co-ordinated H2O ligands by MeOH molecules, forming the compound [MoO2Cl2(MeOH)2]? 18-crown-6 3 which was also investigated by X-ray crystallography.The structure reveals that the extra water molecules (those linked to the crown ether in the structure of 2) have been lost. The pair of cis-MeOH ligands of 3 form Mo–O bonds of 2.240(3) Å which is significantly longer than the Mo–O bond distance of 2.210(3) Å to the H2O ligands of 2. The C–O bond length of the co-ordinated methanol molecule is 1.467(5) Å and the bond angle Mo–O–C measures 133.9(2)8. Other bond distances and angles are given in Table 1 where it will be noted that the Mo]] O and Mo–Cl bonds of 3 are slightly longer than those of 2.The O–H bond of each methanol ligand measures 0.73(5) Å. These hydrogens are involved in hydrogen bonds of 1.97(5) Å to oxygen atoms of the 18-crown-6 molecule. The angle O– H? ? ? O is 167(5)8. Vibrational spectroscopy Table 2 compares the vibrational data for complex 1 with those of 2 and 4.Bands due to the 18-crown-6 component do not interfere with those of the molybdenum species and are omitted. Strong bands from the cis-MoO2 unit are characteristic, and are accompanied by a number of other bands, including those of the Mo–Br or Mo–Cl stretching modes, assigned in Table 2. The spectrum of 2 gives frequencies for [MoO2Cl2- (H2O)2] closely matching those of [MoO2Cl2(H2O)2]?H2O? NEt4Cl where Et4N1 and Cl2 ions accompany the diaquadichlorodioxomolybdenum( VI) complex.8 The present spectra of [MoO2X2(H2O)2] (X = Br 1 or Cl 2) closely resemble those given by acidic molybdenum(VI) solutions and solvent extracts,1 supporting the earlier conclusion these are the principal complexes involved.The 95Mo NMR spectra of the molybdenum(VI) solutions have also been attributed to [MoO2X2(H2O)2],1 although the presence of other halogeno-complexes was not ruled out. Thus complexes such as [MoO2X3(H2O)]2 and [MoO2X4]22 may coexist with [MoO2- X2(H2O)2].In species such as these, bonds opposed to the oxo-groups are expected to be weak because of strong trans influence, allowing facile exchange of the ligands (water molecules or halide ions). [K(18-crown-6)][MoO2Cl3(H2O)]. Initially we thought that the molybdenum(VI) compound from HCl solution might be of Table 2 Vibrational spectral data (cm21) and assignments for crystalline solids [MoO2X2(H2O)2] (X = Br 1 or Cl 2) and [MoO2Cl3(H2O)]2 * 1 2 4 Raman 948vs 914m 372m 247m 205vs 185vs 103m 87m IR 957s 948s 907s 425m 372m 248vs 205w 180w 118m Raman 951vs 919m 435vw 378m 318s 262s 146m 92s IR 957s 951s 909vs 433m 376w 332s 260m 146w 125m Raman 946vs 896s 391s 327m 300m 253s 146w 113m IR 957s 944s 6 894vs 390m 329s 320 (sh) 251s 147m Vibration nsym(MoO2) nasym(MoO2) nsym(Mo–OH2) nasym(Mo–OH2) nasym(Mo–Cl) nasym(Mo–Cl) nsym(Mo–Cl) r(MoO2) nasym(Mo–Br) nsym(Mo–Br) d(MoCl2) d(Cl–Mo–OH2) d(MoBr2) d(Br–Mo–OH2) * The spectrum of complex 3 is almost identical to that of 2 except for an additional IR band, 1002w cm21, due to MeOH.J.Chem. Soc., Dalton Trans., 1998, 3195–3198 3197 Table 3 Data collection and processing parameters for complexes 1–3 Formula M Crystal system Space group a/Å b/Å c/Å b/8 V/Å3 Z m/mm21 T/K Reflections collected/unique (Rint) No. observed reflections [I > 2s(I)] Goodness of fit on F2 R (observed data) wR2 (all data) 1 C12H32Br2MoO12 624.14 Monoclinic C2/c 8.6690(1) 19.7867(2) 13.5329(1) 94.284(1) 2314.82(4) 4 4.069 203 10949/2567 (0.0176) 2445 1.075 0.0267 0.0632 2 C12H32Cl2MoO12 535.22 Orthorhombic Pbcn 8.3364(2) 19.2115(6) 14.3430(4) 2297.10(11) 4 0.854 203 13297/2026 (0.0316) 1619 1.073 0.0374 0.0933 3 C14H32Cl2MoO10 527.24 Monoclinic C2/c 8.7118(1) 19.0183(3) 13.6617(2) 94.949(1) 2255.08(6) 4 0.862 203 6451/2481 (0.0218) 2191 1.131 0.0456 0.1335 R = S||Fo| 2 |Fc||/S|Fo|, wR2 = [Sw(Fo 2 2 Fc 2)2]/Sw(Fo 2)2]� �� .the form [H3O(18-crown-6)]1[MoO2Cl3(H2O)]2 since ionic complexes of this kind are known for other elements.10–13 However, this possibility was ruled out when vibrational and crystallographic evidence for 2 showed it to be a molecular adduct containing [MoO2Cl2(H2O)2]. A diVerent complex, namely [K(18-crown-6)][MoO2Cl3- (H2O)] 4, was obtained when the synthetic procedures (as above) employed potassium molybdate, instead of ammonium molybdate. The IR and Raman spectra of 4 are given in Table 2 and provide firm support for the presence of the [MoO2Cl3- (H2O)]2 anion.The stretching frequencies of the cis-MoO2 group drop from 909 and 951 cm21 in the molecular complex 2 to 894 and 944 cm21 in 4, while that of the MoO2 rocking mode falls from 260 to 251 cm21. The Mo–Cl stretching frequencies of 4, 300, 320 and 329 cm21, are lower than those of 2 which displays nsym at 318 cm21 and nasym at 332 cm21. This comparison also suggests that the molybdenum(VI) species in 4 is anionic.The spectra of Cs2[MoO2Cl4] 1,2 confirm the trend to lower frequencies, with cis-MoO2 bands at 883 and 919 cm21 and Mo–Cl stretches at 308 and 325 cm21. The formation of [K(18-crown-6)][MoO2Cl3(H2O)], rather than the [MoO2Cl2(H2O)2] complex, is attributable to replacement of one of the H2O ligands by Cl2 under the trans influence of the MoO2 oxygen atoms, aided by the tendency of 18-crown- 6 to accommodate K1 as the necessary counter ion.14 Unfortunately, attempts to solve the structure of 4 by X-ray crystallography were defeated by severe disorder problems. Rather surprisingly, the addition of a crown ether to the MoVI–HBr systems in the presence of potassium ions yields crystals of the molecular [MoO2Br2(H2O)2] adduct 1, with no sign of a bromide complex analogous to 4.In like vein, eVorts to prepare Cs2[MoO2Br4], by adding CsBr to MoVI–HBr solutions, were also unsuccessful. Experimental Preparations The compounds 18-crown-6, 15-crown-5 and 12-crown-4 were obtained from Acros Organics.Molybdenum(VI) solutions were prepared using ammonium or potassium molybdate and analytical grade acids, HBr or HCl. The compound K2MoO4?5H2O was added to 6 mol dm23 HBr or 8 mol dm23 HCl to prepare solutions which were 0.5 mol dm23 in MoVI. The solution (10 ml) was shaken with an equal volume of diethyl ether and the upper, ether-extract phase was withdrawn. Crown ether (0.15 g in 1 ml Et2O) was added to 2 ml of the ether extract in which molybdenum(VI) complexes were present.Each extract gave crystals with 18- crown-6, 15-crown-5 and 12-crown-4 which were pale yellow for the bromide samples and colourless for the chlorides. The samples were collected, rinsed with ether and dried in a stream of nitrogen. The IR spectra suggested that the products were likely to be crown ether adducts of [MoO2X2(H2O)2] (X = Cl or Br). Further samples were prepared using 18-crown-6 and characterised, as follows.[MoO2Br2(H2O)2]?2H2O?18-crown-6 1. The addition of 18- crown-6 to the diethyl ether extract from a solution of potassium molybdate in 6 mol dm23 HBr gave pale yellow crystals which were rinsed with ether and dried under nitrogen, mp 115– 118 8C (Found: C, 23.1; H, 4.90. C12H32Br2MoO12 requires C, 23.1; H, 5.16%). [MoO2Cl2(H2O)2]?2H2O?18-crown-6 2. The addition of 18- crown-6 to the ether extract from a solution of potassium molybdate in 8 mol dm23 HCl gave colourless crystals.A further sample, with identical IR spectrum, was obtained when 18-crown-6 was added directly to an aqueous solution of ammonium molybdate in 8 mol dm23 HCl. The crystals were rinsed with acetone and dried under nitrogen, mp 112–114 8C (Found: C, 30.5; H, 6.52. C12H32Cl2MoO12 requires C, 26.2; H, 5.86%). The discrepancy in the analytical results may be due to adherence of some crown ether to the crystals which were identified as 2 by their IR and Raman spectra, and by X-ray crystallography.[MoO2Cl2(MeOH)2]?18-crown-6 3. A sample of complex 2 was recrystallised from hot methanol which yielded colourless crystals, mp 105 8C (decomp.) (Found: C, 29.8; H, 6.18. C14H32Cl2MoO10 requires C, 31.9; H, 6.11%). The constitution of this product was established by X-ray crystallography. K[MoO2Cl3(H2O)]?18-crown-6 4. Extraction of a solution of potassium molybdate in 8 mol dm23 HCl with diethyl ether containing 18-crown-6 led to the formation of colourless crystals with a diVerent IR spectrum from that of complex 2 (whose preparation had involved separating the ether extract from the aqueous phase before introducing the crown ether). The same product 4 was obtained by adding 18-crown-6 directly to the aqueous solution of potassium molybdate in 8 mol dm23 HCl, mp 228–230 8C (Found: C, 26.1; H, 5.35.C12H26Cl3KMoO9 requires C, 25.9; H, 4.71%). Spectroscopy Infrared spectra were recorded as pressed discs in Polythene or KBr, using Bio-Rad FTS6000 or Perkin-Elmer Paragon 1000PC spectromRaman spectra were obtained with a Bio-Rad FT Raman spectrometer, which gave data above 703198 J. Chem. Soc., Dalton Trans., 1998, 3195–3198 cm21, supplemented by scanning the range 10–1200 cm21 using a Jobin–Yvert U1000 system with a Spectra Physics 2016 argonion laser tuned to the green line of wavelength 514 nm. Crystal structure determinations Crystallographic data for complexes 1 to 3 are summarised in Table 3.Suitable crystals were mounted in Paratone oil on glass fibres and frozen to 270 8C for data collection on a Siemens SMART diVractometer. Data collection covered a nominal hemisphere of reciprocal space. Lorentz-polarisation corrections were applied and absorption corrections using the program SADABS.15 The structures were solved by Patterson and Fourier techniques and refined by full matrix least squares. Anisotropic thermal parameters were used for all non-H atoms.Hydrogen atoms of 18-crown-6 were placed in calculated positions and refined as riding atoms with isotropic thermal parameters fixed at 20% greater than those of the host atom. Structures were determined using SHELXS16 for the solution and SHELXL17 for refinement. CCDC reference number 186/1104. See http://www.rsc.org/suppdata/dt/1998/3195/ for crystallographic files in .cif format. Acknowledgements We are grateful for financial aid from the University of Auckland Research Committee, and also thank the Swedish Research Council for support. References 1 J. M. Coddington and M. J. Taylor, J. Chem. Soc., Dalton Trans., 1990, 41. 2 W. P. GriYth, J. Chem. Soc. A, 1969, 211. 3 G. H. Morrison and H. Freiser, Solvent Extraction in Analytical Chemistry, Wiley, New York, 1957. 4 I. Nelidow and R. M. Diamond, J. Phys. Chem., 1955, 59, 710. 5 E. J. Steifel, in Comprehensive Coordination Chemistry, eds. G. Wilkinson, R. D. Gillard and J. A. McClevery, Pergamon, Oxford, 1987, vol. 3, p. 1375. 6 R. H. Fenn, J. Chem. Soc. A, 1969, 1764. 7 J. M. Mayer, Inorg. Chem., 1988, 27, 3899. 8 M. J. Taylor, Wang Jirong and C. E. F. Rickard, Polyhedron, 1993, 12, 1433. 9 B. Kamenar and M. Penavic, Acta Crystallogr., Sect. B, 1976, 32, 3323. 10 J.-P. Behr, P. Dumas and D. Moras, J. Am. Chem. Soc., 1982, 104, 4540. 11 G. S. Heo and R. A. Bartsch, J. Org. Chem., 1982, 47, 3557. 12 R. Chénevert, A. Rodrigue, D. Chamberland, J. Ouellet and R. Savoie, J. Mol. Struct., 1985, 131, 187. 13 R. Chénevert, D. Chamberland, M. Simard and F. Brisse, Can. J. Chem., 1989, 67, 32. 14 C. J. Pedersen, J. Am. Chem. Soc., 1967, 89, 7017. 15 G. M. Sheldrick, SADABS, University of Göttingen, 1996. 16 G. M. Sheldrick, SHELXS 97, University of Göttingen, 1997. 17 G. M. Sheldrick, SHELXL 97, University of Göttingen, 1997. Paper 8/03295E

ISSN:1477-9226    

DOI:10.1039/a803295e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, 3199–3208 3199 The first alkynethiolate derivatives of bis(substituted cyclopentadienyl) titanium(IV) and their role in the synthesis of heterobimetallic compounds. Crystal structures of [Ti(Á5-C5H4SiMe3)2(SC]] CBut)2] and [(Á5-C5H4SiMe3)(SC]] CBut)Ti(Ï-Á5 :Í-P-C5H4PPh2)(Ï-SC]] CBut)- Pt(C6F5)2] † Irene Ara,a Esther Delgado,*b Juan Forniés,a Elisa Hernández, Elena Lalinde,*c Noelia Mansilla b and M.Teresa Moreno c a Departamento de Química Inorgánica, Instituto de Ciencia de Materiales de Aragón, Universidad de Zaragoza-C.S.I.C., 50009 Zaragoza, Spain b Departamento de Química Inorgánica, Facultad de Ciencias, Universidad Autónoma de Madrid, 28049 Madrid, Spain. E-mail: esther.delgado@aumam.es c Departamento de Química, Universidad de la Rioja, 26001 Logroño, Spain Received 20th May 1998, Accepted 13th July 1998 The first thioalkyne derivatives of functionalised titanocene of formula [Ti(h5-C5H4R9)(h5-C5H4R0)(SC]] ] CR)2] (R = But, R9 = R0 = SiMe3 1a; R = Ph, R9 = R0 = SiMe3 1b; R = But, R9 = SiMe3, R0 = PPh2 2a; R = But, R9 = R0 = PPh2 3a) have been prepared by reaction of [Ti(h5-C5H4R9)(h5-C5H4R0)Cl2] and LiSC]] ] CR in diethyl ether.Complexes 1a and 2a have been used as precursors in the synthesis of Ti–M (M = d6 or d8 metal) heteronuclear complexes showing diVerent co-ordination modes. All compounds have been characterised by elemental analysis and 1H, 31P, 19F and 13C NMR and infrared spectroscopy. The crystal structures of two complexes have been solved.Introduction The synthesis and study of early–late heterobimetallic compounds is an active subject of research in organometallic chemistry.1 One of the reasons for this interest is related to some catalytic processes in view of the potential of this type of compound to promote activation of small molecules (e.g. CO).2 Owing to the propensity of sulfur to form M(m-SR)M9 bridges, an appropriate synthetic pathway to such species consists on the use of thiolate derivatives of group 4 metallocenes as metalloligands.Stephan and co-workers 3 have made an important contribution in this area by using diVerent thiolate derivatives of titanocene in their reactions with d10 transition metal species. In the last years we have studied the reactions between d6 and d8 metal fragments and [Ti(h5-C5H4R)2(SR9)2] (R = H, SiMe3 or PPh2; R9 = aryl or alkyl group), yielding bi- and tri-nuclear compounds stabilised by double homo (m-h5 :k-P-C5H4PPh2)2 and (m-SR)2 or hetero (m-h5 :k-PC5H4PPh2)( m-SR) bridging systems.4 On the other hand, the ability of alkynyl ligands to bind several metal centres through s and p bonds is now firmly established.5 In particular in this area we and others have also reported the synthesis of diVerent early–late binuclear doubly alkynyl bridged complexes.6,8 These complexes have been studied in order to gain understanding of the factors that govern the preferred geometries of the C]] ] C groups because of their relevance in C–C coupling alkynide processes,7 as well as C–C bond cleavage on butadiynes.8 By contrast with the amount of work devoted to thiolate and alkynide bridged heterobimetallics and their mononuclear precursors, reports on related alkynethiolates are exceedingly rare.Interestingly the few examples that have been published show a quite versatile co-ordination behaviour (Scheme 1). For † Dedicated to Professor Pascual Royo on the occasion of his 60th birthday.instance, Weigand et al. have reported 9 not only the syntheses of several alkyne thiolate mononuclear complexes of RuII and PtII with these ligands acting as h1-(S) bonded ligands (M–SC]] ] CR9), but also the ability of the phenylalkynethiolate to act as an h1-(C) bonded thioketenyl, [Ru]]] C(Ph)–C]] S, terminal group.9a Recently, the co-ordination as an alkyne thioketenyl h2-(C,C) with the ligand acting as a three electron donor has been also demonstrated,10 but, as far as we are aware, only a diiron carbonyl complex [Fe2(CO)6(m-C]] ] CPh)(m-SC]] ] CPh)] containing a sulfur alkynethiolate bridging group m-(S,S) has been reported.11 In the context of these groups it should be noted Scheme 1 M S C C R M S C C Ph M S C C Ph h1-( S) h1-( C) h2-( C, C) M C S C M' m-( S,S) R M = Pt, Ru Ru Mo, W Fe3200 J.Chem. Soc., Dalton Trans., 1998, 3199–3208 that some additional work has been developed with the isomeric thioacetylide ligands C]] ] CSR.12 In this paper we report on the preparation and properties of several mononuclear alkynethiolate titanocene complexes [Ti(C5H4R9)(C5H4R0)(SC]] ] CR)2] 1–3 and describe their reactivity towards several d6 [Mo(CO)4(nbd)] and [Mo(CO)3- (NCMe)3] and d8 cis-[M(C6F5)2(thf)2] (M = Pt or Pd) metal complexes containing labile ligands.The syntheses of homo bis(m-alkynethiolate) 4a–6b and hetero bis(m-alkynethiolate, m-cyclopentadienyldiphenylphosphine) bridged derivatives 7, 8, 9 and the solid-state structures of [Ti(h5-C5H4SiMe3)2(SC]] ] CBut)2] 1a and [(h5-C5H4SiMe3)(SC]] ] CBut)Ti(m-h5 :k-P-C5H4- PPh2)(m-SC]] ] CBut)Pt(C6F5)2] 8 are presented.Results and discussion Mononuclear derivatives The formation of metallocene alkynethiolate titanium(IV) derivatives [Ti(h5-C5H4R9)(h5-C5H4R0)(SC]] ] CR)2] 1a–3a was accomplished by treatment of [Ti(h5-C5H4R9)(h5-C5H4R0)Cl2] (R9 = R0 = SiMe3; R9 = SiMe3, R0 = PPh2, R9 = R0 = PPh2) with lithium alkynethiolate reagents LiSC]] ] CR13 (2 equivalents) at very low temperature (270 8C) in diethyl ether [eqn. (1)].After conventional work-up complexes 1–3 were isolated as green microcrystalline solids and their spectroscopic (IR, 1H, 13C and 31P NMR) and analytical data unequivocally confirm the structural proposal shown in eqn. (1) with the alkynethiolate ligands h1-(S) bonded. Further confirmation was obtained from the X-ray diVraction study of compound 1a.It should be noted that initial attempts to carry out the former reaction at room temperature, following similar reaction conditions to those reported for ruthenium(II) and platinum(II) complexes,9 failed to yield the alkynethiolate derivatives. The substitution of SiMe3 by PPh2 groups on the cyclopentadienyl rings reduces considerably the stability of these systems.Thus, whereas complexes 1a,1b and 2a show satisfactory elemental analysis, the instability of 3a in solution and in the solid state precludes a good analysis. In the same line we have previously shown that the stability of mixed [Ti(h5-C5- H4SiMe3)(h5-C5H4PPh2)X2] (X = Cl or SPh) derivatives is considerably higher than that of analogous [Ti(h5-C5H4PPh2)2X2].14 The most noticeable fact in the IR spectra of complexes 1–3 is the presence of a weak absorption in the 2129–2145 cm21 region corresponding to the C]] ] C stretching mode, clearly indicating that the acetylenic fragments are not involved in co-ordination.Their NMR data (1H and 13C) indicate that only one of the two expected isomers (syn or anti) is present in solution (see Experimental section).In the 1H NMR spectra the resonances due to cyclopentadienyl protons, two for 1a, 1b and 3a (d 6.40–6.62, 6.01–6.53) and four for complex 2a (6.54, 6.38, 6.34, 6.11) due to the presence of two diVerent substituted rings, are shifted upfield in relation to the dichloride starting precursors. This eVect can be accounted for the lowering in the electronegativity on going from the chloride to the alkynethiolate ligand.As expected, singlet signals are observed for the But or SiMe3 groups in all complexes. The presence of these groups is also confirmed by their characteristic 13C NMR Ti Cl Cl R' R'' RC CSLi Ti R' R'' 1a SiMe3 SiMe3 But 1b SiMe3 SiMe3 Ph 2a SiMe3 PPh2 But 3a PPh2 PPh2 But + -70 oC Et2O R' R" R SC CR SC CR (1) 2 resonances which appear in the expected range.Particularly evident are the acetylenic carbon resonances (d 80.8, 117.5 1a; 93.0, 107.3 1b; 80.2, 117.8 2a) which occur in a similar region to that previously reported for other h1-(S) bonded alkynethiolate (M–SC]] ] CR)9 or alkynyl (M–C]] ] CR)6,8 compounds. Complexes 1a and 1b show only three cyclopentadienyl carbon resonances while the mixed derivative [Ti(h5-C5H4SiMe3)- (h5-C5H4PPh2)(SC]] ] CBut)2] 2a exhibits five resonances for each substituted C5H4 ring suggesting that the five carbon atoms are inequivalent probably due to molecular steric strains.The shielding of the 31P resonances displayed by complexes 2a (d 215.2) and 3a (d 215.5) is typical of this type of compound.4a Crystal structure of [Ti(Á5-C5H4SiMe3)2(SC]] ] CBut)2] 1a This compound crystallises with two crystallographically independent molecules, which have essentially the same structure, in the asymmetric unit.Discussion will therefore be limited to only one of them. The monomeric structure of 1a is shown in Fig. 1 and selected bond distances and angles are listed in Table 1. The compound shows a distorted tetrahedral arrangement around the titanium atom made up of the two centroids of trimethylsilylcyclopentadienyl rings, which adopt a staggered disposition, and the two thiolate ligands.The S(1)–Ti(1)–S(2) angle of 92.30(13)8 as well as the Ti(1)–S(1,2) [2.451(4) Å], Ti(1)–centroid(1) [2.050(2) Å] and Ti(1)–centroid(2) [2.038(3) Å] distances are in the range reported for analogous compounds [Ti{h5-C5H4P(S)Ph2}2- (SPh)2],14b [Ti(h5-C5H4SiMe3)2(SC6F5)2] 14a and [Ti(h5-C5H5)2- (SMe)2].3e Once again the endo (anti) conformation shown by this titanium(IV) derivative confirms the relationship between the type of isomer and the S(1)–Ti–S(2) angle.The bond lengths S–C [1.688(11), 1.712(12) Å] and C]] ] C [1.174(13), 1.144(14) Å] and angles Ti–S–C [107.4(7), 114.2(3)], S–C–C [177.5(12), 174.0(12)] and C–C–C [173.3(13), 169(2)] found Fig. 1 View of molecular structure of [Ti(h5-C5H4SiMe3)2(SC]] ] CBut)2] 1a. Table 1 Selected bond lengths (Å) and angles (8) of complex 1a (molecule 1) Ti(1)–S(1) Ti(1)–S(2) S(1)–C(17) S(2)–C(23) S(2)–Ti(1)–S(1) cent(1)–Ti(1)–cent(2) S(2)–C(23)–C(24) Ti(1)–S(1)–C(17) 2.451(4) 2.451(4) 1.688(11) 1.712(12) 92.30(13) 130.9 174.0(13) 107.7(4) C(17)–C(18) C(23)–C(24) Ti(1)–cent(1) Ti(1)–cent(2) S(1)–C(17)–C(18) C(17)–C(18)–C(19) C(23)–C(24)–C(25) Ti(1)–S(2)–C(23) 1.174(13) 1.144(14) 2.050 2.038 177.5(12) 173.3(13) 169(2) 114.2(3)J.Chem. Soc., Dalton Trans., 1998, 3199–3208 3201 within the alkynethiolate fragments show no unusual features, being quite similar to those found in complexes [Pt(PPh3)2- {SC]] ] CC(Me)}2] 9b and [Fe2(CO)6(m-C]] ] CPh)(m-SC]] ] Ph)] 11 which to our knowledge are the only examples of thioalkyne derivatives of transition metals structurally characterised.Heterobinuclear derivatives We have previously shown that titanocene thiolate derivatives [Ti(h5-C5H4R9)2(SR)2] (R9 = H, SiMe3 or PPh2) can act as either bi- [R9 = H or SiMe3(S,S), PPh2(P,P)] or tetra-dentate [R9 = PPh2, bis(P,S) or P,P; S,S] ligands towards the d6 Mo(CO)4 and d8 M(C6F5)2 (M = Pd or Pt) metal fragments.4 The substitution of arene- or alkene-thiolates by alkynethiolates on the mononuclear titanocene supplies an additional co-ordination position.We have reported several examples illustrating the ability of bis(alkynyl) transition metal complexes [M9Ln- (C]] ] CR)2] (M9 = Pt,15a–d Ir 15e or Ti 6a) to bond “cis-M(C6F5)2” (M = Pt or Pd) metal fragments through h2-acetylenic bonding interactions.Therefore, we considered it of interest to explore the reactivity of the novel bis(alkynethiolate) derivatives 1–3 towards the same substrates: [Mo(CO)4(nbd)] and cis- [M(C6F5)2(thf)2] (M = Pt or Pd, thf = tetrahydrofuran), respectively.The results of this study are summarised in Scheme 2. Treatment of [Ti(h5-C5H4SiMe3)2(SC]] ] CR)2] with either [Mo(CO)4(nbd)] (excess) or cis-[M(C6F5)2(thf)2] (1 equivalent) in toluene at room temperature (for 1a and M = Pt in CH2Cl2) results in the formation of neutral bis(thiolato)bridged heterobinuclear complexes [(h5-C5H4SiMe3)2Ti(m-SC]] ] CR)2- MLn] [ MLn = Mo(CO)4 4a, Pt(C6F5)2 5a, 5b, or Pd(C6F5)2 6a, 6b] in moderate to high yield (60% 4a–88% 6b).Complex 4a is isolated as a green solid after chromatographic purification. Complexes 5a (orange) and 6b (red-garnet) are precipitated as solids by treatment of the residues with n-heptane and n-hexane respectively, while 5b and 6a can be isolated as orange solids only by removing the solvent. These latter compounds are extremely soluble even in hydrocarbon solvents such as nhexane, pentane or n-heptane.In spite of many attempts we have not been able to obtain suitable crystals for X-ray analysis of any of these dinuclear compounds 4–6, however their spectroscopic data are consistent with the S,S co-ordination mode of the difunctional metallocene [Ti](SC]] ] CR)2 chelating ligands.Thus, their IR spectra show a medium n(C]] ] C) absorption in the characteristic region of non-co-ordinated alkynes.13b Compared with the precursors (1a 2129 cm21 and 1b 2134 cm21) the stretching frequency n(C]] ] C) for 5 and 6 is shifted to higher wavenumbers (2168 5a, 2165, 5b, 6b; 2166 cm21 6a) suggesting that co-ordination of the sulfur lone pair to platinum or palladium probably reduces sulfur p-donor interactions with the acetylenic fragment.In marked contrast the solution IR spectrum of the Mo–Ti complex 4a shows the n(C]] ] C) at 2072 cm21. The relative lowering of n(C]] ] C) (ª57 cm21) is considerably smaller than those previously reported for coordinated thioalkynes,16 i.e. [S{(h2-C]] ] CPh)Co2(CO)6}2] 16a 1592 vs. S(C]] ] CPh)2 2180 cm21 and [{Cu(O3SCF3)2S(C]] ] CBut)2] 16b 1988 vs.S(C]] ] CBut)2 2200 cm21, suggesting that acetylenic fragments are not co-ordinated to Mo. The NMR data are consistent with the presence of the dinuclear species in the two isomeric forms syn and anti shown in Scheme 2. This structural feature, which arises from the relative orientation of the alkyne groups on the sulfur atoms, is not unusual and in many cases equilibrium studies find the two conformers to be of similar thermodynamic stability.In fact, a few [(h5-C5H5)2Ti(m-SR)2- Mo(CO)4] compounds have been reported as endo (anti/syn) stereochemically non-rigid mixtures in solution.17 We have previously found that the heterobimetallic Ti–Pt and Ti–Pd [(h5-C5H4R9)2Ti(m-SR)2M(C6F5)2] (M = Pt or Pd, R9 = H or SiMe3, R = Ph or C6F5) systems adopt both in the solid state (X-ray; M = Pd, R9 = SiMe3, R = Ph) and in solution an endo (syn) arrangement with respect to the central TiS2M core.4b The higher preference for the anti isomer found for these Ti–M mixed derivatives, related to the ones mentioned before, could be attributed to the bulkiness of the alkyne fragment on these Scheme 2 (i) [Mo(CO)3(NCMe)3], toluene, room temperature (r.t.); (ii) [M(C6F5)2(thf)2] (M = Pt or Pd), toluene, 220 8C; (iii) [Mo(CO)4(nbd)], toluene, r.t.; (iv) [M(C6F5)2(thf)2] (M = Pt or Pd), toluene, r.t.(for 5a, CH2Cl2). Ti R' R'' SC CR SC CR Ti P SiMe3 Mo OC CO CO Ph Ph SC CBut SC CBut S S SiMe3 SiMe3 S S SiMe3 SiMe3 S S SiMe3 SiMe3 Mo CO CO CO CO S S SiMe3 SiMe3 Mo CO CO CO CO Ti P Ph Ph SC CBut SiMe3 SC CBut C6F5 C6F5 C C6F5 M=Pt 8 M=Pd 9 C C6F5 M CR (iii) CR C C M C6F5 Ti (ii) (iv) 4a ratio syn:anti 1:1 7 C (i) C + R M 5:1 ratio syn: anti 5a But 0.9:1 6a But C Pd + Pt 6b Ph Pd CR M 10:1 5b Ph But Ti C C C Pt C 1:1 Ti C CR Ti But But But C6F53202 J.Chem. Soc., Dalton Trans., 1998, 3199–3208 m-SC]] ] CR bridging ligands. Similar steric considerations have previously been suggested to rationalise the shift of the equilibrium in favour of the anti isomer.18 The preference for the syn isomer is slightly higher for the phenyl derivatives 6b and 5b than for the tert-butyl complexes 6a and 5a respectively.The reason for the fact that the syn conformation seems to be more thermodynamically favoured on palladium than platinum mixed-metal complexes is less clear.According to the presence of an ª1:1 syn : anti mixture, the Ti–Mo complex 4a exhibits in its proton spectrum two singlet resonances (d 1.23, 1.20) due to But groups and, at high field, two signals of equal intensity (d 0.44, 0.33) assigned to the nonequivalent SiMe3 groups in the syn isomer and, a more intense signal at d 0.39 which belongs to the equivalent SiMe3 groups in the anti isomer. The expected three distinct cyclopentadienyl sets of resonances are observed slightly upfield shifted (d 6.31– 5.27) with respect to those seen for 1 (d 6.46, 6.38) indicating an increase of electron density on the Ti.This spectroscopic feature has been previously observed in related bis(alkyl) and aryl bridging thiolate Ti–Mo compounds.17 The proton spectrum is temperature dependent.Thus, on raising the temperature all signals broaden, and at 150 8C a single sharp But (d 1.24) and broad SiMe3 (d 0.41) resonances are observed while in the cyclopentadienyl region only two very broad humps are barely discernible suggesting that both isomers are interconverting on the NMR timescale. When the temperature is lowered the high-field region (But, SiMe3 resonances) does not change indicating a similar syn : anti ratio (1.1 : 1) but, however, the signals in the cyclopentadienyl region clearly broaden.In the lowest temperature spectrum (250 8C) ten distinct proton resonances [d 6.30 (2 H), 6.22 (2 H), 6.12, 5.84, 5.67 (1 H each), 5.52 (2 H), 5.26 (4 H), 5.12, 4.88, 4.79 (1 H each)] are seen implying rigid formulations with the lack of a symmetry plane passing through Ti and Mo atoms at low temperature.This fact could be tentatively related to hindrance of the rotation of either the bulky C]] ] CBut groups around the C(sp)–S bonds or the substituted h5-C5H4SiMe3 rings. As was previously found in related aryl (SPh, SC6F5) thiolate syn isomers [(h5-C5H4SiMe3)2Ti(m-SR)2M(C6F5)2] (M = Pt or Pd), the heterobinuclear Ti–Pt complexes 5 are relatively more rigid in solution than the Ti–Pd ones 6.Thus, both titanium– platinum complexes 5 display in their low (250 8C) and room temperature (20 8C) 19F NMR spectra the expected two diVerent sets (AFMRX systems) of rigid C6F5 fluorine resonances (one set assigned to each isomer), and similar spectra, but with a less defined pattern, were also observed at the highest accessible temperature (150 8C).No significant modification of the ratio of both isomers was observed in the range of temperature explored. Similar results were observed from the variable-temperature 1H NMR spectra. Only at high temperature (150 8C) the cyclopentadienyl and SiMe3 (also But groups for 5a) resonances of both isomers become broad (one SiMe3 is observed for 5b but coalescence of C5H4 signals is not reached) suggesting that the rate of interconversion syn–anti is still slow on the NMR timescale.By contrast, the 1H and 19F NMR spectra of the titanium–palladium complexes 6 at 150 8C show the presence of only one set of resonances for the C6F5, C5H4SiMe3 and But groups (this latter in the case of 6a).Selected ranges of the variable temperature 19F (Fortho) and 1H (C5H4SiMe3) spectra of 6b are shown in Figs. 2 and 3. As can be observed when the temperature is lowered the broad Fortho resonance (Fig. 2) is resolved into four distinct resonances with very diVerent 1 : 10 : 10 : 1 ratio. The signals with lower intensity which exhibit higher d(F2), d(F6) values (at 250 8C, 2115.8; 2117.6) are unequivocally assigned to the anti isomer in accordance with the proton data (Fig. 3). The proton spectrum at low temperature (250 8C) clearly reveals the presence of the two non-equivalent C5H4SiMe3 groups, which is consistent with that expected for the syn isomer (major isomer, d 6.81, 6.57, 6.36 and 6.28 CH; d 0.38, 0.22 SiMe3). The remaining signals of lower intensity (d 6.77, 6.45, 6.21 and 5.92 CH; 0.31 SiMe3) are therefore attributed to the anti isomer. When the temperature is increased the signals broaden and, finally, collapse to only two broad ones for the cyclopentadienyl resonances and one signal for the SiMe3 at ca. 150 8C. This pattern suggests fast interconversion of both isomers on the NMR timescale at this high temperature. Similar behaviour was observed for complex 6a, the most remarkable diVerence being the diVerent syn : anti (ª5 : 1) ratio found at low temperature.The 13C NMR spectra of all complexes have also been recorded (5, 6 at 250 8C, due to their low stability in solution, and room temperature for 4a, see Experimental section for data). A syn : anti mixture in approximately the expected ratio is observed for all complexes, particularly, for the SiMe3 and But (5a, 6a) resonances.Unfortunately, they are not very informative in the C]] ] C region. Only for 5b the expected four alkyne carbon resonances which appear slightly upfield shifted in relation to the starting material are clearly identified. For 6b the acetylenic carbon resonances of the major isomer (syn) are also shifted (d 103.5, 81.6 vs. 107.3, 93.0 1b) and, a small signal at d 99.2 can tentatively assigned to the minor anti isomer. For the remaining complexes, only one (4a) or two signals (5a, 6) in the d 67.78–75.5 range can be assigned. According to previous results 4c the preference for coordination through the phosphorus atom is evidenced by using the mixed-ligand mononuclear complex [Ti(h5-C5H4SiMe3)- (h5-C5H4PPh2)(SC]] ] CBut)2] 2a as precursor.Thus (Scheme 2), by treatment of 2a with [Mo(CO)4(nbd)] in toluene, at room temperature, a single heterodimetallic complex 7 was obtained in very low yield (12%). The IR spectrum (toluene solution) of the isolated material showed, in addition to a band at Fig. 2 Variable temperature 19F NMR spectra (Fortho region) of [(h5- C5H4SiMe3)2Ti(m-SC]] ] CPh)2Pd(C6F5)2] 6b (syn and anti).Fig. 3 Proton NMR spectra of the complex [(h5-C5H4SiMe3)2Ti(m- SC]] ] CPh)2Pd(C6F5)2] 6b (syn and anti) at diVerent temperatures.J. Chem. Soc., Dalton Trans., 1998, 3199–3208 3203 2070 cm21 assignable to n(C]] ] C), a clear CO 1956vs, 1895m, 1879s pattern attributable to a fac-Mo(CO)3 unit suggesting that the organometallic 2a fragment is acting as a tridentate (S,S,P) ligand to the Mo.Further evidence follows from the elemental analysis and the spectroscopic properties. Moreover, when [Mo(CO)3(NCMe)3] was used instead of [Mo(CO)4(nbd)] the reaction proceeded, as expected, in a cleaner way and complex [(h5-C5H4SiMe3)Ti(m-h5 :k-P-C5H4PPh2)(m-SC]] ] CBut)2- Mo(CO)3] 7 was obtained in a higher yield (53%).A similar behaviour has recently been observed by us when using related trifunctional ligand systems [Ti(h5-C5H4SiMe3){h5-C5H4P(E)- PPh2}(SPh)2] (E = O or S) and [W(CO)4(nbd)2].19 It seems that the three potential donor atoms (S, S and P) are well suited for the stabilisation of the fac-Mo(CO)3 fragment. The NMR data reveal that only one of the two expected isomers (syn and anti) is present in solution.A syn orientation is tentatively suggested on the basis of the 1H and 13C NMR spectra which display magnetically equivalent SC]] ] CBut ligands. Thus, the 1H NMR spectrum exhibits, in addition to four cyclopentadienyl proton resonances at d 6.23, 5.17 and 5.61, 5.49 assignable to diVerent C5H4PPh2 and C5H4SiMe3 rings, respectively, a single sharp But signal at d 1.20.The SiMe3 protons are observed at d 0.41. A similar pattern was observed in the 250 to 150 8C temperature range, suggesting the absence of any dynamic process. In the 13C NMR spectrum the proposed formulation is mainly supported by the observation of only one set of acetylenic carbon resonances (d 111.4 and 75.4) and a clear singlet signal at d 31.1 due to methyl carbon resonances of the equivalent But groups.Furthermore, in accordance with the P,S,S, coordination suggested, the 31P NMR spectrum shows the phosphorus resonances (d 39.7) strongly shifted to low field (D = 154.9) relative to the starting material (d 215.2 2a). Similarly, as shown in Scheme 2, treatment of complex 2a with 1 equivalent of cis-[M(C6F5)2(thf)2] (M = Pt or Pd) in toluene at low temperature (220 8C) aVords the heterodinuclear derivatives [(h5-C5H4SiMe3)(SC]] ] CBut)Ti(m-h5 :k-P-C5H4PPh2)- (m-SC]] ] CBut)M(C6F5)2] (M = Pt 8 or Pd 9).These complexes, isolated as violet microcrystalline solids, are moderately airstable in the solid state, but in solution they decompose in a few hours. The dimetallic formulation with an heteromixed bridging system is consistent with their spectroscopic data (IR, NMR) and confirmed by an X-ray diVraction study on the Ti–Pt complex 8 (see below).The presence of non-co-ordinated alkyne fragments is inferred from the IR spectra. Thus, both complexes show n(C]] ] C) absorptions assignable to the alkynethiolate ligands which lie approximately in the same region as for the corresponding mononuclear derivative [2157w, 2141m 8; 2158w, 2141m 9 vs. 2145 cm21 2a]. Moreover, co-ordination of the phosphorus atom is evidenced from their 31P NMR spectra, which show a singlet resonance (d 5.14 8, 10.93 9) shifted to higher frequency relative to that of 2a. For both complexes the signal is somewhat broad probably due to unresolved longrange phosphorus–fluorine couplings and, as expected, for 8 the signal is flanked by 195-platinum satellites [1J(Pt–P) = 2361 Hz].The 1H NMR spectra (at 250 8C and at room temperature) exhibit, in addition to phenyl resonances, two singlets at d 1.20, 1.11 for 8 and 1.22, 1.11 for 9 and another singlet at d 0.13 due to the methyl moieties of the inequivalent tertbutylalkynethiolate and free C5H4SiMe3 ligands, respectively.Seven proton signals (one of them with double intensity) are seen in the cyclopentadienyl region indicating magnetically non-equivalent halves on both substituted cyclopentadienyl rings. The 19F NMR spectra are not temperature dependent either, showing the presence of inequivalent C6F5 rings, for which the platinum co-ordination plane is not a mirror plane (AFMRX systems, see Experimental section).A single crystal X-ray structural determination of complex 8 (Fig. 4) confirmed that the mononuclear precursor acts as a P,S bidentate ligand towards the “cis-Pt(C6F5)2” fragment. The complex crystallises together with one molecule of toluene and 0.5 of hexane. Selected bond lengths and angles are collected in Table 2. The titanium atom is pseudo-tetrahedrally surrounded by two cyclopentadienyl ligands and the sulfur atoms of the two SC]] ] CBut ligands.The platinum centre exhibits a distorted “square-planar” geometry formed by the Cipso atoms of two mutually cis C6F5 groups, a sulfur atom of a m-SC]] ] CBut ligand and a phosphorus atom of the bridging C5H4PPh2 group. The centroid(1)–Ti–centroid(2) angle of 133.58 as well as the titanium–centroid distances (2.039 and 2.049 Å) are in the usual range found for related complexes such as [{(Mo(CO)4}2{m-(PPh2C5H4)2Ti(SPh)2}] (2.065 Å) 4e or [(h5-C5H4SiMe3)2Ti(m-SPh)2Pd(C6F5)2] [2.07(2), 2.05(1) Å] 4b with the cyclopentadienyl rings exhibiting an antiperiplanar (staggered) disposition.As was expected the two titanium to sulfur linkages are very diVerent, the shortest corresponding to the unco-ordinated SC]] ] CBut. The bond between the metal to the sulfur of the terminal thioalkyne ligand [Ti–S(2) 2.366(4) Å] is slightly shorter than that observed in 1a [2.451(4) Å] but in the range found for other mononuclear titanocene dithiolates such as [Ti(h5-C5H5)2(SEt)2] [2.398(3) and 2.387(3) Å].20 The other sulfur atom S(1) is bridging between titanium and platinum.The Ti–S(1) bond distance [2.532(4) Å] is substantially longer than the corresponding Pt–S(1) bond length [2.360(3) Å] and both slightly longer than those previously observed in the trimetallic complex [(OC)4Mo(m-PPh2C5H4)2- Ti(m-SPh)2Pt(C6F5)2] 4c [Ti–S 2.305(1), 2.456(2); Pt–S 2.256(1), 2.347(1) Å]. However, these distances lie in the range of those for other thiolate-bridged containing titanium or platinum centres.3,4,21,22 The Pt–P bond distance of 2.277(3) Å (and also the Pt–S) is comparable with that found in [Pt(SC5H9NMe2)( dppe)].22 The S(1)–Ti–S(2) angle of 89.39(12)8 is slightly Fig. 4 Molecular structure of [(h5-C5H4SiMe3)(SC]] ] CBut)Ti(m-h5 :k-PC5H4PPh2)( m-SC]] ] CBut)Pt(C6F5)2] 8. Table 2 Selected bond lengths (Å) and angles (8) for complex 8 Pt–C(7) Pt–P(1) Pt–C(1) Pt–S(1) Ti–S(2) Ti–S(1) Ti ? ? ? Pt C(1)–Pt–P(1) Ti–S(1)–C(38) C(1)–Pt–S(1) C(7)–Pt–S(1) P(1)–Pt–S(1) Pt–S(1)–Ti S(2)–Ti–S(1) Ti–S(2)–C(44) 2.023(13) 2.277(3) 2.039(12) 2.360(3) 2.366(4) 2.532(4) 3.817(3) 175.7(4) 106.6(4) 94.5(3) 178.1(4) 83.92(11) 102.51(12) 89.39(12) 111.1(4) S(1)–C(38) S(2)–C(44) Ti–cent(1) Ti–cent(2) C(38)–C(39) C(44)–C(45) C(39)–C(38)–S(1) C(38)–C(39)–C(40) C(45)–C(44)–S(2) S(1)–Ti–cent(1) S(1)–Ti–cent(2) C(7)–Pt–C(1) C(7)–Pt–P(1) C(44)–C(45)–C(46) 1.700(12) 1.668(12) 2.039 2.049 1.19(2) 1.20(2) 175.3(11) 178.2(13) 179.3(11) 102.0 109.7 87.0(5) 94.6(4) 178.2(13)3204 J.Chem. Soc., Dalton Trans., 1998, 3199–3208 smaller than that seen in 1a [92.30(13)8] and those observed in related mononuclear titanocene bis(thiolate) complexes [Ti(h5-C5H4SiMe3)2(SC6F5)2] [100.6(1)8],14a [Ti(h5-C5H5)2(SR)2] [R = Ph (99.48),23 or Et (93.88) 20] or titanocene thiolate bridged [(h5-C5H4SiMe3)2Ti(m-SPh)2Pd(C6F5)2] 4b [95.7(2)8] complexes. The internal angles at platinum and at bridging sulfur [P(1)–Pt– S(1) 83.92(11)8 acute and Pt–S(1)–Ti 102.51(12)8 obtuse] are in accordance with the very long Pt ? ? ? Ti distance [3.817(3) Å] found.The acetylenic fragments, C]] ] CBut, are located on the same side of the S(1)–Ti(1)–S(2) plane adopting an endo (syn) conformation. Their structural data, C]] ] C bonds [C(38)–C(39) 1.19(2), C(44)–C(45) 1.20(2) Å] and bond angles [S(1)–C(38)– C(39) 175.3(11); C(38)–C(39)–C(40) 178.2(13), S(2)–C(44)– C(45) 179.3(11), C(44)–C(45)–C(46) 178.2(13)8], are in the usual range and deserve no further comment.As was mentioned before, complex [Ti(h5-C5H4PPh2)2- (SC]] ] CBut)2] 3a is very unstable both in the solid state and in solution. In preliminary experiments it was treated with [Mo(CO)4(nbd)] (1 equivalent) in toluene either at room or lower (240 8C) temperature, but unfortunately the reaction failed, giving just decomposition products and, therefore, no more experiments were made with this precursor.In summary, bis(alkynethiolate)titanium complexes [Ti(h5- C5H4R9)(h5-C5H4R0)(SC]] ] CR)2] 1–3 have been prepared from [Ti(h5-C5H4R9)(h5-C5H4R0)Cl2], by using classical metal– halogen exchange reactions with LiSC]] ] CR reagents. In spite of the presence of two potential bifunctionalities, the lone pair at the sulfur atom and the acetylenic moiety on each SC]] ] CR, the mononuclear [Ti(h5-C5H4SiMe3)2(SC]] ] CR)2] (R = But 1a or Ph 1b) complexes serve only as bidentate (S,S) metalloligands when treated with d6 [Mo(CO)4(nbd)] or d8 cis-[M(C6F5)2(thf)2] (M = Pt or Pd) substrates.The co-ordination of the acetylenic fragments cannot be forced even in presence of an excess of these latter reagents, reactions which lead to the same doubly thiolate-bridged early–late heterodimetallic products 4–6 (see Experimental section).The NMR data reveal that in all cases the products are isolated as a syn : anti mixture of isomers with a clear thermodynamic preference for the syn conformation in the palladium mixed-metal complexes (ª1 : 1 for Ti–Mo 4a and Ti–Pt 5 vs. 5:1 6a, 10 : 1 6b). The variable NMR data confirm that both isomers interconvert on the NMR timescale at the highest accessible temperature (150 8C) (4a and 6 fast 5 slow). Similar to previous observations, a favoured co-ordination through phosphine ligands with these late transition metals is evidenced by the fact that the mixed ligand complex [Ti(h5-C5H4SiMe3)(h5-C5H4PPh2)(SC]] ] CBut)2] 2a acts as bidentate P,S when treated with cis-[M(C6F5)2(thf)2] (M = Pt or Pd) yielding 8 and 9, respectively and, as a tridentate organometallic metallo ligand toward [Mo(CO)4(nbd)] or [Mo(CO)3- (NCMe)3], giving [(h5-C5H4SiMe3)Ti(m-h5 :k-P-C5H4PPh2)- (m-SC]] ] CBut)2Mo(CO)3] 7.Experimental Reactions were carried out under an atmosphere of argon by means of conventional Schlenk techniques.24 Solvents were purified according to standard procedures.25 The complexes [Ti(h5-C5H4SiMe3)2Cl2],26 [Ti(h5-C5H4SiMe3)(h5- C5H4PPh2)Cl2],14b [ Ti(h5-C5H4PPh2)2Cl2],27 [ Mo(CO)4(nbd)],28 [Mo(CO)3(NCMe)3] 29 and cis-[M(C6F5)2(thf)2] 30 (M = Pt or Pd) were prepared as previously published.All other reagents were used as obtained commercially. Microanalyses were determined with Perkin-Elmer 2400 and 240-B microanalysers. Infrared spectra (KBr) were recorded on Perkin-Elmer 1600 FT and FT-IR 1000 spectrophotometers, NMR spectra on Bruker AMX-300 or ARX-300 with chemical shifts reported in ppm relative to external standards (SiMe4 for 1H and 13C, CFCl3 for 19F and H3PO4 for 31P) and mass spectra (FAB1) on a VG Autospec spectrometer. Syntheses [Ti(Á5-C5H4SiMe3)(SC]] ] CBut)2] 1a.To a diethyl ether solution (20 cm3) of LiBun (1.66 cm3, 2.66 mmol) cooled at 220 8C was added ButC]] ] CH (0.32 cm3, 2.66 mmol). After 10 min of stirring S8 (0.085 g, 0.33 mmol) was introduced in the Schlenk and the cooling bath was removed. The mixture was stirred for 45 min at room temperature and subsequently added dropwise to another diethyl ether solution (25 cm3) of [Ti(h5-C5H4SiMe3)2Cl2] (0.50 g, 1.27 mmol) cooled at 270 8C.The bright green solution obtained was kept under nitrogen with continuous stirring for 1 h while the temperature slowly reached 210 8C. The solvent was evaporated to dryness, the residue then extracted with pentane and filtered through a pad of Celite. The resulting solution was concentrated and cooled to 220 8C to yield dark green needles of complex 1a (0.63 g, 85%) (Found: C, 61.13; H, 8.03.C28H44S2Si2Ti requires C, 61.28; H, 8.08%); n& max/cm21 2129 (C]] ] C). MS: m/z 548 {(h5-C5H4SiMe3)2Ti(SC]] ] CBut)2]1, 8}, 435 {[(h5-C5H4SiMe3)2- (SC]] ] CBut)]1, 100} and 322 {[(h5-C5H4SiMe3)2Ti]1, 60%}. 1H NMR (CDCl3): d 6.46 (t, 4 H, C5H4SiMe3), 6.38 (t, 4 H, C5H4SiMe3), 1.35 (s, 18 H, But) and 0.24 (s, 18 H, SiMe3). 13C-{1H} NMR (CDCl3): d 123.3 (s, C1 of C5H4), 122.0 (s, C2,5 of C5H4), 119.4 (s, C3,4 of C5H4), 117.5 (s, C]] ] C), 80.8 (s, C]] ] C), 31.8 (s, But) and 0.17 (s, SiMe3). [Ti(Á5-C5H4SiMe3)2(SC]] ] CPh)2] 1b. This compound was obtained following the above procedure starting from [Ti(h5-C5H4SiMe3)2Cl2] (0.45 g, 0.76 mmol) and LiSC]] ] CPh (1.60 mmol).After 1.5 h of stirring the resulting diethyl ether solution was concentrated and filtered through a pad of Celite. Crystallisation from a saturated diethyl ether solution at 220 8C aVorded dark green crystals of compound 1b (85%) (Found: C, 65.03; H, 6.09. C32H36S2Si2Ti requires C, 65.28; H, 6.16%); n& max/cm21 2134 (C]] ] C). MS: m/z 588 {[(h5-C5H4SiMe3)2- Ti(SC]] ] CPh)2]1, 4}, 455 {[(Ti(SC]] ] CPh)]1, 100} and 322 {[(h5-C5H4SiMe3)2Ti]1, 95%}. 1H NMR (CDCl3): d 7.41–7.37 (m, 4 H, Ph), 7.23–7.19 (m, 6 H, Ph), 6.62 (t, 4 H, C5H4SiMe3), 6.53 (t, 4 H, C5H4SiMe3) and 0.28 (s, 18 H, SiMe3). 13C-{1H} NMR (CDCl3): d 138.8–126.9 (s, C6H5), 124.3 (s, C1 of C5H4), 122.2 (s, C2,5 of C5H4) 119.7 (s, C3,4 of C5H4), 107.3 (s, C]] ] C), 93.0 (s, C]] ] C) and 0.16 (s, SiMe3).[Ti(Á5-C5H4SiMe3)(Á5-C5H4PPh2)(SC]] ] CBut)2] 2a. This compound was obtained following the same procedure as for 1a but the final residue was extracted with heptane (73% yield). The precursors used were [Ti(h5-C5H4SiMe3)- (h5-C5H4PPh2)Cl2] (0.37 g, 0.56 mmol) and LiSC]] ] CBut (1.17 mmol) (Found: C, 66.82; H, 6.90. C37H45PS2SiTi requires C, 67.25; H, 6.86%); n& max/cm21 2145 (C]] ] C).MS: m/z 660 {[(h5-C5H4SiMe3)(h5-C5H4PPh2)Ti(SC]] ] CBut)2]1, 30}, 547 {[(h5- C5H4SiMe3)(h5-C5H4PPh2)Ti(SC]] ] CBut)]1, 100} and 432 {[(h5-C5H4SiMe3)(h5-C5H4PPh2)Ti]1, 50%}. 1H NMR (CDCl3): d 7.41–7.37 (m, 4 H, Ph), 7.23–7.19 (m, 6 H, Ph), 6.54 (m, 2 H, C5H4PPh2), 6.38 (t, 2 H, C5H4SiMe3), 6.34 (t, 2 H, C5H4SiMe3), 6.11 (m, 2 H, C5H4PPh2), 1.31 (s, 18 H, But) and 0.17 (s, 9 H, SiMe3). 31P-{1H} NMR: d 215.2 (s, C5H4- PPh2). 13C-{1H} NMR (CDCl3): d 133.8–128.5 (s, C6H5), 124.5, 123.5, 123.4, 121.8, 121.5, 120.3, 120.1, 120.0, 119.9, 119.1 (s, C5H4), 117.8 (s, C]] ] C), 80.2 (s, C]] ] C), 31.7 (s, But) and 0.11 s, SiMe3). [Ti(Á5-C5H4PPh2)2(SC]] ] CBut)2] 3a. The synthesis was performed as described for complex 1a starting from [Ti(h5-C5H4PPh2)2Cl2] (0.45 g, 0.73 mmol) and LiSC]] ] CBut (1.53 mmol).After 30 min of stirring the resulting diethyl ether solution was concentrated and filtered through a pad of Celite. The solvent was evaporated to dryness aVording 3a as a green solid (70%). n& max/cm21 2129 (C]] ] C). 1H NMR (CDCl3): d 7.27–7.17 (m, 10 H, Ph), 6.40 (t, 4 H, C5H4PPh2), 6.01 (m, 4J. Chem. Soc., Dalton Trans., 1998, 3199–3208 3205 H, C5H4PPh2) and 1.33 (s, 18 H, But). 31P-{1H} NMR: d 215.5 (s, C5H4PPh2). The 13C NMR spectrum could not be recorded due to the low stability in solution. [(Á5-C5H4SiMe3)2Ti(Ï-SC]] ] CBut)2Mo(CO)4] 4a (syn and anti). To a solution of [Ti(h5-C5H4SiMe3)2(SC]] ] CBut)2] 1a (0.20 g, 0.36 mmol) in toluene (25 cm3) was added [Mo(CO)4(nbd)] (0.32 g, 1.08 mmol) and the mixture stirred at room temperature for 30 h.The solvent was removed in vacuo and the solid obtained purified by chromatography on silica gel 100. Elution with hexane–toluene (3 : 1) aVorded a green-blue band of complex 4a (0.16 g, 60%) (syn : anti ratio ª1 : 1). Identical results were obtained starting from 1a and 1 equivalent of [Mo(CO)4(nbd)], but in that case longer periods of stirring (ª72 h) were necessary (Found: C, 50.51; H, 5.73.C32H44MoO4S2Si2Ti requires C, 50.79; H, 5.86%). n& max/cm21 2072 (C]] ] C); (toluene solution) 2019s, 1929s, 1915vs (CO). MS: m/z 756 {[(h5-C5H4SiMe3)2Ti- (m-SC]] ] CBut)2Mo(CO)4]1, <5}, 728 {[(h5-C5H4SiMe3)2Ti(m-SC]] ] CBut)2Mo(CO)3]1, <5%}, 700 {[(h5-C5H4SiMe3)2Ti(m-SC]] ] CBut)2Mo(CO)2]1, <5}, 672 {[(h5-C5H4SiMe3)2Ti(m-SC]] ] CBut)2- Mo(CO)]1, <5}, 644 {[(h5-C5H4SiMe3)2Ti(m-SC]] ] CBut)2Mo]1, 15}, 435 {[(h5-C5H4SiMe3)2Ti(SC]] ] CBut)]1, 50} and 322 {[(h5- C5H4SiMe3)2Ti]1, 100%}. 1H NMR (CDCl3): at 250 8C, d 6.30 (2 H), 6.22 (2 H, 6.12, 5.84, 5.67 (1 H each), 5.52 (2 H), 5.26 (4 H), 5.12, 4.88, 4.79 (1 H each) (C5H4 syn, anti isomers), 1.20s, 1.17s (But), 0.37 (s, SiMe3, anti isomer), 0.42s, 0.22s (SiMe3, syn isomer); at 20 8C, 6.31 (s, br, 2 H, C5H4, syn isomer), 6.21 (s, br, 2 H, C5H4, syn isomer), 5.55 (s, br, 6 H, C5H4, anti and syn isomers), 5.36 (s, br, 2 H, C5H4, syn isomer), 5.27 (s, br, 4 H, C5H4, anti isomer), 1.23 (s, But), 1.20 (s, But), 0.39 (s, 18 H, SiMe3, anti isomer), 0.44, 0.33 (s, SiMe3, syn isomer), (ratio syn : anti ª1 : 1); at 150 8C, cyclopentadienyl region very broad (ª6.2, 5.4 br), 1.24 (s, But), 0.41 (s, br, SiMe3). 13C-{1H} NMR (CDCl3): d 217.9 (s, CO equatorial, syn isomer), 217.2 (s, CO equatorial, anti isomer), 204.7 (s, CO axial, syn isomer), 203.1 (s, CO axial, anti isomer), 201.7 (s, CO axial, syn isomer), 129.3br, 123.4br, 116.8, 114.5, 113.3, 112.4, 106.2, 102.2, 101.2 (s, C5H4, C]] ] C), 75.9 (s, C]] ] C), 31.1 [s, C(CH3)3], 28.9 (s, CMe3) and 0.14 (s, SiMe3).[(Á5-C5H4SiMe3)2Ti(Ï-SC]] ] CBut)2Pt(C6F5)2] 5a (syn and anti). A deep green solution of [Ti(h5-C5H4SiMe3)2(SC]] ] CBut)2] (0.098 g, 0.178 mmol) in CH2Cl2 (10 cm3) was treated with cis-[Pt(C6F5)2(thf)2] (0.120 g, 0.178 mmol) and, immediately, turned red-brown. The mixture was stirred for 5 min and then the solvent was removed in vacuo.Addition of n-heptane (ª5 cm3) to the residue aVorded an orange-brown solid (0.153 g, 80% yield) identified as a mixture of syn and anti isomers of [(h5-C5H4SiMe3)2Ti(m-SC]] ] CBut)2Pt(C6F5)2] 5a. When the reaction was carried out in a molar ratio 1 : 2 using complex 1a (0.010 g, 0.019 mmol) and cis-[Pt(C6F5)2(thf)2] (0.025 g, 0.037 mmol) in CDCl3 (0.6 cm3) and monitored by 1H and 19F NMR spectroscopy at 20 8C the complex 5a was observed (major component) in addition to decomposition products (Found: C, 44.50; H, 3.70; S, 5.95.C40F10H44PtS2Si2Ti requires C, 44.57; H, 4.11; S, 5.48%). n& max/cm21 2168m (C]] ] C), 800vs, (br) (C6F5)X–sens. MS: m/z 1077 (M1, 28), 964 ([M 2 SC]] ] CBut]1, 32), 940 ([M 2 C5H4SiMe3]1, 25), 910 ([M 2 C6F5]1, 94), 631 {[(h5-C5H4SiMe3)2Ti(SC]] ] CBut)Pt]1, 58}, 475 {[(h5-C5H4- SiMe3)(C5H4)Ti(SC]] ] CBut)2]1, 100}, 435 {[(h5-C5H4SiMe3)2- Ti(SC]] ] CBut)]1, 70} and 322 {[(h5-C5H4SiMe3)2Ti]1, 100%}. 1H NMR (CDCl3): at 20 8C, d 6.46, 6.29, 6.23, 6.13, 5.94, 5.87, 5.73 (s, ratio 1:1:1:1:2:1:1, C5H4, syn and anti isomers), 1.21 (s, But), 1.14 (s, But), 0.39 (s, SiMe3, syn isomer), 0.33 (s, SiMe3, anti isomer), 0.25 (s, SiMe3, syn isomer), (syn : anti 0.9 : 1); approximately the same spectra is observed at 250 8C; at 150 8C, the signals are broad, 6.5, 6.3, 6.2, 5.98, 5.90, 5.80 (br, C5H4), 1.21, 1.17 (br, But), 0.35 (br, SiMe3 anti and syn isomers), 0.28 (SiMe3, syn isomer). 19F NMR [CDCl3, 3J(Pt–Fo)/Hz in parentheses]: at 250 8C, d 117.94 [dm (417)], 2118.07 [dm (ª355)], 2118.6 [dm (ª465)], 2119.4 [dm (392)] (Fo syn and anti isomers), 2161.3 (t, Fp, anti isomer), 2161.5 (t, Fp, syn isomer), 2164.2 (m, Fm, syn and anti isomers) (syn : anti 0.9 : 1); at 20 8C, 2117.6 [dm, overlapping of two Fo (ª411, ª337)], 2118.6 [dm (ª455)], 2119.3 [dm (ª385 Hz)] (ratio 2:1:1, Fo, syn and anti), 2162.0m, 2162.25m (ratio 0.9 : 1, Fp, syn and anti), 2164.5, 2165.0 (m, ratio 2 : 2, Fm, syn and anti); at 150 8C, 2117.5 [br (358)], 2118.6 [d, br (446)], 2119.2 [d, br (391)] (ratio 2:1:1, Fo, syn and anti), 2162.3 (m, br, overlapping of two Fp, syn and anti), 2164.6 (m, br), 2165.3m, (ratio 1 : 1, Fm, syn and anti). 13C-{1H} NMR (CDCl3): at 250 8C, d 148.05, 144.96, 138.5, 135.3, 123.3, 116.05 (br, C6F5), 130.1, 122.1, 120.8, 120.2, 120.0, 119.6, 113.1, 112.5, 109.6, 107.0 (s, C5H4 and C]] ] C), 67.98s, 67.78s (C]] ] C, syn and anti isomers), 30.3 [s, C(CH3)3], 29.1 (s, CMe3), 28.9 (s, CMe3), 0.0 (s, syn isomer), 20.14 (s, anti isomer) and 20.39 (s, syn isomer) [Si(CH3)3].[(Á5-C5H4SiMe3)2Ti(Ï-SC]] ] CPh)2Pt(C6F5)2] 5b (syn and anti). A solid mixture of [Ti(h5-C5H4SiMe3)2(SC]] ] CPh)2] (0.131 g, 0.223 mmol) and cis-[Pt(C6F5)2(thf)2] (0.150 g, 0.223 mmol) was treated with toluene (5 cm3).Immediately the resulting brown-red solution was concentrated in vacuo, giving an orange-red residue, identified as [(h5-C5H4SiMe3)2Ti(m- SC]] ] CPh)2Pt(C6F5)2] 5b (0.174 g, 70% yield) (syn : anti ratio at 250 8C, 1 : 1). When the reaction in a molar ratio 1 :2 {0.010 g, 0.0170 mmol of complex 1b and 0.023 g, 0.034 mmol of cis- [Pt(C6F5)2(thf)2] in 0.6 cm3 of CDCl3} was monitored by NMR spectroscopy at 20 8C considerable decomposition took place, with 5b being the major product.After longer periods (ª3 h) more decomposition was observed (Found: C, 47.47; H, 3.38; S, 5.31. C44F10H36PtS2Si2Ti requires C, 47.27; H, 3.24; S, 5.73%). n& max/cm21 2165m (C]] ] C), 801vs, (br) (C6F5)X–sens. MS: m/z 1117 (M1, 10), 619 {[(h5-C5H4SiMe3)2Ti(SC]] ] CPh)- (C6F5) 2 3H]1, 14}, 457 {[(h5-C5H4SiMe3)2Ti(SC]] ] CPh) 1 2H]1, 56} and 322 {[(h5-C5H4SiMe3)2Ti]1, 100%}. 1H NMR (CDCl3): at 250 8C, d 7.37–7.20 (Ph), 6.59, 6.46, 6.41, 6.37, 6.12, 6.09, 6.03, 5.91 (s, identical ratio, C5H4, syn and anti isomers), 0.43 (s, SiMe3, syn isomer), 0.36 (s, SiMe3, anti isomer), 0.28 (s, SiMe3, syn isomer) (syn : anti ª1 : 1); at 20 8C, 7.37–7.16 (Ph), 6.61, 6.48, 6.41, 6.17, 6.12, 6.10, 5.95 (s, ratio 1:1:2:1:1:1:1, C5H4, syn and anti isomers), 0.43 (s, SiMe3, syn isomer), 0.38 (s, SiMe3, anti isomer), 0.30 (s, SiMe3, syn isomer) (syn : anti ª1 : 1); at 150 8C, 7.36–7.15 (Ph), 6.60sh, 6.45br, 6.14br, 6.01sh (C5H4) and 0.39 (s, br, SiMe3). 19F NMR [CDCl3, 3J(Pt–Fo)/Hz in parentheses]: at 250 8C, d 2117.8 [d (430), 2F], 2118.5 [d (451), 2F], 2118.97 [d (365), 2F] 2120.0 [d (398), 2F] (Fo, syn and anti isomer), 2160.6, 2161.5 (t, Fp, syn and anti isomer), 2163.3, 2164.0 (m, Fm, syn and anti isomer) (syn : anti 1 : 1); at 20 8C, 2117.6 [d (408), 2F], 2118.6 [dm, overlapping of two Fo (ª458, ª389), 4F], 2119.8 [d (392), 2F], (Fo, syn and anti isomer), 2161.3, 2161.7 (t, Fp, syn and anti isomer), 2163.8, 2164.6 (m, Fm, syn and anti isomer) (syn : anti ª1 : 1); at 150 8C, 2117.5, 2118.3, 2118.5, 2118.98 (br, Fo), 2161.6, 2161.9 (br, Fp), 2164.0, 2164.9 (br, Fm) (syn and anti isomer). 13C-{1H} NMR (CDCl3): at 250 8C, d 148.2, 145.2, 138.7, 138.2, 135.4–134.0, 116.6, 113.8 (br, C6F5), 138.2, 131.3–112.5 (s, C6H5, C5H4), 99.3s, 96.4s (C]] ] C syn and anti isomers), 79.7s, 79.5s (C]] ] C syn and anti isomers), 20.0 (s, syn and anti isomer) and 20.35 (s, syn isomer) [Si(CH3)3]. [(Á5-C5H4SiMe3)2Ti(Ï-SC]] ] CBut)2Pd(C6F5)2] 6a (syn and anti).This product was prepared in a similar way to complex 5b by using the appropriate starting precursors, [Ti(h5-C5H4- SiMe3)2(SC]] ] CBut)2] (0.141 g, 0.256 mmol) and cis- [Pd(C6F5)2(thf)2] (0.150 g, 0.256 mmol). It was isolated by removing the solvent in vacuo, (yield 0.16 g 63%) (mixture of syn and anti isomers, ratio ª5:1 at 250 8C).When an excess of cis-[Pd(C6F5)2(thf)2] was employed {1 : 2 molar ratio; 0.012 g, 0.021 mmol of 1a and 0.025 g, 0.043 mmol of cis- [Pd(C6F5)2(thf)2] in 0.6 cm3 of CDCl3} a mixture of 6a and3206 J. Chem. Soc., Dalton Trans., 1998, 3199–3208 cis-[Pd(C6F5)2(thf)2] was observed by NMR spectroscopy (Found: C, 48.08; H, 4.21; S, 6.32.C40F10H44PdS2Si2Ti requires C, 48.56; H, 4.48; S, 6.48%). n& max/cm21 2166m (C]] ] C), 786s, 778s (C6F5)X–sens. MS: m/z 1011 ([M 1 Na]1, 2), 541 {[(h5-C5H4- SiMe3)2Ti(SC]] ] CBut)Pd]1, 7}, 492 {[(h5-C5H4SiMe3)2Pd- (SC]] ] CBut) 2 H]1, 7}, 435 {[(h5-C5H4SiMe3)2Ti(SC]] ] CBut)]1, 53} and 322 {[(h5-C5H4SiMe3)2Ti]1, 100%}. 1H NMR (CDCl3): at 250 8C, d 6.45, 6.39, 6.07, 6.00 (s, C5H4, syn isomer), 6.31, 5.72 (C5H4, anti isomer), 1.23 (s, But, syn isomer), 1.14 (s, But, anti isomer), 0.37 (s, SiMe3, syn isomer), 0.27 (s, SiMe3, anti isomer), 0.19 (s, SiMe3, syn isomer) (syn : anti 5 : 1); at 20 8C, 6.48, 6.43, 6.11, 6.07 (s, ratio 1:1:1:1, C5H4, syn isomer), 6.30, 5.81 (br, C5H4, anti isomer), 1.26 (s, But, syn isomer), 1.22 (sh, But, anti isomer), 0.37 (s, SiMe3, syn isomer), 0.30 (s, SiMe3, anti isomer) and 0.23 (s, SiMe3, syn isomer); at 150 8C, 6.42, 6.11 (br, C5H4), 1.25, (s, But) and 0.31 (s, SiMe3). 19F NMR (CDCl3): at 250 8C, d 2115.2 (d, anti isomer), 2115.5 (d, syn isomer), 2115.9 (dm, syn isomer), 2117.0 (d, anti) (Fo, ratio syn : anti 5 : 1), 2160.7 (t, overlapping of two Fp, syn and anti isomer), 2163.1, 2163.8 (br, Fm, syn and anti isomer); at 20 8C, 2114.9 (d, anti isomer), 2115.5 (m, overlapping of two Fo, syn isomer) 2116.8 (d, anti isomer) (Fo, syn : anti 3: 1), 2161.4 (t, Fp), 2163.8, 2164.5 (m, Fm, syn and anti isomer); at 150 8C, 2115.3 (br, Fo), 2161.7 (t, Fp), 2164.1, 2164.7 (br, Fm). 13C- {1H} NMR (CDCl3): at 250 8C, d 147.7, 144.7, 138.5–137.1, 135.3–133.8, 120.4 (br, C6F5), 129.4–110.0 (s, C5H4, C]] ] C), 70.0 (s, C]] ] C, anti isomer), 69.8 (s, C]] ] C, syn isomer), 30.4 [s, C(CH3)3, syn isomer], 29.3 [s, C(CH3)3, anti isomer], 29.1 (s, CMe3, syn isomer), 27.6 (s, CMe3, anti isomer), 20.17 [s, Si(CH3)3, syn isomer], 20.27 [s, Si(CH3)3, anti isomer] and 20.52 [s, Si(CH3)3, syn isomer].[(Á5-C5H4SiMe3)2Ti(Ï-SC]] ] CPh)2Pd(C6F5)2] 6b (syn and anti). The reaction was performed as described for complex 5a in toluene (5 cm3) starting from [Ti(h5-C5H4SiMe3)2(SC]] ] CPh)2] (0.150 g, 0.255 mmol) and cis-[Pd(C6F5)2(thf)2] (0.149 g, 0.255 mmol).In this case 6b was precipitated as a red-garnet solid by adding n-hexane (3 cm3) (0.23 g, 88% yield) (syn : anti at 250 8C, 10 : 1). When the reaction was carried out in a 1 : 2 molar ratio in 0.6 cm3 of CDCl3{0.013 g, 0.021 mmol of 1b and 0.025 g, 0.043 mmol of cis-[Pd(C6F5)2(thf)2]} a mixture of 6b and cis-[Pd(C6F5)2(thf)2] was observed by NMR spectroscopy (Found: C, 50.79; H, 3.61; S, 5.91.C44F10H36PdS2Si2Ti requires C, 51.34; H, 3.52; S, 6.23%). n& max/cm21 2165m (C]] ] C), 789vs, 778vs (C6F5)X–sens. MS: m/z 861 {[(h5-C5H4SiMe3)2Ti(SC]] ] CPh)2- Pd(C6F5)]1, 15}, 619 {[(h5-C5H4SiMe3)2Ti(SC]] ] CPh)- (C6F5) 2 3H]1, 14}, 457 {[(h5-C5H4SiMe3)2Ti(SC]] ] CPh) 1 2H]1, 15} and 322 {[(h5-C5H4SiMe3)2Ti]1, 100%}. 1H NMR (CDCl3): at 250 8C, d 7.29 (m), 7.17 (m) (Ph), 6.81, 6.57, 6.36, 6.28 (br, C5H4, syn isomer), 6.77, 6.45, 6.21 and 5.92 (C5H4, anti isomer), 0.38 (s, SiMe3, syn isomer), 0.31 (s, SiMe3, anti isomer), 0.22 (s, SiMe3, syn isomer) (syn : anti, 10 : 1); at 20 8C, 7.27 (m, Ph), 6.83, 6.59, 6.42, 6.31 (br, C5H4), 0.39 (s, SiMe3), 0.26 (s, SiMe3) (syn and anti isomer); at 150 8C, 7.33 (d), 7.23 (m) (Ph), 6.66 (vbr), 6.39 (vbr) (C5H4) and 0.34 (s, SiMe3). 19F NMR (CDCl3): at 250 8C, d 2115.8 (d, anti isomer), 2115.4 (d, syn isomer), 2116.7 (d, syn isomer), 2117.6 (d, anti isomer) (Fo, syn : anti 10 : 1), 2160.1 (t, Fp, anti isomer), 2160.4 (t, Fp, syn isomer), 2162.2, 2163.5 (m, Fm, syn and anti isomer); at 20 8C, 2114.8 (br, anti isomer), 2115.5 (d, syn isomer), 2116.1 (br, syn isomer), 2117.4 (br, anti isomer) (Fo, syn : anti ª7 : 1), 2161.0 (t, Fp, syn and anti isomer), 2162.9 (br), 2164.2 (m, br) (Fm, syn and anti isomer); at 150 8C, 2115.9 (br, Fo), 2161.2 (t, Fp), 2162.3 (br) and 2163.6 (br, Fm). 13C-{1H} NMR (CDCl3): at 250 8C, d 148.4, 147.6, 138.7, 138.1–137.2, 135.5, 134.9–133.8, 121.09, 118.2 (br, C6F5), 131.16, (s, Co, C6H5), 128.7, 128.4, 128.2 (s, Cm, C6H5), 125.2, 123.4, 123.1, 122.6 (s, C6H5, C5H4, syn), 103.5 (s, C]] ] C syn isomer), 81.6 (s, C]] ] C syn isomer), small signals seen at 121.1 and 99.2 (C]] ] C) tentatively attributed to the anti isomer, 20.13 [s, Si(CH3)3, anti isomer], 20.17 (s) and 20.39 (s) [Si(CH3)3, syn isomer].[(Á5-C5H4SiMe3)Ti(Ï-Á5 :Í-P-C5H4PPh2(Ï-SC]] ] CBut)2- Mo(CO)3] 7. To a toluene solution (25 cm3) of [Ti(h5-C5H4- SiMe3)(h5-C5H4PPh2)(SC]] ] CBut)2] 2a (0.20 g, 0.30 mmol) was added [Mo(CO)3(NCMe)3] (0.11 g, 0.36 mmol). After 3 h of stirring at room temperature the solvent was evaporated to dryness and the solid residue chromatographed on silica gel 100.A violet band was eluted by hexane–toluene (1 : 1) and its recrystallisation from heptane at 220 8C yielded 7 as a dark violet solid (0.13 g, 53%). Complex 7 can also be obtained in very low yield (12%) using 2a (0.23 g, 0.34 mmol) and [Mo(CO)4(nbd)] (0.12 g, 0.42 mmol) as precursors (Found: C, 56.51; H, 5.25. C40H45MoO3PS2SiTi requires C, 57.14; H, 5.39%).n& max/cm21 2070 (C]] ] C); (toluene solution) 1956vs, 1895m, 1879s (CO). MS: m/z 840 {[(h5-C5H4SiMe3)(h5-C5H4- PPh2)Ti(m-SC]] ] CBut)2Mo(CO)3]1, <5}, 784 {[(h5-C5H4SiMe3)- (h5-C5H4PPh2)Ti(m-SC]] ] CBut)2Mo(CO)]1, <5}, 756 {[(h5-C5H4- SiMe3)(h5-C5H4PPh2)Ti(m-SC]] ] CBut)2Mo]1, 100}, 547 {[(h5- C5H4SiMe3)(h5-C5H4PPh2)Ti(m-SC]] ] CBut)]1, 15} and 434 {[(h5- C5H4SiMe3)(h5-C5H4PPh2)Ti]1, 65%}. 1H NMR (CDCl3): d 7.61–7.52 (m, 4 H, Ph), 7.33–7.28 (m, 6 H, Ph), 6.23 (s, br, 2 H, C5H4PPh2), 5.61 (s, br, 2 H, C5H4SiMe3), 5.49 (s, br, 2 H, C5H4SiMe3), 5.17 (s, br, 2 H, C5H4PPh2), 1.20 (s, 18 H, But) and 0.41 (s, 9 H, SiMe3); similar spectra were obtained at low (250 8C) and high (150 8C) temperature. 31P-{1H} NMR: d 39.7 (s, C5H4PPh2). 13C-{1H} NMR (CDCl3): d 213.8 (s, CO), 133.3–128.4 (s, C6H5), 124.3–100.7 (s, C5H4), 111.4 (s, C]] ] C), 75.4 (s, C]] ] C), 31.1 [s, C(CH3)3], 28.9 [C(CH3)3] and 0.30 (s, SiMe3). [(Á5-C5H4SiMe3)(SC]] ] CBut)Ti(Ï-Á5 :Í-P-C5H4PPh2)(Ï-SC]] ] CBut)Pt(C6F5)2] 8. To a toluene solution (20 cm3) of complex 2a (0.120 g, 0.18 mmol) at 220 8C was added cis- [Pt(C6F5)2(thf)2] (0.122 g, 0.18 mmol).The cooling bath was then removed and the mixture stirred for 15 min. The resulting violet solution was subsequently filtered through a pad of Celite and concentrated (ca. 10 cm3). Addition of n-hexane (10 cm3) aVorded complex 8 as a violet crystalline solid (0.150, 70%) (Found: C, 49.85; H, 3.85. C49H45F10PPtS2SiTi requires C, 49.46; H, 3.81%). n& max/cm21 2157w, 2141m (C]] ] C), 800vs, 786vs (C6F5)X–sens.MS: m/z 1077 {[(h5-C5H4SiMe3)(h5-C5H4PPh2)- Ti(SC]] ] CBut)Pt(C6F5)2]1, 22}, 661 {[(h5-C5H4SiMe3)(h5-C5H4- PPh2)Ti(SC]] ] CBut)2]1, 55}, 548 {[(h5-C5H4SiMe3)(h5-C5H4- PPh2)Ti(SC]] ] CBut)]1, 100} and 434 {[(h5-C5H4SiMe3)(h5-C5H4- PPh2)Ti]1, 50%}. 1H NMR (CDCl3): at 250 8C, d 7.63 (m), 7.39–7.18 (m) (Ph), 7.01 (s, 2 H), 6.78, 6.68, 6.48, 6.16, 6.08, 5.90 (s, 1 H each) (C5H4), 1.20 (s, 9 H, But), 1.11 (s, 9 H, But) and 0.13 (s, 9 H, SiMe3); a similar pattern was observed at 20 8C with some of the C5H4 signals slightly displaced. 19F NMR [CDCl3, 3J(Pt–Fo)/Hz in parentheses]: at 20 8C, d 2116.6 [m (343), 1F], 2117.7 [dm (455), 1F], 2118.7 [d (414), 1F], 2120.0 [m, br (326), 1F] (Fo), 2162.7, 2163.4 (t, Fp), 2164.2 (m, 1F), 2164.5 (m, 1F), 2164.9 (m, 2F) (Fm); a similar pattern was observed at 250 8C. 31P-{1H} NMR (CDCl3): d 5.14 [s, C5H4- PPh2, 1J(Pt–P) = 2361 Hz].[(Á5-C5H4SiMe3)(SC]] ] CBut)Ti(Ï-Á5 :Í-P-C5H4PPh2(Ï-SC]] ] CBut)Pd(C6F5)2] 9. The synthesis was performed as described for complex 8 starting from 2a (0.15 g, 0.22 mmol) and cis-[Pd(C6F5)2(thf)2] (0.13 g, 0.22 mmol) (45%) (Found: C, 53.89; H, 4.19. C49H45F10PPdS2SiTi requires C, 53.44; H, 4.12%).n& max/cm21 2158w, 2141m (C]] ] C), 786vs, 776vs (C6F5)X–sens. MS: m/z 540 {[(h5-C5H4SiMe3)(h5-C5H4PPh2)Ti- (SC]] ] CBut)Pd 2 H]1, 42} and 434 {[(h5-C5H4SiMe3)(h5-C5H4- PPh2)Ti]1, 52%}. 1H NMR (CDCl3): at 250 8C, d 7.52–7.18 (m, Ph), 7.03, 7.00, 6.89, 6.71, 6.39 (s, 1 H each), 6.00 (s, 2 H), 5.86 (s, 1 H) C5H4), 1.22 (s, 9 H, But), 1.11 (s, 9 H, But) and 0.13 (s, 9 H, SiMe3); a similar pattern was observed at 20 8C. 19F NMRJ. Chem. Soc., Dalton Trans., 1998, 3199–3208 3207 Table 3 Crystal data and structure refinement for [Ti(h5-C5H4SiMe3)2(SC]] ] CBut)2] 1a and [(h5-C5H4SiMe3)(SC]] ] CBut)Ti(m-h5 :k-P-C5H4PPh2)- (m-SC]] ] CBut)Pt(C6F5)2] 8 Empirical formula Ma /Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/Mg m23 F(000) m/mm21 Crystal size/mm q Range for data collection/8 hkl Index ranges Reflections collected Independent reflections Data/restraints/parameters Goodness of fit on F2 R1, wR2 Final indices [I > 2s(I)] (all data) Largest diVerence peak and hole/e Å23 1a C28H44S2Si2Ti 548.83 11.470(6) 14.170(7) 21.405(11) 94.04(3) 104.40(3) 106.26(4) 3198(3) 4 1.140 1176 0.487 0.46 × 0.13 × 0.08 2.05 to 23.53 212 to 12, 215 to 15, 0–24 9746 9454 [R(int) = 0.1162] 6527/0/547 0.967 0.0902, 0.1122 0.2820, 0.1703 0.655 and 20.665 8 C57H60F10PPtS2SiTi 1307.22 12.904(2) 14.069(1) 18.090(2) 70.18(1) 71.42(1) 74.91(1) 2885.6(6) 2 1.505 1312 2.749 0.34 × 0.30 × 0.12 2.10 to 25.00 214 to 1, 215 to 15, 220 to 20 10495 9885 [R(int) = 0.0788] 8882/0/685 1.049 0.0672, 0.1538 0.1217, 0.2064 2.599 and 1.748 Details in common: l 0.71073 Å; triclinic, space group P1� ; full-matrix least-squares refinement on F2; R1 = S(|Fo| 2 |Fc|)/S|Fo|; wR2 = [Sw(Fo 2 2 Fc 2)2/ SwFo 2]� �� ; goodness of fit = [Sw(Fo 2 2 Fc 2)2/(Nobs 2 Nparam)]; w = [s2(Fo) 1 (g1P)2 1 g2P]21; P = [max(Fo 2, 0 1 2Fc 2)]/3.(CDCl3): at 20 8C, d 2114.1 (d, 1F), 2115.05 (d, 1F), 2115.6 (d, 1F), 2117.5 (m, 1F) (Fo), 2161.95, 2161.99 (overlapping of two triplets, 2Fp), 2163.4 (m), 2163.7 (m) (3F), 2164.1 (m, 1F) (Fm); a similar pattern was observed at 250 8C. 31P-{1H} NMR (CDCl3): d 10.93 (s, C5H4PPh2). X-Ray crystallography Complex 1a. Crystals of compound 1a suitable for X-ray analysis were grown from a saturated pentane solution at 220 8C. A deep brown needle-shaped crystal was fixed with epoxy on top of a glass fiber and transferred to the cold stream of the low temperature device of a Siemens STOE/AED2 automated four circle diVractometer.Crystal data and structure refinement parameters are listed in Table 3. Data were collected at 200 K by the q–2q method. Three check reflections measured at regular intervals showed no loss of intensity at the end of data collection.An empirical absorption correction based on y scans was applied (maximum and minimum transmission factor3, 0.841). The structure was solved by the Patterson method. All non-hydrogen atoms were located in succeeding Fourier diVerence syntheses and refined with anisotropic thermal parameters. Hydrogen atoms were added at calculated positions and assigned isotropic displacement parameters equal to 1.2 or 1.5 times the Uiso value of their respective apparent carbon atoms.Two molecules of the compound were found per asymmetric unit. There was no electron density higher than 1 e Å23 in the final map. Complex 8?0.5 n-hexane?toluene. Suitable crystals of complex 8?0.5 n-hexane?toluene were obtained by slow diVusion of hexane into a toluene solution of 8 at 20 8C.A dark red crystal was mounted in inert oil on top of a glass fiber and transferred to the cold stream of the low temperature device of a Siemens P4 automated four circle diVractometer. Crystal data and structure refinement parameters are listed in Table 3. Cell constants were calculated from 50 well centered reflections with 2q angles ranging from 23 to 268.Data were collected at 173 K by the q–2q method. Three check reflections measured at regular intervals showed no significant loss of intensity at the end of data collection. The data were treated (maximum and minimum transmission factors 0.983 and 0.680) and the structure solved and refined as above. Regions of electron density located at non-bonding distances were modelled as interstitial solvent and refined with anisotropic displacement parameters.In total, there were a quarter of a molecule of n-hexane and a molecule of toluene per formula unit. Three carbon atoms, refined at half occupancy, were found for the hexane molecule, three other carbon atoms being generated by symmetry. The toluene molecule was found in two regions, with half occupancy in each and with the molecule disordered over a symmetry center.There were four peaks of electron density higher than 1 e Å23 in the final map, three located very close to the platinum atom having no chemical meaning and the other in the solvent area. All calculations were carried out using the program SHELXL 93.31 CCDC reference number 186/1093. Acknowledgements We thank the Dirección General de Investigación Científica y Técnica (Spain) (Projects PB93-0250 and PB95-0003-C02- 01-02) and the University of La Rioja (Project API-98/B16) for financial support.References 1 G. L. GeoVroy and D. A. Roberts, in Comprehensive Organometallic Chemistry, eds. G. Wilkinson, F. G. A. Stone and E. W. Abel, Pergamon, Oxford, 1982, vol. 6, ch. 40; D. W. Stephan, Coord.Chem. Rev., 1989, 95, 41; B. Cornils and W. A. Herrmann (Editors), Applied Homogeneous Catalysis with Organometallic Compounds, VCH, Weinheim, 1996, vols. 1 and 2 and refs. therein; B. Bosch, G. Erker and R. Fröhlich, Inorg. Chim. Acta, 1998, 270, 446. 2 K. G. Anderson and M. Lin, Organometallics, 1988, 7, 2285; F. Ozawa, J. W. Park, P. B. Mackenzie, W. P. Schaefer, L. M. Henling and R.H. Grubbs, J. Am. Chem. Soc., 1989, 111, 1389; D. G. Dick and D. W. Stephan, Organometallics, 1990, 9, 1910; D. G. Dick, Z. Hou and D. W. Stephan, Organometallics, 1992, 11, 2378; R. Choukroun, F. Dahan, D. Gervais and C. Rifaï, Organometallics, 1990, 9, 1982; D. Selent, R. Beckhaus and T. Bartik, J. Organomet. Chem., 1991, 405, C15; P. Y. Zheng, T. T. Nadasdi and D. W.Stephan, Organometallics, 1988, 8, 1393.3208 J. Chem. Soc., Dalton Trans., 1998, 3199–3208 3 (a) G. S. White and D. W. Stephan, Inorg. Chem., 1985, 24, 1499; (b) T. A. Wark and D. W. Stephan, Inorg. Chem., 1987, 26, 363; (c) G. S. White and D. W. Stephan Organometallics, 1987, 6, 2169; (d ) G. S. White and D. W. Stephan, Organometallics, 1988, 7, 903; (e) T. A. Wark and D.W. Stephan, Organometallics, 1989, 8, 2836; ( f ) T. A.Wark and D. W. Stephan, Inorg. Chem., 1990, 29, 1731; ( g) T. T. Nadasdi and D. W. Stephan, Organometallics, 1992, 11, 116; (h) Y. Huang, R. J Drake and D. W. Stephan, Inorg. Chem., 1993, 32, 3022. 4 (a) E. Delgado, J. Forniés, E. Hernández, E. Lalinde, N. Mansilla and M. T. Moreno, J. Organomet. Chem., 1995, 494, 261; (b) U.Amador, E. Delgado, J. Forniés, E. Hernández, E. Lalinde and M. T. Moreno, Inorg. Chem., 1995, 34, 5279; (c) I. Ara, E. Delgado, J. Forniés, E. Hernández, E. Lalinde, N. Mansilla and M. T. Moreno, J. Chem. Soc., Dalton Trans., 1996, 3201. 5 J. Forniés and E. Lalinde, J. Chem. Soc., Dalton Trans., 1996, 2587; W. Beck, B. Niemer and M. Wieser, Angew. Chem., Int. Ed. Engl., 1993, 32, 923; S.Lotz, P. H. Van Rooyen and R. Meyer, Adv. Organomet. Chem., 1995, 37, 219 and refs. therein; R. Nast, Coord. Chem. Rev., 1982, 47, 89; A. J. Carty, Pure Appl. Chem., 1982, 54, 113; M. I. Bruce, Pure Appl. Chem., 1986, 58, 553; 1990, 6, 1021; P. R. Raithby and M. J. Rosales, Adv. Inorg. Chem. Radiochem., 1985, 29, 169; E. Sappa, A. Tiripicchio and P. Braunstein, Coord. Chem.Rev., 1985, 65, 219; P. N. V. Pavan Kumar and E. D. Jemmis, J. Am. Chem. Soc., 1988, 110, 125. 6 (a) J. R. Berenguer, L. R. Falvello, J. Forniés, E. Lalinde and M. Tomás, Organometallics, 1993, 12, 6; (b) J. R. Berenguer, J. Forniés, E. Lalinde and A. Martín, Angew. Chem., Int. Ed. Engl., 1994, 33, 2083; (c) H. Lang, K. Köhler and S. Blau, Coord. Chem. Rev., 1995, 143, 113 and refs.therein; (d ) S. Back, H. Printzkow and H. Lang, Organometallics, 1998, 17, 41; (e) K. Köhler, S. J. Silverio, I. Hyla-Kryspin, R. Gleiter, L. Zsolnai, A. Driess, G. Huttner and H. Lang, Organometallics, 1997, 16, 4970; ( f ) M. D. Janssen, K. Köhler, M. Herres, A. Dedieu, W. J. J. Smeets, A. L. Spek, D. M. Grove, H. Lang and G. van Koten, J. Am. Chem. Soc., 1996, 118, 4817. 7 D. G. Sekutowski and G. D. Stucky, J. Am. Chem. Soc., 1976, 981, 1376; T. M. Cuenca, R. Gómez, P. Gómez-Sal, G. M. Rodríguez and P. Royo, Organometallics, 1992, 11, 1229; W. J. Evans, R. A. Keyer and J. W. Ziller, Organometallics, 1993, 12, 2618; C. M. Forsyth, S. P. Nolan, C. L. Stern, T. J. Marks and A. L. Rheingold, Organometallics, 1993, 12, 3618; R. Duchateau, C. T. van Wee and J.H. Teuben, Organometallics, 1996, 15, 2291; T. Takahashi, Z. Xi, Y. Obora and N. Suzuki, J. Am. Chem. Soc., 1995, 117, 2665; P. D. Hsu, W. M. Davis and S. L. Buchwald, J. Am. Chem. Soc., 1993, 115, 10 394. 8 S. Pulst, P. Arndt, B. Heller, W. Baumann, R. Kempe and H. Rosenthal, Angew. Chem., Int. Ed. Engl., 1996, 35, 1112 and refs. therein; V. V. Burlakov, A. OhV, C. Lefeber, A.Tillack, W. Bauman, R. Kempe and U. Rosenthal, Chem. Ber., 1995, 128, 967; V. Varga, K. Mach, J. Hiller, U. Thewalt, P. Sedmera and M. Polásek, Organometallics, 1995, 14, 1410; U. Rosenthal, S. Pulst, P. Arndt, A. OhV, A. Tillack, W. Baumann, R. Kempe and V. V. Burlakov, Organometallics, 1995, 14, 2961 and refs. therein. 9 (a) W. Weigand, Z. Naturforsch., Teil B, 1991, 46, 1333; (b) W.Weigand and C. Robl, Chem. Ber., 1993, 126, 1807; (c) W. Weigand, M. Weishäupl and C. Robl, Z. Naturforsch., Teil B, 1996, 51, 501. 10 (a) T.-Y. Lee and A. Mayr, J. Am. Chem. Soc., 1994, 116, 10 300; (b) A. F. Hill and J. M. Malget, Chem. Commun., 1996, 1177. 11 C. Rosenberg, N. Steunou, S. Jeannin and Y. Jeannin, J. Organomet. Chem., 1995, 494, 17. 12 D. C. Miller and R. J.Angelici, Organometallics, 1991, 10, 79, 89. 13 (a) J. Meijer and L. Brandsma, Recl. Trav. Chim. Pays-Bas, 1971, 97, 1098; (b) L. Brandsma and H. D. Verkruijsse, Synthesis of Acetylenes, Allenes and Cumulenes, Elsevier, Amsterdam, 1981. 14 (a) E. Delgado, E. Hernández, A. Hedayat, J. Tornero and R. Torres, J. Organomet. Chem., 1994, 466, 119; (b) E. Delgado, M. A. García, E. Hernández, N. Mansilla, L. A. Martínez-Cruz, J. Tornero and R. Torres, J. Organomet. Chem., in the press. 15 (a) J. Forniés, M. A. Gómez-Saso, E. Lalinde, F. Martínez and M. T. Moreno, Organometallics, 1992, 11, 2873; (b) J. Forniés, E. Lalinde, A. Martín and M. T. Moreno, J. Chem. Soc., Dalton Trans., 1994, 135; (c) J. R. Berenguer, J. Forniés, E. Lalinde and F. Martínez, Organometallics, 1996, 15, 4537; (d ) I. Ara, L. R. Falvello, S. Fernández, J. Forniés, E. Lalinde, A. Martín and M. T. Moreno, Organometallics, 1997, 16, 5923; (e) J. R. Berenguer, J. Forniés, E. Lalinde and F. Martínez, J. Chem. Soc., Chem. Commun., 1995, 1227. 16 (a) M. Herres, O. Walter, H. Lang, R. Hosch and J. Hahn, J. Organomet. Chem., 1994, 466, 237; (b) G. Schmidt, N. Schittenhelm and U. Behrens, J. Organomet. Chem., 1995, 496, 49; (c) G. Schmidt and U. Behrens, J. Organomet. Chem., 1995, 503, 101; (d ) D. C. Miller and R. J. Angelici, J. Organomet. Chem., 1990, 394, 235. 17 M. Y. Darensbourg, M. Pala, S. A. Houliston, K. P. Kidwell, D. Spencer, S. S. Chojnacki and J. H. Reibenspies, Inorg. Chem., 1992, 31, 1487; C. J. RuYng and T. B. Rauchfuss, Organometallics, 1985, 4, 524; P. S. Braterman, V. A. Wilson and K. K. Joshi, J. Chem. Soc. A, 1971, 191. 18 G. Natile, L. Maresca and G. Bor, Inorg. Chim. Acta, 1977, 23, 37. 19 E. Delgado, M A. García, E. Gutiérrez-Puebla, N. Mansilla and F. Zamora, unpublished work. 20 M. J. Calhorda, M. A. A. F. de C. T. Carrondo, A. R. Dias, C. F. Frazão, M. B. Hursthouse, J. A. M. Simões and C. Teixeira, Inorg. Chem., 1988, 27, 2513. 21 R. Usón, J. Forniés, M. A. Usón, M. Tomás and M. A. Ibáñez, J. Chem. Soc., Dalton Trans.,1994, 401. 22 M. Capdevila, W. Clegg, P. González-Duarte, B. Harris, I. Mira, J. Sola and I. C. Taylor, J. Chem. Soc., Dalton Trans., 1992, 2817. 23 E. G. Muller, S. F. Watkins and L. F. J. Dahl, J. Organomet. Chem., 1976, 14, 73. 24 D. F. Shriver and M. A. Drezdon, The Manipulation of Air Sensitive Compounds, 2nd edn., Wiley, New York, 1986. 25 D. D. Perrin, W. L. F. Armarengo and D. R. Perrin, Purification of Laboratory Chemicals, 2nd edn., Pergamon, Oxford, 1980. 26 M. F. Lappert, C. Pickett, P. I. Riley and P. I. W. Yarrow, J. Chem. Soc., Dalton Trans., 1981, 805. 27 J. C. Leblanc, C. Moise, A. Maisonnat, R. Poilblanc, C. Charrier and F. Mathey, J. Organomet. Chem., 1982, 231, C43. 28 R. B. King, Organomet. Synth., 1965, 1. 29 D. P. Tate, W. R. Knipple and J. M. Augl, Inorg. Chem., 1962, 1, 433. 30 R. Usón, J. Forniés, M. Tomás and B. Menjón, Organometallics, 1985, 4, 1912. 31 G. M. Sheldrick, SHELXL 93, a FORTRAN 77 program for crystal structure determination from diVraction data, University of Göttingen, 1993. Paper 8/03806F

ISSN:1477-9226    

DOI:10.1039/a803806f

出版商:RSC    

年代:1998

数据来源: RSC