摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1998, Pages 2289–2291 2289 Novel redox-active ruthenium cluster crown compounds capable of host–guest chemistry Douglas S. Shephard,*,†,a Brian F. G. Johnson,*,a Justin Matters a and Simon Parsons b a Department of Chemistry, The University of Cambridge, Lensfield Road, Cambridge, UK CB2 1EW b Department of Chemistry, The University of Edinburgh, West Mains Road, Edinburgh, UK EH9 3JJ The new redox-active cluster compounds [Ru6C(CO)14(h6- C6H4C8H16O5)] and [Ru6C(CO)14(h6-C6H4C10H20O6)] have been synthesised by direct reaction of the aryl crown ether with the hexaruthenium carbido cluster [Ru6C(CO)17] and display host–guest type behaviour with cations Na1 and NH4 1 both spectroscopically and electrochemically.In continuation of our development of the chemistry of the ruthenium arene clusters based on the Ru6C unit,1 we have examined ways of incorporating these redox active units 2 into supramolecular assemblies.Given the wealth of host–guest chemistry involving crown compounds3 we elected to concentrate our initial investigations into this area with a view to examining the molecular sensor behaviour of the new materials. Furthermore, incorporation of such recognition characteristics (e.g. for RNH3 1) in the cluster unit, may lead to several applications wherein these metal-rich domains may be directed on surfaces or large biomolecules with appropriate functionality on the nanoscale.Herein we would like to report: (i) the synthesis and characterisation of two new crown ether cluster compounds [Ru6C(CO)14(h6-C6H4C8H16O5)] 2 and [Ru6C(CO)14- (h6-C6H4C10H20O6)] 3 containing an electronic link between the redox-active cluster core and the receptor site; (ii) the structural characterisation of the model host–guest complex [Ru6C(CO)14- (h6-C6H4C10H20O6)]?NH4PF6 3?NH4 1 in the solid state, thus establishing the host capabilities of 3; (iii) the pronounced redox activity of 3 alone and in the presence of guests Na1 and NH4 1.Our synthetic approach to the target compounds is novel for arene cluster chemistry in that it involves the direct thermolytic reaction of the fully aromatised arene with the preformed cluster [Ru6C(CO)17] 1 (cf. ref. 1 for previous synthetic routes). Thermolysis of 1 in the presence of excess benzo-15-crown-5 or benzo-18-crown-6 in di-n-butyl ether for 4.5 h gave a dark redbrown solution together with a small amount of dark precipitate.After cooling, the reaction mixture was initially passed through a short silica column using dichloromethane as eluent and the resulting brown-red solution reduced to dryness in vacuo. After separation by TLC the fully developed plates showed three distinct bands corresponding to unreacted 1, the brown compounds 2 or 3 and unstable red materials 2a or 3a. Preliminary characterisation ‡ of 2 and 3 was achieved on the † E-Mail: dss1001@cus.cam.ac.uk ‡ Spectroscopic data for 2.IR (CH2Cl2/cm21): n(CO) 2074m, 2030w (sh), 2022vs, 1979w (br), 1808w (br); n(COC) 1269s, 1257s. 1H NMR (CDCl3): d 5.66 (m, 2 H), 5.02 (m, 2 H), 3.5–4.2 (m br, 16 H) (Found: Ru6C(CO)14C6H4C18H16O5?CH2Cl2: C, 28.05; H, 1.66. Calc. for 2?CH2Cl2: C, 28.04; H, 1.71%). Spectroscopic data for 3. IR (CH2Cl2/cm21): n(CO) 2073m, 2045w (sh), 2031w (sh), 2022vs, 1982w (br), 1811w (br); n(COC) 1272s, 1258s. 1H NMR (CDCl3): d 5.65 (m, 2 H), 5.02 (m, 2 H), 3.5–4.2 (m br, 20 H) (Found: Ru6C(CO)14C6H4C10H20O6?CH2Cl2: C, 27.3; H, 1.86.Calc. for 3?CH2Cl2: C, 27.5; H, 1.87%). basis of their solution infrared spectra in dichloromethane, which showed absorptions in the nCO region typical of an h6-coordinated arene together with absorptions in the nether region. Positive-ion fast atom bombardment mass spectra revealed parent ions (Found: m/z = 1280. Calc. for 2: m/z = 1279. Found: m/z = 1324. Calc. for 3: m/z 1323) and ions corresponding to the sequential loss of up to 14 carbonyl fragments.An additional peak was found at m/z = 1302 for 2 and is attributable to a Na1 adduct.§ The fractions 2a and 3a were isolated in low yields and in each case appeared to consist of several isomers (shown by broad bands on the TLC plate and complex IR spectra). These materials are unstable and readily disproportionate in dry degassed solvents at ambient temperature to starting materials and product. Bearing in mind the possible co-ordinative diversity and relative instability of the Ru0]OR2 bond, it is tempting to suggest that 2a and 3a are intermediates wherein the cluster is co-ordinated to the ligand via the ether oxygens.Compounds of a similar nature have been isolated in well established reactions of Group 6 metal carbonyls with triglyme [CH3- (OCH2CH2)3OCH3]. The solid-state molecular structure of 3?NH4 1 is shown in Fig. 1 along with some key structural parameters as determined by single-crystal X-ray diVraction.¶ The asymmetric unit of the host–guest complex 3?NH4 1 contains one molecule. The arene ring is co-ordinated h6 to a single ruthenium atom, and the interstitial carbon is displaced slightly from the centre of the ruthenium octahedron towards the arene ring.There is one bridging carbonyl ligand, spanning two ruthenium atoms adjacent to the arene ring. The oxygen atoms in the macrocyclic ring remain approximately planar (mean deviation 0.186 Å). This plane lies at an angle of 27.28 to the arene ring which bonds the macrocycle to the cluster.The catecholic oxygen conformational characteristics tend to suggest that the lone pairs on the oxygens interact with some arene carbons, thereby, producing an electronic link 4 between the receptor site and the redox-active cluster core. The ammonium ion can be clearly seen to act as a guest within the macrocyclic cavity. The nitrogen atom lies 0.805(16) Å above the average plane of the oxygen atoms, with three of the hydrogen atoms (refined riding on the N atom which was § This would require scavenging of the alkali metal from some part of the work up.The silica plates used for the separation, for instance, have an appreciable sodium content (up to 200 ppm according to the suppliers Merck). In a further experiment solutions of 2 and 3 were treated with several alkali metal (Na1, K1 or Cs1) and ammonium salts (NH4 1 or ButNH3 1), after separation FAB1 mass spectra were taken, which revealed peaks in each case corresponding to host–guest pairing ¶ Crystal data for 3?NH4 1: C34.5H32F6NO20Ru6, M = 1532.00, monoclinic, space group P21/c, a = 9.608(2), b = 17.071(2), c = 28.633(4) Å, b = 95.806(14)8, U = 4672.2(12) Å3, Z = 4, m = 2.024 mm21, T = 220(2) K.R1 [= F > 4s(F)] (5460 data) and wR2 (all data) are equal to 0.0545 and 0.1223, respectively. The H atoms were placed in calculated positions and allowed to refine riding on their respective atoms.The lattice contains a disordered toluene solvate molecule (one half per asymmetric unit in two orientations with equal occupancy). CCDC reference number 186/1031.2290 J. Chem. Soc., Dalton Trans., 1998, Pages 2289–2291 Fig. 1 The solid-state molecular structure of 3?NH4 1 with important interatomic distances (Å): Ru(1)]C(1m) 2.288(9), O(7m)]N(1) 2.94(2), Ru(1)]C(2m) 2.246(9), O(10m)]N(1) 2.84(2), Ru(1)]C(3m) 2.216(10), O(13m)]N(1) 2.80(2), Ru(1)]C(4m) 2.203(10), O(16m)]N(1) 2.79(2), Ru(1)]C(5m) 2.255(9), O(19m)]N(1) 2.87(2), Ru(1)]C(6m) 2.291(9), O(22m)]N(1) 3.02(2), C(1m)]C(2m) 1.397(13), N(1)]F(1) 3.05(2), C(2m)]C(3m) 1.420(12), N(1)]F(3) 3.07(2), C(3m)]C(4m) 1.416(14), N(1)]F(6) 2.96(2), C(4m)]C(5m) 1.388(14), C(5m)]C(6m) 1.390(14), C(6m)]C(1m) 1.417(13) allowed to rotate as a rigid body) interacting with two oxygen atoms each, and the remaining hydrogen directed towards the counter ion.The N atom is shifted significantly away from the slightly electron-deficient catecholic oxygens and toward the four remaining ether oxygens by ca. 0.15 Å, this is contrary to the observations of Truter and co-workers 5 who found roughly equal N ? ? ? O distances in a comparable system wherein the arene ring was not co-ordinated. The aliphatic methylene groups of the crown in 3?NH4 1 are orientated such that they encircle a carbonyl oxygen. This serves to support the normally conformationally mobile crown ether oxygens in an orientation resembling that observed in benzo-crown compounds bound to cationic guests.At the same time such an intramolecular interaction may enhance the selectivity of the host compound by inhibiting co-ordination with larger guests.6 An examination of the redox chemistry || of 3 [Fig. 2(a)] revealed an expected irreversible two-electron reductive wave at 20.91 V with a daughter peak on the oxidative sweep at 20.17 V. The peak shapes and relative intensities of both the twoelectron reductive and daughter peaks are very similar to those observed for the parent cluster [Ru6C(CO)17] and [Ru6C- || Electrochemical experiments were carried out using a DSL 286-D PC with General Purpose Electrochemical System (GPES) Version 3 software coupled to an Autolab system containing a PSTAT 10 potentiostat.A conventional three-electrode cell was employed with Pt counter and micro-working electrodes and Ag–AgCl reference electrode against which the Fc–Fc1 couple was measured at 10.55 V.All electrochemical experiments were performed in 0.5 M tetrabutylammonium tetrafluoroborate solution in CH2Cl2 under an atmosphere of dry oxygen-free argon. All cyclic voltammograms were recorded at a scan rate of 200 mV s21 save the experiment on 3 alone (50 mV s21). (CO)14R] (R = arene) in which concomitant CO expulsion is observed with a reductive process.2 The reduction product of 3 has also been synthesised by chemical reduction with sodium naphthalide and isolated as [Ru6C(CO)13(h6-C6H4C10H20- O6)]22.7 Precise addition of the salts NH4BF4 and NaBF4 to the electrolyte solution containing a known amount of 3 produced a marked change in the position of the reduction wave, in particular the appearance of a new peak significantly anodically shifted in both cases.Cyclic voltammograms for these two experiments are shown in Figs. 2(b)–2(d). The cyclic voltammograms obtained during the addition of the Na1 ion to the electrolyte show the gradual evolution of a second peak at 20.66 V which corresponds to a large anodic shift (ca. 250 mV) of the reduction potential of 3 [Fig. 2(b)]. This may be ascribed to the formation of a simple host–guest complex between the cluster’s pendant crown ether and the alkali metal cation. The corresponding experiment wherein ammonium ions were added to the electrolyte also gave a new peak (20.65 V) corresponding to a host–guest complex of the 3?NH4 1 type when the ratio of the two components was unity or more [Fig. 2(c)]. However, when the total amount of ammonium ions was less than that of the host cluster 3 a second new peak was observed at 20.78 V [see Fig. 2(d)] which may correspond to a dimer in which two crown ether moieties complex a single ammonium ion.** In this work we have been able to show that hexaruthenium ** In a similar experiment to those detailed above using the cluster 1 no significant shift in the reduction potential was observed on addition of either alkali metal or ammonium cations. This serves to show that the observed electrochemical behaviour of 3 in the presence of NH4 1/Na is a direct consequence of host–guest type complex formation in solution.J.Chem. Soc., Dalton Trans., 1998, Pages 2289–2291 2291 clusters may be derivatised such that they may participate in host–guest chemistry. We are presently undertaking a range of related experiments to elucidate the nature of the host–guest Fig. 2 Cyclic voltammogram of (a) 3, (b) 3 in the presence of 0.6 equivalent NH4BF4, (c) 3 in the presence of 1.0 equivalent NH4BF4, and (d) 3 in the presence of 0.8 equivalent NaBF4 interactions (including binding constants and selectivity) with these and other cluster derivatised compounds in solution and at surfaces with a view to constructing highly ordered nanostructures containing metal-rich domains. Acknowledgements We gratefully acknowledge the University of Cambridge, the University of Edinburgh and the EPSRC for financial support.We would also like to thank Dr. L. J. Yellowlees (Edinburgh University) for her assistance in processing the electrochemical data. References 1 D. Braga, P. J. Dyson, F. Grepioni and B. F. G. Johnson, Chem. Rev., 1994, 94, 1585. 2 S. R. Drake, B. F. G. Johnson and J. Lewis, J. Chem. Soc., Dalton Trans., 1989, 243; A. J. Blake, A. Harrison, B. F. G. Johnson, E. J. L. McInnes, S. Parsons, D. S. Shephard and L.J. Yellowlees, Organometallics, 1995, 14, 3160; S. R. Drake, Polyhedron, 1990, 9, 455. 3 J. M. Lehn, Supramolecular Chemistry Concepts and Perspectives, VCH, New York, 1995; R. M. Izatt, J. S. Bradshaw and K. Paulak, Chem. Rev., 1991, 91, 1721; B. P. Hay and J. R. Rustad, J. Am. Chem. Soc., 1994, 116, 6316; F. C. J. M. van Veggel, W. Verboom and D. N. Reinhoudt, Chem. Rev., 1994, 94, 1279; A. E. G. Cass (Editor), Biosensors: a Practical Approach, IRL Press, Oxford, 1990; P.D. Beer, Chem. Soc. Rev., 1989, 18, 409; P. D. Beer, H. Sikanyika, C. Blackburn, J. F. McAleer and M. G. B. Drew, J. Organomet. Chem., 1988, 356, C19; K. H. Pannell, D. C. Hambrick and G. S. Lewandos, J. Organomet. Chem., 1975, 99, C19; K. J. Odell, E. M. Hyde, B. L. Shaw and I. Shephard, J. Organomet. Chem., 1979, 168, 103. 4 C. E. Anson, C. S. Creaser and G. R. Stephenson, J. Chem. Soc., Chem. Commun., 1994, 2175. 5 J. A. Bandy, C. H. L. Kennard, D. G. Parsons and M. R. Truter, J.Chem. Soc., Perkin Trans. 2, 1984, 309. 6 R. D. Hancock and A. E. Martell, Chem. Rev., 1989, 89, 1875. 7 D. S. Shephard, J. M. Matters and B. F. G. Johnson, unpublished work. Received 24th April 1998; Communication 8/03084GJ. Chem. Soc., Dalton Trans., 1998, Pages 2289–2291 2291 clusters may be derivatised such that they may participate in host–guest chemistry. We are presently undertaking a range of related experiments to elucidate the nature of the host–guest Fig. 2 Cyclic voltammogram of (a) 3, (b) 3 in the presence of 0.6 equivalent NH4BF4, (c) 3 in the presence of 1.0 equivalent NH4BF4, and (d) 3 in the presence of 0.8 equivalent NaBF4 interactions (including binding constants and selectivity) with these and other cluster derivatised compounds in solution and at surfaces with a view to constructing highly ordered nanostructures containing metal-rich domains. Acknowledgements We gratefully acknowledge the University of Cambridge, the University of Edinburgh and the EPSRC for financial support.We would also like to thank Dr. L. J. Yellowlees (Edinburgh University) for her assistance in processing the electrochemical data. References 1 D. Braga, P. J. Dyson, F. Grepioni and B. F. G. Johnson, Chem. Rev., 1994, 94, 1585. 2 S. R. Drake, B. F. G. Johnson and J. Lewis, J. Chem. Soc., Dalton Trans., 1989, 243; A. J. Blake, A. Harrison, B. F. G. Johnson, E. J. L. McInnes, S. Parsons, D. S. Shephard and L. J. Yellowlees, Organometallics, 1995, 14, 3160; S. R. Drake, Polyhedron, 1990, 9, 455. 3 J. M. Lehn, Supramolecular Chemistry Concepts and Perspectives, VCH, New York, 1995; R. M. Izatt, J. S. Bradshaw and K. Paulak, Chem. Rev., 1991, 91, 1721; B. P. Hay and J. R. Rustad, J. Am. Chem. Soc., 1994, 116, 6316; F. C. J. M. van Veggel, W. Verboom and D. N. Reinhoudt, Chem. Rev., 1994, 94, 1279; A. E. G. Cass (Editor), Biosensors: a Practical Approach, IRL Press, Oxford, 1990; P. D. Beer, Chem. Soc. Rev., 1989, 18, 409; P. D. Beer, H. Sikanyika, C. Blackburn, J. F. McAleer and M. G. B. Drew, J. Organomet. Chem., 1988, 356, C19; K. H. Pannell, D. C. Hambrick and G. S. Lewandos, J. Organomet. Chem., 1975, 99, C19; K. J. Odell, E. M. Hyde, B. L. Shaw and I. Shephard, J. Organomet. Chem., 1979, 168, 103. 4 C. E. Anson, C. S. Creaser and G. R. Stephenson, J. Chem. Soc., Chem. Commun., 1994, 2175. 5 J. A. Bandy, C. H. L. Kennard, D. G. Parsons and M. R. Truter, J. Chem. Soc., Perkin Trans. 2, 1984, 309. 6 R. D. Hancock and A. E. Martell, Chem. Rev., 1989, 89, 1875. 7 D. S. Shephard, J. M. Matters and B. F. G. Johnson, unpublished work. Received 24th April 1998; Communication 8/03084G

ISSN:1477-9226    

DOI:10.1039/a803084g

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2293–2301 2293 The synthesis of C[Si(CH3)2X]3SiX3 compounds (X 5 H, Cl, Br and OH) and the molecular structure of C[Si(CH3)2H]3SiH3 in the gas phase; a study by electron diVraction and ab initio molecular orbital calculations † Carole A. Morrison,a David W. H. Rankin,*a Heather E. Robertson,a Paul D. Lickiss b and Phindile C. Masanganeb ‡ a Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh, UK EH9 3JJ b Chemistry Department, Imperial College of Science, Technology and Medicine, London, UK SW7 2AY Received 13th April 1999, Accepted 3rd June 1999 A series of new hexafunctional tetrasilylmethane derivatives, C[Si(CH3)2X]3SiX3, have been prepared.In addition a comprehensive structural study of the parent compound C[Si(CH3)2H]3SiH3 has been undertaken. A full search of the potential energy surface has been performed ab initio. Results indicate the presence of a total of eleven distinct conformational minima with a total energy range of only 3 kJ mol21, with evidence for appreciable energy barriers to internal rotation.The analysis of gas-phase electron diVraction data has been undertaken for this eleven-conformer model and a satisfactory fit has been obtained. Introduction There has been considerable interest in recent years in the chemistry and structures of tetrasilylmethane derivatives as the steric hindrance at silicon centres in such compounds can lead to unusual reactivity and novel structural features.1,2 The majority of tetrasilylmethanes contain only one or two silicons with substituents other than methyl groups at which reactions can be relatively easily carried out.For example, compounds of the type C[Si(CH3)3]3SiRR9X and C[Si(CH3)3]2(SiRR9X)[Si- (CH3)2Y] (where R and R9 = alkyl or aryl; X and Y = H, halide, or pseudohalide) have been of particular interest from a mechanistic point of view.3 Several studies have led to an interest in compounds in which all four of the silicons bear reactive substituents, for example C(SiH3)4 has been proposed as a precursor to silicon-containing thin films via chemical vapour deposition 4 and C[Si(CH3)2H]4 has been used as a potential free-radical reducing agent.5 The work presented in this paper consists of two parts.First, as part of our continuing interest in polyfunctional tetrasilylmethanes the synthesis of a range of compounds C[Si(CH3)2- X]3SiX3 (where X = H, Cl, Br, and OH) is now reported.Secondly, extensive ab initio calculations over the full potential energy surface for the simple hexahydridosilane C[Si(CH3)2- H]3SiH3 have been performed, in addition to the determination of its gas-phase structure by electron diVraction. Results and discussion Synthesis and characterisation The synthetic route to C[Si(CH3)2X]3SiX3 species is outlined in Fig. 1. The cleavage of Si–CH3 bonds by iodine monochloride has been used previously in the preparation of C[Si(CH3)2Cl]4 † Supplementary data available: GED data analysis model and coordinates for 11 conformers of C[Si(CH3)2H]3SiH3; Brookhaven (.pdb) files of 11 conformers.For direct electronic access see http://www.rsc.org/ suppdata/dt/1999/2293/. ‡ Present address: Chemistry Department, University of Swaziland, P/Bag 4, Kwaluseni, Swaziland, Southern Africa. from C[Si(CH3)3]4,6 and the reaction between C[Si(CH3)3]3SiCl3 and ICl proceeds cleanly to give the hexachloride C[Si- (CH3)2Cl]3SiCl3 as an air-stable, white crystalline solid.A similar reaction of C[Si(CH3)3]3SiCl3 with IBr was found to be much slower. 1H NMR spectroscopy showed that although a stepwise reaction to give C[Si(CH3)2Br][Si(CH3)3]2SiCl3, C[Si- (CH3)2Br]2[Si(CH3)2]SiCl3 and C[Si(CH3)2Br]3SiCl3 appeared to be occurring, the reaction was incomplete after eight weeks. None of the mixed halogen-containing compounds was isolated. Clean reduction of the hexachloride proved diYcult to achieve. The common method for the reduction of chlorosilanes containing the bulky C[Si(CH3)3]3 group,7 (i.e.reduction using LiAlH4 in THF) led to a mixture of the required hexahydrido compound together with the trisilylmethane C[Si(CH3)2H]3H [which has also been prepared from the reaction between CHBr3, Mg and Si(CH3)2ClH] 8 by cleavage of the central C–SiX3 bond and other unidentified compounds. Similar C–Si bond cleavage is seen in the reaction of C[Si(CH3)2Ph]3SiCl3 with LiAlH4, which aVords C[Si(CH3)2Ph]3H,9 and the formation of C[(SiH3)3]2 in the reduction of C(SiBrH2)4 by LiAlH4.4 This presumably reflects the stabilising eVect of silyl groups a to a carbanion and hence the good leaving group ability of trisilylmethyl ions.Variation of the reaction time and the amount Fig. 1 Synthetic routes to new C[Si(CH3)2X]3SiX3 compounds.2294 J. Chem. Soc., Dalton Trans., 1999, 2293–2301 of LiAlH4 used did not allow a clean product to be formed but the use of a LiAlH4/toluene two-phase system containing benzyltriethylammonium chloride as a phase transfer catalyst,10 a method which has successfully been used in the synthesis of C(SiH3)4 from C(SiBrH2)4,4 did allow C[Si(CH3)2H]3SiH3 to be prepared in good yield as a highly volatile solid. The IR spectrum of C[Si(CH3)2H]3SiH3 shows two strong bands in the Si–H stretching region at 2110 cm21 and 2140 cm21, which were assigned to the tertiary Si–H and primary Si–H stretches respectively by comparison with frequencies of 2090 cm21 and 2138 cm21 observed for C[Si(CH3)2H]3H8 and C[Si(CH3)3]3- SiH3,11 respectively.The 29Si NMR data for C[Si(CH3)2H]3SiH3 are also in good agreement with those of C[Si(CH3)2H]3H and C[Si(CH3)3]3SiH3. The SiH3 signal is a quartet at d 261.3 with 1JSi–H = 200 Hz while that of C[Si(CH3)3]3SiH3 is also centred at d 261.3 with a coupling constant of 198 Hz. The Si(CH3)2H signal of C[Si(CH3)2H]3SiH3 is a doublet at d 213.8 with 1JSi–H = 188 Hz, comparing well with d 215.5 and a coupling constant of 185 Hz found in the Si(CH3)2H signal of C[Si(CH3)2H]3H.8 The proton coupled 29Si NMR spectrum of C[Si(CH3)2H]3SiH3 revealed further couplings, the SiH3 group split into a quartet (3JSi–H = 3.5 Hz) by the three equivalent Si(CH3)2H hydrogens, and the Si(CH3)2H signal, split by the three equivalent SiH3 hydrogens (3JSi–H = 3.5 Hz) (as well as by the coupling to the CH3 groups).The bromination of C[Si(CH3)2H]3SiH3 with a 1.0 M solution of Br2 in CCl4 was found to be rapid and gave the expected hexabromide C[Si(CH3)2Br]3SiBr3 in near quantitative yield. The use of a deficiency of Br2 allowed the intermediate bromides to be identified by 1H NMR spectroscopy, which showed that initial bromination occurred at the Si(CH3)2H groups, and that C[Si(CH3)2Br]3SiH3 formed as a significant intermediate in the reaction. Unfortunately, the partially brominated compounds could not be isolated individually. The faster bromination of the more sterically hindered tertiary hydrides, Si(CH3)3H, than the primary hydrides, SiH3, is consistent with the results of El-Durini and Jackson who found the rates of bromination of Si(C2H5)3H, Si(CH3)3H and Si(C6H5)H3 to be 13.4, 1.10 and 0.65 mol21 s21 respectively.12 In such reactions it was concluded that steric eVects are relatively unimportant, but that it is the presence of electron-releasing groups, increasing positive charge build-up at silicon during the reaction that increases the rate.The novel highly brominated silane C[Si(CH3)2Br]3SiBr3 was isolated as colourless crystals. As in the case of C[Si(CH3)3]3SiBr3, it is air stable, a very unusual property for a bromosilane, which would normally be expected to undergo rapid hydrolysis with atmospheric moisture to give HBr, together with silanols and siloxanes. Although air stable, the hexabromide is hydrolysed in aqueous acetone solution to give a mixture of unidentified siloxanes, and the partially brominated species, which are less bulky, react readily with atmospheric moisture, also to give unidentified siloxanes.At room temperature the 1H NMR signal (270 MHz) of C[Si(CH3)2Br]3SiBr3 was found to be a broad resonance between d 1.1 and 1.2. This suggested that a dynamic process was occurring and that hindered rotation around the central Si–C bonds at this temperature was relatively slow.The temperature dependence of the 1H NMR spectrum was then studied between 70 8C and 270 8C on a 270 MHz spectrometer and at 70 8C the spectrum, as expected, showed a sharp singlet at d 1.22 which began to split unsymmetrically into two main peaks as the temperature of the sample was lowered. Expansion of the signal recorded at 270 8C showed the presence of two main resonances in an approximate 1 : 1 ratio, together with about six smaller ones. This is consistent with the presence of one predominant conformer in which the methyl groups within a single Si(CH3)2Br group are inequivalent but the three Si(CH3)2Br groups overall are equivalent.The presence of several smaller resonances is consistent with the presence of at least one less symmetrical conformer. The freezing out of more than one conformer in tetrasilylmethanes at low temperature has also been observed for the related compounds C[Si(CH3)2I]4 and C[Si(CH3)2Br]4,13 and other tetrasilylmethane derivatives have been shown to exhibit hindered rotation at temperatures readily accessible by NMR spectroscopy. 14,15 Using the Eyring equation 16 with values of 298 ± 5 8C for the coalescence temperature and 40 Hz for the separation between the two main peaks at low temperature, an activation energy of 63.5 ± 1 kJ mol21 for methyl group exchange was calculated. Similar studies for C[Si(CH3)3]3SiBr3 have shown two methyl group exchange processes to occur between room temperature and 270 8C with free energies of activation calculated to be 45 ± 0.5 and 52.2 ± 0.5 kJ mol21.15 The related trichlorosilane C[Si(CH3)2Ph]3SiCl3 also shows hindered rotation within a conformer at low temperature, also having two distinct methyl group resonances whose exchange has a free energy of activation of 53.1 kJ mol21 at 210 8C.15 Interestingly, the 29Si NMR spectra of C[Si(CH3)2Br]3SiBr3 showed no resonances when recorded at room temperature at 53.6 MHz and at 99.6 MHz but at 50 8C signals were as expected: a singlet at d 18.66 which was assigned to Si(CH3)2Br and another singlet at d 236.59 assigned to SiBr3.Again this can be attributed to hindered rotation about the Si–C bonds at room temperature. X-Ray diVraction studies on the only crystals of C[Si- (CH3)2Br]3SiBr3 that could be obtained show a twofold disorder of the Si4 tetrahedron generating a “cube” of eight half Si atoms with a carbon atom at the centre. The six bromine atoms are disordered over twelve sites (eVectively bridging the edges of the cube) and obscuring the positions of the methyl groups which must alternate with them in the crystal.This model is consistent with the presence of many isomers co-crystallising or with a single isomer in eight diVerent orientations randomly distributed through the crystal and superimposed to give the observed image.17 Although C[Si(CH3)2H]3SiH3 is a potentially useful precursor to dendritic organosilanes it would also be useful to have precursors containing Si–O functions, particularly silanols, that could be used as cores for dendrimers.Silanes are readily oxidised by dioxiranes to give the corresponding silanols and this provides a mild, high-yield, and eVective route for the synthesis of sensitive, polyfunctional silanols.18 The new silane, C[Si(CH3)2- H]3SiH3, when treated with six equivalents of dimethyldioxirane solution gave the new silanol C[Si(CH3)2OH]3Si(OH)3 in excellent yield.Unfortunately, the silanol tends to undergo condensation reactions in acetone, pentane or chlorinated solvents, but it is stable in the solid state as a fine white powder. Its 29Si NMR spectrum shows two signals; one at d 243.5 assigned to the Si(OH)3 resonance compares well with d 239.9 and 240.1 of the Si(OH)3 group in C[Si(CH3)3]3Si(OH)3 and C[Si(CH3)2Ph]3Si(OH)3, respectively,9 and one at d 10.5 assigned to the Si(CH3)2(OH) resonance is comparable to the d 13.4 assigned for the Si(CH3)2(OH) resonance in C[Si- (CH3)3]3Si(CH3)2OH.19 Silanols are known to crystallise to give a variety of hydrogen bonded structures 18 and it was thought of interest to see if the structure of C[Si(CH3)2OH]3Si(OH)3 might combine the unusual hydrogen-bonded features of the triol C[Si(CH3)3]3Si(OH)3, which crystallises to give discrete three-dimensional hexameric hydrogen bonded cages,20 and C[Si(CH3)2OH]4 which forms an infinite three-dimensional network. 21 Unfortunately all attempts at obtaining crystals suitable for X-ray crystallography have so far proved unsuccessful and are impeded by the silanol’s ease of condensation. Ab initio calculations for C[Si(CH3)2H]3SiH3 A full search of the potential energy surface located eleven different local minima for [(CH3)2SiH]3CSiH3, a surprisingly large number for a molecule comprising only 35 atoms. Moreover, the minima were found to lie at points on the potential energyJ. Chem.Soc., Dalton Trans., 1999, 2293–2301 2295 Table 1 Partial geometries of the eleven conformers of C[Si(CH)3)2H]3SiH3 calculated ab initio at 6-31G*/MP2 (re/Å, </8). Branch types ‘a’, ‘b’ and ‘c’ denotes branch torsional angles Si(2)–C(1)–Si–H of ca. 1608, 408 and 2808, respectively ‘aaa’ ‘bbb’ ‘ccc’ ‘aab’ ‘aca’ ‘abb’ ‘cbb’ ‘cca’ ‘ccb’ ‘acb’ ‘abc’ Bond distances rC(1)–Si(2) rC(1)–Si(3) rC(1)–Si(13) rC(1)–Si(23) rSi(2)–H(33) rSi(2)–H(34) rSi(2)–H(35) rSi(3)–H(4) rSi(3)–C(5) rSi(3)–C(6) rSi(13)–H(14) rSi(13)–C(15) rSi(13)–C(16) rSi(23)–H(24) rSi(23)–C(25) rSi(23)–C(26) 1.877 1.900 —— 1.490 —— 1.496 1.888 1.889 —————— 1.891 1.900 —— 1.489 —— 1.497 1.888 1.889 —————— 1.884 1.897 —— 1.489 —— 1.498 1.888 1.888 —————— 1.881 1.899 1.898 1.900 1.489 1.490 1.490 1.496 1.888 1.889 1.496 1.887 1.889 1.496 1.888 1.888 1.879 1.902 1.895 1.900 1.490 1.490 1.489 1.495 1.888 1.887 1.496 1.889 1.890 1.496 1.887 1.888 1.885 1.897 1.896 1.900 1.488 1.490 1.490 1.497 1.889 1.889 1.497 1.888 1.887 1.497 1.890 1.888 1.888 1.896 1.892 1.901 1.488 1.489 1.489 1.496 1.889 1.889 1.497 1.889 1.889 1.497 1.888 1.888 1.882 1.902 1.897 1.896 1.490 1.489 1.490 1.495 1.888 1.886 1.497 1.889 1.888 1.496 1.890 1.887 1.886 1.900 1.893 1.896 1.489 1.489 1.489 1.496 1.888 1.888 1.497 1.888 1.889 1.497 1.887 1.889 1.884 1.901 1.897 1.896 1.489 1.490 1.490 1.496 1.888 1.887 1.495 1.889 1.888 1.497 1.889 1.889 1.884 1.893 1.902 1.900 1.488 1.489 1.491 1.498 1.890 1.889 1.496 1.887 1.888 1.497 1.889 1.889 Bond angles <Si(2)–C(1)–Si(3) <Si(2)–C(1)–Si(13) <Si(2)–C(1)–Si(23) <C(1)–Si(3)–H(4) <C(1)–Si(3)–C(5) <C(1)–Si(3)–C(6) <H(4)–Si(3)–C(5) <H(4)–Si(3)–C(6) <C(5)–Si(3)–C(6) <C(1)–Si(13)–H(14) <C(1)–Si(13)–C(15) <C(1)–Si(13)–C(16) <H(14)–Si(13)–C(15) <H(14)–Si(13)–C(16) <C(15)–Si(13)–C(16) <C(1)–Si(23)–H(24) <C(1)–Si(23)–C(25) <C(1)–Si(23)–C(26) <H(24)–Si(23)–C(25) <H(24)–Si(23)–C(26) <C(25)–Si(23)–C(26) 110.2 —— 107.6 114.1 112.3 107.7 108.8 106.2 ———————————— 106.7 —— 107.4 112.4 114.0 107.2 107.4 108.0 ———————————— 107.8 —— 107.8 112.3 113.5 108.1 107.9 107.2 ———————————— 108.4 110.4 108.8 108.3 112.8 112.7 108.5 108.4 106.1 107.8 114.2 112.4 107.5 107.9 106.8 107.7 112.6 113.2 108.9 107.1 107.1 109.3 109.4 109.3 107.7 111.8 114.1 107.9 107.5 107.5 108.2 112.4 113.6 108.5 108.2 105.7 107.3 112.6 113.1 107.2 106.9 109.4 108.8 108.0 106.9 108.2 112.7 113.1 108.7 108.0 105.9 107.4 113.5 113.4 108.3 107.2 106.7 107.2 113.8 112.7 107.9 107.1 107.9 107.9 107.8 105.1 107.1 113.2 113.7 107.5 107.7 107.4 107.3 113.9 113.0 107.8 106.9 107.6 106.8 114.0 112.6 107.4 107.2 108.5 107.5 109.1 108.9 107.9 112.3 113.2 107.8 107.4 108.0 107.6 112.8 113.4 107.3 107.4 108.1 108.3 113.4 111.8 108.1 108.2 106.7 106.8 109.1 106.0 107.4 112.4 113.5 108.2 108.1 107.0 107.5 113.0 113.6 107.4 107.3 107.8 106.8 113.9 113.1 107.6 106.9 108.3 107.9 109.0 107.7 108.2 112.9 112.1 108.4 108.0 107.1 107.6 113.1 113.1 107.1 107.1 108.4 107.1 113.3 113.7 106.9 108.0 107.9 109.9 106.2 107.9 108.3 113.5 112.8 108.3 108.3 105.4 106.7 113.9 113.5 108.1 107.4 106.9 107.0 113.5 113.1 107.7 107.2 107.9 Torsional angles tSi(2)–C(1)–Si(3)–H(4) tSi(2)–C(1)–Si(13)–H(14) tSi(2)–C(1)–Si(23)–H(24) 158.8 —— 43.8 —— 280.1 —— 163.4 154.9 45.2 158.1 276.9 161.4 163.2 43.0 44.8 277.9 43.2 40.9 276.4 280.0 164.4 278.2 279.8 40.3 165.0 277.2 243.3 161.3 39.8 277.4 Table 2 Absolute energies (calculated ab initio) and relative abundance of all conformers modelled in the electron diVraction analysis Conformera Multiplicity Energy/Eh 6-31G*/MP2 Zero-point energy (ZPE) correction 6-31G*/HF Absolute energy (ZPE corrected) Population in gas phase b ‘aaa’ ‘bbb’ ‘ccc’ ‘aab’ ‘aca’ ‘abb’ ‘cbb’ ‘cca’ ‘ccb’ ‘acb’ ‘abc’ 11133333333 21436.050072 21436.049974 21436.050837 21436.050232 21436.050361 21436.049731 21436.049891 21436.050462 21436.050241 21436.050298 21436.049575 0.303010 0.302480 0.302932 0.302740 0.302696 0.302597 0.302703 0.302598 0.302832 0.302553 0.302781 21435.747062 21435.747494 21435.747905 21435.747492 21435.747665 21435.747134 21435.747188 21435.747864 21435.747409 21435.747745 21435.746794 0.03 0.04 0.05 0.11 0.13 0.08 0.09 0.16 0.11 0.14 0.06 a See text and Fig. 2 for structural definitions. b Calculated on the basis of a Boltzmann distribution, relative to the lowest energy conformer ‘ccc’ at 373 K. Abundances are then corrected for the eVects of multiplicity and normalised.surface within a range of only ca. 3 kJ mol21, predicting that all eleven conformers will exist in the gas phase, and thus be of comparable importance in the GED refinement. Partial geometries obtained from the 6-31G*/MP2 optimisations only are presented in Table 1. [A full set of Brookhaven (pdb) files are available as Electronic Supplementary Information.] The abso-2296 J. Chem. Soc., Dalton Trans., 1999, 2293–2301 Fig. 2 Molecular structures of the eleven conformers of C[Si(CH3)2H]3SiH3. Torsional angles Si(2)C(1)SiH denoted ‘a’ ca. 1608, ‘b’ ca. 408 and ‘c’ ca. 2808. lute energies obtained in the 6–31G*/MP2 set of calculations, with zero-point energy corrections obtained from the 6-31G*/ HF frequency calculations, are listed in Table 2. This system, comprising three branches with three diVerent possible orientations for each branch, will have a total of 27 (33) diVerent possible conformations, some of which are equivalent.With reference to Fig. 2 and 3 (for atom labelling) and Table 1 it is clear that the geometries calculated fall into tightly defined categories. In compounds of the type A(XY3)4, 1,3-interactions between atoms or groups Y cause the XY3 groups to twist away from perfectly staggered conformations, usually by 15–208. All four groups must twist in the same sense. In this case, therefore, to within a close approximation only three diVerent values of the branch torsional angles [i.e.Si(2)–C(1)–Si(3)–H(4), Si(2)–C(1)–Si(13)–H(14) and Si(2)–C(1)–Si(23)–H(24)] were observed, ca.1608 (labelled type ‘a’), 408 (type ‘b’) and 2808 (type ‘c’). Thus there are three conformers with C3 symmetry (and multiplicities of one) labelled ‘aaa’, ‘bbb’ and ‘ccc’, and eight remaining conformers with C1 symmetry (multiplicities of three) arise from all other possible combinations of the three diVerent branch types (see Fig. 2). Further conformations are possible if the twisting of branch methyl groups relative to the branch hydrogens H(4), H(14) and H(24) is considered.However with all eleven minima found the methyl torsional twists are very slight, with most groups less than 58 away from a perfectly staggered conformation. This structural feature is therefore considered to be negligible, and no further minima have been found. A rigid scan of the potential energy surface linking two minima (corresponding to conformers ‘cca’ to ‘ccb’; 6-31G*/HF level, see Table 3) was undertaken to determine the height of the potential energy barrier to free rotation between the two con-J.Chem. Soc., Dalton Trans., 1999, 2293–2301 2297 formers. Although a rigid scan can only give an approximation of the barrier height, a value of more than 80 kJ mol21 was calculated, indicating that the minima must relate to two very distinct features on the potential energy surface. Perhaps this is not surprising, as a substantial geometry change of ca. 1208 is required to any one branch torsional angle to convert between two conformer types. It can therefore be concluded that in all likelihood all geometries calculated will exist as distinct entities; the groups are not simply freely rotating between the eleven minima found. Gas-phase electron diVraction (GED) GED model. In light of the unexpected prediction from ab initio calculations that eleven independent conformers of near equal energy will exist in the gas phase, several structural assumptions had to be made in order to reduce the number of geometric parameters needed to describe such an exceptionally large system.Full details of the model comprising 45 geometric parameters can be found in the Electronic Supplementary Information. The main structural assumptions made are listed here for ease of reference. Si(2,3,13,23) branches. The biggest structural assumption made was to describe all branches in all eleven conformers by just three diVerent branch types ‘a’, ‘b’ and ‘c’, as detailed above.This greatly simplified the model required but still made it possible to model all Si–C(1)–Si angles and torsions calculated ab initio to within a mean deviation of just 28. Thus three parameters were required to describe the angles Si(2)–C(1)– Si(3,13 or 23), and just seven torsional angles were required to separate the three branch types to give the conformer arrangements shown in Fig. 2. On the basis of geometries calculated ab initio the conformers were assumed to possess only three diVerent Si–C distances: silyl rSi(2)–C(1), middle rSi(3,13 or Fig. 3 Atom numbering scheme adopted for C[Si(CH3)2H]3SiH3. Table 3 Results from ab initio (6-31G*/HF) scan of potential energy surface between minima corresponding to conformers ‘cca’ and ‘ccb’ Dihedral angle branch ‘a’ æÆ ‘b’ Absolute energy/Eh Da 123456789 10 11 12 13 166.6 156.4 146.2 136.0 125.8 115.6 105.4 95.2 85.0 74.8 64.6 54.4 44.2 21434.78839 21434.78662 21434.78031 21434.76989 21434.76098 21434.75875 21434.75860 21434.75594 21434.75632 21434.76463 21434.77495 21434.78197 21434.78487 0.0 14.6 121.2 148.6 180.0 177.8 178.2 185.2 184.2 162.4 135.3 116.9 19.0 a DiVerences quoted in kJ mol21, relative to the minimum ‘cca’. 23)–C(1) and branch rSi–C(methyl). Only two diVerent Si–H distances were modelled [those in the SiH3 groups and those in the Si(CH3)H groups]. One parameter was used to describe all 66 C(1)–Si–C branch angles [e.g.<C(1)Si(3)C(5 or 6) etc.] and one parameter to define all 33 C(1)–Si–H branch angles. Calculated values for both parameters fell over a range of just 28 (see Table 1), justifying the use of one parameter in each case. Methyl groups. All were considered to be identical and possess local C3v symmetry. In light of the ab initio calculations they were also assumed to be perfectly staggered with respect to the branch hydrogens H(4), H(14) and H(24).SiH3 group. This group was assumed to be identical in all eleven conformers and to possess local C3v symmetry. GED refinement. On the basis of the ab initio calculations detailed above, eleven diVerent conformers were modelled in the refinement of the GED data collected for C[Si(CH3)2- H]3SiH3. The relative weightings of the conformers were fixed at values obtained from a consideration of the Boltzmann distribution of population states, relative to the lowest energy conformer ‘ccc’ (6-31G*/MP2; ZPE corrected, see Table 3), at the temperature of the GED data collection.§ These relative weightings were then corrected for the eVects of multiplicity (i.e.the weightings of the C3 conformers reduced to one third of their initial value) and then normalised. The values used in the final GED refinement are given in the last column of Table 2. The presence of a large number of similar interatomic distances and of some parameters involving hydrogen (which is a poor scatterer of electrons) of low multiplicity prevented a complete structure determination for C[Si(CH3)2H]3SiH3 using just experimental data, even with the simplifications built into the model.In such cases it is our practice to include information obtained theoretically to allow complete structural determination using the SARACEN method.22 The essential feature of this method is that information calculated ab initio is introduced into the refinement procedure as additional observations (or restraints), the weight of any observation being assigned according to the level of convergence achieved in a series of graded ab initio calculations. By employing the SARACEN method in the present work it has been possible to refine the values of all structural parameters and all significant amplitudes of vibration.The final refinement is then the best fit to all available information, both experimental and theoretical, and represents the most probable structure, avoiding subjective preference for one particular type of data. The values of all additional observations used in the refinement can be found in Table 4 together with their respective weightings (uncertainties).The results from the SARACEN refinement, based on GED data supplemented with ab initio-based restraints, are given in Table 4 where they are compared with parameters derived computationally. ¶ In general most geometric parameters refined to values in good agreement with those calculated ab initio.Most notably the freely refining parameters, which define the key features on the radial distribution curve (see Fig. 4), refined to values within acceptable agreement with calculated geometries. The distance C–H (p1), which gives an unusually large contribution to the radial distribution curve, refined to 1.089(4) Å, just out within one esd of the average ab initio value of 1.094 Å. The average Si–C distance (p3, peak three in the radial distribution curve) refined to 1.884(1) Å, compared to the average calculated § Calculated changes in entropy (6-31G*/HF) for all eleven conformers fell over a range of just 0.02 kJ K21.The conformation of the gas phase mixture was determined based on DH values, not DG. ¶ For large, floppy molecules, such as those detailed in this paper, it is not realistic to expect to obtain a reliable ra structure based on harmonic rectilinear (parallel and perpendicular) vibrational corrections.In particular the perpendicular corrections are very poorly calculated, and introduce errors greater than those they are meant to solve. The structures presented in this paper are therefore of type ra.2298 J. Chem. Soc., Dalton Trans., 1999, 2293–2301 Table 4 Structural parameters obtained by gas-phase electron diVraction and ab initio calculations (r/Å, </8) Parameter a GED (restrained results) (ra) b Ab initio (6-31G*/MP2) (re) c Independent parameters p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 p41 p42 p43 p44 p45 p46 p47 p48 p49 rC–H <Si–C–H (branch) av.rSi–C (silyl 1 middle 1 branch) D rSi–C [middle 2 av.(branch 1 silyl)] D rSi–C (branch 2 silyl) av. rSi–H av. (branch 1 silyl) D rSi–H (branch 2 silyl) <C(1)–Si–C (branch) <C(1)–Si–H (branch) <H–Si–C (branch) D <Si(2)–C(1)–Si (‘a’ 1 ‘b’ 1 ‘c’) D <Si(2)–C(1)–Si [‘a’ 2 av.(‘b’ 1 ‘c’)] D <Si(2)–C(1)–Si (‘b’ 2 ‘c’) av.<Si(branch)–C(1)–Si(branch) D <Si–C(1)–Si [av.(‘ab’ 1 ‘bb’ 1 ‘cb’) 2 av.(‘aa’ 1 ‘ac’ 1 ‘bc’ 1 ‘cc’)] D <Si–C(1)–Si [av.(‘ab’ 1 ‘bb’) 2 ‘cb’] D <Si–C(1)–Si(‘ab’ 2 ‘bb’) D <Si–C–(1)–Si[av. (‘aa’ 1 ‘cc’) 2 av.(‘ac’ 1 ‘bc’)] D <Si–C(1)–Si[‘cc’ 2 ‘aa’] D <Si–C(1)–Si[‘ac’ 2 ‘bc’] <C(1)–Si(2)–H (centre) twist angle SiH3 ‘aaa’ tC(1)–Si(2)–Si–H ‘bbb’ tC(1)–Si(2)–Si–H ‘ccc’ tC(1)–Si(2)–Si–H ‘aab’ tC(1)–Si(2)–Si(3)–H(4) ‘aab’ tC(1)–Si(2)–Si(3)–H(14) ‘aab’ tC(1)–Si(2)–Si(3)–H(24) ‘aca’ tC(1)–Si(2)–Si(3)–H(4) ‘aca’ tC(1)–Si(2)–Si(13)–H(14) ‘aca’ tC(1)–Si(2)–Si(23)–H(24) ‘abb’ tC(1)–Si(2)–Si(3)–H(4) ‘abb’ tC(1)–Si(2)–Si(13)–H(14) ‘abb’ tC(1)–Si(2)–Si(23)–H(24) ‘cbb’ tC(1)–Si(2)–Si(3)–H(4) ‘cbb’ tC(1)–Si(2)–Si(13)–H(14) ‘cbb’ tC(1)–Si(2)–Si(23)–H(24) ‘cca’ tC(1)–Si(2)–Si(3)–H(4) ‘cca’ tC(1)–Si(2)–Si(13)–H(14) ‘cca’ tC(1)–Si(2)–Si(23)–H(24) ‘ccb’ tC(1)–Si(2)–Si(3)–H(4) ‘ccb’ tC(1)–Si(2)–Si(13)–H(14) ‘ccb’ tC(1)–Si(2)–Si(23)–H(24) ‘abc’ tC(1)–Si(2)–Si(3)–H(4) ‘abc’ tC(1)–Si(2)–Si(13)–H(14) ‘abc’ tC(1)–Si(2)–Si(23)–H(24) ‘acb’ tC(1)–Si(2)–Si(3)–H(4) ‘acb’ tC(1)–Si(2)–Si(13)–H(14) ‘acb’ tC(1)–Si(2)–Si(23)–H(24) 1.089(4) 109.8(5) 1.884(1) 0.013(2) 0.005(2) 1.502(12) 0.010(5) 113.8(4) 106.4(9) 107.1(7) 107.0(6) 1.6(4) 21.5(5) 109.9(9) 2.7(4) 22.6(5) 0.7(5) 1.8(5) 0.2(5) 0.9(5) 109.0(9) 280.5(18) 158.7(18) 43.8(18) 279.8(18) 163.7(18) 154.8(18) 45.6(18) 158.2(18) 276.5(18) 161.2(18) 163.4(18) 43.2(18) 44.8(18) 278.0(18) 43.4(18) 40.7(18) 275.8(17) 279.9(18) 164.4(18) 278.3(18) 279.6(18) 40.4(17) 161.6(18) 39.7(18) 277.4(18) 165.3(18) 277.2(18) 43.3(18) av. 1.094 range 109.0–113.0 1.890 0.012(2) 0.004(2) 1.480(20) 0.010(5) av. 113.2 107.0(10) 108.0(10) 108.3(10) 1.3(5) 21.2(5) 110.6 2.8(5) 22.6(5) 0.8(5) 1.8(5) 0.2(5) 0.9(5) 109.0(10) 280(2) 159(2) 44(2) 280(2) 163(2) 155(2) 45(2) 158(2) 277(2) 161(2) 163(2) 43(2) 45(2) 278(2) 43(2) 41(2) 276(2) 280(2) 164(2) 278(2) 280(2) 40(2) 161(2) 40(2) 277(2) 165(2) 277(2) 43(2) Dependent parameters rSi–C (silyl) rSi–C (middle) rSi–C (branch) rSi–H (silyl) rSi–H (branch) <Si(2)–C(1)–Si (‘a’) <Si(2)–C(1)–Si (‘b’) <Si(2)–C(1)–Si (‘c’) <Si–C(1)–Si (‘aa’) <Si–C(1)–Si (‘ab’) <Si–C(1)–Si (‘ac’) <Si–C(1)–Si (‘bb’) <Si–C(1)–Si (‘bc’) <Si–C(1)–Si (‘cb’) <Si–C(1)–Si (‘cc’) 1.878(1) 1.893(2) 1.883(1) 1.497(12) 1.508(12) 108.1(6) 105.7(6) 107.2(6) 109.6(10) 110.3(10) 107.4(10) 111.0(10) 108.3(10) 113.2(10) 109.8(10) av. 1.883 av. 1.898 av. 1.888 av. 1.490 av. 1.497 av. 109.2 av. 107.0 av. 108.3 av. 110.2 av. 111.0 av. 108.1 av. 111.7 av. 109.0 av. 113.9 av. 110.4 a See text for model description. Note: ‘silyl’ = rSi(2)–C(1), ‘middle’ = rSi(3,13 or 23)–C(1) and ‘branch’ = rSi–C(methyl) distances (see Fig. 3 for atom numbering scheme). Abbreviations used: r = bond distance, < = angle, t = dihedral angle, av. = average, D = diVerence, ‘a,b,c’ = branch types; see the text and Fig. 2 for details. b Estimated standard deviations (e.s.d.s) obtained in the least-squares refinement are given in parentheses, quoted to 1s. c Ab initio data quoted with uncertainties are SARACEN restraints used in the GED refinement; ab initio values quoted as averages are derived from values calculated for each given parameter over all eleven conformers.J. Chem. Soc., Dalton Trans., 1999, 2293–2301 2299 value of 1.890 Å.The Si–C–H (branch) angle (p2), which in conjunction with p1 and p3 defines the position of the fourth peak on the radial distribution curve (labelled rHmethyl ? ? ? Si), refined to 109.8(5)8, which falls within the range of values calculated for this angle (109.0–113.08). The branch C(1)–Si–C angle (p8), which along with p3 directly positions the C ? ? ? C distances under peak 5, refined to 113.8(4)8, compared to the calculated average value of 113.28. The two branch angles, <Si(2)–C(1)–Si and <Si(branch)–C– Si(branch), p11 and p14, which together with the Si–C distances define the Si ? ? ? Si distances under peak 5, were found to be heavily correlated in the refinement (see Table 5).Unrestrained, both parameters may drift to unrealistic values. A restraint was therefore applied to parameter 11 (p14 left unrestrained), resulting in both parameters returning values in the least-squares analysis within ca. 18 of the calculated values. The remaining parameters which could not refine to realistic values, because they refer either to subtle geometry diVerences between correlated bond distances or angles (i.e.parameters 4, 5, 7, 12, 13, and 15–20) or to parameters involving hydrogen (parameters 7, 9, 10, 21, 22 and 23–45) were assigned ab initio based restraints to aid their refinement. All restrained parameters returned values in the least-squares analysis in agreement with their imposed restraints to within one or two e.s.d.s. In addition to all 45 geometric parameters, four amplitudes of vibration, corresponding to groups of similar distances under the four most prominent peaks on the radial distribution curve, were also refined.The groups chosen correspond to all C–H distances for the eleven conformers under peak 1, the Si–C distances under peak 3, rHmethyl ? ? ? Si (peak 4) and the Si ? ? ? Si distances under peak 5. All amplitudes refined to reasonable values, in good agreement with those calculated ab initio.The final RG factor recorded for this eleven-conformer refinement is 0.075, indicating that the data are of good quality and a good fit between model and experiment has been obtained. The final experimental and diVerence radial distribution and molecular scattering curves are shown in Fig. 4 and 5, respectively. A selective listing of bonding distances and amplitudes of vibration common to all eleven conformers (due to the structural approximations made as listed above) is given in Table 5.The final correlation matrix obtained is given in Table 6. A full set of coordinates is available in Table 1 of the Electronic Supplementary Information. Experimental Synthesis Preparation of C[Si(CH3)2Cl]3SiCl3. A solution of C[Si- (CH3)3]3SiCl3 23 (2.30 g, 6.28 mmol) in 5.13 M ICl in CCl4 (30 ml, 153.9 mmol) was stirred under N2 for 6 hours at room temperature. The excess ICl was destroyed by cautiously shaking the mixture with saturated aqueous sodium thiosulfate (50 ml).The organic layer was separated, washed with water, separated, dried over MgSO4 and the solvent removed under vacuum to leave a pale yellow solid. The desired product was extracted into Table 5 Selected distances, common to all level conformer models, from the SARACEN refinement of C[Si(CH3)2H]3SiH3 Position on radial distribution curve Atom pair Amplitude, u/Å peak 1 peak 2 peak 3 peak 4 peak 5 C(5)–H(7) Si(3)–H(4) Si(2)–H(33) Si(3)–C(5) Si(2)–C(1) Si(3)–C(1) Si(3) ? ? ?Hmethyl Si ? ? ? Si 0.066(5) 0.086 (fixed) 0.085 (fixed) 0.039(2) 0.030 {tied to u[Si(2)–C(5)]} 0.041 {tied to u[Si(2)–C(5)]} 0.09(1) 0.118(3) chloroform. Removal of the chloroform left shiny yellow crystals which were sublimed under vacuum (0.2 mmHg, 150– 180 8C) to give white crystals identified as tris(chlorodimethylsilyl) trichlorosilylmethane, C[Si(CH3)2Cl]3SiCl3 (2.4 g, 83% yield). Mp >320 8C. 1H NMR: d 0.91 [s, Si(CH3)2]. Proton coupled 29Si: d 0.18 (s, SiCl3), 22.64 (m, Si(CH3)2Cl, 2JSi–H = 6.7 Hz); m/z (based on 35Cl ) 411 (100%, [M 2 CH3]1), 391 (5%, [M 2 Cl]1), 303 (30%), 283 (45%), 93 (45%, [Si(CH3)2Cl]1). Accurate m/z for [M 2 CH3]1 410.8349 (calc. 410.8352), actual isotope pattern for [M 2 CH3]1 matches computer simulation. Reaction of C[Si(CH3)3]3SiCl3 with IBr. The chlorosilane, C[Si(CH3)3]3SiCl3 (0.81 g, 2.22 mmol), was added to a 7.45 M solution of IBr in CCl4 (10 ml, 74.5 mmol). The mixture was left stirring under N2 for 6 hours after which the 1H NMR spectrum was consistent with the presence of a mixture of C[Si(CH3)3]3SiCl3, C[Si(CH3)2Br][Si(CH3)3]2SiCl3 and CH3I; C[Si(CH3)3]3SiCl3 (80%): d 0.38 [Si(CH3)3]; C[Si(CH3)2Br]- [Si(CH3)3]2SiCl3 (20%): d 0.48 [Si(CH3)3], 0.96 [Si(CH3)2Br)]; CH3I: d 2.66 (CH3). The mixture was left stirring at room temperature for a further five days after which the 1H NMR spec- Fig. 4 Experimental and diVerence (experimental 2 theoretical) radial distribution curves for the multi-conformer analysis of C[Si(CH3)2H]3SiH3. Before Fourier inversion the data were multiplied by s.exp[(20.002s2)/(ZSi 2 fSi)(ZC 2 fC)].Fig. 5 Experimental and final weighted diVerence (experimental 2 theoretical) molecular scattering intensities for the multi-conformer analysis of C[Si(CH3)2H]3SiH3. Table 6 Correlation matrix (× 100) for the SARACEN refinement of C[Si(CH3)2H]3SiH3. Only elements with absolute values greater than 50 are shown.k2 is a scale factor; u is an amplitude of vibration p10 p11 p14 k2 p8 p11 u(C–Si) 255 65 269 286 722300 J. Chem. Soc., Dalton Trans., 1999, 2293–2301 trum indicated the presence of a trace amount of C[Si- (CH3)3]3SiCl3 and a 2 : 1 mixture of C[Si(CH3)2Br][Si(CH3)3]2- SiCl3 and C[Si(CH3)2Br]2[Si(CH3)3]SiCl3; C[Si(CH3)2Br]2- [Si(CH3)3]SiCl3: d 0.57 [Si(CH3)3], 1.05 [Si(CH3)2Br)]. After 28 days the 1H NMR spectrum had changed little from the one observed after 5 days, and so the reaction was not followed up further.Reduction of C[Si(CH3)2Cl]3SiCl3 with LiAlH4. A mixture of C[Si(CH3)2Cl]3SiCl3 (0.98 g, 2.29 mmol) and LiAlH4 (0.50 g, 13.17 mmol) in dry THF under N2 was boiled under reflux for 6 hours. The excess LiAlH4 was destroyed by carefully adding the mixture to 10% aqueous tartaric acid (100 ml) and the product extracted with hexane. Following separation, the solvent was removed under reduced pressure to give an oily liquid which was found by 1H NMR spectroscopy to be a mixture of C[Si(CH3)2H]3SiH3, C[Si(CH3)2H]3H,24 and other unidentified compounds.Reducing the reflux time to 2 hours with the same amount of LiAlH4 gave the same results as above so the reduction was carried out with half the amount of LiAlH4 at room temperature. After 24 hours, the 1H NMR spectrum of a sample of the reaction mixture showed the presence of the starting material, C[Si(CH3)2H]3SiH3 and a small amount of C[Si(CH3)2H]3H. After a further 24 hours of stirring only C[Si(CH3)2H]3H could be identified by 1H NMR spectroscopy. Preparation of C[Si(CH3)2H]3SiH3.A mixture of LiAlH4 (1.88 g, 49.5 mmol), benzyltriethylammonium chloride (0.38 g, 1.67 mmol), C[Si(CH3)2Cl]3SiCl3 (1.782 g, 4.17 mmol) and dry toluene (40 ml) was boiled under reflux under a dry N2 atmosphere for 6 h. The mixture was left to cool and then filtered through a sintered glass funnel to leave the inorganic salts. A vacuum of 2 mmHg was applied to the solution and the desired product collected as white shiny crystals in a liquid nitrogen cooled trap.The crystals were identified as tris(dimethylsilyl)- silylmethane, C[Si(CH3)2H]3SiH3, (0.37 g, 40% yield). Mp 67 8C. 1H NMR: d 0.22 [d, 18H, Si(CH3)2H, 3JMe–H = 3.96 Hz], 3.64 (s, 3H, SiH3), 4.07 [sept, 3H, Si(CH3)2H]. Proton coupled 29Si NMR: d 13.77 (d of m, Si(CH3)2H, 1JSi–H = 188 Hz), 261.28 (qq, SiH3, 1JSi–H = 200 Hz, 3JSi–H = 3.5 Hz); m/z: 219 (5%, [M 2 H]1), 205 (25, [M 2 CH3]1), 173 (10), 115 (20), 113 (25), 95 (30), 93 (95), 73 (55, [Si(CH3)3]1), 65 (45), 63 (40), 59 (30, [Si(CH3)2H]1).Accurate mass for [M 2 H]1 219.086 (calc. 219.088). IR (Nujol): 2140.3 cm21 (s, SiH3), 2110.0 (s, Si(CH3)2H). Preparation of C[Si(CH3)2OH]3Si(OH)3. In a 100 ml two necked round bottom flask cooled to 210–0 8C with the aid of an ice/salt bath, the silane C[Si(CH3)2H]3SiH3 (0.06 g, 0.27 mmol) was stirred with a dimethyldioxirane solution (0.038 M, 50 ml) 25 under N2 for 8 hours.The solvent was removed under vacuum to leave a white solid identified as trihydroxysilyltris[( hydroxy)dimethylsilyl]methane, C[Si(CH3)2OH]3Si(OH)3, (0.081 g, 95% yield). Mp 85 8C (decomp.). 1H NMR (acetoned6): d 0.51 [s, Si(CH3)2OH]; the OH signal could not be assigned. Proton coupled 29Si NMR (acetone-d6): d 10.47 [m, Si(CH3)2OH, 2JSi–H = 6.7 Hz], 242.53 [s, Si(OH)3]; m/z 283 (20%, [M–H2O–CH3]1), 265 (100, [M–2H2O–CH3]1), 249 (95), 205 (40), 125 (40).Accurate mass for [M–H2O–CH3]1 283.033 (calc. 283.031). IR (KBr disc): 3350 cm21 (b, H-bonded SiO–H). Bromination of C[Si(CH3)2H]3SiH3. A 1 M solution of Br2 in benzene was added dropwise to the silane (0.024 g, 0.1 mmol) under an atmosphere of dry N2 and the reaction was monitored by 1H NMR spectroscopy until all the Si–H had disappeared. The solvent was removed under vacuum to leave white shiny crystals identified as tris(bromodimethylsilyl)tribromosilylmethane, C[Si(CH3)2Br]3SiBr3 (0.075 g, 100% yield).Mp >320 8C. 1 H NMR (at 25 8C): d 1.2 [broad s, (CH3)2SiBr.] 29Si NMR (at 50 8C): d 18.66 [s, (CH3)2SiBr], 236.59 (s, SiBr3); m/z (based on 79Br) 679 (30%, [M 2 CH3]1), 615 (100, [M 2 Br]1), 527 (20), 461 (60), 397 (17), 137 (25, [Si(CH3)2Br]1), 73 (35, [Si(CH3)3]1). Accurate mass for [M 2 CH3]1 678.525 (calc. 678.529). Computer simulation of [M 2 CH3]1 ion matches the acquired spectrum. Partial bromination of C[Si(CH3)2H]3SiH3. The silane (0.185 g, 0.84 mmol) was dissolved in benzene (2 ml) and a bromine solution (0.63 M, 2.0 ml in benzene) was added dropwise while the silane solution was being stirred. After an hour of further stirring at room temperature, the 1H NMR spectrum of a sample showed a mixture of C[Si(CH3)2Br][Si(CH3)2H]2SiH3, C[Si(CH3)2Br]2[Si(CH3)2H]SiH3, C[Si(CH3)2Br]3SiH3 and a trace amount of the starting material to be present. 1H NMR (33%): C[Si(CH3)2Br][Si(CH3)2H]2SiH3: d 0.31 [d, Si(CH3)2H, 2JMe–H = 3.96 Hz], 0.69 [s, Si(CH3)2Br], 3.69 (s, SiH), 4.0–4.2 [m, Si(CH3)2H]; C[Si(CH3)2Br]2[Si(CH3)2H]SiH3 (33%): d 0.41 [d, Si(CH3)2H, 2JMe–H = 3.63 Hz], 0.79 [s, Si(CH3)2Br], 3.75 (s, SiH), 4.0–4.2 [m, Si(CH3)2H]; C[Si(CH3)2Br]3SiH3 (33%): d 0.90 [s, Si(CH3)2Br], 3.82 (s, SiH).The Si–H signals for the Si(CH3)2H groups overlap (4.0–4.2 ppm) and cannot be distinguished. More of the bromine solution (0.28 mmol) was added and the solution was left stirring overnight. A 1H NMR spectrum of a sample of the products after that was consistent with the presence of a mixture of C[Si(CH3)2Br]3SiH3, C[Si(CH3)2Br]3Si- BrH2, and C[Si(CH3)2Br]3SiBr2H. 1H NMR: C[Si(CH3)2- Br]3SiH3 as above; C[Si(CH3)2Br]3SiBrH2, (50%): d 1.00 [s, Si(CH3)2Br], 4.82 (s, SiH); C[Si(CH3)2Br]3SiBr2H, (10%): d 1.10 [s, Si(CH3)2Br], 5.65 (s, SiH). Ab initio calculations Calculations were performed on a DEC Alpha APX 1000A workstation using the GAUSSIAN94 program,26 with the larger calculations run on a DEC 8400 superscalar cluster equipped with 10 fast processors, 6 GB of memory and 150 GB disk (resource of the UK Computational Chemistry Facility).Geometry optimisations. An extensive search of the potential energy surface was undertaken at the 3–21G* 27–29/HF level in order to locate all structurally stable conformers. In total eleven minima were found, corresponding to three structures with C3 symmetry and eight with C1 (see Fig. 2). Further geometry optimisations were then undertaken for all minima with the 6-31G* basis set 30–32 at the HF and MP2 levels of theory.Frequency calculations. Vibrational frequencies were calculated from analytic second derivatives at the 3–21G*/HF and 6-31G*/HF levels to confirm all conformers as local minima on the potential energy surface. The force constants obtained from the 6-31G*/HF calculations were subsequently used to construct harmonic force fields using the ASYM40 program.33 As no fully assigned vibrational spectra are available for this compound to scale the force fields, scaling factors of 0.9, 0.85 and 0.8 were adopted for bond stretches, angle bends and torsions, respectively, with values chosen falling within acceptable guidelines as suggested by Rauhut and Pulay.34 The scaled harmonic force fields were then used to provide estimates of amplitudes of vibration (u) for use in the GED refinements.Potential energy surface scan. To establish the eleven minima found on the potential energy surface as distinct features with appreciable barriers to internal rotation, a rigid scan of the potential energy surface connecting the minima corresponding to conformers ‘cca’ and ‘ccb’ was undertaken at the 6-31G*/HF level.The geometry of the molecule was frozen with the exception of the one dihedral angle required to convert from conformer ‘cca’ to ‘ccb’, which was stepped in twelve increments of 10.28. Single-point energy calculations were then performed for each new value of the dihedral angle.J. Chem.Soc., Dalton Trans., 1999, 2293–2301 2301 Table 7 GED data analysis parameters for C[Si(CH3)2H]3SiH3 Camera Weighting functions/Å21 Correlation Scale Electron distance/mm Ds smin sw1 sw2 smax parameter factor, ka wavelengthb/Å 247.59 95.63 0.2 0.4 2.0 8.0 4.0 10.0 10.0 30.4 11.0 35.6 0.3350 0.4011 0.930(6) 0.919(20) 0.06016 0.06016 a Figures in parentheses are the estimated standard deviations. b Determined by reference to the scattering patterns of benzene vapour.Gas-phase electron diVraction (GED) Electron scattering intensities were recorded on Kodak Electron Image photographic plates using the Edinburgh gasphase electron diVraction apparatus,35 operating at ca. 40 kV. Five plates (three from the long camera distance and two from the short distance) were recorded and converted into digital format using a computer-controlled PDS microdensitometer employing a 200 micron pixel size at the Royal Greenwich Observatory, Cambridge.36 The sample and nozzle temperatures were maintained at ca. 373 K during the exposure periods. Standard programs were used for the data reduction with the scattering factors of Ross et al.37 Nozzle-to-plate distances, weighting functions used to set up the oV-diagonal weight matrix, correlation parameters, final scale factors and electron wavelengths for the measurements are collected in Table 7. Acknowledgements We thank the EPSRC for financial support of the Edinburgh Electron DiVraction Service (grant GR/K44411) and for the Edinburgh ab initio facilities (grant GR/K04194).We also thank the UK Computational Chemistry Facility (admin: Department of Chemistry, King’s College London, Strand, London, UK WC2R 2LS) for the computing time on Columbus. P. C. M. wishes to thank the British Council for the award of a Commonwealth Scholarship. References 1 See, for example, C. Eaborn, Y. Y. El-Kaddar and P.D. Lickiss, Inorg. Chim. Acta, 1992, 200, 337. 2 See, for example, P. D. Lickiss, in Comprehensive Organic Functional Group Transformations, ed. A. R. Katritzky, O. Meth-Cohn and C. W. Rees, Pergamon, Oxford, 1995, vol. 6, p. 377. 3 See, for example J. R. Black, C. Eaborn, P. M. Garrity and D. A. R. Happer, J. Chem. Soc., Perkin Trans. 2, 1997, 1633; M. A. M. R. Al-Gurashi, G. A. Ayoko, C. Eaborn and P. D. Lickiss, J. Organomet. Chem., 1995, 499, 57; C. Eaborn, A. Kowalewska and W.A. Stanczyk, J. Organomet. Chem., 1998, 560, 41, and references therein. 4 R. Hager, O. Steigelmann, G. Muller, H. Schmidbaur, H. E. Robertson and D. W. H. Rankin, Angew. Chem., Int. Ed. Engl., 1990, 29, 201. 5 P. Kulpinski, P. D. Lickiss and W. Stanczyk, Bull. Pol. Acad. Sci., 1992, 40, 21. 6 C. Eaborn and P. D. Lickiss, J. Organomet. Chem., 1985, 294, 305. 7 S. S. Dua, C. Eaborn, D. A. R. Happer, S. P. Hopper, K. D. Safa and D. R. M. Walton, J. Organomet. Chem., 1979, 178, 75. 8 C. Eaborn, P. B. Hitchcock and P. D. Lickiss, J. Organomet. Chem., 1983, 252, 281. 9 S. M. Whittaker, PhD Thesis, University of Salford, 1993. 10 V. N. Gevorgyan, L. M. Ignatovich and E. Lukevics, J. Organomet. Chem., 1985, 284, C31. 11 P. C. Masangane, Imperial College, London, unpublished work. 12 N. M. K. El-Durini and R. A. Jackson, J. Chem. Soc., Perkin Trans. 2, 1993, 1275. 13 P. D. Lickiss, D.Phil. Thesis, University of Sussex, 1983. 14 C. Eaborn, P. B. Hitchcock, A.Pidcock and K. D. Safa, J. Chem. Soc., Dalton Trans., 1984, 2015. 15 A. G. Avent, S. G. Bott, J. A. Ladd, P. D. Lickiss and A. Pidcock, J. Organomet. Chem., 1992, 427, 9. 16 W. Kemp, NMR in Chemistry: A Multinuclear Introduction, Macmillan Education Ltd., London, 1986. 17 M. McPartlin, University of North London, personal communication. 18 P. D. Lickiss, Adv. Inorg. Chem., 1985, 42, 147. 19 R. Damrauer and R. J. Linerman, J. Organomet. Chem., 1982, 235, 1. 20 S. S. Al-Juaid, N. N. Buttrus, R. I. Damja, Y. Derouiche, C. Eaborn, P. B. Hitchcock and P. D. Lickiss, J. Organomet. Chem., 1989, 371, 287. 21 S. S. Al-Juaid, C. Eaborn, P. B. Hitchcock and P. D. Lickiss, J. Organomet. Chem., 1988, 353 297. 22 A.J. Blake, P. T. Brain, H. McNab, J. Miller, C. A. Morrison, S. Parsons, D. W. H. Rankin, H. E. Robertson and B. A. Smart, J. Phys. Chem., 1996, 100, 12280; P. T. Brain, C. A. Morrison, S. Parsons and D. W. H. Rankin, J. Chem. Soc., Dalton Trans., 1996, 4589. 23 Z. A. Aiube and C. Eaborn, J. Organomet. Chem., 1984, 269, 217; S. S. Dua, C. Eaborn, D. A. R. Happer, S. P. Hopper, K. D. Safa and D. R. M. Walton, J. Organomet. Chem., 1979, 178, 75. 24 C. Eaborn, P. B. Hitchcock and P. D. Lickiss, J. Organomet. Chem., 1983, 252, 281. 25 R. W. Murray and R. Jeyaraman, J. Org. Chem., 1985, 50, 2847. 26 Gaussian 94 (Revision C.2), M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheesman, T. A. Keith, G. A. Petersson, J. A.Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez and J. A. Pople, Gaussian Inc., Pittsburgh, PA, 1995. 27 J. S. Binkley, J. A. Pople and W. J. Hehre, J. Am. Chem. Soc., 1980, 102, 939. 28 M. S. Gordon, J. S. Binkley, J. A. Pople, W. J. Pietro and W. J. Hehre, J. Am. Chem. Soc., 1982, 104, 2797. 29 W. J. Pietro, M. M. Francl, W. J. Hehre, D. J. Defrees, J. A. Pople and J. S. Binkley, J. Am. Chem. Soc., 1982, 104, 5039. 30 J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56, 2257. 31 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213. 32 M. S. Gordon, Chem. Phys. Lett., 1980, 76, 163. 33 ASYM40 version 3.0, update of program ASYM20. L. Hedberg and I. M. Mills, J. Mol. Spectrosc., 1993, 160, 117. 34 G. Rauhut and P. Pulay, J. Phys. Chem., 1995, 99, 3093. 35 C. M. Huntley, G. S. Laurenson and D. W. H. Rankin, J. Chem. Soc., Dalton Trans., 1980, 954. 36 J. R. Lewis, P. T. Brain and D. W. H. Rankin, Spectrum, 1997, 15, 7. 37 A. W. Ross, M. Fink and R. Hilderbrandt, in International Tables for Crystallography, ed. A. J. C. Wilson, Kluwer Academic Publishers, Dordrecht, 1992, vol. C, p. 245. Paper 9/02915J

ISSN:1477-9226    

DOI:10.1039/a902915j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2303–2310 2303 1,1,2-Tri-tert-butyldisilane, But 2HSiSiH2But: vibrational spectra and molecular structure in the gas phase by electron diVraction and ab initio calculations † Sarah L. Hinchley,a Bruce A. Smart,a Carole A. Morrison,a Heather E. Robertson,a David W. H. Rankin,*a Robert Zink b and Karl Hassler*b a Department of Chemistry, University of Edinburgh, West Mains Rd., Edinburgh, UK EH9 3JJ b Institut für Anorganische Chemie, Technische Universität Graz, Stremayrgasse 16, A-8010 Graz, Austria Received 24th March 1999, Accepted 3rd June 1999 The molecular structure of 1,1,2-tri-tert-butyldisilane, But 2HSiSiH2But, has been determined in the gas phase by electron diVraction (GED) and ab initio molecular-orbital calculations.Vibrational spectra are consistent with a vapour consisting of one conformer, identified by the structural study as a syn arrangement in which each of the butyl groups eclipses an Si–H bond.Important structural parameters (ra) for the conformer are: Si–Si 236.3(8), Si–C (mean) 191.1(3), C–C 154.5(1), C–H 112.4(1) pm, Si(1)–Si(2)–C(21) 116.0(8), Si(2)–Si(1)–C(11) 111.2(10), Si(2)– Si(1)–C(12) 108.7(9), C(11)–Si(1)–C(12) 121.1(11) and C(21)–Si(2)–Si(1)–H(13) 26.2(11)8, where C(11), C(12) and C(21) are the central carbon atoms of the three tert-butyl groups. These experimental observations are supported by theoretical predictions obtained at the D95*/MP2 level, which also identify two higher-energy conformers.Introduction The electronic spectra of peralkylated silicon backbone polymers in the near-UV region have been found to be surprisingly sensitive to conformational properties of the compound under investigation.1 Considerable variations of the absorption bands are observed as a function of conformation about the Si–Si backbone for both polysilanes and short-chain silanes.2 The structures of some simple disilanes including Si2H6 and Si2Cl6 have been determined previously,3 as have the structures of some partially halogenated disilanes such as 1,1,2,2-tetrabromodisilane, 4 1,2-diiododisilane 5 and 1,1,2,2-tetraiododisilane. 5 Recently, more sterically crowded systems containing tert-butyl groups were studied including 1,2-di-tert-butyldisilane, 6 1,2-di-tert-butyltetrafluorodisilane 7 and 1,2-di-tertbutyltetrachlorodisilane. 8 Ab initio computations have been performed on all of these compounds and, as might be expected on steric grounds, the anti conformation is favoured in all cases.The 1,2-di-tertbutyltetrachlorodisilane is predicted to exhibit three local minima on the potential energy surface at the 6-31G*/MP2 level, with C–Si–Si–C dihedral angles of 56, 94 and 1698, the anti conformer being slightly distorted from the idealised structure. In contrast, for 1,2-di-tert-butyldisilane only two conformers were located at the 6-31G*/SCF level, anti and gauche (C–Si– Si–C dihedral angles of 176.8 and 69.08 respectively).The energy minimum for the gauche structure of 1,2-di-tertbutyldisilane was estimated to lie 5.4 kJ mol21 above that for the anti structure on the potential energy surface and therefore the gauche structure should not be observable by electron diVraction in the gas phase. It was not possible to determine from the GED data how much of the gauche conformer was present, † Supplementary data available: Experimental coordinates and parameters for the GED studies and theoretical geometrical parameters.For direct electronic access see http://www.rsc.org/suppdata/dt/1999/2303/, otherwise available from BLDSC (No. SUP 57577, 10 pp) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/ dalton). although it was certainly less than 20%. For 1,2-di-tertbutyltetrafluorodisilane two conformers, gauche and anti, were predicted from calculations at the 6-31G*/SCF level, with vibrational frequency calculations indicating that both forms represent local minima.However, the barrier to interconversion between gauche and anti was predicted to lie just 0.25 kJ mol21 above the gauche isomer, which may therefore represent a quasiminimum on the potential energy surface rather than a distinct conformer. The experimental structure was modelled with two conformers, but as the C–Si–Si–C dihedral angles refined to 184(7) and 152(3)8, the data are consistent with a single conformer with a large-amplitude motion over a torsional range of around 140–2208 rather than with a mixture of two distinct conformers.In view of the interesting conformational behaviour of di-tert-butyl-substituted disilanes, we have now undertaken a further structural study on a disilane with three tert-butyl groups, 1,1,2-tri-tert-butyldisilane, employing the techniques of vibrational spectroscopy, gas-phase electron diVraction and ab initio calculations. This system is more crowded than the di-tert-butyl substituted disilanes and the structure would be expected to be dominated by steric interactions between these groups.Experimental Synthesis A sample of But 2HSiSiH2But was prepared according to the literature method.9 Ab initio calculations All calculations at the 3-21G*/SCF 10–12 and 6-31G*/SCF13–15 levels were performed on a Dec Alpha 1000 4/200 workstation using the Gaussian 94 program.16 Calculations at the MP2 level using the 6-31G* and D95* 17 basis sets were performed using resources of the U.K.Computational Chemistry Facility, on a DEC 8400 superscalar cluster equipped with 10 fast processors,2304 J. Chem. Soc., Dalton Trans., 1999, 2303–2310 6 GB of memory and 150 GB disk. An extensive search of the torsional potential of 1,1,2-tri-tert-butyldisilane was undertaken at the 3-21G*/SCF level in order to locate all local minima. Three conformers, syn, gauche and antiperiplanar, were located and further geometry optimisations were undertaken at the 6-31G*/SCF level and at the MP2 level using the 6-31G* and D95* basis sets.The D95* basis set is a double zeta basis set that forms molecular orbitals from a linear combination of functions for each atomic orbital and gives a good orbital representation of the first and second row atoms in molecules. Vibrational frequencies were calculated from analytic second derivatives at the 3-21G*/SCF and 6-31G*/SCF levels to determine the nature of stationary points for comparison with experimentally determined frequencies, and the force field provided estimates of amplitudes of vibration (u) for use in the GED refinements.Infrared and Raman spectra Infrared spectra in the range 3200–300 cm21 were measured with a Perkin-Elmer 883 spectrometer using a film of pure liquid between CsBr plates. The Raman spectra were recorded with a Jobin Yvon T64000 triple monochromator employing a charge-coupled device (CCD) camera and the 514.5 nm line of an argon-ion laser as the source of excitation.Variabletemperature Raman spectra were obtained by mounting a capillary containing the sample on a copper block equipped with a heater and a thermocouple. Liquid nitrogen was used for cooling the sample. Electron diVraction Electron scattering intensities were recorded on Kodak Electron Image plates using the Edinburgh gas diVraction apparatus operating at ca. 44.5 kV (electron wavelength ca. 5.6 pm).18 Nozzle-to-plate distances for the metal inlet nozzle were ca. 94 and 259 mm yielding data in the s range 20–356 nm21; three plates were exposed at each camera distance. The sample and nozzle temperatures were maintained at ca. 411 K during the exposure periods. The scattering patterns of benzene were also recorded for the purpose of calibration; these were analysed in exactly the same way as those for But 2HSiSiH2But so as to minimise systematic errors in the wavelengths and camera distances.Nozzle-to-plate distances, weighting functions used to set up the oV-diagonal weight matrix, correlation parameters, final scale factors and electron wavelengths for the measurements are collected in Table S1 (SUP 57577). The electron-scattering patterns were converted into digital form using a computer-controlled Joyce Loebl MDM6 microdensitometer with a scanning program described elsewhere.19 The programs used for data reduction 19 and least-squares refinement20 have been described previously; the complex scattering factors were those listed by Ross et al.21 Results Theoretical computations A series of ab initio molecular-orbital calculations was undertaken to investigate the structure of 1,1,2-tri-tert-butyldisilane (Fig. 1). An extensive search of the torsional potential led to the location of three minima, conformers syn, t[C(21)–Si(2)–Si(1)– H(13)] = 24.2, gauche, t[C(21)–Si(2)–Si(1)–H(13)] = 63.4, and antiperiplanar, t[C(21)–Si(2)–Si(1)–H(13)] = 163.88.The twist about the silicon–silicon bond is uniquely described by this torsion angle. Vibrational frequency calculations at the 6-31G*/ SCF level confirm that all three forms represent local minima on the potential energy surface. However, the syn structure was found to be 10.8 kJ mol21 lower in energy than the gauche structure and 10.3 kJ mol21 below the antiperiplanar structure. This would equate to a mixture containing 96.3% of the syn conformer and 2.3% and 1.4% of the gauche and antiperiplanar conformers, respectively, at room temperature.Attention will therefore be paid mainly to the syn structure. The molecular geometry of this conformer for the D95*/MP2 calculation is presented in Table 1; those calculated at the 3-21G*/SCF, 6-31G*/SCF and 6-31G*/MP2 levels of theory are presented in Table S2 (SUP 57577). The molecular geometries of the gauche and antiperiplanar conformers calculated at the 3- 21G*/SCF, 6-31G*/SCF, 6-31G*/MP2 and D95*/MP2 levels are presented in Table S3 (SUP 57577), and relative energies are given in Table 2.The nomenclature used to define the three conformers describes the positions of the tert-butyl groups at the But 2HSi end of the molecule relative to the third tert-butyl group. Fig. 1 (a) Molecular structure and (b) Newman projection of the syn conformer, viewed down the Si(2)–Si(1) bond, of But 2HSiSiH2But. Table 1 Theoretical geometrical parameters (D95*/MP2 level) for the syn conformer of 1,1,2-tri-tert-butyldisilane a Si(1)–Si(2) Si(1)–C(11) C(11)–C(111) C(11)–C(112) C(11)–C(113) Si(1)–C(12) C(12)–C(121) C(12)–C(122) C(12)–C(123) Si(2)–C(21) C(21)–C(211) C(21)–C(212) C(21)–C(213) Si(1)–H(13) Si(2)–H(22) Si(2)–H(23) C–Hb 237.5 192.3 153.8 154.2 154.2 192.4 154.2 153.9 153.9 191.5 153.9 154.0 153.9 150.5 149.8 149.9 110.1 C(11)–Si(1)–C(12) Si(1)–Si(2)–C(21) Si(2)–C(21)–C(211) Si(2)–C(21)–C(212) Si(2)–C(21)–C(213) Si(2)–Si(1)–C(11) Si(1)–C(11)–C(111) Si(1)–C(11)–C(112) Si(1)–C(11)–C(113) Si(2)–Si(1)–C(12) Si(1)–C(12)–C(121) Si(1)–C(12)–C(122) Si(1)–C(12)–C(123) Si(1)–Si(2)–H(22) Si(1)–Si(2)–H(23) Si(2)–Si(1)–H(13) C–C–Hb C(21)–Si(2)–Si(1)–H(13) 118.6 112.6 109.7 109.2 110.6 110.4 112.2 111.3 107.6 108.8 106.8 111.8 112.5 110.4 110.9 107.2 111.1 24.2 a All distances in pm, all angles in degrees.See Fig. 1 for atom numbering. b Average value.J. Chem. Soc., Dalton Trans., 1999, 2303–2310 2305 As expected, since this system contains no multiple bonds or lone pairs of electrons, the molecular geometry of the conformer proved to be insensitive to changes in the theoretical method.For this reason, only the highest level results (D95*/ MP2) will be discussed. The molecular geometry of the conformer appears to be dictated predominately by steric interactions, as evident in the calculated values for the C(11)–Si(1)–C(12) angle. In the syn conformer, the C(11)–Si(1)–C(12) angle is predicted to be 118.6 compared to 109.58 for an ideal tetrahedral geometry.Further evidence of steric repulsion is found in the value of the Si(1)– Si(2)–C(21) angle, 112.68. On the other hand, the calculated Si(2)–Si(1)–C(11) angle shows very little deviation from the parent tetrahedral angle of 109.58 (110.48), as does the Si(2)– Si(1)–C(12) angle (108.88). The Si(1)–Si(2)–C(21) angle is similar to the Si–Si–C angles calculated for 1,2-di-tertbutyldisilane 6 (114.48).These structural changes relative to the idealised tetrahedral angle of 109.58 serve to reduce the steric interactions in these systems; however, resultant nearest neighbour H ? ? ? H distances were still predicted to be 219 pm, as compared to 240 pm for the sum of the van der Waals radii of two hydrogen atoms. Internal C–C–C angles indicate that the tert-butyl groups are not significantly distorted from local C3 symmetry. Bond lengths are generally within the expected range based on the results obtained previously for disilanes with tert-butyl groups.For example, the Si–Si bond length was predicted to be 237.5 as compared to 236.8 pm in 1,2-di-tertbutyldisilane. 6 All the C–C bond lengths fell within the range 153.8–154.2, and the Si–C distances are all in the range 191.5– 192.4 pm. These Si–C bond lengths are longer than those of normal Si–C bonds, for example 188.2(1) and 188.6(1) pm for 1,4-disilabutane and 1,5-disilapentane,22 but compare well with the calculated Si–C bond length in 1,2-di-tert-butyldisilane 6 (191.9 pm), which may be a further demonstration of steric interactions in these crowded disilanes, although it could be an electronic eVect of the electron-releasing tert-butyl groups.The molecular geometries of the gauche and antiperiplanar conformers also appear to be dictated predominately by steric interactions, as evident in the predicted values for the Si(1)– Si(2)–C(21) angles. In the gauche conformer, the Si(1)–Si(2)– C(21) angle is predicted to be 122.08 and, as might be expected, the same angle is predicted to be even wider in the antiperiplanar structure (124.78) since all the tert-butyl groups are in closer proximity in this conformer.Further evidence of steric repulsion is found in the values of the C(11)–Si(1)– C(12) angles in the two conformers, predicted to be 115.88 in the gauche conformer but again, as expected, rather wider at 117.58 in the antiperiplanar conformer.The Si(2)–Si(1)–C(11) angles in both conformers show a much less dramatic deviation from the parent tetrahedral angle of 109.58 (gauche 107.2, antiperiplanar 111.58). The predicted Si(2)–Si(1)–C(12) angles show a larger deviation (gauche 113.4, antiperiplanar 113.78) and, again, these angles are similar to Si–Si–C angles calculated for 1,2-di-tert-butyldisilane 6 (114.48). The Si–Si bond lengths were predicted to be 237.3 and 237.6 pm in the gauche and the antiperiplanar conformers respectively.These bond lengths are very similar to the predicted value for the syn conformer and agree well with those found for 1,2-di-tertbutyldisilane. 6 Table 2 Relative energies (kJ mol21) of the syn, gauche and antiperiplanar conformers of 1,1,2-tri-tert-butyldisilane Basis Set/Level syn gauche antiperiplanar 3-21G*/SCF 6-31G*/SCF 6-31G*/MP2 D95*/MP2 0000 12.7 11.1 10.1 10.8 14.8 12.1 10.2 10.3 (b) Vibrational spectra and rotational isomerism As mentioned above, calculations predict the existence of three conformers, syn, gauche and antiperiplanar, on the potential energy surface of But 2HSiSiH2But with the high energy conformations (antiperiplanar and gauche) lying 10.3 and 10.8 kJ mol21 respectively above the syn structure. The GED data can be fitted with the single syn structure corresponding to the global minimum.Vibrational spectroscopy should be a slightly more sensitive tool for the detection of the less stable rotamers whose presence in the conformational mixture should be very small according to their predicted energies. In particular, variable-temperature Raman spectroscopy has proven to be an extremely useful tool for conformational analyses as Ramanactive skeletal modes are usually very sensitive to the backbone conformation.For example, the energy diVerence between the anti and gauche rotamers of MeCl2SiSiCl2Me has been determined recently from temperature-dependent Raman intensities. 23 In the present study we have recorded the infrared spectrum of liquid But 2HSiSiH2But, variable-temperature Raman spectra in the temperature range from 25 8C to 150 8C, and the Raman spectrum of the solid. Selected vibrational spectra of liquid and solid But 2HSiSiH2But are summarised in Table 3. Calculated and observed vibrational wavenumbers are compared in Table 4. Using the program ASYM40,24 the calculated Cartesian Hessian matrices were converted into symmetry force fields resulting in a description of normal modes in terms of symmetry coordinates according to the potential energy distributions.Most of the normal coordinates of the molecule But 2- HSiSiH2But are dominated by more than just a single symmetry coordinate and the description given in Table 4 is highly approximate. Its primary use is to help with labelling rather than to permit an accurate visualisation of the vibrational motions. For reasons of clarity and simplicity high-frequency vibrations involving the methyl groups (ns,asCH3 and ds,asCH3) are omitted from Table 4.These vibrations are well known, not sensitive to the conformation around the Si–Si bond and therefore unimportant from the viewpoint of rotational isomerism. The following discussion and characterisation of vibrational frequencies refers explicitly to the global minimum, the syn conformer, if not stated otherwise. The eighteen rocking vibrations of the methyl groups occur in three main spectral regions and have been summarised by their respective wavenumber ranges using the labels r1CH3, r2CH3 and r3CH3, respectively.Wavenumber ranges have also been used for the modes ns,asCC3, ds,asCC3, rCC3 and the torsional vibrations about the C–C single bonds (tCC), because of the large number of each type of these vibrational modes. Further, calculations predict that the symmetry coordinates rCC3 and tCC are strongly mixed with each other, implying that the torsional vibrations around the C–C bonds, which usually elude observation in the case of smaller molecules like ButSiX3 (X = halogen),25 gain intensity.Therefore, no attempt was made to describe normal modes dominated by both rCC3 and tCC by a single symmetry coordinate and only one wavenumber range for these vibrations has been given. The three rocking vibrations labelled by r3CH3 correspond to a2 modes if the tert-butyl fragment assumes ideal C3v symmetry and therefore are not intense enough to be observed in the IR and Raman.The assignment of the modes nsCC3 and nasCC3 to experimental wavenumbers is straightforward and agrees well with that made for molecules ButSiX3.25 The mode dSiH2 appears in the same spectral region as the asymmetric C–C stretching modes nasCC3 and could not be distinguished experimentally from the latter. The wagging mode gSiH2 is usually of high intensity in the IR but low intensity in the Raman and is assigned to the shoulder at 840 cm21 (IR) and the weak Raman band at 841 cm21 in the spectrum of the solid.The mode rSiC2 is ascribed to the strong IR band at 793 cm21 (7922306 J. Chem. Soc., Dalton Trans., 1999, 2303–2310 cm21 in the Raman) and appears at a somewhat lower wavenumber than predicted by the calculations with respect to the position of the nsCC3 vibration. The three Si–C stretching modes, nasSiC2, nSiC and nsSiC2 are readily attributed to the vibrational bands at 613, 595 and 575 cm21 (IR) or 617, 594 and 575 cm21 (Raman, solid), respectively.The origin of the strong IR band at 656 cm21 remains unexplained. As can be seen from Table 4, the modes described by dSiSiH (angle bending), tSiH2 (twist), rSiH2 (rocking) and nSiSi are predicted to be highly sensitive to the conformation about the Si–Si bond. The modes dSiSiH and tSiH2 of the syn structure are assigned to the strong and very broad IR peak at 707 cm21 and the Raman band at 710 cm21 which splits into two upon solidification (702 and 711 cm21).The shoulder at 740 cm21 (IR) and the weak shoulder around 730 cm21 in the Raman spectrum (intensity of the shoulder increasing with temperature) might be due to the mode dSiSiH of a high energy conformer, perhaps the antiperiplanar structure. Similarly, the shoulder at 770 cm21 (IR) could Table 3 Vibrational spectra of 1,1,2-tri-tert-butyldisilane (<2250 cm21) a IR (l, 25 8C) Raman (l, 25 8C) Raman (s) 2104vs 2080vs 1210 (sh) 1200s 1188s 1163ms 1089s 1070 (sh) 1035ms 1012vs — 935 (sh) 927vs 890vw (sh) — 840 (sh) 818vs — 793vs 770 (sh) 740 (sh) 707vs, br — 656s 613m 595 (sh) 575s 501s 479m 465vw (sh) 435ms — 410vw 387m — 370m — 349s —————————— 2104vvs 2080s — 1201s 1189 (sh) — 1090vw — 1034vw 1014w 1003vw 939ms 928 (sh) 888vvw 862vvw — 824s — 794w — 730 (sh) 710m — 654vw — 593vs 575ms 502 (sh) 479s 460 (sh) 434w 424w 408vvw — 382m 372 (sh) — 348w 305w 272w 242w 219s — 185vw — 152w 134ms 100w 2108vvs 2076s — 1203vs 1192s 1178 (sh) 1089vw — 1035vw 1015m 1004vw 940s 927m 890vvw 860vvw 841vw 825vs 816 (sh) 792w —— 711mw 702mw 654vw 617w 594s 575m 502w 481s 461w 434w 423vw 407vw 385 (sh) 382m 370w 353 (sh) 348w 314w 281mw, br 245w 220ms 206s 187vvw 165 (sh) 154w 135m 103w a Si–H stretching vibrations are included.Key: vvw = very very weak, vw = very weak, w = weak, mw = medium weak, m = medium, ms = medium strong, s = strong, vs = very strong, vvs = very very strong, sh = shoulder, br = broad. also stem from another rotamer (possibly from the gauche structure).However, due to the high probability of the presence of strong combination bands or overtones in the IR spectrum of a molecule with 126 fundamental modes care must be taken when stating evidence for the presence of more than just a single rotamer. A slightly stronger argument in favour of the presence of high energy rotamers in liquid But 2HSiSiH2But is provided by the appearance of several Raman peaks around 479 cm21.The Raman peaks at 479 and 502 cm21 are assigned to nSiSi and rSiH2 of the syn conformation, respectively. However, the intensity of the weak shoulder at 460 cm21 seems to increase slightly with temperature, as shown in Fig. 2, and could be due to one or both of the modes nSiSi of the high-energy conformations, which are predicted to diVer by approximately 20 cm21 from the value of the syn structure. Vibrations below 460 cm21 do not provide any additional information about rotational isomerism and will not be discussed due to the large number of vibrations and the highly approximate description of these modes by local symmetry coordinates.The lowest wavenumbers corresponding to torsional vibrations around Si–C (tSiC) and Si–Si (tSiSi) bonds elude observation in the vibrational spectra. It can be summarised that the present study of the rotational isomerism of But 2HSiSiH2But employing IR spectroscopy at ambient temperature and variable temperature Raman spectroscopy is consistent with a single conformer, in accordance with the calculations which predict only 2.3% of the gauche and 1.4% of the antiperiplanar conformers at room temperature.A more sensitive technique like matrix-isolation spectroscopy seems more suitable for unambiguously proving the existence of the three backbone conformers. Electron diVraction analysis On the basis of the ab initio calculations described above, electron-diVraction refinements were carried out using a model of the syn conformation (C1 symmetry) to describe the vapour.The conformer is in C1 symmetry rather than Cs due to the twists of the tert-butyl groups in the But 2Si fragment to avoid methyl ? ? ? methyl interactions. The large number of geometric parameters needed to define the model made it necessary to make a number of assumptions including local C3v symmetry for all methyl groups and local C3 symmetry for the tert-butyl groups.Initially, some of the diVerences between similar bond lengths and bond angles were restrained using the SARACEN26 method. However, since many of these diVerence parameters proved to be uncorrelated with other refining parameters, and returned values and e.s.d.s which were close to the restraints, they were fixed in the final refinement. We can therefore be confident that the refined parameters, and their e.s.d.s, are not aVected by the assumptions applied to the molecular model.Fig. 2 Portion of the Raman spectrum of liquid But 2HSiSiH2But at 25 and 150 8C.J. Chem. Soc., Dalton Trans., 1999, 2303–2310 2307 Table 4 Calculated and observed wavenumbers for 1,1,2-tri-tert-butyldisilane Approximate ab initio (unscaled) ab initio (scaled by 0.92) observed (IR, l, 25 8C) observed (Raman, l, description syn gauche antiperiplanar syn gauche antiperiplanar syn 25 8C) syn nsSiH2 nasSiH2 nSiH r1CH3 r2CH3 r3CH3 nasCC3 dSiH2 gSiH2 nsCC3 rSiC2 dSiSiH tSiH2 nasSiC2 nSiC nsSiC2 rSiH2 nSiSi ds,asCC3 rCC3, tCC dSiC2 gSiC2 tSiC2 dSiSiC tSiC tSiSi 2329.1 2320.1 2298.7 1348.9–1316.1 1136.7–1122.7 1052.5–1050.2 1028.3–1017.7 1042.9 923.3 889.0–883.5 878.8 801.4 782.3 649.1 628.7 611.5 559.3 518.3 472.7–372.4 342.0–219.7 159.9 136.6 140.4 96.2 62.6–29.5 48.2 2335.7 2312.6 2300.8 1347.9–1316.7 1136.8–1124.3 1051.6–1048.6 1030.0–1017.8 1027.8 926.3 889.8–883.5 874.0 850.6 772.9 644.7 631.1 610.2 551.5 499.3 459.2–368.6 349.5–218.1 153.1 139.3 128.7 87.6 100.5–44.0 32.5 2326.4 2319.3 2292.5 1348.4–1316.6 1136.4–1123.9 1051.8–1049.0 1030.0–1015.9 1029.0 903.0 889.1–883.9 880.9 825.7 799.2 649.4 634.2 611.9 540.7 500.4 459.7–360.2 343.8–223.1 145.7 146.9 129.5 101.1 76.5–48.8 26.9 2143 2134 2115 1241–1211 1046–1033 968–966 946–936 959 849 818–813 808 737 720 597 578 563 515 477 435–343 315–202 147 126 129 89 58–27 44 2149 2128 2117 1240–1211 1046–1034 967–965 948–936 946 852 819–813 804 783 711 593 581 561 507 459 422–339 322–201 141 128 118 81 92–40 30 2140 2134 2109 1241–1221 1045–1034 968–965 948–935 947 831 818–813 810 760 735 597 583 563 497 460 423–331 316–205 134 135 119 93 70–45 25 2104 2104 2080 1210/1200/1188 1012 — 935/927 935 or 927 840 818 793 707 707 613 595 575 501 479 435/387/ 370/349 — —————— 2104 2104 2080 1201/1189 1014/1003 — 939/928 939 or 928 — 824 794 710 710 593 593 575 502 479 434/382/ 372/348 305/272/ 242/219/ 185 152 134 134 100 —— The structure of But 2HSiSiH2But was finally defined in terms of twenty-seven independent geometric parameters, comprising five bond lengths, six bond angles and sixteen torsion, rock and tilt parameters (Table 5; atom numbering shown in Fig. 1). Table 5 Refined and calculated geometric parameters for 1,1,2-tritert- butyldisilane (distances in pm, angles in 8) from the GED study a No. Parameter b GED (ra) D95*/MP2 (rc) p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 C–H C–C Si–Si Si–C (mean) Si–H (mean) CCH CCC SiSiH average SiSiC average SiSiC diVerence 1 SiSiC diVerence 2 Me twist Me tilt Me rock But twist average But twist diVerence 1 But twist diVerence 2 But rock (gpA) But rock (gpB) But rock (gpC) But tilt (gpA) But tilt (gpB) But tilt (gpC) C twist average C twist diVerence 1 H twist average HSiSiC 112.4(1) 154.5(1) 236.3(8) 191.0(3) 149.7(10) 110.1(6) 108.5(2) 109.3(11) 112.0(6) 4.8(10) 7.3(11) 58.4(22) 24.4(11) 2.0(21) 62.0(14) 212.3(20) 20.3(16) 2.4(11) 4.0(10) 24.7(9) 23.0(10) 22.0(9) 22.4(10) 112.1(7) 20.3(11) 122.0(11) 26.2(11) 110.1 154 237.5 192.1 150.1 111.1 108.7 109.5 110.6 2.2 3.8 61.4 —— 61.5 28.7 21.3 —————— 114.1 0.2 121.3 24.2 a Figures in parentheses are the estimated standard deviations of the last digits.See text for parameter definitions. b gpA = tert-butyl group with C(21) at centre; gpB = tert-butyl group with C(11) at centre; gpC = tert-butyl group with C(12) at centre.The independent parameters include the C–H and C–C bond lengths (p1 and p2). Average bond lengths were used for the Si– Si, Si–C and Si–H bond lengths (p3–p5), with small diVerences between non-equivalent bond lengths fixed at the ab initio values. All C–C–H bond angles (p6) were assumed to be identical, as were all C–C–C bond angles (p7). An average value was adopted for the three Si–Si–H angles (p8), with the small diVerences from the mean being set at the ab initio values.The Si–Si–C angles were defined in terms of an average (p9) of Si(1)–Si(2)–C(21), Si(2)–Si(1)–C(11) and Si(2)–Si(1)–C(12), and two diVerence parameters, which were included in the refinement procedure since the predicted Si–Si–C angles spanned a wide range of values. The diVerences were described as the diVerences between Si(1)–Si(2)–C(21) and Si(2)–Si(1)– C(11) (p10) and between Si(1)–Si(2)–C(21) and Si(2)–Si(1)– C(12) (p11). Of the remaining sixteen parameters, nine represent the tilts, rocks and torsions of the three tert-butyl groups.These groups were generated by initially placing a methyl group carbon at the origin, with its three H atoms arranged with local C3v symmetry about the x-axis and one H in the xy plane in the positive x and y directions. The methyl torsion, tilt and rock parameters, (p12– p14) are rotations about the local x-, z-, and y-axes respectively. The methyl group is then translated along the positive x-axis by the C–C bond length and the central carbon of the tert-butyl group is placed at the origin.The correct C–C–C bond angles are generated by rotating the methyl group about the z-axis, moving the methyl carbon atom in the positive y direction, and then generating the other methyl groups by rotation of the first group about the x-axis by 120 or 21208, respectively. The tertbutyl torsion angle is a rotation of the group about the x-axis. Parameters introduced here for the tert-butyl torsions include an average (p15) of torsions C(211)–C(21)–Si(2)–Si(1), C(111)– C(11)–Si(1)–Si(2), and C(121)–C(12)–Si(1)–Si(2), and two differences.These were the diVerence between torsion C(211)– C(21)–Si(2)–Si(1) and torsion C(111)–C(11)–Si(1)–Si(2) or C(121)–C(12)–Si(1)–Si(2) (p16 and p17). The rock and tilt parameters are rotations of the whole tert-2308 J. Chem. Soc., Dalton Trans., 1999, 2303–2310 butyl groups about the y-axis and the z-axis respectively. Three individual rocks (p18–p20) and three individual tilts (p21–p23) were introduced here for the tert-butyl groups with C(21), C(11) and C(12) as the central atoms of the groups.A positive rock would move the tert-butyl group with C(21) at the centre away from that with C(12) at the centre whilst the tert-butyl with C(11) at the centre would be moved towards C(12) and C(12) would be moved away from C(11), all in the local y direction of the tertbutyl groups. Positive tilts would move the tert-butyl groups at one end of the molecule towards the group at the other end in the local z direction, and vice versa.Having generated the tert-butyl groups in their local coordinate systems, they need to be rotated about the x-axis to put them in the correct position relative to the silicon atoms. The two tert-butyl groups and the hydrogen attached to Si(1) were initially placed in the xy plane, and the tert-butyl groups were then rotated about the x-axis. These rotations are defined in terms of an average of C(11)–Si(1)–Si(2)–H(13) and C(12)– Si(1)–Si(2)–H(13) (p24) and a diVerence between torsion C(11)– Si(1)–Si(2)–H(13) and torsion C(12)–Si(1)–Si(2)–H(13) (p25).The tert-butyl group and H atoms attached directly to Si(2) were placed in the xy plane and the two hydrogen atoms were then rotated about the x-axis in opposite directions by torsions H(22)–Si(2)–Si(1)–C(21) and H(23)–Si(2)–Si(1)–C(21). The average of these two dihedrals is (p26) and the diVerence was set at the ab initio value.Finally, the dihedral angle C(21)–Si(2)–Si(1)–H(13) (p27) described the overall conformation about the Si–Si bond, with a value of zero indicating the conformation in which the hydrogen of the But 2HSi group and the carbon of the ButH2Si group were eclipsing one another. The starting parameters for the ra refinement were taken from the theoretical geometry optimised at the D95*/MP2 level. The Ra structure was not refined due to the fact that the rectilinear vibrational corrections (i.e.parallel and perpendicular correction terms) are known to be unreliable for a molecule this size with many low lying vibrational modes. Theoretical (6- 31G*/SCF) Cartesian force fields were obtained and converted into force fields described by a set of symmetry coordinates using a version of the ASYM40 program24 modified to work for molecules with more than forty atoms. All geometric parameters were then refined.In total twenty-seven geometric parameters and forty-three groups of vibrational amplitudes were refined. Flexible restraints were employed during the refinement using the SARACEN method.26 In total, twenty-one geometric and thirty-seven amplitude restraints were employed. These are listed in Tables S4 and S5 (SUP 57577). The success of the final refinement, for which RG = 0.060 (RD = 0.052), can be assessed on the basis of the molecular scattering intensity curves (Fig. 3) and the radial distribution curve (Fig. 4). Final refined parameters are listed in Table 5, Fig. 3 Experimental and final weighted diVerence (experimental 2 theoretical) molecular-scattering intensities for 1,1,2-tri-tert-butyldisilane. interatomic distances and the corresponding amplitudes of vibration in Table 6, with the least-squares correlation matrix shown in Table S6 and the experimental coordinates from the GED analysis in Table S7 (SUP 57577). In the SARACEN analysis, all correlation between refining parameters is included in the error estimates by the use of flexible restraints.We therefore quote the estimated standard deviations, s, and believe that these are realistic estimates of the uncertanties of the parameters. Fig. 1 shows a perspective view of the syn conformer of But 2HSiSiH2But in the optimum refinement of the GED data with a Newman projection along the Si–Si bond vector showing the syn conformation. Discussion Theoretical and experimental studies show that 1,1,2-tri-tertbutyldisilane exists essentially as a single syn conformer in the gas phase.The electron diVraction data for the compound were fitted using the SARACEN26 method on the basis of such an syn structure. The vibrational spectra do not change significantly with changes in the temperature, indicating the presence of one con- Fig. 4 Experimental and diVerence (experimental 2 theoretical) radial-distribution curves, P(r)/r, for But 2HSiSiH2But.Before Fourier inversion the data were multiplied by s.exp(20.00005s2)/(ZSi 2 fSi)/ (ZC 2 fC). Table 6 Selected interatomic distances and mean amplitudes of vibration for 1,1,2-tri-tert-butyldisilane from the GED study a No. Atom pair ra/pm u/pm 123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Si(1)–Si(2) Si(2)–C(21) Si(1)–C(11) Si(1)–C(12) Si(2)–H(23) Si(2)–H(22) Si(1)–H(13) C–C C–H C(21) ? ? ? H(2111) C(211) ? ? ? C(212) Si(1) ? ? ? C(21) Si(2) ? ? ? C(211) Si(2) ? ? ? C(212) Si(2) ? ? ? C(213) Si(2) ? ? ? C(11) Si(1) ? ? ? C(111) Si(1) ? ? ? C(112) Si(1) ? ? ? C(113) Si(2) ? ? ? C(12) Si(1) ? ? ? C(121) Si(1) ? ? ? C(122) Si(1) ? ? ? C(123) C(11) ? ? ? C(12) 236.3(8) 190.4(3) 191.7(3) 191.3(3) 149.3(10) 149.7(10) 150.0(10) 154.5(1) 112.4(1) 220.7(25) 250.8(3) 362.7(14) 283.3(19) 278.6(17) 289.8(18) 354.1(19) 279.7(18) 292.4(16) 282.8(16) 348.5(18) 293.5(51) 279.4(53) 280.8(106) 333.4(20) 7.2(6) 6.1(7) 6.2(6) 6.2(6) 9.4(2) 9.4(tied to u5) 9.4(tied to u5) 5.2(2) 6.9(2) 9.7(12) 7.3(5) 13.9(12) 10.7(5) 10.3(7) 10.9(7) 14.0(12) 10.3(7) 10.7(7) 10.8(7) 12.8(9) 10.4(7) 10.2(7) 10.0(7) 10.0(11) a See Fig. 1 for atom numbering. (Other atom pairs were also used in the refinement but are not shown here.)J. Chem. Soc., Dalton Trans., 1999, 2303–2310 2309 former. Spectroscopic studies are therefore consistent with the GED experiment and theory, but do not unambiguously prove that there is only one conformer present in the vapour.The final experimental structure is in good agreement with that calculated ab initio at the D95*/MP2 level; computed bond lengths and angles generally fall within 1–2 pm or 1–28 of the GED values (Table 5). For example, the Si–Si bond length refined to 236.3(8) pm as compared to the computed value of 237.5 pm. The mean C–C bond length refined to 154.5(1) pm compared to 154.0 pm (mean) from the calculations and the experimental range of Si–C bond lengths was 190.4–191.7 pm compared to the calculated range of 191.5–192.4 pm.However, the Si(1)–Si(2)–C(21) bond angle refined to 116.0(8)8, a lot wider than the predicted value of 112.68, and the C(11)–Si(1)– C(12) bond angle refined to 121.1(11)8 compared to the calculated value of 118.68. Both these observations serve to highlight the significant steric interactions within the molecule. The torsion about the Si–Si axis, dihedral angle C(21)–Si(2)–Si(1)– H(13), which uniquely describes the position of all the groups about the Si–Si axis, agrees reasonably well with the predicted value; 26.2(11)8 vs. 24.28. Observed geometric parameters are generally consistent with those for a number of other closely related compounds. For example, the Si–Si bond distance in the syn conformer of 1,1,2- tri-tert-butyldisilane [236.3(8) pm] is within the range of values found for other disilanes from GED refinements including 1,2-di-tert-butyldisilane 6 [234.8(3) pm], 1,1,2,2-tetrabromodisilane 4 [ 234.9(19) pm], 1,2-diiododisilane 5 [238.0(34) pm] and 1,1,2,2-tetraiododisilane 5 [238.9(37) pm], but a little longer as might be expected, either on steric or electronic grounds, with the tert-butyl groups being electron-donating.Refined values of the C–C [154.5(1) pm] and Si–H [149.7–150.0 pm] bond lengths are in excellent agreement with calculated values and compare well with other previously reported bond lengths,22 as would be expected. The most striking feature of the structure is the deviation of the C(11)–Si(1)–C(12) bond angle from the “pure” sp3 tetrahedral angle [109.58] by 11.68.This provides evidence of steric strain and the wide angle observed is probably caused by the close proximity of two of the tert-butyl groups at one end of the molecule. It also reflects the easy deformation of angles at silicon, which allows the accommodation of several large substituents. Another structural feature of note is the value obtained for the Si(1)–Si(2)–C(21) bond angle [116.0(8)8].This angle is similar to those previously observed for 1,2-di-tertbutyldisilane 6 [113.7(3)8] and 1,2-di-tert-butyltetrafluorodisilane 7 [ 114.6(7)8] and provides evidence for significant steric interaction in all the tert-butyldisilanes due to the individual tert-butyl groups. Much larger Si(1)–Si(2)–C(21) bond angles are observed in the gauche and antiperiplanar conformers [122.5(8) and 125.2(9)8 respectively]. This can be attributed to the closer interactions between the tert-butyl groups at either end of the molecule.The values of the C(11)–Si(1)–C(12) angles refined to 114.5(9) for the gauche conformer and 117.2(8)8 for the antiperiplanar conformer. These are smaller than that observed for the syn conformer but are still significantly distorted from the idealised tetrahedral angle, again indicating the easy deformation of angles at silicon atoms to accommodate bulky substituents. In the early stages of this analysis, before the existence of the syn conformer had been recognised, the experimental data were also fitted with a mixture of the other two conformers, gauche {t[C(21)–Si(2)–Si(1)–H(13)] = 63.4} and antiperiplanar {t[C(21)–Si(2)–Si(1)–H(13)] = 163.88}, in equal amounts, as predicted ab initio.All geometric parameters were then refined before determining the relative weights of the two conformations. The final weight of the gauche conformer was thus determined to be 50.8% with a standard deviation of 3.2% according to the Hamilton test for this parameter.27 From the final refinement, for which RG = 0.057 (RD = 0.054), it can be seen that this two conformer model fits the experimental data as well as the single syn model.This demonstrates that caution must be exercised when initially exploring the potential energy surface to locate all structurally stable minima and to determine the diVerences in energy between them. The two-conformer model used forty-two independent geometric parameters comprising five bond lengths, nine bond angles and twenty-eight torsion, rock and tilt parameters.This large number of refinable parameters probably contributed to the overall goodness of fit of these two conformers compared to the single syn conformer. Mixtures of all three conformers will also fit the data well. However, we believe that the refinement based on the syn conformer alone is the most satisfactory result, in the light of all available information, both theoretical and experimental.Acknowledgements We thank the EPSRC for financial support of the Edinburgh Electron DiVraction Service (grant GR/K44411), for the provision of microdensitometer facilities at the Daresbury Laboratory and for the Edinburgh ab initio facilities (grant GR/ K04194). We also thank Dr. V. Typke of the University of Ulm for the variable-array version of ASYM40, and the U.K. Computational Chemistry Facility (admin: Department of Chemistry, King’s College London, Strand, London WC2R 2LS) for the computing time on Columbus.References 1 R. D. Miller and J. Michl, Chem. Rev., 1989, 89, 1359. 2 See, for example, B. Albinsson, H. Teramae, J. W. Downing and J. Michl, Chem. Eur. J., 1996, 2, 529. 3 V. S. Mastryukov, in Stereochemical Applications of Gas-Phase Electron DiVraction, eds. I. Hargittai and M. Hargittai, VCH, Weinheim, 1990, vol. B, p. 1. 4 H. Thomassen, K. Hagen, R. Stølevik and K.Hassler, J. Mol Struct., 1986, 147, 331. 5 E. Røhmen, K. Hagen, R. Stølevik, K. Hassler and M. Pöschl, J. Mol Struct., 1991, 244, 41. 6 D. Hnyk, R. S. Fender, H. E. Robertson, D. W. H. Rankin, M. Bühl, K. Hassler and K. Schenzel, J. Mol. Struct., 1995, 346, 215. 7 B. A. Smart, H. E. Robertson, N. W. Mitzel, D. W. H. Rankin, R. Zink and K. Hassler, J. Chem. Soc., Dalton Trans., 1997, 2475. 8 S. L. Hinchley, B. A. Smart, C. A Morrison, H. E. Robertson, D. W. H. Rankin, R.Zink and K. Hassler, unpublished results. 9 B. Reiter and K. Hassler, J. Organomet. Chem., 1994, 467, 21 10 J. S. Binkley, J. A. Pople and W. J. Hehre, J. Am. Chem. Soc., 1980, 102, 939. 11 M. S. Gordon, J. S. Binkley, J. A. Pople, W. J. Pietro and W. J. Hehre, J. Am. Chem. Soc., 1982, 104, 2797. 12 W. J. Pietro, M. M. Francl, W. J. Hehre, D. J. DeFrees, J. A. Pople and J. S. Binkley, J. Am. Chem. Soc., 1982, 104, 5039. 13 W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56, 2257. 14 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213. 15 M. S. Gordon, Chem. Phys. Lett., 1980, 76, 163. 16 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheesman, T. A. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez and J. A. Pople, Gaussian 94 (Revision C.2), Gaussian Inc., Pittsburgh, PA, 1995. 17 T. H. Dunning, Jr. and P. J. Hay, in Modern Theoretical Chemistry, ed. H. F. Schaefer III, Plenum Press, New York, 1976, p. 1. 18 C. M. Huntley, G. S. Laurenson and D. W. H. Rankin, J. Chem. Soc., Dalton Trans., 1980, 954. 19 S. Cradock, J. Koprowski and D. W. H. Rankin, J. Mol. Struct., 1981, 77, 113. 20 A. S. F. Boyd, G. S. Laurenson and D. W. H. Rankin, J. Mol. Struct., 1981, 71, 217.2310 J. Chem. Soc., Dalton Trans., 1999, 2303–2310 21 A. W. Ross, M. Fink and R. Hilderbrandt, in International Tables for Crystallography, ed. A. J. C. Wilson, Vol. C, Kluwer Academic Publishers, Dordrecht, 1992, p. 245. 22 See, for example, N. W. Mitzel, B. A. Smart, A. J. Blake, H. E. Robertson and D. W. H. Rankin, J. Phys. Chem., 1996, 100, 9339. 23 For example, see M. Ernst, K. Schenzel, A. Jähn and K. Hassler, J. Mol. Struct., 1997, 412, 83. 24 L. Hedberg and I. M. Mills, J. Mol. Spectrosc., 1993, 160, 117. 25 R. Zink and K. Hassler, Spectrochimica Acta, Part A, 1999, 55, 333. 26 A. J. Blake, P. T. Brain, H. McNab, J. Miller, C. A. Morrison, S. Parsons, D. W. H. Rankin, H. E. Robertson and B. A. Smart, J. Phys. Chem., 1996, 100, 12280; P. T. Brain, C. A. Morrison, S. Parsons and D. W. H. Rankin, J. Chem. Soc., Dalton Trans., 1996, 4589. 27 W. C. Hamilton, Acta Crystallogr., 1965, 18, 502. Paper 9/02342I

ISSN:1477-9226    

DOI:10.1039/a902342i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2311–2316 2311 Syntheses, crystal structures and magnetic properties of two novel layered compounds: [Fe3(C2O4)3(4,49-bpy)4] and [Co(C2O4)- (4,49-bpy)] (4,49-bpy 5 4,49-bipyridine) † Li-Min Zheng,*a Xia Fang,a Kwang-Hwa Lii,b Hui-Hua Song,a Xin-Quan Xin,a Hoong-Kun Fun,c Kandasamy Chinnakali c and Ibrahim Abdul Razak c a State Key Laboratory of Coordination Chemistry, Coordination Chemistry Institute, Nanjing University, Nanjing 210093, China.E-mail: lmzheng@netra.nju.edu.cn; Fax: 186-25-3314502 b Institute of Chemistry, Academia Sinica, Taipei, China c X-Ray Crystallography Unit, School of Physics, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia Received 15th December 1998, Accepted 1st June 1999 The combination of 4,49-bipyridine (4,49-bpy) and oxalate ligands has resulted in the formation of two novel layered compounds [Fe3(C2O4)3(4,49-bpy)4] 1 and [Co(C2O4)(4,49-bpy)] 2. Both compounds exhibit two-dimensional layer structures.That of 1 consists of iron(II) oxalate chains, with the terminal and the bridging 4,49-bpy stacking alternately between the chains. The structure of 2 contains linear chains of cobalt(II) oxalate which are cross-linked by 4,49-bpy molecules in a perpendicular manner, leading to square grid sheets with rectangular windows. Magnetic measurements reveal that weak antiferromagnetic interactions are mediated in both compounds. In recent years extended networks based on transition metal oxalato complexes have drawn increased attention due to their potential applications as molecular-based magnetic materials.Two- and three-dimensional frameworks bridged purely by oxalate ligand have been designed and synthesized which show tunable ferro- or antiferro-magnetic exchanges by varying the nature of the magnetic centers and charge-compensating cations.1–9 The development of synthetic routes to novel polymeric co-ordination compounds with mixed bridging ligands, however, remains much to be explored.Success has been achieved, by combination of oxalate and 2,29-bipyrimidine (bpym) as chelating bridging ligands, which resulted in layered compounds [Cu2(bpym)(C2O4)2]?5H2O,10 [Mn2(bpym)- (C2O4)2]?6H2O11 and [Cu2(bpym)(C2O4)Cl2].12 The first two contain alternately bridging oxalate and bipyrimidine ligands, thus forming honeycombed layered structures similar to the purely oxalate bridged two-dimensional network.2,4 The latter corrugated network consists of alternately bpym and C2O4 bridged copper(II) chains which are further connected through m-chloro ligands. Such structural features could be related to the striking similarity of 2,29-bipyrimidine and oxalate ligands in their co-ordination modes.The replacement of 2,29- bipyrimidine by pyrazine (pyz), however, leads to the formation of [Cu2(C2O4)2(pyz)3] with a new structure type which is best described as a pleated ribbon of copper(II) oxalate linked by pyz molecules at every two copper atoms.13 Analogous to pyrazine, the rigid 4,49-bipyridine (4,49-bpy) ligand has been a useful building block for the construction of metal–organic co-ordination frameworks.A number of polymeric compounds bridged by 4,49-bpy have been synthesized which range from the 1-D chain compound [Co(SO4)(H2O)3- (4,49-bpy)]?2H2O,14 2-D square grid cationic sheets of [Cd- (4,49-bpy)(NO3)2] to 3-D interpenetrating structures of [Ag(4,49-bpy)(NO3)].15–22 Neutral sheets have also been found in † Supplementary data available: powder XRD pattern, IR spectra, unitcell contents.Available from BLDSC (No. SUP 57574, 6 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http:// www.rsc.org/dalton). M(4,49-bpy)X2 (M = Cu or Ni; X = Cl or Br),23 where linear {M(4,49-bpy)}n and {MX2}n chains are mutually perpendicularly arranged. The remarkable co-ordination abilities of oxalate and 4,49- bpy ligands have prompted us to design and synthesize new polymeric, microporous co-ordination compounds containing both ligands.The eVorts resulted in the preparation of two new compounds [Fe3(C2O4)3(4,49-bpy)4] 1 and [Co(C2O4)(4,49-bpy)] 2, with novel layered structures. Both structures diVer signifi- cantly from those with the combination of bis-chelating bridging ligands oxalate and 2,29-bipyrimidine,10–12 although they may be compared with that of [Cu2(C2O4)2(pyz)3].13 In this paper the crystal structures, characterizations and magnetic properties of both compounds are reported. Experimental Materials and methods All starting materials were reagent grade used as purchased.The elemental analyses were performed on a PE 240C elemental analyzer. The infrared spectra were recorded on a Fourier Nicolet FT-170SX spectrometer with pressed KBr pellets, Mössbauer spectra on an S-600 Mössbauer spectrometer using a 57Co/Pd source. The latter spectrometer was calibrated by a standard sample of Na2[Fe(CN)5(NO)]?2H2O (SNP) at room temperature.The isomer shifts are reported relative to SNP. The variable temperature magnetic susceptibility data were obtained on polycrystalline samples (20.7 mg for compound 1, 15.0 mg for 2) from 2 to 300 K in a magnetic field of 5 kG after zero-field cooling using a SQUID magnetometer. Diamagnetic corrections were estimated from Pascal’s constants.24 Synthesis of [Fe3(C2O4)3(4,49-bpy)4] 1 The compound was prepared as a major phase (in ca. 40% yield based on iron) through hydrothermal reaction of K3[Fe(C2O4)3]? 3H2O (0.5 mmol, 0.2454 g), 4,49-bipyridine (0.5 mmol, 0.0961 g) and water (10 mL) at 180 8C for 20 h. Red needles of 1 were manually picked and used for single crystal structural2312 J. Chem. Soc., Dalton Trans., 1999, 2311–2316 Fig. 1 The asymmetric unit of compound 1 (ellipsoids at 50% probability). The H atoms are omitted. determination and property measurements (Found: C, 52.04; H, 3.13; N, 10.48.Calc. for C46H32Fe3N8O12: C, 52.30; H, 3.03; N, 10.61%). IR (KBr): 3444w (br), 1668m, 1608s, 1535w, 1488w, 1412m, 1357w, 1312m, 1219w, 1078w, 1045w, 1005w, 861w, 810m, 799m, 734w, 629m, 618m, 574w and 506m cm21. Synthesis of [Co(C2O4)(4,49-bpy)] 2 The hydrothermal treatment of a mixture of Co(NO3)2?6H2O (0.2 mmol, 0.0588 g), H2C2O4?2H2O (0.4 mmol, 0.0501 g), 4,49- bpy (0.2 mmol, 0.0389 g) and water (10 mL) at 180 8C for 24 h resulted in orange crystals of compound 2 (in ca. 50% yield based on cobalt). The product is monophasic, as judged by comparison of the powder X-ray diVraction of the bulk product with the pattern simulated from the single crystal data (Found: C, 47.36; H, 2.48; N, 9.23. Calc. for C12H8CoN2O4: C, 47.54; H, 2.60; N, 9.24%). IR (KBr): 3466w (br), 3061w, 1610s, 1537w, 1493w, 1417m, 1357w, 1315m, 1223w, 1079w, 1046w, 1008w, 818m, 808m, 734w, 634m and 488m cm21. The direct reactions between the chemicals used in the hydrothermal syntheses of 1 and 2 were unsuccessful.Crystallography Crystals of dimensions 0.38 × 0.1 × 0.1 (compound 1) and 0.22 × 0.20 × 0.16 mm (2) were used for indexing and intensity data collection on a Siemens Smart-CCD diVractometer equipped with a normal focus, 3 kW sealed tube X-ray source and graphite-monochromated Mo-Ka radiation (l = 0.71073 Å) at 293 K. Empirical absorption corrections were applied for both compounds by using the SADABS program for the Siemens area detector.Lorentz-polarization and secondary extinction corrections were applied for 1. The structures were solved by direct methods and refined by using SHELXTL.25 All non-hydrogen atoms in both structures were refined with anisotropic displacement parameters. The hydrogen atoms in 1 were introduced in idealized positions and refined isotropically with fixed thermal parameters, those in 2 were localized from a Fourier-diVerence map and refined isotropically. Some relevant crystallographic data and structure determination parameters are listed in Table 1, selected bond lengths and angles in Tables 2 and 3 for 1 and 2, respectively.CCDC reference number 186/1483. See http://www.rsc.org/suppdata/dt/1999/2311/ for crystallographic files in .cif format. Results and discussion Infrared spectra The infrared spectra of compounds 1 and 2 (see SUP 57574) exhibit characteristic bands for both oxalate and 4,49- bipyridine ligands. For 1 the peaks at 1668 (nCO), 1357, 1312 (nCC) and 810 cm21 (dOCO) are attributed to the co-ordinated oxalate group.The peak at 506 cm21 is assigned to nFe–O, which is higher than that for [Fe(C2O4)(H2O)2] (490 cm21).26 In addition, the aromatic C–C and C–N stretching vibration absorptions appear at 1608, 1535, 1488 and 1412 cm21. The bands in the region 618–1219 cm21 can be assigned to the CH in-plane or out-of-plane bend, ring breathing and ring deformation absorptions of 4,49-bipyridine.27 As the band at 618 cm21 of the free ligand is very sensitive and shifts to a higher frequency after coordination to a metal ion,27 the appearance of two peaks at 629 and 618 cm21 suggests the presence of two kinds of 4,49-bpy which is consistent with the single crystal structural analysis.The infrared spectrum of 2 is similar to that of 1 except that only one peak (634 cm21) appears near 618 cm21, which indi-J. Chem. Soc., Dalton Trans., 1999, 2311–2316 2313 cates the existence of only one type of 4,49-bpy.The band at 488 cm21 is assigned to nCo-O. Structure of compound 1 The structure of compound 1 is made up of asymmetric units of Fe3(m-C2O4)3(m-4,49-bpy)2(4,49-bpy)2 with 69 non-hydrogen independent atoms (Fig. 1). Three iron centers are crystallographically distinguishable; each co-ordinates to four oxygen atoms from the oxalate anions and two nitrogen atoms from the 4,49-bpy molecules to form a distorted octahedral environment.The oxalate anion behaves as a bridging ligand and links the three iron centers repeatedly to form infinite chains running parallel to the b axis, with zigzag undulation in the (101) plane (Fig. 2). The Fe–O distances [2.102(2)–2.126(2) Å] are comparable to those in [Fe(bipy)3][Fe2(C2O4)3] [2.122(2), 2.128(2) Å] 3 Fig. 2 One layer of compound 1 viewed along the [101] direction. The hydrogen atoms are omitted for clarity. Table 1 Crystal data and structure refinement for compounds 1 and 2 1 2 Empirical formula Formula weight Crystal system Space group a/Å b/Å c/Å b/8 U/Å3 Z Dc/g cm23 m(Mo-Ka)/cm21 F(000) Reflections collected Independent reflections Data, restraints, parameters Goodness of fit of F2 R1, wR2 [I > 2s(I)] (all data) Extinction coeYcient Absolute structure parameter (Dr)max, (Dr)min/e Å23 C46H32Fe3N8O12 1056.35 Monoclinic P21/c 16.2475(2) 15.9840(2) 16.3856(1) 94.2776(3) 4243.49(8) 4 1.653 10.90 2152 24362 9195 [R(int) = 0.0621] 9195, 0, 623 1.023 0.0493, 0.0887 0.1113, 0.1127 0.00022(5) 0.357, 20.541 C12H8CoN2O4 303.13 Orthorhombic Imm2 10.9768(3) 11.4017(3) 5.4032(1) 676.23(3) 2 1.489 12.78 306 2230 628 [R(int) = 0.0241] 628, 1, 55 1.119 0.0257, 0.0666 0.0267, 0.0671 0.08(6) 0.302, 20.238 and [Fe(C2O4)(2,29-bpy)] (2,29-bpy = 2,29-bipyridine) [2.092(3)– 2.162(4) Å].28 The 4,49-bpy molecule serves as either a terminal or bridging ligand.The apical positions of the Fe1 (or Fe3) octahedron are occupied by N atoms from both terminal and bridging 4,49-bpy.The apical positions of the Fe2 octahedron, however, are filled by the N atoms from two bridging 4,49-bpy. The Fe–N bond lengths [2.193(3)–2.290(3) Å] agree well with that in [Fe(SCN)2(H2O)2(4,49-bpy)2] [2.214(2) Å],29 although the Fe–N lengths with terminal 4,49-bpy [2.290(3) and 2.249(3) Å for Fe1–N1 and Fe3–N7, respectively] are slightly Table 2 Selected bond lengths (Å) and angles (8) for compound 1 Fe(1)–O(1) Fe(1)–O(2) Fe(1)–N(1) Fe(2)–O(5) Fe(2)–O(7) Fe(2)–N(4A) Fe(3)–O(9) Fe(3)–O(11) Fe(3)–N(6B) O(1)–C(45C) O(3)–C(41) O(5)–C(41) O(7)–C(43) O(9)–C(43) O(11)–C(45) C(41)–C(42) C(45)–C(46D) O(4)–Fe(1)–O(3) O(3)–Fe(1)–O(2) O(4)–Fe(1)–O(1) O(4)–Fe(1)–N(3) O(2)–Fe(1)–N(3) O(4)–Fe(1)–N(1) O(2)–Fe(1)–N(1) N(3)–Fe(1)–N(1) O(5)–Fe(2)–O(6) O(5)–Fe(2)–O(7) O(6)–Fe(2)–O(7) O(8)–Fe(2)–N(5) O(7)–Fe(2)–N(5) O(8)–Fe(2)–N(4A) O(7)–Fe(2)–N(4A) O(9)–Fe(3)–O(10) O(10)–Fe(3)–O(11) O(10)–Fe(3)–O(12) O(9)–Fe(3)–N(6B) O(11)–Fe(3)–N(6B) O(9)–Fe(3)–N(7) O(11)–Fe(3)–N(7) N(6B)–Fe(3)–N(7) C(46)–O(2)–Fe(1) C(42)–O(4)–Fe(1) C(1)–N(1)–Fe(1) C(15)–N(3)–Fe(1) C(42)–O(6)–Fe(2) C(44)–O(8)–Fe(2) C(25)–N(5)–Fe(2) C(18)–N(4)–Fe(2E) C(44)–O(10)–Fe(3) C(46D)–O(12)–Fe(3) C(28)–N(6)–Fe(3F) C(35)–N(7)–Fe(3) O(5)–C(41)–C(42) O(6)–C(42)–O(4) O(4)–C(42)–C(41) O(7)–C(43)–C(44) O(8)–C(44)–O(10) O(10)–C(44)–C(43) O(11)–C(45)–C(46D) O(12C)–C(46)–O(2) O(2)–C(46)–C(45C) 2.124(3) 2.112(3) 2.290(3) 2.114(2) 2.123(2) 2.256(3) 2.104(2) 2.123(2) 2.193(3) 1.264(4) 1.259(4) 1.254(4) 1.250(4) 1.260(4) 1.246(4) 1.556(5) 1.559(5) 79.58(9) 171.99(9) 172.84(9) 97.61(10) 89.71(10) 87.29(10) 84.75(10) 173.13(12) 79.69(9) 99.66(9) 179.05(9) 92.15(10) 92.13(10) 86.62(10) 86.69(10) 79.47(9) 173.10(9) 101.32(10) 94.71(10) 90.12(10) 90.41(10) 84.97(10) 173.44(12) 113.0(2) 112.4(2) 120.7(2) 120.6(2) 112.3(2) 111.3(2) 119.8(2) 121.8(2) 111.8(2) 112.8(2) 121.0(2) 119.8(2) 117.4(3) 125.5(3) 117.0(3) 117.6(3) 124.9(3) 117.1(3) 117.5(3) 125.9(3) 116.9(3) Fe(1)–O(4) Fe(1)–O(3) Fe(1)–N(3) Fe(2)–O(6) Fe(2)–O(8) Fe(2)–N(5) Fe(3)–O(10) Fe(3)–O(12) Fe(3)–N(7) O(2)–C(46) O(4)–C(42) O(6)–C(42) O(8)–C(44) O(10)–C(44) O(12)–C(46D) C(43)–C(44) O(4)–Fe(1)–O(2) O(3)–Fe(1)–O(1) O(2)–Fe(1)–O(1) O(3)–Fe(1)–N(3) O(1)–Fe(1)–N(3) O(3)–Fe(1)–N(1) O(1)–Fe(1)–N(1) O(5)–Fe(2)–O(8) O(8)–Fe(2)–O(6) O(8)–Fe(2)–O(7) O(5)–Fe(2)–N(5) O(6)–Fe(2)–N(5) O(5)–Fe(2)–N(4A) O(6)–Fe(2)–N(4A) N(5)–Fe(2)–N(4A) O(9)–Fe(3)–O(11) O(9)–Fe(3)–O(12) O(11)–Fe(3)–O(12) O(10)–Fe(3)–N(6B) O(12)–Fe(3)–N(6B) O(10)–Fe(3)–N(7) O(12)–Fe(3)–N(7) C(45C)–O(1)–Fe(1) C(41)–O(3)–Fe(1) C(5)–N(1)–Fe(1) C(11)–N(3)–Fe(1) C(41)–O(5)–Fe(2) C(43)–O(7)–Fe(2) C(21)–N(5)–Fe(2) C(19)–N(4)–Fe(2E) C(43)–O(9)–Fe(3) C(45)–O(11)–Fe(3) C(29)–N(6)–Fe(3F) C(31)–N(7)–Fe(3) O(5)–C(41)–O(3) O(3)–C(41)–C(42) O(6)–C(42)–C(41) O(7)–C(43)–O(9) O(9)–C(43)–C(44) O(8)–C(44)–C(43) O(11)–C(45)–O(1D) O(1D)–C(45)–C(46D) O(12C)–C(46)–C(45C) 2.102(2) 2.106(2) 2.217(3) 2.119(2) 2.117(2) 2.240(3) 2.115(2) 2.126(2) 2.249(3) 1.256(4) 1.261(4) 1.258(4) 1.257(4) 1.267(4) 1.249(4) 1.547(5) 98.91(9) 101.40(9) 79.12(9) 98.29(10) 89.28(10) 87.31(10) 85.68(10) 178.78(9) 100.98(9) 79.65(9) 88.87(10) 88.56(10) 92.34(10) 92.63(10) 178.43(11) 99.67(9) 176.17(9) 79.09(9) 96.77(10) 88.93(10) 88.19(10) 85.87(11) 112.6(2) 112.3(2) 122.9(3) 122.9(2) 112.7(2) 111.2(2) 123.2(2) 120.7(2) 112.8(2) 113.0(2) 122.5(2) 121.7(3) 125.9(3) 116.7(3) 117.5(3) 125.4(3) 117.0(3) 117.9(3) 125.7(3) 116.8(3) 117.2(3) Symmetry transformations used to generate equivalent atoms: A 2x 1 1, y 1 1– 2, 2z 1 1– 2; B 2x, y 1 1– 2, 2z 1 3– 2; C x, y 2 1, z; D x, y 1 1, z; E 2x 1 1, y 2 1– 2, 2z 1 1– 2; F 2x, y 2 1– 2, 2z 1 3– 2.2314 J. Chem.Soc., Dalton Trans., 1999, 2311–2316 Fig. 3 The co-ordination around the Co in compound 2 (ellipsoids at 50% probability). The H atoms are omitted. longer. Consequently, condensed two-dimensional sheets are constructed in the (101) planes (Fig. 2), with the terminal and the bridging 4,49-bpy stacking alternately between the iron(II) oxalate chains. The bridging 4,49-bpy molecules connect the Fe2 atom with both Fe1 and Fe3 atoms to form iron “trimers” in the layer. The two pyridine rings of each 4,49-bpy are twisted with dihedral angles of 28.4–38.98. The intra-layer Fe ? ? ? Fe separations are 5.397–5.431 Å through the oxalate bridges and 11.494–11.543 Å through the 4,49-bpy bridges, respectively. The interlayer distance is 5.55 Å.Weak hydrogen bondings are observed within the layer and between the layers. The four shortest C ? ? ? O distances are 3.030(5), 3.164(5), 3.166(5) and 3.193(5) Å for C35 ? ? ? O12, C29 ? ? ? O12i, C11 ? ? ? O1 and C1 ? ? ? O1 (symmetry code: i 2x, y 2 ��� , 2z 1 ��� ), respectively. The structure of compound 1 is unique compared with that of [Cu2(C2O4)2(pyz)3].13 The latter contains two types of copper atoms with O4N2 co-ordination environments.One has only terminally co-ordinated pyz molecules while the other is linked by bridging pyz molecules. Structure of compound 2 The structure of compound 2 is highly symmetric. Fig. 3 shows the co-ordination around the Co which exhibits an elongated Fig. 4 One layer of compound 2 viewed along the [100] direction. octahedral environment. The basal plane consists of four oxygen atoms from two equivalent oxalate anions, whereas the apical positions are occupied by two nitrogen atoms from the symmetrically equivalent 4,49-bpy molecules.The Co1–O bond length [mean 2.084(6) Å] can be compared with those in [Co- (bpy)3][Co2(C2O4)3]ClO4 [2.070(5), 2.081(5) Å] 9 and [CoCr2- (bipy)2(m-C2O4)2(C2O4)2(H2O)2]?H2O [2.089(6), 2.155(6) Å].30 The Co–N distance [Co1–N 2.153(2) Å] is consistent with those in [Co2(4,49-bpy)3(NO3)4]?4H2O [2.13(1) Å].16 Both the oxalate anion and 4,49-bpy act as bridging ligands. The cobalt–oxalate chains along the [001] direction are linear, which is diVerent from those in compound 1.The adjacent chains are cross-linked by 4,49-bpy molecules in a perpendicular manner, leading to square grid sheets in the bc planes with rectangular windows (Fig. 4). The two pyridine rings of 4,49-bpy are coplanar and symmetrically related to each other. The intra-layer Co ? ? ? Co separations are 5.403 Å through the oxalate bridge and 11.402 Å through the 4,49-bpy bridge, respectively.In the crystal, the sheets are stacked along the [100] direction to give extended one-dimensional channel networks. The interlayer distance is 5.50 Å for 2. Weak hydrogen bonding has been found within the layer and between the layers. The three shortest C ? ? ?O distances are 3.260, 3.283 and 3.283 Å for C2 ? ? ? O2, C2 ? ? ? O1i and C2 ? ? ? O1ii (symmetry codes: i 2x 1 1, 2y, z; ii 2x 1 1, y, z), respectively. Clearly, the structure of compound 2 is quite diVerent from that of 1.This reflects, on the one hand, the coordination flexibilities of 4,49-bpy either as a terminal or bridging ligand with the two pyridine rings twisted or coplanar. On the other hand, the formation of zigzag iron(II) oxalate chains in 1 may originate from the starting materials, wherein a tris-chelated iron compound was employed. It is worth noting that the oxidation state of the iron in 1 is 12, though the starting material is an iron(III) compound. A similar phenomenon has been observed in the hydrothermal synthesis of [Fe2(C2O4)(OH)2].31 Mössbauer spectrum of compound 1 The room-temperature Mössbauer spectrum of compound 1 is shown in Fig. 5. It can be least-squares fitted with one doublet instead of three doublets corresponding to the three types of iron components observed in the crystal structure. The parameters obtained are d (isomer shift) = 0.95 mm s21, DEQ (quadrupole splitting) = 2.22 mm s21. The spectrum is typical of high-spin FeII 5,26 which is consistent with the magnetic susceptibility data of 1.Magnetic properties of compound 1 Fig. 6(a) shows the magnetic behavior of a powdered sample of compound 1 in the form of a cm vs. T plot where cm is the molar magnetic susceptibility. At 300 K the eVective magnetic moment (meff) per Fe, calculated from meff = 2.828(cmT)1/2, is 5.68 mB which is greater than the spin-only value of 4.90 mB for S = 2.J. Chem. Soc., Dalton Trans., 1999, 2311–2316 2315 Table 3 Selected bond lengths (Å) and angles (8) for compound 2 Co(1)–O(2) Co(1)–N(1) O(1)–C(1) N(1)–C(2) C(3)–C(4) O(2)–Co(1)–O(2A) O(2A)–Co(1)–O(1) O(1)–Co(1)–O(1A) O(1)–Co(1)–N(1A) O(1)–Co(1)–N(1) C(1)–O(1)–Co(1) C(2)–N(1)–Co(1) O(2D)–C(1)–C(1A) C(2C)–N(1)–C(2) C(4)–C(3)–C(2) C(3)–C(4)–C(4E) 2.074(7) 2.153(2) 1.293(12) 1.317(3) 1.369(3) 82.4(3) 99.09(8) 79.4(3) 90.0 90.0 115.3(5) 122.04(12) 118.5(5) 115.9(2) 120.1(2) 121.97(13) 2x 2x 2x 2x 2x 2x 2x Co(1)–O(1) C(1)–C(1A) O(2)–C(1B) C(2)–C(3) C(4)–C(4E) O(2)–Co(1)–O(1) O(2A)–Co(1)–O(1A) O(2)–Co(1)–N(1A) O(2)–Co(1)–N(1) N(1A)–Co(1)–N(1) C(1B)–O(2)–Co(1) O(2D)–C(1)–O(1) O(1)–C(1)–C(1A) N(1)–C(2)–C(3) C(3)–C(4)–C(3C) 2.094(5) 1.582(5) 1.206(13) 1.381(3) 1.489(5) 178.5(3) 178.5(3) 90.0 90.0 180.0(3) 110.3(5) 126.5(2) 115.0(4) 123.9(2) 116.1(3) 2x 2x 2x Symmetry transformations used to generate equivalent atoms: A 2x 1 1, 2y, z; B 2x 1 1, 2y, z 2 1; C 2x 1 1, y, z; D 2x 1 1, 2y, z 1 1; E 2x 1 1, 2y 1 1, z.The higher value of meff is attributed to the orbital contribution of the high spin iron(II) center.The maximum around 44 K in Fig. 6(a) suggests the presence of an antiferromagnetic coupling, which is confirmed by a negative Weiss constant (216.0 K) determined in the temperature range 300–100 K using the equation cm = C/(T 2 q). The magnetic behavior of compound 2 is quite similar to that of 1 [Fig. 6(b)]. At 300 K the eVective magnetic moment per Co (5.02 mB) is much greater than the spin-only value (3.87 mB) expected for a high spin (S = 3/2) center, which originates from an orbital contribution of CoII.The maximum around 40 K in Fig. 6(b) indicates that an antiferromagnetic interaction is propagated. The layered structures of compounds 1 and 2 contain metal(II) oxalate chains linked by 4,49-bipyridine molecules. It is well known that an eYcient antiferromagnetic exchange can be transmitted through the oxalate bridge.24 The magnetic interaction through the 4,49-bipyridine bridge, however, is usually very weak considering the long M ? ? ? M distances of ca. 11.5 Å. This interaction could be further reduced because of the twist of the two pyridine rings as in the case of 1. Therefore, it is reasonable to describe the magnetic behaviors of both compounds based on a chain model. The M ? ? ? M separations through the oxalate ligand are 5.397–5.431 Å in 1 and 5.403 Å in 2. The three interaction parameters in 1 may be assumed to be very close to each other.The susceptibility data were then analyzed by using an expression for the magnetic susceptibility of a uniform chain of classical spins derived by Fisher, with the classical spins scaled to a real quantum spin S = 2 for 1 and 3/2 for 2.24,32 The equation, however, does not fit the experimental data well, especially at low temperatures (<44 K). The inclusion of an interchain exchange based on the molecular field approximation does not improve the theoretical fitting distinctly.The bad fits for both compounds may be explained by Fig. 5 Mössbauer spectrum of compound 1 at 298 K. the fact that Fisher’s equation does not take into account the eVects of the zero-field splitting and/or spin–orbital coupling which could be significant for iron(II) and cobalt(II) ions. In conclusion, this paper describes two new layered compounds [Fe3(C2O4)3(4,49[Co(C2O4)(4,49-bpy)], containing both 4,49-bpy and oxalate bridging ligands.The magnetic properties indicate antiferromagnetic exchanges in both compounds which should be mediated mainly through the oxalate bridges. Further work is in progress, using mixed bridging ligands, to build up novel polymeric, microporous co-ordination networks with tunable pore size and interesting properties. Acknowledgements This work is supported by the National Natural Science Foundation, Natural Science Foundation of Jiangsu Province of China, the Education Commission of China and Malaysian Government research grant R&D (No. 190-9609-2801). We are grateful to Professor S.-L. Wang and Ms F.-L. Liao at Tsing Hua University for X-ray intensity data collection for compound 1, to Dr Z. Yu at Nanjing University for assistance in collecting and fitting the Mössbauer data, and to Professor G. X. Wang at Nanjing University for stimulating discussions. References 1 H. Tamaki, Z. J. Zhong, N. Matsumoto, S. Kida, N. Koikawa, Y. Achiwa, Y.Hashimoto and H. Okawa, J. Am. Chem. Soc., 1992, 114, 6974. Fig. 6 Temperature dependent molar magnetic susceptibilities for compounds 1 (a) and 2 (b).2316 J. Chem. Soc., Dalton Trans., 1999, 2311–2316 2 L. O. Atovmyran, G. V. Shilov, R. N. Lyubovskaya, E. I. Zhilyacva, N. S. Ovanesyan, S. I. Pirumova and I. G. Gusakovskaya, JETP Lett., 1993, 58, 766. 3 S. Decurtins, H. W. Schmalle, P. Schneuwly and H. R. Oswald, Inorg. Chem., 1993, 32, 1888. 4 S. Decurtins, H. W.Schmalle, H. R. Oswald, A. Linden, J. Ensling, P. Gutlich and A. Hauser, Inorg. Chim. Acta, 1994, 216, 65. 5 S. Decurtins, H. W. Schmalle, P. Schneuwly, J. Ensling and P. Gutlich, J. Am. Chem. Soc., 1994, 116, 9521. 6 C. Mathonière, C. J. Nuttall, S. G. Carling and P. Day, Inorg. Chem., 1996, 35, 1201. 7 M. Clemente-León, E. Coronado, J.-R. Galán-Mascarós and C. J. Gómez-García, Chem. Commun., 1997, 1727. 8 J. Larionova, B. Mombelli, J. Sanchiz and O. Kahn, Inorg. Chem., 1998, 37, 679. 9 M.Hernández-Molina, F. Lloret, C. Ruiz-Pérez and M. Julve, Inorg. Chem., 1998, 37, 4131. 10 G. De Munno, M. Julve, F. Nicolo, F. Lloret, J. Faus, R. Ruiz and E. Sinn, Angew. Chem., Int. Ed. Engl., 1993, 32, 613. 11 G. De Munno, R. Ruiz, F. Lloret, J. Faus, R. Sessoli and M. Julve, Inorg. Chem., 1995, 34, 408. 12 S. Decurtins, H. W. Schmalle, P. Schneuwly, L.-M. Zheng, J. Ensling and A. Hauser, Inorg. Chem., 1995, 34, 5501. 13 S. Kitagawa, T. Okubo, S. Kawata, M. Kondo, M. Katada and H. Kobayashi, Inorg. Chem., 1995, 34, 4790. 14 J. Lu, C. Yu, T. Niu, T. Paliwala, G. Crisci, F. Somosa and A. J. Jacobson, Inorg. Chem., 1998, 37, 4637. 15 M. Fujita, Y. J. Kwon, S. Washizu and K. Ogura, J. Am. Chem. Soc., 1994, 116, 1151. 16 M. Kondo, T. Yoshitomi, K. Seki, H. Matsuzaka and S. Kitagawa, Angew. Chem., Int. Ed. Engl., 1997, 36, 1725. 17 O. M. Yaghi, H. Li and T. L. Groy, Inorg. Chem., 1997, 36, 4292. 18 O. M. Yaghi and H. Li, J. Am. Chem. Soc., 1995, 117, 10401. 19 O. M. Yaghi and H. Li, J. Am. Chem. Soc., 1996, 118, 295. 20 A. J. Blake, S. J. Hill, P. Hubberstey and W.-S. Li, J. Chem. Soc., Dalton Trans., 1997, 913. 21 S. Subramanian and M. J. Zaworotko, Angew. Chem., Int. Ed. Engl., 1995, 34, 2127. 22 M. L. Tong, B. H. Ye, J. W. Cai, X.-M. Chen and S. W. Ng, Inorg. Chem., 1998, 37, 2645. 23 N. Masciocchi, P. Cairati, L. Carlucci, G. Mezza, G. Ciani and A. Sironi, J. Chem. Soc., Dalton Trans., 1996, 2739. 24 O. Kahn, Molecular Magnetism, VCH, New York, 1993. 25 SHELXTL Version 5.0 Reference Manual, Siemens Analytical X-Ray Systems, Inc., Madison, WI, 1996. 26 J. T. Wrobleski and D. B. Brown, Inorg. Chem., 1979, 18, 2738. 27 J. G. Contreras and C. J. Diz, J. Coord. Chem., 1987, 16, 245. 28 H.-K. Fun, S. S. S. Raj, X. Fang, L.-M. Zheng and X.-Q. Xin, Acta Crystallogr., 1999, in the press. 29 L.-M. Zheng, X. Chen, S. Gao, K. Chinnakali and H.-K. Fun, submitted for publication. 30 F. D. Rochon, R. Melanson and M. Andruh, Inorg. Chem., 1996, 35, 6086. 31 M. Molinier, D. J. Price, P. T. Wood and A. K. Powell, J. Chem. Soc., Dalton Trans., 1997, 4061. 32 M. E. Fisher, Am. J. Phys., 1964, 32, 343. Paper 8/09738K

ISSN:1477-9226    

DOI:10.1039/a809738k

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 2315 Topotactic synthesis of ·-zirconium phenylphosphonate from ·-zirconium phosphate Gary B. Hix, Simon J. Kitchin and Kenneth D. M. Harris * School of Chemistry, University of Birmingham, Edgbaston, Birmingham, UK B15 2TT a-Zirconium phenylphosphonate has been synthesised by a reaction between a-zirconium phosphate and molten phenylphosphonic acid. The reaction produces a physical mixture of the ‘pure’ phases of a-zirconium phenylphosphonate and a-zirconium phosphate, with the ratio dependent on the ratio of phenylphosphonic acid to a-zirconium phosphate in the synthesis mixture.Detailed characterization of the reaction products has been carried out by powder X-ray diVraction, high-resolution solid-state 31P NMR spectroscopy, thermogravimetric analysis and FT-IR spectroscopy, and by considering experiments involving the selective intercalation of n-butylamine. Zirconium phosphate exists in two well characterised solid forms: a-zirconium phosphate 1 [Zr(HPO4)2?H2O] and g-zirconium phosphate 2 [Zr(PO4)(H2PO4)?2H2O]. Organic derivatives of a-zirconium phosphate, known as a-zirconium phosphonates [Zr(RPO3)2?nH2O; usually n = 1], were first prepared 3 by direct synthesis, involving the use of the appropriate phosphonic acid, rather than phosphoric acid, in the synthesis procedure.These methods and some of the structural and chemical aspects of zirconium phosphonates have been discussed in recent review articles.4,5 Several aspects of the chemistry of these a-zirconium phosphonates are well developed; for example, applications have been found as intercalation hosts for systems involving Brønsted and Lewis acidity,6 and investigations have been carried out with regard to their ionic conductivity and ion exchange properties,7–11 photochemical reactions 12,13 and catalysis.14,15 In contrast, there have been no reported preparations of organic derivatives of g-zirconium phosphate by direct synthesis, but preparations employing topotactic reactions between pre-formed g-zirconium phosphate and solutions of phosphonic acids 16,17 or solutions of propylene oxide 18,19 have been reported. Conversely, there have been no reports of the synthesis of a-zirconium phosphonates topotactically from preformed a-zirconium phosphate (we note, however, that a topotactic reaction of this type using a-titanium phosphate has been reported).20 a-Zirconium phosphonates are typically made using hydrothermal techniques or by refluxing solutions of zirconium salts in the presence of HF and a phosphonic acid.As shown recently,21 layered metal phosphonates can be formed by reaction between layered metal hydroxides and molten phosphonic acids. The application of this synthetic method would remove the need to employ either extreme conditions or highly hazardous chemicals in the synthesis of a-zirconium phosphonates.The aim of the present paper has been to explore the synthetic strategy of using molten phosphonic acids in such preparations, particularly with a view to opening up new routes to metal phosphonate materials. In the present paper, we demonstrate the success of this approach for topotactic synthesis of azirconium phenylphosphonate by reaction between pre-formed a-zirconium phosphate and molten phenylphosphonic acid. In principle, a synthetic procedure of this type may lead to a variety of possible products depending on the extent to which the phosphate groups in the a-zirconium phosphate precursor are replaced by phenylphosphonate groups.Some of the pos- * E-Mail: K.D.M.Harris@bham.ac.uk sible types of product are shown schematically in Fig. 1 (see also ref. 22). In addition to the case of a single product phase of a-zirconium phenylphosphonate [Fig. 1(c)], other possibilities include: (i) single phase mixed derivatives, containing both phenylphosphonate and phosphate groups ‘interspersed’ between the layers [an example of this type of product is shown in Fig. 1(b)]; (ii) single phase ‘staged’ products, comprising layers with only phenylphosphonate groups and layers with only phosphate groups [an example is shown in Fig. 1(d)]; or (iii) a physical mixture of the pure phase of a-zirconium phenylphosphonate [Fig. 1(c)] and the pure phase of a-zirconium phosphate [Fig. 1(a)]. Clearly the actual type of product obtained should be readily identified using diVraction techniques.Experimental Preparation of ·-zirconium phosphate a-Zirconium phosphate was prepared by the sol–gel method of Benhamza et al.23 An aqueous solution of phosphoric acid (85%) and a solution of zirconium propoxide in propan- 1-ol (1 mol dm23) were mixed, leading to the formation of a gelatinous white precipitate which was kept in the mother-liquor at 60 8C for 10 d. The semi-crystalline a-zirconium phosphate was recovered by centrifugation, washed with propan-1-ol and water, and finally dried in air at 60 8C. Reaction of ·-zirconium phosphate and molten phenylphosphonic acid a-Zirconium phosphate (ca. 1 g) was mixed with phenylphosphonic acid; diVerent ratios of these reagents were used, as specified in Table 1. The mixture was ground thoroughly and placed in a thick pyrex tube (volume ca. 35 cm3) capable of withstanding pressures up to 200 psi (ª1379 kPa). The tube was sealed with a Teflon screw cap.The reaction mixture was then kept at 166 8C (the melting point of phenylphosphonic acid) for 3 d. Following this period, the products of the reaction were suspended in deionised water, recovered by filtration, and then washed thoroughly to remove any unreacted phenylphosphonic acid. The solid products were then dried in air at 60 8C for about 30 min. The products obtained from the reactions with diVerent ratios of a-zirconium phosphate and a-zirconium phenylphosphonate are denoted samples A–G (see Table 1).Procedure for intercalation of n-butylamine The procedure for attempted intercalation (see below) of n-butylamine within samples A–G was to place the sample2316 J. Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 together with an open vessel containing n-butylamine inside a closed desiccator. After 24 h the sample (which in some cases was visibly swollen) was removed, washed briefly with deionised water, and allowed to dry under ambient conditions.Characterization Powder X-ray diVraction patterns were recorded using Cu-Ka1 radiation on a Siemens D5000 diVractometer, operating in transmission mode with a primary beam germanium monochromator. Fourier-transform IR spectra were recorded using a Perkin-Elmer Paragon 1000 FT-IR spectrometer for materials dispersed in KBr pellets. Thermogravimetric analysis (TGA) was carried out using a Stanton Redcroft 870 instrument referenced against re-calcined alumina; the samples were heated Fig. 1 Possible arrangements of phosphate (filled circles) and phenylphosphonate (open ellipses) groups on the internal layer surfaces of a phosphate/phenylphosphonate material. See text for discussion. The pure phosphate phase is represented in (a) and the pure phenylphosphonate phase is represented in (c). Table 1 Ratio of reactants in the reactions carried out between molten phenylphosphonic acid (H2O3PC6H5) and a-zirconium phosphate [Zr(HPO4)2?H2O] Sample AB C DEF G Mass of phenylphosphonic acid/g 0.1 0.2 0.3 0.4 0.5 1.0 2.0 Molar ratio H2O3PC6H5 :Zr(HPO4)2?H2O 0.38 0.77 1.1 1.5 1.9 3.8 7.6 in air (at 10 8C per min) to a maximum temperature of 1000 8C.High resolution solid-state 31P NMR spectra were recorded at 121.16 MHz on a Chemagnetics CMX-Infinity 300 spectrometer, using a Chemagnetics double resonance probe with magic angle sample spinning (frequency ca. 8 kHz; rotor diameter 3.2 mm) and high-power 1H decoupling. All spectra were recorded at ambient temperature, with the recycle delay (120 s) estimated to be greater than 5 × T1 (to ensure quantitative spectral intensities).Chemical shifts are given relative to the 31P resonance in aqueous H3PO4 solution (42%) as an external standard. Results and Discussion Thermogravimetric analysis (TGA) was carried out for all samples A–G in air, and stepwise mass losses are observed in all cases. A representative TGA curve is shown in Fig. 2, and total mass losses for all samples studied are reported in Table 2.The first step occurs between ambient temperature and 120 8C and is ascribed to the loss of water. The second step in the region 400–550 8C involves oxidation of the organic portion and dehydration to yield zirconium pyrophosphate. These observations are in agreement with results reported previously for mixed phosphate/phosphonate derivatives of zirconium24 and titanium.20 The stoichiometries of samples A–G determined from the TGA data are given in Table 3.As the materials are mixtures of the pure phases of a-zirconium phosphate and azirconium phenylphosphonate, the composition is specified as [Zr(O3POH)2?H2O]x[Zr(O3PC6H5)2]y rather than Zr(O3POH)x- (O3PC6H5)2 2x?x/2H2O (which is how a single phase of a mixed derivative phase would be represented). The stoichiometries determined from the TGA data indicate that the synthesis method is successful in producing a-zirconium phenylphosphonate.In order to achieve a complete exchange, however, an excess of phenylphosphonic acid must be employed in the synthesis. Sample G was prepared using a four-fold excess of Fig. 2 Typical TGA curve for samples obtained in the synthesis involving a-zirconium phosphate and phenylphosphonic acid. The TGA curve shown is for sample C Table 2 Total mass losses, determined by TGA, for samples A–G obtained in the synthesis involving a-zirconium phosphate and phenylphosphonic acid, and for the materials formed following the n-butylamine intercalation procedure on these samples.The corresponding data for a-zirconium phosphate (denoted a-ZrP) are also given Sample a-ZrP AB C DEF G Total mass loss (%) 15.0 19.7 22.5 28.4 30.8 31.3 31.9 34.1 Total mass loss for n-butylamine intercalate (%) 40.4 39.6 38.9 37.8 37.3 37.2 37.1 36.5J. Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 2317 phenylphosphonic acid over that required for a stoichiometric exchange of the phosphate groups, and this sample was indeed the one closest to pure a-zirconium phenylphosphonate (representing 98% exchange of the phosphate groups).Increasing the quantity of phenylphosphonic acid used in the synthesis, or re-immersing the sample in a fresh sample of molten phenylphosphonic acid may be successful in producing a pure phase of a-zirconium phenylphosphonate. When lower molar ratios of phenylphosphonic acid to a-zirconium phosphate are used (samples A–D), there is an almost stoichiometric exchange of phosphate groups by phenylphosphonate groups.Clearly it becomes progressively less easy to replace the final phosphate groups, and a significant excess of phenylphosphonic acid is required. For the materials obtained following the n-butylamine intercalation procedure for samples A–G, the TGA curves again exhibit stepwise mass losses. A representative TGA curve is shown in Fig. 3, and total mass losses for all samples studied are reported in Table 2.The first step occurs between ambient temperature and 150 8C and is due to the loss of water. The second step, from 150 to 360 8C, involves desorption/decomposition of the intercalated n-butylamine molecules. The third step, in the region 420–570 8C, is due to oxidation of the phenyl moiety and dehydration to yield zirconium pyrophosphate. These observations are again in agreement with observations for corresponding materials containing titanium.20 The powder X-ray diVraction patterns of samples A–G are shown in Fig. 4. All materials have a reflection at 2q ª 5.98 (corresponding to d spacing 15.1 Å), assigned as the {002} reflection of a-zirconium phenylphosphonate. This reflection is not present in the powder X-ray diVractogram of a-zirconium phosphate and is in good agreement with results reported previously 3,25 for a-zirconium phenylphosphonate synthesized by other methods. The relative intensity of this reflection increases as the ratio of phenylphosphonic acid to a-zirconium phosphate in the reaction mixture increases (samples A to G), with a concomitant decrease in the relative intensity of the reflection at 2q ª 11.78 and the pair of reflections in the range 2q ª 25–268 Fig. 3 Typical TGA curve for the materials obtained following the n-butylamine intercalation procedure for samples A–G. The TGA curve shown is for sample D Table 3 Stoichiometry [Zr(O3POH)2?H2O]x[Zr(O3PC6H5)2]y of samples A–G determined from TGA data and high-resolution solidstate 31P NMR data TGA 31P NMR Sample AB C DEF G x 1.43 1.20 0.65 0.40 0.34 0.28 0.02 y 0.57 0.80 1.35 1.60 1.66 1.72 1.98 x 1.37 1.17 0.67 0.43 0.36 0.25 0.05 y 0.63 0.83 1.33 1.57 1.67 1.75 1.95 which are assigned to the pure phase of a-zirconium phosphate [the peak at 2q ª 11.78 (corresponding to d spacing 7.56 Å) is the {002} reflection of a-zirconium phosphate]. However, when the ratio of phenylphosphonic acid to a-zirconium phosphate in the reaction mixture is greater than 1.5 (samples E–G), no further decrease in the intensity of the peak at 2q ª 11.78 is observed.This is because, by coincidence, the {004} reflection for a-zirconium phenylphosphonate also occurs at 2q ª 11.78, and therefore overlaps the {002} reflection for a-zirconium phosphate. The powder X-ray diVractograms provide no evidence for the production of a single-phase mixed derivative, for which a peak corresponding to a d spacing of about 11.3 Å may be expected {corresponding, for example, to a single phase containing both phenylphosphonate and phosphate groups ‘interspersed’ between the layers [Fig. 1(b)] or to a single phase with phenylphosphonate and phosphate groups occupying layer faces opposite each other}. There is also no evidence for ‘staged’ products [Fig. 1(d)], for which a peak corresponding to d spacing in the range 22.5–25.4 Å would be expected (the actual value would depend on the level of hydration of the material).24 In summary, the powder X-ray diVraction data suggest that the materials formed in our preparation procedure are mixtures of the pure phases of a-zirconium phosphate and a-zirconium phenylphosphonate. To further confirm this conclusion, samples A–G were subjected to the conditions for intercalation of n-butylamine.n-Butylamine is preferentially intercalated into regions of a host material in which there are opportunities for acid–base interactions, and should therefore be included preferentially into a-zirconium phosphate rather than a-zirconium phenylphosphonate.Following this attempted intercalation procedure on samples A–G, the powder X-ray diVractograms (Fig. 5) of all materials have an additional reflection at 2q ª 4.78, corresponding to a d spacing of about 18.8 Å. This observation is consistent 26 with intercalation of an extended bimolecular layer of n-butylamine within a-zirconium phosphate (with the longitudinal axes of the n-butylamine molecules inclined at ca. 598 with respect to the plane of the layer). Thus, following the n-butylamine intercalation procedure, samples A–G (assigned above as mixtures of a-zirconium phosphate and Fig. 4 Powder X-ray diVractograms of the samples A–G obtained in the synthesis involving a-zirconium phosphate and phenylphosphonic acid. The diVerent samples were prepared from diVerent initial ratios of phenylphosphonic acid to a-zirconium phosphate in the reaction mixture (as detailed in Table 1).The absolute intensities for the diVerent samples are not normalized2318 J. Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 a-zirconium phenylphosphonate) are converted into mixtures of a-zirconium phosphate intercalated with n-butylamine and a-zirconium phenylphosphonate (essentially unaltered by the intercalation procedure). This conclusion is further supported by the fact that the relative intensity of the new peak at 2q ª 4.78, following the n-butylamine intercalation procedure, decreases from sample A to sample G (i.e.as the amount of a-zirconium phosphate decreases). Consistent with our suggestion that the a-zirconium phenylphosphonate does not intercalate n-butylamine is the fact that the peak at 2q ª 5.98 remains unchanged following the n-butylamine intercalation procedure. In summary, all the evidence from our powder X-ray diVraction studies suggests that samples A–G are physical mixtures of the pure phases of a-zirconium phenylphosphonate and a-zirconium phosphate, rather than single phases containing both phosphate and phenylphosphonate groups.The FT-IR spectra of samples A–G (Fig. 6) confirm that they contain both phosphate and phenylphosphonate groups. Thus, characteristic bands 27 for a-zirconium phenylphosphonate and a-zirconium phosphate at 1438 [phenyl n(C]] C)], 748 and 692 [d out of plane], 3440 [n(O]H)], 3055 [aromatic n(C]H)], 1156 [n(P]C)] and 1044 cm21 [n(P]O(H))] are all Fig. 5 Powder X-ray diVractograms of the materials obtained following the n-butylamine intercalation procedure for samples A–G Fig. 6 Fourier-transform IR spectra for samples obtained in the synthesis involving a-zirconium phosphate and phenylphosphonic acid: (a) sample A, (b) sample G present. It is interesting to note that the sharp bands at 3593 and 3510 cm21 decrease in intensity as the ratio of phenylphosphonic acid to a-zirconium phosphate in the synthesis mixture increases (i.e.from sample A to sample G). These bands are attributed to the asymmetric and symmetric stretching modes of the interlayer water molecule of a-zirconium phosphate. High-resolution solid-state 31P NMR spectra (Fig. 7) of samples A–G contain two main peaks at d 25.3 and 218.6, assigned to phenylphosphonate and phosphate groups respectively, in line with results reported previously.24,28 (We note that the 31P NMR spectra contain other peaks of low intensity, suggesting that the samples also contain small amounts of other phosphorus-containing species, perhaps at the surfaces of the crystallites or in less well ordered regions of the samples.At present, a definite assignment of the identity of these minor components of the samples has not been made.) As the ratio of phenylphosphonic acid to a-zirconium phosphate in the synthesis mixture is increased (i.e. from sample A to sample G), the relative intensity of the peak at d 25.3 increases and the relative intensity of the peak at d 218.6 decreases.However, the peak at d 218.6 due to a-zirconium phosphate is always present, even when a large excess of phenylphosphonic acid is used in the synthesis. This observation is in agreement with our conclusions from powder X-ray diVraction discussed above. Integration of the peaks at d 25.3 and 218.6 provides an estimate of the ratio of a-zirconium phenylphosphonate to a-zirconium phosphate (see Table 3) and the stoichiometries determined in this way are in good agreement with those determined from the TGA data.Conclusion In summary, the topotactic synthesis procedure described in this paper has been used successfully to produce a-zirconium phenylphosphonate by reaction of a-zirconium phosphate with molten phenylphosphonic acid. It is clear from powder X-ray diVraction that the materials formed are not single phases containing both phenylphosphonate and phosphate groups, but rather they are physical mixtures of the pure phases of a-zirconium phosphate and a-zirconium phenylphosphonate.The TGA data and high-resolution solid-state 31P NMR data indicate clearly that the quantity of a-zirconium phenylphosphonate formed increases as the ratio of phenylphosphonic acid to a-zirconium phosphate in the reaction mixture is increased. However, all the materials synthesized contain some Fig. 7 High-resolution solid-state 31P NMR spectra for samples A–G obtained in the synthesis involving a-zirconium phosphate and phenylphosphonic acidJ.Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 2319 amount of a-zirconium phosphate, even when a large excess of phenylphosphonic acid is used in the synthesis procedure. The production of mixtures of pure a-zirconium phosphate and pure a-zirconium phenylphosphonate, rather than a single phase containing both phosphate and phenylphosphonate groups, is interesting {although we note that the synthesis of a staged material [Fig. 1(d)] has been reported 24 using hydrothermal methods}.Evidently for the synthetic method reported here, the phenylphosphonate groups become preferentially incorporated into certain layers, arising from the mobility of the phenylphosphonate anions in the interlayer region, which ultimately become pure phenylphosphonate layers in the a-zirconium phenylphosphonate phase of the product. The results reported here have several parallels with a previous report 20 of the topotactic formation of a-titanium phenylphosphonate from a-titanium phosphate. This reaction involved repeated contact of a-titanium phosphate with an aqueous solution of phenylphosphonic acid, and resulted in the formation of mixtures of a-titanium phosphate and a-titanium phenylphosphonate.The extent of formation of a-titanium phenylphosphonate was found to be related to the contact time between solid and solution. The incomplete replacement of the phosphate groups by phenylphosphonate groups was attributed to the establishment of an equilibrium between the a-titanium phosphate and a-titanium phenylphosphonate. It is plausible that a similar equilibrium may be set up during the reaction reported here between a-zirconium phosphate and molten phenylphosphonic acid. Thus, a-zirconium phosphate may be partially hydrolysed under the conditions of the experiment to give phosphoric acid and ZrIV, which then reacts with the phenylphosphonic acid to form a-zirconium phenylphosphonate.Equilibrium (1) between the a-zirconium phenyl- Zr(HPO4)2 1 2 RPO3H2 Zr(RPO3)2 1 2 H3PO4 (1) phosphonate and a-zirconium phosphate phases may then be established. Clearly the position of this equilibrium will be altered by changing the ratio of phenylphosphonic acid to a-zirconium phosphate in the initial reaction mixture. Although the results presented here would be consistent with the existence of such an equilibrium, they do not prove that this is actually the mechanism that operates.Further investigations are clearly required to understand in detail, from both chemical and thermodynamic viewpoints, the mechanism of the synthetic reaction for the preparation of a-zirconium phenylphosphonate reported here. Ultimately, such an understanding may be applied to determine the conditions required to produce a single pure phase of a-zirconium phenylphosphonate via this synthetic procedure. Acknowledgements We are grateful to the NuYeld Foundation and EPSRC for financial support.References 1 A. Clearfield and G. D. Smith, Inorg. Chem., 1969, 8, 431. 2 A. Clearfield, R. H. Blessing and J. A. Stynes, J. Inorg. Nucl. Chem., 1968, 30, 2249. 3 G. Alberti, U. Costantino, S. Alluli and N. Tomassini, J. Inorg. Nucl. Chem., 1978, 40, 1113. 4 A. Clearfield, Curr. Opin. Solid State Chem., 1996, 1, 268. 5 G. Alberti, Adv. Mater., 1996, 8, 291. 6 G. L. Rosenthal and J. Caruso, Inorg.Chem., 1992, 31, 3104. 7 G. Alberti, M. Casciola, U. Costantino, A. Peraio and E. Monerini, Solid State Ionics, 1992, 50, 315. 8 G. Alberti, M. Casciola, R. Palombari and A. Peraio, Solid State Ionics, 1992, 58, 339. 9 R. C. T. Slade, C. R. M. Forano and A. Peraio, Solid State Ionics, 1993, 61, 23. 10 M. Casciola, U. Costantino, A. Peraio and T. Rega, Solid State Ionics, 1995, 77, 229. 11 A. Clearfield, Chem. Rev., 1988, 88, 125. 12 G. Alberti, Solid State Ionics, 1996, 97, 177. 13 M. Ogawa and K. Kuroda, Chem. Rev., 1995, 95, 399. 14 G. Alberti and U. Costantino, J. Mol. Catal., 1994, 27, 235. 15 A. Clearfield, in Surface Organometallic Chemistry: Molecular Approaches to Surface Catalysis, ed. J. M. Basset, Kluwer Academic Publishers, Norwell, MA, 1988, pp. 269–271. 16 S. Yamanaka and M. Hattori, Inorg. Chem., 1981, 20, 1929. 17 S. Yamanaka, M. Matsunaga and M. Hattori, J. Inorg. Nucl. Chem., 1978, 43, 1343. 18 S. Yamanaka, Inorg. Chem., 1976, 15, 2811. 19 S.Yamanaka, M. Tsujimoto and M. Tanaka, J. Inorg. Nucl. Chem., 1978, 41, 605. 20 E. Jaimez, A. Bortun, G. B. Hix, J. R. Garcia, J. Rodriguez and R. C. T. Slade, J. Chem. Soc., Dalton Trans., 1996, 2285. 21 G. B. Hix and K. D. M. Harris, J. Mater. Chem., 1998, 8, 579. 22 J. D. Wang, A. Clearfield and G. Z. Peng, Mater. Chem. Phys., 1993, 35, 187. 23 H. Benhamza, P. Barboux, F.-A. Josien and J. Livage, J. Mater. Chem., 1991, 1, 681. 24 A. Clearfield, J. D. Wang, Y. Tian, E.Stein and C. Bhardwaj, J. Solid State Chem., 1995, 117, 275. 25 M. D. Poojary, H.-L. Hu, F. L. Campbell and A. Clearfield, Acta Crystallogr., Sect. B, 1993, 49, 996. 26 A. Clearfield, in Inorganic Ion Exchange Materials, ed. A. Clearfield, CRC Press Inc., Boca Raton, FL, 1991. 27 N. B. Colthup, L. H. Daly and S. E. Wiberley, Introduction to Infrared and Raman Spectroscopy, Academic Press Inc., San Diego, CA, 1990. 28 N. J. Clayden, J. Chem. Soc., Dalton Trans., 1987, 1877.Received 8th April 1998; Paper 8/02673DJ. Chem. Soc., Dalton Trans., 1998, Pages 2315–2319 2319 amount of a-zirconium phosphate, even when a large excess of phenylphosphonic acid is used in the synthesis procedure. The production of mixtures of pure a-zirconium phosphate and pure a-zirconium phenylphosphonate, rather than a single phase containing both phosphate and phenylphosphonate groups, is interesting {although we note that the synthesis of a staged material [Fig. 1(d)] has been reported 24 using hydrothermal methods}. Evidently for the synthetic method reported here, the phenylphosphonate groups become preferentially incorporated into certain layers, arising from the mobility of the phenylphosphonate anions in the interlayer region, which ultimately become pure phenylphosphonate layers in the a-zirconium phenylphosphonate phase of the product. The results reported here have several parallels with a previous report 20 of the topotactic formation of a-titanium phenylphosphonate from a-titanium phosphate.This reaction involved repeated contact of a-titanium phosphate with an aqueous solution of phenylphosphonic acid, and resulted in the formation of mixtures of a-titanium phosphate and a-titanium phenylphosphonate. The extent of formation of a-titanium phenylphosphonate was found to be related to the contact time between solid and solution. The incomplete replacement of the phosphate groups by phenylphosphonate groups was attributed to the establishment of an equilibrium between the a-titanium phosphate and a-titanium phenylphosphonate.It is plausible that a similar equilibrium may be set up during the reaction reported here between a-zirconium phosphate and molten phenylphosphonic acid. Thus, a-zirconium phosphate may be partially hydrolysed under the conditions of the experiment to give phosphoric acid and ZrIV, which then reacts with the phenylphosphonic acid to form a-zirconium phenylphosphonate.Equilibrium (1) between the a-zirconium phenyl- Zr(HPO4)2 1 2 RPO3H2 Zr(RPO3)2 1 2 H3PO4 (1) phosphonate and a-zirconium phosphate phases may then be established. Clearly the position of this equilibrium will be altered by changing the ratio of phenylphosphonic acid to a-zirconium phosphate in the initial reaction mixture. Although the results presented here would be consistent with the existence of such an equilibrium, they do not prove that this is actually the mechanism that operates.Further investigations are clearly required to understand in detail, from both chemical and thermodynamic viewpoints, the mechanism of the synthetic reaction for the preparation of a-zirconium phenylphosphonate reported here. Ultimately, such an understanding may be applied to determine the conditions required to produce a single pure phase of a-zirconium phenylphosphonate via this synthetic procedure. Acknowledgements We are grateful to the NuYeld Foundation and EPSRC for financial support.References 1 A. Clearfield and G. D. Smith, Inorg. Chem., 1969, 8, 431. 2 A. Clearfield, R. H. Blessing and J. A. Stynes, J. Inorg. Nucl. Chem., 1968, 30, 2249. 3 G. Alberti, U. Costantino, S. Alluli and N. Tomassini, J. Inorg. Nucl. Chem., 1978, 40, 1113. 4 A. Clearfield, Curr. Opin. Solid State Chem., 1996, 1, 268. 5 G. Alberti, Adv. Mater., 1996, 8, 291. 6 G. L. Rosenthal and J. Caruso, Inorg. Chem., 1992, 31, 3104. 7 G. Alberti, M.Casciola, U. Costantino, A. Peraio and E. Monerini, Solid State Ionics, 1992, 50, 315. 8 G. Alberti, M. Casciola, R. Palombari and A. Peraio, Solid State Ionics, 1992, 58, 339. 9 R. C. T. Slade, C. R. M. Forano and A. Peraio, Solid State Ionics, 1993, 61, 23. 10 M. Casciola, U. Costantino, A. Peraio and T. Rega, Solid State Ionics, 1995, 77, 229. 11 A. Clearfield, Chem. Rev., 1988, 88, 125. 12 G. Alberti, Solid State Ionics, 1996, 97, 177. 13 M. Ogawa and K. Kuroda, Chem. Rev., 1995, 95, 399. 14 G. Alberti and U. Costantino, J. Mol. Catal., 1994, 27, 235. 15 A. Clearfield, in Surface Organometallic Chemistry: Molecular Approaches to Surface Catalysis, ed. J. M. Basset, Kluwer Academic Publishers, Norwell, MA, 1988, pp. 269–271. 16 S. Yamanaka and M. Hattori, Inorg. Chem., 1981, 20, 1929. 17 S. Yamanaka, M. Matsunaga and M. Hattori, J. Inorg. Nucl. Chem., 1978, 43, 1343. 18 S. Yamanaka, Inorg. Chem., 1976, 15, 2811. 19 S. Yamanaka, M. Tsujimoto and M. Tanaka, J. Inorg. Nucl. Chem., 1978, 41, 605. 20 E. Jaimez, A. Bortun, G. B. Hix, J. R. Garcia, J. Rodriguez and R. C. T. Slade, J. Chem. Soc., Dalton Trans., 1996, 2285. 21 G. B. Hix and K. D. M. Harris, J. Mater. Chem., 1998, 8, 579. 22 J. D. Wang, A. Clearfield and G. Z. Peng, Mater. Chem. Phys., 1993, 35, 187. 23 H. Benhamza, P. Barboux, F.-A. Josien and J. Livage, J. Mater. Chem., 1991, 1, 681. 24 A. Clearfield, J. D. Wang, Y. Tian, E. Stein and C. Bhardwaj, J. Solid State Chem., 1995, 117, 275. 25 M. D. Poojary, H.-L. Hu, F. L. Campbell and A. Clearfield, Acta Crystallogr., Sect. B, 1993, 49, 996. 26 A. Clearfield, in Inorganic Ion Exchange Materials, ed. A. Clearfield, CRC Press Inc., Boca Raton, FL, 1991. 27 N. B. Colthup, L. H. Daly and S. E. Wiberley, Introduction to Infrared and Raman Spectroscopy, Academic Press Inc., San Diego, CA, 1990. 28 N. J. Clayden, J. Chem. Soc., Dalton Trans., 1987, 1877. Received 8th April 1998; Paper 8/02673D

ISSN:1477-9226    

DOI:10.1039/a802673d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2317–2321 2317 Synthesis, structure and characterization of novel nickel(II) and iron(II) complexes with a 5,59-bis[2-(2,29-bipyridin-6-yl)-ethyl]-2,29- bipyridine ligand Mou-Hai Shu, Chun-Ying Duan, Wei-Yin Sun,* You-Jun Fu, Dao-Hua Zhang, Zhi-Ping Bai and Wen-Xia Tang * State Key Laboratory of Coordination Chemistry, Coordination Chemistry Institute, Nanjing University, Nanjing 210093, China Received 9th March 1999, Accepted 28th May 1999 A new tris-bipyridine ligand 5,59-bis[2-(2,29-bipyridin-6-yl)-ethyl]-2,29-bipyridine (L) was synthesized, and complexes [ML3][PF6]2?2EtOH?0.5H2O [M = Ni(II), 1; M = Fe(II), 2] were obtained by reaction of the ligand L with Ni(II) and Fe(II) ions, respectively.X-Ray data of complex 1 show that the central 5,59-disubstituted 2,29-bipyridine units of each ligand L coordinate to the Ni(II) ion to give a distorted octahedral environment, while all the terminal 6-monosubstituted 2,29-bipyridine groups of the ligand keep free of coordination.Complex 2 is isomorphic to 1. The solution behavior of complex 2 was investigated by 1H NMR spectroscopy. Complexes 1 and 2 were also characterized by ES-MS spectrometry and cyclic voltammetry. The ES-MS spectral data indicate that only the mononuclear complexes formed in the reaction mixtures of the ligand L and M(ClO4)2 (M = Fe and Ni) even in the presence of excess metal ion. The results illustrate that the central 5,59-disubstituted 2,29-bipyridine moiety of each ligand L is selectively coordinated by octahedral geometric metal ions, whereas the 6-substituted 2,29-bipyridine unit does not participate in any metal ion coordination.Introduction Oligobipyridine ligands with diVerent spacer groups and diVerent linkage modes have been widely used in the study of the assembly of supramolecules in recent years.1–4 For example, 1,2- bis(2,29-bipyridin-6-yl)ethane coordinates to Cu(II) in a tetradentate mode to give a mononuclear complex.3 In the case of a tris-2,29-bipyridine ligand with 4,49- and 6,69-disubstituted 2,29-bipyridine units, the terminal 4,49-disubstituted 2,29- bipyridine moieties are selectively coordinated by Ru(bipy)2 (bipy = 2,29-bipyridine), whereas the central 6,69-disubstituted 2,29-bipyridine units do not participate in any metal ion coordination.4 However, oligobipyridine ligands containing 6,69- or 5,59-disubstituted 2,29-bipyridine result in the formation of double- and triple-stranded helixes 5–10 or novel circular double helical complexes.11,12 Therefore it appears that the linkage mode of the spacer group to the bipyridine units in the oligobipyridine ligands plays an important role in the binding behavior of the 2,29-bipyridine group. In order to further illustrate this argument, we designed and synthesized a new ligand L, in which three bipyridine units are bridged by a CH2CH2 group at the 5 and 69-positions, i.e.the central bipyridine unit is 5,59-disubstituted and terminal bipyridine units are 6-monosubstituted as shown schematically below (Scheme 1), and the Scheme 1 N N N N N N N N N N N N L L¢ reactions between L and Ni(II) and Fe(II) were investigated. The results show that the oligobipyridine ligand L coordinates to Ni(II) and Fe(II) with the central 5,59-disubstituted 2,29- bipyridine moiety to form mononuclear complexes and the terminal 6-monosubstituted 2,29-bipyridine units are pendant.The results indicate that the linkage mode of the spacer group to the bipyridine units in the oligobipyridine ligands influence the assembly of the supramolecules as well as the binding of the 2,29-bipyridine group. Experimental Materials Anhydrous THF was refluxed with sodium–benzophenone and distilled before use. 6-Methyl-2,29-bipyridine 3 and 5,59-bis- (bromomethyl)-2,29-bipyridine 13 were synthesized according to the literature methods.Acetonitrile was purified by treatment with KMnO4 and then distilled over P2O5 and K2CO3. All other chemicals were of reagent grade quality obtained from commercial sources and used without further purification. Preparation of 5,59-bis[2-(2,29-bipyridin-6-yl)ethyl]-2,29- bipyridine (L) The ligand L was prepared by the reaction of 6-lithiomethyl- 2,29-bipyridine with 0.5 equivalents of 5,59-bis(bromomethyl)- 2,29-bipyridine in anhydrous THF at 278 8C under an argon atmosphere.8 The crude product was washed with methanol, recrystallized from MeOH–CHCl3 (1/1, v/v), and purified by chromatography on silica with light petroleum (bp range: 60– 90 8C)–acetone (5 : 1, v/v) as eluting agent, after recrystallizing from acetonitrile, the ligand L was obtained as a pure white powder in 40% yield, mp 168–170 8C, ES-MS: m/z, 521.3 (M 1 H)1, 261.5 (M 1 2H)21; 1H NMR (CD3CN, 298 K): d 3.23 (s, 8 H); 7.22 (d, 2 H); 7.36 (td, 2 H); 7.70 (dd, 2 H); 7.77 (t, 2H); 7.85 (td, 2H); 8.22 (d, 2 H); 8.24 (d, 2 H), 8.43 (d, 2H); 8.47 (s, 2 H), 8.65 (d, 2 H).Found: C 78.49; H 5.52; N 15.78. Calc. for C34H28N6: C 78.43, H 5.42, N 16.15%.2318 J. Chem. Soc., Dalton Trans., 1999, 2317–2321 Preparation of [NiL3][PF6]2?2EtOH?0.5H2O 1 60 mg of L (0.115 mmol) was suspended in ethanol (30 ml) and a solution containing 28.8 mg of Ni(OAc)2?4H2O (0.115 mmol) in water (4 ml) was added dropwise. The reaction mixture was refluxed with stirring for 4 hours, [NiL3][PF6]2 was precipitated by addition of excess of KPF6.The product was isolated by filtration, washed with water and diethyl ether, dried in vacuum, yield 64 mg (83%). Pink single crystals were obtained by slow evaporation from a solution of acetonitrile and methanol. Found: C 63.85; H 4.72; N 12.83. Calc. for C106H97F12N18- NiO2.5P2: C 63.29; H 4.86; N 12.53% (the deviation may be caused by loss of the ethanol solvent molecule). Preparation of [FeL3][PF6]2?2EtOH?0.5H2O 2 The complex was synthesized by reaction of FeSO4?7H2O (0.115 mmol) with the ligand L (0.115 mmol) following the same procedures as used for the synthesis of complex 1.The product was isolated as a dark red powder, washed with water and diethyl ether, dried in vacuum, yield 65 mg (85%). Red single crystals were obtained by slow evaporation from a solution of acetonitrile and toluene. 1H NMR (CD3CN, 298 K): d 2.73 (m, 2 H), 2.85 (m, 2 H); 6.25 (s, 2 H); 6.59 (d, 2 H); 7.37 (d, 2 H); 7.45 (t, 2 H); 7.47 (t, 2 H); 7.56 (d, 2 H); 7.85 (td, 2 H); 8.13 (d, 2 H), 8.19 (d, 2 H); 8.80 (s, 2 H).Found: C 63.76; H 4.81; N 12.90. Calc. for C106H97F12N18FeO2.5P2: C 63.38; H 4.87; N 12.55%. Crystallography The diVraction intensities of complexes 1 and 2 were collected on a Siemens P4 four-circle diVractometer with graphitemonochromated Mo-Ka radiation (l=0.71073 Å) at 293(2) K using the w/2q scan mode. Data were corrected for Lorentzpolarization eVects during data reduction using XSCANS,14 and semi-empirical absorption correction from psi-scans was applied.The structure of complex 1 was solved by direct methods and refined on F2 by full-matrix least square methods using SHELXTL version 5.0.15 The PF6 2 anions are disordered, P–F distances were fixed at 1.540(10) Å, the site occupancy factors (s.o.f.) of these disordered fluorine atoms were refined by setting the free variable as 0.53 for F(1), F(2), F(3), F(4), F(5), F(6), and 0.47 for F(19), F(29), F(39), F(49), F(59), F(69), respectively.The hydrogen atoms of the water molecule were found from the diVerence Fourier map and refined using a riding model. The oxygen atom of the water molecule at the special position (0, 0.5, 1.0) was also refined disordered with s.o.f. = 0.25. The ethanol molecules are also disordered, the C–C distances were fixed at 1.500(20) and the C–O distances were fixed at 1.500(20) Å, the site occupancy factors were refined by setting the free variable as 0.64 for C(52), C(53), O(1), and 0.36 for C(529), C(539), O(19), respectively. The other hydrogen atoms were placed in calculated positions (C–H, 0.96) assigned fixed isotropic thermal parameters at 1.2 times the equivalent isotropic U of the atoms to which they are attached and allowed to ride on their respective parent atoms.The contributions of these hydrogen atoms were included in the structure-factor calculations. All computations were carried out on a PC-586 computer using the SHELXTL-PC program package.15 Parameters for data collection and refinement of 1 are summarized in Table 1 and selected bond lengths and bond angles are listed in Table 2.Details of the structure of complex 2 are not reported because the diVraction intensities are weak and attempts to refine the structure were unsuccessful, however, the same space group and the similar cell parameters [a = 22.473(3), b = 16.880(2), c = 25.901(3) Å, b = 95.899(10)8, and V = 9773(2) Å3 at 293(2) K] indicate that complex 2 is an isomorph of complex 1.CCDC reference number 186/1477. See http://www.rsc.org/suppdata/dt/1999/2317/ for crystallographic files in .cif format. ES-MS spectral measurement Complexes 1 and 2 were dissolved in 1 : 1 (v/v) acetonitrile– water, with the approximate concentration 0.01 mmol L21. The other samples were prepared by the following procedures: 1.00 ml of ligand L in acetonitrile (1.0 mg ml21) was mixed with a measured portion of freshly prepared standardised M(ClO4)2 solutions (M = Fe and Ni) in water, and the resulting mixtures were stirred for 4 hours at 70 8C to give clear solutions.Electrospray mass spectra (ES-MS) were measured on a LCQ system (Finngan MAT, USA) using methanol as the mobile phase. The spray voltage, tube lens oVset, capillary voltage and capillary temperature were set at 4.50 kV, 0 V, 17.00 V and 150 8C, respectively. The quoted m/z values represent the major peaks in the isotopic distribution. Molecular modeling Cerius 2 (Version 3.5)16 software program from Molecular Simulation Inc.was employed to do the molecular modeling. Dynamic calculations and energy minimization were carried out using the Universal forcefield 17 which is an open forcefield (OFF). Other physical measurements 1H NMR spectra were recorded on a Bruker AM-500 spectrometer using the residual proton in deuterated solvent as internal reference (CD3CN, 1.93 ppm).Electrochemical measurements were performed on an EG&G M273 potentiostat/ galvanostat system using a platinum disk working electrode (0.008 cm2), which was polished prior to use and rinsed thoroughly with water and acetone, a platinum wire counter electrode and an Ag–AgCl electrode used as reference. All the measurements were carried out at 298 K in argon-purged (dioxygen-free) acetonitrile solutions with 0.1 M freshly Table 1 Summary of crystal data, data collection and structure refinement for complex 1 Empirical formula Formula weight T/K Crystal system Space group a/Å b/Å c/Å b/8 U/Å3, Z m/mm21 Measured/Independent reflections R1 wR2, wR2 (for all data) C106H97F12N18NiO2.5P2 2011.67 293(2) Monoclinic C2/c 22.705(8) 16.885(2) 25.917(4) 96.44(2) 9856(4), 4 0.313 9717/8272, [Rint = 0.0519] 0.0720 0.1735, 0.2258 Table 2 Selected bond lengths (Å) and angles (8) for complex 1 Ni(1)–N(3) Ni(1)–N(4) Ni(1)–N(9) N(3)–Ni(1)–N(4) N(3)–Ni(1)–N(9) N(3)–Ni(1)–N(3) a N(3)–Ni(1)–N(4) a N(3)–Ni(1)–N(9) a N(4)–Ni(1)–N(9) N(4)–Ni(1)–N(3) a N(4)–Ni(1)–N(4) a 2.104(3) 2.083(3) 2.100(3) 78.88(14) 172.30(13) 91.3(2) 94.43(13) 95.29(13) 96.60(13) 94.43(13) 170.5(2) Ni(1)–N(3) a Ni(1)–N(4) a Ni(1)–N(9) a N(4)–Ni(1)–N(9) a N(9)–Ni(1)–N(3) a N(9)–Ni(1)–N(4) a N(9)–Ni(1)–N(9) a N(3) a–Ni(1)–N(4) N(3) a–Ni(1)–N(9) N(4) a–Ni(1)–N(9) 2.104(3) 2.083(3) 2.100(3) 90.76(12) 95.29(13) 90.76(12) 78.4(2) 78.88(14) 172.29(13) 96.60(13) Symmetry transformations used to generate equivalent atoms: a 2x, y, 2z 1 3/2.J. Chem.Soc., Dalton Trans., 1999, 2317–2321 2319 Fig. 1 The structure of the cation [NiL3]21 with atom numbering scheme. All the hydrogen atoms were omitted for clarity. recrystallized tetra-n-butylammonium perchlorate [(TBA)ClO4] as supporting electrolyte. The scan rate was 100 mV s21. Ferrocene was added at the end of each experiment as an internal reference. Results and discussion The crystal structure of complex 1 Fig. 1 shows the structure of cation part [NiL3]21 of complex 1 with the atom numbering scheme. The central 5,59-disubstituted bipyridine units of each ligand coordinate the Ni(II) ion, while all the terminal 6-monosubstituted bipyridine units of the ligand keep free of coordination. The central Ni(II) ion has a distorted octahedral coordination sphere. The Ni–N distances vary from 2.083(3) to 2.104(3) Å. The cis-N–Ni–N bond angles vary from 78.4(2) to 96.60(13)8 and from 170.5(2) to 172.29(13)8 for the trans as listed in Table 2.The Ni(II) center is a tris- (bipyridine) complex with D3 local symmetry ignoring the terminal substituted 2,29-bipyridine groups of each ligand. The two-fold rotation axis passes through the Ni(II) ion and the center between N(9) and N(9A) atoms, the corresponding twofold rotation symmetry operation can generate the other half of the molecule. Each pyridine ring is planar with a maximum deviation of 0.03 Å, and the bipyridine fragments in each ligand L are almost planar as reflected by their dihedral angles [N(1)/N(2): 6.3; N(3)/N(4): 1.9; N(5)/N(6): 8.0; N(7)/N(8): 4.6; N(9)/N(9A): 4.78 the pyridine rings are numbered according to the nitrogen atom as shown in Fig. 1]. The dihedral angles between the bipyridine units are 59.88 for N(1),N(2)/N(3),N(4), 63.78 for N(3),N(4)/N(5),N(6), and 69.78 for N(7),N(8)/N(9),N(9A), or N(7A),N(8A)/N(9),N(9A), respectively (the bipyridine planes were named after the nitrogens they contained).The torsion angles are 268.88 for C(10)–C(11)–C(12)–C(14), and 254.78 for C(21)–C(23)–C(24)–C(25) whereas the C–CH2–CH2–C torsion angle in the strand containing N(9), N(9A) is 2171.88 for C(44)–C(45)–C(46)–C(48). It is diVerent from those in the other two ligand strands as just mentioned. It can be seen from Fig. 1 that the strands containing N(3),N(4), and N(3A),N(4A) are intertwined around the Ni(II) ion while the strand containing N(9),N(9A) adopts an almost linear arrangement.In complex 1, all the terminal 6-monosubstituted bipy units of the three ligands keep free of coordination. However, in the case of the 5,59-disubstituted tris-bipyridine ligand (L9) (see Scheme 1) the triple-stranded helical complex has been obtained with the same nickel(II) ion. The absence of the trinuclear triple-stranded helicate species M3L3 with the ligand L is considered to be caused by the steric hindrance of the methylene substituents alpha to the chelating nitrogens of the terminal bipyridine units.Balzani and co-workers have obtained similar results 4 in which the central 6,69-substituted bipyridine unit does not participate in the coordination with [Ru(bpy)2]21 (bpy = bipyridine), PdCl2, ZnBr2 and Hg(CF3- CO2)2 and only the terminal 4,49-disubstituted 2,29-bipyridine moieties coordinate to the metal ions. Electrospray mass spectra of the complexes The electrospray mass spectra (ES-MS) of complexes 1 and 2 are shown in Fig. 2. The spectrum of complex 1 displays three peaks [Fig. 2(a)], a base peak at m/z 809.9 in which the isotopic distribution patterns separated by 0.5 ± 0.1 dalton correspond Fig. 2 (a) Electrospray mass spectrum of complex 1, (b) Electrospray mass spectrum of complex 2, the insets show the comparison of the observed (traces) and calculated (bars) isotopic distributions for the base peaks observed in the ES-MS spectra of the complexes.2320 J.Chem. Soc., Dalton Trans., 1999, 2317–2321 to the major species present: [NiL3]21, a small peak at m/z 1764.2 with 1.0–1.1 dalton separated isotopic distribution patterns results from the cation {[NiL3][PF6]}1, and another small peak at m/z 549.4 corresponds to [NiL2]21. As shown in Fig. 2(b), the ES-MS spectrum of 2 includes four main peaks, a peak at m/z 809.1 corresponding to the most abundant double cation [FeL3]21, a peak at m/z 1761.7 resulting from the monocation {[FeL3][PF6]}1 while the peaks at m/z 548.5 and 521.5 correspond to the species [FeL2]21 and LH1, respectively.The isotopic patterns of the signals at m/z 809.1 in complex 2 show peaks separated by 0.5 ± 0.1 dalton confirming the existence of a doubly charged species, and the 1.0–1.1 dalton separated isotopic distribution patterns of the peaks at m/z 1761.7 in complex 2 indicate the existence of a singly charged cation. The insets show the comparison of observed (traces) and calculated (bars) isotopic distributions for the major species.In all cases the agreement between experimental and calculated isotopic distribution patterns is excellent and supported the above assignments. The results indicate that the structure of complexes 1 and 2 in solution are the same as those in the solid state which were determined by X-ray diVraction studies. ES-MS spectroscopy was employed to detect the species produced in the reaction of the ligand L with the Ni(II) and Fe(II) ions in solution (at various molar ratios).The ES-MS spectra were recorded for the reaction mixtures of the ligand L with M(ClO4)2 at the ratios M(II):L = 1 : 4, 1 : 3, 2 : 3, 3 : 3, and 3 : 2, respectively. The ES-MS spectra show that there is no iron(II) or nickel(II) polynuclear complex formed in the reaction mixtures in the presence of an excess of the metal ions, and the species formed in the reaction mixtures are mononuclear complexes. Thus the results show that the linkage of the two 6-substituted 2,29-bipyridine units at the 5,59-position of the central bipyridine prevents the formation of polynuclear helical complexes.Molecular modeling In order to investigate the possibility of forming a trinuclear complex [Ni3L3]61 with the ligand L molecular dynamic calculations were carried out. The structure of [Ni3L3]61 (A) was built either by shifting the connection of the spacer groups from the b to the a position of the coordinating nitrogens of the terminal bipyridines and all the terminal methyl groups were replaced by hydrogen atoms in structure [Ni3L93]61 which was obtained by accessing the X-ray crystal structure data from the Cambridge database or by accessing the crystal structure of complex [NiL3]21 and adding two Ni(II) ions which were coordinated by the pendant bipyridine units by the molecular modeling 3D-builder of Cerius 2.The total energy and the parameters for the resulting minimized structures are almost the same no matter how the original structures were entered.For comparison, the structure of the reported complex [Ni3L93]61 (B) in which the ligand L9 contains three bipyridine groups connected by a CH2CH2 spacer group at the 5,59- positions, was built on the basis of the X-ray crystal structure data from the Cambridge database. Since the numbers of the atoms in structure B do not equal those in structure A, another structure (C) which has the same atoms and charge as the structure A was built by replacing all the terminal methyl groups with hydrogen atoms in the structure of [Ni3L93]61 (B), the total energy E and the parameters for the resulting minimized structures A, B, and C were calculated.EA is about 750 kJ mol21 (50%) higher than EB which is about 80 kJ mol21 higher than EC, this implies that it is energetically disadvantageous to form the structure A with the ligand L compared with the structures B and C. Furthermore, building a trinuclear triple helix with the ligand L resulted in loss of the planarities of all the 6-monosubstituted pyridine groups due to the strain caused by the CH2CH2 spacer group, and the largest deviation of the atom from the pyridine plane is 0.2 Å, whereas the deviation is less than 0.03 Å for the complex [Ni3L93]61 in both the crystal and the modeling structures.In addition the CH2 carbon atoms of the adjacent terminal bipyridine unit is also out of the parent pyridine plane with distance longer than 0.3 Å.The results of the modeling and calculations imply that it is geometrically dif- ficult to form a trinuclear complex through linkage of the two 6-monosubstituted bipyridines in the 5,59-position of the central bipyridine. All the above mentioned results are not unexpected in view of the fact that bipyridines substituted at the 6-position do not easily complex with octahedral coordination centers, but rather with tetrahedral metals like copper(I) and silver(I).Electrochemical properties of the complexes Complex 1 shows no voltammetric response in the potential region from 0.0 to 2.0 V in acetonitrile, but exhibits a reversible metal-based redox at E1/2 = 21.30 V (DE = 64 mV) vs. Ag–AgCl in CH3CN solution which is assigned as the (NiII/I) couple. This is somewhat diVerent from the metal-based redox potential of complex [NiII(mt-sexpy)]2[PF6]4 (E1/2 = 21.0 V) in acetonitrile [(mt-sexpy) = 49,40-bis(methylthio)-2,29:69,20:60,2-:6-,2+,6+, 20--sexipyridine].18 The diVerence between the E1/2 values for these complexes is likely to be due to the diVerent coordination environments.Complex 2 displays four apparent cyclic voltammetric responses as shown in Fig. 3, a reversible metal-based (FeIII/II) redox potential at E1/2 = 11.03 V (DE = 56 mV) and three ligand-based redox processes at 21.45, 21.70, and 21.96 V vs. Ag–AgCl in CH3CN. The E1/2 value of metal-based (FeIII/II) redox potential for complex 2 is similar to those of the reported analogs 10 (E1/2 values are in the range of 0.923 to 1.062 V) due to the fact that central Fe(II) ions in complex 2 and the reported complexes have a similar N6 coordination environment and D3 local symmetry. 1H NMR Spectra of complex 2 1H NMR spectra of the ligand L and complex 2 in acetonitriled3 at 298 K are shown in Fig. 4. Only one kind of ligand proton signals were observed for 2 within the NMR timescale, although there is a symmetric center for 2 in the solid state.For resolving all the protons of 2, COSY and NOESY experiments were employed in this work, and peak attributions of the aromatic proton have been assigned as shown in Fig. 4(b). The chemical shifts of the aromatic proton in the complex are diVerent from those in the free ligand. The protons a and b shift downfield by less than 0.16 ppm, while the other protons shift upfield by more than 0.30 ppm, g, j, and h especially shift upfield by 0.63, 0.68 and 2.22 ppm, respectively.Also significant changes occur for the CH2CH2 signals, these protons, which appear as a singlet in the free ligand, shift upfield by more than 0.36 ppm and display an ABCD pattern in the complex. Clearly, the protons of each CH2 group have become nonequivalent in the complex due to the coordination of the central bipyridine group to the Fe(II) ion. Conclusions The present study shows that the coordination to metal occurs only at the 5,59-disubstitued 2,29-bipyridine moiety of the Fig. 3 Cyclic voltammogram of [FeL3][PF6]2?2EtOH?0.5H2O [vs. Ag– AgCl, 100 mV s21, acetonitrile–0.10 M (TBA)ClO4].J. Chem. Soc., Dalton Trans., 1999, 2317–2321 2321 ligand L and supports the idea that 2,29-bipyridine units substituted in the 6-position do not easily form complexes with octahedral coordination centers, but rather with tetrahedral metals like copper(I) and silver(I). The results also indicate that the self-assembly process of a helicate not only is related to the coordination geometry of the metal ions and the ligand donor set, but also can be partially controlled by the linkage position of the spacer group to the bipyridine units in oligobipyridine ligands.Fig. 4 (a) 1H NMR spectra of the ligand L in acetonitrile-d3 at 298 K. (b) COSY (left) and NOESY (right, mixing time: 400 ms) spectra of complex 2 in acetonitrile-d3 at 298 K. Acknowledgements We are grateful to the National Nature Science Foundation of China for financial support.We thank Prof. Dr. Hoong Kun Fun of X-ray Crystallography Unit, University Sains Malasia for his help in X-ray structural analysis. References 1 (a) F. Vögtle, Supramolecular Chemistry, Wiley, Chichester, 1993; (b) J.-M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1995. 2 (a) E. C. Constable, Tetrahedron, 1992, 48 10013; (b) C. Piguet, G. Bernardinelli and G. Hopfgartner, Chem. Rev., 1997, 97, 2005. 3 T. Garber, S.V. Wallendael, D. P. Rillema, M. Kirk, W. E. Hatfield, J. H. Welch and P. Singh, Inorg. Chem., 1990, 29, 2863. 4 B. König, M. Zieg, M. Cumtz, L. D. Cola and V. Balzani, Chem. Ber., 1997, 130, 529. 5 M.-T. Youinou, R. Ziessel and J.-M. Lehn, Inorg. Chem., 1991, 30, 2144. 6 (a) J.-M. Lehn, A. Rigault, J. Siegel, J. Harrowfield, B. Chevrier and D. Morras, Proc. Natl. Acad. Sci. U.S.A., 1987, 84, 2565; (b) U. Koert, M. M. Harding and J.-M. Lehn, Nature, 1990, 346, 339; (c) A. Pfeil and J.-M. Lehn, J. Chem. Soc., Chem. Commun., 1992, 838; (d ) C. R. Woods, M. Benaglia, F. Cozzi and J. S. Seigel, Angew. Chem., Int. Ed. Engl., 1996, 35, 1830; (e) A. Rigault and J.-M. Lehn, Angew. Chem., Int. Ed. Engl., 1988, 27, 1095. 7 T. M. Garrett, U. Koert, J.-M. Lehn, A. Rigault, D. Meyer and J. Fischer, J. Chem. Soc., Chem. Commun., 1990, 557. 8 R. Kramer, J.-M. Lehn, A. D. Cian and J. Fischer, Angew. Chem., Int. Ed. Engl., 1993, 32, 703. 9 B. R. Serr, K. A. Anderson, C. M. Elliott and O. P. Anderson, Inorg. Chem., 1988, 27, 4499. 10 S. Ferrere and C. M. Elliott, Inorg. Chem., 1995, 34, 5818. 11 B. Hasenknopf, J.-M. Lehn, N. Boumediene, A. Dupont, G, A. V. Dorsselaer, B. Kneisel and D. Fenske, J. Am. Chem. Soc., 1997, 119, 10956. 12 B. Hasenknopf, J.-M. Lehn, B. O. Kneisel, G. Baum and D. Fenske, Angew. Chem., Int. Ed. Engl., 1996, 35, 1838. 13 J.-C. Rodriguez-U, B. Alpha, D. Plancherel and J.-M. Lehn, Helv. Chim. Acta., 1984, 67, 2264. 14 XSCANS Version 2.1, Siemens Analytical X-ray Instruments Inc., Madison, WI, 1994. 15 Siemens SHELXTL Version 5.0, Siemens Industrial Automation, Inc., Analytical Instrumentation, Madison, WI, 1995. 16 Cerius 2, version 3.5, Molecular Simulations Inc., 1997. 17 A. K. Rappe, C. J. Casewit, K. S. Colwell, W. A. Goddard III and W. M. SkiV, J. Am. Chem. Soc., 1992, 114, 10024. 18 K. T. Potts, M. Keshavarz, K, F. S. Tham, H. D. Abruna and C. Arana, Inorg. Chem., 1993, 32, 4436. Paper 9/01835B

ISSN:1477-9226    

DOI:10.1039/a901835b

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J. Chem. Soc., Dalton Trans., 1998, Pages 2321–2325 2321 DALTON Structural and 31P CP MAS NMR spectroscopic studies of the P2CuN2 copper(I) complexes [Cu(PPh3)2(MeCN)2]X for X 5 PF6, BF4 and ClO4 John V. Hanna,a Robert D. Hart,b Peter C. Healy,*,†,c Brian W. Skelton b and Allan H. White b a CSIRO North Ryde NMR Facility, North Ryde, 2113, Australia b Department of Chemistry, University of Western Australia, Perth, 6907, Australia c School of Science, Griffith University, Brisbane, 4111, Australia The mixed ligand P2CuN2 copper(I) complexes [Cu(PPh3)2(MeCN)2]X have been studied by one- and twodimensional 31P CP MAS NMR spectroscopy for X = PF6, BF4 or ClO4 and single crystal X-ray diVraction for X = PF6 and ClO4, completing availability of precise structural data for this isomorphous series.The compounds crystallise as discrete cations and anions in space group P21/n with a ª 15, b ª 27, c ª 9 Å, b = 958, Z = 4. The anion is located ca. 6 Å from the copper atom and adjacent to a cleft formed between the acetonitrile ligands and phosphine ligand 2 while the crystallographically independent PPh3 ligands adopt staggered three-bladed propeller-type conformations of opposite chirality.The geometric symmetry of the P2CuN2 co-ordination sphere is low with Cu]P(1) 2.276(4)–2.287(2), Cu]P(2) 2.258(4)–2.269(1) Å, Cu]N 2.023(9)–2.053(3) Å, P]Cu]P 126.82(4)–127.73(5), N]Cu]N 99.5(4)–100.3(1), P(1)]Cu]N 100.87(8)–102.34(9) and P(2)]Cu]N 110.4(1)– 111.9(3)8.One- and two-dimensional solid state 31P CP MAS NMR spectra of the compounds at 9.40 T show chemical shift diVerences of 6 ppm between the signals arising from the two P sites which form part of an ABX spin system with 1J[31P(1)]63Cu] 1.13–1.14 kHz, 1J[31P(2)]63Cu] 1.30 kHz and 2J(31P]31P) 75 Hz. The copper quadrupolar induced distortion of the line spacings is diVerent for the two sites and is postulated to be a consequence of variation in the angle between the Cu]P vectors and the z axis of the electric field gradient tensor.The magnitude of the distortion is relatively small and consistent with small copper quadrupolar coupling constants for the compounds and a balanced electronic charge distribution about the copper(I) site in spite of the low geometric symmetry of the P2CuN2 co-ordination sphere. Bis(triphenylphosphine)copper(I) compounds with monovalent anions, Cu(PPh3)2X, have been shown by single crystal structure determinations to crystallize from polar organic solvents as discrete monomeric [Cu(PPh3)2X] or dimeric [{Cu(PPh3)2X}2] molecules with X acting as monodentate, bidentate or bridging ligands and the copper site(s) three- or four-co-ordinate, depending on the donor and steric properties of the anion.1–9 For the weakly co-ordinating anions PF6 2, BF4 2 and ClO4 2, however, recrystallization from acetonitrile results instead in the formation of mixed ligand ionic compounds [Cu(PPh3)2- (MeCN)2]X in which the anion is displaced from the copper co-ordination sphere by a pair of solvent molecules.10–13 Single crystal structure determinations for the ClO4 10 and BF4 12 compounds show the overall structure of the P2CuN2 copper coordination sphere to be similar and considerably distorted from tetrahedral symmetry.However, line spacing distortions in the solid state 31P CP MAS NMR spectra of the perchlorate compound10 arising from perturbation of the spectra by copper quadrupolar interactions 14–18 were found to be unusually small and consistent with a relatively balanced charge distribution about the copper site.In order to improve the quality of the NMR data for these compounds, and because the cation has been shown to be an active catalyst in cyclopropanation reactions, 12 we recorded one- and two-dimensional (COSY) solid state 31P CP MAS NMR parameters at 9.40 T for all three compounds, together with a determination of the structure of the PF6 compound, completing the availability of structural data for the series.As part of this work we also redetermined the structure of the ClO4 compound as the initial structure determination was completed on a crystal of marginal quality. The results of this work form the basis of the present report. † E-Mail: P.Healy@sct.gu.edu.au Experimental Synthesis The compounds [Cu(PPh3)2(MeCN)2]X, for X = PF6, BF4, ClO4, were prepared according to established procedures.10–13 Dissolution of [Cu(MeCN)4]X (2 mmol) and PPh3 (4 mmol) in warm acetonitrile (20 ml) followed by slow cooling and partial evaporation of the solvent gave well formed air stable crystals of the desired complex.Melting points: X = PF6, 168–172; BF4, 171–176; ClO4, 182–186 8C (decomp.). Crystallography Unique diVractometer data sets were measured at ca. 293 K (2q–q scan mode, monochromatic Mo-Ka radiation, l = 0.710 73 Å) for [Cu(PPh3)2(MeCN)2]X, X = PF6 or ClO4. N Independent reflections were obtained, No with I > 3s(I) being considered ‘observed’ and used in the full matrix least squares refinements after absorption correction.Anisotropic thermal parameters were refined for the non-hydrogen atoms; (x, y, z, Uiso)H were included, constrained at estimated values. Conventional residuals at convergence, R, R9 on |F| are recorded, statistical reflection weights derivative of s2(I) = s2(Idiff) 1 0.0004s4(Idiff) being used. Neutral atom complex scattering factors were used, computation with the XTAL 3.2 program system implemented by S.R. Hall.19 The phenyl rings of the PPh3 ligands are labelled nm where n is the ligand number 1 or 2 and m is the ring number 1, 2 or 3. Phenyl carbons are labelled C(nml), l = 1–6 where 1 is the ipso- and 2 the ortho-carbon that points towards the P atom. Crystal/refinement data. [Cu(PPh3)2(MeCN)2]PF6 º C40H36- CuF6N2P3, M = 814.9, monoclinic, space group P21/n, (C5 2h, no. 14, variant), a = 15.616(3), b = 27.38(1), c = 9.194(7) Å, b =2322 J.Chem. Soc., Dalton Trans., 1998, Pages 2321–2325 95.32(5), U = 3913 Å3, Dc (Z = 4) = 1.38 g cm23, F(000) = 1672, mMo = 7.7 cm21, specimen 0.40 × 0.37 × 0.32 mm, 2qmax = 508, A*min,max = 1.23, 1.30, N = 6830, No = 4039, R = 0.047, R9 = 0.045. [Cu(PPh3)2(MeCN)2]ClO4 º C40H36ClCuN2O4P2, M = 769.7, monoclinic, space group P21/n, a = 15.434(2), b = 26.958(3), c = 9.199(2) Å, b = 94.68(5), U = 3815 Å3, Dc (Z = 4) = 1.34 g cm23, F(000) = 1592, mMo = 7.4 cm21, specimen 0.62 × 0.45 × 0.60 mm, 2qmax = 508, A*min,max = 1.17, 1.37, N = 6692, No = 4669, R = 0.039, R9 = 0.044.CCDC reference number 186/1013. See http://www.rsc.org/suppdata/dt/1998/2321/ for crystallographic files in .cif format. Spectroscopy Solid state 31P CP MAS NMR spectra were acquired at room temperature on a Bruker MSL-400 high field (Bo = 9.40 T) spectrometer operating at a 31P frequency of 161.92 MHz. Conventional cross-polarization and magic-angle-spinning techniques, coupled with spin temperature alternation to eliminate spectral artefacts were implemented using a Bruker 4 mm double air bearing probe in which MAS frequencies of 10 kHz were achieved.A recycle delay of 15 s, contact period of 10 ms and 1H p/2 pulse length of 3 ms were common to all spectra. No spectral smoothing was employed prior to Fourier transformation. Two-dimensional 31P CP MAS correlation spectroscopy (COSY) experiments were implemented at high field with the time proportional phase incrementation (TPPI) method for acquisition of phase sensitive data in both F1 and F2 dimensions.The recycle delay, contact period, 1H p/2 pulse length and MAS spinning rate were the same as those implemented for the one-dimensional experiments. The 31P chemical shifts were externally referenced to solid triphenylphosphine which has a chemical shift of d 29.9 with respect to 85% H3PO4. Results and Discussion Structural data The isomorphous series of compounds [Cu(PPh3)2(MeCN)2]X for X = PF6, BF4 or ClO4 crystallize in space group P21/n with a ª 15, b ª 27, c ª 9 Å, b = 958 and Z = 4.The molecular packing viewed down the c axis is shown in Fig. 1 while a representative view of the cation and an associated anion is shown in Fig. 2. Relevant geometric parameters are listed in Table 1. The anion is located ca. 6 Å from the nearest copper site and adjacent to a cleft formed between the acetonitrile ligands and phosphine ligand 2 with H ? ? ? F/O contact distances of 2.8–3.0 Å between Fig. 1 Molecular packing of [Cu(PPh3)2(MeCN)2]ClO4 viewed down the c axis the anion and phenyl ring (21) and 3.0–3.2 Å between the anion and the methyl group of acetonitrile ligand 2. Weak HPh ? ? ? O/F interactions with contact distances of ca. 2.8 Å also exist between the anion and the cation located at (��� 2 x, ��� 1 y, ��� 2 z). The cations related by the crystallographic centre of symmetry are linked through N ? ? ?HMe interactions between acetonitrile ligands with contact distances of ca. 2.8 Å (Fig. 1). Accommodation of these crystal packing forces results in considerable angular distortion of the P2CuN2 copper coordination sphere of the cation away from C2v symmetry. In particular, the two acetonitrile ligands bend by 108 or more out of the plane orthogonal to the CuP2 plane and towards phenyl rings 11 and 12. The two PPh3 ligands are staggered with respect to each other and adopt three-bladed propeller conformations of opposite chirality with Cu]P(1)]C(1n1)] C(1n2) ª 260, 240 and 2158 and Cu]P(2)]C(2n1)]C(2n2) ª 155, 150 and 1308.This conformational structure diVers from that found for the majority of [Cu(PPh3)2X] complexes in which X? ? ?HPh interactions between the co-ordinated anion and one phenyl ring of each phosphine ligand lock the structure into an Fig. 2 Molecular structure of [Cu(PPh3)2(MeCN)2]1 with its associated perchlorate anion; 20% thermal ellipsoids are shown for the nonhydrogen atoms, hydrogen atoms having arbitrary radii of 0.1 Å Table 1 Core geometries (distances in Å, angles in 8) for [Cu(PPh3)2- (MeCN)2]X, X = PF6, BF4 or ClO4 X Cu]P(1) Cu]P(2) Cu]N(1) Cu]N(2) P(1)]Cu]P(2) N(1)]Cu]N(2) P(1)]Cu]N(1) P(1)]Cu]N(2) P(2)]Cu]N(1) P(2)]Cu]N(2) Cu]N(1)]C(1) Cu]N(2)]C(2) Cu]P(1)]C(111)]C(112) Cu]P(1)]C(121)]C(122) Cu]P(1)]C(131)]C(132) Cu]P(2)]C(211)]C(212) Cu]P(2)]C(221)]C(222) Cu]P(2)]C(231)]C(232) PF6 2.287(2) 2.269(1) 2.049(4) 2.039(4) 127.73(5) 100.1(2) 101.5(1) 101.7(1) 110.4(1) 111.7(1) 167.8(4) 166.4(4) 259.6(4) 2 241.7(5) 2 216.9(5) 2 55.4(4) 48.5(4) 35.4(5) BF4 * 2.276(4) 2.258(4) 2.032(9) 2.023(9) 127.1(1) 99.5(4) 101.1(3) 102.3(3) 111.0(3) 111.9(3) 165.3(9) 168.0(10) 61.9(9) 2 41(1) 2 16(1) 2 55.7(8) 48.2(6) 32(1) ClO4 2.284(1) 2.265(1) 2.053(3) 2.037(3) 126.82(4) 100.3(1) 100.87(8) 102.34(9) 111.42(8) 111.60(9) 166.2(3) 165.7(3) 62.2(3) 40.9(3) 17.2(3) 56.3(3) 49.7(3) 32.3(3) * Ref. 12.J. Chem. Soc., Dalton Trans., 1998, Pages 2321–2325 2323 eclipsed conformation with the same chirality for both ligands.7 For each compound, Cu]P(1) is found to be consistently ca. 0.02 Å longer than Cu]P(2) in accord with the diVerences in steric crowding observed about the Cu]P(1) and Cu]P(2) bonds. Comparison of the geometry of the P2Cu molecular core with those of other four-co-ordinate Cu(PPh3)2X compounds (Table 2 1,6,7,20–23) shows that the Cu]P distances of 2.258(4)–2.287(2) Å lie at the upper end of the range and similar to values found for other Cu(PPh3)2N2 compounds (Table 2) while the P]Cu]P angles of 126.82(4)–127.73(5)8 are signifi- cantly greater than these, reflecting the smaller steric profile of the acetonitrile ligands.Despite the angular distortion of the P2CuN2 core, the Cu]N bond lengths of 2.023(9)–2.053(3) Å are short and indicative of relatively strong co-ordination of the acetonitrile ligands to the copper.Solid state CP MAS 31P NMR spectroscopy Under high field (Bo > ª5 T) and slow 63,65Cu relaxation conditions, solid state CP MAS 31P NMR spectroscopy of copper(I) phosphine complexes yields signals for each crystallographically independent phosphorus atom which are split into pairs of closely spaced asymmetric quartets in an intensity ratio of ª2 : 1 as a result of scalar spin–spin coupling and quadrupolar interactions between the phosphorus and the spin ��� 63Cu (g = 7.0965 × 107 rad T21 s21, natural abundance 69.09%) and 65Cu (g = 7.6018 × 107 rad T21 s21, natural abundance 30.91%) isotopes.The linewidths of the peaks are such that the two quartets are not generally resolved and observed spectra are dominated by the 63Cu spectrum, with the 65Cu spectrum identified (when the linewidth is suYciently narrow) as a splitting of the outer peaks of the quartet.14–18 Spectra recorded for the present series of complexes at 9.40 T are reproduced in Fig. 3 and consist of overlapping quartets of doublets representing the AB portion of a total ABX spin system arising from spin– spin coupling between the 31P and 63Cu nuclei and smaller geminal 2J(31P]31P) coupling, with satellite signals from the 65Cu spectrum also observable.The overlap of the peaks in each quartet precludes accurate measurement of the chemical shift and line spacing data from these spectra, but these may be extracted from the two-dimensional 31P CP MAS COSY spectra which separate the two 1J(P]Cu) coupled quartets above and below the main diagonal (Fig. 4). These parameters are in Table 3. The average chemical shift for each P site is located at d 13–14 and 7–8 [d(PPh3) 0] and assigned to P(1) and P(2) on the basis of the values of the 1J(31P]63Cu) coupling constants 1.14(1) kHz for the signals centred at d 13–14 and 1.30(1) kHz for the signals centred at d 7–8 corresponding to the longer and shorter Cu]P bond lengths respectively.These coupling constants are comparable to values of 1.08(1) and 1.25(2) kHz for [{Cu(PPh3)2(4-Mepc)}2] (pc = phenylcyanamide),1,16 1.15 kHz for [Cu(PPh3)2(BH4)] 3 and 1.17 kHz for [Cu(PPh3)2(S4CPh)],24 but smaller than the 1.40 kHz found for [Cu(PPh3)2(NO3)] 3,25,26 reflecting relatively weaker covalent bonding character for the copper–oxygen bonds in this compound. Table 2 P2Cu Geometries (distances in Å, angles in 8) for selected [Cu(PPh3)2X2] species Compound [Cu(PPh3)2(O2CCF3)] [Cu(PPh3)2(NO3)] [Cu(PPh3)2(O2CPh)] [Cu(PPh3)2(S2CPh)] [Cu(PPh3)2(bpy)]ClO4 [Cu(PPh3)2(BH4)] [{Cu(PPh3)2(4-Me-pc)}2] [Cu(PPh3)2(py)2]NO3 [Cu(PPh3)2(phen)]NO3 Cu]P(1) 2.235(2) 2.249(1) 2.238(1) 2.267(2) 2.256(3) 2.276(1) 2.284(1) 2.295(3) 2.271(1) Cu]P(2) 2.228(2) 2.249(1) 2.228(1) 2.257(2) 2.246(3) 2.276(1) 2.257(1) 2.271(4) 2.245(10) P]Cu]P 136.7(1) 130.87(7) 126.78(4) 125.6(1) 125.4(1) 123.26(6) 121.7(1) 115.85(9) 115.44(4) Ref. 677 20 21 22 1 21 23 py = Pyridine, bpy = 2,29-bipyridyl, phen = 1,10-phenanthroline.Geminal 2J(31P]31P) coupling constants for [Cu(PPh3)2X(X9)] compounds have been previously determined only for [{Cu(PPh3)2(4-Mepc)}2] 16 (122 Hz), [Cu(PPh3)2(BH4)] 26 (140 Hz) and [Cu(PPh3)2(NO3)] 26 (157 Hz). The value of 75 Hz recorded for each of the present compounds is, therefore, by far the smallest value yet measured. The reasons for this large decrease in magnitude are not readily apparent. Studies on the eVect of change in the donor ligand X on 2J(31P]31P) for a series of mercury complexes [Hg(PPh3)2X2] complexes 27,28 show a correlation between the co-ordinating ability of X and decreasing 2J(31P]31P) which would suggest an expectation of comparable values of 2J(31P]31P) for the acetonitrile, phenylcyanamide and tetrahydroborate complexes. A characteristic feature of the quartets observed in the spectra of all copper(I) phosphine complexes under high field conditions (Bo > ª5 T) is that the chemical shifts d1, d2, d3 and d4 of the peaks of the J quartet show diVerential chemical shifts as a result of perturbation of the spectra by copper quadrupolar interactions.Analysis of this interaction as a first order perturbation of the J spectrum 17 predicts the outer two lines Fig. 3 One-dimensional 31P CP MAS NMR spectra of [Cu(PPh3)2- (MeCN)2]X at 9.40 T for X = PF6 (a), BF4 (b) and ClO4 (c). Resolution of the outer components of the 65Cu multiplet is apparent fom. Soc., Dalton Trans., 1998, Pages 2321–2325 (d1, d4) to shift to more negative ppm (upfield) and the inner two lines (d2, d3) to more positive ppm (downfield) by a parameter d such that (D32 2 D21) = (D43 2 D32) = 2d, from which d = (D43 2 D21)/4; d is found to be inversely dependent on the field strength and multiplication by the 63Cu Zeeman frequency, nCu, yields a Fig. 4 Two-dimensional 31P CP MAS COSY NMR spectra of [Cu- (PPh3)2(MeCN)2]X at 9.40 T for X = PF6 (a), BF4 (b) and ClO4 (c) field independent parameter d nCu that is suitable for comparison of results obtained from spectra recorded at diVerent field strengths.The first order theory allows d nCu to be calculated according to expression (1) where cCu = e2qQ/h is the 63Cu quadrupolar d nCu = (3cCuDeff/20)(3 cos2 bD 2 1 1 h sin2 bD cos 2aD) (1) coupling constant, Deff = (D 2 DJ/3) [where D = (mo/4p)gPgCuh/ 4p2r3 is the Cu]P dipolar coupling constant and DJ is the ansiotropy in the J tensor], h is the asymmetry parameter of the electric field gradient (EFG) tensor, and aD, bD are the polar angles defining the direction of the Cu]P internuclear vector with respect to the principal axial system (PAS) of the EFG tensor.In practice, the observed spectra are found to deviate from this first order model with (D32 2 D21) > 2d > (D43 2 D32). This result has been ascribed to higher order quadrupole eVects with the sense of the distortion indicative of a positive sign for J.14,18 These latter eVects, however, are generally quite small and are recorded for the present compounds as an upfield shift of peaks 1 and 3 and a downfield shift of peaks 2 and 4 by a correction term of magnitude d1 such that D21 = J 2 2d 2 2d1, D32 = J 1 2d1 and D43 = J 1 2d 2 2d1, from which d = (D43 2 D21)/4 as before, d1 = (D32 2 J)/2 while J = (D21 1 2D32 1 D43)/4. For the present complexes, the measured values of d nCu for each phosphorus site diVer by ca. 50%, ranging from (2.2– 2.5) × 109 Hz2 for P(1) and (3.3–4.0) × 109 Hz2 for P(2). As the contribution of the copper quadrupolar coupling constant, cCu, to the determination of the value of d nCu is the same for both sites, these diVerent values must arise from diVerences in other factors such as Deff, the orientation of the Cu]P vectors with respect to the PAS of the EFG tensor and/or higher order eVects. While the relative contributions of these eVects await measurement of further data from (for example) single crystal NMR studies, it is interesting that the observed diVerences can be accounted for by a decrease in the angle between Cu]P(2) and the z axis of the EFG tensor [and an increase in the angle between Cu]P(1) and the same z axis] by ca. 58, corresponding to a shift of the z axis towards the bond of greater electron density. Comparison of these values of d nCu with those calculated from published NMR data for other four-co-ordinate [Cu- (PPh3)2X(X9)] compounds shows this parameter to increase from (1.5–2.0) × 109 Hz2 for [{Cu(PPh3)2(4-Mepc)}2] 16 to (2.2– 4.0) × 109 Hz2 for the present series to 4.5 × 109 Hz2 for [Cu(PPh3)2(S4CPh)],24 6.0 × 109 Hz2 for [(Ph3P)2CuBr2Cu- (PPh3)],29 7.0 × 109 Hz2 for [Cu(PPh3)2(BH4)] 3 and 10.0 × 109 Hz2 for [Cu(PPh3)2(NO3)].3 Rearrangement of equation (1) to the expression (2) where G = (3 cos2 bD 2 1 1 h sin2 bD cos 2aD) cCu/d nCu = (20G/3Deff) (2) shows the ratio between cCu and d nCu to be dependent on a number of unknown factors including the magnitude of DJ and h, and the orientation of the EFG tensor in the molecular frame and, as well as higher order quadrupolar eVects, precluding direct estimation of cCu from the measured values of d nCu.However, for the present series of complexes, the 63Cu nuclear quadrupolar resonance frequency for tetrahedral copper in [(Ph3P)2CuBr2Cu(PPh3)] has been found to be 14.31 MHz,30 yielding cCu of the order of 30 MHz and cCu/d nCu ª 5.Using this ratio, and assuming an uncertainty of the order of 20%, yields estimates of cCu of the order of 5–10 MHz for [{Cu(PPh3)2(4-Mepc)}2], 10–20 MHz for the present series of complexes, 20–30 MHz for [Cu(PPh3)2(S4CPh)], 30–40 MHz for [Cu(PPh3)2(BH4)] and 45–55 MHz for [Cu(PPh3)2(NO3)]. While these estimates are very approximate in absolute terms, the trend that is observed is likely to be maintained as more accurate data become available from, for example, NQR or solid state 63,65Cu NMR experiments, and supports the notion that evenJ.Chem. Soc., Dalton Trans., 1998, Pages 2321–2325 2325 Table 3 31P CP MAS NMR data for [Cu(PPh3)2(MeCN)2]X, X = PF6, BF4 or ClO4 X = PF6 BF4 ClO4 Cu]P/Å ·dÒ D21/Hz D32/Hz D43/Hz 1J(31P]63Cu)/kHz P(1) 2.287 13.4 1080 1136 1174 1.13 P(2) 2.269 7.8 1230 1305 1362 1.30 P(1) 2.276 14.2 1094 1149 1176 1.14 P(2) 2.258 7.0 1217 1313 1368 1.30 P(1) 2.284 14.0 1080 1136 1163 1.13 P(2) 2.265 6.9 1217 1327 1341 1.30 2J(31P]31P)/Hz 75 75 75 d/Hz d1/Hz d nCu/109 Hz2 23 2.5 2.5 33 2.5 3.5 20 3.5 2.2 38 5.0 4.0 21 3.5 2.2 31 12.0 3.3 ·dÒ = Average chemical shift of the four lines of each quartet with respect to PPh3; 1J(31P]63Cu) = (D21 1 2D32 1 D43)/4; d = (D43 2 D21)/4, d1 = (D32 2 1J)/2; nCu (9.40 T) = 106.1 MHz.Estimated errors in Dij = 20 Hz, d 5 Hz, d1 = 5 Hz and d nCu = 1 × 109 Hz2. quite small changes in the nature of the ligands or anions coordinated to [Cu(PPh3)2]1 can have a significant influence on the distribution of charge about the copper site.For the present acetonitrile complexes, the results suggest this distribution to be relatively symmetric, despite the low geometric symmetry of the copper co-ordination sphere. Finally, we note that in terms of the catalytic properties of these compounds 12 the results of this study are of interest in that they illustrate, first, the conformational flexibility of the [Cu(PPh3)2]1 cation in interacting with co-ordinating bases and, secondly, the capacity of the cation to redistribute electron density about the copper site in response to changing donor properties of these bases. Acknowledgements J.V. H. thanks the Cooperative Research Centre for Molecular Engineering and the Australian National Nanofabrication Facility for funding of the NMR Facility. References 1 E. W. Ainscough, E. N. Baker, M. L. Brader, A. M. Brodie, S. L. Ingham, J.M. Waters, J. V. Hanna and P. C. Healy, J. Chem. Soc., Dalton Trans., 1991, 1243. 2 A. Avdeef and J. P. Fackler, J. Coord. Chem., 1975, 4, 211. 3 G. A. Bowmaker, J. C. Dyason, P. C. Healy, L. M. Engelhardt, C. Pakawatchai and A. H. White, J. Chem. Soc., Dalton Trans., 1987, 1089. 4 M. A. Cabras, L. Naldini, M. A. Zoroddu, F. Cariati, F. Demartin, N. Masciocchi and M. Sansoni, Inorg. Chim. Acta, 1985, 104, L19. 5 D. J. Darensbourg, M. W. Holtcamp, B. Khandelwal and J. H. Reibenspies, Inorg.Chem., 1995, 34, 5390. 6 R. D. Hart, P. C. Healy, G. A. Hope, D. W. Turner and A. H. White, J. Chem. Soc., Dalton Trans., 1994, 773. 7 R. D. Hart, P. C. Healy, M. L. Peake and A. H. White, Aust. J. Chem., 1998, 51, 67. 8 S. J. Lippard and G. J. Palenik, Inorg. Chem., 1971, 10, 1322. 9 I. G. Dance, M. L. Scudder and L. J. Fitzpatrick, Inorg. Chem., 1985, 24, 2547. 10 P. F. Barron, J. C. Dyason, L. M. Engelhardt, P. C. Healy and A. H. White, Aust. J. Chem., 1985, 38, 261. 11 J. Diez, S. Falagan, P. Gamasa and J. Gimeno, Polyhedron, 1988, 7, 37. 12 J. Green, E. Sinn and S. Woodward, Polyhedron, 1993, 12, 991. 13 A. M. Leiva, L. Rivera and B. Loeb, Polyhedron, 1991, 10, 347. 14 E. M. Menger and W. S. Veeman, J. Magn. Reson., 1982, 46, 257. 15 R. K. Harris and A. C. Olivieri, Prog. Nucl. Magn. Reson. Spectrosc., 1992, 24, 435. 16 J. V. Hanna, M. E. Smith, S. N. Stuart and P. C. Healy, J. Phys. Chem., 1992, 96, 7560. 17 A. Olivieri, J. Am. Chem. Soc., 1992, 114, 5758. 18 S. H. Alarcon, A. C. Olivieri and R. K. Harris, Solid State Nucl. Magn. Reson., 1993, 2, 325. 19 S. R. Hall, H. D. Flack and J. M. Stewart, The Xtal 3.2 Reference Manual, Universities of Western Australia, Geneva and Maryland, 1992. 20 C. Bianchini, C. A. Ghilardi, D. Masi, A. Meli and A. Orlandini, Cryst. Struct. Commun., 1982, 11, 1495. 21 L. M. Engelhardt, C. Pakawatchai, A. H. White and P. C. Healy, J. Chem. Soc., Dalton Trans., 1985, 125. 22 S. J. Lippard and K.M. Melmed, J. Am. Chem. Soc., 1967, 89, 3929. 23 J. R. KirchhoV, D. R. McMillin, W. R. Robinson, D. R. Powell, A. T. McKenzie and S. Chen, Inorg. Chem., 1985, 24, 3928. 24 F. Asaro, A. Camus, R. Gobetto, A. C. Olivieri and G. Pellizer, Solid State NMR, 1997, 8, 81. 25 J. W. Diesveld, E. M. Menger, H. T. Edzes and W. S. Veeman, J. Am. Chem. Soc., 1980, 102, 7935. 26 G. Wu and R. E. Wasylishen, Inorg. Chem., 1996, 35, 3113. 27 D. Dakternieks, Inorg. Chim. Acta, 1984, 89, 209. 28 T. Allman and R. E. Lenkinski, Inorg. Chem., 1986, 25, 3202. 29 G. A. Bowmaker, B. W. Skelton, A. H. White and P. C. Healy, J. Chem. Soc., Dalton Trans., 1988, 2825. 30 H. Negita, M. Hiura, K. Yamada and T. Okuda, J. Mol. Struct., 1980, 58, 205. Received 27th March 1998; Paper 8/02383BJ. Chem. Soc., Dalton Trans., 1998, Pages 2321–2325 2325 Table 3 31P CP MAS NMR data for [Cu(PPh3)2(MeCN)2]X, X = PF6, BF4 or ClO4 X = PF6 BF4 ClO4 Cu]P/Å ·dÒ D21/Hz D32/Hz D43/Hz 1J(31P]63Cu)/kHz P(1) 2.287 13.4 1080 1136 1174 1.13 P(2) 2.269 7.8 1230 1305 1362 1.30 P(1) 2.276 14.2 1094 1149 1176 1.14 P(2) 2.258 7.0 1217 1313 1368 1.30 P(1) 2.284 14.0 1080 1136 1163 1.13 P(2) 2.265 6.9 1217 1327 1341 1.30 2J(31P]31P)/Hz 75 75 75 d/Hz d1/Hz d nCu/109 Hz2 23 2.5 2.5 33 2.5 3.5 20 3.5 2.2 38 5.0 4.0 21 3.5 2.2 31 12.0 3.3 ·dÒ = Average chemical shift of the four lines of each quartet with respect to PPh3; 1J(31P]63Cu) = (D21 1 2D32 1 D43)/4; d = (D43 2 D21)/4, d1 = (D32 2 1J)/2; nCu (9.40 T) = 106.1 MHz.Estimated errors in Dij = 20 Hz, d 5 Hz, d1 = 5 Hz and d nCu = 1 × 109 Hz2. quite small changes in the nature of the ligands or anions coordinated to [Cu(PPh3)2]1 can have a significant influence on the distribution of charge about the copper site. For the present acetonitrile complexes, the results suggest this distribution to be relatively symmetric, despite the low geometric symmetry of the copper co-ordination sphere. Finally, we note that in terms of the catalytic properties of these compounds 12 the results of this study are of interest in that they illustrate, first, the conformational flexibility of the [Cu(PPh3)2]1 cation in interacting with co-ordinating bases and, secondly, the capacity of the cation to redistribute electron density about the copper site in response to changing donor properties of these bases.Acknowledgements J. V. H. thanks the Cooperative Research Centre for Molecular Engineering and the Australian National Nanofabrication Facility for funding of the NMR Facility.References 1 E. W. Ainscough, E. N. Baker, M. L. Brader, A. M. Brodie, S. L. Ingham, J. M. Waters, J. V. Hanna and P. C. Healy, J. Chem. Soc., Dalton Trans., 1991, 1243. 2 A. Avdeef and J. P. Fackler, J. Coord. Chem., 1975, 4, 211. 3 G. A. Bowmaker, J. C. Dyason, P. C. Healy, L. M. Engelhardt, C. Pakawatchai and A. H. White, J. Chem. Soc., Dalton Trans., 1987, 1089. 4 M. A. Cabras, L.Naldini, M. A. Zoroddu, F. Cariati, F. Demartin, N. Masciocchi and M. Sansoni, Inorg. Chim. Acta, 1985, 104, L19. 5 D. J. Darensbourg, M. W. Holtcamp, B. Khandelwal and J. H. Reibenspies, Inorg. Chem., 1995, 34, 5390. 6 R. D. Hart, P. C. Healy, G. A. Hope, D. W. Turner and A. H. White, J. Chem. Soc., Dalton Trans., 1994, 773. 7 R. D. Hart, P. C. Healy, M. L. Peake and A. H. White, Aust. J. Chem., 1998, 51, 67. 8 S. J. Lippard and G. J. Palenik, Inorg. Chem., 1971, 10, 1322. 9 I. G. Dance, M. L. Scudder and L. J. Fitzpatrick, Inorg. Chem., 1985, 24, 2547. 10 P. F. Barron, J. C. Dyason, L. M. Engelhardt, P. C. Healy and A. H. White, Aust. J. Chem., 1985, 38, 261. 11 J. Diez, S. Falagan, P. Gamasa and J. Gimeno, Polyhedron, 1988, 7, 37. 12 J. Green, E. Sinn and S. Woodward, Polyhedron, 1993, 12, 991. 13 A. M. Leiva, L. Rivera and B. Loeb, Polyhedron, 1991, 10, 347. 14 E. M. Menger and W. S. Veeman, J. Magn. Reson., 1982, 46, 257. 15 R. K. Harris and A. C. Olivieri, Prog. Nucl. Magn. Reson. Spectrosc., 1992, 24, 435. 16 J. V. Hanna, M. E. Smith, S. N. Stuart and P. C. Healy, J. Phys. Chem., 1992, 96, 7560. 17 A. Olivieri, J. Am. Chem. Soc., 1992, 114, 5758. 18 S. H. Alarcon, A. C. Olivieri and R. K. Harris, Solid State Nucl. Magn. Reson., 1993, 2, 325. 19 S. R. Hall, H. D. Flack and J. M. Stewart, The Xtal 3.2 Reference Manual, Universities of Western Australia, Geneva and Maryland, 1992. 20 C. Bianchini, C. A. Ghilardi, D. Masi, A. Meli and A. Orlandini, Cryst. Struct. Commun., 1982, 11, 1495. 21 L. M. Engelhardt, C. Pakawatchai, A. H. White and P. C. Healy, J. Chem. Soc., Dalton Trans., 1985, 125. 22 S. J. Lippard and K. M. Melmed, J. Am. Chem. Soc., 1967, 89, 3929. 23 J. R. KirchhoV, D. R. McMillin, W. R. Robinson, D. R. Powell, A. T. McKenzie and S. Chen, Inorg. Chem., 1985, 24, 3928. 24 F. Asaro, A. Camus, R. Gobetto, A. C. Olivieri and G. Pellizer, Solid State NMR, 1997, 8, 81. 25 J. W. Diesveld, E. M. Menger, H. T. Edzes and W. S. Veeman, J. Am. Chem. Soc., 1980, 102, 7935. 26 G. Wu and R. E. Wasylishen, Inorg. Chem., 1996, 35, 3113. 27 D. Dakternieks, Inorg. Chim. Acta, 1984, 89, 209. 28 T. Allman and R. E. Lenkinski, Inorg. Chem., 1986, 25, 3202. 29 G. A. Bowmaker, B. W. Skelton, A. H. White and P. C. Healy, J. Chem. Soc., Dalton Trans., 1988, 2825. 30 H. Negita, M. Hiura, K. Yamada and T. Okuda, J. Mol. Struct., 1980, 58, 205. Received 27th March 1998; Paper 8/02383B

ISSN:1477-9226    

DOI:10.1039/a802383b

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2323–2328 2323 Nickel complexes with tris(2-aminoethyl)amine (tren): [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O (ox 5 oxalate), {[Ni2(tren)3][ClO4]4?H2O}n, and [Ni2(tren)2(aepd)][ClO4]4?2H2O (aepd 5 N-(2-aminoethyl)pyrrolidine-3,4-diamine). Synthesis, structure and magnetism† Vanessa M. Masters,a Paul V. Bernhardt,a Lawrence R. Gahan,*a Boujemaa Moubaraki,b Keith S. Murray b and Kevin J. Berry c a Department of Chemistry, The University of Queensland, Brisbane, QLD 4072, Australia b Chemistry Department, Monash University, Clayton, Victoria 3168, Australia c Westernport Secondary College, Hastings, Victoria 3915, Australia Received 2nd March 1999, Accepted 21st May 1999 Reaction of K3[Cr(ox)3] (ox = oxalate) with nickel(II) and tris(2-aminoethyl)amine (tren) in aqueous solution resulted in isolation of the bimetallic assembly [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O.The polymeric complex {[Ni2(tren)3]- [ClO4]4?H2O}n has been prepared by reaction of nickel(II) perchlorate and tren in aqueous solution.From the same reaction mixture the complex [Ni2(tren)2(aepd)][ClO4]4?2H2O (aepd = N-(2-aminoethyl)pyrrolidine-3,4-diamine), in which a bridging tren ligand contains a carbon–carbon bond between two arms forming a substituted pyrrolidine, has been isolated. The complexes have been characterized by X-ray crystallography. The magnetic susceptibility (300– 4.2 K) and magnetization data (2, 4 K, H = 0–5 T) for {[Ni2(tren)3][ClO4]4?H2O}n (300 K , 4.23 mB) exhibit evidence of weak antiferromagnetic coupling and zero field splitting (2J = 21.8 cm21; |D| = 2 cm21) at low temperature. For [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O the susceptibility data at 300 K are indicative of uncoupled nickel(II) and chromium(III) sites with zero-field splitting and intramolecular antiferromagnetic coupling predicted at low temperature. Introduction Magnetic interactions between complex cations and anions, propagated by intermolecular eVects such as those occurring though hydrogen bonded water molecules, might be expected to be weak.1 For example, the complex [Cu(tpm)2]3[Cr(ox)3]2? 20H2O (S1 = ��� , S2 = ��� ; tpm = tris(pyrazolyl)methane; ox = oxalate) shows weak ferromagnetic coupling below 20 K.1 However, the potential for variation in magnetic characteristics of both the complex cation and the complex anion makes the resulting possibility of ferro- or ferri-magnetic interactions attractive,2 and the search for examples of bimetallic complexes of some interest.1–5 We have chosen to explore the synthesis of complex bimetallic salts of chromium(III) and nickel(II).Bimetallic complexes of this type have been reported, e.g. [Cr(NH3)6][Ni(H2O)6]Cl5? ��� NH4Cl,6 the interest being in the combination of S = ��� and 1 spin states in octahedral fields.7–10 In the present case [Cr(ox)3]32 is employed as the complex anion and the complex cation results from the interaction of nickel(II) with tris(2-aminoethyl)- amine (tren).In addition, two complexes of tren with nickel(II) have been identified, the first the polymeric species {[Ni2(tren)3]- [ClO4]4?H2O}n and the second [Ni2(tren)2(aepd)][ClO4]4?2H2O, in which a bridging tren ligand contains a carbon–carbon bond between two arms forming the substituted pyrrolidine (aepd = N-(2-aminoethyl)pyrrolidine-3,4-diamine). Experimental Materials Tris(2-aminoethyl)amine (tren), classified as 96%, was obtained † Non-SI unit employed: mB ª 9.27 × 10224 J T21.from Aldrich Chemical Company and either used without further purification or first distilled and fractionated. Samples of tren from two separate bottles were employed. For GC-MS a Hewlett Packard 5890 gas chromatograph, 5970 Series mass selective detector and a 25 m BPS column (0.25 mm i.d., 0.25 mm phase thickness) were employed. The temperature was raised from 100 to 270 8C at a rate of 168 min21; the oven was held at 100 8C for 2 min, raised at 168 min21 to 270 8C, then held at that temperature for 10 min.Helium was used as carrier at a head pressure of 20 psi. 13C-{1H} NMR spectra were recorded with a JEOL GX400 spectrometer. Syntheses CAUTION: perchlorate salts of metal complexes are potentially explosive and should be handled in small quantities. [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O. The salt Ni(NO3)2?6H2O (1.0 g; 3.44 mmol) was dissolved in a small volume of ethanol (10 cm3) and tren (3 cm3) was added dropwise until a lilac precipitate appeared.The mixture was filtered, and the product washed with ethanol (20 cm3) diethyl ether (20 cm3) and dried in air. The resulting purple solid was dissolved in water (20 cm3) and an aqueous solution (5 cm3) of K3[Cr(ox)3]?3H2O (0.56 g; 1.15 mmol) added. Ethanol (5 cm3) was added slowly until the solution became cloudy, at which time water (1 cm3) was added until the solution clarified. Within several hours small deep purple crystals suitable for X-ray crystallographic studies had formed (20 mg) (Found: C, 27.1; H, 5.84; N, 14.2.[Ni3(tren)4- (H2O)2][Cr(ox)3]2?6H2O requires C, 28.1; H, 5.78 ; N, 14.6%). {[Ni2(tren)3][ClO4]4?H2O}n. The salt Ni(ClO4)2?6H2O (1.0 g; 2.73 mmol) dissolved in a small volume of ethanol (20 cm3) and tren (3 cm3; 0.02 mol) added dropwise. A lilac precipitate2324 J. Chem. Soc., Dalton Trans., 1999, 2323–2328 appeared immediately. The solution was filtered, and the product washed with ethanol and dried in air (1.60 g).The precipitate was dissolved in water (100 cm3) and the solution left to stand on the bench. Lilac crystals, suitable for X-ray crystallographic studies, grew over a period of a week (0.5 g) (Found: C, 21.5; H, 5.9; N, 16.8. {[Ni2(tren)3][ClO4]4?H2O}n requires C, 22.2; H, 5.8; N, 17.3%). Small violet crystals which were separated manually were also observed (Found: C, 21.4; H, 5.7; N, 16.6. [Ni2(tren)2(aepd)][ClO4]4?2H2O requires C, 21.9; H, 5.7; N, 17.0%).The same products were obtained using tren which had been distilled and fractionated. Crystallography Data collection, structure solution and refinement. Cell constants were determined by a least-squares fit to the setting parameters of 25 independent reflections measured on an Enraf-Nonius CAD4 four circle diVractometer employing graphite monochromated Mo-Ka radiation (l 0.71073 Å) and operating in the w–2q scan mode.Data reduction and empirical absorption corrections (y scans) were performed with the XTAL11 package. Structures were solved by heavy atom methods with SHELXS 8612 and refined by full-matrix leastsquares analysis on F2 with SHELXL 93.13 All non-H atoms were refined with anisotropic thermal parameters except those mentioned below. Crystal data appear in Table 1. The atomic nomenclature is defined in Figs. 1–4 drawn with PLATON.14 Abnormal features. {[Ni2(tren)3][ClO4]4?H2O }n.This polymeric structure exhibited disorder in the positions both the bridging and terminally bound tren ligands. The refined model comprised independent polymeric chains with occupancies restrained to 50%. The co-ordination sphere of Ni(1) comprises two monodentate bridging tren ligands, N(2nA)/C(2nA) and N(2nB)/C(2nB), in addition to a tetradentate terminally bound tren, N(1n)/C(1n). All of these ligands are disordered about the mirror plane upon which Ni(1) is situated.Centrosymmetrically related pairs of tridentate tren ligands are bound to Ni(2) in an alternating array along the chain, i.e. N(2nA)/C(2nA):N(2nA)/ C(2nA) then N(2nB)/C(2nB) :N(2nB)/C(2nB). The C and N atoms in these bridging ligands were refined with isotropic thermal parameters. Positional disorder in all ClO4 2 anions necessitated refinement of all O atoms with isotropic thermal parameters. [Ni2(tren)2(aepd)][ClO4]4?2H2O. The bridging rac-aepd ligand was found to exhibit disorder between its enantiomeric RR and SS isomers.The occupancies of the resulting diastereomeric complexes shown in Fig. 4 were 60 and 40% respectively. Perchlorate disorder was also identified in this structure and the O atoms connected to Cl(3) were refined with isotropic thermal parameters. [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O. Two of the waters of crystallisation (O(15) ardered over three sites of occupancies of 40 : 40 :20%. CCDC reference number 186/1473. See http://www.rsc.org/suppdata/dt/1999/2323/ for crystallographic files in .cif format.Magnetic studies Magnetic susceptibility measurements were made using a Quantum Design MPMS SQUID magnetometer with an applied field of 1 T. The powdered sample was contained in a calibrated gelatine capsule which was held in the centre of a drinking straw fixed to the end of the sample rod. For high-field magnetization work (0 to 5 T) the samples were also contained in the gelatine capsule.The magnetization values of the instrument were calibrated against a standard palladium sample, supplied by Quantum design, and also against chemical calibrants such as CuSO4?5H2O and [Ni(en)3][S2O3] (en = ethylenediamine). EVective magnetic moments, per mol, were calculated using the relationship meff = 2.828(cmT)� �� where cm is the susceptibility per mol of complex. Results and discussion Crystal structures Deep purple crystals of the complex double salt [Ni3(tren)4- (H2O)2][Cr(ox)3]2?6H2O crystallized from aqueous solution after an excess of tren was treated with a nickel(II) salt and K3[Cr(ox)3] added.The structure consists of a complex cation ([Ni3(tren)4(H2O)2]61), two complex anions ([Cr(ox)3]32), and water molecules. The structure of the complex cation is depicted in Fig. 1 whilst Fig. 2 shows the packing diagram. Selected bond distances and angles are listed in Table 2. The three nickel(II) ions in the [Ni3(tren)4(H2O)2]61 cation do not have the same co-ordination sphere. The central nickel(II) ion is situated at a centre of symmetry and is bonded to two facially co-ordinated tridentate tren ligands.Each tren ligand bears a pendant primary amine which is co-ordinated to a terminal nickel(II) ion. The terminal enantiomeric pair of octahedral nickel(II) ions are each bound to a single tetradentate tren ligand, a pendant amine from the central Ni(tren)2 21 unit, and an aqua ligand. Each of the nickel(II) ions exhibits a distorted octahedral geometry, with the angle between two cis nitrogen atoms (N–Ni–N) in the range 80.1–99.98.The Ni–N bond lengths are in the range 2.052(9)–2.111(9) Å for the terminal nickel and 2.099(7)–2.152(8) Å for the central nickel. The Ni–N bond lengths for the terminal nickel ions in [Ni3- (tren)4(H2O)2]61 were slightly diVerent to the range observed for [Ni(tren)Cl(H2O)]Cl?H2O (2.047–2.172 Å).15 The structure of the [Cr(ox)3]32 anion is essentially the same as seen previously Fig. 1 A PLATON plot of the complex cation [Ni3(tren)4(H2O)2]61 with relevant atoms labelled; 30% probability ellipsoids are shown. Fig. 2 Packing diagram for [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O.J. Chem. Soc., Dalton Trans., 1999, 2323–2328 2325 Table 1 Crystal data for [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O, {[Ni2(tren)3][ClO4]4?H2O}n and [Ni2(tren)2(aepd)][ClO4]4?2H2O [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O [Ni2(tren)2(aepd)][ClO4]4?2H2O {[Ni2(tren)3][ClO4]4?H2O}n Empirical formula Formula weight T/K Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 V/Å3 Z Dc/g cm3 m/mm21 F(000) Reflections collected Independent reflections Data/restraints/parameters Goodness of fit on F2 Final R1, wR2 [I > 2s(I)] (all data) Residual extrema/e Å23 C36H88Cr2N16Ni3O32 1537.35 293(2) Monoclinic P21/n 9.0430(10) 24.151(3) 15.318(2) 96.619(7) 3323.1(7) 2 1.536 1.245 1608 6220 5829 [R(int) = 0.0868] 5829/0/411 1.005 0.0775, 0.2013 0.2008, 0.2746 0.952, 20.549 C18H56Cl4N12Ni2O18 987.97 293(2) Triclinic P1� 9.968(2) 15.480(1) 15.684(2) 118.552(7) 104.25(1) 97.47(1) 1971.5(5) 2 1.664 1.31 1032 7353 6916 [R(int) = 0.0262] 6916/0/504 1.02 0.0470, 0.1283 0.0838, 0.1503 1.463, 20.577 C18H56Cl4N12Ni2O17 971.97 293(2) Monoclinic P21/m 8.784(3) 23.414(2) 10.310(3) 108.06(1) 2016.0(9) 2 1.601 1.278 1016 3893 3647 [R(int) = 0.0505] 3647/0/311 1.048 0.0902, 0.2435 0.1990, 0.3210 1.683, 20.595 for K3[Cr(ox)3]?3H2O.16 The bimetallic complex has a number of waters of crystallization, two of which are disordered over three sites.The most notable hydrogen bonding interaction is that between a chelating oxygen of [Cr(ox)3]32 and a water molecule co-ordinated to nickel(II) (O(4) ? ? ? H(13B)–O(13) 2.179 Å, 174.88). The distances Cr ? ? ? Ni(2), Cr ? ? ? Ni(1), and Ni(1) ? ? ? Ni(2) are 7.618, 14.252 and 7.431 Å, respectively. The complex {[Ni2(tren)3][ClO4]4?H2O}n consists of a polymeric cation, perchlorate anions and a water molecule.The structure of the cation [Ni2(tren)3]41 is depicted in Fig. 3, and selected bond distances and angles are listed in Table 3. There are two distinct geometries in which the nickel(II) ions in {[Ni2(tren)3][ClO4]4?H2O}n exist. In the first instance, two tren molecules act as tridentate ligands and co-ordinate to a single nickel ion, with each tren ligand having a pendant arm ending with a primary amine. In the second case, a single tren molecule co-ordinates to a single nickel(II) ion in a tetradentate fashion Table 2 Selected bond distances (Å) and bond angles (8) for [Ni3- (tren)4(H2O)2][Cr(ox)3]2?6H2O Cr(1)–O(5) Cr(1)–O(6) Cr(1)–O(3) C(1)–C(2) C(5)–C(6) Ni(1)–N(11) Ni(2)–N(23) Ni(2)–N(24) Ni(2)–O(13) O(5)–Cr(1)–O(1) O(1)–Cr(1)–O(6) O(1)–Cr(1)–O(2) O(5)–Cr(1)–O(3) O(6)–Cr(1)–O(3) O(5)–Cr(1)–O(4) O(6)–Cr(1)–O(4) O(3)–Cr(1)–O(4) N(12)–Ni(1)–N(13) N(14)–Ni(2)–N(23) N(23)–Ni(2)–N(24) N(23)–Ni(2)–N(21) N(24)–Ni(2)–N(21) N(23)–Ni(2)–N(22) N(21)–Ni(2)–N(22) N(23)–Ni(2)–O(13) N(21)–Ni(2)–O(13) 1.959(7) 1.971(7) 1.975(7) 1.57(2) 1.55(2) 2.139(8) 2.088(9) 2.089(11) 2.229(9) 93.5(3) 175.1(3) 82.4(3) 91.5(3) 93.2(3) 172.9(3) 93.1(3) 83.2(3) 82.8(3) 176.6(4) 84.3(4) 82.7(4) 95.3(4) 82.5(3) 160.9(4) 95.8(4) 84.8(4) Cr(1)–O(1) Cr(1)–O(2) Cr(1)–O(4) C(3)–C(4) Ni(1)–N(12) Ni(1)–N(13) Ni(2)–N(22) Ni(2)–N(21) Ni(2)–N(14) O(5)–Cr(1)–O(6) O(5)–Cr(1)–O(2) O(6)–Cr(1)–O(2) O(1)–Cr(1)–O(3) O(2)–Cr(1)–O(3) O(1)–Cr(1)–O(4) O(2)–Cr(1)–O(4) N(12)–Ni(1)–N(11) N(11)–Ni(1)–N(13) N(14)–Ni(2)–N(24) N(14)–Ni(2)–N(21) N(14)–Ni(2)–N(22) N(24)–Ni(2)–N(22) N(14)–Ni(2)–O(13) N(24)–Ni(2)–O(13) N(22)–Ni(2)–O(13) 1.966(7) 1.972(8) 1.980(7) 1.55(2) 2.099(7) 2.152(8) 2.111(9) 2.100(10) 2.052(9) 82.3(3) 93.7(3) 95.3(3) 89.5(3) 170.6(3) 91.2(3) 92.2(3) 92.8(3) 80.1(3) 93.4(4) 95.1(4) 100.2(4) 95.2(4) 86.4(4) 179.8(3) 84.7(4) with the two remaining co-ordination sites occupied by the pendant primary amines from tren ligands bound tridentate to neighbouring nickel(II) ions.The [Ni(tren)2]21 fragment is located on a centre of symmetry, but disordered over two positions. The [Ni(tren)(RNH2)2]21 moiety is situated on a mirror plane that passes through the atoms Ni(1), N(13), C(14) and C(15). However, this plane does not coincide with any local symmetry element of the complex, so all atoms in this fragment not lying on this mirror plane are disordered. These two disordered components comprise a pair of [Ni(tren)(RNH2)2]21 enantiomers. The centre of symmetry at Ni(2) results in an alternating array of the two enantiomeric [Ni(tren)(RNH2)2]21 Fig. 3 A PLATON plot of the complex cation [Ni2(tren)3]41 with relevant atoms labelled; 30% probability ellipsoids are shown. Table 3 Selected bond lengths (Å) and angles (8) for {[Ni2(tren)3]- [ClO4]4?H2O}n Ni(1)–N(11) Ni(1)–N(12) Ni(1)–N(14) Ni(2)–N(23A) Ni(2)–N(21B) Ni(2)–N(21A) N(11)–Ni(1)–N(23B3) N(23B2)–Ni(1)–N(13) N(13)–Ni(1)–N(12) N(11)–Ni(1)–N(14) N(23B3)–Ni(1)–N(14) N(12)–Ni(1)–N(14) N(22B)–Ni(2)–N(21B) N(21B)–Ni(2)–N(24B) N(24A3)&nd)–N(21A) 2.08(2) 2.17(2) 2.22(3) 2.13(2) 2.14(2) 2.16(2) 99.7(10) 83.0(7) 161.2(5) 83.0(11) 177.2(11) 90.1(9) 84.3(7) 81.9(7) 97.8(7) Ni(1)–N(13) Ni(1)–N(22A) Ni(2)–N(24A) Ni(2)–N(22B) Ni(2)–N(24B) N(11)–Ni(1)–N(13) N(11)–Ni(1)–N(12) N(13)–Ni(1)–N(22A) N(23B2)–Ni(1)–N(14) N(13)–Ni(1)–N(14) N(24A)–Ni(2)–N(23A) N(22B)–Ni(2)–N(24B) N(24A)–Ni(2)–N(21A) N(23A)–Ni(2)–N(21A) 2.11(2) 2.17(2) 2.09(2) 2.14(2) 2.14(2) 80.5(7) 83.7(9) 102.2(8) 82.4(10) 98.1(8) 92.3(8) 93.6(9) 82.2(7) 81.9(7) Symmetry relations: 2 2x 1 1, y 2 1– 2, 2z 1 1. 3 2x 1 1, 2y 1 1, 2z 1 1.2326 J. Chem. Soc., Dalton Trans., 1999, 2323–2328 fragments along the polymer bridged by centrosymmetric [Ni(tren)2]21 units. When disorder at both nickel(II) centres is considered, the polymeric structure may be separated into two similar, achiral strands, which are ‘out of phase’ with respect to the position of the enantiomeric [Ni(tren)(RNH2)2]21 moieties. Despite the disorder, the structures of {[Ni2(tren)3][ClO4]4? H2O}n and [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O are actually quite similar.The cationic portion of the latter structure may be thought of as being derived from {[Ni2(tren)3][ClO4]4?H2O}n by terminating the polymeric chain with aqua ligands. Refinement in the more common acentric space group P21 was tried but this led to a worse R parameter and unreasonable interatomic distances and angles.All of the nickel ions are in a slightly distorted octahedral geometry with the angles between two cis N atoms (N–Ni–N) in the range 80.5(7)–98.1(7)8. The Ni–N distances are in the range 2.08–2.22(2) Å and the Ni ? ? ? Ni separation is 7.5 Å. Complexes of the type [Ni2(tren)3 ]41 or [Ni(tren)2 ]21 have been reported,17–19 and the [Ni(tren)2]21 unit has been structurally characterized in the complexes [Ni(tren)2][BF4]2 and [Ni(tren)2][Ni(mnt)2] (mnt = maleonitriledithiolate dianion).20 In both [Ni(tren)2][Ni(mnt)2] and [Ni(tren)2][BF4]2 the tridentate tren ligands are facially co-ordinated with pendant primary amines and the geometries of the cations are nearly identical with Ni–N bond lengths in the range 2.129(2)– 2.172(2) Å,20 although the bond lengths for the Ni(tren)2 21 centre of Ni3(tren)4 61 are shorter than seen in [Ni(tren)2][BF4]2.The structure of [Ni2(tren)2(aepd)][ClO4]4?2H2O consists of the complex cation, four perchlorate anions and two water molecules.Fig. 4 shows a PLATON drawing of the cation which has two tren ligands bound in a tetradentate manner capping each metal ion. The two metal centres are joined by a bridging ligand which is an analogue of tren but has a carbon–carbon bond (1.50(2) Å, Table 4) joining two arms of the tren ligand, resulting in the substituted pyrrolidine, aepd. The geometries about each metal ion are that of a distorted octahedron with N–Ni–N angles in the range 81.2(3)–99.6(2)8.The octahedral environment about one nickel(II) site is composed of four nitrogen donors from a tren ligand and two primary amine donors in a five membered chelate ring formed by the pyrrolidine. The second nickel(II) site is again composed of a tetradentate tren ligand, the remaining two co-ordination sites occupied by a tertiary nitrogen donor from the pyrrolidine and the pendant primary amine from the 2-aminoethyl moiety.The Ni–N bond lengths associated with the tetradentate tren ligands lie within the range 2.087(7)–2.164(4) Å, a larger range than reported for other nickel(II) complexes of tren.15,20–23 For the pyrrolidine ligand the Ni–N (primary amine) distances are Fig. 4 A PLATON plot of the complex cation [Ni2(tren)2(aepd)]41 (lel, ob) with relevant atoms labelled; 30% probability ellipsoids are shown. significantly diVerent (2.202(4) and 2.132(4) Å) with the Ni–N (tertiary) distance (2.165(4) Å) longer than that for the analogous bond in the capping tren ligands (2.087(4) and 2.103(4) Å).The complex crystallizes in diasteroisomeric forms, differentiated by the chirality around carbon atoms C(33) and C(34). The diasteroisomer can be diVerentiated by the conformations of C(33B)–C(34B)/C(33C)–C(34C) relative to C(13A)-C(13B) (lel 60%: ob 40%,24 respectively). The distance between the two nickel ions is 6.79 Å. A 13C-{1H} NMR spectrum of a concentrated sample of the commercial tren, recorded in CDCl3, revealed two major resonances at d 56.32 and 38.45 assigned to the methylene carbon atoms adjacent to the tertiary nitrogen and the primary amine of tren, respectively. Sets of lower intensity (<3%) resonances at d 38.96, 57.87, 59.46 and 60.80 were also evident, with indications of even weaker resonances at d 48.4, 51.52, 58.02, and 60.22.The DEPT spectra indicated that for the weaker signals the resonances at d 38.96, 57.87 and 60.80 arose from methylene carbon atoms whilst that at d 59.46 arose from a methine carbon suggesting that compounds such as aepd were possible contaminants.The GC-MS analysis of samples of tren also indicated the presence of a number of components, although they were poorly separated under the column conditions. The mass spectra of the two most significant components of the mixture corresponded to tren (M1, m/z 147, calc. 146.24; m/z 129, 116, 112, 99, 87, 73 and 70) and a minor fraction with M1 at m/z 145 which exhibited a mass spectral breakdown pattern identical to that of tren, but displaced by two mass units suggestive of, but not unambiguously assigned to, aepd (calc.for M1, m/z 144.23; m/z 127, 114, 97, 85, 71 and 68). The chromatographic analysis indicated that the compounds with M1 at m/z 145 and 147 had approximately the same boiling point, and subsequent distillation and fractionation of tren failed to remove the m/z 145 peak in the GC-MS.It would seem therefore that the origin of the aepd is from the commercial sample of tren, which is catalogued as of 96% purity. Magnetic susceptibility The magnetic properties of powder samples of {[Ni2(tren)3]- [ClO4]4?H2O}n and [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O are reported in the form of eVective magnetic moment per mol versus T plots measured in a field of 1 T (Figs. 5 and 6, respectively). For {[Ni2(tren)3][ClO4]4?H2O}n at 300 K the moment (4.23 mB) is essentially that expected (4.00–4.60 mB) for two uncoupled nickel(II) ions (S = 1, g = 2.0–2.3).25 The meff values decrease a little in the region 300 (meff = 4.23 mB, cmT = 2.23 cm3 K mol21) to 30 K (meff = 4.03 mB, cmT = 2.03 cm3 K mol21) and then drop more rapidly to meff = 3.19 mB (1.27 cm3 K mol21) at 4.5 K.The decrease in meff below 30 K is indicative of zero field splitting (D) and/or weak antiferromagnetic coupling (J) between the two nickel(II) (S = 1) centres. The small but gradual decrease in meff above 30 K is due to weak antiferromagnetic coupling and second order Zeeman (Na) eVects.Two approaches were used in order to explain the magnetic data. In the first the data were fitted using the expression for a classical-spin Heisenberg chain (S = 1),26–28 based on a Table 4 Selected bond lengths (Å) and angles (8) for [Ni2(tren)2- (aepd)][ClO4]4?2H2O Ni(1)–N(21) Ni(1)–N(23) Ni(1)–N(31) Ni(2)–N(11) Ni(2)–N(14) Ni(2)–N(12) C(33B)–C(34B) N(21)–Ni(1)–N(31) N(11)–Ni(2)–N(34) 2.103(4) 2.144(4) 2.165(4) 2.087(4) 2.113(4) 2.147(4) 1.50(2) 177.3(2) 177.3(2) Ni(1)–N(22) Ni(1)–N(24) Ni(1)–N(32) Ni(2)–N(13) Ni(2)–N(34) Ni(2)–N(33) N(32)–Ni(1)–N(31) N(34)–Ni(2)–N(33) 2.118(4) 2.164(4) 2.143(4) 2.096(4) 2.132(4) 2.202(4) 82.7(2) 83.0(2)J.Chem. Soc., Dalton Trans., 1999, 2323–2328 2327 2JSi?Si 11 Hamiltonian (1) where x = |J|/kT and the term Na cm = S Ng2b2 k(T 2 q)DF 2.0 1 0.0194x 1 0.777x2 3.0 1 4.346x 1 3.232x2 1 5.834x3G1 Na (1) accounting for temperature-independent paramagnetism, resulting in J = 21.2 cm21, q (Weiss constant) = 21.2 K, g = 2.28, Na = 5.0 × 1024 cm3 mol21 and R = 6.1 × 1025 (the function minimized in curve fitting was R = S (cm obs 2 cm calc)2/S (cm obs)2).Calculations to incorporate the local anisotropy are neglected in this approach.28 The g and Na values are correlated to some extent and are a bit higher than normal. Incorporation of the small negative Weiss constant might imply the existence of very weak chain–chain interactions.The second approach was an attempt to assess the contribution of both J and D to the low temperature susceptibility. Thus the polymer was considered in terms of the dimeric [Ni2(tren)3]41 moiety. The Hamiltonian (2) was employed.28 The H� = 22JS� 1?S� 2 1 D(S� 1 2 1 S� 2) 1 gbH?S (2) energies of the ground state manifold for a dinuclear nickel(II) moiety were obtained by diagonalization of the appropriate 9 × 9 matrix under the above Hamiltonian operator.We have assumed that the g and D values of the two metal ions are the same even though the symmetries are diVerent (Fig. 3). Further, the g values were assumed to be isotropic. The temperature dependence of the magnetic susceptibilities was calculated using the thermodynamic expression for the susceptibility rather than the Van Vleck equation.28–30 This is particularly important for the low temperature magnetisation data, described below.A temperature independent susceptibility term (Na) was also included; 28 J, D, and g were permitted to vary for best fit and Na was set at 2.5 × 1024 cm3 mol21, which is the Fig. 5 Plot of meff (per mol) vs. temperature for {[Ni2(tren)3][ClO4]4? H2O}n measured in a field of 1 T. The solid line represents the best fit to the experimental data (2J = 21.8 cm21, |D| = 2 cm21, g = 2.065, Na = 2.5 × 1024 cm3 mol21); j from temperature vs. meff data, e from magnetization data. Fig. 6 Plot of meff (per mol) vs. temperature for [Ni3(tren)4(H2O)2]- [Cr(ox)3]2?6H2O measured in a field of 1 T. normal value for two mols of NiII.27 A very good fit, shown in Fig. 5, was obtained for the following parameter set: 2g = 2.06, 2J = 21.8 cm21 and D = 2.0 cm21. This set of parameter values was strongly supported when they were used to generate plots of magnetization versus applied field, in fields of 0 to 5 T. In Fig. 7 it can be seen that the 4 K data are very well simulated, as are the 2 K data in terms of the gentle S-shape, but with small discrepancies at intermediate field values.If J or D are set to zero there are large diVerences between observed and calculated magnetization plots. The g value obtained by matrix diagonalization is slightly less than normal values for NiII (2–2.1) while that obtained from the chain model (2.28) is high because of correlation eVects and limitations of this chain model. Whichever of the two approaches is employed suggests that the two nickel(II) sites are weakly antiferromagnetically coupled through the diamagnetic atoms joining them (Ni ? ? ? Ni 7.5 Å).There are no significant intermolecular Ni ? ? ? Ni pathways evident from the crystal structure and so the q value obtained from the chain model, if real, may relate to dipolar eVects. There are a number of reported instances of polynuclear nickel(II) complexes in which weak antiferromagnetic coupling occurs through diamagnetic bridges and these can be compared to the present system.26,31–33 For [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O, the cmT values (per mol) in the region 300 (m = 7.70 mB, cmT = 7.41 cm3 K mol21) to 30 K (m = 7.46 mB, cmT = 6.96 cm3 K mol21) decrease a little in a linear fashion (Fig. 6). At 300 K for uncoupled nickel(II) and chromium(III) ions {3meff 2 [Ni3(tren)4(H2O)2]61 1 2meff 2 [Cr(ox)3]32} the theoretical eVective magnetic moment is 7.35 mB (S = 1, ��� , g = 2).34 The value of cmT decreases rapidly below 30 K (m = 6.83 mB, cmT= 5.84 cm3 K mol21, 4.5 K).The general slope of the curve is similar to that in Fig. 5 for the [Ni2(tren)3]41 complex. The magnetic behaviour is also similar to that reported for the double salt [Cu(tpm)2]3[Fe(ox)3]2?20H2O with meff (300 K) 9.85 mB with Curie like behaviour between 300 and 4 K and weak antiferromagnetic coupling and/or zero field splitting at low temperature.1 Contrastingly [Cu(tpm)2]3[Cr(ox)3]2? 20H2O exhibits Curie like behaviour between 300 and 20 K and weak ferromagnetic coupling below 20 K, the CrIII ? ? ? CrIII interactions being mediated by hydrogen bonding of water to neighbouring oxalate groups.The intermolecular contacts include an interaction between a chelating oxygen of an oxalate ligand and a water molecule (O(1) ? ? ? O(w) 2.90(1) Å),1 similar to that observed for [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O (O(4) ? ? ? H(13B)–O(13) 2.179 Å, 174.88). Thus the weak antiferromagnetic behaviour of [Ni3(tren)4(H2O)2][Cr(ox)3]2?6H2O could arise either from intramolecular coupling within the Ni2(tren)3 moiety and zero Ni ? ? ? Cr coupling, or intermolecular interaction between [Cr(ox)3]32 species as discussed above.Zero-field splitting from both the cation and, to a lesser extent, the anionic centres will contribute to the decrease occur- Fig. 7 Plots of magnetization M, vs field H for {[Ni2(tren)3][ClO4]4? H2O}n. Solid lines are calculated using 2J = 21.8 cm21, |D| = 2 cm21 and g = 2.065 (see text).2328 J.Chem. Soc., Dalton Trans., 1999, 2323–2328 ring below 30 K. In view of these ambiguities quantitative analysis was not pursued. Concluding remarks The isolation of the polymeric{[Ni2(tren)3][ClO4]4?H2O}n complex and the bimetallic species [Ni3(tren)4(H2O)2][Cr(ox)3]2? 6H2O although unanticipated is, in retrospect, not surprising. Complexes of the type [Ni2(tren)3 ]41 or [Ni(tren)2 ]21 have been reported previously, and indeed the results of a potentiometric titration of tren in the presence of nickel(II) can be fitted using a model which includes species such as [Ni(tren)2]21 and [Ni(tren)(Htren)]31.35 Further, the weak magnetic interactions observed appear typical for these types of complexes. However, the isolation of [Ni2(tren)2(aepd)][ClO4]4?2H2O was remarkable given the previous extensive chemistry of tren and nickel(II)– tren systems. It is possible that the ligand has always been present in reaction mixtures, albeit in small amounts, but it presumably cannot compete with the strongly co-ordinating tetradentate tren.In the present case, an apparently ideal combination of tetradentate co-ordinating tren ligands coupled with the bridging bis-didentate aepd, which cannot bind to the same metal ion through more than two nitrogen atoms because of stereochemical constraints enforced by the pyrrolidine ring, results in the observed dinuclear nickel(II) hexaamine.Acknowledgements The financial support of the Australian Research Council is gratefully acknowledged. References 1 K. S. Murray, G. D. Fallon, D. C. R. Hockless, K. D. Lu, B. Moubaraki and K. Van Langenberg, ACS Symp. Ser., 1996, 644, 201. 2 M. C. Morón, F. Palacio, J. Pons, J. Casabó, X. Solans, K. E. Merabet, D. Huang, X. Shi, B. K. Teo and R. L. Carlin, Inorg. Chem., 1994, 33, 746. 3 R. L. Carlin, Comments Inorg. Chem., 1991, 11, 215. 4 S. Decurtins, H. W. Schmalle, P.Schneuwly and H. R. Oswald, Inorg. Chem., 1993, 32, 1888. 5 V. A. Grillo, L. R. Gahan, G. R. Hanson and T. W. Hambley, Polyhedron, 1996, 15, 559. 6 M. C. Morón, A. Le Bail and J. Pons, J. Solid State Chem., 1990, 88, 498. 7 Y. Pei, Y. Journaux and O. Kahn, Inorg. Chem., 1989, 28, 100. 8 T. Mallah, C. Auberger, M. Verdaguer and P. Veillet, J. Chem. Soc., Chem. Commun., 1995, 61. 9 V. Gadet, T. Mallah, I. Castro and M. Verdaguer, J. Am. Chem. Soc., 1992, 114, 9213. 10 A.P. Ginsberg, Inorg. Chim. Acta Rev., 1971, 5, 45. 11 S. R. Hall, H. D. Flack and J. M. Stewart (Editors), The XTAL3.2 User’s Manual, Universities of Western Australia, Geneva and Maryland, 1992. 12 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467. 13 G. M. Sheldrick, SHELXL 93, A program for crystal structure determination, University of Göttingen, 1993. 14 A. L. Spek, PLATON, A Thermal ellipsoid plotting program, University of Utrecht, 1994. 15 A. Marzotto, D. A. Clemen Acta Crystallogr., Sect. C, 1993, 49, 1252. 16 D. Taylor, Aust. J. Chem., 1978, 31, 1455. 17 F. G. Mann, J. Chem. Soc., 1929, 409. 18 F. G. Mann and W. J. Pope, J. Chem. Soc., 1926, 482. 19 C. K. Jørgensen, Acta Chem. Scand., 1956, 10, 887. 20 G. J. Colpas, M. Kumar, R. O. Day and M. J. Maroney, Inorg. Chem., 1990, 29, 4779. 21 M. Salah El Fallah, E. Rentschler, A. Caneschi, R. Sessoli and D. Gatteschi, Angew. Chem., Int. Ed. Engl., 1996, 35, 1947. 22 R. D. Willett, Acta Crystallogr., Sect. C, 1987, 43, 1494. 23 M. L. Calatayud, I. Castro, J. Sletten, J. Cano, F. Lloret, J. Faus, M. Julve, G. Seitz and K. Mann, Inorg. Chem., 1996, 35, 2858. 24 The lel and ob nomenclature is described in Inorg. Chem., 1970, 9, 1. 25 M. Ohba, N. Fukita and H. O—kawa, J. Chem. Soc., Dalton Trans., 1997, 1733. 26 H. L. Shyu, H. H. Wei and Y. Wang, Inorg. Chim. Acta, 1997, 258, 81. 27 A. Meyer, A. Gleizes, J.-J. Girerd, M. Verdaguer and O. Kahn, Inorg. Chem., 1982, 21, 1729. 28 O. Kahn, Molecular Magnetism, VCH, New York, 1993, p. 257. 29 K. S. Murray, Adv. Inorg. Chem., 1995, 43, 261. 30 J. Glerup, P. A. Goodson, D. J. Hodgson and K. Michelson, Inorg. Chem., 1995, 34, 6255. 31 Y. Journaux, J. Sletten and O. Kahn, Inorg. Chem., 1986, 25, 439. 32 P. Chaudhuri, M. Winter, B. P. C. Della Vdova, E. Bill, A. Trautwein, S. Gehring, P. Fleischhauer, B. Nuber and J. Weiss, Inorg. Chem., 1991, 30, 2148. 33 R. Ruiz, M. Julve, J. Faus, F. Lloret, M. C. Muñoz, Y. Journaux and C. Bois, Inorg. Chem., 1997, 36, 3434. 34 L. N. Mulay, Magnetic Susceptibility, Interscience Publishers, New York, 1963, p. 1751. 35 L. R. Gahan and V. M. Masters, unpublished work. Paper 9/01659G

ISSN:1477-9226    

DOI:10.1039/a901659g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 2327 Thioalcohols as bridging ligands in polynuclear PrIII/CuII and BaII/CuII complexes. Syntheses, structures and magnetic properties of Pr2Cu4(tde)3(Htde)2(hfacac)4(Ï6-O), Ba2Cu2(Htde)2(hfacac)4 and Cu4(tde)2(hfacac)4, (H2tde 5 HOCH2CH2SCH2CH2OH)‡ Steven R. Breeze,a Suning Wang,*,†,a John E. Greedan b and N. P. Raju b a Department of Chemistry, Queen’s University, Kingston, Ontario, K7L 3N6, Canada b Department of Chemistry, McMaster University, Hamilton, Ontario, L8S 4M1, Canada The syntheses of heterometallic compounds involving praseodymium, barium and copper ions by using 2,29-thiodiethanol (H2tde) as the bridging ligand have been investigated.Three new polynuclear compounds, Pr2Cu4(tde)3(Htde)2(hfacac)4(m6-O) 1, Ba2Cu2(Htde)2(hfacac)4 2 and Cu4(tde)2(hfacac)4 3 have been obtained and characterized structurally. The metal atoms in 1 have an octahedral arrangement with the two PrIII occupying two opposite vertices.In the center of the octahedron is a m6-oxygen atom. The metal atoms in 2 have a rhombic arrangement with a long Cu ? ? ? Cu separation distance (6.082 Å). The metal ions in 3 also have a rhombic arrangement with the copper atoms in close proximity to each other. The sulfur atoms of the Htde2 or tde22 ligand are bound exclusively to copper centers in all three compounds. The Htde2 and tde22 ligands display versatile bonding modes in the three compounds.Antiferromagnetic exchanges dominate in these compounds. Heterometallic polynuclear lanthanide–copper and alkaline earth metal–copper compounds have attracted much research attention recently because of their potential applications in copper oxide based superconductors and molecular magnetism.1–3 Our earlier research eVorts focused on the syntheses of heterometallic Ln]Cu and MII]Cu (MII = alkaline earth metal) compounds by using bifunctional ligands such as aminoalcohols.A variety of polynuclear Ln]Cu and MII]Cu compounds were obtained successfully by employing the aminoalcohol ligands.3 Further to explore the chemistry of polynuclear Ln]Cu and MII]Cu compounds, we investigated the utility of thio-containing alcohol ligands in the synthesis of mixed metal Ln]Cu and MII]Cu compounds. The Ln]Cu compounds with thioalcohol ligands are not useful as precursors for superconductors due to the problem of sulfide contamination. They may, however, find use in the field of molecular magnetism.One advantage provided by thioalcohol ligands over aminoalcohol is that, as a soft donor, the sulfur atom has a high aYnity for the copper center, making them better bifunctional ligands than aminoalcohols for selective binding to lanthanide or alkaline earth metal and copper centers. The thioalcohol ligand chosen for our study is 2,29-thiodiethanol (H2tde) for its simplicity. This ligand has two acidic protons which can be removed readily to form either the monoanion, Htde2 or the dianion, tde22. The alkoxo oxygen atom of Htde2 or tde22 was anticipated to bridge a hard metal ion such as lanthanide(III) or barium(II) and the copper(II) ion in the same manner as those in aminoalcohol ligands.Indeed, we have succeeded in synthesizing several interesting polynuclear complexes by using the H2tde ligand. We report herein the syntheses, structures and magnetic properties of two new heterometallic compounds, a m6-oxygen bridged Pr2Cu4 compound and a rhombic Ba2Cu2 compound and a tetranuclear Cu4 compound as well, obtained by using the H2tde ligand.S OH HO H2tde † E-Mail: wangs@chem.queensu.ca ‡ Non-SI units employed: G = 1024 T, mB ª 9.27 × 10224 J T21. Experimental All reactions and manipulations were performed under an atmosphere of nitrogen using either a Vacuum Atmospheres glove-box or standard Schlenk-line techniques. Solvents were reagent grade and distilled from appropriate drying agents under nitrogen prior to use.Copper(II) methoxide and 2,29- thiodiethanol were purchased from the Aldrich Chemical Company and used as received. Praseodymium, barium and yttrium hexafluoroacetylacetonates were purchased from Strem Chemicals. The IR spectra were recorded on a Bomen FTIR spectrometer. The KBr used in the pellets for the IR studies was dried in an oven for several hours prior to use. Elemental analyses were performed by Canadian Microanalytical Service, Delta, British Columbia. Variable temperature magnetic susceptibility data were collected on a SQUID magnetometer at 1 kG.All molar susceptibility data were corrected for diamagnetism by using Pascal constants.4 Syntheses Pr2Cu4(tde)3(Htde)2(hfacac)4(Ï6-O) 1. The compound Cu(OCH3)2 (100 mg, 0.80 mmol) was treated with H2tde (122 mg, 1.00 mmol) in CH2Cl2 (3 mL) at 25 8C, which yielded an insoluble purple solid. After allowing the mixture to stir for approximately 20 h, Pr(hfacac)3 (Hhfacac = 1,1,1,5,5,5- hexafluoroacetylacetonate) (303 mg, 0.40 mmol), water (3.6 mg, 0.20 mmol) and CH2Cl2 (20 mL) were added.The solution became dark green and was stirred for 4 h before being filtered to remove a trace amount of precipitate. The filtrate was concentrated in vacuo to approximately 5 mL and hexane (2 mL) was added to crystallize the product. The crystals were dark green (153 mg, 0.079 mmol, 40%), m.p. 215–216 8C (Found: C, 24.17; H, 2.36.Calc. for C40H46Cu4F24O19Pr2S5: C, 24.23; H, 2.34%). IR (KBr, cm21): 1660s, 1528m, 1259s, 1201s, 1142vs and 1073m. Magnetic moment at 300 K: 2.65 mB, Pascal constant = 8.10 × 1024 cm3 mol21. Ba2Cu2(Htde)4(hfacac)4 2. The compound Cu(OCH3)2 (50 mg, 0.40 mmol) was treated with 2 equivalents of H2tde (97 mg, 0.80 mmol) in CH2Cl2 (3 mL) at 25 8C. After allowing the mixture to stir for approximately 20 h, Ba(hfacac)2 (220 mg,2328 J. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 Table 1 Crystallographic data for compounds 1–3 Formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z m/cm21 Reflections measured Reflections used (Rint) Final R1, wR2 [I > 2s(I)] (all data) 1 C40H46Cu4F24O19Pr2S5 1983.04 Triclinic P1� 12.081(2) 14.493(3) 20.640(4) 72.69(3) 75.30(3) 76.72(3) 3290.6(11) 2 30.1 12 546 12 439 (0.001) 0.0537, 0.1553 0.0732, 0.1726 2 C36H40Ba2Cu2F24O16S4 1714.68 Monoclinic P21/n 12.7094(2) 9.2893(2) 25.6763(4) 100.391(1) 2981.66(9) 2 22.8 22 885 5879 (0.045) 0.0460, 0.1223 0.0626, 0.1391 3 C28H20Cu4F24O12S2 1322.72 Triclinic P1� 10.693(2) 10.8750(13) 11.130(2) 66.214(13) 70.73(2) 72.356(9) 1096.0(3) 1 21.6 2943 2762 (0.026) 0.0567, 0.1265 0.0899, 0.1479 R1 = (S|Fo| 2 |Fc|)/S|Fo|, wR2 = [Sw(Fo 2 2 Fc 2)2/Sw(Fo 2)2]� �� , w = 1/[s2(Fo 2) 1 (0.075P)2], where P = [max (Fo 2, 0) 1 2Fc 2]/3. 0.40 mmol) and CH2Cl2 (20 mL) were added. The solution was stirred for 4 h before being filtered to remove a blue precipitate.The filtrate was concentrated in vacuo to approximately 5 mL and hexane (2 mL) was added to crystallize the product. The crystals of compound 2 were light blue (87 mg, 0.051 mmol, 25.3%), m.p. 162–165 8C (Found: C, 25.22; H, 2.35. Calc. for C18H20BaCuF12O8S2: C, 25.22; H, 2.46%). IR (KBr, cm21): 1667s, 1537s, 1262s, 1203s, 1131vs and 1071m. Magnetic moment at 300 K: 1.50 mB, Pascal constant = 7.13 × 1024 cm3 mol21. Cu4(tde)2(hfacac)4 3. The compound Cu(OCH3)2 (100 mg, 0.80 mmol) was treated with H2tde (122 mg, 1.00 mmol) in CH2Cl2 (3 mL) at 25 8C.After allowing the mixture to stir for approximately 20 h Y(hfacac)3 (283 mg, 0.40 mmol), water (3.6 mg, 0.20 mmol) and CH2Cl2 (20 mL) were added. The solution was stirred for 4 h before being filtered to remove a trace amount of precipitate. The filtrate was concentrated in vacuo to approximately 5 mL and hexane (2 mL) was added to crystallize the product. The crystal were dark green (187 mg, 0.141 mmol, 71%), m.p. 185–187 8C (Found: C, 25.47; H, 1.55. Calc. for C14H10Cu2F12O6S: C, 25.46; H, 1.53%). IR (KBr, cm21): 1648s, 1561w, 1468m, 1262s, 1211s and 1148vs. After the green crystals were collected, blue crystals were isolated from the remaining solution (50 mg) (Found: C, 22.93; H, 1.28%)position matches well with the formula of Cu2Y2(tde)2(hfacac)3- (O2CCF3)2(OH) (C, 22.99; H, 1.41%). IR (KBr, cm21): 1663s, 1529w, 1503m, 1254s, 1142vs and 1077m. Magnetic moment at 300 K: 3.47 mB, Pascal constant = 5.30 × 1024 cm3 mol21.Single-crystal X-ray diVraction analysis X-Ray quality crystals were obtained from concentrated hexane–CH2Cl2 solutions at 0 8C. For compounds 1 and 2 the data were collected on a Siemens CCD X-ray diVractometer. For compound 3 a suitable crystal was mounted on a glass fiber with epoxy resin and the data were collected on a conventional Siemens P4 diVractometer with a Mo-Ka radiation source operating at 60 kV and 40 mA.Data were collected at 23 8C over the range 2 < 2q < 538 for 1, 3.2 < 2q < 548 for 2 and 4 < 2q < 458 for 3. Three standard reflections were measured every 197. The data for all the compounds were processed on a Pentium personal computer using the Siemens SHELXTL crystallographic package (version 5).5 Data were corrected for Lorentz-polarization eVects and empirical absorption corrections were applied in all cases. No significant decay was observed. Neutral atomic scattering factors were taken from the literature.6 Some of the fluorine atoms in compounds 1–3 displayed a two-fold rotational disorder, which was successfully modeled and refined with 50% occupancy factors for each site.All of the non-hydrogen atoms were refined anisotropically except the disordered fluorine atoms in 1. The positions of the hydrogen atoms bound to carbon atoms were calculated and included in the structure factor calculations. The crystal data are given in Table 1.CCDC reference number 186/1009. Results and Discussion Synthesis and structure of Pr2Cu4(tde)3(Htde)2(hfacac)4(Ï6-O) 1 The reaction between copper methoxide and H2tde resulted in a sparingly soluble purple compound. Following the addition of Pr(hfacac)3 the mixture immediately changed to dark green. When this solution was concentrated and a small amount of hexane was added dark green crystals were obtained. Attempts to determine the structure of the compound by single-crystal X-ray diVraction experiments on a conventional diVractometer were unsuccessful due to the insuYcient number of data.The dark green color of the compound indicated the presence of copper(II). The IR spectrum of 1 showed that the hexafluoroacetylacetonato ligand is present. The structure and composition of 1 was ultimately established by a single-crystal X-ray diVraction experiment performed on a CCD diVractometer. The oxo ligand was believed to come from the trace amount of water present in the reaction medium.Indeed, the yield of compound 1 was improved when a stoichiometric amount of water was added to the reaction mixture. An ORTEP7 drawing of compound 1 with the fluorine atoms omitted for clarity is shown in Fig. 1. Selected bond lengths and angles are provided in Table 2. Compound 1 consists of four copper atoms and two praseodymium atoms in an octahedral arrangement, with the two praseodymium atoms occupying opposite vertices of the octahedron.The PrIII and the CuII are linked together by an oxo ligand and eight oxygen atoms from the deprotonated 2,29-thiodiethanol ligands. The Pr]Cu distances range from 3.285(1) [Pr(2)]Cu(1)] to 3.408(1) Å [Pr(2)]Cu(2)], comparable to those found in previously reported Ln]Cu complexes where the LnIII and CuII are linked together by oxygen atoms from either hydroxypyridine or aminoalcohol ligands.2,3 Several Ln]Cu compounds containing an octahedral Ln2Cu4 unit have been reported previously where hydroxypyridine ligands are used as the bridging ligands.2c,3d Compound 1 is, however, the first example of octahedral Ln]Cu complexes using thioalcohol ligands and the first example of Ln2Cu4 complexes containing a m6-oxo ligand [O(19)].The m6 bridging mode of the oxo ligand, albeitJ. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 2329 Table 2 Selected bond lengths (Å) and angles (8) Compound 1 Pr(1)]O(11) Pr(1)]O(14) Pr(1)]O(16) Pr(1)]O(9) Pr(1)]O(4) Pr(1)]O(1) Pr(1)]O(2) Pr(1)]O(3) Pr(1)]O(19) Pr(1)]Cu(2) Pr(1)]Cu(4) Pr(1)]Cu(1) O(12)]Pr(2)]O(18) O(12)]Pr(2)]O(13) O(18)]Pr(2)]O(13) O(12)]Pr(2)]O(7) O(18)]Pr(2)]O(7) O(13)]Pr(2)]O(7) O(12)]Pr(2)]O(5) O(18)]Pr(2)]O(5) O(13)]Pr(2)]O(5) O(7)]Pr(2)]O(5) O(12)]Pr(2)]O(15) O(18)]Pr(2)]O(15) O(13)]Pr(2)]O(15) O(7)]Pr(2)]O(15) O(5)]Pr(2)]O(15) O(12)]Pr(2)]O(8) O(18)]Pr(2)]O(8) O(13)]Pr(2)]O(8) O(7)]Pr(2)]O(8) O(5)]Pr(2)]O(8) O(15)]Pr(2)]O(8) O(12)]Pr(2)]O(6) O(18)]Pr(2)]O(6) O(13)]Pr(2)]O(6) O(7)]Pr(2)]O(6) O(5)]Pr(2)]O(6) O(15)]Pr(2)]O(6) O(8)]Pr(2)]O(6) O(12)]Pr(2)]O(19) O(18)]Pr(2)]O(19) 2.394(5) 2.438(5) 2.453(5) 2.464(5) 2.492(6) 2.521(6) 2.525(6) 2.548(6) 2.654(4) 3.352(2) 3.370(2) 3.3740(14) 127.4(2) 80.2(2) 77.4(2) 145.8(2) 79.1(2) 86.5(2) 79.7(2) 141.6(2) 139.4(2) 90.8(2) 74.0(2) 80.9(2) 124.7(2) 137.6(2) 82.3(2) 78.2(2) 135.7(2) 72.3(2) 67.7(2) 69.3(2) 143.3(2) 133.8(2) 74.7(2) 145.2(2) 68.3(2) 67.2(2) 70.5(2) 115.9(2) 64.66(14) 62.7(2) Pr(2)]Cu(4) Cu(1)]O(15) Cu(1)]O(12) Cu(1)]O(16) Cu(1)]O(19) Cu(1)]S(4) Cu(2)]O(9) Cu(2)]O(15) Cu(2)]O(19) Cu(2)]S(1) Cu(2)]Cu(1) Cu(2)]Cu(3) O(14)]Pr(2)]O(16) O(11)]Pr(1)]O(9) O(14)]Pr(1)]O(9) O(16)]Pr(1)]O(9) O(11)]Pr(1)]O(4) O(14)]Pr(1)]O(4) O(16)]Pr(1)]O(4) O(9)]Pr(1)]O(4) O(11)]Pr(1)]O(1) O(14)]Pr(1)]O(1) O(16)]Pr(1)]O(1) O(9)]Pr(1)]O(1) O(4)]Pr(1)]O(1) O(11)]Pr(1)]O(2) O(14)]Pr(1)]O(2) O(16)]Pr(1)]O(2) O(9)]Pr(1)]O(2) O(4)]Pr(1)]O(2) O(1)]Pr(1)]O(2) O(11)]Pr(1)]O(3) O(14)]Pr(1)]O(3) O(16)]Pr(1)]O(3) O(9)]Pr(1)]O(3) O(4)]Pr(1)]O(3) O(1)]Pr(1)]O(3) O(2)]Pr(1)]O(3) O(11)]Pr(1)]O(19) O(14)]Pr(1)]O(19) O(16)]Pr(1)]O(19) O(9)]Pr(1)]O(19) 3.396(2) 2.412(5) 1.892(5) 1.898(5) 2.034(4) 2.336(2) 1.902(5) 1.914(4) 2.039(4) 2.358(2) 2.7674(13) 2.973(2) 126.3(2) 127.5(2) 79.4(2) 77.0(2) 81.8(2) 81.6(2) 141.7(2) 139.3(2) 140.3(2) 140.9(2) 81.0(2) 81.2(2) 91.8(2) 73.7(2) 141.0(2) 73.4(2) 139.5(2) 69.2(2) 67.5(2) 140.0(2) 74.2(2) 139.4(2) 73.3(2) 67.0(2) 67.8(2) 114.4(2) 64.94(14) 62.91(14) 63.36(14) 62.56(14) Pr(2)]O(12) Pr(2)]O(18) Pr(2)]O(13) Pr(2)]O(7) Pr(2)]O(5) Pr(2)]O(15) Pr(2)]O(8) Pr(2)]O(6) Pr(2)]O(19) Pr(2)]Cu(1) Pr(2)]Cu(3) O(13)]Pr(2)]O(19) O(7)]Pr(2)]O(19) O(5)]Pr(2)]O(19) O(15)]Pr(2)]O(19) O(8)]Pr(2)]O(19) O(6)]Pr(2)]O(19) O(11)]Pr(1)]O(14) O(11)]Pr(1)]O(16) O(12)]Cu(1)]O(16) O(12)]Cu(1)]O(19) O(16)]Cu(1)]O(19) O(12)]Cu(1)]S(4) O(16)]Cu(1)]S(4) O(19)]Cu(1)]S(4) O(12)]Cu(1)]O(15) O(16)]Cu(1)]O(15) O(19)]Cu(1)]O(15) S(4)]Cu(1)]O(15) O(9)]Cu(2)]O(15) O(9)]Cu(2)]O(19) O(15)]Cu(2)]O(19) O(9)]Cu(2)]S(1) O(15)]Cu(2)]S(1) O(19)]Cu(2)]S(1) O(18)]Cu(3)]O(14) O(18)]Cu(3)]O(19) O(14)]Cu(3)]O(19) O(18)]Cu(3)]S(5) O(14)]Cu(3)]S(5) O(19)]Cu(3)]S(5) O(11)]Cu(4)]O(13) O(11)]Cu(4)]O(19) 2.396(4) 2.415(5) 2.453(5) 2.455(6) 2.474(5) 2.479(4) 2.553(5) 2.596(5) 2.669(4) 3.2848(11) 3.3295(14) 63.05(14) 134.6(2) 134.3(2) 61.75(13) 125.1(2) 119.00(14) 77.1(2) 80.4(2) 172.8(2) 87.6(2) 86.2(2) 98.7(2) 88.5(2) 154.71(13) 85.1(2) 96.4(2) 72.5(2) 83.58(12) 168.9(2) 84.9(2) 84.0(2) 88.4(2) 102.3(2) 165.07(13) 171.2(2) 85.5(2) 85.8(2) 87.8(2) 100.9(2) 173.30(13) 167.7(2) 84.6(2) Cu(3)]O(18) Cu(3)]O(14) Cu(3)]O(19) Cu(3)]S(5) Cu(4)]O(11) Cu(4)]O(13) Cu(4)]O(19) Cu(4)]S(3) Cu(4)]S(2) Cu(4)]Cu(3) Cu(4)]Cu(1) O(4)]Pr(1)]O(19) O(1)]Pr(1)]O(19) O(2)]Pr(1)]O(19) O(3)]Pr(1)]O(19) O(11)]Cu(4)]S(3) O(13)]Cu(4)]S(3) O(13)]Cu(4)]O(19) O(19)]Cu(4)]S(3) O(11)]Cu(4)]S(2) O(13)]Cu(4)]S(2) O(19)]Cu(4)]S(2) S(3)]Cu(4)]S(2) Pr(1)]O(19)]Pr(2) Cu(2)]O(9)]Pr(1) Cu(4)]O(11)]Pr(1) Cu(3)]O(19)]Pr(1) Cu(1)]O(19)]Pr(1) Cu(2)]O(19)]Pr(1) Cu(4)]O(19)]Pr(1) Cu(1)]O(16)]Pr(1) Cu(4)]O(13)]Pr(2) Cu(3)]O(18)]Pr(2) Cu(3)]O(19)]Pr(2) Cu(1)]O(19)]Pr(2) Cu(2)]O(19)]Pr(2) Cu(4)]O(19)]Pr(2) Cu(3)]O(19)]Cu(1) Cu(3)]O(19)]Cu(2) Cu(1)]O(19)]Cu(2) Cu(3)]O(19)]Cu(4) Cu(1)]O(19)]Cu(4) Cu(2)]O(19)]Cu(4) 1.904(5) 1.905(5) 2.006(4) 2.335(2) 1.893(5) 1.895(5) 2.137(4) 2.490(2) 2.573(2) 2.8968(12) 2.980(2) 135.1(2) 133.0(2) 123.4(2) 122.2(2) 102.7(2) 85.4(2) 83.2(2) 140.05(12) 83.8(2) 106.2(2) 133.61(12) 86.34(8) 177.5(2) 99.5(2) 103.1(2) 91.71(14) 91.10(14) 90.20(14) 88.75(13) 100.9(2) 102.0(2) 100.2(2) 89.66(14) 87.52(14) 91.75(14) 89.22(14) 177.2(2) 94.6(2) 85.6(2) 88.7(2) 91.2(2) 176.6(2) Compound 2 Ba(1)]O(5) Ba(1)]O(4) Ba(1)]O(2) O(5)]Ba(1)]O(4) O(5)]Ba(1)]O(2) O(4)]Ba(1)]O(2) O(5)]Ba(1)]O(1) O(4)]Ba(1)]O(1) O(2)]Ba(1)]O(1) O(5)]Ba(1)]O(3) O(4)]Ba(1)]O(3) O(2)]Ba(1)]O(3) 2.760(3) 2.761(4) 2.780(4) 116.08(13) 169.61(13) 73.1(2) 108.57(11) 100.69(14) 63.41(13) 75.14(10) 62.92(11) 106.69(13) Ba(1)]O(59) Ba(1)]O(7) Cu(1)]O(7) O(1)]Ba(1)]O(59) O(3)]Ba(1)]O(59) O(79)]Ba(1)]O(59) O(5)]Ba(1)]O(7) O(4)]Ba(1)]O(7) O(2)]Ba(1)]O(7) O(1)]Ba(1)]O(7) O(3)]Ba(1)]O(7) O(79)]Ba(1)]O(7) 2.834(3) 2.841(3) 1.937(3) 124.76(12) 163.30(12) 43.14(9) 54.77(9) 170.85(13) 116.00(13) 83.30(12) 111.41(10) 94.03(9) Ba(1)]O(1) Ba(1)]O(3) Ba(1)]O(79) O(1)]Ba(1)]O(3) O(5)]Ba(1)]O(79) O(4)]Ba(1)]O(79) O(2)]Ba(1)]O(79) O(1)]Ba(1)]O(79) O(3)]Ba(1)]O(79) O(5)]Ba(1)]O(59) O(4)]Ba(1)]O(59) O(2)]Ba(1)]O(59) 2.784(4) 2.787(4) 2.830(3) 70.77(13) 69.43(9) 81.90(12) 118.34(12) 177.30(11) 110.04(11) 92.57(9) 114.89(10) 87.45(12) Cu(1)]O(5) Cu(1)]S(1) Cu(1)]S(2) O(59)]Ba(1)]O(7) Cu(1)]O(7)]Ba(1) Ba(19)]O(7)]Ba(1) O(7)]Cu(1)]O(5) O(7)]Cu(1)]S(1) O(5)]Cu(1)]S(1) O(7)]Cu(1)]S(2) O(5)]Cu(1)]S(2) S(1)]Cu(1)]S(2) 1.944(3) 2.348(2) 2.349(2) 68.25(9) 96.88(12) 85.97(9) 83.23(14) 171.25(11) 88.30(11) 88.83(11) 171.67(11) 99.52(6) Symmetry transformation used to generate equivalent atoms: 9 2x 1 1, 2y, 2z.Compound 3 Cu(1)]O(5) Cu(1)]O(6) Cu(1)]O(4) O(5)]Cu(1)]O(6) O(5)]Cu(1)]O(4) O(6)]Cu(1)]O(4) O(5)]Cu(1)]O(3) O(6)]Cu(1)]O(3) O(4)]Cu(1)]O(3) O(5)]Cu(1)]O(69) 1.941(5) 1.970(5) 1.984(6) 96.1(2) 89.3(2) 173.6(2) 175.6(2) 84.5(2) 89.9(2) 71.6(2) Cu(1)]S Cu(2)]O(59) Cu(2)]O(6) O(3)]Cu(1)]S O(69)]Cu(1)]S O(59)]Cu(2)]O(6) O(59)]Cu(2)]O(1) O(6)]Cu(2)]O(1) O(59)]Cu(2)]O(2) O(6)]Cu(2)]O(2) 2.654(3) 1.897(6) 1.928(5) 103.2(2) 150.09(13) 85.2(2) 91.8(2) 175.3(2) 171.5(3) 93.4(2) Cu(1)]O(3) Cu(1)]O(69) O(6)]Cu(1)]O(69) O(4)]Cu(1)]O(69) O(3)]Cu(1)]O(69) O(5)]Cu(1)]S O(6)]Cu(1)]S O(4)]Cu(1)]S 1.985(5) 2.434(6) 86.2(2) 92.2(2) 104.1(2) 81.3(2) 84.4(2) 99.9(2) Cu(2)]O(1) Cu(2)]O(2) O(1)]Cu(2)]O(2) Cu(29)]O(5)]Cu(1) Cu(1)]O(6)]Cu(19) Cu(2)]O(6)]Cu(1) Cu(2)]O(6)]Cu(19) 1.935(6) 1.952(6) 90.1(2) 108.8(3) 93.8(2) 111.4(2) 90.6(2) Symmetry transformation used to generate equivalent atoms: 9 2x, 2y, 2z 1 1.2330 J.Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 uncommon, has been observed recently in Na2Fe6(m6-O)- (OCH3)18(HOCH3)6,8 H4Ba6(m6-O)(OCH2CH2OCH3)14,9 and Y4Ba2(m6-O)(m3-OC2H5)8[ButC(O)CHC(O)But]6.10 The average praseodymium–oxo distance of 2.661 Å and the average copper–oxo distance of 2.054 Å are longer than the praseodymium–alkoxo and copper–alkoxo distances averaging 2.478 and 1.900 Å, respectively. There are three doubly deprotonated 2,29-thiodiethanol and two singly deprotonated 2,29- thiodiethanol ligands.As anticipated, the sulfur atoms are bound exclusively to the copper centers with the Cu]S bond lengths ranging from 2.335(2) to 2.573(2) Å which are within the normal range of CuII]S bond lengths.11 The Cu]O]Cu angles between the oxo ligand and the copper atoms range from 85.6(2) to 94.6(2)8. The Cu]Cu separation distances in 1 range from 2.7674(13) Å between Cu(1) and Cu(2) to 2.980(2) Å between Cu(1) and Cu(4).One of the alkoxo oxygen atoms of the tde22 ligand chelates to the copper center while the other functions as a bridging ligand between the PrIII and the second CuII. The two nondeprotonated hydroxyl oxygen atoms, O(10) and O(17), are not bound to any metal ions, but form hydrogen bonds with the oxygen atoms of the hfacac ligands, O(6) and O(4), respectively, as indicated by the distances of O(6) ? ? ? O(10) 2.925 and O(4) ? ? ? O(17) 2.971 Å.Another important feature of compound 1 is that it is a chiral molecule with each of the copper(II) ions having a distinct chemical environment. As shown in Fig. 2, the Cu(1) and Cu(2) ions have a distorted square planar geometry while Cu(3) has a nearly ideal square planar geometry [S(5)]Cu(3)]O(19) 173.30(13), O(18)]Cu(3)]O(14) 171.2(2)8]. In contrast, the Cu(4) ion is co-ordinated by three oxygen atoms and two sulfur atoms in an approximate trigonal bipyramidal geometry.There are weak axial bonds associated with the Cu(1), Cu(2) and Cu(3) centers, as evidenced by the distances of Cu(1)]O(15) 2.412(5), Cu(2)]O(18) 2.749(6) and Cu(3)]O(13) 2.574(5) Å. The O(12)]Cu(4) distance, 2.734(6) Å, could also be considered as a weak Cu]O bond. Interestingly, the four copper(II) units in 1 have a propeller arrangement around the Pr(1)]O(19)]Pr(2) axis, as shown in Fig. 2. We Fig. 1 Molecular structure of compound 1 with 50% thermal ellipsoids and labeling scheme.For clarity, fluorine and hydrogen atoms are omitted and only the metal atoms and sulfur atoms are shown as anisotropic thermal ellipsoids (50%) believe that such a propeller arrangement is likely dictated by non-bonding interactions between the 2,29-thiodiethanolato ligands. Polynuclear metal complexes with propeller structures are rare. One of the examples is Cu6(m3-O)(m3-OH)(bdmap)3Cl6 [bdmap = 1,3-bis(dimethylamino)propan-2-olate], reported recently by our group, where the six copper(II) ions are in a propeller arrangement with a C3 symmetry.12 Synthesis and structure of Ba2Cu2(Htde)4(hfacac)4 2 The Ba2Cu2 compound was initially isolated from an analogous synthetic procedure as used for compound 1.Compound 2 was subsequently synthesized using a logical procedure in which Cu(OCH3)2, H2tde and Ba(hfacac)2 reacted in a 1:2:1 ratio in CH2Cl2. The composition and structure of the compound was determined by single-crystal X-ray diVraction and elemental analysis.An ORTEP diagram displaying the structure of this Fig. 2 View of the propeller core structure of compound 1 projected down the Pr]O]Pr vector Fig. 3 Molecular structure of compound 2 with 50% thermal ellipsoids and labeling scheme. For clarity, hydrogen and fluorine atoms are omittedJ. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 2331 compound is shown in Fig. 3. Selected bond lengths and angles are provided in Table 2.Compound 2 consists of two copper(II) and two barium ions in a rhombic arrangement with the Ba(1) ? ? ? Cu(1) and Ba(1) ? ? ? Cu(19) distances being 3.5817(6) and 3.6251(6) Å, respectively, and the Ba(1)]Cu(1)]Ba(19) and Cu(1)]Ba(1)] Cu(19) angles being 64.89(1) and 115.11(1)8, respectively. The molecule of 2 has an inversion center and contains four Htde2 ligands. The four alkoxo oxygen atoms of the four Htde2 ligands function as triply bridging atoms to the copper and barium centers.The barium ion is surrounded by eight oxygen atoms, four alkoxo oxygen atoms and four from the two hfacac ligands, with the Ba]O bond lengths ranging from 2.760(3) to 2.841(3) Å, comparable to those reported previously. The geometry of the barium ion is almost an ideal square prism, which resembles that found in YBa2Cu3O7 2 x superconductor.1 The sulfur atoms co-ordinate exclusively to the copper(II) centers in a similar manner as was observed 1.Each copper center is chelated by two Htde2 ligands via the alkoxo oxygen and sulfur atoms at the equatorial sites with the O(7)]Cu(1)]S(1) and O(5)]Cu(1)]S(2) angles being 171.3(1) and 171.7(1)8, respectively. As observed in 1, the four nondeprotonated hydroxyl oxygen atoms form four intramolecular hydrogen bonds with the oxygen atoms of the hfacac ligands as evidenced by the distances of O(8) ? ? ? O(1) 2.891(5) and O(6) ? ? ? O(49) 2.919(6) Å. However, unlike compound 1, where the non-deprotonated hydroxyl oxygen atoms are not bound to any metal ions, those of the Htde2 ligands in 2 occupy the axial positions of the copper(II) centers as evidenced by the long bond lengths of O(8)]Cu(1) 2.523(5) and O(6)]Cu(1) 2.542(5) Å.The O(8)]Cu(1)]O(6) bond angle of 150.3(5)8 is significantly oV linearity, perhaps due to steric interactions. The geometry of the copper(II) center can be therefore best described as a distorted elongated octahedron. The distance between the two copper centers is 6.082(5) Å.The arrangement of the four metal ions in 2 resembles that in the Ba2Cu2 compound Ba2Cu2- (acac)4(tme)4?2(Htme) (tme = 2-methoxyethoxide), reported by Ryan and co-workers.2g Compound 2 is, however, the first BaCu bimetallic complex employing thioalcohol as the bridging ligand. Synthesis and structure of Cu4(tde)2(hfacac)4 3 Compound 3 was obtained from a reaction intended to produce a structural analogue of 1 with the non-paramagnetic yttrium ion in place of the praseodymium ion.However, instead of obtaining the Y2Cu4 compound, dark green crystals of 3 were isolated along with a blue microcrystalline compound. The structure and composition of 3 were determined by single crystal X-ray diVraction and elemental analyses. The blue compound did not form crystals suitable for X-ray diVraction analysis. The fact that it has a similar color as compound 2 and was obtained in good yield makes us suggest that the blue compound likely contains both copper and yttrium.The results of elemental analysis appeared to match well the formula of Cu2Y2(tde)2(hfacac)3(O2CCF3)2(OH), where the trifluoroacetate ligand would be originated from the decomposition of the hfacac ligand.3f However, the exact nature of this blue compound still remains a mystery. An ORTEP diagram showing the structure of 3 is given in Fig. 4. Selected bond lengths and angles are provided in Table 2. Compound 3 consists of four Cu(hfacac)1 units which are linked together by the alkoxo oxygen atoms of two tde22 ligands.Two of the oxygen atoms act as double bridges while the remaining two function as triple bridges. In contrast to compounds 1 and 2, where the tde22 and Htde2 ligands function as a bidentate chelate to the copper(II) ion, in 3 the tde22 chelates to the Cu(1) center as a tridentate ligand such that the alkoxo oxygen atoms bind to the equatorial positions while the sulfur donor occupies one of the axial positions [Cu(1)]S 2.654(3) Å].The second axial positions of Cu(1) is occupied by an inversion center symmetry related O(69) atom with a long bond length of 2.434(6) Å. The S]Cu(1)]O(69) angle of 150.09(13)8 is significantly deviated from linearity. The geometry of Cu(1) can be therefore best described as a distorted elongated octahedron. The Cu(2) center is co-ordinated by two alkoxo oxygen atoms from the bridging tde22 ligands and a chelating hexafluoroacetylacetonato ligand in a square-planar geometry. As shown in Fig. 5, the fifth position of the Cu(2) atom is occupied by a weakly bound O(3) atom from a neighbouring hfacac ligand [O(5)]Cu(2) 2.521(6) Å]. The geometry of Cu(2) is therefore a square pyramid. The four copper ions are coplanar with the separation distances being Cu(1) ? ? ? Cu(2) 3.220(5), Cu(1) ? ? ? Cu(29) 3.121(5), Cu(1) ? ? ? Cu(19) 3.230(5) and Cu(2) ? ? ? Cu(29) 5.458(6) Å, respectively. The arrangement of the copper atoms in 3 can be therefore described as a rhombus or two oxygen-capped copper triangles sharing an edge which is unusual for tetranuclear copper(II) compounds.A tetrahedral arrangement is most common for tetranuclear copper(II) compounds.13 Rectangular or square arrangements of cyclic tetranuclear copper(II) complexes, albeit rare, have been reported in compounds such as Cu4Zr4O3(OPri)18,14 [Cu2(bpim)(im)]2(NO3)4?3H2O [bpim = 4,5-bis{[(2-pyridin-2-ylethyl)imino]methyl}-2H-imidazole],15 Cu4(MPZ)4(acMPZ)2(ONO2)2 [HMPZ = 3,5-dimethyl-1H-pyrazole, HacMPZ = 1-(5-methyl-1H-pyrazol-1-yl)ethan-1-ol],16a and [Cu4(bdmap)2(O2CCH3)4][PF6]2.17 Compound 3 is one of the rare examples of rhombic CuII 4 compounds.A rhombic Cu4 arrangement has also been observed in a heteropolyoxotungstate anion,16b [Cu4(H2O)2(PW9O34)2]2.10 Despite the Fig. 4 Molecular structure of compound 3 with 50% thermal ellipsoids and labeling scheme. The hydrogen and fluorine atoms are omitted for clarity Fig. 5 Core structure and the geometry surrounding the copper centers in compound 32332 J. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 coplanarity of the four copper centers in 3, the four squareplanar CuO4 units are not coplanar. The two CuO4 units involving Cu(1) are parallel to each other. The dihedral angle between the CuO4 unit of Cu(1) and that of Cu(2) is 75.28, almost orthogonal. As soft donors, thiol ligands are often associated with copper(I) compounds.Copper(II) compounds containing sulfur donors are much less common than those of oxygen or nitrogen as donor atoms due to the reducing property of the sulfur atom. Most of the previously known sulfur-containing copper(II) compounds are limited to thiocarbamate-type ligands and dimercaptomalionitrile.18 One of the rare examples of copper(II) complexes, {Cu(R-sno)}2][ClO4]2, containing a thioalcohol ligand was reported by Kida and co-workers 19 where R-sno = N-(2-alkylsulfanylethyl)-3-aminopropanolate.The stability of the copper(II) ions in compounds 1–3 can be attributed to the chelate eVect of the tde22 or Htde2 ligand. Magnetic properties Compound 1. Compound 1 contains six paramagnetic metal ions linked by a central oxo ligand and alkoxo oxygen atoms as well. Substantial magnetic interactions between the metal ions are therefore anticipated. The susceptibility data at 2 to 300 K are shown in Fig. 6. Compound 1 appears to be dominated by antiferromagnetic exchanges, since the room temperature magnetic moment (2.64 mB) is much less than that expected for the non-interacting four copper(II) and two praseodymium centers (ª7 mB).4 In addition, the magnetic moment decreases slowly with decreasing temperature from 300 to 40 K, and drops rapidly after 40 K, which is again consistent with the dominance of antiferromagnetic exchange in the system.If one assumes that magnetic interactions between Pr and Pr and Cu and Pr are negligible, the magnetic exchange behavior of 1 Fig. 6 Plots of magnetic moment (diamond) and molar susceptibility (square) versus temperature for compounds 1 (top), 2 (middle) and 3 (bottom) would be dictated by the four copper(II) centers. To confirm it, an analogous compound, Y2Cu4 or La2Cu4, where no paramagnetic lanthanides are present, is required. Unfortunately, we have not been able to synthesize these analogues. The Cu4O portion in 1, however, resembles that in Cu4Zr4O3(OPri)18, reported by Caulton and co-workers,14 which has been shown to have a singlet ground state.The magnetic behavior of 1 appears to be similar to that of Caulton’s compound. The fact that the magnetic moment of 2.64 mB at 300 K is also much smaller than that of two non-interacting praseodymium(III) ions (ª4.9 mB) further suggests that Pr]Pr and Cu]Pr magnetic interactions are also present and likely dominated by antiferromagnetic exchange as well, which has been observed frequently in Ln]Cu complexes reported previously.2,3 Compound 2.As shown in Fig. 3, the copper(II) ions in compound 2 are far apart [6.082(5) Å]. In addition, there is no bridging ligand linking these two copper(II) ions directly. One therefore would anticipate that there is either a very weak magnetic interaction or no interaction at all between the two copper(II) ions. To find out the truth, we measured magnetic susceptibility data for this compound over the temperature range of 2–300 K.The plots of magnetic susceptibility and the eVective magnetic moment versus temperature for 2 are shown in Fig. 6 (middle). The susceptibility data have a maximum at 35 K, which is an indication of the presence of a fairly strongly coupled antiferromagnetic ground state.1c,4 To determine the magnitude of the exchange constant J, the susceptibility data were fitted by a modified Bleaney–Bowers equation (1) 1c,4 that c = (1 2 r){(2Ng2b/3kT) [1 1��� exp(22J/kT)]21} 1 rcmonomer 1 ctip (1) takes into account the eVects of paramagnetic impurities (r) and temperature independent paramagnetism (tip) using a nonlinear regression program.The results of the fitting yielded J = 219.2(1) cm21, r = 0.012, g = 2.25, tip = 20.000542 and R = 0.0056. The J value of 2, although not large, is appreciable, considering that the two copper centers are far apart with no direct ligand links. It is likely that the dipolar interaction is operative in 2 and responsible for the sizable J value.1c,4 Exchange pathways involving Cu]O]Ba]O]Cu are also possible. Compound 3.Compound 3 contains four copper in close proximity to each other. One might therefore expect the presence of significant magnetic interactions between the copper centers. However, the fact that there is no bridging ligand linking the two dx2 2 y2 orbitals of the two Cu(1) centers and the two dx2 2 y2 orbitals of the Cu(1) and Cu(2) [or Cu(1a) and Cu(2)] centers are 75.28 with respect to each other, albeit linked by the O(6) atom [O(5) in the case of Cu(1a) and Cu(2)], is not in favor of a strong magnetic exchange via superexchange pathways.1c,4 To establish the nature of magnetic exchanges in 3, magnetic susceptibility measurements were performed over the temperature range 2–300 K.The susceptibility data were corrected for diamagnetic contributions using Pascal constants. Plots of the molar susceptibility and the eVective magnetic moment versus temperature are shown in Fig. 6 (bottom). The magnetic moment of 3.48 mB at 300 K is slightly less than that of four non-interacting copper(II) ions (ª3.80 mB, assuming g = 2.2, similar to that of compound 2), suggesting that the magnetic exchange between the copper centers is fairly weak. The decreasing trend of magnetic moment with decreasing temperature implies that this compound is dominated by antiferromagnetic exchanges. The drastic decrease of magnetic moment from 1.5 mB at about 15 K to 0.5 mB at about 2 K appears to suggest that the ground state of this compound is a singlet.To obtain the magnitude of exchange constants, we attempted to fit the data by a tetramer model13 where fourJ. Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 2333 independent J values are employed (JCu1]Cu2, JCu1a]Cu2, JCu1]Cu1a, JCu2]Cu2a). Unfortunately, we were unable to obtain satisfactory fitting of the susceptibility data. Conclusion It has been shown that the 2,29-thiodiethanol ligand is eVective in forming heterometallic complexes containing copper(II) and lanthanide or alkaline earth metal ions.The sulfur atoms of the tde22 and Htde2 bind to the copper center exclusively. Despite their simplicity, the Htde2 and tde22 ligands display versatile bonding patterns to metal ions as summarized in Scheme 1. Compounds 1 and 2 are the first examples of thioalcohol bridged Ln]Cu and Ba]Cu compounds with unusual structural features.These compounds may not be useful as precursors for oxides due to the sulfur contamination, but they provide valuable information on the construction of polynuclear heterometallic compounds using thioalcohols as ligands. The environments surrounding the copper(II) centers in compounds 1–3 are distinctively diVerent: in 1 all four copper(II) ions are linked together by a central oxo ligand and alkoxo ligands as well; in 2 there is no direct ligand bridge between the two copper centers; and in 3 the copper centers are linked by alkoxo oxygen atoms in either a nearly orthogonal fashion or a face-to-face parallel manner, leading to their distinct magnetic behavior.Although quantitative analyses on the magnetic exchange of compounds 1 and 3 could not be achieved, the experimental data show that antiferromagnetic exchanges dominate in all three compounds. Acknowledgements We thank the Natural Sciences and Engineering Research Council of Canada for financial support, Professor C.V. Stager for use of the magnetometer and Jim Britten for crystal data collection for compounds 1 and 2. References 1 (a) Ceramic Superconductors, ed. M. F. Yan, American Ceramic Society, Westerville, OH, 1988; (b) Chemistry of High Temperature Superconductors, ACS Symp. Ser., 1987, 351; (c) O. Kahn, Molecular Magnetism; VCH, Weinheim, 1993; (d ) D. Gatteschi, O. Kahn, J. S. Miller and F. Palacio (Editors), Magnetic Molecular Materials, Kluwer, Dordrecht, 1991. 2 (a) O. Guillou, R. L. Oushoorn, O. Khan, K. Boukekeur and P. Betail, Angew. Chem., Int. Ed. Engl., 1992, 31, 626; (b) A. J. Blake, P. E. Y. Milne, P. Thornton and R. E. P. Winpenny, Angew. Chem., Scheme 1 Bonding modes of the Htde2 and tde22 ligands; M = PrIII or BaII M O S Cu OH M O S M Cu HO M O S O Cu Cu M Cu O S Cu O Cu Cu Int. Ed. Engl., 1991, 30, 1139; (c) D. M. L. Goodgame, D. J. Williams and R. E. P. Winpenny, Polyhedron, 1989, 8, 1531; (d ) C.P. Love, C. C. Torardi and C. J. Page, Inorg. Chem., 1992, 31, 1784; (e) W. Bidell, J. Doring, H. W. Bosch, H.-U. Hund, E. Plappert and H. Berke, Inorg. Chem., 1993, 32, 502; ( f ) W. Bidell, H. W. Bosch, D. Veghini, H.-U. Hund, J. Doring and H. Berke, Helv. Chim. Acta, 1993, 76, 596; ( g) N. N. Sauer, E. Garcia, K. V. Salazar, R. R. Ryan and J. A. Martin, J. Am. Chem. Soc., 1990, 112, 1524; (h) C. Benelli, A. Caneschi, D. Gatteschi, O. Guilou and L. Pardi, Inorg.Chem., 1990, 29, 1750; (i) M. Andruh, I. Ramade, E. Codjovi, O. Guillou, O. Kahn and J. C. Trombe, J. Am. Chem. Soc., 1993, 115, 1822; ( j) J. L. Sanz, R. Ruiz, A. Gleizes, F. Lloret, J. Fans, M. Julve, J. J. Borrás-Almenar and Y. Journaux, Inorg. Chem., 1996, 35, 7384. 3 (a) S. Wang, S. J. Trepanier and M. J. Wagner, Inorg. Chem., 1993, 32, 833; (b) S. R. Breeze and S. Wang, Inorg. Chem., 1994, 33, 5114; (c) S. R. Breeze and S. Wang, Inorg. Chim. Acta, 1996, 250, 163; (d ) S.Wang, Z. Pang and M. J. Wagner, Inorg. Chem., 1992, 61, 1613; (e) L. Chen, S. R. Breeze, R. Rousseau, S. Wang and L. K. Thompson, Inorg. Chem., 1995, 34, 454; ( f ) S. Wang, Z. Pang, K. D. L. Smith, C. Deslippe and Y. Hua, Inorg. Chem., 1995, 34, 908; ( g) S. Wang, Z. Pang, K. D. L. Smith and M. J. Wagner, J. Chem. Soc., Dalton Trans., 1994, 955. 4 R. Carlin, Magnetochemistry, Springer, Berlin, 1986; R. S. Drago, Physical Methods in Chemistry, W. B. Saunders Co., Philadelphia, 1977, W.E. Hatfield, in Magneto-Structural Correlations in Exchange Coupled Systems, eds. R. D. Willett, D. Gatteshi and O. Kahn, NATO ASI, Reidel, Dordrecht, 1985, p. 555. 5 SHELXTL, version 5, crystal structure analysis package, Bruker Axs, Analytical X-ray System, Madison, WI, 1995. 6 D. T. Cromer and J. T. Waber, International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.2A. 7 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 8 K. Hegetschweiler, H. W. Schmalle, H. M. Streit, V. Gramlich, H.-U. Hund and I. Erni, Inorg. Chem., 1992, 31, 1299. 9 K. G. Caulton, M. H. Chisholm, S. R. Drake and J. C. HuVman, J. Chem. Soc., Chem. Commun., 1990, 1498. 10 P. Miele, J. D. Foulon, N. Hovnanian and L. Cot, J. Chem. Soc., Chem. Commun., 1993, 29. 11 E. Bouwman, R. Day, W. L. Driessen, W. Tremel, B. Brebs, J. S. Wood and J. Reedijk, Inorg. Chem., 1988, 27, 4614; M. Zoeteman, E. Bouwman, R.A. G. de GraV, W. L. Driessen, J. Reedijk and P. Zanello, Inorg. Chem., 1990, 29, 3487. 12 S. Wang, Z. Pang, J. C. Zheng and M. J. Wagner, Inorg. Chem., 1993, 32, 5975. 13 S. R. Breeze, S. Wang and L. Chen, J. Chem. Soc., Dalton Trans., 1996, 1341; S. Teiplel, K. Griesar, W. Haase and B. Krebs, Inorg. Chem., 1994, 33, 456; J. W. Hall, E. D. Estes, R. P. Scaringe and W. E. Hatfield, Inorg. Chem., 1977, 16, 1527; R. Mergehenn, L. Merz and W. Haase, J. Chem. Soc., Dalton Trans., 1980, 1703; E.D. Estes and D. J. Hodgson, Inorg. Chem., 1975, 14, 334; S. Wang, J. C. Zheng and J. R. Hall, Polyhedron, 1994, 13, 1039. 14 J. A. Samuels, B. A. Vaartstra, J. C. HuVman, K. L. Trojan, W. E. Hatfield and K. G. Caulton, J. Am. Chem. Soc., 1990, 112, 9623. 15 G. Kolks and S. J. Lippard, Acta Crystallogr., Sect. C, 1984, 40, 261. 16 (a) R. W. M. Ten Hoedt, F. B. Hulsbergen, G. C. Verschoor and J. Reedijk, Inorg. Chem., 1982, 21, 2369; (b) J. Gómez-Garcia, E.Coronado and J. J. Borrás-Almenar, Inorg. Chem., 1992, 31, 1667. 17 S. Wang, S. J. Trepanier, J. C. Zheng, Z. Pang and M. J. Wagner, Inorg. Chem., 1992, 31, 2118. 18 M. Bonamico, G. Dessy, A. Mugnoli, A. Vaciago and L. Zambonelli, Acta Crystallogr., 1965, 19, 886; C. K. Schauer, K. Akabori, C. M. Elliot and O. P. Anderson, J. Am. Chem. Soc., 1984, 106, 1127. 19 M. Mikuriya, H. Okawa and S. Kida, Inorg. Chim. Acta, 1979, 34, 13. Received 25th March 1998; Paper 8/02333FJ.Chem. Soc., Dalton Trans., 1998, Pages 2327–2333 2333 independent J values are employed (JCu1]Cu2, JCu1a]Cu2, JCu1]Cu1a, JCu2]Cu2a). Unfortunately, we were unable to obtain satisfactory fitting of the susceptibility data. Conclusion It has been shown that the 2,29-thiodiethanol ligand is eVective in forming heterometallic complexes containing copper(II) and lanthanide or alkaline earth metal ions. The sulfur atoms of the tde22 and Htde2 bind to the copper center exclusively.Despite their simplicity, the Htde2 and tde22 ligands display versatile bonding patterns to metal ions as summarized in Scheme 1. Compounds 1 and 2 are the first examples of thioalcohol bridged Ln]Cu and Ba]Cu compounds with unusual structural features. These compounds may not be useful as precursors for oxides due to the sulfur contamination, but they provide valuable information on the construction of polynuclear heterometallic compounds using thioalcohols as ligands. The environments surrounding the copper(II) centers in compounds 1–3 are distinctively diVerent: in 1 all four copper(II) ions are linked together by a central oxo ligand and alkoxo ligands as well; in 2 there is no direct ligand bridge between the two copper centers; and in 3 the copper centers are linked by alkoxo oxygen atoms in either a nearly orthogonal fashion or a face-to-face parallel manner, leading to their distinct magnetic behavior.Although quantitative analyses on the magnetic exchange of compounds 1 and 3 could not be achieved, the experimental data show that antiferromagnetic exchanges dominate in all three compounds. Acknowledgements We thank the Natural Sciences and Engineering Research Council of Canada for financial support, Professor C.V. Stager for use of the magnetometer and Jim Britten for crystal data collection for compounds 1 and 2. References 1 (a) Ceramic Superconductors, ed. M. F. Yan, American Ceramic Society, Westerville, OH, 1988; (b) Chemistry of High Temperature Superconductors, ACS Symp.Ser., 1987, 351; (c) O. Kahn, Molecular Magnetism; VCH, Weinheim, 1993; (d ) D. Gatteschi, O. Kahn, J. S. Miller and F. Palacio (Editors), Magnetic Molecular Materials, Kluwer, Dordrecht, 1991. 2 (a) O. Guillou, R. L. Oushoorn, O. Khan, K. Boukekeur and P. Betail, Angew. Chem., Int. Ed. Engl., 1992, 31, 626; (b) A. J. Blake, P. E. Y. Milne, P. Thornton and R. E. P. Winpenny, Angew. Chem., Scheme 1 Bonding modes of the Htde2 and tde22 ligands; M = PrIII or BaII M O S Cu OH M O S M Cu HO M O S O Cu Cu M Cu O S Cu O Cu Cu Int.Ed. Engl., 1991, 30, 1139; (c) D. M. L. Goodgame, D. J. Williams and R. E. P. Winpenny, Polyhedron, 1989, 8, 1531; (d ) C. P. Love, C. C. Torardi and C. J. Page, Inorg. Chem., 1992, 31, 1784; (e) W. Bidell, J. Doring, H. W. Bosch, H.-U. Hund, E. Plappert and H. Berke, Inorg. Chem., 1993, 32, 502; ( f ) W. Bidell, H. W. Bosch, D. Veghini, H.-U. Hund, J.Doring and H. Berke, Helv. Chim. Acta, 1993, 76, 596; ( g) N. N. Sauer, E. Garcia, K. V. Salazar, R. R. Ryan and J. A. Martin, J. Am. Chem. Soc., 1990, 112, 1524; (h) C. Benelli, A. Caneschi, D. Gatteschi, O. Guilou and L. Pardi, Inorg. Chem., 1990, 29, 1750; (i) M. Andruh, I. Ramade, E. Codjovi, O. Guillou, O. Kahn and J. C. Trombe, J. Am. Chem. Soc., 1993, 115, 1822; ( j) J. L. Sanz, R. Ruiz, A. Gleizes, F. Lloret, J. Fans, M. Julve, J. J. Borrás-Almenar and Y. Journaux, Inorg.Chem., 1996, 35, 7384. 3 (a) S. Wang, S. J. Trepanier and M. J. Wagner, Inorg. Chem., 1993, 32, 833; (b) S. R. Breeze and S. Wang, Inorg. Chem., 1994, 33, 5114; (c) S. R. Breeze and S. Wang, Inorg. Chim. Acta, 1996, 250, 163; (d ) S. Wang, Z. Pang and M. J. Wagner, Inorg. Chem., 1992, 61, 1613; (e) L. Chen, S. R. Breeze, R. Rousseau, S. Wang and L. K. Thompson, Inorg. Chem., 1995, 34, 454; ( f ) S. Wang, Z. Pang, K. D. L. Smith, C. Deslippe and Y. Hua, Inorg. Chem., 1995, 34, 908; ( g) S.Wang, Z. Pang, K. D. L. Smith and M. J. Wagner, J. Chem. Soc., Dalton Trans., 1994, 955. 4 R. Carlin, Magnetochemistry, Springer, Berlin, 1986; R. S. Drago, Physical Methods in Chemistry, W. B. Saunders Co., Philadelphia, 1977, W. E. Hatfield, in Magneto-Structural Correlations in Exchange Coupled Systems, eds. R. D. Willett, D. Gatteshi and O. Kahn, NATO ASI, Reidel, Dordrecht, 1985, p. 555. 5 SHELXTL, version 5, crystal structure analysis package, Bruker Axs, Analytical X-ray System, Madison, WI, 1995. 6 D. T. Cromer and J. T. Waber, International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.2A. 7 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 8 K. Hegetschweiler, H. W. Schmalle, H. M. Streit, V. Gramlich, H.-U. Hund and I. Erni, Inorg. Chem., 1992, 31, 1299. 9 K. G. Caulton, M. H. Chisholm, S. R. Drake and J. C. HuVman, J. Chem. Soc., Chem. Commun., 1990, 1498. 10 P. Miele, J. D. Foulon, N. Hovnanian and L. Cot, J. Chem. Soc., Chem. Commun., 1993, 29. 11 E. Bouwman, R. Day, W. L. Driessen, W. Tremel, B. Brebs, J. S. Wood and J. Reedijk, Inorg. Chem., 1988, 27, 4614; M. Zoeteman, E. Bouwman, R. A. G. de GraV, W. L. Driessen, J. Reedijk and P. Zanello, Inorg. Chem., 1990, 29, 3487. 12 S. Wang, Z. Pang, J. C. Zheng and M. J. Wagner, Inorg. Chem., 1993, 32, 5975. 13 S. R. Breeze, S. Wang and L. Chen, J. Chem. Soc., Dalton Trans., 1996, 1341; S. Teiplel, K. Griesar, W. Haase and B. Krebs, Inorg. Chem., 1994, 33, 456; J. W. Hall, E. D. Estes, R. P. Scaringe and W. E. Hatfield, Inorg. Chem., 1977, 16, 1527; R. Mergehenn, L. Merz and W. Haase, J. Chem. Soc., Dalton Trans., 1980, 1703; E. D. Estes and D. J. Hodgson, Inorg. Chem., 1975, 14, 334; S. Wang, J. C. Zheng and J. R. Hall, Polyhedron, 1994, 13, 1039. 14 J. A. Samuels, B. A. Vaartstra, J. C. HuVman, K. L. Trojan, W. E. Hatfield and K. G. Caulton, J. Am. Chem. Soc., 1990, 112, 9623. 15 G. Kolks and S. J. Lippard, Acta Crystallogr., Sect. C, 1984, 40, 261. 16 (a) R. W. M. Ten Hoedt, F. B. Hulsbergen, G. C. Verschoor and J. Reedijk, Inorg. Chem., 1982, 21, 2369; (b) J. Gómez-Garcia, E. Coronado and J. J. Borrás-Almenar, Inorg. Chem., 1992, 31, 1667. 17 S. Wang, S. J. Trepanier, J. C. Zheng, Z. Pang and M. J. Wagner, Inorg. Chem., 1992, 31, 2118. 18 M. Bonamico, G. Dessy, A. Mugnoli, A. Vaciago and L. Zambonelli, Acta Crystallogr., 1965, 19, 886; C. K. Schauer, K. Akabori, C. M. Elliot and O. P. Anderson, J. Am. Chem. Soc., 1984, 106, 1127. 19 M. Mikuriya, H. Okawa and S. Kida, Inorg. Chim. Acta, 1979, 34, 13. Received 25th March 1998; Paper 8/02333F

ISSN:1477-9226    

DOI:10.1039/a802333f

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2329–2336 2329 Mono- and di-nuclear complexes of (trpy)MII (M 5 Pd, Pt) with the model nucleobase 1-methylcytosine. Crystal structure and NMR solution studies † Sultan Cos�ar,a Matthias B. L. Janik,a Matthias Flock,a Eva Freisinger,a Etelka Farkas b and Bernhard Lippert *a a Fachbereich Chemie, University of Dortmund, D-44221 Dortmund, Germany b Department of Inorganic and Analytical Chemistry, L. Kossuth University, Debrecen, Hungary Received 12th April 1999, Accepted 2nd June 1999 Reactions of (trpy)MII (M = Pd and Pt; trpy = 2,29:69,20-terpyridine) with the model nucleobase 1-methylcytosine (Hmcyt) have been performed in water and studied by 1H and 195Pt NMR spectroscopy.Mononuclear [(trpy)M(Hmcyt-N3)]21 1 (M = Pd) and 3 (M = Pt) and dinuclear [{(trpy)M}2(mcyt-N3,N4)]31 2 (M = Pd) and 4 (M = Pt) complexes are formed. The two (trpy)M entities in the dinuclear species 2 and 4 are arranged syn to each other in the solid state.X-Ray crystal structures have been performed for [(trpy)Pd(Hmcyt-N3)][NO3]2?5H2O 1, [{(trpy)Pd}2(mcyt-N3,N4)][ClO4]3?H2O 2b, [(trpy)Pt(Hmcyt-N3)][NO3]2?5H2O 3 and [{(trpy)Pt}2(mcyt- N3,N4)][ClO4]3?H2O 4b. The eVect of the M–trpy entity on the resonances of the Hmcyt nucleobase in 1 and 3 (large downfield shifts of H5 and H6) is possibly related to the p-acceptor capacity of the trpy ligand. Mono- and di-nuclear 2,29:69,20-terpyridine (trpy) complexes of PdII and PtII have been the subject of numerous studies, e.g.with regard to photophysical 1,2 and electrochemical 3 properties as well as substitution-mechanistic aspects.4 The mononuclear compounds have also been applied in biochemical and biochemistry- related studies as DNA intercalators 5,6 or metal probes for the imidazole moieties of histidine 7,8 and arginine residues.9 Reactions of [(trpy)PtCl]1 with thiols 10 and the unexpected easy cleavage of the Pt–S bond by CuII and ZnII in phosphate buVers 11 have likewise been investigated.More recently, supramolecular aggregation of [(trpy)Pt(Me)]1 to the a-helix of poly(L-glutamic acid) has been reported.12 [(trpy)- PtCl]1,13 [(trpy)Pt(OH)]1,14 and [{(trpy)Pt}2L]41 (L = 4,49- vinylenedipyridine 15a or 4-picoline 15b) have been found to initially intercalate into DNA before forming covalent adducts with bases, apparently preferentially with guanine. We have recently carried out a solution study on the interaction of [(trpy)PdCl]1 with model nucleobases such as 1-methylcytosine (Hmcyt).16 Formation of mononuclear [(trpy)Pd(Hmcyt-N3)]21 as well as dinuclear [{(trpy)Pd}2(mcyt- N3,N4)]31 was clearly evident from potentiometric and proton NMR studies, but there was conflicting evidence as to the spatial orientation of the two metal entities in the dinuclear complex.The ambiguity was the result of rather unusual downfield shifts of the aromatic protons of the cytosine nucleobase, and in particular of H5, which previously had been observed only in cases where a metal ion was attached to the exocyclic 4-position and oriented anti with respect to N3.Lowe and Vilaivan,15b on the other hand, have assigned for a dinuclear cytidine complex of (trpy)PtII a stacked conformation on the basis of NMR and UV-vis spectra. We therefore decided to isolate and characterise the above complexes and to also study the PtII analogues. † Supplementary data available: structures of cations 3 and 4b; 1H NMR and 2D DQF COSY spectra of 2 and 4; chemical shifts and coupling constants of 1–4.For direct electronic access see http:// www.rsc.org/suppdata/dt/1999/2329/, otherwise available from BLDSC (No. SUP 57571, 8 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton). Experimental Synthesis of complexes [(trpy)PdCl]Cl?2H2O,17 [(trpy)PtCl]Cl?2H2O18 and Hmcyt19 were prepared as described in the literature.[(trpy)Pd(Hmcyt-N3)][NO3]2?5H2O 1. To an aqueous solution of Hmcyt (0.025 g, 0.2 mmol) and AgNO3 (0.068 g, 0.4 mmol) was added [Pd(trpy)Cl]Cl?2H2O (0.089 g, 0.2 mmol). The reaction mixture was stirred in the dark at 40 8C for 1 d. After filtration of AgCl the pH of the yellow solution (3.0) was adjusted to 4.0 by addition of 0.1 M NaOH. The solution was concentrated at 35 8C to about half of the original volume (5–7 ml), and then left for crystallisation for 7 d.The product was filtered oV and air-dried. Yield: 89.50 mg (67%) of yellow cubic crystals of 1. A suitable crystal was characterised by X-ray crystallography (Found: C, 35.3; H, 3.9; N, 16.6. Calc. for C20H28N8O12Pd: C, 35.4; H, 4.2; N, 16.5%). IR (KBr, n& max/ cm21): 1674s, 1628s (CO), 1374s (NO3 2). [{(trpy)Pd}2(mcyt-N3,N4)][NO3]3?7.5H2O 2a. [Pd(trpy)Cl]Cl? 2H2O (0.179 g, 0.4 mmol) was added to a solution of Hmcyt (0.025 g, 0.2 mmol) and AgNO3 (0.136 g, 0.8 mmol) in water (15 ml).The resultant mixture was stirred in the dark at 40 8C for 1 d. The formed AgCl was filtered oV and the pH of the resulting yellow solution (1.9) was adjusted to 7.9 by addition of 0.1 M NaOH solution. The volume was reduced at 30–35 8C to 5–6 ml and after several days at room temperature orange needles of 2a formed. The yield of the product was 184.60 mg (41%) (Found: C, 37.3; H, 3.3; N, 15.0. Calc. for C35H43N12O17.5Pd2: C, 37.4; H, 3.8; N, 15.0%). IR (KBr, n& max/cm21): 1629s, 1655s (CO), 1383s (NO3 2).UV-vis (H2O): n/cm21 (e/l mol21 cm21) 263 (25030), 332 (8900), 347 (12270), 364 (9330). [{(trpy)Pd}2(mcyt-N3,N4)][ClO4]3?H2O 2b. To a solution of [{(trpy)Pd}2(mcyt-N3,N4)][NO3]3 2a (0.067 g, 0.07 mmol) in water (8 ml) was added a NaClO4 solution (0.35 M, 2 ml). A yellow amorphous precipitate was filtered oV and recrystallised from acetonitrile (6 ml) to give orange-brownish crystals suit-2330 J. Chem. Soc., Dalton Trans., 1999, 2329–2336 able for X-ray crystallography.Yield: 43.80 mg (58%) (Found: C, 37.6; H, 2.7; N, 11.4. Calc. for C35H30N9O14Cl3Pd2: C, 37.5; H, 2.7; N, 11.3%). IR (KBr, n& max/cm21): 1628s, 1653s (CO), 1089s (ClO4 2), 624s (ClO4 2). [(trpy)Pt(Hmcyt-N3)][NO3]2?5H2O 3. To a solution of Hmcyt (0.025 g, 0.2 mmol) and AgNO3 (0.068 g, 0.4 mmol) in water was added [Pt(trpy)Cl]Cl?2H2O (0.107 g, 0.2 mmol). The reaction mixture was stirred in the dark at 65 8C for 3 d. The red suspension was cooled to room temperature and the pH was raised from 3.8 to 4.0 by addition of 0.1 M NaOH solution.The reaction mixture was stirred for a further 2 d at 65 8C. After cooling to room temperature the AgCl was filtered oV, and the volume was reduced to 5–6 ml on a rotavapor. After several days yellow cubic crystals of compound 3 had formed. The crystals were filtered oV and air-dried. The product was identi- fied by X-ray crystallography. Yield: 25.10 mg (17%) (Found: C, 31.4; H, 3.4; N, 14.9.Calc. for C20H28N8O12Pt: C, 31.3; H, 3.7; N, 14.6%). IR (KBr, max/cm21): 1671s, 1628s (CO), 1381s (NO3 2). [{(trpy)Pt}2(mcyt-N3,N4)][NO3]3?4H2O 4a. [Pt(trpy)Cl]Cl? 2H2O (0.200 g, 0.4 mmol) was added to a solution of Hmcyt (0.023 g, 0.2 mmol) and AgNO3 (0.127 g, 0.7 mmol) in water (15 ml) and stirred at 40 8C for 1 d. The orange-red suspension was heated to 90 8C and stirred for 1 d. Then the red reaction mixture was cooled to room temperature, the pH brought from 1.8 to 8.7 with 0.1 M NaOH solution, and the suspension stirred at 70 8C for another 2 d.After cooling to room temperature AgCl was filtered oV and the pH of 6.3 was adjusted to 9.3. Following reduction of the volume to 4 ml, 4a crystallised as dark red crystals after a few days. The product was filtered oV and airdried. Yield: 192.20 mg (43%) (Found: C, 33.6; H, 3.1; N, 13.4. Calc. for C35H36N12O14Pt2: C, 33.9; H, 3.1; N, 13.6%). IR (KBr, n& max/cm21): 1636s, 1658s (CO), 1384s (NO3 2).UV-vis (H2O): n/cm21 (e/l mol21 cm21) 280 (25450), 340 (13940), 432 (1710), 462 (2380), 492 (2870). [{(trpy)Pt}2(mcyt-N3,N4)][ClO4]3?H2O 4b. To an aqueous solution (20 ml) of Hmcyt (0.025 g, 0.2 mmol) and AgClO4? H2O (0.1803 g, 0.8 mmol) was added [Pt(trpy)Cl]Cl?2H2O (0.214 g, 0.4 mmol). The red suspension was stirred in the dark at 65 8C for 1 d, the pH was raised from 2.5 to 8.9 by addition of 0.1 M NaOH and stied for another 2 d. After cooling to room temperature AgCl was filtered oV and the pH was again adjusted to 9.2.From this orange solution, X-ray quality crystals of 4b (37.50 mg, 7%) precipitated over several days at 24 8C (Found: C, 32.1; H, 2.3; N, 9.7. Calc. for C35H30O14- N9Cl3Pt2: C, 32.4; H, 2.3; N, 9.7%). IR (KBr, nmax/cm21): 1685s, 1656s (CO), 1087s, 622s (ClO4 2). Physical measurements Fourier-transform infrared spectra (KBr pellets) were recorded on a Bruker IFS 113 v FTIR instrument. A Hitachi U-2000 spectrophotometer was used to record UV-vis spectra.Proton NMR spectra were recorded on Bruker AC 200, DPX 300 and DRX 400 spectrometers at ambient temperature. Sodium 3-(trimethylsilyl)propanesulfonate, TSP (1H, D2O) or tetramethylsilane, TMS (1H, DMSO[D6], DMF[D7]) were used as internal references. 2D DQF COSY, NOESY (tm 1.5 s) and TOCSY (tm 150 ms) spectra were recorded with 1k or 2k data points in f2 and 256 or 512 experiments in f1 and apodized with 908 phase-shifted squared sine bell functions in both dimensions.For 195Pt{1H} NMR spectra, the 195Pt edited 1H NMR spectrum and the 195Pt, 1H HMQC spectrum were recorded on the AC 200 spectrometer operating at 42.9 MHz at 323 K. 195Pt chemical shifts were referenced against external Na2PtCl6. For the HMQC 64 experiments with 128 transitions were collected in f1 with 2k data points in f2. A 908 shifted squared sine bell function in f2 and an exponential multiplication in f1 were applied prior to Fourier transformation.t1 Noise was removed by background subtraction. The sequence was optimized for a J value of 25 Hz. pD values were obtained by adding 0.4 to the pH meter reading (Metrohm 6321; combination glass electrode). pH* Values represent uncorrected pH meter readings in D2O solutions. Crystallography Intensity data for 1, 2b, 3 and 4b were collected on an Enraf- Nonius Kappa CCD20 (Mo-Ka, l = 0.71069 Å, graphite monochromator) with sample-to-detector distances of 28.7 (1, 2b) and 30.2 mm (3, 4b). They covered the whole sphere of reciprocal space by measurement of 360 frames rotating about w in steps of 18 with scan times of 60 (1, 3), 22 (2b), and 20 s (4b) per frame.Preliminary orientation matrices and unit cell parameters were obtained from the peaks of the first ten frames, respectively, and refined using the whole data set. Frames were integrated and corrected for Lorentz and polarization eVects using DENZO.21 The scaling and the global refinement of crystal parameters were performed by SCALEPACK.21 Reflections, which were partly measured on previous and following frames, were used to scale these frames on each other.This procedure in part eliminates absorption eVects and also considers crystal decay if present. The structures were solved by standard Patterson methods 22 and refined by full-matrix least-squares based on F2 using the SHELXTL-PLUS23 and SHELXL-93 programs.24 The scattering factors for the atoms were those given in the SHELXTLPLUS program.Transmission factors were calculated with SHELXL-97.25 Hydrogen atoms were placed at calculated positions and refined with a common isotropic temperature factor, except for those in 4b, which could be localised with diVerence Fourier syntheses and were not further refined. None of the structures show any disorder besides 2b, where the oxygens of two perchlorate anions were spread over eight and seven positions, respectively. All non-hydrogen atoms were refined anisotropically with the following exceptions in order to save parameters: the atoms of the Hmcyt ring, some of the atoms of the trpy ligand, the disordered perchlorate oxygens and the water molecule O(1w) in 2b as well as some of the trpy atoms in 4b.Crystal data and data collection parameters are summarised in Table 1. CCDC 186/1486. See http://www.rsc.org/suppdata/dt/1999/2329/ for crystallographic files in .cif format. Results and discussion Solid state structures of [(trpy)M(Hmcyt-N3)][NO3]2?5H2O (M 5 Pd (1), Pt (3)) The two compounds 1 and 3 are isostructural.As an example, the cation of 1 and the atom numbering scheme is given in Fig. 1. The cation of 3 is depicted in the Supplementary Material (SUP 57571). Selected interatomic distances and angles of the two compounds are listed in Table 2. Metal binding is through the N3 position of the neutral Hmcyt nucleobase. The metal coordination sphere is square planar, with the expected deviations from right angles about the heavy metal.The M–N bond length to the central N atom of the trpy ligand is well below 2 Å and significantly shorter than any of the three other M–N bonds. The trpy ring and the Hmcyt base are almost perpendicular to each other (dihedral angle 84.2(2)8, av. of 1 and 3). There are no unusual structural features of the (trpy)MII entity when compared with other [(trpy)MX]n1 species.26,27 Comparison of 1 and 3 with the Hmcyt complexes trans- [M(NH3)2(Hmcyt-N3)2][NO3]2 (M = Pd,28 Pt 29) and trans-[PdCl2- (Hmcyt-N3)2] 30 reveals diVerences only in the cases of the PdJ.Chem. Soc., Dalton Trans., 1999, 2329–2336 2331 Table 1 Crystallographic data for compounds 1, 2b, 3 and 4b 1 2b 3 4b Chemical formula Formula weight T/K Crystal system Space group a/Å b/Å c/Å b/8 V/Å3 Z m(Mo-Ka)/mm21 No. reflections measured No. reflections observed R1 (obs. data) wR2 (obs. data) C20H28N8O12Pd 678.90 293(2) Monoclinic Cc 11.104(2) 10.193(2) 24.365(5) 99.34(3) 2721.1(9) 4 0.758 7866 3427 I > 2s(I) 0.0366 0.0641 a C35H30N9O14Cl3Pd2 1119.83 293(2) Monoclinic P21/c 14.309(3) 10.961(2) 25.975(5) 100.95(3) 3999.8(14) 4 1.181 11271 2980 I > 2s(I) 0.0439 0.0954 b C20H28N8O12Pt 767.59 126(2) Monoclinic Cc 11.061(2) 9.993(2) 24.211(5) 98.70(3) 2645.3(9) 4 5.382 8522 4699 I > 2s(I) 0.0238 0.0579 c C35H30N9O14Cl3Pt2 1297.21 163(2) Monoclinic P21/c 14.066(3) 10.988(2) 25.783(5) 101.70(3) 3902.2(13) 4 7.451 10867 3277 I > 2s(I) 0.0409 0.0747 d R1 = S Fo| 2 |Fc /S|Fo|, wR2 = [Sw(Fo 2 2 Fc 2)2/Sw(Fo 2)2]� �� .a w = 1/[s2(Fo 2) 1 0.0239P2 1 0.00P]. P = [Max(Fo 2, 0) 1 2Fc 2]/3. b w = 1/[s2(Fo 2) 1 0.0553P2 1 0.00P]. P = [Max(Fo 2, 0) 1 2Fc 2]/3. c w = 1/[s(Fo 2) 1 0.0351P2 1 0.00P]. P = [Max(Fo 2, 0) 1 2Fc 2]/3. d w = 1/[s2(Fo 2) 1 0.0327P2 1 0.00P]. P = [Max(Fo 2, 0) 1 2Fc 2]/3. complexes: thus in 1 the internal ring angle at N(3) (119.0(7)8) is somewhat smaller (3.8s, with s = (s1 2 1 s2 2)1/2) than that found in the two other Pd compounds, whereas the internal ring angle at C(2) (120.5(6)8) is larger (4.8–5.7s) in 1.The angle at N(3) is therefore comparable to that of free Hmcyt (120.0(1)8) 31 and much smaller than that of N3-protonated cytosine, H2mcyt1 (124.7(3)8,32 8.5s). There are no statistically significant diVerences in bond lengths between any of the PdII or PdII compounds discussed here and Hmcyt or H2mcyt1.This applies in particular for bond lengths involving the C(5) atom (see below). The packing patterns of 1 and 3 are dominated by the following motifs: cations are arranged like tiles of a roof with pyridine Fig. 1 Molecular cation of [(trpy)Pd(Hmcyt-N3)][NO3]2?5H2O 1 with atom numbering scheme. The Pt analogue 3 is similar and not shown. Table 2 Selected distances (Å) and angles (8) for 1 and 3 1 3 M–N(3) M–N(1a) M–N(1b) M–N(1c) N(1a)–M–N(3) N(1c)–M–N(3) N(1a)–M–N(1b) N(1c)–M–N(1b) C(2)–N(3)–C(4) C(4)–N(4) C(2)–O(2) trpyM/Hmcyt 2.028(7) 2.013(4) 1.932(7) 2.011(4) 96.8(2) 100.8(2) 82.0(3) 80.5(3) 119.0(7) 1.328(8) 1.234(6) 84.0(2) 2.056(8) 2.014(4) 1.941(8) 2.025(4) 96.6(3) 101.0(3) 81.9(3) 80.7(3) 120.8(7) 1.327(7) 1.229(6) 84.4(2) rings a and c of each trpy ligand stacking with their respective neighbours.Cytosine rings, which are approximately perpendicular to the trpy ligands, are likewise parallel, but 7.0 Å apart. In between two adjacent cytosine rings a NO3 2 anion is sandwiched.Individual rows of tiles are interconnected by a hydrogen bonding pattern (Fig. 2, compound 1) which involves a water molecule, O2 of one nucleobase and N4 of another, as well as a NO3 2 anion which is sandwiched between trpy ligands.estingly there is also a relatively short contact of 3.305(4) Å between the water molecule (O(1w) in Fig. 2) and one of the Pd ions (Pd(1a)). In the case of the Pt complex 3 this distance is 3.339(4) Å.The other water molecules form an ordered spine of hydration which occasionally involves oxygen atoms of nitrate anions and, with the exception of N(4) (O(3w)–N(4a), 3.12(1) Å), no other atoms of the cations. Solid state structures of [{(trpy)M}2(mcyt-N3,N4)][ClO4]3?H2O (M 5 Pd (2b), Pt (4b)) The dinuclear complexes [{(trpy)M}2(mcyt-N3,N4)]31 (M = Pd (2), Pt (4)) were prepared from [(trpy)M(H2O)]21/[(trpy)- M(OH)]1 and Hmcyt in slightly alkaline solution (pH 8–9).Previous work16 had shown (in the case of M = Pd) that formation of 2 under these conditions is quantitative. 2 and 4 were isolated as nitrate and perchlorate salts, but only crystals of the ClO4 2 salts 2b and 4b proved suitable for X-ray analysis. Fig. 3 gives a view of the dinuclear cation of 2b. The Pt2 complex 4b is very similar to that of 2b and therefore is not Fig. 2 Section of crystal packing of 1 with H2O molecules connecting the nucleobases via H bond formation indicated (O(2)–O(1w), 2.772(5) Å; O(1w)–N(4a), 2.933(6) Å).O(1w) also forms a short contact (3.305(4) Å) with Pd(1a). Pd ? ? ? Pd distances are 7.537(1) Å. Each trpy ring is part of a tile-like arrangement of (trpy)Pd running approximately perpendicular to the plane of the paper. DiVerentiation of light atoms: C, shaded; N, empty; O, hatched.2332 J. Chem. Soc., Dalton Trans., 1999, 2329–2336 discussed in detail. Selected interatomic distances and angles of 2b and 4b are listed in Table 3.As can be seen, the two (trpy)PdII entities in 2b adopt a syn orientation when binding to N3 and the deprotonated N4 position of the cytosine base. Both (trpy)Pd entities are again close to perpendicular to the cytosine plane. The two (trpy)PdII planes are not exactly parallel (dihedral angle 17.58) and are also slightly twisted (188) with respect to each other. Pd–N distances are in the expected range, with the Pd–N(4) bond being significantly (8.5s) shorter than the Pd–N(3) bond.Again, of the M–N bonds those involving N atoms of the trpy ligands (trans to N(3) and trans to N(4)) are significantly shorter than any of the six other M–N bonds. The Pd–Pd distance within the cation of 2b is 3.0431(11) Å. This Fig. 3 Dinuclear cation of [{(trpy)Pd}2(mcyt-N3,N4)][ClO4]3?H2O 2b. The Pt compound has an analogous structure and is not shown. The numbering scheme of the trpy ligands is diVerent from that used for 1. DiVerentiation of light atoms as in Fig. 2. Table 3 Selected distances (Å) and angles (8) for 2b and 4b 2b 4b M(1)–N(3) M(1)–N(11) M(1)–M(12) M(1)–N(13) N(3)–M(1)–N(11) N(3)–M(1)–N(13) N(11)–M(1)–N(12) N(13)–M(1)–N(12) M(2)–N(4) M(2)–N(21) M(2)–N(22) M(2)–N(23) N(4)–M(2)–N(21) N(4)–M(2)–N(23) N(21)–M(2)–N(22) N(23)–M(2)–N(22) C(2)–N(3)–C(4) C(4)–N(4) C(2)–O(2) M(1)–M(2) wa trpy/trpy trpy/M(1)–mcyt trpy/M(2)–mcyt M(1)–M(1a) 2.059(5) 2.033(5) 1.951(5) 2.024(5) 101.2(2) 98.0(2) 80.7(2) 80.3(2) 1.996(6) 2.014(6) 1.928(6) 2.005(7) 97.8(3) 100.6(3) 80.9(3) 80.7(3) 123.4(6) 1.308(8) 1.225(7) 3.0431(11) 18 17.5(3) 89.8(2) 90.0(2) 5.996(2) b 2.050(8) 2.028(7) 1.936(8) 2.023(7) 101.3(3) 97.7(3) 80.3(3) 81.0(3) 1.980(9) 2.019(8) 1.935(9) 2.040(8) 98.2(3) 100.1(4) 80.5(3) 81.2(3) 121.3(9) 1.341(12) 1.220(12) 3.0350(10) 18 17.1(4) 87.7(2) 87.6(2) 5.834(2) c a Torsional angle about M(1)–M(2) vector.b Symmetry operation 2x 1 1; 2y 1 1; 2z 1 1. c Symmetry operation 2x; 2y 1 2; 2z. value is shorter than twice the van der Waals radius of Pd (3.2 Å),33 but much longer than Pd–Pd separations observed in dinuclear PdII complexes containing two bridging mcyt nucleobases (2.948(1) Å 28) or two bridging 1-methylthyminato ligands (2.848(1) Å 34).The imido proton at the N4 position of mcyt in 2b was located in the structure determination (N(4)–H, 0.81(6) Å; C(4)–N(4)–H, 111(5)8). Its position is consistent with sp2 hybridization of N(4). The distance between this proton and H(5) of mcyt is 2.34(7) Å (c.f.NMR section below). Dinuclear cations of 2b are stacked to give centrosymmetric pairs, as shown in Fig. 4. The stacking distance between the trpy rings of Pd(1) and Pd(1a) (symmetry operation 2x11; 2y11; 2z 1 1) is @ 3.4 Å. Each O(2) oxygen atom of the bridging cytosinato ligands forms two short intercationic contacts with the opposite trpy, e.g. O(2)–C(12d), 3.258(9) Å, and O(2)–C(13b), 3.187(9) Å. The intercationic Pd(1)–Pd(1a) separation is 5.996(2) Å. 2b and 4b have a close structural similarity with other diplatinum(II) complexes containing two trpy ligands and a bridging NCN type ligand such as canavanine,9 guanidine,10 or formamidine.35 Similarities include the nearly eclipsed arrangement of the two trpy ligands in the dinuclear complex, and to similar metal–metal distances. These distances in 2b (3.0431(11) Å) and 4b (3.0350(10) Å) are at the lower end of this series, the shortest ones being 2.9884(7) and 2.9872(8) Å for two crystallographically independent cations of the canavanine 9 complex of PtII.They are significantly shorter than those observed in dinuclear complexes containing diVerent types of bridges (pyrazole 36 or azaindole 2b) as well as within pairs of unbridged [(trpy)PtX]n11b,2a,5a or [(trpy)Ag(MeCN)]1 cations (3.1698(12) Å).37 However, they are larger than the distance observed in a metal–metal bonded, acetato-bridged dirhodium(II) complex (2.6341(9) Å).38 Solution NMR spectra of mononuclear Pt compound 3 Fig. 5 gives the low field portion of the 1H NMR spectrum of [(trpy)Pt(Hmcyt-N3)]21 3 in D2O. Individual trpy resonances have been assigned by a 2D DQF COSY experiment. Chemical shifts and coupling constants are close to previously found values, except the upfield shifted resonance of H69 which is known to be strongly concentration dependent.5a What are unusual are the shifts of the H5 and H6 doublets of the Hmcyt Fig. 4 Centrosymmetric dimer-of-dimers arrangement of 2b.Short contacts between O(2) of mcyt and heteroaromatic trpy C atoms are indicated. DiVerentiation of light atoms as in Fig. 2.J. Chem. Soc., Dalton Trans., 1999, 2329–2336 2333 nucleobase (d 6.32 and 7.91, 3J 7.5 Hz). As with the Pd analogue 1 (see below), the downfield shifts of these resonances relative to the free nucleobase as a consequence of (trpy)MII binding to N3 exceed those of the N3 protonated nucleobase (Table 4). This situation appears to be unique for the trpy ligand in that it violates the “rule of thumb” according to which the eVect of a metal ion such as PdII or PtII carrying Cl, H2O, NH3, or aliphalic amine ligands is usually less than that of a proton sitting at the same donor atom.This rule holds up even for metal ions in a higher oxidation state, e.g. PtIV, having a coordination sphere of similar ligands (H2O, OH, NH3, or amine).39 There are no changes in the 1H NMR spectrum with time (days), indicating that 3 is kinetically inert.In DMF[D7] the N4H2 resonances of the Hmcyt ligand in 3 occur at d 9.04 (NH syn to N3) and d 9.19 (NH anti to N3) as two singlets, which compares with d 6.85 and d 7.10 for free Hmcyt in the same solvent. This dramatic downfield shift reflects a substantial increase in the acidity of this group due to (trpy)PtII binding at the N3 site. An additional eVect of the anion, as observed with protonated cytosine,40 may also contribute to this downfield shift.Attempts to diVerentiate the two N4 protons on the basis of NOE cross-peaks with H5 of the Hmcyt ligand were at first sight not fully conclusive, in that cross-peaks between H5 and both amino protons, albeit of diVerent intensities, were observed. However, a diVerentiation of the two amino protons is possible by the following arguments. First, the NH syn to N3 should resonate at higher field than the proton anti to N3 as a consequence of the ring current of trpy.Second, the cross-peak of the syn proton with H5 should be weaker than that of the anti proton because of the larger distance to H5. Both criteria are fulfilled. The 195Pt NMR resonance of 3 at d 22791 (D2O, 25 8C) is in agreement with a N4Pt coordination sphere. Solution spectra of dinuclear Pt compound 4 The 1H NMR spectrum of [{(trpy)Pt}2(mcyt-N3,N4)]31 4 in Fig. 5 1H NMR spectrum of 3 in D2O, pD 4.8 (c = 1.6 × 1022 M). H5 and H6 resonances are due to the Hmcyt ligand.Table 4 Comparison of chemical shifts of protons of Hmcyt, H2mcyt1, and cytosine ligands in 1–4 in D2O H6 a H5 a CH3 pD Ref. Hmcyt [(dien)Pd(Hmcyt-N3)]21 [H2mcyt]1 1324 7.54 7.59 7.81 7.87 7.91 7.50 7.53 5.95 5.95 6.13 6.26 6.32 6.38 6.49 3.36 3.40 3.44 3.54 3.57 3.49 3.53 9 10.6 2 3.7 4.8 6.2 8.8 This work 16 This work This work This work This work This work a Doublet, 3J @ 7.5 Hz. D2O, pD 8.8 reveals in the low field region an additional down- field shift of the cytosine H5 resonance (d 6.49) relative to the free base, yet an upfield shift (as compared to 3) of H6 (d 7.53, d, 3J 7.5 Hz).The latter probably reflects deprotonation of the cytosine base at N4. The trpy resonances are strongly superimposed and therefore not further distinguished, with the exception of individual resonances of the H69 protons. Overall, an upfield shift of 0.1–0.4 ppm of the trpy resonances as compared to 3 is observed, which points toward stacking interactions of the aromatic rings, hence suggesting a syn orientation of the two (trpy)PtII entities as seen in the solid state.More direct evidence—interligand NOE’s between the two trpy entities—is not available due to extreme signal overlap and insignificant chemical shift dispersion. 1H NMR spectra (2D NOESY, 2D DQF COSY; see also SUP 57571) permit the assignment of a number of individual resonances of 4 (Fig. 6). Cross-peaks of the proton of the exocyclic amino group N4 (d 8.22) with H5 of mcyt and H69 of the trpy ligand bound to N4 are observed. While not inconsistent with a syn orientation of the two trpy ligands, and hence a stacked conformation, the cross-peak between N4H and H5 of mcyt is no proof of such an arrangement (see above).In the 195Pt NMR spectrum (DMSO[D6], 50 8C) of 4 two signals at d 22630 and 22540 are observed. Based on the 1H– 195Pt coupling patterns (Fig. 7), the assignment of the N3-bound Pt to the d 22540 resonance and the N4-bound Pt to the d 22630 resonance is straightforward.41,42 Thus Pt at N3 couples strongly with H5 of mcyt (4J ª 20 Hz), whereas Pt at N4 couples with N4H (2J ª 15 Hz) and weakly with H6 of mcyt (5J ª 7 Hz).In addition, cross-peaks of both Pt signals with trpy resonances H69 (3J ª 40 Hz), H59 and H39 (both 3J ª 20 Hz) are observed. As compared to the mononuclear Pt compound 3, the 195Pt NMR resonances of the diplatinum species 4 are shifted by some 200 ppm (see above). A similar trend has been reported in other complexes of dimetallated mcyt, especially in cases with metal– metal interactions occurring.42–44 However, a 1J coupling between the two Pt atoms in 3, which would be the ultimate proof of a syn orientation of the two trpy ligands, is not seen, even at a very good signal to noise ratio.Similar observations have, however, also been made in other diplatinum(II) complexes having Pt–Pt distances as short as 2.9–2.95 Å and the two Pt dx2 2 y2 orbitals parallel.45 Crystals of 4, which are reddish-black in the solid state, dissolve to give an orange-red color in water.In the UV-vis spectrum, multiple absorptions occur between 246 and 492 nm. The absorption bands in the visible region, at 462 and 492 nm, are similar to those reported by Lowe et al. for the analogous complex of 29-deoxycytidine,15b but relative intensities of these two bands are reversed in the case of 4, viz. 492 nm (e 2870 l mol21 cm21) and 462 nm (e 2380 l mol21 cm21). The Lambert–Beer law is obeyed, proving that stacking association of dinuclear species 4 is not playing a major role in solution.Che and co-workers 1 have assigned a 480 nm band of comparable intensity in a guanidinate-bridged dinuclear Pt–trpy complex to a 1[ds*- (Pt2) æÆ s(p*)(trpy)] transition, with ds* representing an orbital formed by the antibonding interaction of the dz2 orbitals of the two Pt atoms. If applicable to our system, this interpretation lends support to a syn orientation of the two Pt(trpy) entities in 4 since any interaction between Pt orbitals requires a stacked conformation within the cation. Solution behaviour of mononuclear Pd (1) and dinuclear Pd (2) compounds The 1H NMR spectrum of redissolved crystals of [(trpy)- Pd(Hmcyt-N3)]21 1 in D2O is more complicated than that of the Pt1 species 3 in that three sets of cytosine resonances are observed, and are assigned to 1, 2 and free Hmcyt (in equilibrium with H2mcyt1, depending on pH*).16 Mononuclear 1 is the dominant component in weakly acidic solution, consistent2334 J.Chem. Soc., Dalton Trans., 1999, 2329–2336 with the species distribution established by potentiometry and 1H NMR.16 The following equilibria exist: [(trpy)Pd(Hmcyt)]21 1 H2O [(trpy)Pd(H2O)]21 1 Hmcyt [(trpy)Pd(Hmcyt)]21 1 [(trpy)Pd(H2O)]21 [{(trpy)Pd}2(mcyt)]31 1 H3O1 As with 3, it is striking that the H5 and H6 signals of the Hmcyt nucleobase of 1 are downfield as compared to the (dien)PdII complex containing Hmcyt-N3 but no trpy ligands (Table 4).Fig. 6 Downfield regions of the 1H NMR spectra of 4 recorded in DMF[D7]: (a) 1D NMR spectrum, (b, c) portions of a 2D NOESY spectrum (c = 1.5 × 1022 M, tm 1.5 s). N4H reveals cross-peaks to H5 of mcyt and H59 of the N4 coordinated trpy ligand. In DMF[D7], 1 also equilibrates with 2 and free Hmcyt. Resonances of the amino protons of the Hmcyt ligand in 1 are observed at d 9.14 and 9.50.The assignment of the two N4H2 protons is similar to that of 3 (see above). The 1H NMR spectra of the dinuclear Pd species 2a and 2b in D2O are identical, as expected. There is no indication from 1H NMR spectroscopy that 2 undergoes any dissociation in water that leads to individual resonances. The 1H NMR spectrum of 2b in DMF[D7] (Fig. 8) diVers from the D2O spectrum and from that of 4 in DMF[D7] in the following points. First, there is dissociation of 2b in DMF[D7] according to [{trpy)Pd}2(mcyt)]31 1 H2O 1 DMF æÆ [(trpy)Pd(Hmcyt)]21 1 [(trpy)Pd(DMF)]21 1 OH2 with H1 for nucleobase protonation probably originating from water of crystallization.There is yet another minor species present containing cytosine, which will be discussed below. Second, the proton at N4 (d 7.40) occurs upfield in 2b as compared to 4. Third, the trpy proton pattern of 2b is somewhat diVerent from that of 4. Fourth, trpy resonances of 1 are observed around d 8.8. The second minor species present in DMF solution has been identified by TOCSY as a species containing cytosine with H5 and H6 doublets at d 6.99 and 7.89, respectively.The rather spectacular extra downfield shift of the H5 doublet strongly points to an anti orientation of the (trpy)PdII entity at N4.42,46 Fig. 7 195Pt edited 1H NMR spectrum of 4 (a) and a 2D 1H,195Pt HMQC spectrum (b) (DMSO[D6], c = 5.3 × 1022 M, 50 8C). The N4 coordinated Pt is identified by its coupling to N4H (2JPt,N4H ª 15 Hz) and H6 (5JPt,H6 ª 7 Hz) of the mcyt ligand.J.Chem. Soc., Dalton Trans., 1999, 2329–2336 2335 Whether the resonances are due to a dinuclear complex (II, Scheme 1) or a mononuclear species (III, Scheme 1) with the (trpy)PdII at N4 in an anti orientation, is not clear at present. However, a trinuclear species (IV), which would be analogous to a CH3HgII species previously reported,47 can be ruled out, since addition of [(trpy)Pd(DMF)]21 to a solution of 2 in DMF[D7] does not lead to an increase in the signal intensity of the resonances at d 6.99 and 7.89.Summary Mono- and dinuclear (trpy)MII (M = Pd, Pt) complexes containing the model nucleobase 1-methylcytosine have been prepared, X-ray structurally characterised and their solution behaviour studied by 1H NMR spectroscopy. The 1H NMR Fig. 8 Downfield region of the 1H NMR spectrum of 2 (DMF[D7], c = 1.8 × 1022 M) (a) and H5, H6 cross-peak section of 2D TOCSY spectrum (s = 1, 1 = anti conformer of 2) (b).chemical shifts of the cytosine resonances H5 and H6 do not match expectations based on known findings of the eVect of M entities carrying aliphatic amines or other nucleobases or simple ligands such as NH3, H2O, OH2, Cl2. Rather the eVect of the (trpy)M entity bound to N3 of Hmcyt in shifting the heteroaromatic protons of the nucleobases to lower field is more pronounced than that of a proton. This is obviously a consequence of the p-acceptor properties of the trpy ligand, which causes a significant deshielding of the H5 and H6 atoms.Binding of a second (trpy)MII entity to [(trpy)M(Hmcyt- N3)]21 does not require strongly alkaline conditions and starts even at acidic pH. It is accompanied by loss of a proton from the exocyclic amine group and a further downfield shift of H5 of the cytosine nucleobase in the 1H NMR spectrum. As pointed out there is circumstantial evidence only for a syn orientation of the two trpy ligands: neither Pt–Pt coupling nor clear-cut inter-residue NOE cross-peaks between the two trpy ligands in the dinuclear complexes are observed, and the NOE cross-peak between N4H and H5 of mcyt is ambiguous.On the other hand, the upfield shift of several of the trpy resonances in dinuclear 2 and 4 as compared to mononuclear 1 and 3 (at comparable concentrations) points to a stacked conformation rather than an anti orientation of the two (trpy)M entities, and the visible spectrum with its absorptions at 492 nm and 462 nm (sh) is likewise in agreement with such a structure. The solid state structures of 2b and 4b further confirm that a stacked conformation is possible, in principle.Acknowledgements This work has been supported by the Fonds der Chemischen Industrie (FCI) and the DAAD (fellowship for E. Farkas). References 1 (a) H.-K. Yip, C.-M. Che, Z.-Y. Zhou and T. C. W. Mak, J. Chem. Soc., Chem. Commun., 1992, 1369; (b) H.-K. Yip, L.-K. Cheng, K.-K.Cheung and C.-M. Che, J. Chem. Soc., Dalton Trans., 1993, 2933. 2 (a) J. A. Bailey, M. G. Hill, R. E. Marsh, V. M. Miskowski, W. P. Schaefer and H. B. Gray, Inorg. Chem., 1995, 34, 4591; (b) J. A. Bailey, V. M. Miskowski and H. B. Gray, Inorg. Chem., 1993, 32, 369. 3 M. G. Hill, J. A. Bailey, V. M. Miskowski and H. B. Gray, Inorg. Chem., 1996, 35, 4585 and refs. cited therein. 4 R. J. Mureinik and M. Bidani, Inorg. Chim. Acta, 1978, 29, 37. 5 (a) K. W. Jennette, J. T. Gill, J.A. Sadownick and S. J. Lippard, J. Am. Chem. Soc., 1976, 98, 6159; (b) A. H.-J. Wang, J. Nathans, Scheme 12336 J. Chem. Soc., Dalton Trans., 1999, 2329–2336 G. A. van der Marel, J. H. van Boom and A. Rick, Nature, 1978, 276, 471. 6 G. Arena, L. Monsú Scolaro, R. F. Pasternack and R. Romeo, Inorg. Chem., 1995, 34, 2994. 7 (a) H. M. Brothers II and N. M. Kostic, Biochemistry, 1990, 29, 7468; (b) H. M. Brothers II and N. M. Kostic, Inorg. Chem., 1988, 27, 1761. 8 T.G. Appleton, F. J. Pesch, M. Wienken, S. Menzer and B. Lippert, Inorg. Chem., 1992, 31, 4410. 9 E. M. A. Ratilla, B. K. Scott, M. S. Moxness and N. M. Kostic, Inorg. Chem., 1990, 29, 918. 10 A. K. Fazlur-Rahman and J. G. Verkade, Inorg. Chem., 1992, 31, 2064. 11 C.-C. Cheng and Y.-K. Lu, Chem. Commun., 1998, 253. 12 L. Monsú Scolaro, A. Romeo and A. Terracina, Chem. Commun., 1997, 1451. 13 M. Howe-Grant, K. C. Wu, W. R. Bauer and S. J. Lippard, Biochemistry, 1976, 15, 4339. 14 C.S. Peyratout, T. K. Aldridge, D. K. Crites and D. R. McMillin, Inorg. Chem., 1995, 34, 4484. 15 (a) G. Lowe, J. A. McCloskey, J. Ni and T. Vilaivan, Bioorg. Med. Chem., 1996, 4, 1007; (b) G. Lowe and T. Vilaivan, J. Chem. Soc., Perkin Trans. 1, 1996, 1499. 16 A. Kiss, E. Farkas, I. Sóvágó, B. Thormann and B. Lippert, J. Inorg. Biochem., 1997, 68, 85. 17 G. T. Morgan and F. H. Burstall, J. Chem. Soc., 1932, 20. 18 G. Annibale, M. Brandolisio and B. Pitteri, Polyhedron, 1995, 14, 451. 19 T. J. Kistenmacher, M. Rossi, J. P. Caradonna and L. G. Marzilli, Adv. Mol. Relax. Interact. Processes, 1979, 15, 119. 20 NONIUS BV, KappaCCD package, Röntgenweg 1, P.O. Box 811, 2600 AV Delft, Netherlands. 21 Z. Otwinowsky and W. Minor, DENZO and SCALEPACK, Methods Enzymol., 1997, 276, 307. 22 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467. 23 G. M. Sheldrick, SHELXTL-PLUS (VMS), Siemens Analytical X-Ray Instruments, Inc., Madison, WI, 1990. 24 G. M. Sheldrick, SHELXL-93, Program for crystal structure refinement, University of Göttingen, Germany, 1993. 25 G. M. Sheldrick, SHELXL-97, Program for the Refinement of Crystal Structures, University of Göttingen, Germany, 1997. 26 G. M. Intille, C. E. Pfluger and W. A. Baker, J. Cryst. Mol. Struct., 1973, 3, 47. 27 W. Zhang, C. Bensimon and R. J. Crutchley, Inorg. Chem., 1993, 32, 5808. 28 M. Krumm, I. Mutikainen and B. Lippert, Inorg. Chem., 1991, 30, 884. 29 B. Lippert, C. J. L. Lock and R. A. Speranzini, Inorg. Chem., 1981, 20, 808. 30 E. Sinn, C. M. Flynn, Jr. and R. B. Martin, Inorg. Chem., 1977, 16, 2403. 31 M. Rossi and T. J. Kistenmacher, Acta Crystallogr., Sect. B, 1977, 33, 3962. 32 M. Rossi, J. P. Caradonna, L. G. Marzilli and T. J. Kistenmacher, Adv. Mol. Relax. Interact. Processes, 1979, 15, 103. 33 A. Bondy, J. Phys. Chem., 1964, 68, 441. 34 W. Micklitz, W. S. Sheldrick and B. Lippert, Inorg. Chem., 1990, 29, 211. 35 J. A. Bailey, V. M. Miskowski and H. B. Gray, Acta Crystallogr., Sect. C, 1993, 49, 793. 36 J. A. Bailey and H. B. Gray, Acta Crystallogr., Sect. C, 1992, 42, 1420. 37 E. C. Constable, A. J. Edwards, G. R. Haire, M. J. Hannon and P. R. Raithby, Polyhedron, 1998, 17, 243. 38 F. P. Pruchnik, F. Robert, Y. Jeannin and S. Jeannin, Inorg. Chem., 1996, 35, 4261. 39 H. Schöllhorn, R. Beyerle-Pfnür, U. Thewalt and B. Lippert, J. Am. Chem. Soc., 1986, 108, 3680. 40 L. G. Marzilli, C.-H. Chang, J. P. Caradonna and T. J. Kistenmacher, Adv. Mol. Relax. Interact. Processes, 1979, 15, 85. 41 B. Lippert, Gazz. Chim. Ital., 1988, 118, 153. 42 J. Müller, E. Zangrando, N. Pahlke, E. Freisinger, L. Randaccio and B. Lippert, Chem. Eur. J., 1998, 3, 397. 43 C. Mealli, F. Pichierri, L. Randaccio, E. Zangrando, M. Krumm, D. Holthenrich and B. Lippert, Inorg. Chem., 1995, 34, 3418. 44 F. Pichierri, E. Chiarparin, E. Zangrando, L. Randaccio, D. Holthenrich and B. Lippert, Inorg. Chim. Acta, 1997, 264, 109. 45 T. V. O’Halloran and S. J. Lippard, Inorg. Chem., 1989, 28, 1289. 46 F. Pichierri, D. Holthenrich, E. Zangrando, B. Lippert and L. Randaccio, J. Biol. Inorg. Chem., 1996, 1, 439. 47 J.-P. Charland, M. Simard and A. L. Beauchamp, Inorg. Chim. Acta, 1983, 80, L57. Paper 9/02862E

ISSN:1477-9226    

DOI:10.1039/a902862e

出版商:RSC    

年代:1999

数据来源: RSC