摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3081�C3086 3081 Structural and electronic changes accompanying reduction of Cr(CO)4(bpy) to its radical anion: a quantum chemical interpretation of spectroelectrochemical experiments Stanislav Z�¢lis¡¦,a Chantal Daniel b and Anton�ªn Vlc¡¦ek, Jr. *a,c a J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejs¡¦kova 3, CZ-182 23 Prague, Czech Republic b Laboratoire de Chimie Quantique, UMR 7551 du CNRS et de l¡�Universit�¦ Louis Pasteur, 4, rue Blaise Pascal, Strasbourg, F-67008, France c Dept.of Chemistry, Queen Mary and Westeld College, Mile End Road, London, UK E1 4NS. E-mail: a.vlcek@qmw.ac.uk Received 5th May 1999, Accepted 16th June 1999 Optimised molecular structures and charge distributions within Cr(CO)4(bpy) and its radical anion were calculated using density functional theory (DFT). It was found that, although reduction predominantly concerns the bpy ligand, its structural and electronic eects extend to the Cr(CO)4 fragment.Each equatorial and axial CO ligand was calculated to accept 7.1 and 4.8%, respectively, of the extra electron density in Cr(CO)4(bpy). This is in accordance with the IR spectroelectrochemical results which show that the corresponding CO stretching force constants decrease by 68 and 21 N m1, respectively. The calculated spin density in Cr(CO)4(bpy) resides predominantly on the bpy ligand which behaves spectroscopically as bpy. The spin density is delocalised to both axial and equatorial pairs of CO ligands by mixing of ¦�*(C O) orbitals with the, predominantly ¦�*(bpy), SOMO. In addition, part of the spin density is delocalised selectively to the axial CO ligands by an admixture of their ¦Ò orbitals into the SOMO.This ¦Ò�C¦�* contribution is responsible for isotropic EPR hyperne splitting which was observed from the axial 13C(CO) atoms only. Accordingly, the isotropic hyperne splitting constants correlate with calculated Fermi contact terms instead of total spin densities.Complete active space self-consistent eld (CASSCF)-calculated changes in charge distribution upon a Cr¡úbpy MLCT excitation show that the electron density localised on the bpy ligand increases by about the same amount upon reduction or MLCT-excitation of Cr(CO)4(bpy). The axial CO ligands are depopulated by MLCT excitation ca. 1.6 times more than the equatorial ones.These conclusions can be generalised and applied to other coordination and organometallic complexes of low-valent metals which contain a reducible or radical-anionic ligand. Introduction The Cr(CO)4(bpy) complex is widely studied as a prototype of an inert organometallic molecule which can be activated by an electron transfer or metal to ligand charge transfer (MLCT) excitation.1�C15 Both one-electron reduction and irradiation into the Cr¡úbpy MLCT absorption band labilise the axial Cr�CCO bond, making the CO ligand susceptible to a facile substitution by a solvent molecule or by a Lewis base, L (e.g.phosphine) present in solution (Scheme 1). The reductive activation,8 described by Scheme 1, eqn. 1(a)�C (c), amounts to an electron transfer catalysed reaction since the radical-anionic product reacts with the starting molecule, regenerating the active Cr(CO)4(bpy) species [Scheme 1, eqn. 1(c)]. Both the reactive MLCT excited state and Cr(CO)4(bpy) involved in the photochemical and reductive activation, respectively, contain a coordinated bpy radical anion.Hence, some specic interaction between the Cr(CO)4 moiety and the singly occupied ¦�* orbital of the bpy ligand Scheme 1 can be suspected of labilising the axial Cr�CCO bonds of Cr(CO)4(bpy). However, the analogy 16 between reduction and MLCT excitation has obvious limitations manifested by the fact that photochemical CO dissociation is an ultrafast process occurring in <400 fs while the Cr(CO)4(bpy) radical anion reacts on a time scale of minutes.7,8 Recently, we have studied the localisation of the extra electron in Cr(CO)4(bpy) by a combination of IR and EPR spectroelectrochemistry 9 while the MLCT excitation was investigated by resonance Raman spectroscopy.10 Spectroelectrochemical studies took advantage of a 13CO isotopic enrichment which allowed us to determine the EPR hyperne interaction with the C atoms of the CO ligands and the changes in the force constants of the stretching vibrations of axial and equatorial CO ligands. EPR spectra of Cr(CO)4(bpy) and its 13CO-enriched isotopomers have shown that the extra electron is predominantly localised on the bpy ligand.Importantly, a hyperne interaction was found to occur only with the axial 13C atoms, the equatorial ones being EPR silent. Simpleminded interpretation of this observation would indicate much larger spin density on the axial than equatorial CO ligands.Seemingly in contradiction, reduction of Cr(CO)4(bpy) causes the stretching force constant of the equatorial CO ligands to decrease about three times more than that of the axial CO¡�s, suggesting that the ¦�-back donation to the equatorial CO ligands of the Cr(CO)4(bpy) radical anion is much stronger than to the axial ones. Experimental results thus demonstrated that the reduction of the bpy ligand in Cr(CO)4(bpy) has a3082 J. Chem. Soc., Dalton Trans., 1999, 3081–3086 profound in.uence on the bonding within the Cr(CO)4 fragment. At the same time, the IR and EPR data showed that the axial and equatorial CO ligands are a.ected by very di.erent mechanisms.Herein, we report results of DFT calculations of the electronic and molecular structure of Cr(CO)4(bpy), and its radical anion, which reveal how bpy-localised reduction a.ects the electron density distribution and bonding within the Cr(CO)4 moiety. Moreover, a comparison of these results with the CASSCF wave function of the dxz.p*(bpy) 1bA1 MLCT excited state 6,11 is used to compare the electronic e.ects of reduction and MLCT excitation.Using Cr(CO)4(bpy) as a prototypical example, this study addresses more general questions of the localisation of redox changes in complexes containing redoxactive ligands, the role of ‘spectator’ ligands, and relations between the extent of ligand reduction caused by electron addition or MLCT excitation.It has been found that neither the reduction nor MLCT excitation can be regarded as localised solely on the acceptor ligand or the metal–ligand moieties, respectively, since the bonding within the remaining part of the coordination sphere is strongly a.ected as well. Calculations Gaussian 98 17 and MOLCAS18 program packages were used for DFT and CASSCF calculations, respectively. Density functional theory (DFT) was employed to calculate the ground state electronic structure of Cr(CO)4(bpy) and its radical anion.DFT calculations of the latter were spin unrestricted. Calculations were performed within the constraint of C2v symmetry, the bpy ligand being located in the xy plane, with the C2 symmetry axis coincident with the x axis (Fig. 1). B3LYP hybrid functionals 19 were used. Cr valence double-. plus polarisation basis set designed for DFT calculations was taken from Godbout et al.20 Dunning’s 21 valence double-. with polarisation functions were used for C, N and H atoms (= basis set I) for geometry optimisation, calculations of charge density and spin distribution, and Fermi contact terms.Isotropic hyper.ne coupling constants were calculated at the optimised geometry. In order to describe the e.ect of basis variation on the spin density distribution and Fermi terms, several di.erent basis sets were used for C, N and H atoms within single point calculations: double-. wave functions with polarisation functions of Adamo and Barone22 designed for EPR calculations (EPR-II basis, = basis set II), and 6-311G**23,24 (= basis set III).The description of lowest excited states of Cr(CO)4(bpy) is based on CASSCF calculations. Here, the generally contracted atomic natural orbital (ANO) type basis sets were used: for the .rst-row atoms a (or hydrogen atoms a (7) set contracted to (2). These calculations were performed using an idealised molecular geometry taken from ref. 15.Results Molecular geometry Calculated important bond lengths and angles of Cr(CO)4(bpy) and its reduced form are summarised in Table 1. The reduction a.ects mainly the geometry of the bpy ligand. The lengthening Fig. 1 The Cr(CO)4(bpy) molecule and chosen orientation of axes. N N Cr CO CO OC OC z x y 2 3 4 5 6 2' 3' 4' 5' 6' of the N1–C2 and shortening of the C2–C2 bonds are the most signi.cant changes. The Cr–N bonds are elongated. To a lesser extent, reduction also in.uences the calculated structure of the Cr(CO)4 moiety.Addition of an electron is accompanied by contraction of Cr–CO bonds and elongation of C–O bonds. Bond length changes within the equatorial Cr(CO)2 fragment are about twice as large as those in the axial OC–Cr–CO moiety. Charge distribution Table 2 summarises calculated charges on the Cr atom, bpy, axial and equatorial CO ligands in Cr(CO)4(bpy) and its radical anion. It follows that the extra electron density in the latter is mostly localised on the bpy ligand.However, a signi.cant charge delocalisation to the CO ligands occurs. The equatorial CO ligands are better electron acceptors than the axial ones in both the neutral and anionic forms of Cr(CO)4(bpy), as is manifested by much larger negative charges. Upon reduction, the equatorial pair of CO ligands accepts about 14.2% of the extra electron density in the radical anion while the two axial CO ligands accommodate only 9.6% (Table 2).The charge on the Cr atom is virtually unchanged by the reduction. The change in the charge distribution in Cr(CO)4(bpy) upon vertical excitation to the b1A1 MLCT excited state,6 which originates predominantly in a dxz.p*(bpy) excitation, was calculated at a CASSCF level and results are summarised in Table 3. The MLCT excitation increases the electron density at the bpy ligand by 79% cf. a value of 76% (Table 2) calculated for the reduction. Molecular orbitals DFT-calculated one-electron orbital energies of Cr(CO)4(bpy) and its radical anion are qualitatively depicted in Fig. 2. The Table 1 Selected DFT calculated bond lengths (Å) and angles ( ) of Cr(CO)4(bpy) and its radical anion Cr(CO)4(bpy) Cr(CO)4(bpy) Change on reduction Cr–N Cr–Cax Cr–Ceq N1–C2 C2–C3 C3–C4 C4–C5 C5–C6 C6–N1 C2–C2 (C–O)ax (C–O)eq Cr–N1–C2 Cr–N1–C6 N1–C2–C2 Cax–Cr–Cax Ceq–Cr–Ceq N–Cr–N 2.127 1.915 1.861 1.361 1.404 1.394 1.401 1.394 1.350 1.476 1.163 1.170 117.5 124.6 114.6 173.6 93.1 75.7 2.147 1.909 1.847 1.394 1.431 1.380 1.423 1.396 1.342 1.427 1.168 1.179 115.6 126.0 116.1 176.5 92.6 76.4 0.020 0.006 0.014 0.033 0.025 0.014 0.022 0.002 0.008 0.049 0.005 0.009 1.9 1.4 1.5 2.9 0.5 0.7 Table 2 Calculated charges (e) on subsystems of Cr(CO)4(bpy) and its radical anion Cr(CO)4(bpy) Cr(CO)4(bpy) Change on reduction Cr bpy (CO)ax (CO)eq 0.043 0.263 0.012 0.098 0.051 0.493 0.061 0.169 0.008 0.756 0.048 a 0.071 b a Out of this, the charge on the O atom changes by 0.046.b Out of this, the charge on the O atom changes by 0.056.J. Chem. Soc., Dalton Trans., 1999, 3081–3086 3083 a1, a2 and b2 set of the highest occupied molecular orbitals (HOMOs) of both the neutral and anionic species have a large contribution from the Cr dx2 y2, dxy and dxz orbitals, respectively. The b2 HOMO orbital is shown in Fig. 3. It has ca. 50% dxz character in both species. The bpy ligand in the neutral and anionic species contributes by 23.5 and 21%, respectively.Each axial and equatorial CO ligand contributes to the b2 HOMO of Cr(CO)4(bpy) by 8.75 and 4.20%, respectively. These contributions increase to 9.65 and 5.01%, respectively, upon reduction. For symmetry reasons, the equatorial CO ligands contribute to the b2 HOMO only by their �*(pz) orbitals. The axial CO ligands participate in the b2 HOMO by both their �*(px) and ó(s, pz) orbitals. The � contribution predominates.It amounts to 8.09 and 8.41% per axial CO in the neutral and anionic species, respectively. The ó-contribution per axial CO increases from 0.66 to 1.24% on going from Cr(CO)4(bpy) to the radical anion. (Note that the out-of-phase combination of ó orbitals of the two axial CO ligands, ó1 ó2, belongs to the same b2 symmetry representation of the C2v point group of the Cr(CO)4(bpy) molecule as the lowest �*(bpy) orbital, allowing for a ó–�* mixing in both HOMO and LUMO.) Fig. 2 Qualitative MO schemes of the Cr(CO)4(bpy) (left) and Cr(CO)4(bpy) (right) complexes based on DFT calculations. The HOMOs of the neutral and anionic complexes are set at the same energy value. Fig. 3 The composition of the b2 HOMO orbital of Cr(CO)4(bpy). Table 3 The CASSCF-calculated charge distribution in the ground state and b1A1 MLCT excited state of Cr(CO)4(bpy). Calculations were performed with 10 correlated electrons in 11 active orbitals G.S. E.S. Di.erence Cr bpy (CO)ax (CO)eq 0.062 0.219 0.001 0.139 0.419 0.570 0.132 0.057 0.357 0.789 0.131 0.082 The b2 lowest lying molecular orbital (LUMO) of Cr(CO)4- (bpy) is half-.lled during the reduction process, becoming the singly occupied molecular orbital (SOMO) of Cr(CO)4(bpy). The plot of the b2 SOMO of Cr(CO)4(bpy), depicted in Fig. 4, and the fragment contributions (Table 4) clearly show that the SOMO is composed mainly of the .rst antibonding �* MO of the bpy ligand. The contribution from the Cr atom decreases from 2.52 to 1.94% on going from the neutral complex to the anion. The SOMO is partially delocalised onto the CO ligands of the Cr(CO)4 moiety, the contribution from the axial CO ligands being almost twice as large as that from the equatorial ones (Table 4).The mechanisms of SOMO delocalisation on the axial and equatorial CO ligands are, however, very di.erent. The SOMO plot (Fig. 4) demonstrates that the equatorial CO ligands lie in the SOMO nodal plane.Hence, they participate in the SOMO only by their �*(pz) orbitals, 0.79% each. Analogously, the �*(px) orbital of each axial CO ligand participates by 0.54%. More interestingly, the calculations show that the ó-orbital of each axial CO ligand is admixed to the SOMO by 0.86%. Speci.cally, it is the ó1 ó2 out-of-phase combination of the lone electron pairs of the two axial CO ligands which contributes to the SOMO by 1.72% (i.e. 0.86% per axial CO). The axial Cr 4pz orbital contributes very little, by 0.078% only.Spin densities and EPR spectra The calculated distribution of total spin density in Cr(CO)4- (bpy) is shown in Fig. 5 and summarised in Table 5, which also compares the e.ect of the basis set on the calculated spin densities and Fermi contact terms. Although there are quantitative di.erences, the choice of the basis set does not a.ect the qualitative conclusions outlined below. The calculated total spin density on C atoms of the axial and equatorial CO ligands is about the same, while the contributions due to the unpaired electron in the SOMO di.er signi.cantly: 1.16% on each axial C atom and 0.43% one each equatorial one (Table 4). Fig. 5 Fig. 4 The composition of the b2 SOMO orbital of Cr(CO)4(bpy). The ó-contribution to the SOMO is clearly visible along the axial Cr–CO bond. Table 4 Calculated contributions (%) of subsystems of Cr(CO)4(bpy) and its radical anion to the redox orbital, i.e. the LUMO and SOMO, respectively Cr(CO)4(bpy) Cr(CO)4(bpy) Change on reduction Cr bpy (CO)ax (CO)eq 2.52 92.14 1.69 0.98 1.94 93.74 1.40 a 0.79 b 0.58 1.60 0.29 0.19 a The ó component accounts for 0.86%, with a 0.84% participation of the 2pz and 2s orbitals at each C atom. The � component accounts for 0.54% with a 0.32% participation of the C 2px orbital. b �-Component only.The C 2pz orbital contributes by 0.43%.3084 J. Chem. Soc., Dalton Trans., 1999, 3081–3086 Table 5 Experimental EPR hyper.ne splitting constants (hfs) and calculated isotropic Fermi contact couplings (104 cm1is sets Experiment Basis set I Basis set II Basis set III Fermi contact Spin Fermi contact Spin Fermi contact Spin Atom hfs term density term density term density 53Cr 14N 1H3,3 1H4,4 1H5,5 1H6,6 13Cax 13Ceq 1.197 3.469 0.982 1.159 4.216 0.701 5.619 — 0.677 4.065 1.590 1.090 4.975 0.732 4.926 0.252 0.0142 0.1557 0.0040 0.0031 0.0139 0.0020 0.0078 0.0070 0.853 3.387 1.164 0.792 4.503 0.405 5.760 0.557 0.0402 0.1373 0.0045 0.0022 0.0136 0.0015 0.0136 0.0123 0.682 2.520 1.310 0.823 4.731 1.318 5.627 0.037 0.0096 0.1662 0.0032 0.0025 0.0138 0.0031 0.0066 0.0063 also demonstrates that the spin density drops to zero in the horizontal (x,y) symmetry plane of the molecule.On the other hand, a large part of the spin density on the axial C atoms is oriented along the z axis, i.e. the axial C–Cr–C s-bond. Calculated Fermi contact terms, which amount to the theoretical values of isotropic EPR hyper.ne splitting constants, compare well with the experimental values (Table 5).Notably, the experimental EPR spectra of 13CO-enriched Cr(CO)4(bpy) revealed only splitting from the two axial 13C atoms, splitting from the equatorial ones being too small to be resolved. Indeed, the calculated Fermi contact term is much larger for the axial 13C(CO) atom than for its equatorial counterpart, despite nearly identical values of the total spin density.Discussion In agreement with all available experimental data, our DFT calculations picture the reduction of Cr(CO)4(bpy) as predominantly bpy-localised but strongly a.ecting the bonding within the Cr(CO)4 fragment. Speci.cally, it was calculated that the bpy ligand accommodates ca. 76% of the extra electron density in Cr(CO)4(bpy). The unpaired electron is localised on the bpy ligand to an extent of 94%, in the b2 SOMO which is, essentially, a p*(bpy) orbital.The calculated bond length changes within the Cr(bpy) fragment re.ect the respective p bonding and antibonding properties of the b2 SOMO. It follows that Cr(CO)4(bpy) can be formally viewed as a d6 chromium(0) complex, with a radical anionic ligand bpy. Accordingly, experimental UV–VIS absorption 3,6,25,26 and EPR9 spectra of Cr(CO)4(bpy) resemble closely those of bpy, being only slightly perturbed by interaction with the Cr(CO)4 moiety.The remaining 24% of the extra electron density in Cr(CO)4(bpy) resides on the CO ligands. The charge on the Cr atom is almost the same in the neutral and anionic complexes. Reduction involves addition of an electron to the b2 LUMO of Cr(CO)4(bpy). This redox orbital becomes singly occupied (SOMO) in the Cr(CO)4(bpy) radical anion, without any signi.cant change in composition. The SOMO is localised on the bpy and the four CO ligands from ca. 94 Fig. 5 The spin density distribution in Cr(CO)4(bpy).and 4%, respectively. Comparison between the charge and SOMO distribution reveals that the electronic e.ects of Cr(CO)4(bpy) reduction are much more delocalised than the SOMO of the radical anion. Namely, the electron density in lower-lying occupied orbitals is redistributed toward the CO ligands, apparently compensating for increased electron donation from the bpy ligand. The two equatorial CO ligands accept much more (14.2%) of the extra electron density in the radical anion than the axial CO ligands, which accommodate 9.6%.This computational .nding accounts well for the experimentally observed changes of the CO stretching force constants which decrease upon reduction by 68 and 21 N m1 for the equatorial and axial CO ligands, respectively.9,10 This weakening of the equatorial and, to a lesser extent, axial C O bonds is manifested also by calculated elongations of C–O bonds, accompanied by shortenings of Cr–CO bonds, on going from the neutral species to the radical anion (Table 1).These structural e.ects are much larger for the equatorial Cr–CO units than for the axial ones. The p back bonding to the equatorial CO ligands is stronger than to the axial ones in both the neutral and anionic forms of Cr(CO)4(bpy). Mixing between an occupied s-non-bonding orbital of the axial OC–Cr–CO fragment with the p*(bpy) orbital in the b2 LUMO and SOMO of Cr(CO)4(bpy) and its radical anion, respectively, is another interesting feature emerging from the DFT calculations.Its contribution is about the same, ca. 1.7%, in both the neutral and anionic complexes. This s–p* interaction explains straightforwardly the huge di.erence in the EPR hyper.ne splitting from 13C nuclei of axial and equatorial CO ligands. Table 5 shows that the isotropic EPR hyper.ne splitting constant is much larger for the axial than equatorial 13C(CO) nuclei despite nearly identical spin densities calculated at both positions.This is caused by the fact that the isotropic EPR splitting constants are determined by corresponding Fermi contact terms,27–29 that is by the spin density at the 13C(CO) nuclei due to electrons in s orbitals. Relatively large observed 9 hyper.ne splitting, 6.1 G, from axial 13C nuclei originates in the axial s–p* interaction which mixes the axial C 2s orbital into the SOMO, giving rise to a substantial Fermi contact. On the other hand, no hyper.ne splitting was observed from the equatorial 13C(CO) nuclei since they lie in the SOMO nodal plane, without any direct s-orbital participation.Hyper- .ne splitting from equatorial 13C atoms could arise only from a spin polarisation of C 1s and 2s electrons by the unpaired electron density in the C 2p p* orbital. Data in Table 5 clearly show that such a spin polarisation produces Fermi contact term values which are 10–150 times smaller than those of the axial 13C atoms for which the s–p* mechanism is available.The above analysis of the EPR hyper.ne splitting from the 13C(CO) nuclei points to two conclusions: (i) the EPR hyper.ne splitting from the 13C donor atoms of the CO ligands in Cr(CO)4(bpy) ligand re.ects the extent of s–p* delocalisation, instead of the total spin-density distribution, and (ii) the direct admixture of the C 2s orbital into the SOMO by s–p* delocalisation contributes toward the isotropic 13C(CO) hyper.ne splitting constantJ. Chem.Soc., Dalton Trans., 1999, 3081�C3086 3085 at least 10 times more than the spin polarisation of the C 1s and 2s electrons by the unpaired electron density in the C 2p ¦�* orbital. These conclusions may be extended to the whole family of transition metal complexes with a radical-anionic ligand which often show large EPR hfs from the nuclei of donor atoms in the cis-position.30�C34 This eect was traditionally attributed to a ¡®¦Ò�C¦� hyperconjugation¡�, a term that is often used in a qualitative sense but seldom analysed in detail and explained in terms of specic orbital interactions.28�C32,35 Similar ¦Ò�C¦�* interaction was recently found to occur in ¦Á- diimine complexes containing axial ligands which are bound to the metal atom by high-lying ¦Ò-orbitals, e.g.metal fragments or alkyls.36�C39 The complexes Ru(SnPh3)2(CO)2(¦Á-diimine) or PtMe4(¦Á-diimine) are typical examples. The extent of the ¦Ò�C¦�* interaction in these complexes is, however, much greater than in Cr(CO)4(bpy)0/, ranging from 14 to 27%.It appears to be the most important factor determining the ground state structural, spectroscopic and electrochemical properties, as well as the reactivity, emission and lifetimes of the excited states of dinuclear or alkyl complexes with an ¦Á-diimine ligand in a cis position. Because of its much smaller extent in the Cr(CO)4- (bpy)0/ complexes studied herein, the ¦Ò�C¦�* interaction does not have such dramatic consequences.Nevertheless, its presence is clearly manifested experimentally by the large EPR hyperne splitting from the axial 13C(CO) atoms. Parallels are often drawn between MLCT excited states and reduced forms of low-valent metal complexes with reducible ligands.16 Namely, the resemblance of the spectroscopic (UV�C VIS, IR, Raman) properties of an excited molecule to erated reduced form are deemed diagnostic for the MLCT character of the excited state in question. Indeed, the calculations reported above show that the spectroscopically 6 and photochemically 3,4,11 important ¡®dxz¡ú¦�*(bpy)¡� MLCT excitation increases the electron density at the bpy ligand to the same extent as reduction, by 79 and 76%, respectively.It follows that the charge transfer in the MLCT state is incomplete, partly compensated for by an opposite drift in the electron density distribution.40 Contrary to a conventional view, the ¡®hole¡� created by excitation is delocalised over the whole Cr(CO)4 fragment, instead of being conned to the Cr atom.In fact, the electron density on the Cr atom decreases by 0.36, less than on the four CO ligands which are depopulated, on the total, by 0.426 e. Axial CO ligands are depopulated ca. 1.6 times more than their equatorial counterparts. This computational result agrees with the conclusions based on the Raman spectra measured in resonance with the lowest allowed MLCT transition.10 The selective enhancement of the Raman band due to the in-phase totally symmetric CO stretching vibrations implies that the axial C O bonds are elongated upon the MLCT excitation 1.2�C3.0 times more than the equatorial ones.This eect has been ascribed to the depopulation of the Cr dxz orbital which diminishes the ¦�-back bonding to the axial CO ligands more than to the equatorial ones. Moreover, the observed strong intensity enhancement of Raman bands due to the bpy vibrations and deformation vibrations of the Cr(bpy) chelate ring agrees well with the calculated charge redistribution between the Cr atom and bpy ligand.Conclusions The DFT calculations on Cr(CO)4(bpy) and its radical anion reproduce well the experimentally observed spectroscopic eects of Cr(CO)4(bpy) reduction and enable us to explain them on a structural and electronic basis. The reduction is predominantly localised on the bpy ligand, its structural and electronic eects spreading over the Cr(CO)4 moiety, mostly the CO ligands.The structural changes of the Cr(bpy) fragment and principal features displayed by electronic absorption and EPR spectra of Cr(CO)4(bpy) are mostly accounted for by the properties of the b2 SOMO, which is 94% ¦�*(bpy) in character. The Cr(CO)4(bpy) thus spectroscopically behaves as containing a bpy ligand. Analysis of the SOMO alone is insucient to describe changes in the overall charge distribution in Cr(CO)4(bpy) upon its reduction.Electronic eects of the reduction are much more delocalised than the SOMO. In particular, the bpy-localised reduction is accompanied by an electron density redistribution toward the equatorial and, to a lesser extent, axial CO ligands. This conclusion explains the experimental observation that the equatorial CO stretching force constant decreases on reduction three times more than the equatorial one. Two principal mechanisms are responsible for the spin density delocalisation to the CO ligands: (i) a direct participation of ¦�*(CO) orbitals of both equatorial and axial CO ligands in the SOMO, and (ii) an admixture of axial ¦Ò(CO) orbitals to the SOMO.The latter mechanism operates only for axial CO ligands. It amounts to a ¦Ò�C¦�* interaction between an orbital that has a ¦Ò character with respect to the axial OC�CCr�CCO bond and the ¦�*(bpy) orbital. It provides the mechanism for a selective increase of spin density in the 1s and 2s orbitals of the axial C atoms.This is experimentally manifested by the isotropic 13C(CO) EPR hyperne splitting constants which are relatively large for the axial 13C(CO) but unobservable for the equatorial ones. The virtual contradiction between the IR and EPR spectroelectrochemical results 9 was resolved: the IR data, i.e. the changes in the CO stretching force constants on reduction, reect the absolute changes of the total electron density on the CO ligands, which are much larger in the equatorial than axial positions.On the other hand, isotropic 13C EPR hyperne splitting constants of the CO ligands in Cr(CO)4(bpy) do not reect the total spin densities at respective 13C nuclei but the mechanism of the spin delocalisation, namely the extent of the ¦Ò�C¦�* interaction. Obviously, the isotropic hyperne splitting constants need not to be a good measure of the total spin density distribution over the donor atoms of ¡®spectator¡� ligands in Cr(CO)4(bpy) and analogous complexes with a radical anionic ligand.30,31,33,34 The amount of the extra electron density localised on the bpy ligand in a MLCT excited state of Cr(CO)4(bpy) is about the same as in Cr(CO)4(bpy), underlying the spectroscopic analogy between MLCT and reduced states of complexes containing reducible ligands.MLCT excitation was found to diminish the electron density on the axial and, to a lesser extent, equatorial CO ligands.Obviously, the, so called MLCT excitation, concerns the whole Cr(CO)4(bpy) molecule and not only the Cr(bpy) fragment, as is usually assumed. The results and conclusions discussed above can easily be extended to other coordination and organometallic complexes of low-valent metals which contain a reducible or radicalanionic ligand. Acknowledgements This work was undertaken as a part of the European collaborative COST projects D4/0001/94 and D14/0001/99. Financial support from the Ministry of Education of the Czech Republic (OC.D4.20 and OC.D14.20) and from the Granting Agency of the Czech Republic (203/97/1048) is gratefully appreciated.References 1 D. J. Stufkens, Coord. Chem. Rev., 1990, 104, 39. 2 R. W. Balk, T. Snoeck, D. J. Stufkens and A. Oskam, Inorg. Chem., 1980, 19, 3015. 3 J. V�ªchov�¢, F. Hartl and A. Vlc¡¦ek, Jr., J. Am. Chem. Soc., 1992, 114, 10903.3086 J. Chem. Soc., Dalton Trans., 1999, 3081.3086 4 A. Vlc¢§ek, Jr., J. Vichova and F.Hartl, Coord. Chem. Rev., 1994, 132, 167. 5 I. G. Virrels, M. W. George, J. J. Turner, J. Peters and A. Vlc¢§ek, Jr., Organometallics, 1996, 15, 4089. 6 D. Guillaumont, C. Daniel and A. Vlc¢§ek, Jr., Inorg. Chem., 1997, 36, 1684. 7 I. R. Farrell, P. Matousek and A. Vlc¢§ek, Jr., J. Am. Chem. Soc., 1999, 121, 5296. 8 D. Miholova and A. A. Vlc¢§ek, J. Organomet. Chem., 1985, 279, 317. 9 A. Vlc¢§ek, Jr., F. Baumann, W. Kaim, F.-W. Grevels and F. Hartl, J. Chem. Soc., Dalton Trans., 1998, 215. 10 A.Vlc¢§ek, Jr., F.-W. Grevels, T. L. Snoeck and D. J. Stufkens, Inorg. Chim. Acta, 1998, 278, 83. 11 D. Guillaumont, C. Daniel and A. Vlc¢§ek, Jr., manuscript in preparation. 12 S. Wieland, K. B. Reddy and R. van Eldik, Organometallics, 1990, 9, 1802. 13 W.-F. Fu and R. van Eldik, Inorg. Chem., 1998, 37, 1044. 14 W. F. Fu and R. van Eldik, Inorg. Chim. Acta, 1996, 251, 341. 15 H. Kobayashi, Y. Kaizu, H. Kimura, H. Matsuzawa and H. Adachi, Mol. Phys., 1988, 64, 1009. 16 A. Vlc¢§ek, Jr., Chemtracts.Inorg. Chem., 1993, 5, 1. 17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Cli.ord, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A.D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople, Gaussian 98, Revision A.6, Gaussian, Inc., Pittsburgh, PA, 1998. 18 K. Andersson, M. R. A. Blomberg, M.P. Fluscher, G. Karstrom, V. Kello, R. Lindh, P.-A. Malmqvist, J. Noga, J. Olsen, B. O. Roos, A. J. Sadlej, P. E. M. Siegbahn, M. Urban and P.-O. Widmark, Lund, Sweden, 1994. 19 A. D. Becke, J. Chem. Phys., 1993, 98, 5648. 20 N. Godbout, D. R. Salahub, J. Andzelm and E. Wimmer, Can. J. Chem., 1992, 70, 560. 21 T. H. Dunnning and P. J. Hay, in Modern Theoretical Chemistry, ed. H. F. Schaefer, III, Plenum, New York, 1976. 22 C. Adamo and V. Barone, Chem. Phys. Lett., 1997, 274, 242. 23 W. J. Hehre, R. Ditch.eld and J. A. Pople, J. Chem. Phys., 1972, 56, 2257. 24 T. Clark, J. Chandrasekhar and P. v. R. Schleyer, J. Comput. Chem., 1983, 4, 294. 25 H. Saito, J. Fujita and K. Saito, Bull. Chem. Soc. Jpn., 1968, 41, 863. 26 H. Saito, J. Fujita and K. Saito, Bull. Chem. Soc. Jpn., 1968, 41, 359. 27 B. A. Goodman and J. B. Raynor, Adv. Inorg. Chem. Radiochem., 1970, 13, 135. 28 M. Symons, Chemical and Biochemical Aspects of Electron-Spin Resonance Spectroscopy, Van Nostrand Reinhold Company Ltd., Wokingham, England, 1978. 29 J. E. Wertz and J. R. Bolton, Electron Spin Resonance. Elementary Theory and Practical Applications, Chapman and Hall, New York, 1986. 30 W. Kaim, Coord. Chem. Rev., 1987, 76, 187. 31 W. Kaim and S. Kohlmann, Inorg. Chem., 1990, 29, 2909. 32 W. Kaim, Inorg. Chem., 1984, 23, 3365. 33 A. Klein, C. Vogler and W. Kaim, Organometallics, 1996, 15, 236. 34 F. Hartl and A. Vlc¢§ek, Jr., Inorg. Chem., 1996, 35, 1257. 35 W. Kaim, B. Olbrich-Deussner, R. Gross, S. Ernst, S. Kohlmann and C. Bessenbacher, in Importance of Paramagnetic Organometallic Species in Activation, Selectivity and Catalysis, eds. M. Chanon, M. Julliard and J.-C. Poite, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1989. 36 M. P. Aarnts, D. J. Stufkens, M. P. Wilms, E. J. Baerends, A. Vlc¢§ek, Jr., I. P. Clark, M. W. George and J. J. Turner, Chem. Eur. J., 1996, 2, 1556. 37 M. P. Aarnts, M. P. Wilms, K. Peelen, J. Fraanje, K. Goubitz, F. Hartl, D. J. Stufkens, E. J. Baerends and A. Vlc¢§ek, Jr., Inorg. Chem., 1996, 35, 5468. 38 W. Kaim, A. Klein, S. Hasenzahl, H. Stoll, S. Zalis¢§ and J. Fiedler, Organometallics, 1998, 17, 237. 39 S. Hasenzahl, H.-D. Hausen and W. Kaim, Chem. Eur. J., 1995, 1, 95. 40 B. S. Brunschwig, C. Creutz and N. Sutin, Coord. Chem. Rev., 1998, 177, 61. Paper 9/03560E

ISSN:1477-9226    

DOI:10.1039/a903560e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3087–3093 3087 Extended multidecker sandwich architecture of Cs–18-crown-6 complexes stabilized in the environment of novel large iodocuprate(I) clusters obtained from zerovalent copper Konstantin V. Domasevitch,a Julia A. Rusanova,a Olga Yu. Vassilyeva,*a Vladimir N. Kokozay,a Philip J. Squattrito,b Joachim Sieler c and Paul R. Raithby *d a Department of Inorganic Chemistry, Shevchenko Kyiv University, Volodimirska St. 64, Kyiv 252033, Ukraine.E-mail: vassil@inorg.chem.univ.kiev.ua or kokozay@chem.kiev.ua b Department of Chemistry, Central Michigan University, Mt. Pleasant, MI 48859, USA c Institut für Anorganische Chemie, Universität Leipzig, Linnéstraße 3, D-04103 Leipzig, Germany d Department of Chemistry, Lens.eld Road, Cambridge, UK CB2 1EW. E-mail: prr1@cus.cam.ac.uk Received 24th April 1999, Accepted 12th July 1999 The interaction of zerovalent copper with a proton-donating agent (NH4I) in acetonitrile solutions of CsX (X = Cl or I) and 18-crown-6 (18c6), in air, resulted in the formation of the novel mixed-metal complexes [Cs(18c6)2][Cu5I6(MeCN)2] 1, [Cs2(18c6)3][Cu8I10(MeCN)2] 2 and [{Cs(18c6)}6Cu4I7][Cs3(18c6)3][Cu7I10][Cu13I14]I2 3.The main structural feature of the compounds is the multidecker sandwich Cs–18c6 cations and .nite iodocuprate(.) anions combined to form ionic lattices. It is proposed that the Cu oxidation state is stabilized in the presence of the Cs–18c6 moieties in acetonitrile, and the overall formation of the solid phase is a template process of mutual stabilization of Cs as bulky [Csn(18c6)m]n cations and Cu as iodocuprate anions.Introduction Employing zerovalent metals in the synthesis of co-ordination compounds has proven to be a fascinating and versatile route to various metal complexes and there is still great scope for novel chemical and structural investigations of such systems.1 In studying the interaction of metal powders, in particular copper and nickel, with non-aqueous solutions of ammonium salts we showed that metal oxidation and complex formation were conditioned by the presence of a proton-donating agent and dioxygen from the air via reaction (1) in which copper oxidation M 2NH4 ��� O2 ..M2 H2O 2NH3 (1) occurred in the two-step process Cu .. Cu .. Cu2.2 The concept of necessity of a proton-donating agent in syntheses of co-ordination compounds from zerovalent metals in air was further developed into a methodology for the preparation of mixed-metal complexes.The strategy consisted of treating copper powder with a salt of another metal in a non-aqueous solution of an aminoalcohol that acts as a proton-containing agent. We believed that co-ordinatively unsaturated copper aminoalkoxide generated in situ could easily interact with other metal ions present in solution to a.ord formation of a mixedmetal compound due to the established ability of aminoalcohols to form polynuclear metal complexes.3 This approach yielded various Cu/M complexes, including Cu/Pb,4a and those containing Ni, Co, Zn, Mn as the second metal.4b In view of these facts, it was of interest to consider whether aggregation of metal ions with distinct co-ordination and donor atom requirements in the presence of cation selective ligands can be accomplished starting from zerovalent metal.The interaction of copper powder with ammonium iodide in the presence of CsX (X = Cl or I) and 18-crown-6 (18c6) resulted in the formation of novel mixed-metal complexes possessing unusual structures.Herein we report the synthesis and crystal structures of [Cs(18c6)2][Cu5I6(MeCN)2] 1, [Cs2- (18c6)3][Cu8I10(MeCN)2] 2 and [{Cs(18c6)}6Cu4I7][Cs3(18c6)3]- [Cu7I10][Cu13I14]I2 3. Experimental All chemicals were of reagent grade and used as received, and all experiments were carried out in air. Elemental analyses were performed by standard titrimetric methods (for Cu and I) and with a Carlo Erba Strumentazion Analyzer (for C, H and N) by the Institute of Organic Chemistry, National Academy of Sciences of Ukraine microanalytical service.Infrared spectra were recorded as KBr discs on a UR-10 spectrophotometer in the 4000–400 cm1 region using conventional techniques. Syntheses [Cs(18c6)2][Cu5I6(MeCN)2] 1 and [Cs2(18c6)3][Cu8I10(Me- CN)2] 2. Copper powder (0.064 g, 1 mmol), NH4I (0.145 g, 1 mmol), CsCl (0.17 g, 1 mmol), 18c6 (0.26 g, 1 mmol) and MeCN (20 cm3) were heated to 60 C, and re.uxed with stirring for 4–5 h.Then the mixture was allowed to stand at room temperature for 12 h, after which it was heated, re.uxed and stirred again until total dissolution of Cu was observed (1 h). After cooling pale yellow crystals precipitated from the clear dark orange solution. Those were .ltered o., washed with PriOH and dried at room temperature to give an inseparable mixture of complexes 1 and 2 in �70% yield (per copper) (Found: C, 17.5; H, 2.1; Cu, 17.7; I, 42.1; N, 0.9. C28H54CsCu5I6N2O12 requires C, 18.4; H, 3.0; Cu, 17.4; I, 41.8; N, 1.5.C20H21- CsCu4I5NO9 requires C, 16.5; H, 1.4; Cu, 17.4; I, 43.5; N,3088 J. Chem. Soc., Dalton Trans., 1999, 3087.3093 1.0%). The IR spectrum shows the typical absorptions corresponding to the 18-crown-6 ligand and the acetonitrile with the characteristic ¥í(C N) at 2265 cm1. The substance is insoluble in water and soluble in acetonitrile.[{Cs(18c6)}6Cu4I7][Cs3(18c6)3][Cu7I10][Cu13I14]I2 3. Copper powder (0.064 g, 1 mmol), NH4I (0.145 g, 1 mmol), CsI (0.26 g, 1 mmol), 18c6 (0.26 g, 1 mmol) and MeCN (15 cm3) were heated to 60 C, re.uxed and stirred for 4.5 h. Then the mixture was allowed to stand at room temperature for 12 h, after which it was heated, re.uxed and stirred again until total dissolution of Cu was observed (1 h). Pale yellow crystals of the product precipitated from the clear dark orange solution over 1.2 d.The crystals were .ltered o., washed with PriOH and dried at room temperature. Mass collected 0.30 g, yield .80% (per copper) (Found: C, 14.2; H, 2.1; Cu, 16.2; I, 45.5. C36H72Cs3Cu8I11O18 requires C, 14.0; H, 2.3; Cu, 16.4; I, 45.1%). The IR spectrum shows the typical absorptions corresponding to the 18-crown-6 ligand. The compound is insoluble in water and soluble in acetonitrile. Crystallography Details of the data collection and processing, structure analysis and re.nement are summarized in Table 1.Di.raction experiments were performed on a Rigaku AFC6S di.ractometer (complex 1), a STOE four-circle di.ractometer using an Oxford Cryostream low temperature attachment (2) and a Siemens SMART area-detector di.ractometer (¥ø rotation scans with narrow frames) (3) equipped with graphite monochromated Mo-K¥á radiation (¥ë = 0.71073 A). Data sets were corrected for Lorentz-polarization e.ects and for the e.ects of absorption.The structures were solved by direct methods using the SHELXS 865 computer program, and re.ned by full-matrix least-squares methods on F2 using SHELXL 93.6 In general all non-hydrogen atoms were re.ned anisotropically, hydrogen atoms were .xed in idealized positions and allowed to ride on the atom to which they were attached. Each hydrogen atom was assigned an isotropic thermal parameter of 1.2 or 1.5 times that of the attached C atom for CH2 and CH3 groups, respectively (2) or .xed Uiso = 0.08 A2 (1, 3).Compound 1 exhibited disorder of three copper atoms which was modelled successfully in terms of two sets of positions, common occupancy re.ned to 40 and 60%. In the case of complex 3 a range of copper and iodide atoms as well as the atom Cs(3) and two unique crown ethers of the [Cs3(18c6)3]3 moiety were disordered. Disorder of copper and iodide atoms of the [Cu4I7]3 cubic core was modelled in terms of two sets of positions so that the cluster may adopt any of the two equivalent orientations about the center of inversion which lies between Cu(7) and its symmetry equivalent, common occupancy of Cu(6), Cu(7), I(8), I(9) and I(10) re.ned to 50%.Disorder of the Cu(4), Cu(5), I(5) and I(6) atoms of the [Cu13I14] cluster was modelled in terms of two sets of positions and is described in detail below; common occupancy of the copper and iodide .ned to 50% with copper atoms left isotropic.Unreasonably short Cu.Cu distances (2.35(4) and 2.43(4) A) in the [Cu4I7]3 and [Cu13I14] cores were consequences of disorder. Modelling two di.erent conformations for the crown molecules was not successful and one of them was re.ned isotropically only, with C.O and C.C bond lengths being constrained to 1.43(2) and 1.48(2) A, respectively.7 Large thermal parameters of the two iodide anions, I(7) and its symmetry equivalent, co-ordinated to the Cs(3) atom were consequences of the disorder of these atoms between six crystallographic positions.A subsequent data set collected at low temperature on a CCD di.ractometer gave worse results. CCDC reference number 186/1565. See http://www.rsc.org/suppdata/dt/1999/3087/ for crystallographic .les in .cif format. Fig. 1 Molecular structures of the [Cs(18c6)2] and [Cu5I6(MeCN)2] constituting complex 1 (H atoms omitted for clarity). Only one of the two possible sets of the Cu(3), Cu(4) and Cu(5) atoms is shown.Table 1 Crystal data for [Cs(18c6)2][Cu5I6(MeCN)2] 1, [Cs2(18c6)3][Cu8I10(MeCN)2] 2 and [{Cs(18c6)}6Cu4I7][Cs3(18c6)3][Cu7I10][Cu13I14]I2 3 1 2 3 Formula M Crystal system Space group a/A b/A c/A ¥á/ ¥â/ ¥ã/ U/A3 ZT /K Total number of re.ections Number of unique re.ections (Rint) Observed re.ections [I > ¥ò(I)] R1 (obs.) wR2 (obs.) C28H54CsCu5I6N2O12 1822.74 Triclinic P1. 10.891(2) 10.959(2) 21.645(4) 88.93(3) 89.58(3) 82.31(3) 2559.7(8) 2 293 9530 7794 (0.020) 5137 0.037 0.102 C40H42Cs2Cu8I10N2O18 2914.16 Triclinic P1. 10.501(5) 13.849(7) 14.153(7) 64.28(3) 89.16(3) 80.49(3) 1825(2) 1 180 6145 4742 (0.028) 4170 0.029 0.074 C108H216Cs9Cu24I33O54 9287.66 Cubic Pa3. 28.2558(1) 28.2558(1) 28.2558(1) 22559.2(1) 4 293 82209 4624 (0.078) 3555 0.096 0.197J. Chem. Soc., Dalton Trans., 1999, 3087¡V3093 3089Table 2 Geometry of the Cs¡V18-crown-6 systemsSeparations/Cation CompoundCs Cs Cs¡VO Cs deviation fromfrom the mean oxygenatoms plane Ref.[Cs(18c6)2][Cs2(18c6)3]2[Cs(18c6)][Cs3(18c6)3]3[Cs2(18c6)3]2[Cs(18c6)]¡Û¡Û[Cs(18c6)] and[Cs2(18c6)3]2[Cs(18c6)2][Cu5I6(MeCN)2] a[Cs2(18c6)3][Cu8I10(MeCN)2][{Cs(18c6)}6Cu4I7][Cs3(18c6)3][Cu7I10][Cu13I14]I2[{Cs(18c6)}6Cu4I7][Cs3(18c6)3][Cu7I10][Cu13I14]I2[Cs2(18c6)3][HX2]22HX2H2Ob[Cs(18c6)][TcNCl4][Cs9(18c6)14]9[Rh22(CO)35Hx]5[Rh22(CO)35Hx 1]4¡X4.501(3)¡X3.844(7)4.289(7)4.335(1)4.275(4)4.6583.194(6)¡V3.719(7)3.261(5)¡V3.700(6)3.204(5)¡V3.486(5)3.431(5)¡V3.795(5)Cs(1): 3.06(2)¡V3.23(2)Cs(2): 3.32(2)¡V3.64(4)3.40(3)¡V3.69(3)Cs(3): 3.12(4)¡V3.47(5)3.152(6)¡V3.413(8)3.393(5)¡V3.636(5)3.34(1)¡V3.68(1)3.29(8)¡V4.25(8)3.51(8)¡V3.67(8)3.35(8)¡V3.96(8)3.76(8)¡V4.32(8)2.051(3)2.067(3)1.850(2)2.249(1)1.401(9)2.0842.14(1)1.7601.640(3)2.167(1)2.138(4)2.351.782.291.88This workThis workThis workThis work111213a There are two unique Cs atoms in the unit cell.b X = £\-Cyanobenzothiazole-£\-carbaldehyde oximate.Results and discussionSynthesisThe compounds were prepared employing zerovalent copper,ammonium iodide, 18-crown-6 and caesium iodide/chloride asstarting materials in acetonitrile.The interaction in such a systemis complex and the composition of the products appearedto depend mainly on the initial Cu : I ratio, so that the largeramount of iodide in the initial mixture in the case of 3 aordeda more complicated composition and, as was revealed by X-raystudies, the bulkier structural moieties.The pale yellow prismatic crystals of the product of theinteraction of copper, NH4I, CsCl and 18c6 gave no satisfactoryelemental analyses and were considered to be a mixture of severalcaesium iodocuprate complexes with 18-crown-6.We werenot able to separate this mixture either by recrystallization fromacetonitrile or by manual separation due to the visual uniformityof the crystals. Examination of the lattice parameters for theseparate single crystals showed the presence of at least two differentspecies. Their composition was established from the Xrayinvestigations as [Cs(18c6)2][Cu5I6(MeCN)2] 1 and[Cs2(18c6)3][Cu8I10(MeCN)2] 2.Crystal structure of complex 1The compound consists of sandwich [Cs(18c6)2] cations andnovel pentanuclear iodocuprate() [Cu5I6(MeCN)2] anions(Fig. 1). The Cs cations are in typical ¡§sunrise¡� co-ordinationto 18c6, which has D3d symmetry with characteristic C¡VO andC¡VC distances of 1.40(1)¡V1.44(1) and 1.47(1)¡V1.52(1) ,respectively. The distances and angles in [Cs(18c6)2] (Table 2)are comparable to those found in related systems.7 In the clusteranion [Cu5I6(MeCN)2] four copper atoms are arranged in atetrahedron bridged on all edges by iodides and with one facecapped by the fth copper atom.Three of the copper atoms,Cu(3), Cu(4), Cu(5), are disordered over two positions each insuch a way that Cu(1) caps the face of the tetrahedron builtfrom Cu(2), Cu(3), Cu(4) and Cu(5), while for the tetrahedronbuilt from Cu(1), Cu(3a), Cu(4a) and Cu(5a) the capping atomis Cu(2). The Cu Cu distances vary from 2.565(3) to 2.860(2)and the Cu¡VI bond lengths from 2.484(2) to 2.719(4) (Table3).Three of the ve copper() ions, Cu(3), Cu(4), Cu(5), have anearly planar trigonal co-ordination (the maximum shift of acopper atom from the plane of three iodides is 0.231(3) ). TheCu(1) and Cu(2) atoms are tetrahedrally co-ordinated by iodideanions and nitrogen atoms, N(1) and N(2), from the acetonitrilemolecules. The cluster is reminiscent of the copper()framework seen for the [Cu4I6]2 unit in [K7(12-crown-4)6][Cu4I6][Cu8I13] 8a or in [MePPh3]2[Cu4I6],8b that have theclosest stoichiometric ratio to [Cu5I6(MeCN)2].Crystal structure of complex 2The compound is built up from cations [Cs2(18c6)3]2 and niteiodocuprate() anions [Cu8I10(MeCN)2]2.Both these largemoieties have comparable sizes (Fig. 2) and co-operate to forma simple ionic lattice. The structure of the anion isunprecedented and can be viewed as a S-shape chain of sevenCu2I2 rhombohedrons, that share opposite edges with two additionaliodide atoms, I(1) and I(1a), bridging the copper atomof the ¡§central¡� and ¡§end¡� rhombohedrons.These bridges leadto the formation of two six membered Cu3I3 rings of chairconformation (Fig. 2), thus both characteristic motifs of theTable 3 Selected bond distances () and angles () for the [Cu5I6-(MeCN)2] anion of complex 1 aCu(1)¡VN(1)Cu(2)¡VN(2)Cu(1)¡VI(3)Cu(1)¡VI(1)Cu(1)¡VI(2)Cu(2)¡VI(2)Cu(2)¡VI(4)Cu(2)¡VI(1)Cu(3a)¡VI(5)Cu(3a)¡VI(1)Cu(3a)¡VI(4)Cu(4a)¡VI(6)Cu(4a)¡VI(2)I(3)¡VCu(1)¡VI(1)I(3)¡VCu(1)¡VI(2)I(1)¡VCu(1)¡VI(2)I(2)¡VCu(2)¡VI(4)I(2)¡VCu(2)¡VI(1)I(4)¡VCu(2)¡VI(1)I(5)¡VCu(3a)¡VI(1)I(5)¡VCu(3a)¡VI(4)I(1)¡VCu(3a)¡VI(4)I(6)¡VCu(4a)¡VI(2)I(6)¡VCu(4a)¡VI(4)I(2)¡VCu(4a)¡VI(4)2.000(8)1.982(8)2.655(2)2.682(2)2.709(2)2.671(2)2.683(2)2.684(2)2.547(2)2.567(2)2.646(2)2.546(3)2.549(3)112.14(6)114.64(6)114.77(5)110.55(6)115.97(5)110.32(6)123.72(9)117.86(8)115.33(8)126.79(9)117.4(1)113.90(9)Cu(4a)¡VI(4)Cu(5a)¡VI(3)Cu(5a)¡VI(6)Cu(5a)¡VI(5)Cu(3b)¡VI(5)Cu(3b)¡VI(1)Cu(3b)¡VI(3)Cu(4b)¡VI(6)Cu(4b)¡VI(2)Cu(4b)¡VI(3)Cu(5b)¡VI(4)Cu(5b)¡VI(5)Cu(5b)¡VI(6)I(3)¡VCu(5a)¡VI(6)I(3)¡VCu(5a)¡VI(5)I(6)¡VCu(5a)¡VI(5)I(5)¡VCu(3b)¡VI(1)I(5)¡VCu(3b)¡VI(3)I(1)¡VCu(3b)¡VI(3)I(6)¡VCu(4b)¡VI(2)I(6)¡VCu(4b)¡VI(3)I(2)¡VCu(4b)¡VI(3)I(4)¡VCu(5b)¡VI(5)I(4)¡VCu(5b)¡VI(6)I(5)¡VCu(5b)¡VI(6)2.700(2)2.484(2)2.565(2)2.593(2)2.531(3)2.562(3)2.668(3)2.517(3)2.591(4)2.719(4)2.509(3)2.568(4)2.601(3)122.40(9)122.05(9)115.40(9)124.6(1)117.3(1)115.7(1)126.2(2)115.3(2)116.5(1)122.3(1)122.6(1)115.0(1)a Atoms Cu(3), Cu(4) and Cu(5) are disordered on two positions ¡¥a¡¦excl;¥b¡¦ with population 0.6 and 0.4, respectively.3090 J.Chem. Soc., Dalton Trans., 1999, 3087–3093 iodocuprate species (Cu2I2 and Cu3I3) 9 are present in the structure of the anion. The copper atoms Cu(2) and Cu(3) adopt triangular planar and Cu(1) and Cu(4) tetrahedral coordination (Cu–I 2.512(2)–2.802(2) Å) (Table 4).The tetrahedral environment of Cu(4) is completed by the nitrogen atom of the acetonitrile molecule (Cu(4)–N(1) 1.988(7) Å); this coordination apparently prevents polymerization of the [Cu8I10]2 units into the usual one dimensional in.nite [Cu2I3] n or [Cu3I4] n chains.8a,10 The cation in complex 2 is a unique example of such a large multidecker 18c6 complex cation being free from disorder or extremely high anisotropy in the thermal motion of the atoms. The Cs Cs separations of 4.501(3) Å as well as the deviation of the caesium atom from the mean plane of the “central” 18c6 molecule (2.249(1) Å) are greater than those observed for [Cs2(18c6)3][HX2]22HX2H2O (X = a-cyanobenzothiazole- a-carbaldehyde oximate) 11 or the “in.nite sandwich” [Cs(18c6)]8 8 12 (Table 2).Thus the interaction between the Cs ions and the crown ligands in the present structure is somewhat weaker (Cs–O 3.204(5)–3.795(5) Å), than the majority of those listed in Table 2.This may be attributed to the adaptability of the club sandwich structure to steric demands of the anionic counterpart. The geometry of the crown ligand is Fig. 2 Molecular structures of the [Cs2(18c6)3]2 and [Cu8I10- (MeCN)2]2 constituting complex 2 (H atoms omitted for clarity). Table 4 Selected bond distances (Å) and angles () for the [Cu8I10- (MeCN)2]2 anion of complex 2 a Cu(1)–I(1) Cu(1)–I(2a) Cu(1)–I(5a) Cu(1)–I(2) Cu(2)–I(2) Cu(2)–I(3) Cu(2)–I(5a) Cu(3)–I(1) Cu(3)–I(4) Cu(3)–I(3a) I(1)–Cu(1)–I(2a) I(1)–Cu(1)–I(5a) I(2a)–Cu(1)–I(5a) I(1)–Cu(1)–I(2) I(2a)–Cu(1)–I(2) I(5a)–Cu(1)–I(2) I(2)–Cu(2)–I(3) I(2)–Cu(2)–I(5a) I(3)–Cu(2)–I(5a) 2.568(2) 2.624(2) 2.701(2) 2.739(2) 2.546(2) 2.564(2) 2.568(2) 2.512(2) 2.530(2) 2.629(2) 119.52(5) 106.93(6) 107.81(5) 108.55(5) 106.02(5) 107.48(5) 119.98(5) 118.13(5) 119.92(5) Cu(4)–N(1) Cu(4)–I(4) Cu(4)–I(5) Cu(4)–I(3a) I(2)–Cu(1a) I(3)–Cu(3a) I(3)–Cu(4a) I(5)–Cu(2a) I(5)–Cu(1a) I(1)–Cu(3)–I(4) I(1)–Cu(3)–I(3a) I(4)–Cu(3)–I(3a) N(1)–Cu(4)–I(4) N(1)–Cu(4)–I(5) I(4)–Cu(4)–I(5) N(1)–Cu(4)–I(3a) I(4)–Cu(4)–I(3a) I(5)–Cu(4)–I(3a) 1.988(7) 2.590(2) 2.658(2) 2.802(2) 2.624(2) 2.629(2) 2.802(2) 2.568(2) 2.701(2) 126.70(5) 117.68(5) 113.80(5) 112.0(2) 102.4(2) 119.57(5) 106.9(2) 106.52(5) 108.90(5) a Atoms designated a are related by the symmetry operation 1 x, 2 y, 1 z.sensitive to such an elongation of the M–O contacts, the OCCO torsion angles (for the central 18c6 molecule ±59.9(8)–64.8(7), average ±61.8(8)), corresponding to a gauche conformation, are less than the usual values of ±65–75.7 Both the independent macrocycles have a typical D3d conformation with characteristic C–O and C–C distances of 1.407(9)–1.436(9) and 1.47(1)–1.50(1) Å, respectively.Crystal structure of complex 3 The structure was modeled to include several di.erent moieties: three iodocuprate(.) clusters [Cu4I7]3, [Cu7I10]3 and [Cu13I14], additional iodide anions, novel 3: 3 club sandwiches [Cs3(18c6)3]3 and the usual half-sandwich encapsulates [Cs(18c6)] (Figs. 3 and 4). The [Cu4I7]3 cluster is made up of a copper tetrahedron interlocked with an iodine tetrahedron, plus three iodine atoms bonded to the three metal atoms, and is closely related to the cubane Cu4I4L4 archetype 14 (Fig. 3); selected bond parameters are listed in Table 5. All faces of the heterocubane core are remarkably non-planar, their edges are of di.erent length and the least-squares rhombohedral planes of the faces are mutually non-orthogonal.Six caesium atoms, each bound in the sunrise geometry by the crown macrocycle, are positioned at the faces of this virtual cube (Cs–I 3.699(7) and 3.82(1) Å), forming the supramolecular cation [{Cs(18c6)}6Cu4I7]3 (Fig. 3). Examples of self-assembled cations in which anions exhibit an unambiguous organizational role are quite rare in the literature, with only a few examples being reported: [{K(15-crown- 5)}4Br]3,15a [{A(dibenzo-18c6)}3AgX3],15b and [{A(18c6)}4- Fig. 3 View of the [{Cs(18c6)}6Cu4I7]3 unit in complex 3 (H atoms omitted for clarity), showing the arrangement of six Cs(18c6) around the central iodocuprate(.) core. Also shown at the bottom is one of the two possible orientations of the cubic core.J. Chem. Soc., Dalton Trans., 1999, 3087¡V3093 3091MX4]2 (A = monocation, M = 3d element).15c Thus in the formationof the [{Cs(18c6)}6Cu4I7]3 moiety the [Cu4I7]3 anionmay be recognized as a templating agent, that possesses theshape, charge and size suitable for the assembly of the bulkycrown ether ¡§supercomplex¡�.The triple decker 3 : 3 club sandwich [Cs3(18c6)3]3 is a novelstructural unit; its main geometrical parameters are summarizedin Table 2.The Cs(2) Cs(2) {x, y, 1 z} separationof 4.289(7) agrees well with the data for Cs(-18c6)Cs systems,while the Cs(2) Cs(3) contact is considerably shorter(3.844(7) ) and, in fact, is the shortest such separationobserved so far in structures of this type.7 A very shortCs Cs separation of the same nature (3.923(4) ) was discussedin terms of a ¡§short bond of order zero¡� for[Cs2(18c6)][SO4(AlMe3)3].16The [Cs3(18c6)3]3 cation contains metal atoms on a threefoldaxis and in the lattice has iodide and iodocuprate() anions atboth ends along the threefold axis.The anions are related by theoperation of an inversion centre, which is located in the centreof the central crown macrocycle.The Cu, I and Cs(3) atoms,that occupy partly populated positions, are disordered in such away that the ¡§crown¡� end of the triple decker sandwich isFig. 4 The arrangement of Cu and I atoms in the [Cu13I14] cluster ofcomplex 3. (a) View of the cluster down the I(6)¡VCu(2)¡VI(3) axis. (b)Truncated tetrahedron of 12 Cu atoms centered around the thirteencopper atom in an ¡§adamantane-like¡� disposition of 10 iodide atoms; 4iodide atoms center hexangulated faces of the copper cluster.The linesconnecting Cu¡VCu and I¡VI atoms do not represent bonds.adjacent to the [Cu13I14] cluster while the caesium atom at theother end of the cation co-ordinates two iodide anions andabuts with the [Cu7I10]3 moiety. Thus the smaller cluster[Cu7I10]3 is a symmetry equivalent of the larger anion[Cu13I14]. The disorder in the system is illustrated in Scheme 1.The novel polynuclear iodocuprate() cluster [Cu13I14] has anearly perfect spherical structure, in which twelve copper atomsdescribe a truncated tetrahedron, centered by the thirteenthmetal atom (Cu Cu 2.542(6)¡V3.032(7) ) (Fig. 4). In thepolyhedron all tri- and hex-angulated faces as well as six commonedges of the latter ones are centered by iodides, so that tenof the iodide atoms exhibit an ¡§adamantane-like¡� disposition.The copper atoms adopt tetrahedral co-ordination, six iodideatoms serve as 2 bridges (Cu¡VI 2.52(3)¡V2.636(5) ), four as3-bridges (Cu¡VI 2.534(4)¡V2.58(3) ), and four are 7 bridges(Cu¡VI 2.643(3)¡V3.099(6) ) (Table 5).The highest condensed iodocuprate() ion observed so far isthe [Cu36I56]20 polyanion in the compound [(Hpy)]2[Cu3I5] inwhich 24 Cu atoms occupy the corners of a cube with octahedralhabit, i.e. the positions formed by cutting o the cornersof a cube; the 12 remaining Cu atoms lie at the middle ofthe edges of this cube.17ConclusionThis study has demonstrated that interaction of zerovalentcopper with ammon in presence of Cs¡V18c6 cationsleads to the formation of iodocuprate() species of dierentTable 5 Selected bond distances () and angles () for iodocuprate()clusters of complex 3 a[Cu13I14]Cu(1)¡VI(1)Cu(1)¡VI(2)Cu(1)¡VI(4)Cu(1)¡VI(3)Cu(2)¡VI(41)Cu(2)¡VI(4)Cu(2)¡VI(42)Cu(2)¡VI(3)Cu(3)¡VI(11)Cu(3)¡VI(2)Cu(3)¡VI(4)Cu(3)¡VI(3)Cu(4)¡VI(5)Cu(4)¡VI(1)Cu(4)¡VI(4)Cu(4)¡VI(42)Cu(5)¡VI(5)Cu(5)¡VI(6)Cu(5)¡VI(42)Cu(5)¡VI(4)I(1)¡VCu(1)¡VI(2)I(1)¡VCu(1)¡VI(4)I(2)¡VCu(1)¡VI(4)I(1)¡VCu(1)¡VI(3)I(2)¡VCu(1)¡VI(3)2.534(4)2.602(5)2.692(5)3.099(6)2.643(3)2.643(3)2.643(3)2.82(1)2.540(4)2.636(5)2.738(5)2.953(6)2.52(3)2.57(3)2.76(3)2.93(3)2.54(3)2.58(3)2.76(3)2.90(3)125.8(2)118.1(2)106.0(2)106.6(2)95.9(2)I(4)¡VCu(1)¡VI(3)I(41)¡VCu(2)¡VI(4)I(41)¡VCu(2)¡VI(42)I(4)¡VCu(2)¡VI(42)I(41)¡VCu(2)¡VI(3)I(4)¡VCu(2)¡VI(3)I(42)¡VCu(2)¡VI(3)I(11)¡VCu(3)¡VI(2)I(11)¡VCu(3)¡VI(4)I(2)¡VCu(3)¡VI(4)I(11)¡VCu(3)¡VI(3)I(2)¡VCu(3)¡VI(3)I(4)¡VCu(3)¡VI(3)I(5)¡VCu(4)¡VI(1)I(5)¡VCu(4)¡VI(4)I(1)¡VCu(4)¡VI(4)I(5)¡VCu(4)¡VI(42)I(1)¡VCu(4)¡VI(42)I(4)¡VCu(4)¡VI(42)I(5)¡VCu(5)¡VI(6)I(5)¡VCu(5)¡VI(42)I(6)¡VCu(5)¡VI(42)I(5)¡VCu(5)¡VI(4)I(6)¡VCu(5)¡VI(4)I(42)¡VCu(5)¡VI(4)98.1(2)112.1(2)112.1(2)112.1(2)106.7(2)106.7(2)106.7(2)122.1(2)117.3(2)103.8(2)110.8(2)98.8(2)100.6(2)125(1)105(1)114(1)99.4(9)109.8(9)100.7(8)121(1)103.6(9)116(1)100.1(9)111.6(9)101.6(8)[Cu4I7]3Cu(6)¡VI(10)Cu(6)¡VI(9)Cu(6)¡VI(85)Cu(6)¡VI(84)Cu(7)¡VI(84)Cu(7)¡VI(83)Cu(7)¡VI(85)Cu(7)¡VI(9)I(10)¡VCu(6)¡VI(9)I(10)¡VCu(6)¡VI(85)2.21(4)2.60(3)2.92(4)3.03(4)2.646(9)2.646(9)2.647(9)2.96(3)125(2)118(1)I(9)¡VCu(6)¡VI(85)I(10)¡VCu(6)¡VI(84)I(9)¡VCu(6)¡VI(84)I(85)¡VCu(6)¡VI(84)I(84)¡VCu(7)¡VI(83)I(84)¡VCu(7)¡VI(85)I(83)¡VCu(7)¡VI(85)I(84)¡VCu(7)¡VI(9)I(83)¡VCu(7)¡VI(9)I(85)¡VCu(7)¡VI(9)105(1)105(1)102(1)98(1)115.6(4)115.6(4)115.6(4)102.3(6)102.3(6)102.3(6)a Symmetry transformations used to generate equivalent atoms (indicatedby superscript): 1 y, z 0.5, x 0.5; 2 z 0.5, x, y 0.5;3 x, y, z; 4 y, z, x; 5 z, x, y.3092 J.Chem. Soc., Dalton Trans., 1999, 3087¡V3093Scheme 1stoichiometry. This is unlike the Cu¡VNH4I interaction in thepresence of such ligands as pyridine, ethylenediamine, aminoalcoholsthat usually result in Cu2 co-ordination compounds.18It may reasonably be proposed that the intermediate Cu oxidationstate is stabilized in presence of the Cs¡V18c6 moietiesin acetonitrile, a solvent which is known to stabilize the Cuby itself.19 The intricate nature of the interaction apparentlyresults in the equilibrium mixture of various iodocuprates()present in solution and the overall formation of the solid phaseis a template process of mutual stabilization of Cs as bulky[Csn(18c6)m]n cations and Cu as iodocuprate anions.Despite the complicated molecular architecture of the tripledecker club sandwiches [Csn(18c6)3]n (n = 2 or 3), formation ofsuch complexes, in the light of the present work and our previouspublication,11 should be considered as a general feature of18-crown-6 co-ordination chemistry.AcknowledgementsThe work was in part supported by the International SorosScience and Education Program (grants PSU083062 (J.A. R.),YSU083008 (O. Yu. V.), YSU083019 (K. V. D.) andQSU083090 (V. N. K.)). O. Yu. V. thanks the CambridgeColleges Hospitality Scheme for a short study visit atCambridge University.References1 See, for example: V. V. Skopenko (Editor), Direct synthesis ofcoordination compounds, Venturi, Kyiv, 1997; S.R. Petrusenko, V. N.Kokozay, O. Yu. Vassilyeva and B. W. Skelton, J. Chem. Soc., DaltonTrans., 1997, 1793; S. M. Godfrey, C. A. McAulie and R. G.Pritchard, J. Chem. Soc., Dalton Trans., 1993, 2875.2 O. Yu. Vassilyeva and V. N. Kokozay, Ukr. Khim. Zh. (Russ. Ed),1993, 59, N2, 176; O. Yu. Vassilyeva, N. D. Nevesenko, V. V.Skopenko, Yu. S. Gerasimenko and V. N. Kokozay, Dopov. Akad.Nauk Ukr. RSR, Ser. B, 1988, N3, 36.3 See, for example: K.Smolander, Ann. Acad. Sci. Fenn., Ser. A2,1983, 1; V. N. Kokozay and A. Sienkiewicz, Polyhedron, 1993, 12,2421; A. Sienkiewicz and V. N. Kokozay, Polyhedron, 1994, 13, 1439.4 (a) L. A. Kovbasyuk, O. Yu. Vassilyeva, V. N. Kokozay, W. Linert,J. Reedijk, B. W. Skelton and A. G. Oliver, J. Chem. Soc., DaltonTrans., 1998, 2735; L. A. Kovbasyuk, O. Yu. Vassilyeva, V. N.Kokozay, W. Linert, B. W. Skelton and A. G. Oliver, New J. Chem.,1998, 931 and refs. therein; (b) O.Yu. Vassilyeva, V. N. Kokozay,V. G. Makhan¡¦kova, E. Vinogradova, W. Linert and P. R. Raithby,unpublished work.5 G. M. Sheldrick, SHELXS 86, Program for the Solution of CrystalStructure, University of Gttingen, 1986.6 G. M. Sheldrick, SHELXL 93, Program for the Renement ofCrystal Structure, University of Gttingen, 1993.7 A. V. Bajaj and N. S. Poonia, Coord. Chem. Rev., 1988, 87, 55; F. R.Fronczek and R. D. Gandour, in Cation binding by Macrocycles:Complexation of Cationic Species by Crown Ethers, eds.Y. Inoue andG. W. Gokel, Marcel Dekker, New York, 1991, p. 311.8 (a) N. P. Rath and E. M. Holt, J. Chem. Soc., Chem. Commun., 1985,665; (b) G. A. Bowmaker, G. R. Clark and D. K. P. Yuen, J. Chem.Soc., Dalton Trans., 1976, 2329.9 S. Jagner and G. Helgesson, Adv. Inorg. Chem., 1991, 37, 1.10 A. K. Nurtaeva and E. M. Holt, Acta Crystallogr., Sect. C, 1998, 54,594; A. K. Nurtaeva, G. Hu and E. M. Holt, Acta Crystallogr., Sect.C, 1998, 54, 597.11 K. V. Domasevitch, V. V. Ponomareva and E. B. Rusanov, J. Chem.Soc., Dalton Trans., 1997, 1177.12 J. Baldas, S. F. Colmanet and G. A. Williams, J. Chem. Soc., Chem.Commun., 1991, 954.13 J. L. Vidal, R. C. Schoening and J. M. Troup, Inorg. Chem., 1981, 20,227.14 A. F. Wells, Structural Inorganic Chemistry, Clarendon Press,Oxford, 1984, ch. 25, p. 1126.15 (a) N. S. Fender, I. A. Kahwa, A. J. P. White and D. J. Williams,J. Chem. Soc., Dalton Trans., 1998, 1729; (b) G. Helgesson and S.Jagner, J. Chem. Soc., Dalton Trans., 1993, 1069; (c) I. A. Kahwa,D. Miller, M. Mitchell, F. R. Fronczek, R. G. Goodrich, D. J.Williams, C. A. O¡¦Mahoney, A. M. Z. Slawin, S. V. Ley and C. J.Groombridge, Inorg. Chem., 1992, 31, 3963; N. S. Fender, F. R.Fronczek, V. John, I. A. Kahwa and G. L. McPherson, Inorg. Chem.,1997, 36, 5539.16 C. M. Means, N. C. Means, S. G. Bott and J. L. Atwood, J. Am.Chem. Soc., 1984, 106, 7627.J. Chem. Soc., Dalton Trans., 1999, 3087–3093 3093 17 H. Hartl and J. Fuchs, Angew. Chem., Int. Ed. Engl., 1986, 25, 569. 18 See, for example: O. Yu. Vassilyeva, V. N. Kokozay and Yu. A. Simonov, Zh. Neorg. Khim., 1991, 36, 3119; V. N. Kokozay, A. A. Dvorkin, O. Yu. Vassilyeva, A. V. Sienkiewicz and O. N. Rebrova, Zh. Neorg. Khim., 1991, 36, 1446; O. Yu. Vassilyeva and V. N. Kokozay, Ukr. Khim. Zh. (Russ. Ed.), 1994, 60, 227; Koord. Khim., 1991, 17, 968. 19 I. D. MacLeod, D. M. Muir, A. J. Parker and P. Singh, Aust. J. Chem., 1977, 30, 1423; A. J. Parker, D. A. Clarke, R. A. Couche, G. Miller, R. I. Tilley and W. E. Waghorne, Aust. J. Chem., 1977, 30, 1661. Paper 9/03225H

ISSN:1477-9226    

DOI:10.1039/a903225h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3095–3101 3095 New ruthenium(II) heteroleptic complexes containing the 4-(2- pyridyl)pyrimidine ligand with amine and amino acid substituents Hassan Aït-Haddou,*a Elena Bejan,a Jean-Claude Daran,a Gilbert G. A. Balavoine,a Florence Berruyer-Penaud,b Lydia Bonazzola,b Henda Smaoui-Chaabouni b and Edmond Amouyal *b a Laboratoire de Chimie de Coordination, CNRS UPR 8241, 205 route de Narbonne, 31077 Toulouse Cedex, France. E-mail: aithad@lcc-toulouse.fr b Laboratoire de Physico-Chimie des Rayonnements, CNRS UMR 8610, Bat. 350, Université Paris-Sud, 91405 Orsay, France. E-mail: edmond.amouyal@lpcr.u-psud.fr Received 9th June 1999, Accepted 13th July 1999 New 4-(2-pyridyl)pyrimidines (L) have been synthesized in high yields by condensing enaminones with the appropriate carboxamidine or guanidine under basic conditions. These asymmetric bidentate diimine ligands coming from a one-step functionalization of amine and amino acid groups were complexed to ruthenium to obtain new heteroleptic complexes of type [Ru(bpy)2(L)]2.The ligands and complexes have been characterized by usual analytical methods, and the crystallographic study of one complex has been performed. Their spectroscopic and electrochemical properties have been investigated. ZINDO Calculations show that in the MLCT excited state the electron is mainly localized on the pyridylpyrimidine ligand. On the basis of electrochemical data, the .rst reduction potential of the complexes has been assigned to the redox couple involving this ligand.Introduction Polypyridine complexes of ruthenium(..) have stimulated considerable interest and activity as photosensitizers for photochemical and photoelectrochemical conversion of solar energy.1 These complexes might also .nd application as components of molecular electronics devices 2 and as photoactive DNA cleavage agents for phototherapeutic purposes.3 Symmetric polypyridine ligands such as 2,2-bipyridine (bpy) have been extensively utilized as chelating agents,4 but relatively little attention has been directed towards complexes that possess asymmetric bidentate diimine ligands.5 It is clear that the replacement of one of the pyridine rings by other nitrogencontaining heterocycles o.ers the possibility of tuning the redox and photophysical properties of the complexes.As part of our programme of designing new nitrogen bidentate ligands,6 we were interested by the synthesis of newly designed 4-(2-pyridyl)pyrimidine asymmetric ligands (L) and their heteroleptic ruthenium(..) complexes of type [Ru(bpy)2(L)]2 (Scheme 1).On the other hand, it was interesting to introduce amine and amino acid groups at position 2 of the pyrimidine ring. Such substituents should increase the a.nity of the corresponding complexes to DNA.7 By the way, it should be emphasized that we herein report a one-step functionalization of an amino acid bearing an asymmetric bidentate diimine which was subsequently and successfully complexed to ruthenium.In this work the spectroscopic and electrochemical properties of this novel series of substituted pyridylpyrimidine ligands as well as the corresponding heteroleptic complexes have been investigated. ZINDO Calculations have been performed and support the assignment of electronic absorption spectra. Scheme 1 Synthesis of the 4-(2-pyridyl)pyrimidines L1–L5 [(a) 1.25 equivalents of (CH3)2NCH(OCH3)2 at 100 C, 16 h, 99%, (b) Guanidine or carboxamidine, 1 or 3 equivalents of sodium, EtOH, re.ux] and the corresponding heteroleptic [Ru(bpy)2(L)]2 complexes.3096 J.Chem. Soc., Dalton Trans., 1999, 3095–3101 Experimental General All reactions were carried out under an argon atmosphere. Nuclear magnetic resonance spectra were recorded with Bruker AM-250 (250 MHz) and AC-200 (200 MHz) spectrometers for 1H. Chemical shifts are reported in ppm and the solvent was used as internal reference.These instruments were also used for 13C spectra. All melting points are uncorrected. The CI and FAB mass spectra (m-nitrobenzyl alcohol matrix) were recorded with a quadripolar Nermag R10-10H instrument. Elemental analyses were performed by the “Service de Microanalyse” of the LCC (Laboratoire de Chimie de Coordination). Column chromatography puri.cations were performed with Merck aluminium (deactivated with 8% water) or with silica gel (35–70 mesh).The complex [Ru(bpy)2Cl2]2H2O, 2-acetylpyridine and dimethylformamide dimethyl acetal were purchased from Fluka, 2,2-bipyridine (99.5%) from Aldrich and [Ru(bpy)3]Cl26H2O (99%) from Strem Chemicals. The guanidines and carboxamides were purchased from Aldrich. For the preparation of compounds 2, L1 and L3 see ref. 6. Spectroscopic grade ethanol (Carlo Erba) was used as supplied. For cyclic voltammetry, acetonitrile (Aldrich, spectrophotometric grade, 99.5%) was used as solvent and 0.1 mol l1 tetrabutylammonium tetra.uoroborate (Janssen, 99%) as supporting electrolyte.Synthesis of the 4-(2-pyridyl)pyrimidine ligands 2-Methyl-4-(2-pyridyl)pyrimidine L2. A solution of acetamidinium chloride (2.14 g, 22.6 mmol) in absolute ethanol (75 ml) was added to a stirred solution of 2-[3-(dimethylamino)- 1-oxoprop-2-en-1-yl]pyridine 2 (2.0 g, 11.3 mmol) in boiling absolute EtOH (50 ml) and stirring was continued for 20 min. To this mixture was added Na (0.78 g, 33.9 mmol, 3 equivalents) in absolute EtOH (50 ml) and the reaction mixture re.uxed for 16 h.The solution was allowed to cool to room temperature and then concentrated under reduced pressure. The residue was dissolved in dichloromethane followed by removal of the precipitate by .ltration. The .ltrate was concentrated and the residue puri.ed by column chromatography on .ash silica gel (ethyl acetate) to give 1.88 g of L2 as a white microcrystalline powder.Yield: 97%. 1H NMR (CDCl3, 250 MHz): d 8.7 (d, 1 H, J = 4.45), 8.63 (m, 1 H), 8.43 (d, 1 H, J = 5.24), 8.1 (d, 1 H, J = 8.01), 7.8 (t, 1 H, J = 7.85 Hz), 7.35 (m, 1 H) and 2.8 (s, 3 H). 13C NMR (CDCl3): d 167.87, 162.80, 157.82, 154.01, 149.44, 136.98, 125.20, 121.61, 114.22 and 26.08. MS (CI, NH3): m/z = 172 (MH, 100%). Calc. for C10H9N3: C, 70.16; H, 5.30; N, 24.54. Found: C, 70.41; H, 5.62; N, 24.12%. 2-Ethylamino-4-(2-pyridyl)pyrimidine L4. This compound was prepared in the same fashion as for L2 by condensation of 3.0 g (16.95 mmol) of 2 with 1.23 g (16.95 mmol) of ethyl guanidine sulfate {[C2H5NHC( NH)NH2]2H2SO4} in the presence of 3 equivalents of sodium in absolute EtOH.The desired product was puri.ed by column chromatography on silica gel (ethyl acetate–n-pentane, 80 : 20) to give 3.12 g of L4 as a microcrystalline white powder. Yield: 92%. 1H NMR (CDCl3, 250 MHz): d 8.64 (dd, 1 H, J = 6.33, 1.6), 8.38 (d, 1 H, J = 5.15), 8.34 (d, 1 H, J = 8.0), 7.77 (ddd, 1H, J = 7.78, 6.0, 1.76), 7.31 (m, 1 H), 5.48 (m, 1 H), 3.5 (m, 2 H) and 1.23 (t, 3 H, J = 7.2 Hz). 13C NMR (CDCl3): d 163.0, 162.0, 158.5, 154.3, 148.8, 136.3, 124.4, 120.9, 106.0, 35.8 and 14.4. MS (CI, NH3): m/z = 201 (MH, 100%). Calc. for C11H12N4: C, 65.98; H, 6.04; N, 27.98. Found: C, 66.21; H, 6.35; N, 27.64%. 2-(4-tert-Butoxycarbonylamino-4-carboxybutylamino)-4-(2- pyridyl)pyrimidine L5. To a stirred solution of compound 2 (0.5 g, 2.8 mmol) in 5 mL of boiling absolute EtOH was added a solution of Na-Boc-.-Arg (0.77 g, 2.8 mmol; Boc = COOC(CH3)3) in 10 mL of absolute EtOH.After 10 min of stirring, 0.195 g (8.4 mmol, 3 equivalents) of sodium in 5 mL of absolute EtOH was added and the re.ux maintained for 2 h. The solution was cooled to room temperature and concentrated under vacuum. The residue was puri.ed by column chromatography on deactivated aluminium (methanol–diethyl ether, 97 : 3) to give 1.1 g of L5 as a yellow powder.Yield: 100%. [a]25 D 13.5 deg. cm3 g1 dm1 (c [g per 100 ml] 1.0, CH3OH). 1H NMR (CDCl3, 250 MHz): d 8.6 (d, 1 H, J = 4.4), 8.27 (d, 1 H, J = 5.2), 7.7 (dd, 1 H, J = 7.7, 1.5), 7.4 (d, 1 H, J = 4.67 Hz), 7.3 (m, 1 H), 6.5 (m, 3 H), 5.9 (m, 1 H), 4.1 (s, 1 H), 3.4 (m, 2 H), 1.7 (m, 4 H) and 1.3 (s, 9 H). 13C NMR (CDCl3): d 178.1, 163.9, 161.7, 157.3, 156.0, 154.2, 149.1, 136.8, 125.0, 121.6, 106.0, 79.1, 52.0, 50.3, 30.3, 28.3 and 25.8. MS (CI, NH3): m/z = 388 (MH, 100%).Calc. for C19H25N5O4: C, 58.9; H, 6.5; N, 18.08. Found: C, 58.21; H, 6.35; N, 17.74%. Synthesis of [Ru(bpy)2(L)][PF6]2 complexes [Ru(bpy)2(L1)]2. The complex cis-[Ru(bpy)2Cl2]2H2O (0.21 g, 0.4 mmol) was suspended in a mixture of ethanol and water (50 mL, 75 : 25) and heated under argon for 30 min. Ligand L1 (0.075 g, 0.48 mmol, 1.2 equivalent) was added, and the mixture re.uxed for 16 h. The resulting solution was cooled to room temperature and the ethanol removed under pressure.After .ltration, a saturated solution of ammonium hexa- .uorophosphate was added dropwise to the .ltrate to complete precipitation. The precipitate was collected by .ltration, washed by water and diethyl ether, and dried to give 0.292 g of the desired complex. TLC (alumina, acetone–water–saturated aqueous potassium nitrate, 90 : 10 : 1) showed the material to be pure. Yield: 85%. 1H NMR (CD3CN): d 9.18 (1 H, d, J = 5.28), 8.9 (1 H, d, J = 8.06), 8.7 (5H, m), 8.55 (1 H, s), 8.3 (5 H, m), 7.99 (1 H, d, J = 5.43), 7.97 (1 H, d, J = 5.44), 7.87 (1 H, d, J = 5.33 Hz), 7.82 (2 H, m), 7.5 (4 H, m) and 7.2 (1 H, m). 13C NMR (CD3CN): d 161.5, 159.24, 158.5, 153.7, 153.2, 152.8, 139.5, 130.9, 129.2, 129.1, 127.8, 125.9, 125.8 and 120.6. FAB MS: m/z = 716 (100, M PF6 ), 571 (47%, M 2PF6 ). Calc. for C29H23F12N7P2Ru: C, 40.48; H, 2.69; N, 11.39. Found: C, 40.75; H, 2.54; N, 11.51%. [Ru(bpy)2(L2)]2. This complex was obtained by reaction of 0.168 g (0.32 mmol) of cis-[Ru(bpy)2Cl2]2H2O with 1.2 equivalents of L2 using the procedure described for [Ru(bpy)2(L1)]2.The complex was puri.ed as a hexa.uorophosphate salt, 0.258 g. Yield: 92%. 1H NMR (CD3CN): d 9.05 (2 H, m), 8.84 (4 H, m), 8.72 (1 H, d, J = 7.85), 8.55 (1 H, d, J = 7.3), 8.25 (6 H, m), 8.02 (1 H, d, J = 8.1), 7.96 (1 H, d, J = 5.54), 7.85 (1 H, d, J = 7.36 Hz), 7.65 (5 H, m) and 2.35 (3H, s). 13C NMR (CD3CN): d 174.13, 165.41, 158.75, 157.94, 157.70, 157.45, 157.19, 154.32, 152.94, 152.56, 152.21, 139.05, 138.80, 138.60, 129.68, 128.76, 128.52, 127.35, 125.37, 125.16, 117.04 and 27.79.FAB MS: m/z = 730 (56.95, M PF6 ) and 585 (100%, M 2PF6 ). Calc. for C30H25F12N7P2Ru: C, 41.20; H, 2.88; N, 11.21. Found: C, 40.85; H, 3.04; N, 11.35%. [Ru(bpy)2(L3)]2. This complex was obtained by reaction of 0.084 g (0.16 mmol) of cis-[Ru(bpy)2Cl2]2H2O with 1.2 equivalents of L3 using the procedure described for [Ru(bpy)2(L1)]2. The complex was puri.ed as a hexa.uorophosphate salt, 0.126 g.Yield: 90%. 1H NMR (CD3CN): d 9.08 (1 H, dd, J = 7.22, 1.12), 8.9 (1 H, d, J = 8.18), 8.8 (1 H, d, J = 8.12), 8.65 (3 H, m), 8.55 (1 H, t, J = 7.95 Hz), 8.25 (5 H, m), 8.05 (2 H, m), 7.8 (3 H, m) and 7.55 (5 H, m). 13C NMR (CD3CN): d 162.9, 160.9, 159.0, 158.8, 158.7, 158.2, 157.7, 155.1, 153.9, 153.7, 153.1, 151.9, 141.2, 140.0, 139.9, 139.7, 129.9, 129.6, 129.1, 128.6, 128.1, 127.1, 126.5, 126.3, 125.9, 125.8, 125.6 and 125.1.FAB MS: m/z = 731 (56.95, M PF6 ) and 585 (100%, M 2PF6 ). Calc. for C29H24F12N8P2Ru 4H2O: C, 36.76; H, 3.40; N, 11.82. Found: C, 36.39; H, 3.51; N, 12.11%.J. Chem. Soc., Dalton Trans., 1999, 3095–3101 3097 [Ru(bpy)2(L4)]2. This complex was obtained by reaction of 0.150 g (0.286 mmol) of cis-[Ru(bpy)2Cl2]2H2O with 1.2 equivalents of L4 using the procedure described for [Ru(bpy)2- (L1)]2. The complex was puri.ed as a hexa.uorophosphate salt, 0.219 g.Yield: 85%. 1HMR (CD3CN): d 8.75 (6 H, m), 8.45 (1 H, d, J = 5.3), 8.2 (5 H, m), 8.0 (1 H, d, J = 5.58), 7.9 (1 H, d, J = 5.06), 7.65 (8 H, m), 5.25 (1 H, m, NH), 3.15 (2 H, m) and 0.63 (3 H, t, J = 7.25 Hz). 13C NMR (CD3CN): d 165.8, 165.3, 161.3, 159.1, 158.8, 158.6, 158.5, 155.1, 153.2, 153.0, 152.7, 152.6, 130.0, 129.9, 129.4, 129.3, 129.2, 127.6, 126.5, 126.0, 125.9, 109.8, 38.0 and 13.1. FAB MS: m/z = 759 (30.29, M PF6 ) and 613 (73.53%, M 2PF6 ). Calc.for C31H28F12N7P2Ru: C, 41.21; H, 3.12; N, 12.40. Found: C, 41.28; H, 3.25; N, 12.56%. [Ru(bpy)2(L5)]2. This complex was obtained by reaction of 0.150 g (0.286 mmol) of cis-[Ru(bpy)2Cl2]2H2O with 1.2 equivalents of L5 using the procedure described for [Ru(bpy)2- (L1)]2. The complex was puri.ed as a hexa.uorophosphate salt, 0.219 g. Yield: 85%. 1H NMR(CD3CN): d 8.75 (6 H, m), 8.45 (1 H, m), 8.2 (5 H, m), 8.0 (2 H, m), 7.65 (8 H, m), 5.25 (1 H, m, NH), 4.1 (1 H, m), 3.15 (2 H, m), 1.8 (1 H, m), 1.51 (1 H, m), 1.45 (9 H, s) and 1.05 (2 H, m). 13C NMR (CD3CN): d 177.5, 165.9, 165.6, 161.5, 159.3, 158.9, 158.7, 158.6, 155.3, 155.0, 153.3, 153.0, 152.8, 152.6, 150.8, 140.5, 140.0, 139.7, 139.4, 139.2, 130.1, 129.6, 128.8, 128.3, 127.8, 126.9, 126.2, 125.8, 125.2, 124.8, 110.0, 80.2, 65.2, 45.0, 43.0, 26.4 and 25.9. FAB MS: m/z = 946 (9.16, M PF6 ) and 800 (13.79%, M 2PF6 ). Calc. for C39H41F12N9O4P2Ru 2H2O: C, 41.57; H, 4.03; N, 11.19. Found: C, 41.68; H, 4.21; N, 11.26%.Absorption spectroscopy and cyclic voltammetry The UV-visible electronic absorption spectra were recorded on a Perkin-Elmer Lambda 9 spectrophotometer. Cyclic voltammetry curves were recorded with a Wenking system (model 81 potentiostat) using a platinum button (Solea Tacussel EDI 101T) as the working electrode and a platinum wire of 1 mm diameter as the counter electrode. The working electrode was carefully polished with diamond sprays (Struers) and rinsed with ethanol before each potential run.Experiments were performed at a scan rate of 0.2 V s1 on Ar-purged acetonitrile solutions, containing 0.1 mol l1 tetrabutylammonium tetra- .uoroborate as the supporting electrolyte. Concentrations of 103 and 5 × 104 mol l1 were used for ligands and complexes respectively. Structure determination of [Ru(bpy)2(L2)]2, 2PF6 The data were collected on a Stoe IPDS (Imaging Plate Di.raction System) equipped with an Oxford Cryosystems cooler device.Coverage of the unique set was over 96% complete to at least 24.1. Crystal decay was monitored by measuring 200 re.ections per image. The .nal unit cell parameters were obtained by the least-squares re.nement of 5000 re.ections. Only statistical .uctuations were observed in the intensity monitors over the course of the data collection. Owing to the rather low µx value, 0.24 (µx is the product of the mid size of the crystal and the absorption coe.cient), no absorption correction was considered.On the basis of the systematic absences, the space group (P21/c or P21) could not be unambiguously de.ned. Indeed if the absences corresponding to the twofold screw axis 21 were veri.ed, the systematic absences related to the c glide plane (h0l, l = 2n 1) were not fully satis.ed. The structure could be solved by direct methods (SIR 92) 8 in P21/c but the re.nement was unstable with large discrepancies between the anisotropic thermal parameters.A much better re.nement was obtained in P21 with two molecules in the asymmetric unit. Although these two molecules are closely related by a pseudo inversion centre, no unusual parameter correlations were observed. Considering Flack’s enantiopole parameter, 0.50, its rather good standard deviation (0.04), and the agreement between related distances in two molecules, the occurrence of a racemic twin might be considered. The structure was re.ned by least-squares procedures on F 2.The H atoms were introduced as a riding model and given isotropic thermal parameters 20% higher than those of the carbon to which they are attached. Details of data collection and re.nement are given in Table 2. Selected bond lengths and angles for the two molecules are listed in Table 3. The calculations were carried out with the help of the SHELXL 97 programs9 running on a PC. The drawing of the molecule was realized with the help of CAMERON.10 CCDC reference number 186/1571. See http://www.rsc.org/suppdata/dt/1999/3095/ for crystallographic .les in .cif format.ZINDO calculations The optical absorption spectra of the ligands and of the [Ru- (bpy)2(L2)]2 complex were calculated by using the ZINDO semiempirical program.11 This Intermediate Neglect of Di.erential Overlap (INDO) model, adapted for spectroscopy and extended to second transition metal series, has been used for some time for studying the spectroscopy of large complexes.12 The SCF calculation is followed by a con.guration interaction (CI) calculation; a Rumer diagram is used to generate the CI matrix.The CI contains all singly excited con.gurations generated by removing electrons from the ten highest occupied MOs and placing them into the ten lowest unoccupied MOs. The geometries of the cis and trans conformers of all the ligands were optimized by using the AM1 method. Concerning the complex [Ru(bpy)2(L2)]2, the crystallographic data determined in this paper for N, C and Ru atoms were used for the ZINDO calculation. Hydrogen atoms were placed at a distance of 1.09 Å from their bonding partner.Results and discussion Ligands Synthesis. The synthesis of ligands L1–L5 is outlined in Scheme 1 (i). The enaminone 2 6a was obtained with quantitative yield by reaction of 2-acetylpyridine 1 with 1.2 equivalents of the dimethylformamide dimethyl acetal at 100 C.13 The condensation of 2 with the appropriate guanidine or carboxamidine under basic conditions yielded the 4-(2-pyridyl)- pyrimidine ligands L1–L5 in good to excellent yields.The functionalized amino acids with metal-binding sites such as L5 are attractive building blocks for the construction of synthetic peptides.14 Since only a few examples of functionalized amino acids with bidentate metal-binding sites have been described, it was interesting to use this straightforward procedure in the preparation of new functionalized amino acids using the Na- Boc-.-Arg as the guanidine reagent.Thus, the reaction of 2 with 1 equivalent of Na-Boc-.-Arg in hot absolute ethanol in the presence of 2 equivalents of sodium ethoxide yielded L5 without loss of the tert-butoxycarbonyl group, in quantitative yield.15 Spectroscopy. Electronic absorption spectra of ligands L1, L3, L4 and L5 are shown in Fig. 1. The corresponding experimental and calculated absorption maxima for all ligands are presented in Table 1. The geometries of cis and trans conformers of bpy have been determined at the MP2/6-31G**//HF/6-31G** level by Howard.16 The trans conformer is predicted to be lower in energy but the cis–trans interconversion barrier is only of 6 kJ mol1.In organic solvents, dipole measurements indicate a trans non-planar con.guration.16 The calculated spectra are in good agreement with previous INDO/S CI calculations 17 and with experimental results if we assume that bpy is in the trans3098 J.Chem. Soc., Dalton Trans., 1999, 3095¡V3101conformation. The experimental and calculated absorptionspectra of L1 and L2 are similar to the bpy spectrum. FromZINDO calculations we have attributed the absorption bandsto strong £k ¡÷ £k* transitions. Absorption spectra of L3, L4and L5 present a supplementary band at wavelengths above 320nm. For L4 and L5 this band is very large and is characterizedby a molar absorption coecient £` less intense than that of the£k¡V£k* bands. It can be attributed to a n¡V£k* transition implying alone pair of electrons located on one of the nitrogen atoms.ForL3, the band at 324 nm is relatively ne and it seems dicultwith the present data to conclude about its nature (n¡V£k* or£k¡V£k* transition). The absence of any calculated absorptionband at a wavelength longer than 300 nm does not agreewith experimental data. A possible explanation would be thepresence of a species due to hydrogen-bonded interactionswith ethanol in the rst solvation shell of 2-aminopyrimidinecompounds.18Redox properties. Peak potentials of the ¡§free¡� ligands aregiven in Table 1.We observed that L1 is easier to reduce thanbpy. In fact, the mesomeric donor eect of the nitrogen atom atthe 1 position of the pyrimidine induces an electronic depletionaround this nitrogen. This increases the electron attractorpower of L1. The reduction potentials of ligands L3, L4 andL5 are similar and more negative than that of L1. Indeed, thelone pair of the nitrogen atom of the substituent at the 2position to the pyrimidine increases the electron density via adonor mesomeric eect making the reduction of the ligandsmore dicult.The reduction potential of L2 is intermediate(1.85 V) between those of L1 and L3. This is in line with thewell known inductive eect of a methyl group which is lessintense than the mesomeric eect.ComplexesSynthesis. Ruthenium() complexes containing the novelligands were synthesized by utilizing [Ru(bpy)2Cl2] as the sourceof the metal fragment as shown in Scheme 1 (ii).The complex[Ru(bpy)2Cl2] reacted for 16 h with the respective ligands in aTable 1 Experimental and calculated absorption maxima in ethanoland redox potentials in acetonitrile for the dierent ligands at 298 K£fmax/nm (£`/l mol1 cm1)calculatedEred/V vs.Ligand a b experimental SCEbpyL1L2L3L4 cL5273235229284240278234280235296289235296290237283 (14480)244 (sh) (9950)236 (12240)279 (12820)242 (sh) (6880)235 (7620)280 (15240)244 (sh) (7360)237 (8710)324 (7660)286 (sh) (7280)276 (9380)241 (23010)352 (sh) (2110)338 (2430)287 (sh) (2550)273 (sh) (4760)253 (13100)242 (sh) (11390)342 (3730)288 (sh) (4440)277 (sh) (7260)253 (19830)2.181.781.851.901.981.95a cis Conformer. b trans Conformer.c Calculation performed on NHCH3instead of NHCH2CH3.boiling 75 :25 solution of EtOH¡Vwater.19 After removal ofethanol under reduced pressure and ltration of the aqueoussolution, the complexes were precipitated with an excess ofNH4PF6.They were obtained in excellent yield with high purityas their hexauorophosphate salts and characterized by NMRspectroscopy, mass spectrometry, electrochemistry and absorptionspectroscopy. An X-ray structural analysis of [Ru(bpy)2-(L2)]2 was carried out.Structure of [Ru(bpy)2(L2)]2, 2PF6. The asymmetric unitcontains two cations and four anions. The two cations arenearly identical; only one (molecule 1) is shown on Fig. 2 withthe atom labelling scheme.The ruthenium¡Vnitrogen bondlengths range from 2.022(7) to 2.137(6) (Table 3). The longestis observed for the pyrimidine ring 2.137(6) [2.109(7) ].Similar lengthening were observed in related complexes with£\ substituted bpy or related ligands.6b,20 As suggested by theFig. 1 Absorption spectra of ligands L1 (a), L3 (b), L4 (c) and L5 (d),in ethanol (105 mol l1).J. Chem. Soc., Dalton Trans., 1999, 3095.3101 3099 Table 2 Crystallographic data for [Ru(bpy)2(L2)][PF6]2 recorded at 160 K Empirical formula C30H25F12N7P2Ru M Shape (color) Crystal system Space group a/A b/A c/A ¥â/ V/A3 ZR (int) R wR2 Goodness of .t 874.58 Box (dark red) Monoclinic P21 12.2871(17) 10.5692(12) 25.100(4) 95.266(17) 3245.9(8) 4 0.0285 0.0348 0.0841 1.024 short contact between one H of the methyl substituent and the nitrogen N(16), 2.563 A [2.558 A], this increase in bond length might result from the steric interaction between this methyl substituent on the pyrimidine ring and the bpy ring 5.It is indeed this bpy ligand which exhibits the larger twisting 8.4(4) [8.9(5)] about the interannular C.C bond. The other bpy ligand displays an inter-ring angle of only 5.6(1) [5.6(2)], whereas the two rings of the (pyridyl) pyrimidine are almost planar, interplanar angle 1.2(5) [2.7(5)]. However, an electronic in.uence of the pyrimidine ring itself could not be ruled out. Spectroscopy. The absorption data for all the complexes are summarized in Table 4.The absorption spectra of [Ru(bpy)2- (L1)]2 and [Ru(bpy)2(L5)]2 determined in ethanol solutions are presented in Fig. 3. By comparison with the reference product, i.e. [Ru(bpy)3]2, the absorption band in the visible region has been assigned to a metal-to-ligand charge transfer (MLCT) Table 3 Selected interatomic distances [A] and bond angles [] for [Ru(bpy)2(L2)][PF6]2 Molecule 1 Molecule 2 Ru(1).N(11) Ru(1).N(12) Ru(1).N(13) Ru(1).N(14) Ru(1).N(15) Ru(1).N(16) N(11).C(111) N(11).C(115) N(12).C(121) N(12).C(125) N(13).C(131) N(13).C(135) N(14).C(141) N(14).C(145) N(15).C(151) N(15).C(155) N(16).C(161) N(16).C(165) N(114).C(113) N(114).C(115) N(14).Ru(1).N(13) N(14).Ru(1).N(16) N(13).Ru(1).N(16) N(14).Ru(1).N(12) N(13).Ru(1).N(12) N(16).Ru(1).N(12) N(14).Ru(1).N(15) N(13).Ru(1).N(15) N(16).Ru(1).N(15)(12).Ru(1).N(15) N(14).Ru(1).N(11) N(13).Ru(1).N(11) N(16).Ru(1).N(11) N(12).Ru(1).N(11) N(15).Ru(1).N(11) C(111).N(11).C(115) C(111).N(11).Ru(1) C(115).N(11).Ru(1) C(121).N(12).C(125) C(121).N(12).Ru(1) C(125).N(12).Ru(1) C(135).N(13).C(131) C(135).N(13).Ru(1) C(131).N(13).Ru(1) C(141).N(14).C(145) C(141).N(14).Ru(1) C(145).N(14).Ru(1) C(155).N(15).C(151) C(155).N(15).Ru(1) C(151).N(15).Ru(1) C(165).N(16).C(161) C(165).N(16).Ru(1) C(161).N(16).Ru(1) C(113).N(114).C(115) 2.137(6) 2.060(4) 2.051(6) 2.022(7) 2.088(7) 2.051(5) 1.356(8) 1.372(10) 1.321(9) 1.369(10) 1.355(10) 1.354(9) 1.318(10) 1.375(9) 1.356(10) 1.315(10) 1.377(8) 1.333(8) 1.301(11) 1.353(10) 78.5(3) 84.8(3) 96.2(2) 93.9(3) 88.3(3) 174.91(19) 96.05(17) 173.4(3) 79.5(3) 95.8(3) 170.4(3) 96.4(2) 103.9(2) 77.8(2) 89.5(3) 117.4(6) 113.1(4) 129.5(5) 117.2(5) 117.6(5) 125.1(5) 117.2(7) 126.5(5) 116.3(4) 118.4(6) 117.3(5) 123.8(5) 121.3(7) 124.6(6) 113.8(5) 118.4(5) 126.3(4) 115.2(4) 118.2(7) Ru(2).N(21) Ru(2).N(22) Ru(2).N(23) Ru(2).N(24) Ru(2).N(25) Ru(2).N(26) N(21).C(211) N(21).C(215) N(22).C(221) N(22).C(225) N(23).C(231) N(23).C(235) N(24).C(241) N(24).C(245) N(25).C(251) N(25).C(255) N(26).C(261) N(26).C(265) N(214).C(213) N(214).C(215) N(24).Ru(2).N(23) N(24).Ru(2).N(26) N(26).Ru(2).N(23) N(22).Ru(2).N(24) N(22).Ru(2).N(23) N(22).Ru(2).N(26) N(25).Ru(2).N(24) N(25).Ru(2).N(23) N(25).Ru(2).N(26) N(25).Ru(2).N(22) N(24).Ru(2).N(21) N(23).Ru(2).N(21) N(26).Ru(2).N(21) N(22).Ru(2).N(21) N(25).Ru(2).N(21) C(215).N(21).C(211) C(211).N(21).Ru(2) C(215).N(21).Ru(2) C(225).N(22).C(221) C(221).N(22).Ru(2) C(225).N(22).Ru(2) C(235).N(23).C(231) C(235).N(23).Ru(2) C(231).N(23).Ru(2) C(245).N(24).C(241) C(241).N(24).Ru(2) C(245).N(24).Ru(2) C(251).N(25).C(255) C(255).N(25).Ru(2) C(251).N(25).Ru(2) C(261).N(26).C(265) C(265).N(26).Ru(2) C(261).N(26).Ru(2) C(215).N(214).C(213) 2.109(7) 2.050(5) 2.061(7) 2.054(8) 2.038(8) 2.060(5) 1.349(10) 1.337(11) 1.398(9) 1.328(10) 1.372(10) 1.344(9) 1.400(11) 1.308(10) 1.340(11) 1.396(11) 1.347(9) 1.362(8) 1.341(13) 1.331(13) 80.1(3) 87.7(3) 96.7(2) 92.3(3) 87.8(3) 175.5(2) 95.09(19) 173.1(3) 78.1(3) 97.4(3) 170.5(3) 96.3(2) 101.5(2) 78.8(3) 89.1(3) 114.4(7) 114.5(5) 131.1(6) 117.7(5) 115.8(5) 126.4(5) 119.2(7) 127.0(6) 113.8(5) 118.5(7) 113.7(5) 127.5(6) 116.2(8) 126.2(7) 117.5(6) 117.9(6) 125.3(4) 116.8(5) 117.2(8)3100 J.Chem. Soc., Dalton Trans., 1999, 3095¡V3101Table 4 Absorption parameters in ethanol solution for ruthenium() complexes (105 mol l1)Complex £fmax/nm (£`/l mol1 cm1)[Ru(bpy)3]2[Ru(bpy)2(L1)]2[Ru(bpy)2(L2)]2[Ru(bpy)2(L3)]2[Ru(bpy)2(L4)]2[Ru(bpy)2(L5)]2482 (sh) (5310)480 (sh) (3620)483 (sh) (11250)487 (sh) (7950)450 (13700)441 (7210)442 (5140)449 (6610)447 (17290)448 (10950)287 (80290)286 (52260)286 (44940)289 (31390)288 (85500)289 (61890)245 (25620)245 (15960)243 (20870)236 (13210)244 (47300)245 (29500)Table 5 Redox potentials in acetonitrile for ruthenium() complexesE1/2/V vs.SCEComplex OxidationReductionE1/2/V[Ru(bpy)3]2[Ru(bpy)2(L1)]2[Ru(bpy)2(L2)]2[Ru(bpy)2(L3)]2[Ru(bpy)2(L4)]2[Ru(bpy)2(L5)]21.301.331.381.341.341.341.301.011.121.111.081.101.481.401.451.431.321.421.731.701.701.671.551.652.602.342.502.452.422.44transition.21 The ZINDO calculation performed on [Ru(bpy)2-(L2)]2 suggests several bands (from 446 to 400 nm) in the visibleregion which can be assigned to a metal-to-ligand charge transferassociated with the promotion of a ruthenium d electroninto a £k* orbital of bpy and (or) pyridylpyrimidine ligands.In the concerned excited state the electron is mainly localizedon the pyridylpyrimidine ligand.A more intense band appears in the region 280¡V320 nmwhich can be associated with a £k ¡÷ £k* ligand-centred (LC)transition 22 in agreement with the ZINDO calculation.Moreoverthe calculation results in this region lead to the attributionof several bands to d ¡÷ d* transitions (MC). A third absorptionband appearing in the 230¡V260 nm region of the experimentalspectra can be assigned to MLCT transitions accordingto ZINDO calculations.Redox properties.The redox potentials of all the complexesas determined by cyclic voltammetry in acetonitrile are inFig. 2 Molecular view of the cation [Ru(bpy)2(L2)]2 (molecule 1).Ellipsoids represent 50% probability. Hydrogen atoms have been omittedfor clarity.Table 5. The voltammograms exhibit, in each case, onereversible oxidation wave and three reversible reduction waves.By comparison with the oxidation potential of the referencecomplex, we attribute the potentials in the range 1.33 to1.38 V vs.SCE to the redox couple Ru3¡VRu2of the heterolepticcomplexes. These potentials are related to the energy ofthe HOMO orbital of the metal.23 The potentials of the secondand third reduction of the heteroleptic complexes are similarto the corresponding potentials of the reference complex. Thisindicates that these two reductions take place on the bpyligands of the heteroleptic complexes. The rst reductionpotentials ranging from 1.01 to 1.12 V vs.SCE are lessnegative than that of [Ru(bpy)3]2 (1.30 V vs. SCE).Consequently, it is clear that the rst reduction can only beattributed to the redox couple involving the pyridylpyrimidineligand. The corresponding potential is related to the energy ofthe £k* (LUMO) ligand orbital. The successive reductions of theheteroleptic complexes take place according to eqns. (1)¡V(3).[RuII(bpy)2(L)]2 e ¡÷ [RuII(bpy)2(L)] (1)[RuII(bpy)2(L)] e ¡÷ [RuII(bpy)(bpy)(L)] (2)Fig. 3 Absorption spectra of complexes [Ru(bpy)2(L1)]2 (full line)and [Ru(bpy)2(L5)]2 (dotted line) in ethanol (105 mol l1).J. Chem. Soc., Dalton Trans., 1999, 3095.3101 3101 [RuII(bpy)(bpy)(L)] e .¡æ [RuII(bpy)(bpy)(L)] (3) The comparison between the .rst reduction potential values shows that the non-substituted ligand L1 in the complex is easier to reduce than the substituted one (by about 100 mV). Consequently, the ¥�* antibonding orbital (LUMO) of L1 is that of lowest energy.This result is in agreement with those obtained for the ¡°free¡± ligand. The ZINDO calculated levels of the orbitals for the ¡°free¡± ligands in their trans conformations are correlated with the electrochemical data. The di.erence .E1/2 between the oxidation potential Eox and the .rst reduction Ered1 of the heteroleptic complexes is correlated with the di.erence between the HOMO and LUMO energy, i.e. to the energy at the maximum of the absorption band located in the visible region.This result reinforces the attribution of this band mainly to the transition S0 .¡æ 1MLCT(L). Thus, the optical and redox orbitals are of similar nature, and the MLCT excited state of the heteroleptic complexes [Ru(bpy)2(L)]2 corresponds essentially to the Ru¡æL transition. In summary, we have prepared, characterized and studied the spectroscopic and electrochemical properties of 4-(2- pyridyl)pyrimidine asymmetric ligands (L) and their heteroleptic ruthenium(..) complexes [Ru(bpy)2(L)]2.The presence of amine and/or amino acid functions on the 4-(2-pyridyl)- pyrimidine ligand makes the corresponding complexes interesting candidates for immobilization on solid supports or incorporation into dendrimers. It is interesting that ligand L5 which is a functionalized amino acid with metal-binding sites provides new opportunities for the construction of synthetic peptides. The synthesis of such new peptides containing L5 or its corresponding complex [Ru(bpy)2(L5)]2 is currently under investigation.Acknowledgements F. B.-P., L. B. and E. A. are indebted to Dr B. Levy for very helpful discussions and valuable suggestions regarding the theoretical calculations. Financial support from the Centre National de la Recherche Scienti.que and for a doctoral fellowship (E. B.) from the ¡°Ministere des A.aires Etrangeres¡± are gratefully acknowledged. References 1 E. Amouyal, Sol. Energy Mater.Sol. Cells, 1995, 38, 249; in Homogeneous Photocatalysis, ed. M. Chanon, Wiley, Chichester, 1997, ch. 8, p. 263; A. Kay and M. Gratzel, Sol Energy Mater. Sol. Cells, 1996, 44, 9 R. Argazzi, C. A. Bignozzi, G. M. Hasselman and G. J. Meyer, Inorg. Chem., 1998, 37, 4533. 2 V. Balzani and F. Scandola, Supramolecular Photochemistry, Ellis Horwood, New York, 1991; A. Hatzidimitriou, A. Gourdon, J. Devillers, J. P. Launay, E. Mena and E. Amouyal, Inorg. Chem., 1996, 35, 2212. 3 J. K. Barton, A. T. Danishefksy and J. M. Goldberg, J. Am. Chem. Soc., 1984, 106, 2172; J. Kelly, A. Tossi, D. McConnel and C. OhUigin, Nucleic Acids Res., 1985, 13, 6017. 4 K. Kalyanasundaram, Coord. Chem. Rev, 1982, 46, 159; A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser and A. Von Zelewsky, Coord. Chem. Rev., 1988, 84, 85; E. C. Constable, Adv. Inorg. Chem., 1989, 34, 1; Prog. Inorg. Chem., 1994, 42, 67. 5 Y. Kawanishi, N. Kitamura and S. Tazuke, Inorg. Chem., 1989, 28, 2968; F.Casalboni, Q. G. Mulazzani, C. D. Clark, M. Z. Ho.man, P. L. Orizondo, M. W. Perkovic and D. P. Rillema, Inorg. Chem., 1997, 36, 2252. 6 (a) E. Bejan, H. Ait-Haddou, J.-C. Daran and G. G. A. Balavoine, Synthesis, 1996, 1012; (b) E. Amouyal, F. Penaud-Berruyer, D. Azhari, H. Ait-Haddou, C. Fontenas, E. Bejan, J.-C. Daran and G. A. A. Balavoine, New J. Chem., 1998, 22, 373. 7 I. Sasaki, M. Imberdis, A. Gaudemer, B. Drahi, D. Azhari and E. Amouyal, New J. Chem., 1994, 18, 759; D.Esposito, G. Del Vecchio and G. Barone, J. Am. Chem. Soc., 1997, 119, 2606 and refs. therein; R. H. Terbrueggen, T. W. Johann and J. K. Barton, Inorg. Chem., 1998, 37, 6874. 8 A. Altomare, G. Cascarano, G. Giacovazzo, A. Guagliardi, M. C. Burla, G. Polidori and M. Camalli, SIR 92, a program for automatic solution of crystal structures by direct methods, J. Appl. Crystallogr., 1994, 27, 435. 9 G. M. Sheldrick, SHELXL 97, program for crystal structure re.nement, University of Gottingen, 1997. 10 D. J. Watkin, C. K. Prout and L. J. Pearce, CAMERON, Chemical Crystallography Laboratory, University of Oxford, Oxford, 1996. 11 A. D. Bacon and M. C. Zerner, Theor. Chim. Acta, 1979, 53, 21. 12 W. P. Anderson, T. R. Cundari and M. C. Zerner, Int. J. Quantum Chem., 1991, 39, 31. 13 R. F. Abdulla and R. V. Brinkmeyer, Tetrahedron , 1979, 35, 1675. 14 B. Imperiali and S.L. Fisher, J. Am. Chem. Soc., 1991, 113, 8527; B. Imperiali, T. J. Prins and S. T. Fisher, J. Org. Chem., 1993, 58, 1613; M. R. Ghadiri, C. Soares and C. Choi, J. Am. Chem. Soc., 1992, 114, 825; S. R. Wilson, A. Yasmin and Y. Wu, J. Org. Chem., 1992, 57, 6941; P. R. Cheng, S. L. Fisher and B. Imperiali, J. Am. Chem. Soc., 1996, 118, 11349. 15 E. Bejan, H. Ait-Haddou, J.-C. Daran and G. G. A. Balavoine, Eur. J. Org. Chem., 1998, 2907. 16 S. T. Howard, J. Am. Chem. Soc., 1996, 118, 10268. 17 S. J. Milder, Inorg. Chem., 1989, 28, 868. 18 J. N. Spencer, S. W. Barton, K. A. Smith, W. S. Wolbach, J. F. Powell, M. R. Kirschenbaum and D. W. Firth, Can. J. Chem., 1982, 61, 194. 19 R. P. Thummel and F. Lefoulon, Inorg. Chem., 1987, 26, 675. 20 D. P. Rillema, D. G. Taghdiri, D. S. Jones, C. D. Keller, L. A. Worl, T. J. Meyer and H. A. Levy, Inorg. Chem., 1987, 26, 578; H. Ichida, S. Tachiyashiki and Y. Sasaki, Chem. Lett., Chem. Soc. Jpn., 1990, 63, 1299; E. Kimura, S. Wada, M. Shionoya, T. Takahashi and Y. Iitaka, J. Chem. Soc., Chem. Commun., 1990, 397; R. Chotalia, E. C. Constable, M. J. Hannon and D. A. Tocher, J. Chem. Soc., Dalton Trans., 1995, 3571; D. A. Bardwell, F. Barigelletti, R. L. Cleary, L. Flamigni, M. Guardigli, J. C. Je.ery and M. D. Ward, Inorg. Chem., 1995, 34, 2438. 21 G. Calzaferri and R. Rytz, J. Phys. Chem., 1995, 99, 12141; B. J. Coe, D. W. Thompson, C. T. Culbertson, J. R. Schoonover and T. J. Meyer, Inorg. Chem., 1995, 34, 3385. 22 R. P. Thummel, F. Lefoulon and J. D. Korp, Inorg. Chem., 1987, 26, 2370. 23 Y. Kawanishi, N. Kitamura, Y. Kim and S. Tazuke, Sci. Pap. Inst. Phys. Chem. Res. (Jpn.), 1984, 78, 212; A. Vlcek, Chemtracts-Inorg. Chem., 1991, 5, 2144. Paper 9/04617H

ISSN:1477-9226    

DOI:10.1039/a904617h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3103–3113 3103 Extending knowledge on the nucleophilicity of the {Pt2S2} core: Ph2PCH2CH2PPh2 as an alternative terminal ligand in [L2Pt(-S)2PtL2] metalloligands † Mercè Capdevila,a Yolanda Carrasco,a William Clegg,b Robert A. Coxall,b Pilar González-Duarte,*a Agustí Lledós a and José Antonio Ramírez c a Departament de Química, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain. E-mail: Pilar.Gonzalez.Duarte@uab.es b Department of Chemistry, University of Newcastle, Newcastle upon Tyne, UK NE1 7RU c Departament de Química Inorgànica, Facultat de Química, Universitat de València, Dr.Moliner 50, 46100 Burjassot, València, Spain Received 17th May 1999, Accepted 7th July 1999 The reaction of [Pt2(dppe)2(µ-S)2] 1, with metal complexes or metal salts gave di.erent types of complexes depending on the nature of the heterometal and of the stoichiometric ratios employed. Thus, a trinuclear complex of formula [Pt(dppe){Pt2(dppe)2(µ3-S)2}]Cl2 2 and an apparently mixed PdxPt3-x product 3 have been prepared and characterised. Alternatively, dissolution of 1 in chlorinated solvents a.ords 2 easily and 3 is formed from 1 with [PdCl2(dppe)].The pentanuclear complexes of formula [M{Pt2(dppe)2(µ3-S)2}2]X2 (M = Zn 4 or Cd 5, X = ClO4; M = Cd, X2 = [CdCl4] 5; M = Hg, X2 = [PF6][HgCl4]0.5 6) have been obtained. The structures of complexes 2, 3, 4, 5, 5 and 6 have been determined crystallographically.Complex 2 comprises three slightly distorted square-planar cis-PtP2S2 co-ordination planes sharing two µ3-S ligands. X-Ray data and NMR studies in solution of di.erent crops of crystals support that 3 is essentially a simple solid-solution mixture of pure complexes [Pd(dppe){Pt2(dppe)2(µ3-S)2}][BPh4]2 3 and [Pt(dppe)- {Pd2(dppe)2(µ3-S)2}][BPh4]2 3 in variable proportions with at most a minor component of the homometallic complex 2. The structure of the cations of complexes 4, 5, 5 and 6 comprises two {Pt2S2} butter.ies linked through sulfur to the metal(..) ion, which shows a signi.cantly distorted tetrahedral environment. All complexes have been fully characterised by multinuclear NMR techniques and the corresponding parameters are reported.Introduction The signi.cant number of known homo- and hetero-metallic derivatives of [L2Pt(µ-S)2PtL2], L = phosphine, shows that this species is one of the most e.ective metalloligands identi.ed to date.This can mainly be attributed to the geometric features of the central Pt2S2 ring with a .exible hinge angle between the two PtIIS2 planes, and to the ability of the bridging sul.de ligands to co-ordinate additional metal ions. At the same time the wide range of co-ordination environments o.ered by main group or transition metals together with the nature of their accompanying species, either ligands or counter ions, enhances the structural diversity of this family of derivatives.An extensive and rigorous account on the synthesis, structures and reactivities of aggregates containing the {M2S2} core (M = Pd or Pt) has recently been provided by Fong and Hor.1 Derivatives of [L2- Pt(µ-S)2PtL2] of known structure are summarised in Scheme 1, which includes varied co-ordination geometries about the heterometal: linear,2 angular,3 T-shaped,4 Y-shaped,5 tetrahedral,6 square-planar,7 square-pyramidal,5c,6b distorted trigonal prismatic 8 and others,9 and displays the ability of the [L2Pt(µ-S)2- PtL2] metalloligand to function as a unidentate,2 bridging,4,9 or chelating ligand.3,5–8 Despite the high number of derivatives based on the Pt(µ-S)2- Pt core,2–10 the synthetic routes reported for complexes of formula [L2Pt(µ-S)2PtL2] are not straightforward,1 often leading to mixtures of products, and reports on their crystal † Supplementary data available: NMR spectra for complex 2.For direct electronic access see http://www.rsc.org/suppdata/dt/1999/3103/, otherwise available from BLDSC (No.SUP 57611, 2 pp.) or the RSC Library. See Instructions for authors, 1999, Issue 1 (http://www.rsc.org/dalton). structures are scarce. We note that synthesis of most of the sul.de-bridged aggregates with the {M2S2} core has been achieved using monodentate phosphines, especially PPh3, as a terminal ligand in the [L2M(µ-S)2ML2] unit. Our group showed the possibilities o.ered by the use of bidentate phosphines.Indeed, we recently obtained signi.cant yields of pure [Pt2- (dppe)2(µ-S)2] by monitoring the reaction by means of 31P NMR and determined its structure.11 This showed that the central Pt2S2 ring is hinged in agreement with our previous theoretical ab initio studies for the [Pt2(PH3)4(µ-S)2] complex.12 In addition, the preparation of [Pt2(dppe)2(µ-S)2] allowed the synthesis and structure determination of pentanuclear [Cu{Pt2(dppe)2(µ3-S)2}2]2, which constituted the .rst example of an homoleptic copper(..) sul.de complex.11 No example structurally characterised by X-ray di.raction where the heterometal is tetrahedrally co-ordinated to two [L2Pt(µ-S)2PtL2] molecules had previously been reported.In this work we report on the synthesis and characterisation of homo- and hetero-nuclear complexes derived from [Pt2- (dppe)2(µ-S)2] 1. The crystalline nature of the trinuclear complex [Pt(dppe){Pt2(dppe)2(µ-S)2}]Cl2 2 (i.e. [Pt3(dppe)3(µ3-S)2]- Cl2) and the pentanuclear complexes [M{Pt2(dppe)2(µ-S)2}2]X2 (M = Zn 4 or Cd 5, X = ClO4; M = Cd, X2 = [CdCl4] 5; M = Hg, X2 = [PF6][HgCl4]0.5 6) has enabled us to determine their molecular structures, which are preserved in solution according to multinuclear 31P-{1H}, 195Pt-{1H}, 113Cd-{1H}, 111Cd-{1H} and 199Hg-{1H} NMR data.The reaction of 1 with [PdCl2(dppe)] a.orded apparently a series of complexes of formula [PdxPt3-x(dppe)3(µ3-S)2][BPh4]2 3, the values of x being determined by the solubility of pure [Pd(dppe){Pt2(dppe)2- (µ3-S)2}][BPh4]2 3 and [Pt(dppe){Pd2(dppe)2(µ3-S)2}][BPh4]2 3 and the concentration of the reactants.Single-crystal di.raction3104 J. Chem. Soc., Dalton Trans., 1999, 3103–3113 Scheme 1 studies together with analytical and 31P-{1H} and 195Pt-{1H} NMR data indicate that the cation of 3 is essentially a mixture of those in 2, 3 and 3. Experimental General remarks Metal complexes of formula [MCl2(dppe)] were prepared according to published methods, M = Pd13 or Pt,14 with minor modi.cations.The synthesis of complex 1 has been described.11 In the following preparations conventionally dried and degassed solvents were used and the manipulations were carried out under an atmosphere of pure dry nitrogen. CAUTION: perchlorate salts of metal complexes with organic ligands are potentially explosive. Microanalyses were performed with a Carlo-Erba NA-1500 analyser. Infrared spectra in the range 4000–400 cm1 were recorded from KBr discs on a Perkin-Elmer 1710 spectrophotometer, all NMR spectra in (CD3)2SO solution at room temperature (31P-{1H}, 195Pt-{1H}, 113Cd-{1H} and 111Cd-{1H} at 121.4, 64.2, 66.5 and 63.6 MHz) on a Varian UNITY300 spectrometer.Some 31P-{1H} NMR spectra have been recorded at 101.2 MHz on a Bruker AM 250 MHz, and some 195Pt-{1H} and 199Hg-{1H} on a Bruker AM-400 spectrometer at 85.6 and 71.6 MHz, respectively. The best 113Cd-{1H} and 111Cd-{1H} NMR spectra were obtained at 44.4 and 42.4 MHz, respectively, on a Bruker AM 200 MHz spectrometer.The 113Cd and 111Cd nuclei, I = 1/2, have similar abundance and neither is especially di.cult to observe; 113Cd is somewhat less abundant (12.26%) but has ca. 10% more sensitivity, and 111Cd (12.75%) has a resonance frequency very close to that of 195Pt, so that both can be measured simultaneously without readjusting the spectrometer probe. The 31P and 195Pt chemical shifts are relative to external 85% H3PO4 and 0.1 mol dm3 Na2PtCl6, respectively, 113Cd and 111Cd to external 0.1 mol dm3 Cd(NO3)2 aqueous solutions [d(Cd(NO3)2) = d(Cd(ClO4)2) 12.7] and 199Hg to 1 mol dm3 HgI2 as reference but the values given are referenced to HgMe2 [d(HgMe2) = d(HgI2) 3106].15 The NMR spectra were simulated on a Pentium-200 computer using the gNMR V4.0.1 program16 which allows iteration from the experimental spectra, considers all isotopomers and works very well for species containing at most twelve NMR active nuclei.Preparations [Pt3(dppe)3(3-S)2Cl2 2. To a solution of complex 1 (0.100 g, 0.08 mmol) in acetone (75 ml) solid [PtCl2(dppe)] (0.053 g, 0.08 mmol) was added with stirring. After 24 h the yellow solution was concentrated to dryness a.ording the expected product (Found: C, 38.20; H, 3.00; S, 2.45. Calc. for C78H72Cl2P6Pt3S2 6CHCl3: C, 38.35; H, 3.00; S, 2.45%). Alternatively, complex 2 was obtained by slow evaporation of a solution of 1 in CHCl3 at room temperature.This second method yielded yellow crystals, which were suitable for X-ray analysis. [PdxPt3-x(dppe)3(3-S)2][BPh4]2 3. To a solution of complex 1 (0.110 g, 0.09 mmol) in acetonitrile (50 ml) was added [PdCl2(dppe)] (0.050 g, 0.09 mmol). After 24 h of stirring the mixture became an almost clear solution. The stoichiometrically required amount of sodium tetraphenylborate (0.060 g, 0.18 mmol) was then added. The white solid formed, NaCl, was .ltered o.and the .ltrate allowed to stand at 4 C for 12 h. At this stage several crops of crystals were subsequently separated from the mother solution. Analytical data for the di.erent crops showed that the value of x in the formula of this material was variable. A single crystal resulting from a middle crop was chosen for X-ray di.raction. [Zn{Pt2(dppe)2(3-S)2}2][ClO4]2 4. A solution of Zn(ClO4)2 (0.027 g, 0.10 mmol) in 15 ml methanol was added to a solution of complex 1 (0.220, 0.18 mmol) in the same solvent (100 ml).After 30 min of stirring the resultant pale yellow solution was concentrated. Addition of diethyl ether resulted in the appearance of a white solid that was .ltered o., washed with cold methanol and dried. Yield 60% (Found: C, 43.55; H, 3.60; S, 4.70. Calc. for C104H96Cl2O8P8Pt4S4Zn: C, 45.15; H, 3.60; S, 4.65%). Recrystallisation of 4 in methanol gave rise to yellow crystals adequate for X-ray di.raction. Alternatively, the same cationic [Zn{Pt2(dppe)2(µ3-S)2}2]2 species with [ZnCl4]2-J.Chem. Soc., Dalton Trans., 1999, 3103–3113 3105 Scheme 2 instead of ClO4 counter ions was obtained by treating 1 with ZnCl2 in the same solvent at an 1 : 1 stoichiometric ratio. [Cd{Pt2(dppe)2(3-S)2}2][ClO4]2 5. By the same procedure as that indicated for complex 4, a white solid separated from a methanolic solution (100 ml) containing Cd(ClO4)26H2O (0.034 g, 0.08 mmol) and 1 (0.195 g, 0.16 mmol).Yield 85% (Found: C, 43.70; H, 3.45; S, 4.35. Calc. for C104H96CdCl2O8- P8Pt4S4: C, 44.40; H, 3.45; S, 4.55%). Recrystallisation of 5 in acetone allowed isolation of colourless X-ray quality crystals. [Cd{Pt2(dppe)2(3-S)2}2][CdCl4] 5. By an analogous procedure, the reaction of CdCl22½H2O (0.046 g, 0.20 mmol) and complex 1 (0.250 g, 0.20 mmol) in methanol (100 ml) a.orded a white solid. Yield: 75% (Found: C, 41.50; H, 3.50; S, 4.40. Calc. for C52H48CdCl2P4Pt2S2: C, 43.50; H, 3.35; S, 4.45%).Recrystallization of 5 in acetone a.orded colourless single crystals. [Hg{Pt2(dppe)2(3-S)2}2][PF6][HgCl4]0.5 6. Solid HgCl2 (0.016 g, 0.06 mmol) was added to a solution of complex 1 (0.150 g, 0.11 mmol) in chloroform (50 ml) and the suspension stirred for 1 h. The salt KPF6 (0.022 g, 0.11 mmol) was then added to the resultant yellow solution and the reaction mixture slowly evaporated in air and room temperature. After several days a microcrystalline orange solid was collected (Found: C, 39.95; H, 3.00; S, 3.80.Calc. for C104H96Cl2F6Hg1.5P9Pt4S4: C, 41.30; H, 3.20; S, 4.20%). Recrystallisation of 6 from acetone gave orange crystals suitable for X-ray di.raction. X-Ray crystallography Crystals of complexes 2, 3, 5, 5 and 6 were examined on Bruker AXS SMART CCD di.ractometers. The small crystal size and weak X-ray scattering of complex 4 made synchrotron data collection necessary; a conventional Mo-Ka X-ray source was used for the other complexes.Methods were as described previously. 17,18 Semi-empirical absorption corrections were applied. The structures were solved by direct methods, and were re.ned on F2 values for all unique re.ections. Disorder in phenyl rings and anions was modelled with the aid of restraints on geometry and displacement parameters. The disorder of Pt and Pd atoms on common sites in the structure of compound 3 was freely re.ned subject only to a total of one metal atom on each of the three sites; no overall Pt : Pd ratio was assumed.The largest residual electron density features (ranging from just over 1 e Å3 in compounds 1 and 2 to 4 e Å3 in 6) were close to metal atoms and disordered groups. Programs were standard manufacturers’ control and data processing software, together with SHELXTL19 and local programs. Crystal data are listed in Table 6, together with other information on the data collection and structure determination. CCDC reference number 186/1559.See http://www.rsc.org/suppdata/dt/1999/3103/ for crystallographic .les in .cif format. Results and discussion Synthesis of the complexes The procedures followed in the synthesis of compounds 2 to 6 are summarised in Scheme 2. The synthesis of the precursor species, [Pt2(dppe)2(µ-S)2] 1, has been reported.11 It requires careful monitoring of the reaction by 31P NMR in order to achieve a pure product in a reasonable yield. This diplatinum complex 1 is soluble in common organic solvents, but rapidly converts into the triplatinum complex 2 when dissolved in3106 J.Chem. Soc., Dalton Trans., 1999, 3103–3113 chlorinated solvents. With the palladium analogue of 1 the tendency to form the trinuclear species is signi.cantly enhanced.20 This could well explain why the complexes of formula [Pd2L4- (µ-S)2], L4 = 4 unidentate or 2 bidentate phosphine ligands, are poorly characterised and its chemistry virtually unknown.21 The crystal structure of the triplatinum complex 2 shows the presence of chloride counter ions. As this complex forms easily by dissolving 1 in CHCl3 or CH2Cl2 the only possible source of the chloride ions is the solvent molecules.This should involve the attack of the nucleophilic sul.do bridges of 1 on the chlorinated solvents, which act as electrophiles. The high nucleophilicity of the µ-thio ligands in [Pt2L4(µ-S)2] complexes, L4 = 4 unidentate or 2 bidentate phosphine ligands, towards halogenated solvents is well estabished.1 Thus, it has been reported that upon standing in CH2Cl2 complexes [Pt2L4(µ-S)2] give rise either to [PtL2(SCH2Cl)2], L2 = dppf 22 or (PPh3)2,10b or to [PtL2(S2CH2)], L = dppy (dppy = 2-diphenylphosphanopyridine). 23 However, on the one hand in these cases formation of the corresponding trinuclear derivatives was not described and on the other we have not identi.ed any secondary product accompanying the trinuclear compound 2. These data seem mutually complementary and suggest that dissolution of complexes of formula [Pt2L4(µ-S)2] in chlorinated solvents leads to formation of the [Pt3L6(µ3-S)2] trinuclear complexes together with the products of alkylation at sulfur, viz.[PtL2(SCH2Cl)2] or [PtL2(S2CH2)]. As an example, in the case of CH2Cl2 it seems reasonable to propose the following sequence of reactions: Probably, it is the solubility of each compound that determines its separation from the reaction mixture.A mechanism for the formation of various thiolato complexes from the disintegration of the {Pt2S2} core in CH2Cl2 has recently been proposed.1 However, in this mechanism the concomitant formation of triplatinum species has not been considered. It is likely that the presence of terminal bidentate phosphines makes the nucleophilic attack of the sul.de centre on solvent molecules more di.cult, thus preventing disintegration of the {Pt2S2} core. Alternatively, the triplatinum complex 2 can be obtained from 1 and [PtCl2(dppe)] in non-halogenated solvents.Interestingly, none of the structurally known trinuclear [M3L6(µ3-S)2] complexes, M = Ni,24 Pd25 or Pt,7d,e has been obtained from the dinuclear [M2L4(µ-S)2] unit. By analogy with the synthesis of the homonuclear Pt3 complex 2, it was expected that the reaction of 1 with [PdCl2(dppe)] would produce the heteronuclear Pt2Pd complex. However, this reaction apparently a.orded a range of products of formula [PdxPt3-x(dppe)3(µ3-S)2][BPh4]2 3, x being in no case exactly equal to 1 (Scheme 2).This observation can be explained on the assumption that 3 is largely a mixture of Pt3 2, PdPt2 3 and Pd2Pt 3, and is in accord with the fact that these pure complexes have di.erent solubilities in the reaction medium. Single-crystal X-ray analysis of a middle crop of crystals of 3 gives x = 1.32 as one of the re.nement results. Moreover, 31P-{1H} and 195Pt-{1H} NMR data, discussed later, indicate that di.erent solutions of 3 contain essentially 2, 3 and 3, those with higher content of Pt being the most soluble.The lack of structurally known pentanuclear aggregates of formula [M{Pt2L4(µ-S)2}2]2 with tetrahedral co-ordination around M led us previously to the Cu2 derivative.11 To extend this family to diamagnetic metal ions, the synthesis of the analogues of Zn, Cd and Hg was undertaken. The reaction of 1 with the corresponding metal perchlorates in a 2 to 1 molar ratio led to the expected pentanuclear complexes 4 and 5.However, the reaction of 1 with MCl2 in a 1 to 1 molar ratio a.orded the same [M{Pt2(dppe)2(µ-S)2}2]2 cations with [MCl4]2 (M = Zn or Cd 5) as counter ions instead of the trinuclear [MCl2{Pt2(dppe)2(µ-S)2}] species. Adducts of this formula with PPh3 instead of dppe ligands for M = Zn, Cd or Hg have been obtained very recently.26 As found in the reaction of 1 with ZnCl2 and CdCl2 in a 2 to 1 molar ratio, that with HgCl2 led also to the pentanuclear complex 6, but in this case the charge of two [Hg{Pt2(dppe)2(µ-S)2}2]2 units is compensated by one [HgCl4]2 and two PF6 anions.All these previous reactions occurred not only in MeOH but also in other organic solvents such as CH3CN or CHCl3. Nuclear magnetic resonance data In order to analyse the NMR data of complexes 2–6 it is convenient to consider those of 1 .rst because it is their basic structural entity. In 1, of known structure,11 both platinum atoms are square planar with a dihedral angle of 140 between their planes (Scheme 3).In this complex all the 31P as well as all the 195Pt nuclei show the same chemical shift and thus have become chemically equivalent in solution. However, neither the Scheme 3J. Chem. Soc., Dalton Trans., 1999, 3103–3113 3107 Table 1 The 31P and 195Pt NMR parameters for complexes 1, 2, 3, 3, 4, 5 and 6 Compound ä(195Pt) ä(31P) 1JPt-P/ Hz 2JPt-Pt/ Hz 3JPt-P/ Hz 4JP-P/ Hz 2JPt-M/ Hz 3JP-M/ Hz 123 3 456 [Pt2(dppe)2S2] [Pt3(dppe)3S2]2 [Pt2Pd(dppe)3S2]2 [PtPd2(dppe)3S2]2 [Zn{Pt2(dppe)2S2}2]2 [Cd{Pt2(dppe)2S2}2]2 [Hg{Pt2(dppe)2S2}2]2 4298 4571 4538 4510 4578 4467 4376 40.5 38.3 40.5 a 42.4 a 37.0 38.3 39.6 2740 3248 3166 3131 3120 3062 3067 680 740 600 360 180 200 44 22 23 36 32 34 25 <10 <10 <10 10 10 10 345 b 460 15 b 40 a Refers to the phosphorus atoms bound to platinum.b Refers to 113Cd coupling constants. 31P nor the 195Pt nuclei are magnetically equivalent and thus the total spin system is (AAIAIIAIII)XXI. For complex 1 the experimental 31P and 195Pt NMR spectra are shown in Fig. 1D and 1G, respectively. They can be interpreted with the aid of computer simulations. Initially, a .rst- Fig. 1 The 31P-{1H} (bottom) and 195Pt-{1H} (top) NMR spectra for complex 1, at 101.2 and 85.6 MHz, respectively: (A) 31P and (E) 195Pt computer simulation with no 2JPt-Pt coupling; (B) 31P and (F) 195Pt computer simulation including 2JPt-Pt coupling; (C) 31P computer simulation including 4JP-P couplings; (D) 31P and (G) 195Pt experimental spectra.order analysis of the 31P-{1H} spectrum (Fig. 1A) shows three distinct regions: down.eld and up.eld satellites due to 1JPt-P and a central signal including the singlet corresponding to the phosphorus bonded to magnetically inactive platinum atoms together with sidebands due to 3JPt-P. The up.eld satellite is a mirror image of the down.eld, and the relative intensities of these regions agree with the relative abundance of the isotopomers (PtPt 43.8; *PtPt 44.8; *Pt*Pt 11.4%).Second order e.ects due to 2JPt-Pt coupling, which is only e.ective for the *Pt*Pt isotopomer, give rise to new lines separated by 2JPt-Pt in the down.eld and up.eld satellites (Fig. 1B). Consideration of an additional 4JP-P coupling (4JPA-PAIII = 4JPAI-PAII and 4JPA-PAII = 4JPAI-PAIII) leads to a general broadening of the spectral linewidth.However, for complex 1 inclusion of these long-range couplings is necessary to obtain a good .t between simulated and experimental spectra (Fig. 1C and 1D). Furthermore, the large 4JP-P value of 25 Hz found with arbitrary sign indicates a signi.cant interaction between 31P nuclei. The 195Pt-{1H} spectrum of complex 1 is expected to consist basically of a triplet of triplets due to 1JPt-P and 3JPt-P couplings from the Pt*Pt isotopomer (Fig. 1E). Consideration of the 2JPt-Pt coupling from the less abundant isotopomer causes the appearance of additional weak sidebands in all multiplets (Fig. 1F) and allows a good match with the experimental spectrum (Fig. 1G). The directly deduced coupling constants together with the found chemical shift values have been used for a full computer simulation of the experimental spectra. Re.nement of the parameters, including the contributions from all isotopomers leads to the .nal values given in Table 1.The previous analysis of the 31P and 195Pt NMR data of complex 1 facilitates that of the trinuclear complexes 2, 3 and 3. Complex 2 (Scheme 3) can be described as a AAIAIIAIIIAIVAVXXIXII spin system, where all 31P and 195Pt nuclei are not magnetically equivalent but show a common chemical shift value. In this complex, the presence of three platinum atoms implies the existence of four possible isotopomers (PtPtPt 29.0; PtPt*Pt 44.4; Pt*Pt*Pt 22.7; *Pt*Pt*Pt 3.9%). The analysis and interpretation of the experimental 31P and 195Pt NMR spectra of complex 2 (SUP 57611) is complicated by the second order e.ects appearing from virtual couplings between all three platinum nuclei and from the 31P–31P couplings through three or more bonds.The simulated and the experimental spectra are in good concordance. However, the linewidth does not allow a good resolution of the observed signals and thus it has only been possible to obtain partial results for the coupling constants (Table 1).In any case, the broadening of the bands suggests an upper limit of ca.10 Hz for 4JP-P. The 31P and 195Pt NMR spectra for complex 3 have allowed determination of the reaction products obtained by treatment of 1 with [PdCl2(dppe)] (Scheme 2). Discussion of these results and comparison with X-ray data are given after the Molecular structures sub-section. The NMR parameters of 3 and 3 are given below (Table 7). Those required for comparison with the other complexes are included in Table 1.3108 J.Chem. Soc., Dalton Trans., 1999, 3103–3113 Comparison of the coupling constants of the trinuclear complexes 2, 3 and 3 with those of the dinuclear 1 indicates that 1JPt-P increases while 3JPt-P and 4JP-P decrease. The values of 2JPt-Pt are signi.cant in both 1 and 2, 3, 3 and fall into the observed range for non-bonded metal–metal atoms.7d,25 It is noteworthy that the structures of dinuclear and trinuclear species are strongly related, the main di.erence lying in the dihedral angle between platinum co-ordination planes, which is 140 in 1 11 but reduces to 120 in 2, 3 and 3 (this work).The pentanuclear complexes (3 in Scheme 3) should show more complicated NMR spectra as a consequence of possible interactions between the two {Pt2S2} entities and of the presence of an heterometal with active isotopes in signi.cant abundance as occurs in the cadmium (5 and 5) and mercury (6) complexes.To analyse the relevance of the former factor, the zinc analogue 4 was synthesized and fully characterised by X-ray di.raction data (see below) and NMR in solution. Complex 4, whose complete spin system is AAIAIIAIIIAIVAVAVIAVIIXXIXIIXIII (12 active nuclei), has six di.erent isotopomers: PtPtZnPtPt 19.2; PtPtZnPt*Pt 39.4; PtPtZn*Pt*Pt 10.0; Pt*PtZnPt*Pt 20.0; Pt*PtZn*Pt*Pt 10.2 and *Pt*PtZn- *Pt*Pt 1.3%. The experimental 31P-{1H} and 195Pt-{1H} NMR spectra are shown in Fig. 2B and 2D, respectively.The 31P-{1H} spectrum is comparable to that for complex 1 with the intense central signal due to the PtPtZnPtPt isotopomer, and with down.eld and up.eld signals separated by 1JPt-P = 3120 Hz. Although the broadening of signals makes it di.cult to obtain the 3JPt-P coupling values, from the separation of the weak sidebands the value of 2JPt-Pt = 360 Hz can be deduced. On the other hand, the 195Pt-{1H} spectrum appears more resolved than that of 31P-{1H} and con.rms the previous coupling constants with additional information on the value of the 3JPt-P coupling from the minor triplet separations.The negative sign of 3JPt-P is clearly corroborated in the simulated spectrum Fig. 2 The 31P-{1H} (bottom) and 195Pt-{1H} (top) NMR spectra for complex 4, at 121.4 and 64.2 MHz, respectively: (A) 31P and (C) 195Pt computer simulations; (B) 31P and (D) 195Pt experimental spectra, respectively. (Fig. 2C) by the relative intensity of the down.eld and up.eld satellite triplets, since only this sign combination .ts well to the experimental spectrum (Fig. 2D). Both 31P and 195Pt NMR spectra are insensitive to the choice of sign for 2JPt-Pt, and a better .t occurs when a long range coupling (4JP-P = 10 Hz) is included. The good match between the simulated and experimental spectra, particularly in the case of 195Pt NMR (Fig. 2C and 2D), does not require consideration of additional couplings between {Pt2S2} entities. Accordingly, the complete spin system (AAIAIIAIIIAIVAVAVIAVIIXXIXIIXIIIY) for the cadmium (5, 5) and mercury (6) analogues can be reduced to [AAIAIIAIIIXXI] 2Y by introducing the assumption that all couplings between nuclei of di.erent {Pt2S2} units, such as 4JPt-Pt, 5JP-Pt and 6JP-P, are e.ectively zero.Both 113Cd-{1H} and 111Cd-{1H} NMR spectra of complexes 5 and 5 in solution, as a consequence of the relative abundance of di.erent isotopomers (Table 2), should be a nonet of nonets, the .rst multiplet being due to coupling with the four platinum nuclei (with a 2JPt-Cd separation and relative intensities 1, 16, 98, 292, 1021, 292, 98, 16, 1), and the second from long range coup- Fig. 3 The 113Cd-{1H} (bottom) and 195Pt-{1H} (top) NMR spectra for complex 5, at 44.4 and 64.2 MHz, respectively: (A) simulated and (B) experimental cadmium spectra; (C) simulated and (D) experimental platinum spectra. Table 2 Relative abundance (%) of the twelve isotopomers of the pentanuclear complexes 5, 5 and 6 Isotopomer X = Cd X = Hg PtPtXPtPt PtPt*XPtPt PtPtXPt*Pt PtPt*XPt*Pt PtPtX*Pt*Pt PtPt*X*Pt*Pt Pt*PtXPt*Pt Pt*Pt*XPt*Pt Pt*PtX*Pt*Pt Pt*Pt*X*Pt*Pt *Pt*PtX*Pt*Pt *Pt*Pt*X*Pt*Pt 14.8 5.0 29.6 9.9 7.4 2.5 14.8 4.9 7.4 2.5 0.9 0.3 16.4 3.4 32.7 6.6 8.2 1.7 16.4 3.4 8.2 1.7 1.0 0.2J.Chem. Soc., Dalton Trans., 1999, 3103.3113 3109 ling with the eight phosphorus nuclei (with a 3JCd-P separation and relative intensities 1, 8, 28, 56, 70, 56, 28, 8, 1).Disregarding the less intense bands the expected spectra should become a quintet. For both 113Cd and 111Cd nuclei we have observed a similar splitting of their resonance signal at about ¥ä 501, which consists of a quintet of septets (Fig. 3B) and shows the expected relative intensities. From simulation the following coupling constants have been deduced: 2J(113Cd-Pt) = 345, 3J(113Cd-P) = 15, 2J(111Cd-Pt) = 329 and 3J(111Cd-P) = 14.3 Hz. The coupling constants involving 113Cd and 111Cd isotopes follow the relation J113 = 1.0461 J111 in agreement with the literature.28 The experimental 199Hg NMR spectrum for complex 6 (Fig. 4B) shows a resonance at ¥ä 922 and consists of a multiplet similar to that observed for complexes 5 and 5.The simulated spectrum (Fig. 4A) allows determination of the 2JHg-Pt = 460 and 3JHg-P = 40 Hz coupling values. The 31P and 195Pt NMR parameters of complexes 5, 5 and 6 given in Table 1 have been obtained through computer simulation of the experimental data.The best .t has required consideration not only of 2JPt-Pt but also of 2JPt-X (X = Cd or Hg). For 5 the value of 2JPt-Pt = 180 Hz has been deduced from the 31P NMR spectrum and con.rmed from the platinum spectrum, on the basis of computer simulations with di.erent 2JPt-Pt couplings. Interestingly, the 2JPt-Cd ca. 340 Hz is similar to 2JPt-Pt observed for complex 4. The Pt-Pt coupling has a signi.cant e.ect on satellite signals of the 31P NMR spectrum, while in the 195Pt NMR spectrum both 2JPt-Cd and 2JPt-Pt couplings produce similar changes.Thus, both couplings can interfere if only .rst order analysis was considered. For complex 6 the 2JPt-Pt = 200 Hz deduced by simulation of both 31P and 195Pt NMR spectra is also less than the 2JPt-Hg = 460 Hz. Fig. 4D and 4E show the experimental 195Pt NMR spectra and Fig. 4C the simulated spectrum, which corroborates the proposed .nal values (Table 1). Interpretation Fig. 4 The 199Hg-{1H} (bottom) and 195Pt-{1H} (top) NMR spectra for complex 6, at 71.6 and 64.2 MHz, respectively: (A) simulated and (B) experimental mercury spectra; (C) simulated, (D) and (E) experimental platinum spectra with di.erent resolution enhancement Fourier transform. of the decrease in the Pt.Pt coupling when another signi.cant Pt.M coupling exists is not simple because of the structural similarity among complexes 4, 5, 5 and 6. The previous NMR data allow us to establish interesting structure.spectroscopy correlations.The structure of the dinuclear complex 1 is known11 and that of the trinuclear 2, 3 and 3 and pentanuclear complexes 4, 5, 5 and 6 is reported in this work. A graphical representation of ¥ä(31P) vs. ¥ä(195Pt) gives two straight lines depending on the nuclearity of the complexes (Fig. 5). Thus, the pentanuclear family of complexes 4, 5, 5 and 6 de.nes that with the smaller slope and includes the precursor species 1, while the trinuclear species 2, 3 and 3 determine a second line.Similar correlations have been reported for series of compounds and analysed through the paramagnetic term in Ramsey¡�s equation.29 On the other hand, another correlation between the 2JPt-Pt and 1JPt-P values is also observed. As shown in Fig. 5, the only species with a dihedral angle of 140 di.ers from the rest of the complexes, where this angle reduces to 120. Interpretation of these observations is complicated by the di.erent factors contributing to both the chemical shifts and the coupling constants.However, experimental data on these systems may be of use in future studies on NMR data.structure relationships. Molecular structures Complex 2 consists of discrete trinuclear [Pt3(dppe)3(¥ì3-S)2]2 cations (Fig. 6), Cl anions and chloroform solvent molecules. It is isostructural with its palladium analogue.20 The cation has exact crystallographic C3 symmetry with an essentially D3h Pt3S2 core.The central unit Pt3S2 consists of an equilateral triangle of platinum atoms capped above and below by two sulfur atoms thus describing a regular trigonal bipyramid. These two triply bridging sul.do ligands lie equidistant, 1.532 A above and 1.531 A below, the Pt3 plane. The Pt Pt, S S distances and the dihedral angles between P2PtS2 planes are shown in Table 5. Each platinum atom has square-planar coordination (Table 3), distorted by a reduction of S.Pt.S and P.Pt.P angles from ideal 90 , and by a twist of 11.2 between the PtS2 and PtP2 planes.Similar structures have been observed for P6M3S2 cores with monodentate and chelating phosphine ligands, for M = Ni,24 Fig. 5 Graphical representation of ¥ä(31P) vs. ¥ä(195Pt) chemical shifts, and 2JPt-Pt vs. 1JPt-P coupling constants, from Table 1, for all complexes.3110 J. Chem. Soc., Dalton Trans., 1999, 3103–3113 Pd,25 and Pt,7d,e including another complex containing the same cation as 2;7e none of these has crystallographic C3 symmetry, but deviations from the ideal symmetry are not great.Complex 3 is a mixed metal Pd/Pt analogue of 2, but with di.erent coter ions (BPh4 ), and incorporating acetonitrile solvent in the crystal structure. The main geometrical results for the cation in 3 (Table 3) are essentially the same as those of 2 and its all-Pd analogue,20 the two metals having the same atomic radii. In contrast, they have very di.erent atomic scattering factors for X-rays, allowing a clear distinction between them in an ordered structure, and a relatively precise determination of the Pd : Pt occupancy ratio for each metal site in a substitutionally disordered structure.We .nd in the case of 3 that all three metal sites in the cation, which has no crystallographic symmetry, are disordered, but unequally so, with platinum occupancies of 59.1(4), 44.8(4), and 64.6(5)%, giving an overall formulation of Pd1.32Pt1.68 (56.2% Pt, 43.8% Pd) rather than PdPt2 for the intended and expected trinuclear product.This empirically determined “analysis” applies only to the particular single crystal selected for X-ray di.raction study, these mixedmetal products not necessarily being homogeneous. A few heteronuclear PdPt complexes are known, in which Pd and Pt occupy chemically equivalent positions; in all cases for which crystal structures have been reported there is disorder of the metal atom sites.26,30 In connection with the NMR studies, it was considered particularly important to obtain a crystal structure of a Zncentred complex such as 4, and to demonstrate that the complex cations of Zn, Cd, and Hg had analogous structures, in order to have a reference compound with a magnetically silent heterometal. Despite repeated attempts, only very small and poorly di.racting crystals could be obtained, but these were successfully studied with synchrotron radiation facilities.Although the results are the least precise of all these structures (a re.ection of the crystal quality rather than the synchrotron data collection itself), they are more than adequate for their purpose. Fig. 6 The structure of the cation of complex 2, with the labeling of independent atoms. Table 3 Selected distances (Å) and angles ( ) for complexes 2 and 3 2 3 M M M–S M–P S–M–S P–M–P 3.1185(4) 2.3634(14) 2.3638(14) 2.2497(15) 2.2463(15) 80.76(6) 86.13(6) 3.0690(5), 3.2159(5), 3.0878(5) 2.3529(17), 2.3687(18), 2.3434(17) 2.3806(17), 2.3858(17), 2.3791(17) 2.2603(19), 2.2560(18), 2.2702(18) 2.2496(18), 2.2739(18), 2.2499(18) 80.73(6), 80.30(6), 80.95(6) 85.19(7), 85.81(7), 85.66(7) Complexes 4, 5, 5 and 6 all contain [M{[Pt(dppe)]2(µ3- S)2}2]2 cations (M = Zn in 4, Cd in 5 and 5, Hg in 6) together with various anions; 5 also has solvent molecules (both acetone and water) present in the crystal structure.Complexes 4 and 5, both with ClO4 anions, are isostructural.The cations are all very similar (Fig. 7), consisting of two {Pt2S2} butter.ies linked through sulfur to the heterometal, Zn, Cd or Hg, which shows a distorted tetrahedral co-ordination. They all have essentially D2 (222) symmetry for the central core excluding C and H atoms, one of these twofold rotation axes being crystallographic for 4, 5, and 6. The detailed geometry around the heterometal for 4, 5, 5 and 6 is given in Table 4. The .exibility of 1 when acting as a metalloligand can be inferred from the data collected in Table 5.In all the derivatives of 1 there is a signi.cant reduction of the dihedral angle between PtS2 planes (.) if compared with the unbound metalloligand (140.2 ). This decrease entails a concomitant shortening of the Pt Pt and smaller changes in the S S distances. Concerning the pentanuclear species, the high degree of distortion from the ideal tetrahedral co-ordination about the heterometal is surprising.We found a similar distortion around CuII in the complex cation [Cu{[Pt(dppe)]2(µ3-S)2}2]2,11 but such a distortion is not to be expected for metals with d10 electronic con.guration, as ZnII, CdII and HgII. In all cases, the deviation from the regular tetrahedral geometry consists of a reduction of two opposite S–M–S angles (involving the chelating metalloligands) and a substantial twist of these two S–M–S planes away from the ideal 90 dihedral angle.This is shown in Fig. 8 for complex 6 as a representative example of all the pentanuclear species here reported. We do not have a ready Fig. 7 Molecular structure of the cation of complex 6, with the labeling of independent atoms. Fig. 8 Detail of the core of the cation of complex 6.J. Chem. Soc., Dalton Trans., 1999, 3103¡V3113 3111Table 4 Selected distances () and angles () for complexes 4, 5, 5 and 64 5 5 6M¡VSPt¡VSPt¡VPS¡VM¡VSS¡VPt¡VSP¡VPt¡VP2.412(6)2.360(6)2.374(7)2.379(5)2.356(6)2.378(7)2.243(7)2.262(6)2.268(9)2.274(8)81.8(3), 81.4(3)114.0(2) ¡Ñ 2137.7(2) ¡Ñ 282.9(2)81.5(2)85.7(3)85.5(3)2.561(2)2.553(3)2.367(2)2.374(2)2.367(2)2.377(3)2.230(2)2.254(2)2.246(3)2.247(3)77.36(12), 76.85(10)117.14(7) ¡Ñ 2140.09(7) ¡Ñ 284.36(8)84.58(10)86.10(9)85.51(12)2.579(2), 2.615(2)2.555(2), 2.565(2)2.366(2), 2.374(2)2.382(2), 2.379(2)2.382(2), 2.376(2)2.370(2), 2.371(2)2.245(2), 2.265(2)2.260(2), 2.246(2)2.254(2), 2.264(2)2.244(2), 2.256(2)75.55(7), 76.39(7)113.96(7), 120.22(7)143.59(8), 139.85(8)84.09(7), 83.66(8)83.69(7), 84.03(8)85.15(8), 85.61(9)85.45(9), 85.62(9)2.599(2)2.585(2)2.373(2)2.381(2)2.367(2)2.362(2)2.242(2)2.246(2)2.259(2)2.242(2)75.62(6) ¡Ñ 2117.57(9), 109.34(8)147.96(2) ¡Ñ 283.93(7)84.46(7)86.98(8)86.19(8)Table 5 Structural parameters (distances in , angles in ) describing the M{Pt2S2} core in adducts of 1Complex Pt Pt S S £c a M Pt £sb1 [Pt2(dppe)2(-S)2] c2 [Pt3(dppe)3(3-S)2]24 [Zn{Pt2(dppe)2(3-S)2}2]25 [Cd{Pt2(dppe)2(3-S)2}2]25 [Cd{Pt2(dppe)2(3-S)2}2]2[Cu{Pt2(dppe)2(3-S)2}2]2 d6 [Hg{Pt2(dppe)2(3-S)2}2]23.2923.1193.0833.1093.1263.0893.1483.2573.0723.1293.1253.1343.0623.1443.0903.1843.1923.1753.1743.0933.0763.178140.2120.0119.8120.2123.1126.0125.9134.4118.4121.1125.3¡X3.1193.1293.0963.2363.1983.295; 3.1603.262; 3.0963.059; 3.0573.001; 3.0763.153; 3.348¡X¡X67.066.464.160.053.0a Dihedral angle between two {PtS2} planes. b Dihedral angle between two {MS2} planes.c Ref. 11. d Ref. 20.Table 6 Crystallographic data for complexes 2¡V6Compound 2 3 4 5 5 6FormulaMCrystal systemSpace groupa/b/c/£\/£]/£^/U/3ZDc/g cm3/mm1T/KData measuredR(F ) (¡¥observed¡¦data)Rw(F2 ) (all data)Data, parametersC78H72Cl2P6Pt3S26CHCl32631.7RhombohedralR3c18.0036(10)18.0036(10)105.495(6)29613(3)121.7714.96160489220.0325 (5086)0.09155749, 411C126H112B2P6Pd1.32-Pt1.68S2MeCN2407.4TriclinicP114.0572(9)15.1519(9)29.5450(19)90.343(2)99.877(2)115.552(2)5570.6(6)21.4352.49160356210.0624 (19221)0.130024671, 1284C104H96P8Pt4S4-ZnCl2O82766.4OrthorhombicC222113.649(8)33.654(19)23.209(12)10661(10)41.7245.75160245910.1052 (8358)0.229510357, 593C104H96CdCl2O8-P8Pt4S42813.5OrthorhombicC222113.590(3)33.847(6)22.927(4)10546(3)41.7725.79160288420.0485 (9640)0.102812427, 591C104H96Cd2Cl4P8-Pt4S44Me2CO2H2O3137.1TriclinicP116.7851(8)16.9057(8)22.1911(10)73.093(2)86.300(2)75.436(2)5831.0(5)21.7875.46160425630.0533 (15645)0.110526388, 1297C104H96Cl2F6Hg1.5-P9Pt4S43018.9TetragonalP421c19.4566(7)19.4566(7)27.2440(10)10313.5(6)41.9447.96160638160.0359 (10766)0.098412245, 603explanation for the twist angles (£s) found but they might berelated to phenyl¡Vphenyl interactions between the two {Pt2S2}units. In fact, the phenyl rings from the two sides of the cationinterlock quite nicely (Fig. 7), although they could probablyadapt fairly easily to a dierent twist angle leading to greaterdistance between them.The pentanuclear complexes 4, 5, 5, 6 and their copper()analogue are the rst examples where the heterometal is tetrahedrallyco-ordinated to four suldes from two {Pt2S2} units.The only other comparable complex with a M{Pt2S2}2 core is[Pd{Pt2(PPh3)4(3-S)2}2]2. By contrast, it has an exactly centrosymmetric,distorted square-planar palladium() co-ordinationin keeping with the d 8 conguration of the central metal.7c Similarto the palladium() complex, other pentanuclear arrays withfour 3-S2 ligands are found in [Ni{Ni2Cp2(3-S)2}2] 31 and in[Ni{Pd2Cl(PPh3)3(3-S)2}2].323112 J.Chem. Soc., Dalton Trans., 1999, 3103–3113 Determination of the composition of the solid solution of overall formula [PdxPt3-x(dppe)3(3-S)2][BPh4]2 3 by NMR The di.erent composition of the crystalline solids isolated in the reaction leading to complex 3, which was expected to a.ord 3, .nds an explanation from the analysis of the NMR spectra in solution. These data show that the reaction occurs according to two simultaneous pathways a and b (Scheme 4).Pathway a leads to the expected trinuclear PdPt2 complex (3) while b gives rise to two trinuclear, Pt3 (2) and Pd2Pt (3), species. This second 2 [(dppe)Pt(µ-S)2Pt(dppe)] 2 [Pd(dppe)Cl2] 4 NaBPh4 a b 2 [PdPt2(dppe)3(µ3-S)2][BPh4]2 3 [Pt3(dppe)3(µ-S)2][BPh4]2 2 4 NaCl [Pd2Pt(dppe)3(µ3-S)2][BPh4]2 3 4 NaCl Scheme 4 Fig. 9 The 31P-{1H} (bottom) and 195Pt-{1H} (top) NMR spectra for complex 3, at 101.2 and 85.6 MHz, respectively: (A) 31P and (E) 195Pt computer simulation for complex 3; (B) 31P and (F) 195Pt computer simulation for 3; (C) 31P and (G) 195Pt computer simulation for 2; (D) 31P and (H) 195Pt experimental spectra.pathway can be explained by considering the ease of formation of 2 from 1 in solution, which allows the reaction of the remaining S2Pt(dppe) fragment with [PdCl2(dppe)] in a 1 : 2 molar ratio. The 31P-{1H} NMR spectrum of a solution of subsequent crops of the crystalline solid isolated from the reaction mixture shows formation of 3, but also that of 3 and 2.Assignment of the peaks was essentially based on NMR data for pure trinuclear 2 and its palladium analogue.20 As observed in Fig. 9D, the experimental 31P-{1H} NMR spectrum is the sum in di.erent proportions of three individual spectra. The .rst one corresponds to complex 3 and shows two di.erent 31P resonances, with relative intensity 1 : 2, at d 49.5 (singlet) and at 40.5 (triplet), which can be assigned to phosphorus bound to palladium and platinum atoms, respectively (Fig. 9A). The second NMR spectrum (Fig. 9B) also displays two 31P signals, at d 51.4 (singlet corresponding to (31P)Pd) and at 42.4 (triplet assigned to (31P)Pt), with relative intensity 2 : 1, respectively. These features allow us to assign this spectrum to complex 3. Finally, the minor additional 31P resonances .t very well with the presence of 2 in small proportion (Fig. 9C). The 31P NMR data for 3 and 3 are given in Table 7.The remaining NMR parameters as well as those of 2 are included in Table 1. The relative intensities of the three previous components in the experimental 31P NMR spectrum of complex 3 (Fig. 9D) allow an estimation of the abundance of 2, 3 and 3 in the crystalline product (Table 7) and thus show that 3 is predominant in the solid mixture while 3 and 2, particularly the latter, form in small proportion. As, according to reaction pathway b, 2 and 3 should be in a 1 : 1 ratio, it can be deduced that the solubility of 2 is greater than that of 3.In order to corroborate the presence in solution of the unexpected heterometallic complex 3, the reaction of [Pd2- (dppe)2(µ-S)2] 20 and [PtCl2(dppe)] was carried out. The NMR study of the samples obtained in this reaction fully agreed with data given in Table 7. The 195Pt-{1H} NMR spectrum of complex 3 (Figure 9H) con.rms the coexistence of 3, 3 and 2 in solution. The chemical shift values for 3 and 3 as well as the di.erent coupling constants, which are in agreement with those found by 31P NMR, are given in Table 1.Based on the relative abundance of the three complexes present in solution, the overall formulation for the [PdxPt3-x- (dppe)3(µ3-S)2][BPh4]2 complex can easily be established. The value found for x is very close to that calculated from the Pd :Pt occupation ratio found for each metal site by X-ray di.raction. The good concordance between NMR and X-ray data suggests that 3 is basically a solid-solution mixture of heterotrinuclear complexes.To our knowledge, this constitutes one of the few examples where it has been possible to establish the exact nature of polynuclear heterometallic PdPt complexes.30e,j Conclusion The results reported in this paper extend the knowledge on the behaviour of the [Pt2(dppe)2(µ-S)2] metalloligand to metal ions with preference for tetrahedral co-ordination. Di.erent counter ions and stoichiometric metalloligand to heterometal molar ratios in di.erent solvents have always led to the pentanuclear [M{Pt2(dppe)2(µ3-S)2}2]2 cations (M = Zn, Cd or Hg), which are structurally similar and show a distorted tetrahedral coordination around the heterometal.In solution, the [Pt2(dppe)2(µ-S)2] metalloligand a.ords the trinuclear [Pt3(dppe)3(µ3-S)2]2 cation easily. We propose that this expansion, Pt2 .. Pt3, is due to the nucleophilic attack of the bridging sulfur ligands on the halogenated solvents, which involves the formation of undetected sulfur alkylation products.In the presence of [PdCl2(dppe)], the [Pt2(dppe)2(µ-S)2] metalloligand gives rise to a solid-solution mixture of the pureJ. Chem. Soc., Dalton Trans., 1999, 3103�C3113 3113 Table 7 The 31P NMR parameters for complexes 3 and 3 Composition (%) Compound ¦Ä(31P)Pd ¦Ä(31P)Pt 1JPt-P/Hz Intensity ratio (P)Pd : (P)Pt Solution 1 a Solution 2 b 3 [Pt2Pd(dppe)3S2]2 3 [PtPd2(dppe)3S2]2 49.5 51.4 40.5 (t) 42.4 (t) 3166 3131 1:2 2:1 73.4 24.6 75.4 21.5 a Obtained from a rst crop of crystals. The remaining percentage (2.0%) corresponds to complex 2.These data allow calculation of the overall platinum (59.1%) and palladium (40.9%) content. b Obtained from a second crop of crystals. The remaining percentage (3.2%) corresponds to complex 2. These data allow calculation of the overall platinum (60.5%) and palladium (39.4%) content. trinuclear complexes [PdPt2(dppe)3(¦Ì3-S)2]2, [PtPd2(dppe)3- (¦Ì3-S)2]2 and [Pt3(dppe)3(¦Ì3-S)2]2.All three complexes have been extensively characterised by NMR spectroscopy, unambiguously conrming the genuine heteronuclear nature of the mixed-metal cations. The composition of the solid as deduced from NMR is in accordance with that calculated from X-ray diraction data. The pentanuclear cations have also been fully characterised by 31P, 195Pt, 111Cd, 113Cd and 199Hg NMR and interesting NMR�Cstructure correlations including the penta- and the trinuclear complexes have been found.Acknowledgements This research was supported by the Ministerio de Educaci�®n y Cultura (Spain, Grants PB97-0216 and PB95-0639-C02-01) and the UK EPSRC; CASE support for R. A. C. from Siemens plc (now Bruker AXS) is gratefully acknowledged. References 1 S.-W. A. Fong and T. S. A. Hor, J. Chem. Soc., Dalton Trans., 1999, 639. 2 W. Bos, J. J. Bour, P. P. J. Schlebos, P. Hageman, W. P.Bosman, J. M. M. Smits, J. A. C van Wietmarschen and P. T. Beurskens, Inorg. Chim. Acta, 1986, 119, 141. 3 M. Zhou, Y. Xu, L.-L. Koh, A. L. Tan, P.-H. Leung and T. S. A. Hor, Inorg. Chem., 1993, 32, 1875; M. Zhou, Y. Xu, A.-M. Tan, P.-H. Leung, K. F. Mok, L.-L. Koh and T. S. A. Hor, Inorg. Chem., 1995, 34, 6425. 4 C. E. Briant, T. S. A. Hor, N. D. Howells and D. M. P. Mingos, J. Organomet. Chem., 1983, 256, C15. 5 H. Liu, A. L. Tan, Y. Xu, K. F. Mok and T. S. A. Hor, Polyhedron, 1997, 16, 377. 6 (a) H. Liu, A. L. Tan, K. F. Mok and T. S. A. Hor, J. Chem. Soc., Dalton Trans., 1996, 4023; (b) M. Zhou, Y. Xu, C.-F. Lam, P.-H. Leung, L.-L. Koh, K. F. Mok and T. S. A. Hor, Inorg. Chem., 1994, 33, 1572; (c) M. S. Zhou, A. L. Tan, Y. Xu, C.-F. Lam, P.-H. Leung, K. F. Mok, L.-L. Koh and T. S. A. Hor, Polyhedron, 1997, 16, 2381. 7 (a) D. I. Gilmour, M. A. Luke and D. M. P. Mingos, J. Chem. Soc., Dalton Trans., 1987, 335; (b) C. E. Briant, D. I. Gilmour, M. A. Luke and D.M. P. Mingos, J. Chem. Soc., Dalton Trans., 1985, 851; (c) C. E. Briant, T. S. A. Hor, N. D. Howells and D. M. P. Mingos, J. Chem. Soc., Chem. Commun., 1983, 1118; (d) G. W. Bushnell, K. R. Dixon, R. Ono and A. Pidcock, Can. J. Chem., 1984, 62, 696; (e) M. J. Pilkington, A. M. Z. Slawin, D. J. Williams D. Woollins, J. Chem. Soc., Dalton. Trans., 1992, 2425. 8 M. Zhou, Y. Xu, C.-F. Lam, L.-L. Koh, K. F. Mok, P.-H. Leung and T. S. A. Hor, Inorg. Chem., 1993, 32, 4660. 9 V. W.-W. Yam, P. K.-Y. Yeung and K.-K. Cheung, Angew. Chem., Int. Ed. Engl., 1996, 35, 739. 10 (a) B. H. Aw, K. K. Looh, H. S. O. Chan, A. L. Tan and T. S. A. Hor, J. Chem. Soc., Dalton Trans., 1994, 3177; (b) M. Zhou, P.-H. Leung, K. F. Mok and T. S. A. Hor, Polyhedron, 1996, 15, 1737. 11 M. Capdevila, Y. Carrasco, W. Clegg, R. A. Coxall, P. Gonz�¢lez- Duarte, A. Lled�®s, J. Sola and G. Ujaque, Chem. Commun., 1998, 597. 12 M. Capdevila, W. Clegg, P. Gonz�¢lez-Duarte, A. Jarid and A.Lled�®s, Inorg. Chem., 1996, 35, 490. 13 A. D. Westland, J. Chem. Soc., 1965, 3060. 14 M. P. Brown, R. J. Puddephatt, M. Rashidi and K. R. Seddon, J. Chem. Soc., Dalton Trans., 1977, 951. 15 R. K. Harris and B. E. Mann; NMR and the Periodic Table, Academic Press, London, 1978. 16 P. H. M. Budzelaar, gNMR V4.01, Cherwell Scientic, 1997. 17 C. Redshaw, V. C. Gibson, W. Clegg, A. J. Edwards and B. Miles, J. Chem. Soc., Dalton Trans., 1997, 3343. 18 W. Clegg, M. R. J. Elsegood, S. J.Teat, C. Redshaw and V. C. Gibson, J. Chem. Soc., Dalton Trans., 1998, 3037. 19 SHELXTL, Bruker AXS, Madison, WI, 1994. 20 M. Capdevila, W. Clegg, R. A. Coxall, P. Gonz�¢lez-Duarte, M. Hamidi, A. Lled�®s and G. Ujaque, Inorg. Chem. Commun., 1998, 1, 466. 21 G. Li, S. Li, A. L. Tan, W.-H. Yip, C. W. Mak and T. S. A. Hor, J. Chem. Soc., Dalton Trans., 1996, 4315. 22 M. Zhou, C. F. Lam, K. F. Mok, P.-H. Leung and T. S. A. Hor, J. Organomet. Chem, 1994, 476, C32. 23 V. W. W. Yam, P. K. Y. Yeung and K. K. Cheung, J. Chem. Soc., Chem. Commun., 1995, 267. 24 K. Matsumoto, N. Saiga, S. Tanaka and S. Ooi, J. Chem. Soc., Dalton Trans., 1991, 1265; C. A. Ghilardi, S. Midollini and L. Sacconi, Inorg. Chim. Acta, 1978, 31, L431; C. A. Ghilardi, S. Midollini, A. Orlandini, C. Battistoni and G. Mattogno, J. Chem. Soc., Dalton Trans., 1984, 939. 25 H. Werner, W. Bertle and U. Schubert, Inorg. Chim. Acta, 1980, 43, 199. 26 Ref. 27 (b) of ref. 1. 27 N. M. Boag, J. Browning, C. Crocker, P. L. Goggin, R. J. Goodfellow, M. Murray and J. L. Spencer, J. Chem. Res. (S), 1978, 228. 28 P. S. Pregosin (Editor), Transition Metal Nuclear Magnetic Resonance, Elsevier, Amsterdam, 1991. 29 M. P. Guy, J. L. Coer, J. S. Rommel and D. W. Bennett, Inorg. Chem., 1988, 27, 2942; S. F. Gheller, T. W. Hambley, J. R. Rodgers, R. T. C. Brownlee, M. J. O¡�Connor, M. R. Snow and A. G. Wedd, Inorg. Chem., 1984, 23, 2519; A. F. Masters, G. E. Bossard, T. A. George, R. T. C. Brownlee, M. J. O¡�Connor and A. G. Wedd, Inorg. Chem., 1983, 22, 908. 30 (a) H. C. Clark, G. Ferguson, V. K. Jain and M. Parvez, Inorg. Chem., 1985, 24, 1477; (b) Inorg. Chem., 1986, 25, 3808; (c) H. C. Clark, G. Ferguson, P. N. Kapoor and M. Parvez, Inorg. Chem., 1985, 24, 3924; (d ) R. J. H. Clark, V. B. Croud, R. J. Wills, P. A. Bates, H. M. Dawes and M. B. Hursthouse, Acta Crystallogr., Sect. B, 1989, 45, 147; (e) M. Capdevila, W. Clegg, P. Gonz�¢lez-Duarte, B. Harris, I. Mira, J. Sola and I. C. Taylor, J. Chem. Soc., Dalton Trans., 1992, 2817; ( f ) T. Suzuki, N. Itaka, S. Kurachi, M. Kita, K. Kashiwavara, S. Ohba and J. Fujita, Bull. Chem. Soc. Jpn., 1992, 65, 1817; ( g) H. Kurosawa, K. Hirako, S. Natsuma, S. Ogoshi, N. Kanehisa, Y. Kai, S. Sakaki and K. Takeuchi, Organometallics, 1996, 15, 2089; (h) E. Colacio, R. Cuesta, M. Ghazi, M. A. Huertas, J. M. Moreno and A. Navarrete, Inorg. Chem., 1997, 36, 1652; (i) G. Reusmann, M. Grehl, W. Reckordt and B. Krebs, Z. Anorg. Allg. Chem., 1994, 620, 199; ( j) W. Clegg, M. Capdevila, P. Gonz�¢lez- Duarte and J. Sola, Acta Crystallogr., Sect. B, 1996, 52, 270. 31 H. Vahrenkamp and L. F. Dahl, Angew. Chem., 1969, 81, 152. 32 D. Fenske, H. Fleischer, H. Krautscheid, J. Magull, C. Oliver and S. Weisgerber, Z. Naturforsch., Teil B, 1991, 46,

ISSN:1477-9226    

DOI:10.1039/a903899j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3115–3119 3115 Dinuclear nickel(II) complexes with a tridentate nitrito bridge and terminal thiocyanato ligands. Crystal structure and magnetic properties Albert Escuer,a Mercè Font-Bardía,b Evaristo Peñalba,a Núria Sanz,a Xavier Solans b and Ramon Vicente *a a Departament de Química Inorgànica, Universitat de Barcelona, Diagonal 647, 08028-Barcelona, Spain. E-mail: rvicente@kripto.qui.ub.es b Departament de Cristal.logra.a i Mineralogia, Universitat de Barcelona, c/Martí Franques s/n, 08028-Barcelona, Spain Received 11th May 1999, Accepted 9th July 1999 The µ-(.1-N:.2-O,O)-nitrito dinuclear compounds [Ni2(µ-NO2)(NCS)3(Medpt)2]H2O 1 and [Ni2(µ-NO2)- (NCS)3(dpt)2] 2, and the mononuclear nitrito compounds [Ni(NO2)(NCS)(Medpt)] 3 and [Ni(NO2)(NCS)(Medien)] 4, where Medpt = bis(3-aminopropyl)methylamine, dpt = bis(3-aminopropyl)amine and Medien = bis(2-aminoethyl) methylamine, have been synthesized and characterised.The crystal structures of 1–4 have been determined by single-crystal X-ray analysis. The thiocyanate ligand appears to stabilise the tridentate co-ordination mode of the nitrito ligand. The magnetic behaviour of the dinuclear compounds 1 and 2 was recorded between 300 and 4 K, showing antiferromagnetic coupling in both cases. The magnetic susceptibility data were .tted by the expression for a dinuclear nickel(..) compound giving the parameters J = 12.8 cm1, g = 2.17 and J = 12.2 cm1, g = 2.27 for 1 and 2 respectively.In a previous study on the bridging carbonate ligand we structurally and magnetically characterised one new pentadentate co-ordination mode for this ligand in the trinuclear compound [Ni3(Medpt)3(NCS)4(µ3-CO3)]. This was obtained by mixing with the carbonate ligand, the nickel(..) salt and the Medpt ligand [bis(3-aminopropyl)methylamine], another potentially bridging ligand, thiocyanate, which in this case acts as a terminal ligand.1 In this compound the thiocyanate anion is critical to the synthesis: all the attempts to synthesize the similar trinuclear compound from halides or di.erent pseudohalide ligands were unsuccessful, because extremely soluble compounds (oils or gums) were obtained.The very similar selenocyanate anion also allows isolation of the compound [Ni3(Medpt)3(NCSe)4(µ3-CO3)].2 Can the thiocyanate anion play the same synthetic role with other potentially bridging ligands? In an attempt to answer this question we used the same synthetic strategy for the nitrito anion.We obtained two new dinuclear µ-NO2 compounds: [Ni2(µ-NO2)(NCS)3(Medpt)2]H2O 1 and [Ni2(µ-NO2)(NCS)3- (dpt)2] 2 [dpt = bis(3-aminopropyl)amine]. Here the thiocyanate also acts as a terminal ligand and the nitrito as a tridentate bridging ligand by using a co-ordination mode not found previously in a dinuclear nickel(..) compound, but recently reported in one mixed valence dicopper(.,..) compound, one copper(.)–zinc(..) analogue 3 and in a nickel–manganese bimetallic chain.4 This bridging network involves a monodentate( N) nitro co-ordination with respect to one NiII and a bidentate( O,O) nitrito co-ordination with respect to the other NiII.The octahedral co-ordination polyhedron for each nickel(..) is completed by the three nitrogen atoms of one tridentate amine and, due to the asymmetrical co-ordination mode of the bridging ligand, by two thiocyanate terminal ligands in the nitronickel(..) and one thiocyanate terminal ligand in the nitritonickel(..).In the synthesis the triamine used and the stoichiometric ratio between the nickel(..) salt, triamine, thiocyanate salt and sodium nitrite in.uence the .nal product: with bis(3-aminopropyl) methylamine only the ratio 1:1:1:1 allowed the synthesis of the dinuclear compound 1, whereas the theoretical ratios 2:2:3:1 and 1:1:2:1 allowed the synthesis of the mononuclear compound [Ni(NO2)(NCS)(Medpt)] 3.With bis(3-aminopropyl)amine the dinuclear compound 2 was obtained with di.erent ratios, but with Medien [bis(2- aminoethyl)methylamine] it was not possible to obtain the dinuclear product and only the mononuclear compound [Ni(NO2)(NCS)(Medien)] 4 was obtained. Experimental Synthesis [Ni2(-NO2)(NCS)3(Medpt)2]H2O 1. This was prepared by mixing 2.65 mmol of nickel(..) nitrate hexahydrate in 15 ml of water, 2.65 mmol of sodium nitrite in 5 ml of water, 2.65 mmol of potassium thiocyanate in 5 ml of water and 2.65 mmol of Medpt.After 1 h of stirring the resulting blue solution was left to evaporate in air. Three days later violet monocrystals of complex 1 suitable for X-ray determination were collected (Found: C, 31.6; H, 6.2; N, 21.9; S, 14.6. Calc. for C17H38N10- Ni2O3S3: C, 31.6; H, 6.2; N, 21.7; S, 14.9%). [Ni2(-NO2)(NCS)3(dpt)2] 2. This was prepared by mixing 5.5 mmol of nickel(..) perchlorate hexahydrate, 5.5 mmol of dpt, 5.5 mmol of sodium nitrite and 5.5 mmol of ammonium thiocyanate in 150 ml of water.After 1 h of stirring the solution was .ltered and the resulting blue solution left to evaporate in air. Several days later, violet monocrystals of complex 2 suitable for X-ray determination were collected (Found: C, 30.3; H, 5.8; N, 23.5; S, 16.3. Calc. for C15H34N10Ni2O2S3: C, 30.0; H, 5.7; N, 23.3; S, 16.0%). [Ni(NO2)(NCS)(Medpt)] 3. This was prepared by mixing 2.65 mmol of nickel(..) nitrate hexahydrate in 15 ml of water, 5.303116 J.Chem. Soc., Dalton Trans., 1999, 3115–3119 Table 1 Crystal data and structure re.nement for [Ni2(µ-NO2)(NCS)3(Medpt)2]H2O 1, [Ni2(µ-NO2)(NCS)3(dpt)2] 2, [Ni(NO2)(NCS)(Medpt)] 3 and [Ni(NO2)(NCS)(Medien)] 4 1 2 3 4 Formula Formula weight T/K Crystal system space group a/Å b/Å c/Å a/ ß/ ./ V/Å3 Z Dc/g cm3 µ(Mo-Ka)/cm1 Data/restraints/parameters R1 wR2 C17H38N10Ni2O3S3 644.17 293(2) Orthorhombic Pbc21 8.501(3) 14.73(2) 24.099(4) 3017(5) 4 1.418 14.91 3274/7/351 0.0518 0.1007 C15H34N10Ni2O2S3 600.12 293(2) Monoclinic P21/c 12.5568(6) 14.1199(6) 15.5747(7) 103.763(4) 2682.1(2) 4 1.486 16.69 4714/0/426 0.0337 0.0851 C8H19N5NiO2S 308.05 293(2) Triclinic P1� 7.000(11) 8.555(8) 10.769(9) 88.17(7) 88.08(9) 86.58(9) 643.1(13) 2 1.591 16.70 3372/0/193 0.0466 0.1200 C6H15N5NiO2S 280.0 293(2) Orthorhombic Pbca 11.401(6) 10.956(3) 18.876(12) 2358(2) 8 1.578 18.13 3196/0/197 0.0317 0.0744 mmol of sodium nitrite in 5 ml of water, 2.65 mmol of potassium thiocyanate in 5 ml of water and 2.65 mmol of Medpt.After 1 h of stirring the resulting deep blue solution was left to evaporate in air. Three days later deep blue monocrystals of complex 3 suitable for X-ray determination were collected (Found: C, 30.8; H, 6.2; N, 22.7; S, 10.2. Calc. for C8H19N5- NiO2S: C, 31.2; H, 6.2; N, 22.7; S, 10.4%). [Ni(NO2)(NCS)(Medien)] 4. This was prepared by mixing 1.45 mmol of nickel(..) perchlorate hexahydrate in 20 ml of water, 1.45 mmol of Medien, 5.80 mmol of sodium nitrite in 5 ml of water and 1.45 mmol of ammonium thiocyanate in 5 ml of water.After 1 h of stirring the resulting blue solution was left to evaporate in air. Six days later blue crystals of complex 4 were obtained. Blue monocrystals suitable for X-ray determination were obtained from recrystallisation in DMF (Found: C, 25.3; H, 5.4; N, 25.2; S, 10.9. Calc. for C6H15N5NiO2S: C, 25.7; H, 5.4; N, 25.0; S, 11.5%).Magnetic measurements Magnetic susceptibility measurements in the temperature range 300–4 K were carried out on polycrystalline samples with a pendulum type magnetometer (MANICS DSM8) equipped with a helium continuous-.ow cryostat and a Brucker B E15 electromagnet. The magnetic .eld was ca. 1.5 T. Diamagnetic corrections were estimated from Pascal’s constants. X-Ray crystallography Prismatic violet crystals for complexes 1 and 2, and prismatic blue crystals for 3 and 4, were selected and mounted on an Enraf-Nonius CAD4 di.ractometer for 1, 3 and 4, and on a STOE STADI4 di.ractometer for 2.Unit cell parameters were determined from automatic centring of 25 re.ections (12 < . < 21) for 1, 3 and 4, and of 56 re.ections (12.5 < . < 16.6) for 2, and re.ned by the least-squares method. Innsities were collected with graphite monochromatised Mo-Ka radiation, using the .–2. scan technique. For 1 3432 re.ections were measured and 2779 assumed as observed, [I > 2s(I)], for 2 4945 re.ections measured and 3435 observed [I > 4s(I)], for 3 3448 re.ections measured and 3263 observed [I > 2s(I)], and for 4, 5974 re.ections measured and 2979 observed [I > 2s(I)]. Three re.ections were measured every two hours as orientation and intensity control; no signi.- cant intensity decay was observed.The crystallographic data, conditions used for the intensity data collection and some features of the structure re.nement are listed in Table 1.Lorentz-polarization, but not absorption, corrections were made for 1, 3 and 4, and absorption corrections with the . scan method were made for 2. The crystal structures were solved by direct methods for complexes 1, 3 and 4 and by Patterson synthesis for 2 using the SHELXS 86 computer program5 and re.ned by the full-matrix least-squares method with the SHELXL 93 computer program. 6 The values of f, f and f were taken from ref. 7. Six H atoms for 1, 34 for 2, 11 for 3 and 15 for 4 were located from a di.erence synthesis, and 25 for 1 and 8 for 3 were computed; all of them were re.ned with an overall isotropic factor using a riding model. CCDC reference number 186/1567. See http://www.rsc.org/suppdata/dt/1999/3115/ for crystallographic .les in .cif format. Results and discussion Crystal structure [Ni2(-NO2)(NCS)3(Medpt)2]H2O 1. Selected bond lengths and bond angles are given in Table 2. An ORTEP8 drawing of the dinuclear unit with atom-labelling scheme is presented in Fig. 1. Complex 1 is a µ-(.1-N:.2-O,O)-nitrito dinuclear compound: the tridentate nitrito bridging ligand comprises a bidentate nitrito(O,O) co-ordination with respect to Ni(1) and a monodentate nitro(N) co-ordination with respect to Ni(2). The octahedral co-ordination polyhedron for Ni(1) is completed by the three nitrogen atoms of one Medpt amine and by the nitrogen atom of one thiocyanate ligand. For Ni(2) the octahedral co-ordination polyhedron is completed by the three nitrogen atoms of one Medpt amine and by the nitrogen atoms of two thiocyanate ligands.For Ni(1) two of the co-ordination sites are Fig. 1 An ORTEP drawing with the atom-labelling scheme for [Ni2(µ-NO2)(NCS)3(Medpt)2]H2O 1.J. Chem. Soc., Dalton Trans., 1999, 3115–3119 3117 occupied by two oxygen atoms of the nitrito bridging ligand: due to the O(2)–N(5)–O(1) angle of 113.9(6), Ni(1) is in a very distorted octahedral environment: O(1)–Ni(1)–O(2) 57.7(2), N(2)–Ni(1)–O(2) 101.8(2), N(4)–Ni(1)–O(1) 104.8(2). Atom Ni(2) has the monodentate nitro(N) co-ordination of the nitrito bridging ligand and, for this reason, the octahedral distortion is signi.cantly smaller. It is in an elongated octahedral coordination with the four planar Ni–N distances in the range 2.068(6)–2.080(6) Å and Ni(2)–N(5) and Ni(2)–N(7) 2.230(7) and 2.146(7) Å respectively.The thiocyanate ligands are trans in the basal plane.In complex 1 there is a non-co-ordinated water molecule. The shortest distances to this water molecule are intramolecular: O(3) N(1), O(3) S(2) and O(3) S(1) are 3.034(6), 3.447(7) and 3.334(7) Å respectively. [Ni2(-NO2)(NCS)3(dpt)2] 2. An ORTEP drawing of the dinuclear compound 2 with the atom-labelling scheme is shown in Fig. 2. The main bond lengths and bond angles are given in Table 3. The structure of 2 is like that of 1. Only slight di.erences can be seen: in the bridging region the Ni(2)–N(1) 2.278(3) Å is longer than the Ni–N(nitro) distance in 1 (2.230(7) Å).The Fig. 2 An ORTEP drawing with the atom-labelling scheme for [Ni2(µ-NO2)(NCS)3(dpt)2] 2. Table 2 Selected bond lengths [Å] and angles [] for [Ni2(µ-NO2)- (NCS)3(Medpt)2]H2O 1 Ni(1)–N(4) Ni(1)–N(3) Ni(1)–O(2) Ni(2)–N(10) Ni(2)–N(6) Ni(2)–N(7) S(1)–C(8) S(3)–C(17) O(2)–N(5) N(4)–Ni(1)–N(1) N(1)–Ni(1)–N(3) N(1)–Ni(1)–N(2) N(4)–Ni(1)–O(2) N(3)–Ni(1)–O(2) N(4)–Ni(1)–O(1) N(3)–Ni(1)–O(1) O(2)–Ni(1)–O(1) N(10)–Ni(2)–N(6) N(10)–Ni(2)–N(9) N(6)–Ni(2)–N(9) N(8)–Ni(2)–N(7) N(9)–Ni(2)–N(7) N(8)–Ni(2)–N(5) N(9)–Ni(2)–N(5) N(5)–O(1)–Ni(1) C(1)–N(1)–Ni(1) O(2)–N(5)–O(1) O(1)–N(5)–Ni(2) C(17)–N(10)–Ni(2) N(9)–C(16)–S(2) 2.022(7) 2.077(7) 2.173(6) 2.068(6) 2.079(7) 2.146(7) 1.628(8) 1.645(7) 1.253(7) 94.0(3) 168.5(3) 93.5(3) 162.4(2) 87.6(3) 104.8(2) 82.4(2) 57.7(2) 91.2(3) 170.8(3) 88.6(3) 95.5(3) 97.5(2) 86.7(3) 86.1(2) 93.0(4) 117.3(5) 113.9(6) 122.6(4) 152.5(6) 178.9(8) Ni(1)–N(1) Ni(1)–N(2) Ni(1)–O(1) Ni(2)–N(8) Ni(2)–N(9) Ni(2)–N(5) S(2)–C(16) O(1)–N(5) N(4)–Ni(1)–N(3) N(4)–Ni(1)–N(2) N(3)–Ni(1)–N(2) N(1)–Ni(1)–O(2) N(2)–Ni(1)–O(2) N(1)–Ni(1)–O(1) N(2)–Ni(1)–O(1) N(10)–Ni(2)–N(8) N(8)–Ni(2)–N(6) N(8)–Ni(2)–N(9) N(10)–Ni(2)–N(7) N(6)–Ni(2)–N(7) N(10)–Ni(2)–N(5) N(6)–Ni(2)–N(5) N(7)–Ni(2)–N(5) N(5)–O(2)–Ni(1) C(8)–N(4)–Ni(1) O(2)–N(5)–Ni(2) C(16)–N(9)–Ni(2) N(4)–C(8)–S(1) N(10)–C(17)–S(3) 2.077(7) 2.103(7) 2.208(5) 2.078(7) 2.080(6) 2.230(7) 1.638(7) 1.268(8) 91.0(3) 95.8(3) 96.3(3) 84.5(3) 101.8(2) 86.3(2) 159.4(2) 90.9(3) 172.2(3) 88.1(3) 91.7(3) 92.0(3) 84.7(3) 86.0(3) 175.9(2) 95.2(4) 156.2(6) 123.1(5) 172.1(6) 178.4(7) 178.6(7) Ni–O(nitrito) distances are Ni(1)–O(1) 2.204(2) and Ni(1)–O(2) 2.291(2) Å (in 1 the same distances are 2.208(5) and 2.173(6) Å).The O(1)–N(1)–O(2) and O(1)–Ni(1)–O(2) angles are 113.6(3) and 55.87(8) respectively (in 1 the same angles are 113.9(6) and 57.7(2)). [Ni(NO2)(NCS)(Medpt)] 3.Selected bond lengths and bond angles are given in Table 4. An ORTEP8 drawing of the dinuclear unit with atom-labelling scheme is presented in Fig. 3. The nickel atom is in a distorted octahedral environment due to the geometry imposed by the bidentate nitrito-O,O ligand. The O(1)–Ni–O(2) angle is 57.48(11). Consequently, the other X–Ni–X (X = O or N) angles in the plane N(4)–N(5)–Ni–O(1)– O(2) are greater than 90: O(1)–Ni–N(5) 106.13(12), O(2)–Ni– N(4) 103.03(12), N(5)–Ni–N(4) 93.33(12).If we consider the Ni(O2N) entity, all distances and angles are similar to the published values in analogous monuclear compounds.9–15 The nitrito ligand and the terminal and central N-co-ordinated atoms of the fac-Medpt ligand are in the same plane. [Ni(NO2)(NCS)(Medien)] 4. Selected bond lengths and bond angles are given in Table 5. An ORTEP drawing of the dinuclear unit with atom-labelling scheme is presented in Fig. 4. As in complex 3 the nickel atom is in a distorted octahedral environment due to the geometry imposed by the bidentate nitrito-O,O ligand. The O(1)–Ni–O(2) angle is 58.29(7). Consequently, the other X–Ni–X (X = O or N) angles in the plane N(3)–N(5)–Ni–O(1)–O(2) are greater than 90: O(1)–Ni–N(3) 99.63(7), O(2)–Ni–N(5) 97.47(7) and N(3)–Ni–N(5) 104.30(8). If we consider the Ni(O2N) entity, all distances and angles are similar to the published values in analogous compounds9–15 and the bond values found in 3.In 4, in contrast to 3, the nitrito ligand and the two terminal N-co-ordinated atoms of the fac-Medien ligand are in the same plane. Magnetic results The plots of the magnetic susceptibility values per dimeric unit, Table 3 Selected bond lengths [Å] and angles [] for [Ni2(µ-NO2)- (NCS)3(dpt)2] 2 Ni(1)–N(2) Ni(1)–N(5) Ni(1)–N(3) Ni(1)–N(4) Ni(1)–O(1) Ni(1)–O(2) S(1)–C(1) S(2)–C(8) S(3)–C(9) O(1)–N(1) N(2)–Ni(1)–N(5) N(2)–Ni(1)–N(3) N(5)–Ni(1)–N(3) N(2)–Ni(1)–N(4) N(5)–Ni(1)–N(4) N(3)–Ni(1)–N(4) N(2)–Ni(1)–O(1) N(5)–Ni(1)–O(1) N(3)–Ni(1)–O(1) N(4)–Ni(1)–O(1) N(2)–Ni(1)–O(2) N(5)–Ni(1)–O(2) N(3)–Ni(1)–O(2) N(4)–Ni(1)–O(2) N(8)–Ni(2)–N(1) N(1)–O(1)–Ni(1) N(1)–O(2)–Ni(1) C(1)–N(2)–Ni(1) C(8)–N(6)–Ni(2) C(9)–N(7)–Ni(2) O(2)–N(1)–O(1) 2.007(3) 2.056(4) 2.075(3) 2.086(3) 2.204(2) 2.291(2) 1.624(4) 1.634(4) 1.637(3) 1.264(3) 95.4(2) 95.9(2) 166.6(2) 97.25(14) 94.21(14) 91.52(14) 97.42(11) 85.15(12) 86.24(12) 165.31(12) 153.26(11) 84.47(13) 82.20(12) 109.45(11) 86.66(11) 97.2(2) 93.2(2) 167.8(3) 157.2(3) 177.9(3) 113.6(3) Ni(2)–N(6) Ni(2)–N(10) Ni(2)–N(7) Ni(2)–N(8) Ni(2)–N(9) Ni(2)–N(1) N(2)–C(1) N(6)–C(8) N(7)–C(9) O(2)–N(1) O(1)–Ni(1)–O(2) N(6)–Ni(2)–N(10) N(6)–Ni(2)–N(7) N(10)–Ni(2)–N(7) N(6)–Ni(2)–N(8) N(10)–Ni(2)–N(8) N(7)–Ni(2)–N(8) N(6)–Ni(2)–N(9) N(10)–Ni(2)–N(9) N(7)–Ni(2)–N(9) N(8)–Ni(2)–N(9) N(6)–Ni(2)–N(1) N(10)–Ni(2)–N(1) N(7)–Ni(2)–N(1) N(2)–C(1)–S(1) N(6)–C(8)–S(2) N(7)–C(9)–S(3) N(9)–Ni(2)–N(1) O(2)–N(1)–Ni(2) O(1)–N(1)–Ni(2) 2.059(3) 2.081(3) 2.085(3) 2.089(3) 2.095(3) 2.278(3) 1.151(5) 1.157(4) 1.150(4) 1.255(3) 55.87(8) 92.70(14) 176.49(12) 87.51(13) 92.09(13) 169.39(12) 87.14(12) 90.98(12) 96.97(12) 92.48(12) 92.41(12) 83.57(11) 84.47(11) 92.97(11) 179.7(4) 178.1(3) 177.5(3) 174.42(11) 120.5(2) 124.3(2)3118 J.Chem. Soc., Dalton Trans., 1999, 3115.3119 ¥öm vs. T of [Ni2(¥ì-NO2)(NCS)3(Medpt)2]H2O 1 and [Ni2(¥ì- NO2)(NCS)3(dpt)2] 2, are shown in Fig. 5. The shape of the plots is consistent with antiferromagnetic [NiNi] entities: the ¥öm plots show maximum susceptibility at 19 K for 1 and at 15 K for 2, whereas ¥ömT values decrease continuously from room temperature (¥ömT values 2.4 cm3 K mol1, for 1 and 2 respectively) and tend to zero at low temperatures. Experimental data were .tted by the isotropic expression derived from the Hamiltonian H = JS1S2 for [NiNi] pairs, eqn. (1), where f(J,T) is (2exp(J/kT) 10exp(3J/kT))/(1 3exp(J/kT) 5exp(3J/kT)).¥öm = N¥â2g2/kTf(J,T) (1) Fig. 3 An ORTEP drawing with the atom-labelling scheme for [Ni(NO2)(NCS)(Medpt)] 3. Fig. 4 An ORTEP drawing with the atom-labelling scheme for [Ni(NO2)(NCS)(Medien)] 4. Table 4 Selected bond lengths [A] and angles [] for [Ni(NO2)(NCS)- (Medpt)] 3 Ni.N(2) Ni.N(3) Ni.O(1) S.C(1) O(2).N(1) N(2).Ni.N(5) N(5).Ni.N(3) N(5).Ni.N(4) N(2).Ni.O(1) N(3).Ni.O(1) N(2).Ni.O(2) N(3).Ni.O(2) O(1).Ni.O(2) N(1).O(2).Ni C(1).N(2).Ni 2.042(3) 2.062(3) 2.156(3) 1.623(3) 1.244(3) 90.46(13) 94.51(13) 93.33(12) 84.45(13) 84.07(12) 85.76(13) 86.51(12) 57.48(11) 94.7(2) 161.5(2) Ni.N(5) Ni.N(4) Ni.O(2) O(1).N(1) N(2).C(1) N(2).Ni.N(3) N(2).Ni.N(4) N(3).Ni.N(4) N(5).Ni.O(1) N(4).Ni.O(1) N(5).Ni.O(2) N(4).Ni.O(2) N(1).O(1).Ni O(2).N(1).O(1) N(2).C(1).S 2.045(3) 2.087(3) 2.166(4) 1.251(4) 1.148(4) 168.36(9) 94.59(12) 95.62(11) 106.13(12) 160.52(9) 163.44(9) 103.03(12) 94.9(2) 112.8(2) 179.1(2) The criterion of best .t was the minimum value of R = ¥Òi(¥öi calc ¥öi obsd)2/(i n), where n is the number of free parameters (3).The results of the .t, shown as the solid lines in Fig. 5, were J = 12.8 cm1, g = 2.17 for complex 1 and J = 12.2 cm1, g = 2.27 for 2. It is interesting that the interaction through the same kind of ¥ì-(¥ç1-N:¥ç2-O,O)-nitrito bridge in the Ni.Mn bimetallic chain [{MnNi(NO2)4(en)2}n] is ferromagnetic.2 The similar values of the coupling constants found for 1 and 2 can be explained in terms of similar bond parameters in the bridging region.Magneto-structural correlations The lower value of the superexchange coupling constant for the O,O,N nitrito bridge described in this paper, close to 12 cm1, in comparison with the well established value of J, close to 30 cm1, reported for the O,N nitrito bridge 16 is surprising. The co-ordination of the second oxygen atom seems to a.ord an increase in the overlap in the bridging region and this suggests a strong interaction, in contrast with experimental results.This experimental .nding can be explained by reference to the Hay.Thibeault.Ho.mann relationship 17 between the ¥Ò.2 and the antiferromagnetic contribution of J. The . values (difference of energy between active MOs of the same symmetry) Fig. 5 Molar magnetic susceptibility vs. T plots of a polycrystalline sample of [Ni2(¥ì-NO2)(NCS)3(Medpt)2]H2O 1 () and [Ni2(¥ì-NO2)- (NCS)3(dpt)2] 2 (). Solid line shows the best .t using the expression for the magnetic susceptibility of isotropically coupled dinuclear S = 1 ions.Table 5 Selected bond lengths [A] and angles [] for [Ni(NO2)(NCS)- (Medien)] 4 Ni.N(1) Ni.N(5) Ni.N(4) S.C(1) O(2).N(2) N(1).Ni.N(3) N(3).Ni.N(5) N(3).Ni.O(1) N(1).Ni.N(4) N(5).Ni.N(4) N(1).Ni.O(2) N(5).Ni.O(2) N(4).Ni.O(2) N(2).O(2).Ni O(2).N(2).O(1) 2.0416(15) 2.0741(18) 2.1439(14) 1.6309(17) 1.255(2) 92.63(6) 104.30(8) 99.63(7) 175.22(7) 83.33(5) 90.43(7) 97.47(7) 94.21(5) 94.60(12) 111.87(17) Ni.N(3) Ni.O(1) Ni.O(2) O(1).N(2) N(1).C(1) N(1).Ni.N(5) N(1).Ni.O(1) N(5).Ni.O(1) N(3).Ni.N(4) O(1).Ni.N(4) N(3).Ni.O(2) O(1).Ni.O(2) N(2).O(1).Ni C(1).N(1).Ni N(1).C(1).S 2.0575(17) 2.1366(16) 2.1567(17) 1.269(2) 1.149(2) 94.95(6) 88.82(7) 155.56(6) 83.50(5) 94.59(5) 157.66(6) 58.29(7) 95.13(12) 168.23(17) 178.47(16)J.Chem. Soc., Dalton Trans., 1999, 3115.3119 3119 may easily be obtained by means of MO extended Huckel calculations by the CACAO program.18 Calculations were performed on a dimeric model using as input parameters Ni.O and Ni.N (nitrito) distances 2.25 A, O.N.O angle 114, and the remaining co-ordination sites occupied by NH3 molecules at a Ni.N distance of 2.075 A.The result of the calculation indicates that the two oxygen atoms mainly interact with the dx2 y2 atomic orbital of one of the nickel atoms, whereas the nitrogen interacts with the dz2 orbital of the second nickel atom. In Fig. 6 the plot of one of the antibonding MOs involved in the superexchange pathway is shown; this plot indicates strict orthogonality between the corresponding dx2 y2 atomic orbitals, that is to say .2 = 0, and the only active pathway for the superexchange interaction should be between the two reversed dz2 orbitals. If we compare the shape and orientation of the dz2 orbitals for the two kinds of nitrito bridges, the lower ¥Ò.2 and J values found for the O,O,N co-ordination as a consequence of the reduction of the overlap are evident.Acknowledgements Financial support for this work was generously given by Direccion General de Investigacion Cienti.ca y Tecnica through Grant PB96/0163. Fig. 6 Plot of one of the antibonding MOs involved in the superexchange pathway showing the axial.equatorial dz2 interaction. References 1 A. Escuer, R. Vicente, S. B. Kumar, X. Solans, M. Font-Bardia and A. Caneschi, Inorg. Chem., 1996, 35, 3094. 2 A. Escuer, M. S. El Fallah, S.B. Kumar, F. Mautner and R. Vicente, Polyhedron, 1998, 18, 377. 3 J. A. Halfen, S. Mahapatra, E. C. Wilkinson, A. J. Gengenbach, V. G. Young, L. Que and W. B. Tolman, J. Am. Chem. Soc., 1996, 118, 763. 4 O. Kahn, E. Bakalbassis, C. Mathoniere, M. Hagiwara, K. Katsumata and L. Ouahab, Inorg. Chem., 1997, 36, 1530. 5 G. M. Sheldrick, SHELXS 86, Acta Crystallogr., Sect. A, 1990, 46, 467. 6 G. M. Sheldrick, SHELXL 93, University of Gottingen, 1993. 7 International Tables of X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. IV, pp. 99.100 and 149. 8 C. K. Johnson, ORTEP, Report ORNL-3794, Oak Ridge National Laboratory, Oak Ridge, TN, 1965. 9 A. J. Finney, M. A. Hitchman, D. L. Kepert, C. L. Raston, G. L. Rowbotton and A. H. White, Aust. J. Chem., 1981, 34, 2177. 10 A. J. Finney, M. A. Hitchman, C. L. Raston, G. L. Rowbotton and A. H. White, Aust. J. Chem., 1981, 34, 2159. 11 A. J. Finney, M. A. Hitchman, C. L. Raston, G. L. Rowbotton and A. H. White, Aust. J. Chem., 1981, 34, 2113. 12 A. Escuer, R. Vicente and J. Ribas, Transition Met. Chem., 1993, 18, 478. 13 M. G. B. Drew, D. M. L. Goodgame, M. A. Hitchman and D. Rogers, Chem. Commun., 1965, 477. 14 R. Birdy, D. M. L. Goodgame, J. C. McConway and D. Rogers, J. Chem. Soc., Dalton Trans., 1977, 1730. 15 M. J. Goldberg and R. E. Marsh, Acta Crystallogr., Sect. B, 1979, 35, 960. 16 A. Escuer, R. Vicente and X. Solans J. Chem. Soc., Dalton Trans., 1997, 531 and refs. therein. 17 J. P. Hay, J. C. Thibeault and R. Ho.mann, J. Am. Chem. Soc., 1975, 97, 4884. 18 CACAO, Computed Aided Composition of Atomic Orbitals, version 4.0, C. Mealli and D. M. Proserpio, J. Chem. Educ., 1990, 67, 3399. Paper 9/03756J

ISSN:1477-9226    

DOI:10.1039/a903756j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3121.3131 3121 Co-ordination engineering: when can one speak of an ¡°understanding¡±? Case study of the multidentate ligand 2,2-dimethyl-4,4-bipyrimidine ¢Ó Christoph Janiak,*a Lars Uehlin,a He-Ping Wu,a Peter Klufers,*b Holger Piotrowski b and Tobias G. Scharmannc a Institut fur Anorganische und Analytische Chemie, Universitat Freiburg, Albertstr. 21, D-79104 Freiburg, Germany. E-mail: janiak@uni-freiburg.de b Institut fur Anorganische Chemie, Universitat Karlsruhe, Kaiserstr. 12, D-76131 Karlsruhe, Germany c Institut fur Anorganische und Analytische Chemie, Technische Universitat Berlin, Stra©�e des 17. Juni 135, D-10623 Berlin, Germany Received 17th June 1999, Accepted 7th July 1999 The mode of co-ordination of the multidentate ligand 2,2-dimethyl-4,4-bipyrimidine (L) was found to depend on the metal ion, the crystallization conditions, the metal-to-ligand ratio, and the anion. With nickel a chelating co-ordination through the endo-dentate nitrogen donor set is observed in the molecular complex [NiCl2(L)(H2O)]CH3NO2, derived from hot CH3NO2.With Cu(NO3)2 and CuI.CH3CN one- and two-dimensional (1-D and 2-D) co-ordination polymers of formula ¡Ä 1[Cu(NO3)2(¥ì-L)] and ¡Ä 2[Cu2(¥ì3-I)2(¥ì-L)] are obtained, where the bipyrimidine ligand is solely bridging through the two exo-dentate nitrogen atoms. On the other hand, a synthesis from CuI and crystallization from hot dimethyl sulfoxide leads to a 1-D iodide-bridged co-ordination polymer ¡Ä 1[Cu(¥ì-I)(L)] with a chelating ligand. With AgNO3 two di.erent types of co-ordination polymers were found, depending on the silver-to-ligand ratio.At a 1 : 1 ratio in the presence of a co-ordinating anion a 2-D network, ¡Ä 2[Ag(¥ì-NO3)2(¥ì-L)], with only bridging bipyrimidine ligands is observed. At a metal excess, a 3-D framework, ¡Ä 3[Ag3(¥ì3-NO3)3(¥ì3-L)2], forms where L functions both as a chelating and as bridging ligand. A tetradentate co-ordination mode of L towards silver is also found with non-co-ordinating anions, such as BF4 and PF6 , and gives rise to the isostructural 2-D co-ordination polymers of formula ¡Ä 2[Ag3(CH3CN)3(¥ì3-L)2]X3 (X = BF4 or PF6).Introduction Metal complexes of chelating 2,2-bipyridine or bridging 4,4- bipyridine or hydrogen-substituted derivatives thereof are of constant and general interest in metal co-ordination chemistry.1 The latter has recently gained considerable interest in the synthesis of (rectangular) two-dimensional network structures.2 The generation of such frameworks is a promising path in the search for stable microporous metal.organic networks that exhibit reversible guest exchange and possibly selective catalytic activity.3 In view of the large use of heterocyclic bipyridine ligands, it is remarkable that the related 2,2-bipyrimidine ligand or its derivatives, such as 2,2-dimethyl-4,4-bipyrimidine (L), have so far rarely been employed in metal co-ordination chemistry.4.6 Dedicated to Professor Dr Heinrich Vahrenkamp on the occasion of his 60th birthday.¢Ó Supplementary data available: X-ray powder di.ractograms. Available from BLDSC (No. SUP 57608, 4 pp.). See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton). These ambi- or multi-dentate bipyrimidine ligands can be thought of combining the ligating properties of the chelating and bridging bipyridine ligands.The idea behind the use of such ambidentate ligands is to have tetrahedral building blocks for co-ordination polymers7 based on the endo-chelation of two ligands with an appropriate metal center (1) or to supply functional donor atoms within the walls of the co-ordination polymer when the ligands are solely exo-bridging (2). Such functionalities should eventually interact with organic guest molecules, e.g. through hydrogen bonding, or allow for the anchoring of additional metal ions.Our research has been concerned with the utilization of multidentate endo-chelating/exo-bridging modi.ed 2,2-bipyridine ligands such as 2,2-bi-1,6-naphthyridine 8 or 5,5-dicyano- 2,2-bipyridine 9 and ligands of the tris(pyrazolyl)borate type for the assembly of metal co-ordination polymers.10 In this paper the ligand 2,2-dimethyl-4,4-bipyrimidine (L) serves as an example to derive at a more profound understanding in the co-ordination engineering with multidentate ligands, as a part of crystal engineering.We describe the results of synthetic and structural studies of metal salts with L and try to elucidate the factors which lead to bridging or chelating or simultaneous bridging/chelating co-ordination. Through the use of a variety of metal centers and crystallization conditions we want to explore a large section of the energy hypersurface of ligands in co-ordination chemistry. An understanding of the self-assembly process between metal ions and .exible, multidentate ligands is a contemporary goal in supramolecular co-ordination chemistry.11 While rigid bidentate ligands, for example, allow only for one result in the metal co-ordination process, .exible and multidentate ligands3122 J.Chem. Soc., Dalton Trans., 1999, 3121¡V3131can give two or more metal structures with the same stoichiometry.This has been termed an ¡§inorganic supramolecularlibrary¡� or ¡§virtual combinatoric library¡� by Lehn and coworkers.12 It is not yet possible to predict the product in thesecases. The structure can be determined by subtle factors such as£k¡V£k stable interactions between the ligands or by the anions.Results and discussionThe ligand 2,2-dimethyl-4,4-bipyrimidine (L) was preparedfollowing a slightly modied literature procedure by Eenberger.13,14 The reaction of acetamidinium hydrochloride with1,6-bis(ethoxy)hexa-1,5-diene-3,4-dione gave L in 42% yield,eqn.(1). The structure of L is shown in the s-cis conformationwhich is the appropriate conformer for chelate formation.A structure determination in the solid state revealed thes-trans conformation of the pyrimidine rings.A semiempiricalPM3 calculation in vacuum and in chloroform supported theassignment of the s-trans form as the preferred conformation.14The aim in the syntheses of the transition metal complexes ofL was to obtain single crystals suitable for X-ray diraction.The diusion technique was generally used: a solution of theligand was overlayered with a solution of the metal salt.Scheme 1 summarizes the reactions and products which aredescribed in more detail below.Fig. 1 gives a schematic overview on the structural resultswhich are systematically arranged according to the followinginterpretation which we propose. There are four principal coordinationmodes of the ambidentate bipyrimidine ligand:bidentate chelating, bidentate bridging, tridentate, and tetradentate.The tridentate mode has been observed before,6 theother three are presented here.Generally, a bridging coordinationmode gives rise to a one-dimensional metal¡Vligandsubstructure. Bridge formation together with metal chelationresults in a two-dimensional metal¡Vligand substructure.Depending on the anion, both types of substructures can thenbe connected further to a two- or three-dimensional framework,respectively.The major determinator for the co-ordination mode of Lappears to be the type of metal ion. With nickel() only thebidentate chelating mode was elucidated (complex 4, [NiCl2-(L)(H2O)]CH3NO2).Copper() was only found in the bidentatebridging mode (5, ¡Û1[Cu(NO3)2(-L)]). Towards copper()and silver() more than one co-ordination mode was observedhere, so that there was an opportunity to check for additionalvariables.Bidentate chelating versus bidentate bridging in the case ofcopper(I): inuence of crystallization conditions (cf. Fig. 1)For copper() the co-ordination mode can be controlledthrough the thermodynamic crystallization conditions. Crystallizationfrom a hot solution or after thermal treatment leads toa chelate complex 8, ¡Û1[Cu(-I)(Ln the whole synthesisand crystallization procedure is carried out at room temperaturea ligand-bridged co-ordination polymer is obtained incomplex 7, ¡Û2[Cu2(3-I)2(-L)]. These conclusions for the case ofcopper() are supported through the crystallization conditionsof the other examples.The nickel chelate structure 4 was crystallizedfrom hot CH3NO2. The copper() and silver() coordinationpolymers 5 and 9, respectively, were crystallized atroom temperature. The origin of these dierences may lie in theconformational s-trans preference of L together with a chargedierence of the nitrogen donors. An AM1 or PM3 theoreticalcalculation 15 assigns a higher negative charge to the endodentatenucleophiles. Other orbital interactions of the nitrogendonors being equal, a higher negative charge leads to a thermodynamicallymore stable electrostatic interaction with themetal.10,16 At the same time, the resulting chelate co-ordinationrequires the ligand to assume the less preferred s-cis conformation.A competitive thermodynamic versus kinetic control mayalso be invoked in the assembly process.17A tridentate co-ordination mode of copper() with L wasdescribed earlier in the compound ¡Û1[Cu2(CH3CN)2(-L)][PF6]2.6There, the participation of one exo- and both endo-nitrogendonors yields an innite helicoidal polymeric chain around a21 axis.This dierence to the copper()¡Vbipyrimidine¡Viodinestructures 7 and 8 can be traced to the change in the anion:co-ordinating iodine versus ¡§non¡�-co-ordinating hexauorophosphate(see below).Bidentate bridging versus tetradentate in the case of silver(I):inuence of the metal-to-ligand ratio and of the anion (cf. Fig. 1)Among the metals examined, only silver() was found capableof occupying all four donor atoms of L in metal co-ordination,as shown by the structures of 10 through 12.At the same timethis appears to be the favored co-ordination mode for silver. Inorder to have only the exo-dentate nitrogen donor set of Lco-ordinating to silver() a 1 : 1 metal-to-ligand ratio has to beemployed in the presence of a co-ordinating (nitrate) anion.J. Chem. Soc., Dalton Trans., 1999, 3121–3131 3123 Fig. 1 Schematic presentation of the co-ordination modes of 2,2-dimethyl-4,4-bipyrimidine (L) towards di.erent metal centers (a) together with a brief indication on the proposed underlying determining variables for copper(.) (b) and silver(.) (c).Scheme 1 Overview of the metal complex formation reactions of L with nickel, copper, and silver salts. Then one derives the structure of 9, 8 2[Ag(µ-NO3)(µ-L)], with solely bridging co-ordination. Otherwise, if an excess of silver was used in the case of the NO3 anion, tetradentate coordination was achieved.The structure of 10, 8 3[Ag3(µ3-NO3)3- (µ3-L)2], is the example. An obvious variable in the silver structures is the anion.18,19 In the networks of 9 and 10 the co-ordinating nitrate anion plays a dominant role in the connectivity of the framework. To check for the in.uence of the anion, the “non”-coordinating counter ions BF4 and PF6 were employed as their silver salts as starting materials. The crystalline products 11 and 12, 8 2[Ag3(CH3CN)3(µ3-L)2]X3, from AgBF4 and AgPF6, respectively, were found to be isostructural.9,18b,20 The structures exhibit the tetradentate co-ordination mode.The result did not depend on the metal-to-ligand ratio here, which was 1 : 1 in the case of X = BF4 (11) and 2 : 1 for X = PF6 (12). Thus, for a 1 : 1 metal-to-ligand ratio the co-ordination of L to silver can be controlled by the counter ion. The above relations are not meant to imply that we have reached a full understanding. Matters may still be more complicated in view of the large number of parameters in a crystallization experiment.However, we feel that we can provide some guidelines and rationalizations which should be applicable at least for similar tetradentate chelating/bridging ligands. The individual structures are brie.y described in more detail in the following. [NiCl2(L)(H2O)]CH3NO2 4. Fig. 2 illustrates the molecular structure of aquadichloro(2,2-dimethyl-4,4-bipyrimidine)- nickel(..): a chelate complex with s-cis conformation of the pyrimidine rings.A somewhat unusual trigonal-bipyrimidal geometry around nickel(..) is found,21 with the stronger nitrogen and oxygen donor atoms in the equatorial plane. The weaker chloride ions occupy the axial sites. Apparently, the methyl groups prevent the formation of an octahedral coordination polyhedron. A nitromethane molecule is incorporated in the lattice but with distances larger than 3 Å from the metal atom.3124 J.Chem. Soc., Dalton Trans., 1999, 3121.3131 ¡Ä 1[Cu(NO3)2(-L)] 5. A crystal structure determination showed that the bipyrimidine ligand bridges between two copper centers with its exo-dentate nitrogen atoms to give a linear .Cu.L.Cu.L. chain, which is indicated in Fig. 3. The endo-dentate donor atoms of L are not involved in metal coordination and the ligand assumes the s-trans conformation. Two chelating nitrate ions complete the Jahn.Teller-distorted pseudo-octahedral co-ordination sphere of the copper centers.The Jahn.Teller distortion is along two trans Cu.O contacts, one to each of the nitrato ligands. The small bite angle of the nitrate anions leads to a considerable distortion of the octahedral co-ordination polyhedron. ¡Ä 2[Cu2(3-I)2(-L)] 7. Fig. 4 illustrates the two-dimensional coordination polymer of this compound.22 The ligand assumes the s-trans conformation and bridges with its exo-dentate nitrogen atoms between two kinked copper iodide double strands.The CuI substructure is an in.nite stair polymer in which each iodide is surrounded by three metal atoms and each copper in turn by three iodide ions. Analogous structures can be found in CuI compounds with pyridine, 2-methylpyridine and 2,4- dimethylpyridine.23,24 The stairs run parallel to the crystallographic c axis. The double strands can be thought of being built from planar parallelograms which are formed from two copper and two iodine atoms. The co-ordination polyhedron at copper is a distorted tetrahedron.The bipyrimidine ligands have a role Fig. 2 Molecular structure of [NiCl2(L)(H2O)] 4. Selected distances [A] and angles []: Ni.N(1) 2.010(2), Ni.N(3) 2.010(2), Ni.O(91) 1.957(3), Ni.Cl(1) 2.3441(7) and Ni.Cl(2) 2.3234(8); O(91).Ni.N(1) 146.00(14), O(91).Ni.N(3) 133.45(14), O(91).Ni.Cl(1) 85.86(7), O(91).Ni.Cl(2) 87.59(7), N(1).Ni.N(3) 80.40(9), N(1).Ni.Cl(1) 90.36(6), N(1).Ni.Cl(2) 93.16(6) and Cl(1).Ni.Cl(2) 172.60(3).Fig. 3 Section of the linear chain structure of ¡Ä 1 [Cu(NO3)2(¥ì-L)] 5. Selected distances [A] and angles []: Cu.N(1),N(1)_2 2.051(4), Cu. O(1),O(1)_2 2.032(4) and Cu.O(2),O(2)_2 2.531(6); O(1).Cu.O(1)_2 180, N(1).Cu.N(1)_2 180, O(1).Cu.N(1) 91.0(2), O(1)_2.Cu.N(1) 89.0(2), O(1).Cu.N(1)_2 89.0(2), O(1)_2.Cu.N(1)_2 91.0(2), O(1). Cu.O(2)_2 123.3(2), O(1).Cu.O(2) 56.7(2), N(1).Cu.O(2)_2 90.9(2), N(1)_2.Cu.O(2)_2 89.1(2), O(1)_2.Cu.O(2)_2 56.7, O(1)_2.Cu.O(2) 123.3(2), N(1).Cu.O(2) 89.1(2), N(1)_2.Cu.O(2) 90.9(2), O(2)_2.Cu.O(2) 180, O(2).N(10).O(1) 117.5(4), O(3).N(10).O(1) 118.4(5) and O(3).N(10).O(2) 124.2(5) (symmetry equivalent position _2 x 1, y 1, z 1). of a spacer between the stairs and bridge the strands to a layer structure. ¡Ä 1 [Cu(-I)(L)] 8. Fig. 5 shows the structure of this onedimensional co-ordination polymer22 which was obtained from hot DMSO solution. The bipyrimidine ligand is in its s-cis conformation and chelates the copper center, as is usually found for bipyridine copper complexes.25 The exo-dentate nitrogen atoms are not involved in metal co-ordination.The pseudo-tetrahedral co-ordination sphere of the copper ions is completed by two iodide ions so that a Cu.I zig zag chain is formed. The chain runs parallel to the crystallographic c axis which is also the needle axis. Similar Cu-I chains are observed with the pyridine derivatives acridine 26 and collidine (2,4,6-trimethylpyridine). 27 We note that dimeric (Cu2I2) or tetrameric cubane or chair/stepped cubane (Cu4I4) structures mit have been another possibility for the arrangement of the CuI substructure. 26,28.31 ¡Ä 2 [Ag(-NO3)(-L)] 9. Fig. 6 depicts the structure of the twodimensional network of complex 9.32 The framework can be thought of being built up from zig zag Ag.NO3.Ag.NO3 chains which are bridged by the bipyrimidine ligands in their s-trans form through the exo-dentate donor atoms. This bridging action of a metal.anion strand is akin to the structural motif found above in 7.¡Ä 3 [Ag3(3-NO3)3(3-L)2] 10. The structure of the threedimensional framework of complex 10 is shown in Fig. 7. Fig. 4 Layer structure of the two-dimensional co-ordination polymer ¡Ä 2 [Cu2(¥ì3-I)2(¥ì-L)] 7 with an in.nite one-dimensional CuI stair substructure bridged by bipyrimidine ligands, viewed perpendicular to the ab plane. Selected distances [A] and angles []: Cu.I 2.6014(7), Cu.I_4 2.6532(7), Cu.I_1 2.7347(8), Cu.N(1) 2.030(3), Cu Cu_4 2.649(1), Cu.I.Cu_4 60.53(2), Cu_3.I_4.Cu_4 103.80(2), Cu.I_4.Cu_3 82.61(2), N(1).Cu.I 113.3(1), N(1).Cu.I_4 114.5(1), N(1).Cu.I_1 105.03(9), I.Cu.I_4 119.47(2), I.Cu.I_1 103.80(2) and I_1.Cu.I_4 97.39(2) (symmetry equivalent positions _1 x, y, z 1; _2 x 1, y 1, z; _3 x 2, y, z; _4 x 2, y, z 1). Fig. 5 Chain structure along c of the one-dimensional co-ordination polymer ¡Ä 1 [Cu(¥ì-I)(L)] 8 with a CuI zig zag chain substructure.Selected distances [A] and angles []: Cu.I 2.5374(5), Cu.I_1 2.5937(5), Cu. N(13) 2.091(2) and Cu.N(23) 2.079(2); Cu.I_1.Cu_1 = I.Cu.I_1 104.049(14), N(13).Cu.N(23) 79.45(10), N(13).Cu.I 127.00(8), N(13).Cu.I_1 104.84(8), N(23).Cu.I 122.05(10) and N(23).Cu.I_1 118.27(10) (symmetry equivalent position: _1 x, y, z 1).J. Chem. Soc., Dalton Trans., 1999, 3121–3131 3125 The structure is an example where L serves both as a chelating and as a bridging ligand at the same time, thereby utilizing all four donor atoms in metal co-ordination (cf. 1). The structure contains three crystallographically di.erent types of silver centers. Two metal atoms, Ag1 and Ag2, are similarly chelated by two bipyrimidine ligands, each, in a four-co-ordinate strongly distorted environment in between a tetrahedral and a square-planar co-ordination sphere (cf. Fig. 7a). The dihedral angles between the two .ve-membered chelate rings are 48.7(4) for Ag1 and 51.7(5) for Ag2.The remaining type of silver atom, Ag3, is .ve-co-ordinate (cf. Fig. 7b). The axial positions of the trigonal-bipyramidal co-ordination polyhedron are occupied by two exo-dentate nitrogen atoms of two bridging bipyrimidine ligands. The equatorial positions are .lled by oxygen atoms of three nitrato groups which bridge between three such silver centers of the same type. The structure of 10 can be viewed as being built by the distorted tectons 1 which are bridged again by metal centers.The ligating action of L towards Ag1 and Ag2 together with the exo co-ordination towards Ag3 yields bipyrimidine–silver–bipyrimidine sandwich layers which are connected to a three-dimensional network through the nitrate bridges between the Ag3 centers. In an alternative view, the structure can be thought of being built from linear Ag3–bipyrimidine chains. These chains run crisscross parallel to the ab plane and are connected by the Ag1 and Fig. 6 Two-dimensional network of 8 2 [Ag(µ-NO3)(µ-L)] 9 built from the bridging of in.nite Ag-NO3-Ag-NO3 strands by the bipyrimidine ligands. (a) View perpendicular to a layer section (along b), (b) view along the layers (along a). Selected distances [Å] and angles []: Ag– N(1),N(1)_3 2.246(2) and Ag–O(82) 2.449(3); N(1)–Ag–N(1)_3 137.0(1), N(1)–Ag–O(82) 120.03(5) and N(1)_3–Ag–O(82) 87.59(7) (symmetry equivalent positions _3 x, y, z ½; _5 x 1, y 1, z 1). Ag2 centers as well as by the nitrate bridges (cf.Fig. 7c). Two additional nitrate positions complete the structure; they have only long-distance contacts to the co-ordination-polymeric part of the structure, though, and, therefore, are strongly disordered. 8 2[Ag3(CH3CN)3(3-L)2]X3 with X BF4 (11) or PF6 (12). The two-dimensional grid structures are illustrated in Figs. 8 and 9. The counter ions are in part disordered in both cases. As in the structure of 10, also in 11 and 12 each bipyrimidine ligand is again serving as a chelating and as a bridging group using all four of its nitrogen donors.The complexes contain two di.erent sets of silver centers. The co-ordination sphere of the fourco- ordinated Ag1 atoms is constructed from two endo-dentate nitrogen donors of a chelating bipyrimidine ligand, an exonitrogen donor of a bridging group and an acetonitrile molecule. The environment of Ag2 is three-co-ordinate and made up of two exo-dentate nitrogen donor atoms from bridging bipyrimidine ligands and a CH3CN molecule.In a di.erent perspective, the structures of can also be thought of as being built up from a helicoidal one-dimensional co-ordination polymer based on the Ag1–ligand units (see shaded ligands in Fig. 8). This helix follows the 21 screw axis parallel to b (running vertical in Fig. 8). At the rims of these ribbons there are Ag2 centers, which link the adjacent helical strands. NMR spectroscopy The nickel and copper(..) compounds 3, 4 and 5 as well as the silver compounds 9 and 10 were soluble in dimethyl sulfoxide and the solutions investigated by proton and carbon-13 NMR.The NMR data are included in Table 1. The spectra are similar in their chemical shifts and resemble the spectra of the “free” ligand. This is evidence that the co-ordination polymeric framework degrades. Especially, since the spectra with the paramagnetic metal centers NiII and CuII neither show a pronounced broadening nor a considerable chemical shift di.erence, as would be expected if the ligand remained bound to the metal ions.The slight broadening of the signal of H6 and the disappearance of the coupling to H5 and H6 is probably due to the presence of the paramagnetic ions in solution. In the spectra of the diamagnetic silver salts the coupling is retained. The peak assignment is as given for the free 2,2-dimethyl-4,4- bipyrimidine ligand L (cf. footnote to Table 1). Infrared spectroscopy The vibrational spectroscopy data was interpreted with the help of results from pyrimidine,33 2-methylpyrimidine,34 and bipyridine 35 ligands and complexes.Most bands of 2,2- dimethyl-4,4-bipyrimidine L show a small shift upon complexation. The ring vibrations at 1575 and 1445 cm1 show a bathochromic shift (to smaller wavenumbers) of up to 40 cm1, in part together with band splitting. The out-of-plane deformation at 770 cm1 experiences a bathochromic shift between 8 and 20 wavenumbers.The in-plane vibrations at 840, 655 and 588 cm1 are shifted from 5 to 15 cm1 to higher wavenumbers. Similar shifts of these aforementioned so-called metal-sensitive vibrations were observed for pyrimidine and bipyridine metal complexes.33,35 Conclusions In the multidentate ligand 2,2-dimethyl-4,4-bipyrimidine (L) both the endo- and exo-dentate nitrogen atoms can function as donors towards metal centers. The choice of co-ordination was found to depend .rst on the type of metal.For a given metal center the mode of co-ordination can further be in.uenced through the thermal crystallization conditions, the anion and the metal-to-ligand ratio as was exempli.ed for copper(.) and silver(.). Solvent control might be another unexplored variable. 36 For copper(.) it is shown how the thermal crystallization3126 J. Chem. Soc., Dalton Trans., 1999, 3121.3131 Fig. 7 (a) Sections of the co-ordination polymer of ¡Ä 3 [Ag3(¥ì3-NO3)3(¥ì3-L)2] 10 showing one of the two bis-chelated silver centers.(b) Section of 10 which details the bridging action of Ag3 between the bis-chelate tectons of Ag1 and Ag2. Only the exo-dentate nitrogen donor atoms are shown for clarity. Also indicated is the connection of three silver atoms by a nitrate group and the surrounding of each Ag3 center by three nitrato ligands in the equatorial plane of the trigonal bipyramid. (c) Stereoscopic view along a. Selected distances [A] and angles []: Ag(1).N(11) 2.40(1), Ag(3).N(22) 2.21(1), Ag(3).N(12) 2.23(1), Ag(3).O(1) 2.91(1), Ag(3).O(2)_8 2.87(1) and Ag(3).O(3)_7 2.78(1); N(11)_1.Ag(1).N(11)_2 150.3(6), N(11)_1.Ag(1).N(11) 116.1(5), N(11)_2.Ag(1).N(11) 72.1(5), N(11).Ag(1).N(11)_3 150.3(6), N(12).Ag(3).N(22) 178.5(4), O(1).Ag(3).O(2)_8 134.3(5), O(1).Ag(3).O(3)_7 112.8(5) and O(2)_8.Ag(3).O(3)_7 112.9(5). For the bis-chelated silver center Ag2 (similar to Ag1), which is not shown here: Ag(2).N(21) 2.43(1), N(21)_4.Ag(2).N(21) 71.8(6), N(21).Ag(2).N(21)_5 148.7(7) and N(21).Ag(2).N(21)_6 117.2(6) (symmetry equivalent positions _1 x �ö, y, z �ö; _2 x �ö, y, z; _3: x, y, z �ö; _4 x �ö, y 1, z; _5 x, y 1, z �ö; _6 x �ö, y 1, z; _7 x �ö, y, z; _8 x 1, y, z).conditions allow one to steer between bidentate chelating or bridging. The case of silver(.) illustrates the e.ect of the anion and the metal-to-ligand ratio. The combination of both factors allows for the choice between bidentate bridging or tetradentate. Depending on ones viewpoint the above results may be viewed as a hopeful step towards a certain understanding or as a still disappointing high number of variables.We acknowledge a high number of variables and do not pretend that we have reached a full understanding. With respect to the title of the manuscript, the individual reader according to his expectations may say if a sense of understanding has been reached.Structure.variable relations are not always as simple as we would like to have them. Yet, we feel that we could pinpoint certain correlations which should be taken into account in structure design. In co-ordination or crystal engineering we are likely to see a higher degree of complexity as we turn to multidentate ligands or multicomponent metal.ligand systems. For this, it will be useful to develop a deeper understanding for the structure control through such factors as type of metal, the counter ion, the metal-to-ligand ratio and others.Within the concept of an ¡°inorganic supramolecular library¡±12 we plan eventually to be able to choose in an assembly or crystallization process those conditions which most likely lead to the target structural motif. Experimental The NMR spectra were collected on a Bruker ARX 200 (200.1 MHz for 1H, 50.3 MHz for 13C) or a Varian O-300 instrument (300.0 MHz for 1H, 75.4 MHz for 13C) and calibrated against the solvent signal (d6-DMSO, 1H ¥ä 2.53, 13C ¥ä 39.5), IR spectra on a Perkin-Elmer 783 spectrophotometer as KBr disks or Nujol mulls.Elemental analyses were carried out with aJ. Chem. Soc., Dalton Trans., 1999, 3121–3131 3127 Perkin-Elmer Elemental Analyzer E 240 C. X-Ray powder diffractograms (SUP 57608) were obtained with a Siemens powder di.ractometer D5000 using Cu-Ka radiation. All crystallizations of the silver complexes were carried out in the dark.The reactions with copper(.) iodide were carried out with Schlenk techniques using .ame-dried glassware and argon as inert gas. Acetonitrile was dried by re.uxing for 2 d over CaH2 followed by distillation and storage under inert gas. Preparations 2,2-Dimethyl-4,4-bipyrimidine L. To a solution of sodium (14 g, 0.61 mol) in dry ethanol (300 ml) was added under argon acetamidinium hydrochloride (56 g, 0.59 mol) at 0 C. The pale yellow solution was stirred for 1 h, cooled to Fig. 8 Section of the two-dimensional co-ordination polymeric network of isostructural 8 2 [Ag3(CH3CN)3(µ3-L)2][BF4]3 11 and 8 2 [Ag3- (CH3CN)3(µ3-L)2][PF6]3 12, viewed along a. The acetonitrile ligand which points in the framework opening lies on an twofold axis; the hydrogen atoms are thus disordered. The BF4 and PF6 anions were omitted for clarity. One of the helicoidal one-dimensional substructures is highlighted by shading the bipyrimidine ligands. Selected distances [Å] and angles [] for 11 [12]: Ag(1)–N(11) 2.229(2) [2.248(2)], Ag(1)– N(31)_6 2.383(2) [2.391(2)], Ag(1)–N(32)_6 2.301(2) [2.312(2)], Ag(1)– N(18) 2.442(3) [2.406(3)], Ag(2)–N(19) 2.272(4) [2.240(4)] and Ag(2)– N(12) 2.279(2) [2.278(2)]; N(11)–Ag(1)–N(32)_6 148.71(8) [143.30(6)], N(11)–Ag(1)–N(31)_6 126.59(7) [126.05(7)], N(31)_6–Ag(1)–N(32)_6 71.51(7) [71.58(6)], N(11)–Ag(1)–N(18) 93.49(9) [98.81(9)], N(32)_6– Ag(1)–N(18) 111.25(10) [110.47(9)], N(31)_6–Ag(1)–N(18) 95.45(10) [99.07(8)], N(19)–Ag(2)–N(12) 114.98(5) [114.65(5)] and N(12)– Ag(2)–N(12)_2 130.05(11) [130.7(1)] (symmetry equivalent positions _2 x ½, y, z 1; _6 x, y ½, z ½).10 C and then 1,6-bis(ethoxy)hexa-1,5-diene-3,4-dione 13 (39.6 g, 0.20 mol) in dry ethanol (100 ml) was added dropwise. After complete addition, stirring was continued for 1 h at 10 C, then the solution was allowed to warm to room temperature and stirred for 4 d. The solvent was removed in vacuum and the residue suspended in acetone (600 ml).The red-brown solid was .ltered o., sublimed in vacuum and recrystallized from ethanol to give 16 g (42%) of colorless needles, mp 134–135 C (lit. 133–135 C13), bp. 229 C (from DTA/TG). 1H NMR (d6-DMSO): d 2.71 (s, 6 H, CH3,CH3), 8.13 (dd, 2 H, J = 0.5, 5.0, H5,5) and 8.90 (d, 2 H, J = 5.0 Hz, H6,6). 13C NMR (d6-DMSO): d 26.1 (CH3,CH3), 114.6 (C5,5), 158.5 (C4,4), 161.1 (C2,2) and 168.3 (C6,6). IR: 3060w, 2980w, 2950w, 1575s, 1545s, 1445s, 1390s, 1357s, 1185m, 1140w, 1100w, 1050m, 1000m, 840s, 770s, 655s, 588s, 427w and 404s cm1.[Ni3Cl6(1)2] 3. The ligand L (186 mg, 1.00 mmol) was dissolved in warm ethanol (40 ml) and added to a solution of NiCl26H2O (357 mg, 1.5 mmol) in dry ethanol (40 ml). After several days at room temperature a yellow precipitate had formed, which was .ltered o., washed with a little ice-cold ethanol and dried for 5 min in vacuum. Yield 316 mg (83%). Further analytical data are in Table 1.The di.usion technique did not produce single crystals here, even when under rather dilute conditions. Compound 3 was slightly soluble in hot nitromethane and slow cooling gave crystals of 4. Aquadichloro(2,2-dimethyl-4,4-bipyrimidine)nickel(II)– nitromethane adduct, [NiCl2(L)(H2O)]CH3NO2 4. The ligand L (186 mg, 1.00 mmol) was dissolved in warm CH3NO2 (50 ml). To this solution was added NiCl26H2O (238 mg, 1.00 mmol) and the mixture heated to re.ux for 5 min.The solution turned red and was .ltered hot from the yellow precipitate, which by IR was shown to be complex 3. The hot .ltrate was slowly cooled to room temperature and within 3 d transparent light green rhombohedral platelets formed. Upon prolonged drying in air the crystals lose solvent of crystallization. The pink solution could be used for another extraction of the yellow precipitate to yield additional crystal batches. Each crystallization a.orded 20 mg (5%) of product.A thermogravimetric analysis showed the loss of CH3NO2 at 77 C and Fig. 9 Stereoscopic view of the two-dimensional co-ordination polymeric network of complexes 11 and 12, viewed along c to show the folding of the layers and the helicoidal nature of the Ag1–ligand ribbon substructure (see text). Two BF4 anions (for 11), one of them disordered, are included in the picture. Similarly, one of the PF6 anions in 12 is twofold disordered. The disordered BF4 or PF6 moiety is located inside a hydrophobic channel without well de.ned bonding sites for an anion.3128 J.Chem. Soc., Dalton Trans., 1999, 3121¡V3131Table 1 Elemental analyses, NMR and IR spectroscopic data of the metal complexes of L aCalc., Found (%)1H NMR (d6- 13C NMRComplex C H N DMSO)b, £_(J/Hz)- (d6-DMSO)b, £_ IR (KBr, cm1) c3 [Ni3Cl6(L)2] 31.5631.832.652.6514.7214.762.58 (s), 8.12 (s),9.12 (s, br)23.9, 113.3, 156.1,158.1, 166.73400m (br), 3065m, 1603s, 1570s, 1560s,1470s, 1450s, 1410m, 1395m, 1155s, 1040m,880m, 865s, 767m, 745m, 670m, 595w4 [NiCl2(L)(H2O)]CH3NO233.4633.103.833.6917.7417.252.60 (s), 4.30 (s,H2O), 8.12 (s),9.18 (s, br)3400m (br), 1613s, 1600s, 1590s, 1557s, 1547s,1457s, 1406s, 1377m1110m,875m, 855s, 757s, 740m, 660s, 595m5 1¡Û [Cu(NO3)2-(-L)]32.1432.152.672.6722.4422.482.61 (s, br), 8.18(s, br), 9.18 (s, br)1601m, 1578w, 1510m, 1458m, 1390s, 1294m,1270m, 1022m, 873w, 811w, 760w, 688m,599w6 [Cu3I3(L)] 15.8615.961.331.297.407.401580s, 1550m, 1445s, 1392m, 1155m, 843s,750s, 661s, 595w7 2¡Û [Cu2(3-I)2-(-L)]21.1821.071.761.729.889.841580s, 1534s, 1433s, 1410m, 1350m, 1269m,1190m, 1100m, 1038m, 1010s, 755s, 660s,590w, 410m8 1¡Û [Cu(-I)(L)] 31.8931.782.683.5314.8817.651582s, 1560s, 1550s, 1447s, 1398m, 1320m,1190m, 1145m, 1110m, 835s, 760s, 649s,588m, 424s9 2¡Û [Ag(-NO3)-(-L)]33.7333.752.832.7119.6719.672.80 (s), 8.30 (dd,J = 5.1, 0.5), 9.01(d, J = 5.1)26.4, 115.4, 159.1,159.9, 167.81580s, 1540s, 1434s, 1410s, 1390s, 1316s,1275s, 1190m, 1037m, 850s, 830w, 820m,760m, 660s, 588w, 410m10 3¡Û [Ag3(3-NO3)3(3-L)]17.6417.322.282.5617.4717.182.81 (s), 8.12 (dd,J = 5.1, 0.5), 9.03(d, J = 5.1)26.6, 115.7, 158.7,160.1, 167.81770m, 1580s, 1540s, 1440s, 1390s (br),1275m, 1190m, 1100m, 1040m, 1010m, 853s,830s, 762s, 662s, 593m, 410s11 2¡Û [Ag3(CH3CN)3(3-L)2][BF4]328.9228.052.712.6014.2713.841579s, 1569s, 1538s, 1435s, 1137s, 1087s,1036s, 1006m, 850m, 833m, 758m, 658m,649m, 583w, 538m, 525m12 2¡Û [Ag3(CH3CN)3(3-L)2][PF6]324.9024.882.331.9612.2912.601580s, 1570s, 1540s, 1435s, 1385m, 1352m,850vs, 832s, 760m, 740w, 658w, 650w, 565sa Color, yield and thermal behavior are described in the individual preparative procedures. b The peak assignment is as given for free 2,2-dimethyl-4,4-bipyrimidine:c Mostly medium and stronger IR frequencies reported, only.of one third of the bipyrimidine ligands at 231 C, possibly toyield 3.(-2,2-Dimethyl-4,4-bipyrimidine)dinitratocopper(II),¡Û1[Cu(NO3)2(-L)] 5.A solution of L (186 mg, 1.00 mmol) inCHCl3 (10 ml) was overlayered in a test-tube rst with CH2Cl2(2 ml) and then with a solution of Cu(NO3)26H2O (296 mg,1.00 mmol) in ethanol (5 ml). Within several hours the contactarea of the solutions turned red and after 2 weeks deep bluevioletcrystals began to grow within the red zone. After 2months a homogeneous red solution had formed with crystalsat the bottom of the test-tube. They were collected, washedwith a little ice-cold ethanol and air-dried.The yield could beincreased by concentrating the mother-liquor; 307 mg (82%).A thermogravimetric analysis showed an exothermic decompositionat 198 C to volatile components and CuO. The crystalcomposition did not depend on the concentrations employedor on the metal-to-ligand ratio. The latter was varied from 10: 1to 1 : 10.[Cu3I3(L)] 6. A 50 ml Schlenk tube was lled with water-freeCuI (571 mg, 3.0 mmol), followed by dry CH3CN (10 ml).Thesolution was then carefully overlayered with a solution of L(186 mg, 1.0 mmol) in dry CH3CN (20 ml). A ne black precipitateoccurred immediately. Within 3 d CuI was completelytransformed into strongly intergrown needle-shaped crystals.These were separated by ltration, washed with a few mlof acetonitrile and dried for 5 min in vacuum. Yield 705 mg(93%). A thermogravimetric analysis indicated the loss of Land the transformation into CuI at 336 C. When stored inthe mother-liquor, the black needles of complex 6 started toturn into bright red cubes of 7 within a few days.When theblack-red needles of 6 were heated in dimethyl sulfoxide, thecompound could be transformed into the co-ordination polymer8.(-2,2-Dimethyl-4,4-bipyrimidine)di(3-iodo)dicopper(I),¡Û 2[Cu2(3-I)2(-L)] 7. Copper() iodide (391 mg, 2.0 mmol) inCH3CN (10 ml) and a solution of L (186 mg, 1.0 mmol) inacetonitrile (20 ml) gave a black precipitate of complex 6 whichwas kept under the mother-liquor. After 2 to 3 weeks cubic redcrystals had grown which were separated from some of theremaining black precipitate by suspension and decantation.The product was dried for 5 min in vacuum.Yield 442 mg(78%). A thermogravimetric analysis showed the loss of onethird of the bipyrimidine ligands at 273 C, possibly yielding 6,followed by the complete loss of L at 334 C to give CuI.(2,2-Dimethyl-4,4-bipyrimidine)(-iodo)copper(I), ¡Û1[Cu-(-I)(L)] 8. Compound 6 (126 mg, 0.167 mmol) was dissolved inDMSO (100 ml) upon heating.Starting from 120 C the solutionwas cooled to room temperature at a rate of 2 C h1. Thedeep red, metallic lustrous needles were ltered o, washed witha little DMSO and dried in vacuum for 15 min. Yield 51 mg(81%). The needles appeared black metallic under the microscopewhen their thickness extended over 20 m.Regardless of the metal-to-ligand ratio, complex 6 wasalways obtained as the rst phase in the synthesis of 7 and 8.J.Chem. Soc., Dalton Trans., 1999, 3121.3131 3129 Table 2 Crystal data for compounds 4, 5, 7.12 4 5 7 8 9 10 11 12 Formula MT /K Crystal system Space group a/A b/A c/A ¥á/ ¥â/ ¥ã/ V/A3 Z D/g cm3 ¥ì/cm1 Measured re.ections Unique re.ections (Rint) Parameters re.ned Maximum, minimum .¥ñ/e A3 R1; wR2 [I > 2¥ò(I )] all re.ections C11H15Cl2N5NiO3 406.88 200(2) Monoclinic P21/n 13.506(3) 7.6665(10) 15.374(3) 100.69(2) 1564.3(5) 4 1.668 15.98 9078 3507 (0.0685) 248 0.0890; 0.968 0.0409; 0.0985 0.0552; 0.1042 C10H10CuN6O6 373.78 294(2) Triclinic P1. 6.111(8) 7.445(8) 9.335(8) 82.29(8) 83.07(8) 67.23(8) 387.0(7) 1.604 14.51 2009 1759 (0.0677) 107 0.943; 1.302 0.0726; 0.1784 0.0899; 0.2087 C10H10Cu2I2N4 567.10 200(2) Triclinic P1. 7.751(1) 10.5927(8) 4.2001(7) 91.81(2) 101.99(2) 88.54(2) 337.1(1) 2.794 77.19 3494 1492 (0.0343) 97 1.601; 0.714 0.0231; 0.0544 0.0265; 0.0553 C10H10CuIN4 376.66 293(2) Orthorhombic Fdd2 33.436(2) 35.0765(18) 4.0449(2) 4747.8(4) 16 2.108 44.18 10123 2811 (0.0405) 163 1.179; 0.327 0.0215; 0.0494 0.0239; 0.0501 C10H10AgNO3 356.09 200(3) Orthorhombic Pbcn 3.822(5) 15.405(5) 19.739(5) 1162.2(16) 4 2.035 18.10 5798 1336 (0.0381) 108 0.485; 0.709 0.0236; 0.0559 0.0345; 0.0586 C20H20Ag3N11O9 882.08 293(2) Orthorhombic Pcna 7.068(8) 21.81(2) 19.41(2) 2992(5) 4 1.958 20.09 3483 3453 212 1.295; 1.759 0.0973; 0.2336 0.2154; 0.3477 C26H29Ag3B3F12N11 1079.60 200(3) Monoclinic I2/a 17.1706(11) 10.8107(7) 21.3415(13) 109.697(7) 3729.7(4) 4 1.923 16.58 21256 4440 (0.0412) 275 0.963; 0.670 0.0291; 0.0755 0.0325; 0.0771 C26H29Ag3F18N11P3 1254.08 200(3) Monoclinic I2/a 17.3846(11) 10.9689(10) 21.9716(14) 105.148(7) 4044.2(5) 4 2.060 16.78 17161 4568 (0.0524) 284 1.001; 0.964 0.0364; 0.0974 0.0403; 0.09963130 J.Chem. Soc., Dalton Trans., 1999, 3121–3131 An X-ray powder di.raction study showed that 6 was di.erent from both 7 and 8. (-2,2-Dimethyl-4,4-bipyrimidine)(-nitrato)silver(I), 8 2[Ag(-NO3)(-L)] 9.A solution of L (186 mg, 1.0 mmol) in CHCl3 (10 ml) was overlayered with a solution of AgNO3 (170 mg, 1.0 mmol) in water (10 ml) in a crystallization dish. The reaction mixture was stored in the dark. Within a few hours colorless crystals started to grow at the phase boundary. When the organic chloroform solvent had evaporated the crystals were collected by .ltration, washed with some water and ethanol and dried in air. They were of needle shape but intergrown to bushels and had a mother-of-pearl like shine. Yield 324 mg (91%).A thermogravimetric analysis showed loss of 50% of the bipyrimidine ligands at 239 C and at 273 C an exothermic decomposition to volatile products and Ag2O. Bis(3-2,2-dimethyl-4,4-bipyrimidine)tri(3-nitrato)- trisilver(I)], 8 3Ag3(3-NO3)3(3-L)2] 10. A solution of L (112 mg, 0.60 mmol) in CHCl3 (6 ml) was overlayered in a test tube .rst with CH2Cl2 (2 ml) and then with a solution of AgNO3 (225 mg, 1.50 mmol) in water (6 ml).Within 12 h yellow crystals formed at the phase boundary. The crystals were collected, washed with a few milliliters of water and air-dried. Preparation and storage was mostly carried out in the dark. Yield 217 mg (88%). A thermogravimetric analysis indicated decomposition to Ag2O and volatile oxidation products at 272 C. Tris(acetonitrile)bis(3-2,2-dimethyl-4,4-bipyrimidine)- trisilver tris(tetra.uoroborate), 8 2[Ag3(CH3CN)3(3-L)2][BF4]3 11.A solution of L (9 mg, 0.05 mmol) in CH2Cl2 (5 ml) was overlayered in a test-tube with a solution of AgBF4 (10 mg, 0.05 mmol) in ethanol (5 ml). After 8 d at room temperature the yellow crystals formed were collected, washed with water and dichloromethane and dried under vacuum. Yield 9 mg (47%). According to IR spectroscopy the same complex was also obtained from a metal-to-ligand ratio of 2 : 1. Tris(acetonitrile)bis(3-2,2-dimethyl-4,4-bipyrimidine)- trisilver tris(hexa.uorophosphate), 8 2[Ag3(CH3CN)3(3-L)2][PF6]3 12.A solution of L (9 mg, 0.05 mmol) in CH2Cl2 (5 ml) was overlayered in a test-tube with a solution of AgPF6 (27 mg, 0.11 mmol) in ethanol (5 ml). After 8 d at room temperature the yellow crystals formed were collected, washed with water and dichloromethane and dried under vacuum. Yield 13 mg (50%). Crystal structure determinations Data were collected with Mo-Ka radiation (. = 0.71073 Å) and the use of a graphite monochromator.Structure solution was performed by direct methods using SHELXS 97 or SHELXS 8637 except for complexes 7 and 8 where SIR 97 38 was employed. Re.nement: full-matrix least squares on F2 [SHELXL 97 (version 97-2), SHELXS 93];37 all non-hydrogen positions found and re.ned with anisotropic thermal parameters. Crystal data are listed in Table 2. Compound 8 was re.ned as a racemic twin. Graphics were computed with ORTEP 3 for Windows.39 CCDC reference number 186/1560.See http://www.rsc.org/suppdata/dt/1999/3121/ for crystallographic .les in .cif format. Acknowledgements This work was supported by the Alexander von Humboldt Foundation (fellowship for H.-P. Wu), the Fonds der Chemischen Industrie, the Deutsche Forschungsgemeinschaft (grant Ja466/10-1), and the Graduate College “Unpaired Electrons” at the University Freiburg. We thank one of the referees for his critical and helpful comments. References 1 A. E. Martell and R. D.Hancock, Metal Complexes in Aqueous Solutions, Plenum, New York, 1996; A. v. Zelewsky, Stereochemistry of Co-ordination Compounds, Wiley, Chicester, 1996; J. M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1995; E. C. Constable, Prog. Inorg. Chem., 1994, 42, 67. 2 C. V. K. Sharma and R. D. Rogers, Chem. Commun., 1999, 83; M.-L. Tong, B.-H. Ye, J.-W. Cai, X.-M. Chen and S. W. Ng, Inorg. Chem., 1998, 37, 2645; L. R. MacGillivray, R. H. Groeneman and J. L. Atwood, J. Am. Chem.Soc., 1998, 120, 2676; B. Olenyuk, A. Fechtenkötter and P. J. Stang, J. Chem. Soc., Dalton Trans., 1998, 1707; A. J. Blake, S. J. Hill, P. Hubberstey and W.-S. Li, J. Chem. Soc., Dalton Trans., 1998, 909; M.-L. Tong, X.-M. Chen, X.-L. Yu and T. C. W. Mak, J. Chem. Soc., Dalton Trans., 1998, 5; K. N. Power, T. L. Hennigar and M. J. Zaworotko, New. J. Chem., 1998, 177; P. Stang and B. Olenyuk, Acc. Chem. Res., 1997, 30, 502; O. M. Yaghi, H. Li and T. L. Groy, Inorg. Chem., 1997, 36, 4292; J.Lu, T. Paliwala, S. C. Lim, C. Yu, T. Niu and A. J. Jacobson, Inorg. Chem., 1997, 36, 923; L. Carlucci, G. Ciani, D. M. Proserpio and A. Sironi, J. Chem. Soc., Dalton Trans., 1997, 1801; P. Lossier and M. J. Zaworotko, Angew. Chem., Int. Ed. Engl., 1996, 35, 2779; O. M. Yaghi and H. Li, J. Am. Chem. Soc., 1996, 118, 295; C. A. Hunter, Angew. Chem., Int. Ed. Engl., 1995, 34, 1079. 3 C. Janiak, Angew. Chem., Int. Ed. Engl., 1997, 36, 1431; O. M. Yaghi, H. Li, C.Davis, D. Richardson and T. L. Groy, Acc. Chem. Res., 1998, 31, 474. 4 G. de Munno, G. Bruno, F. Nicoló, M. Julve and J. A. Real, Acta Crystallogr., Sect. C, 1993, 49, 457 and refs. therein. 5 N.-E. Ghermani, N. Bouhmaida, C. Lecomte, A.-L. Papet and A. Marsura, J. Phys. Chem., 1994, 98, 6287. 6 F. Bodar-Houillon, T. Humbert, A. Marsura, J.-B. Regnouf de Vains, O. Dusausoy, N. Bouhmaida, N. E. Ghermani and Y. Dusausoy, Inorg. Chem., 1995, 34, 5205. 7 L. Tei, V. Lippolis, A. J.Blake, P. A. Cooke and M. Schröder, Chem. Commun., 1998, 2633; F.-Q. Liu and T. D. Tilley, Chem. Commun., 1998, 103; H. Li, C. E. Davis, T. L. Groy, D. G. Kelley and O. M. Yaghi, J. Am. Chem. Soc., 1998, 120, 2186; C. J. Kepert and M. J. Rosseinsky, Chem. Commun., 1998, 31; B. F. Hoskins, R. Robson and D. A. Slizys, Angew. Chem., Int. Ed. Engl., 1997, 36, 2752; R. W. Saalfrank, O. Struck, M. G. Davidson and R. Snaith, Chem. Ber., 1994, 127, 2489. 8 H.-P. Wu, C. Janiak, L. Uehlin, P.Klüfers and H. Piotrowski, Chem. Commun., 1998, 2637. 9 H.-P. Wu, C. Janiak, G. Rheinwald and H. Lang, J. Chem. Soc., Dalton Trans., 1999, 183. 10 See for example, C. Janiak, T. G. Scharmann, J. C. Green, R. P. G. Parkin, M. J. Kolm, E. Riedel, W. Mickler, J. Elguero, R. M. Claramunt and D. Sanz, Chem. Eur. J., 1996, 2, 992; C. Janiak, T. G. Scharmann, W. Günther, F. Girgsdies, H. Hemling, W. Hinrichs and D. Lentz, Chem. Eur. J., 1995, L, 637. 11 J. S. Fleming, K. L.V. Mann, C.-A. Carraz, J. C. Je.ery, E. Psillakis, J. A. McCleverty and M. D. Ward, Angew. Chem., Int. Ed., 1998, 37, 1279. 12 P. N. W. Baxter, J.-M. Lehn and K. Rissanen, Chem. Commun., 1997, 1323; D. P. Funeriu, J.-M. Lehn, G. Baum and D. Fenske, Chem. Eur. J., 1997, 3, 99. 13 F. E.enberger, Chem. Ber., 1965, 98, 2260. 14 J.-B. Regnouf de Vains, J.-M. Lehn, N. E. Ghermani, O. Dusausoy, Y. Dusausoy, A.-L. Papet, A. Marsura, P. Friant and J. L. Rival, New J. Chem., 1994, 18, 701. 15 HYPERCHEM, Version 4.5. AM1, J. J. P. Stewart, Reviews in Computational Chem., eds. K. B. Kipkowitz and D. B. Boyd, VCH, New York, 1990, ch. 2, p. 45.; PM3, J. J. P. Stewart, J. Comput. Chem., 1989, 10, 209, 221. 16 C. Janiak and H. Hemling, J. Chem. Soc., Dalton Trans., 1994, 2947; C. Janiak, Chem. Commun., 1994, 545. 17 B. Hasenknopf, J.-M. Lehn, N. Boumediene, E. Leize and A. Van Dorsselaer, Angew. Chem., Int. Ed., 1998, 37, 3268. 18 (a) J. A. R. Navarro, J. M. Salas, M.A. Romero and R. Faure, J. Chem. Soc., Dalton Trans., 1998, 901; (b) M. A. Withersby, A. J. Blake, N. R. Champness, P. Hubberstey, W.-S. Li and M. Schröder, Angew. Chem., Int. Ed. Engl., 1997, 36, 2327. 19 M. C. Muñoz, M. Julve, F. Lloret, J. Faus and M. Andruh, J. Chem. Soc., Dalton Trans., 1998, 3125. 20 S. Lopez, M. Kahraman, M. Harmata and S. W. Keller, Inorg. Chem., 1997, 36, 6138. 21 G. V. Long, S. E. Boyd, M. M. Harding, I. E. Buys and T. W. Hambley, J. Chem. Soc., Dalton Trans., 1993, 3175. 22 I. M. Müller, T. Röttgers and W. S. Sheldrick, Chem. Commun., 1998, 823; Y. Suenaga, S. G. Yan, L. P. Wu, I. Ino, T. Kuroda-Sowa, M. Maekawa and M. Munakata, J. Chem. Soc., Dalton Trans.,J. Chem. Soc., Dalton Trans., 1999, 3121–3131 3131 1998, 1121; O. M. Yaghi and H. Li, J. Am. Chem. Soc., 1995, 117, 10401. 23 E. Eitel, D. Oelkrug, W. Hiller and J. Strähle, Z. Naturforsch., Teil B, 1980, 35, 1247. 24 N. P. Rath, J. L. Maxwell and E. M. Holt, J. Chem.Soc., Dalton Trans., 1986, 2449. 25 B. W. Skelton, A. F. Water and A. H. White, Aust. J. Chem., 1991, 44, 1207. 26 P. C. Healy, C. Pakawatchai, C. L. Raston, B. W. Skelton and A. H. White, J. Chem. Soc., Dalton Trans., 1983, 1905. 27 J. A. Campbell, C. L. Raston and A. H. White, Aust. J. Chem., 1977, 30, 1937. 28 L. M. Engelhardt, P. C. Healy, J. D. Kildea and A. H. White, Aust. J. Chem., 1989, 42, 107. 29 N. P. Rath, E. M. Holt and K. Tanimura, J. Chem. Soc., Dalton Trans., 1986, 2303. 30 P. C. Healy, C. Pakawatchai and A. H. White, J. Chem. Soc., Dalton Trans., 1985, 2531. 31 M. Munakata, T. Kuroda-Sowa, M. Maekawa, A. Honda and S. Kitagawa, J. Chem. Soc., Dalton Trans., 1994, 2771. 32 C. B. Aakeröy and A. M. Beatty, Chem. Commun., 1998, 1067; C. Kaes, M. W. Hosseini, C. E. F. Rickard, B. W. Skelton and A. H. White, Angew. Chem., Int. Ed. Engl., 1998, 37, 920; P. K. Bowyer, K. A. Porter, A. D. Rae, A. C. Willis and S. B. Wild, Chem. Commun., 1998, 1153; K. A. Hirsch, S. R. Wilson and J. S. Moore, Chem. Commun., 1998, 13; Chem. Eur. J., 1997, 3, 765; Inorg. Chem., 1997, 36, 2960; L. Carlucci, G. Ciani, D. W. v. Gudenberg, D. M. Proserpio and A. Sironi, Chem. Commun., 1997, 631; A. J. Blake, N. R. Champness, A. Khlobystov, D. A. Lemenovskii, W.-S. Li and M. Schröder, Chem. Commun., 1997, 2027; L. Carlucci, G. Ciani, D. M. Proserpio and A. Sironi, Chem. Commun., 1996, 1393; D. Venkataraman, S. Lee, J. S. Moore, P. Zhang, K. A. Hirsch, G. B. Gardner, A. C. Covey and C. L. Prentice, Chem. Mater., 1996, 8, 2030; C. Janiak, T. G. Scharman, P. Albrecht, F. Marlow and R. Macdonald, J. Am. Chem. Soc., 1996, 118, 6307; L. Carlucci, G. Ciani, D. M. Proserpio and A. Sironi, J. Am. Chem. Soc., 1995, 117, 4562. 33 F. Milani-Nejad and H. Stidham, Spectrochim. Acta, Part A, 1975, 31, 1433; R. Foglizzo and A. Novak, Spectrochim. Acta, Part A, 1970, 26, 2281. 34 E. Allenstein, W. Podszun, P. Kiemle, H.-J. Mauk, E. Schlipf and J. Weidlein, Spectrochim. Acta, Part A, 1976, 32, 777. 35 D. Thornton and G. Watkins, J. Coord. Chem., 1992, 24, 299. 36 G. Baum, E. C. Constable, D. Fenske, C. E. Housecroft and T. Kulke, Chem. Commun., 1998, 2659. 37 G. M. Sheldrick, SHELXL 97 and SHELXS 93, Programs for Crystal Structure Re.nement, University of Göttingen, 1997 and 1993, respectively; SHELXS 97 and SHELXS 86, Programs for Crystal Structure Solution, University of Göttingen, 1997 and 1986, respectively. 38 G. Cascarano, A. Altomare, C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, D. Siliqi, M. C. Burla, G. Polidori and M. Camalli, Acta Crystallogr., Sect. A, 1996, 52, C79. 39 M. N. Burnett and C. K. Johnson, ORTEP III, Oak Ridge Thermal Ellipsoid Plot Program for Crystal Structure Illustrations, Report ORNL-6895, Oak Ridge National Laboratory, Oak Ridge, TN, 1996; L. J. Farrugia, ORTEP 3 for Windows, version 1.0.1ß, J. Appl. Crystallogr., 1997, 30, 565. Paper 9/04829D

ISSN:1477-9226    

DOI:10.1039/a904829d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 3133–3136 3133 A new route to tris(pyrazolyl)borate ligands and new structural variations in TlTp complexes † Christoph Janiak,*a Lothar Braun b and Frank Girgsdies b a Institut für Anorganische und Analytische Chemie, Universität Freiburg, Albertstr. 21, D-79104 Freiburg, Germany. E-mail: janiak@uni-freiburg.de b Institut für Anorganische und Analytische Chemie, Technische Universität Berlin, Strasse des 17. Juni 135, D-10623 Berlin, Germany Received 22nd March 1999, Accepted 7th July 1999 Tris(pyrazolyl)borate ligands were synthesized by a new route from MeBBr2 and pyrazole derivatives under very mild conditions at room temperature to give TlL complexes.For L = [MeB(3,5-Me2pz)3] a bridging co-ordination of the ligand is found, interpreted as sterically enforced upon comparison with the structures for L = [MeB(3-Mepz)3] and [HB(3,5-Me2pz)3]. Introduction The tris(pyrazolyl)borate (Tp) ligand together with various substituted forms has developed into one of the most versatile tripodal auxiliary ligands in (bio)inorganic co-ordination chemistry. 1 The standard or so-called Tro.menko method of synthesis is the reaction of substituted pyrazoles with KBH4 at elevated temperatures above 190 C, eqn. (1) (Hpz = pyrazole KBH4 3 Hpz 190 C K[HB(pz)3] 3 H2 (1) or derivative).2 However, this route fails in the case of thermally sensitive pyrazole derivatives.3 A few other syntheses to tris(pyrazolyl)borates have been described.They start from monoorganylboron compounds and are given in eqns. (2) 4,5 (R = iPr, nBu, Ph or 4-BrC6H4) and (3) (pyrazole only).4 The RB(OH)2 Na(pz) 2 Hpz 185–220 C Na[RB(pz)3] 2 H2O (2) PhBCl2 4 pzH 55 C [H2pz][PhB(pz)3] 2 HCl (3) route outlined in eqn. (2) also requires high temperatures. Neither of the synthetic schemes in eqns. (2) and (3) appears to have found widespread applications for the preparation of Tp ligands. In part, this may be due to the low yield of the products and the noted di.culty to crystallize the boron substituted Tp– metal complexes.4,5 Here, we report the results of our search for yet another route to Tp ligands.Results and discussion Following a procedure to ferrocene-based Tp ligands by Wagner and co-workers,6 we found that dibromo(methyl)borane, MeBBr2,7 reacts at room temperature with pyrazole derivatives in the presence of NEt3 and thallium ethoxide to form methyltris( pyrazolyl)boratothallium complexes, eqn.(4). The reaction is demonstrated with pyrazole, 3,5-dimethylpyrazole and 3- † Supplementary data available: IR data. For direct electronic access see http://www.rsc.org/suppdata/dt/1999/3133/, otherwise available from BLDSC (No. SUP 57609, 2 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton). MeBBr2 3 Hpz 2 NEt3 TlOEt toluene, RT Tl[MeB(pz)3] 2 [NHEt3]Br EtOH (4) 1 pz = pz 2 pz = 3,5-Me2pz 3 pz = 3-Mepz methylpyrazole and can easily be extended to other pyrazole derivatives.The boratothallium complexes are obtained in good to high yield. Often, TlTp compounds are valued as a means of isolation and characterization of a (new) Tp ligand.8 Moreover, just like cyclopentadienylthallium,9 TlTp is also a milder (less reducing) and mostly more stable ligand transfer reagent in place of Tp alkali-metal salts. Hence, the initial KTp salt from the Tro.menko route, eqn.(1), is occasionally transformed into the TlTp complex for further reactions. Despite its toxicity, TlTp is a common reagent for Tp-ligand transfer and ligand characterization in the case of the more sterically demanding or so-called “second-generation Tro.menko ligands”.8 This is the basis of interest in TlTp structural chemistry. In addition to the usual spectroscopic methods, identi.cation of compounds 2 and 3 was also based on X-ray crystallography.There is interest in thallium(.) structures because of their diversity and theoretical aspects of the in/active lone pair of electrons. 10,11 So far, TlTp compounds have shown little variation. All, but four, are clearly built up from molecular units with a trihapto, C3-symmetrical metal co-ordination. The exceptions are Tl[HB(3-C3H5pz)3] (C3H5 = cyclopropyl)],10 Tl[HB{3-(4- MeC6H4)pz}3],12 Tl[HB{2,4-(MeO)2pz}3] 13 and Tl[HB(pz)3].14 The .rst complex forms a stable tetramer with a perfect tetrahedron of Tl atoms.The next two are “dimeric” with Tl Tl contacts of 3.86 and 3.995 Å, respectively. The last has metal– ligand strands based on eletrostatic thallium–pyrazolyl p interactions. Despite these di.erences in the molecular packing, every TlTp structure hitherto reported shows the expected threefold co-ordination of the Tp ligand to the metal. The structure of compound 2 represents a remarkable exception to the above generalization.Fig. 1 illustrates that the ligand [MeB(3,5-Me2pz)3] bridges between two thallium atoms. One thallium atom is co-ordinated by two pyrazolyl rings. The third pyrazolyl ring binds the adjacent symmetry related thallium center in a monodentate fashion. The plane of this ring assumes an angle of 81.3(4) to the B–Me axis. The bridging3134 J. Chem. Soc., Dalton Trans., 1999, 3133¡V3136action of the [MeB(3,5-Me2pz)3] ligand leads to a 21-helicoidalchain. Why is there no triphapto metal co-ordination? A visualinspection of the structure suggests that the space requirementsof the methyl groups on the boron and the pyrazolyl-C5 atomsdo not allow the simultaneous trihapto co-ordination of allthree rings to one thallium center.The thallium, boron and(B)carbon atom do not lie on a straight line but form an angleof 159.6. Both, the thallium atom as well as the boron-bondedmethyl group appear to be moved away from the imaginaryaxis. Only one of the chelating pyrazolyl rings [N(3)¡VN(4)]coincides with the B¡VMe axis, the other is tilted by 27.8(3).This interpretation of a methyl¡Vmethyl repulsion was testedwith a structural investigation of Tl[MeB(3-Mepz)3] 3 and ofTl[HB(3,5-Me2pz)3] 4.The prototypical complex 4 with theTp* ligand has been used as a ligand transfer reagent,15 but hasapparently never been characterized. Both these complexes lacka methyl group either on the pz 5 position or on the boronatom, so that there is no repulsive methyl¡Vmethyl interaction.Hence, the expected trihapto, C3-symmetrical thallium coordinationis found in their molecular structures (Figs. 2 and 3).Selected bond distances and angles are collected in Table 1.Fig. 1 (a) Repeat unit of 1¡ÛTl[MeB(3,5-Me2pz)3] 2 and (b) section of the21-helicoidal co-ordination polymer. In (b) the hydrogen atoms havebeen omitted for clarity. Symmetry relation: _2 0.5 x, 0.5 y,0.5 z.Fig. 2 Molecular structure of Tl[MeB(3-Mepz)3] 3.The Tl¡VN distances in 2 are longer than usual for TlTpcompounds.8It may be noted that the NMR spectra of the polymericcompound 2 are rather simple, suggesting a C3 symmetricalstructure in solution. They show that all three rings are equivalentin solution, since only one signal is observed for each typeof proton or carbon atom.This is the typical NMR pattern forTlTp complexes and is also seen in the spectra of 1, 3 and 4. Ofcourse, a C3v symmetrical structure of a thallium¡Vligand complex2 in solution would be a contradiction to the above argumenton the solid-state arrangement. Here we note that theTlTp complexes are seldom retained intact in solution.Usually,very loose or solvent-separated cation¡Vanion pairs are formed.If the TlTp complex is present in solution Tl¡VH and Tl¡VC couplingwould be observed. Both natural thallium isotopes 205Tland 203Tl have spin . The absence of such coupling to thalliumis indicative of either a predominantly ionic thallium¡Vringinteraction or fast intermolecular exchange processes.8 Then,the solution NMR spectra correspond to the more-or-less freeand anionic Tp ligand.Even in view of the steric interactions inthe free Tp ligand of 2, a C3 symmetrical structure can still beassumed in solution by having the three ring planes canted allin the same direction with respect to the B¡VMe bond. Fig. 4presents a space-lling drawing of [MeB(3,5-Me2pz)3] obtainedfrom a molecular mechanics optimization.ExperimentalThe NMR spectra were collected on a Bruker ARX200 spectrometer(200.1 MHz for 1H, 50.3 MHz for 13C) and calibratedagainst the solvent signal (CDCl3, 1H £_ 7.26, 13C £_ 77.0), IRspectra on a Nicolet-Magna Spectrometer 750 as KBr disks(only major peaks are listed) and mass spectra with a VarianMAT 311 A/AMD spectrometer and electron-impact (EI)Fig. 3 Molecular structure of Tl[HB(3,5-Me2pz)3] 4.Table 1 Selected bond lengths () and angles () in compounds 2¡V42 3 4Tl¡VN(2)Tl¡VN(4)Tl¡VN(6)B¡VCB¡VNN(2)¡VTl¡VN(4)N(2)¡VTl¡VN(6)N(4)¡VTl¡VN(6)N¡VB¡VN2.638(3)2.760(4)2.876(4) a1.602(6)1.557(6)1.577(6)68.9(1)96.5(1) a96.6(1) a106.5(3)108.5(3)2.547(4)2.504(4)2.499(4)1.587(7)1.555(7)1.654(7)75.4(1)74.4(1)73.9(1)107.6(4)109.4(4)2.534(6)2.499(6)2.515(6)¡X1.555(10)1.566(10)75.7(2)74.4(2)73.9(2)109.2(6)111.2(6)a Symmetry related atom generated by the transformation 0.5 x,0.5 y, 0.5 z.J.Chem. Soc., Dalton Trans., 1999, 3133.3136 3135 Table 2 Crystal data for compounds 2.4 2 3 4 Formula M Crystal system Space group a/A b/A c/A ¥â/ V/A3 Z Dc/g cm3 F(000) ¥ì/cm1 Measured re.ections Unique re.ections (Rint) Observed re.ections [I > 2¥ò(I)] Parameters re.ned .¥ñ a/e A3 R1, wR2 [I > 2¥ò(I)] (all re.ections) C16H24BN6Tl 515.59 Monoclinic P21/n 11.4656(1) 9.9676(1) 17.0414(3) 96.125(1) 1936.45(4) 4 1.769 992 83.50 14261 4432 (0.0604) 3304 225 0.651, 1.211 0.0312, 0.0566 0.0560, 0.0625 C13H18BN6Tl 473.51 Monoclinic P21/n 8.1756(1) 14.1921(1) 13.9746(1) 96.771(1) 1610.15(3) 4 1.953 896 100.3 12008 3685 (0.0549) 2800 194 0.661, 0.974 0.0311, 0.0617 0.0502, 0.0688 C15H22BN6Tl 501.57 Monoclinic C2/c 30.6275(5) 8.6168(1) 15.6585(2) 119.812(1) 3585.57(9) 8 1.858 1920 90.16 13372 4110 (0.0758) 2909 214 1.350, 2.862 0.0479, 0.0948 0.0790, 0.1060 a Largest di.erence peak and hole.sample ionization at 70 eV. Elemental analyses were done with a Perkin-Elmer 2400 Series II CHNS/O Analyzer. The reactions were carried out with Schlenk techniques using .ame-dried glassware and argon as inert gas.The solvent CH2Cl2 was dried by re.uxing over CaH2 followed by distillation and storage under inert gas; MeBBr2 was prepared according to ref. 7 and K[HB(3,5-Me2pz)3] according to ref. 2. Pyrazoles and TlOEt were purchased from Aldrich. Preparations [Methyltris(pyrazol-1-yl)borato]thallium(I), Tl[MeB(pz)3] 1. The compound MeBBr2 (1.67 g, 9.0 mmol) was added to a solution of pyrazole (1.83 g, 20.9 mmol) in toluene (20 ml).After stirring for 1 h, NEt3 (1.82 g, 18.0 mmol) was added and stirring continued for 12 h. A white precipitate was removed by .ltration. The .ltrate was cooled to 78 C and TlOEt (2.24 g, 9.0 mmol) added. The mixture was stirred for 4 h at room temperature, the solvent then removed in vacuum and the residue Fig. 4 Molecular mechanics optimized structure of free [MeB(3,5- Me2pz)3] in complex 2 viewed along the Me.B bond; dark spheres are the nitrogen atoms.extracted in a Soxhlet apparatus with CH2Cl2. Removal of CH2Cl2 in vacuum left the product as a white powder (yield 1.82 g, 47%). 1H NMR (CDCl3): ¥ä 1.09 (s, 3 H, B-CH3), 6.30 (s, 3 H, H4), 7.58 (s, 3 H, H5) and 7.82 (s, 3 H, H3). 13C NMR (CDCl3): ¥ä 4.16 (B-CH3), 103.84 (C4), 133.10 (C5) and 138.76 (C3). MS (65 C): m/z 432 (2) [M]; 417 (43), [M Me]; 365 (16), [M pz]; 350 (2), [M Me pz]; and 205 (100%), [Tl]. Calc. for C10H12BN6Tl: C, 27.84; H, 2.80; N, 19.48.Found: C, 28.01; H, 2.57; N, 19.38%. [Tris(3,5-dimethylpyrazol-1-yl)methylborato]thallium(I), Tl[MeB(3,5-Me2pz)3] 2. The compound MeBBr2 (1.84 g, 9.9 mmol) was added to a solution of 3,5-dimethylpyrazole (2.86 g, 29.7 mmol) in toluene (40 ml). After stirring for 1 h, NEt3 (2.00 g, 19.8 mmol) was added and stirring continued for 12 h. A white precipitate was removed by .ltration, TlOEt (1.70 g, 6.8 mmol) added and the reaction mixture stirred for 4 h. The solvent was removed in vacuum and the residue extracted in a Soxhlet apparatus with CH2Cl2.Removal of CH2Cl2 in vacuum left the product as a white powder (yield 2.71 g, 53%). A crystalline sample was obtained from CH2Cl2, mp 232 C. 1H NMR (CDCl3): ¥ä 0.83 (s, 3 H, B-CH3), 2.17 (s, 9 H, pz 5-CH3), 2.23 (s, 9 H, pz 3-CH3) and 5.87 (s, 3 H, H4). 13C NMR (CDCl3): ¥ä 1.01 (B-CH3), 13.36 (pz 5-CH3), 14.14 (pz 3-CH3), 107.56 (C4), 146.15 (C5) and 147.62 (C3). MS (119 C): m/z 516 (14), [M]; 501 (56), [M Me]; 421 (100), [M Me2pz]; and 205 (42%), [Tl].IR (strong signals only): 3061, 2738, 2520, 2420, 2359, 2290, 2236, 2129, 1105, 1057, 766, 760, 748, 703, 671, 655, 605, 593, 590, 513, 452 and 440 cm1. Calc. for C16H24BN6Tl: C, 37.27; H, 4.69; N, 16.30. Found: C, 36.91; H, 3.96; N, 16.27%. [Methyltris(3-methylpyrazol-1-yl)borato]thallium(I), Tl[MeB(3-Mepz)3] 3. The compound MeBBr2 (1.27 g, 6.8 mmol) was added to a solution of 3-methylpyrazole (1.68 g, 20.5 mmol) in toluene (20 ml).After stirring for 1 h, NEt3 (1.38 g, 13.6 mmol) was added and stirring continued for 12 h. The white precipitate was removed by .ltration, TlOEt (1.70 g, 6.8 mmol) added and the reaction mixture stirred for 12 h. The solvent was removed to leave the product as a white powder (yield 2.95 g, 91%). A crystalline sample was obtained from CH2Cl2. 1H NMR (CDCl3): ¥ä 0.97 (s, 3 H, B-CH3), 2.43 (s, 9 H, pz CH3), 5.98 (s, 3 H, H4) and 7.62 (s, 3 H, H5). 13C NMR (CDCl3): ¥ä 5.16 (B-CH3), 13.20 (pz CH3), 104.11 (C4), 133.61 (C5) and 148.25 (C3).MS (104 C): m/z 474 (4), [M]; 4593136 J. Chem. Soc., Dalton Trans., 1999, 3133–3136 (100), [M Me]; 393 (56), [M Mepz]; 295 (28), [M Me 2(Mepz) 2H]; and 205 (48%), [Tl]. IR (strong signals only): 3141, 2855, 1094, 884, 855, 846, 835, 520, 253, 207, 197 and 192 cm1. Calc. for C13H18BN6Tl: C, 32.98; H, 3.83; N, 17.75. Found: C, 32.77; H, 3.43; N, 17.85%. [Tris(3,5-dimethylpyrazol-1-yl)hydroborato]thallium(I), Tl[HB(3,5-Me2pz)3] 4.The compound K[HB(3,5-Me2pz)3] (5.00 g, 14.9 mmol) and TlNO3 (3.96 g, 14.9 mmol) were stirred in CH2Cl2 (20 ml) for 6 h. The precipitate was separated by .ltration to give a clear, colorless solution. Removal of the solvent in vacuum left a white solid (4.58 g, 61.4%). Free 3,5-dimethylpyrazole which had been formed during the reaction has to be removed by sublimation. The residue was dissolved again in CH2Cl2 and .ltered.Cooling of the solution together with slow concentration gave clear crystals, Mp >240 C. 1H NMR (CDCl3): d 2.31 (s, 9 H, pz 3-CH3), 2.39 (s, 9 H, pz 5-CH3) and 5.76 (s, 3 H, H4). 13C NMR (CDCl3): d 12.85 (3,5-CH3), 105.15 (C4), 144.36 (C5) and 147.43 (C3). MS (125 C): m/z 502 (22), [M], 407 (100), [M 3,5Me2pz]; and 205 (55%), [Tl]. IR (strong peaks only): 2812, 2729, 2646, 2512 [.(BH)], 2372, 2356, 2237, 1139, 858, 811 and 456 cm1. Calc. for C15H22BN6Tl: C, 35.92; H, 4.42; N, 16.76; Found: C, 36.12; H, 4.49; N, 16.26%.Structure determinations Data were collected by the .-scan method with graphite monochromated Mo-Ka radiation (. = 0.71073) at 293 K on a Siemens Smart CCD di.ractometer. Structure solution was by direct methods (SHELXS 97) 16 and re.ned by full-matrix least squares on F2 (SHELXL 97);16 all non-hydrogen positions were found and re.ned with anisotropic thermal parameters. Crystal data are listed in Table 2.Graphics were obtained with ORTEP 3 for Windows.17 CCDC reference number 186/1561. See http://www.rsc.org/suppdata/dt/1999/3133/ for crystallographic .les in .cif format. Acknowledgements We appreciate the support by Deutsche Forschungemeinschaft (grant JA466/4-2 and 4-3), the Fonds der Chemischen Industrie and the graduate college “Metal catalysts” at the TU Berlin. References 1 S. Tro.menko, Chem. Rev., 1993, 93, 943; P. K. Byers, A. J. Canty and R. T. Honeyman, Adv.Organomet. Chem., 1992, 34, 1; N. Kitajima and W. B. Tolman, Prog. Inorg. Chem., 1995, 43, 419; G. Parkin, Adv. Inorg. Chem., 1995, 42, 291; D. L. Reger, Coord. Chem. Rev., 1996, 147, 571; M. Etienne, Coord. Chem. Rev., 1997, 156, 201. 2 S. Tro.menko, Inorg. Synth., 1970, 99, 12. 3 C. Janiak and L. Esser, Z. Naturforsch., Teil B, 1993, 48, 394. 4 S. Tro.menko, J. Am. Chem. Soc., 1967, 89, 6288. 5 D. L. White and J. W. Faller, J. Am. Chem. Soc., 1982, 104, 1548; D. L. Reger and M.E. Tarquini, Inorg. Chem., 1982, 21, 840. 6 F. Jäkle, K. Polborn and M. Wagner, Chem. Ber., 1996, 129, 603; F. F. de Biani, F. Jäkle, M. Spiegler, M. Wagner and P. Zanello, Inorg. Chem., 1997, 36, 2103. 7 T. E. Cole, R. Quinanilla, B. M. Smith and D. Hurst, Tetrahedron Lett., 1992, 33, 2761. 8 C. Janiak, Main Group Met. Chem., 1997, 21, 33; Coord. Chem. Rev., 1997, 163, 107. 9 C. Janiak, J. Prakt. Chem., 1998, 340, 181. 10 A. L. Rheingold, L. M. Liable-Sands and S. Tro.menko, Chem. Commun., 1997, 1691. 11 C. Janiak and R. Ho.mann, J. Am. Chem. Soc., 1990, 112, 5924. 12 G. Ferguson, M. C. Jennings, F. J. Lalor and C. Shanahan, Acta Crystallogr., Sect. C, 1991, 47, 2079. 13 S. Tro.menko, University of Delaware, personal communication, October 1998. 14 C. Janiak, S. Temizdemir and T. G. Scharmann, Z. Anorg. Allg. Chem., 1998, 624, 755. 15 A. Looney, R. Han, I. B. Gorrell, M. Cornebise, K. Yoon, G. Parkin and A. L. Rheingold, Organometallics, 1995, 14, 274; D. L. Reger, S. S. Mason and A. L. Rheingold, J. Am. Chem. Soc., 1993, 115, 10406; D. L. Reger and S. S. Mason, Organometallics, 1993, 12, 2600. 16 G. M. Sheldrick, SHELXS 97, SHELXL 97, Programs for Crystal Structure Analysis, University of Göttingen, 1997. 17 M. N. Burnett and C. K. Johnson, ORTEP III, Oak Ridge Thermal Ellipsoid Plot Program for Crystal Structure Illustrations, Report ORNL-6895, Oak Ridge National Laboratory, Oak Ridge, TN, 1996; L. J. Farrugia, ORTEP 3, windows version 1.0.4ß, J. Appl. Crystallogr., 1997, 30, 565. Paper 9/02264C © Copyright 1999 by the Royal Society of Chemistry

ISSN:1477-9226    

DOI:10.1039/a902264c

出版商:RSC    

年代:1999

数据来源: RSC