摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2065–2069 2065 Hole states in Eu0.92xPrxCa0.1BaSrCu3O72‰ studied by X-ray absorption spectroscopy Ponnusamy Nachimuthu,a Jin-Ming Chen,b Ru-Shi Liu,a Tryambak Bhimsen Waje,c Iyyani Kunjappu Gopalakrishnan d and Jatin`der Vir Yakhmid a Department of Chemistry, National Taiwan University, Taipei, Taiwan, ROC b Synchrotron Radiation Research Center, Hscinchu, Taiwan, ROC c Computer Division, Bhabha Atomic Research Center, Mumbai 400 085, India d Chemistry Division, Bhabha Atomic Research Center, Mumbai 400 085, India Received 5th October 1998, Accepted 15th April 1999 The hole states located on diVerent oxygen sites have been investigated in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d with increasing Pr31 substitution by high-resolution oxygen K-edge and copper L-edge X-ray absorption spectra.The results reveal that the hole states are depleted systematically by the substitution and in turn lead to a reduction in Tc.However, the rate at which the depletion occurs is lower in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d compared to that in Y1 2 xPrxBa2Cu3O7 2 d and Dy1 2 xPrxBa2Cu3O7 2 d. This is due to the partial substitution of Ba21 by Sr21 ions and R (rare earth element) by Ca21 ions in the present case. The depletion of hole states gives evidence in support of the hole-depletion models based on Pr 4f–O 2pp hybridization. The possible reasons for the anomalous behavior of Pr31 ions in RBa2Cu3O7 2 d are also discussed.Introduction The materials RBa2Cu3O7 2 d (R123) where R = rare earth element, are known to show a superconducting transition temperature (Tc) above 90 K except where R = Ce, Pr or Tb.1,2 Though isomorphic structure is not formed where R = Ce or Tb unlike the other rare earth based R123 compounds, 2 ideal orthorhombic structure similar to that of other superconducting R123 compounds is formed but no superconductivity is exhibited by PrBa2Cu3O7 2 d (Pr123).3–9 This anomalous behavior of PrBa2Cu3O7 2 d is not fully understood although much has been published about the aspects related to nonsuperconducting nature of this compound.3–9 Recently, there have been some reports suggesting superconductivity in Pr123 but these have met with considerable skepticism owing to the lack of reproducibility of data and the existence of an anomalously long c axis reported for the structure of Pr123.6,10–15 Several models have been proposed during the past few years to explain the anomalous behavior of PrBa2Cu3O7 2 d.3 The most widely discussed deal with the hole filling 16–18 or pair breaking mediated by the hybridization of praseodymium 4f and oxygen 2p states of the CuO2 planes.19 The hole filling model assumes that Pr has a valence of more than 13 whereby the extra electrons from Pr neutralize the mobile holes which in turn brings the compound to a near insulating regime.However, this model has not met with great success since a valence exceeding 13 for Pr is not supported by many experimental findings.20–24 On the other hand, pair breaking appeared to be a more promising model as it can explain the suppression of superconductivity in R1 2 xPrxBa2Cu3O7 2 d through the hybridization of praseodymium 4f and oxygen 2p states of the CuO2 planes leading to the localization of the holes, although pair breaking by itself cannot account for the insulating behavior of PrBa2Cu3O7 2 d.19 Recently, two models have been proposed by Fehrenbacher and Rice (FR) 25 and Liechtenstein and Mazin (LM),26 both of which involve the transfer of holes from the pds state to the pdp state.In the FR model the 4fz(x2 2 y2) orbital of Pr31 ion is hybridized with pp orbitals of neighboring planar oxide ions. The pp holes are treated as planar (pxy) in the FR model, whereas they have comparable amounts of px,y and pz character in the LM model. These models are well supported by a polarization dependent X-ray absorption study on detwinned Y1 2 xPrxBa2Cu3O7 2 d single crystals.7 Although both FR and LM models can explain the insulating behavior of PrBa2- Cu3O7 2 d, they do not account for a high value of nearly 17 K for the antiferromagnetic ordering of praseodymium moments (Neel temperature TN) for this compound.6,9 Earlier experimental reports have demonstrated that the hole states play a vital role in superconductivity in the p-type cuprate superconductors.Therefore, knowledge of the electronic structure near the Fermi level of these compounds is important to understand the mechanism of superconductivity.The X-ray absorption spectra are determined by electronic transitions from a selected atomic core level to the unoccupied electronic states near the Fermi level. X-Ray absorption near edge structure (XANES) is therefore a direct probe of the character and local density of the hole states responsible for the high temperature superconductivity.It has been generally accepted that the hole states in p-type cuprate superconductors are located on oxygen sites. Moreover, there are several non-equivalent oxygen sites in these materials. Therefore it becomes necessary to understand the distribution of hole states among diVerent oxygen sites and their role in superconductivity as well. The recent works on bulk and thin films of PrBa2 2 xSrx- Cu3O7 2 d show that strontium doping at the barium site increases the distance between Pr31 and O22 ions in the CuO2 plane and consequently leads to a dramatic decrease of resistivity in doped samples.9 Very recently, Liu et al.5 also suggested in support of the above observations that chemical substitution of Sr for Ba in YBa2 2 xSrxCu3O7 2 d gives rise to higher hole concentrations, leading to an overdoped state.As a result, a decrease in Tc was found from 92 K when x = 0.0 to 84 K when x = 0.8, which was the maximum solubility of Sr without modifying the phase.Similarly, the holes induced by the Ca21 ions in Pr0.5Ca0.5Ba2Cu3O7 2 d thin films lead to the recovery of the high temperature superconductivity for the Pr123 phase.14 Recently, this was also well brought out by Merz et al.8 based on their polarization dependent X-ray absorption spectral studies on Y1 2 xCaxBa2Cu3O7 2 d single crystals. They found that the maximum hole counts correspond to the composition2066 J. Chem. Soc., Dalton Trans., 1999, 2065–2069 Y0.9Ca0.1Ba2Cu3O6.91.All these reports clearly indicate that the substitution of Ba21 partially by Sr21 ions and R by Ca21 ions induces and creates the holes, in contrast to the substitution of Pr31 ions at the R site which always depletes the hole content. In addition, it has been reported that Eu1 2 xPrxBa2Cu3O7 2 d becomes an insulator when x � 0.5. However, Eu1 2 xPrxBaSr- Cu3O7 2 d and Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d do show superconductivity unlike Eu1 2 xPrxBa2Cu3O7 2 d even when x = 0.5 (Table 1).X-Ray absorption studies on Y1 2 xPrxBa2Cu3O7 2 d and Dy1 2 xPrxBa2Cu3O7 2 d indicate that the depression of Tc is due to the depletion of the hole states by Pr31 substitution and the rate at which the depletion occurs is comparable for both systems.4,7 However, there has been no report of X-ray absorption spectroscopy on R123 by substituting Ba21 partially by Sr21 ions and R by Ca21 and Pr31 ions. Therefore, we have chosen a series of compounds corresponding to the composition Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d, where x = 0.1–0.5, for the present study. Preliminary data for the present and the related systems, viz.Eu1 2 xPrxBa2Cu3O7 2 d and Eu1 2 xPrxBaSr- Cu3O7 2 d where x = 0.0–0.5, have been reported elsewhere.27,28 The corresponding Tc values are reproduced in Table 1. It has been shown in the present study that the hole states are depleted systematically which in turn leads to reduction in Tc by the substitution of Pr31 ions in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d.However, the rate at which the depletion occurs is lower when compared to those of the Y1 2 xPrxBa2Cu3O7 2 d and Dy1 2 xPrx- Ba2Cu3O7 2 d systems. Experimental The samples Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d (x = 0.1, 0.2, 0.3, 0.4 or 0.5) were prepared by solid state reactions of the respective oxides or nitrates at 1250 K. Stoichiometric amounts of Eu2O3, Pr6O11, Ba(NO3)2, Sr(NO3)2, CaCO3nd CuO obtained from Aldrich Chem.Co. (99.9%) were weighed and mixed thoroughly in an agate pestle and mortar. The mixtures were then transferred to a platinum crucible and placed in a furnace. Initially the mixture was calcined at 1225 K for 24 h. The mixture was then pressed into pellets and sintered at 1250 K for 78 h. The heat treatment was interrupted every 24 h for grinding and pelletizations. The sintered samples were then annealed in flowing argon gas for 12 h at 1100 K, the furnace was cooled to 725 K and reannealed in oxygen for 72 h.The samples were then cooled to room temperature at the rate of 1 8C min21. The powder X-ray diVractograms were recorded on a Philips X-ray diVractometer (Model PW 1071) with nickel filtered Cu-Ka radiation. The analyses of powder X-ray diVractograms showed that all the samples were single phase. Iodometric titration was used to determine the oxygen contents for the samples.28 The Tc values were obtained for the above compounds by using an APD cryocooler with a Meissner coil attachment in conjunction with an E.G.and G.P.A.R. twophase lock-in amplifier (Model 5280). The X-ray absorption measurements for the polycrystalline samples were carried out at the 6 m high energy spherical grating monochromator (HSGM) beam line of the Synchrotron Radiation Research Center (SRRC), Taiwan, ROC with an electron beam energy of 1.5 GeV and a maximum stored current of 240 mA. The X-ray fluorescence yield (XFY) technique was adopted for recording the spectra by using a micro channel plate (MCP) as a detector.The MCP detector is composed of a dual set of micro channel plates with an electrically isolated grid mounted in front of them. The X-ray fluorescence yield technique is strictly bulk sensitive with a probing depth of thousands of angstroms in contrast to the electron yield technique. During the X-ray fluorescence yield measurements the grid was set to a voltage of 100 V, while the front of the MCP was set at 22000 V and the rear at 2200 V.The grid bias insured that positive ions would not be detected while the MCP bias insured that no electrons were detected. The MCP detector was located at ª2 cm from the sample and oriented parallel to the sample surface. The photons were incident at an angle of 458 with respect to the sample normal. The incident photon intensity (I0) was monitored simultaneously by a nickel mesh located after the exit slit of the monochromator.All the measurements were normalized to I0. The photon energies were calibrated using the known oxygen K-edge and copper L3-edge absorption peaks of CuO. The energy resolution of the monochromator was set to ª0.22 and ª0.45 eV for the oxygen K-edge and copper L3-edge X-ray absorption spectral measurements, respectively. All the measurements were performed at room temperature. Results and discussion The oxygen contents of Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d where x = 0.1, 0.2, 0.3, 0.4 or 0.5 as determined by the iodometric titrations are given in Table 1.These values remain unaltered at ª6.97 ± 0.02 with increasing praseodymium substitution, indicating that Pr is in trivalent state and is substituting at the europium site. The ionic size of Ca21 is closer to that of Eu31 ion, rather than that of Ba21/Sr21,30 but its valence would favor its going to the Ba/Sr site. However, an earlier study on Eu1 2 x- CaxBaSrCu3O7 2 d by Waje et al.31 found that the lattice parameter ‘a’ increased with increasing calcium concentration ‘x’ indicating that Ca substitutes at the europium site.The oxygen K-edge X-ray absorption spectra of Eu0.9 2 xPrx- Ca0.1BaSrCu3O7 2 d where x = 0.1, 0.2, 0.3, 0.4 or 0.5 obtained by the X-ray fluorescence yield technique are shown in Fig. 1. All these spectra were normalized to their respective absorption cross-sections of the actual oxygen contents given in Table 1. The salient features of these spectra are as follows.There are two prepeaks at ª528.3 and ª529.4 eV with a shoulder at 527.5 eV, and a broad peak at ª537 eV. The low-energy transitions with energy below 532 eV are ascribed to transitions from the oxygen 1s core electrons to holes with predominant 2p character on the oxygen sites. The transitions with energy above 532 eV are attributed to the transitions to oxygen 2p states, which are hybridized with barium-4d, strontium-3d, praseodymium- 5d or 4f states.24 The system Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d is isomorphic with YBa2Cu3O7 2 d (orthorhombic structure with space group pmmm) at lower Pr31 concentrations. However, with increasing Pr31 concentration, i.e.when x = 0.3 and above, the phase changes from orthorhombic structure to tetragonal structure. 27,28 The tetragonal structure is again isomorphic with YBa2Cu3O6.32 We therefore adopted the same scheme of assignments for the present oxygen 1s absorption spectra as the samples belong to both orthorhombic as well as tetragonal phases.A study on R(Ba1 2 zRz)2Cu3O7 2 d (R = Nd or Pr) by soft X-ray absorption spectroscopy also supports the present assignments.33 The orthorhombic crystal structure of Eu0.9 2 x- PrxCa0.1BaSrCu3O7 2 d has four non-equivalent oxygen sites, viz. O(2) and O(3) within the Cu(2)O2 layers, O(4) in the BaO and SrO planes, and O(1) in the Cu(1)O chains along the b axis. Table 1 The Tc values for Eu12xPrxBa2Cu3O72d, Eu12xPrxBaSrCu3O72d and Eu0.92xPrxCa0.1BaSrCu3O72d, and the oxygen contents (7 2 d) for Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d Tc/K ± 1 Composition Eu1 2 xPrxBa2Cu3O7 2 d Eu1 2 xPrxBaSrCu3O7 2 d Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d x 0.0 94.8 84.0 77.0 0.1 79.5 68.0 69.0 0.2 57.9 62.0 65.0 0.3 38.8 53.0 60.0 0 .4 15.0 37.5 33.7 0.5 0.0 26.7 21.4 Oxygen content (7 2 d) ± 0.02 Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d — 6.95 6.97 6.98 6.97 6.99J.Chem. Soc., Dalton Trans., 1999, 2065–2069 2067 The observed multiple transitions at energy below 532 eV are related to the chemical shifts of oxygen 1s binding energies for diVerent oxygen sites.The diVerences in the chemical shifts of oxygen 1s core levels can be obtained by using local density approximation (LDA) band-structure calculations. Based on LDA calculations for YBa2Cu3O7 2 d by Krakauer et al.,34 the oxygen 1s energy levels of the sites O(2) and O(3) in the Cu(2)O2 planes were very close to each other and found to be 0.29 eV higher than that of the O(1) site in the Cu(1)O chains which in turn was 0.4 eV higher than the energy level of the O(4) site in the BaO planes.Based on these LDA calculations and earlier reports on isomorphic systems, viz. YBa2- Cu3O7 2 d, PrBa2Cu3O7 2 d and DyBa2Cu3O7 2 d, the following assignment scheme for the oxygen 1s absorption spectra of Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d is made.4,7,8,24,34–37 The shoulder at 527.5 eV is assigned to the transition from the oxygen 1s core level to the superposition of oxygen 2p hole states originating from the apical oxygen sites, viz.O(4) sites in the BaO and SrO planes, and oxygen sites in Cu(1)O chains, viz. O(1). The high energy prepeak at 528.3 eV is due to the transition into oxygen 2p holes on O(2) and O(3) sites situated in the Cu(2)O2 planes. The peak at 529.4 eV is ascribed to the transition into the conduction band, in other words the upper Hubbard band (UHB) which is predominantly formed by copper 3d states with some admixture of oxygen 2p states.38,39 As a consequence of the strong on-site correlation eVects on the copper sites in cuprate superconductors, the upper Hubbard band is always assumed to be present.40 The spectral weight of the component corresponding to the transition to the hole states on O(2) and O(3) sites situated in Cu(2)O2 planes decreases, while the component corresponding to the transition to the upper Hubbard band increases, with increasing Pr31 substitution for Eu31 in Eu0.9 2 x- PrxCa0.1BaSrCu3O7 2 d (Fig. 1). These observations clearly demonstrate that the chemical substitution of Pr31 for Eu31 in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d reduces the hole states in Cu(2)O2 planes. To estimate the hole states on diVerent oxygen sites and to understand the variation of hole states for diVerent praseodymium substitution, the oxygen K-edge X-ray absorption spectra as shown in Fig. 1 were resolved into diVerent Gaussian components by the following procedure.By fixing the band positions obtained by plotting the second derivative form, the bandwidths were evaluated over a series. It was found that the Fig. 1 Oxygen K-edge X-ray absorption spectra for Eu0.9 2 xPrx- Ca0.1BaSrCu3O7 2 d. All these spectra were normalized to their respective absorption cross-sections of the actual oxygen contents given in Table 1. bandwidths obtained from the fits were almost constant for a particular band over a series. These bandwidths were averaged over a series for a particular band.Then the resultant bandwidths were fixed for the final fit and the band positions allowed to vary. The goodness of fit was judged by the c2 value. The relative intensities of each Gaussian component were obtained by integrating the area under the band. The errors in the relative intensities of each Gaussian component were obtained as standard errors (the square root of the mean square error) for the best fits. As an illustrative example, the pre-edge structure of oxygen 1s X-ray absorption spectrum of Eu0.8Pr0.1Ca0.1BaSr- Cu3O7 2 d including the Gaussian components along with their assignments is shown in Fig. 2(a). The resultant spectral weights for diVerent Gaussian components are plotted as a function of praseodymium substitution in Fig. 3. All these spectral weights were normalized to that of the intense band at Fig. 2 Pre-edge structure of the oxygen K-edge (a) and copper L3-edge (b) X-ray absorption spectra for Eu0.8Pr0.1Ca0.1BaSrCu3O7 2 d.The dashed lines represent the resolved Gaussian components of the spectra. d, Experimental; ——, fit. Fig. 3 Plots of the relative intensities of hole states in Eu0.9 2 xPrx- Ca0.1BaSrCu3O7 2 d on oxygen sites originating from the (a) CuO3 ribbons, (b) CuO2 planes and (c) the upper Hubbard band (UHB) as a function of Pr31 substitution. The solid lines are guides to the eyes.2068 J. Chem. Soc., Dalton Trans., 1999, 2065–2069 ª537 eV. The spectral weight at 527.5 eV corresponding to oxygen 2p hole states on the apical oxygen sites, viz.O(4), and on the Cu(1)O chains, viz. O(1), is constant with increasing praseodymium substitution (Fig. 3). On the other hand, the spectral weight at 528.3 eV corresponding to oxygen 2p hole states on the O(2) and O(3) sites situated in the Cu(2)O2 planes decreases systematically, while the spectral weight of the upper Hubbard band at 529.4 eV is found to increase at the cost of hole states on the O(2) and O(3) sites in the Cu(2)O2 planes with increasing praseodymium substitution (Fig. 3). The praseodymium M45-edge and copper L23-edge X-ray absorption spectra of Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d where x = 0.1, 0.2, 0.3, 0.4 or 0.5 obtained by the X-ray fluorescence yield technique are shown in Fig. 4. The bands centered at ª931 and ª951 eV are called copper L3 and L2 white lines, and are assigned to the transitions from the ground states of Cu21, Cu(2p3/2,1/2)3d9–O 2p6, into the excited states, Cu(2p3/2,1/2)213d10– O 2p6, respectively, where (2p3/2,1/2)21 denotes a hole containing 2p3/2 or 2p1/2 states.41,42 The shoulders situated to high energy of each main band are assigned to the transitions from the ground states of Cu31, Cu(2p3/2,1/2)3d9L21 into the excited states, Cu(2p3/2,1/2)213d10L21, respectively, where L21 denotes a ligand hole on the oxygen 2p orbital.These high-energy shoulders are therefore identified as the holes in the CuO2 layers and CuO3 ribbons.41,42 The shoulders to low energy of both copper Ledges are due to praseodymium M5 and M4 white lines respectively and are assigned to the transitions from 3d5/2 electrons to 4f states of Pr31. Attempts were made to resolve the copper L3-edge and praseodymium M5-edge X-ray absorption spectra and as a representative example the results of the spectrum corresponding to the composition Eu0.8Pr0.1Ca0.1BaSrCu3O7 2 d including the resolved Gaussian components and their assignments are shown in Fig. 2(b). The resultant spectral weights for Gaussian components corresponding to the transitions, viz. the hole contents in the CuO2 layers and CuO3 ribbons, and praseodymium M5 white line, are plotted as a function of praseodymium substitution in Fig. 5. All these spectral weights are normalized to that corresponding to the transition due to Cu21 at ª931 eV. The spectral weight corresponding to the total hole contents decreases while that corresponding to the praseodymium M5 white line increases with increasing praseodymium concentration.The increase in the spectral weight corresponding to the M5 white line confirms that Eu31 ions in the Eu0.9 2 x- PrxCa0.1BaSrCu3O7 2 d system are partially substituted by Pr31 ions. The decrease in hole content with increasing Pr31 is also Fig. 4 Praseodymium M45-edge and copper L23-edge X-ray absorption spectra for Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d. evidenced by the oxygen K-edge X-ray absorption spectra as discussed earlier.The present experimental results based on oxygen K-edge and copper L-edge X-ray absorption spectra for Eu0.9 2 xPrx- Ca0.1BaSrCu3O7 2 d clearly demonstrate that the hole states responsible for superconductivity get depleted progressively by the praseodymium substitution.43 These should in principle reduce the Tc, which is indeed found to be the case. Thus these results give evidence in support of the hole-depletion models based on Pr 4f–O 2pp hybridization.25,26 By comparing the depletion of hole states upon praseodymium substitution in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d with those in Y1 2 xPrxBa2Cu3- O7 2 d and Dy1 2 xPrxBa2Cu3O7 2 d, it is found that the rate at which the depletion of hole states occurs in the present case is lower than in Y1 2 xPrxBa2Cu3O7 2 d and Dy1 2 xPrxBa2- Cu3O7 2 d.4,7 This is due to the partial substitution of Ba21 by Sr21 and Eu by Ca21 ions which keeps inducing and creating the holes which in turn leads to a decrease in the rate of hole depletion in the present case.It is well known that the phenomenological Judd–Ofelt model can account for the observed intensities for 4f–4f transitions of all the rare earth ions except Pr31.44–46 In the case of Pr31 ions, the 5d level is relatively low lying which results in a strong mixing with the 4f orbital. This is reflected by the abnormal intensity reported for the transition 3H4 æÆ 3P2 in the absorption spectra measured in the uv-visible region for Pr31 ion in diVerent host materials.44–46 The models proposed by FR and LM consider that the 4fz(x2 2 y2) orbital of Pr31 is strongly localized and hybridized with pp orbitals of neighboring planar oxide ions.25,26 These models do not consider 4f–5d mixing for Pr31 ion.The abnormal intensity reported for the transition 3H4 æÆ 3P2 and the non-applicability of Judd–Ofelt theory for the Pr31 ions suggest that the 4f orbital is strongly perturbed by the 5d orbital.One should take this eVect into account when the 4f orbital of Pr31 ion is considered for hybridization. This may be one of the reasons that both FR and LM models could not account for a high value of nearly 17 K for the antiferromagnetic ordering of praseodymium moments (Neel temperature, TN) for PrBa2Cu3O7 2 d.6,9 Conclusion High resolution oxygen K-edge and copper L23-edge X-ray Fig. 5 Plots of the relative intensities of (a) the band at ª930 eV originating from 3d5/2 to 4f of Pr31 and (b) the defect state at ª932 eV on the copper sites, for Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d as a function of Pr31 substitution.The solid lines are guides to the eyes.J. Chem. Soc., Dalton Trans., 1999, 2065–2069 2069 absorption spectra for Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d as a function of Pr31 substitution have been obtained by using a bulk sensitive X-ray fluorescence yield technique. The spectral weight of the pre-edge structure in the oxygen 1s X-ray absorption at 527.5 eV corresponding to oxygen 2p hole states on apical oxygen sites and on Cu(1)O chains is constant with increasing praseodymium substitution. On the other hand, the spectral weight at 528.3 eV corresponding to oxygen 2p hole states situated in the Cu(2)O2 planes decreases systematically, while the spectral weight of the upper Hubbard band at 529.4 eV is found to increase at the cost of hole states in the Cu(2)O2 planes with increasing praseodymium substitution.The analyses of the copper L3-edge also show that the spectral weight of the high energy shoulder at 932.3 eV corresponding to the hole contents in the CuO2 layers and CuO3 ribbons decreases with increasing praseodymium concentration. Both the oxygen Kedge and copper L-edge X-ray absorption spectra for Eu0.9 2 x- PrxCa0.1BaSrCu3O7 2 d clearly demonstrate that the hole states are depleted by the praseodymium substitution, which explains the reduction in the Tc.However, the rate at which the depletion occurs is lower in Eu0.9 2 xPrxCa0.1BaSrCu3O7 2 d compared to that in Y1 2 xPrxBa2Cu3O7 2 d and Dy1 2 xPrxBa2Cu3O7 2 d. This is due to the partial substitution of Ba21 by Sr21 ions and Eu by Ca21 ions, which induces and creates the holes. Our results give evidence in support of the hole-depletion models based on Pr 4f–O 2pp hybridization. Further the anomalous behavior of Pr31 can be well explained if one takes into account the 4f–5d mixing in which the 5d level is low lying and which strongly mixes with the 4f orbital of Pr31 ions unlike those of other rare earth metal ions.References 1 Z. Fisk, J. D. Thompson, E. Zirngiebl, J. L. Smith and S. W. Cheong, Solid State Commun., 1987, 62, 743. 2 P. H. Hor, R. L. Meng, Y. Q. Wang, L. Gao, Z. J. Huang, J. Bechtold, K. Forster and C. W. Chu, Phys. Rev. Lett., 1987, 58, 1891. 3 H. B. Radousky, Mater. Res. Bull., 1992, 7, 1917. 4 J. M. Chen, R. S. Liu, J.G. Lin, C. Y. Huang and J. C. Ho, Phys. Rev. B, 1997, 55, 14586 and refs. therein. 5 R. S. Liu, C. Y. Cheng and J. M. Chen, Inorg. Chem., 1998, 37, 5527. 6 V. P. S. Awana, S. X. Dou, R. Singh, A. V. Narlikar, S. K. Malik and W. B. Yelon, J. Appl. Phys., 1998, 83, 7315 and refs. therein. 7 M. Merz, N. Nucker, E. Pellegrin, P. Schweiss, S. Schuppler, M. Kielwein, M. Knupfer, M. S. Golden, J. Fink, C. T. Chen, V. Chakarian, Y. U. Idzerda and A. Erb, Phys. Rev. B, 1997, 55, 9160. 8 M. Merz, N. Nucker, P. Schweiss, S. Schuppler, C. T. Chen, V. Chakarian, J. Freeland, Y. U. Idzerda, M. Klaser, G. Muller-Vogt and Th. Wolf, Phys. Rev. Lett., 1998, 80, 5192. 9 Y. G. Zhao, Z. W. Dong, M. Rajeswari, R. P. Sharma and T. Venkatesan, Phys. Rev. B, 1998, 58, 1068. 10 H. A. Blackstead and J. D. Dow, Phys. Rev. B, 1995, 51, 11830. 11 Z. Zou, K. Oka, T. Ito and Y. Nishihara, Jpn. J. Appl. Phys., 1997, 36, L18. 12 Z. Zou, J. Ye, K. Oka and Y. Nishihara, Phys. Rev. Lett., 1998, 80, 1074. 13 J. Ye, Z. Zou, A. Matsushita, K. Oka, Y. Nishihara and T. Matsumoto, Phys. Rev. B, 1998, 58, R619. 14 D. P. Norton, D. H. Lowndes, B. C. Sales, J. D. Budai, B. C. Chakoumakos and H. R. Kerchner, Phys. Rev. Lett., 1991, 66, 1537. 15 Y. F. Xiong, Y. S. Yao L. F. Xu, F. Wu, D. Jin and Z. X. Zhao, Solid State Commun., 1998, 107, 509. 16 A. Matsuda, K. Kinoshita, T. Ishii, H. Shibata, T. Watanabe and T. Yamada, Phys. Rev. B, 1988, 38, 2910. 17 Z. Quirui, Z. H. Zenhui, Z.Han, X. Jiansheng, W. Shenxi and F. Minghu, Physica C, 1989, 162–164, 963. 18 B. Fisher, J. Genossar, L. Patlagan and J. Ashkenazi, Phys. Rev. B, 1991, 43, 2821. 19 A. Kebede, C. S. Jee, J. Schwegler, J. E. Crow, T. Mihalisin, G. H. Myer, R. E. Salomon, P. Schlottmann, M. V. Kuric, S. H. Bloom and R. P. Guertin, Phys. Rev. B, 1989, 40, 4517. 20 L. Soderholm, C. K. Loong, G. I. Goodman and B. D. Dabrowski, Phys. Rev. B, 1991, 43, 7923. 21 G. Hilscher, E. Holland-Moritz, T.Holubar, H. D. Jostarndt, V. Nekvasil, G. Schaudy, U. Walter and G. Fillion, Phys. Rev. B, 1994, 49, 535. 22 J. S. Kang, J. W. Allen, Z. X. Shen, W. P. Ellis, J. J. Yeh, B. W. Lee, M. B. Maple, W. E. Speicer and I. Lindau, J. Less-Common Met., 1989, 148, 121. 23 U. Neukrich, C. T. Simmons, D. Sladeczek, C. Laubschat, O. Strebel, G. Kaindl and D. D. Sarma, Europhys. Lett., 1988, 5, 567. 24 J. Fink, N. Nucker, H. Romberg, M. Alexander, M. B. Maple, J. J. Neumeier and J. W.Allen, Phys. Rev. B, 1990, 42, 4823. 25 R. Fehrenbacher and T. M. Rice, Phys. Rev. Lett., 1993, 70, 3471 and refs. therein. 26 A. I. Liechtenstein and I. I. Mazin, Phys. Rev. Lett., 1995, 74, 1000 and refs. therein. 27 I. K. Gopalakrishnan, T. B. Waje and J. V. Yakhmi, Physica C, 1999, 311, 246. 28 T. B. Waje, Ph.D. Thesis, University of Bombay, 1998. 29 E. H. Appelmann, L. R. Morss, A. M. Kini, U. Geiser, A. Umezawa, G. W. Crabtree and K. D. Carlson, Inorg. Chem., 1987, 26, 3237. 30 R. D. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751. 31 T. B. Waje, I. K. Gopalakrishnan and J. V. Yakhmi, J. Supercond., 1998, 10, 645. 32 P. K. Gallagher, M. M. OBryan, S. A. Sunshine and D. W. Murphy, Mater. Res. Bull., 1987, 22, 995. 33 J. M. Chen, R. S. Liu, P. Nachimuthu, M. J. Kramer, K. W. Dennis and R. W. McCallum, Phys. Rev. B, 1999, 59, 3855 and refs. therein. 34 H. Krakauer, W. F. Pickett and R. E. Cohen, J. Supercond., 1988, 1, 111. 35 N. Nucker, E. Pellegrin, P. Schweiss, J. Fink, S. L. Molodtsov, C. T. Simmons, G. Kaindl, W. Frentrup, A. Erb and Muller-Vogt, Phys. Rev. B, 1995, 51, 8529. 36 J. Fink, N. Nucker, E. Pellegrin, H. Romberg, M. Alexander and M. Knupfer, J. Electron Spectrosc. Relat. Phenom., 1994, 66, 395. 37 J. G. Lin, Y. Y. Xue, C. W. Chu, X. W. Cao and J. C. Ho, J. Appl. Phys., 1993, 73, 5871. 38 H. Romberg, M. Alexander, N. Nucker, P. Adelmann and J. Fink, Phys. Rev. B, 1990, 42, 8768. 39 C. T. Chen, F. Sette, Y. Ma, M. S. Hybertsen, E. B. Stechel, W. M. C. Foulkes, M. Schluter, S. W. Cheong, A. S. Cooper, L. W. Rupp, Jr., B. Batlogg, Y. L. Soo, Z. H. Ming, A. Krol and Y. H. Kao, Phys. Rev. Lett., 1991, 66, 104. 40 D. Vaknin, S. K. Shiha, D. E. Moneton, D. C. Johnston, J. M. Newsam, C. R. Safiva and H. E. King, Jr., Phys. Rev. Lett., 1987, 58, 2802. 41 A. Bianconi, M. De Santis, A. Di Ciccio, A. M. Flank, A. Fronk, A. Fontaine, P. Legarde, H. K. Yoshida, A. Kotani and A. Marcelli, Phys. Rev. B, 1988, 38, 7196. 42 A. Bianconi, M. De Santis, A. Di Ciccio, A. M. Flank, A. Fronk, A. Fontaine, P. Legarde, H. K. Yoshida, A. Kotani and A. Marcelli, Physica C, 1988, 153–155, 1760. 43 A. Krol, C. S. Lin, Y. L. Soo, Z. H. Ming, Y. H. Kao, J. H. Wang, M. Qi and G. C. Smith, Phys. Rev. B, 1992, 45, 10051. 44 R. Resifeld and C. K. Jorgensen, Lasers and Excited States of Rare Earths, Springer, New York, 1977. 45 R. D. Peacock, Struct. Bonding (Berlin), 1975, 22, 83. 46 M. Malinowski, R. Wolski and W. Wolinski, Solid State Commun., 1990, 74, 17. Paper 9/02994J

ISSN:1477-9226    

DOI:10.1039/a902994j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 2067 Deoxyribonucleic acid binding and photocleavage studies of rhenium(I) dipyridophenazine complexes Vivian Wing-Wah Yam,*,a Kenneth Kam-Wing Lo,a Kung-Kai Cheung a and Richard Yuen-Chong Kong b a Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong b Department of Biology and Chemistry, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong The intercalative binding interaction of [Re(dppz)(CO)3(py)][O3SCF3] (dppz = dipyrido[3,2-a:29,39-c]phenazine, py = pyridine) and [Re(dppn)(CO)3(py)][O3SCF3] (dppn = benzo[i]dipyrido[3,2-a:29,39-c]phenazine) with doublestranded calf thymus DNA, and synthetic oligonucleotides poly(dA)?poly(dT) and poly(dC)?poly(dG) has been studied with spectroscopic methods.The complexes have been found to promote cleavage of plasmid pBR322 DNA from the supercoiled form I to the open circular form II upon irradiation. The crystal structure of [Re(dppz)(CO)3(py)][O3SCF3] has also been established.A number of transition-metal complexes have been utilized to probe nucleic acid structures and in the development of DNAcleaving agents.1–12 Barton and co-workers 3c–f employed a number of chiral ruthenium(II) polypyridine complexes as DNA chirality probes. The complex [Ru(bipy)2(dppz)]21 (bipy = 2,29- bipyridine, dppz = dipyrido[3,2-a:29,39-c]phenazine) termed as a molecular ‘light-switch’ for DNA is non-emissive in aqueous buffer, but a luminescence enhancement of >104 is observed in the presence of double-stranded DNA with the appropriate chirality.3g–h The interaction is intercalative in nature, with the extended planar dppz ligand stacking into the double helix.Although the photophysics and photochemistry of rhenium(I) complexes have been well documented, reports on the interactions of related complexes with DNA have been very limited.11,12 We recently communicated DNA-binding studies of two rhenium(I) complexes [Re(dppz)(CO)3(py)][O3SCF3] 1 and [Re(dppn)(CO)3(py)][O3SCF3] 2 (dppn = benzo[i]dipyrido- [3,2-a:29,39-c]phenazine, py = pyridine) with extended diimine ligands dppz and dppn.12 Herein we describe DNA-binding studies of these complexes with double-stranded calf thymus DNA and synthetic oligonucleotides poly(dA)?poly(dT) and poly(dC)?poly(dG) (dA, dT, dC, dG = deoxy-adenosine, -ribosylthymine, -cytidine, -guanosine).The photoexcited complexes have also been found to cleave plasmid pBR322 DNA.Moreover, the crystal structure of 1 has also been determined. Experimental Complexes 1 and 2 were prepared as described.12 Autoclaved Milli-Q water was used for the preparation of the aqueous solutions. Calf thymus DNA was obtained from Sigma Chemical Company and purified by phenol extraction as described.13 Synthetic polynucleotides poly(dA)?poly(dT) and poly(dC)? poly(dG) were obtained from Pharmacia Biotech Ltd. and used as received.The UV/VIS spectra were obtained on a Hewlett-Packard 8452A diode-array spectrophotometer, steady-state excitation and emission spectra on a Spex Fluorolog 111 spectrofluorometer. All spectroscopic titrations were carried out in aerated 5% aqueous buffered [20 mmol dm23 tris(hydroxymethyl)methylamine (Tris)–HCl, pH 7.0] methanolic solutions. The DNA concentrations per nucleotide were determined by absorption spectroscopy using the following molar absorption coefficients 14 (dm3 mol21 cm21): calf thymus DNA, 6600 at 260 nm; poly(dA)?poly(dT), 6000 at 260 nm; and poly(dC)? poly(dG), 7400 at 253 nm. Plasmid pBR322 DNA was extracted from pBR322- transformed Escherichia coli DH5a and the supercoiled DNA (form I) purified with a Qiagen Plasmid Maxi Kit.Superoxide dismutase (SOD) from bovine erythrocyte was obtained from Sigma Chemical Company. Solutions for photocleavage studies were irradiated at room temperature with RPR-3500 Å lamps (Rayonet photochemical chamber reactor, model RPR-100) in Pyrex tubes immersed in a water-bath to cut off both UV and IR radiation. Solutions were electrophoresed for 3 h at 40 V on a 0.8% agarose gel in Tris–acetate buffer, pH 8.The gel was stained with ethidium bromide (3,8-diamino-5-ethyl-6-phenylphenanthridinium bromide) and photographed under UV light. Crystallography Single crystals of complex 1 were grown from diffusion of diethyl ether vapour into a concentrated acetonitrile solution of the complex.Crystal data. 2C26H15N5O3Re1 2CF3SO3 2?H2O, Mr = 1579.43, monoclinic, space group P21/n (no. 14), a = 14.361(6), b = 20.878(7), c = 18.906(6) Å, b = 90.99(3)8, U = 5667(3) Å3, Z = 4, Dc = 1.851 g cm23, m(Mo-Ka) = 44.36 cm21, F(000) = 3064, T = 301 K. A yellow crystal of dimensions 0.20 × 0.15 × 0.35 mm was used for data collection at 28 8C on a Rigaku AFC7R diffractometer with graphite-monochromatized Mo-Ka radiation (l = 0.710 73 Å) using w–2q scans with w-scan angle (0.73 1 0.35 tan q)8 at a scan speed of 16.08 min21 [up to six scans for reflection I < 10s(I)].Intensity data (2qmax = 458; h 0–15, k 0–22, l 220 to 20; three standard reflections measured after every 300 showed decay of 10.64%), were corrected for decay and Lorentz-polarization effects, and empirical absorption corrections based on the y scan of four strong reflections (minimum and maximum transmission factors 0.936 and 1.000). Upon averaging the 8037 reflections, 7678 of which were uniquely measured (Rint = 0.027), 4450 with I > 3s(I) were considered observed and used in the structural analysis.The space group was uniquely determined from systematic absences and the structure was solved by Patterson N N N N N N N N dppz dppn2068 J. Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 methods and expanded using Fourier methods 15 and refinement by full-matrix least squares using the software package TEXSAN16 on a Silicon Graphics Indy computer.A crystallographic asymmetric unit consists of two complex cations and two CF3SO3 2 anions and one water molecule. All non-H atoms of the complex cations and the S atoms of the anions were refined anisotropically; C, F and O atoms of the anions and the O atom of the water molecule have large thermal motions and were refined isotropically. The hydrogen atoms of the water molecule were not located. The other 30 H atoms at calculated positions with thermal parameters equal to 1.3 times that of the attached C atoms were not refined.Convergence for 709 variable parameters by least-squares refinement on F with w = 4Fo 2/ s2(Fo 2), where s2(Fo 2) = [s2(I ) 1 (0.035Fo 2)2] for 4450 reflections with I > 3s(I) was reached at R = 0.049 and R9 = 0.065 with a goodness of fit of 1.93; (D/s)max = 0.04 for atoms of the complex cation. The final Fourier-difference map was featureless, with maximum positive and negative peaks of 1.63 and 0.91 e Å23 respectively.Selected bond distances and angles are summarized in Table 1. Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC). See Instructions for Authors, J. Chem. Soc., Dalton Trans., 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/490. Results and Discussion Even though there have been reports on the crystal structures of rhenium(I) diimine complexes,17 those of the molecular structures of related complexes with extended planar ligands have been exceptionally rare.12 The perspective drawings of the complex cation [Re(dppz)(CO)3(py)]1 with atomic numbering schemes are illustrated in Fig. 1(a). It is interesting that although the molecules of complexes 1 and 2 are similar, the unit cell of 1 contains two independent cations and anions while there is only one of each for that of 2.12 The perspective view of the unit cell of 1 is illustrated in Fig. 1(b). Selected bond distances and angles of 1 are listed in Table 1. The rhenium(I) centres adopt a distorted octahedral geometry with the carbonyl groups co-ordinated in a facial manner. The average bond distances and angles are comparable to typical values for similar rhenium(I) diimine complexes.12,17 As expected, all the atoms of the dppz ligand lie on an essentially perfect plane, with the distance from the phenazine edge of the plane to the Table 1 Selected geometric data (bond lengths in Å, angles in 8) for [Re(dppz)(CO)3(py)][O3SCF3] Re(1)]N(1) Re(1)]N(3) Re(1)]C(2) Re(2)]N(6) Re(2)]N(8) Re(2)]C(28) 2.18(1) 2.20(1) 1.85(2) 2.18(1) 2.17(1) 1.89(2) Re(1)]N(2) Re(1)]C(1) Re(1)]C(3) Re(2)]N(7) Re(2)]C(27) Re(2)]C(29) 2.16(1) 1.91(1) 1.92(2) 2.19(1) 1.90(1) 1.89(2) N(1)]Re(1)]N(2) N(1)]Re(1)]C(1) N(1)]Re(1)]C(3) N(2)]Re(1)]C(1) N(2)]Re(1)]C(3) N(3)]Re(1)]C(2) C(1)]Re(1)]C(2) C(2)]Re(1)]C(3) N(6)]Re(2)]N(8) N(6)]Re(2)]C(28) N(7)]Re(2)]N(8) N(7)]Re(2)]C(28) N(8)]Re(2)]C(27) N(8)]Re(2)]C(29) C(27)]Re(2)]C(29) 87.3(4) 175.9(5) 92.4(6) 92.2(5) 175.6(5) 171.1(6) 89.6(7) 88.7(7) 86.0(4) 91.8(5) 75.3(4) 172.4(5) 93.1(5) 173.6(6) 88.7(7) N(1)]Re(1)]N(3) N(1)]Re(1)]C(2) N(2)]Re(1)]N(3) N(2)]Re(1)]C(2) N(3)]Re(1)]C(1) N(3)]Re(1)]C(3) C(1)]Re(1)]C(3) N(6)]Re(2)]C(7) N(6)]Re(2)]C(27) N(6)]Re(2)]C(29) N(7)]Re(2)]C(27) N(7)]Re(2)]C(29) N(8)]Re(2)]C(28) C(27)]Re(2)]C(28) C(28)]Re(2)]C(29) 85.1(4) 94.5(6) 75.4(4) 95.8(6) 90.8(5) 100.2(5) 87.9(6) 84.8(4) 178.2(5) 92.0(6) 93.5(5) 98.4(6) 97.8(5) 89.9(6) 88.4(7) metal centre being ca. 10.1 Å. The distances between the ideal ring planes of the two dppz ligands on adjacent cations are found to be ca. 3.47 Å, showing some stacking interaction [Fig. 1(b)]. These values are similar to the base-pair stacking distance in DNA18a and the intercalator–base pair stacking distance in oligonucleotide intercalator complexes.18b,c Similar interplanar separations have also been observed in other dppz complexes such as [Ru(OH2)(dppz)(terpy)]21 (terpy = 2,29:69,20- terpyridine)4d and [Ru(h5-C5Me5)(NO)(dppz)]21.18d DNA Binding Electronic absorption titrations.In 5% aqueous buffered (20 mmol dm23 Tris–HCl, pH 7.0) methanolic solutions, the lowenergy absorption bands of 1 at 366 and 384 nm and 2 at 320, 400 and 422 nm exhibit hypochromism upon addition of double-stranded calf thymus DNA. The electronic absorption spectral traces for the titration are illustrated in Fig. 2.A small bathochromic shift is observed. These findings suggest the binding of the complexes to the biopolymers, most likely through a non-covalent intercalative mode.19 Similar observations have also been reported for other platinum(II),2c,10a,c,20 copper(I),9 ruthenium(II) 3c,d,f–h,5–7 and osmium(II) 3k metallointercalators. The intrinsic binding constants K of the rhenium(I) complexes with calf thymus DNA have been determined from the equation 21 D/Deap = (D/De) 1 (DeK)21 through a plot of D/ Deap vs.D, where D is the concentration of DNA in base pairs, Deap = (ea 2 ef) and De = (eb 2 ef). The apparent absorption coefficient, ea, was obtained by calculating Aobs/[Re]; eb and ef are the absorption coefficients of the bound and free form of the rhenium(I) complex, respectively. The slope and y intercept Fig. 1 (a) Perspective views of the two independent complex cations of complex 1.Thermal ellipsoids are shown at the 35% probability level. (b) Perspective view of the unit cell of 1J. Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 2069 of the linear fit of D/Deap vs. D give (1/De) and 1/(DeK), respectively. The intrinsic binding constant K can be obtained from the ratio of the slope to the y intercept. An intrinsic binding constant K of 6.4 × 104 dm3 mol21 was determined from the decay of the absorbance of complex 2 monitored at 310 nm upon addition of double-stranded calf thymus DNA.This value is comparable to those observed for [Pt(phen)(en)]21 (5 × 104 dm3 mol21) 22 and [Ru(phen)2(phi)]21 (4.7 × 104 dm3 mol21) 3f (phen = 1,10-phenanthroline, en = ethane-1,2-diamine and phi = 9,10-phenanthrenequinonediimine). However, for 1, a linear fit was not obtained for the absorption data. Emission titrations. In aqueous MeOH–buffer (20 mmol dm23 Tris–HCl, pH 7.0) solutions the low-energy emission of complex 1 is enhanced upon addition of double-stranded calf thymus DNA.Despite the structural similarity of the complexes, their interactions with double-stranded calf thymus DNA are significantly different, as revealed by the luminescence behaviour in the presence of DNA. The results of the emission titrations for both complexes with DNA are illustrated with the titration curves in Fig. 3. Upon addition of calf thymus DNA the emission intensity of 1 grows steadily to around 13 times Fig. 2 Electronic spectral traces of complexes 1 (a) and 2 (b) in 5% aqueous buffered (20 mmol dm23 Tris–HCl, pH 7.0) methanolic solution upon addition of double-stranded calf thymus DNA larger and saturates at a [DNA phosphate] : [Re] ratio of ca. 4.5 : 1. However, in the case of 2, the emission intensity drops at low [DNA phosphate] : [Re] ratios (minimum at around 0.4 : 1) before it gradually approaches saturation with an overall gain of approximately 1.3 times. Similar variation in emission intensity in the presence of double-stranded DNA has been observed for the related complex [Ru(phen)2(dppn)]21.3h The enhancement in the luminescence intensities of the complexes, together with the hypochromicity observed in the electronic absorption spectra, can be ascribed to intercalation of the rhenium(I) complexes to the double helix.This is in line with the DNA-binding affinities of other transition-metal complexes with a planar heterocyclic ligand.3c,d,f–i,k,5–7,9–11,20 Data from the emission titrations were also employed to determined the binding constants of the rhenium(I) complexes with DNA.The concentration of the free complex, cF, was obtained using equation (1) where cT is the sum of the cF = cT[(I/I0) 2 P]/(1 2 P) (1) concentrations of the free and bound forms of the complex, I and I0 are the emission intensities in the presence and absence of DNA and P is the ratio of the observed emission intensity of the bound complex to that of the free one.The limiting emission intensity is the y intercept of a plot of I/I0 vs. 1/[DNA phosphate] and the value of P can then be determined. The concentration of the bound complex, cB is equal to cT–cF. A plot of r/cF vs. r, where r is cB/[DNA], was constructed according to the modified Scatchard equation by McGhee and von Hippel,23 equation (2) where K is the intrinsic binding constant r/cF = K(1 2 nr){(1 2 nr)/[1 2 (n 2 1)r]}n21 (2) of the complex with the DNA and n is the binding site size in base pairs.The binding data were fitted using the above equation to obtain the binding parameters. An intrinsic binding constant K of 4.2 × 104 dm3 mol21 for complex 1 was determined. The size of the binding site, n, was 2. This indicates that two base pairs of calf thymus DNA are occupied after binding of a single [Re(dppz)(CO)3(py)]1 unit. This value is close to those reported for other copper porphyrin intercalators.24a Synthetic DNA binding In order to gain more insight into the possibility of preferential binding of the DNA with the rhenium(I) diimine complexes, absorption and emission titrations using synthetic oligonucleotides poly(dA)?poly(dT) and poly(dC)?poly(dG) have also been carried out.The low-energy absorption bands of both complexes experience a hypochromism and small red shift in the presence of the synthetic oligonucleotides. This is similar to the case of double- Fig. 3 Emission titration curves for complexes 1 (j) and 2 (d) with calf thymus DNA in aqueous buffered methanol. The emission intensities were monitored at 560 and 603 nm, respectively2070 J.Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 stranded calf thymus DNA and suggests a binding mode of intercalation. In addition, emission titrations have also been carried out for the complexes with poly(dA)?poly(dT) and poly(dC)?poly(dG), respectively. Fig. 4 illustrates the emission spectral traces for complex 1 upon addition of poly(dA)?poly(dT) in 5% methanol–buffer (20 mmol dm23 Tris–HCl, pH 7.0).The emission intensity shows a dramatic enhancement and saturates at a very low [DNA phosphate] : [Re] ratio of around 1.3 : 1, with an overall 13-fold gain. Fig. 5 displays the titration curves for 1 with poly(dA)?poly(dT) and poly(dC)?poly(dG), respectively. In sharp contrast, with poly(dC)?poly(dG), the emission intensity of 1 does not show any considerable enhancement.Owing to the similar luminescence behaviour of the complex in the presence of calf thymus DNA and poly(dA)?poly(dT), it is suggested that the complex has a higher affinity towards the AT sites of double-stranded calf thymus DNA. Similar specificity at the adenine and thymine pairs has also been observed for [Cu(bcp)2]19b (bcp = bathocuproine = 2,9-dimethyl-4,7-diphenyl- 1,10-phenanthroline) and [Ru(phen)3]21.24b This preferential binding for poly(dA)?poly(dT) polymer can be explained by the fact that propeller twisting of the base pairs is relatively facile for a dA?dT sequence,24c and this could alleviate steric effects associated with the non-intercalated ligands about the metal centre.Surprisingly, in contrast to the case with calf thymus DNA, complex 2 reveals a considerable enhancement in luminescence intensity in the presence of poly(dA)?poly(dT) (Fig. 6). The intensity of the band at 603 nm shows a steady increase and saturates at [DNA phosphate] : [Re] of ca. 5 : 1 with a total intensity gain of ca. five-fold. The titration curves for 2 with poly- Fig. 4 Emission spectral traces of complex 1 (62 mmol dm23) in aqueous MeOH–buffer (20 mmol dm23 Tris–HCl, pH 7.0) at 298 K in the presence of 0, 11, 22, 33, 45, 56, 67 and 134 mmol dm23 poly(dA)?poly(dT) Fig. 5 Emission titration curves for complexes 1 with poly(dA)? poly(dT) (j) and poly(dC)?poly(dG) (d), respectively, in aqueous buffered methanol. The emission intensities were monitored at 560 nm.(dA)?poly(dT) and poly(dC)?poly(dG) are illustrated in Fig. 7. It is interesting that in the presence of poly(dC)?poly(dG) the emission of the complex is quenched. The emission intensity becomes steady at [DNA phosphate] : [Re] = 3 : 1 and the final intensity is ca. 0.2 times the original. Similar observations have also been reported for other ruthenium(II) metallointercalators such as [Ru(tap)3]21 (tap = 1,4,5,8-tetraazaphenanthrene) and the luminescence quenching may be attributable to photooxidation of the guanine by the excited complex.6a–c,e It appears that the combined effects of (1) quenching of the luminescence associated with the poly(dC)?poly(dG) sites and (2) emission enhancement occurring with the poly(dA)? poly(dT) sites can explain the characteristic emission titration curve of complex 2 with calf thymus DNA (Fig. 3). Photocleavage Irradiation of complex 1 or 2 and plasmid pBR322 DNA in 1.7% methanolic buffer (20 mmol dm23 Tris–HCl, pH 7.0) at l > 350 nm for 30 min under aerobic conditions results in cleavage of the supercoiled form (I) of the plasmid pBR322 DNA to the nicked form (II) [Fig. 8(a), 8(b)]. No DNA cleavages are observed for controls in which the complexes are absent [lanes A, Fig. 8(a), 8(b)] or incubation of the plasmid with either complex in the dark [lanes H, Fig. 8(a), 8(b)]. In the case of 2, a [DNA] : [Re] ratio of 10 : 1 (or smaller) causes degradation of the plasmid [lanes D–G, Fig. 8(b)] as a result of non-specific multiple cuts. In order to establish the reactive species responsible for the photoinduced cleavage of the plasmid, the following experiments have been carried out. Irradiation of the plasmid pBR322 DNA in the presence of complex 1 under anaerobic Fig. 6 Emission spectral traces of complex 2 (40 mmol dm23) in aqueous MeOH–buffer (20 mmol dm23 Tris–HCl, pH 7.0) at 298 K in the presence of 0, 33, 67, 100, 134, 167 and 201 mmol dm23 poly(dA)? poly(dT) Fig. 7 Emission titration curves for complex 2 with poly(dA)? poly(dT) (j) and poly(dC)?poly(dG) (d), respectively, in aqueous buffered methanol. The emission intensities were monitored at 603 nmJ. Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 2071 conditions by degassing the solution with purified nitrogen for 25 min does not cause appreciable changes (data not shown). It appears that oxygen is not involved in the photolytic cleavage. This finding is also demonstrated in Fig. 8(c). The cleavage of the plasmid is not inhibited in the presence of a singlet oxygen (1O2) scavenger, histidine 25 (1.20 mmol dm23) [lane D, Fig. 8(c)], as well as hydroxyl radical (OH?) quenchers such as mannitol 26a (50.0 mmol dm23) [lane F, Fig. 8(c)], ethanol 26 (1.7 mol dm23) [lane G, Fig. 8(c)] and sodium formate26b (100.0 mmol dm23) [lane H, Fig. 8(c)]. Furthermore, no enhancement in photocleavage activity is observed for the reaction carried out in D2O, in which singlet oxygen has a prolonged lifetime.27 These findings suggest that the photoinduced cleavage is a consequence of the oxidation of the plasmid DNA biopolymer by the excited complex 1*, probably via oxidation at the guanine site.6a–c,e In contrast, photocleavage of the plasmid pBR322 DNA is inhibited in degassed buffer solution for complex 2.Superoxide anion radical (O2~2) may be the reactive species as the cleavage is slightly inhibited in the presence of superoxide dismutase (SOD)26 (1.0 mmol dm23) (data not shown), a facile superoxide radical quencher.The role played by reactive oxidants such as singlet oxygen and hydroxyl radical has been revealed by the following experiments. Fig. 8 0.8% Agarose gel showing the results of electrophoresis of pBR322 plasmid DNA (8.7 mmol dm23) photolysed for 30 min: (a) in the presence of complex 1 at 0 (A), 0.25 (B), 0.50 (C), 1.00 (D), 1.49 (E), 1.98 (F) and 2.45 mmol dm23 (G), 2.45 mmol dm23 in the dark (H); (b) in the presence of 2 at 0 (A), 0.22 (B), 0.43 (C), 0.87 (D), 1.30 (E), 1.73 (F) and 2.17 mmol dm23 (G), 2.17 mmol dm23 in the dark (H); (c) in the absence of complex (A); in the presence of 0.50 mmol dm23 1 (B), in the dark (C), in the presence of histidine (1.20 mmol dm23) (D); D2O used instead of buffer (E); in the presence of mannitol (50.0 mmol dm23) (F), ethanol (1.70 mol dm23) (G), sodium formate (100.0 mmol dm23) (H); (d ) in the absence of complex (A); in the presence of 0.22 mmol dm23 2 (B), in the dark (C), in the presence of histidine (1.20 mmol dm23) (D), D2O used instead of buffer (E); in the presence of mannitol (50.0 mmol dm23) (F), ethanol (1.70 mol dm23) (G), sodium formate (100.0 mmol dm23) (H) In the presence of histidine (1.2 mmol dm23) [lane D, Fig. 8(d )] no noticeable inhibition in the cleavage activity of complex 2 is observed, indicating that singlet oxygen is not likely to be the cleaving agent. Besides, no enhancement but inhibition in the photocleavage of the plasmid DNA is observed for 2 in D2O [lane E, Fig. 8(d )], which further confirms that singlet oxygen is not involved in the cleavage. On the other hand, different degrees of inhibition in the photoinduced cleavage of the plasmid by complex 2 are observed in the presence of hydroxyl radical scavengers such as mannitol (50.0 mmol dm23) [lane F, Fig. 8(d )], ethanol (1.7 mol dm23) [lane G, Fig. 8(d )] and sodium formate (100.0 mmol dm23) [lane H, Fig. 8(d )]. This suggests that, in addition to superoxide anion radical, hydroxyl radical is likely to be one of the reactive species for the cleavage. In conclusion, the complexes bind to the biopolymer by intercalation, with the planar diimine ligands stacked in between the base pairs of the DNA. It is this close proximity which renders cleavage of the DNA effective by the photoactivated complexes. The cleavage mechanism for 1 and pBR322 is likely to be direct oxidation of the DNA by the excited complex.For that between 2 and pBR322, the complex acts as an oxygen sensitizer and superoxide and hydroxyl radicals are likely to be the reactive species responsible for cleavage of the plasmid. Similar observations have also been reported for other ruthenium(II) and cobalt(III) systems.28 Acknowledgements V. W.-W. Y. acknowledges financial support from the Research Grants Council, The Croucher Foundation and The University of Hong Kong, K. K.-W.L. the receipt of a Sir Edward Youde Postgraduate Fellowship, administered by the Sir Edward Youde Memorial Fund Council and a Postgraduate Studentship, administered by The University of Hong Kong. References 1 R. P. Hertzberg and P. B. Dervan, J. Am. Chem. Soc., 1982, 104, 313. 2 (a) B. E. Bowler, K. J. Ahmed, W. I. Sundquist, L. S. Hollis, E. E. Whang and S. J. Lippard, J. Am. Chem. Soc., 1989, 111, 1299; (b) M. V. Keck and S. J. Lippard, J. Am. Chem. Soc., 1992, 114, 3386; (c) T.Ren, D. P. Bancroft, W. I. Sundquist, A. Masschelein, M. V. Keck and S. J. Lippard, J. Am. Chem. Soc., 1993, 115, 11 341. 3 (a) J. K. Barton, J. J. Dannenberg and A. L. Raphael, J. Am. Chem. Soc., 1982, 104, 4967; (b) J. K. Barton and A. L. Raphael, J. Am. Chem. Soc., 1984, 106, 2466; (c) J. K. Barton, A. T. Danishefsky and J. M. Goldberg, J. Am. Chem. Soc., 1984, 106, 2172; (d ) C. V. Kumar, J. K. Barton and N. J. Turro, J. Am. Chem. Soc., 1985, 107, 5518; (e) A. M.Pyle and J. K. Barton, Inorg. Chem., 1987, 26, 3820; ( f ) A. M. Pyle, J. P. Rehmann, R. Meshoyrer, C. V. Kumar, N. J. Turro and J. K. Barton, J. Am. Chem. Soc., 1989, 111, 3051; (g) A. E. Friedman, J. C. Chambron, J. P. Sauvage, N. J. Turro and J. K. Barton, J. Am. Chem. Soc., 1990, 112, 4960; (h) R. M. Hartshorn and J. K. Barton, J. Am. Chem. Soc., 1992, 114, 5919; (i ) A. Sitlani, E. C. Long, A. M. Pyle and J. K. Barton, J. Am. Chem. Soc., 1992, 114, 2303; ( j) A. H. Krotz, L.Y. Kuo, T. P. Shields and J. K. Barton, J. Am. Chem. Soc., 1993, 115, 3877; (k) R. E. Holmlin, E. D. A. Stemp and J. K. Barton, J. Am. Chem. Soc., 1996, 118, 5236. 4 (a) H. H. Thorp, N. J. Turro and H. B. Gray, New J. Chem., 1991, 15, 601; (b) W. A. Kalsbeck, N. Grover and H. H. Thorp, Angew. Chem., Int. Ed. Engl., 1991, 30, 1517; (c) N. Grover and H. H. Thorp, J. Am. Chem. Soc., 1991, 113, 7030; (d) N. Gupta, N. Grover, G. A. Neyhart, W. Liang, P. Singh and H. H. Thorp, Angew.Chem., Int. Ed. Engl., 1992, 31, 1048; (e) N. Grover, N. Gupta, P. Singh and H. H. Thorp, Inorg. Chem., 1992, 31, 2014; ( f ) N. Grover, N. Gupta and H. H. Thorp, J. Am. Chem. Soc., 1992, 114, 3390; (g) N. Gupta, N. Grover, G. A. Neyhart, P. Singh and H. H. Thorp, Inorg. Chem., 1993, 32, 310; (h) G. A. Neyhart, N. Grover, S. R. Smith, W. A. Kalsbeck, T. A. Fairley, M. Cory and H. H. Thorp, J. Am. Chem. Soc., 1993, 115, 4423; (i) W. A. Kalsbeck and H. H. Thorp, J. Am. Chem.Soc., 1993, 115, 7146; ( j ) T. W. Welch, G. A. Neyhart, J. G. Goll, S. A. Ciftan and H. H. Thorp, J. Am. Chem. Soc., 1993, 115, 9311; (k) S. R. Smith, G. A. Neyhart, W. A. Kalsbeck and H. H. Thorp, New J. Chem., 1994, 18, 397; (l) W. A. Kalsbeck and2072 J. Chem. Soc., Dalton Trans., 1997, Pages 2067–2072 H. H. Thorp, Inorg. Chem., 1994, 33, 3427; (m) K. M. Breiner, M. A. Daugherty, T. G. Oas and H. H. Thorp, J. Am. Chem. Soc., 1995, 117, 11 673. 5 R. J. Morgan, S.Chatterjee, A. D. Baker and T. C. Strekas, Inorg. Chem., 1991, 30, 2687; A. D. Baker, R. J. Morgan and T. C. Strekas, J. Am. Chem. Soc., 1991, 113, 1411; J. Chem. Soc., Chem. Commun., 1992, 1099; T. C. Strekas, A. D. Baker, O. Harripersad-Morgan and R. J. Morgan, J. Coord. Chem., 1996, 34, 77. 6 (a) J. M. Kelly, D. J. McConnell, C. OhUigin, A. B. Tossi, A. Kirsch- De Mesmaeker, A. Masschelein and J. Nasielski, J. Chem. Soc., Chem. Commun., 1987, 1821; (b) A. Kirsch-De Mesmaeker, G.Orellana, J. K. Barton and N. J. Turro, Photochem. Photobiol., 1990, 52, 461; (c) G. Orellana, A. Kirsch-De Mesmaeker, J. K. Barton and N. J. Turro, Photochem. Photobiol., 1991, 54, 499; (d ) J. P. Lecomte, A. Kirsch-De Mesmaeker, M. Demeunynck and J. Lhomme, J. Chem. Soc., Faraday Trans., 1993, 3261; (e) M. M. Feeney, J. M. Kelly, A. B. Tossi, A. Kirsch-De Mesmaeker and J. P. Lecomte, J. Photochem. Photobiol. B: Biol., 1994, 23, 69; ( f ) J. P. Lecomte, A. Kirsch-De Mesmaeker and G.Orellana, J. Phys. Chem., 1994, 98, 5382; (g) F. O’Reilly, J. Kelly and A. Kirsch-De Mesmaeker, Chem. Commun., 1996, 1013. 7 C. Hiort, B. Nordén and A. Rodger, J. Am. Chem. Soc., 1990, 112, 1971; M. Eriksson, M. Leijon, C. Hiort, B. Nordén and A. Gräslund, J. Am. Chem. Soc., 1992, 114, 4933; C. Hiort, P. Lincoln and B. Nordén, J. Am. Chem. Soc., 1993, 115, 3448; I. Haq, P. Lincoln, D. Suh, B. Nordén, B. Z. Chowhry and J. B. Chaires, J. Am. Chem. Soc., 1995, 117, 4788; P.Lincoln, A. Broo and B. Nordén, J. Am. Chem. Soc., 1996, 118, 2644; P. Lincoln and B. Nordén, Chem. Commun., 1996, 2145. 8 D. S. Sigman, D. R. Graham, V. D’Aurora and A. M. Stern, J. Biol. Chem., 1979, 254, 12 269; D. R. Graham, L. E. Marshall, K. A. Reich and D. S. Sigman, J. Am. Chem. Soc., 1980, 102, 5419; K. A. Reich, L. E. Marshall, D. R. Graham and D. S. Sigman, J. Am. Chem. Soc., 1981, 103, 3582; D. S. Sigman, Acc. Chem. Res., 1986, 19, 180; T. B. Thederahn, M.D. Kuwabara, T. A. Larsen and D. S. Sigman, J. Am. Chem. Soc., 1989, 111, 4941. 9 (a) R. Tamilarasan, S. Ropartz and D. R. McMillin, Inorg. Chem., 1988, 27, 4082; (b) R. Tamilarasan, S. Ropartz and D. R. McMillin, Inorg. Chem., 1990, 29, 2798; (c) B. P. Hudson, J. Sou, D. J. Berger and D. R. McMillin, J. Am. Chem. Soc., 1992, 114, 8997; (d ) K. A. Meadows, F. Liu, J. Sou, B. P. Hudson and D. R. McMillin, Inorg. Chem., 1993, 32, 2919; (e) K. A. Meadows, F. Liu, B. P. Hudson and D.R. McMillin, Inorg. Chem., 1993, 32, 4663; ( f) F. Liu, K. A. Meadows and D. R. McMillin, J. Am. Chem. Soc., 1993, 115, 6699. 10 (a) H. Q. Liu, S. M. Peng and C. M. Che, J. Chem. Soc., Chem. Commun., 1995, 509; (b) H. Q. Liu, T. C. Cheung, S. M. Peng and C. M. Che, J. Chem. Soc., Chem. Commun., 1995, 1787; (c) H. Q. Liu, T. C. Cheung and C. M. Che, Chem. Commun., 1996, 1039. 11 N. B. Thornton and K. S. Schanze, Inorg. Chem., 1993, 32, 4994; H. D. Stoeffler, N. B. Thornton, S.L. Temkin and K. S. Schanze, J. Am. Chem. Soc., 1995, 117, 7119; N. B. Thornton and K. S. Schanze, New J. Chem., 1996, 20, 791. 12 V. W. W. Yam, K. K. W. Lo, K. K. Cheung and R. Y. C. Kong, J. Chem. Soc., Chem. Commun., 1995, 1191. 13 T. Maniatis, E. F. Fritsch and J. Sambrook, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York, 1982, p. 458. 14 C. V. Kumar and E. H. Asuncion, J. Am. Chem. Soc., 1993, 115, 8547. 15 PATTY and DIRDIF 92: P. T.Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smits and C. Smykalla, The DIRDIF program system, Technical Report of the Crystallography Laboratory, University of Nijmegen, 1992. 16 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, TX, 1985 and 1992. 17 E. Horn and M. R. Snow, Aust. J. Chem., 1980, 33, 2369; W. Tikkanen, W. C. Kaska, S. Moya, T. Layman and R. Kane, Inorg. Chim. Acta, 1983, 76, L29; J. C. Calabrese and W. Tam, Chem. Phys. Lett., 1987, 133, 244; R. Lin, Y. Fu, C. P. Brock and T. F. Guarr, Inorg. Chem., 1992, 31, 4346; T. A. Oriskovich, P. S. White and H. H. Thorp, Inorg. Chem., 1995, 34, 1629; L. Wallace, C. Woods and D. Rillema, Inorg. Chem., 1995, 34, 2875; V. W. W. Yam, V. C. Y. Lau and K. K. Cheung, J. Chem. Soc., Chem. Commun., 1995, 259. 18 (a) L. Stryer, Biochemistry, Freeman, New York, 1988, p. 76; (b) A. H. J. Wang, J. Nathans, G. van der Marel, J. H. van Boom and A. Rich, Nature (London), 1978, 276, 471; (c) A. H. J. Wang, G. J. Quigley and A. Rich, Nucleic Acids Res., 1979, 6, 3879; (d ) T. K Schoch, J. L. Hubbard, C. R. Zoch, G. B. Yi and M. Sørlie, Inorg. Chem., 1996, 35, 4383. 19 V. A. Bloomfield, D. M. Crothers and I. Tinoco, jun., Physical Chemistry of Nucleic Acids, Harper and Row, New York, 1974, p. 432. 20 S. J. Lippard, P. J. Bond, K. C. Wu and W. R. Bauer, Science, 1976, 194, 726; G. Arena, L. M. Scolaro, R. F. Pasternack and R. Romeo, Inorg. Chem., 1995, 34, 2994. 21 A. Wolfe, G. H. Shimer, jun. and T. Meehan, Biochemistry, 1987, 26, 6392. 22 M. Howe-Grant and S. J. Lippard, Biochemistry, 1979, 18, 5762. 23 J. D. McGhee and P. H. von Hippel, J. Mol. Biol., 1974, 86, 469. 24 (a) R. F. Pasternack, E. J. Gibbs and J. J. Villafranca, Biochemistry, 1983, 22, 2466; (b) A. B. Tossi and J. M. Kelly, Photochem. Photobiol., 1989, 49, 545; (c) A. Sarai, J. Mazur, R. Nussinov and R. L. Jernigan, Biochemistry, 1969, 28, 7842. 25 R. Nilsson, P. B. Merkel and D. R. Kearns, Photochem. Photobiol., 1972, 16, 117. 26 C. C. Cheng, S. E. Rokita and C. J. Burrows, Angew. Chem., Int. Ed. Engl., 1993, 32, 277; S. A. Lesko, R. J. Lorentzen and P. O. Ts’o, Biochemistry, 1980, 19, 3023. 27 A. U. Khan, J. Phys. Chem., 1976, 80, 2219. 28 M. B. Fleisher, K. C. Waterman, N. J. Turro and J. K. Barton, Inorg. Chem., 1986, 25, 3549. Received 5th February 1997; Paper 7/00828G

ISSN:1477-9226    

DOI:10.1039/a700828g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2071–2076 2071 Platinum metal ditelluroether complexes: synthesis, spectroscopic and structural studies of [M(L–L)2][PF6]2 [M 5 Pd or Pt, L–L 5 RTe(CH2)3TeR (R 5 Me or Ph) or C6H4(TeMe)2-o], [Rh(L–L)2Cl2]PF6, [Ru(L–L)2X2] (X 5 Cl, Br or I) and [Ru(L–L)2(PPh3)Cl]PF6 William Levason, Simon D. Orchard, Gillian Reid and Vicki-Ann Tolhurst Department of Chemistry, University of Southampton, Southampton, UK SO17 1BJ Received 11th March 1999, Accepted 7th May 1999 The reaction of [MCl2(MeCN)2] (M = Pd or Pt) with L–L (RTe(CH2)3TeR (R = Me or Ph) or C6H4(TeMe)2-o) and TlPF6 in a 1 : 2 : 2 ratio in MeCN gave planar [M(L–L)2][PF6]2 which was confirmed by an X-ray crystallographic study of [Pd{meso-C6H4(TeMe)2-o}2][PF6]2.The complexes trans-[Ru(L–L)2X2] (X = Cl, Br or I) were prepared from [Ru(dmf)6][CF3SO3]3, L–L and LiX, whilst [Ru(PPh3)3Cl2], L–L and NH4PF6 gave trans-[Ru(L–L)2(PPh3)Cl]PF6. The crystal structure of [Ru{MeTe(CH2)3TeMe}2(PPh3)Cl][PF6] revealed one ditelluroether in the meso form and the other in the DL, the first time both stereoisomers have been crystallographically identified in the same compound.In contrast in trans-[Ru{PhTe(CH2)3TePh}2Cl2] both ditelluroethers are meso forms. For Pd, Pt and Ru multinuclear NMR spectroscopy (1H, 125Te-{1H}, 195Pt) showed a mixture of stereoisomeric forms in solution, and for the compounds of Pd and Pt pyramidal inversion at Te is fast on the NMR timescales at ambient temperatures.Reaction of RhCl3?3H2O, L–L and NH4PF6 gave [Rh(L–L)2Cl2]PF6 which appear to be predominantly trans isomers based upon their NMR and UV/vis spectra. Attempts to oxidise the complexes of PtII or RuII (to PtIV or RuIII) with halogens or electrochemically were unsuccessful, contrasting with the successful oxidation of analogous complexes with dithioether or diselenoether ligands. Introduction Chelating ditelluroethers including RTe(CH2)3TeR (R = Me or Ph) and C6H4(TeMe)2-o (L–L) were first reported about 10 years ago,1,2 and a range of platinum metal halide complexes with a 1 : 1 metal : ditelluroether ratio were subsequently characterised, including [M(L–L)X2] (M = Pd or Pt, X = Cl, Br or I),3,4 [Ir(L–L)X4]2,4 and [{Ir(L–L)Cl3}n].5 Far fewer complexes with a 2 : 1 ditelluroether : metal ratio are known, examples being limited to some unstable cobalt(III) complexes 6 and homoleptic compounds of CuI and AgI.7 More recently, ditelluroether complexes of Group 6 metal carbonyls and Group 7 carbonyl halides have been obtained.8 However it remains true that far less is known about tellurium donor ligands than their thio- or seleno-ether analogues. Here we report the synthesis of some complexes of Pd, Pt, Rh and Ru with a 2 : 1 ditelluroether :M ratio, their spectroscopic and structural characterisation, and an exploration of the redox chemistry of the platinum and ruthenium complexes.The ligand properties of the ditelluroethers are compared with those of Se, S and Group 15 donor analogues. Results and discussion Palladium and platinum The reaction of [MCl2(MeCN)2] (M = Pd or Pt) with a large excess of the ditelluroether aVords only the [M(L–L)Cl2] complexes. 3,4 However, if the halide is removed with TlPF6 the products are the yellow or orange [M(L–L)2][PF6]2 (L–L = RTe(CH2)3TeR, R = Me or Ph, or C6H4(TeMe)2-o). Coordinated ditelluroethers exist as two diastereoisomers, meso (with syn R groups) and DL (anti R groups).9 Proton and especially 125Te-{1H} NMR spectroscopies have proved very useful in assigning structures to many ditelluroether complexes,3–8 but are rather less useful for the bis(ditelluroether) complexes in the present study.The possible combinations of meso or DL ditelluroethers for planar M(L–L)2 or trans M(L–L)2X2 moieties result in five possible isomers (invertomers), containing eight distinct tellurium centres, although all possible isomers need not be present in significant amounts.For lower symmetry trans M(L–L)2XY or cis M(L–L)2X2 even more resonances are predicted. The isomers interconvert by pyramidal inversion at Te, a process whose energy depends upon the metal centre present, the ligand structure, chelate ring size, and ligands trans to Te.9 The complexity of the spectra and in some cases the consequences of the onset of pyramidal inversion means that assignment of resonances to particular invertomers is not possible, although usually the geometric isomer(s) present can be identified.For the planar [M(L–L)2]21 (M = Pd or Pt) the Te-trans-Te arrangement lowers inversion barriers due to the high trans influence of tellurium, and at room temperature the 1H NMR spectra of all the complexes show broad features, sometimes with ill defined splittings, showing that inversion processes are present. Similarly at 300 K the 125Te-{1H} NMR spectra show very broad features typical of systems near to coalescence.On cooling the spectra sharpen and for the platinum complexes individual resonances resolve, but even at 210 K inversion still leads to significant broadening and 195Pt satellites are not resolved. Consistent with this, none of the platinum complexes exhibited a 195Pt NMR spectrum at ambient temperatures, but on cooling a solution of [Pt{C6H4(TeMe)2-o}2]21 to 210 K broad resonances appeared at d 24790 and 24760, which are entirely reasonable shifts for a PtIITe4 centre.The structure of [Pd{C6H4(TeMe)2-o}2][PF6]2?MeCN reveals a square planar cation with the Pd on an inversion centre, coordinated to two meso ditelluroether ligands (Fig. 1, Table 1). The Te–Pd–Te angles are very close to 908, with Pd–Te 2.5716(4)–2.5789(5) Å, markedly longer than Pd–Te in2072 J. Chem. Soc., Dalton Trans., 1999, 2071–2076 [Pd{PhTe(CH2)3TePh}Br2] 3 (2.528(1), 2.525(1) Å) and consistent with the relative trans influence Te > Br.Attempts to oxidise the [Pt(L–L)2]21 to PtIV using Cl2 were unsuccessful, causing decomposition of the complexes, and treatment of either the palladium or platinum cations with LiCl in MeCN resulted in displacement of one ditelluroether and the formation of the corresponding [M(L–L)Cl2].3,4 Ruthenium Direct reaction of the ditelluroethers with RuCl3?nH2O proved generally unsatisfactory, although one example trans-[Ru{C6H4- (TeMe)2-o}2Cl2] has been obtained by this route.10 Entry into the ruthenium chemistry was achieved by reaction of the ligands with [Ru(PPh3)3Cl2] in the presence of NH4PF6, [Ru- (dmso)4Cl2] or [Ru(dmf)6][CF3SO3]3.11 The first reaction gave trans-[Ru(L–L)2(PPh3)Cl]PF6 as orange-brown powders, whilst use of [Ru(dmso)4Cl2] aVorded trans-[Ru(L–L)2Cl2].A more general route to trans-[Ru(L–L)2X2] (X = Cl, Br or I) was reaction of [Ru(dmf)6][CF3SO3]3 with L–L and LiX in EtOH. The ruthenium(II) complexes are air-stable in the solid state; solubility in organic solvents decreases in the order Cl > Br > I with the iodides in particular very poorly soluble in chlorocarbons or MeCN, which limited solution spectroscopic studies.The structure determination of a yellow crystal of [Ru{Me- Te(CH2)3TeMe}2(PPh3)Cl]PF6 revealed a trans cation with one meso and one DL form of the ditelluroether (Fig. 2, Table 2), the first time both forms have been identified crystallographically in one complex.The d(Ru–P) 2.304(4) Å and d(Ru–Cl) 2.467(4) Å are similar to those found12 in trans-[Ru{[16]aneSe4}- (PPh3)Cl]PF6 ([16]aneSe4 = 1,5,9,13-tetraselenacyclohexadecane), 2.307(6), 2.499(5) Å respectively. The d(Ru–Te) lie in the range 2.636(1)–2.655(1) Å, the first examples of RuII–Te bond lengths reported. The 31P-{1H} NMR of this complex shows a strong resonance at d 51.5 and the corresponding 125Te- {1H} NMR has four major resonances of similar intensities at d 177, 262, 274 and 371 which show doublet splittings of 30–60 Fig. 1 View of the structure of one of the two independent [Pd{C6H4- (TeMe)2-o}2]21 cations with numbering scheme adopted (the other cation is essentially indistinguishable). Atoms marked * are related by a centre of inversion at Pd. Ellipsoids are drawn at 40% probability. Table 1 Selected bond lengths (Å) and angles (8) for [Pd{C6H4- (TeMe)2-o}2][PF6]2?MeCN Te(1)–Pd(1) Te(1)–C(4) Te(2)–C(5) Te(3)–Pd(2) Te(3)–C(13) Te(4)–C(10) Pd(1)–Te(1)–C(3) C(3)–Te(1)–C(4) Pd(1)–Te(2)–C(6) Pd(2)–Te(3)–C(12) C(12)–Te(3)–C(13) Pd(2)–Te(4)–C(11) Te(1)–Pd(1)–Te(2) 2.5716(4) 2.132(8) 2.112(9) 2.5732(5) 2.120(7) 2.118(7) 102.3(2) 93.6(3) 102.4(2) 99.4(3) 92.8(3) 99.7(2) 89.98(1) Te(1)–C(3) Te(2)–Pd(1) Te(2)–C(6) Te(3)–C(12) Te(4)–Pd(2) Te(4)–C(11) Pd(1)–Te(1)–C(4) Pd(1)–Te(2)–C(5) C(5)–Te(2)–C(6) Pd(2)–Te(3)–C(13) Pd(2)–Te(4)–C(10) C(10)–Te(4)–C(11) Te(3)–Pd(2)–Te(4) 2.120(7) 2.5789(5) 2.116(7) 2.112(9) 2.5781(5) 2.136(8) 98.3(2) 100.5(2) 92.8(3) 103.2(2) 102.7(2) 93.6(3) 89.67(1) Hz attributable to 2J(31P–125Te).Much weaker peaks at 53.1 (31P) and 249 (125Te) are assigned to a second isomer. The spectra are consistent with the form identified in the solid state being the major invertomer in solution: although there is no requirement that the form in the crystal should be the major solution form, studies of several diselenoether systems have shown this is often true in practice.13 The trans-[Ru{PhTe- (CH2)3TePh}2(PPh3)Cl]PF6 shows two major 31P NMR resonances of approximately equal intensity (d 44.8, 46.0) and five 125Te resonances in the range d 500–550, suggesting two invertomers are present in significant amounts.In the case of trans- [Ru{C6H4(TeMe)2-o}2(PPh3)Cl]PF6 two 31P resonances (d 52.0, 44.8) are associated with five 125Te resonances (d 870–785), one of which (d 842) is very much more intense than the rest, suggesting this is an invertomer with both ligands in the meso form.This complex also decomposes slowly on standing in CH2Cl2 solution in air, developing new 31P resonances at d 26.0 and 27.2. This is consistent with the co-ordinated PPh3 being oxidised to OPPh3, a reaction observed in other ruthenium complexes.14 Cyclic voltammetry revealed that the complexes undergo irreversible oxidation at ca. 1.2 V in CH2Cl2 solution (vs. Fc– Fc1 at 0.49 V), showing that RuIII is not stable when surrounded by tellurium donors (Te4PCl donor set). Fig. 2 View of the structure of [Ru{MeTe(CH2)3TeMe}2(PPh3)Cl]1 with numbering scheme adopted. Ellipsoids are drawn at 40% probability. Table 2 Selected bond lengths (Å) and angles (8) for [Ru{MeTe- (CH2)3TeMe}2(PPh3)Cl]PF6 Te(11)–Ru(1) Te(11)–C(12) Te(12)–C(14) Te(21)–Ru(1) Te(21)–C(22) Te(22)–C(24) Ru(1)–Cl(1) Ru(1)–Te(11)–C(11) C(11)–Te(11)–C(12) Ru(1)–Te(12)–C(15) Ru(1)–Te(21)–C(21) C(21)–Te(21)–C(22) Ru(1)–Te(22)–C(25) Te(11)–Ru(1)–Te(12) Te(11)–Ru(1)–Te(22) Te(11)–Ru(1)–P(1) Te(12)–Ru(1)–Te(22) Te(12)–Ru(1)–P(1) Te(21)–Ru(1)–Cl(1) Te(22)–Ru(1)–Cl(1) Cl(1)–Ru(1)–P(1) 2.650(1) 2.14(1) 2.18(1) 2.647(1) 2.17(1) 2.12(1) 2.467(4) 105.2(5) 90.7(6) 108.0(4) 114.3(4) 91.8(6) 109.2(4) 89.00(4) 177.59(5) 89.61(9) 89.96(4) 92.09(9) 78.21(9) 90.35(9) 176.3(1) Te(11)–C(11) Te(12)–Ru(1) Te(12)–C(15) Te(21)–C(21) Te(22)–Ru(1) Te(22)–C(25) Ru(1)–P(1) Ru(1)–Te(11)–C(12) Ru(1)–Te(12)–C(14) C(14)–Te(12)–C(15) Ru(1)–Te(21)–C(22) Ru(1)–Te(22)–C(24) C(24)–Te(22)–C(25) Te(11)–Ru(1)–Te(21) Te(11)–Ru(1)–Cl(1) Te(12)–Ru(1)–Te(21) Te(12)–Ru(1)–Cl(1) Te(21)–Ru(1)–Te(22) Te(21)–Ru(1)–P(1) Te(22)–Ru(1)–P(1) 2.13(1) 2.655(1) 2.16(2) 2.16(1) 2.636(1) 2.12(1) 2.304(4) 101.2(3) 107.0(4) 96.2(6) 109.0(4) 106.9(4) 91.6(6) 92.25(4) 87.47(9) 168.15(5) 90.09(9) 88.33(4) 99.69(9) 92.60(9)J.Chem. Soc., Dalton Trans., 1999, 2071–2076 2073 The orange [Ru(L–L)2Cl2] are air-stable solids, moderately soluble in chlorocarbons and poorly soluble in MeCN.The crystal structure of triclinic crystals of [Ru{PhTe(CH2)3TePh}2- Cl2] (Fig. 3, Table 3) revealed a trans geometry with the Ru on an inversion centre and the ditelluroether ligands have meso arrangements. The d(Ru–Cl) 2.4389(13) Å is shorter than that in the chlorophosphine complex (above), consistent with the expected trans influence order P > Cl, but is similar to those in other trans Cl–RuII–Cl arrangements in for example trans-[Ru(PhSeCH2CH2SePh)2Cl2] (2.413(1), 2.444(1) Å) 10 or trans-[Ru{H2C]] C(PPh2)2}2Cl2] (2.431(1) Å).15 The d(Ru–Te) (2.6194(3), 2.6247(3) Å) are slightly shorter than in the chlorophosphine.The 1H and 125Te-{1H} NMR spectra show that several diastereoisomers are present in solution, in particular the latter has a strong resonance at d 638 due to a high symmetry form with all four tellurium sites identical (possibly the form present in the crystal), and six weaker resonances. The UV/vis spectra show two bands in the region 18 000–25 000 cm21 as expected for low spin d6 metal centres in local D4h symmetry.16 The characterisation of [Ru(L–L)2Cl2] (L–L = C6H4(TeMe)2-o or MeTe(CH2)3TeMe) and the corresponding bromo- and iodo-complexes was achieved by analysis and ES1 mass spectrometry, and their assignment as trans isomers follows from the similar spectra to those of the dichlorocompounds.Multinuclear NMR spectroscopy was less successful due to their poor solubility, and convincing 125Te NMR spectra were not observed from saturated CH2Cl2 solutions of the bromo- or iodo-complexes after ca. 2 × 106 scans. Cyclic voltammetry was used to study the oxidation of these complexes. In CH2Cl2 containing 0.1 mol dm23 NBun 4BF4 freshly prepared solutions of the dichloro-complexes showed irreversible or quasi-reversible oxidations at ca. 10.5–0.8 V (vs. Fc–Fc1 at 0.49 V), the waves diminishing rapidly on repetitive scans consistent with deposition of material onto the electrodes.This behaviour contrasts with [Ru(L–L)2Cl2] (L–L = various dithioether or diselenoether ligands) 10,17 which show reversible RuII–RuIII couples at less positive potentials, and for which the ruthenium(III) complexes can be isolated by halogen oxidation of the corresponding ruthenium(II) complex in CH2Cl2. Treatment of CH2Cl2 solutions of [Ru(ditelluroether)2Cl2] with Cl2 gave dark brownish solutions, which rapidly decomposed. Our Fig. 3 View of the structure of [Ru{PhTe(CH2)3TePh}2Cl2] with numbering scheme adopted.Atoms marked * are related by a centre of inversion at Ru. Ellipsoids are drawn at 40% probability. Table 3 Selected bond lengths (Å) and angles (8) for [Ru{PhTe- (CH2)3TePh}2Cl2] Te(1)–C(11) Te(1)–Ru(1) Te(2)–C(3) Ru(1)–Cl(1) C(11)–Te(1)–C(1) C(1)–Te(1)–Ru(1) C(21)–Te(2)–Ru(1) Cl(1)–Ru(1)–Te(2) Te(2)–Ru(1)–Te(1) 2.142(5) 2.6247(3) 2.150(6) 2.4389(13) 89.9(2) 105.15(17) 110.93(14) 87.62(3) 87.255(11) Te(1)–C(1) Te(2)–C(21) Te(2)–Ru(1) C(11)–Te(1)–Ru(1) C(21)–Te(2)–C(3) C(3)–Te(2)–Ru(1) Cl(1)–Ru(1)–Te(1) 2.153(6) 2.147(6) 2.6194(3) 111.38(15) 91.2(2) 104.46(16) 87.71(3) inability to isolate ruthenium(III) ditelluroether complexes is a further example of the reduced ability of these ligands to stabilise higher oxidation states of transition metals compared to their lighter analogues. Phosphorus or arsenic donor ligands form quite stable ruthenium(III) complexes as trans- [Ru(bidentate ligand)2X2]1 cations, and most exhibit electrochemically reversible RuII–RuIII couples.17 Rhodium The reaction of RhCl3?3H2O, L–L and NH4PF6 in EtOH gave deep orange [Rh(L–L)2Cl2]PF6.The ES1 mass spectra show ion multiplets with the correct isotope patterns for [Rh(L–L)2Cl2]1, and the presence of PF6 2 anions is confirmed by the IR spectra which show the expected bands at ca. 840 (n(P–F)) and ca. 560 cm21 (d(PF2)). We have been unable to grow crystals of any of these rhodium compounds, either with PF6 2 or ClO4 2 anions, suitable for an X-ray study.The UV/vis spectra each show a d–d band at ca. 21 000 cm21 which is consistent with trans geometric isomers (cis isomers have d–d bands at higher energy, often obscured by charge-transfer transitions).18 The 1H NMR spectra are complex revealing a number of invertomers are present. Poor solubility made recording the 125Te-{1H} NMR spectra diYcult, but after very long accumulations the spectrum of [Rh{MeTe(CH2)3TeMe}2Cl2]1 showed a number of doublets with d 250–300 and with the 1J(125Te–103Rh) couplings of 50–70 Hz.The spectrum of [Rh{PhTe(CH2)3TePh}2Cl2]1 was similar. However, for [Rh{C6H4(TeMe)2-o}2Cl2]1 the 125Te- {1H} NMR spectrum showed, in addition to seven major doublets, some evidence for a number of weaker features which may be due to the second geometric isomer (cis). It is notable that, for dithioether or diselenoether (L9–L9) ligands, direct reaction with RhCl3?3H2O leads to mixtures of trans- and cis- [Rh(L9–L9)2Cl2]Cl and [{Rh(L9–L9)Cl3}n] 18 and pure trans isomers were obtained by oxidative addition to rhodium(I) analogues.18,19 Experimental Infrared spectra were measured as CsI discs using a Perkin- Elmer 983 spectrometer over the range 200–4000 cm21, UV/vis spectra were recorded in solution using 1 cm path length quartz cells on a Perkin-Elmer Lambda19 spectrometer.Mass spectra were run by fast-atom bombardment (FAB) using 3-nitrobenzyl alcohol as matrix on a VG Analytical 70-250-SE Normal Geometry Double Focusing Mass Spectrometer or by positive electrospray (ES1) using a VG Biotech Platform.The 1H NMR spectra were recorded using a Bruker AM300 spectrometer operating at 300 MHz, 125Te-{1H}, 31P-{1H} and 195Pt NMR spectra on a Bruker AM360 spectrometer operating at 113.6, 145.8 and 77.4 MHz respectively and referenced to neat Me2Te, 85% H3PO4 and aqueous [PtCl6]22. Microanalyses were performed by the microanalytical service of Strathclyde University.Electrochemical studies used an Eco Chemi PGstat20 with Pt working and auxiliary electrodes. Preparations [Pd{MeTe(CH2)3TeMe}2][PF6]2. The complex [PdCl2- (MeCN)2] (82 mg, 0.32 mmol) and TlPF6 (226 mg, 0.65 mmol) were stirred in MeCN (30 cm3) for 30 min under a dinitrogen atmosphere. The compound MeTe(CH2)3TeMe (221 mg, 0.68 mmol) in MeCN (5 cm3) was then added and the reaction stirred at room temperature for 18 h to give a yellow solution and fine white precipitate (TlCl). The solution was filtered to remove the TlCl, reduced to ca. 2 cm3 in vacuo and diethyl ether (10 cm3) added to precipitate a yellow solid. Yield: 220 mg, 65% (Found: C, 11.4; H, 2.3. C10H24F12P2PdTe4 requires C, 11.4; H, 2.3%). 1H NMR (CD3CN): (CH2CH2) d 2.3 (br) [1H]; (TeCH3) 2.4 (s) [3H]; and (CH2Te) 2.9 (m) [2H]. 125Te-{1H} NMR (Me2CO–CDCl3, 300 K): d 198–219 (vbr). IR/cm21: 2957w,2074 J. Chem. Soc., Dalton Trans., 1999, 2071–2076 2924w, 1425w, 1357s, 1279s, 1212w, 1095m, 987m, 832s, 739w, 709m, 613w and 556s.UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 26 660 (19 800). The following complexes were prepared similarly. [Pd{PhTe(CH2)3TePh}2][PF6]2. Yield 70% (Found: C, 27.9; H, 2.3. C30H32F12P2PdTe4 requires C, 27.7; H, 2.5%). 1H NMR (CD3CN): (CH2CH2) d 2.6 (br) [1H]; (CH2Te) 3.1 (br) [2H]; and (TePh) 7.58 (m) [5H]. 125Te-{1H} NMR (Me2CO–CDCl3, 300 K): d ca. 485, 580 and 605. IR/cm21: 1570w, 1470w, 1432w, 1357s, 1260w, 1210w, 1093s, 1018w, 996m, 840s, 728m, 686m, 615w, 557s and 452w.UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 26 380 (33 300). [Pd{C6H4(TeMe)2-o}2][PF6]2. Yield 80% (Found: C, 17.6; H, 1.8. C16H20F12P2PdTe4 requires C, 17.2; H, 1.8%). 1H NMR (CD3CN): (TeCH3) d 2.6 (s) [3H] and (C6H4) 7.8 (m) [2H]. 125Te- {1H} (Me2CO–CDCl3, 300 K): d 770–825 (br). IR/cm21: 1356s, 1093s, 985m, 834s, 756m, 613w and 556m. UV/vis (MeCN)/ cm21 (emol/dm3 mol21 cm21): 26 600 (23 900). [Pt{MeTe(CH2)3TeMe}2][PF6]2.Yield 80% (Found: C, 11.0; H, 2.0. C10H24F12P2PtTe4 requires C, 10.5; H, 2.1%). 1H NMR (CD3CN): (CH2CH2) d 2.24 (q) [1H]; (TeCH3) 2.43 (s) [3H]; and (CH2Te) 3.02 (t) [2H]. 125Te-{1H} NMR (Me2CO–CDCl3, 300 K): d 195, 196, 200, 201 and 202. IR/cm21: 2922w, 2853w, 1357s, 1262vw, 1228w, 1205w, 1092s, 986w, 834s, 613w and 557s. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 30 770 (sh) (8300). [Pt{PhTe(CH2)3TePh}2][PF6]2. Yield 76% (Found: C, 26.0; H, 2.1.C30H32F12P2PtTe4 requires C, 26.0; H, 2.3%). 1H NMR (CD3CN): (CH2CH2) d 2.5 (br) [1H]; (CH2Te) 3.0 (br) [1H]; 3.26 (br) [1H]; and (TePh) 7.58 (m) [5H]. 125Te-{1H} NMR (Me2CO– CDCl3, 300 K): d 570–580 (br). IR/cm21: 3070w, 1569w, 1471w, 1357m, 1210w, 1093m, 1015w, 996m, 838s, 732m, 689m, 613w, 557s and 453w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 29 950 (8400) and 33 330 (16 600). [Pt{C6H4(TeMe)2-o}2][PF6]2. Yield 75% (Found: C, 15.9; H, 1.6. C16H20F12P2PtTe4 requires C, 15.9; H, 1.7%). 1H NMR (CD3CN): (TeCH3) d 2.6 (m) [3H]; and (C6H4) 7.84 (m) [2H]. 125Te-{1H} NMR (MeCN–CD3CN, 300 K): d 692 and 720 (m). 195Pt-{1H} (Me2CO–CDCl3, 210 K): d 24790 and 24760. IR/cm21: 1357s, 1261w, 1092s, 987m, 839s, 743m, 613m and 557s. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 32 800 (14 700). [Ru{MeTe(CH2)3TeMe}2(PPh3)Cl][PF6]. The complex [RuCl2(PPh3)3] (372 mg, 0.39 mmol) and MeTe(CH2)3TeMe (274 mg, 0.84 mmol) were refluxed for 3 h under a dinitrogen atmosphere in EtOH (40 cm3).The orange solution was cooled to room temperature after which NH4PF6 (221 mg, 1.36 mmol) was added. A light orange precipitate was formed immediately. The reaction mixture was refluxed for 10 min. After cooling, the solution was reduced to ca. 2 cm3 in vacuo and the yelloworange precipitate collected. Yield 428 mg, 92% (Found: C, 28.0; H, 3.3. C28H39ClF6P2RuTe4 requires C, 28.1; H, 3.3%). 1H NMR (CDCl3): d 1.2–2.6 (CH2 1 CH3) and 7.4–7.6 (Ph). 31P- {1H} NMR (CDCl3–CH2Cl2): d 51.5 (br, s), 53.1 (PPh3) and 2143 (septet, PF6). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 177 (2), 249 (2J(31P–125Te) = 50), 262 (2J = 30), 274 (2J = 60) and 371 (2J = 55 Hz). ES1 (MeCN): m/z = 1055, 833 and 792; calc. for [102Ru{Me130Te(CH2)3 130TeMe}2 35Cl(PPh3)]1 1053, [102Ru{Me- 130Te(CH2)3 130TeMe}2 35Cl 1 MeCN]1 842 and [102Ru{Me- 130Te(CH2)3 130TeMe}2 35Cl]1 801. IR/cm21: 1581w, 1478w, 1430m, 1412m, 1357m, 1274w, 1219w, 1086m, 995w, 960w, 840s, 753w, 704m, 614w, 557s, 516s, 480w, 423w, 279w, 246w and 204w.UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 25 200 (630) and 37 740 (22 440). [Ru{PhTe(CH2)3TePh}2(PPh3)Cl][PF6]. Yield 89% (Found: C, 39.9; H, 3.4. C48H47Cl2F6P2RuTe4 requires C, 39.8; H, 3.3%). 1H NMR (CDCl3): d 7.88–6.25 (Ph) and 2.96–2.23 (CH2). 31P-{1H} NMR (CDCl3–CH2Cl2): d 44.8, 46.0 (PPh3) and 2145 (septet, PF6). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 500, 508.5, 511.5, 525.5 and 544 (2J(31P–125Te) couplings poorly resolved). ES1 (MeCN): m/z = 933, 893 and 852; calc.for [102Ru{Ph130Te- (CH2)3 130TePh}35Cl(PPh3) 1 2MeCN]1 937, [102Ru{Ph130Te- (CH2)3 130TePh}35Cl(PPh3) 1 MeCN]1 896 and [102Ru{Ph- 130Te(CH2)3 130TePh}35Cl(PPh3)]1 855. IR/cm21: 1576w, 1479m, 1435m, 1358m, 1261w, 1085m, 1018w, 997w, 835s, 732m, 690m, 555m, 530m, 450w, 428w, 280w and 256w. UV/vis (MeCN)/ cm21 (emol/dm3 mol21 cm21): 23 000 (300), 26 200 (1055) and 32 500 (6490). [Ru{C6H4(TeMe)2-o}2(PPh3)Cl][PF6]. Yield 37% (Found: C, 32.0; H, 2.5.C34H35ClF6P2RuTe4 requires C, 32.2; H, 2.8%). 1H NMR (CDCl3): d 1.95–2.5 (m, CH3) and 7.3–8.0 (m, C6H4). 31P-{1H} NMR (CDCl3–CH2Cl2): d 44.8, 52.0 (PPh3) and 2143 (septet, PF6). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 785, 797, 842 (major), 854 and 870 (2J(31P–125Te) couplings of ca. 30–55 Hz poorly resolved). ES1 (MeCN): m/z = 1123, 902 and 862; calc. for [102Ru{C6H4(130TeMe)2-o}2 35Cl(PPh3)]1 1131, [102Ru- {C6H4(130TeMe)2}2 35Cl 1 MeCN]1 910 and [102Ru{C6H4- (130TeMe)2}2 35Cl]1 869.IR/cm21 1478m, 1431m, 1408m, 1356m, 1256w, 1217w, 1087s, 1024w, 998w, 957w, 840s, 747m, 697m, 615w, 557m, 526m, 462w, 428w, 327w, 298w, 216w and 201w. [Ru{MeTe(CH2)3TeMe}2Cl2]. Method 1. The complex [Ru- (dmf)6][CF3SO3]3 (151 mg, 0.155 mmol), MeTe(CH2)3TeMe (100 mg, 0.307 mmol) and LiCl (42 mg, 0.991 mmol) in EtOH (70 cm3) were refluxed for 4 h. The solvent was removed in vacuo and CH2Cl2 added (2 cm3). The orange solution was filtered and Et2O added to precipitate an orange solid.Yield: 40 mg, 32% (Found: C, 13.9; H, 3.1. C10H24Cl2RuTe4 requires C, 14.5; H, 2.9%). 1H NMR (CDCl3): d 2.0–3.32 (m). ES1 (MeCN): m/z = 834, 793 and 754; calc. for [102Ru{Me130Te(CH2)3 130TeMe}2- 35Cl 1 MeCN]1 842, [102Ru{Me130Te(CH2)3 130TeMe}2 35Cl]1 801 and [102Ru{Me130Te(CH2)3 130TeMe}2]1 766. IR/cm21: 2918w, 1356s, 1270m, 1227w, 1151m, 1090m, 1028m, 834w, 636m, 516w and 216w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 21 000 (180), 24 600 (540) and 38 300 (18 235).Method 2. The complex [Ru(dmso)4Cl2] (71 mg, 0.129 mmol) and MeTe(CH2)3TeMe (87 mg, 0.267 mmol) in MeOH were refluxed for 3 h. The solvent was reduced to ca. 1 cm3 in vacuo and Et2O added to aVord a light orange precipitate. Yield 84 mg, 78%. Unless indicated otherwise, the ruthenium complexes below were made by method 1 using the appropriate LiX. [Ru{MeTe(CH2)3TeMe}2Br2]. Yield 49% (Found: C, 12.7; H, 2.5. C10H24Br2RuTe4 requires C, 13.1; H, 2.6%). 1H NMR (CDCl3): d 1.98–2.93 (CH2, CH3).ES1 (MeCN): m/z = 878 and 839; calc. for [102Ru{Me130Te(CH2)3 130TeMe}2 79Br 1 MeCN]1 886 and [102Ru{Me130Te(CH2)3 130TeMe}2 79Br]1 845. IR/cm21 2918w, 1357s, 1272m, 1092s, 1028m, 987w, 834m, 636m, 534w, 248w and 229w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21) 19 200 (570), 24 700 (1575) and 36 250 (1978). [Ru{MeTe(CH2)3TeMe}2I2]. Yield 16% (Found: C, 11.3; H, 2.5. C10H24I2RuTe4 requires C, 11.9; H, 2.4%). 1H NMR (CDCl3): d 2.02–2.32 (CH2, CH3). ES1 (MeCN): m/z = 1053 and 884; calc.for [102Ru{Me130Te(CH2)3 130TeMe}2I 1 MeCN]1 1061 and [102Ru{Me130Te(CH2)3 130TeMe}2I]1 893. IR/cm21: 2918w, 1358s, 1092m, 834m, 614w, 511w and 217w. [Ru{PhTe(CH2)3TePh}2Cl2]. From [Ru(dmso)4Cl2] as above (82%) (Found: C, 33.0; H, 3.0. C30H32Cl2RuTe4 requires C, 33.5;J. Chem. Soc., Dalton Trans., 1999, 2071–2076 2075 Table 4 Crystallographic data collection and refinement parameters [Pd{C6H4(TeMe)2-o}2]- [PF6]2?MeCN [Ru{PhTe(CH2)3- TePh}2Cl2] [Ru{MeTe(CH2)3TeMe}2- (PPh3)Cl]PF6 Formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z Dc/g cm23 m(Mo-Ka)/cm21 Absorption correction (maximum and minimum transmission factors) Unique obs.reflections Obs. reflections with [Io > 2s(Io)] No. parameters RR 9 C18H23F12NPdTe4 1160.11 Triclinic P1� 8.9645(6) 18.896(2) 8.9325(5) 94.536(6) 95.649(5) 99.533(7) 1478.0(2) 2 2.607 46.94 — 5223 4284 355 0.031 0.035 C30H32Cl2RuTe4 592.22 Triclinic P1� 10.3523(4) 10.7860(4) 11.1139(5) 114.60(4) 107.14(4) 102.42(4) 993.33(7) 1 2.203 38.35 0.859, 0.458 5594 206 0.050 a 0.118 a C28H39ClF6P2RuTe4 1198.48 Monoclinic P21/c 9.296(3) 30.811(2) 12.732(3) 105.40(2) 3515(1) 4 2.264 39.18 1.000, 0.865 6316 3723 379 0.045 0.051 a R1, wR2 (I > 2s(I )) 0.076, 0.013 (all data).H, 3.0%). 1H NMR (CDCl3): d 1.86–3.75 (12 H, m, CH2) and 6.91–7.92 (10 H, m, C6H5). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 545.6, 587.7, 592.6, 606.9, 637.6, 644.5 and 653.4. ES1 (MeCN): m/z = 1038; calc.for [102Ru{Ph130Te(CH2)3 130TePh}2- 35Cl]1 1049. IR/cm21: 3049w, 2929w, 1572m, 1474m, 1432s, 1413m, 1359s, 1261w, 1195w, 1098m, 1018m, 997m, 875w, 793w, 727s, 690s, 530w, 453m, 305w, 258w and 225w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 20 600 (240) and 24 300 (750). [Ru{PhTe(CH2)3TePh}2Br2]. Brown powder. Yield: 21% (Found: C, 31.3; H, 2.9. C30H32Br2RuTe4 requires C, 31.0; H, 2.8%). 1H NMR (CDCl3): d 2.65–3.93 (12 H, m, CH2) and 6.91–7.87 (20 H, m, C6H5). ES1 (MeCN): m/z = 1083; calc.for [102Ru{Ph130Te(CH2)3 130TePh}2 79Br]1 1093. IR/cm21: 1473w, 1433m, 1358m, 1198w, 1061m, 998w, 740m, 689m, 671m, 610m, 454m and 223w. [Ru{C6H4(TeMe)2-o}2Cl2]. Yield 78% (Found: C, 21.2; H, 2.4. C16H20Cl2RuTe4 requires C, 21.5; H, 2.3%). 1H NMR (CDCl3): d 1.20–2.65 (6 H, m, CH3) and 7.38–7.93 (4 H, m, C6H5). 125Te- {1H} NMR (CDCl3–CH2Cl2): d 804, 885, 890, 893, 894, 914 and 917. ES1 (MeCN): m/z = 901 and 859; calc. for [102Ru- {C6H4(130TeMe)2}2 35Cl 1 MeCN]1 910 and [102Ru{C6H4- (130TeMe)2}2 35Cl]1 869.IR/cm21: 1438w, 1410w, 1358s, 1260w, 1098m, 831w, 803w, 749m, 594s, 248w and 209w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 20 300 (110) and 25 900 (1300). [Ru{C6H4(TeMe)2-o}2Br2]. Yield 45% (Found: C, 18.9; H, 2.0. C16H20Br2RuTe4 requires C, 19.5; H, 2.1%). 1H NMR (CDCl3): d 2.13–2.97 (6 H, m, CH3) and 7.40–7.97 (4 H, m, C6H5). ES1 (MeCN): m/z = 946 and 904; calc. for [102Ru{C6H4(130Te- Me)2}2 79Br 1 MeCN]1 954 and [102Ru{C6H4(130TeMe)2}2 79Br]1 913.IR/cm21: 2929w, 2852w, 1356s, 1256w, 1089m, 991w, 834w, 746w, 638w and 542w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 21 500 (800) and 25 840 (2520). [Ru{C6H4(TeMe)2-o}2I2]. Yield 32% (Found: C, 13.9; H, 3.1. C16H20I2RuTe4 requires C, 14.5; H, 2.9%). 1H NMR (CDCl3): d 2.31–2.54 (6 H, CH3) and 7.30–7.44 (4 H, m, C6H5). ES1 (MeCN): m/z = 998 and 950; calc. for [102Ru{C6H4(130Te- Me)2}2I 1 MeCN]1 1002 and [102Ru{C6H4(130TeMe)2}2I]1 961. IR/cm21: 2929w, 1593s, 1358s, 1076m, 883w, 747w, 598w, 511w and 254m. UV/vis (diVuse reflectance)/cm21: 20 900, 25 200 and 31 200.[Rh{MeTe(CH2)3TeMe}2Cl2]PF6. The compounds RhCl3? 3H2O (51 mg, 0.194 mmol) and MeTe(CH2)3TeMe (125 mg, 0.384 mmol) in EtOH (30 ml) were refluxed for 30 min, NH4PF6 (41 mg, 0.252 mmol) was then added and stirred at room temperature for 1 h. The resulting orange precipitate was collected and washed with diethyl ether. Yield: 76 mg, 41% (Found: C, 12.3; H, 1.7.C10H24Cl2F6PRhTe4 requires C, 12.4; H, 2.5%). 1H NMR (CD3CN): d 2.7–2.9 (CH3 1 CH2) and 3.0–3.1 (CH2). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 250.5 (1J(125Te–103Rh) 60), 259.5 (60), 283.4 (50), 292 (70), 300 (2) and 300.5 (68). ES1 (MeCN): m/z = 831; calc. for [Rh{Me130Te(CH2)3 130TeMe}2- 35Cl2]1 837. IR/cm21: 1359s, 1262w, 1220w, 1203m, 1090m, 994w, 837s, 745w, 612w, 558m, 331m and 203w. UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 21 800 (1250). The following complexes were made similarly.[Rh{PhTe(CH2)3TePh}2Cl2]PF6. Yield 84% (Found: C, 29.7; H, 1.5. C30H32Cl2F6PRhTe4 requires C, 29.5; H, 2.6%). 1H NMR ((CD3)2SO): d 2.9–3.5 (CH2) and 7.1–7.8 (Ph). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 481.5 (1J(125Te–103Rh) 75), 492 (72), 534 (70), 540 (2), 542 (75) and 558 (80). ES1 (MeCN): m/z = 1077; calc. for [Rh{Ph130Te(CH2)3 130TePh}2 35Cl2]1 1085. IR/cm21: 1571w, 1474m, 1434m, 1358m, 1264w, 1206w, 1091m, 1017w, 997m, 837s, 731m, 690m, 654w, 613w, 557m, 451m, 339w, 255w and 226w.UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 21 350 (1700). [Rh{C6H4(TeMe)2-o}2Cl2]PF6. Yield 65% (Found: C, 18.6; H, 1.7. C16H20Cl2F6PRhTe4 requires C, 18.45; H, 1.9%). 1H NMR ((CD3)2SO): d 2.3–2.7 (Me) and 7.4–7.8 (C6H4). 125Te-{1H} NMR (CDCl3–CH2Cl2): d 799 (1J(125Te–103Rh) 90), 814 (2), 840.3 (70), 840.5 (70), 848.5 (70), 861 (55), 869 (80) and 891.5 (70). ES1 (MeCN): m/z = 898; calc. for [Rh{C6H4(130TeMe)2}2- 35Cl2]1 905. IR/cm21: 2935w, 1665w, 1355m, 1259w, 1087s, 992sh, 842s, 751m, 561m, 421m, 327w, 303w, 251w and 208m.UV/vis (MeCN)/cm21 (emol/dm3 mol21 cm21): 21 400 (1380).2076 J. Chem. Soc., Dalton Trans., 1999, 2071–2076 Crystal structures of [Pd{C6H4(TeMe)2-o}2][PF6]2?MeCN, [Ru{PhTe(CH2)3TePh}2Cl2] and [Ru{MeTe(CH2)3TeMe}2- (PPh3)Cl]PF6 Details of the crystallographic data collection and refinement parameters are given in Table 4. For [Pd{C6H4(TeMe)2-o}2]- [PF6]2?MeCN and [Ru{MeTe(CH2)3TeMe}2(PPh3)Cl]PF6 data collection used a Rigaku AFC7S four-circle diVractometer operating at 150 K and graphite-monochromated Mo-Ka X-radiation (l = 0.71073 Å).The data for [Ru{MeTe(CH2)3- TeMe}2(PPh3)Cl]PF6 were corrected for absorption using y-scans. The structures were solved by heavy atom methods 20 and developed by iterative cycles of full-mx least-squares refinement21 and Fourier-diVerence syntheses. All non-H atoms were refined anisotropically and H atoms were placed in fixed, calculated positions.The complex [Pd{C6H4(TeMe)2-o}2]- [PF6]2?MeCN shows two independent half-cations with inversion symmetry, two PF6 2 anions on general positions and two disordered half MeCN solvent molecules in the asymmetric unit. The latter are disordered across inversion centres such that the methyl C atom of one form is superimposed on the cyano C atom of the other form and vice versa, with the inversion centre at the midpoint of this C–C vector. The H atoms associated with the MeCN molecules were not located from the diVerence map and therefore were omitted from the final structure factor calculation.For [Ru{PhTe(CH2)3TePh}2Cl2] data collection used an Enraf-Nonius Kappa CCD diVractometer operating at 150 K and Mo-Ka X-radiation (l = 0.71073 Å). The data were corrected for absorption using SORTAV.22 The structures were solved by heavy atom methods20 and refined using SHELXL 97.23 The molecule has crystallographic i symmetry (at Ru). All non-H atoms were refined anisotropically and H atoms were placed in fixed, calculated positions.CCDC reference number 186/1454. See http://www.rsc.org/suppdata/dt/1999/2071/ for crystallographic files in .cif format. Acknowledgements We thank EPSRC for support (to V.-A. T. and S. D. O.) and Johnson Matthey plc for generous loans of platinum metal salts. We also thank Professor M. B. Hursthouse and Dr S. Coles for collecting X-ray data for [Ru{PhTe(CH)3TePh}2Cl2]. References 1 E. G. Hope, T.Kemmitt and W. Levason, Organometallics, 1988, 7, 78. 2 T. Kemmitt and W. Levason, Organometallics, 1989, 8, 1303. 3 T. Kemmitt, W. Levason and M. Webster, Inorg. Chem., 1989, 28, 692. 4 T. Kemmitt and W. Levason, Inorg. Chem., 1990, 29, 731. 5 R. A. Cipriano, L. R. Hanton, W. Levason, D. Pletcher, N. A. Powell and M. Webster, J. Chem. Soc., Dalton Trans., 1988, 2483. 6 J. L. Brown, T. Kemmitt and W. Levason, J. Chem. Soc., Dalton Trans., 1990, 1513. 7 J. R. Black and W. Levason, J. Chem. Soc., Dalton Trans., 1994, 3225; J. R. Black, N. R. Champness, W. Levason and G. Reid, J. Chem. Soc., Dalton Trans., 1995, 3439; Inorg. Chem., 1996, 35, 1820. 8 W. Levason, S. D. Orchard and G. Reid, Organometallics, 1999, 18, 1275; A. J. Barton, W. Levason and G. Reid, J. Organomet. Chem., in the press. 9 E. W. Abel, K. G. Orrell, S. P. Scanlan, D. Stevenson, T. Kemmitt and W. Levason, J. Chem. Soc., Dalton Trans., 1991, 591; E. W. Abel, S. K. Bhargava and K. G. Orrell, Prog. Inorg. Chem., 1984, 32, 1. 10 N. R. Champness, W. Levason, S. R. Preece and M. Webster, Polyhedron, 1994, 13, 881. 11 R. J. Judd, R. Cao, M. Biner, T. Armbruster, H.-B. Burgi, A. E. Merbach and A. Ludi, Inorg. Chem., 1995, 34, 5080. 12 W. Levason, J. J. Quirk, G. Reid and S. M. Smith, J. Chem. Soc., Dalton Trans., 1997, 3719. 13 E. W. Abel, I. Moss, K. G. Orrell, V. Sik, D. Stephenson, P. A. Bates and M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1988, 521. 14 B. A. Moyer, B. K. Sipe and T. J. Meyer, Inorg. Chem., 1981, 20, 1475. 15 F. A. Cotton, M. Diebold and M. Matusz, Polyhedron, 1987, 6, 2164. 16 D. M. Klassen and G. A. Crosby, J. Mol. Spectrosc., 1968, 25, 398. 17 N. R. Champness, W. Levason, D. Pletcher and M. Webster, J. Chem. Soc., Dalton Trans., 1992, 3243. 18 D. J. Gulliver, E. G. Hope, W. Levason, S. G. Murray and G. L. Marshall, J. Chem. Soc., Dalton Trans., 1985, 1265. 19 R. A. Walton, J. Chem. Soc. A, 1967, 1852; J. Chatt, G. J. Leigh and A. P. Storace, J. Chem. Soc. A, 1971, 899. 20 PATTY, The DIRDIF Program System, P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smits and C. Smykalla, Technical Report of the Crystallography Laboratory, University of Nijmegen, 1992. 21 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, Texas, 1995. 22 R. H. Blessing, Acta Crystallogr., Sect. A, 1995, 51, 33; J. Appl. Crystallogr., 1997, 30, 421. 23 G. M. Sheldrick, SHELXL-97, University of Göttingen, 1997. Paper 9/01942A

ISSN:1477-9226    

DOI:10.1039/a901942a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2073–2077 2073 Reduction of trans-dichloro- and trans-dibromo-tetracyanoplatinate( IV) by L-methionine † Tiesheng Shi, Johan Berglund and Lars I. Elding * Inorganic Chemistry 1, Chemical Center, Lund University, PO Box 124, S-221 00 Lund, Sweden Reduction of trans-[Pt(CN)4X2]22 (X = Cl or Br) [as model compounds for antitumour-active platinum(IV) prodrugs] to [Pt(CN)4]22 by L-methionine, MeSR, has been studied at 25 8C in the range 0 < pH < 12 (X = Cl) and 0 < pH < 6 (X= Br) by use of stopped-flow spectrophotometry.The stoichiometry is [PtIV] : [MeSR] ª 1:1; the reaction products are methionine S-oxide and [Pt(CN)4]22 as identified by NMR and UV spectroscopies, respectively. The kinetics is first order with respect to the platinum(IV) and methionine concentrations and the second-order rate constants have a small pH dependence. In analogy with reduction of platinum(IV) complexes by thioglycolic acid, cysteine, penicillamine and glutathione, a mechanism is postulated in which [Pt(CN)4X2]22 is reduced by the various protolytic species of methionine in parallel reactions.In the transition state the thioether group of methionine is assumed to interact with co-ordinated halide, mediating the electron transfer to the platinum(IV) centre. The transition states for previously studied reactions between [Pt(CN)4X2]22 and thiols are discussed in view of these results. It is concluded that methionine-containing biomolecules may compete with thiol compounds for reduction of platinum(IV) pro-drugs under acidic conditions, and also in neutral solutions with low concentrations of thiol-containing biomolecules.Antitumour-active platinum(IV) drugs are assumed to be reduced to their platinum(II) analogues within the cell prior to interaction with DNA.1–9 Potential intracellular reducing agents are ascorbate and thiol-containing biomolecules, in particular glutathione (g-glutamylcysteinylglycine) which is frequently present in high concentrations.10,11 We have recently reported on the kinetics and mechanism for the reduction of trans-[Pt(CN)4Cl2]22 to [Pt(CN)4]22 with the thiols thioglycolic acid, L-cysteine, DL-penicillamine (3-sulfanyl-D-valine) and glutathione. 12 The platinum complex was chosen as a convenient model compound for the study of reactions of various platinum( IV) pro-drugs, since the redox process is not disturbed by subsequent substitution processes in the platinum(II) product formed, [Pt(CN)4]22. The reaction kinetics can be interpreted in terms of a mechanism involving chloride-mediated electron transfer from the different protolytic species of the reductant to the platinum centre.‡ A Brønsted correlation of the type log kRS2 = 0.82 pKRSH 1 1.1 between the rate constants for oxidation of the four thiolate anions, RS2, and the pKRSH of the thiol was observed, indicating that the basicity of the reductant is a predominant factor in determining the reactivity towards the platinum(IV) complex.Interestingly, the same type of correlation is also found for the different species within each single protolytic thiol system as shown in Fig. 1. Unfortunately, it was not possible to determine the rates under conditions where the amino groups are deprotonated because the reactions become too fast to be monitored by the conventional stopped-flow technique already at pH ca. 5. Methionine, MeS(CH2)2CH(NH3 1)CO2H, as part of some intracellular proteins, is also a potential reductant for platinum( IV) pro-drugs. Thus, it is interesting to compare the reactivity of this thioether-containing amino acid with the thiolcontaining biomolecules studied previously. Thioethers are less † Supplementary data available (No. SUP 57246, 6 pp.): pseudo-firstorder and second-order rate constants. See Instructions for Authors, J. Chem. Soc., Dalton Trans., 1997, Issue 1.‡ The thiols contain two to four protolytic groups, glutathione [HO2CCH(NH3 1)CH2CH2CONHCH(CH2SH)CONHCH2CO2H] for example contains two carboxylic, one thiol, and one amine group. The structures of the other three thiols are given in ref. 12. reactive than thiols and we were, therefore, able to study the reduction of the complexes trans-[Pt(CN)4X2]22 (X = Cl or Br) with methionine in a much broader pH range than with the thiol compounds. Contrary to the large variations in rate with pH observed in the thiol systems, with methionine there is only a minor change of reduction rate in the range 0 < pH < 12.Experimental Chemicals and solutions The salt K2[Pt(CN)4Cl2] was synthesized by oxidation of K2[Pt(CN)4]?3H2O with chlorine as described previously.13 The UV/VIS spectrum of the sample solution agreed with that reported earlier for [Pt(CN)4Cl2]22.14 The salt K2[Pt(CN)4Br2] was prepared according to the literature.15 Stock solutions of 5.0 mmol dm23 K2[Pt(CN)4X2] in 10 mmol dm23 HClO4, 10 Fig. 1 Correlations between log k for reduction of trans-[Pt(CN)4- Cl2]22 and pKa for the various species within thiol protolytic systems (glutathione, cysteine, penicillamine and thioglycolic acid) based on data from ref. 12 (n is an arbitrary constant added to separate the lines for clarity)2074 J. Chem. Soc., Dalton Trans., 1997, Pages 2073–2077 mmol dm23 NaX and 0.98 mol dm23 NaClO4 ionic medium were stable for several months.Deuterium oxide (Janssen, 99.5%), L-methionine (Merck, pa, >99%) and DL-methionine S-oxide (Acros Organics, >99%) were used as received. Stock solutions of methionine were prepared fresh daily. Acetate (0.2 mol dm23), phosphate (0.05–0.10 mol dm23), borate (0.05 mol dm23) and carbonate (0.10 mol dm23) buffers were used in the range 3.5 < pH < 12 and perchloric acid was used to adjust the pH in the range pH < 2.4. The ionic strength, I, was adjusted to 1.00 mol dm23 with sodium perchlorate, and all experiments were run at sufficiently high halide concentrations (usually 0.10 mol dm23) to suppress hydrolysis of [Pt(CN)4X2]22.Water was doubly distilled from quartz. Apparatus and measurements The UV/VIS spectra were recorded by use of a Milton Roy 3000 diode-array spectrophotometer and 1.00 cm quartz Suprasil cells and proton NMR spectra on a Varian UNITY 300 MHz spectrometer working at room temperature. Proton NMR chemical shifts were measured in ppm relative to the residual solvent peak calibrated by SiMe4.Kinetic traces were collected using an Applied Photophysics Bio Sequential SX-17 MX stopped-flow ASVD spectrofluorimeter, and rate constants (reported as average values from five to seven independent runs) were evaluated with the Applied Photophysics software package.16 Hydrogen-ion concentrations were calculated from equation (1) which is based on a mean activity coefficient of 0.630 for 1.00 mol kg21 NaClO4.17 2log [H1] = p[H1] = pH 2 0.20 (1) Results and Discussion Stoichiometry Ten solutions with [Pt(CN)4Cl2 22] = 1.0 × 1024 mol dm23, [HClO4] = 0.10 mol dm23, [Cl2] = 0.10 mol dm23, I = 1.00 mol dm23 and 0 < [MeSR]tot < 2.0 × 1024 mol dm23 were aged for ca. 18 h and their absorbances at the typical 255 nm maximum of [Pt(CN)4]22 were recorded. The stoichiometry was determined as [PtIV] : [MeSR]tot = 1 : 0.93 from the plot in Fig. 2. In order to check the reaction products further, 1H NMR spectra of a solution of ca. 1.0 × 1022 mol dm23 methionine, 1.0 × 1022 mol dm23 [Pt(CN)4Br2]22 in D2O and standard samples of DLmethionine S-oxide and L-methionine in D2O were recorded. Comparison of those spectra shows that methionine is consumed almost completely when reduced by platinum(IV) as indicated by the decrease of the signal at d(CH3S) 1.98. The appearance of a sharp signal at d[CH3S(O)] 2.62 indicates that methionine S-oxide is the oxidation product.From the UV/VIS and NMR measurements we conclude that the stoichiometry can be expressed by equation (2). No subsequent reaction [Pt(CN)4X2]22 1 MeSR 1 H2O [Pt(CN)4]22 1 MeS(O)R 1 2X2 1 2H1 (2) between methionine S-oxide and [Pt(CN)4Cl2]22 could be detected under similar conditions during 18 h. Kinetics Reduction of [Pt(CN)4X2]22 was studied under pseudo-firstorder conditions with methionine in at least a 10-fold excess. It was monitored by following the increase of absorbance at the 255 nm maximum of [Pt(CN)4]22 or the decrease of the 240 nm peak of [Pt(CN)4Br2]22, respectively.Single-exponential kinetics traces were obtained in both cases and the concentration of the excess of halide had no influence on the rate of reaction. Time-resolved spectra for the reaction with [Pt(CN)4Cl2]22 are shown in Fig. 3, indicating that there is no accumulation of intermediates during reduction to [Pt(CN)4]22.§ Slight systematic deviations from the isosbestic points at 285 and, in particular, 246 nm are due to the absorbance contribution from the methionine S-oxide product.The observed rate constants, kobs, are proportional to the excess concentration of methionine in the whole pH region studied (cf. Fig. 4), giving overall secondorder kinetics according to equation (3), where k9 denotes the 2d[Pt(CN)4X2 22]/dt = d[Pt(CN)4 22]/dt = k9[MeSR]tot[Pt(CN)4X2 22] (3a) kobs = k9[MeSR]tot (3b) pH-dependent second-order rate constants. Values of k9, calculated by the use of equation (3b) from the data summarised in SUP 57246 are shown as a function of p[H1] in Fig. 5.¶ Values of kobs and k9 are summarised in SUP 57246. Stoichiometric reaction mechanism Platinum(IV) complexes are substitution inert and initial complex formation prior to electron transfer does not normally take Fig. 2 Absorbance at 255 nm for solutions with constant [PtIV] and increasing [MeSR]tot. Conditions: [Pt(CN)4Cl2 22] = 1.0 × 1024 mol dm23, [Cl2] = 0.10 mol dm23, [H1] = 0.10 mol dm23, I = 1.00 mol dm23 and room temperature Fig. 3 Time-resolved spectra for reduction of [Pt(CN)4Cl2]22 by methionine. Conditions: [PtIV]tot = 1.25 × 1024 mol dm23, [MeSR]tot = 2.0 mmol dm23, [HClO4] = 0.10 mol dm23, [Cl2] = 0.10 mol dm23, 25 8C and I = 1.00 mol dm23; 1.00 mol dm23 sodium perchlorate was used as reference. Time interval between scans 50 s; first scan obtained ca. 10 s after mixing § This has been shown to be true also when thiols 12 and sulfite 13 are used as reductants.¶ Reduction of [Pt(CN)4Br2]22 was only studied in the interval 0 < pH < 6 since this complex is not stable at pH > 7 (see ref. 15).J. Chem. Soc., Dalton Trans., 1997, Pages 2073–2077 2075 place in reductive-elimination reactions.13–15,18–24 The simple and well defined monophasic kinetics observed for reduction of [Pt(CN)4Cl2]22 is consistent with that general behaviour. In analogy with the previously studied thiol systems,12 we propose a mechanism (Scheme 1) where all protolytic species of methionine reduce [Pt(CN)4X2]22 in parallel reactions, giving MeS(X)R as short-lived intermediates; in a subsequent rapid step, MeS(X)R will hydrolyse to give MeS(O)R as the final oxidation product.Similar short-lived intermediates were also suggested previously for reduction of gold(III) complexes by dimethyl sulfide.25,26 The second-order rate constants k9 defined in equation (3) can be derived as in (4).A weighted non-linear k9 = k1[H1]2 1 k2K1[H1] 1 k3K1K2 [H1]2 1 K1[H1] 1 K1K2 (4) least-squares fit of equation (4) to the experimental data for the chloride complex with the rate and protolysis constants as adjustable parameters gave pK1 = 1.7 ± 0.2 and pK2 = 8.9 ± 0.1, and the values of the rate constants listed in Table 1, cf. Fig. 5. The protolysis constants derived from the curve fit are in reasonably good agreement with those reported in the literature (pKa1 = 2.22 ± 0.04 and pKa2 = 9.02 ± 0.02),27 if the difference in ionic medium used is taken into account.At pH < 6, where the reduction of [Pt(CN)4Br2]22 was studied, the data indicate that the pH is sufficiently far from the pK2 value of methionine to assure that the concentration of fully deprotonated methionine can be neglected; equation (4) can Fig. 4 Plots of kobs as a function of [MeSR]tot and temperature for reduction of [Pt(CN)4X2]22 (X = Cl or Br). Conditions: p[H1] = 4.22, 0.20 mol dm23 acetic acid–acetate buffer, [X2] = 0.10 mol dm23.Temperatures are 20.0 (a), 25.0 (b), 30.0 (c), 35.0 (d) and 40.1 8C (e) then be simplified to (5). A weighted non-linear least-squares fit of equation (5) to the data for the bromide complex gave k9 = (k1[H1] 1 k2K1)/(K1 1 [H1]) (5) pK1 = 1.85 ± 0.11 and the values of k1 and k2 listed in Table 1 (Fig. 5). In the interval 4 < p[H1] < 7, MeSCH2CH2CH(NH3 1)CO2 2 is the only contributing reductant. Activation enthalpies and entropies for the reduction of [Pt(CN)4X2]22 by this protolyte were calculated by fitting the Eyring equation to the variable-temperature data for k2 at p[H1] = 4.22, cf.Table 2. Transition states for reactions with methionine and thiols Halide-mediated reductive-elimination reactions of platinum- (IV) complexes involving various reductants have been suggested to take place via an attack by the reductant on coordinated halide.12–15,22–24 In the present case the transition state may be formulated as in I.The reductive elimination might be visualized as a classical SN2 attack of the reductant on coordinated halide with the platinum moiety as the leaving group.12 The increase in rate with increasing pH as shown in Fig. 5 could easily be accounted for by the observed decrease in redox potential of methionine as the pH is increased.28 Thus, it is likely that it is the thioether group of the various protolytic species of methionine that interacts with the platinum complex in all cases, as indicated above.Reduction of the bromide complex with methionine is faster than reduction of the chloride complex, consistent with bromide being a better bridging ligand than chloride.|| In addition, the negative activation entro- Fig. 5 Second-order rate constants k9 defined by equation (3), as a function of p[H1] at 25.0 8C for reduction of [Pt(CN)4X2]22 (X = Cl or Br) with methionine. The solid lines represent the best fits of equations (4) and (5), respectively, to the experimental data by use of a weighted least-squares regression analysis || The kBr :kCl ratios of 6–8 : 1 observed in the present study are signifi- cantly smaller than those of about 400 : 1 observed previously for reduction of [Pt(CN)4X2]22 with SCN2, I2, CN2 and SO3 22.232076 J.Chem. Soc., Dalton Trans., 1997, Pages 2073–2077 Scheme 1 pies derived for reductions with MeSCH2CH2CH(NH3 1)CO2 2 given in Table 2 are very similar to those calculated previously for reduction of gold(III) complexes by dimethyl sulfide,26 also supporting a transition state of the type suggested above.The kinetics for reduction with thiols studied previously12 is very different from the present findings. The Brønsted correlation observed between the rate of reduction of [Pt(CN)4Cl2]22 and the protolysis constants of the thiol groups in thioglycolic acid, L-cysteine, DL-penicillamine, and glutathione indicates that the basicity of the reductant is a crucial parameter in determining the reactivity towards the platinum complex.12 Within each protolytic thiol system, there are carboxylate, thiol and amine groups which may act as nucleophiles towards co-ordinated chloride, depending on the pH.The similar Brønsted correlations observed for the four different thiolate anions are also observed within all four protolytic thiol systems (log k = a pKa 1 b with 0.93 < a < 1.0 and small intercepts, b), cf.Fig. 1. Eventually, these correlations might indicate that the protolytic group with the highest basicity interacts with the platinum(IV) complex in the rate-limiting step. Thus, at low pH, where the thiol groups are protonated, the carboxylate groups should be able to attack [Pt(CN)4Cl2]22 forming a bridged intermediate in which electron transfer from the thiol group to the platinum(IV) centre could take place rapidly. This mechanistic interpretation Table 1 Second-order rate constants, kn (n = 1–3), for reduction of trans-[Pt(CN)4X2]22 (X = Cl or Br) by the different protolytic species of methionine at 25 8C and 1.00 mol dm23 ionic strength Reductant kn Cl/dm3 mol21 s21 kn Br/dm3 mol21 s21 kn Br :kn Cl MeSCH2CH2CH(NH3 1)CO2H MeSCH2CH2CH(NH3 1)CO2 2 MeSCH2CH2CH(NH2)CO2 2 2.54 ± 0.01 4.47 ± 0.01 27.5 ± 0.1 14.8 ± 0.2 36.9 ± 0.2 5.8:1 8.3:1 Table 2 Rate constants, k2, as a function of temperature and activation parameters for the reduction of trans-[Pt(CN)4X2]22 by MeSCH2CH2CH(NH3 1)CO2 2 at 1.00 mol dm23 ionic strength [Pt(CN)4X2]22 T/ 8C k2/dm3 mol21 s21 X = Cl 20.0 25.0 30.0 35.0 40.1 3.28 ± 0.01 4.47 ± 0.01 5.71 ± 0.04 7.49 ± 0.05 9.5 ± 0.2 DH‡ 37.7 ± 0.7 kJ mol21, DS‡ 2106 ± 2 J K21 mol21 X = Br 20.0 25.0 30.0 35.0 24.1 ± 0.1 36.9 ± 0.2 51.4 ± 0.4 70.6 ± 0.2 DH‡ 51 ± 2 kJ mol21, DS‡ 244 ± 7 J K21 mol21 would imply that the rate of oxidation should continue to increase in alkaline solutions, where the amine groups are deprotonated.We were not able to verify if this is the case, since the reactions become too fast to be followed by the stopped- flow technique at pH ca. 5. At the moment, it is not possible to conclude with certainty whether the carboxylate and amine groups in the previously studied thiols in fact do attack [Pt(CN)4X2]22. It may be possible that only the thiol group is interacting with the complex, as seems to be the case with the methionine thioether moiety, and that the strong Brønsted correlations and large variations in rates observed for the thiol protolytic systems have alternative explanations.For example, in the transition state formed between [Pt(CN)4Cl2]22 and the cysteine protolyte HSCH2- CH(NH3 1)CO2 2, the carboxylate group might form an intramolecular hydrogen bond with the thiol (SH),29 resulting in an increased electron density on the sulfur atom, facilitating electron transfer. Similar intramolecular interactions within the other polyfunctional protolytic thiol systems might change the electronic properties of the thiol moieties and the redox rates.Implications for platinum(IV) antitumour drugs Reduction of the platinum(IV) model complex by methionine is considerably slower than reduction by cysteine, penicillamine and glutathione in neutral solutions, but the rates are comparable under acidic conditions. The conclusion must be, then, that in neutral solutions, methionine-containing biomolecules can compete for reduction of platinum(IV) pro-drugs only under conditions of relatively low concentrations of thiols; under acidic conditions, however, methionine-containing biomolecules may be as competitive as thiols for reduction of platinum- (IV) pro-drugs to antitumor-active platinum(II) complexes.Acknowledgements Financial support from the Swedish Natural Science Research Council is gratefully acknowledged. References 1 A. Eastman, Biochem. Pharmacol., 1987, 36, 4177. 2 W. K. Anderson, D. A. Quagliato, R. D. Haugwitz, V. L. Narayanan and M. K. Wolpert-DeFilippes, Cancer Treat. Rep., 1986, 70, 997. 3 L. J. Wikoff, E. A. Dulmadge, M. W. Trader, S. D. Harrison and D. P. Griswold, Cancer Chemother. Pharmacol., 1987, 20, 96.J. Chem. Soc., Dalton Trans., 1997, Pages 2073–2077 2077 4 G. R. Gibbons, S. D. Wyrick and S. G. Chaney, Cancer Res., 1989, 49, 1402. 5 P. C. Dedon and R. F. Borch, Biochem. Pharmacol, 1987, 36, 1955. 6 L. Pendyala, J.W. Cowens, G. B. Chheda, S. P. Dutta and P. J. Creaven, Cancer Res., 1988, 48, 3533. 7 E. E. Blatter, J. F. Vollano, B. S. Krishnan and J. C. Dabrowiak, Biochemistry, 1984, 23, 4817. 8 M. Laverick, A. H. W. Nias, P. J. Sadler and I. M. Ismail, Br. J. Cancer, 1981, 43, 732. 9 J. L. van der Veer, A. R. Peters and J. Reedijk, J. Inorg. Biochem., 1986, 26, 137. 10 D. L. Rabenstein, R. Guevremont and C. A. Evans, Metal Ions Biol. Systems, 1979, 9, 103. 11 G. B. Henderson, A.H. Fairlamb, P. Ulrich and A. Cerami, Biochemistry, 1987, 26, 3023. 12 T. Shi, J. Berglund and L. I. Elding, Inrog. Chem., 1996, 35, 3498. 13 J. Berglund, R. Voigt, S. Fronaeus and L. I. Elding, Inorg. Chem., 1994, 33, 3346. 14 L. Drougge and L. I. Elding, Inorg. Chim. Acta., 1986, 121, 175. 15 C. E. Skinner and M. M. Jones, J. Am. Chem. Soc., 1969, 91, 1984. 16 Applied Photophysics Bio Sequential SX-17 MV, Sequential Stopped-Flow ASVD Spectroflorimeter, software manual, Applied Photophysics Ltd., Leatherhead. 17 D. R. Lide, Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL, 75th edn., 1994, pp. 5–97. 18 F. Basolo, P. H. Wilks, R. G. Pearson and R. G. Wilkins, J. Inorg. Nucl. Chem., 1958, 6, 161. 19 F. Basolo, M. L. Morris and R. G. Pearson, Discuss. Faraday Soc., 1960, 29, 80. 20 W. R. Mason, Coord. Chem. Rev., 1972, 7, 241. 21 L. I. Elding and L. Gustafson, Inorg. Chim. Acta., 1971, 5, 643; 1976, 19, 31, 165; 1977, 22, 201. 22 W. K. Wilmarth, Y.-T. Fanchiang and J. E. Byrd, Coord. Chem. Rev., 1983, 51, 141. 23 P. Chandayot and Y.-T. Fanchiang, Inorg. Chem., 1985, 24, 3532. 24 P. Chandayot and Y.-T. Fanchiang, Inorg. Chem., 1985, 24, 3535. 25 G. Annibale, L. Canovese, L. Cattalini and G. Natile, J. Chem. Soc., Dalton Trans., 1981, 1093. 26 A. Ericson, L. I. Elding and S. K. C. Elmroth, J. Chem. Soc., Dalton Trans., 1997, 1159. 27 R. M. Smith and A. E. Martell, Critical Stability Constants, Plenum, New York, 1989, vol. 6, Suppl. 2; M. Jawaid and F. Ingman, Talanta, 1981, 28, 137. 28 Sanaullah, G. S. Wilson and R. S. Glass, J. Inorg. Biochem., 1994, 55, 87. 29 M. R. Crampton, in The Chemistry of the Thiol Group, ed. S. Patai, Wiley, New York, 1974, ch. 8. Received 19th December 1996; Paper 6/08507E

ISSN:1477-9226    

DOI:10.1039/a608507e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2077–2086 2077 Synthesis, structural characterization and reactivities of hexaosmium carbonyl clusters with six-membered cyclic thioether and thioxane ligands Kelvin Sze-Yin Leung and Wing-Tak Wong * Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, P. R. China Received 29th January 1999, Accepted 30th April 1999 Cluster complexes [Os6(CO)15{S(CH2)4CH2}{m-S(CH2)4CH2}] 1 and [Os5(CO)15{S(CH2)4CH2}] 2 were isolated from the reaction of [Os6(CO)16(MeCN)2] with S(CH2)4CH2 (L1) in 24 and 22% yields respectively.Cluster degradation from 1 to 2 can be achieved by addition of a stoichiometric quantity of L1 in CHCl3 under reflux. Monosubstituted cluster complex [Os6(CO)16{m-S(CH2)3SCH2}] 3 containing a S(CH2)3SCH2 (L2) ligand bridging across an Os–Os edge via one of the sulfur atoms was isolated in good yield from a similar reaction. However, similar reaction between [Os6(CO)16(MeCN)2] and S(CH2)2SCH2CH2 (L3) gave [Os6(CO)16{S(CH2)2SCH2CH2}2] 4 as the major product instead.The two thio ligands are found to co-ordinate in terminal fashion to two diVerent vertices of the Os6 core. Interaction of tridentate sulfur donor ligand SCH2SCH2SCH2 (L4) with [Os6(CO)16(MeCN)2] aVorded [Os6(CO)14- (m-CO)(SCH2SCH2SCH2)] 5 in which the ligand L4 was found to cap over a triangular face of the Os6 skeleton. Treatment of mixed-donor ligand S(CH2)2OCH2CH2 (L5) yielded a pair of isomeric complexes [Os6(CO)15- {S(CH2)2OCH2CH2}{m-S(CH2)2OCH2CH2}] 6 and [Os6(CO)15{O(CH2)2SCH2CH2}{m-S(CH2)2OCH2CH2}] 7.Carboxylation of 6 gave [Os6(CO)16{m-S(CH2)2OCH2CH2}] 8 and [Os6(CO)18]. Hydrogenation of 6 led to the formation of dihydrido cluster [Os6(CO)15(m-H)2{S(CH2)2OCH2CH2}{m-S(CH2)2OCH2CH2}] 9. The metal skeleton of 9 can be described as two fused tetrahedra sharing a common edge. All new compounds were fully characterized by spectroscopic and analytical techniques.In addition the structures of 1, 2, 3, 4, 5, 6, 8 and 9 were established by X-ray crystallography. Introduction Following the industrial applications of hydrodesulfurization in fossil fuel purification,1 much interest has been shown in cluster compounds bearing sulfur heteroatoms. A new class of cluster derivatives containing sulfido ligands has been extensively studied and reported in tri-,2 tetra-,3 and penta-4 osmium systems by Adams and co-workers over the past decades. These have been widely used in the systematic build-up of high nuclearity clusters 5 as exemplified by the thermally activated coupling of [Os4(CO)13(m3-S)2] and [Os3(CO)10(MeCN)2] to yield the heptanuclear cluster compound [Os7(CO)20(m4-S)2].6 The ring-opening oligomerization of thiirane 7 and thietane 8 have been shown to be catalysed by [Os6(CO)16(MeCN)2] and [Os4(CO)11(SCH2CH2CH2)(m-H)4] respectively.Recently we have also reported some triosmium alkylidyne clusters containing cyclic thioether ligands that undergo facile isomerization.9 However, further explorations on these cyclic sulfur donor ligands in larger cluster systems are relatively rare.Herein we report the syntheses, spectroscopic studies and reactivities of a series of cyclic sulfur-containing hexaosmium carbonyl clusters. Results and discussion The reaction (Scheme 1) of preformed labile cluster [Os6(CO)16- (MeCN)2] with L1 yielded the hexanuclear [Os6(CO)15- {S(CH2)4CH2}{m-S(CH2)4CH2}] 1 and pentanuclear [Os5(CO)15- {S(CH2)4CH2}] 2 species together with small amount of the known cluster [Os3(CO)12].10 Facile replacement of the labile MeCN ligand to yield mono- and di-substituted cluster derivatives has previously been reported in the triosmium system.11,12 Cluster 1 consists of an identical metal-core architecture to that in the parent cluster [Os6(CO)18].A perspective drawing of cluster 1 with the atomic numbering scheme is shown in Fig. 1. Selected bond parameters are presented in Table 1. The metal–metal bond distances are comparable to the corresponding values in [Os6(CO)18].13 However, the Os(1)– Os(4) vector [2.931 Å] is significantly longer than Os(1)–Os(3) [2.823 Å], leading to a ‘distorted’ bicapped tetrahedron. The observed bond elongation may be attributed to steric influence between the bulky L1 and equatorial CO on Os(4). The Os(2)– Os(5) bond distance is the shortest M–M bond length [2.679 Å] which might be due to the ‘clamping’ eVect of the bridging ligand.Two ligands L1 are co-ordinated to the Os6 core in a terminal and a bridging mode respectively. According to the eighteen electron rule, the vertices Os(1) and Os(6) are electron deficient and hence susceptible to nucleophilic attack.14,15 Similar edge bridging of heteroatoms across a central tetra- S S S S S S O S S S L1 L2 L3 L4 L52078 J. Chem. Soc., Dalton Trans., 1999, 2077–2086 Scheme 1 Os Os Os Os Os S S Os S Os Os Os Os Os Os Os Os Os Os S S Os S S Os Os Os Os Os Os S S Os Os Os Os Os Os S S S Os Os Os Os Os Os S O S O Os Os Os Os Os Os S O Os Os Os Os Os Os S O S O H H [Os6(CO)16(MeCN)2] [Os6(CO)15{O(CH2)2SCH2CH2}{m-S(CH2)2OCH2CH2}] + + L1 L2 L3 L4 L5 1 2 3 4 5 6 8 9 7 L1 H2 Refluxing CHCl3 CO Refluxing CHCl3 Refluxing CHCl3 [Os6(CO)18] + hedron is also observed in [Os6(CO)16(m-H)(m-C8H11)].16 However, the Os(2)–S(2) and Os(5)–S(2) distances [2.269(4) and 2.276(5) Å, respectively] are considerably shorter than those found in other thietane and thiolate bridging analogues [Os3(CO)10(m-SCH2CMe2CH2)] 17 and [Os3(CO)10(m-H)(m- Fig. 1 Molecular structure of cluster 1 showing the atom-labelling scheme for non-hydrogen atoms.Table 1 Selected bond distances (Å) and angles (8) for cluster 1 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) 2.740(1) 2.823(1) 2.931(1) 2.859(1) 2.753(1) 2.679(1) 2.766(1) 2.845(1) 2.838(1) Os(4)–Os(5) Os(4)–Os(6) Os(5)–Os(6) Os(1)–S(1) Os(2)–S(2) Os(5)–S(2) Os(2)–S(2)–Os(5) Os(5)–Os(2)–S(2) S(2)–Os(5)–Os(2) 2.795(1) 2.848(1) 2.714(1) 2.368(5) 2.269(4) 2.276(5) 72.2(1) 54.0(1) 53.8(1) SCH2CMe2CH2Cl)],18 respectively.Both ligands are in the stable chair conformation. Cluster 2 was found to be a pentanuclear osmium compound having a trigonal bipyramidal metal core arrangement with fifteen terminally bonded carbonyl ligands (Fig. 2). Selected bond parameters are in Table. 2. The Os(2) atom, which has four terminally bonded ligands, is considerably electron-rich and all Os–Os vectors involving Os(2) are significantly longer than other metal–metal bonds in the structure. This phenomenon is Fig. 2 Molecular structure of cluster 2. Table 2 Selected bond distances (Å) and angles (8) for cluster 2 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) 2.893(1) 2.765(1) 2.776(1) 2.832(1) 2.828(1) 2.833(1) 2.790(1) Os(3)–Os(5) Os(4)–Os(5) Os(2)–S(1) Os(4) ? ? ? C(4) Os(2)–C(4)–O(4) 2.773(1) 2.813(1) 2.430(4) 2.49(1) 158(1)J.Chem. Soc., Dalton Trans., 1999, 2077–2086 2079 generally in line with those observed in other Os5(CO)15(L) [L = CO,19 PMe3 20 or P(OMe)3 21] analogues. Besides, this kind of electronic imbalance within a metal core leads to the formation of ‘semibridging’ carbonyl ligands. Cotton 22 has suggested that this type of interaction could alleviate the polar nature of donor–acceptor metal–metal bonds.Cluster 2 also manifests this eVect in C(4)–O(4) [Os(2)–C(4)–O(4) 1588; Os(4) ? ? ? C(4) 2.49 Å]. On the electronic grounds, as in the rationalization of site preference in cluster 1, nucleophilic attack would be expected on the apical atoms which are relatively electron deficient. However, the solid state structure of compound 2 revealed the co-ordination of L1 in the equatorial plane of the metal core. This observation can be attributed to the better s-donating ability of L1, together with three carbonyl ligands, stabilizing the dative metal–metal bonds originated from electron rich Os(2).23 Using an excess of ligand in the reaction leads to higher yields of compounds 1 and 2 and also the disappearance of the known cluster [Os3(CO)12]. However, a number of additional products in relatively low yields are also observed and are proposed to be multi-substituted hexaosmium species based on the spectroscopic and elemental analyses.Treatment of 1 with a stoichiometric quantity of L1 in refluxing CHCl3 leads to the formation of 2. Therefore, we believe 1 is likely to be an intermediate for the formation of 2. Unfortunately, we are not able to isolate and characterize the mononuclear osmium fragment formed in this conversion. Under similar reaction conditions, both 1,3-dithiane (L2) and 1,4-dithiane (L3) reacted with [Os6(CO)16(MeCN)2]. However, the major products isolated from these reactions are rather diVerent from that with L1. Clusters 3 and 4 were isolated in moderate yields.Positive FAB MS of 3 revealed an intense molecular envelope centre at m/z 1709 (Table 3), which along with the 1H NMR spectroscopy suggested that it consists of a monosubstituted ligand in a Os6 core. Its molecular structure was ascertained by a single crystal X-ray analysis. A perspective drawing of it with atomic numbering scheme is shown in Fig. 3 and important bond parameters are summarized in Table 4.Its metal core arrangement is similar to that of 1, comprising six osmium atoms in a bicapped tetrahedral mode. The Os–Os bond lengths in the structure span a range [2.713 to 2.867 Å] which is commonly observed for Os–Os single bonds except for the one bridged by the thio ligand [Os(2)–Os(5) 2.662 Å]. The 1H NMR spectra can be interpreted with reference to the solid state structure of compound 4. The spectrum of 4 exhibits Fig. 3 Molecular structure of cluster 3.four groups of well defined multiplets in close proximity due to two individual ligands in the same magnetic environment. The molecular structure of cluster 4 established by X-ray analysis is depicted in Fig. 4 and some important bond parameters are given in Table 5. The Os6 metal skeleton contains a more symmetrical ligand deposition with a non-crystallographic twofold axis passing through the midpoints of Os(3)–Os(4) and Os(2)– Os(5). Two ligands are terminally bonded to the two vertices of the polyhedron.Both S(2) and S(4) are unco-ordinated and might be able to act as donor groups for further cluster buildup. However, cluster 4 does not react with labile clusters such as [Os6(CO)16(MeCN)2] and [Os3(CO)10(MeCN)2] 24 to give linked clusters. Attempts to generate a pentanuclear species as in compound 1 using an excess of ligand at elevated reaction temperature only led to cluster decomposition. It is also noteworthy that no C–S bond cleavages were observed under forcing thermolytic conditions which contrasts to those findings in triosmium analogues.25 Treatment of ligand L4 with an equivalent of [Os6(CO)16- (MeCN)2] gave a moderate yield of cluster 5 as the major product isolated upon TLC purification.The molecular structure of 5 is shown in Fig. 5 and selected bond lengths and angles are summarized in Table 6. The tridentate ligand asymmetrically caps the triangular face defined by Os(1)– Os(2)–Os(4) forming a cage and slightly tilted towards Os(1)– Os(2) edge as is evident from the non-orthogonal bond angles S(1)–Os(1)–Os(4), S(2)–Os(2)–Os(4) and S(3)–Os(4)– C(9).Facial cappings by this type of ligand have also been observed in [Ru3(CO)9(SCH2SCH2SCH2)],26 [Rh4(CO)9- (SCH2SCH2SCH2)] 27 and [Ir4(CO)9(SCH2SCH2SCH2)].28 The Fig. 4 Molecular structure of cluster 4. Fig. 5 Molecular structure of cluster 5.2080 J. Chem. Soc., Dalton Trans., 1999, 2077–2086 Table 3 Spectroscopic data for clusters 1 to 9 Complex IR, n& (CO)a/cm21 1H NMR, d(J/Hz) b Mass, m/z c 2073s, 2018s, 2005s, 1983m 3.07–3.12 (m, 4 H, Ha, Hb) 2.76–2.83 (m, 4 H, Ha9 or Hb9) 1.41 (m, 1 H, He9 or Hf9) 1.35 (m, 1 H, He or Hf) 1.11 (m, 2 H, Hc9 or Hd9) 0.96 (m, 2 H, Hc or Hd) 0.87 (m, 2 H, Hc or Hd9) 0.77 (m, 2 H, Hc or Hd) 0.57 (m, 1 H, He9 or Hf9) 0.35 (m, 1 H, He or Hf) 1765 (1765) 2089w, 2068m, 2055s, 2032vs, 2010m, 1979w 2.86 (m, 2 H, Ha or Hb) 2.47 (m, 2 H, Ha or Hb) 1.35 (m, 1 H, He or Hf) 0.96 (m, 2 H, Hc or Hd) 0.77 (m, 2 H, Hc or Hd) 0.35 (m, 1 H, He or Hf) 1473 (1473) 2080w, 2068w, 2024 (br) 4.34 (m, 2 H, Ha, Ha9) 3.57 (m, 2 H, Hb, Hb9) 2.05 (m, 2 H, Hd, Hd9) 1.96 (m, 2 H, Hc, Hc9) 1709 (1709)J.Chem. Soc., Dalton Trans., 1999, 2077–2086 2081 Table 3 (Contd.) Complex IR, n& (CO)a/cm21 1H NMR, d(J/Hz) b Mass, m/z c 2093w, 2080w, 2064s, 2032s, 2024vs, 2001s, 1941w 3.78–3.80 (m, 4 H, Ha, Ha9 or Hb, Hb9) 2.86–2.92 (m, 4 H, Hb, Hb9 or Ha, Ha9), 1.92–2.02 (m, 4 H, Hc, Hd or Hc9, Hd9) 1.67–1.71 (m, 4 H, Hc9, Hd9 or Hc, Hd) 1829 (1829) 2084s, 2039s, 2020vs, 1999m, 1968w 2.86–2.92 (m, 2 H, Ha, Hb) 1.92–2.02 (m, 2 H, He, Hf) 1.67–1.71 (m, 2 H, Hc, Hd) 1699 (1699) 2078s, 2020vs, 2006vs, 1985 (br) 3.34–3.43 (m, 4 H, Ha, Hb, Ha9, Hb9) 2.45–2.51 (m, 4 H, Hc, Hd, Hc9, Hd9) 1769 (1769) 2081s, 2019vs, 2005vs, 1983 (br) 4.67–4.78 (m, 2 H, Ha9 or Hb9) 4.41–4.44 (m, 2 H, Hc9 or Hd9) 3.38–3.41 (m, 2 H, Ha or Hb) 2.62–2.78 (m, 2 H, Hc or Hd) 1769 (1769)2082 J.Chem. Soc., Dalton Trans., 1999, 2077–2086 Table 3 (Contd.) Complex IR, n& (CO)a/cm21 1H NMR, d(J/Hz) b Mass, m/z c 2064s, 2037m, 2024s 3.91–3.94 (m, 2 H, Ha or Hb) 2.67–2.69 (m, 2 H, Hc or Hd) 1693 (1693) 2084m, 2033m, 2020s, 2006vs, 1979w 3.68–3.75 (m, 4 H, Ha, Hb, Ha9, Hb9) 2.76–2.85 (m, 4 H, Hc, Hd, Hc9, Hd9) 211.98 (s, 1 H, OsH) 215.98 (s, 1 H, OsH) 1770 (1770) a In CH2Cl2. b In C6D6. c Simulated values given in parentheses. metal–metal bond lengths in the triangle lie in the range of 2.793 to 2.836 Å which are comparable to the corresponding Os–Os vectors in [Os6(CO)18].13 All the carbonyl ligands are terminally bonded except the one bridging the polyhedral edge Os(2)–Os(5) which leads to the shortest metal–metal bond distance in the structure.Cluster 5 is rather stable and does not react with molecular hydrogen or carbon monoxide even under forcing conditions (in refluxing CHCl3 for 24 h).The strong and rigid ligand chelation of ligand L4 may account for the relative inertness of the complex towards hydrogenation and carboxylation. Compounds 6 and 7 were isolated as a pair of isomers from Table 4 Selected bond distances (Å) and angles (8) for cluster 3 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) 2.713(2) 2.861(2) 2.836(2) 2.865(2) 2.771(3) 2.662(2) 2.755(2) 2.852(2) 2.849(2) Os(4)–Os(5) Os(4)–Os(6) Os(5)–Os(6) Os(2)–S(1) Os(5)–S(1) Os(2)–S(1)–Os(5) Os(5)–Os(2)–S(1) S(1)–Os(5)–Os(2) 2.759(2) 2.867(2) 2.714(2) 2.27(1) 2.26(1) 71.9(3) 53.9(3) 54.3(3) the reactions with 1,4-thioxane (L5) containing hetero-donor atoms S and O.The molecular structure of 6 is illustrated in Fig. 6 and pertinent bond parameters are given in Table 7. The bonding architecture of cluster 6 is similar to that of 1 with all their bonding parameters comparable. All attempts to obtain suitable single crystals of 7 were not successful owing to its instability in solution, however it is thought to be of similar structure to 6 according to FAB MS and 1H NMR spectroscopies.In fact, the high resemblances between the IR patterns of 6 and 7, along with their identical molecular ion envelopes exhibited in mass spectra, imply that both structures have similar carbonyl ligand depositions with disubstituted ligand L5. Table 5 Selected bond distances (Å) for cluster 4 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) 2.869(2) 2.829(2) 2.808(2) 2.794(2) 2.779(2) 2.756(2) 2.787(2) 2.786(2) Os(3)–Os(6) Os(4)–Os(5) Os(4)–Os(6) Os(5)–Os(6) Os(1)–S(1) Os(6)–S(3) 2.797(2) 2.799(2) 2.828(2) 2.876(2) 2.36(1) 2.39(1)J.Chem. Soc., Dalton Trans., 1999, 2077–2086 2083 The only diVerence lies in their 1H NMR spectra, with a considerably downfield shift shown by two groups of multiplets. With reference to the 1H NMR spectrum of cluster 6, this is unambiguously assigned to the terminally bonded ligand L5.We believe the co-ordination of Os(1) via O atom in L5 results in such changes in chemical shift. In the literature, the occurrence of terminally co-ordinated O-donors is not uncommon and there are many alkoxytriosmium clusters such as [Os3(CO)10- (m-H)(m-OR)], R = H,29 Me30 or nBu.31 However, the diVerence in chemical reactivities between 1 and 6 towards neutral molecules such as CO and H2 can be attributed to the presence of the reactive oxygen atom in L5, which conferred higher reactivities to 6.Facile displacement of a terminally bonded atom S(1) by a Fig. 6 Molecular structure of cluster 6. Table 6 Selected bond distances (Å) and angles (8) for cluster 5 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) Os(4)–Os(5) Os(4)–Os(6) 2.793(3) 2.740(2) 2.836(2) 2.845(3) 2.824(3) 2.663(3) 2.733(3) 2.791(3) 2.863(3) 2.835(3) 2.812(3) Os(5)–Os(6) Os(1)–S(1) Os(2)–S(2) Os(4)–S(3) S(1)–Os(1)–Os(4) S(1)–Os(1)–Os(2) S(2)–Os(2)–Os(1) S(2)–Os(2)–Os(4) S(3)–Os(4)–Os(2) S(3)–Os(4)–Os(1) 2.704(3) 2.38(1) 2.42(1) 2.43(1) 94.6(3) 95.0(3) 92.3(3) 95.9(3) 89.7(3) 90.1(3) Table 7 Selected bond distances (Å) and angles (8) for cluster 6 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) 2.733(1) 2.843(1) 2.924(1) 2.850(1) 2.764(1) 2.679(1) 2.754(1) 2.855(1) 2.848(1) Os(4)–Os(5) Os(4)–Os(6) Os(5)–Os(6) Os(1)–S(1) Os(2)–S(2) Os(5)–S(2) Os(2)–S(2)–Os(5) Os(5)–Os(2)–S(2) S(2)–Os(5)–Os(2) 2.795(1) 2.837(1) 2.731(1) 2.385(4) 2.272(4) 2.284(4) 72.0(1) 54.2(1) 53.8(1) Table 8 Selected bond distances (Å) and angles (8) for cluster 8 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(2)–Os(5) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) 2.708(1) 2.880(1) 2.857(1) 2.788(1) 2.865(1) 2.666(1) 2.769(1) 2.757(1) 2.884(1) Os(4)–Os(5) Os(4)–Os(6) Os(5)–Os(6) Os(2)–S(1) Os(5)–S(1) Os(2)–S(1)–Os(5) Os(5)–Os(2)–S(1) S(1)–Os(5)–Os(2) 2.878(1) 2.841(1) 2.726(1) 2.281(5) 2.263(5) 71.8(2) 53.8(1) 54.4(1) CO ligand from the metal framework aVorded compound 8.Complete conversion into the parent cluster, [Os6(CO)18], could be achieved upon further refluxing of compound 6 in CHCl3 under a stream of CO for 48 h. Indeed, the metal–ligand architecture of 8 can be obtained in a similar reaction with L2, as discussed previously. A perspective view of 8 is illustrated in Fig. 7 with selected bond parameters in Table 8. As in 3, all sixteen carbonyl ligands are terminally bonded in cluster 8 with a thioxane ligand L5 bridging the edge Os(2)–Os(5) of the central tetrahedron. The bonding parameters of both structures are in good agreement, except for the Os(3)–Os(6) vector in cluster 8 which is slightly elongated, leading to a distorted bicapped tetrahedron. Hydrogenation of compound 6 induced bond cleavage along the Os(2)–Os(5) edge and gave a new cluster 9.Its molecular structure is presented in Fig. 8. Important bond parameters are summarized in Table 9. The metal core consists of six osmium atoms that are arranged in the form of two fused tetrahedra sharing a common edge. This kind of metal skeleton arrangement is rather rare in hexaosmium systems and a similar core is reported for [Os6(CO)12(m-CNMe2)2(m3-SMe)2(m-H)2].32 The cleavage of the Os(2)–Os(5) bond not only gives a relatively long non-bonding Os ? ? ? Os distance [3.50 vs. 2.679 Å in 6] but also results in longer average Os–S distances [2.300 vs. 2.273 in 1 and 2.278 Å in 6 respectively]. Both 1H NMR and a potentialenergy calculation 33 suggested the presence of a pair of bridging hydrides across the Os(1)–Os(5) and Os(5)–Os(6) edges so as to complete the co-ordination sphere by 86 CVE which the edge-fused bi-tetrahedron should attain.34 Experimental All reactions and manipulations were carried out in an Fig. 7 Molecular structure of cluster 8. Table 9 Selected bond distances (Å) and angles (8) for cluster 9 Os(1)–Os(2) Os(1)–Os(3) Os(1)–Os(4) Os(2)–Os(3) Os(2)–Os(4) Os(3)–Os(4) Os(3)–Os(5) Os(3)–Os(6) Os(4)–Os(5) 2.896(1) 2.808(1) 2.869(2) 2.888(1) 2.841(2) 2.771(2) 2.879(2) 2.872(2) 2.826(1) Os(4)–Os(6) Os(5)–Os(6) Os(1)–S(1) Os(2)–S(2) Os(5)–S(2) Os(2)–S(2)–Os(5) S(2)–Os(5)–Os(3) Os(5)–Os(3)–Os(2) Os(3)–Os(2)–S(2) 2.778(2) 2.857(2) 2.380(7) 2.321(7) 2.278(6) 99.1(2) 88.8(2) 74.7(4) 87.7(2)2084 J.Chem. Soc., Dalton Trans., 1999, 2077–2086 atmosphere of dry argon using standard Schlenk techniques. All solvents were purified and dried by standard methods prior to use.35 Chemicals were purchased from Aldrich chemicals and used as received. The compound [Os6(CO)18] 36 was obtained from vacuum pyrolysis of [Os3(CO)12] and the activated cluster [Os6(CO)16(MeCN)2] was prepared by the literature method.37 Infrared spectra were recorded on a Bio-Rad FTS-7 spectrometer, using 0.5 mm thick calcium fluoride solution cells, proton NMR spectra on a Bruker DPX 300 spectrometer using C6D6 with reference to SiMe4 (d = 0) and mass spectra on a Finnigan MAT 95 instrument by the fast atom bombardment technique, using m-nitrobenzyl alcohol or a-thioglycerol as the matrix solvent.Elemental analyses were conducted by Butterworth Laboratories, UK. Routine separation of products in air was performed by thin-layer chromatography (TLC) on plates coated with Merck Kieselgel 60 GF254.Syntheses [Os6(CO)15{S(CH2)4CH2}{Ï-S(CH2)4CH2}] 1 and [Os5(CO)15- {S(CH2)4CH2}] 2. The complex [Os6(CO)16(MeCN)2] (112 mg, 0.067 mmol) was dissolved in CH2Cl2 (40 cm3) and stirred with dropwise addition of one equivalent of L1 (0.70 cm3 diluted in 10 cm3 CH2Cl2) under ambient conditions. After the reaction had proceeded for 4 h, the mixture gradually changed from blackish brown to turbid brown.It was then filtered and the volume reduced to 5 cm3 in vacuo. The residue was subsequently purified by TLC using hexane–CH2Cl2 (20 : 10 v/v) as eluent. The first yellow band was found to be [Os3(CO)12] (< 5%, as confirmed by IR spectroscopy). Two major bands of brown cluster 1 (Rf 0.70, 28 mg, 0.016 mmol, 24%) and red cluster 2 (Rf 0.65, 22 mg, 0.015 mmol, 22%) were then eluted consecutively (Found for 1: C, 16.9; H, 1.0; S, 3.6. Calc. for C25H20- O15Os6S2 C, 17.0; H, 1.1; S, 3.6.Found for 2: C, 14.4; H, 0.7; S, 2.1. Calc. for C20H10O15Os5S: C, 14.5; H, 0.7; S, 2.2%). Conversion of compound 1 into 2. A deep brown solution of compound 1 (40 mg, 0.02 mmol) in refluxing CHCl3 was added to an equivalent of L1 (0.23 cm3 diluted in 5 cm3 CH2Cl2). Two bands were eluted [hexane–CH2Cl2 (10 : 10 v/v)] which were identified by solution IR as 2 (Rf 0.60, 18 mg, 0.012 mmol, 54%) and unchanged 1 (Rf 0.45, 10 mg, 0.006 mmol, 24%). [Os6(CO)16{Ï-S(CH2)3SCH2}] 3.Treatment of equimolar L2 (7.2 mg, 0.06 mmol) with [Os6(CO)16(MeCN)2] (100 mg, 0.06 Fig. 8 Molecular structure of cluster 9. mmol) in CH2Cl2 (35 cm3) under ambient conditions over a period of 24 h aVorded a deep brown reaction mixture. Purifi- cation by TLC using hexane–CH2Cl2 (10 : 10 v/v) gave the brown cluster 3 (Rf 0.45, 30 mg, 0.018 mmol, 29%) together with several uncharacterized, low-yield products (Found: C, 14.1; H, 0.5; S, 1.8. Calc. for C10H4O8Os3S: C, 14.3; H, 0.5; S, 1.9%).[Os6(CO)16{S(CH2)2SCH2CH2}2] 4. To a solution of [Os6- (CO)16(MeCN)2] (92 mg, 0.055 mmol) in CH2Cl2 (30 cm3) was added L3 (13 mg, 0.108 mmol). The reaction mixture was stirred at room temperature for 6 h during which time it darkened. Excess of solvent was then removed under reduced pressure, yielding a deep brown residue. This was then dissolved in CH2Cl2 (5 cm3) and subjected to preparative TLC on silica using hexane–CH2Cl2 (10 : 30, v/v) as eluent.The brown cluster 4 was isolated as the major product (Rf 0.78, 32 mg, 0.017 mmol, 32%) (Found: C, 15.9; H, 0.9; S, 6.9. Calc. for C12H8O8- Os3S2: C, 15.7; H, 0.9; S, 7.0%). [Os6(CO)14(Ï-CO)(SCH2SCH2SCH2)] 5. A solution of [Os6- (CO)16(MeCN)2] (75 mg, 0.045 mmol) was stirred with L4 (5.4 mg, 0.045 mmol) in CH2Cl2 (30 cm3) at ambient temperature for 3 h. After reduction in solvent volume, the filtrate was separated by preparative TLC on silica, with an eluent of hexane–CH2Cl2 (10 : 30 v/v).Compound 5 was isolated as the major product (Rf 0.35, 24 mg, 0.014 mmol, 31%) (Found: C, 12.5; H, 0.3; S, 5.5. Calc. for C6H2O5Os2S: C, 12.7; H, 0.4; S, 5.7%). [Os6(CO)15{S(CH2)2OCH2CH2}{Ï-S(CH2)2OCH2CH2}] 6 and [Os6(CO)15{O(CH2)2SCH2CH2}{Ï-S(CH2)2OCH2CH2}] 7. To a CH2Cl2 solution (40 cm3) of [Os6(CO)16(MeCN)2] (100 mg, 0.060 mmol), a dilute solution of an equivalent of L5 (0.56 cm3 diluted in 10 cm3 CH2Cl2) was added dropwise with stirring for 24 h under ambient conditions. After reduction of solvent to ca. 5 cm3, the residue was subjected to preparative TLC for purifi- cation using hexane–CH2Cl2 as eluent (10 : 10 v/v). Two consecutive bands of nearly equal abundance were then eluted: compound 6 (Rf 0.65, 20 mg, 0.011 mmol, 19%) and 7 (Rf 0.55, 18 mg, 0.010 mmol, 17%) (Found for 6: C, 15.8; H, 0.8; S, 3.7. Calc. for C23H16O17Os6S2: C, 15.6; H, 0.9; S, 3.6%). Carboxylation of compound 6. Compound 6 (30 mg, 0.017 mmol) was dissolved in CHCl3 (10 cm3) to give a pale brown solution.It was purged by a stream of carbon monoxide at 1 atm continuously for 3 h and the reaction monitored by spot TLC and IR spectroscopy. Following evaporation of most of the solvent, the residue was purified by preparative TLC using hexane–CH2Cl2 (10 : 10 v/v) as eluent. Two brown bands were eluted, namely [Os6(CO)18] (Rf 0.85, 6 mg, 0.004 mmol, 22%) and cluster 8 (Rf 0.55, 8 mg, 0.005 mmol, 28%) (Found for 8: C, 14.5; H, 0.4; S, 1.9.Calc. for C20H8O17Os6S: C, 14.2; H, 0.5; S, 1.9%). Hydrogenation of compound 6. The procedure described above was followed but using hydrogen instead of carbon monoxide. Compound 9 (Rf 0.40, 8 mg, 0.005 mmol, 27%) was isolated as the only major product along with traces of unchanged 6 (Found for 9: C, 15.8; H, 0.9; S, 3.7. Calc. for C23H18O17Os6S2: C, 15.6; H, 1.0; S, 3.6%). Crystallography Single crystals suitable for X-ray crystallographic analyses for clusters 1–6, 8 and 9 were mounted on a glass fibre (except for 4 and 5) or a Lindermann glass capillary (4 and 5) using epoxy resin.All samples except for 5 were obtained by slow evaporation of a saturated toluene–CHCl3 solution at room temperature for several days. Brown crystals of cluster 5 were obtainedJ. Chem. Soc., Dalton Trans., 1999, 2077–2086 2085 Table 10 Summary of crystal data and data collection parameters for clusters 1–6, 8 and 9 1 2 3 4?0.5C7H8 5?0.5C7H8 6 8 9 Empirical formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 Z m(Mo-Ka)/cm21 No.reflections collected No. unique reflections RR 9 T of data collection/8C R(int) C25H20O15Os6S2 1765.74 Orthorhombic Pbca (no. 61) 20.998(1) 15.341(2) 21.779(2) ——— 7015.7(9) 8 218.18 59908 7090 0.050 0.051 25 0.134 C20H10O15Os5S 1473.35 Monoclinic P21/c (no. 14) 9.062(1) 16.484(2) 18.892(2) — 94.76(1) — 2812.3(5) 4 226.37 23912 5330 0.045 0.053 25 0.079 C20H8O16Os6S2 1709.59 Triclinic P1� (no. 2) 11.976(3) 13.059(4) 9.823(2) 90.43(2) 90.44(2) 87.36(2) 1534.6(6) 2 249.32 4245 4013 0.067 0.067 25 0.064 C27.5H20O16Os6S4 1875.89 Monoclinic P21/n (no. 14) 9.181(1) 28.787(1) 16.533(1) — 102.38(1) — 4268.0(6) 4 180.38 17557 6612 0.065 0.070 25 0.115 C21.5H10O15Os6S3 1745.69 Orthorhombic P212121 (no. 19) 10.120(1) 11.170(1) 29.792(3) ——— 3367.7(5) 4 227.84 21402 2904 0.062 0.050 25 0.100 C23H16O17Os6S2 1769.69 Triclinic P1� (no. 2) 10.314(1) 12.160(2) 14.420(2) 108.01(2) 92.80(1) 102.84(2) 1663.4(5) 2 230.10 14561 5736 0.048 0.062 25 0.077 C20H8O17Os6S 1693.53 Monoclinic P21/c (no. 14) 15.251(1) 12.358(2) 16.422(2) — 102.29(1) — 3024.1(5) 4 252.38 20502 5863 0.053 0.063 25 0.089 C23H18O17Os6S2 1769.69 Triclinic P1� (no. 2) 8.992(1) 10.348(2) 18.637(2) 99.00(2) 94.67(1) 104.25(1) 1647.0(5) 2 232.39 10408 5630 0.061 0.065 25 0.0802086 J. Chem. Soc., Dalton Trans., 1999, 2077–2086 as a solvate of stoichiometry 5?0.5C6H5Me at 210 8C for 2 d. DiVraction data were collected at room temperature on a Rigaku AFC7R diVractometer (for cluster 3) using graphitemonochromated Mo-Ka radiation (l = 0.71073 Å) and w–2q scan technique.Unit-cell parameters were determined from 25 accurately centred reflections. The stability of the crystal was monitored at regular intervals using three standard reflections and no significant decay was observed. For clusters 1, 2, 4–6, 8 and 9, data were collected on a MAR research image plate scanner using graphite-monochromated Mo-Ka radiation (l = 0.71073 Å) and the w scan technique.A summary of the crystallographic data and structure refinement is listed in Table 10. All diVracted intensities were corrected for Lorentzpolarization eVects. Absorption correction by the y-scan method was applied for structure 3. An approximate absorption correction by interimage scaling was applied for 1, 2, 4–6, 8 and 9. Space groups of all the crystals were determined from a combination of Laue symmetry check, and their systematic absences were confirmed by successful refinement of the structures.The structures were solved by a combination of direct methods: SIR 88 38 for 9, SIR 92 39 for 1–6 and 8 along with Fourier-diVerence techniques. Structure refinements were made on F by full-matrix least-squares analysis. The hydrogen atoms of the organic moieties were generated in their idealized positions whereas all metal hydrides were estimated by potentialenergy calculations.33 All calculations were performed on a Silicon-Graphics computer using the program package TEXSAN.40 CCDC reference number 186/1449.See http://www.rsc.org/suppdata/dt/1999/2077/ for crystallographic files in .cif format. Acknowledgements We gratefully acknowledge financial support from the Hong Kong Research Grants Council and the University of Hong Kong. K. S.-Y. L acknowledges the receipt of a postgraduate studentship and a scholarship, administered by the University of Hong Kong and the Epson Foundation respectively.References 1 R. J. Angelici, Acc. Chem. Res., 1988, 21, 387; C. M. Friend and J. T. Roberts, Acc. Chem. Res., 1988, 21, 394; E. J. Markel, G. L. Schrader, N. N. Sauer and R. J. Angelici, J. Catal., 1989, 116, 11; N. N. Sauer, E. J. Markel, G. L. Schrader and R. J. Angelici, J. Catal., 1989, 117, 295; W. R. Moser, G. A. Rossetti, J. T. Gleaves and J. R. Ebner, J. Catal., 1991, 127, 190. 2 R. D. Adams, H.-S. Kim and S. Wang, J.Am. Chem. Soc., 1985, 107, 6107; R. D. Adams, T. S. A. Hor and I. T. Horvath, Inorg. Chem., 1984, 23, 4733; R. D. Adams, J. E. Babin, H.-S. Kim, J. T. Tanner and T. A. Wolfe, J. Am. Chem. Soc., 1990, 112, 3426; R. D. Adams and J. E. Babin, New J. Chem., 1988, 12, 641; R. D. Adams, I. T. Horvath and H.-S. Kim, Organometallics, 1984, 3, 548; R. D. Adams, X. Qu and W. Wu, Organometallics, 1993, 12, 4117. 3 R. D. Adams, I. T. Horvath and K. Natarajan, Organometallics, 1984, 3, 1540; R.D. Adams and L.-W. Yang, J. Am. Chem. Soc., 1983, 105, 235; R. D. Adams, G. Chen, S. Sun, J. T. Tanner and T. A. Wolfe, Organometallics, 1990, 9, 251; R. D. Adams and W. Wu, Inorg. Chem., 1991, 30, 3605. 4 R. D. Adams, I. T. Horvath and L.-W. Yang, Organometallics, 1983, 2, 1257; R. D. Adams, Babin, R. Mathab and S. Wang, Inorg. Chem., 1986, 21, 1623; R. D. Adams, J. E. Babin and K. Natarajan, J. Am. Chem. Soc., 1986, 108, 3518; R. D. Adams, J. E. Babin, M.Tasi and J.-G. Wang, Inorg. Chem., 1987, 26, 3708. 5 (a) R. D. Adams, J. E. Babin and H.-S. Kim, Inorg. Chem., 1986, 25, 1122; (b) R. D. Adams, I. T. Horvath and P. Mathur, Organometallics, 1984, 3, 623; (c) R. D. Adams and J.-G. Wang, Polyhedron, 1989, 8, 1437; (d) R. D. Adams, I. T. Horvath and L.-W. Yang, J. Am. Chem. Soc., 1983, 105, 1533; (e) R. D. Adams, Z. Dawoodi, D. F. Foust and B. E. Segmuller, J. Am. Chem. Soc., 1983, 105, 831. 6 R. D. Adams, I. T. Horvath, P. Mathur, B.E. Segmuller and L.-W. Yang, Organometallics, 1983, 2, 1078. 7 R. D. Adams, G. Chen, S. Sun and Thomas A. Wolfe, J. Am. Chem. Soc., 1990, 112, 868. 8 R. D. Adams and S. B. Falloon, Organometallics, 1995, 14, 4594. 9 W. Y. Wong and W. T. Wong, J. Chem. Soc., Dalton Trans., 1995, 2735. 10 M. R. Churchill and B. G. DeBoer, Inorg. Chem., 1977, 16, 878. 11 R. D. Adams, J. E. Cortopassi, J. H. Yamamoto and W. Wu, Organometallics, 1993, 12, 4955. 12 A. J. Deeming, S. Donovan-Mtunzi, S.E. Kabir and P. J. Manning, J. Chem. Soc., Dalton Trans, 1985, 1037. 13 R. Mason, K. M. Thomas and D. M. P. Mingos, J. Am. Chem. Soc., 1973, 95, 3802. 14 B. F. G. Johnson and J. Lewis, Adv. Inorg. Radiochem., 1980, 24, 225. 15 C. Coutoure, D. H. Farrar, M. P. Gomez-Sal, B. F. G. Johnson, R. A. Kamarudin, J. Lewis and P. R. Raithby, Acta Crystallogr., Sect. C, 1986, 42, 163. 16 C. Coutoure and D. H. Farrar, J. Chem. Soc., Dalton Trans., 1987, 2253. 17 R. D. Adams and M.P. Pompeo, J. Am. Chem. Soc., 1991, 113, 1619. 18 R. D. Adams, J. A. Belinski and M. P. Pompeo, Organometallics, 1991, 10, 2539. 19 C. R. Eady, B. F. G. Johnson, J. Lewis, B. R. Reichert and G. M. Sheldrick, J. Chem. Soc., Chem. Commun., 1976, 271. 20 W. Wang, R. J. Batchelor, F. W. B. Einstein, C.-Y. Lu and R. K. Pomeroy, Organometallics, 1993, 12, 3598. 21 R. Khattar, B. F. G. Johnson, J. Lewis, P. R. Raithby and M. J. Rosales, J. Chem. Soc., Dalton Trans., 1990, 2167. 22 F. A. Cotton, Prog. Inorg. Chem., 1976, 21, 1. 23 G. R. John, B. F. G. Johnson and J. Lewis, J. Organomet. Chem., 1979, 181, 143. 24 J. N. Nicholls and M. D. Vargas, Inorg. Synth., 1989, 26, 989. 25 R. D. Adams, L. Chen and J. H. Yamamoto, Inorg. Chim. Acta, 1995, 229, 47. 26 S. Rossi, K. Kallinen, J. Pursianinen, T. T. Pakkanen and T. A. Pakkanen, J. Organomet. Chem., 1992, 440, 367. 27 R. J. Crowte, J. Evans and M. Webster, J. Chem. Soc., Chem. Commun., 1984, 1344. 28 G. Suardi, A. Strawczynski, R. Ros, R. Roulet, F. Grepioni and D. Braga, Helv. Chim. Acta, 1990, 73, 154. 29 M. G. Karpov, S. P. Tunik, V. R. Denisov, G. L. Starova, A. B. Nikol’skii, F. M. Dolgushin, A. I. Yanovsky and Yu. T. Struchkov, J. Organomet. Chem., 1995, 219, 485. 30 M. R. Churchill and H. J. Wasserman, Inorg. Chem., 1980, 19, 2391. 31 W. T. Wong, unpublished results. 32 R. D. Adams and J. E. Babin, Inorg. Chem., 1987, 26, 980. 33 A. G. Orpen, J. Chem. Soc., Dalton Trans., 1980, 2509. 34 M. McPartlin and D. M. P. Mingos, Polyhedron, 1984, 3, 1321. 35 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory Chemicals, 3rd edn., Pergamon, Oxford, 1988. 36 C. R. Eady, B. F. G. Johnson and J. Lewis, J. Chem. Soc., Dalton Trans., 1975, 2606. 37 B. F. G. Johnson, R. A. Kamarudin, F. J. Lahoz, J. Lewis and P. R. Raithby, J. Chem. Soc., Dalton Trans., 1988, 1205. 38 SIR 88, M. C. Burla, M. Camalli, G. Cascarano, C. Giacovazzo, G. Polidor, R. Spagna and D. Viterbo, J. Appl. Crystallogr., 1989, 22, 389. 39 SIR 92, A. Altomare, M.C. Burla, M. Camalli, M. Cascarano, C. Giacovazzo, A. Guagliardi and G. Polidori, J. Appl. Crystallogr, 1994, 27, 435. 40 TEXSAN, Crystal Structure Analysis Package, Molecular Structure Corporation, Houston, TX, 1985 and 1992. Paper 9/00808J

ISSN:1477-9226    

DOI:10.1039/a900808j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 2079 Lanthanide complexes of the tetradentate N-donor ligand dihydrobis[3-(2-pyridyl)pyrazolyl]borate and the terdentate N-donor ligand 2,6-bis(1H-pyrazol-3-yl)pyridine: syntheses, crystal structures and solution structures based on luminescence lifetime studies David A. Bardwell, John C. JeVery, Peter L. Jones, Jon A. McCleverty,* Elefteria Psillakis, Zoe Reeves and Michael D. Ward * School of Chemistry, University of Bristol, Cantock’s Close, Bristol, UK BS8 1TS Lanthanide complexes of two polydentate N-donor ligands containing a mixture of pyridyl and pyrazolyl donors have been prepared.Dihydrobis[3-(2-pyridyl)pyrazolyl]borate (L1)2 is a tetradentate ligand with two bidentate chelating pyridyl/pyrazolyl arms linked by an apical BH2 group; 2,6-bis(1H-pyrazol-3-yl)pyridine (L2) is a terdentate chelating ligand reminiscent of terpyridine. Reaction of L1 with lanthanide salts gave complexes of the type [M(L1)2X]n1; the crystal structures of [Eu(L1)2(dmf)][ClO4]?2.5CH2Cl2, [Tb(L1)2(NO3)]?2CH2Cl2 and [Tb(L1)2(H2O)][L1]?H2O?0.5CH2Cl2 were determined and all contain two tetradentate ligands L1 and an ancillary ligand X [dimethylformamide (dmf), nitrate or water] whose nature depends on the reaction/recrystallisation conditions to complete the co-ordination sphere.Luminescence studies of [Tb(L1)2(NO3)] in water or D2O and MeOH or CD3OD showed that in methanol the solvation number q is ª1.8, consistent with displacement of nitrate by the solvent; however in water q ª 4.5, indicating additional displacement of some of the N-donor heterocyclic rings of L1 by co-ordinating water molecules.Reaction of L2 with lanthanide salts afforded [M(L2)3]31, all isolated as their hexafluorophosphate salts. The crystal structures of three of these (M = Eu, Gd or Ho) showed that they are isostructural and isomorphous, with tricapped trigonal-prismatic nine-co-ordinate geometries similar to that of [M(terpy)3]31 (terpy = 2,29:69,20-terpyridine).Luminescence studies of [Tb(L2)3][PF6]3 gave a solvation number q of 0.6 in methanol, which is small enough to be accounted for by second-sphere solvation effects alone and therefore suggests that the nine-co-ordinate structure is retained in methanol solution. However in water, q is again ª4.5, due to displacement of some of the donor groups of the L2 ligands by water. The co-ordination chemistry of lanthanides has become of increasing significance in the last few years due to the wide variety of potential applications of lanthanide complexes.Luminescent complexes of EuIII and TbIII may be useful in medicine as luminescent probes,1 and highly paramagnetic complexes, generally of GdIII, are used as contrast agents to enhance the output from magnetic resonance imaging (MRI) scanners.2 In the area of supramolecular co-ordination chemistry, lanthanide ions are used as templates for the assembly of triple helicate complexes using compartmental ligands with terdentate N-donor binding sites, because of their tendency to form nine-co-ordinate complexes with three terdentate fragments. 3 The occurrence of an antenna effect, which allows sensitisation of the metal-centred luminescence via energy transfer from an aromatic ligand which acts as a light harvester, means that lanthanide complexes with aromatic ligands are of interest for supramolecular light-conversion devices.4 Detailed NMR and luminescence studies have allowed determination of the relationship between solid-state and solution properties, in particular the extent of solvation.5,6 There is therefore considerable interest in the development of new multidentate ligands containing an aromatic chromophore for lanthanide co-ordination chemistry.To this end we recently described the co-ordination chemistry of lanthanide ions with the hexadentate N-donor podand ligand hydrotris[3-(2-pyridyl)- pyrazolyl]borate (L);7 we found that the combination of a multidentate N-donor ligand set and a negative charge made this compound a highly effective ligand, forming complexes with both 1 : 1 and 1: 2 (12-co-ordinate) metal : ligand ratios depending on the conditions.We describe in the first part of this paper the synthesis of the simpler tetradentate analogue dihydrobis[3-(2-pyridyl)pyrazolyl]borate (L1), in which two bidentate chelating arms are linked at the apical anionic BH2 2 group, and the crystal structures of some of its lanthanide complexes together with some preliminary luminescence properties.In the second part of this paper we describe some homoleptic nine-co-ordinate lanthanide complexes of the terdentate N-donor ligand 2,6-bis(1H-pyrazol-3-yl)pyridine (L2). As mentioned above, terdentate N-donor ligands such as 2,29 : 69,20-terpyridine (terpy) 8 and various structural analogues 9–11 have been of particular interest in lanthanide co-ordination chemistry because of their structural and photophysical properties.The preparation of L2 was reported a while ago,12 but as far as we are aware the only co-ordination chemistry that has been described with it was the study of a mononuclear iron(II) complex showing spin-crossover behaviour.13 The crystal structures of the complexes and some preliminary luminescence studies are described.2080 J. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 Experimental Instrumentation used for standard spectroscopic and analytical studies has been described previously.7 3-(2-Pyridyl)pyrazole and 2,6-bis(pyrazol-3-yl)pyridine (L2) were prepared according to the literature method.12 Luminescence spectra were recorded using a Perkin-Elmer LS-50B spectrofluorimeter equipped with a Hamamatsu R928 photomultiplier tube, using excitation and emission slit widths of 5–12 nm depending on the intensity of the emission. Phosphorescence lifetimes (t) were measured with the instrument in time-resolved mode, and are the average of at least three independent measurements which were made by monitoring the decay at a wavelength corresponding to the maximum intensity of the emission spectrum (546 nm for Tb), following pulsed excitation. The intensity of the emission after the pulsed excitation was monitored after 20 different delay times spanning at least two lifetimes.The resulting first-order decay curves gave linear plots of ln I vs.t from which the lifetime was calculated by t = (ln 2)/slope. The number of co-ordinated solvent molecules (q) for the terbium(III) complexes was calculated from the Horrocks equation, q = n(tH 21 2 tD 21), where tH is the lifetime in the protonated solvent, (water or MeOH), tD that lifetime in the corresponding deuteriated solvent, and the values of n are 4.2 (Tb in water–D2O) or 8.4 (Tb in MeOH–CD3OD).6 Preparations Potassium dihydrobis[3-(2-pyridyl)pyrazolyl]borate KL1.A mixture of 3-(2-pyridyl)pyrazole (3.50 g, 24.1 mmol) and KBH4 (0.49 g, 9.0 mmol) was ground together finely with a mortar and pestle and then gradually heated to 150 8C. Melting occurred at ca. 120 8C, at which point evolution of H2 commenced. The temperature was maintained at 150 8C for 30 min, after which time evolution of H2 had ceased. The solid white mixture was cooled and warm toluene (100 cm3) was added. The suspension was stirred vigorously overnight to allow the excess of unreacted 3-(2-pyridyl)pyrazole to dissolve.The product was filtered off as a white solid, which was washed further with several portions of toluene and then hexane and dried. Yield of KL1: 2.48 g (81%). Negative-ion FAB mass spectrum: m/z 301 [(L1)2]. NMR [(CD3)2CO] 1H, d 8.51 (1 H, ddd, J 4.9, 1.7, 1.0, pyridyl H6), 7.8–7.7 (2 H, m, pyridyl H3, H4), 7.54 (1 H, d, J 2.0, pyrazolyl H5), 7.13 (1 H, ddd, J 7.3, 4.9, 1.5, pyridyl H5) and 6.59 (1 H, d, J 2.0 Hz, pyrazolyl H4); 13C, d 155.1, 151.7, 150.1, 137.3, 136.3, 122.0, 120.8 and 102.6.nBH(KBr disc): 2378, 2349, 2318 and 2254 cm21 (Found: C, 55.8; H, 4.2; N, 24.7. Calc. for C16H14BN6K: C, 56.5; H, 4.1; N, 24.7%). [Eu(L1)2(dmf)][ClO4]. To a solution of EuCl3?6H2O (0.055 g, 0.15 mmol) in methanol (20 cm3) was added a solution of KL1 (0.1 g, 0.29 mmol) in methanol (20 cm3). After stirring for 1 h, aqueous NaClO4 (excess) was added to give a white precipitate which was filtered off and dried.Recrystallisation from dimethylformamide (dmf)–diethyl ether afforded 0.064 g (46%) of the product as a white powder. X-Ray-quality crystals of [Eu(L1)2(dmf)][ClO4]?2.5CH2Cl2 were grown from CH2Cl2– hexane mixtures. [Tb(L1)2(NO3)]. To a solution of Tb(NO3)3?5H2O (0.128 g, 0.30 mmol) in methanol (20 cm3) was added a solution of KL1 (0.20 g, 0.59 mmol) in methanol (20 cm3). The mixture was allowed to stir for 1 h, and then water was added until precipitation of the product as a white solid was complete.The fine white precipitate was extracted with several portions of CH2Cl2 which were combined, dried, and evaporated to dryness to give [Tb(L1)2(NO3)] in 0.20 g (81%) yield. X-Ray-quality crystals of [Tb(L1)2(NO3)]?2CH2Cl2 were grown from CH2Cl2–hexane. [Tb(L1)2(H2O)][L1]?H2O. To a solution of TbCl3?6H2O (0.055 g, 0.15 mmol) in methanol (20 cm3) was added a solution of KL1 (0.1 g, 0.29 mmol) in methanol (20 cm3). After stirring for 1 h, the product was precipitated by addition of water.The fine white precipitate was extracted with several portions of CH2Cl2 which were combined, dried, and evaporated to dryness. Recrystallisation from CH2Cl2–hexane afforded [Tb(L1)2(H2O)]- [L1]?H2O?0.5CH2Cl2 as colourless X-ray-quality crystals (0.063 g, 39%). [M(L2)3][PF6]3 (M = Eu, Gd, Tb or Ho). A solution of the appropriate MCl3?6H2O (0.33 mmol) and L2 (0.21 g, 1 mmol) in methanol (10 cm3) was stirred for 30 min.A solution of an excess of KPF6 in water was then added, resulting in a white precipitate. After overnight cooling to allow complete precipitation, the solid was filtered off and dried in vacuo to afford [M(L2)3][PF6]3 in 75–90% yield. The crude materials were readily crystallised by slow evaporation from methanol to give X-rayquality crystals. Mass spectra and elemental analyses for all of the complexes are summarised in Table 1. Crystallography Suitable crystals were quickly transferred from the motherliquor to a stream of cold N2 at 2100 8C on a Siemens SMART diffractometer fitted with a CCD-type area detector.A detailed experimental description of the methods used for data collection and integration using the SMART system has been published. 7 Table 2 contains a summary of the crystal parameters, data collection and refinement. In all cases the structures were solved by conventional heavy-atom or direct methods and refined by the full-matrix least-squares method on all F2 data using the SHELXTL 5.03 package on Silicon Graphics Indigo- R4000 or Indy computers.14 Non-hydrogen atoms were refined with anisotropic thermal parameters; hydrogen atoms were included in calculated positions and refined with isotropic thermal parameters.Crystals of [Eu(L1)2(dmf)][ClO4]?2.5CH2Cl2 were needles for which an accurate absorption correction was difficult, resulting in significant residual peaks in the final electron-density map close to the metal centre.Apart from one complex formula unit, the asymmetric unit contains 2.5 molecules of CH2Cl2 of which the half molecule is disordered across an inversion centre. The disorder is such that the two Cl atoms are ordered, but the carbon atom [C(30)] occurs in two positions with equal site occupation factors. Hydrogen atoms were not included for the disordered carbon atom. The structure of [Tb(L1)2(H2O)][L1]?H2O?0.5CH2Cl2 showed a remarkable amount of disorder in the positions of many of the atoms of the L1 ligands (see Figs. 2 and 3). This arises from two significantly different conformers which are present in the crystal. The disordered atoms could be separated into two sets; refinement of the fractional site occupancies of the two sets converged at approximately 50 : 50 so the site occupancies were fixed at that value in the final refinement. The coordinated N atoms of the ligands L1 were common to both conformers; the two conformations are generated by movement of the aromatic rings about their point of attachment to the metal.The structural determination of [Tb(L1)2(NO3)]?2CH2Cl2 presented no problems. The three compounds [M(L2)3][PF6]3 (M = Eu, Gd or Ho) are isostructural and isomorphous, all crystallising in space group C2/c such that the complex cation and one of the hexafluorophosphate anions lie astride C2 axes. The hexafluorophosphate anions were disordered in the same manner in every case, but the two disordered components could be successfully resolved.In [Gd(L2)3][PF6]3 isotropic restraints were applied to fluorine atoms F(119), F(129) and F(149) (where the prime denotes the minor component of the disorder) to keep the refinement stable. Atomic coordinates, thermal parameters, and bond lengthsJ. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 2081 Table 1 Analytical and mass spectroscopic data for the new complexes Elemental analyses (%) a FAB mass spectral data [m/z, relative intensity (%), assignment] Complex [Eu(L1)2(dmf)][ClO4] [Tb(L1)2(NO3)] [Tb(L1)2(H2O)][L1]?H2O [Eu(L2)3][PF6]3 [Gd(L2)3][PF6]3 [Tb(L2)3][PF6]3 [Ho(L2)3][PF6]3 45.0 (45.4) 46.0 (46.6) 48.8 (48.5) 33.4 (32.5) 31.3 (32.3) 31.7 (32.3) 32.8 (32.1) 3.4 (3.7) 3.4 (3.4) 3.7 (4.0) 2.4 (2.2) 2.3 (2.2) 2.1 (2.2) 2.2 (2.2) 19.6 (19.7) 21.6 (22.1) 20.5 (20.6) d 18.0 (17.2) 16.7 (17.1) 16.7 (17.1) 17.7 (17.0) 906 755 597 454 297 761 604 447 912 761 784 613 593 573 1081 935 789 618 598 578 1083 790 619 599 579 796 625 605 585 3 100 30 25 32 100 30 24 3 100 28 52 38 100 4 10 64 50 38 100 4 56 45 35 100 22 60 34 100 b[ Eu(L1)2]1 [EuL1(pypz)]1c [EuL1]1 [Eu(pypz)]1 [Tb(L1)2]1 [TbL1(pypz)]1 [Tb(pypz)2]1 b[ Tb(L1)2]1 [EuL2(L2 2 H)2]1 [Eu(L2)2F2]1 [EuL2(L2 2 H)F]1 [Eu(L2 2 H)2]1 [Gd(L2)3(PF6)2]1 [Gd(L2)2(L2 2 H)(PF6)]1 [GdL2(L2 2 H)2]1 [Gd(L2)2F2]1 [GdL2(L2 2 H)F]1 [Gd(L2 2 H)2]1 [Tb(L2)3(PF6)2]1 [TbL2(L2 2 H)2]1 [Tb(L2)2F2]1 [TbL2(L2 2 H)F]1 [Tb(L2 2 H)2]1 [HoL2(L2 2 H)2]1 [Ho(L2)2F2]1 [HoL2(L2 2 H)F]1 [Ho(L2 2 H)2]1 a Calculated values in parentheses.b This peak is at 151 mass units above the parent ion [M(L1)2]1. The matrix (3-nitrobenzyl alcohol) has a relative molecular mass of 153; the peaks are therefore most likely due to a matrix adduct of some sort. c pypz = 3-(2-Pyridyl)pyrazole, the bidentate ‘arm’ arising from fragmentation of the ligand L1 (see text). d Analysis calculated with 1.5 molecules of CH2Cl2; the sample for analysis was crystallised from this solvent. and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC).See Instructions for Authors, J. Chem. Soc., Dalton Trans., 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/500. Results and Discussion Synthesis of L1 The new ligand L1 was prepared by reaction of 3-(2-pyridyl)- pyrazole with KBH4, in a 2.5 : 1 ratio and at a temperature (150 8C) such that formation of the bis(pyrazolyl)borate occurred readily but further reaction to give the tris(pyrazolyl)- borate did not.This synthesis follows the usual route for preparing bidentate bis(pyrazol-1-yl)borates from substituted pyrazoles.15,16 Compound L1 therefore has two bidentate arms linked by a flexible BH2 2 spacer, and we can envisage two different co-ordination modes for this ligand. Depending on the charge and stereoelectronic preferences of the metal ion, it could (i) co-ordinate as a tetradentate ligand to a single metal ion, or (ii) act as a bis-bidentate bridge, co-ordinating each arm to a different metal ion, which generally leads to formation of polynuclear helicates.16,17 For example with the related ligand bis[3-(2-pyridyl)pyrazol-1-yl]phosphinate (L3), which is similar to L1 but contains a phosphinate PO2 2 bridge instead of BH2 2, both types of co-ordination have been observed.17 Syntheses and structures of lanthanide complexes of L1 Reaction of L1 (as its potassium salt) with lanthanide(III) salts (nitrate, chloride) resulted in complexes of the form [LnL1 2- X]n1. In all cases two ligands L1 are co-ordinated to a single metal centre, which accounts for eight co-ordination sites.The remaining site(s) and the nature of the anion varied with the details of the synthesis and recrystallisation. Reaction of EuCl3?6H2O with 2 equivalents of KL1 in methanol, followed by precipitation as the perchlorate salt and recrystallisation from dmf–ether, afforded [Eu(L1)2(dmf)]- [ClO4].Initially the ninth co-ordination site is probably occupied by water, which is replaced by dmf during crystallisation. Further subsequent recrystallisation from the non-coordinating solvent mixture CH2Cl2–hexane did not result in displacement of the dmf ligand. In contrast, reaction of Tb(NO3)3?5H2O with 2 equivalents of KL1 under the same conditions resulted in formation of neutral [Tb(L1)2(NO3)] in which the nitrate counter ion is retained in the co-ordination sphere of the TbIII.We were then interested to see what would happen in the absence of both nitrate and any other added anions. Reaction of TbCl3?6H2O with 2 equivalents of KL1 in methanol followed by the addition of water to precipitate a white solid yielded [Tb(L1)2(H2O)][L1]?H2O, in which a ‘free’ molecule of L1 (associated with a water molecule) is acting as the counter ion.All of these complexes were characterised by FAB mass spectroscopy, elemental analyses, and finally X-ray crystallography. In all of the FAB mass spectra (Table 1) the most intense peak corresponded to [Ln(L1)2]1, in which the ancillary ligands (nitrate, water, dmf) have dissociated. Lower-mass peaks were generally observed due to fragmentation of L1 by cleavage of the B]N bonds, leaving bidentate 3-(2-pyridyl)pyrazole fragments attached to the metal centre. The crystal structure of [Eu(L1)2(dmf)][ClO4]?2.5CH2Cl2 is shown in Fig. 1; selected bond lengths and angles are in Table 3. The two ligands L1 are acting as tetradentate chelates, in the manner of the hexadentate podand L but with one bidentate arm missing. The ninth co-ordination site is occupied by an Odonor dmf molecule. As with the lanthanide complexes of L,72082 J. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 the bonds from Eu to the pyrazolyl N donors (2.52–2.60 Å) are rather shorter than those to the pyridyl N donors (2.63–2.72 Å).Each ligand L1 is not planar, but folded due to the tetrahedral geometry of the apical boron atoms, which have N]B]N angles very close to the tetrahedral ideal. The geometry about the metal ion approximates to that of a capped square antiprism [Fig. 2(a)], with N(51), N(71), N(41) and O(101) forming the top plane, N(21), N(61), N(31) and N(81) forming the bottom plane and N(11) capping the bottom plane. For the N(51), N(71), N(41), O(101) set the displacements from the mean plane through them are 20.168, 10.173, 20.170 and 10.165 Å Fig. 1 Crystal structure of the complex cation of [Eu(L1)2(dmf)]- [ClO4]?2.5CH2Cl2 Fig. 2 Co-ordination geometries of the metal centres in (a) [Eu(L1)2- (dmf)][ClO4]?2.5CH2Cl2 and (b) [Tb(L1)2(H2O)][L1]?H2O?0.5CH2Cl2 respectively. The N(21), N(61), N(31), N(81) set forms a more accurate plane, with the displacements of the atoms from the mean plane through them being 10.076, 20.055, 10.055 and 20.076 Å respectively.The angle of intersection between the two mean planes is 98. The structure (Figs. 3 and 4, Table 4) of the [Tb(L1)2(H2O)]1 cation of [Tb(L1)2(H2O)][L1]?H2O?0.5CH2Cl2 is similar to that of [Eu(L1)2(dmf)]1 (above). Again it contains two tetradentate L1 ligands and one monodentate O-donor (water) making a similar nine-co-ordinate geometry, which again may be described as a capped square antiprism [Fig. 2(b)]. In this case the ‘top’ plane is described by the donor atoms O(1), N(61), N(81), N(31), which have displacements of 10.137, 20.133, 10.139 and 20.143 Å respectively from the mean plane through them.The ‘bottom’ plane is described by N(11), N(71), N(41), N(51), which have displacements of 10.077, 20.077, 10.058 and 20.058 Å respectively from the mean plane through them. Atom N(21) caps the bottom plane; the angle of intersection between the two mean planes is 78. In this structure the anion is of particular interest, being a free L1 which encloses a water molecule hydrogen bonded to the two pyrazolyl nitrogen atoms (Fig. 4). The non-bonded distances O(2) ? ? ? N(101) and O(2) ? ? ? N(121) are 2.943 and 2.899 Å respectively which are typical distances for O]H? ? ? N hydrogen bonds.18,19 This is a nice example of a host–guest type interaction via multipoint hydrogen bonding, with an ideal complementarity between the geometric arrangement of the hydrogen-bond donors of the guest (the water O]H bonds) and the hydrogen-bond acceptors of the host (the pyrazolyl N atoms).One of the pyridyl rings of the complex anion [L1?H2O]2 is involved in a hydrogenbonding interaction with the co-ordinated water on the complex cation, i.e. Tb]OH2 ? ? ? N (pyridyl), with a non-bonded O(1) ? ? ? N(111) separation of 2.770 Å. A particular problem with this structure was the presence of a 50 : 50 disorder between two different conformations of Fig. 3 Crystal structure of the complex cation of [Tb(L1)2(H2O)][L1]? H2O?0.5CH2Cl2, showing the two disordered components Fig. 4 Crystal structure of the complex anion of [Tb(L1)2(H2O)][L1]? H2O?0.5CH2Cl2, showing the two disordered componentsJ. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 2083 Table 2 Summary of crystal parameters, data collection and refinement for the six crystal structures a [Eu(L1)2(dmf)]- [ClO4]?2.5CH2Cl2 [Tb(L1)2(NO3)]? 2CH2Cl2 [Tb(L1)2(H2O)][L1]? H2O?0.5CH2Cl2 [Eu(L2)3][PF6]3 [Gd(L2)3][PF6]3 [Ho(L2)3][PF6]3 Formula M Space group a/Å b/Å c/Å b/8 U/Å3 Dc/g cm23 m/mm21 F(000) Crystal size/mm Reflections collected: total, independent, Rint Data, restraints, parameters Final R1, wR2b,c Weighting factors c Largest peak, hole/e Å23 C37.5H40B2Cl6- EuN13O4 1139.10 P21/n 12.207(2) 12.124(2) 31.498(6) 98.02(2) 4616.0(14) 1.639 1.764 2284 0.4 × 0.3 × 0.1 27 933, 10 398, 0.061 10 396, 0, 586 0.066, 0.202 0.1180, 0 2.492, 23.671 C34H32B2Cl4N13O3- Tb 993.07 P21/n 10.3520(7) 21.792(2) 17.818(2) 99.971(7) 3959.0(6) 1.666 2.110 1976 0.45 × 0.3 × 0.3 25 068, 9000, 0.025 8999, 0, 514 0.023, 0.059 0.0319, 0 10.874, 20.746 C48.5H46B3ClN18- O2Tb 1139.83 P21/c 12.167(2) 18.901(2) 22.224(2) 91.207(10) 5109.8(8) 1.482 1.495 2304 0.4 × 0.2 × 0.15 51 516, 11 696, 0.038 11 695, 27, 847 0.063, 0.134 0, 25.5765 11.209, 21.591 C33H27EuF18N15P6 1220.57 C2/c 16.150(2) 21.274(2) 12.710(3) 95.396(14) 4347.5(5) 1.865 1.679 2400 0.25 × 0.2 × 0.1 12 384, 4726, 0.051 4726, 0, 374 0.038, 0.071 0.0293, 0 10.669, 21.107 C33H27F18GdN15P6 1225.86 C2/c 16.134(2) 21.251(3) 12.697(2) 95.42(2) 4334.0(10) 1.879 1.77 2404 0.3 × 0.2 × 0.1 13 794, 4964, 0.108 4962, 48, 359 0.058, 0.129 0.0541, 0 11.436, 21.853 C33H27F18HoN15P6 1233.54 C2/c 16.142(3) 21.210(2) 12.685(2) 95.470(12) 4323.3(1) 1.895 2.068 2416 0.2 × 0.15 × 0.1 13 630, 4940, 0.038 4940, 0, 374 0.028, 0.055 0.0218, 0 10.480, 20.594 a Details in common: graphite-monochromatised Mo-Ka radiation, l “ = 0.710 73 Å; 2q limits for data collection, 3–558; temperature for data collection, 173 K; monoclinic; Z = 4.b Structure was refined on Fo 2 using all data; the value of R1 is given for comparison with older refinements based on Fo with a typical threshold of F > 4s(F). c wR2 = [Sw(Fo 2 2 Fc 2)2/Sw(Fo 2)2]� �� where w21 = [s2(Fo 2) 1 (aP)2 1 bP] and P = [max(Fo 2, 0) 1 2Fc 2]/3. several of the heterocyclic rings in both the complex cation and the complex anion.This is illustrated in Figs. 3 and 4. In one of the co-ordinated L1 ligands of the cation one pyridyl ring has two slightly different orientations with the site of attachment to the metal ion [N(51)] being common to both. The second coordinated L1 is more disordered with both of the pyrazolyl rings, the apical boron atom and one of the pyridyl rings exhibiting disorder over two positions (Fig. 3). It should be emphasised that since the sites of attachment of the heterocyclic rings to the metal ion are common to both components the geometry about the metal ion is the same in each case.In the complex anion [L1?H2O]2 one of the pyridyl rings exhibits this twodisorder (Fig. 4). Despite the extensive disorder, the components could be satisfactorily resolved and the overall level of refinement is quite reasonable (R1 = 0.063). The crystal structure of [Tb(L1)2(NO3)]?2CH2Cl2 is in Fig. 5; selected bond lengths and angles are in Table 5. The structure of the complex is similar to those above with the exception that a bidentate nitrate ion has replaced the monodentate (dmf or water) ligand.This has a noticeable effect on the complex struc- Fig. 5 Crystal structure of the complex unit of [Tb(L1)2(NO3)]? 2CH2Cl2 ture, as the bidentate nitrate ligand has a greater sterically hindering effect close to the metal ion than do the monodentate ligands of the first two complexes. This is illustrated in Fig. 6, which shows alternative views of all three complexes: the two bidentate arms of the L1 ligands which lie near the ancillary ligand are obviously repelled from the nitrate ion more than they are from the monodentate ligands.This effect may be quantified by measuring the angle between the mean planes of these two bidentate (pyridylpyrazolyl) arms. In the above two structures, with monodentate dmf and H2O ligands, the angles between these two ligand fragments are 150 and 1648 respectively; in [Tb(L1)2(NO3)] it is 1738.Similarly, the bonds from Tb Table 3 Selected bond lengths (Å) and angles (8) for [Eu(L1)2(dmf)]- [ClO4]?2.5CH2Cl2 Eu]O(101) Eu]N(11) Eu]N(71) Eu]N(51) Eu]N(31) N(11)]Eu]N(21) N(31)]Eu]N(41) N(51)]Eu]N(61) N(71)]Eu]N(81) O(101)]Eu]N(61) 2.360(5) 2.516(6) 2.531(5) 2.587(6) 2.596(6) 64.3(2) 61.9(2) 62.4(2) 64.1(2) 70.2(2) Eu]N(81) Eu]N(21) Eu]N(61) Eu]N(41) N(11)]Eu]N(31) N(21)]Eu]N(41) N(71)]Eu]N(51) N(81)]Eu]N(61) O(101)]Eu]N(41) 2.627(6) 2.651(6) 2.692(5) 2.718(6) 72.1(2) 140.6(2) 71.3(2) 137.1(2) 73.0(2) Table 4 Selected bond lengths (Å) and angles (8) for [Tb(L1)2(H2O)]- [L1]?H2O?0.5CH2Cl2 Tb]O(1) Tb]N(21) Tb]N(81) Tb]N(41) Tb]N(61) N(41)]Tb]N(31) N(21)]Tb]N(11) N(61)]Tb]N(51) N(81)]Tb]N(71) O(1)]Tb]N(31) 2.353(5) 2.520(7) 2.524(6) 2.564(8) 2.595(6) 62.8(3) 64.6(2) 62.4(2) 64.7(2) 71.2(2) Tb]N(11) Tb]N(71) Tb]N(31) Tb]N(51) N(21)]Tb]N(41) N(11)]Tb]N(31) N(81)]Tb]N(61) N(71)]Tb]N(51) O(1)]Tb]N(51) 2.599(6) 2.622(6) 2.648(8) 2.686(7) 71.6(3) 140.5(2) 71.8(2) 138.9(2) 71.6(2)2084 J.Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 to the pyridyl N atoms N(11) and N(71) are significantly longer (0.14 Å on average) than those to N(31) and N(51), which may be ascribed to their proximity to the bidentate nitrate ligand. Solution luminescence properties of [TbL1 2(NO3)] The luminescence properties of lanthanide complexes in solution, particularly those of TbIII and EuIII, can give useful information regarding their solution structures.5,6 To this end we examined [Tb(L1)2(NO3)] as a representative member of this series of complexes.The results are summarised in Table 6. We showed earlier from conductivity studies with [TbL(NO3)2] that the neutral molecule remained intact in CH2Cl2 solution, but dissociated in water to give [TbL(H2O)q]21 (q ª 3.6) and two nitrate ions, and was therefore soluble in both hydrophobic and hydrophilic solvents.7 We would therefore expect similar behaviour from [Tb(L1)2(NO3)], and indeed found it to be significantly soluble in both CH2Cl2 and water.The complex dissolves in CH2Cl2 solution as a neutral molecule with retention of the co-ordinated nitrate. The electronic spectrum contains ligand-centred p æÆ p* transitions at lmax = 295 and 254 nm, and excitation at either of these wavelengths gives an entirely typical and extremely intense terbium(III) emission spectrum containing the expected sequence of 5D4 æÆ 7Fn transitions, with the n = 6, 5, 4 and 3 components being visible.7 The half-life t of this emission is 1.32 ms.Comparison of the luminescence lifetimes of [Tb(L1)2(NO3)] in water and D2O allows an estimation of the value of q, the number of co-ordinated water molecules in the hydrated species, because of the differing abilities of the O]H and O]D oscillators to quench the metal-based excited state.6 Similarly, measurements in MeOH and MeOD allow estimation of the number of co-ordinated methanol molecules.Considering the methanol solutions first, the difference between the t values (1.18 ms in MeOH and 1.57 ms in CD3OD) gives q = 1.8 ± 0.5 (the error of ±0.5 is generally assumed to be reasonable for this calculation).6 This is consistent with either two co-ordinated methanol molecules (10-co-ordinate metal centre), or with one directly co-ordinated methanol (nine-co-ordinate metal centre) and additional contributions to solvent-based quenching arising from second-sphere interactions. It has been shown by Parker and co-workers 20 that, even in complexes with no solvent molecules directly co-ordinated to the lanthanide centre, second-sphere co-ordination and/or hydrogen bonding of solvent molecules to the ligand can still lead to significant lifetime differences between protonated and deuteriated solvents, leading to apparent q values anywhere between 0 and 1.The complex [Tb(L1)2(NO3)] therefore dissociates in methanol to give [Tb(L1)2(MeOH)q]1 where the value of 1.8 ± 0.5 for q may reasonably be ascribed to one directly co-ordinated methanol molecule with additional second-sphere solvent effects also providing significant quenching.The results in water and D2O were significantly different. The Table 5 Selected bond lengths (Å) and angles (8) for [Tb(L1)2(NO3)]? 2CH2Cl2 Tb]O(1) Tb]N(41) Tb]N(61) Tb]O(2) Tb]N(21) N(21)]Tb]N(11) N(41)]Tb]N(31) N(61)]Tb]N(51) N(81)]Tb]N(71) O(1)]Tb]N(71) O(1)]Tb]N(11) O(1)]Tb]O(2) 2.494(2) 2.512(2) 2.521(2) 2.538(2) 2.602(2) 60.39(6) 64.23(6) 64.32(6) 60.37(6) 91.14(6) 66.26(6) 50.46(6) Tb]N(81) Tb]N(51) Tb]N(31) Tb]N(71) Tb]N(11) N(41)]Tb]N(21) N(31)]Tb]N(11) N(61)]Tb]N(81) N(51)]Tb]N(71) O(2)]Tb]N(71) O(2)]Tb]N(11) 2.622(2) 2.624(2) 2.637(2) 2.736(2) 2.797(2) 71.45(6) 132.17(6) 70.56(6) 132.43(6) 65.92(6) 93.12(6) difference between the t values (0.58 ms in water and 1.56 ms in D2O) gives q = 4.5 ± 0.5.Such a large extent of hydration cannot be accounted for just by dissociation of one nitrate ion, which would permit co-ordination of one or two water molecules (cf.the results in methanol, above). It therefore appears that dissolution of [Tb(L1)2(NO3)] in water results in dissociation of one or more of the heterocyclic rings in addition to nitrate dissociation. Such behaviour has been established in [Eu(terpy)3]31, which in MeCN solution can undergo a ligandbased conformational rearrangement involving rotation of terminal pyridyl rings about the interannular C]C bonds.21 The cis,cis conformation (terdentate terpyridine) can become cis, trans (bidentate terpyridine) or even trans,trans (monodentate Fig. 6 Alternative views of the three structures, emphasising the effect of the steric bulk of the ancillary ligands on the geometries of the complexesJ. Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 2085 terpyridine), allowing co-ordination of acetonitrile molecules to the metal.Acetonitrile is a relatively poor ligand for lanthanide( III) ions, although adducts are known;22 it would therefore be expected on the basis of the known oxophilicity of lanthanide(III) ions that potentially co-ordinating O-donor solvents might also induce partial dissociation of polydentate N-donor ligands. The q value obtained for [Tb(L1)2(NO3)] in water suggests four co-ordinated water molecules plus an additional small contribution to quenching from outer-sphere effects; this could be attained by (for example) dissociation of a bidentate arm of one L1 ligand, or dissociation of some of the terminal pyridyl rings by the conformational rearrangement discussed above, in addition to dissociation of the nitrate ion.The other types of complex cation based on L1 that were crystallographically characterised, [Tb(L1)2(H2O)]1 and [Eu- (L1)2(dmf)]1, will show similar solution behaviour given the labile nature of the monodentate solvent ligands.A more detailed study of the photophysical properties of these complexes is in progress and will be reported separately. Fig. 7 Crystal structure of the complex cation of [Eu(L2)3][PF6]3; the complexes of Gd and Ho are isostructural Fig. 8 Co-ordination geometry of the metal centre in [Eu(L2)3][PF6]3 Table 6 Luminescence lifetimes (ms) of terbium(III) complexes in different solvents a Complex CH2Cl2 MeOH CD3OD Water D2O [Tb(L1)2(NO3)] 1.32 1.18 1.57 0.58 1.56 [Tb(L2)3][PF6]3 1.46 1.09 1.18 0.45 b 0.88 b a Estimated error on lifetimes ±0.02 ms, except where stated otherwise.b Decay not single exponential: estimated error in lifetimes ±0.1 ms. Synthesis and structures of lanthanide complexes with L2 Reaction of L2 with a range of lanthanide salts in methanol, followed by treatment of the resultant solutions with aqueous KPF6, resulted in formation of the homoleptic complexes [M(L2)3][PF6]3 in good yields. These were characterised on the basis of their elemental analyses and FAB mass spectra (Table 1).The FAB spectra were interesting in that as well as the expected ML2 n fragments, co-ordinated fluoride ions (arising from the hexafluorophosphate) could be seen, with one or more of the acidic pyrazole groups losing protons as required to give a 11 charge. Thus, peaks corresponding to [M(L2)2F2]1, [ML2(L2 2 H)F]1 and [M(L2 2 H)2]1 occurred in every case. We have previously structurally characterised a europium(III) complex in which a co-ordinated fluoride was extracted from a hexafluorophosphate ion.7 Three of these complexes (M = Eu, Gd or Ho) were structurally characterised, and are essentially isostructural and isomorphous.Bond lengths and angles for all three are summarised in Table 7. The structure of one example (M = Eu) is shown in Fig. 7. The metal ion is nine-co-ordinate from three terdentate ligands, with a tricapped trigonal-prismatic geometry (Fig. 8). The two triangular faces of the trigonal prism are formed by the two sets of three pyrazolyl N-donor atoms, with the pyridyl donor atoms being the three caps.These planes are not exactly eclipsed but are slightly staggered; however they are almost exactly parallel, with angles between them of 0.68 in each case. All three complexes crystallise in the space group C2/c, with a C2 axis along the N(51)]M bond. The terdentate ligands are not exactly planar, but have slight twists between the adjacent aromatic rings.For M = Eu the dihedral angles between the mean planes of adjacent aromatic rings are as follows: between rings 1 and 2, 13.58; between rings 2 and 3, 11.68 in the same sense; between rings 4 and 5, 8.78 [where ring 1 means atoms N(11) to C(15), and so on]. The absence of any stereoelectronic requirement for octahedral or planar geometry means that the ligands coordinate in a strain-free manner with bite angles of ca. 61– 638 between adjacent rings, in contrast to the values of 70– 808 that would be expected with most d-block metal ions.These structures are very similar to those of the homoleptic nine-co-ordinate lanthanide complexes that have been charac- Table 7 Selected bond lengths (Å) and angles (8) for [M(L2)3][PF6]3 (M = Eu, Gd or Ho) M = Eu M = Gd M = Ho M]N(11) M]N(21) M]N(31) M]N(41) M]N(51) N(41)]M]N(41A) N(41)]M]N(31A) N(41)]M]N(31) N(31A)]M]N(31) N(41)]M]N(11) N(41A)]M]N(11) N(31A)]M]N(11) N(31)]M]N(11) N(11)]M]N(11A) N(41)]M]N(51) N(31)]M]N(51) N(11)]M]N(51) N(41)]M]N(21A) N(31)]M]N(21A) N(11)]M]N(21A) N(41)]M]N(21) N(31)]M]N(21) N(11)]M]N(21) N(51)]M]N(21) N(21A)]M]N(21) 2.548(3) 2.599(3) 2.534(3) 2.516(3) 2.577(4) 126.51(14) 149.53(10) 77.51(10) 88.35(14) 86.98(10) 78.54(9) 79.51(10) 125.64(10) 147.56(14) 63.26(7) 135.82(7) 73.78(7) 135.20(9) 73.91(9) 137.81(9) 75.62(9) 62.75(10) 62.96(9) 120.82(6) 118.36(13) 2.526(5) 2.593(5) 2.530(5) 2.504(5) 2.571(7) 127.6(2) 148.8(2) 77.4(2) 87.9(2) 86.8(2) 78.6(2) 79.9(2) 126.1(2) 146.6(2) 63.80(11) 136.07(11) 73.32(11) 135.0(2) 73.4(2) 138.2(2) 75.3(2) 62.9(2) 63.3(2) 120.95(11) 118.1(2) 2.497(2) 2.554(2) 2.482(2) 2.464(2) 2.535(3) 128.01(10) 147.88(7) 77.58(7) 87.77(11) 86.44(7) 78.78(7) 79.73(8) 126.88(7) 145.96(10) 64.00(5) 136.12(5) 72.98(5) 135.25(7) 72.92(7) 138.29(7) 74.96(7) 63.42(7) 63.54(7) 120.94(5) 118.12(10)2086 J.Chem. Soc., Dalton Trans., 1997, Pages 2079–2086 terised with architecturally similar ligands such as terpyridine and its relatives.8–11 Solution luminescence properties of [Tb(L2)3][PF6]3 The luminescence lifetimes t for [Tb(L2)3][PF6]3 in various solvents are summarised in Table 6.In CH2Cl2 (in which the compound is only sparingly soluble) the electronic spectrum showed the expected ligand-centred p æÆ p* transitions at lmax = 300 (maximum) and 260 nm (shoulder). Irradiation at either of these wavelengths produced the typical sequence of 5D4 æÆ 7Fn transitions in the emission spectrum, with the n = 6, 5, 4 and 3 components being visible, and a lifetime t of 1.46 ms.The difference between the lifetimes in MeOH and CD3OD (t = 1.09 and 1.18 ms respectively) gives q = 0.6 from the Horrocks equation. This is consistent with the nine-co-ordinate structure being retained in solution, and second-sphere coordination of methanol providing the limited amount of solvent-based quenching that occurs. We note that the nonco- ordinated pyrazolyl NH fragments at position 1 of the pyrazolyl rings provide sites for hydrogen bonding with solvents, which could contribute significantly to second-sphere solvation effects.In water and D2O however the behaviour is very different. The emission decays are not exactly single exponential [plots of ln(intensity) vs. time are slightly curved], indicating a mixture of at least two species with different lifetimes. Consequently the values of t derived from these data are rather approximate, but the substantial difference between them (0.45 and 0.88 ms in water and D2O respectively) gives q ª 4.6.As with [Tb(L1)2- (NO3)] (above) and [Eu(terpy)3]31,21 dissolution of [Tb(L2)3]- [PF6]3 in water results in partial dissociation of some of the chelating ligands to allow co-ordination of about four or five water molecules. The non-single-exponential decay observed suggests that two or more species, with different extents of hydration and different emission lifetimes, are present and that interconversion between them is slow on the timescale of the luminescence experiment.The dramatic differences between the behaviour of [Tb(L2)3][PF6]3 in water and methanol mirrors that of [Tb(L1)2(NO3)] (above). Clearly water is a much better donor to the metal centre in these complexes than is methanol, with water able to effect partial displacement of the N-donor chelating ligands from the co-ordination sphere of the metal, but methanol not able to do so.Acknowledgements We thank the EPSRC for financial support. References 1 E. Soini, I. Hemmilä and P. Dhalen, Ann. Biol. Chem., 1990, 48, 567; J.-C. G. Bünzli, in Lanthanide Probes in Life, Chemical and Earth Sciences, Theory and Practice, eds. J.-C. G. Bünzli and G. R. Choppin, Elsevier, Amsterdam, 1989, p. 219; A. K. Saha, K. Kross, E. D. Kloszewski, D. A. Upson, J. L. Toner, R. A. Snow, C. D. V. Black and V. C. Desai, J. Am. Chem. Soc., 1993, 115, 11 032; V.M. Mukkala, M. Heleniu, I. Hemmilä, J. Kankare and H. Takalo, Helv. Chim. Acta, 1993, 76, 1361; M. J. P. Leiner, Anal. Chim. Acta, 1991, 255, 209; A. P. de Silva, H. Q. N. Gunaratne and T. E. Rice, Angew. Chem., Int. Ed. Engl., 1996, 35, 2116; M. A. Mortellaro and D. G. Nocera, J. Am. Chem. Soc., 1996, 118, 7414. 2 S. H. Koenig and R. D. Brown, Prog. Nucl. Magn. Reson. Spectrosc., 1990, 22, 487; R. B. Lauffer, Chem. Rev., 1987, 87, 901; D. Parker, Chem. Br., 1994, 818; A.D. Watson, J. Alloys Compd., 1994, 207/208, 14; S. Aime, M. Botta, S. G. Crich, G. B. Giovenzana, G. Jommi, R. Pagliarin and M. Sisti, J. Chem. Soc., Chem. Commun., 1995, 1885; M. Inoue, R. E. Navarro, M. Inoue and Q. Fernando, Inorg. Chem., 1995, 34, 6074; D. E. Reichert, R. D. Hancock and M. J. Welch, Inorg. Chem., 1996, 35, 7013; E. Tóth, S. Vauthey, D. Pubanz and A. E. Merbach, Inorg. Chem., 1996, 35, 3375; D. H. Powell, O. M. Ni Dhubhghaill, D. Pubanz, L. Helm, Y. S. Lebedev, W.Schlaepfer and A. E. Merbach, J. Am. Chem. Soc., 1996, 118, 9333; G. R. Choppin and K. M. Schaab, Inorg. Chim. Acta, 1996, 252, 299. 3 C. Piguet, J.-C. G. Bünzli, G. Bernardinelli, G. Hopfgartner and A. F. Williams, J. Am. Chem. Soc., 1993, 115, 8197; C. Piguet, J.-C. G. Bünzli, G. Bernardinelli and A. F. Williams, Inorg. Chem., 1993, 32, 1237; C. Piguet, J.-C. G. Bünzli, G. Bernardinelli, C. G. Bochet and P. Froidevaux, J. Chem. Soc., Dalton Trans., 1995, 83; C. Piguet, G.Hopfgartner, A. F. Williams and J.-C. G. Bünzli, J. Chem. Soc., Chem. Commun., 1995, 491; C. Piguet, E. Rivara- Minten, G. Bernardinelli, J.-C. G. Bünzli and G. Hopfgartner, J. Chem. Soc., Dalton Trans., 1997, 421. 4 J.-M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1995, pp. 92–95 and refs. therein; P. R. Selvin, J. Jancarik, M. Li and L.-W. Hung, Inorg. Chem., 1996, 35, 700; A. Døssing, H. Toftlund, A. Hazell, J. Bourassa and P. C. Ford, J. Chem. Soc., Dalton Trans., 1997, 335. 5 D. Parker and J. A. G. Williams, J. Chem. Soc., Dalton Trans., 1996, 3613. 6 W. D. Horrocks and D. R. Sudnick, Acc. Chem. Res., 1981, 14, 384; N. Sabbatini, M. Guardigli and J.-M. Lehn, Coord. Chem. Rev., 1993, 123, 201. 7 P. L. Jones, A. J. Amoroso, J. C. Jeffery, J. A. McCleverty, E. Psillakis, L. H. Rees and M. D. Ward, Inorg. Chem., 1997, 36, 10. 8 G. H. Frost, F. A. Hart and N. B. Hursthouse, Chem. Commun., 1969, 1421; D. A. Durham, G. H. Frost and F. A. Hart, J. Inorg. Nucl. Chem., 1969, 31, 833. 9 C. Piguet, J.-C. G. Bünzli, G. Bernardinelli and A. F. Williams, Inorg. Chem., 1993, 32, 4139. 10 C. Mallet, R. P. Thummel and C. Hery, Inorg. Chim. Acta, 1993, 210, 223. 11 C. Yang, X.-M. Chen, W.-H. Zhang, J. Chen, Y.-S. Yang and M.-L. Gong, J. Chem. Soc., Dalton Trans., 1996, 1767. 12 Y. Lin and S. A. Lang, J. Heterocycl. Chem., 1977, 14, 345. 13 T. Buchen, P. Gütlich, K. H. Sugiyarto and H. A. Goodwin, Chem. Eur. J., 1996, 2, 1134. 14 SHELXTL 5.03 program system, Siemens Analytical X-Ray Instruments, Madison, WI, 1995; Software package for use with the SMART diffractometer; Siemens Analytical X-Ray Instruments, Madison, WI, 1995. 15 S. Trofimenko, J. C. Calabrese and J. S. Thompson, Inorg. Chem., 1992, 31, 974 and refs. therein. 16 E. Psillakis, J. C. Jeffery, J. A. McCleverty and M. D. Ward, Chem. Commun., 1997, 479. 17 E. Psillakis, J. C. Jeffery, J. A. McCleverty and M. D. Ward, J. Chem. Soc., Dalton Trans., in the press. 18 A. Novak, Struct. Bonding (Berlin), 1974, 18, 177. 19 D. A. Bardwell, J. C. Jeffery, P. L. Jones, J. A. McCleverty and M. D. Ward, J. Chem. Soc., Dalton Trans., 1995, 2921. 20 S. Aime, M. Botta, D. Parker and J. A. G. Williams, J. Chem. Soc., Dalton Trans., 1996, 17. 21 R. D. Chapman, R. T. Loda, R. W. Riehl and R. W. Schwartz, Inorg. Chem., 1984, 23, 1652. 22 J.-C. G. Bünzli, J.-R. Yersin and C. Mabillaud, Inorg. Chem., 1982, 21, 1471; J.-C. G. Bünzli and A. Nilicic-Tang, Inorg. Chim. Acta, 1996, 252, 221. Received 24th February 1997; Paper 7/01297G

ISSN:1477-9226    

DOI:10.1039/a701297g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 2083 Electrochemical and binding properties of a novel ferrocenecontaining redox-active basket-shaped host molecule Georg C. Dol,a Paul C. J. Kamer,a Frantisek Hartl,a Piet W. N. M. van Leeuwen *,†,a and Roeland J. M. Nolte b a J. H. van ’t Hoff Research Institute, University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands b Department of Organic Chemistry, NSR Centre, University of Nijmegen, Toernooiveld, 6525ED Nijmegen, The Netherlands A novel basket-shaped host based on the rigid molecule diphenylglycoluril {3a,6a-diphenyltetrahydroimidazo- [4,5-d ]imidazole-2,5(1H,3H)-dione} has been synthesized and characterised.It is able to bind cations in its crown ether rings and neutral organic substrates in its cavity. DiVerential pulse voltammetry experiments showed that the host is redox-responsive to cations. It forms 1 : 1 complexes with K1 and NH4 1 ions and 1 : 2 host–guest complexes with Na1 ions.On addition of (di)ammonium salts, protonation of the host occurs. A complex was formed between the host and model substrate olivetol (5-pentylbenzene-1,3-diol). In the absence of additives, this complex is stabilised via hydrogen bonds and p–p stacking interactions. In the presence of Na1 ions a complex consisting of the host, the diol, and 2 Na1 ions was formed, in which hydrogen bonds are no longer present. In the presence of 2 Na1 ions a four-fold increase in association constant has been found.Spectroscopic (NMR and IR) experiments have been used to elucidate the mode of co-operative co-ordination between the host, diol substrate and Na1 ions. Supramolecular chemistry has undergone a tremendous growth in the last three decades. After the early work of Pedersen,1 Cram2–4 and Lehn5,6 and co-workers macrocyclic crown ether rings have been applied as building blocks to construct large molecular systems with specific properties.7 Supramolecular chemistry has led to new applications, one of which is the development of ion-selective sensors.8 Combinations of redox reactive centres and crown ethers have been used to design these chemical sensors.The first examples of redox-responsive molecules were (aza)crown ethers linked to a ferrocene or nitrobenzene unit.9,10 These molecules showed diVerent redox behaviour in the presence of positively charged molecules. Recently more elegant responsive compounds have been developed, containing a ferrocene or tetrathiafulvalene unit as the redox-active centre.11–13 In general, these compounds exhibit a low response towards ammonium cations; only the triaza-crown-6 based sensors developed by Beer et al.14 are very responsive to NH4 1 ions.Sensors responsive to alkylsubstituted ammonium cations, however, have yet to be reported. This paper describes the synthesis and properties of a new redox-active metallohost based on diphenylglycoluril {3a,6adiphenyltetrahydroimidazo[ 4,5-d ]imidazole-2,5-(1H,3H)- dione}, to which a ferrocene unit has been connected.Host molecules derived from diphenylglycoluril are capable of binding organic substrates like dihydroxybenzenes and hard cations like Na1 or K1. Other molecules that form complexes with these host molecules are substituted ammonium salts. Hydrogen bonds between the urea carbonyl groups of the host molecule and p–p stacking interactions between the aromatic walls of the host and the aromatic ring of the guest molecule contribute to the process of binding.The alkali metal ions are bound to the crown ether parts of the host molecule, whereas neutral molecules are clammed between the aromatic walls. In this paper we present electrochemical studies and a study towards the binding of diVerent guest molecules in the host. Cyclic voltammetry, diVerential pulse voltammetry, IR and † E-Mail: pwnm@anorg.chem.uva.nl NMR measurements were used to elucidate the nature of the host–guest binding processes taking place.A novel Na1- promoted host–guest complex is presented as well. Results and Discussion Synthesis The most frequently used method to produce tertiary aminomethylferrocene derivatives is the reaction of a primary amine with either chlorocarbonylferrocene,11,13–15 ferrocenylmethylpyridinium toluene-p-sulfonate 16,17 or trimethylammoniomethylferrocene iodide 1a (Scheme 1).14,18 This method could not be applied to the synthesis of our target molecule because the preparation of the unprotected amine (4b) from the benzyl analogue (4a)19 turned out to be rather troublesome and diYcult to reproduce.Therefore a diVerent reaction procedure for the synthesis of substituted aminomethylferrocenes was Scheme 1 Synthesis of compound 2: (a) NH3–MeOH, 80 8C, 24 h; (b) (i) potassium phthalimide, dmf, 90 8C; (ii) N2H4–EtOH; (iii) 10% HCl, 100 8C 1 h Fe NMe3I Fe NH2 Fe N O O ( b)( ii,iii) 2 1a ( a) 1b ( b)( i)2084 J.Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 developed. First aminomethylferrocene 2 was synthesized, which was then made to react with 3 to aVord host 4 (Scheme 2). Trimethylammoniomethylferrocene iodide 20 1a was used as the starting compound in the production of aminomethylferrocene. Two routes towards the desired product 2 were investigated: (a) direct conversion of 1a into the amine by reaction with ammonia and (b) a classical Gabriel 21 synthesis using potassium phthalimide and subsequently hydrazine (see Scheme 1).To avoid the use of hazardous azidomethylferrocene, this method was not used.22 Route (a). Several conditions were investigated to convert trimethylammoniomethylferrocene iodide 1a into aminomethylferrocene 2 in one step. Only the reaction of 1a with NH3 in MeOH at temperatures between 70 and 80 8C aVorded 2 in acceptable yields (60% at 80 8C). This reaction was performed in an autoclave to prevent evaporation of NH3.Reactions of 1a performed at lower temperatures gave low conversions and in liquid NH3 no reaction occurred. Lower yields were also obtained when the reaction was conducted at temperatures higher than 80 8C because undefined side products were formed. At the optimum temperature this route proved to be a very smooth one-step synthesis towards aminomethylferrocene and is preferred to the Gabriel synthesis or the method employing hazardous azides.Route (b). Trimethylammoniomethylferrocene iodide was made to react under standard conditions with potassium phthalimide yielding 66% of 1-(phthalimidomethyl)ferrocene 1b (lit.,21 99%). After isolation and reaction with hydrazine, aminomethylferrocene was obtained in 63% yield. The overall yield for the conversion of 1 into 2 was 42%. Routes (a) and (b) result in the formation of the desired product 2 according to 1H, 13C NMR and GC-MS measurements. Compound 2 is relatively stable, but it is recommended to use it in further Scheme 2 Synthesis of compound 4: (i) NaI–MeCN, 24 h reflux then 2, 3 d reflux N N N N O O O O O O O O O O N N Ph Ph 4 Fe Fe ( i) 3 N N N N O O O O O O Ph Ph O O O O Cl Cl Cl Cl reactions shortly after isolation.When stored over longer periods, decomposition occurs and the product liquefies with formation of volatile amines and cyclopentadiene. The final reaction in the synthesis of target molecule 4 consisted of a ring closure under high dilution conditions.This procedure was developed and optimised previously by Nolte and co-workers.23,24 Compound 3 was converted in situ into its iodide analogue via a Finkelstein 25 reaction. When all the chlorine atoms had been replaced, amine 2 was slowly added allowing the ring closure reaction to take place. After standard work-up and column chromatography to remove side products, compound 4 was isolated in 45% yield. The product could be purified by column chromatography using silica gel and TLC plates impregnated with NaBr.26 Electrochemical experiments The electrochemical properties of compound 4 were studied using cyclic voltammetry (CV) and diVerential pulse voltammetry (DPV).Solvent mixtures of acetonitrile and dichloromethane were used in combination with NBun 4PF6 as the electrolyte. Ferrocene could not be used as the internal reference in our experiments because of interference with the oxidation of compound 4.The salt [Co(h5-C5H5)2]PF6 was therefore used as the internal reference.16 In separate experiments, the reference potential E2� 1 {[Co(h5-C5H5)2]1/0} was determined to be 21.33 V vs. ferrocene–ferrocenium. Owing to the high molecular weight and thus small diVusion coeYcient of compound 4, only current responses of moderate intensity could be observed in the CV experiments. After additions of cations the anodic peaks became broad and new signals arose leading to a drop in current intensity, which made the exact determination of E2� 1 values from CV measurements diYcult.Cyclic voltammetry was therefore only used to check the chemical reversibility of the processes and DPV was chosen to determine the exact values of E2� 1 . The cationic guest molecules used in these experiments are only soluble in MeCN whereas compound 4 is hardly soluble in this solvent, therefore CH2Cl2 was added to obtain a homogeneous solution. In all experiments oxidation of a species adsorbed at the surface of the electrode was observed at 275 mV (vs.ferrocene–ferrocenium). At high cation concentrations this species disappeared, probably due to strong complexation to 4. The DPV responses were recorded after progressively adding aliquots of stock solutions in MeCN containing substoichiometric equivalents of Na1, K1, NH4 1, 4,49-bipyridinium bis(tetrafluoroborate) 5 and 1,4-phenylenediammonium bis- (trifluoromethanesulfonate) 6. In the absence of additives a single anodic peak was observed for 4.This implies that both ferrocene moieties are oxidised in one step. As compared to ferrocene itself, the oxidation of 4 is shifted to a slightly more negative value (E2� 1 = 30 mV vs. ferrocene–ferrocenium), which is in agreement with previous observations on multiple-ferrocenecontaining crown ethers.14 Significant anodic perturbations could be observed after addition of any of the cationic species mentioned above; the data obtained are summarised in Table 1. Ammonium cations possess the most pronounced influence N NH H H3 N NH3 C5 H11 HO OH 5 6 7 2 BF4 – 2 CF3SO3 –J.Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 2085 on the oxidation potential of compound 4, followed by sodium and potassium cations. The addition of ammonium ions resulted in separate DPV responses for 4 and 4?NH4 1, respectively. After the addition of 1 equivalent of NH4 1 ions the response of the free host had disappeared.The addition of sodium and potassium ions resulted in a continuously shifting DPV response to the final E2� 1 value of the host–guest complexes, as can be seen from Figs. 1 and 2, respectively. The diVerence between Na1 (DE2� 1 = 88 mV) and K1 (DE2� 1 = 64 mV) might be caused by the higher polarisability 13 of the latter cation, which causes a smaller anodic shift of the oxidation potential. Another explanation might be that K1 forms a 1 : 1 complex with 4 whereas Na1 aVords a 1 : 2 host–guest complex, i.e.a species with a higher positive charge per ferrocenyl.19,27,28 The larger positive shift caused by addition of NH4 1 ions (DE2� 1 = 106 mV) is probably caused by a stronger interaction between the nitrogen atom of the aminomethylferrocene residue and NH4 1 ions compared to Na1 and K1. The experiments were also performed using an excess of 5 equivalents of cation, resulting in unaVected oxidation potentials for the Na1- and K1-containing complexes.This confirms a very strong preference for the formation of a 1 : 1 complex of 4 with K1 ions and the formation of a 1 : 2 host–guest complex with Na1 ions. Addition of 4 equivalents or more of NH4 1 ions, however, resulted in an additional anodic peak at a more positive oxidation potential (DE2� 1 = 220 mV). This is probably caused by protonation of part of the aminomethylferrocene nitrogen atoms (see below). The stoichiometries found for Na1, K1 and NH4 1 are in good agreement with values obtained from Fig. 1 The DPV responses of a titration of compound 4 with Na1 ions, producing 4?2Na1; platinum disc (1 mm2 apparent surface area) in CH2Cl2–MeCN (3: 1, v/v), [4] = 1023 M, [NBun 4PF6] = 2 × 1021 M, n = 10 mV s21, 293 K. Complex 4 plus a 0, b 0.5, c 1.0, d 1.5, e 2.0 and f 2.5 equivalents of Na1 ions Table 1 Cation-dependent oxidation potentials of compound 4 a DE2� 1 /mV E2� 1 b/mV 230 c,d 2 Na1 88 c 1 K1 64 c 1 NH4 1 106,c 220 e 1 (521) 232 d 0.3 (621) 240 d a Obtained from DPV measurements using a 1023 M solution of compound 4; Na1, K1 and NH4 1 were added as their trifluoromethanesulfonate salts.b vs. Ferrocene–ferrocenium. c Measured in CH2Cl2–MeCN (1: 1, v/v). d Measured in CH2Cl2–MeCN (3: 1, v/v). e After addition of 3.5 equivalents of NH4 1. picrate extraction experiments by Smeets et al.19,27,28 These experiments were performed using 4a, which has the same properties as those of 4 with regard to binding guest molecules.Addition of ammonium salt 5 or 6 to host 4 resulted in the formation of an additional DPV response at a more anodic potential of E2� 1 ª 200 mV. Titration of 4 with the respective ammonium salts resulted in a gradual decrease of the parent signal whereas the signal of the host–guest complexes increased. The stoichiometry at which the original signal had disappeared diVered for these two molecules. Beforehand it was expected that these dications (5 and 6) would give rise to a 1 : 1 adduct stoichiometry with 4, and this indeed was found after titration of 4 with 1 equivalent of 5.When 6 was used, however, a host–guest stoichiometry of 1 : 0.25 was observed. This diVerence in stoichiometry is not fully understood and is currently under investigation. The nearly identical oxidation potentials after addition of these two substrates indicate that the same product is formed after the addition of either 5 or 6. This could be confirmed by NMR and IR experiments as is described below.The electrochemical experiments, as well as the NMR and IR data, suggest that after addition of either NH4 1 ions, 5 or 6 protonation of the aminomethylferrocene nitrogen takes place. Plenio et al.16 reported an anodic shift of DE2� 1 = 250 mV after protonation of one of the nitrogen atoms in a ferrocenecontaining cryptand [based on 1,19-bis(methylene)ferrocene and 1,4,10,13-tetraoxa-7,16-diazacyclooctadecane, which is in good agreement with our findings.A competition experiment between Na1 ions and compound 5 was also performed (see Fig. 3). As can clearly be seen, after Fig. 2 The DPV responses of a titration of compound 4 with K1 ions, producing 4?K1. Conditions as in Fig. 1. Complex 4 plus a 0, b 0.5, c 1.0, d 1.5 and e 2.5 equivalents of K1 ions 4a, R = CH2Ph 4b, R = H N N N N O O O O O O O O O O N N Ph Ph R R2086 J. Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 addition of 2 equivalents of Na1 ions the oxidation potential of 4 shifts to a more positive value of E2� 1 = 54 mV.After addition of 0.25 equivalent of 5 to this host–guest complex 4?2Na1 a new peak at E2� 1 = 200 mV appears. This new signal grows on addition of more equivalents of 5 whereas a decrease in intensity of the original signal of 4?2Na1 is observed. After the addition of 1.25 equivalents of 5 the original signal has completely vanished. The E2� 1 value of the final product is in agreement with that found for the complex 4?5, an indication that the Na1 ions are not bound to 4 after protonation of the nitrogen atoms. When the order of addition was reversed, i.e.first substoichiometric amounts of 5 were added to 4 followed by the addition of Na1 ions, the DPV response of the host–guest complex 4?5 was observed only. Recently, Carr et al.29 reported a sensor for neutral molecules. This ferrocene-containing molecule was responsive to the addition of carboxylic acids.We also investigated the influence of neutral guest molecules on the oxidation potential of host 4. Substituted resorcinol derivatives were added to solutions containing 4, but no eVect could be detected. Resorcinol derivatives were also added to the preformed host–guest complex 4?2Na1, but the EThese experiments indicate that resorcinol derivatives do not influence the oxidation potential of host 4. Complexation of guest molecules in solution Compounds 4a and 4b are able to bind hydroxy-substituted aromatic guest molecules and cations.27,28 A particularly suitable guest molecule is olivetol (5-pentylbenzene-1,3-diol) 7, because it is well soluble in organic solvents like CH2Cl2 and CHCl3. It can easily be used in the determination of association constants (K) via NMR titration, which is the most often applied technique nowadays to determine such constants of host–guest complexes.The NMR titrations were carried out by monitoring the signals of both the host and the guest molecule.The method developed by Granot 30 was applied, and after curve fitting of the NMR data the association constants were obtained. Compound 4 was found to bind olivetol with an association constant of K = 1724 M21 which is comparable with literature values of similar compounds.31 The association con- Fig. 3 The DPV responses of 1023 M solutions of compound 4 and its complex with 2 equivalents of Na1 and the response to subsequent additions of various amounts of 5, producing 4?5.Platinum disc (1 mm2 apparent surface area) in CH2Cl2–MeCN (3: 1, v/v), [NBun 4PF6] = 2 × 1021 M, n = 10 mV s21, 293 K. Complex 4 plus a 0 and b 2 equivalents of Na1 ions; 4?2Na1 plus c 0.25, d 0.5, e 0.75, f 1.0 and g 1.25 equivalents of 5 stant was also determined in the presence of 10 equivalents of Na1 ions (added as 4-dodecylbenzenesulfonate). As shown above, host 4 forms a complex in which 2 Na1 ions are bound in its crown ether moieties.Interestingly, this experiment aVorded an association constant which was approximately 4 times higher (K = 7897 M21) than that obtained from the experiment without Na1 ions present. This is in contrast to the expectation that binding of organic molecules would be blocked when cations are bound in the host, as suggested by Coolen et al.32 A comparison of the chemically induced shift (CIS) values obtained from the NMR titration experiments revealed that those (20.36 ppm) of the protons of the aromatic sidewalls of compound 4 in the presence of Na1 are much lower than those without Na1 (20.51 ppm) present, suggesting that the olivetol molecule is bound less deeply33 in the cavity of 4.Although the NMR data clearly showed that a diVerent association process takes place between 4 and 7 in the presence of Na1 ions, they did not reveal what type of interactions are involved between host and guest. Therefore IR experiments were performed in order to monitor the stretching vibration of the urea carbonyl groups of the host molecule. Several combinations of host, guest and Na1 ions were studied and the values obtained are summarised in Table 2.When host 4 was measured without additives n(C]] O) was located at 1708 cm21. When 8 equivalents of olivetol were added to 4 a shift of n(C]] O) to 1685 cm21 was observed. This shift of 223 cm21 points to the formation of hydrogen bonds 33 between the carbonyl groups of 4 and the hydroxyl groups of the olivetol molecule.On the other hand a solution containing 4 and 10 equivalents of Na1 aVorded an IR spectrum in which n(C]] O) was similar to that of 4 without additives. The same unchanged carbonyl stretching frequency was found when the IR spectrum of a solution containing 4, 8 equivalents of olivetol 7 and 10 equivalents of Na1 was recorded. These results indicate that the hydroxyl groups of olivetol are not involved in hydrogen bonding with the carbonyl groups of the host when Na1 ions are present, thus demonstrating a diVerent type of binding between 4 and this guest in the presence of Na1 ions.Based on NMR and IR data we propose that the Na1 ions partly block the entrance to the lower part of the cavity of compound 4, making hydrogen bonding with the carbonyl groups impossible. However, the two-electron-poor alkali-metal cations are readily available for an interaction with the oxygen atoms of the incoming olivetol molecule, allowing them to be six-co-ordinated by these donor atoms, viz.four oxygen atoms and one nitrogen atom of 4 and one oxygen atom of olivetol. A computer-generated drawing (Cache molecular modelling program34,35 using extended MM2 force field parameters) of the proposed binding modes is given in Fig. 4. Table 2 Association constants and IR shift data of host–guest complexes of compound 4 Guest 77 1 10 equivalents Na1c 7 1 5 equivalents 8 d 5 equivalents 8 5-Cyanoresorcinol Ka/M21 1724 (35) 7897 (1650) —— n.d.e CIS/ppm 20.51 20.36 —— n.d.e Dn(C]] O)b/ cm21 223 22 227 226 n.d.f a Obtained from NMR titrations using (0.5–0.75) × 1023 M solutions of host and a (5–7.5) × 1023 M solution of guest in CDCl3.Association constants were calculated using the Granot 30 procedure; errors are given in parentheses. b Measured in CH2Cl2 solution; cell optical path 0.5 mm; c = 1.5 × 1023 M, host : guest mole ratio 1 : 8. c Sodium dodecylbenzenesulfonate was used as the co-guest. d The NMR signals of compounds 4 and 7 remained unaVected on the addition of benzyldimethylammonium tetrafluoroborate 8.e Not determined; K is estimated to be 1 × 105 M21 based on a literature value of a similar compound.31 f Not determined.J. Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 2087 Fig. 4 Drawings of the host–guest complexes of compound 4: A host 4 with resorcinol in the absence of Na1; B host 4 with resorcinol in the presence of Na1 Studies to investigate the eVect of K1 ions on the binding aYnity between the host and olivetol were unsuccessful, probably because of the low solubility of the potassium salts tried.The salt having the highest solubility was potassium dodecylsulfate but even this resulted in turbid solutions unsuitable for NMR studies. The NMR and IR experiments were also carried out to study the mode of binding of ammonium salts 5 and 6 in host 4. Proton NMR experiments on 4, performed in a mixture of CDCl3 and CD3CN, showed changes on the addition of substoichiometric amounts of these ammonium salts.The resonances of the CH2N groups at d 2.66 (in this solvent mixture) broadened and shifted to d 3.23 on addition of either 5 or 6. Additionally, a new relatively broad signal appeared at d 4.4. Furthermore, shifting and some broadening of the large multiplet between d 3.8 and 3.4 (CH2O residues) occurred. Changes in the NMR signals of the wall protons and the guest protons were expected, due to mutual anisotropic ring eVects, but significant changes could not be detected.The new signal at d 4.4, the shift of the CH2N protons and the lack of further changes in the NMR spectrum suggest that proton transfer takes place between 4 and the ammonium salts instead of the formation of a host–guest complex. The NMR experiments using organic acids and ammonium salts were performed to investigate the changes that take place in the NMR spectrum of 4.Trifluoroacetic acid, 2,4,6-trichlorobenzoic acid and benzyldimethylammonium tetrafluoroborate 8 were used as the proton donors. When any of these three substrates was added to 4 the changes in the NMR spectra were identical to those observed for the ammonium salts 5 and 6, thus implying that protonation of the nitrogen atoms of 4 takes place. To investigate if resorcinol derivatives are still bound in a protonated host, these NMR experiments were also performed in the presence of 6 equivalents of 5-cyanoresorcinol. According to our findings this is the only resorcinol derivative that still binds to 4 in solvent mixtures with CD3CN.The cyano group increases the strength of the hydrogen bonds with the carbonyl groups as well as the p-stacking interactions between host and guest, resulting in an overall high binding aYnity. In the presence of solvents capable of formation of hydrogen bonds like CD3CN the association constant will drop to lower values.31 When 5-cyanoresorcinol is used, however, binding still occurs.The NMR signals of the aromatic walls of the host as well as signals of the guest can be monitored to determine whether a host–guest complex is formed in the presence of a proton source. All these experiments showed that after protonation of the nitrogen atom the diol is not bound in the host molecule. An additional IR experiment revealed a change in the n(C]] O) of 4 in the presence of 4 equivalents of benzyldimethylammonium tetrafluoroborate 8 (see Table 2), indicative of intramolecular hydrogen bonds between the carbonyl groups and the protonated nitrogen atoms of the host.Hence, it is concluded that these intramolecular hydrogen bonds prevent the formation of a host–guest complex with resorcinol derivatives. Attempts were made to investigate the association process with the two ammonium salts 5 and 6. In order to obtain a homogeneous solution these experiments had to be performed in a solvent mixture of CH2Cl2 and MeCN.In this solvent mixture an intermolecular hydrogen bond between the proton donor and MeCN is favoured over the intramolecular hydrogen bond, therefore the carbonyl stretching frequency remains unaVected. Conclusion A novel redox-responsive host molecule based on diphenylglycoluril has been synthesized. This host molecule is able to bind both neutral organic guest molecules and positively charged ions.Cations like Na1, K1 and NH4 1 have pronounced eVects on the oxidation potential of compound 4. Protonation of the aminomethylferrocenyl group of 4 also results in a shift of the oxidation potential of 4. Binding studies in combination with electrochemical experiments show that 4 binds Na1 in an 1 : 2 host–guest ratio, whereas K1 and NH4 1 are bound in a 1 : 1 ratio. Sodium ions have a pronounced eVect on the association constant of the complexes between 4 and olivetol. An enhanced binding has been shown for this novel four-component (4, 7 and 2Na1) host–guest complex.A new co-ordination mode assisted by these Na1 ions is presented, which can explain the relatively large association constant found for this multicomponent host–guest complex. Host 4 is responsive to the addition of primary ammonium salts, but instead of binding the ammonium salts protonation of the CH2N groups of the host occurs.2088 J. Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 Experimental General Proton (300.13 MHz) and 13C-{1H} NMR (75.48 MHz) spectra were measured on Bruker AMX 300 and DRX 300 machines, with SiMe4 as the external reference, IR spectra on a Nicolet 510m FT-IR spectrophotometer.Melting points were determined on a Gallenkamp MFB-595 apparatus; the values are uncorrected. The GC-MS measurements were done on a Hewlett-Packard gas chromatograph, equipped with a DB- 5MS column (length 12 m, inner diameter 0.2 mm, film thickness 0.33 mm).Column chromatography was performed with silica gel 60, 70–230 mesh ASTM (Merck). Analytical TLC was performed on TLC aluminium foil, silica gel 60 F254 (Merck). Silica gel and TLC plates were impregnated with NaBr according to a literature procedure.26 Microanalyses were carried out in our own laboratory on an Elementar Vario EL apparatus (Foss Electric). Cyclic voltammetry and diVerential pulse voltammetry were performed using a EG&G PAR model 283 potentiostat.The electrochemical samples were 1023 M in redoxactive host and 2 × 1021 M in NBun 4PF6 as supporting electrolyte. The working electrode (1 mm disc) and the counter electrode (gauze) were made of Pt. The reference system consisted of a silver wire as the pseudo-reference electrode and [Co(h5-C5H5)2]PF6 was used as standard internal reference.16,36 The reported electrode potentials are given relative to the ferrocene–ferrocenium redox couple. Chemicals Acetonitrile and CH2Cl2 were distilled from CaH2; CDCl3 used in the association constant determinations was distilled from P2O5 before use.The salt NBun 4PF6 was recrystallised twice from absolute ethanol and dried overnight at 80 8C in vacuo before use. Trimethylammoniomethylferrocene iodide 20 and compound 337 were prepared according to literature procedures. Ammonium salts 5 and 6 were obtained after reaction of their amines with HCl and exchange with the respective NaX or AgX salts in MeOH or MeCN. NMR experiments Titration experiments were performed by monitoring the singlet at d 6.73 in the 1H NMR spectrum; on addition of olivetol the chemical shift of the protons shift to a lower value.From curve fitting assuming a 1 : 1 complex, K and CIS values were calculated according to literature procedures.30,31,33 Protonation experiments using various proton sources were performed as follows: compound 4 (5 mg, 4 mmol), 5-cyanoresorcinol (0.5– 0.6 mg, 40 mmol) and stoichiometric amounts of the various proton sources were measured separately and in combination.The solvent was CDCl3–CD3CN (1: 1, v/v). IR experiments A number of 4 × 1023 M solutions of host 4 in CH2Cl2 were prepared, followed by separate additions of 8 equivalents of 7, 10 equivalents of dodecylbenzenesulfonic acid sodium salt and 5 equivalents of 8, and the IR spectra were recorded. Combinations of the substrates 7, 8, and Na1 were also added followed by recording of an IR spectrum.A mixture of CH2Cl2 and MeCN (1: 1, v/v) had to be used in the experiments with 5 or 6. These solutions had to be kept in an ultrasonic bath for prolonged periods in order to dissolve all components. Electrochemical experiments A three-electrode vacuum-tight voltammetric cell was loaded with host 4 (5 mg, 4 mmol), a small amount of [Co(h5-C5H5)2]- PF6 and electrolyte (NBun 4PF6). The cell was evacuated and flushed with dry nitrogen twice, followed by addition of CH2Cl2 (2 cm3) and MeCN (2 cm3).Voltammetric (CV, DPV) responses were recorded and subsequently substoichiometric aliquots of guest molecules in MeCN (typically 100 ml) were added. In between all measurements the working electrode was polished using a 0.25 mm diamond paste. Syntheses Aminomethylferrocene 2. Route (a). A stainless-steel autoclave was loaded with methanol (30 cm3), saturated with NH3 and compound 1a (0.5 g, 1.4 mmol) was added. The autoclave was closed and heated at 80 8C overnight.After cooling the autoclave was opened and the solution evaporated to dryness. The remaining solid was redissolved in CH2Cl2 (30 cm3) and water (30 cm3) was added. The organic layer was washed twice with 1 M NaOH (10 cm3) and water (10 cm3) and dried over MgSO4. Finally the solvent was evaporated to aVord 2 as an orange-red sticky solid, that turned into an oil on contact with air (0.14 g, 60% yield). NMR (CDCl3): dH 4.14 (9 H, m, ferrocene H, NCH2) and 3.5 (2 H, br, NH2); dc 68.4, 68.1, 67.4, 66.9 and 37.EI mass spectrum: m/z 215 (M1). Route (b) (i) 1-(Phthalimidomethyl)ferrocene 1b. This compound was prepared according to a literature procedure 21 from 1a (0.5 g, 1.4 mmol) and potassium phthalimide (0.26 g, 1.4 mmol). An orange powder was obtained (0.3 g, 66% yield), m.p. 203 8C (decomp.) [lit.,21 201–202 8C (decomp.)]. IR (CDCl3): n& max/cm21 1716s (C]] O), 1432m, 1393m and 1331m. NMR (CDCl3): dH 7.80 (2 H, dd, 3J = 5.4, 4J = 3.2), 7.67 (2 H, dd, 3J = 5.4, 4J = 3.2 Hz), 4.61 (2 H, s, CH2), 4.37 (2 H, s), 4.20 (5 H, s) and 4.10 (2 H, m); dC 167.7 (C]] O), 133.7, 131.9, 123.0, 117.6, 69.3, 68.4, 68.1, 66.0 and 37.0.(ii) Aminomethylferrocene 2. This compound was prepared according to a literature procedure 21 from 1b (0.3 g, 0.87 mmol) and hydrazine (50 ml). Compound 2 was obtained as an orange solid (95 mg, 63% yield). For identification see above. Compound 4. A mixture of compound 3 (1.3 g, 1.31 mmol), NaI (1.76 g, 11.7 mmol) and Na2CO3 (2.1 g, 19.6 mmol) in MeCN (500 cm3) was heated at reflux for 24 h.A solution of 2 (0.5 g, 2.5 mmol) in MeCN (300 cm3) was then added to this refluxing mixture at a rate of 0.25 cm3 min21. After 3 d the reaction mixture was cooled to room temperature, filtered and concentrated. After addition of CH2Cl2 (100 cm3) and water (100 cm3), the two layers were separated and the water layer was washed with CH2Cl2 (50 cm3). The collected organic layers were washed three times with demineralised water (100 cm3) and dried over MgSO4.The solvent was evaporated and the product purified by column chromatography over silica gel impregnated with NaBr (gradient eluent 1% NEt3–3–6% MeOH–CH2Cl2) to aVord an orange-brown powder (0.7 g, 45%), m.p. 183 8C (decomp.). IR (KBr): n& max/cm21 3447w, 2923–2870w, 1713s (C]] O), 1459s, 1427m, 1258m, 1128m and 1105m. NMR (CDCl3): dH 7.0 (10 H, m, aryl), 6.73 (4 H, s, aryl sidewall), 5.63 [4 H, d, 2J = 16, (CO)NCHH], 4.23–3.67 [50 H, m, OCH2, ferrocenyl NCH2, (CO)NCHH, ferrocenyl H] and 2.81 (8 H, br t, 3J = 5.0 Hz, CH2N); dC 157.6, 151.2, 135.8, 134.4, 129.1, 128.7, 107.8, 85.3, 70.6, 70.4, 69.7, 68.8, 68.3, 55.2, 53.3 and 37.2.High-resolution mass spectrum (FAB): Found 1273.4386, C70H77Fe2N6O10 requires 1273.4405 (M 1 H) (Found: C, 64.6; H, 5.9; N, 6.5. C70H76Fe2N6O10?1.5H2O requires C, 64.7; H, 6.1; N, 6.5%). Acknowledgements This work was supported by the Netherlands Foundation for Chemical Research (SON) with financial aid from the Netherlands Organisation for Scientific Research (NWO).References 1 C. J. Pedersen, J. Am. Chem. Soc., 1967, 89, 7017. 2 L. A. Singer and D. J. Cram, J. Am. Chem. Soc., 1965, 85, 1080.J. Chem. Soc., Dalton Trans., 1998, Pages 2083–2089 2089 3 D. J. Cram and R. H. Bauer, J. Am. Chem. Soc., 1959, 81, 5971. 4 D. J. Cram, Angew. Chem., 1988, 100, 1041. 5 B. Dietrich, J.-M.Lehn and J. P. Sauvage, Tetrahedron Lett., 1969, 34, 2885. 6 J.-M. Lehn, Angew. Chem., 1988, 100, 91. 7 F. C. J. M. van Veggel, W. Verboom and D. N. Reinhoudt, Chem. Rev., 1994, 94, 279. 8 A. E. Kaifer and S. Mendoza, Comprehensive Supramolecular Chemistry, ed. G. W. Gokel, Pergamon, Oxford, 1996, vol. 1, p. 701. 9 S. R. Miller, D. A. Gustowski, Z. Chen, G. W. Gokel, L. Echegoyen and A. E. Kaifer, Anal. Chem., 1988, 60, 2021. 10 P. D. Beer, H. Sikanyika, C. Blachburn, J. F. McAleer and M.G. B. Drew, J. Organomet. Chem., 1988, 356, C19. 11 P. D. Beer, A. D. Keefe and H. Sikanyika, J. Chem. Soc., Dalton Trans., 1990, 3289. 12 J. C. Medina, T. T. Goodnow, S. Bott, J. L. Atwood, A. E. Kaifer and G. W. Gokel, J. Chem. Soc., Chem. Commun., 1991, 290. 13 J. C. Medina, T. T. Goodnow, M. T. Rojas, J. L. Atwood, B. C. Lynn, A. E. Kaifer and G. W. Gokel, J. Am. Chem. Soc., 1992, 114, 10 583. 14 P. D. Beer, D. B. Crowe, M. I. Ogden, M. G. B. Drew and B. Main, J.Chem. Soc., Dalton Trans., 1993, 2107. 15 C. D. Hall, I. P. Danks, P. D. Beer, S. Y. F. Chu and S. C. Nyburg, J. Organomet. Chem., 1994, 468, 196. 16 H. Plenio, H. El-Desoky and J. Heinze, Chem. Ber., 1993, 126, 2403. 17 H. Plenio and R. Diodone, J. Organomet. Chem., 1995, 492, 73. 18 P. D. Beer, Z. Chen and M. G. B. Drew, J. Chem. Soc., Chem. Commun., 1993, 1046. 19 J. W. H. Smeets, R. P. Sijbesma, L. van Dalen, A. L. Spek, W. J. J. Smeets and R. J. M. Nolte, J. Org. Chem., 1989, 54, 3710. 20 D. Lednicer and C. R. Hauser, Org. Synth., 1960, 40, 31. 21 A. N. Mesmeyanova, E. G. Perevalova, L. S. Shilovtseva and V. D. Tyurin, Izv. Akad. Nauk USSR, Otd. Khim. Nauk, 1962, 1997. 22 D. E. Bublitz, J. Organomet. Chem., 1970, 23, 225. 23 R. P. Sijbesma, W. P. Bosman and R. J. M. Nolte, J. Chem. Soc., Chem. Commun., 1991, 885. 24 R. P. Sijbesma, A. M. Kentgens and R. J. M. Nolte, J. Org. Chem., 1991, 56, 3122. 25 S. Kulstad and L. Å. Malmsten, Acta Chem. Scand., Ser. B, 1979, 33, 469. 26 C. F. Martens, R. J. M. Klein Gebbink, M. C. Feiters and R. J. M. Nolte, J. Am. Chem. Soc., 1994, 116, 5667. 27 J. W. H. Smeets, H. C. Visser, V. E. M. Kaats-Richters and R. J. M. Nolte, Recl. Trav. Chim. Pays-Bas, 1990, 54, 3710. 28 J. W. H. Smeets, L. van Dalen, V. E. M. Kaats-Richters and R. J. M. Nolte, J. Org. Chem., 1990, 55, 454. 29 J. D. Carr, L. Lambert, D. E. Hibbs, M. B. Hursthouse, K. M. A. Malik and J. H. R. Tucker, Chem. Commun., 1997, 1649. 30 J. Granot, J. Magn. Reson., 1983, 55, 216. 31 J. N. H. Reek, A. H. Priem, H. Engelkamp, A. E. Rowan, J. A. A. W. Elemans and R. J. M. Nolte, J. Am. Chem. Soc., 1997, 119, 9956. 32 H. K. A. C. Coolen, J. A. M. Meeuwis, P. W. N. M. van Leeuwen and R. J. M. Nolte, J. Am. Chem. Soc., 1995, 117, 11 906. 33 R. P. Sijbesma, A. P. M. Kentgens, E. T. G. Lutz, J. H. van der Maas and R. J. M. Nolte, J. Am. Chem. Soc., 1993, 115, 8999. 34 M. Kranenburg, J. P. G. Delis, P. C. J. Kamer, P. W. N. M. van Leeuwen, K. Vrieze, N. Veldman, A. L. Spek, K. Goubitz and J. Fraanje, J. Chem. Soc., Dalton Trans., 1997, 1839. 35 U. Burkert and N. L. Allinger, Molecular Mechanics, American Chemical Society, Washington, DC, 1982. 36 R. S. Stojanovic and A. M. Bond, Anal. Chem., 1993, 65, 56. 37 R. P. Sijbesma and R. J. M. Nolte, Recl. Trav. Chim. Pays-Bas, 1993, 112, 643. Received 19th January 1998; Paper 8/00493E

ISSN:1477-9226    

DOI:10.1039/a800493e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2087–2095 2087 Electronic structures of copper(II) complexes of tetradentate hydroquinone-containing SchiV bases † Elizabeth H. Charles,a Li Mei Lindy Chia,a Joanne Rothery,a Emma L. Watson,a Eric J. L. McInnes,b Robert D. Farley,c Adam J. Bridgeman,a Frank E. Mabbs,b Christopher C. Rowlands c and Malcolm A. Halcrow*d a Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, UK CB2 1EW b EPSRC CW EPR Service Centre, Department of Chemistry, University of Manchester, Oxford Road, Manchester, UK M13 9PL c EPSRC ENDOR Service Centre, Department of Chemistry, University of Wales at CardiV, PO Box 912, CardiV, UK CF1 3TB d School of Chemistry, University of Leeds, Woodhouse Lane, Leeds, UK LS2 9JT.E-mail: M.A.Halcrow@chem.leeds.ac.uk Received 23rd March 1999, Accepted 23rd April 1999 Reaction of 2,5-dihydroxybenzaldehyde with 0.5 equivalent of 1,2-diaminoethane, trans-1,2-diaminocyclohexane or 1,2-diaminobenzene in the absence or presence of hydrated Cu(O2CMe)2?H2O in refluxing MeOH respectively aVorded the ligands H2L and complexes [Cu(L)] [H2L = N,N9-bis(2,5-dihydroxybenzylidene)-1,2-diaminoethane (H2L1), N,N9-bis(2,5-dihydroxybenzylidene)-trans-1,2-diaminocyclohexane (H2L2), N,N9-bis(2,5-dihydroxybenzylidene)- 1,2-diaminobenzene (H2L3)] in yields of 66–86%.Using X- and Q-band EPR and 1H and 14N X-band ENDOR data, the following fractional unpaired spin densities were calculated: r{Cu} = 0.75, r{N} = 0.07, r{O} £ 0.04, r{H} = 0.01.Density functional (DF), intermediate neglect of diVerential overlap with spectroscopic parameterisation (INDO/S) and extended Hückel calculations of [Cu(MeOsalen)] broadly reproduced these results, the DF calculations demonstrating that the phenoxide oxygen lone pair is misdirected away from the Cu–O bond. The cyclic voltammograms of the ligands and complexes in dmf–0.1 M NBun 4PF6 at 293 K showed a single oxidation of the two ligand hydroquinone groups, and two principal daughter processes: an irreversible reduction of the initial oxidised quinone, probably in a monoprotonated or metallated form; and a more cathodic reduction and associated reoxidation indicative of a proton- and metal ion-induced electrochemical step–chemical step–electrochemical step reaction.Introduction The redox, spectroscopic, structural and photochemistries of catecholate and ortho-semiquinone complexes continue to be very well studied.1 However, a much smaller literature exists for metal complexes of ligands containing s-co-ordinated parahydroquinone or para-benzoquinone groups,2–12 despite the potential for novel redox chemistry in such compounds.Indeed, we are aware of only three electrochemical studies of parahydroquinonate complexes, one of which appeared during the course of this work.4,5,12 We note that H2L1, H2L2 and a small number of their complexes have been briefly described, although in all cases only basic analytical data (IR, CHN) were reported.7 Since H2L1 can be thought of as a “non-innocent” analogue of the classic SchiV base ligand H2salen, we were interested to examine in more detail the co-ordination chemistry of hydroquinone-containing SchiV ligands of this type.We report here the syntheses, and a preliminary voltammetric study, of the copper(II) complexes of H2L1–H2L3. Also described is a comprehensive EPR, ENDOR and theoretical characterisation of these complexes, which has allowed us to define in some detail the spin distribution within this ubiquitous class of compounds.† Abbreviations used: 2,5-dhb = 2,5-dihydroxybenzaldehyde; H2MeOsalen = N,N9-bis(5-methoxysalicylidene)-1,2-diaminoethane; H2MeOsalophen = N,N9-bis(5-methoxysalicylidene)-1,2-diaminobenzene; H2salophen = N,N9-bis(salicylidene)-1,2-diaminobenzene. Results and discussion Syntheses and characterisation of the SchiV base ligands and complexes Following the usual method for the synthesis of H2salen and its analogues,13 treatment of 2,5-dhb with 0.5 molar equivalent of the appropriate diamine in refluxing MeOH for 3 h aVorded H2L1, H2L2 or H2L3 as analytically pure solids in 72–86% yield.The 1H NMR spectra of H2L1–H2L3 in (CD3)2SO are consistent with the proposed molecular structures, showing resonances for the 2- and 5-hydroxyl groups close to d 12.5 and 9.0, respectively, and for the aldimine proton at d 8.4–8.8, in addition to the appropriate pattern of aryl and alkyl resonances.In keeping with the microanalysis of H2L3, one mole equivalent of MeOH is also revealed in the 1H spectrum of this compound. Attempted syntheses of H2L3 in CHCl3 or MeCN, intended to give solvent-free ligand, aVorded mixtures of products that we were unable to separate. The monoimine HL4 was also prepared as a model compound for the electrochemical studies, being obtained in analytical purity by reaction of 2,5-dhb with MeNH2 in MeOH.The complexes [Cu(L)] (L22 = [L1]22, [L2]22 or [L3]22) were synthesized under the same conditions employed for the “free” ligands, with the addition of 0.5 molar equivalent of Cu(O2CMe)2?H2O as template ion. These reactions proceeded cleanly, aVording brown precipitates, which were filtered oV, washed with MeOH and Et2O and dried in vacuo. Microanalysis was consistent with the formulation of these solids as2088 J. Chem.Soc., Dalton Trans., 1999, 2087–2095 the desired compounds as hydrates or methanol solvates, while FAB mass spectrometry confirmed the presence of the intact complex molecule while showing no significant fragmentation or impurity peaks. As for the “free” ligands, attempted template syntheses in CHCl3 or MeCN did not yield pure products. However, mixing of MeCN solutions of Cu(O2CMe)2?H2O and preformed H2L2 or H2L3 resulted in the precipitation of brown microcrystals, both of which reproducibly analysed for [Cu(L)]?0.5H2O?0.5MeCN (L22 = [L2]22 or [L3]22).The ligands H2L2 and H2L3 are sparingly soluble in MeCN and MeNO2. However, H2L1 and the complexes are insoluble in all common solvents except dmf, dma (dimethylacetamide) and dmso, in which they have only sparing solubilities. This low solubility has severely hampered our eVorts to recrystallise these complexes, and we have thus far been unable to obtain X-ray quality crystals of any of the compounds in this study.The UV/visible spectra of [Cu(L1)] and [Cu(L2)] in dmf at 298 K are barely distinguishable, these two complexes showing a d–d maximum close to nmax 17500 cm21 with emax = 420–430 M21 cm21. This band is at the high end of the range of energies reported for other copper(II) complexes of salen-type ligands, which lie between nmax 15200 and 17900 cm21.14–19 For [Cu(L3)] this peak is obscured by more intense ligand-derived absorptions, occurring as a shoulder close to nmax = 15400 cm21.The remaining bands in all the complexes can be assigned to p æÆ p* transitions within the co-ordinated SchiV bases.14,20 These p æÆ p* absorptions lie up to 5000 cm21 lower in energy than those previously reported for copper(II) complexes of salen and salophen,14,16,18,20–22 consistent with the increased electron-richness of the dihydroxyphenyl rings in [L1]22–[L3]22. EPR and ENDOR spectroscopy of [Cu(L)] (L22 5 [L1]22– [L3]22) The X-band EPR spectra of [Cu(L)] (L22 = [L1]22–[L3]22) in 10 : 1 dmf: toluene solution at 293 K are essentially identical, the isotropic EPR parameters derived from them closely matching those reported previously for other copper(II) complexes of bis-salicylaldimine ligands.19,21–25 The spectra consist of the expected 4-line pattern from coupling to 63,65Cu (I = ��� ), with ·gÒ = 2.100 (L22 = [L1]22), 2.096 (L22 = [L2]22) and 2.097 {L22 = [L3]22; Fig. 1(a)}. Although suVering from mI-dependent line broadening, the mI = 2��� line in each spectrum shows superhyperfine coupling to two 14N and two 1H nuclei, which correspond to the ligand aldimino N]] CH protons.26 The spectra for all three complexes were simulated assuming ·A{63,65Cu}Ò = 85.2 × 1024, &miOgrave; = 13.7 × 1024 and ·A{1H}Ò = 7.4 × 1024 cm21 with errors of ±0.5 × 1024 cm21 [Fig. 1(b)]. In frozen dmf: toluene solution the X-band EPR spectra exhibit the axial pattern typical of tetragonal copper(II) complexes with an approximate {dxy}1 or {dx22y2}1 ground state,27 showing g|| ª 2.21, g^ ª 2.05, A|| {63,65Cu} ª 200 × 1024 cm21 [Fig. 2(a)]. Interestingly, the perpendicular regions of these spectra show clearly resolved hyperfine and superhyperfine interactions. The axial symmetry of the system was confirmed by obtaining spectra at Q band, which reproduced the g values derived from X-band data well but no longer showed resolved coupling in the perpendicular region [Fig. 2(c)]. No spin-triplet signal suggesting partial dimerisation of the complexes in solution 25 was detected.Given the unusually well resolved anisotropic X-band spectra shown by these copper(II) complexes, [Cu(L1)] and [Cu(L3)] were further investigated by X-band ENDOR spectroscopy; no study of [Cu(L2)] was attempted because of its lower solubility. All ENDOR measurements were performed in frozen dmf : toluene solution at 10 K; 1H and 14N spectra at 8 ([Cu(L1)]) and 10 Fig. 1 Second-derivative EPR spectrum of [Cu(L3)] in 10 : 1 dmf: toluene solution at 298 K: (a) experimental X-band spectrum (n = 9.445 GHz); (b) simulation using ·gÒ = 2.100, ·A{63,65Cu}Ò = 85.2 × 1024, ·A{14N}Ò = 13.7 × 1024, ·A{1H}Ò = 7.4 × 1024 cm21 and linewidth coeYcients A = 19, B = 15 and C = 5 G for the expression DB = A 1 BmI 1 CmI 2, where DB is the peak-to-peak linewidth of the individual copper hyperfine lines.J.Chem. Soc., Dalton Trans., 1999, 2087–2095 2089 ([Cu(L3)]) diVerent applied fields were measured, covering the complete EPR absorption for the two compounds.In the following discussion it is assumed that the molecular, g and A{63,65Cu} axes in the ENDOR and EPR experiments are coincident, with the parallel direction corresponding to the molecular z axis [Scheme 1(a)]. The local 14N tensor is assumed to be axial, with A||{14N} aligned along the corresponding Cu–N bond [Scheme 1(b)]. Both compounds gave single crystal-type 14N ENDOR spectra at a field corresponding to the lowest-field A||{63,65Cu} hyperfine transition, exhibiting well resolved A^{14N} couplings. Fig. 2 First derivative EPR spectra of [Cu(L3)] in 10 : 1 dmf: toluene solution at 120 K: (a) experimental X-band spectrum (n = 9.445 GHz); (b) simulated X, using fitting parameters described in the text; (c) experimental Q-band spectrum (n = 34.200 GHz). The * shows the field position for the ENDOR spectrum shown in Fig. 3. Scheme 1 Molecular axes for copper(II) complexes of salen-type ligands with C2v symmetry: (a) true molecular axes, and axes for the g and A{63,65Cu} tensors; (b) local axes for the A{14N} tensor.At higher fields both A||{14N} and A^{14N} features were observed (Fig. 3), since the mI = 2��� A||{63,65Cu} feature overlays the EPR g^ region. Importantly both these couplings were essentially invariant at diVerent applied fields across the perpendicular region of the spectrum, showing that the spectra had true powder character reflecting the selection of the vast majority of molecular orientations within the EPR linewidth.The following coupling constants were derived: for [Cu(L1)], |A||{14N}| = 46.5 (14.3 × 1024) and |A^{14N}| = 38.9 MHz (13.0 × 1024 cm21); for [Cu(L3)], |A||{14N}| = 46.3 (14.3 × 1024) and |A^{14N}| = 36.9 MHz (12.3 × 1024 cm21). These must correspond to the true values, given the relative alignments of the axes for the A{14N} and A{63,65Cu} tensors (Scheme 1).28 All the 14N ENDOR spectra were significantly broadened, presumably by unresolved quadrupolar couplings, which prevented the detection of second-order splittings that might be expected at these frequencies.The 1H ENDOR spectra obtained at the same field employed for the 14N parallel spectra (Fig. 2) were also well resolved. For [Cu(L1)] 5 of the 6 expected proton environments were detected, with |A{1H}| = 4.5, 2.3, 1.7, 0.7 and 0.2 MHz (1.5, 0.7, 0.5, 0.2 and 0.1 × 1024 cm21); we assign the strongest of these couplings to the alkyl CH2 protons of the [L1]2 ligand, and the other values to the C–H and O–H environments of the hydroquinonate rings.For [Cu(L3)], 4 distinct proton coupling constants were observed at |A{1H}| = 1.7, 1.3, 0.7 and 0.3 MHz (0.5, 0.4, 0.2 and 0.1 × 1024 cm21). A weak low frequency feature corresponding to |A{1H}| = 17.5 and 15.6 MHz for [Cu(L1)] and [Cu(L3)], respectively, might arise from one peak of the aldimine proton resonance; however, the putative high frequency component of this doublet is obscured by the stronger 14N peaks, preventing corroboration of this assignment.Unfortunately, at higher fields extremely complex powder-type 1H spectra were obtained, which could not be unambiguously interpreted. The orientations and symmetries of the local axes for these protons are therefore unknown, and no more detailed analysis is possible. The frozen solution X-band EPR spectra of [Cu(L1)] and [Cu(L3)] were simulated using the ENDOR results for A{14N}.In the absence of conclusive 1H ENDOR data, the superhyper- fine interaction to the aldimine N]] CH nuclei was treated as isotropic, the simulations proving sensitive to the value of this coupling. Good simulations were obtained for both [Cu(L1)] and [Cu(L3)] using the following parameters [Fig. 2(b)], which diVer only in the linewidths employed for the two complexes: g|| = 2.210, g^ = 2.045, A||{63,65Cu} = 203 × 1024, A^{63,65Cu} = 27 × 1024, A||{14N} = 12 × 1024, A^{14N} = 15 × 1024, A||{1H} = A^{1H} = 7 × 1024 cm21 W|| = 15, W^ = 14 ([Cu(L1)]) or 10 G ([Cu(L3)]).Note that these EPR simulations fit all the g and tensors to the true molecular axes [Scheme 1(a)], so that Fig. 3 X-Band 14N (2nN = 2.04 MHz) ENDOR spectrum of [Cu(L3)] in 10 : 1 dmf: toluene solution at 10 K. The field position is marked on Fig. 2. Features ‘x’ are radiofrequency artifacts.2090 J. Chem. Soc., Dalton Trans., 1999, 2087–2095 A||{14N} from the EPR simulations is equivalent to the true A^{14N} value measured by ENDOR.EPR Calculations The spin-Hamiltonian parameters from the EPR study were employed to calculate covalency parameters according to the ligand-field method of Kivelson and Neiman.29,30 Unfortunately, the original description of this approach29 contains several misprinted equations and constants, which do not seem to have been corrected in later reports. We will therefore discuss our calculations in detail.Metal–ligand bonding about a D4h copper(II) centre is described according to eqns. (1)–(4): where l0 is the spin–orbit g|| 2 ge = 28r{ab 2 a9bS 2 [a9(1 2 b2)� �� T(n)/2]} (1) g^ 2 ge = 22m{ab 2 a9dS 2 [a9(1 2 d2)� �� T(n)/2� �� ]} (2) A||{63,65Cu} = PF2 4 7 a2 2 k0 1 (g|| 2 ge) 1 3 7 ( g^ 2 ge) 2 8r{a9bS 1 [a9(1 2 b2)� �� T(n)/2]} 2 6 7 m{a9dS 1 [a9(1 2 d2)� �� T(n)/2� �� ]}G (3) A^{63,65Cu} = PF2 7 a2 2 k0 1 1 4 ( g^ 2 ge) 2 m{a9dS 1 [a9(1 2 d2)� �� T(n)/2� �� ]}G (4) coupling constant for the free copper(II) ion (2828 cm21 31), k0 is the Fermi contact constant for the free ion (0.33 26,29,31), S is the Cu–N/O overlap integral (0.093 for N, 0.076 for O) assuming Cu–N/O bonds of 1.9 Å,26,29 the constant T(n) has the value 0.333 for nitrogen ligation and 0.220 for O, the constant P has the value 0.036 cm21 31 and m and r have the definitions 29 in eqns.(5) and (6). Here, a2 and a92 correspond respectively to r = l0ab/DEx2 2 y2 (5) m = l0ad/DExz (6) covalency factors at Cu{dxy} and the ligand donor atoms for metal–ligandSimilarly, b and d are covalency factors at Cu for a2 1 a92 2 2aa9S = 1 (7) in-plane metal–ligand p bonding (Cu{dx2 2 y2}) and out-ofplane p bonding (Cu{dxz/dyz}), respectively. Although these equations do not allow for d-orbital mixing, which is possible in the C2v point group exhibited by [Cu(L1)] and [Cu(L3)] (Scheme 1), the axial symmetry of the g and A tensors for our compounds suggests that this approximation is justified. Several calculations of this type for [Cu(salen)] and related SchiV base complexes have been reported previously,21,24,26,29,30 with somewhat contradictory results which reflect diVerent values of DEx2 2 y2, DExz, S and T(n), and/or an incorrect value of k0, used by each group.We employed average values of S = 0.085 and T(n) = 0.276 in eqns. (1)–(4), to reflect the N2O2 ligation at Cu. The d–d transition energies are more problematic, however, particularly since no single crystal UV/vis data are available for [Cu(salophen)].For [Cu(salen)], the DEx2 2 y2 energy has been shown to lie close to the d–d maximum.14,32–34 Therefore, we estimated DEx2 2 y2 = 17500 cm21 for [Cu(L1)] and 16000 cm21 for [Cu(L3)] (from the UV/vis of [Cu(MeOsalophen)], which shows a maximum here in dmf 21). The DExz transition has not been unambiguously assigned for [Cu(salen)], and may be obscured by DEx2 2 y2 or by charge transfer bands with nmax £ 22000 cm21.14,34 We therefore performed calculations using DExz values corresponding to both of the possible ranges (21000– 25000 cm21 for [Cu(salen)] 14,32–34), and assuming that the ratio DEx2 2 y2 :DExz is the same for [Cu(L1)] and [Cu(L3)] (Table 1).The covalency factors were derived from eqns. (1)–(4) by an iterative procedure.29 The results of these calculations, together with the estimated transition energies employed, are listed in Table 1.Of our calculated parameters, only b 2 and d 2 are sensitive to the values of the estimated d–d transition energies. Our values for a2 are close to values previously calculated by others for [Cu(salen)] from less complete EPR data,21,24,30 while our values of b 2 are similar to those recently derived by Srinivas and co-workers.21,24 Our values of d 2 seem anomalously low in the light of our MO calculations which imply negligible out-ofplane p bonding (see below); this suggests that we may have significantly underestimated DExz for the two compounds.Nonetheless, the insensitivity of a2 to DEx2 2 y2 and DExz implies that, despite the lack of accurate transition energies, we have a consistent description of the metal–ligand s bonding in these complexes. The a92 parameters calculated above correspond to a cumulative value for the 2 N- and 2 O-donor atoms of the chelate ligands.35 A value a92{N} for the 2 N-donors only can be derived from A{14N}, using eqns.(8) and (9) which assume coupling to 2 sp2-hybridised N atoms; 26‡ where gL is the magnetogyric ratio of the 14N nucleus (0.404), d(r) is the value of the 2s function at the nitrogen nucleus (33.4 × 1024) and ·r23Òp for nitrogen 2p orbitals has the value 21.1 × 1024.29 Using this method we obtain a92{N} = 0.16 (i.e. 0.08 per N atom) for both complexes. We can also derive a covalency factor for the ligand hydrogen nuclei, by dividing S·A{1H}Ò by the hyperfine constant of the free H atom (1420.4 MHz).36 Assuming ·A{1H} Ò ª A||{1H}, this equates to a total a92{H} = 0.04, of which ca. 80% derives from the aldimine group protons. Finally, using eqn. (10), which assumes zero overlap between non-bonded atoms, we obtain a92{O} £ 0.08. This represents an upper limit for a92{O} since eqn. (10) does not allow for delocalisation of the unpaired spin onto the ligand C nuclei (a92{C}), which should be of a similar magnitude to a92{H}.Nonetheless, it is apparent that for [Cu(L1)] and [Cu(L3)] greater s covalency exists within the copper–ligand bonds to the ‘softer’ sp2 N-donors compared to the ‘harder’ anionic O-donors. There is good agreement between values of a92 calculated from eqns. (1)–(5) (0.31) and a92{N} 1 a92{H} 1 a92{O} from eqns. (8)–(10) (0.28). A||{14N} = Sa9{14N}2 2 D(2gLbObN)F2 8p 9 d(r) 1 8 15 ·r23 ÒpG (8) A^{14N} = Sa9{14N}2 2 D (2gLbObN)F2 8p 9 d(r) 2 1 15 ·r23ÒpG (9) Table 1 Covalency factors for [Cu(L1)] and [Cu(L3)] calculated from eqns.(1)–(7), showing the estimated d–d transition energies employed in the calculations [Cu(L1)] [Cu(L3)] DEx2 2 y2/cm21 DExz/cm21 a2 a92 b2 d2 17500 21000 0.84 0.24 0.72 0.73 17500 25000 0.83 0.25 0.72 0.86 16000 19000 0.84 0.24 0.68 0.67 16000 23000 0.83 0.25 0.68 0.79 ‡ Earlier EPR calculations of copper(II) salicylaldimine complexes in refs. 24, 29 and 30 did not correct eqns. (8) and (9) to describe coupling to 2 rather than 4 equivalent ligand donors.Values of a92 calculated from A{14N} in these references are therefore misleading.J. Chem. Soc., Dalton Trans., 1999, 2087–2095 2091 a2 1 a92{N} 1 a92{O} 1 a92{H} 2 2aa9{N}S{Cu,N} 2 2aa9{O}S{Cu,O} = 1 (10) MO Calculations of [Cu(MeOsalen)] For comparison with our EPR results, calculations were carried out using crystal structure coordinates for [Cu(MeOsalen)] 24 at the Extended Hückel (EH), Intermediate Neglect of DiVerential Overlap with Spectroscopic parameterisation (INDO/S) and Density Functional (DF) levels.The three methods gave similar descriptions of bonding within the complex, with only slight variations in the orbital ordering. Since the DF calculation is likely to aVord the most accurate description of the molecule only this will be discussed in detail. Fig. 4 shows the eigenvalues and approximate descriptions of the highest lying orbitals in the complex from the DF study. The calculations predict a 2B2 ground state in the approximate C2v symmetry of the copper ion, with a singly occupied orbital (SOMO) of mainly dxy character.The low energy of the copper dxz and dyz orbitals suggests that the out-of-plane p interaction is small. This is confirmed by the very small changes in the Mulliken populations for these two orbitals compared to their values in a purely ionic crystal field. The low energy of the copper dz2 orbital is mainly a result of s–d hybridisation.37 The copper dx2 2 y2 orbital suVers a large ligand-field eVect, predominantly due to interaction with the in-plane lone pairs of the co-ordinated oxygen atoms. Analysis of the density matrix leads to Cu–N and Cu–O bond orders of 0.68 and 0.57 respectively, reflecting the poor donor ability of the more electronegative oxygen and the antibonding role of its lone pairs.Interestingly, the DF calculation shows that the O(1) lone pair is misdirected away from the Cu–O vector (Fig. 5), as previously proposed for [Co(salen)].38 This feature is not reproduced by the EH or INDO/S calculations, which show the O(1) lone pair aligned along the Cu–O bond. In order to compare the results of the EPR and MO calculations the EPR-derived covalency factors must be converted into fractional spin densities according to eqns. (11) and (12) (E = N, O or H). The resultant experimental composition of the r{Cu} = a2 [a2 1 a92{N} 1 a92{O} 1 a92{H}] (11) r{E} = a92{E} 2[a2 1 a92{N} 1 a92{O} 1 a92{H}] (12) Fig. 4 Kohn–Sham eigenvalue diagram for the 2B2 ground state of [Cu(MeOsalen)], as calculated by the DF method (1 eV ª 8065.5 cm21). The percentage character of the metal-dominated functions is given with symmetry labels corresponding to the approximate C2v symmetry of the first co-ordination sphere. The unlabelled functions are all ligand-based p orbitals with varying amounts of metal dxz and dyz contributions. SOMO of [Cu(L1)], and those predicted by the EH, INDO/S and DF calculations, are summarised in Table 2, the SOMO derived from the DF method being displayed in Fig. 5. Only Cu, N, O(1) and H(7) (Fig. 5) contribute a fractional spin density � 0.01 to the SOMO; there is an additional cumulative contribution from the carbon content of the molecule (Sr{C}, Table 2) that approximately equally distributed between C(1), C(6) and C(7). Each level of theory predicts a cumulative fractional spin density of 0.90–0.91 on the [CuN2O2] core of the molecule, compared to the experimental value of 0.83–0.97, depending on the true value of r{O}.However, the EH and INDO/S calculations predict a reduced metal contribution to this orbital, and correspondingly increased contributions from the N- and O-donors, compared to the DF results. Given the approximations in our EPR calculations, there is reasonable agreement between the experimental and DF fractional spin densities.39 Electrochemical studies § Cyclic voltammograms of the ligands and complexes were run in dmf–0.1 M NBun 4PF6 at 293 K; the results thus obtained are summarised in Table 3.Unless otherwise stated, all potentials quoted refer to measurements run at a scan rate (‘n’) of 100 mV s21, and are quoted against an internal ferrocene–ferrocenium standard. Under these conditions, 2,5-dhb exhibits a single irreversible oxidation (P1) at Epa = 10.90 V with a single broad irreversible daughter (P2) at Epc = 20.52 V. Addition of up to 5 molar equivalents of 2,6-dimethylpyridine to the sample causes a broadening and a 220 mV cathodic shift of P1, reflecting deprotonation of 2,5-dhb, without aVecting the potential of the P2 process.By comparison with the known voltammetric behaviour of hydroquinones in non-aqueous solvents,40,41 we assign P1 and P2 to the processes in Scheme 2. The nonappearance of a reversible Q–Q22 couple upon addition of base 41 suggests that the QH1 species has an unusually high pKa, which may reflect intramolecular hydrogen bonding to the carbaldehyde substituent (Scheme 2).42 The cyclic voltammograms of H2L1–H2L3 exhibit, in addition to P1 and P2, an additional irreversible daughter reduction P3 Fig. 5 Composition of the SOMO for [Cu(MeOsalen)], as calculated by the DF method. The atom numbering scheme is that employed in Table 1. The molecule has approximate C2v symmetry. Table 2 Experimental fractional unpaired spin densities for [Cu(L1)], and calculated fractional spin densities for the SOMO of [Cu(MeOsalen)].The atom numbering scheme is shown in Fig. 5 Experimental EH INDO/S DF r{Cu} r{N} r{O(1)} r{H(7)} Sr{C}a 0.75 0.07 £0.04 0.01 £0.04 0.44 0.16 0.07 0.01 0.04 0.52 0.10 0.09 0.01 0.04 0.63 0.07 0.07 0.01 0.03 a Cumulative fractional spin density for the carbon atoms C(1)–C(8). § The nomenclature for describing the diVerent redox states of quinones is described in ref. 39.2092 J. Chem. Soc., Dalton Trans., 1999, 2087–2095 Table 3 Cyclic voltammetric data for the ligands and complexes (dmf–0.1 M NBun 4PF6, 298 K, n = 100 mV s21).All potentials quoted vs. the ferrocene–ferrocenium couple. For the complexes, metal-centered processes are not listed, but are discussed in the text P1/P19 Epa/V P2/P29 Epc/V P3/P39 Epc/V P4/P49 Epa/V Other daughters Epc/V 2,5-dhb H2L1 H2L2 H2L3 HL4 [Cu(L1)] [Cu(L2)] [Cu(L3)] 10.90 10.37 10.4 a 10.43 10.51 10.16 10.18 10.28 20.52 20.76 20.68 20.59 20.63 20.23 20.28 20.25 b — 21.24 21.15 21.00 21.41 20.96 20.97 20.93 — 20.83 20.86 20.84 — 20.77 20.79 20.70 ————— 20.7 a 20.7 a 20.7,a 21.3 a a Broad shoulder.b n = 500 mV s21. Not detected at n = 100 mV s21. with an associated reoxidation P4 [Fig. 6(a)]. At increased scan rates P2 becomes more intense relative to P3 and P4, and vice versa. Upon addition of 3 equivalents of 2,6-dimethylpyridine to all the ligands, P3 and P4 are no longer observed while P2 is enhanced in intensity [Fig. 6(b)]; the potential of P1 is essentially unaVected.Conversely, addition of up to 5 equivalents of MeOH or water to H2L1 or H2L2 results in diminution of P2 relative to P3/P4; the CV of H2L3 is similarly aVected by water [Fig. (6c)], but is unchanged upon treatment with MeOH. Under conditions where P2 is not observed, the ratio of Ipa{P1}: Ipc{P3} is 3.5–4.0 :1. The cyclic voltammogram of HL4 is Fig. 6 Cyclic voltammograms (dmf–0.1 M Bun 4NPF6, 298 K, 100 mV s21) of: (a) H2L3?MeOH; (b) H2L3?MeOH 1 3 mole equivalents of 2,6- dimethylpyridine; (c) H2L3?MeOH 1 5 mole equivalents of water.Scheme 2 X = O or NR. similar to those of the bis-imines, although a wave corresponding to P4 was not detected under any of the above conditions. The disappearance of P3 and P4 upon addition of base [Fig. 6(b)] rules out their potential assignment as a 1,4-benzoquinone carbaldimine Q–Q22 couple.41 Given the ª4 : 1 ratio for Ipa{P1}: Ipc{P3}, P3 and P4 probably correspond to the products of a proton-catalysed ECE (electrochemical step–chemical step–electrochemical step) reaction, which will be induced by the liberation of protons at the anode following oxidation (Scheme 2).Precedent suggests three plausible pathways for the chemical step in this process: radical coupling or atom abstraction by QH? or QH2~1;40 hydrolysis of the QH1 imine moiety by adventitious moisture, followed by Michael addition of liberated amine to the quinone ring; 43 or nucleophilic attack of H2O or, where present, MeOH at QH1 or QH21.44¶ We are presently unable to distinguish between these possibilities, however, and the origin of the instability of the 1,4-benzoquinone carbaldimine function, compared to 1,4-benzoquinone carbaldehyde, is unclear.The three copper(II) complexes show similar ligand-based redox chemistry to that of the “free” ligands, exhibiting one ligand-based oxidation (P19) and three principal daughter peaks P29–P49; one or more weaker daughter reductions close to P39 are also evident, however (Fig. 7). The P19 oxidation shows pronounced asymmetry, which at n £ 25 mV s21 resolves itself into a low-potential shoulder occurring at approximately 0.5 of the total Ipa for this process. The behaviour of P29–P49 upon varying n is identical to P2–P4 for the uncomplexed ligands. However, addition of 2,6-dimethylpyridine or MeOH to the samples did not change the relative intensities of P29–P49. Decomposition of 1,4-benzoquinone carbaldimines hence appears to be promoted by co-ordination to a metal ion, although it is unclear whether the P3/P4 and P39/P49 species result from the same ECE reaction.Fig. 7 Cyclic voltammogram of [Cu(L2)], showing the ligand-centered processes only (dmf–0.1 M Bun 4NPF6, 298 K, 100 mV s21). ¶ The presence of moderate hydrogen bond donors or acceptors such as MeOH or water has a negligible eVect on the oxidation potentials of hydroquinones.45J. Chem.Soc., Dalton Trans., 1999, 2087–2095 2093 The instability of our complexed ligands to electrooxidation contrasts with other voltammetric studies of s-hydroquinonate complexes, for which chemically reversible ligand redox cycles have been reported.4,12 Our results also contrast with the only other electrochemical study of a hydroquinone-containing SchiV base, namely that of Feringa and co-workers,5 who prepared a dicopper complex of a bicompartmental ligand with a 2,6-di(formaldimino)-4-hydroxyphenoxy bridging group.In this case no ligand oxidation was detected within the window of the MeCN solvent, which they attributed to the presence of 2 electron-withdrawing carbaldimine substituents on the hydroquinonyl group. A more detailed study of H2L1–H2L3, their substituted derivatives and complexes with other transition ions, designed to identify the final products of ligand oxidation, is in progress and will be reported separately. In addition to the ligand-centred processes, [Cu(L1)]– [Cu(L3)] exhibit a reduction with Ipc = (0.2–0.3)Ipa{P19}, which we assign as a CuII–CuI reduction.For [Cu(L1)] and [Cu(L2)] this process is irreversible, occurring at Epc = 21.9 V; for [Cu(L3)] a reversible CuII–CuI couple at E2� 1 = 21.70 V is observed. Compound [Cu(L3)] also exhibits an irreversible peak at Epc = 22.56 V with no detectable daughters, which we attribute to a ligand reduction associated with the [L3]22 phenylenediimine moiety.Experimental Unless stated otherwise, all manipulations were performedn air using commercial grade solvents. Electrochemical studies employed anhydrous 99.8% dmf (Aldrich). 2,5-Dihydroxybenzaldehyde, methylamine (1.0 M solution in MeOH), trans- 1,2-diaminocyclohexane (Aldrich), 1,2-phenylenediamine, 1,2- ethylenediamine and Cu(O2CMe)2?H2O (Avocado) were used as supplied; H2L1 was prepared by the literature method.7 Syntheses N,N9-Bis(2,5-dihydroxybenzylidene)-trans-1,2-diaminocyclohexane (H2L2). 2,5-Dihydroxybenzaldehyde (2.0 g, 1.45 × 1022 mol) and trans-1,2-diaminocyclohexane (0.83 g, 7.25 × 1023 mol) were refluxed in MeOH (50 cm3) for 3 h, aVording a bright yellow solution which was evaporated to a yellow oil. Trituration of this oil with Et2O yielded a mustard yellow microcrystalline solid, which was filtered oV, washed with Et2O and dried in vacuo. Yield 2.2 g, 86% (Found: C, 67.7; H, 6.3; N, 7.9. Calc. for C10H11NO2: C, 67.8; H, 6.3; N, 7.8%), mp 195 8C (decomp.).FAB mass spectrum: m/z 355 [M 1 H]1; and 354, [M]1. 1H NMR [(CD3)2SO]: d 12.47 [s, 2 H, phenyl OH2], 8.96 [s, 2 H, OH5], 8.36 [s, 2 H, CH]] N], 6.72 [dd, J = 2.8, 9.1, 2 H, phenyl H4], 6.72 [d, J = 2.8, 2 H, phenyl H6], 6.64 [d, J = 9.1 Hz, 2 H, phenyl H3], 3.33 [m, 2 H, cyclohexyl CHN] and 1.42–1.86 [m, 8 H, cyclohexyl CH2]. N,N9-Bis(2,5-dihydroxybenzylidene)-1,2-diaminobenzene (H2L3). 2,5-Dihydroxybenzaldehyde (2.0 g, 1.45 × 1022 mol) and 1,2-phenylenediamine (0.78 g, 7.25 × 1023 mol) were refluxed in MeOH (50 cm3) for 3 h.The resultant brick red precipitate was filtered oV, washed with MeOH until the washings were colourless, washed with Et2O and dried in vacuo. Yield 1.8 g, 72% (Found: C, 66.4; H, 5.3; N, 7.5. Calc. for C20H16NO4?CH3OH: C, 66.3; H, 5.3; N, 7.4%), mp 140 8C (decomp.). FAB mass spectrum: m/z 349, [M 1 H]1; and 348, [M]1. 1H NMR [(CD3)2SO]: d 12.18 [s, 2 H, phenyl OH2], 9.14 [s, 2 H, phenyl OH5], 8.82 [s, 2 H, CH]] N], 7.41 [m, 4 H, phenylene CH], 7.06 [d, J = 2.9, 2 H, phenyl H6], 6.89 [dd, J = 2.9, 8.8, 2 H, phenyl H4], 6.81 [d, J = 8.8, 2 H, phenyl H3], 4.13 [q, J = 5.3, 1 H, CH3OH] and 3.17 [d, J = 5.3 Hz, 3 H, CH3OH]. 2,5-Dihydroxybenzylidenemethylamine (HL4). 2,5-Dihydroxybenzaldehyde (0.5 g, 3.63 × 1023 mol) and methylamine (3.63 cm3 of a 1.0 M solution in MeOH, 3.63 × 1023 mol) were stirred in MeOH (20 cm3) for 1 h. The resultant orange solution was evaporated to an orange microcrystalline residue, which was washed with the minimum volume of Et2O and dried in vacuo.Yield 0.48 g, 88% (Found: C, 63.3; H, 6.0; N, 9.0. Calc. for C8H9NO2: C, 63.6; H, 6.0; N, 9.3%), mp 105 8C (decomp.). FAB mass spectrum: m/z 152, [M 1 H]1; 151, [M]1. 1H NMR [(CD3)2SO]: d 12.55 [s, 1 H, phenyl OH2], 8.93 [s, 1 H, OH5], 8.44 [s, 1 H, CH]] N], 6.79 [d, J = 2.9, 1 H, phenyl H6], 6.72 [dd, J = 2.9, 8.7, 1 H, phenyl H4], 6.58 [d, J = 8.7 Hz, 1 H, phenyl H3] and 3.41 [s, 3 H, CH3].[N,N9-Bis(2,5-dihydroxybenzylidene)-1,2-diaminoethanato]- copper(II) ([Cu(L1)]). A MeOH solution of 2,5-dihydroxybenzaldehyde (0.20 g, 1.45 × 1023 mol), 1,2-ethylenediamine (0.044 g, 7.25 × 1024 mol) and Cu(O2CMe)2?H2O (0.14 g, 7.25 × 1024 mol) was refluxed for 2 h, to give a brown precipitate. This was filtered oV, washed with MeOH until the washings were colourless, washed with Et2O and dried in vacuo. Yield 0.17 g, 66% (Found: C, 50.0; H, 4.2; N, 7.2. Calc.for C16H14CuN2O4?H2O: C, 50.0; H, 4.2; N, 7.4%). FAB mass spectrum: m/z 362, [63Cu(L1) 1 H]1; and 361, [63Cu(L1)]1. UV/vis spectrum (dmf): nmax 17500 (emax = 420), 24800 (9800 M21 cm21), and 36600 (sh) cm21. [N,N9-Bis(2,5-dihydroxybenzylidene)-1,2-diaminocyclohexanato] copper(II) ([Cu(L2)]). Method A. As for [Cu(L1)], employing trans-1,2-diaminocyclohexane (0.083 g, 7.25 × 1024 mol). Yield 0.22 g, 73% (Found: C, 56.0; H, 5.3; N, 6.3. Calc. for C20H20CuN2O4?CH3OH: C, 56.3; H, 5.4; N, 6.2%).FAB mass spectrum: m/z 416, [63Cu(L2) 1 H]1; and 415, [63Cu(L2)]1. UV/ vis spectrum (dmf): nmax 17700 (emax = 410), 24800 (10500 M21 cm21) and 36000 (sh) cm21. Method B. Solutions of H2L2 (0.50 g, 1.41 × 1023 mol) and Cu(O2CMe)2?H2O (0.28 g, 1.41 × 1023 mol) were mixed at room temperature and allowed to stand for 1 h. The resultant deep brown microcrystals were filtered oV, washed with MeCN and EtO, and dried in vacuo. Yield 0.48 g, 82% (Found: C, 56.6; H, 5.0; N, 8.0.Calc. for C20H20CuN2O4?0.5H2O?0.5CH3CN: C, 56.6; H, 5.1; N, 7.9%). [N,N9-Bis(2,5-dihydroxybenzylidene)-1,2-diaminobenzenato]- copper(II) ([Cu(L3)]). Method A. As for [Cu(L1)], employing 1,2- diaminobenzene (0.078 g, 7.25 × 1024 mol). Yield 0.23 g, 78% (Found: C, 56.6; H, 4.2; N, 6.4. Calc. for C20H14CuN2O4? CH3OH: C, 57.0; H, 4.1; N, 6.3%). FAB mass spectrum: m/z 410, [63Cu(L3) 1 H]1; 409, [63Cu(L3)]1. UV/vis spectrum (dmf): nmax 15400 (sh), 21200 (emax = 16500), 24200 (sh), 28400 (sh), 29900 (19800), 31800 (22700 M21 cm21), 32500 (sh) and 36600 (sh) cm21.Method B. As for [Cu(L2)], using H2L3?CH3OH (0.53 g, 1.41 × 1023 mol). Yield 0.48 g, 82% (Found: C, 57.3; H, 3.7; N, 7.8. Calc. for C20H14CuN2O4?0.5H2O?0.5CH3CN: C, 57.4; H, 3.8; N, 8.0%). Other measurements Infrared spectra were obtained as Nujol mulls pressed between KBr windows at 400–4000 cm21 using a Perkin-Elmer Paragon 1000 spectrophotometer, UV/visible spectra with a Perkin- Elmer Lambda 12 spectrophotometer operating between 1100 and 200 nm in 1 cm quartz cells, positive ion fast atom bombardment mass spectra on a Kratos MS890 spectrometer, employing a 3-nitrobenzyl alcohol matrix, and NMR spectra on a Bruker DPX250 spectrometer, operating at 250.1 MHz (1H).The CHN microanalyses were performed by the University of Cambridge Department of Chemistry microanalytical service. Melting points are uncorrected. The EPR spectra were obtained using a Bruker ESP300E spectrometer; X-band spectra employed an ER4102ST resonator and ER4111VT2094 J.Chem. Soc., Dalton Trans., 1999, 2087–2095 cryostat, while for Q-band spectra an ER5106QT resonator and an ER4118VT cryostat were used. Spectral simulations were performed using in-house software which has been described elsewhere.44 ENDOR Spectra were recorded on a Bruker ESP300E X-band EPR spectrometer fitted with an ESP360 DICE ENDOR unit coupled to an EIN A-300 RF power amplifier operating between 300 and 500 W (8 dB).A Bruker EN 801 ENDOR cavity fitted with an Oxford Instruments ESR 900 liquid helium cryostat was used for the experiments, employing a modulation frequency of 12.5 kHz. Electrochemical measurements were carried out using an Autolab PGSTAT20 voltammetric analyser, in dmf containing 0.1 M NBun 4PF6 (prepared from aqueous NBun 4OH and HPF6, recrystallised twice from MeOH) as supporting electrolyte. Cyclic voltammetric experiments employed a double platinum working/counter electrode and a silver wire reference electrode; all potentials are referenced to an internal ferrocene– ferrocenium standard and were obtained at a scan rate of 100 mV s21.EHMO Calculations were carried out using the CACAO package,46 while INDO/S calculations were performed using the Argus package written by Thompson.47 Density functional calculations were performed using the DEFT code written by St-Amant48 in the linear combination of Gaussian-type orbitals framework.The calculations used the Vosko–Wilk– Nusair 49 local spin density approximation of the correlation part of the exchange-correlation potential with the Becke 50 non-local functional for exchange and the Perdew51 non-local functional for correlation. The all-electron treatment used Gaussian basis functions of double-zeta quality with contraction patterns (721/51/1*) for carbon, nitrogen and oxygen, (63321/531*/411) for copper and (41/1*) for hydrogen.All calculations employed the crystal structure coordinates for [Cu(MeOsalen)] 24 oriented with the global z axis lying perpendicular to the approximately planar CuO2N2 unit and the x axis bisecting the co-ordinated oxygen atoms. The symmetry of the first co-ordination sphere of the copper ion is approximately C2v. Acknowledgements The authors thank the Royal Society (London) for a University Research Fellowship to M. A. H., the Government of Singapore (L. M. L. C.), the EPSRC and the University of Cambridge for financial support.The authors would like to thank Dr D. M. Murphy (EPSRC ENDOR centre) for useful discussions, and Dr A. St-Amant (University of Ottawa) for making the DEFT software publically available. References 1 C. G. Pierpont and R. M. Buchanan, Coord. Chem. Rev., 1981, 38, 45; C. G. Pierpont and C. W. Lange, Prog. Inorg. Chem., 1994, 41, 331. 2 C. Floriani, G. Fachinetti and F. Calderazzo, J. Chem. Soc., Dalton Trans., 1973, 765; L.Kessel and D. N. Hendrickson, Inorg. Chem., 1978, 17, 2630; R. H. Heistand II, A. L. Roe and L. Que, Jr., Inorg. Chem., 1982, 21, 676. 3 M. Handa, H. Sono, K. Kasamatsu, K. Kasuga, M. Mikuriya and S. Ikenoue, Chem. Lett., 1992, 453; M. Handa, A. Takata, T. Nakao, K. Kasuga, M. Mikuriya and T. Kotera, Chem. Lett., 1992, 2085; M. Handa, H. Matsumoto, T. Namura, T. Nagaoka, K. Kasuga, M. Mikuriya, T. Kotera and R. Nukada, Chem. Lett., 1995, 903; M. Handa, M. Mikuriya, Y.Sato, T. Kotera, R. Nukada, D. Yoshioka and K. Kasuga, Bull. Chem. Soc. Jpn., 1996, 69, 3483; M. Handa, T. Nakao, M. Mikuriya, T. Kotera, R. Nukada and K. Kasuga, Inorg. Chem., 1998, 37, 149. 4 S. Ernst, P. Hänel, J. Jordanov, W. Kaim, V. Kasack and E. Roth, J. Am. Chem. Soc., 1989, 111, 1733. 5 O. J. Gelling, A. Meetsma and B. L. Feringa, Inorg. Chem., 1990, 29, 2816. 6 G. Asgedom, A. Sreedhara, E. Kolehmainen and C. P. Rao, J. Chem. Soc., Dalton Trans., 1996, 93. 7 M. Calvin and C.H. Barkelew, J. Am. Chem. Soc., 1946, 68, 2267; T. Tanaka, Bull. Chem. Soc. Jpn., 1960, 259; T. D. ShaVer and K. A. Sheth, Mol. Cryst. Liq. Cryst., 1989, 172, 27; S. E. Byström, E. M. Larsson and B. Åkermark, J. Org. Chem., 1990, 55, 5674. 8 R. A. Berthon, S. B. Colbran and D. C. Craig, Polyhedron, 1992, 11, 243; R. A. Berthon, S. B. Colbran and G. M. Moran, Inorg. Chim. Acta, 1993, 204, 3; S. B. Sembiring, S. B. Colbran and D. C. Craig, Inorg. Chem., 1995, 34, 761; S.B. Sembiring, S. B. Colbran, R. Bishop, D. C. Craig and A. D. Rae, Inorg. Chim. Acta, 1995, 228, 109; S. B. Sembiring, S. B. Colbran, D. C. Craig and M. L. Scudder, J. Chem. Soc., Dalton Trans., 1995, 3731. 9 C. D. Stevenson, R. C. Reiter, R. D. Burton and T. D. Halvorsen, Inorg. Chem., 1995, 34, 1368. 10 J. Li, V. Katti, A. A. Pinkerton, H. Nar and R. G. Cavell, Can. J. Chem., 1996, 74, 2378. 11 P. Cornago, C. Escolástico, M. D. Santa María, R. M. Claramunt, D. Carmona, M.Esteban, L. A. Oro, C. Foces-Foces, A. L. Llamas- Saiz and J. Elguero, J. Organomet. Chem., 1994, 467, 293. 12 T. E. Keyes, P. M. Jayaweera, J. J. McGarvey and J. G. Vos, J. Chem. Soc., Dalton Trans., 1997, 1627. 13 R. H. Bailes and M. Calvin, J. Am. Chem. Soc., 1947, 69, 1886. 14 R. S. Downing and F. L. Urbach, J. Am. Chem. Soc., 1969, 91, 5977. 15 T. N. Waters and D. Hall, J. Chem. Soc., 1959, 1200. 16 R. Atkins, G. Brewer, E. Kokot, G. M. Mockler and E. Sinn, Inorg. Chem., 1985, 24, 127. 17 D. F. Rohrbach, W. R. Heineman and E. Deutsch, Inorg. Chem., 1979, 18, 2536. 18 K. Bernardo, S. Leppard, A. Robert, G. Commenges, F. Dahan and B. Meunier, Inorg. Chem., 1996, 35, 387. 19 T. Tanaka, K. Tsurutani, A. Komatsu, T. Ito, K. Iida, Y. Fujii, Y. Nakano, Y. Usui, Y. Fukuda and M. Chikira, Bull. Chem. Soc. Jpn., 1997, 70, 615. 20 B. Bosnich, J. Am. Chem. Soc., 1968, 90, 627; A. C. Braithwaite, P. E. Wright and T. N. Waters, J. Inorg. Nucl. Chem., 1975, 37, 1669. 21 E. Suresh, M. M. Bhadbhade and D. Srinivas, Polyhedron, 1996, 15, 4133. 22 G. M. Larin, V. M. Dziomko, K. A. Dunaevskaya and Y. K. Syrkin, Zh. Strukt. Khim., 1965, 6, 391; G. M. Larin, G. V. Panova and E. G. Rukhadze, Zh. Strukt. Khim., 1965, 6, 699; Dokl. Akad. Nauk SSSR, 1966, 166, 363. 23 E. Hasty, T. J. Colburn and D. N. Hendrickson, Inorg. Chem., 1973, 12, 2414. 24 M. M. Bhadbhade and D. Srinivas, Inorg. Chem., 1993, 32, 6122. 25 T. Lund and W. E. Hatfield, J. Chem. Phys., 1973, 59, 885; A.D. Toy, M. D. Hobday, P. D. W. Boyd and T. D. Smith, J. Chem. Soc., Dalton Trans., 1973, 1259; M. Chikira, H. Yokoi and T. Isobe, Bull. Chem. Soc. Jpn., 1974, 47, 2208. 26 A. H. Maki and B. R. McGarvey, J. Chem. Phys., 1958, 29, 31, 35. 27 B. A. Goodman and J. B. Raynor, Adv. Inorg. Chem., 1970, 13, 135. 28 B. M. HoVman, J. Martinsen and R. A. Venters, J. Magn. Reson., 1984, 59, 110. 29 D. Kivelson and R. Neiman, J. Chem. Phys., 1961, 35, 149. 30 V. G. Swett and E. P. Dudek, J. Phys. Chem., 1968, 72, 1244; I. Adato, A. H. I. Ben-Bassat and S. Sarel, J. Phys. Chem., 1971, 75, 3828. 31 A. Abragam and M. H. L. Pryce, Proc. R. Soc. London, Ser. A, 1951, 206, 164. 32 J. Ferguson, J. Chem. Phys., 1961, 34, 2206; R. L. Belford and T. Piper, Mol. Phys., 1962, 5, 251. 33 C. D. Olson, G. Basu and R. L. Belford, J. Coord. Chem., 1971, 1, 176. 34 M. A. Hitchman, Inorg. Chem., 1977, 16, 1985. 35 A. G. Prudell and P. Kusch, Phys. Rev., 1952, 88, 184. 36 A. J. Bridgeman and M. Gerloch, Prog. Inorg. Chem., 1996, 45, 179. 37 M. J. Duer, N. D. Fenton and M. Gerloch, Int. Rev. Phys. Chem., 1990, 9, 227. 38 J. E. Huyett, S. B. Choudhury, D. M. Eichhorn, P. A. Bryngelson, M. J. Maroney and B. M. HoVman, Inorg. Chem., 1998, 37, 1361. 39 J. Q. Chambers, The Chemistry of the Quinoid Compounds, ed. S. Patai, Wiley, New York, 1974, ch. 14, pp. 737–791; J. Q. Chambers, The Chemistry of the Quinoid Compounds, vol. 2, eds. S. Patai and Z. Rappoport, Wiley, New York, 1988, ch. 12, pp. 719– 757. 40 B. R. Eggins and J. Q. Chambers, Chem. Commun., 1969, 232; V. D. Parker, Chem. Commun., 1969, 716; B. R. Eggins and J. Q. Chambers, J. Electrochem. Soc., 1970, 117, 186. 41 See e.g. R. Wang, T. E. Keyes, R. Hage, R. H. Schmehl and J. G. Vos, J. Chem. Soc., Chem. Commun., 1993, 1652. 42 M. Z. Barakat, S. K. Shebab and M. M. El-Sadr, J. Chem. Soc., 1958, 901; W. Schäfer and A. Aguado, Angew. Chem., Int. Ed. Engl., 1971, 10, 405; G. D. Storrier, S. B. Colbran and D. C. Craig, J. Chem. Soc., Dalton Trans., 1997, 3011.J. Chem. Soc., Dalton Trans., 1999, 2087–2095 2095 43 L. Papouchado, G. Petrie, J. H. Sharp and R. N. Adams, J. Am. Chem. Soc., 1968, 90, 5620. 44 F. E. Mabbs and D. Collison, Electron Paramagnetic Resonance of d Transition Metal Compounds, Elsevier, Amsterdam, 1992, ch. 7. 45 See e.g. J. A. Lanning and J. Q. Chambers, Anal. Chem., 1973, 45, 1010; N. Guptan and H. Lipschitz, J. Am. Chem. Soc., 1997, 119, 6384. 46 C. Mealli and D. M. Proserpio, J. Chem. Educ., 1990, 67, 399. 47 M. A. Thompson, Battelle Pacific Northwest Laboratories, 1992. 48 A. St-Amant, DEFT, A FORTRAN program, University of Ottawa, 1994. 49 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200. 50 A. D. Becke, Phys. Rev. A, 1988, 38, 3098. 51 J. P. Perdew, Phys. Rev. B, 1986, 33, 8822. Paper 9/02294E

ISSN:1477-9226    

DOI:10.1039/a902294e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 2091–2095 2091 Carbonyl ligand substitution of an aminerhenium benzyl complex. Interconversion of Á1- and Á3-benzyl complexes Tein-Fu Wang,*,† Chong-Chen Hwu, Chia-Wen Tsai and Yuh-Sheng Wen Institute of Chemistry, Academia Sinica, Taipei, Taiwan Removal of a CO ligand from the h1-benzyl complex [Re(CO)2(h1-CH2Ph){NH(Me)CH2CH2(h5-C5H4)}]1Br2 yielded an h3-benzyl complex. The h3 co-ordination mode was converted into h1 when the compound was dissolved in acetonitrile.Evaporation of acetonitrile removed the labile acetonitrile ligand and regenerated the h3-benzyl complex. The h3-benzyl complex [Re(CO)(h3-CH2Ph){NH(Me)CH2CH2(h5-C5H5)}]1BF4 2 is stable to air and moisture. The corresponding perrhenate salt has been characterized crystallographically. The tetrafluoroborate reacted with bromide, chloride and acetate anion to give the corresponding neutral h1-benzyl complexes in excellent yield. When treated with two-electron donor ligands, such as tert-butyl isocyanide and pyridine, it was converted into the corresponding (h1-benzyl)(isocyanide) and pyridine complexes.Amines are the classic ligands of co-ordination chemistry, but they have not often been used with organotransition-metal compounds.1 Unlike ligands containing the heavier Group 15 atoms, they show no p-acceptor ability. Therefore, they coordinate only weakly to the low-valent transition metals and form relatively labile complexes.2 However, through intramolecular chelation, stable amine complexes of manganese,3 molybdenum4 and rhenium5,6 have been prepared.Strong s donation of amine ligands to low-valent transition metals makes the metal centre much more electron-rich, thereby enhancing its ability to undergo nucleophilic reactions. We have reported such reactions of an aminerhenium complex with various electrophiles.6 Complex 1 reacts with electrophiles to give exclusively the alkyl complexes 2 (see Scheme 1).Benzyl complex 2a has been characterized by X-ray crystallography. The benzyl group and the amine ligand are mutually trans in a four-legged piano-stool structure. The carbonyl stretching bands of the aminerhenium complex 1 appear at 1893 and 1816 cm21. Upon alkylation at Re to give cationic 2a these bands shift to higher wavenumbers at 2034 and 1966 cm21, suggesting that the carbonyl ligands should be relatively labile. Ligand substitution of this type of labile carbonyl ligand is not uncommon.7 However, it is also of interest to see whether the h1-benzyl group forms the corresponding h3-benzyl complex when a carbonyl ligand is ejected.8 Here, we report the results of ligand substitution reactions of the benzyl complex 2a, and the formation of a stable h3-benzyl complex.Results and Discussion The carbonyl groups of complex 2a are tightly bonded to the metal and are not displaced by triphenylphosphine and isocyanides at ambient temperature through a dissociation– association mechanism.7 Extrusion of a carbonyl ligand by use of N-oxides was therefore investigated.9 Treatment of 2a with either Me3NO or PhIO in MeCN did cause extrusion of CO, only decomposition resulted.When 1 equivalent of the reagent was used similar decomposition was observed and some 2a was recovered. These results suggest that the expected acetonitrile complex reacts much faster with the oxidizing agent than does 2a.Nevertheless, when the reaction was carried out in acetone, the bromide 3 (66%) and the h3-benzyl complex 4 (5%) were obtained (see Scheme 2). The latter precipitated from the reaction mixture. † E-Mail: tfwang@chem.sinica.edu.tw The characterization of bromide 3 is based on spectroscopy. In the carbonyl stretching region of the infrared spectrum only one absorption at 1877 cm21 was observed, indicating a terminal CO ligand. In the 1H NMR spectrum the N-methyl group appeared at d 3.12 as a doublet, suggesting that the Re]N bond is intact.The doublet signal of the methyl group, arising from the coupling with the N]H proton, is a general feature for amines ligated to rhenium.6 The diastereotopic benzylic protons appeared at d 4.38 and 3.98. The four inequivalent C5H4 protons were observed at d 5.79, 5.48, 4.91 and 3.47. These spectroscopic data and the presence of a molecular ion peak in the mass spectrum suggest the formulation shown. The relative stereochemistry of the N-methyl and the CO ligand is not defined.The terminal carbonyl stretching of complex 4 appeared at 1927 cm21. The phenyl protons of 4 give rise to five distinctive signals in the 1H NMR spectrum, suggesting that the benzene ring is not freely rotating. Definitive structural characterization of 4 was carried out by a single-crystal X-ray analysis. Fig. 1 shows that the benzyl group binds to the rhenium in an h3 fashion with exo orientation. Bond distances (Table 1) from rhenium to the h3-carbons are 2.197 Å for Re]CH2, 2.333 Å for Re]Cipso and 2.535 Å for Re]Cortho.The latter is 0.338 Å longer than the Re]CH2 bond length, compatible with those reported for h3-benzyl complexes.11 Being co-ordinated to the rhenium, Cipso]Cortho [C(11)]C(16)] is lengthened from the average value of 1.38 Å of benzene to 1.426 Å. Other carbon–carbon bonds of the six-membered ring show alternation in lengths, consistent with some localization of single and double bonds.Average distances are 1.426 Å for the ‘single bonds’ C(11)]C(12), C(13)]C(14) and C(15)]C(16) and 1.35 Å for the ‘double bonds’ C(12)]C(13) and C(14)]C(15). The N-methyl group and the CO ligand are mutually syn. The perrhenate anion of the h3-benzyl complex 4 is apparently derived from over oxidation of some rhenium complexes by Me3NO. The h3-benzyl complex 4 could be converted into the bromide 3 by excess of LiBr. Abstraction of the bromide anion from 3 with silver tetrafluoroborate in acetonitrile gave Scheme 1 Re OC N CO Me H Re N CO R OC Me H a: RX = PhCH2Br b: RX = MeI c: RX = d: RX = HC CCH2Br e: RX = MeO2CCH2Br f: RX = EtO2CCH2Br 2 + RX X– 1 CH2 CH2Br2092 J.Chem. Soc., Dalton Trans., 1998, Pages 2091–2095 Scheme 2 (i) Me3NO, acetone, room temperature (r.t.); (ii) LiBr, MeOH; (iii) AgBF4, MeCN, 0 8C; (iv) KReO4, acetone; (v) CD3CN; (vi) LiCl, MeOH; (vii) NaO2CMe, MeOH; (viii) ButNC, CH2Cl2; (ix) pyridine, CH2Cl2 Re N OC Me H Re N CO OC Me H Ph Re N Br OC Me H Ph Re N NCCD3 OC Me H Ph Re N Cl OC Me H Ph Re N O2CMe OC Me H Ph Re N CNBut OC Me H Ph Re N N OC Me H Ph Re N OC Me H 2a 3 + 8 _ 9 _ BF4 _ + Br _ + 10 Br BF4 7 + 4 ReO4 _ 6 + BF4 _ + 5 + i ii iii iv v vi ii vii viii ix the corresponding acetonitrile complex, which showed a CO stretching band at 1908 cm21 in the infrared spectrum.However, upon removal of acetonitrile and dissolving the residue in CH2Cl2, the CO band shifts to 1927 cm21, which is the same frequency as that for the h3-benzyl complex 4.The 1H NMR spectrum also shows similarity with that of 4. Therefore, the final product is believed to be the h3-benzyl complex 5. Anion exchange of 5 with potassium perrhenate gave 4, thus confirming the assignment. The ipso- and the co-ordinated orthocarbons of the benzene ring of 5 resonate at d 105.1 and 92.1 respectively in the 13C NMR spectrum, consistent with the h3 bonding mode. Solid 5 is stable at room temperature in air. In solution [(CD3)2CO] there is no visible decomposition and the 1H NMR spectrum remained unchanged after the solution was kept at room temperature in air overnight. In the presence of acetonitrile the h3-benzyl complex 5 is in equilibrium with the corresponding (h1-benzyl)(acetonitrile) Fig. 1 An ORTEP10 drawing of [Re(CO)(h3-CH2Ph){NH(Me)- CH2CH2(h5-C5H4)}]1ReO4 2 4; ReO4 2 is omitted for simplicity complex 6. The 1H NMR spectrum of 5 in CD3CN showed that 5 and 6 reach equilibrium in a ratio of 1 : 6 in less than 5 min.Evaporation of CD3CN and dissolving the residue in (CD3)2CO regenerates 5 completely, as shown by the 1H NMR spectrum. The ease of formation of the h3-benzyl complex 5 over the acetonitrile complex 6 is interesting, because in the h3-binding mode the aromaticity of the six-membered ring is destroyed. Although the h3-benzyl complex 5 is much more stable than the corresponding (h1-benzyl)(acetonitrile) complex, addition of ligands to the former proceeded smoothly to give the corresponding h1-benzyl complexes.12 Treatment of 5 with an excess of lithium bromide provided bromide 3 in quantitative yield.Similarly, 5 reacts with an excess of lithium chloride and sodium acetate, to provide chloride 7 and acetate 8 respectively. Characterization of 7 and 8 was based upon their spectral data and elemental analyses. Complex 5 also reacts with two-electron donor ligands, such as tert-butyl isocyanide and pyridine, to give the cationic complexes 9 and 10 respectively.Isocyanide complex 9 shows characteristic isocyanide and carbonyl stretchings at 2176 and 1927 cm21 respectively. The terminal CO stretching of the pyridine complex 10 appeared at 1893 cm21 in the infrared spectrum. The pyridine is co- Table 1 Selected bond lengths (Å) and angles (8) for complex 4 Re(1)]Cp* Re(1)]N Re(1)]C(9) Re(1)]C(10) Re(1)]C(11) Re(1)]C(16) C(10)]C(11) Cp]Re(1)]N Cp]Re(1)]C(9) Cp]Re(1)]C(10) Cp]Re(1)]C(11) Cp]Re(1)]C(16) Cp]C(3)]C(2) N]Re(1)]C(9) N]Re(1)]C(10) N]Re(1)]C(11) 1.918(1) 2.189(8) 1.866(11) 2.197(10) 2.333(10) 2.535(10) 1.454(15) 107.2(2) 123.3(3) 115.9(3) 138.9(3) 116.0(2) 169.8(9) 90.4(4) 132.5(3) 96.0(3) C(11)]C(12) C(11)]C(16) C(12)]C(13) C(13)]C(14) C(14)]C(15) C(15)]C(16) N]Re(1)]C(16) Re(1)]C(9)]O(9) C(9)]Re(1)]C(10) C(9)]Re(1)]C(11) C(9)]Re(1)]C(16) C(10)]Re(1)]C(11) C(10)]Re(1)]C(16) C(11)]Re(1)]C(16) C(10)]C(11)]C(16) 1.443(15) 1.426(13) 1.339(18) 1.428(17) 1.369(17) 1.417(16) 82.1(3) 173.3(9) 82.5(4) 89.1(4) 119.7(4) 37.3(4) 62.0(3) 33.7(3) 116.9(3) *Cp means the centre of the C5H4 ring.J.Chem. Soc., Dalton Trans., 1998, Pages 2091–2095 2093 ordinated in an h1-N mode, as suggested by the 1H and 13C NMR spectra. In conclusion, a stable h3-benzyl complex has been prepared from the dicarbonyl 2a in two steps. In acetonitrile solution, the h3-benzyl 5 and the h1-benzyl 6 are both observed, in a ratio of 1 : 6.Upon removal of acetonitrile the h3-benzyl 5 is regenerated. Preference for the h3-benzyl 5 over the h1-benzyl acetonitrile complex 6 has been testified by adding 1 equivalent of acetonitrile to a (CD3)2CO solution of 5. No resonances of the corresponding h1-benzyl acetonitrile complex were observed by 1H NMR spectroscopy. The thermodynamic stability of 5 is noteworthy, because h3 co-ordination of the benzyl group requires loss of aromaticity. Experimental Infrared solution spectra were recorded on a Perkin-Elmer 882 spectrophotometer using 0.1 mm cells with CaF2 windows.Melting points were determined by using a Yanaco model MP micro melting point apparatus and were uncorrected. Proton (300 MHz) and 13C NMR (75 MHz) spectra were obtained with a Bruker AC-300 FT spectrophotometer. For the assignments the carbon bound to the nitrogen is designated as C1 and the hydrogens on it as H1a and H1b. The next carbon is designated as C2 and the hydrogens on it as H2a and H2b.All chemical shifts are expressed in d relative to SiMe4 (d 0.0). Elemental analyses were obtained on a Perkin-Elmer 2400 CHN elemental analyzer. Mass spectra were recorded on a VG 70-250S mass spectrometer. Syntheses [Re(CO)(Á1-CH2Ph)Br{NH(Me)CH2CH2(Á5-C5H4)}] 3. To a stirred solution of complex 2a 6 (560 mg, 1.04 mmol) in acetone (30 cm3) was added trimethylamine N-oxide dihydrate (134 mg, 1.2 mmol) in one portion at room temperature. The progress of the reaction was monitored by infrared spectroscopy. After 30–60 min the CO bands associated with 2a disappeared.The solvent was evaporated. To the residue was added CH2Cl2 (3 cm3). The resulting mixture was flashed through a column packed with silica gel (10 g), using ethyl acetate as eluent. The yellow band was collected and concentrated. The residue was dissolved in CHCl3 (30 cm3). The resultant yellow solution was allowed to stand in a refrigerator at about 5 8C for 16 h.Some crystalline compound was collected and washed twice with CHCl3 to give 4 (35 mg, 5% yield). The combined CHCl3 solution was treated with LiBr (600 mg) at room temperature for 1 h. The resulting mixture was filtered through Celite. Filtrates were concentrated and dissolved in CH2Cl2 (5 cm3). The CH2Cl2 solution was added to hexane (50 cm3). The resultant orange powder was collected and washed twice with hexane to give 3 (350 mg, 66% yield). M.p. 123–129 8C (decomp.).IR (CH2Cl2): 1877s cm21. 1H NMR (CDCl3, 300 MHz): d 7.19– 7.11 (4 H, m, Ph), 6.94–6.88 (1 H, m, Ph), 5.80–5.78 (1 H, m, C5H4), 5.49–5.47 (1 H, m, C5H4), 4.92–4.90 (1 H, m, C5H4), 4.74 (1 H, br, NH), 4.38 (1 H, d, J = 10.3, benzylic Ha), 3.98 (1 H, d, J = 10.3, benzylic Hb), 3.83–3.74 (1 H, m, H1a), 3.48–3.46 (1 H, m, C5H4), 3.34–3.22 (1 H, m, H1b), 3.12 (3 H, d, J = 6.1, NCH3), 2.23 (1 H, ddd, J = 14.4, 10.3, 5.7, H2a) and 2.07 (1 H, dt, J = 14.4, 4.9 Hz, H2b). 13C NMR (CDCl3, 75 MHz): d 220.6 (CO), 152.4 (C, Ph), 128.0 (CH × 2, Ph), 127.5 (CH × 2, Ph), 123.3 (CH, Ph), 123.2 (C, C5H4), 99.5 (CH, C5H4), 97.9 (CH, C5H4), 78.9 (CH, C5H4), 73.1 (CH, C5H4), 69.2 (CH2, C1), 47.3 (NCH3), 26.6 (CH2, C2) and 7.1 (CH2, benzylic CH2). Mass spectrum (FAB, 187Re81Br): m/z [relative intensity (%)] 509 (M1, 5), 428 (M1 2 Br, 100) (Found: C, 37.98; H, 3.69; N, 2.52. C16H19BrNORe requires C, 37.87; H, 3.77; N, 2.76%). [Re(CO)(h3-CH2Ph){NH(Me)CH2CH2(h5-C5H4)}]1ReO4 2 4. Orange crystals.M.p. 199–203 8C (decomp.). IR (CH2Cl2): 1927s cm21. 1H NMR (CDCl3, 300 MHz): d 8.19 (1 H, t, J = 7.6), 7.85 (1 H, d, J = 8.6), 7.43 (1 H, t, J = 7.5), 7.18 (1 H, t, J = 8.2), 7.11 (1 H, d, J = 6.4), 6.65–6.63 (1 H, m, C5H4), 6.00– 5.98 (1 H, m, C5H4), 5.28–5.26 (1 H, m, C5H4), 4.93–4.91 (1 H, m, C5H4), 3.87 (1 H, d, J = 3.7, H10a), 3.69–3.60 (1 H, m, H1a), 3.16 (1 H, d, J = 3.7, H10b), 2.97–2.85 (1 H, m, H1b), 2.77–2.67 (1 H, m, H2a), 2.17–2.10 (1 H, m, H2b) and 2.05 (3 H, d, J = 5.8 Hz, NCH3).Mass spectrum (FAB, 187Re): m/z [relative intensity (%)] 428 (M1, 100) (Found: C, 28.12; H, 2.95; N, 2.35. C16H19NO5Re2 requires C, 28.36; H, 2.82; N, 2.07%). [Re(CO)(Á3-CH2Ph){NH(Me)CH2CH2(Á5-C5H4)}]1BF4 2 5. To a stirred solution of complex 3 (467 mg, 0.92 mmol) in MeCN (40 cm3) at 0 8C was added a MeCN solution of AgBF4 (0.25 M, 3.7 cm3, 0.92 mmol) over 3 min. After stirring for 10 min, the reaction was complete as indicated by the infrared spectrum (1908 cm21).White precipitates were removed by filtering through Celite. The residue after concentration was dissolved in CH2Cl2 and filtered again through Celite. After removal of solvent 5 was obtained as an orange powder (471 mg, 99% yield). An analytically pure sample was obtained by recrystallization from acetone and hexane. M.p. 198–203 8C (decomp.). IR (CH2Cl2): 1926s cm21. 1H NMR [(CD3)2CO, 300 MHz]: d 8.09 (1 H, t, J = 7.5, H14), 7.97 (1 H, d, J = 8.4, H12), 7.27 (1 H, d, J = 5.6, H16), 7.18–7.13 (2 H, m, H13 and H15), 6.80–6.78 (1 H, m, C5H4), 6.48–6.46 (1 H, m, C5H4), 5.54–5.52 (1 H, m, C5H4), 5.36–5.34 (1 H, m, C5H4), 3.81 (1 H, d, J = 3.3, H10a), 3.63–3.55 (1 H, m, H1a), 3.40–3.30 (1 H, m, H1b), 3.36 (1 H, d, J = 3.3 Hz, H10b), 2.50–2.36 (2 H, m, H2) and 2.31 (3 H, br s, NCH3). 13C NMR [(CD3)2CO, 75 MHz]: d 204.0 (CO), 138.7 (CH, C12), 134.5 (CH, C14), 132.6 (CH, C13 or C15), 129.9 (CH, C15 or C13), 126.2 (C, C5H4), 105.1 (C, C11), 97.9 (CH, C5H4), 94.7 (CH, C5H4), 92.1 (CH, C16), 79.3 (CH, C5H4), 74.6 (CH, C5H4), 73.6 (CH2, C1), 50.1 (NCH3), 26.1 (CH2, C2) and 21.1 (CH2, C10).Mass spectrum (FAB, 187Re): m/z [relative intensity (%)] 428 (M1, 100) (Found: C, 37.10; H, 3.68; N, 2.57. C16H19BF4NORe requires C, 37.36; H, 3.72; N, 2.72%). Interconversion of complexes 5 and 6. A 5 mm NMR tube was charged with complex 5 (3.1 mg), followed by CD3CN (0.5 cm3). The resultant yellow solution was examined by NMR spectroscopy immediately.The 1H NMR spectrum showed two complexes, 5 and 6, in a ratio of 1 : 6. The ratio remained steady during a period of 24 h. The major resonances of 5 and 6 were well separated. Partial assignments: 5, d 8.09–8.04 (1 H, m), 7.92 (1 H, d, J = 8.7), 6.48–6.46 (1 H, m, C5H4), 6.20–6.18 (1 H, m, C5H4), 5.40–5.38 (1 H, m, C5H4) and 5.09–5.07 (1 H, m, C5H4); 6, d 7.21–7.12 (3 H, m, Ph), 7.02–6.95 (2 H, m, Ph), 6.11–6.09 (1 H, m, C5H4), 5.74–5.72 (1 H, m, C5H4), 5.01–4.99 (2 H, m, C5H4), 3.64–3.57 (1 H, m), 3.36–3.27 (1 H, m), 2.96 (3 H, d, J = 5.9 Hz, NCH3) and 2.30–2.12 (2 H, m).General procedure for [Re(CO)(Á1-CH2Ph)Cl{NH(Me)CH2- CH2(Á5-C5H4)}] 7 and [Re(CO)(Á1-CH2Ph)(MeCO2){NH(Me)- CH2CH2(Á5C5H4)}] 8. To a stirred solution of complex 5 (120 mg, 0.23 mmol) in methanol (5 cm3) was added an excess of LiCl (200 mg) for the preparation of 7 or sodium acetate (300 mg) for 8. The mixture was stirred at room temperature and the reaction monitored by IR spectroscopy.After reaction was complete (about 30 min) the solvent was evaporated. The product was taken up in CHCl3 and filtered through Celite. Addition of the CHCl3 solution to hexane gave a yellow precipitate, which was collected and washed with hexane. Analytically pure samples were obtained by recrystallization of the compounds from CH2Cl2 and hexane. Complex 7. Yellow powder (107 mg, 99% yield). M.p. 181– 183 8C. IR (CH2Cl2): 1872s cm21. 1H NMR (CDCl3, 300 MHz): d 7.23–7.13 (4 H, m, Ph), 6.93 (1 H, t, J = 7, Ph), 5.73–5.71 (1 H, m, C5H4), 5.49–5.47 (1 H, m, C5H4), 4.98–4.96 (1 H, m, C5H4), 4.63 (1 H, br, NH), 4.16 (1 H, d, J = 10.3, benzylic Ha), 3.91 (1 H, d, J = 10.3, benzylic Hb), 3.81–3.73 (1 H, m, H1a), 3.56–2094 J. Chem. Soc., Dalton Trans., 1998, Pages 2091–2095 3.53 (1 H, m, C5H4), 3.37–3.24 (1 H, m, H1b), 3.12 (3 H, d, J = 6.1 Hz, NCH3), 2.25 (1 H, ddd, J = 14.4, 10.3, 5.7, H2a) and 2.12 (1 H, dt, J = 14.4, 4.9 Hz, H2b). 13C NMR (CDCl3, 75 MHz): d 220.5 (CO), 152.3 (C, Ph), 128.0 (CH × 2, Ph), 127.6 (CH × 2, Ph), 123.3 (CH, Ph), 123.3 (C, C5H4), 99.8 (CH, C5H4), 98.8 (CH, C5H4), 78.9 (CH, C5H4), 73.3 (CH, C5H4), 68.8 (CH2, C1), 47.1 (NCH3), 26.6 (CH2, C2) and 10.6 (CH2, benzylic CH2). Mass spectrum (FAB, 187Re35Cl): m/z [relative intensity (%)] 463 (M1, 35), 429 (M1 1 1 2 Cl, 100) (Found: C, 41.31; H, 3.97; N, 2.87. C16H19ClNORe requires C, 41.51; H, 4.14; N, 3.02%). Complex 8.Hygroscopic yellow powder (104 mg, 92% yield). IR (CH2Cl2): 1865s and 1595w cm21. 1H NMR (CDCl3, 300 MHz): d 7.21–7.12 (4 H, m, Ph), 6.95–6.90 (1 H, m, Ph), 6.67 (1 H, br, NH), 6.11–6.09 (1 H, m, C5H4), 5.36–5.34 (1 H, m, C5H4), 4.89–4.87 (1 H, m, C5H4), 3.72 (1 H, d, J = 10.1, benzylic Ha), 3.62 (1 H, d, J = 10.1, benzylic Hb), 3.53–3.47 (1 H, m, H1a), 3.46–3.44 (1 H, m, C5H4), 3.11–3.00 (1 H, m, H1b), 2.95 (3 H, d, J = 6.0, NCH3), 2.47 (1 H, ddd, J = 14.0, 12.4, 5.6, H2a), 2.20 (3 H, s, O2CCH3) and 1.98 (1 H, ddd, J = 14.0, 5.4, 2.0 Hz, H2b). 13C NMR (CDCl3, 75 MHz): d 215.7 (CO), 180.3 (C, O2CCH3), 152.0 (C, Ph), 127.8 (CH × 2, Ph), 127.5 (CH × 2, Ph), 123.0 (CH, Ph), 122.9 (C, C5H4), 99.6 (CH, C5H4), 96.8 (CH, C5H4), 78.7 (CH, C5H4), 73.1 (CH, C5H4), 67.5 (CH2, C1), 46.0 (NCH3), 26.3 (CH2, C2), 24.2 (CH3, O2CCH3) and 17.4 (CH2, benzylic CH2). Mass spectrum (FAB, 187Re): m/z [relative intensity (%)] 429 (M1 1 1 2 O2CCH3, 100) (Found: C, 44.62; H, 4.74; N, 2.59.C18H22NO3Re requires C, 44.43; H, 4.56; N, 2.88%). General procedure for [Re(CO)(Á1-CH2Ph)(ButNC){NH(Me)- CH2CH2(Á5-C5H4)}]1BF4 2 9 and [Re(CO)(Á1-CH2Ph)(C5H5N)- {NH(Me)CH2CH2(Á5-C5H4)}]1BF4 2 10. To a stirred solution of complex 5 (120 mg, 0.23 mmol) in CH2Cl2 (15 cm3) was added ButNC (0.15 cm3) for the preparation of 9 or pyridine (0.5 cm3) for 10. The solution was stirred at room temperature and the reaction monitored by IR spectroscopy. After reaction was complete (0.5–4 h) the solvents were evaporated.The residue was dissolved in CH2Cl2 (3 cm3) and added to diethyl ether (30 cm3). Yellow precipitates were collected and washed twice with ether to give 9 and 10 respectively. Analytically pure samples were obtained by recrystallization of the compounds from CH2Cl2 and hexane. Complex 9. Yellow powder (135 mg, 97% yield). M.p. 139– 141 8C. IR (CH2Cl2): 2176s and 1927s cm21. 1H NMR (CDCl3, 300 MHz): d 7.21–7.15 (2 H, m, Ph), 7.05–6.97 (3 H, m, Ph), 6.26–6.24 (1 H, m, C5H4), 5.98 (1 H, br, NH), 5.81–5.79 (1 H, m, C5H4), 4.59–4.57 (1 H, m, C5H4), 3.94–3.88 (1 H, m, H1a), 3.51–3.49 (1 H, m, C5H4), 3.45 (1 H, d, J = 10.5, benzylic Ha), 3.19 (1 H, d, J = 10.5, benzylic Hb), 3.15–3.00 (1 H, m, H1b), 2.96 (3 H, d, J = 5.8, NCH3), 2.67 (1 H, td, J = 14.0, 5.7, H2a), 1.99 (1 H, dd, J = 14.0, 5.4 Hz, H2b) and 1.67 (9 H, s, But). 13C NMR (CDCl3, 75 MHz): d 206.2 (CO), 150.1 (C, Ph), 136.9 (C, isocyanide), 127.8 (CH × 2, Ph), 127.4 (CH × 2, Ph), 127.1 (C, C5H4), 124.3 (CH, Ph), 91.6 (CH, C5H4), 90.4 (CH, C5H4), 86.7 (CH, C5H4), 83.4 (CH, C5H4), 74.4 (CH2, C1), 59.5 (C, But), 48.9 (CH3, NCH3), 30.1 (CH3 × 3, But), 25.2 (CH2, C2) and 22.1 (CH2, benzylic CH2) (Found: C, 42.07; H, 4.49; N, 4.43.C21H28BF4N2ORe requires C, 42.22; H, 4.72; N, 4.69%). Complex 10. Yellow powder (133 mg, 96% yield). M.p. 106– 110 8C (decomp.). IR (CH2Cl2): 1893s cm21. 1 H NMR (CDCl3, 300 MHz): d 9.19 (1 H, br, pyridine), 8.87 (1 H, br, pyridine), 8.01 (1 H, tt, J = 7.7, 1.5, pyridine), 7.67 (1 H, br, pyridine), 7.53 (1 H, br, pyridine), 7.18–7.12 (2 H, m, Ph), 7.01–6.95 (3 H, m, Ph), 6.27–6.25 (1 H, m, C5H4), 6.56–6.54 (1 H, m, C5H4), 5.06 (1 H, br, NH), 5.02–5.00 (1 H, m, C5H4), 3.88–3.81 (1 H, m, H1a), 3.52–3.60 (1 H, m, C5H4), 3.44 (1 H, d, J = 10.0, benzylic Ha), 3.31 (1 H, d, J = 10.0, benzylic Hb), 3.27–3.13 (1 H, m, H1b), 2.93 (1 H, td, J = 13.7, 5.6, H2a), 2.76 (3 H, d, J = 5.7, NCH3) and 2.08 (1 H, dd, J = 13.7, 4.8 Hz, H2b). 13C NMR (CDCl3, 75 MHz): d 208.4 (CO), 160.5 (CH, pyridine), 157.6 (pyridine), 150.3 (C, Ph), 139.1 (CH × 2, pyridine), 128.3 (CH, pyridine), 128.2 (C, C5H4), 127.8 (CH × 2, Ph), 127.7 (CH × 2, Ph), 124.1 (CH, Ph), 98.9 (CH, C5H4), 98.3 (CH, C5H4), 81.7 (CH, C5H4), 75.8 (CH, C5H4), 71.8 (CH2, C1), 47.2 (CH3, NCH3), 25.5 (CH2, C2) and 15.9 (CH2, benzylic CH2). Mass spectrum (FAB, 187Re): m/z [relative intensity (%)] 428 (M1 2 pyridine, 100) (Found: C, 42.67; H, 3.90; N, 4.43.C21H24BF4N2ORe requires C, 42.50; H, 4.08; N, 4.72%). Crystallography A single crystal of complex 4 was obtained by slow diVusion of a CH2Cl2 solution of 4 into hexane at 25 8C. DiVraction measurement was made on an Enraf-Nonius CAD-4 automated diVractometer by use of graphite-monochromated Mo-Ka radiation (l = 0.710 69 Å) with the q–2q scan mode at 25 8C. The unit cell was determined and refined using 25 randomly selected reflections obtained with the automatic search, centre, index, and least-squares routines.Lorentzpolarization and empirical absorption corrections based on three azimuthal scans were applied to the data. The space group (P21/n) was determined from the systematic absences observed during data collection. All data reduction and refinements were carried out on a DecAlpha 3400/400 computer using the NRCVAX program.13 The structure was solved by direct methods and refined by a full-matrix least-squares routine 14 with anisotropic thermal parameters for all non-hydrogen atoms.The structure was refined by minimizing Sw|Fo 2 Fc|2, where w = (1/s2)Fo was calculated from the counting statistics. Hydrogen atoms were included in the structure factor calculations in their expected positions based on idealized bonding geometry but not refined. The final cell parameters and data collection parameters are listed in Table 2. CCDC reference number 186/975. Acknowledgements We are grateful to the National Science Council of Taiwan, ROC for financial support.Table 2 Crystallographic data and structure refinements for complex 4 Formula M Crystal size/mm Crystal system Space group a/Å b/Å c/Å b/8 U/Å3 Z Dc/g cm23 F(000) m(Mo-Ka)/cm21 Transmission Scan speed/8 min21 q–2q scan width/8 2qmax/8 h, k, l Ranges Collected reflections Unique reflections Observed reflections [I > 2.0s(I)] Refined parameters R, R9 Goodness of fit Weight modifier K in KFo 2 (Dr)max,min/e Å23 C16H19NO5Re2 677.74 0.19 × 0.28 × 0.50 Monoclinic P21/n 7.842(2) 22.687(3) 10.110(3) 108.67(3) 1703.9(7) 4 2.642 1248 144.180 0.329, 1.000 1.37–8.24 2(0.65 1 0.35 tan q) 50.0 29 to 8, 0–26, 0–12 3161 2988 2498 218 0.033, 0.039 1.95 0.000 100 2.740, 21.630 R = S(Fo 2 Fc)/S(Fo), R9 = [Sw(Fo 2 Fc)2/SwFo 2]� �� .J.Chem. Soc., Dalton Trans., 1998, Pages 2091–2095 2095 References 1 J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, 2nd edn., University Science Books, Mill Valley, CA, 1987, p. 64; F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, 5th edn., Wiley, New York, 1988, p.36. 2 P. Jutzi, J. Dahlhaus and M. O. Kristen, J. Organomet. Chem., 1993, 450, C1–C3; D. Sellmann and J. Mueller, J. Organomet. Chem., 1985, 281, 249; D. Sellmann, J. Mueller and P. Hofmann, Angew. Chem., Int. Ed. Engl., 1982, 21, 691. 3 T. F. Wang, T. Y. Lee, Y. S. Wen and L. K. Liu, J.Organomet. Chem., 1991, 403, 353. 4 T. F. Wang and Y. S. Wen, J. Organomet. Chem., 1992, 439, 155; T. F. Wang, T. Y. Lee, J. W. Chou and C. W. Ong., J. Organomet. Chem., 1992, 423, 31. 5 T. F. Wang, J. P. Juang and Y. S. Wen, Bull. Inst. Chem., Acad. Sin., 1995, 42, 41. 6 T. F. Wang, C. Y. Lai, C. C. Hwu and Y. S. Wen, Organometallics, 1997, 16, 1218. 7 J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, 2nd edn., University Science Books, Mill Valley, CA, 1987, p. 247. 8 M. Brookhart, R. C. Buck and E. Danielson, III, J. Am. Chem. Soc., 1989, 111, 567. 9 M. O. Albers and N. J. Coville, Coord. Chem. Rev., 1984, 53, 227. 10 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 11 J. R. Bleeke, R. R. Burch, C. L. Coulman and B. C. Schardt, Inorg. Chem., 1981, 20, 1316; R. R. Burch, E. L. Muetterties and V. W. Day, Organometallics, 1982, 1, 188; L. E. Crascall, S. A. Litster, A. D. Redhouse and J. L. Spencer, J. Organomet. Chem., 1990, 394, C35; H. Adams, N. A. Bailey, M. J. Winter and S. Woodward, J. Organomet. Chem., 1991, 418, C39; M. A. Bennett, L. Y. Goh, I. J. McMahon, T. R. B. Mitchell, G. B. Robertson, T. W. Turney and W. A. Wickramasinghe, Organometallics, 1992, 11, 3069; M. D. Fryzuk, D. H. McConville and S. J. Rettig, J. Organomet. Chem., 1993, 445, 245; H. Wadepohl, G. P. Elliott, H. Pritzkow, F. G. A. Stone and A. Wolf, J. Organomet. Chem., 1994, 482, 243. 12 J. S. Roberts and K. J. Klabunde, J. Am. Chem. Soc., 1977, 99, 2509. 13 E. J. Gabe, Y. LePage, J. P. Charland, F. L. Lee and P. S. White, J. Appl. Crystallogr., 1989, 22, 384. 14 P. Main, in Crystallographic Computing 3: Data Collection, Structure Determination, Proteins and Databases, eds. G. M. Sheldrick, C. Krueger and R. Goddard, Clarendon Press, Oxford, 1985, pp. 206–215. Received 9th February, 1998; Paper 8/

ISSN:1477-9226    

DOI:10.1039/a801151f

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 2097–2102 2097 Conformational choice in disilver cryptates; an 1H NMR and structural study Oliver W. Howarth,a Grace G. Morgan,b,c Vickie McKee *b and Jane Nelson *b,c a Chemistry Department, University of Warwick, Coventry, UK CV4 7AL b School of Chemistry, Queens University, Belfast, UK BT9 5AG c Chemistry Department, Open University, Milton Keynes, UK MK7 6AA Received 8th February 1999, Accepted 29th April 1999 Disilver complexes of three iminocryptands, with either tris(aminopropylamine)- or tris(aminoethylamine)-derived caps {N[(CH2)nNCHRCHN(CH2)n]3N (n = 2 or 3, R = 1,3-(CH2)2C6H4 or 2,5-(CH2)2C4H4O)} have been structurally characterised, and show relatively close approach of the silver(I) ions (3.08–3.77 Å). DiVerences in the geometry of the coordination site adopted can be related to the flexibility of the cryptand host.This is also important in solution, where the smaller cap-size cryptates are restricted to one conformation while the larger cap-size cryptates appear ready to adopt a range of conformations.The coordinative plasticity of Ag(I) is of considerable advantage in its role as a templating cation in the assembly of macrocyclic or cryptand ligands. Although the textbook view of Ag(I) is that its preferred mode of coordination is linear twocoordinate, this is not a mode that is commonly found in macrocyclic or cryptate complexes 1–3. Thus in macrocyclic silver complexes, coordination modes from near square planar to distorted tetrahedral are often found,1 and higher coordination modes such as pentagonal pyramidal or trigonal bipyramidal occasionally appear.2 In azacryptates where we have found silver(I) a valuable templating cation, the trigonal pyramidal four-coordinate mode is predominant,3 but not by any means exclusive.Where steric factors operate, the silver cation shows itself readily satisfied with a coordination number less than four; 4–7 distorted trigonal planar and non-linear two-coordinate geometries have also been observed.The energetic similarity of the alternative coordination environments is well-illustrated in the crystallography of the disilver cryptate of the propylenecapped ligand L6,6 where minor and major components of the disorder use respectively: (a) a pair of trigonal pyramidal inclusive sites for both Ag(I) ions, and (b) an unsymmetric conformation where one Ag(I) has this inclusive four-coordinate site while the other occupies a facial three-coordinate site.These Ag(I) cations are separated by a relatively long distance (7.09 or 7.44 Å) where there can be no possibility of a stabilising interaction between the silver ions. In solution, 1H NMR spectra of [Ag2L6]21 show fluxionality 8 which may arise, among other things, from interconversion of the conformers in solution. In the disilver cryptate of an analogous thiophene-linked tris(propylene) (trpn)-capped ligand,9 1H NMR studies have shown that two forms are present in solution.The major conformer derives from an unsymmetric pair of coordination sites, despite X-ray crystallographic structure determination which shows the form isolated in the solid state to be symmetric. Once more the separation of coordination sites is large enough to dismiss any chance that interactions between silver(I) cations contribute to the stabilization of particular conformers. Such interactions may exist, however, in disilver cryptates of the small hosts imBT (1,4,7,10,13,16,21,24-octaazabicyclo[8.8.8]- hexacosa-4,6,13,15,21,23-hexaene) and amBT (= imBT 1 12 H) where Ag ? ? ? Ag distances of the order of 2.8 Å are observed 4,5 and in other disilver complexes with even shorter Ag ? ? ? Ag distances.10 In the light of the increasingly well-demonstrated 11 tendency of Ag(I) cations to aggregate, we wish to examine the solid state structures and solution conformations of our cryptates wherever short internuclear distances between silver(I) ions exist, for comparison with those of dicopper(I) analogues, in an attempt to investigate the relative degree of interaction present in the diVerent d10-d10 systems.Results and discussion The cryptand hosts, L1 and L2 are suYciently flexible to adopt whichever conformation best meets the preferences of the cationic guest. These cryptands were generated by template synthesis on Ag(I), and are isolated as the disilver cryptates [Ag2L1][ClO4]2?3H2O 1, and [Ag2L2][ClO4]2 2.For comparison, and because few cryptate structures of this ligand were known, the disilver cryptate of L3, [Ag2L3][BF4]2?2H2O 3 (made by treatment of the readily available free cryptand with silver tetrafluoroborate) was also synthesised and recrystallised to give X-ray quality crystals. The 1H NMR spectra of both disilver, 3, and dicopper, [Cu2L3][(ClO4)]2 4, cryptates of L3 have already been reported,12 and give no indication of the existence of unsymmetric or additional conformers in solution. The main diVer- A B A L3 L4 L5 L7 Nbr N R N R N R N Nbr N N N R N Nbr Nbr R R N N N O S R = B L2 L1 L6 L82098 J.Chem. Soc., Dalton Trans., 1999, 2097–2102 ence between the spectra of disilver and dicopper cryptates of this ligand is in the splitting, in 3, of the (C)H imino signal by coupling 3J(1H109,107Ag) to the spin-half silver isotopes, which testifies to the absence of rapid exchange on the NMR time scale, between solvated and cryptated situations for the Ag1 cation.Although 63,65Cu–15N coupling is observable 13 in the solid state MAS spectrum, the quadrupolar nature of the copper nuclei prohibits observation, in solution phase spectra, of coupling to the copper nuclei. So information on dynamic exchange processes in solution cannot be derived from 1H– 63,65Cu coupling, but the sharply resolved nature of the 1H NMR spectrum for 4 suggests the absence of rapid decomplexation equilibria.The crystal structure of the dicopper salt of L2, 5, has been described 14 and the well-resolved simple 1H NMR spectrum, reported for the first time here, suggests that the complex is non-labile and that the solution conformation mirrors that seen in the solid state. X-Ray crystallographic studies Suitable crystals for X-ray study of the nitrate salts of [Ag2L1]21 1a and [Ag2L2]21 2a were obtained by recrystallisation from ethanol–acetonitrile.These structures are shown in Fig. 1 and 2 Fig. 1 Structure of [Ag2L1]21, 1a. Fig. 2 Structure of [Ag2L2]21, 2a. and their dimensions are listed in Table 1 together with those of the tetrafluoroborate salt of [Ag2L3]21?2MeCN 3a (Fig. 3). Table 2 extends the comparison to include other disilver cryptates structurally characterised to date. While all three cryptates retain a trigonal pyramidal N4 cap derived site for Ag1 coordination, distinct diVerences between the coordination sites used in tren- and trpn-capped systems are evident.Comparing the thiophene-spaced analogues 9 [Ag2L8]21/[Ag2L7]21 with the m-xylyl-spaced pair of cryptates [Ag2L2]21 and [Ag2L3]21, we see that in both cases, the trenderived site presents all four N-atoms well to one side of the cation, the distance from cation to the imino N3 plane in 3a being 0.64 Å. In the trpn-derived site of 2a the Ag1 cation lies almost coplanar with, i.e. within 0.085 Å of, the N3 plane.The other consequence of the greater flexibility of the trpn cap is that regular trigonal pyramidal angles are achievable within the trigonal plane, and axial to it, in contrast to the tren-capped cryptate 3a where significant deviations from the ideal 1208 and 908 values are seen. The Ag–N distances, however, are not dissimilar for the two cryptates, Ag–Nimine falling close to 2.3 Å, Fig. 3 Structure of [Ag2L3]21, 3a. Table 1 Selected interatomic distances (Å) and angles (8) [Ag2(L1)]21 1a Ag–N(1) Ag–N(2) Ag–Ag(A)a 2.506(5) 2.338(3) 3.0480(10) N(2)–Ag–N(2D)b N(1)–Ag–N(2) 119.04(2) 84.35(6) [Ag(L3)]21 3a Ag–N(1) Ag–N(5) Ag–N(4A)c Ag–N(3) Ag–Ag(A)c 2.553(2) 2.293(2) 2.298(2) 2.313(2) 3.4491(5) N(5)–Ag–N(3) N(3)–Ag–N(4A)c N(5)–Ag–N(4A)c N(1)–Ag–N(5) N(1)–Ag–N(4A)c N(1)–Ag–N(3) 114.26(9) 109.21(9) 114.46(9) 73.94(8) 73.57(8) 74.31(8) [Ag2(L2)]21 2a Ag(1)–N(1) Ag(2)–N(2) Ag(1)–N(3) Ag(2)–N(4) Ag(1)–Ag(2) 2.41(3) 2.43(2) 2.291(12) 2.358(13) 3.775(2) N(3)–Ag(1)–N(3A)d N(1)–Ag(1)–N(3) N(4)–Ag(2)–N(4A)d N(4)–Ag(2)–N(2) 119.87(4) 87.9(3) 119.86(5) 87.8(3) Symmetry transformations used to generate equivalent atoms: a y 1 1 3 – , x 2 1 3 – , 2z 1 1 6 – ; b 2y 1 1, x 2 y, z; c 1 2 x, y, 1– 2 2 z; d 2y, x 2 y, z.J.Chem. Soc., Dalton Trans., 1999, 2097–2102 2099 Table 2 Comparisons of dimensions in disilver cryptates d [Ag2L1]21 1a [Ag2L2]21 2a [Ag2L3]21 3a [Ag2L4]21 6 [Ag2L5]21 [Ag2L6]c 21 [Ag2L7]21 [Ag2L8]21 Space group Ag(1)–Ag(2)/Å Nbr–Nbr/Å Ag–X/Åa Ag–Nimine/Å Ag–imine plane/Å b Ag–Nbr/Å Nim–Ag–Nim/8 R3� c 3.05 8.06 2.91 2.34 10.23 2.51 119.0 R3c 3.77 8.61 2.56 2.31 10.08 2.42 119.9 C2/c 3.45 8.56 2.65 2.28 10.64 2.55 112.6 P21/n 3.11 8.53 3.15 2.30 10.76 2.68 110.1 P21/a 6.06 10.85 2.47 2.31 10.51 2.40 115.3 P1� 7.09 11.97 3.15 2.32 20.23 2.47 119.5 P21/c 4.65 9.69 3.24 2.29 10.62 2.68 110.2 P21/c 5.00 9.96 3.22 2.31 10.03 2.49 119.8 a Distance to central hydrogen atom(s) for phenyl spacers, to central O or S atom for the other systems. b 1 Indicates displacement towards central cavity, 2 indicates displacement towards bridgehead nitrogen atoms.c One of the silver ions is disordered over two positions. d Average distances given for lower symmetry structures, rounded to two decimal places. and Ag–Nbr in the range 2.4–2.7 Å in all cases. The observation of less regular trigonal pyramidal geometry in tren-capped cryptates in comparison with the trpn-derived analogue is general for dinuclear cryptates 8,9,15 and extends even to the analogous podate systems.16 The Ag ? ? ? Ag separation in 2a, at 3.775(2) Å the longest in the disilver L1–L4 series, exceeds that in 3a by ª0.33 Å and both exceed those in the furan spaced analogues, presumably in response to steric constraints associated with the relatively bulky m-xylyl spacer.The relatively long internuclear distance in 2a would appear to rule out any possibility of Ag ? ? ? Ag bonding.Comparing the disilver cryptate [Ag2L1]21 1a of the furanspaced trpn-derived [Ag2L4]21 ligand, L1, with its structurally characterised 17 tren-derived L4 analogue 6, we see similar eVects on the geometry of the coordination site (Table 1). The metal–ligand distances (Ag–Nimine at 2.338 for 1a vs. 2.265, 2.326 for 6 and Ag–Nbr at 2.506 for 1a vs. 2.656, 2.708 Å for 6) are similar in tren- and trpn-derived cryptates, but the less regular trigonal pyramidal coordination geometry of the trenderived structure is demonstrated in N–Ag–N angles far from 908/1208 and in the larger Ag–imine plane distances of: 0.713, 0.80 Å for 6 against 0.230 Å for the trpn-capped cryptate, 1a.The Nbr–Nbr distance in 1a is >0.4 Å shorter than in the trencapped analogue 6 mainly because of the tighter helical pitch possible in the trpn-capped host. The Ag ? ? ? Ag distance this time, at 3.048(1) Å is just under 0.07 Å shorter in the trpncapped cryptate 1a than in 6.So in this case, both shorter cavity length and Ag ? ? ? Ag separation derive in the larger host from the increased flexibility which allows the development of a tighter pitch in the triple-helical cryptand strands. In comparison with the dicopper(I) analogues,12,14,15,18 M–Nimine distances show an expected extension of around 0.25– 0.35 Å in the disilver analogues 1a–3a, where M–Nbr distances are also longer (by ª0.1–0.2 Å in the trpn- and 0.2–0.3 Å in the tren-capped pairs).What is less intuitively obvious, however, is that Ag ? ? ? Ag distances should be markedly shorter (by >1 Å) than Cu ? ? ? Cu distances in the analogous cryptates. This observation contrasts with results for more rigid cryptand hosts studied earlier 9,19 where dicopper(I) and disilver(I) structures are isomorphous or nearly so. One possible driving force for the M ? ? ? M contraction in disilver(I) relative to the dicopper(I) cryptates might be the existence, in the silver(I) cryptates only, of Ag–p interactions involving aromatic rings or other, e.g.Ag–H, agostic interactions. Table 2 shows that both p- and m-xylyl-spaced hosts exhibit relatively close approach (of the order of 2.4–2.7 Å between Ag1 and aromatic protons). However, the contraction of intercationic distances in the m-xylyl versus p-xylyl spaced analogues is much larger than can be accounted for simply by removal of the additional C–C bond (i.e. ª1.4 Å) in the spacer link.Also, comparison with the dicopper(I) analogues 15,20 reveals similar M–CHar distances. Although these results appear to rule out indirect interaction as a rationale for the short Ag1 ? ? ? Ag1 distances, there remains the hypothesis of direct interaction, of whatever kind, between Ag1 cations, which is at least conceivable on the results of calorimetric measurements 17 in the Ag–L4 system, where positive enthalpic cooperativity for complexation of a second Ag1 cation is exhibited.The enthalpy of complexation of two Ag1 ions within this host is more (by 13 kJ) than double that for complexation of a single Ag1 ion. However this 13 kJ excess cannot be interpreted as solely due to Ag1 ? ? ? Ag1 interaction as reorganisation costs involved in complexation of the first Ag1 are likely to be appreciable. 1H NMR measurements While the 1H NMR spectra of the dicopper(I) cryptates readily demonstrate that both ends and all strands of the cryptand host are equivalent, the situation with the disilver cryptates is more complex (Table 3).The tren-capped disilver(I) cryptates [Ag2L3]21 3 12 and [Ag2L4]21 6,18 when conformationally frozen out at temperatures below ambient, show simple spectra, suggesting retention of the symmetric solid state conformation in solution. The complexity of the low temperature (233 K) 1H NMR spectra of [Ag2L1]21 1 and [Ag2L2]21 2, on the other hand, can only be explained by assuming the presence of more than one species in solution for these trpn-capped cryptates.For [Ag2L1]21 1 (Fig. 4) the appearance of three separate signals in CD3CN solution in the imino and furano region at 233 K demonstrates the existence of at least three separate conformers. The aromatic resonances consist of two singlets and an AB quartet, the latter suggesting an asymmetric conformation, with inequivalent aromatic protons, while the singlets derive from two diVerent but symmetric dinuclear conformers.The Fig. 4 Aromatic and imino resonances in the d 7.0–8.5 region of the 400 MHz 1H NMR spectrum of 1 at 230 K.2100 J. Chem. Soc., Dalton Trans., 1999, 2097–2102 Table 3 Proton NMR spectra of disilver and dicopper cryptates g n/ Imino Methylene CHs Complex MHz T/K H(C]] N)f Aromatic CH ar1 a b g Ref. 1 major sym 1 minor sym 1 unsym 2 average 2 major 2 minor 3456 400 400 400 400 400 400 400 400 500 400 230 230 230 298 233 233 233 233 300 233 8.22,d a (6.5) 8.17,d a9 (7.5) 8.19,m a0,h 8.46,d (9.2) 8.42,d (9.4) 8.47,d (ª9.5) 8.63,d (8.0) 8.50,s 8.29,s 8.24,d (7.9) 7.14,s a 7.07,d a9 7.12,q a0 7.80,t;7.78,d;7.73,m 7.78,m;7.73,m b e,h 7.84,d;7.76,t 7.79,d;7.70,t 7.66,m 7.17,s ——— 9.92,s 9.87,s 9.99,s 9.59,s 9.89,s 10.09,s — e e e 3.35br,s 3.28,d;2.57,t e,h 3.51,t;3.31,d 3.25,m b 3.11,d;3.01,t 3.54,t;3.19,d e e e 1.69br,s 1.81,q; h 1.68,t h e,h 3.08,d;2.65,t 3.13,d;2.64,m 2.33,m; h,c 1.58,d c 2.96,d;2.55,t e e e 2.66;1.90vbr 2.99,t;1.70,d e,h —— 2.51,t;1.84,d — This work This work This work This work This work This work 12 12 This work 9,18 a,a9,a0 Related via NOE enhancement; b Not resolvable as separate signals; c Related by coupling (Jax,ax9 ª Jax,eq ª 15 Hz).e Complex uninterpretable spectra; f 3J(1H,109.107Ag) in ptheses; g Shifts in ppm from TMS; s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet; h Obscured. ratio of these diVerent conformers appears as major symmetrical ª unsymmetrical @ minor symmetrical.NOE experiments relate the most intense (d 7.14) furano signal to the d 8.22 imino doublet having 3J(1H,109,107Ag) = 6.5 Hz (average) and the weak d 7.07 furano-singlet to the d 8.17 imino doublet with 3J(1H,109,107Ag) = 7.5 Hz, while irradiation of the quartet at d 7.12 appears to enhance, though only weakly, imino resonances around d 8.19. As expected for overlap of three sets of methylene signals, one of them of low symmetry, the methylene region of the spectrum is highly complex and can only be approximately assigned on the data available.NOE and COSY experiments have enabled tentative identification of groups of signals to protons a, b or g to the imino N in the regions d 3.5– 2.6, ª1.7 and 2.7–2.3 respectively. For [Ag2L2]21 2, also, the complexity of the d 1.5–4 region of the spectrum discourages attempts at detailed assignment of the methylene protons, at least in the first instance. Instead we initially concentrate on the aromatic and imine regions of the 1H NMR spectrum.Both ligands L3 and L2 contain a unique aromatic proton ar1, which is subject to the eVects of ring currents from aromatic residues in the pair of adjacent cryptand strands. In the tren-capped free ligand, L3, the relative disposition of ring current with respect to this proton is shielding and accordingly ar1 appears at d 5.3, well removed from the normal aromatic region.21 In the disilver cryptate, 3, the eVect of the ring current on this unique proton is reversed, and the ar1 resonance is deshielded 12 relative to the normal aromatic range, appearing around d 9.6. Similar deshielding of ar1 is observed for [Ag2L2]21, 2, where at 298 K a time-averaged singlet resonance is observed at around d 9.92, which first broadens as temperature is decreased before freezing out (by 230 K) to a pair of singlets at d 9.87 and 9.99.Fig. 5 shows the behaviour in the low field (high frequency) region as temperature is varied.The unequal intensity ratio of the resonances points to the presence of a major and a minor solution conformer. Fig. 5 Aromatic and imino resonances in the 400 MHz 1H NMR spectrum of 2 at (i) 298 K and (ii) 230 K. The imino resonance region (d 8.4–8.5) is consistent with this interpretation, consisting at 233 K of a closely overlapped pair of doublets centred at d 8.47 and 8.42, with 3J(1H,109,107Ag) for both close to 9.5 Hz. At 298 K an average coupling between 1H and 109,107Ag of 9.2 Hz is seen; that this coupling persists while the methylene coupling is lost due to time-averaging indicates that the Ag1 ion remains coordinated during whatever dynamic process is responsible for the loss of coupling in the methylene resonances.The greater 3J(1H,109,107Ag) coupling in 2 versus 1 suggests stronger interaction of Ag1 with coordinating imino- N in 2, which does not appear to support any hypothesis of additional Ag1 involvement in bonding via e.g.Ag1–Har agostic interaction or incipient Ag1 ? ? ? Ag1 bonding. The remaining aromatic resonances are less easily interpreted, particularly at low temperature. At 298 K doublet and triplet signals around d 7.8 and a second order multiplet centered around d 7.7 (arising from a combination of o-coupling of around 6 Hz with m-coupling close to 2 Hz) correspond to aromatic protons other than ar1. On freezing out the separate conformers at low temperature, this d 7.7–7.8 aromatic resonance becomes an overlapped multiplet; however it does not become markedly more complex as might be expected if one of the solution conformers corresponded to an unsymmetrically coordinated situation.The methylene region of the spectrum at 298 K is broad and fluxional while the 233 K spectrum is once more too complex to be completely assignable, although outline information supported by NOE and COSY experiments allows tentative assignment of resonances due to the major component (Table 3).These assignments are supported by comparison with the spectrum of the dicopper analogue 5 which is relatively simple as it does not suVer from the complication of having more than one conformation in solution. The evidence of solution NMR studies is that dissolution of the symmetric form of these disilver trpn-capped cryptates results in the generation of additional dinuclear conformer(s), a second symmetric conformer involving a diVerent pair of identical sites, and/or an unsymmetric conformer possibly involving a pair of N4 (inclusive) and N3 (facial) coordination sites.However, in neither of the present examples is there evidence for rapid exchange of Ag1 between the diVerent coordination situations, or with the CD3CN solvent. Whatever the rationale for the readiness of the silver(I) cations to approach closely in the solid state, their propensity to adopt other conformations in solution where silver ions may be more widely separated and/or some solvation of Ag1 may occur, suggests that any Ag ? ? ? Ag interaction cannot exceed by much MeCN solvation enthalpies.Two (not necessarily mutually exclusive) explanations of the relatively short [vs. dicopper(I) analogues] Ag ? ? ? Ag separations present themselves. One is that the larger Ag1 radius allows for tighter pitch in the triple helical strands than isJ. Chem. Soc., Dalton Trans., 1999, 2097–2102 2101 Table 4 Data collection and structure refinement parameters [Ag2(L1)][NO3]2?H2O [Ag2(L3)][BF4]2?2MeCN [Ag2(L2)][NO3]2 Formula MT /K Crystal system a/Å b/Å c/Å b/8 Space group Z m/mm21 Reflections measured Independent reflections (Rint) Goodness of fit on F 2 Final R indices [I > 2s(I)] Final R indices (all data) C36H60Ag2N10O13 1056.68 293(2) Rhombohedral 15.627(3) 15.627(3) 31.583(7) —R 3� c 6 0.951 3011 1316 (0.0151) 1.049 R1 = 0.0281 wR2 = 0.0820 R1 = 0.0372 wR2 = 0.0890 C40H48Ag2B2F8N10 1058.24 153(2) Monoclinic 22.869(2) 11.134(1) 18.937(2) 112.520(6) C2/c 4 0.955 4466 3921 (0.0131) 1.044 R1 = 0.0282 wR2 = 0.0618 R1 = 0.0367 wR2 = 0.0651 C42H54Ag2N10O6 1010.69 293(2) Rhombohedral 15.821(3) 15.821(3) 34.58(1) —R 3c 6 0.835 3309 2957 (0.0530) 1.094 R1 = 0.0635 wR2 = 0.1777 R1 = 0.1017 wR2 = 0.2088 possible where the smaller Cu1 cations are coordinated.The second is that non-zero bond order between the Ag1 cations is responsible for the relatively close approach.The truth may lie somewhere between these steric and electronic alternatives; we are currently engaged 15 in ADF modelling calculations involving disilver(I) and dicopper(I) cryptates, which will enable us to estimate bond order eVects in these diVerent d10 cations and thus lead to a better understanding of the observed behaviour. Conclusion Two comparisons are emphasised as a result of this study: (i) the diVerence between tren-capped and trpn-capped hosts, and (ii) the diVerence between disilver(I) and dicopper(I) cryptates, particularly in respect of cation internuclear distances.In the case of (i), it is seen that the greater flexibility of the trpn-capped ligand allows for adoption of more regular geometry by the encapsulated cations, but this greater flexibility in turn allows competition in solution by several alternative coordination sites for the coordinatively undemanding Ag1 cation. In the case of (ii), it appears that the flexible cryptand skeleton, by utilising a triple helical twist mechanism, allows the pair of cations to select from a range of internuclear distances.The separation adopted may derive from steric factors (eVect of radius of the coordinated guest on the helical pitch) and/or from electronic eVects viz. weak bonding between d10 systems. In the latter model, Ag1 ? ? ? Ag1 interactions would appear to be weak but significant in comparison with the Cu1 ? ? ? Cu1 situation. Our current modelling studies 15 may explain the origin of the diVerent d10 ? ? ?d10 cation separations.Experimental Synthesis of cryptates and precursors Trpn was prepared as described in ref. 22, and used in synthesis of 1, 2 and 5 as described below. 2,5-Dormylfuran was prepared as described in earlier papers.18 Isophthalaldehyde was purchased from Aldrich and used as supplied. [Ag2L1][ClO4]2?3H2O 1. 2,5-Diformylfuran, (0.009 mol) was dissolved in 300 cm3 methanol and filtered into a dropping funnel.It was added slowly overnight to a solution of silver nitrate (0.006 mol) in methanol (100 cm3), concurrently with a solution of trpn (0.006 mol) in methanol (300 cm3) with vigorous stirring at 40 8C. A solution of sodium perchlorate (0.009 mol) in 20 cm3 methanol was then added dropwise, resulting finally in a beige precipitate. The filtrate was left standing overnight to yield a second crop, and finally reduced in volume on a rotary evaporator to give a third crop.Total yield 73%. FABMS m/z (%): 955 (43) Ag2L1ClO4 1; 856 (37) Ag2L11; 747 (40) AgL11. %C H N (calculated percentages in parentheses) C, 39.02 (38.97); H, 4.80 (4.91); N, 10.08 (10.10). The nitrate salt 1a used for X-ray crystallography constituted the final crop of crystals obtained in this preparation. [Ag2L2][ClO4]2 2. Silver nitrate (0.005 mol) and isophthalaldehyde (0.006 mol) were dissolved in methanol (200 cm3) to give a pinky-brown solution which was stirred at 30 8C for 10 min before trpn (0.004 mol) in methanol (100 cm3) was added dropwise over 1.5 h, during which time the solution became dark brown.The solution was stirred at room temperature overnight before being filtered to remove the brown-black silver residues, and excess silver perchlorate (0.012 mol) in 80 cm3 methanol was poured in to produce an instant yellow precipitate which was recovered by gravity filtration. Yield 57%. FAB-MS m/z (%): 985 (12) Ag2L2ClO4 1; 886 (4) Ag2L21; 777 (18) AgL21.%C H N (calculated percentages in parentheses) C, 46.25 (46.47); H, 4.64 (5.01); N, 10.03 (10.32). The sample 2a used for crystallography was obtained by concentrating the solution from template condensation on silver nitrate, with repeated filtration to remove silver residues. [Ag2L3][BF4]2?2H2O 3. To 0.001 mol L3 dissolved in 2 cm3 CHCl3 was added a solution of 0.002 mol of the silver tetrafluoroborate salt in 6 cm3 EtOH. A white precipitate was filtered oV in around 80% yield which could be recrystallised from acetonitrile; slow evaporation in the dark followed by diethyl ether diVusion yields colorless rhombic crystals suitable for X-ray crystallography.FAB-MS m/z (%): 693 (100), AgL3; 802 (30) Ag2L31; 889 (15) Ag2L3BF4 1. %C H N (calculated percentages in parentheses) C, 42.77 (42.77); H, 4.09 (4.48); N, 11.03 (11.08). The sample 3a used for crystallography was recrystallised from MeCN by diethyl ether diVusion. X-Ray crystallography The data sets were all collected on a Siemens P4 diVractometer using graphite-monochromated Mo-Ka radiation (l = 0.71073 Å). [Ag2(L1)][NO3]2?H2O and [Ag2(L3)][BF4]2?2MeCN were solved by direct methods and [Ag2(L2)][NO3]2 was solved using Patterson methods.All three were refined on F 2 using all the data (ref. 23), full-occupancy non-hydrogen atoms were refined anisotropically and hydrogen atoms were included at calculated positions. The data collection and refinement parameters are listed in Table 4.Although the cations (Fig. 1, 2 and 3) all have2102 J. Chem. Soc., Dalton Trans., 1999, 2097–2102 triple helical structures the imposed crystallographic symmetry is diVerent in each case. The [Ag2(L1)]21 cation shows 32 point symmetry, the silver ions lie on the threefold axis and each furan oxygen atom is bisected by a twofold axis. The two nitrate ions associated with each cation are disordered over three equivalent positions on the twofold axes.In [Ag2(L3)][BF4]2?2MeCN the asymmetric unit contains half of the cation, one BF4 2 anion and one acetonitrile solvate molecule. The cation lies on a twofold axis which passes through C(18), C(19) and the mid-point of the Ag–Ag(A) vector. The [Ag2(L2)][NO3]2 structure was initially solved in R3� c which revealed the cation but not the anions; the nitrate anions were only located in the lower symmetry space group R3c. The cation lies on the threefold axis, which passes through the silver ions.As was the case for [Ag2(L1)]- [NO3]2?H2O, the symmetry of the space group requires that the two nitrate ions associated with each cation are disordered over three equivalent positions about the threefold axis but these sites are further disordered. Consequently, each asymmetric unit contains one third of a cation and two thirds of a nitrate ion disordered over two positions. CCDC reference number 186/1445. See http://www.rsc.org/suppdata/dt/1999/2097/ for crystallographic files in .cif format.NMR measurements 1H NMR spectra were obtained on a Bruker ACP 400 spectrometer, under standard conditions, with acetonitrile-d3 as solvent. A 2 s presaturation delay was used for NOE-diVerence spectra. 2D COSY spectra required 256 time-domain points zero filled to 2k × 2k spectrum points after use of sinebell windows. Acknowledgements We thank OURC for support (GGM) and EPSRC for access to services, FAB-MS at Swansea, and high field NMR at Warwick.References 1 S. M. Nelson, Pure Appl. Chem., 1980, 52, 2461. 2 S. M. Nelson, S. G. McFall, M. G. B. Drew and A. H. bin Othman, J. Chem. Soc., Chem. Commun., 1977, 370; M. G. B. Drew, S. G. McFall, C. P. Waters and S. M. Nelson, J. Chem. Res., 1979, 16. 3 J. Nelson, G. Morgan and V. McKee, Prog. Inorg. Chem., 1998, 47, 167. 4 J. Coyle, V. McKee and J. Nelson, Chem. Commun., 1998, 709. 5 J. Nelson, V. McKee and R. M. Town, unpublished results. 6 G. G. Morgan, V. McKee and J. Nelson, Inorg. Chem., 1994, 33, 4427. 7 S.-Y. Yu, Q.-H. Luo, B. Wu, X.-Y. Huang, T.-L. Sheng, X.-T. Wu and D.-X. Wu, Polyhedron, 1997, 16, 453. 8 G. G. Morgan, PhD thesis, Open University, 1996. 9 M. G. B. Drew, O. W. Howarth, G. G. Morgan, D. J. Marrs, C. J. Harding and J. Nelson, J. Chem. Soc., Dalton Trans., 1996, 3021. 10 D. E. Fenton and G. Rossi, Inorg. Chim. Acta, 1985, 98, L29; F. A. Cotton, X. Feng, M. Matusz and R. Poliz, J. Am. Chem. Soc., 1988, 110, 7077; J. Beck and J. Strahle, Z. Naturforsch., Teil B, 1986, 41. 11 P. Pyykko, Chem. Rev., 1997, 97, 597. 12 Q. Lu, J. F. Malone, V. McKee, N. Martin, C. J. Harding and J. Nelson, J. Chem. Soc., Dalton Trans., 1995, 1739. 13 D. Apperley, S. Coles, W. Clegg, J. Coyle, B. Maubert, N. Martin, V. McKee and J. Nelson, J. Chem. Soc., Dalton Trans., 1999, 229. 14 M. P. Ngwenya, J. Ribenspeis and A. E. Martell, J. Chem. Soc., Chem. Commun., 1990, 1207. 15 M. G. B. Drew, D. Farrell, G. G. Morgan, V. McKee and J. Nelson, unpublished results. 16 J. L. Coyle, PhD thesis, Open University, 1999. 17 R. Abidi, F. Arnaud-Neu, M. G. B. Drew, S. Lahely, D. Marrs, J. Nelson and M.-J. Schwing-Weil, J. Chem. Soc., Perkin Trans. 2, 1996, 2747. 18 Q. Lu, J.-M. Latour, C. J. Harding, N. Martin, D. J. Marrs, V. McKee and J. Nelson, J. Chem. Soc., Dalton Trans., 1994, 1471. 19 D. J. Marrs, J. Hunter, C. J. Harding, M. G. B. Drew and J. Nelson, J. Chem. Soc., Dalton Trans., 1992, 3235. 20 M. G. B. Drew, D. McDowell and J. Nelson, Polyhedron, 1998, 21, 2229. 21 D. McDowell, V. McKee, J. Nelson and W. T. Robinson, Tetrahedron Lett., 1987, 7453; M. G. B. Drew, V. Felix, V. McKee, G. G. Morgan and J. Nelson, Supramol. Chem., 1995, 5, 281. 22 J. Chin, M. Banasczyk, V. Jubian and X. Zou, J. Am. Chem. Soc., 1989, 103, 4073. 23 G M Sheldrick, SHELXL-97, Universitat Göttingen, 1997. Paper 9/01041F © Copyright 1999 by the Royal Society of Chemist

ISSN:1477-9226    

DOI:10.1039/a901041f

出版商:RSC    

年代:1999

数据来源: RSC