摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 21–29 21 Anodic oxidation of molybdenum and tungsten in alcohols: isolation and X-ray single-crystal study of side products Vadim G. Kessler,*,†,a Andrei N. Panov,a Nataliya Ya. Turova,a Zoya A. Starikova,b Alexandr I. Yanovsky,b Fedor M. Dolgushin,b Alexandr P. Pisarevsky b and (the late) Yuri T. Struchkov b a Chemistry Department, Moscow State University, 119899 Moscow, Russia b A. N. Nesmeyanov Institute of Organoelement Compounds, Vavilov St. 28, 117813 Moscow, Russia The study of the side products of the anodic dissolution of molybdenum and tungsten metals in alcohols in the presence of LiCl showed them to be [LiMo2O2(OMe)7(MeOH)] 1 in the case of MeOH and [LiMo2O4(OEt)5(EtOH)] 2 in EtOH. Treatment of 2 with an excess of PriOH gave [LiMo2O4(OPri)5(PriOH)] 3, the structure of which was confirmed by a study of [{LiMo2O4(OPri)4(OC2H4OMe)}2] 4, the product of partial substitution of OR groups in 3 by 2-methoxyethoxide ligands.Reaction of 2 with an excess of MeOC2H4OH led to an equimolar mixture of [MoO2(OC2H4OMe)2] and [LiMoO2(OC2H4OMe)3] 5. In PriOH a crystalline product identified as [Mo6O10(OPri)12] 6 was isolated. Anodic oxidation of tungsten in MeOH gave a mixture of homometallic W(OMe)6 and [WO(OMe)4]. Electrosynthesis in EtOH gave as major product an amorphous glass-like mass {after separation of crystalline [WO(OEt)4] by filtration and subsequent drying of the filtrate in vacuo}.Treatment of the latter with an excess of HOC2H4OMe led to crystallization of [{LiWO2(OC2H4OMe)3}2? 2Li(HOC2H4OMe)2]21[W6O19]22 7. Complexes 1, 4 and 7 were characterized by X-ray single-crystal studies. A GLC-mass spectrometric study of the composition of organic side products indicated that the processes were associated with formation of ethers, alkyl halides, aldehydes or ketones and their derivatives. The nature of the possible side reactions was deduced on the basis of the data obtained.The chemistry of metal alkoxides is of interest because of the prospect of their large-scale application as precursors of oxide materials for modern technology.1 The anodic oxidation of metals in anhydrous alcohols in the presence of conductive additives (alkali-metal halides) was developed to meet its requirements.2,3 Earlier we reported the application of this method for the synthesis of the alkoxides of molybdenum and tungsten.4–7 The main products of this reaction were [MO(OR)4].The formation of oxo complexes was due mainly to the propensity of molybdenum and tungsten derivatives M(OR)6 to decomposition with elimination of ethers,4,7 equation (1). When [NBu4]Br was taken in high concentration M(OR)n heat MO(OR)n22 1 R2O (1) as conductive additive after a long electrolysis of tungsten in EtOH the product of complete decay, [NBu4]2[W6O19], was isolated. 5 The formation of oxo complexes was considered to be caused also by cathodic reduction of the initially formed hexavalent alkoxide and subsequent oxidation of reduction products by traces of oxygen (from the atmosphere or dissolved in the solvents applied),8,9 equations (2) and (3).The reddish Cathode: M(OR)n 1 e M(OR)n21 1 RO2 (2) M(OR)n21 1 ��� O2 MoO(OR)n22 1 RO? (3) brown coloration observed in the course of electrolysis was consistent with this supposition. Its intensity under comparable conditions increased with the size and branching of the radical in the series Me ! Et < Pri, while the speed of discoloration of electrolytes in dry air decreased in the same series.The formation of a red crystalline product was observed on anodic dis- † Permanent address: Chemistry Department, Swedish University of Agricultural Sciences, Box 7015, S-75007 Uppsala, Sweden. solution of the metal in PriOH. It was described as [Mo5O7- (OPri)8] {by analogy with [Mo5O7(OH)8], ‘molybdenum blue’, possessing the same color}.4 Anodic dissolution of both molybdenum and tungsten metals in a functional alcohol, 2-methoxyethanol, led to the formation of brightly colored products, which were practically stable to oxidation.7,10 The yields of [MO(OR)4] (M = Mo, R = Me, Et or Pri; 4 M = W, R = Me or Et 5) were in all cases far from quantitative and therefore it was interesting to examine the composition of the side products and try to understand the mechanisms of processes taking place and improve the synthesis of pure products. Experimental All compounds described in this work are highly sensitive to moisture and oxygen and therefore all operations with them were carried out either using a vacuum line or a dry-box.Dehydration of alcohols, ROH, used was performed by boiling them under reflux over magnesium (R = Me or C2H4OMe), calcium (R = Et) and aluminium (R = Pri) alkoxides with subsequent distillation. Toluene and hexane were refluxed with sodium wire and then distilled over LiAlH4. Pyridine was stored over anhydrous KOH and distilled over a fresh portion of KOH prior to use.The carbon and hydrogen contents of the samples were determined using the conventional combustion technique. Molybdenum and tungsten contents were determined gravimetrically as MO3. Hydrated oxides were precipitated from solutions by hot 1 : 1 diluted HNO3, washed with deionized water on the filter and heated to constant weight at 400 8C. The lithium content was determined by flame photometry with a FLAPHO-4 device.Infrared spectra were registered with a Specord IR 75 for Nujol or hexachlorobutadiene mulls. Crystalline solids were identified using powder diffractometry (DRON 3M). The GLC–mass spectrometry tests were carried out using a Varian MAT 311A device (source temperature 200 8C, SE-30 separation column).22 J. Chem. Soc., Dalton Trans., 1998, Pages 21–29 Table 1 Main parameters of the electrochemical synthesis of molybdenum and tungsten alkoxides Alcohol MeOH EtOH PriOH MeOH EtOH Material of anode/cathode Mo/Mo Mo/Pt Mo/Mo W/Pt W/Pt E/V 110 110 220 110 110 [LiCl]/mol dm23 0.076 0.50 0.05 0.021 0.47 t/h 12 18 12 12 30 Final Li :M ratio in the electrolyte 1 : 4.92 1 : 0.87 1 : 0.49 1 : 8.46 1 : 0.83 Products, isolable from electrolyte [MoO(OMe)4] [LiMo2O2(OMe)7(MeOH)] [MoO(OEt)4] [LiMo2O4(OEt)5(EtOH)] [MoO(OPri)4] [Mo6O10(OPri)12] W(OMe)6 [WO(OMe)4] [WO(OEt)4] Yield (%) in relation to metal dissolved 71 17 82 15 46 17 81 17 28 Organic side products detected None EtCl, Et2O MeCH(OEt)2 PriCl, Me2CO Pri 2O None EtCl, Et2O, MeCHO Synthesis and isolation of the products obtained The anodic oxidation of metals was carried out in a standard cell with an undivided cathodic and anodic space, supplied with water cooling; the anodes were plates of molybdenum or tungsten (ª20 cm2) and the cathodes were plates of the same metals or platinum having the same size.3 The parameters of the processes are summarized in Table 1. [LiMo2O2(OMe)7(MeOH)] 1.The wine-red electrolyte prepared by dissolution of molybdenum in MeOH (ª100 cm3) containing LiCl was concentrated to 1/10 of the initial volume and cooled to 0 8C. The precipitate formed was separated by decantation and washed twice with cold MeOH (5 cm3). From the orange crystalline product obtained (8.64 g, 88% in relation to molybdenum metal dissolved), [MoO(OMe)4] was extracted by three portions (each 50 cm3) of hot (50 8C) hexane. The extracts were mixed with each other and after evaporation in vacuum gave a yellow powder (6.26 g, 71%), identified as [MoO(OMe)4] using IR and X-ray powder data (see ref. 6). The residue after extraction, a bright reddish orange powder, was recrystallized from the minimum volume of MeOH, which was then cautiously evaporated in vacuum to dryness giving bright red needles. Yield 2.37 g (17%) (Found: C, 19.71; H, 5.06; Li, 1.38; Mo, 40.8. Calc. for C8H25LiMo2O10: C, 20.02; H, 5.21; Li, 1.36; Mo, 40.0%).IR (cm21): 3533m, 3360s (br), 1453s, 1412m, 1351m, 1154w, 1140 (sh), 1080m, 1046s, 1005s, 916s,* 894s,* 569s and 500s (br).‡ [LiMo2O4(OEt)5(EtOH)] 2. The attempts to crystallize the ethoxide analog of compound 1 being unsuccessful and observing the continuous decoloration of the electrolytes even in an inert atmosphere, we decided to subject the solutions to oxidation. Tdish brown electrolyte obtained was thus left for 2 d in a vessel connected to the atmosphere via a column filled with dry molecular sieves (4 Å).Its color then slowly changed to yellowish brown. The solution obtained was concentrated in vacuum and [MoO(OEt)4] was extracted by hexane from the liquid residue formed, using the technique described 4 (10.42 g, 82%). The residue from the extraction, a yellowish brown viscous liquid, was left for crystallization. After a week the formation of slightly yellowish transparent crystals was observed.They were separated from the mother-liquor by decantation and dried in vacuum. Yield 1.91 g (15%) (Found: C, 26.06; H, 5.93; Li, 1.23; Mo, 36.5. Calc. for C12H31LiMo2O10: C, 26.99; H, 5.81; Li, 1.22; Mo, 36.0%). IR (cm21): 3406s (br), 1273w, 1160m, 1097s, 1052s (br), 954 (sh),* 940s,* 920s,* 887s (br),* 814w,* 614w, 558s (br), 471m and 427w. [LiMo2O4(OPri)5(PriOH)] 3. To a portion of the crystals of compound 2 (ª0.5 g) was added PriOH (20 cm3) and after the ‡ Characteristic bands in the M]O and C]C vibration region are indicated by an asterisk.dissolution was completed the solvent was removed in vacuum. The operation was repeated twice. The residue finally obtained was a viscous yellowish brown matter displaying rather high solubility in hexane and PriOH. After approximately 3 d of storage in a refrigerator at 0 8C the formation of nearly colorless transparent crystals was observed. It was practically impossible to separate them properly from the surrounding amorphous matrix, which eventually precluded reliable microanalysis data or determination of the yield.The IR spectrum was registered for a single crystal cleaned manually (cm21): 3400s (br), 1170m, 1125 (sh), 1110s, 1067m, 1035m, 1010w, 980s,* 930s (br),* 901s,* 885s,* 845m,* 826m,* 817m,* 600s and 475m. [LiMo2O4(OPri)4(OC2H4OMe)] 4. To a portion (1 g, ª1.6 mmol) of the product obtained above was added hexane (10 cm3) and then MeOC2H4OH (0.2 g, 2.6 mmol).The system immediately separated into two liquid phases and in a week colorless transparent crystals grew from the lower layer. The yield was 0.82 g (88%) (Found: C, 32.16; H, 6.29; Li, 1.1; Mo, 33.1. Calc. for C15H35LiMo2O10: C, 31.38; H, 6.10; Li, 1.13; Mo, 33.5%). IR (cm21): 1366s, 1348w, 1330w, 1264w, 1239w, 1199w, 1167m, 1129s, 1108vs, 1060s, 1015m, 964 (sh),* 936vs (br),* 908s,* 893s,* 878 (sh),* 848m,* 833m,* 822m,* 612s, 602s, 551w, 481m, 455m (br), 421w and 394m.[LiMoO2(OC2H4OMe)3] 5. To this residue from the [MoO(OEt)4] extraction from the corresponding electrolyte, containing an estimated 2.5 g of compound 2, was added MeOC2H4OH (20 cm3) and the mixture was heated to 60 8C, and then the solvent was evaporated to dryness in vacuum. This operation was repeated twice and the residual viscous dark matter left for crystallization which occured in 2–3 h with the formation of crystals of two types, bulky prisms and thin needles. On addition of MeOC2H4OH (4 cm3) and heating to 40 8C the needles dissolved completely. The residue consisted of 0.61 g of prismatic crystals found to be [MoO2(OC2H4OMe)2] from the IR and microanalysis data (see ref. 10). On cooling the above-mentioned solution the precipitation of needle-shaped crystals of compound 5 occurred (Found: C, 20.64; H, 4.06; Li, 1.91; Mo, 26.1. Calc. for C9H21LiMoO8: C, 20.00; H, 3.89; Li, 1.94; Mo, 26.67%). IR (cm21): 1105s, 1052s, 1024s, 1000s, 964w,* 898s,* 870s,* 846s,* 820s,* 800s,* 600s, 550s, 515m, 485s and 425s.Compound 5 was also synthesized on saturation of 10% LiOC2H4OMe solution (3 g) (prepared by dissolution of lithium metal in MeOC2H4OH) with solid [MoO2(OC2H4OMe)2] (1.02 g) (prepared according to ref. 10). Compound 5 (0.81 g, 61%) precipitated after the mixture had been kept overnight at room temperature. [Mo6O10(OPri)12] 6. The mixture obtained by dissolution of molybdenum in PriOH consisted of two different products, a yellowish brown solution {from which, after the evaporation ofJ.Chem. Soc., Dalton Trans., 1998, Pages 21–29 23 solvent in vacuum, [MoO(OPri)4] containing variable amounts of Li according to microanalysis and thus polluted by compound 3 can be extracted by hexane} and reddish orange crystals. The latter are practically insoluble in the electrolyte on heating. They do not possess any noticeable solubility either in PriOH or hexane and are relatively stable to oxidation in a dry atmosphere.We have earlier erroneously described them as [Mo5O7(OPri)8] by analogy with the [Mo5O7(OH)8] ‘molybdenum blue’ possessing the same color.4 In the present work we compared the microanalytical and IR spectral data for this product with those given9 for [Mo6O10(OPri)12] (Found: C, 28.85; H, 5.61; Mo, 40.4. Calc. for C36H84Mo6O22: C, 29.92; H, 5.82; Mo, 39.9%). IR (cm21): 1319m, 1140m, 1118m, 1099s, 986s,* 953s,* 930vs,* 850m,* 842m,* 800s (vbr),* 618 (sh), 605s, 555m, 500 (sh) and 473m.The obtained crystals of compound 6 demonstrate the same chemical properties as those described for [Mo6O10(OPri)12] in ref. 9: they are perfectly soluble in toluene, but cannot be crystallized out again from this solution by either cooling or evaporation of the solvent; the latter yields a viscous reddish brown liquid. The addition of dry pyridine (py) to the mentioned solution leads to precipitation of a red crystalline powder of [Mo4O8(OPri)4(py)4] as has been described 9 for the discussed compound.Anodic oxidation of tungsten in methanol. The reddish orange electrolyte (ª100 cm3) was evaporated in vacuum practically to dryness and the residue extracted by hexane (50 cm3). The extract had a dark blue color which soon disappeared. When dried in vacuum it gave a colorless crystalline product identical with that described 5 and thus being a mixture of W(OMe)6 and [WO(OMe)4] (see Table 1). Anodic dissolution of tungsten in ethanol; isolation of [{LiWO2(OC2H4OMe)3}2?2Li(MeOC2H4OH)2]21[W6O19]22 7.The wine-red electrolyte obtained was left for 2 d in contact with the atmosphere via a column filled with molecular sieves which made it colorless. From this colorless solution thin needle-shaped crystals crystallized in 2 d. They were filtered off, washed on the filter by two portions of EtOH (each 10 cm3) and dried in vacuum. This gave 4.16 g (28%) of a product identified as [WO(OEt)4] by IR and microanalysis (see ref. 11). Evaporation of the filtrate gave ª11 g of a colorless amorphous glasslike matter. It was mixed with MeOC2H4OH (40 cm3), heated to 60 8C and then dried in vacuum. This procedure was repeated twice and the colorless liquor obtained was layered with hexane (30 cm3). After approximately 18 h large irregularly shaped crystals crystallized; yield 6.41 g, ª50% (Found: C, 13.81; H, 2.87; Li, 1.01; W, 55.9. Calc. for C30H74Li4O43W8: C, 13.73; H, 2.82; Li, 0.99; W, 56.15%).IR (cm21): 1352s, 1290m, 1280w, 1265 (sh), 1245m, 1204s, 1130s (br), 1110m, 1092 (sh), 1070s, 1036s, 1016s, 982m,* 940s,* 918vs,* 900vs,* 876s*, 820s,* 625s, 578s, 566s, 535s, 508s, 465s and 410m. Crystallography All compounds studied are extremely sensitive to the ambient atmosphere and therefore were placed in glass capillaries, sealed under an inert atmosphere for data collection. The crystal data and experimental conditions are presented in Table 2. All calculations were performed on an IBM personal computer using SHELXTL PLUS programs (Version 4 for 4, Version 5 for 1 and 7).12 All structures were solved by direct methods and refined by the full-matrix least-squares technique.All H atoms in 1 were located in the Fourier-difference synthesis and refined in the isotropic approximation. The positions of hydrogen atoms in 4 and 7 [with the exception of those of hydroxyl groups, i.e. atoms H(10) and H(12) in 7] were calculated geometrically and included in the refinement in isotropic approximation for 4 and using the riding motion model for 7; the thermal parameters for H atoms in 7 were taken as Uiso = 1.2Ueq(C) for those in CH and CH2 groups and 1.5Ueq(C) for those of CH3 groups, where Ueq(C) was the equivalent parameter for the carbon atom to which the hydrogen atom is attached.Atoms H(10) and H(12) in 7 were located in the Fourier-difference syntheses and thereafter refined using the riding motion model; the Uiso values were taken as 1.2Ueq(O), where Ueq(O) was the equivalent parameter for the O(10) and O(12) atom respectively.CCDC reference number 186/752. For crystallographic files in CIF format see http:// www.rsc.org/suppdata/dt/1998/21/. Results and Discussion Molecular and crystal structures The structure of compound 1 can be most naturally considered as consisting of anionic [Mo2O2(OMe)7]2 and cationic [Li(Me- OH)]1 fragments bound in infinite chains parallel to the a axis; a fragment of the chain is shown in Fig. 1; see also Table 3. The hydrogen bonds which occur between the OH group H(1)]O(10) of the solvating MeOH molecule and one of the oxygen atoms O(4) connect the chains into layers parallel to the ac plane [O(10) ? ? ? O(4D) 2.911(5), O(10)]H(1) 0.91(11), H(1) ? ? ? O(4D) 2.17 Å, O(10)]H(1)]O(4D) 139(9)8]. The anionic fragment in the crystal of compound 1 is an example of the M2X9 group which is very widespread in the chemistry of metal alkoxides and halides and consists of two octahedra sharing a common face.Such an isolated fragment has been most frequently observed in the chemistry of molybdenum derivatives, e.g. in heterometallic 13,14 anionic [Mo2O2- (OMe){S(CH2)3S}3]215 or neutral homo- [Cl2OMo(m-OEt)2- (m-EtOH)MoOCl2] 16 or hetero-metallic [(MeO)2OMo(m-OMe)3- ReO(OMe)2] 17 cluster fragments. As a particular feature of 1 the slightly shorter Mo ? ? ? Mo distance [2.6545(7) Å] in comparison with those observed for the above-mentioned analogs [2.875(1),15 2.683–2.697 16 and 2.658(2) 17 Å] should be mentioned. It can be explained by the fact that in this case the cluster fragment is incorporated into the structure of a polymer which in its turn should reduce the contribution of additional p interaction of the molybdenum atom orbitals with those of oxygen atoms of terminal alkoxide groups.The presence of this Fig. 1 Structure of a fragment of the polymeric chain in the crystal of compound 1.Atoms derived from the reference atoms by symmetry operations are denoted by capital letters: A x 1 1, y, z; B x 2 1, y, z; C x, 0.5 2 y, z 2 0.5; D x, 0.5 2 y, z 1 0.524 J. Chem. Soc., Dalton Trans., 1998, Pages 21–29 Table 2 Crystal data and the details of diffraction experiments for compounds 1, 4 and 7 Formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 T/K Z Dc/g cm23 F(000) m/cm21 Tmin, Tmax Absorption correction Diffractometer 2qmax/8 Scan mode h,k,l Range Check reflection variation (%) Number of measured reflections Number of independent reflections Number of observed reflections Number of parameters refined R1 a wR2b R9 c Goodness of fit Maximum D/s ratio Maximum, minimum residual electron density/e Å23 1 C8H25LiMo2O10 480.1 Monoclinic P21/c 8.449(2) 14.256(3) 14.223(3) 97.33(3) 1699.1(6) 293 4 1.877 960 15.15 Enraf-Nonius CAD4 58 q 25 to 3q 1h, 1k, ±l 2.0 5049 4496 3085 [I > 2s(I)] 290 0.0337 0.1030 — 1.24 0.19 0.90, 20.91 4 C30H70Li2Mo4O20 1148.6 Monoclinic P21/n 11.535(6) 18.802(8) 11.673(5) 111.11(2) 2362(2) 157 2 1.615 1168 11.05 Siemens P3/PC 40 q–2q 1h, 1k, ±l 1.5 5601 5297 4305 [I > 3s(I)] 393 0.0299 —d 0.0420 1.05 2.03 0.87, 20.94 7 C30H74Li4O43W8 2621.5 Triclinic P1� 11.461(5) 11.828(4) 12.623(5) 89.90(3) 110.73(3) 103.01(3) 1554(1) 293 1 2.802 1202 148.4 0.246, 0.983 y Scans Enraf-Nonius CAD4 60 q–5/3q 1h, ±k, ±l 2.0 8628 8192 5988 [I > 2s(I)] 388 0.0591 0.1761 — 0.98 0.01 1.705, 21.568 Graphite-monochromated Mo-Ka radiation (l 0.710 73 Å).a S|Fo 2 |Fc /S(Fo) for observed reflections. b [Sw(Fo 2 2 Fc 2)2/Sw(Fo 2)2]� �� for all reflections. c [Sw(|Fo| 2 |Fc|)2/SwFo 2]� �� for observed reflections. d Structure was refined against F; no wR2 value was calculated. effect is reflected in the values of the Mo]O]C bond angles for the ‘terminal alkoxo groups’ [O(5), O(6), O(7), O(9)] which in the structure of 1 are regularly lower [128.1(3)–130.2(4)8] than those, for example, in the structure of [MoO(OMe)4] [130.7(6)– 137.2(6)8],6 where the additional p interaction plays a very significant role.As well as in the structures of the analogous fragments 15–17 all the Mo(1)]m2-O(R)]Mo(2) bridges in compound 1 are practically symmetric with their lengths falling in the interval usually observed. The lengths of the bonds formed by oxygen atoms not affected by trans influence such as O(1) [Mo(1)]O(1) 2.108(3), Mo(2)]O(1) 2.102(3) Å] and O(3) [Mo(1)]O(3) 2.070(3), Mo(2)]O(3) 2.072(3) Å] are slightly shorter than those in the Mo(1)]O(2)]Mo(2) bridge [Mo(1)]O(2) 2.132(3), Mo(2)]O(2) 2.154(3) Å] elongated due to the trans influence of the oxygens O(4) and O(8).The Mo(1)]m-O(R)]Mo(2) angles are smaller in 1 (76.6–79.78) than in the structures of the analogs (for example, 84.48 in ref. 15) due, presumably, to the shorter Mo ? ? ? Mo distance. The bonds which can be considered as ‘terminal’ for the Mo2O2(OMe)7 fragment [Mo(1)]O(5) 1.963(3), Mo(1)]O(6) 1.944(3), Mo(2)]O(7) 1.961(3), Mo(2)]O(9) 1.949(3) Å] are noticeably longer than terminal bonds in the structures of stereochemically isolated fragments (1.895–1.904 Å in that of an isoelectronic molecule 17) which demonstrates the relative character of the subdivision of the structure of 1 into ionic fragments and indicates predominantly the covalent nature of bonding within the polymer, in spite of the fact that Li]O(Mo) bond lengths in 1 (2.075–2.130 Å) are close to the sum of the correspondent ionic radii (2.08 Å for five-co-ordinated lithium according 18).The co-ordination polyhedron of the lithium atom in compound 1 is a tetragonal pyramid. The Li(1) atom is displaced by 0.53(9) Å from the basal plane of the pyramid, O(5)O(6)O(7B)- O(9B), towards the oxygen atom of the MeOH molecule O(10), occupying the axial position and forming the shortest [Li(1)]O(10) 1.972(9) Å] of all Li]O contacts; other Li]O distances are in the range 2.075–2.130 Å.The lithium atom bonded only to the O atoms of alkoxide groups and solvating alcohol molecules usually displays a tetrahedral co-ordination as observed in the structures of [LiNbO(OEt)4(EtOH)],19 [LiNb- (OEt)6],20 [Li2Ti2(OPri)10],21 [LiZr2(OPri)9(PriOH)].22 The higher co-ordination of lithium in 1 is apparently due to the lower size of the methoxide ligand. An interesting feature of the structure is that the polymeric molecule is built up of alternating M(m- OR)2M and M(m-OR)3M elements.The majority of the known polymeric structures of bimetallic alkoxides contain fragments of only one type, for example [{LiNb(OEt)6}n] features only M(m-OR)2M,20 whereas [{KSn(OBut)3}n] only M(m-OR)3M.23 In combination, apart from the structure of 1, these two different fragments are present also in the structure of [LaNb2(OPri)13],19 where the lanthanum and one niobium atom form a binuclear {LaNbO9} fragment, while the other niobium atom is connected to lanthanum via two Nb]O]La bridges thus forming a four-membered LaNbO2 ring.The structure of compound 4 is built up of centrosymmetric dimeric [{LiMo2O4(OPri)4(OC2H4OMe)}2] molecules (Fig. 2, Table 4), and is a combination of a pair of identical triangular fragments that belong to a very widespread [M3(m3-L)2(m-L9)223] (L, L9 = O or OR) type known for both homo-, as [Ti3O(OMe)( OPri)9] 24 and [Ce3O(OPri)10],25 and hetero-metallic alkoxides, [BaTi2(OEt)10(EtOH)5],26 [LiZr2(OPri)9(PriOH)],22 etc.The skeleton of the triangular fragment is severely distorted. Thus for the Mo]m-O(R) distances no decrease is observed when we go from the tridentate [Mo(1)]O(5) 2.055(2), Mo(1)]O(7) 2.175(2), Mo(2)]O(5) 2.273(2), Mo(2)]O(7) 2.245(2) Å] to the bidentate groups [Mo(1)]O(6) 2.280(2), Mo(2)]O(6) 1.996(3), Mo(2)]O(10) 1.948(2) Å], while the Mo]O(R) terminal bonds [Mo(1)]O(9), 1.878(2) Å] are nevertheless shorter than any ofJ.Chem. Soc., Dalton Trans., 1998, Pages 21–29 25 the Mo]O(R) bridgiThe difference in the lengths of the bonds of the same kind is caused apparently by the trans influence of the oxygen atoms which appear as terminal ligands for the triangular fragment, and the small size of the lithium atom. The same kind of distortion is observed in the structure of [{NaMo2O4(OPri)5(PriOH)}2],13 which is the closest structural analog of 4.The Mo]O distances in 4 fall within the range usually observed and the MoO6 octahedra show minor distor- Fig. 2 Structure of the centrosymmetric dimer in the crystal of compound 4. Only one carbon atom is shown for each of the Pri groups; letter A denotes atoms related to the corresponding reference atoms by inversion Table 3 Selected bond lengths (Å) and angles (8) in compound 1* Mo(1)]O(4) Mo(1)]O(6) Mo(1)]O(5) Mo(1)]O(3) Mo(1)]O(1) Mo(1)]O(2) Mo(1) ? ? ? Mo(2) Mo(2)]O(8) Mo(2)]O(9) Mo(2)]O(7) O(4)]Mo(1)]O(6) O(4)]Mo(1)]O(5) O(6)]Mo(1)]O(5) O(4)]Mo(1)]O(3) O(6)]Mo(1)]O(3) O(5)]Mo(1)]O(3) O(4)]Mo(1)]O(1) O(6)]Mo(1)]O(1) O(5)]Mo(1)]O(1) O(3)]Mo(1)]O(1) O(4)]Mo(1)]O(2) O(6)]Mo(1)]O(2) O(5)]Mo(1)]O(2) O(3)]Mo(1)]O(2) O(1)]Mo(1)]O(2) O(8)]Mo(2)]O(9) O(8)]Mo(2)]O(7) O(9)]Mo(2)]O(7) O(8)]Mo(2)]O(3) O(9)]Mo(2)]O(3) O(8)]Mo(2)]O(1) O(9)]Mo(2)]O(1) O(3)]Mo(2)]O(2) O(1)]Mo(2)]O(2) 1.692(3) 1.944(3) 1.963(3) 2.070(3) 2.108(3) 2.132(3) 2.6545(7) 1.681(3) 1.949(3) 1.961(3) 106.7(2) 104.9(2) 78.94(14) 93.97(14) 87.55(13) 159.24(14) 90.4(2) 160.74(14) 88.31(13) 100.26(11) 152.66(14) 95.30(13) 94.93(14) 70.50(11) 71.26(12) 105.6(2) 105.5(2) 78.69(14) 92.3(2) 88.26(12) 91.0(2) 161.01(14) 70.02(12) 70.93(12) Mo(2)]O(3) Mo(2)]O(1) Mo(2)]O(2) Li(1)]O(10) Li(1)]O(6) Li(1)]O(5) Li(1)]O(7B) Li(1)]O(9B) Li(1A)]O(7) Li(1A)]O(9) O(10)]Li(1)]O(6) O(10)]L(1)]O(5) O(6)]Li(1)]O(5) O(10)]Li(1)]O(7B) O(6)]Li(1)]O(7B) O(5)]Li(1)]O(7B) O(10)]Li(1)]O(9B) O(6)]Li(1)]O(9B) O(70)]Li(1)]O(9B) C(1)]O(1)]Mo(2) C(1)]O(1)]Mo(1) Mo(2)]O(1)]Mo(1) C(2)]O(2)]Mo(1) Mo(1)]O(2)]Mo(2) C(3)]O(3)]Mo(1) C(3)]O(3)]Mo(2) Mo(1)]O(3)]Mo(2) C(5)]O(5)]Mo(1) C(5)]O(5)]Li(1) Mo(1)]O(5)]Li(1) C(6)]O(6)]Mo(1) C(6)]O(6)]Li(1) Mo(1)]O(6)]Li(1) C(7)]O(7)]Mo(2) 2.072(3) 2.102(3) 2.154(3) 1.972(9) 2.075(9) 2.080(9) 2.103(9) 2.130(9) 2.103(9) 2.130(9) 103.1(4) 107.1(4) 73.4(3) 106.4(4) 150.5(5) 99.4(4) 102.0(4) 100.5(4) 71.7(3) 123.6(3) 124.0(3) 78.17(10) 124.0(3) 76.55(10) 122.5(3) 123.9(3) 79.71(10) 129.8(4) 126.8(5) 103.2(3) 128.5(3) 121.8(4) 104.0(3) 130.2(4) * Letters A and B denote the atoms derived from the reference atoms by the symmetry operations x 1 1, y, z and x 2 1, y, z respectively.tion, the sums of the bond lengths in trans position to each other differing insignificantly: 3.871–4.005 Å for Mo(1) and 3.941–3.978 Å for Mo(2). The stronger distortion of the Mo(1) octahedron is apparently associated with the co-ordination of O(2) by the lithium atom of the neighbouring fragment.The Mo(1)]O(2) bond length in 4 [1.725(2) Å] practically coincides with that observed for the sodium analog [Mo]O(Na) 1.722 Å], which in combination with the observed short alkali metal atom–oxo-oxygen atom distances [Li]O(2) 1.903(5), Na]O 2.256(3) Å], being significantly shorter than the sums of corresponding atomic radii (2.08 and 2.49 Å 18 respectively), demonstrates that the Mo]] O bond is sufficiently active to co-ordinate the sodium and lithium atoms.The multiple character of bonding is preserved which is shown not only by the short Mo]O distance but also by the decreased value of the Li(1)]O(2)] Mo(1) angle value, 156.6(2)8. In the structure of [{ZnTa2IO2- (OPri)7}2], where the triangular IZnTa2(m3-O)(m-OR)3(OR)4O fragments are also connected via an oxygen atom which is in this case disposed symmetrically relative to the tantalum atoms of different fragments, the Ta](m-O)]Ta angle is as large as 175.3(5)8.27 The lithium atom co-ordination in 4 seems to be rather irregular and even if the atom O(5), forming the weakest and presumably ‘forced’ contact with it, is excluded from consideration, the environment of Li(1) can hardly be treated as a distorted tetrahedron.The O]Li]O angles vary within the limits of 63.4–146.28 (see Table 4), the Li]O bond lengths falling into the range 1.903–2.020 Å. Only one lithium atom contact in the structure of 4 [Li(1)]O(7)] is close to the sum of ionic radii (for tetrahedrally co-ordinated lithium), while two contacts [Li(1)]O(2A) and Li(1)]O(10)] are significantly shorter.This fact in combination with the influence on the bonding parameters of the oxygen atoms connected to lithium indicates Table 4 Selected bond lengths (Å) and angles (8) for compound 4 * Mo(1)]O(1) Mo(1)]O(2) Mo(1)]O(9) Mo(1)]O(5) Mo(1)]O(6) Mo(1)]O(7) Mo(2)]O(3) Mo(2)]O(4) Mo(2)]O(5) O(1)]Mo(1)]O(2) O(1)]Mo(1)]O(5) O(2)]Mo(1)]O(5) O(1)]Mo(1)]O(6) O(2)]Mo(1)]O(6) O(5)]Mo(1)]O(6) O(1)]Mo(1)]O(7) O(5)]Mo(1)]O(7) O(6)]Mo(1)]O(7) O(1)]Mo(1)]O(9) O(2)]Mo(1)]O(9) O(5)]Mo(1)]O(9) O(3)]Mo(2)]O(4) O(3)]Mo(2)]O(5) O(4)]Mo(2)]O(5) O(3)]Mo(2)]O(6) O(4)]Mo(2)]O(6) O(5)]Mo(2)]O(6) O(6)]Mo(2)]O(7) O(3)]Mo(2)]O(7) O(4)]Mo(2)]O(7) O(5)]Mo(2)]O(7) O(5)]Mo(2)]O(10) O(6)]Mo(2)]O(10) O(7)]Mo(2)]O(10) O(5)]Li(1)]O(7) O(5)]Li(1)]O(8) 1.696(2) 1.725(2) 1.878(2) 2.055(2) 2.280(2) 2.175(2) 1.705(3) 1.696(2) 2.273(2) 103.7(1) 98.8(1) 95.2(1) 88.6(1) 162.3(1) 70.0(1) 158.3(1) 68.6(1) 70.8(1) 100.1(1) 103.1(1) 149.6(1) 104.3(1) 157.6(1) 97.5(1) 98.6(1) 102.2(1) 71.1(1) 74.7(1) 94.7(1) 161.1(1) 63.7(1) 80.8(1) 147.2(1) 78.0(1) 63.4(2) 146.2(3) Mo(2)]O(6) Mo(2)]O(7) Mo(2)]O(10) Li(1)]O(8) Li(1)]O(10) Li(1)]O(5) Li(1)]O(7) Li(1)]O(2A) O(2)]Li(1A) O(7)]Li(1)]O(8) O(5)]Li(1)]O(10) O(7)]Li(1)]O(10) O(8)]Li(1)]O(10) O(5)]Li(1)]O(2A) O(7)]Li(1)]O(2A) O(8)]Li(1)]O(2A) O(10)]Li(1)]O(2A) Mo(1)]O(2)]Li(1A) Mo(1)]O(5)]Mo(2) Mo(1)]O(5)]Li(1) Mo(2)]O(5)]Li(1) Mo(1)]O(5)]C(1) Mo(2)]O(5)]C(1) Li(1)]O(5)]C(1) Mo(1)]O(6)]Mo(2) Mo(1)]O(6)]C(4) Mo(2)]O(6)]C(4) Mo(1)]O(7)]Mo(2) Mo(1)]O(7)]Li(1) Mo(2)]O(7)]Li(1) Mo(1)]O(7)]C(7) C(8)]O(8)]C(9) Mo(1)]O(9)]C(10) Mo(2)]O(10)]Li(1) Mo(2)]O(10)]C(13) Li(1)]O(10)]C(13) 1.996(3) 2.245(2) 1.948(2) 2.020(5) 1.969(5) 2.479(5) 1.990(6) 1.903(5) 1.903(5) 82.8(2) 75.3(2) 84.0(2) 103.9(2) 105.5(2) 142.6(3) 100.6(2) 129.7(3) 156.6(2) 97.5(1) 93.6(1) 78.0(1) 121.8(2) 130.7(2) 122.9(2) 99.0(1) 127.2(2) 131.8(2) 94.9(1) 105.6(2) 89.8(1) 124.6(2) 113.5(3) 129.8(2) 99.7(2) 125.8(2) 132.4(3) * Letter A denotes the atoms derived from the reference atoms by the symmetry operation 2x, 1 2 y, 2z.26 J.Chem. Soc., Dalton Trans., 1998, Pages 21–29 Table 5 Selected bond lengths (Å) and angles (8) in compound 7 * W(1)]O(2) W(1)]O(1) W(1)]O(7A) W(1)]O(5) W(1)]O(3A) W(1)]O(3) W(2)]O(18) W(2)]O(15) W(2)]O(17B) O(2)]W(1)]O(1) O(2)]W(1)]O(7A) O(1)]W(1)]O(7A) O(2)]W(1)]O(5) O(1)]W(1)]O(5) O(7A)]W(1)]O(5) O(2)]W(1)]O(3A) O(1)]W(1)]O(3A) O(7A)]W(1)]O(3A) O(5)]W(1)]O(3A) O(2)]W(1)]O(3) O(1)]W(1)]O(3) O(7A)]W(1)]O(3) O(5)]W(1)]O(3) O(3A)]W(1)]O(3) O(18)]W(2)]O(15) O(18)]W(2)]O(17B) O(15)]W(2)]O(17B) O(18)]W(2)]O(20) O(15)]W(2)]O(20) O(17B)]W(2)]O(20) O(18)]W(2)]O(21) 1.720(8) 1.725(8) 1.952(8) 1.971(8) 2.140(7) 2.179(7) 1.695(9) 1.916(9) 1.925(10) 103.7(4) 98.1(4) 92.5(4) 92.9(4) 97.9(4) 162.6(3) 95.5(3) 160.1(3) 79.2(3) 86.4(3) 163.1(3) 92.7(3) 85.0(3) 80.7(3) 68.7(3) 104.1(5) 103.8(5) 152.1(4) 103.6(4) 86.7(4) 86.4(4) 103.4(4) W(4)]O(13) W(4)]O(17) W(4)]O(16) W(4)]O(14) W(4)]O(15) W(4)]O(22) Li(1)]O(9) Li(1)]O(10) Li(1)]O(11) O(11)]Li(1)]O(12) O(2)]Li(1)]O(12) O(4)]Li(2)]O(7) O(4)]Li(2)]O(6) O(7)]Li(2)]O(6) O(4)]Li(2)]O(5) O(7)]Li(2)]O(5) O(6)]Li(2)]O(5) O(4)]Li(2)]O(8) O(7)]Li(2)]O(8) O(6)]Li(2)]O(8) O(5)]Li(2)]O(8) O(4)]Li(2)]O(3) O(7)]Li(2)]O(3) O(6)]Li(2)]O(3) O(5)]Li(2)]O(3) O(8)]Li(2)]O(3) W(1)]O(2)]Li(1) C(1)]O(3)]W(1A) C(1)]O(3)]W(1) W(1A)]O(3)]W(1) C(1)]O(3)]Li(2) 1.701(11) 1.899(9) 1.904(9) 1.909(10) 1.929(9) 2.317(1) 2.00(2) 2.02(3) 2.04(2) 79.2(8) 94.4(9) 127.7(11) 117.6(11) 113.5(10) 105.3(11) 93.4(9) 80.3(8) 91.3(9) 78.1(9) 89.9(9) 163.3(11) 75.5(7) 67.0(6) 148.1(11) 68.0(6) 120.0(9) 141.9(7) 125.1(6) 121.5(5) 111.3(3) 99.7(7) W(2)]O(20) W(2)]O(21) W(2)]O(22) W(3)]O(19) W(3)]O(21) W(3)]O(14B) W(3)]O(20B) W(3)]O(16) W(3)]O(22) O(15)]W(2)]O(21) O(17B)]W(2)]O(21) O(20)]W(2)]O(21) O(18)]W(2)]O(22) O(15)]W(2)]O(22) O(17B)]W(2)]O(22) O(20)]W(2)]O(22) O(21)]W(2)]O(22) O(19)]W(3)]O(21) O(9)]Li(1)]O(10) O(9)]Li(1)]O(11) O(10)]Li(1)]O(11) O(9)]Li(1)]O(2) O(10)]Li(1)]O(2) O(11)]Li(1)]O(2) O(9)]Li(1)]O(12) O(10)]Li(1)]O(12) C(7)]O(7)]W(1A) C(7)]O(7)]Li(2) W(1A)]O(7)]Li(2) C(9)]O(8)]Li(2) 1.929(8) 1.935(8) 2.323(1) 1.708(9) 1.922(8) 1.926(10) 1.926(8) 1.929(10) 2.327(1) 86.9(4) 87.1(4) 152.9(3) 179.5(4) 76.3(3) 75.7(3) 76.5(2) 76.4(2) 102.9(4) 80.7(9) 97.8(11) 122.9(11) 96.0(9) 124.5(12) 112.5(12) 169.5(12) 92.5(10) 127.8(8) 106.8(9) 121.9(7) 131.5(11) Li(1)]O(2) Li(1)]O(12) Li(2)]O(4) Li(2)]O(7) Li(2)]O(6) Li(2)]O(5) Li(2)]O(8) Li(2)]O(3) O(3)]C(1) W(1A)]O(3)]Li(2) W(1)]O(3)]Li(2) C(3)]O(4)]C(2) C(3)]O(4)]Li(2) C(2)]O(4)]Li(2) C(4)]O(5)]W(1) C(4)]O(5)]Li(2) W(1)]O(5)]Li(2) C(6)]O(6)]C(5) C(6)]O(6)]Li(2) C(5)]O(6)]Li(2) C(8)]O(8)]Li(2) C(11)]O(9)]C(12) C(11)]O(9)]Li(1) C(12)]O(9)]Li(1) C(10)]O(10)]Li(1) C(14)]O(11)]Li(1) C(15)]O(11)]Li(1) C(13)]O(12)]Li(1) W(2)]O(15)]W(4) W(4)]O(22)]W(2) 2.09(2) 2.10(2) 1.91(2) 1.98(2) 2.05(2) 2.10(2) 2.19(2) 2.64(2) 1.458(11) 91.4(5) 92.6(5) 114.4(10) 125.0(11) 108.8(9) 128.5(8) 104.9(9) 118.5(6) 113.8(11) 122.5(11) 110.7(10) 109.8(9) 116.2(12) 113.9(11) 123.7(12) 110.0(12) 114.9(10) 124.4(11) 105.6(10) 117.3(4) 90.11(3) * Letters A and B denote the atoms derived from the reference atoms by the symmetry operations 2x 2 1, 2y 1 1, 2z and 2x, 2y, 2z 2 1 respectively.the predominantly covalent character of its bonding with the neighbouring atoms. In the structure of compound 7, [{LiWO2(OC2H4OMe)3}2? 2Li(HOC2H4OMe)2]21[W6O19]22, one neutral centrosymmetric {Li2W2O4(OR)6} and two cationic {Li(ROH)2}1 fragments are bound by W]] O]Li bonds [Li(1)]O(2) 2.09(2) Å] into hexanuclear cationic aggregates of [{Li2W2O4(OR)6}{Li(ROH)2}2]21 composition, R = OC2H4OMe (Fig. 3, Table 5), which are linked via O(10)]H(10) ? ? ? O(10) (2x, 1 2 y, 2z) hydrogen bonds into infinite zigzag chains stretching along the a axis [O(10) ? ? ? O(10) 2.75(1) Å].There are also two more rather long contacts involving the H(12) atom, one of them corresponding to the intramolecular [O(12) ? ? ? O(1) 3.02(1) Å] and another Fig. 3 Structure of the cationic aggregates [{LiWO2(OC2H4OMe)3}2? 2Li(HOC2H4OMe)2]21 in the crystal of compound 7. The atoms derived from the corresponding reference atoms by x 2 1, y, z and 2x, 1 2 y, 2z symmetry operations are primed and doubly primed respectively; those that are related to the corresponding atoms by inversion are denoted by letter A corresponding to the intermolecular [O(12) ? ? ? O(100) 3.12(1) Å] hydrogen bonds.The [W6O19]22 ion has a centrosymmetric structure (Fig. 4), tungsten atoms forming an octahedral {W6} framework, centered with the m6-oxygen atom O(22) and made up of six edgesharing WO6 octahedra. The W]m6-O distances [W(2)]O(22) 2.323(1), W(3)]O(22) 2.327(1), W(4)]O(22) 2.317(1) Å] and overall geometry of the anion are in good agreement with Fig. 4 Structure of the [W6O19]22 anion in the crystal of compound 7. Letters A and B denote atoms derived from the reference atoms by the symmetry operations 2x 2 1, 2y 1 1, 2z and 2x, 2y, 2z 2 1 respectivelyJ. Chem. Soc., Dalton Trans., 1998, Pages 21–29 27 Scheme 1 Mo electrolyte [MoO(OMe)4] MeOH LiCl concentration [LiMo2O2(OMe)7(MeOH)] 1 + electrolyte evaporation to dryness viscous liquid extraction with hexane [MoO(OEt)4] (extract) O2 [MoO(OPri)4](in solution) + [Mo6O10(OPri)12] [LiMo2O4(OEt)5(EtOH)] 2 (in residue) [LiMo2O4(OPri)5(PriOH)] 3 [LiMo2O4(OPri)4(OC2H4OMe)] 4 PriOH (excess) 1-2 equivalents MeOC2H4OH [LiMoO2(OC2H4OMe)3] 5 + [MoO2(OC2H4OMe)2] W electrolyte MeOH LiCl evaporation to dryness solid extraction with hexane W(OMe)6 + [WO(OMe)4] electrolyte evaporation to dryness glass-like solid excess of MeOC2H4OH [{LiWO2(OC2H4OMe)3}2• 2Li(MeOC2H4OH)2]2+ [W6O19]2– 7 [WO(OEt)4 O2 EtOH LiCl EtOH LiCl PriOH LiCl excess MeOC2H4OH those found in the structure of [NBu4]2[W6O19] 28 [W]m6-O 2.311(1)–2.328(1) Å].The lengths of terminal W]] O bonds (1.695–1.708 Å) in the structure of the anion of compound 7 are as usual shorter than those of W]m-O bonds (1.899–1.929 Å) and W]m6-O bonds (2.317–2.327 Å). The W]] O and W]m-O bond lengths are similar to the statistical average values for such bonds.29 The most important feature of the structure of the [{Li2W2O4(OR)6}{Li(ROH)2}2]21 cationic aggregate is the W]] O]Li bond between the neutral and the cationic fragments. The length of the corresponding bond W(1)]O(2) [1.720(8) Å] practically coincides in this case with that of W(1)]O(1) [1.725(8) Å], involving the O(1) atom participating in a moderately strong hydrogen O(1) ? ? ? O(10) bond.The average W]] O bond length (1.722 Å) for the dicationic fragment of 1 is smaller than that found in WO2 groups of binuclear anions [W2O6L9]52 {H5L9 = MeC(OH)[PO(OH)2]2} 30 (1.745 and 1.739 Å), [W2O5L02]62 [ H4L0 = HOC(CH2CO2H)2CO2H] 31 (1.753 and 1.762 Å), [W2O5(OMe)4]2232 (average value 1.739 Å).It is important to note that the tungsten atoms in all these compounds are co-ordinated exclusively by oxygen atoms. This should be taken into consideration for correct comparison of terminal W]] O bond lengths because the different redistributions of the electron density within the WO6 co-ordination octahedra of the W atom in various compounds lead to considerable variations in the lengths of bonds which should formally be considered as having the same structural functions.Thus, for example, in the structures of monomeric octahedral complexes [WO2Cl2(OPPh3)2],33 [WO2Cl2(OMe2)] 34 and [WO2- F2(bipy)] (bipy = 2,29-bipyridine) 35 the W]] O bond lengths in WO2 groups are significantly shortened being equal to 1.704, 1.67 and 1.667 Å. At the same time, the W]] O bond length in a ‘wolframyl’ group is to a certain extent dependent upon the nature of other bonds in the co-ordination polyhedron of the tungsten atom.Thus in ref. 29 the elongation of W]] O bonds in [W2O6L9]52 was explained by formation of a pseudo-trioxo group WO3 (taking into account that one of the W]m-O bridging bonds is strongly shortened in comparison with the other analogous bridging bonds equal to 1.828 Å), as the W]O bonds should logically be longer in WO3 groups than in WO2 due to delocalization of p bonding over three bonds (rather than two bonds as in WO2).The co-ordination polyhedron of Li(1) is a trigonal bipyramid, its equatorial plane being formed by O(2), O(10) and O(11) and apical positions occupied by atoms O(9) and O(12) [O(9)]Li(1)]O(12) 169.5(12)8]. The Li(1)]O(2) bond length [2.09(2) Å], responsible for the formation of the cationic aggregate, as well as other Li]O bond lengths (2.00–2.10 Å) are close to the sum of ionic radii for five-co-ordinated Li1 cation and two-co-ordinate O22 anion (2.04 Å 12). In contrast to the co-ordination polyhedron of Li(1) with long bonds indicating their predominantly ionic character, in the Li(2) co-ordination polyhedron there is one considerably shortened Li(2)]O(4) bond [1.91(2) Å].The co-ordination polyhedron can hardly be associated with any of the regular geometries known for six-co-ordinated atoms: the minimum and maximum bond angles are 67.0(6) and 163.3(11)8 respectively (see Table 5).This is apparently due to both very mild coordination requirements of Li and the steric tension arising on formation of the chelate rings. The structure of the {Li2M2O6} core of the neutral bimetallic [Li2W2O4(OR)6] fragment belongs to the [{Ti(OR)4}4] type (R = Me or Et36) quite frequently found in the metal alkoxide structures. The core formed by atoms Li(2), Li(2A), W(1), W(1A), O(3), O(3A), O(5), O(5A), O(7), O(7A) is not planar: the central W(1)O(3)W(1A)O(3A) ring is rotated by 62.7(3)8 relative to the W(1)Li(2)W(1A)Li(2A) plane.Among the known structures of the bimetallic alkoxides the analogous metal–oxygen frameworks have been observed for NaMO2(OC2H4OMe) 3, M = Mo10 or W,14 where the dimeric molecules were also built up of two octahedra sharing a common edge. In contrast to compound 7 where the W]O]W bridges are nearly symmetric (W]O 2.140 and 2.179 Å), the Mo]O]Mo bridges in [NaMoO2(OC2H4OMe)3] (Mo]O 2.255 and 2.162 Å) are significantly asymmetric. The apparent analogy in structure and composition of the considered fragment and the known bimetallic 2-methoxyethoxides allows to suggest that the complex 5 possesses an analogous molecular structure.Conditions of formation and isolation of the side products of electrosynthesis The present study showed cathodic reduction to be a general characteristic feature of electrosynthesis (see Scheme 1), having a more pronounced effect in case of tungsten than of molybdenum.Its extent increases if the same metal is used as cathode28 J. Chem. Soc., Dalton Trans., 1998, Pages 21–29 and at higher concentrations of conductive additive. The product obtained in the anodic oxidation of molybdenum in PriOH revealed the highest stability to oxidation. Its IR spectrum and chemical properties testified to its identity with [Mo6O10(OPri) 12] 6, earlier obtained by Chisholm et al.9 by oxidation of Mo2(OPri)8 with molecular oxygen. This was quite surprising taking into account the usual stability of methoxides and the reluctance of methanol to participate in side reactions that in the case of molybdenum dissolution in methanol resulted in isolation of a comparably stable compound 1, which proved to be a bimetallic compound of lithium and MoV formally derived from the MoO(OR)3 series, which has not been described in the literature as far as we know.It is interesting that the product of cathodic reduction of tungsten methoxide in MeOH is immediately oxidized on attempted isolation and the resulting mixture of products consists of W(OMe)6 and [WO(OMe)4].We failed, unfortunately, to identify the bimetallic complexes formed in situ on dissolution of metals in EtOH because of the extremely strong trend of ethoxo-molybdates and -tungstates(V) to oxidation. Therefore, the products of their reaction with molecular oxygen were analysed. In the case of molybdenum, after extraction of [MoO(OEt)4], from the residue the product to which the composition [LiMo2O4(OEt)5(EtOH)] 2 was ascribed was practically quantitatively crystallized (unfortunately, twinning problems hindered X-ray single-crystal studies).The composition of 2 was confirmed indirectly by the results of an X-ray study of the product of its treatment with an excess of PriOH and controlled amounts of 2-methoxyethanol, equation (4). [LiMo2O4(OEt)5(EtOH)] PriOH (excess) 2 [LiMo2O4(OPri)6(PriOH)] 1–2 equivalents MeOC2H4OH 3 [LiMo2O4(OPri)4(OC2H4OMe)] (4) 4 Compound 4, easily crystallizable from hexane, results from substitution of one of the OPri groups in 3 with a 2-methoxyethoxide group. The IR spectra of 2–4, and those of [NaMo2- O4(OR)5(ROH)] (R = Et or Pri) 7,13 are similar to each other in the n(M]O) region, which also testifies to the analogy in their structures.It is interesting that the complete substitution of OR groups in compound 2 with 2-methoxyethoxide ones leads to an equimolar mixture of [MoO2(OC2H4OMe)2] and [LiMoO2- (OC2H4OMe)3] 5, equation (5).The isolated complex 5 can also [LiMo2O4(OEt)5(EtOH)] 1 MeOC2H4OH 2EtOH (excess) [MoO2(OC2H4OMe)2] 1 [LiMoO2(OC2H4OMe)3] (5) 5 be prepared by direct interaction of 2-methoxyethoxides of lithium and molybdenum in 2-methoxyethanol, equation (6). [MoO2(OC2H4OMe)2] 1 LiOC2H4OMe ROH [LiMoO2(OC2H4OMe)3] (6) Its IR spectrum is very similar to that provided for [NaMO2- (OC2H4OMe)3] (M = Mo10 or W 14) thus indicating the analogy in their structure.§ Compound 5 turned out to be rather less § It should be mentioned that the metal–oxygen core analogous to those of Na]Mo(W) 2-methoxyethoxides was also observed in the structure of a Li]W 2-methoxyethoxide cation (see above).stable in comparison with the sodium derivative: it can be present unchanged in solution only in the presence of an excess of [MoO2(OC2H4OMe)2]. If dissolved in pure ROH, or in a solution containing LiOR (R = C2H4OMe), it undergoes complete decomposition leading to formation of a mixture of Li2MoO4 and Li2Mo2O7.In contrast to the situation observed for molybdenum, the anodic dissolution of tungsten in EtOH leads to two main products, crystalline [WO(OEt)4] and glass-like [WOn(OEt)622n], where n > 2. The latter is also contaminated by lithium. On the reaction of an excess of 2-methoxyethanol with the residue upon evacuation of the electrolyte [{LiWO2(OC2H4OMe)3}2? 2Li(MeOC2H4OH)2]21[W6O19]22 7 was obtained in high yield (>50% in relation to the amount of metal consumed).So the considerable decomposition with formation of oxo complexes is presumably due to the higher extent of cathodic reduction and easier oxidation of the reduced products for tungsten alkoxides in comparison with molybdenum ones. As the most important result of the present work we consider the identification of bimetallic lithium derivatives among the side products of electrosynthesis.None of the isolated complexes contained chlorine atoms, i.e. they were formed due to complexation between molybdenum (or tungsten) alkoxides and LiOR. The latter originated from the following electrochemical reaction (7) analogous to that known for water LiCl 1 ROH LiOR 1 ��� H2­ 1 Cl? (7) solutions. Oxidation of alcohols by released chlorine radicals should lead, for example, to reactions (8) and (9). A GLC–mass EtOH 1 2Cl? MeCHO 1 2HCl (8) EtOH 1 HCl EtCl 1 H2O (9) spectrometric study of methanol-based electrolytes has not indicated the presence of such derivatives (presumably because of their high volatility) but in the case of electrolytes based on EtOH and PriOH they were represented by alkyl halides and aldehydes or ketones (see Table 1).The presence in the electrolyte obtained by anodic oxidation of molybdenum in EtOH of such a specific product as the diethyl acetal of acetaldehyde is presumably due to condensation of the initially formed aldehyde with alcohol, promoted by the catalytic action of [MoO(OEt)4], equation (10). MeCHO 1 2EtOH [MoO(OEt)4] MeCH(OEt)2 1 H2O (10) The results obtained permit us also to propose a general preparation scheme for different homologues of the [MoO(OR)4] series, based on the electrochemical technique: a first step in all cases is the anodic oxidation of metal, for example, in ethanol.It should always be followed by separation of the alkoxide oxide by extraction {as the side products, i.e.bimetallic alkoxides of lithium and molybdenum, are more easily separable from the product, [MoO(OR)4], for R = Et than for R = Pri, C2H4OMe} and the final step should be the alcohol interchange reaction of the pure [MoO(OEt)4]. An analogous approach can be applied for the preparation of different homologues of the W(OR)6 series, but in this case the reagent for the alcohol-interchange reaction should be W(OMe)6, prepared by anodic oxidation of tungsten in methanol.Alcohol-interchange techniques for preparation of molybdenum and tungsten alkoxides have been described.7,10,14 Acknowledgements The authors express their gratitude to International Science Foundation, Grants MPR 000 and MPR 300, and European Program of Intellectual and Technical Assistance to the FormerJ. Chem. Soc., Dalton Trans., 1998, Pages 21–29 29 States of the Soviet Union, Grants 93-2792 and 94-771, and the Russian Foundation for Basic Research, Project 97-03-33783, and Russian Foundation for Fundamental Research for Technological Applications for financial support of this work. References 1 N.Ya. Turova, E. P. Turevskaya, V. G. Kessler and M. I. Yanovskaya, J. Sol-Gel Sci. Technol., 1994, 2, 18. 2 E. P. Kovsman, S. I. Andruseva, L. I. Slovjeva, V. I. Fedyaev, M. N. Adamova and T. V. Rogova, J. Sol-Gel Sci. Technol., 1994, 2, 61. 3 V. A. Shreider, E. P. Turevskaya, N. I. Kozlova and N.Ya. Turova, Inorg. Chim. Acta, 1981, 53, L73. 4 N. Ya. Turova and V. G. Kessler, Russ. J. Gen. Chem., 1990, 60, 113. 5 S. I. Kucheiko, N. Ya. Turova and V. A. Shreider, Russ. J. Gen. Chem., 1985, 55, 2396. 6 V. G. Kessler, A. V. Mironov, N. Ya. Turova, A. I. Yanovsky and Yu. T. Struchkov, Polyhedron, 1993, 12, 1573. 7 N. Ya. Turova, Kessler and S. I. Kucheiko, Polyhedron, 1991, 10, 2617. 8 V. G. Kessler, S. Yu. Vasilyev, A. I. Belokon’ and N. Ya. Turova, Russ.J. Gen. Chem., 1990, 60, 2629. 9 M. H. Chisholm, K. Folting, J. C. Huffman and C. C. Kirkpatrick, Inorg. Chem., 1984, 23, 1021. 10 V. G. Kessler, N. Ya. Turova, A.V. Korolev, A. I. Yanovsky and Yu. T. Struchkov, Mendeleev Commun., 1991, 89. 11 S. I. Kucheiko, N. Ya. Turova and N. I. Kozlova, Sov. J. Coord. Chem., 1985, 11, 1521, 1656. 12 SHELXTL PLUS PC, Version 4, Siemens Analytical X-Ray Instruments, Madison, WI, 1990; Version 5, Siemens Industrial Automation, Madison, WI, 1994. 13 V. G. Kessler, N. Ya. Turova, A. I. Yanovsky and Yu. T. Struchkov, Russ. J. Gen. Chem., 1990, 60, 2769. 14 V. G. Kessler, D. E. Chebukov and N. Ya. Turova, Russ. J. Inorg. Chem., 1993, 18, 23. 15 L. Ma, S. Liu, H. Zhu and J. Zubieta, Polyhedron, 1989, 8, 669. 16 C. Limberg, S. Parsons, A. J. Downs and D. J. Watkin, J. Chem. Soc., Dalton Trans., 1994, 1169. 17 V. G. Kessler, G. A. Seisenbaeva, A. V. Shevelkov and G. V. Khvorykh, J. Chem. Soc., Chem. Commun., 1995, 1779. 18 R. D. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751. 19 E. P. Turevskaya, N. Ya. Turova, A. V. Korolev, A. I. Yanovsky and Yu. T. Struchkov, Polyhedron, 1995, 14, 1531. 20 D. Eichorst, D. A. Payne, S. R. Wilson and K. E. Howard, Inorg. Chem., 1990, 29, 1458. 21 M. J. Hampden-Smith, D. S. Williams and A. L. Rheingold, Inorg. Chem., 1990, 29, 4076. 22 B. A. Vaarstra, J. C. Huffman, W. E. Streib and K. G. Caulton, J. Chem. Soc., Chem. Commun., 1990, 1750. 23 M. Veith and R. Rosler, Z. Naturforsch., Teil B, 1986, 41, 1071. 24 V. W. Day, D. A. Ebershpaher, Y. Chen, J. Hao and W. J. Klemperer, Inorg. Chim. Acta, 1995, 229, 391. 25 L. G. Hubert-Pfalzgraf and C. Sirio, personal communication. 26 A. I. Yanovsky, E. P. Turevskaya, M. I. Yanovskaya, V. G. Kessler, N. Ya. Turova, A. P. Pissarevsky and Yu. T. Struchkov, Russ. J. Inorg. Chem., 1995, 40, 355. 27 S. Boulmaaz, L. G. Hubert-Pfalzgraf, S. Halut and J.-C. Daran, J. Chem. Soc., Chem. Commun., 1994, 601. 28 A. I. Yanovsky, S. I. Kucheiko and Yu. T. Struchkov, Sov. J. Coord. Chem., 1987, 13, 694. 29 Structure Correlation, eds. H.-B. Burgi and J. D. Dunitz, VCH, Weinheim, New York, 1994, vol. 2, p. 815. 30 V. S. Sergiyenko, E. O. Tolkacheva and A. B. Ilukhim, Russ. J. Inorg. Chem., 1994, 39, 243. 31 E. Llopis, J. A. Ramirez, A. Domenech and A. Cervilla, J. Chem. Soc., Dalton Trans., 1993, 1121. 32 W. Clegg, R. J. Errington, K. A. Fraser and D. G. Richards, J. Chem. Soc., Chem. Commun., 1993, 1105. 33 J. F. Wet, M. R. Caira and B. J. Gellatly, Acta Crystallogr., Sect. B, 1978, 34, 762. 34 M. G. B. Drew, G. W. A. Fowles, D. A. Rice and K. J. Shanton, J. Chem. Soc., Chem. Commun., 1974, 64. 35 P. Schreiber, K. Weighardt, B. Nuber and J. Weiss, Z. Naturforsch., Teil B, 1990, 45, 619. 36 J. A. Ibers, Nature (London), 1963, 197, 686. Received 16th June 1997; Paper 7/04198E

ISSN:1477-9226    

DOI:10.1039/a704198e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 25–29 25 Copper(II) complexes of a novel ligand 4,5-dicyanoimidazole: structural and magnetic studies B. L. V. Prasad,a Hirohiko Sato,b Toshiaki Enoki,b Shmuel Cohen c and T. P. Radhakrishnan *a a School of Chemistry, University of Hyderabad, Hyderabad 500 046, India. E-mail: tprsc@uohyd.ernet.in; Fax: 91-40-3010120 b Department of Chemistry, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo 152, Japan c Department of Inorganic and Analytical Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904, Israel Received 20th July 1998, Accepted 30th October 1998 The potential use of 4,5-dicyanoimidazolate ion (DCI2) as a bridging ligand for magnetically interesting co-ordination polymers has been explored. New CuII–bipy complexes containing DCI2 were prepared and characterised; crystal structure analysis revealed DCI2 co-ordinated CuII in one of the complexes. The ESR and magnetic susceptibility studies showed that the compounds are nearly Curie paramagnetic or weakly antiferromagnetic.A polymeric structure is proposed for one of the complexes. Introduction Molecule-based magnetic materials is an area of immense research activity today. Based on theoretical ideas such as the negative spin density product model,1 configuration interaction in charge transfer complexes 2 and topological models,3 several experimental investigations 4–7 have been carried out to fabricate molecular magnetic materials. Extensive work carried out on high spin species based on carbene and nitrene radical centres 8 indicates that the high chemical reactivity often leads to low spin concentrations in the bulk material.One of the strategies to overcome this problem is to replace the reactive radical sites based on light atoms by metal ion spin centres. Ferromagnetic coupling between the metal spin centres can be achieved using superexchange interactions 9 or by the orthogonality of the metal ion spin orbitals through proper design of the bridging ligands.10 Magnetic materials based on transition metal co-ordination polymers with molecular bridging ligands investigated in earlier studies 11 were mostly based on one-dimensional systems.Copper(II) ions assembled in 2-D and 3-D networks have also been reported 12 taking cognisance of the necessity to go beyond one-dimensional systems for achieving phase transitions at finite temperatures.13 However these systems have shown antiferromagnetic interactions.An interesting case of ferromagnetic spin coupling between vanadyl centres mediated by pyrimidine ligands has been reported by Iwamura and coworkers. 14 The current interest in the exploration of diVerent molecular species as mediators of magnetic interactions between metal ion spins may be juxtaposed with the extensive studies of magnetic coupling units, motivated by topological models for ferromagnetism in organic radical systems.In the spirit of the simple spin coupling schemes we have developed,15 it may be envisaged that, if a molecular ligand bridging two paramagnetic metal ions provides a short odd p-electron path, it could serve as a ferromagnetic coupling unit for the metal ion spins, provided the metal ion spin orbitals are in conjunction with the ligand p system. The pyrimidine ligands mentioned above are an illustrative example. Imidazole, pyrazine, etc. are popular bridging ligands for metal ions.16 However, these ligands provide 4-p-electron pathways which probably lead to the dominant antiferromagnetic coupling of the paramagnetic metal ions observed in such systems.17 A ligand such as 4,5-dicyanoimidazole (HDCI) would be an interesting variant to investigate, since the cyano groups strongly withdraw the charge on the easily formed 4,5-dicyanoimidazolate anion bridge and diminish the electron count on the p pathway between the metal centres.Our ab initio calculations at the MP2/6-3111G** level indicate that the net charge on the N–CH–N path is 20.82 in imidazolate whereas it reduces to 20.36 in 4,5-dicyanoimidazolate anion, supporting this contention. The compound HDCI could also support 2-D metal co-ordination networks. Surprisingly there are no reports of structural or magnetic characterisation of dicyanoimidazole ligated paramagnetic metal systems. The only structural study that we are aware of, involving a derivative of DCI2 as a ligand, is that of a complex 18 of CuI wherein one of the cyano groups of HDCI is converted into an iminoester.This paper also reported a compound with a possible polymeric structure involving HDCI-bridged copper(I) ions. Earlier studies in our laboratory 19 have shown that water soluble copper(II) salts readily react with HDCI to give highly insoluble complexes. Variable temperature ESR studies showed antiferromagnetic coupling between copper(II) spins in these materials. These were amorphous and no structural information could be gathered. Structural studies on some paramagnetic complexes with DCI2 as the ligand could provide insight into its potential use in the design of co-ordination polymers with desired magnetism.We report here the synthesis and first structural characterisation of some DCI2 complexes of CuII; [Cu(bipy)2]21 (bipy = 2,29- bipyridyl) was used as the precursor to limit the DCI2 ligation facilitating the formation of monomeric or controlled polymeric systems.We also present magnetic studies which indicate that the magnetic interactions are either very weak or antiferromagnetic in the present systems. Results and discussion Initially we carried out the simple reaction of copper(II) nitrate with bipy and HDCI. The product 1 could be grown as large prismatic crystals. The IR spectrum revealed that HDCI was indeed included in this compound. However, the cyano stretch frequency at 2239 cm21 indicated that the HDCI is not deprotonated in this complex (cyano stretch appears as a doublet at 2245 and 2258 cm21 for HDCI and at ª2224 cm21 for DCI2).26 J.Chem. Soc., Dalton Trans., 1999, 25–29 Single crystal analysis revealed the molecular structure shown in Fig. 1 which resembles the structure of [Cu(dmphen)2- (O2NO)][CCl3CO2]?CCl3CO2H (dmphen = 2,9-dimethyl-1,10- phenanthroline),21 but now incorporates 4,5-dicyano-imidazole; the molecular formula is [CuII(bipy)2(NO3)]NO3?HDCI? H2O.It is clearly seen that the HDCI does not co-ordinate to the CuII, but is present because of a hydrogen bond that links its acidic H atom to the unligated NO3 2. The small shift of the cyano stretch frequency with respect to free HDCI appears to be the consequence of the hydrogen-bonding interaction. Since the initial attempt failed to achieve co-ordination of DCI2 to CuII we tried two other procedures. In the first of these methods we proceeded on the basis that a strongly co-ordinating ligand such as NO3 2 in the starting copper salt would hinder the co-ordination of HDCI.Hence we used CuSO4 as the precursor material. As reported earlier,22 a polymeric material containing the sulfate ion precipitated upon addition of bipy. However, using the remaining solution we were able successfully to prepare complex 2 wherein the HDCI is co-ordinated to CuII. The molecular structure from the single crystal analysis is shown in Fig. 2 and corresponds to the molecular formula, [CuII(bipy)2(DCI)][DCI2]?(HDCI). The Fig. 1 An ORTEP20 diagram of the molecular structure of complex 1; 50% probability thermal ellipsoids are indicated. Only one of the disordered positions of the water oxygen atom is shown. Table 1 Significant bond lengths (Å) and bond angles (8) in (a) complex 1 and (b) complex 2; standard deviations are provided in parentheses (a) Cu–O(51) Cu–O(52) Cu–N(1) O(51)–Cu–N(1) O(51)–Cu–N(2) O(51)–Cu–N(3) O(51)–Cu–N(4) N(1)–Cu–N(2) 2.078(3) 2.639(4) 1.980(4) 92.5(1) 161.5(1) 88.1(1) 90.8(1) 81.1(1) Cu–N(2) Cu–N(3) Cu–N(4) N(1)–Cu–N(3) N(1)–Cu–N(4) N(2)–Cu–N(3) N(2)–Cu–N(4) N(3)–Cu–N(4) 2.032(3) 2.008(4) 2.185(3) 176.7(1) 104.2(1) 97.4(1) 107.5(1) 79.0(1) (b) Cu–N(1) Cu–N(2) Cu–N(3) N(1)–Cu–N(2) N(1)–Cu–N(3) N(1)–Cu–N(4) N(1)–Cu–N(5) N(2)–Cu–N(3) 1.993(2) 2.043(2) 2.003(2) 80.79(7) 173.04(7) 96.08(7) 94.16(7) 95.92(7) Cu–N(4) Cu–N(5) N(2)–Cu–N(4) N(2)–Cu–N(5) N(3)–Cu–N(4) N(3)–Cu–N(5) N(4)–Cu–N(5) 2.162(2) 2.001(2) 98.83(7) 145.28(7) 78.29(7) 92.01(7) 115.88(7) significant feature is that the sulfate ion is absent in this complex and the charge balance is maintained by two DCI2, one coordinated and the other unco-ordinated to Cu21; interestingly, a neutral HDCI is also present hydrogen bonded with the latter DCI2.The IR cyano stretch frequency of 2 appears at 2226 cm21 indicating the formation of DCI2. The co-ordination around CuII in complex 1 is best described as cis-distorted octahedral 21,23 with one elongated (2.639 Å) and one normal Cu–O bond and four normal Cu–N bonds [Table 1(a)].Complex 2 has five N atoms co-ordinated to it [Table 1(b)] and the co-ordination may be described as halfway between square pyramidal and trigonal bipyramidal (t value based on the conventional description 24 of distorted five-coordinated geometries is ª46%). Examination of the packing diagrams of 1 and 2 shows that the CuII ? ? ? CuII distances are all very large, the closest being 7.1 and 8.2 Å respectively.Since the by-product during the reaction of copper salts with HDCI is the corresponding acid, in the second strategy that we adopted to get HDCI to co-ordinate to CuII we used an acid buVer (pH ª 4.0) following similar procedures reported earlier.12 We were not able to grow single crystals of the product 3. However, the cyano stretch frequency at 2227 cm21 indicates that DCI2 is formed and the stoichiometry from elemental analysis implies that DCI2 must be co-ordinated to CuII.Based on elemental analysis the molecular formula is assigned as CuII(bipy)(DCI)2?3H2O. The peculiar hemispherical beadlike morphology and the low solubility in water and organic solvents displayed by 3 are suggestive of a polymeric structure. To gain further insight into the local structure around CuII in complex 3, we have investigated the ESR spectra of all three complexes as microcrystalline solids and in acetonitrile solution.The relevant characteristics at room temperature are collected in Table 2. The powder ESR spectra of 1 and 2 are very similar with g^ > g|| ª 2.0, indicative of an axially compressed geometry and the spin residing in the dz2 orbital. The line shape and g values are similar to those found for [Cu- (bipy)2(ONO2)]NO3?H2O which has been described as weakly six-co-ordinated bicapped square pyramidal 25 as well as fiveco- ordinated distorted trigonal bipyramidal.26 The powder ESR line shape of 3 is quite diVerent with g|| > g^ indicating an axially elongated case with the dx2 2 y2 orbital bearing the spin.Though the solution ESR of 1 did not exhibit any hyperfine structure, the solution spectra of 2 and 3 showed the typical four line hyperfine pattern expected from 65Cu. This can be taken as indicative of similar co-ordination around CuII in the last two cases. We believe that as in the case of 2, DCI2 is co- Fig. 2 An ORTEP diagram of the molecular structure of complex 2; 50% probability thermal ellipsoids are indicated.J. Chem. Soc., Dalton Trans., 1999, 25–29 27 Table 2 Room temperature magnetic moment and ESR spectral data of complexes 1–3 Microcrystals Solution (CH3CN) Complex 123 m/mB 1.66 1.78 1.99 ESR Linewidth/G 145 180 205 g|| 2.027 2.014 2.190 g^ 2.167 2.172 2.065 giso 2.109 2.099 2.095 Aiso (×1024 cm21) — ca. 53 ca. 56 ordinated to CuII in 3 as well. This is further supported by the similar infrared spectra of these two complexes mentioned above.The diVerence in the ESR spectra of 2 and 3 in the microcrystalline state may arise from the diVerent exchange mechanisms operative in the solid state; the broader signals of 3 could be taken as evidence for strong exchange interactions, Fig. 3 Plots of 1/c vs. T of complexes (a) 1, (b) 2 and (c) 3; the lines are Curie–Weiss law fits. The Weiss constants are 15.4, 10.9 and 219.2 K for 1, 2 and 3 respectively (1 m3 mol21 = 4p × 1026 emu mol21). which is supported by the magnetic data presented below.The ESR spectra measured at low temperature (150 K) did not provide any further insight into the structure of the complexes. Magnetic susceptibilities of these complexes (as powder samples) were measured on a SQUID magnetometer, from 300 to 3 K at a field of 10 kG. The room temperature magnetic moments are indicated in Table 2. The data are plotted as 1/c against T in Fig. 3. It is seen that the behaviours in the case of 1 and 2 [Fig. 3(a) and 3(b) respectively] are very close to Curie paramagnetic. The data could be fit quite well to the Curie– Weiss law with Curie constants of 0.343 and 0.394 and Weiss constants of 15.4 and 10.9 K respectively. The data analysis indicates that the copper(II) ions in both 1 and 2 are coupled by very weak interactions, obviously a consequence of the large Cu ? ? ? Cu distances in these two complexes. The magnetic susceptibility of 3 also follows a Curie–Weiss law [Fig. 3(c)] with a Curie constant of 0.532 and Weiss constant of 219.2 K; the Curie constant was estimated using the molecular formula derived from elemental analysis. The larger negative Weiss constant of 3 is indicative of moderate antiferromagnetic interactions. This possibly arises from copper(II) interactions through a bridging ligand. In view of the composition, ESR and magnetic data on 3 we speculate that it has a polymeric structure involving the CuII–DCI2–CuII structural unit.It is quite possible that, due to the steric repulsions of the cyano groups on the DCI, these chains have strongly non-planar structures. Further, as the ESR data showed, the spin on CuII does not reside in a d orbital that is in conjugation with the p system of DCI2. Hence the simple topological models for spin coupling through p-electron pathways may not be realised here. Further work including elucidation of the structure of a coordination polymer would be required to extend this scheme for the fabrication of magnetic materials.Conclusion We have presented the structural characterisation of two copper(II) complexes containing 4,5-dicyanoimidazole. Ligation of DCI2 to CuII is demonstrated for the first time in one of these complexes. A third complex synthesized indicates that DCI2 may be useful as a bridging ligand for co-ordination polymers. The ESR studies and magnetic susceptibility measurements on the new complexes reveal paramagnetic or weak antiferromagnetic interactions.Experimental Syntheses The copper salts were purchased from Loba Chemie (India) whilst 4,5-dicyanoimidazole and 2,29-bipyridyl were purchased from the Aldrich Chemical Company (USA). Complex 1. The salt Cu(NO3)2?3H2O (0.10 g, 0.41 mmol) was dissolved in 5 ml of distilled water and warmed. A warm solution of 0.13 g (0.83 mmol) of bipy in 10 ml of ethanol was added. The pale blue copper nitrate turned to dark blue immediately. A hot solution of 0.05 g (0.42 mmol) HDCI in 5 ml of ethanol was added and stirred for 10 min.The solution was allowed to cool and then subjected to slow evaporation. Large blue prism-shaped crystals separated in 2–3 d. They were28 J. Chem. Soc., Dalton Trans., 1999, 25–29 filtered oV, washed with cold ethanol and water and dried thoroughly. It was found that complex 1 could be obtained also by treating HDCI with independently prepared [Cu(bipy)2]- [NO3]2 [Found (Calc.for C25H20CuN10O7): C, 47.20 (47.14); H, 3.15 (3.21); N, 22.03 (22.07)%]. IR (KBr, cm21): 3474, 3074, 2239, 1599, 1444, 1354, 1026 and 777. Complex 2. The salt CuSO4?5H2O (0.50 g, 2 mmol) was dissolved in 10 ml of water and warmed. A hot solution of 0.63 g (4 mmol) of bipy dissolved in 10 ml of distilled water was added. The pale blue precipitate which formed immediately was filtered out. Its IR spectrum matches well with the one reported 22 for polymeric [{Cu(bipy)(SO4)}n].A warm solution of 0.24 g (2 mmol) of HDCI in 5 ml of ethanol was added to the filtrate and stirred for 10 min. On cooling a crystalline precipitate formed. This product was recrystallised from an acetonitrile –water mixture to give pale blue platelets. These were filtered oV, washed with plenty of water and dried thoroughly [Found (Calc. for C35H20CuN16): C, 57.72 (57.78); H, 2.75 (2.76); N, 30.79 (30.97)%]. IR (KBr, cm21): 3109, 2226, 1601, 1442, 1109 and 769.Complex 3. The salt Cu(NO3)2?3H2O (0.10 g, 0.41 mmol) was dissolved in 5 ml of distilled water and warmed; 2 ml of a freshly prepared acetic acid–sodium acetate buVer were added. The solution immediately turned dark blue. A warm solution of 0.13 g (0.83 mmol) of bipy in 10 ml of ethanol was added followed by 0.05 g (0.42 mmol) of HDCI dissolved in 10 ml of warm ethanol. The solution was cooled to room temperature and the solvents were slowly evaporated.Pale blue hemispherical beads separated over a period of 2–3 d. The product was filtered oV, washed with water and dried thoroughly. This compound showed very poor solubility in most solvents and all eVorts to grow single crystals were unsuccessful [Found (Calc. for C20H16CuN10O3): C, 47.28 (47.33); H, 3.18 (2.66); N, 27.58 (27.02)%]. IR (KBr, cm21): 3117, 2227, 1604, 1442, 1300, 1111 and 769. Magnetic and crystallographic studies Magnetic susceptibility measurements were performed on a SQUID magnetometer (QUANTUM DESIGN MPMS5). Correction for the diamagnetism was estimated using the Pascal constants for complexes 1–3; values are 2326.52 × 1026, 2396.80 × 1026 and 2261.58 × 1026 emu mol21 respectively.The ESR studies were carried out on a JEOL JES-FE3X XTable 3 Crystallographic data for complexes 1 and 2 Chemical formula Formula weight Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 V/Å3 T/K Z m/cm21 Number of reflections measured Number of independent reflections Number of reflections with I > 3sI RR9 1 C25H20N10CuO7 636.1 Triclinic P1� 12.398(2) 13.679(3) 8.837(2) 100.77(1) 100.39(1) 103.40(2) 1392.4(8) 293 2 8.45 (Mo-Ka) 4900 4900 4249 0.051 0.066 2 C35H20N16Cu 728.3 Triclinic P1� 9.803(2) 19.666(6) 9.276(1) 101.45(2) 100.11(1) 99.36(2) 1689(1) 293 2 13.09 (Cu-Ka) 5033 5033 4550 0.034 0.054 band spectrometer employing 100 kHz modulation.Diphenylpicrylhydrazyl was used for g value calibration. Crystal structure data (Table 3) were collected on Philips PW1100 and Enraf-Nonius CAD4 computer-controlled diVractometers using graphite-monochromated Mo-Ka (l = 0.71073 Å) and Cu-Ka (l = 1.54178 Å) radiation respectively.Intensities were corrected for Lorentz-polarisation eVects. All non-hydrogen atoms were found using direct method analysis. After several cycles of refinements the positions of the hydrogen atoms were calculated and added to the refinement process. Refinement proceeded to convergence by minimising the function Sw(|Fo| 2 |Fc|)2.CCDC reference number 186/1230. Acknowledgements T. P. R. and B. L. V. P. thank the CSIR (Council for Scientific and Industrial Research), New Delhi for financial assistance and a Senior Research Fellowship respectively. B. L. V. P. thanks the DST (Department of Science and Technology), New Delhi and JSPS (Japan Society for Promotion of Science), Tokyo for the support of a visit to Tokyo Institute of Technology in 1996 under the India–Japan Cooperative Science Program.Fruitful discussions with Drs. M. V. Rajasekharan and Bhasker G. Maiya are gratefully acknowledged. References 1 H. M. McConnell, J. Chem. Phys., 1963, 39, 1910. 2 H. M. McConnell, Robert. A. Welch Found. Conf. Chem. Res., 1967, 11, 144. 3 N. Mataga, Theor. Chim. Acta, 1968, 10, 372; A. A. Ovchinnikov, Theor. Chim. Acta, 1978, 47, 297. 4 S. Chittipeddi, K. R. Cromack, J. S. Miller and A. J. Epstein, Phys. Rev. Lett., 1987, 58, 2695. 5 A. Zheludev, A. Grand, E. Ressouche, J. Schweizer, B. G. Morin, A. J. Epstein, D. A. Dixon and J. S. Miller, J. Am. Chem. Soc., 1994, 116, 7243. 6 K. Awaga and Y. Maruyama, J. Chem. Phys., 1989, 91, 2743. 7 P.-M. Allemand, K. C. Khemani, A. Koch, F. Wudl, K. Holczer, S. Donovan, G. Gruner and J. D. Thompson, Science, 1991, 253, 301. 8 H. Iwamura and N. Koga, Acc. Chem. Res., 1993, 26, 346. 9 W. E. Hatfield, ACS Symp. Ser., 1974, 5, 108. 10 D. J. Hodgson, Prog. Inorg.Chem., 1975, 19, 173; O. Kahn, J. Galy, Y. Journaux, J. Jaud and I. Morgenstern-Badarau, J. Am. Chem. Soc., 1982, 104, 2165; P. De Loth, P. Karafiloglou, J. P. Daudey and O. Kahn, J. Am. Chem. Soc., 1988, 110, 5676; O. Kahn, R. Prins, J. Reedijk and J. S. Thompson, Inorg. Chem., 1987, 26, 3557. 11 Organic and Inorganic Low Dimensional Crystalline Materials, eds. P. Delhaes and M. Drillon, Reidel, Dordrecht, 1987; W. E. Hatfield, W. E. Estes, W. E. Marsh, M. W. Pickens, L. W.ter Haar and R. R. Weller, in Extended Linear Chain Compounds, ed. J. S. Miller, Plenum, New York, 1982, vol. 3, p. 43; R. D. Willet, R. M. Gaura and C. P. Landee, in Extended Linear Chain Compounds, ed. J. S. Miller, Plenum, New York, 1982, vol. 3, p. 143. 12 S. Kawata, S. Kitagawa, M. Kondo, I. Furuchi and M. Munakata, Angew. Chem., Int. Ed. Engl., 1994, 33, 1759. 13 F. Palacio, in Magnetic Molecular Materials, eds. D. Gatteschi, O. Kahn, J. S. Miller and F. Palacio, NATO ASI Series 198, Kluwer, Dordrecht, 1991, p. 1; H. Iwamura, K. Inoue and T. Hayamizu, Pure Appl. Chem., 1996, 68, 243; H. O. Stumpf, Y. Pei, O. Kahn, J. Sletten and J. P. Renard, J. Am. Chem. Soc., 1993, 115, 6738. 14 S. Mitsubori, T. Ishida, T. Nogami, H. Iwamura, N. Takeda and M. Ishikawa, Chem. Lett., 1994, 685; S. Mitsubori, T. Ishida, T. Nogami and H. Iwamura, Chem. Lett., 1994, 285. 15 T. P. Radhakrishnan, Chem. Phys. Lett., 1991, 181, 455. 16 J. Avrey, in Understanding Molecular Properties, ed. J. P. Dahl, Reidel, Dordrecht, 1987, p. 187; P. G. Rasmussen, J. B. Kolowich and J. C. Bayón, J. Am. Chem. Soc., 1988, 110, 7042; G. Kolks, S. J. Lippard, J. V. Waszczak and H. R. Lilienthal, J. Am. Chem. Soc., 1982, 104, 717; M. Z. Wan, H. Q. Wei, T. W. Xia, L. S. Xiaung, W. Z. Min and H. J. Ling, Polyhedron, 1996, 15, 321; P. G. Rasmussen and J. E. Anderson, Polyhedron, 1983, 2, 547; M. S. Haddad, D. N. Hendrickson, J. P. Cannady, R. S. Drago and D. S. Bieksza, J. Am. Chem. Soc., 1979, 101, 898.J. Chem. Soc., Dalton Trans., 1999, 25–29 29 17 H. W. Richardson and W. E. Hatfield, J. Am. Chem. Soc., 1976, 98, 835. 18 P. G. Rasmussen, L. Rongguang, W. M. Butler and J. C. Bayón, Inorg. Chim. Acta, 1986, 118, 7. 19 Sathya Prasanna, Ph.D. Thesis, University of Hyderabad, 1996. 20 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 21 M. van Meerssche, G. Germain, J. P. Declercq and L. Wilputte- Steinert, Cryst. Struct. Commun., 1981, 10, 47. 22 W. R. McWhinnie, J. Inorg. Nucl. Chem., 1964, 26, 21. 23 B. J. Hathaway, in Comprehensive Coordination Chemistry, eds. G. Wilkinson, R. D. Gillard and J. A. McCleverty, Pergamon, Oxford, 1987, vol. 5, p. 612. 24 B. J. Hathaway, in Comprehensive Coordination Chemistry, eds. G. Wilkinson, R. D. Gillard and J. A. McCleverty, Pergamon, Oxford, 1987, vol. 5, p. 608. 25 R. J. Fereday, P. Hodgson, S. Tyagi and B. J. Hathaway, J. Chem. Soc., Dalton Trans., 1981, 2070. 26 H. Nakai, Bull. Chem. Soc. Jpn., 1980, 53, 1321. Paper 8/05

ISSN:1477-9226    

DOI:10.1039/a805623d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 31–41 31 Further attempts to rationalise the co-ordination chemistry of manganese with SchiV base ligands and supplementary carboxylate donors Michael Watkinson,a Matilde Fondo,b Manuel R. Bermejo,*b Antonio Sousa,*b Charles A. McAuliVe,*c Robin G. Pritchard,c Nongnuj Jaiboon,c Nadeem Aurangzeb c and Mohammed Naeemc a Department of Chemistry, Queen Mary and Westfield College, University of London, London, UK E1 4NS b Departamento de Química Inorgánica, Facultad de Química, Universidad de Santiago, 15706 Santiago de Compostela, Galicia, Spain.E-mail: qiansoal@usc.es c Department of Chemistry, University of Manchester Institute of Science and Technology (UMIST), Sackville Street, Manchester, UK M60 1QD Received 17th July 1998, Accepted 27th October 1998 Some manganese(III) complexes of SchiV base ligands with ancillary carboxylate donors have been found to exhibit structural diversity, although some patterns emerged.Thus, when the ligands 3CH3O-salen and 3CH3O-salpn [3CH3O-salen = dianion of N,N9-bis(3-methoxysalicylidene)ethane-1,2-diamine, 3CH3O-salpn = dianion of N,N9- bis(3-methoxysalicylidene)propane-1,3-diamine] are used in conjunction with carboxylates RCO2 2 (R = Me, Et, Prn or CH2Ph) unidentate carboxylate bonding occurs as in the crystallographically observed [Mn(3CH3O-salen)(O2CMe)- (H2O)]?2H2O 1, [Mn(3CH3O-salen)(O2CCH2Ph)(H2O)]?H2O 4, and [Mn(3CH3O-salpn)(O2CCH2Ph)(H2O)] 5.On the other hand, employing the more sterically encumbered carboxylates ButCO2 2 and PriCO2 2, bidentate chelating binding of the carboxylate occurs, as in [Mn(3CH3O-salpn)(O2CBut)] 2 and [Mn(3CH3O-salpn)(O2CPri)] 3. The reactivity of [Mn(salpn)(acac)] (acac = acetylacetonate) with Me3SiCl and aliphatic carboxylic acids, RCO2H (R = Me, Et, Prn, Bun, Pri or But), has also been investigated. A dimer species [{Mn(salpn)Cl}2]?CH3CN 6 was isolated from the reaction of Me3SiCl with [Mn(salpn)(acac)] while the carboxylic acids seem to lead to the isolation of monomers, such as [Mn(salpn)(O2CPri)] 7.This synthetic route has also been applied to the preparation of related complexes, in which the manganese(III) centre is not attainable with reliability via the aerobic oxidation of a manganese(II) precursor, although some rare examples have been obtained by the latter method, such as [Mn(5NO2-salen)(O2CMe)(H2O)], 8, [5NO2-salen = dianion of N,N9-bis(5-nitrosalicylidene)ethane-1,2-diamine]). Single crystals were grown from a dimethylformamide solution of the material of stoichiometry Mn(3Br,5NO2-salpn)(O2CMe)?H2O [3Br,5NO2- salpn = dianion of N,N9-bis(3-bromo-5-nitrosalicylidene)propane-1,3-diamine], isolated from this route and found to consist of the unexpected [{Mn(m-3Br,5NO2-salpn)(m-O)}2]?3DMF 9, apparently containing a manganese(IV) species, in spite of the electron withdrawing nature of the substituents on the aromatic rings of the ligand.Introduction Much of the recent interest in the co-ordination chemistry of manganese has been driven by the involvement of manganese in a number of biological systems.1–5 Owing to the strong experimental evidence associated with the presence of a multinuclear cluster of manganese ions within the Oxygen Evolving Complex (OEC) of Photosystem II, and the intrinsic interest in such compounds, inorganic chemists have been enticed into the area, resulting in the preparation of a great number of extremely elegant biomimetic compounds.6 Such complexes have not been limited to multinuclear cluster complexes, since enzymes such as manganese catalase, ribonucleotide reductase and the xylose isomerases involve two manganese ions which are thought to be associated in a dimer.Thus, great interest has also surrounded the preparation of dinuclear complexes 7 in addition to the higher nuclearity cluster compounds. Dinuclear units in which the manganese ions are linked by two or three ions that include O22 in a bis(m-oxo) bridge 8 or O22 together with two carboxylates, 9 alkoxides or a combination of alkoxide/phenoxide and carboxylate donors are widely observed and have been well characterised.10 Our recent investigations in this area have centred on the preparation of manganese complexes of tetradentate (N2O2 donor set) SchiV base ligands with ancillary carboxylate ligands to see whether diVerent structural behaviour in model complexes results through the variation of the carboxylate used, in order to model the aspartate and glutamate side chains of proteins.We have found that changing these carboxylate donors can have a profound eVect on the structural chemistry in the resultant complexes. When acetate was employed with the salpn ligand the polymeric complex [{Mn(salpn)(O2CMe)]2- (H2O)3}n] was isolated.11 Unlike the other rare examples of such polymeric compounds12 in which the ligands stack on top of one another, there was twist of 908 between adjacent ligands in the polymeric backbone.This results in the polymer having a dimeric, rather than the more usual monomeric, repeat unit. Consequently we extended our investigations to other linear chain carboxylates, butyrate (O2CPrn) and valerate (O2CBun), and found that a diverse range of structural chemistry was available.13 With butyrate and the salen ligand a dinuclear species was characterised, [Mn2(salen)2(O2CPrn)(H2O)(EtOH)]- [O2CPrn]; an apparent ‘snapshot’ of polymer formation (the non-co-ordinated carboxylate, when viewed in the crystal packing diagram, appears ideally positioned to bind to the manganese centres and hence allow polymer formation to occur) which, on the basis of cyclic voltammetry studies, maintains its structural integrity in acetonitrile solution.In complete32 J. Chem. Soc., Dalton Trans., 1999, 31–41 contrast, when the valerate counter ion was used the monomeric species [Mn(salpn)(O2CBun)] was isolated.We have since isolated other monomeric species, together with dimeric and polymeric species, by using other linear, branched chain and aromatic carboxylates. Interestingly, the polymeric species we have isolated can contain either monomeric or dimeric repeat units together with ligation of the carboxylate in the anti-anti or syn-anti mode.14 Despite the large number of species we have now structurally characterised, we have been unable, as yet, to establish any pattern in the structural chemistry obtained.We were therefore keen to rationalise this behaviour through the structural characterisation of additional compounds of this type and to establish whether we could prepare a specific structural motif to order. An obvious means of achieving this goal was by attempting to replace a ligand already co-ordinated to a manganese centre whilst maintaining the geometric integrity of that centre.An ideal candidate appeared to be the monomeric species [Mn(salpn)(acac)] 15 and herein we report our findings. Results and discussion Preparation of complexes of 3CH3O-salen and 3CH3O-salpn with auxiliary carboxylate ligands All of the complexes were prepared as previously reported 11–14 by the reaction of the ligand with a stoichiometric quantity of the appropriate manganese(II) carboxylate salt in ethanol under aerobic conditions, yielding complexes of stoichiometry MnL(O2CR) (see Table 1).All complexes show characteristic shifts in their infrared spectra to lower frequency of n(C]] N), cf. ‘free’ ligand values, upon co-ordination to manganese, which together with parent ion peaks in their FAB mass spectra corresponding to [MnL]1 and room temperature magnetic moments typical of such compounds 14 indicate complete coordination of the SchiV base ligand. Unfortunately little other information can be deduced without resorting to single crystal X-ray diVraction. In view of our previous interesting findings and our desire to rationalise the structural chemistry in these species further we sought crystals suitable for X-ray diVraction studies.Crystal structure of [Mn(3CH3O-salen)(O2CMe)(H2O)]? 2H2O 1 Previously we had proposed that complexes prepared by this method containing the 3CH3O-salen ligand and aliphatic carboxylate anions were isostructural.14 This was based on a comparison of a variety of behaviour, in particular the almost identical fragmentation patterns observed by FAB mass spectroscopy; such a phenomenon has not been observed by us in any other series of these compounds.The crystallographic characterisation of the monomeric species, 1, indicates that although the co-ordination mode of the carboxylate remains as the unusual unidentate mode when this ligand is used, a slight variation in structure is possible, Fig. 1. The structure is found to consist of a monomeric manganese(III) centre which is ligated in the xy plane by the tetradentate SchiV base ligand.The axial co-ordination sites of manganese are filled by a water molecule and a unidentate acetate ligand. This mode of co- N OH HO N R1 R2 R1 R2 (CH2) n H23CH3O-salen: n = 2, R1 = CH3O, R2 = H H23CH3O-salpn: n = 3, R1 = CH3O, R2 = H H25NO2-salpn: n = 3, R1 = H, R2 = NO2 H23Br,5NO2-salpn: n = 3, R1 = Br, R2 = NO2 ordination of the carboxylate anion was also observed in both of our previous examples of these materials, [{Mn(3CH3Osalen)( O2CEtn)}2] and [{Mn(3CH3O-salen)(O2CBun)}2],14 and this perhaps accounts for the unusual similarity in the fragmentation pattern observed in the FAB mass spectra of these complexes.Unlike the previous two structurally characterised complexes, in which the co-ordination sphere was completed by weak m-phenoxy interactions [cf. Mn–O(1*) = 2.559 and 2.771 Å in the propionate and valerate, respectively], the manganese co-ordination sphere in 1 is completed by a water molecule [Mn–O(1w) 2.334(4) Å].In all other respects the structures are identical and it is perhaps not surprising that we were previously unable to distinguish between these structural motifs. Such subtle diVerences in solid state structure are also observed in [Mn(salpn)Cl] (see below) and it seems possible that the monomeric versus dimeric structure is aVected by the solvent used for crystallisation. It is important to stress though that, with the exception of iPr and tBu (see below), the carboxylate is always co-ordinated in the unidentate mode when a methoxy group is ortho to the phenolic oxygen donor.All other bond lengths and angles are typical of such complexes. Crystal structures of [Mn(3CH3O-salpn)(O2CBut)] 2 and [Mn(3CH3O-salpn)(O2CPri)] 3 Both complexes 2 and 3 were found to exist as monomeric species in which the SchiV base ligand aVords cis sites at the manganese(III) centre for bidentate chelation of the carboxylate ligands.They consist (Fig. 2) of extremely distorted octahedral centres, typical in such complexes.13,14 The asymmetric nature of the Mn–Nimine and Mn–Ocarboxylate bond lengths is particularly striking, Mn–N(1) 2.150(5) and 2.150(13), Mn–N(2) 1.986(5) and 1.939(15), Mn–O(5) 2.329(5) and 2.372(12) Å; Mn–O(6) 2.071(4) and 2.026(13) Å for 2 and 3, respectively. The distortion in the octahedral centres is further exemplified by the following angles for 2 and 3, respectively: O(2)–Mn(1)–O(6) 155.5(2) and 154.8(5), O(2)–Mn(1)–N(1) 110.1(2) and 113.3(5), O(6)–Mn(1)–O(5) 58.7(2) and 58.5(5) and N(1)–Mn(1)–O(5) 153.1(2) and 150.1(5)8.Such co-ordination is very rare in these compounds with only one example of such a monomeric SchiV base complex in the Cambridge Crystallographic Database,13 whilst two further examples have also recently been reported by us.14 The presence of cis sites at the metal centre for chelation of the carboxylate results from the flexibility in the trimethylene backbone between the imine nitrogens of the SchiV base ligand; however, such ’flexing’ of the ligand backbone is not guaranteed.It is apparent that when sterically demanding carboxylates are employed, such as trimethylacetate and isobutyrate, monomers such as 2 and 3 will prevail, as only monomeric species have been observed with these carboxylates. It therefore seems reasonable to propose that with these sterically demanding carboxylates and a tetradentate N2O2 donor set SchiV base ligand with a trimethylene backbone a monomeric complex will result.Nevertheless, steric congestion cannot be the sole reason for the ligand to adopt the strained cis conformation, as we have also observed monomers with valerate (O2CBun) 13 and propionate (O2CEt)14 anions; the former is the exclusive product whilst a polymeric species, with trans sites ligated by an anti-anti bridging carboxylates, is also observed for the latter.Crystal structures of [Mn(3CH3O-salen)(O2CCH2Ph)(H2O)]? H2O 4 and [Mn(3CH3O-salpn)(O2CCH2Ph)(H2O)] 5 The structural characterisation of complex 4, Fig. 3(a), serves to substantiate further our proviso that only unidentate coordination of a carboxylate ligand should be observed with the 3CH3O-salen ligand. As in 1 the co-ordination sphere of the octahedral manganese centre consists of the tetradentate SchiV base ligand in the xy plane, with the axial sites filled with the unidentate carboxylate O(5) and a water molecule O(1w).AJ. Chem. Soc., Dalton Trans., 1999, 31–41 33 Table 1 Analytical and some selected data for the manganese(III) carboxylate complexes containing 3CH3O-salen and 3CH3O-salpn as ligands Analysis (%) a IR/cm21 Complex [Mn(3CH3O-salen)(O2CPh)]?H2O [Mn(3CH3O-salen)(O2CCH2Ph)]?2H2O [Mn(3CH3O-salpn)(O2CMe)] [Mn(3CH3O-salpn)(O2CEt)] [Mn(3CH3O-salpn)(O2CPrn)] [Mn(3CH3O-salpn)(O2CBun)] [Mn(3CH3O-salpn)(O2CPri)] [Mn(3CH3O-salpn)(O2CBut)] [Mn(3CH3O-salpn)(O2CPh)] [Mn(3CH3O-salpn)(O2CCH2Ph)] C 57.1 (57.7) 56.9 (56.5) 55.2 (55.5) 56.1 (56.4) 56.9 (57.3) 57.7 (58.1) 57.4 (57.3) 58.3 (58.1) 60.5 (60.5) 60.5 (61.1) H 4.2 (4.8) 5.0 (5.3) 4.8 (5.1) 5.2 (5.4) 5.1 (5.6) 6.0 (5.9) 5.6 (5.6) 6.3 (5.9) 4.8 (4.9) 4.8 (5.1) N 5.3 (5.4) 5.1 (5.1) 6.1 (6.2) 6.1 (6.0) 5.8 (5.8) 5.5 (5.6) 5.7 (5.8) 5.4 (5.6) 5.3 (5.4) 5.1 (5.3) Mn 11.8 (12.1) 11.6 (11.7) 10.9 (11.4) 11.6 (11.1) n(C]] N) 1625 1624 1616 1616 1616 1616 1617 1614 1613 1616 meff /mB 4.8 4.8 4.8 4.6 4.4 4.7 4.8 4.6 4.9 4.9 FABb m/z 381 381 395 395 395 395 395 395 395 395 Yield c (%) 95 82 86 86 71 68 76 94 78 78 a Found (calculated).b Peaks corresponding to [ML]1. c Based on manganese. further water molecule of crystallisation O(2w) is hydrogen bonded to O(6). The structural characterisation of 5, Fig. 3(b), as a related monomer was somewhat surprising. We have only structurally characterised one material containing this carboxylate, which was polymeric,14 where the carboxylate bridged planar manganese SchiV base centres are in the syn-anti mode.The reasons for this bridging mode lay in the steric demands associated with the carboxylate, and we therefore expected 5 to be either a monomeric complex as observed for 2 and 3, typical of sterically demanding carboxylates, or a polymeric complex as we have previously reported.14 The isolation of a monomeric species in which the carboxylate is unidentate appears to indicate that the governing factor when the carboxylate is of an intermediate steric nature is the ortho-methoxy group, and that provided tBu and iPr groups are avoided on the carboxylate the unidentate mode of binding will always prevail.Comparison of the bond lengths about the metal centres in complexes 1, 4 and 5 reveals some interesting structural diVerences between the three monomers. In all three structures the Mn–Ocarboxylate bond lengths are identical (within their e.s.d.s): Mn–O(5) 2.130(3), 1; Mn(1)–O(5) 2.135(10), 4; and Mn(1)– O(5) 2.139(7) Å, 5.Perhaps more interesting are the diVerences in bond lengths between the manganese ions and the donor atoms of the SchiV base ligands. For 1 and 4 these bond lengths are virtually identical when viewed within the extent of the data limits, with the exception of Mn–N(2) which shows a slight lengthening in 1. For 5 it is immediately obvious that all these bond lengths are longer than those observed for 1 and 4: Fig. 1 Molecular structure of [Mn(3CH3O-salen)(O2CMe)(H2O)]? 2H2O 1. Selected bond lengths (Å) and angles (8) with estimated standard deviations (e.s.d.s) in parentheses: Mn–O(1) 1.891(3), Mn–O(2) 1.885(3), Mn–N(1) 1.988(4), Mn–N(2) 1.986(4), Mn–O(5) 2.130(3) and Mn–O(1w) 2.334(4); O(1)–Mn–O(2) 95.36(13), O(1)–Mn–N(1) 91.0(2), O(1)–Mn–N(2) 171.6(2), O(1)–Mn–O(5) 96.96(14), O(1)–Mn–O(1w) 89.0(2), O(2)–Mn–N(1) 172.9(2), O(2)–Mn–N(2) 91.2(2), O(2)–Mn– O(5) 93.69(13), O(2)–Mn–O(1w) 90.6(2), O(5)–Mn–O(1w) 172.28(14), N(1)–Mn–N(2) 82.2(2), N(1)–Mn–O(5) 88.7(2), N(1)–Mn–O(1w) 86.3(2), N(2)–Mn–O(5) 87.8(2) and N(2)–Mn–O(1w) 85.7(2).Mn(1)–O(1) 1.912(6), Mn(1)–N(1) 2.026(8), Mn(1)–N(2) 2.050(9) Å for 5, cf. Mn–O(1) 1.891(3), Mn–N(1) 1.988(4), Mn– N(2) 1.986(4) for 1 and Mn–O(1) 1.880(9), Mn–N(1) 1.995(11) and Mn–N(2) 1.971(11) Å for 4. The lengthening observed in the Mn–Ophenolic bond lengths in 5 is not nearly as pronounced as the diVerences in the Mn–Nimine bond length, that of Mn(1)– N(2) being particularly marked.These diVerences must be accounted for by the change of ligand backbone bridges in 1 and 4. Preparation of monomeric manganese SchiV base complexes with chelating carboxylate ligands The reaction of Me3SiCl with [Mn(salpn)(acac)]. Our initial reasons for investigating the reactivity of [Mn(salpn)(acac)] with Me3SiCl were twofold. First, we wished to discover whether it was a straightforward procedure to replace the acetylacetonate ligand, and, secondly, to see whether such a reaction could lead to the formation of cluster compounds. This reagent has previously been employed as an abstractor of carboxylate ligands co-ordinated to metal centres 16 and has subsequently been employed in this manner to prepare manganese clusters.17 The reaction of [Mn(salpn)(acac)] with Me3SiCl in dry degassed acetonitrile resulted in the formation of a compound of stoichiometry Mn(salpn)Cl in 61% yield (see Table 2).The positive ion FAB mass spectrum of the complex shows a parent ion peak at m/z 335 which corresponds to a cation of the form [Mn(salpn)]1. No fragments involving chloride co-ordination were observed, indicating that if chloride is ligated to the metal centre in the solid state that this ligation is weak. The absence of bands in the infrared spectrum at 1635 and 1600 cm21, which we attribute to the acetylacetonate group of the starting material, together with a shift in the imine band from 1615 to 1623 cm21, indicate that complete displacement of acetylacetonate has occurred, apparently via a simple counter ion exchange.In the light of these data, and the only slightly lowered room temperature magnetic moment of 4.5 mB, relative to that expected for magnetically dilute d4 manganese(III), it is reasonable to propose the formation of a five-co-ordinate monomeric species as has previously been observed in related compounds.18 Crystals suitable for single crystal X-ray diVraction studies were mounted in an atmosphere of the mother-liquor (since decomposition occurs in the absence of solvent vapour).These were found to consist of the weakly m-phenoxy bridged dimeric species [{Mn(salpn)Cl}2]?CH3CN 6, shown in Fig. 4. The structure is essentially analogous to that found for the related [Mn(salen)Cl] 18b and other species 18a,c–e however due to the much shorter m-phenoxy bonds Mn(1)–O(1*) 2.487(4) Å (cf. 2.897 Å in a related monomer18c) and the roughly octahedral manganese(III) geometry, we thus prefer to present this structure as a dimer rather than a monomer. Indeed such a structure has been postulated by Pecoraro and co-workers,1934 J. Chem. Soc., Dalton Trans., 1999, 31–41 who suggested that the absence of solvent molecules capable of co-ordinating to the manganese(III) centre would lead to the formation of 6, rather than the monomeric octahedral species that have been reported.The remaining bond lengths and angles are unexceptional for such compounds. The characterisation of complex 6 clearly demonstrates that it is a simple procedure to replace the acetylacetonate ligand. Fig. 2 (a) Molecular structure of [Mn(3CH3O-salpn)(O2CBut)] 2. Selected bond lengths (Å) and angles (8) with e.s.d.s in parentheses: Mn(1)–O(1) 1.868(4), Mn(1)–O(2) 1.918(4), Mn(1)–N(1) 2.150(5), Mn(1)–N(2) 1.986(5), Mn(1)–O(5) 2.329(5) and Mn(1)–O(6) 2.071(4); O(1)–Mn(1)–O(2) 91.0(2), O(1)–Mn(1)–N(1) 87.9(2), O(1)–Mn(1)– N(2) 177.3(2), O(1)–Mn(1)–O(5) 94.0(2), O(1)–Mn(1)–O(6) 91.2(2), O(2)–Mn(1)–N(1) 110.1(2), O(2)–Mn(1)–N(2) 88.9(2), O(2)–Mn(1)– O(5) 96.8(2), O(2)–Mn(1)–O(6) 155.5(2), O(5)–Mn–O(6) 58.7(2), N(1)– Mn(1)–N(2) 89.6(2), N(1)–Mn(1)–O(5) 153.1(2), N(1)–Mn(1)–O(6) 94.4(2), N(2)–Mn(1)–O(5) 88.7(2) and N(2)–Mn–O(6) 90.0(2).(b) Molecular structure of [Mn(3CH3O-salpn)(O2CPri)], 3. Selected bond lengths (Å) and angles (8) with e.s.d.s in parentheses: Mn(1)–O(1) 1.867(12), Mn(1)–O(2) 1.875(11), Mn(1)–N(1) 2.150(13), Mn(1)–N(2) 1.939(15), Mn(1)–O(5) 2.372(12) and Mn(1)–O(6) 2.026(13); O(1)– Mn(1)–O(2) 90.0(5), O(1)–Mn(1)–N(1) 89.3(6), O(1)–Mn(1)–N(2) 179.0(6), O(1)–Mn(1)–O(5) 93.8(5), O(1)–Mn(1)–O(6) 92.5(5), O(2)– Mn(1)–N(1) 113.3(5), O(2)–Mn(1)–N(2) 89.0(6), O(2)–Mn(1)–O(5) 96.4(5), O(2)–Mn(1)–O(6) 154.8(5), O(5)–Mn–O(6) 58.5(5), N(1)– Mn(1)–N(2) 89.7(6), N(1)–Mn(1)–O(5) 150.1(5), N(1)–Mn(1)–O(6) 91.7(5), N(2)–Mn(1)–O(5) 87.3(5) and N(2)–Mn–O(6) 88.2(6).Of interest is the change of geometry at the metal centre in that the cis-chelated sites of the “pretemplated” complex do not remain. In 6 trans sites are open at the metal centre due to the Fig. 3 (a) Molecular structure of [Mn(3CH3O-salen)(O2CCH2Ph)- (H2O)]?H2O 4. Selected bond lengths (Å) and angles (8) with e.s.d.s in parentheses: Mn–O(1) 1.880(9), Mn–O(2) 1.887(9), Mn–N(1) 1.995(11), Mn–N(2) 1.971(11), Mn–O(5) 2.135(10) and Mn–O(1w) 2.353(9); O(1)–Mn–O(2) 93.5(4), O(1)–Mn–N(1) 172.5(5), O(1)–Mn–N(2) 92.0(4), O(1)–Mn–O(5) 98.7(4), O(1)–Mn–O(1w) 91.8(4), O(2)–Mn– N(1) 91.7(4), O(2)–Mn–N(2) 172.7(5), O(2)–Mn–O(5) 94.5(4), O(2)– Mn–O(1w) 87.9(4), O(5)–Mn–O(1w) 169.1(3), N(1)–Mn–N(2) 82.4(5), N(1)–Mn–O(5) 86.2(4), N(1)–Mn–O(1w) 83.0(4), N(2)–Mn–O(5) 89.5(4) and N(2)–Mn–O(1w) 87.1(4).(b) Molecular structure of [Mn(3CH3O-salpn)(O2CCH2Ph)(H2O)] 5. Selected bond lengths (Å) and angles (8) with e.s.d.s in parentheses: Mn(1)–O(1) 1.912(6), Mn(1)– O(2) 1.914(6), Mn(1)–N(1) 2.026(8), Mn(1)–N(2) 2.050(9), Mn(1)–O(5) 2.139(7) and Mn(1)–O(6) 2.316(6); O(1)–Mn(1)–O(2) 87.0(3), O(1)– Mn(1)–N(1) 89.7(3), O(1)–Mn(1)–N(2) 175.2(3), O(1)–Mn(1)–O(5) 94.3(3), O(1)–Mn(1)–O(6) 87.5(2), O(2)–Mn(1)–N(1) 173.9(3), O(2)– Mn(1)–N(2) 89.8(3), O(2)–Mn(1)–O(5) 100.4(3), O(2)–Mn(1)–O(6) 89.9(3), O(5)–Mn–O(6) 169.6(3), N(1)–Mn(1)–N(2) 93.1(3), N(1)– Mn(1)–O(5) 85.0(3), N(1)–Mn(1)–O(6) 84.7(3), N(2)–Mn(1)–O(5) 89.8(3) and N(2)–Mn–O(6) 88.9(3).J.Chem. Soc., Dalton Trans., 1999, 31–41 35 Table 2 Analytical and some selected data for the manganese complexes containing salpn, 5NO2-salpn and 3Br,5NO2-salpn as ligands Analysis (%) a IR/cm21 Complex [Mn(salpn)Cl] [Mn(salpn)(O2CMe)]?H2O [Mn(salpn)(O2CEt)] [Mn(salpn)(O2CPrn)] [Mn(salpn)(O2CBun)] [Mn(salpn)(O2CPri)] [Mn(salpn)(O2CBut)] Mn(5NO2-salpn)(H2O) Mn(3Br,5NO2-salpn) [Mn(5NO2-salpn)(O2CMe)]?H2O [Mn(5NO2-salpn)(O2CEt)] [Mn(5NO2-salpn)(O2CPrn)]?2H2O [Mn(5NO2-salpn)(O2CPri)] [Mn(5NO2-salpn)(O2CBut)] [Mn(3Br,5NO2-salpn)(O2CMe)]?2H2O [Mn(3Br,5NO2-salpn)(O2CEt)] [Mn(3Br,5NO2-salpn)(O2CPrn)]?H2O C 55.0 (55.1) 55.2 (55.3) 58.5 (58.8) 60.0 (59.7) 60.3 (60.6) 59.2 (59.7) 60.2 (60.6) 45.9 (46.0) 37.0 (36.8) 45.4 (45.4) 47.6 (48.2) 45.8 (46.0) 48.5 (49.2) 49.1 (50.2) 33.6 (33.6) 36.1 (36.6) 36.2 (36.6) H 4.5 (4.3) 5.0 (5.1) 5.2 (5.2) 5.4 (5.5) 5.8 (5.7) 5.5 (5.5) 5.5 (5.7) 3.6 (3.6) 2.6 (2.5) 3.8 (3.8) 3.8 (3.8) 3.9 (4.6) 3.9 (4.1) 4.2 (4.4) 2.6 (2.8) 2.6 (2.6) 2.6 (3.0) N 7.2 (7.6) 6.6 (6.8) 6.7 (6.9) 6.5 (6.6) 6.4 (6.4) 6.2 (6.6) 6.2 (6.4) 12.2 (12.6) 10.8 (10.1) 11.4 (11.2) 11.0 (11.2) 10.4 (10.2) 11.3 (10.9) 10.5 (10.6) 8.3 (8.3) 8.5 (8.5) 8.7 (8.1) Mn 14.3 (14.8) 13.8 (13.4) 13.0 (13.5) 13.1 (13.0) 12.9 (12.6) 13.1 (13.0) 12.8 (12.6) n(C]] N) 1623 1615 1614 1614 1616 1616 1614 1622 1598 1622 1622 1620 1623 1622 1620 1622 1626 meff/mB 4.5 4.8 4.8 4.8 4.5 4.5 4.7 4.8 4.7 4.9 4.9 5.0 5.0 4.7 4.8 FABb m/z 335 335 335 335 335 335 335 425 425 425 425 425 583 583 581 Yield c (%) 61 57 58 79 57 66 72 46 40 59 54 47 62 58 54 57 54 a Found (calculated). b Peaks corresponding to [ML]1.c Based on manganese. SchiV base ligand adopting an equatorial position. Larson and Pecoraro15 have previously employed [Mn(salpn)(acac)] in the preparation of the dinuclear manganese(IV) species [{Mn- (salpn)O}2] through peroxide oxidation of the manganese(III) species [Mn(salpn)(acac)] in acetonitrile.In this instance the cis sites at the manganese centre, occupied by the acetylacetonate ligand, were replaced directly by the bridging m-oxo groups in the manganese(IV) complex; this was attributed directly to the pretemplated nature of the manganese(III) species. Refined mechanistic interpretations of this reaction were made which invoked the initial process to be deprotonation of hydrogen peroxide by the acetylacetonate ligand and showed that both the bridging m-oxo groups originated from the same molecule of hydrogen peroxide.From these results it was therefore apparent that if we used the co-ordinated acetylacetonate as a base, and avoided high thermodynamic driving forces (e.g. the formation of Si–O bonds), we would be able directly to replace the acetylacetonate with carboxylate groups through the reaction of carboxylic acids with [Mn(salpn)(acac)] and produce manganese(III) complexes. The reaction of carboxylic acids with [Mn(salpn)(acac)].A series of straight- and branched-chain aliphatic carboxylic acids, RCO2H (R = Me, Et, Prn, Bun, Pri or But), were treated Fig. 4 Molecular structure of [{Mn(salpn)Cl}2]?CH3CN 6. Lattice acetonitrile is not depicted. Selected bond lengths (Å) and angles (8) with e.s.d.s: Mn(1)–Cl(1) 2.4521(18), Mn(1)–O(1) 1.919(4), Mn(1)– O(2) 1.868(4), Mn(1)–N(1) 2.026(5), Mn(1)–N(2) 2.039(5) and Mn(1)– O(1*) 2.487(4); O(1)–Mn(1)–Cl(1) 99.45(13), O(2)–Mn(1)–Cl(1) 96.78(14), N(1)–Mn(1)–Cl(1) 90.62(14), N(2)–Mn(1)–Cl(1) 91.92(15), O(1)–Mn(1)–O(2) 86.51(17), O(1)–Mn(1)–N(1) 87.16(18), O(1)– Mn(1)–N(2) 168.47(18), O(2)–Mn(1)–N(1) 170.96(19), O(2)–Mn(1)– N(2) 90.36(19) and N(1)–Mn(1)–N(2) 94.6(2).with [Mn(salpn)(acac)] in ethanol aVording a succession of compounds of stoichiometry [Mn(salpn)(O2CR)], see Table 2. All compounds exhibit magnetic properties typical of manganese( III) SchiV base compounds of this type 11,13,14 with most magnetic moments being close to the spin only value of 4.9 mB expected for magnetically dilute high spin d4 systems, indicating little or no antiferromagnetic interactions in the solid state.The positive ion FAB mass spectra of the compounds all show parent ion peaks relating to the fragment [Mn(salpn)]1 at m/z 335 and no other fragments relating to carboxylate (or acetylacetonate) co-ordination to a monomeric manganese species. This behaviour is consistent with the other monomeric manganese( III) SchiV base complexes reported herein and those that we have previously reported with bidentate and chelating carboxylates. 13,14 All compounds show a strong band between 1614 and 1616 cm21 in their infrared spectra attributable to n(C]] N). These values are comparable with those of the starting material; however, the complete absence of strong bands at 1635 and 1600 cm21 of the acetylacetonate species indicate that this anion has been replaced by a carboxylate anion.This is substantiated by the presence of bands between 1543 and 1548, 1466 and 1470 and 616 and 622 cm21 which can be assigned to nasym(OCO), nsym(OCO) and d(OCO), respectively. The diVerence between the asymmetric and symmetric stretching modes of the carboxylate of ca. 80 cm21 can be tentatively assigned to the presence of bidentate and chelating carboxylate groups.13,14 Electrochemistry of the complexes. The electrochemical behaviour of the complexes was investigated in dichloromethane solution using tetrabutylammonium hexafluorophosphate as a supporting electrolyte.All compounds formulated as [Mn(salpn)(O2CR)] show essentially identical behaviour which is typified by the voltammogram of [Mn(salpn)(O2CBun)], in Fig. 5(a). There are two diVerent redox processes occurring. The first process (O1/R1) corresponds to the quasi-reversible manganese(II/III) couple, E1/2 = 20.25 to 20.30 V (see Table 3 for electrochemical data) whilst the second redox process (O2/ R2) is assigned to the sequential oxidation of the species to the manganese(IV) state (Epa = 0.97 to 1.00 V).This second process does not show reversible electrochemistry. However, the reductive waves between 0.18 and 0.25 V can be associated with this oxidative process, as the reductive wave is absent when the electrode is rotated and also through alteration of the voltage window, since when a window up to 0.9 V is employed the reductive wave is not observed.It is interesting that cyclic voltammetry serves as a very sensitive test of the purity of these compounds. Despite the apparent purity of the compounds36 J. Chem. Soc., Dalton Trans., 1999, 31–41 Fig. 5 Cyclic voltammograms of (a) [Mn(salpn)(O2CBun)], (b) [Mn(salpn)(O2CBut)], (c) [Mn(salpn)(O2CMe)]2n(H2O)3n and (d) [{Mn(salpn)Cl}2] in dichloromethane. Calomel electrode all potentials are versus the saturated calomel electrode. Table 3 Electrochemical data for the complexes versus saturated calomel electrode Assigned couple (V) Compound [{Mn(salpn)Cl}2] [Mn(salpn)(O2CMe)]?H2O [Mn(salpn)(O2CEt)] [Mn(salpn)(O2CPrn)] [Mn(salpn)(O2CBun)] [Mn(salpn)(O2CPri)] [Mn(salpn)(O2CBut)] [{[Mn(salpn)(O2CMe)]2(H2O)3}n] MnII/III 0.0 20.19 20.19 20.17 20.19 20.18 20.21 20.18 MnIII/II 20.31 20.35 20.35 20.34 20.32 20.42 20.40 20.34 E1/2 20.16 20.27 20.27 20.25 20.26 20.30 20.30 20.26 MnIII/IV 1.21 0.98 0.98 1.00 0.97 1.00 0.97 1.00 MnIV/III 1.18 0.22 0.22 0.22 0.25 0.17 0.18 — E1/2 1.20 ——————— (viz.as judged by elemental analyses and infrared spectroscopy, see Table 2) cyclic voltammetry is able to detect the presence of starting material as this species shows a quasi-reversible couple at Epa 0.53 V and Epc 0.45 V. A typical example is given in Fig. 5(b) in which the starting material impurity (labelled SM) is clearly present, in spite of the apparent “purity” of the complex. In these cases further reaction with the appropriate acid until this wave is no longer observed indicates complete displacement of acetylacetonate has occurred.These voltammetric studies appear to suggest that similar solution species are present. However, as yet we are unable to predict the form these species take in solution, i.e. whether they exist as the desired monomers with cis chelated carboxylates or as ionised species of the form Mn(salpn)1. We have four reasons to suppose that the solution species are the desired cis chelated monomers.First, the subtle changes in the E1/2 values of the manganese(II/III) couples indicate that changing the carboxylate auxiliary ligand aVects, albeit slightly, the potential at which this process occurs. If the solution species was the cationic Mn(salpn)1 these values would be expected to be identical. Secondly, the cyclic voltammogram of the polymeric species [Mn(salpn)(O2CMe)]2n(H2O)3n whose structure we have previously reported,11 in which the manganese centre contains trans co-ordinated carboxylates, shows subtle diVerences electrochemically, Fig. 5(c). As for the species previously described a quasi-reversible manganese(II/III) couple is observed {Epa 20.18 V, Epc 20.34 V, E1/2 20.26 V, cf. 20.27 V for [Mn(salpn)- (O2CMe)]}. However, unlike the previous complexes no reversible process relating to the oxidative wave at Epa 1.0 V is observed. Thirdly, when the cyclic voltammogram of 6 was recorded diVerent behaviour was again observed, Fig. 5(d). Two quasi-reversible processes are observed which we assign to the manganese-(II/III) and -(III/IV) couples with E1/2 values of 20.16 and 1.20 V, respectively. Li and Pecoraro18c have previously reported that the only solution species observed for the structurally related [Mn(salen)Cl] 18b was the Mn(salen)1 cation. It seems reasonable to propose that in solution 6 will behave in a similar manner and that the diVerences between the voltammogram of 6 and those of the carboxylate complexes prepared are as a result of such ionisation.Finally dichloromethane solutions of the complexes formulated as [Mn(salpn)(O2CR)] show no conductivity. X-Ray studies. X-Ray powder diVraction studies of [Mn- (salpn)(O2CBun)] and [Mn(salpn)(O2CBut)] indicate that these species are monomeric with cis-chelated carboxylates in the solid state. This is based on the identical nature of the powder diVraction patterns with those of the crystallographically characterised monomers derived from our conventional reaction of the appropriate manganese(II) carboxylate salt with the SchiV base ligand.13,14 These powder diVraction patterns are shown in Fig. 6. Crystals suitable for single crystal X-ray dif-J. Chem. Soc., Dalton Trans., 1999, 31–41 37 fraction studies of [Mn(salpn)(O2CPri)] 7 were found to consist, as expected, of a monomeric manganese(III) SchiV base complex with a cis-chelated carboxylate ligand, Fig. 7. The distorted N2O4 octahedral co-ordination sphere of the manganese centre is essentially identical to those that we have previously characterised in other related monomers.13,14 The distortion of the manganese centre is as a result of the chelation of the carboxylate. These structures are all typified by asymmetric coordination of the carboxylate [in this case Mn(1)–O(3) 2.304(4) Fig. 6 Powder X-ray diVraction patterns of (a) [Mn(salpn)(O2CBun)] and (b) [Mn(salpn)(O2CBut)] prepared in this study compared with those of their crystallographically characterised monomers.13,14 Fig. 7 Molecular structure of [Mn(salpn)(O2CPri)] 7. Selected bond lengths (Å) and angles (8) with e.s.d.s: Mn(1)–O(1) 1.899(3), Mn(1)– O(2) 1.872(3), Mn(1)–O(3) 2.304(4), Mn(1)–O(4) 2.099(4), Mn(1)–N(1) 1.990(4) and Mn(1)–N(2) 2.128(4), O(1)–Mn(1)–O(2) 89.73(15), O(1)– Mn(1)–N(1) 89.59(15), O(2)–Mn(1)–N(1) 176.21(15), O(1)–Mn(1)– N(2) 112.59(15), O(2)–Mn(1)–N(2) 87.51(15), N(1)–Mn(1)–N(2) 89.31(16), O(3)–Mn(1)–N(2) 153.48(14), O(4)–Mn(1)–N(2) 94.28(14), O(1)–Mn(1)–O(3) 93.90(15), O(1)–Mn(1)–O(4) 153.03(15) and O(4)– Mn(1)–O(3) 59.21(13).and Mn(1)–O(4) 2.099(4) Å]. It is such co-ordination that we have attributed to the absence of fragments relating to the monomeric manganese species [Mn(salpn)(O2CR)]1 in the FAB mass spectra of the complexes. All other bond lengths and angles are unexceptional. Preparation of manganese(III) complexes of SchiV base ligands containing electron-withdrawing substituents with ancillary carboxylate donors In view of the apparent success of the reaction of carboxylic acids with [Mn(salpn)(acac)] for the preparation of complexes of the form [Mn(salpn)(O2CR)], we wondered whether this synthetic method might access a route to manganese(III) complexes of this stoichiometry with electron withdrawing substituents about the aromatic rings of the SchiV base ligands, e.g. 5NO2- salpn and 3Br,5NO2-salpn. Our desires for such a route were based on our experiences with the related salen ligands with these substituents 20,21 and the inconsistent nature of the products obtained under our standard aerobic reaction conditions. 11,13,14 We found that aerial oxidation under Boucher conditions 22 does not yield any manganese(III) complex with 3Br, 5NO2 substituents and that for 5NO2 substituents, although some manganese(III) species could be detected by cyclic voltammetry, the main product of the reaction was a manganese(II) species.Manganese(III) species could be isolated, however, by the chemical oxidation of the manganese(II) species using ferrocenium tetrachloroferrate(III).20 Furthermore, we have even found that it is possible to isolate manganese(II) species from the reduction of manganese(III) acetate in the presence of 3Br,5NO2-salen.21 Our attempts to isolate manganese( III) complexes with ancillary carboxylate ligands, as indicated, have been largely unsuccessful, manganese(II) complexes of stoichiometry Mn(L)?nH2O being isolated routinely.However, on one occasion, in the reaction of 5NO2-salpn with manganese(II) acetate tetrahydrate, we have been able to isolate a manganese(III) species from the reaction liquor after removal of the precipitated manganese(II) complex and slow evaporation yielding [Mn(5NO2-salpn)(O2CMe)(H2O)]. This is consistent with the isolation of [Mn(5NO2-salen)(O2CMe)(H2O)] 8, Fig. 8, using 5NO2-salen. The structure consists of an octahedrally co-ordinated manganese(III) centre, analogous with 1, 4 and 5, in which the xy plane of the manganese ion is coordinated by the SchiV base ligand and axial sites are filled by water and the acetate ligated in the unidentate mode.The Mn–Ophenolic bond lengths are identical with those observed in 1 and 4 indicating that the ortho-methoxy group and para-nitro group do not aVect this bond length in any way. The Mn–Nimine bonds are shorter than all of the other bonds in 4 and 5.The Mn–Ocarboxylate distance is the shortest of any of these Fig. 8 Molecular structure of [Mn(5NO2-salen)(O2CMe)(H2O)] 8. Selected bond lengths (Å) and angles (8) with e.s.d.s: Mn(1)–O(1) 1.882(4), Mn(1)–O(2) 1.890(4), Mn(1)–N(1) 1.977(4), Mn(1)–N(2) 1.970(4), Mn(1)–O(7) 2.347(5) and Mn(1)–O(8) 2.123(4); O(1)–Mn(1)– O(2) 91.67(14), O(1)–Mn(1)–N(2) 172.31(17), O(2)–Mn(1)–N(2) 92.81(16), O(1)–Mn(1)–N(1) 91.25(15), O(2)–Mn(1)–N(1) 173.09(17), N(2)–Mn(1)–N(1) 83.63(16), O(1)–Mn(1)–O(8) 92.86(15), O(2)– Mn(1)–O(8) 95.39(16), N(2)–Mn(1)–O(8) 92.93(17), N(1)–Mn(1)–O(8) 90.71(18), O(1)–Mn(1)–O(7) 92.17(17), O(2)–Mn(1)–O(7) 89.94(17), N(2)–Mn(1)–O(7) 81.68(18), N(1)–Mn(1)–O(7) 84.69(19), O(8)– Mn(1)–O(7) 173.26(16).38 J.Chem. Soc., Dalton Trans., 1999, 31–41 monomers, Mn–O(8) 2.123(4) Å, whilst the manganese–water distance is intermediate between 1 and 4 at 2.347(5) Å. In view of the success in the displacement of acetylacetonate with a carboxylate in [Mn(salpn)(acac)] we attempted to apply the same reaction to [Mn(5NO2-salpn)(acac)] and [Mn- (3Br,5NO2-salpn)(acac)] prepared by the same method as [Mn(salpn)(acac)] (see Experimental section).15 It appears that this does provide a satisfactory route to compounds of stoichiometry MnL(O2CR), but in much lower yields than the unsubstituted analogue, due to the precipitation of yellow materials of composition MnL(H2O)n, n = 0,2, in up to 50% yield.We have only been able to isolate crystals from one of these complexes, Mn(3Br,5NO2-salpn)(O2CMe)?2H2O. These were not the expected complex, but rather a di-m-oxo manganese(IV) species, [{Mn(m-3Br,5NO2-salpn)(m-O)}2]?5DMF, 9, Fig. 9, which is structurally analogous with a related dimer that we have previously reported.20 In the previous case the remarkable feature of the structure lay in the non-planar nature of the salen ligand. Although this twist is still inherent in 9 it is perhaps less surprising since there is much greater conformational freedom between the imino nitrogen atoms N1 and N2 in this ligand than in a salen-based ligand.The most surprising feature of the compound is the apparent attainment of the 14 oxidation state with this ligand. This assignment is based on the similarity between the bond lengths and angles in the central Mn–O2–Mn of 9 and other well characterised di-m-oxo-bridged manganese( IV) species.20 The Mn–O(3) bridging bonds of 1.797(4) Å and 1.809(5) Å, the Mn–O(3)–Mn9 angle of 98.7(2)8 and the Mn? ? ? Mn separation of 2.736(2) Å are all typical of such systems.Fig. 9 Molecular structure of [{Mn(m-3Br,5NO2-salpn)(m–O)}2]? 5DMF 9. Selected bond lengths (Å) and angles (8) with e.s.d.s: Mn(1)– Mn(1*) 2.736(2), Mn(1)–O(1) 1.933(5), Mn(1)–O(1*) 1.952(5), Mn(1)– O(3) 1.809(5), Mn(1)–O(3*) 1.797(4), Mn(1)–N(1) 1.997(6) and Mn(1)– N(1*) 1.993(6); O(1)–Mn(1)–O(1*) 93.1(2), O(1)–Mn(1)–O(3) 173.6(2), O(1)–Mn(1)–O(3*) 92.5(2), O(1)–Mn(1)–N(1) 88.4(2), O(1)– Mn(1)–N(1*) 87.5(2), O(1*)–Mn(1)–O(3) 93.2(2), O(1*)–Mn(1)–O(3*) 174.5(2), O(1*)–Mn(1)–N(1) 86.0(2), O(1*)–Mn(1)–N(1*) 87.1(2), O(3)–Mn(1)–O(3*) 81.3(2), O(3)–Mn(1)–N(1) 93.3(2), O(3)–Mn(1)– N(1*) 91.6(2), O(3*)–Mn(1)–N(1) 94.7(2), O(3*)–Mn(1)–N(1*) 92.7(2) and N(1)–Mn(1)–N(1*) 171.7(2). Conclusion The co-ordination chemistry of manganese(III) SchiV base complexes with ancillary carboxylate ligands is an area of tremendous variety.Although at times capricious, some obvious patterns are developing. When sterically restricting carboxylates 2O2CBut and 2O2CPri are employed with salpn based ligands monomers ensue, in which the carboxylate asymmetrically chelates cis sites at the metal centre as is observed in complexes 2, 3, 7 and previous studies.13,14 This behaviour appears to be independent of ligand substituents. When tetradentate SchiV base ligands with methoxy groups ortho to the phenolic donor atoms are used and sterically very demanding carboxylates are avoided the unique mode of co-ordination of the carboxylate is unidentate, 1, 4 and 5.Unlike our previous studies 14 in which the co-ordination spheres of the manganese centres were completed by weak m-phenoxy bridges to a second metal centre forming dimers, the complexes reported herein are all monomers in which water molecules complete the co-ordination sphere. The diVerences between these two structural motifs are relatively minor and of greatest significance is the unidentate binding mode of the carboxylate which appears to result exclusively for these systems.A similar binding mode is also observed in 8. It is apparent that we are able to prepare monomeric species of stoichiometry [Mn(salpn)(O2CR)] by the reaction of [Mn- (salpn)(acac)] with a variety of aliphatic carboxylic acids. This is substantiated by cyclic voltammetry and X-ray studies. X-Ray powder diVraction shows the products of these reactions with valeric and trimethylacetic acid to be the same as the monomers that we have crystallographically characterised from the reaction of the manganese(II) carboxylate salt and the SchiV base ligand.13,14 Single crystal studies of 7 further substantiate this view.This methodology is also a relatively satisfactory means of preparing complexes of manganese(III) with electron withdrawing substituents on the aromatic rings of the SchiV base ligands which normally prevent the aerial oxidation of manganese(II) salts.20 Although manganese(II) side products are also isolated a number of manganese(III) complexes have been isolated in fair yield.On one occasion we have been able to isolate from one of these products single crystals of a manganese(IV) di-m-oxo dimer in which the manganese centres are also bridged by two ligands, 9, a further remarkable example of oxidation of manganese to the 14 level with concomitant formation of di-m-oxo bridges upon recrystallisation from DMF.We believe these results develop the chemistry of these systems considerably. Clear and predictable patterns have developed, and we hope will continue to develop yet unpredictable and fascinating structures also persist within this niche. Experimental All experiments, except for the preparation of complex 6, were carried out in air. All reagents (Aldrich and Maybridge Chemicals) and solvents were used as received except acetonitrile (for the preparation of 6) and dichloromethane (for cyclic voltammetry) which were freshly distilled from CaH2 prior to use.Physical measurements Infrared spectra were recorded on a Bruker IFS66V spectrophotometer as KBr discs, FAB mass spectra on a Kratos MS-45 spectrometer with a FAB probe and a xenon reaction gas using a m-nitrobenzyl alcohol matrix. Room temperature magnetic moments were measured on a Faraday balance which was calibrated using HgCo(NCS)4.Elemental analyses were performed by the in-house services of the University of Santiago and UMIST. Cyclic voltammetry was performed with a Princeton Applied Research model 270 potentiostat and potential scan-J. Chem. Soc., Dalton Trans., 1999, 31–41 39 Table 4 Crystal data and details of refinement Formula M Crystal dimensions/mm Crystal system Space group a/Å b/Å c/Å b/8 V/Å3 Z Dc/g cm23 F(000) m/cm21 2qmax/8 Number of unique data Number of observed data a Number of variables Rb R9 c 1 C20H21MnN2O9 488.33 0.45 × 0.38 × 0.12 Monoclinic P21/c (no. 14) 12.053(2) 19.874(4) 9.312(2) 99.97(2) 2196.9 4 1.476 1008 6.55 50.10 2166 2166 382 0.0433 0.1186 2 C25H31Cl2MnN2O6 581.36 0.56 × 0.45 × 0.33 Monoclinic P21/n (no. 14) 13.034(3) 12.111(4) 18.446(5) 102.760(10) 2839.9 4 1.360 1208 6.93 45.00 3679 3679 325 0.0748 0.2098 3 C23H27MnN2O6 482.41 0.52 × 0.16 × 0.10 Monoclinic P21/c (no. 14) 9.692(2) 12.088(2) 20.027(4) 99.55(3) 2313.8 4 1.385 1008 6.11 44.84 1097 1097 134 0.1063 0.2957 4 C26H27MnN2O8 550.44 0.40 × 0.27 × 0.12 Monoclinic P21/c (no. 14) 9.927(2) 13.495(3) 18.761(4) 95.88(3) 2500.1 4 1.462 1160 5.82 44.90 888 888 184 0.0541 0.1385 5 C28H32MnN2O8 579.50 0.62 × 0.42 × 0.21 Monoclinic P21/c (no. 14) 10.440(6) 13.647(5) 19.300(11) 99.33(6) 2713 4 1.419 1212 5.40 45.02 3547 3544 343 0.0946 0.2518 6 C19H19ClMnN3O2 411.76 0.40 × 0.25 × 0.25 Monoclinic P21/n (no.14) 10.849(4) 12.409(5) 13.610(4) 95.54(3) 1823.6 4 1.500 848 8.89 50.00 3161 3161 236 0.0444 0.1057 7 C21H23MnN2O4 422.35 0.35 × 0.30 × 0.20 Orthorhombic Pbca (no.61) 17.844(4) 18.881(5) 12.337(3) 90.0 4156.5 8 1.350 1760 6.63 50.00 3643 3643 323 0.048 0.1226 8 C18H17MnN4O9 488.30 0.35 × 0.30 × 0.15 Monoclinic P21/c (no. 14) 14.014(2) 12.086(2) 13.370(2) 115.71(2) 2040.4 4 1.590 1000 7.07 50.00 3573 3573 357 0.0461 0.1096 9 C49H59Br4Mn2N13O19 1563.61 0.25 × 0.25 × 0.20 Monoclinic P21/n (no. 14) 12.592(2) 12.103(3) 20.266(2) 92.11(2) 3086.7(10) 2 1.682 1572 3.079 50.00 5398 5398 418 0.0430 0.0817 a I > 2.00s(I).b R = S Fo| 2 |Fc /S|Fo|. c R9 = [Sw(|Fo|2 2 |Fc|2)2/SwFo 2)]� �4 .40 J. Chem. Soc., Dalton Trans., 1999, 31–41 ning unit with the operating program ECHEM. The electrolytic cell consisted of a METROHM model 6.12404 carbon disc working electrode, a saturated calomel reference electrode and a platinum wire auxiliary electrode. Voltammograms were obtained of ca. 1 mM dichloromethane solutions of the complexes using 0.1 M Bu4NPF6 as a supporting electrolyte.Preparations SchiV base ligands. All ligands were prepared in virtually quantitative yields as previously reported and satisfactorily characterised by elemental analysis, 1H NMR, infrared and FAB mass spectroscopy.20 Compounds of stoichiometry [Mn(3CH3O-salen)(O2CR)] and [Mn(3CH3O-salpn)(O2CR)]. All comps were prepared by the same method as previously reported in high yields (based on manganese).11,13,14 They were characterised by elemental analysis, infrared, mass spectroscopy, room temperature magnetic moment measurement and FAB mass spectroscopy.[MnL(acac)] complexes. All complexes were prepared as previously reported 15 by the reaction of [Mn(acac)3] (Aldrich) with the appropriate ligand. The [MnL(acac)] were characterised satisfactorily. [Mn(salpn)Cl]. Acetonitrile (25 cm3) was freshly distilled from CaH2 and added to [Mn(salpn)(acac)] (0.5 g, 1.15 mmol) under nitrogen. Chlorotrimethylsilane (0.15 cm3, 1.15 mmol) was added to the solution which resulted in the immediate formation of an emerald green suspension which rapidly darkened to form a dark solid.The reaction was stirred for 5 min before cooling to 5 8C. The resultant microcrystalline precipitate formed was collected by filtration, washed copiously with diethyl ether and dried in vacuo. Monomeric complexes [Mn(salpn)(O2CR)], by the reaction of carboxylic acids with [Mn(salpn)(acac)]. All the monomeric manganese(III) complexes were prepared similarly, typified by that of [Mn(salpn)(O2CPri)].To an ethanolic solution (25 cm3) of [Mn(salpn)(acac)] (0.568 g, 1.3 mmol) was added isobutyric acid (1.27 g, 1.4 mmol). The solution was heated to reflux with stirring for 2 h then reduced in volume (ca. 4 cm3) and the complex precipitated in diethyl ether (100 cm3) from this solution. The green solid thus formed was collected by filtration, washed with diethyl ether (2 × 5 cm3) and dried in vacuo.Reaction of [Mn(5NO2salpn)(acac)] and [Mn(3Br,5NO2- salpn)(acac)] with carboxylic acids These reactions were performed in an analogous way to the reactions of carboxylic acids with [Mn(salpn)(acac)] with one modification. Prior to volume reduction the precipitated yellow material was removed by filtration. This material was washed with diethyl ether (20 cm3) and dried in vacuo yielding complexes of stoichiometry MnL(H2O)n (n = 0–2) in 35–50% yield. X-Ray diVraction studies Powder studies.X-Ray powder diVraction patterns were recorded using a Scintag XRD2000 powder diVractometer using Cu-Ka radiation of l = 1.5481 Å. Single crystal studies. Crystals suitable for X-ray studies were obtained in the following ways. Those of complexes 1 and 4 were obtained by slow evaporation of acetonitrile solutions of the complexes. Crystals of 2 and 3 were obtained by layering a concentrated dichloromethane solution of the complexes with n-hexane.Crystals of 5 were obtained by slow evaporation of an acetonitrile–ethanol solution of the complexes. Cooling of a saturated acetonitrile solution of the complex at 5 8C for 3 weeks produced crystals of it as its acetonitrile solvate, 6. Crystals of 7 and 8 were obtained by layering a concentrated ethanolic solution of the complex with n-hexane. Crystals of 9 were obtained from a DMF solution of [Mn(3Br,5NO2- salpn)(O2CMe)]?2H2O after several weeks. Crystal data and experimental conditions are listed in Table 4.All data were collected at ambient temperature with Mo-Ka radiation of l = 0.71069 Å on a Nicolet P3, 1 and 3, a Siemens P4, 2 and 5, CAD-4, 4, or Rigaku AFC6S diVractometer, 6–9. The intensities of three representative reflections declined by 14.4% for 7. A linear correction factor was applied to account for this. All data were corrected for Lorentz-polarisation eVects. The structures were solved by direct methods 24,25 and refined using full matrix least squares on F2.Non-hydrogen atoms were refined anisotropically. Hydrogen atoms were included in idealised positions. All calculations were performed using the TEXSAN26 crystallographic package. CCDC reference number 186/1217. See http://www.rsc.org/suppdata/dt/1999/31/ for crystallographic files in .cif format. Acknowledgements The authors thank The Royal Society for the award of a Postdoctoral Research Fellowship from the European Scientific Exchange Programme (M.W.) and Ministerio de Educación y Cultura (Grant PB95-0827). References 1 V. L. Pecoraro, Manganese Redox Enzymes, VCH, New York, 1992. 2 W. F. Beyer and I. Fridovich, Biochemistry, 1985, 24, 6460. 3 A. Willing, H. Follman and G. Auling, Eur. J. Biochem., 1988, 170, 603. 4 M. L. Ludwig, K. A. Pattridge and W. C. Stallings, Manganese in Metabolism and Enzyme Function, Academic Press, New York, 1986, ch. 21, p. 405. 5 K. T. Govindjee and W. Coleman, Photochem. Photobiol., 1985, 42, 187; J.Amesz, J. Biochem. Biophys. Acta, 1983, 726, 1. 6 G. Christou, Acc. Chem. Res., 1989, 22, 328; S. Wang, H.-L. Tsai, K. S. Hagen, D. N. Hendrickson and G. Christou, J. Am. Chem. Soc., 1994, 116, 8376; R. C. Squire, S. M. J. Aubin, K. Folting, W. E. Streib, D. N. Hendrickson and G. Christou, Angew. Chem., Int. Ed. Engl., 1995, 34, 887. 7 H. Matsushima, E. Ishiwa, M. Koikawa, M. Nakashima and T. Tokii, Chem. Lett., 1995, 129; T. Tanase and S. J. Lippard, Inorg.Chem., 1995, 34, 4682; J. E. Sheats, R. S. Czernuszewicz, G. C. Dismukes, A. L. Rheingold, V. Petrouleas, J. Stubbe, W. H. Armstrong, R. H. Beer and S. J. Lippard, J. Am. Chem. Soc., 1987, 109, 1435; S. K. Mandal and W. H. Armstrong, Inorg. Chim. Acta, 1995, 229, 261; H. Adams, N. A. Bailey, N. Debaecker, D. E. Fenton, W. Kanda, J. M. Latour, H. Okawa and H. Sakiyama, Angew. Chem., Int. Ed. Engl., 1995, 34, 2535. 8 M. K. Chan and W. H. Armstrong, J. Am. Chem. Soc., 1991, 113, 5055. 9 A Pal, J. W. Gohdes, C. C. Wolf, A. Wilsch and W. H. Armstrong, Inorg. Chem., 1992, 31, 713. 10 V. L. Pecoraro, Photochem., Photobiol., 1988, 48, 249; K. Wieghardt, Angew. Chem., Int. Ed. Engl., 1989, 28, 1153; 1994, 33, 725. 11 N. Aurangzeb, C. E. Hulme, C. A. McAuliVe, R. G. Pritchard, M. Watkinson, A. Garcia-Deibe, M. R. Bermejo and A. Sousa, J. Chem. Soc., Chem. Commun., 1992, 1524. 12 J. E. Davies, B. M. Gatehouse and K. S. Murray, J. Chem. Soc., Dalton Trans., 1973, 2523; F. Akhtar and M. G. B. Drew, Acta Crystallogr., Sect. B, 1982, 38, 612; J. A. Bonadies, M. L. Kirk, M. S. Lah, D. P. Kessissoglou, W. E. Hatfield and V. L. Pecoraro, Inorg. Chem., 1989, 28, 2037. 13 N. Aurangzeb, C. E. Hulme, C. A. McAuliVe, R. G. Pritchard, M. Watkinson, M. R. Bermejo and A. Sousa, J. Chem. Soc., Chem. Commun., 1994, 2193. 14 C. E. Hulme, M. Watkinson, M. Haynes, C. A. McAuliVe, R. G. Pritchard, A. Sousa, M. R. Bermejo and M. Fondo, J. Chem. Soc., Dalton Trans., 1997, 1805. 15 E. J. Larson and V. L. Pecoraro, J. Am. Chem. Soc., 1991, 113, 3810. 16 M. L. H. Green, G. Parkin, J. Bashkin, J. Fail and K. Prout, J. Chem. Soc., Dalton Trans., 1982, 2519.J. Chem. Soc., Dalton Trans., 1999, 31–41 41 17 Q. Li, J. B. Vincent, E. Libby, H.-R. Chang, J. C. HuVman, P. D. W. Boyd, G. Christou and D. N. Hendrickson, Angew. Chem., Int. Ed. Engl., 1988, 27, 1731. 18 (a) L. J. Boucher and V. W. Day, Inorg. Chem., 1977, 16, 1360; (b) V. L. Pecoraro and W. M. Butler, Acta Crystallogr., Sect. C, 1986, 42, 1151; (c) X. Li and V. L. Pecoraro, Inorg. Chem., 1989, 28, 3403; (d) C. A. McAuliVe, R. G. Pritchard, L. Luaces, J. A. Garcia- Vazquez, J. Romero, M. R. Bermejo and A. Sousa, Acta Crystallogr., Sect. C, 1993, 49, 587; (e) C. Fraser, L. Johnston, A. L. Rheingold, B. S. Haggerty, G. K. Williams, J. Whelan and B. Bosnich, Inorg. Chem., 1992, 31, 1835. 19 N. A Law, T. E. Machonkin, J. P. McGorman, E. J. Larson, J. W. Kampf and V. L. Pecoraro, J. Chem. Soc., Chem. Commun., 1995, 2015. 20 M. R. Bermejo, A. Castineiras, J. C. Garcia-Monteagudo, M. Rey, A. Sousa, M. Watkinson, C. A. McAuliVe, R. G. Pritchard and Roy L. Beddoes, J. Chem. Soc., Dalton Trans., 1996, 2935. 21 M. R. Bermejo, A. Garcia-Deibe, J. Sanmartin, A. Sousa, N. Aurangzeb, C. E. Hulme, C. A. McAuliVe, R. G. Pritchard and M. Watkinson, J. Chem. Soc., Chem. Commun., 1994, 645. 22 L. J. Boucher and C. G. Coe, Inorg. Chem., 1975, 14, 1289. 23 N. Aurangzeb and C. A. McAuliVe, unpublished results. 24 G. M. Sheldrick, SHELXS 86, in Crystallographic Computing 3, eds. G. M. Sheldrick, C. Krueger and R. Goddard, Oxford University Press, 1985, pp. 175–189. 25 P. T. Beurskens, DIRDIF, Direct methods for diVerence structures, an automatic procedure for phase extension and refinement of diVerence structures, Technical Report 1984/1, Crystallography Laboratory, Toernooiveld, 1984. 26 TEXSAN TEXRAY, Structure Analysis Package, Molecular Structure Corporation, The Woodlands, TX, 1985. Paper 8/05555F

ISSN:1477-9226    

DOI:10.1039/a805555f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1996, Pages 35–38 35 A new one-dimensional system starting from a trinuclear copper(II) complex and selenocyanate as bridging ligand. Comparison with the thiocyanate analogue Joan Ribas,*,a Carmen Diaz,a Xavier Solans b and Mercé Font-Bardía b a Departament de Química Inorgànica, Universitat de Barcelona, Diagonal 647, 08028-Barcelona, Spain b Departament de Cristallografia i Mineralogia, Universitat de Barcelona, Martí i Franqués s/n, 08028-Barcelona, Spain A new copper(II) bridging-selenocyanate complex has been synthesized starting from the known trinuclear copper(II) complex, [{Cu(tmen)}2{m-Cu(pba)}][ClO4]2 [tmen = N,N,N9,N9-tetramethylethane-1,2-diamine and H4pba = N,N9-propane-1,3-diylbis(oxamic acid)].The reaction of an aqueous solution of this complex with KSeCN gave good crystals of the new one-dimensional compound in which two selenocyanate ligands are linked to each terminal copper(II) ion. Both these ligands are weakly Se-coupled to the central copper(II) atom of another trinuclear entity giving a pseudo one(two)-dimensional structure.The magnetic behaviour was recorded between 300 and 4 K, showing strong intramolecular antiferromagnetic coupling (J = 2356 cm21, gC = 2.20 and gT = 2.21; C = central, T = terminal Cu atom) and very weak intermolecular ferromagnetic coupling. In a previous study 1 we reported a new one-dimensional system made by treating the known2 trinuclear copper(II) complex [{Cu(tmen)}2{m-Cu(pba)}][ClO4]2 with NaSCN [tmen = N,N,N9,N9-tetramethylethane-1,2-diamine and H4pba = N,N9- propane-1,3-diylbis(oxamic acid)].This new complex is a chain of trinuclear entities with two N-thiocyanato terminal ligands linked by the sulfur atom to the central copper(II) atom of another trinuclear entity. The interest in it was the possibility of ferromagnetic interactions between the trinuclear entities, bearing in mind the known tendency of thiocyanate anion to give ferromagnetic coupling when it bridges in end-to-end mode.This tendency is very clear for nickel(II) 3–10 complexes but less pronounced for copper(II) due to the low number of complexes reported.11 Consequently the behaviour of this system could be the opposite to that reported by Drillon et al.,12,13 where copper( II) trimers were coupled in an antiferromagnetic arrangement to give a new ferrimagnetic model: A2Cu3(PO4)4 (A = Ca or Sr). As indicated in our previous work,1 the same reaction with NaN3 allows us to prepare only a dinuclear copper(II) complex, already reported by Kahn and co-workers,14 and with NaNCO the reaction did not take place.Working with KSeCN we were able to prepare a new complex, very similar to the thiocyanato analogue, in which the only difference lies in the distance between the selenium atom of the terminal selenocyanate ligand and the copper(II) central atom of the neighbouring entity, compared with the distance from the sulfur atom to a copper(II) ion: Se ? ? ? Cu (central) 3.06 and 3.465; S ? ? ? Cu (central) 3.014 and 3.617 Å.This paper describes the first selenocyanato-bridged one-dimensional complex prepared from trinuclear copper(II) entities with the aim of studying the co-ordination and magnetic properties of the end-to-end selenocyanato bridges. Experimental Synthesis The salt Na2[Cu(pba)]?6H2O was synthesized as previously described.15 [{Cu(tmen)(SeCN)}2{Ï-Cu(pba)}] 1.An ethanolic solution of N,N,N9,N9-tetramethylethane-1,2-diamine (0.48 g, 4.14 mmol) was added to a stirred solution of copper(II) nitrate hexahydrate (1 g, 4.14 mmol) in ethanol (50 cm3). Then solutions of Na2[Cu(pba)]?6H2O (0.89 g, 2.07 mmol) in water (50 cm3) and KSeCN (0.3 g, 2.07 mmol) in water (5 cm3) were added consecutively. The resultant blue solution was filtered to remove any impurity and left to evaporate slowly at room temperature. Blue monocrystals suitable for X-ray determination were collected after 2 weeks (yield ca. 60%) (Found: C, 28.5; H, 4.9; N, 12.6. Calc. for C21H42Cu3N8O8Se2: C, 28.6; H, 4.8; N, 12.7%). Spectral and magnetic measurements Infrared spectra were recorded on a Nicolet 520 FT-IR spectrometer. Magnetic susceptibility measurements in the range 300–4 K were made on polycrystalline samples with a pendulum-type magnetometer (Manics DSM8) equipped with a helium continuous-flow cryostat under a magnetic field of approximately 1.5 T.Diamagnetic corrections were estimated from Pascal tables. The ESR spectra were recorded on a Bruker ES200 spectrometer at X-band frequency, with an Oxford liquid-helium cryostat for variable temperatures. Crystallography A prismatic crystal (0.20 × 0.10 × 0.10 mm) of complex 1 was selected and mounted on an Enraf-Nonius CAD4 four-circle diffractometer. Unit-cell parameters were determined from automatic centring of 25 reflections (12 < q < 21) and refined by least-squares methods.Intensities were collected with graphite-monochromatized Mo-Ka radiation (l = 0.710 69 Å), using the w–2q scan technique. 5505 Reflections were measured in the range 1.77 < q < 29.948, 3095 of which were assumed as observed [I > 2s(I)]. Three reflections were measured every 2 h as orientation and intensity control; no significant decay was observed. The crystallographic data and some features of the structure refinement are listed in Table 1.Lorentz-polarization but not absorption corrections were made. The structure was solved by Patterson synthesis using the SHELXS 86 computer program16 and refined (on F2) by full-matrix least squares using SHELXL 9317 with 5455 reflections (very negative intensities were not employed). The function minimized was36 J. Chem. Soc., Dalton Trans., 1997, Pages 35–38 Fig. 1 Atom-labelling scheme for [{Cu(tmen)(SeCN)}2{m-Cu(pba)}] 1 Sw(|Fo|2 2 |Fc|2)2, where w = [s2(I) + (0.2335P)2 + P]21, and P = (|Fo|2 + 2|Fc|2)/3.Values of f, f 9 and f 0 were taken from ref. 18. The extinction coefficient was 0.0003(4). The chirality of the structure was defined from the Flack 19 coefficient, which is equal to 0.01(3) for the results given. All H atoms were computed and refined with an overall isotropic thermal parameter, Fig. 2 View of the pseudo-two-dimensional entity for complex 1 and its thiocyanato analogue; X represents S or Se. The distances between X and Cu(2) are: Cu(2) ? ? ? S(1) 3.014, Cu(2) ? ? ? S(2) 3.617; Cu(2) ? ? ? Se(1) 3.059, Cu(1) ? ? ? Se(2) 3.465 Å Fig. 3 View of the packing of the layers in complex 1 using a riding model. Maximum shift/e.s.d. = 0.33, mean shift/ e.s.d. = 0.08. Maximum and minimum peaks in final difference synthesis 0.679 and 20.542 e Å23 respectively. The high equivalent anisotropic thermal parameters for O(8), C(12), C(15), C(16) and C(18) show a possible disorder of these atoms. 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., 1996, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/280. Results and Discussion Infrared spectrum The IR spectrum of complex 1 shows the typical and strong band corresponding to the selenocyanate ligand at 2075 cm21, two other very strong bands centred at 1600 and 1620 cm21 due to the co-ordinated oxamato group and between 1500 and 400 cm21 there are many bands attributable to the co-ordinated amines.Crystal structure The trinuclear entity of compound 1 is shown in Fig. 1. The terminal copper(II) ions Cu(1) and Cu(3) have 4 + 1 coordination, their basal planes being formed by two nitrogen atoms from the tmen ligand, two oxygen atoms from the oxamate and one nitrogen from the selenocyanate.The central Cu(2) atom has quasi-square-planar co-ordination, with a weak tetrahedral distortion calculated as 0.98. Selected bond Table 1 Crystallographic data for [{Cu(tmen)(SeCN)}2{m-Cu(pba)}] 1 Formula M Crystal system Space group T/8C a/Å b/Å c/Å U/Å3 Z Dc/g cm23 m(Mo-Ka)/cm21 F(000) No. parameters refined Ra wR2 b C21H42Cu3N8O8Se2 883.17 Orthorhombic P212121 25 12.444(6) 14.309(3) 19.303(5) 3437(2) 4 1.707 4.009 1772 381 0.052 0.147 a S Fo| 2 |Fc /S|Fo|. b {S[(Fo)2 2 (Fc)2]2/S(Fo)4}� �� .J.Chem. Soc., Dalton Trans., 1997, Pages 35–38 37 distances and angles are given in Table 2. All are very similar to those reported for the perchlorate 2 and thiocyanato 1 analogues. The Cu ? ? ? Cu separations within the trinuclear unit are Cu(1) ? ? ? Cu(2) 5.21 and Cu(2) ? ? ? Cu(3) 5.22 Å. The selenium atoms are weakly linked to a central Cu(2) of other trinuclear entities: Cu(2) ? ? ? Se(1) 3.06 and Cu(2) ? ? ? Se(2) 3.46 Å giving a one(two)-dimensional structure shown in Fig. 2. The angles in the bridge Cu(2)]SeCN]Cu(1,3) are as follows: Cu(3)]N]C 155.1 and Cu(2)]Se(2)]C 110.98 (in the moiety with short distance Cu ? ? ? Se 3.06 Å) and Cu(1)]N]C 143.9 and Cu(2)]Se(1)]C 130.98 (in the moiety with long distance Cu ? ? ? Se 3.46 Å). These pseudo-twodimensional layers are linked in the crystal (Fig. 3) by van der Waals forces. Magnetic properties Susceptibility measurements of [{Cu(tmen)(SeCN)}2{m- Cu(pba)}] are plotted in Fig. 4 as cmT vs. T. From room temperature down to 25 K there is a clear decrease in cmT values, followed by a plateau close to 0.45 cm3 K mol21, a typical value for an isolated copper(II) trimer (S = ��� ), assuming antiferromagnetic coupling between the copper(II) ions.2 At low temperature cmT decreases, as with the similar thiocyanato analogue.1 Taking into account the structure, also very similar, we can propose the same hypothesis: a small ferro- Fig. 4 Magnetic susceptibility of a polycrystalline sample of complex 1.The solid line shows the best fit from the expression for the magnetic susceptibility of an isotropically coupled copper(II) trimer Table 2 Selected bond distances (Å) and angles (8) for compound 1 Cu(1)]O(2) Cu(1)]O(1) Cu(1)]N(4) Cu(1)]N(3) Cu(1)]N(8) Cu(2)]N(2) Cu(2)]N(1) 1.988(7) 2.000(7) 2.016(9) 2.051(8) 2.23(2) 1.952(8) 1.957(7) Cu(2)]O(5) Cu(2)]O(6) Cu(3)]O(4) Cu(3)]O(3) Cu(3)]N(6) Cu(3)]N(5) Cu(3)]N(7) 1.992(7) 2.000(7) 1.989(8) 2.020(8) 2.030(8) 2.060(9) 2.180(11) O(2)]Cu(1)]O(1) O(1)]Cu(1)]N(4) O(2)]Cu(1)]N(3) O(1)]Cu(1)]N(3) N(4)]Cu(1)]N(3) N(3)]Cu(1)]N(8) O(2)]Cu(1)]N(8) O(1)]Cu(1)]N(8) N(4)]Cu(1)]N(8) O(2)]Cu(1)]N(4) N(2)]Cu(2)]N(1) N(2)]Cu(2)]O(5) O(5)]Cu(2)]O(6) 85.6(3) 93.5(3) 90.9(3) 167.3(4) 87.6(4) 98.0(4) 94.2(5) 94.5(4) 96.6(5) 169.2(4) 94.4(3) 170.5(4) 94.5(3) N(1)]Cu(2)]O(5) N(2)]Cu(2)]O(6) N(1)]Cu(2)]O(6) O(4)]Cu(3)]O(3) N(5)]Cu(3)]N(7) O(4)]Cu(3)]N(7) O(3)]Cu(3)]N(6) O(4)]Cu(3)]N(5) O(3)]Cu(3)]N(5) O(4)]Cu(3)]N(6) O(3)]Cu(3)]N(7) N(6)]Cu(3)]N(5) N(6)]Cu(3)]N(7) 84.7(3) 85.0(3) 171.7(4) 83.8(3) 99.1(4) 96.5(5) 155.5(4) 164.4(4) 91.9(3) 91.4(3) 106.2(4) 86.4(3) 98.2(4) magnetic coupling between trimers due to the selenocyanate bridging ligands.1,3–11 Indeed, in all complexes with thiocyanate or selenocyanate as bridging ligands the bond angles M]N]C and M]S(Se)]C are relatively close to 180 and 908 respectively.In our case these angles are 155.1 and 1118 respectively (considering that a bond forms when the distance Cu ? ? ? Se is 3.0 Å, but not at 3.47 Å).Ginsberg et al.7 and Duggan and Hendrickson 8 developed a valence bonding model by applying the Goodenough–Kanamori rules 20 or Anderson’s expanded orbital theory 21 to demonstrate the ferromagnetism of these pseudohalide complexes. The closer to those values (180 and 908) the stronger is the ferromagnetic coupling. In our case this coupling should be weak. With this hypothesis the total cmT value should tend to zero at 0 K, as is observed experimentally.The experimental data (from room temperature to 20 K) were fitted using the theoretical expression deduced from the spin Hamiltonian (1) with T1 and T2 being the terminal copper(II) H = 2J[ST1SC + SCST2] + bH[gT(ST1 + ST2) + gCSC] (1) ions and C the central one. In this Hamiltonian the interaction between the terminal ions is assumed to be nil. The mathematical expression is given in ref. 2.Minimizing R = S[(cmT)obs 2 (cmT)calc]2/S[(cmT)obs]2 leads to the values J = 2355.7 cm21, gC = 2.20 and gT = 2.21. The g value obtained from the powder EPR spectra at room temperature is 2.13. The band is very broad and isotropic; at lower temperature the signal becomes sharp and anisotropic, with g^ = 2.30 and g|| = 2.15 at 4 K. As previously pointed out 2 it is highly probable that the intermolecular exchange-averaging condition is always fulfilled for trinuclear copper(II) compounds.This shows that special care should be exercised when interpreting powder spectra of such complexes. Comparison with perchlorate and thiocyanate analogues. Comparing J values for the three complexes (perchlorate, thiocyanate and selenocyanate), small (but significant) differences are found: J = 2380, 2332 and 2356 cm21, respectively. There Fig. 5 Schematic representation (taken from crystal data) of the three complexes, perchlorate (top), selenocyanate (middle) and thiocyanate (bottom)38 J.Chem. Soc., Dalton Trans., 1997, Pages 35–38 Table 3 Main molecular parameters which affect the antiferromagnetic coupling for the three similar complexes (perchlorate, selenocyanate and thiocyanate) Deviation c/Å J/cm21 Cu]Cu]Cu/8 Planes angle a/8 Torsion at Cu(2) b/8 Cu(2) Cu(1) Cu(3) ClO4 2 SeCN2 SCN2 2380 2356 2332 169.5 165.6 163.3 164.7 163.3 160.7 6.29 0.9 1.2 0.034 20.153 20.153 20.107 20.204 0.363 20.004 0.351 0.128 a Formed by the two mean oxamato-like planes which contain the copper(II) ions.b Square planar �Æ tetrahedral distortion. c Deviation of the copper(II) ion from the mean plane created by the four basal atoms (in square-planar or square-pyramidal co-ordination). is, thus, a gradual decrease in the order ClO4 2 > SeCN2 > SCN2. The J variation from one complex to another is ca. 30 cm21. Fig. 5 shows a schematic representation (taken from crystal data) of the three complexes, and in Table 3 we have gathered the most significant differences between them.As previously reported 2 the most antiferromagnetic coupling occurs when the trinuclear entity is completely planar assuming that all copper(II) ions are in the centre of the square-planar coordination. This is an ideal case, not found even in the perchlorate. From Table 3 it is seen that the perchlorate has the smallest deviations from the co-ordination planes, but the torsion (square planar–tetrahedral) in Cu(2) is the greatest (6.298); in contrast, this torsion for the thiocyanate and selenocyanate is less pronounced (only 18) but the deviations from the mean planes for Cu(1), Cu(2) and Cu(3) are more marked and almost equal for both pseudohalides.Thus, the most important difference between the three complexes lies in the angles formed by the three copper(II) ions and/or in the angles formed by the two mean oxamato-like planes [which contain the copper(II) ions]. These angles create a deviation from planarity in the order SCN2 > SeCN2 > ClO4 2, which may be the main factor that diminishes the antiferromagnetic coupling because it reduces the overlap between magnetic orbitals.The experimental J values are consistent with this explanation. Acknowledgements Financial support for this work was given by the Dirección General de Investigación Científica y Técnica through Grant PB93/0772. References 1 J. Ribas, C. Diaz, X. Solans and M. Font-Bardía, Inorg.Chim. Acta, 1995, 231, 229. 2 R. Costa, A. García, J. Ribas, T. Mallah, Y. Journaux, J. Sletten, X. Solans and V. Rodríguez, Inorg. Chem., 1993, 32, 3733. 3 J. G. Haasnoot, W. L. Driessen and J. Reedijk, Inorg. Chem., 1984, 23, 2803. 4 R. Vicente, A. Escuer, J. Ribas and X. Solans, J. Chem. Soc., Dalton Trans., 1994, 259. 5 B. W. Dockum and W. M. Reiff, Inorg. Chem., 1982, 21, 2613. 6 T. Rojo, R. Cortés, L. Lezama, M. I. Arriortua, K. Urtiaga and G. Villeneuve, J. Chem.Soc., Dalton Trans., 1991, 1779. 7 A. P. Ginsberg, R. L. Martin, R. W. Brookes and R. C. Sherwood, Inorg. Chem., 1972, 11, 2884. 8 D. M. Du Hendrickson, Inorg. Chem., 1974, 13, 2929. 9 M. Monfort, J. Ribas and X. Solans, Inorg. Chem., 1994, 33, 4271. 10 M. Monfort, C. Bastos, C. Diaz, J. Ribas and X. Solans, Inorg. Chim. Acta, 1994, 218, 185. 11 R. Vicente, A. Escuer, E. Peñalba, X. Solans and M. Font-Bardía, Inorg. Chim. Acta, in the press and refs. therein 12 M. Drillon, E. Coronado, M. Belaiche and R. L. Carlin, J. Appl. Phys., 1988, 63, 3551. 13 M. Drillon, M. Belaiche, J. M. Heintz, G. Villeneuve, A. Boukhari and J. Aride, in Organic and Inorganic Low-Dimensional Crystalline Materials, eds. P. Delhaes and M. Drillon, Plenum, New York, 1987, p. 421. 14 I. Bkouche-Waksman, S. Sikorav and O. Kahn, J. Crystallogr. Spectrosc. Res., 1983, 13, 60. 15 K. Nonoyama, H. Ojima and M. Nonoyama, Inorg. Chim. Acta, 1976, 20, 127. 16 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467. 17 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993. 18 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, pp. 99, 100, 149. 19 H. D. Flack, Acta Crystallogr., Sect. A, 1983, 39, 876. 20 J. B. Goodenough, Magnetism and the Chemical Bond, Interscience, New York, 1963, pp. 165–185 21 P. W. Anderson, Magnetism, Academic Press, New York, 1963, vol. 1, ch. 2. Received 25th June 1996; Paper 6/04430A

ISSN:1477-9226    

DOI:10.1039/a604430a

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 37–43 37 Solution behaviour, kinetics and mechanism of the acid-catalysed cyclopalladation of imines * Montserrat Gómez, Jaume Granell and Manuel Martinez Department de Química Inorgànica, Facultat de Química, Universitat de Barcelona, Martí i Franquès 1-11, E-08028 Barcelona, Spain The cyclometallation reactions of N-benzylidene-benzylamines, -anilines and -propylamine with palladium acetate have been studied in acetic acid solution.Carbon–hydrogen electrophilic bond activation occurs to produce different types of metallacycles, given the polyfunctional nature of the ligands selected. The cyclometallated compounds formed indicate that the stability of the final species is, with respect to the activated C]H bond, in the order: five-membered aromatic endo > six-membered aliphatic endo > five-membered aromatic exo, >>> five-membered exo, four-membered. The nature of the final cyclometallated compounds in acetic acid solution has been ascertained via 1H NMR spectroscopy; as a whole the spectra are complex, indicating that the nature of the cyclometallated species in solution is not simple, and that a wide variety of compounds is present depending on the imine used.The metallation reactions have been monitored kinetically via UV/VIS spectroscopy at different temperatures and pressures in order to establish the mechanism through which these acid-assisted reactions occur.Although the thermal activation parameters cover a wide range of values (DH‡ = 49 to 73 kJ mol21, DS‡ = 252 to 2138 J K21 mol21), the activation volume is in a very narrow range, 215 ± 3 cm3 mol21. The results are interpreted as the formation of a highly ordered four-centred transition state, involving the C]H and Pd]O (acetato) bonds, which is found to be very sensitive to the presence of any protons that could enhance the leaving-group characteristics of the MeCO2H ligand, converting it into its protonated MeCO2H2 1 form.Although cyclometallation reactions on palladium(II) complexes have been thoroughly studied by a number of research groups in view of their interest in organic synthesis,1 design of new metallomesogens 2 and antitumoral drugs,3 usage for enantiomeric excess determination,4 etc., the number of these studies dealing with kinetic and mechanistic information is very limited.5 Very little information is available about the nature of the species existing in the reaction solutions.6 Although it is generally assumed that cyclopalladated compounds maintain their dimeric structure in solution, it is difficult to think in these terms when the reactions are carried out in acetic acid, one of the standard solvents for such cyclopalladation reactions.In this respect, especially enlightening is the characterization of a number of intermediate species arising from the cyclometallation reaction on primary amines in non-protic solvents.7 Our interests have been centred on this type of reaction on imine ligands as well as in the mechanisms operating in organometallic reactions involving the activation of C]X bonds on platinum(II) and dinuclear rhodium(II) complexes.8,9 The final goal of all these investigations is the study of the importance of the relative influences of steric and electronic factors that could tune the reactivity and thermodynamic preferences of the reactions involved in different processes.10,11 In this paper we report a kinetic study of the influence of temperature and pressure on the cyclometallation of a wide variety of imines, X (Scheme 1), by palladium acetate in acetic acid as solvent.These imines have been selected to allow comparison of the metallation of aromatic versus aliphatic carbon atoms, formation of endo (1x) versus exo (2x) metallacycles, and formation of five- or six-membered metallacycles (Scheme 2). By doing so, the electronic and steric influence of the substituents in these processes has been examined.A complete study of * Supplementary data available: observed pseudo-first order rate constants. For direct electronic access see http://www.rsc.org/suppdata/dt/ 1998/37/, otherwise available from BLDSC (No. SUP 57315, 5 pp.) or the RSC Library. See Instructions for Authors, 1998, Issue 1 (http:// www.rsc.org/dalton). the nature of the final cyclometallated compounds in acetic acid solution has also been carried out; our findings indicate that important differences exist in the final species in solution, depending on the imine ligand.Results and Discussion Compounds The reaction of palladium acetate with imines A–N in acetic acid has been studied. With A–J and L, dinuclear, acetatobridged, complexes 1x are obtained. Despite the fact that some of the ligands can undergo metallation at different carbon atoms, endo-metallacycles (Scheme 2) were selectively formed in all these cases (Schemes 3 and 4).It should be noted that, although imines E and L, those containing a nitro substituent in meta position, could afford two different five-membered endo-metallacycles, only the metallation of the less hindered Caromatic]H bond was observed. When the cyclopalladation reaction was performed with imine K a mixture of endo- and exo-palladacycles was obtained, as found in toluene solution. The formation of the exo-cyclic compound with this imine can be explained by steric effects.The presence of a nitro group in the ortho position of the aromatic ring undergoing metallation (if an endo-metallacycle is formed) prevents the adoption of a planar conformation between the imine moiety and this aromatic ring, somehow hindering the formation of the proposed four-centered transition state (see below), in consequence the exo-metallacycle is also obtained. Imines M and N could afford both five-membered exometallacycles, by activation of a Caromatic]H bond, and sixmembered endo-metallacycles, by activation of a Caliphatic]H bond.Although, in general, a strong tendency to form fivemembered metallacycles and preferential activation of aromatic over aliphatic C]H bonds is widely accepted,12 some exceptions to these rules are known.13 In this case and for the reactions performed under mild conditions (40 8C, 3 h), the fivemembered exo-metallacycles, 2x, were selectively obtained; nevertheless, when the same reactions were carried out under38 J.Chem. Soc., Dalton Trans., 1998, Pages 37–43 Scheme 1 Arrows indicate the activated position CH N Prn CH N CH N MeO CH N NO2 CH N O2N CH N O2N CH N Me Me Me CH N Me Me Me Me CH N CH2 CH N CH2 Cl CH N CH2 Cl NO2 CH N CH2 O2N CH N Me CH2 Me Me CH N Me CH2 Me Me Me A B C D E F G H N M L K J I more energetic conditions (80 8C, 1 h), the six-membered endometallacycles 1x were obtained via activation of Caliphatic]H bonds.All the final isolated compounds have been previously described, and characterized by comparison with their 1H NMR spectra already published.8b,e,11,13c,14 Nevertheless, as regards the structure in solution of the acetate- or halidebridged cyclopalladated compounds, very little data are available. It has been generally assumed, by analysis of the NMR spectra, that these compounds maintain their dimeric structure in solution, but they are usually recorded only in poorly coordinating, non-protic, solvents such as CDCl3.Recently, it has been shown by 1H NMR spectroscopy that even acetone is able to break the bromo bridge of the cyclopalladated complex [{PdBr[C6H4CH(Me)NH2]}2] leading to mononuclear species.15 In this respect, 1H NMR spectra of acetato-bridged complexes derived from imines similar to those used in this study indicate the existence, in solution, of two isomeric forms. When C5D5N is added to these solutions, only the mononuclear complex trans-[Pd(O2CMe)L(C5D5N)] (L = metallated imine) is observed, indicating that the two isomeric forms of these compounds are solely related by their dinuclear core.8c In view of these data, we have studied the structure of the cyclopalladated compounds in solution by 1H NMR spectroscopy both in nonprotic solvents and in acetic acid.From the NMR data in different deuteriated solvents of the five-membered metallacycles containing a Caromatic]Pd bond (Table 1, compounds 1i and 1l–1n) the following conclusions can be drawn: (i) in solution only one isomer can be observed; (ii) the acetate CH3 signal appears as a singlet, which indicates a trans arrangement of the C,N chelate around the {Pd2- (m-O2CMe2)2} core; (iii) the two methylenic protons (CH2N) are different chemically in all the studied solvents, two doublets, AB spin system, in accordance with a folded open-book dimeric structure, which has been previously found by X-ray diffraction studies of related compounds;16 (iv) the chemical shift of the imine proton indicates that the imine conformation is E (upfield from the free imine) in all the endo-cyclic compounds and Z (downfield from the free imine) in the exo-cyclic derivatives.8a In contrast, the six-membered derivatives, those containing CH2]Pd bonds, have different structures in solution depending on the nature of the solvent (Table 1, compounds 1g, 1m and 1n).In aprotic solvents, such as CDCl3, [2H8]toluene or (CD3)2CO, broad signals for the methylenic protons (those bonded to palladium and nitrogen) are observed.This is characteristic of the hampered movement of the six-membered metallacycles, as a consequence of their folded structure; in some cases even two isomers can be observed. When C5D5N is added the NMR spectra show only one compound in solution, obviously the mononuclear complex trans-[Pd(O2CMe)L(py)],J. Chem. Soc., Dalton Trans., 1998, Pages 37–43 39 and all the signals became narrow.These results indicate that the different isomers found are directly related with the dinuclear core of these complexes in solution. When the 1H NMR spectra of these compounds are recorded in deuteriated acetic acid or in the presence of CF3CO2H they are very similar to those of the pyridine derivatives; in acidic media the acetatobridged core seems to be completely broken and the mononuclear complexes trans-[Pd(O2CMe)L(solv)] (where solv is a solvent molecule) are formed.Moreover, the substitution of the acetate, assisted by the acidic medium, by the poorly coordinating CF3CO2 2 group in the six-membered metallacycles is very fast; showing the important bridge lability of the dinuclear compounds. This behaviour of the six-membered metallacycles in solution can be explained by their structural features; in these compounds the six-membered metallacycle is non-planar, adopting a half-skew-chair conformation, as has been shown by crystal structure determination of [Pd{1-CH2-2- (HC]] NPh)-3,5-Me2C6H2}Br(PPh3)].13c This important distor- Scheme 2 CH N Pd R¢ O O 2 five-membered endocycle CH2 N CH Pd R O O 2 O O 2 six-membered endocycle five-membered exocycle H2C R¢¢ CH N Pd Scheme 3 (i) Palladium acetate, MeCO2H, 80 8C, 1 h.The exocyclic compound 2k is also formed with imine K C H N R' C H N R' Pd R R O O 1f R= 5-NO2, R'= Ph 1i R= H, R'= CH2Ph 1j R= 5-Cl, R'= CH2Ph B R= H, R'= Ph 2 1b R= H, R'= Ph 1c R= 5-MeO, R'= Ph C R= 4-MeO, R'= Ph 1a R= H, R'= Prn 1e R= 4-NO2, R'= Ph 1d R= 3-NO2, R'= Ph 1k R= 3-NO2, R'= CH2Ph A R= H, R'= Prn D R= 2-NO2,R'= Ph E R= 3-NO2, R'= Ph F R= 4-NO2, R'= Ph I R= H, R'= CH2Ph J R= 4-Cl, R'= CH2Ph K R= 2-NO2, R'= CH2Ph L R= 3-NO2, R'= CH2Ph 1l R= 4-NO2, R'= CH2Ph ( i ) tion of the metallacycle increases the steric congestion between both moieties of the molecule so decreasing the stability of the dinuclear core.In the same context we have also studied the 1H NMR spectra of palladium acetate in acetic acid solution at different concentrations in order to establish the nature of the starting material of the cyclometallation reaction in solution.Although palladium acetate is a trinuclear compound in the solid state, with all the acetato groups acting as bridging ligands,17 in solution mixtures of trinuclear closed compounds (only with acetato bridges) and open complexes (containing terminal and bridging acetate ligands) have been observed.6,11 Only one signal at d 2.05 is observed for a saturated solution of palladium acetate in CD3CO2D, but when the spectra are recorded at lower concentrations new signals of low intensity appear, their number increasing with dilution of the sample.Some of the signals appeared at high fields, d 1.3–1.1, which could be assigned to terminal acetato groups.8c All these results suggest that the formation of solvato complexes with palladium acetate have occurred in solution, affording polynuclear species with both bridging and terminal acetate ligands.In this respect cyclopalladation of primary amines has, recently been proposed to occur via [Pd(O2CMe)2L92] and [{Pd(O2CMe)- (m-O2CMe)L9}2] (L9 = primary amine) intermediates, both isolated in the solid state; the crystal structure of the latter has been determined.7 Mechanism The reaction of palladium acetate with imines A–N in acetic acid has been studied kinetically by means of UV/VIS spectroscopy.The reactions were followed via 1H NMR spectroscopy under the same kinetic conditions in order to establish the presence of the cyclometallated compound as the product of the absorbance change monitored. The products isolated from the reaction mixture once C]H activation is achieved are in excellent agreement with those obtained when the reaction was carried out in toluene solution at 60 8C.11 Scheme 4 (i) Pd(O2CMe2)2, MeCO2H, 40 8C, 3 h; (ii) Pd(O2CMe2)2, MeCO2H, 80 8C, 1 h C H N Me Me Me C N H2C Me Me Me H R' R Pd O O C H N R' CH2 Me Me Pd O O C H N R' CH2 Me Me Pd O O 2n R= Me 2m R= H 2 1g R'= Ph M R'= CH2Ph 2 N R'= CH2C6H4Me-2 G R'= Ph H R'= 2-MeC6H4 2 1h R'= 2-MeC6H4 1n R= Me 1m R= H ( i ) ( ii ) (ii )40 J.Chem. Soc., Dalton Trans., 1998, Pages 37–43 Table 1 Proton NMR data a of selected cyclometallated compounds in different deuteriated solvents CH N (CH2) n Pd Pd 6 7 8 9 10 1 2 3 4 5 Compound 1g (CD3CO2D) CH3 2.46 (s, 3 H, Me10) 2.39 (s, 3 H, Me8) MeCO2 b CH2 3.35 (s, 2 H) Aromatic 7.60–7.25 (br m, 5 H) 7.20 (s, 1 H, H9) 6.90 (s, 1 H, H7) HC]] N 8.20 (s, 1 H) 1g 1 C5D5N (CDCl3) 2.30 (s, 3 H, Me10) 2.20 (s, 3 H, Me8) 1.40 (s, 3 H, MeCO2) 2.85 (s, 2 H) 7.50–7.16 (m, 5 H) 6.85 (s, 1 H, H9) 6.72 (s, 1 H, H7) 7.97 (s, 1 H) 1i (CD3CO2D) MeCO2 b 4.65 [br d, 2 H, 2J(HH) = 15.8] 4.16 [br d, 2 H, 2J(HH) = 15.8] 7.60–7.0 (br m, 20 H) c (CDCl3) 2.18 (s, 6 H, MeCO2) 4.58 [d, 2 H, 2J(HH) = 15.2] 4.04 [d, 2 H, 2J(HH) = 15.2] 7.30–6.80 (br m, 20 H) c 1i 1 C5D5N (CDCl3) 1.85 (s, 3 H, MeCO2) 4.80 (s, 2 H) 7.33–7.20 (m, 5 H) 7.09 [d, 1 H, 3J(HH) = 7.2, H10] 6.90–6.80 (m, 2 H, H9, H8) 6.17 [d, 1 H, 3J(HH) = 7.2, H7] 7.62 (s, 1 H) 1l (CD3CO2D) MeCO2 b 4.65 (br d, 2 H) 4.14 (br d, 2 H) 8.10–7.00 (br m, 18 H) c (CDCl3) 2.20 (s, 6 H, MeCO2) 4.58 [d, 2 H, 2J(HH) = 15.2] 4.04 [d, 2 H, 2J(HH) = 15.2] 7.30–6.80 (br m, 18 H) c 1m (CD3CO2D) 2.28 (s, 3 H, Me10) 2.23 (s, 3 H, Me8) MeCO2 b 3.05 (s, 2 H) Pd]CH2 4.93 (s, 2 H) CH2N 7.30–7.20 (br m, 5 H) 7.07 (s, 1 H, H9) 6.80 (s, 1 H, H7) 7.96 (s, 1 H) 1m 1 C5D5N (CDCl3) 2.24 (s, 6 H, Me10, Me8) 1.94 (s, 3 H, MeCO2) 2.51 (s, 2 H) Pd]CH2 5.05 (s, 2 H) CH2N 7.65 [d, 3J(HH) = 7.2, 2 H] 7.40–7.32 (m, 3 H) 6.72 (s, 1 H, H9) 6.61 (s, 1 H, H7) 7.50 (s, 1 H) 1m 1 CF3CO2H ([2H8]toluene) 2.03 (s, 3 H, Me10) 1.87 (s, 3 H, Me8) 2.92 (s, 2 H) Pd]CH2 4.49 (s, 2 H) CH2N 7.25–7.05 (br m, 5 H) 6.81 (s, 1 H, H9) 6.50 (s, 1 H, H7) 7.36 (s, 1 H) 1n (CD3CO2D) 2.32 (s, 3 H, Me10) 2.28 (s, 3 H, Me8) 2.11 (s, 3 H, MeCO2) 2.06 (s, 3 H, Me5) 3.20 (s, 2 H) Pd]CH2 4.96 (s, 2 H) CH2N 7.26 (br m, 4 H) 7.10 (s, 1 H, H9) 6.77 (s, 1 H, H7) 7.70 (s, 1 H) 1n 1 C5D5N (CDCl3) 2.37 (s, 3 H, Me10) 2.27 (s, 3 H, Me8) 2.08 (s, 3 H, Me5) 1.91 (s, 3 H, MeCO2) 2.77 (s, 2 H) Pd]CH2 5.15 (s, 2 H) CH2N 7.34–7.25 (br m, 4 H) 6.90 (s, 1 H, H9) 6.72 (s, 1 H, H7) 7.60 (s, 1 H) 2m (CDCl3) 2.24 (s, 6 H, MeCO2) 2.14 (s, 6 H, Me8) 2.09 (br s, 6 H, Me10) 1.48 (br s, 6 H, Me6) 4.22 [br d, 2 H, 2J(HH) = 17.5] 3.94 [br d, 2 H, 2J(HH) = 17.5] 6.6–6.9 (m, 12 H) 8.53 (s, 2 H) ([2H8]toluene) 2.20 (s, 6 H, MeCO2) 2.04 (s, 6 H, Me8) 1.75 (br s, 6 H, Me10) 1.44 (br s, 6 H, Me6) 4.18 [br d, 2 H, 2J(HH) = 18.0] 4.05 [br d, 2 H, 2J(HH) = 18.0] 7.2–7.1 (m, 2 H) 6.90–6.80 (m, 4 H) 6.6–6.4 (m, 6 H) 8.65 (s, 2 H) (CD3CO2D) 2.23 (s, 6 H, Me8) 2.10 (br s, 6 H, Me10) 2.06 (s, 6 H, MeCO2) 1.49 (br s, 6 H, Me6) 4.37 [br d, 2 H, 2J(HH) = 18.0] 3.98 [br d, 2 H, 2J(HH) = 18.0] 7.0–6.6 (m, 12 H) 8.61 (s, 2 H) 2m 1 C5D5N (CDCl3) 2.29 (s, 3 H, Me8) 2.20 (s, 6 H, Me6, Me10) 1.88 (s, 3 H, MeCO2) 4.56 (s, 2 H) 6.92 [t, 1 H, 3J(HH) = 7.5, H3] 6.89 (s, 2 H, H7, H9) 6.79 [d, 1 H, 3J(HH) = 7.5, H5] 6.74 [t, 1 H, 3J(HH) = 7.5, H4] 6.11 [t, 1 H, 3J(HH) = 7.5, H2] 8.84 (s, 1 H) 2n (CDCl3) 2.26 (s, 6 H, Me8) 2.15 (s, 6 H, MeCO2) 2.07 (br s, 6 H, Me10) 1.84 (s, 6 H, Me5) 1.69 (br s, 6 H, Me6) 4.15 [br d, 2 H, 2J(HH) = 18.0] 3.45 [dd, 2 H, 2J(HH) = 18.0] 6.9–6.6 (m, 10 H) 8.55 (s, 2 H) (CD3CO2D) 2.24 (s, 6 H, Me8) 2.10 (br s, 6 H, Me10) 2.06 (s, 6 H, MeCO2) 1.84 (s, 6 H, Me5) 1.64 (br s, 6 H, Me6) 4.25 [br d, 2 H, 2J(HH) = 17.5] 3.57 [br d, 2 H, 2J(HH) = 17.5] 7.0–6.6 (m, 5 H) 8.61 (s, 1 H) (CD3COCD3) 2.25 (s, 3 H, Me8) 2.12 (br s, 3 H, Me10) 2.06 (s, 3 H, MeCO2) 1.86 (s, 3 H, Me5) 1.69 (br s, 3 H, Me6) 4.27 [br d, 2 H, 2J(HH) = 18.0] 3.73 [br d, 2 H, 2J(HH) = 18.0] 7.0–6.6 (m, 5 H) 8.60 (s, 1 H) ([2H8]toluene) 2.22 (s, 6 H, MeCO2) 1.96 (s, 6 H, Me8) 1.74 (br s, 6 H, Me10) 1.66 (s, 6 H, Me5) 1.64 (br, s, 6 H, Me6) 4.22 [br d, 2 H, 2J(HH) = 18.0] 4.19 [br d, 2 H, 2J(HH) = 18.0] 7.2–7.0 [d, 2 H, 2J(HH) = 18.0] 6.90–6.45 (m, 8 H) 8.68 (s, 1 H) a d in ppm with respect to internal SiMe4; coupling in Hz; see figure for hydrogen labels.Abbreviations: br, broad; s, singlet; d, doublet; t, triplet; m, multiplet. b Resonance obscured by solvent. c Obscured by aromatic.J.Chem. Soc., Dalton Trans., 1998, Pages 37–43 41 Table 2 Kinetic and activation parameters for the cyclometallation reaction of palladium acetate with the imines in Scheme 1 in neat acetic acid solution Imine A B C D E F G H I J K L M N Metallated compound 1a 1b 1c 1d 1e 1f 1g 1h 1i 1j 1k 1 2k 1l 2m 2n 103k298/s21 1.2 1.1 1.5 3.0 a 2.6 a 2.3 1.3 2.7 0.15 1.5 0.16 b 0.47 a 1.3 0.22 a DH‡/kJ mol21 65 ± 7 73 ± 6 73 ± 1 73 ± 4 73 ± 5 66 ± 1 58 ± 3 65 ± 12 69 ± 6 59 ± 5 60 ± 7 65 ± 13 49 ± 5 69 ± 6 DS‡/J K21 mol21 287 ± 21 260 ± 8 256 ± 4 252 ± 12 255 ± 5 275 ± 3 2108 ± 10 275 ± 30 290 ± 18 299 ± 17 2123 ± 22 299 ± 40 2138 ± 16 291 ± 18 DV‡/cm3 mol21 (T/K) 217 ± 1 (298) 218 ± 1 (298) 215 ± 2 (293) 212 ± 1 (293) 216 ± 3 (293) 217 ± 3 (293) 217 ± 2 (298) 211 ± 1 (293) 215 ± 1 (308) 216 ± 2 (318) 217 ± 2 (293) 217 ± 1 (318) 216 ± 3 (318) 215 ± 4 (303) 211 ± 1 (293) 213 ± 2 (303) a A statistical factor of 2 has been applied.b A statistical factor of 1.1 has been applied according to the ratio [1k] : [2k] = 2.5 : 1 determined under these conditions in toluene solution. Blank experiments run in the absence of palladium acetate demonstrated the stability of the selected imines under the monitoring conditions.Imines A–F, I, J and L produce the endo five-membered metallacycles (1x) by metallation of the Caromatic]H bond; the cyclopalladation reaction of imine K produced a mixture of the five-membered endo- and exometallacycles 1k, 2k as in toluene solution. Imines G and H produce also endo (1g, 1h) six-membered metallacycles via activation of a Caliphatic]H bond, while M and N produce fivemembered exo (2m, 2n) compounds (Schemes 3 and 4).Further reaction of all the series of exo-metallacycles 2k, 2m and 2n in acetic acid rapidly produces a deep red solution that further evolves to produce the corresponding endo (1k, 1m and 1n) metallacycles in quantitative yield. This reaction, which has not been detected in toluene solution, is currently under study.Parallel 1H NMR monitoring of the kinetically studied reaction mixtures enabled us to insure that under the conditions described in this paper only the first reaction has been followed; by doing so a clear comparison with the previously studied cyclometallation reactions in toluene solution has been achieved.11 Two reaction steps are involved in the overall reaction studied: first co-ordination of the imine to the palladium acetate, and secondly C]H bond activation leading to the formation of the final cyclometallated product.The formation of the coordination complex is believed to be fast and not detectable under the monitoring condition used in the cyclometallation kinetic study,18 consequently all the values of the pseudo-firstorder rate constants, kobs, correspond to the bond-activation step (SUP 57315). From these observed rate constants, the firstorder constants at 298 K, thermal activation parameters, and activation volumes collected in Table 2 were derived.Both the enthalpy and entropy of activation are within the range of values observed for other C]H bond activation via electrophilic substitution in the presence or absence of protic solvents.9c–e,11,12e The values determined for DV‡ (Fig. 1) are in perfect agreement with those found for acid-assisted electrophilic C]H bond-activation reactions on rhodium(II) dimers,9c–e indicative of a compressed arrangement in the transition state.All the results obtained are consistent with the presence of a highly ordered transition state as that shown in Scheme 5; in this transition state the neighbouring terminal acetato group, protonated by the acidic medium, accepts a proton from the imine to produce the MeCO2H2 1 species that acts as an excellent leaving group. In good agreement with the mechanism here proposed are the recently found intermediates for the cyclopalladation of primary amines, [Pd(O2CMe)2L92] and [{Pd(O2CMe)- (m-O2CMe)L9}2] (L9 = primary amine), both isolated in the solid state.7 The crystal structure of the latter has also been determined and the distance between the oxygen atom of the monodentate acetate and the o-hydrogen of the aromatic ring of the amine is shorter than the sum of their van der Waals radii.This suggests that cyclopalladation reactions could occur through an intramolecular process involving interactions between the monodentate acetate ligand and the o-hydrogen atom of the N-donor ligand.Table 3 collects all the relevant previously published data from the same cyclometallation reactions carried out in toluene, that is in the absence of protons, in order clearly to establish the extreme difference found for the same systems depending on the reaction media. First of all a dramatic acceleration of the reaction rate in acetic acid medium is observed in all cases indicating that the transition state must have a structure with much lower energy than in the case of the spontaneous reac- Fig. 1 Plots of the variation with pressure of the cyclometallation rate constants for some of the systems studied Scheme 5 N Pd H O O H42 J. Chem. Soc., Dalton Trans., 1998, Pages 37–43 Table 3 Kinetic and activation parameters for the cyclometallation reaction of palladium acetate with the imines in Scheme 1 in toluene solution (from ref. 11) Imine A B C D E F G H I J K L M N Metallated compound 1a 1b 1c 1d 1e 1f 1g 1h 1i 1j 1k 1 2k 1l 2m 2n 104k323/s21 4.7 1.4 1.2 1.3 a 1.6 a 0.58 4.7 0.59 5.9 3.8 5.5 b 3.3 a 7.0 9.2 a DH‡/kJ mol21 63 ± 5 71 ± 6 73 ± 10 66 ± 10 67 ± 12 75 ± 18 48 ± 3 49 ± 11 52 ± 3 65 ± 1 46 ± 4 45 ± 2 67 ± 5 62 ± 2 DS‡/J K21 mol21 2115 ± 16 2102 ± 18 297 ± 30 2123 ± 33 2113 ± 36 296 ± 50 2167 ± 9 2177 ± 33 2150 ± 10 2110 ± 2 2168 ± 13 2180 ± 6 2100 ± 15 2120 ± 9 DV‡/cm3 mol21 (T/K) 212 ± 3 (318) 223 ± 3 (343) 224 ± 5 (323) 223 ± 4 (343) 221 ± 2 (343) 225 ± 2 (343) 224 ± 3 (333) 220 ± 1 (343) 215 ± 2 (323) 212 ± 1 (333) 211 ± 1 (323) 215 ± 4 (323) 212 ± 1 (323) 217 ± 1 (323) a A statistical factor of 2 has been applied.b A statistical factor of 1.1 has been applied according to the ratio [1k] : [2k] = 2.5 : 1 determined under these conditions. tion. According to Scheme 5 this fact has to be related to the extremely good leaving ligand characteristics of the MeCO2H2 1 group; consequently the bond regime in the transition state has to be much more similar to that in the final cyclometallated complex than is the case for the non-H1-assisted reaction.Such observations have also been made for a large number of electrophilic substitution activations of C]H bonds in RhII 2 core compounds.9c–e As for the thermal activation parameters, DH‡ and DS‡, they are both spread over a large range. In this case, though, no important differences can be found between the activation of C]H bonds corresponding to imine ligands with large differences in the N-centred cone angle.Even so, a certain grouping for the aniline derivatives exists with high values (DH‡ = 66–73 kJ mol21, DS‡ = 252 to 275 J K21 mol21), while for the endo benzylamine and propylamine derivatives the values are lower (DH‡ = 59–69 kJ mol21, DS‡ = 287 to 2123 J K21 mol21). Given the wide range of the values determined and the lack of a trend for the reactions in toluene solution, it seems clear that the thermal activation parameters follow a uniform trend not observed for the reaction carried out in toluene solution.Finally, with reference to the volumes of activation, DV‡, extracted from the slope of ln k versus P plots (Fig. 1), all fall in a rather narrow range around 215 ± 3 cm3 mol21. This is the most dramatic difference with respect to the available data for the reactions carried out in toluene solution. In acetic acid solution (i.e. acid-assisted process) no differences are detected between the sets of imines that could be separated according to the steric demands of the substituents on the central N, while for the spontaneous reactions (i.e.in toluene solution) the values of DV‡ fall in two ranges, 223 ± 2 and 214 ± 3 cm3 mol21 for the large (imines B to H) and small (imines A and I to N) nitrogen cone angles respectively (Table 3). Somehow, it seems clear that for the acid-assisted reaction the compression to form the transition state is practically independent of the activated imine ligand.The fact that the transition state for this path is more advanced along the reaction co-ordinate (see above) has to be somehow matched by a lesser degree of organization and contraction from the starting materials. That is, the transition state, being more advanced, already involves significant release of the MeCO2H2 1 group, with consequently a smaller degree of overall compression. Furthermore, given the fact that the transition state is a late one, the influence of the imine steric backbone has to be much less important once the right positioning has taken place for the C]H activation to occur.Conclusion The mechanism for the cyclopalladation of imines in acetic acid can be proposed to take place via a first fast step, the formation of an imine–palladium co-ordination compound containing terminal and bridging acetate ligands, followed by a second rate-determining step. The formation of a highly ordered transition state in which there is a four-centred interaction between the carbon and the hydrogen atoms of the C]H bond to be activated, the oxygen atom of the monodentate acetate ligand and the metal atom, explains the experimental data both from a synthetic and a kinetic point of view.The evolution of this transition state to the formation of the final Pd]C bond is favoured in protic media, the leaving acetic acid being able to afford the poorly co-ordinating MeCO2H2 1 group, thus favouring the formation of the final cyclopalladated species.Experimental Instruments and materials Proton NMR spectra were recorded on Varian XL-200 (200), VXR-500 (500) and Bruker DRX-250 (250 MHz) spectrometers, UV/VIS spectra on a HP8452A diode-array instrument and on a Beckmann UV5420 instrument equipped with a highpressure cell.19 All the acetato-bridged cyclometallated compounds have been characterized previously.8b,e,11,13c,14 Kinetic measurements The reactions at atmospheric pressure were followed by UV/ VIS spectroscopy in the full 750–300 nm range on a HP8452A instrument equipped with a multicell transport, thermostatted (±0.1 8C) with a circulation bath.Observed rate constants were derived from the absorbance versus time traces at wavelengths where a maximum increase and/or decrease of absorbance was observed. No dependence of the values on the selected wavelengths was detected, as expected for reactions where a good retention of isosbestic points is observed.The general kinetic technique was that previously described.19 Solutions for the kinetic runs were prepared by dissolving calculated amounts of the compounds (palladium acetate and imine) in acetic acid. In all cases no dependence on the concentration of palladium acetate or imine was detected, and a [Pd] : [imine] ratio within the range 0.7–1.3 : 1 was maintained to insure the nonappearance of the insoluble well known [Pd(O2CMe)2(imine)2] species.For runs at elevated pressure a previously described pressurizing system and high-pressure cell were used.19 In these cases the absorbance versus time traces were recorded on a Beckmann UV5420 instrument at a fixed wavelength chosen from the atmospheric pressure experiments. Rate constants were derived from exponential least-squares fitting by standard routines.J. Chem. Soc., Dalton Trans., 1998, Pages 37–43 43 Least-squares errors for the rate constants were always in the range 10–15% of the calculated value. All post-run fitting by rate laws were done by standard fitting programs commercially available.Acknowledgements We acknowledge financial support from the Dirección General de Investigación Científica y Técnica (Projects PB96-0164, PB94-0857) and Comissió Interdepartamental de Recerca i Innovació Tecnològica (Project QFN95-4708). References 1 A. D. Ryabov, Synthesis, 1985, 233; M. Pfeffer, Recl. Trav. Chim.Pays Bas, 1990, 109, 567. 2 N. J. Thompson, J. L. Serrano, M. J. Baena and P. Espinet, Chem. Eur. J., 1996, 2, 214. 3 C. Navarro-Ranninger, I. López-Solera, J. M. Pérez, J. R. Masaguer and C. Alonso, Appl. Organomet. Chem., 1993, 7, 57; C. Navarro- Ranninger, I. López-Solera, V. M. González, J. M. Pérez, A. Alvarez-Valdés, A. Martín, P. R. Raithby, J. R. Masaguer and C. Alonso, Inorg. Chem., 1996, 35, 5181. 4 S. Y. M. Chooi, P. H. Leung, C. C. Lim, K. F. Mok, G. H. Quek, K.Y. Sim and M. K. Tan, Tetrahedron: Asymmetry, 1992, 3, 529; E. P. Kyba and S. P. Rines, J. Org. Chem., 1982, 47, 4800; J. Albert, J. Granell, G. Muller, D. Sainz, M. Font-Bardia and X. Solans, Tetrahedron: Asymmetry, 1995, 6, 325; N. W. Alcock, J. M. Brown and D. I. Hulmes, Tetrahedron: Asymmetry, 1993, 4, 743; N. Gabbitas, G. Salem, M. Sterns and A. C. Willis, J. Chem. Soc., Dalton Trans., 1993, 3271; S. Y. M. Chooi, S. Y. Siah, P. H. Leung and K. F. Mok, Inorg. Chem., 1993, 32, 4812; S.Gladiali, A. Dore, D. Fabbri, O. De Lucchi and M. Manassero, Tetrahedron: Asymmetry, 1994, 5, 511; M. Pabel, A. C. Willis and S. B. Wild, Angew. Chem., Int. Ed. Engl., 1994, 33, 1835; C. E. Barclay, G. Deeble, R. J. Doyle, S. A. Elix, G. Salem, T. L. Jones, S. B. Wild and C. Willis, J. Chem. Soc., Dalton Trans., 1995, 57. 5 A. D. Ryabov, I. K. Sakondinskaya and A. K. Yatimirski, J. Chem. Soc., Dalton Trans., 1985, 2629; A. D. Ryabov and A. K. Yatimirski, Inorg.Chem., 1984, 23, 789; A. D. Ryabov, Inorg. Chem., 1987, 26, 1252. 6 I. P. Romm, S. V. Kravtsova, T. I. Perepelkova, E. S. Petrov, I. O. Kalinovski and T. M. Buslaeva, Russ. J. Coord. Chem., 1995, 21, 740. 7 J. Vicente, I. Saura-Llamas and M. G. Palin, Organometallics, 1997, 16, 826. 8 (a) J. Albert, R. M. Ceder, M. Gómez, J. Granell, J. Sales and X. Solans, Organometallics, 1990, 9, 1405; (b) J. Albert, R. M. Ceder, M. Gómez, J. Granell and J. Sales, Organometallics, 1992, 11, 1536; (c) J.Albert, J. Granell, R. Moragas, J. Sales, M. Font-Bardia and X. Solans, J. Organomet. Chem., 1995, 494, 95; (d ) J. Albert, J. Granell, J. Sales, M. Font-Bardia and X. Solans, Organometallics, 1995, 14, 1393; (e) J. Granell, D. Sainz, J. Sales, X. Solans and M. Font-Altaba, J. Chem. Soc., Dalton Trans., 1986, 1785. 9 (a) M. Crespo, M. Martinez and J. Sales, Organometallics, 1992, 11, 1288; (b) M. Crespo, M. Martinez and J. Sales, Organometallics, 1993, 12, 4297; (c) G.González, P. Lahuerta, M. Martinez, M. Sanau and E. Peris, J. Chem. Soc., Dalton Trans., 1994, 545; (d) F. Estevan, P. Lahuerta, E. Peris, M. A. Ubeda, S. García- Granda, F. Gómez-Beltrán, E. Pérez Carreño, G. González and M. Martinez, Inorg. Chim. Acta, 1994, 218, 189; (e) F. Estevan, G. González, P. Lahuerta, M. Martinez, E. Peris and R. van Eldik, J. Chem. Soc., Dalton Trans., 1996, 1045. 10 M. Crespo, M. Martinez and J. Sales, J. Chem. Soc., Chem. Commun., 1992, 822; M. Schmülling, A. D. Ryabov and R. van Eldik, J. Chem. Soc., Chem. Commun., 1992, 1609; G. González, B. Moullet, M. Martinez and A. E. Merbach, Inorg. Chem., 1994, 33, 2330; G. González, M. Martinez and E. Rodriguez, J. Chem. Soc., Dalton Trans., 1995, 891. 11 M. Gómez, J. Granell and M. Martinez, Organometallics, 1997, 16, 2539. 12 (a) M. I. Bruce, Angew. Chem., Int. Ed. Engl., 1977, 16, 73; (b) G. R. Newkome, W. E. Puckett, W. K. Gupta and G. A. Kiefer, Chem. Rev., 1986, 86, 451; (c) I. Omae, Coord. Chem. Rev., 1988, 83, 137; (d) V. V. Dunina, O. A. Zalevskaya and V. M. Potatov, Russ. Chem. Rev., 1988, 57, 250; (e) A. D. Ryabov, Chem. Rev., 1990, 90, 403. 13 (a) P. L. Alsters, P. F. Engel, M. P. Hogerheide, M. Copjin, A. L. Spek and G. van Koten, Organometallics, 1993, 12, 1831; (b) G. De Munno, M. Ghedini and F. Neve, Inorg. Chim. Acta, 1995, 239, 155; (c) J. Albert, J. Granell, J. Sales, X. Solans and M. Font, Organometallics, 1986, 5, 2567. 14 J. Albert, J. Granell and J. Sales, J. Organomet. Chem., 1984, 273, 393. 15 J. Vicente, I. Saura-Llamas and P. G. Jones, J. Chem. Soc., Dalton Trans., 1993, 3619. 16 A. Albinati, P. S. Pregosin and R. Ruedi, Helv. Chim. Acta, 1985, 68, 2046; J. Selbin, K. Abboud, S. F. Watkins, M. A. Gutiérrez and F. R. Fronczek, J. Organomet. Chem., 1983, 241, 259; G. Balavoine, J. C. Clinet, P. Zerbib and K. Boubeukur, J. Organomet. Chem., 1990, 389, 259; J. L. García-Ruano, I. López-Solera, J. R. Massaguer, C. Navarro-Ranninger, J. H. Rodríguez and S. Martínez-Carrera, Organometallics, 1992, 11, 3013. 17 A. C. Skapski and M. L. Smart, Chem. Commun., 1970, 658; F. A. Cotton and S. Han, Rev. Chim. Miner., 1983, 20, 496; A. Mawby and G. E. Pringle, J. Inorg. Nucl. Chem., 1971, 33, 1989; F. A. Cotton and S. Han, Rev. Chim. Miner., 1985, 22, 277. 18 H. A. Zhong and R. A. Winderhoefer, Inorg. Chem., 1977, 36, 2610. 19 M. Crespo, M. Martinez and E. de Pablo, J. Chem. Soc., Dalton Trans., 1997, 1321. Received 28th July 1997; Paper 7/05435A

ISSN:1477-9226    

DOI:10.1039/a705435a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1996, Pages 39–45 39 Synthesis, stereochemistry and crystal structures of cobalt(III) complexes containing 5,8-diphenyl-2,11-dithia-5,8-diphosphadodecane or 5,9-diphenyl-2,12-dithia-5,9-diphosphatridecane Kazuo Kashiwabara,*,a Yasuhiro Ito,a Masakazu Kita,b Junnosuke Fujita c and Kiyohiko Nakajima d a Department of Chemistry, Faculty of Science, Nagoya University, Nagoya 464-01, Japan b Naruto University of Education, Takashima, Naruto 772, Japan c Division of Natural Sciences, International Christian University, Mitaka 181, Japan d Department of Chemistry, Aichi University of Education, Kariya 448, Japan Twelve new cobalt(III) complexes containing a tetradentate phosphine ligand of the type MeS(CH2)2PPh(CH2)nPPh(CH2)2SMe (n = 2 L1 or 3 L2) were prepared.Their structures were assigned on the basis of electronic absorption and 1H NMR spectra and the molecular structures of cis-a-[CoCl2{rac(P)-L1}]BF4 and trans-[CoCl2{meso(P)-L2}]BF4 were determined by single-crystal X-ray diffraction.The last two complexes comprise a slightly distorted octahedron with bond distances Co]P 2.210(1), Co]S 2.254(1) and Co]Cl 2.281(1) and Co]P 2.227(3), Co]S 2.327(3) and Co]Cl 2.241(3) Å (all averages) respectively. The differences in Co]S and Co]Cl bond distances are attributable to the strong trans influence of the phosphino group. Acetylacetonate complexes of rac(P)-SPPS in organic solvents form equilibrium mixtures of the cis-a and cis-b isomers, the molar ratio in MeNO2 solution at 60 8C being 2 : 1 for the L1 complex and 3 : 2 for the L2 complex.Linear tetradentate tetraphosphine ligands exist as racemic (RR/SS) and meso (RS) diastereomers by virtue of the absolute configurations of the two internal phosphorus atoms in the skeletons.1 Each of the diastereomers generates stereospecific co-ordination modes upon complexation.2 We have prepared a large number of linear tetradentate phosphine(soft base)– amine(hard base) hybrid compounds where the phosphorus donor atoms are chiral, examined the complexation to cobalt(III) of a typical Lewis hard acid, and found that these complexes have quite different properties from those of analogous tetradentate tetramines.3–6 In a previous paper 7 we extended the study to a tetradentate phosphorus–sulfur hybrid bearing only soft donor groups, 5,8-diphenyl-2,11-dithia-5,8- diphosphadodecane (L1), and obtained [Co(acac)2L1]+ (acac = pentane-2,4-dionate) in which L1 acts only as a didentate ligand through two phosphorus atoms.The purpose of this paper is to examine more thoroughly the complexation ability of L1 and the related L2 (5,9-diphenyl-2,12-dithia-5,9- diphosphatridecane) to cobalt(III) and clarify the correlations between the co-ordination modes and the absolute configurations of the chiral phosphorus atoms and/or the conformations (or size) of the central P]Co]P chelate rings.Bosnich et al.8,9 reported the preparation of a series of cobalt(III) complexes with the linear tetradentate tetraarsine [Me2As(CH2)3As(Ph)CH2]]2, and clarified the optical stability of the chiral arsine atoms and topological stabilities of various geometrical isomers of the complexes. However, this arsine differs in the kind of donor atoms and the skeletal structure from our present phosphines: the number of methylene groups of the arsine is 3,2,3, while our phosphines have 2,2,2 and 2,3,2.Such differences often result in different topological preferences and stabilities upon complexation. To our knowledge, only two papers have appeared on metal complexes of SPPS-type tetradentate ligands. Issleib and Gans10 prepared [MX(NH3)(SCH2CH2PHCH2CH2PHCH2- CH2S)] (M = Co, X = Br; M = Rh, X = Cl) and [M{SCH(R)- CH2PHCH2CH2PHCH2CH(R)S}] (M = Ni, Pd or Pt; R = H or Me). However, the stereochemistry of the complexes was not well defined. Schmelzer and Schwarzenbach11 reported the crystal structure of racemic(P) and meso(P) isomers of [Ni- (SCH2CH2PPhCH2CH2CH2PPhCH2CH2S)], but experimental details and chemical properties were not described.Experimental The phosphines were handled under an atmosphere of nitrogen using Schlenk techniques until the cobalt(III) complexes were formed. All of the solvents used for the preparation of ligands and complexes were made oxygen-free by bubbling nitrogen for 20 min immediately before use. Absorption spectra were recorded on a Hitachi U3400 spectrophotometer and NMR spectra on Hitachi R-90H and Varian INOVA 500 spectrometers.Preparation of phosphines L1. This compound was prepared from 1,2-bis(phenylphosphino) ethane and 2-chloroethyl methyl sulfide according to a previous method.7 L2. A 15% hexane solution of butyllithium (30 cm3, 48.6 mmol) was added dropwise with stirring to a tetrahydrofuran solution (400 cm3) of 1,3-bis(phenylphosphino)propane (Strem Chem.Inc.) (5 g, 19.2 mmol). After stirring for 30 min, 2- chloroethyl methyl sulfide (4.25 g, 38.4 mmol) was added portionwise to the resulting yellow solution. The solution was stirred for 1 h at 50 8C, and then overnight at room temperature. Water (50 cm3) was added dropwise with stirring. After a while the ethereal layer was separated, dried over MgSO4 (5 g) overnight, filtered, and the filtrate evaporated to give a viscous syrup. The product could not be distilled because of the very high boiling point, and was used as such for the preparation of cobalt(III) complexes. Yield: 6.92 g (88%).It was found to be a mixture of racemic and meso isomers (ca. 1 : 1 according to the NMR spectra. (CDCl3): 31P-{1H} (external 85% H3PO4), d 225.64 and 225.69; 13C-{1H}, d 15.40 (s, SCH3), 15.41 (s, SCH3), 22.34 (t, J = 14.96, CH2), 22.44 (t, J = 14.96, CH2), 28.20 (d, J = 13.24, PCH2), 28.31 (d, J = 14.39, PCH2), 29.34 (t, J = 12.09, PCH2), 29.41 (t, J = 12.09, PCH2), 30.58 (d,40 J.Chem. Soc., Dalton Trans., 1997, Pages 39–45 J = 17.84, SCH2), 30.62 (d, J = 17.84, SCH2), 128.42 (d, J = 6.91, m-C), 128.44 (d, J = 7.48, m-C), 129.01 (s, p-C), 132.33 (d, J = 18.99, o-C), 132.34 (d, J = 18.99, o-C), 137.19 (d, J = 11.51, ipso-C) and 137.31 (d, J = 11.51 Hz, ipso-C); 1H, d 1.45 (m, CH2), 1.80 (m, PCH2), 1.92 (m, PCH2), 2.01 (s, SCH3), 2.02 (s, SCH3), 2.42 (m, SCH2), 7.31 (m, C6H5) and 7.43 (m, C6H5). Preparation of complexes trans-[CoCl2{meso(P)-L1}2]BF4 1 and cis-·-[CoCl2{rac(P)- L1}]BF4 2. A methanol solution (120 cm3) containing trans- [CoCl2(py)4]Cl?6H2O12 (py = pyridine) (2.45 g, 4.13 mmol) and L1 (1.63 g, 4.13 mmol) was stirred overnight at room temperature, and then concentrated to a small volume.The concentrate was chromatographed with a column (3 × 35 cm) of Toyopearl HW-40 and methanol as eluent. Two large green and red bands were obtained separately, the former being eluted faster. Each eluate was evaporated to dryness under reduced pressure, and the residue mixed with a small amount of methanol to extract the complex. On addition of an excess of LiBF4 the methanol extract gave a green or a red precipitate, which was filtered off and recrystallized from acetonitrile and diethyl ether to afford green (1) and red (2) crystals, respectively.Yield: 0.48 (12) for 1 and 0.79 g (31%) for 2 (Found: C, 47.5; H, 5.5. Calc. for C40H56BCl2CoF4P4S4 1: C, 47.75; H, 5.6. Found: C, 39.5; H, 4.8.Calc. for C20H28BCl2CoF4P2S2 2: C, 39.3; H, 4.6%). Both complexes 1 and 2 are soluble in nitromethane, acetonitrile, acetone, chloroform, or dichloromethane, but insoluble in water or diethyl ether; 1 is slightly soluble in methanol or ethanol. cis-‚-[Co(acac){rac(P)-L1}][SbF6]2 3 and cis-·-[Co(acac){rac- (P)-L1}][SbF6]2 4. A methanol solution (50 cm3) of Li(acac) (0.087 g, 0.82 mmol) was added to an acetonitrile solution (60 cm3) of cis-a-[CoCl2{rac(P)-L1}]BF4 (0.50 g, 0.82 mmol). The solution was stirred overnight at room temperature, then diluted ten times with water, and applied to a column (3 × 130 cm) of SP-Sephadex C-25.By elution with an aqueous 0.15 mol dm23 NaCl solution a small dark red band of [Co(acac)2{rac(P)-L1}]+ 7 and then two large red-orange and red bands were eluted separately. Each eluate of the two large bands was evaporated to dryness under reduced pressure at 20 8C. The residue was mixed with a small amount of ethanol to extract the complex, and the extract evaporated again to dryness under reduced pressure at 20 8C.The residue was dissolved in a small amount of water. On addition of an excess of NaSbF6 the solution yielded a red-orange (3) or red (4) precipitate, which was filtered off and recrystallized from methanol and diethyl ether to afford the crystals, respectively. Yields: 0.084 (10) for 3 and 0.352 g (42%) for 4 (Found: C, 29.55; H, 3.2 for 3. C, 29.2; H, 3.45 for 4.Calc. for C25H35CoF12O2- P2S2Sb2: C, 29.3; H, 3.45%). Both complexes 3 and 4 are soluble in nitromethane, acetonitrile, acetone, methanol or ethanol, slightly soluble in chloroform, dichloromethane or water, but insoluble in diethyl ether. [Co(acac){meso(P)-L1}2][SbF6]2?3H2O 5. A methanol solution (30 cm3) of Li(acac) (0.023 g, 0.22 mmol) was added to an acetonitrile solution (50 cm3) of trans-[CoCl2{meso(P)- L1}2]BF4 (0.217 g, 0.22 mmol). The solution was stirred for 5 h at room temperature, diluted ten times with water, and filtered. The filtrate was applied to a column (3 × 60 cm) of SPSephadex C-25, and the adsorbed products were eluted with an aqueous 0.15 mol dm23 NaCl solution to give two large dark red and red-orange bands.The eluate of the second red-orange band was evaporated to dryness under reduced pressure, and the complex extracted with ethanol from the residue. The extract was evaporated again to dryness under reduced pressure, and the residue was dissolved in a small amount of water.On addition of an excess of NaSbF6 a red-orange precipitate was obtained, filtered off and recrystallized from nitromethane and diethyl ether to afford the crystals. Yield: 0.042 g (14%) (Found: C, 36.5; H, 4.2. Calc. for C45H69CoF12O5P4S4Sb2: C, 36.7; H, 4.7%). The complex is soluble in nitromethane, acetonitrile, acetone, chloroform, dichloromethane, methanol or ethanol, slightly soluble in water, but insoluble in diethyl ether.From the eluate of the first dark red band, [Co(acac)2{meso- (P)-L1}]SbF6 7 was obtained in 15% yield by the same method as that for complex 5. trans-[CoCl2{meso(P)-L2}]BF4 6 and cis-·-[CoCl2{rac(P)- L2}]BF4 7. Complexes 6 (green) and 7 (red) were obtained by methods similar to those for the corresponding L1 complexes 1 and 2, respectively, using L2.Yields: 36% for 6 and 40% for 7 (Found: C, 40.3; H, 4.9 for 6. C, 40.4; H, 4.9 for 7. Calc. for C21H30BCl2CoF4P2S2: C, 40.35; H, 4.85%).Both complexes are soluble in nitromethane, acetonitrile, acetone, chloroform or dichloromethane, less soluble in methanol or ethanol, but insoluble in diethyl ether. cis-‚-[Co(acac){rac(P)-L2}][SbF6]2 8 and cis-·-[Co(acac)- {rac(P)-L2}][SbF6]2 9. Complexes 8 (red-orange) and 9 (red) were prepared from cis-a-[CoCl2{rac(P)-L2}]BF4 and Li(acac) by methods similar to those for the corresponding L1 complexes 3 and 4, respectively. In contrast to the case of L1, [Co(acac)2{rac(P)-L2}]+ was not formed.Yields: 24% for 8 and 48% for 9 (Found: C, 29.8; H, 3.6 for 8. C, 29,3; H, 3.45 for 9. Calc. for C26H37CoF12O2P2S2Sb2: C, 30.1; H, 3.6%). The solubilities of 8 and 9 are similar to those of 3 and 4. cis-‚-[Co(acac){meso(P)-L2}][SbF6]2 10. To an acetonitrile solution (20 cm3) of trans-[CoCl2{meso(P)-L2}]BF4 (0.20 g, 0.32 mmol) were added a methanol solution (40 cm3) of Li(acac) (0.034 g, 0.32 mmol) and active charcoal (0.05 g). The mixture was stirred for 24 h at room temperature, and then filtered to remove charcoal. The filtrate was diluted ten times with water, and applied to a column (3 × 60 cm) of SP-Sephadex C-25.By elution with an aqueous 0.15 mol dm23 NaCl solution, a redorange band was developed. The eluate of the band was evaporated to dryness under reduced pressure, and the residue mixed with a small amount of ethanol to extract the complex. The extract was evaporated again to dryness under reduced pressure, and the residue dissolved in a small amount of water. On addition of an excess of NaSbF6 the solution gave a red precipitate, which was filtered off and recrystallized from methanol and diisopropyl ether to afford the crystals.Yield: 0.05 g (15%) (Found: C, 30.1; H, 3.55. Calc. for C26H37Co- F12O2P2S2Sb2: C, 30.1; H, 3.6%). The solubility of the complex is similar to those of 8 and 9. [Co(acac)2{meso(P)-L2}]SbF6 11 and ƒ(RR)/L(SS)-[Co- (acac)2{rac(P)-L2}]SbF6 12.A mixture of [Co(acac)3]13 (0.44 g, 1.22 mmol), L2 (0.50 g, 1.22 mmol), and active charcoal (0.05 g) in methanol (50 cm3) was stirred for 15 h at room temperature, and then filtered to remove charcoal. The filtrate was diluted ten times with water, and applied to a column (3 × 20 cm) of SP-Sephadex C-25. By elution with an aqueous 0.05 mol dm23 NaCl solution, two large dark red bands of complexes 11 and 12 appeared, the former being eluted faster. Each eluate of these bands was evaporated to dryness under reduced pressure, and the complex extracted from the residue with dichloromethane.The extract was evaporated again to dryness under reduced pressure, and the residue dissolved in a small amount of water. On addition of an excess of NaSbF6 the solution yielded orange-brown crystals, which were filtered off and recrystallized from hot methanol. Yields: 0.22 (20) for 11 and 0.18 g (16%) for 12 (Found: C, 41.3; H, 4.9 for 11. C, 41.2; H, 4.9 for 12.Calc. for C31H44CoF6O4P2S2Sb: C, 41.3; H, 4.9%). The complexes are soluble in methanol, ethanol, chloroform or acetone, and slightly soluble in water or diethyl ether.J. Chem. Soc., Dalton Trans., 1997, Pages 39–45 41 Complex 11 was also prepared by the following method. To an acetonitrile solution (20 cm3) of trans-[CoCl2{meso(P)- L2}]BF4 (0.15 g, 0.24 mmol) were added a methanol solution (30 cm3) of Li(acac) (0.077 g, 0.72 mmol) and active charcoal (0.03 g). The mixture was stirred for 24 h at room temperature, and then filtered to remove charcoal.The filtrate was diluted ten times with water, applied to a column (3 × 60 cm) of SPSephadex C-25, and the adsorbed products were eluted with an aqueous 0.05 mol dm23 NaCl solution. A large orange band appeared, and the eluate was treated as described above to give orange crystals of 11. Yield: 0.11 g (49%). [Co(acac)2{meso(P)-L1}]SbF6 13 and ƒ(RR)/L(SS)- [Co(acac)2{rac(P)-L1}]SbF6 14. These complexes were prepared by a previous method.7 Crystallography Single crystals of complexes 2 (0.25 × 0.25 × 0.40 mm) and 6 (0.20 × 0.30 × 0.40 mm) were fixed on the end of a glass fibre with epoxy resin.They were mounted on a Rigaku AFC-5 diffractometer individually, and the diffraction data collected at 298 K with graphite-monochromated Mo-Ka radiation (l = 0.710 73 Å) using the w–2q scan mode for 2 and the w scan mode for 6. Cell dimensions were determined by least-squares refinement of the angular positions of 25 independent reflections in the range 25 < 2q < 308 for each sample.Crystallographic data and experimental details are listed in Table 1. The position of the cobalt was determined by direct methods (SHELXS 86 14) for each complex and the remaining nonhydrogen atoms were located by subsequent Fourier syntheses. The structure was refined on F by full-matrix least-squares techniques with anisotropic thermal parameters for nonhydrogen atoms.The disordered tetrafluoroborate anions of 6 were refined with isotropic thermal parameters. All the hydrogen atoms were placed at calculated positions with isotropic displacement parameters of their parent carbon atoms. The calculations were carried out with the XTAL 3.2 15 software, and the refinement of positional and thermal parameters finally converged to R = 0.047 (R9 = 0.055) for 2 and R = 0.058 (R9 = 0.059) for 6. 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., 1996, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/289. Results and Discussion The synthetic routes for twelve new complexes are shown in Scheme 1. Dichloro complexes For L1 and L2 green and red dichlorocobalt(III) complexes were obtained by the reaction of [CoCl2(py)4]Cl?6H2O in a molar ratio of 1 : 1 in methanol (green 1 and 6, and red 2 and 7).Elemental analyses of these complexes show that 1 involves two L1 ligands, while 2, 6 and 7 have a tetradentate L1 or L2 ligand. The structures of 2 and 6 were determined by X-ray diffraction. Perspective views of the complex cations are shown in Figs. 1 and 2, respectively, and selected bond distances and angles in Tables 2 and 3. The complex cations form an octahedron with two chloride ions and rac(P)-L1 in a cis-a configuration for 2, and with two chloride ions and meso(P)-L2 in a trans one for 6.Compound 2 crystallizes with molecular C2 symmetry. The cation lies with Co and the centre of the C(10)]C(109) bond on the two-fold axis along a while the boron atom of the anion lies on the two-fold axis along c. The Co]P bond distance of 2.210(1) Å in 2 is relatively short compared with those found in related cobalt(III) phosphine complexes (2.194–2.353 Å).16 The Co]Cl bond distance 2.281(1) Å is appreciably longer than those not only in 6 [2.243(3) and 2.239(3) Å] but also the related dichlorocobalt( III) complexes trans(Cl,Cl)cis(P,P)-[CoCl2(NH2CH2CH2- PPh2)2]?0.5[CoCl4] [average 2.240(3) Å],17 trans(Cl,Cl)cis(P,P)- [CoCl2(NH2CH2CH2PBuPh)2]ClO4 [average 2.238(3) Å] 17 and trans(Cl,Cl)cis(P,P)-[CoCl2(NH2CH2CH2PMe2)2]PF6?0.5MeOH [average 2.242(3) Å].18 The elongation of the Co]Cl bond in 2 may be attributed to the stronger trans influence of the phosphine donor group relative to that of the chlorine atom in 6.On the other hand, the Co]S bond distances of 6 [2.325(3) and 2.329(3) Å] are longer than that [2.252(1) Å] of 2. The elongation in 6 is also interpreted by the stronger trans in- fluence of the phosphine group than that of the thioether group in 2. Several examples of Co]S bond distances in cobalt(III)–thioether complexes are [Co{(R)-NH2CH(CO2)- CH2SMe}2]ClO4?H2O [average 2.272(2) Å],19 [Co(NH2CH2- CH2SMe)(NH2CH2CH2NH2)2][Fe(CN)6]?4H2O [2.268(10) Å]20 and [Co{(R)-[MeSCH2CH(CO2)NHCH2]]2}]ClO4 [average 2.261(4) Å].21 The chelate angles of L1 [86.79(4), 87.18(4)8] in complex 2 are typical for a five-membered chelate ring.17,18,21 No large deviation from an octahedral angle was observed at the Co atom.The complex ion in Fig. 1 has a L configuration. In this con- figuration both the P and S donor atoms take an S configuration, and the P]P chelate ring a l gauche conformation, while the two P]S chelate rings take an envelope one.The 1H and 13C NMR spectra of this complex in CD3NO2 at 30 8C show a singlet signal for the SMe group at d 2.35 and 20.1, respectively (Table 4),which is retained even at 280 8C in (CD3)2CO. The singlet indicates a rapid configurational inversion or a preferential configuration of the (S)-sulfur atom in the L isomer as shown in Fig. 1. Although we have no evidence for either, molecular models indicate that the methyl group on the sulfur Scheme 1 (i) Li(acac); (ii) n Li(acac), activated charcoal; (iii) L2, activated charcoal trans -[CoCl2(py)4]Cl•6H2O + L1 or L2 cis -a-[CoCl2{rac (P)-L}]+ (i ) cis -a-[Co(acac){rac (P)-L}]2+ 2, 7 cis -b-[Co(acac){rac (P)-L}]2+ 4, 9 trans -[CoCl2{meso (P)-L1}2]+ 1 trans -[CoCl2{meso (P)-L2}]+ 6 [Co(acac){meso (P)-L1}2]2+ (i ) 3, 8 5 (ii ) n = 1 n = 3 cis -b-[Co(acac){meso (P)-L2}]2+ [Co(acac)2{meso (P)-L2}]+ 10 11 (iii ) [Co(acac)2{rac (P)-L2}]+ 12 [Co(acac)3] D6/04759I/A642 J.Chem. Soc., Dalton Trans., 1997, Pages 39–45 Table 1 Crystallographic data for cis-a-[CoCl2{rac(P)-L1}]BF4 2 and trans-[CoCl2{meso(P)-L2}]BF4 6 * 2 6 Formula M Space group a/Å b/Å c/Å U/Å3 m(Mo-Ka)/cm21 Crystal colour Dc/g cm23 Dm/g cm23 Scan range/8 Reflections measured No. reflections measured No. reflections observed [|Fo| > 3s(|Fo|)] RR 9 S Largest difference peak, hole/e Å23 C20H28BCl2CoF4P2S2 611.16 Pnna (no. 52) 12.392(1) 18.808(4) 10.900(1) 2540.4(7) 12.06 Red 1.60 1.59 1.15 + 0.50 tan q 0 < h < 17, 0 < k < 26, 0 < l < 15 3812 2297 0.047 0.055 1.72 0.93, 20.82 C21H30BCl2CoF4P2S2 625.19 Pbc21 (no. 29) 9.914(2) 29.945(4) 8.946(1) 2655.8(9) 11.55 Green 1.56 — 0.735 + 0.50 tan q 0 < h < 13, 0 < k < 42, 0 < l < 12 4203 2161 0.058 0.059 1.61 0.85, 20.78 * Details in common: orthorhombic; Z = 4; prismatic; scan speed 88 min21; 2qmax 608; R = o ||Fo| 2 |Fc||/o |Fo|, R9 = (o w||Fo| 2 |Fc||2/o w|Fo|2)� �� , w = [s2(Fo) + (0.015Fo)2]21.atom in the R configuration points to the phenyl group to form a crowded structure, suggesting the preferential co-ordination of the S (or R)-sulfur atom in the L (or D) isomer. For complex 6 the five- and six-membered chelate rings of the meso(P)-L2 ligand take a gauche and a chair conformation, respectively. The chelate angles of L2 are almost 908 and no Fig. 1 Perspective view of cis-a-[CoCl2{rac(P)-L1}]+, which crystallizes with molecular C2 symmetry Fig. 2 Perspective view of trans-[CoCl2{meso(P)-L2}]+ large deviation from an octahedron was observed at the Co atom. Complex 7 exhibits a singlet SMe signal in the 1H and 13C NMR spectra at d 2.28 and 18.6, respectively, indicating a trans or a cis-a configuration. The complex shows an absorption spectral pattern very similar to that of 2 [Fig. 3(a)], and is assigned to cis-a-[CoCl2{rac(P)-L2}]+. The meso(P)-SPPS ligand cannot form a cis-a isomer since the two terminal P]S chelate arms point to the same apical site with respect to the P]Co]P plane.Table 2 Selected bond distances (Å) and angles (8) for complex 2 Co]Cl Co]S 2.281(1) 2.254(1) Co]P 2.210(1) Cl]Co]S Cl]Co]P Cl]Co]Cl9 Cl]Co]S9 Cl]Co]P9 S]Co]S9 S]Co]P S]Co]P9 P]Co]P9 91.88(4) 89.72(3) 93.93(4) 89.82(4) 175.35(4) 177.51(5) 87.18(4) 91.01(4) 86.79(4) Co]S]C(1) Co]S]C(2) C(1)]S]C(2) Co]P]C(3) Co]P]C(4) Co]P]C(10) C(3)]P]C(4) C(3)]P]C(10) C(4)]P]C(10) 109.6(2) 102.3(1) 100.7(2) 107.0(1) 119.4(1) 108.3(1) 107.7(2) 105.5(2) 108.1(2) Table 3 Selected bond distances (Å) and angles (8) for complex 6 Co]Cl(1) Co]Cl(2) Co]S(1) 2.243(3) 2.239(3) 2.325(3) Co]S(2) Co]P(1) Co]P(2) 2.329(3) 2.223(3) 2.230(3) Cl(1)]Co]Cl(2) Cl(1)]Co]S(1) Cl(1)]Co]S(2) Cl(1)]Co]P(1) Cl(1)]Co]P(2) Cl(2)]Co]S(1) Cl(2)]Co]S(2) Cl(2)]Co]P(1) Cl(2)]Co]P(2) S(1)]Co]S(2) S(1)]Co]P(1) S(1)]Co]P(2) S(2)]Co]P(1) S(2)]Co]P(2) P(1)]Co]P(2) Co]S(1)]C(1) Co]S(1)]C(2) 174.5(1) 84.6(1) 92.1(1) 86.3(1) 87.9(1) 93.7(1) 82.7(1) 98.8(1) 93.9(1) 90.7(1) 88.3(1) 172.5(1) 178.2(1) 90.0(1) 90.8(1) 111.1(4) 104.9(4) C(1)]S(1)]C(2) Co]S(2)]C(11) Co]S(2)]C(12) C(11)]S(2)]C(12) Co]P(1)]C(3) Co]P(1)]C(4) Co]P(1)]C(10) C(3)]P(1)]C(4) C(3)]P(1)]C(10) C(4)]P(1)]C(10) Co]P(2)]C(13) Co]P(2)]C(14) Co]P(2)]C(20) C(13)]P(2)]C(14) C(13)]P(2)]C(20) C(14)]P(2)]C(20) 101.7(6) 115.0(5) 100.6(5) 99.9(6) 105.2(4) 120.8(4) 114.3(4) 104.5(5) 104.2(5) 106.3(5) 104.2(4) 121.7(4) 114.1(4) 105.9(5) 108.4(6) 101.8(5)J.Chem. Soc., Dalton Trans., 1997, Pages 39–45 43 Table 4 Absorption and 1H NMR spectral data 1H NMR (d) b Complex Absorption a CH3 ]] CH] 16273 45 8 9 10 11 12 13 14 trans-[CoCl2{meso(P)-L1}2]+ trans-[CoCl2{meso(P)-L2}]+ cis-a-[CoCl2{rac(P)-L1}]+ cis-a-[CoCl2{rac(P)-L2}]+ cis-b-[Co(acac){rac(P)-L1}]2+ cis-a-[Co(acac){rac(P)-L1}]2+ [Co(acac){meso(P)-L1}2]2+ cis-b-[Co(acac){rac(P)-L2}]2+ cis-a-[Co(acac){rac(P)-L2}]2+ cis-b-[Co(acac){meso(P)-L2}]2+ [Co(acac)2{meso(P)-L2}]+ [Co(acac)2{rac(P)-L2}]+ [Co(acac)2{meso(P)-L1}]+ [Co(acac)2{rac(P)-L1}]+ 16.4 (2.15) 16.1 (2.25) 20.0 (3.19) 19.4 (2.91) 20.5 (3.12) 19.5 (3.04) 20.2 (3.28) 19 (sh) (3.0) 20.7 (3.14) 18.9 (2.86) 19 (sh) (2.8) 21.3 (2.98) 19 (sh) (2.7) 21.6 (2.87) 19 (sh) (2.6) 21.6 (2.82) 21.3 (2.98) 21.2 (3.07) 1.92 2.37 2.35 2.28 1.76, 1.89 2.00, 2.07 1.95, 2.29 0.96, 1.72, 1.80 1.89, 2.06, 2.21 1.79, 2.11 2.56, 2.60 1.73, 2.20 1.07, 1.75 2.08, 2.41 1.52, 1.64, 1.67 1.90, 2.07, 2.11 1.64, 1.78 1.97 0.96, 1.58, 1.94 2.03, 2.06, 2.16 1.34, 1.98, 2.04 4.97 5.84 4.68 5.24 5.694.72 5.34 5.06 4.74 5.51 4.86 a First absorption bands: n/103 cm21 (log e), solvent MeCN, sh = shoulder.b Solvents: CD3NO2 for complexes 1–10 and CDCl3 for 11–14. Complex 1 which has the composition [CoCl2L1 2]+ was obtained by the reaction of [CoCl2(py)4]+ and L1 in a molar ratio of 1 : 1. It shows a singlet signal for the SMe group in the 1H and 13C NMR spectra, and the absorption spectral pattern is similar to those of 6 [Fig. 3(b)] and trans-[CoCl2- (Bu2PCH2CH2PBu2)2]+ (first absorption peak: 16 610 cm21 with log e = 2.05).22 Thus complex 1 is a trans-dichloro isomer with two didentate L1 ligands chelated through two phosphorus atoms. The reaction of 1 with Li(acac) yielded only [Co(acac)2{meso(P)-L1}]+ 13 and no rac(P)-L1 complex was observed. Thus 1 is trans-[CoCl2{meso(P)-L1}2]+, although no assignment can be made for two diastereomers, trans[P(R)P(R) or P(S)P(S)] and trans[P(R)P(S)].When meso(P)-L1 acts as a Fig. 3 Absorption spectra of (a) cis-a-[CoCl2{rac(P)-L1}]+ (——) and cis-a-[CoCl2{rac(P)-L2}]+ (– – – – –), (b) trans-[CoCl2{meso(P)-L1}2]+ (——) and trans-[CoCl2{meso(P)-L2}]+ (– – – –) in MeCN tetradentate ligand to form a trans isomer the central fivemembered P]Co]P chelate ring is forced to take an envelope conformation, and the complex will be unstable. Acetylacetonato complexes By the reaction with an equimolar amount of Li(acac), cis-a- [CoCl2{rac(P)-L}]+ (L = L1 or L2) yielded both cis-a and cis-b- [Co(acac){rac(P)-L}]2+.Since the cis-a and cis-b complexes have C2 and C1 symmetry, respectively (Fig. 4), the structures can be assigned easily from the 1H NMR spectra (Table 4). The cis-b-[Co(acac){meso(P)-L2}]2+ complex was prepared from trans-[CoCl2{meso(P)-L2}]+ by a similar method. However, neither cis-b-[Co(acac){meso(P)-L1}]2+ nor related complexes in which meso(P)-L1 acts as a tetradentate ligand were obtained.The reaction of trans-[CoCl2{meso(P)-L1}2]+ with an equimolar amount of Li(acac) afforded [Co(acac){meso(P)-L1}2]2+ and [Co(acac)2{meso(P)-L1}]+ in similar yields. When the tetradentate meso(P)-L1 forms a cis-b structure the remaining two coordination sites are surrounded by the two bulky phenyl groups of L1, and seem to hinder the co-ordination of a six-membered Fig. 4 Three possible isomers of cis-[Co(acac)L]2+44 J.Chem. Soc., Dalton Trans., 1997, Pages 39–45 acac chelate ring. An analogous ligand meso(P)-NH2CH2- CH2PPhCH2CH2PPhCH2CH2NH2(L3) yielded a cis-b isomer with acac in small yield, but the complex decomposed slowly in water.4 The 1H NMR spectra of the acac ligand in cis-a and cis-b- [Co(acac){rac(P)- or meso(P)-L}]2+ (L = L1 or L2) reflect the structures of three geometrical isomers shown in Fig. 4. The resonance peaks of the methine proton of cis-b- [Co(acac){rac(P)-L1}]2+ and the methine and one methyl group protons of cis-b-[Co(acac){meso(P)-L2}]2+ are observed at a fairly high field compared with those of the corresponding protons of other isomers (Table 4).The high-field shifts of these methine and methyl proton signals are attributed to the shielding effect of the phenyl group located near these protons. The complex [Co(acac){meso(P)-L1}2]2+ 5 shows six singlet methyl proton signals, one of which and the methine proton signal of acac are shifted to high field.This spectral pattern is consistent only with the molecular model of the trans[P(R)P(S)] isomer. Fig. 5 shows the absorption spectra of cis-a- and cis-b- [Co(acac){rac(P)-L2}]2+ and cis-b-[Co(acac){meso(P)-L2}]2+. These are similar to those of the corresponding isomers of [Co(acac)L4]2+ (L4 = NH2CH2CH2PPhCH2CH2CH2PPh- CH2CH2NH2); the first d–d bands of the cis-b isomers are broader than that of the cis-a isomer and have a shoulder to lower energy.4 The similarity in spectra between the L2 and the L4 complexes indicates that the ligand-field strength of SMe is similar to that of NH2.23 The spectra of the L1 complexes are quite similar to those of the corresponding L2 complexes.Both cis-a and cis-b isomers of [Co(acac){rac(P)-L}]2+ (L = L1 or L2) change in absorption spectra in organic solvents at elevated temperatures with isosbestic points, the final spectra being the same for both isomers. These results indicate that these two isomers isomerize to each other in solution to give an equilibrium mixture. The isomerization reactions were also monitored by the 1H NMR spectral changes with time, and the molar ratios of the isomers at equilibrium in MeNO2 solutions were obtained; cis-a: cis-b = 2 : 1 for the L1 complex and 3 : 2 for the L2 one.Such an isomerization reaction was not observed for the corresponding NPPN complexes.4 The reaction of L2 [a mixture of meso(P) and rac(P) isomers] with [Co(acac)3] in methanol in the presence of active charcoal afforded [Co(acac)2{meso(P)-L2}]+ and D(RR)/L(SS)- [Co(acac)2{rac(P)-L2}]+, and no D(SS)/L(RR) isomer of the rac(P)-L2 complex was formed.The same reaction with L1 did not yield the D(SS)/L(RR) isomer.7 However, reactions of cisa-[ CoCl2{rac(P)-L}]+ (L = L1 or L2) with Li(acac) in a molar ratio of 1 : 3 in the absence of active charcoal yielded a mixture Fig. 5 Absorption spectra of cis-a-[Co(acac){rac(P)-L2}]2+ (——), cisb-[ Co(acac){rac(P)-L2}]2+ (– – – –), and cis-b-[Co(acac){meso(P)-L2}]2+ (–?–?–?–) in MeCN of D(RR)/L(SS)- and D(SS)/L(RR)-[Co(acac)2{rac(P)-L}]+. Although the two isomers were not separated by column chromatography, their molar ratio was estimated to be D(RR)L(SS) :D(SS)/L(RR) = 2 : 1 for both the L1 and L2 complexes from the intensity ratio of the methine proton of acac in the 1H NMR spectra.Thus, [Co(acac)2{rac(P)-L}]+ seems to be more stable in the D(RR)/L(SS) than the D(SS)/L(RR) isomer.In the D(RR)L(SS) isomer the phenyl group of the SPPS ligand is located over the acac chelate ring as indicated by the high-field shift of the methine proton of acac. This structure may be more stable than that of the other D(SS)/L(RR) isomer where the CH2CH2SMe group is located over the acac chelate ring. Conclusion This study has revealed that L1 and L2 bearing only soft donor groups can function as a tetradentate ligand to a hard cobalt(III) ion to afford complexes of various geometrical isomers (trans, cis-a, and cis-b).The co-ordination modes are governed specifically by the absolute configurations of the inner chiral phosphorus atoms and the conformations of the P]Co]P chelate rings. Several complexes in which the SPPS compounds co-ordinate as didentate ligands through two phosphorus atoms were obtained (complexes 1, 5, 11 and 12), such co-ordination modes not being observed for the corresponding NPPN compounds. The isomerization equilibrium between the cis-a and cis-b isomers of acac complexes was observed also only for the SPPS-type ligands and not the NPPN-type ones.These differences in behaviour may be caused by the weaker co-ordination ability of thioether than amino groups to a cobalt(III) ion. Acknowledgements This work was partially supported by Grants-in-Aid for Scientific Research Nos. 06453048 and 08640709, Development Scientific Research No. 07554061, and Scientific Research on Priority Areas No. 08220230 from the Ministry of Education, Science and Culture of Japan. We thank the Institute for Molecular Science (Okazaki) for the use of X-ray facilities. References 1 F. A. Cotton and B. Hong, Prog. Inorg. Chem., 1992, 40, 179. 2 A. Airey, G. F. Swiegers, A. C. Willis and S. B. Wild, J. Chem. Soc., Chem. Commun., 1995, 693 and refs. therein. 3 K. Kashiwabara, M. Jung and J. Fujita, Bull. Chem. Soc. Jpn., 1991, 64, 2372. 4 M. Jung, M. Atoh, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1990, 63, 2051. 5 M. Atoh, H. Sugiura, Y. Seki, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1987, 60, 1699. 6 M. Atoh, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1986, 59, 1001. 7 T. Kitagawa, M. Kita, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1991, 64, 2942. 8 B. Bosnich, W. G. Jackson and S. B. Wild, J. Am. Chem. Soc., 1973, 95, 8269. 9 B. Bosnich, W. G. Jackson and S. B. Wild, Inorg. Chem., 1974, 13, 1121. 10 K. Issleib and W. Gans, Z. Anorg. Allg. Chem., 1982, 491, 163. 11 R. Schmelzer and D. Schwarzenbach, Cryst. Struct. Commun., 1981, 10, 1317. 12 J. Glerup, C. E. Schaffer and J. Springburg, Acta Chem. Scand., Ser. A, 1978, 32, 673. 13 B. E. Bryant and W. C. Fernelius, Inorg. Synth., 1957, 5, 188. 14 G. M. Sheldrick, SHELXS 86, Program for Crystal Structure Determination, University of Göttingen, 1986. 15 S. R. Hall, H. D. Flack and J. M. Stewart, XTAL 3.2, Program for X-Ray Crystal Structure Analysis, Universities of Western Australia, Geneva and Maryland, 1992. 16 T. Ando, M. Kita, K. Kashiwabara, J. Fujita, S. Kuracha and S. Ohba, Bull. Chem. Soc. Jpn., 1982, 65, 2748 and refs. therein.J. Chem. Soc., Dalton Trans., 1997, Pages 39–45 45 17 I. Kinoshita, Y. Yokota, K. Matsumoto, S. Ooi, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1983, 56, 1067. 18 M. Kita, K. Kashiwabara, J. Fujita, H. Tanaka and S. Ohba, Bull. Chem. Soc. Jpn., 1994, 67, 2457. 19 P. de Meester and D. J. Hodgson, J. Chem. Soc., Chem. Commun., 1976, 618. 20 R. C. Elder, G. J. Kennard, M. D. Payne and E. Deutsch, Inorg. Chem., 1978, 17, 1296. 21 K. Okamoto, T. Isago, M. Ohmasa and J. Hidaka, Bull. Chem. Soc. Jpn., 1982, 55, 1077. 22 T. Ohishi, K. Kashiwabara and J. Fujita, Bull. Chem. Soc. Jpn., 1987, 60, 575. 23 Y. Shimura, Bull. Chem. Soc. Jpn., 1988, 61, 693. Received 8th July 1996; Paper 6/04759I

ISSN:1477-9226    

DOI:10.1039/a604759i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 43–49 43 Cationic zirconocene complexes with benzyl and Si(SiMe3)3 substituted cyclopentadienyl ligands Manfred Bochmann,*a Malcolm L. H. Green,*b Annie K. Powell,c Jörg Saßmannshausen,a Michael U. Triller b and Sigrid Wocadlo c a School of Chemistry, University of Leeds, Leeds, UK LS2 9JT b Inorganic Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, UK OX1 3QR c School of Chemical Sciences, University of East Anglia, Norwich, UK NR4 7TJ Received 23rd September 1998, Accepted 5th November 1998 Alkyl zirconocenes [Zr(h-C5H4R)2X2] (where R = CH2Ph, X = Cl 1a or Me 1b; R = CHPh2, X = Cl 2a or Me 2b; R = Si(SiMe3)3, X = Cl 4a or Me 4b) and for comparison [Zr(h-C5H5)(h-C5H4CH2Ph)Cl2] 3a were prepared and characterised.The reactions of these compounds with the methide abstracting reagents B(C6F5)3, B(o-C6F5C6F4)3 and [Ph3C]1[B(C6F5)4]2 were investigated by low temperature NMR spectroscopy.Compound 1b reacts with [Ph3C]1[B(C6F5)4]2 to form the homodinuclear complex [{Zr(h-C5H4CH2Ph)2Me}2(m-Me)]1[B(C6F5)4]2. The related compound [{Zr(C5H4CH2Ph)2Me}2(m-Me)]1[MeB(C6F5)3]2 5a was formed from the reaction of 1b with 0.5 equivalent of B(C6F5)3. Reaction between 1b and 1 equivalent B(C6F5)3 gave [Me(h-C5H4CH2Ph)2Zr(m-Me)B(C6F5)3] 6a and the ion pair [Zr(h-C5H4CH2Ph)2Me][MeB(C6F5)3] 6b which are in equilibrium with each other. A similar observation was made when 2b was used instead of 1b.The sterically more demanding 4b does not show this behaviour. The role of the ligands in ethylene polymerisation was investigated. Introduction The activities and stereoselectivities of metallocene-based olefin polymerisation catalysts are known to be highly susceptible to the influence of substituents on the cyclopentadienyl ligand framework.1,2 The factors that aVect the polymer molecular weight have, by comparison, been less well studied. The average molecular weight of a polymer chain depends on the ratio of the rate of chain propagation, kp, and the rate of chain termination, kt, and since this ratio is temperature dependent the modification of polymer molecular weights may within certain limits be achieved by varying the reaction temperature.Similarly, the presence of weak donor ligands such as aromatic ethers and amines can convert a polymerisation catalyst into a system for the production of oligomers, albeit not without a significant reduction in activity since donor ligands eVectively compete with the olefinic substrate for the co-ordination site required for chain growth.3,4 Some eVorts to control the polymer molecular weight of polyethylenes by substituting the Cp ligands with alkyl groups of varying degrees of steric hindrance have been reported, although the eVect was not pronounced,5 and eVorts to modify the electronic characteristics of the h-cyclopentadienyl ligands by introducing electron withdrawing or donating substituents such as CF3 and NMe2 appear to aVect primarily the activity of the catalysts rather than the polymer molecular weight.6 We wished therefore to explore the possibility of controlling the kp :kt ratio by introducing substituents on the h-cyclopentadienyl which had the potential of acting as weakly and reversibly co-ordinating ligands to the metal centre in the catalytically active 14-electron species [Zr(h-C5H4R9)2R]1 and chose to investigate the influence of benzyl substituents.During the course of this work a related study has been reported.7 We report here the synthesis of cationic zirconocene alkyl species [Zr(h-C5H4R9)2R]1 where R9 = CH2Ph or CHPh2 and, for comparison, the bulky group Si(SiMe3)3 and their role in ethylene polymerisation. Results and discussion The known compound bis(benzylcyclopentadienyl)zirconium dichloride 1a was prepared as described (Scheme 1).8 The diphenylmethylcyclopentadienyl analogue 2a was prepared similarly as a white solid in 35% yield.The compound [Zr(h-C5- H5)(h-C5H4CH2Ph)Cl2] 3a was prepared as described as a light yellow powder.8 The bulky cyclopentadiene C5H5Si(SiMe3)3 was prepared in high yield from LiCp and SiBr(SiMe3)3 in THF as a yellow waxy material. Deprotonation of this compound with butyllithium followed by the addition of 0.5 equivalent of [ZrCl4- (thf)2] aVorded [Zr(CpSi)2Cl2] 4a, where CpSi = h-C5H4Si- (SiMe3)3, as a yellow solid in modest yield (Scheme 2).The Scheme 1 (i) Li[CpCH2Ph]; (ii) LiMe; (iii) Li[CpCHPh2]; (iv) K- [CpCH2Ph]; (v) MgMeCl. Ph ZrCl2 Ph ZrMe2 ZrCl4(thf)2 (CpCH Ph)2ZrCl2 (CpCH Ph)2ZrMe2 (CpCHPh )2ZrCl2 (CpCHPh )2ZrMe2 ( ii) 1b 1a 2a 2b CpZrCl3(dme) 3a 3b ( i) ( iii) ( ii) ( iv) ( v) 2 2 2 2 Scheme 2 (i) LiCp; (ii) LiBu, thf, 278 8C; (iii) [ZrCl4(thf)2], room temperature; (iv) LiMe, Et2O. Si(SiMe3)3 ZrCl2 Si(SiMe3)3 Si(SiMe3)3 ZrMe2 Si(SiMe3)3 4a Si(SiMe3)3 BrSi(SiMe3)3 4b ( i) ( ii), ( iii) ( iv)44 J.Chem. Soc., Dalton Trans., 1999, 43–49 Table 1 Proton and 13C NMR data Compound 1a [Zr(h-C5H4CH2Ph)2Cl2] Data a (d, J/Hz) 1H NMR (CDCl3, 270 MHz, 20 8C): 3.99 (s, 4 H, CH2Ph); 6.18 (t, 4 H, CpH, JHH = 1.65); 6.21 (t, 4 H, CpH, JHH = 1.97); 7.17–7.28 (m, 10 H, Ph) 13C NMR (CDCl3, 67.80 MHz, 20 8C): 36.10 (CH2Ph); 112.72 (Cp); 115.71 (Cp); 126.43 (p-C of Ph); 128.52 (m-C of Ph); 128.81 (o-C of Ph); 133.50 (ipso-C of Cp); 139.71 (ipso-C of Ph) 2a [Zr(h-C5H4CHPh2)2Cl2] 1H NMR (CDCl3, 270 MHz, 20 8C): 5.56 (s, 2 H, CHPh2); 5.72 (t, 4 H, JHH = 2.64, Cp); 5.97 (t, 4 H, JHH = 2.64, Cp); 7.02–7.04 (m, Ph) 13C NMR (CDCl3, 20 8C): 51.52 (CPh2); 115.09 (Cp); 116.09 (Cp); 126.73 (p-C of Ph); 128.41 (m-C of Ph); 129.06 (o-C of Ph); 136.33 (ipso-C of Cp); 143.48 (ipso-C of Ph) 4a [Zr{h-C5H4Si(SiMe3)3}2Cl2] 1H NMR (CDCl3, 90 MHz, 20 8C): 0.14 (s, 54 H, SiMe3); 6.3–6.5 (m, 8 H, Cp) 1b [Zr(h-C5H4CH2Ph)2Me2] 1H NMR (CDCl3, 270 MHz, 20 8C): 20.32 (s, 6 H, CH3, JCH = 117.5); 3.83 (s, 4 H, CH2Ph); 5.81 (t, 4 H, JHH = 2.64, Cp); 5.97 (s, 4 H, JHH = 2.64, Cp); 7.21–7.35 (m, 10 H, Ph) 13C NMR (CDCl3, 20 8C): 30.05 (CH3); 36.08 (CH2Ph); 107.84 (Cp); 111.48 (Cp); 126.13 (p-C of Ph); 126.47 (ipso-C of Cp); 128.44 (m-C of Ph); 128.52 (o-C of Ph); 141.15 (ipso-C of Ph) 2b [Zr(h-C5H4CHPh2)2Me2] 1H NMR (CDCl3, 270 MHz, 20 8C): 20.34 (s, 6 H, CH3); 5.32 (s, 2 H, CHPh2); 5.54 (t, 4 H, JHH = 2.64, Cp); 5.72 (t, 4 H, JHH = 2.64, Cp); 7.12–7.28 (m, 10 H, Ph) 13C NMR (CDCl3, 67.80 MHz, 20 8C): 30.75 (CH3); 51.81 (CHPh2); 108.86 (Cp); 111.56 (Cp); 126.43 (p- C of Ph); 128.32 (m-C of Ph); 128.84 (o-C of Ph); 129.47 (ipso-C of Cp); 144.58 (ipso-C of Ph) 3b [Zr(h-C5H5)(h-C5H4CH2Ph)Me2] 1H NMR (CDCl3, 270 MHz, 20 8C): 0.39 (s, 6 H, ZrMe); 5.14 (d, 1 H, CH2; JHH = 1.08); 5.41 (d, 1 H, CH2; JHH = 1.08); 5.98 (s., 5 H, Cp); 6.04 (t, 2 H, C5H4); 6.12 (t, 2 H, C5H4); 7.32 (“s”, 5 H, Ph) 13C NMR (CDCl3, 20 8C): 31.37 (ZrMe); 109.47 (C5H4); 109.68 (C5H4); 110.74 (Cp); 112.79 (CCH2); 124.16 (CCH2); 127.70 (p-C of PH); 128.17 (o-C of Ph); 128.39 (m-C of Ph); 141.54 (ipso-C of Cp9); 143.79 (ipso-C of Ph) 4b [Zr{h-C5H4Si(SiMe3)3}2Me2] 1H NMR (CDCl3, 270 MHz, 20 8C): 20.39 (s, 6 H, CH3); 0.0 (s, 54 H, SiCH3); 5.98 (t, 4 H, JHH = 2.3, Cp); 6.01 (t, 4 H, JHH = 2.30, Cp) 13C NMR (CDCl3, 67.80 MHz, 20 8C): 1.78 (SiCH3); 32.71 (CH3); 112.11 (Cp); 118.51 (Cp) a Cp indicates hydrogens attached to C5-ring carbons.spectroscopic data of all new compounds are collected in Table 1. Treatment of the dichloride compounds 1a–4a with either methyllithium or methylmagnesium chloride aVords the corresponding zirconocene dimethyl complexes 1b, 2b and 4b as colourless solids. The compound 3b could not be obtained pure and appears to be thermally sensitive. Slow cooling of a light petroleum solution of compound 1b to 5 8C forms colourless plates suitable for X-ray crystallography.The molecular structure of 1b is shown in Fig. 1 and is similar to that of the known dichloride 1a.9 Selected bond distances and angles are shown in Table 2. The structure of 1b shows that the benzyl substituent lies approximately in the plane bisecting the Me–Zr–Me angle, with the phenyl substituents bent away to occupy the sterically least encumbered position. This behaviour is similar to that of substituted titanocenes.10 The crystal packing indicates no close intermolecular contacts such as p stacking of the phenyl rings that would aVect the preferential conformation of the h-C5H4- CH2Ph groups.Synthesis of cationic complexes The reaction of compound 1b with 0.5 equivalent of B(C6F5)3 in dichloromethane was monitored by NMR spectroscopy. It is slow at 278 8C but quantitative at 220 8C to give as a single product the homodinuclear complex [{Zr(h-C5H4CH2Ph)2- Me}2(m-Me)]1[MeB(C6F5)3]2 5a. Similar methyl-bridged zirconocenes have been observed before.11–14 Brintzinger and co-workers 14 observed a tight ion pair [{[Zr(h-C5H5)2Me]2- (m-Me)}1MeB(C6F5)3 2] in C6D6 solutions.Solutions of 5a in CD2Cl2 show only solvated ion pairs (Scheme 3). When compound 1b was treated with 1 equivalent or a slight excess of B(C6F5)3 two mononuclear products were formed Fig. 1 Crystal structure of complex 1b. which appear to be in equilibrium. One compound is characterised by two singlets at d 0.20 and 0.75, assigned to MeB and ZrMe groups, respectively, and the NMR data are consistent with the formulation as the zwitterionic complex [Me(h-C5H4- CH2Ph)2Zr(m-Me)B(C6F5)3] 6a.This assignment was supported by 1H–13C COSY experiments. The corresponding MeB signal for the second compound is observed at d 0.48 and is assigned to the solvent separated isomer of 6a, namely the ion pair [Zr(h-C5H4CH2Ph)2Me][MeB(C6F5)3] 6b. In agreement with this assignment, the 13C NMR spectrum of 6b shows a broadened singlet at d 10.2, typical for a separated, non- Scheme 3 (i) 0.5 B(C6F5)3; (ii) B(C6F5)3; (iii) 0.5 PBB–C6D6 or PBB– CD2Cl2; (iv) Ph3C1. 1b (CpCH Ph)2Zr Me Me Zr(CpCH Ph)2 Me (CpCH Ph)2Zr Me Me B(C6F5)3 [MeB(C6F5)3]– + 6a (CpCH Ph)2Zr Me 5a 6b [MeB(C6F5)3]– (iii) (iv) (CpCH Ph)2Zr Me Me Zr(CpCH Ph)2 Me 5b or 5c [MePBB]– d 0.20 1H d »24 13C d 0.481H d »10 13C (CpCH Ph)2Zr Me Me Zr(CpCH Ph)2 Me [B(C6F5)4] – 5d (i) (ii) + + + 2 2 2 2 2 2 2 2 Table 2 Selected bond distances (Å) and angles (8) for compound 1b Zr(1)–C(10) Zr(1)–C(13) Zr(1)–C(30) Zr(1)–C(33) Zr(1)–C(11) Zr(1)–C(14) C(10)–Zr(1)–C(30) C(11)–Zr(1)–C(31) C(12)–Zr(1)–C(35) 2.277(3) 2.494(3) 2.283(3) 2.483(3) 2.570(3) 2.485(3) 98.08(13) 171.11(8) 150.16(10) Zr(1)–C(31) Zr(1)–C(34) Zr(1)–C(12) Zr(1)–C(15) Zr(1)–C(32) Zr(1)–C(35) C(13)–Zr(1)–C(34) C(14)–Zr(1)–C(33) C(15)–Zr(1)–C(32) 2.587(3) 2.487(3) 2.526(3) 2.533(3) 2.546(3) 2.546(3) 86.75(10) 94.37(11) 159.00(10)J.Chem. Soc., Dalton Trans., 1999, 43–49 45 bridging [MeB(C6F5)3]2 anion.The chemical shift diVerence between the m- and p-19F resonances of the anion 15 can also be indicative of anion co-ordination. Thus values of Dd(m,p-F) between 3 and 6 ppm indicate co-ordination, values <3 ppm indicate non-co-ordination. We observe 2 sets of signals for the 2 diVerent anions formed, one which has Dd(m,p-F) of 4.85 ppm, another which has Dd(m,p-F) of 2.81 ppm. However, the ZrMe signal for 6b could not be found (Scheme 3). It has been shown that the bulky Lewis acid B(o-C6F4C6F5)3 (PBB)11,12 reacts with group 4 metallocene dimethyls to form cationic, dimeric complexes, even when an excess of PBB is present.Treatment of 1b with 0.5 equivalent of PBB in C6D6 at room temperature gave the homodinuclear complex [{Zr- (h-C5H4CH2Ph)2Me}2(m-Me)]1[MePBB]2 5b, characterised in the 1H NMR spectrum by two singlets at d 20.1 and 21.07 which may be assigned to the terminal and bridging Me (Scheme 3).Treatment of compound 1b in dichloromethane with 1 equivalent of PBB at 260 8C gave an incomplete reaction which becomes quantitative at 230 8C to give the single compound 5c (Scheme 3). The same reaction in benzene gave decomposition of the metallocene, suggesting a crucial role for the solvent in these reactions. It may be that solvent separation of the anion and the cation is enhanced in dichloromethane and whereas for benzene this separation might not occur if the reaction product is confined within a solvent cage subsequent decomposition reaction might occur more readily.The reaction of 1b with 1 equivalent of [Ph3C]1[B(C6F5)4]2 in CD2Cl2 at 260 8C proceeds cleanly to the homodinuclear complex [{Zr(h-C5H4CH2Ph)2Me}2(m-Me)]1[B(C6F5)4]2 5d (Scheme 3). The reaction between compound 2b and 1 equivalent of B(C6F5)3 in dichloromethane gave rise to two compounds which are in equilibrium with each other. However, here the solvent separated compound is the dominant product, as indicated by the integration of the signals at d 0.16 and 0.53 assigned to the MeB signals of the zwitterionic complex [Me- (h-C5H4CHPh2)2Zr(m-Me)B(C6F5)3] 7a and the solvent separated complex [Zr(h-C5H4CHPh2)2Me]1[MeB(C6F5)3]2 7b, respectively (Scheme 4).The ratio between 7a and 7b is approximately 1 : 2 (ratio 6a : 6b is 3 : 1). The ZrMe signals were observed at d 0.75 (7a) and 0.84 (7b). The CH coupling constant of the Zr-CH3 is around 120 Hz, and there was no indication of an a agostic interaction.It is tempting to speculate that the formation of the ion separated compounds [Zr(h-C5H4CH2Ph)2Me]1[MeB(C6F5)3]2 6b and [Zr(h-C5H4CHPh2)2Me]1[MeB(C6F5)3]2 7b is facilitated by co-ordination of the phenyl rings to the metal centre and thereby assisting the dissociation of the anion. This may account for the fact that other substituted zirconocenes react with B(C6F5)3 to form only zwitterionic complexes, similar to 6a and 7a, under otherwise identical conditions.For example, we have studied the reaction with the sterically more encumbered 4b with 1 equivalent of B(C6F5)3. The reaction proceeded cleanly in dichloromethane to give the zwitterionic complex [Me{h-C5H4- Si(SiMe3)3}2Zr(m-Me)B(C6F5)3] 8. The VT-NMR spectroscopic data studies are collected in Table 3. Polymerisation of ethene with compounds 1b, 2b and 4b The ethene polymerisation reactions were performed as previously described.16 Catalyst systems were prepared by mixing 1b Scheme 4 (i) B(C6F5)3.(CpCHPh )2Zr Me Me B(C6F5)3 (CpCHPh )2Zr Me 7a 7b [MeB(C6F5)3]– + ( i) 2b d 0.16 1H d 0.53 1H d »10 13C 2 2 or 2b with either B(C6F5)3 or [Ph3C]1[B(C6F5)4]2, and in the case of 4b was mixed with methylaluminoxane (MeAlO)n (MAO). The data for the polymerisation experiments are given in Table 4. Polydispersivity values lie between 2 and 4 consistent with single-site catalysts. There is a decrease in polymer molecular weights from 4b via 1b to 2b.The compound 1b gives a more active catalyst system than 4b or 2b. The compounds 4b and 2b give catalyst systems with similar activities but the resulting polymers have diVerent molecular weights. Assuming the same amount of active catalyst centres are formed in both cases it appears that 2b enhances b-H elimination without reduction of activity compared with 4b. This implies that kp is the same but kt is approximately 8 times greater for 2b than 4b.It may be that the phenyl group acts as a weak Lewis base and enhances b-H elimination. Conclusion The new zirconocenes 1a, 2a and 4a and 1b, 2b and 4b are described. Reaction of 1b and 2b with methide abstracting agents gives unusual behaviour which may reflect participation of the phenyl rings, by co-ordination to the zirconium centre. This hypothesis is strengthened by comparison with the sterically more demanding 4b, which does not show any solvent separated species.However, the NMR data do not provide evidence for the co-ordination of the phenyl ring to the cationic metal centres of 1b and 2b. Experimental General procedures All experiments were carried out under a nitrogen atmosphere using standard Schlenk techniques. Solvents were dried over sodium (toluene, low in sulfur), sodium–potassium alloy (diethyl ether; light petroleum, bp 40–60 8C) sodium–benzophenone (thf) and calcium hydride (dichloromethane). The NMR solvents were dried over activated molecular sieves, freeze thawed and stored in Young’s-Tap sealed ampoules.The spectra were recorded on a Bruker ARX250 or DPX300, a VARIAN UnityPlus 500 or a JEOL EX90FT or EX270FT machine and referenced to the residual solvent peak for 1H. Chemical shifts are quoted in ppm relative to tetramethylsilane. The 13C NMR spectra were proton decoupled using the standard program installed on the spectrometer. CH-Coupling constants were measured by INEPT (insensitive nuclei enhanced by polarisation transfer).Ethylene (BOC) was purified by passing through CaCl2, P2O5, activated molecular sieves and AlEt3–silicone oil columns. The ligands C5H5R (R = CH2Ph or CHPh2) were prepared in moderate yields by treating the corresponding benzyl chlorides with sodium cyclopentadienide in thf.17,18 The known compounds bis(benzylcyclopentadienyl)- zirconium dichloride 1a and [Zr(h-C5H5)(C5H4CH2Ph)Cl2] 3a were prepared essentially as described.8 Crystal structure determination of compound 1b Crystal data.C26H28Zr, M = 431.70, triclinic, a = 9.857(2), b = 13.365(3), c = 8.244(2) Å, a = 103.48(2), b = 101.19(2), g = 87.95(2)8, U = 1036.0(4) Å3, T = 293(2) K, space group P1� , Z = 2, l = 0.71073 Å, m = 0.538 mm21, Dc = 1.384 Mg m23, F(000) = 448, crystal size 0.2 × 0.2 × 0.1 mm, q range for data collection 2.59 to 23.038, limiting indices 0 £ h £ 10, 214 £ k £ 14, 29 £ l £ 9, reflections collected 3085, independent 2888 (Rint = 0.0274).Absorption correction: none. Refinement method: full matrix least squares on F2, Data/restraints/ parameters: 2888/0/329. Goodness-of-fit on F2: 1.012. Final R indices [I > 2s(I)] R1 = 0.0231, wR = 0.0568; (all data) R1 = 0.0331, wR = 0.0604. Extinction coeYcient: 0.0005(6). Largest diVerence peak and hole: 0.249 and 20.379 e Å23.46 J. Chem. Soc., Dalton Trans., 1999, 43–49 Table 3 Proton and 13C NMR data of cationic complexes 5–10 Compound 5a in CD2Cl2, 220 8C Zr PhCH2 PhCH2 Me Zr CH2Ph CH2Ph Me Me B(C6F5)4 – 1H NMR (d, J/Hz) 20.83 (s, 3 H) a 0.38 (s, 6 H) 0.48 (s, 3 H) 3.74 (dd, 8 H) 5.97 (d, 4 H, JHH = 2.5) 6.07 (d, 4 H, JHH = 3.5) 6.08 (d, 4 H, JHH = 2.5) 6.31 (d, 4 H, JHH = 2.5) 7.25 (m, 20 H) m-Me ZrMe CH3B CH2Ph Cp H Cp H Cp H Cp H Ph 13C NMR (d) 23.05 b 32.82 36.38 42.05 110.33 112.29 112.87 115.05 127.03 128.36 128.92 131.73 139.14 m-CH3 CH3B CH2Ph ZrCH3 Cp Cp Cp Cp p-C Ph o-C Ph m-C Ph i-C Cp i-C Ph 19F NMR (d) 5b in C6D6, 20 8C Zr PhCH2 PhCH2 Me Zr CH2Ph CH2Ph Me Me MePBB– 21.07 (s, 3 H) a 20.88 (s, 3 H) 20.11 (s, 6 H) 3.4 (dd, 8 H) 5.38 (d, 8 H, JHH = 2.5) 5.65 (d, 8 H, JHH = 2.5) 7.0 (m, 20 H) m-CH3 CH3B ZrCH3 CH2Ph Cp Cp Ph 2123.58 c (d, 3 F, JFF = 17.9) 2139.05 (d, 3 F, JFF = 23.5) 2139.32 (d, 3 F, JFF = 22.6) 2139.40 (d, 3 F, JFF = 21.6) 2155.50 (t, 3 F, JFF = 21.2) 2159.30 (t, 3 F, JFF = 22.8) 2162.70 (t, 3 F, JFF = 21.2) 2163.30 (t, 3 F, JFF = 22.6) 2163.90 (t, 3 F, JFF = 22.6) 5c in CD2Cl2, 230 8C Zr PhCH2 PhCH2 Me Zr CH2Ph CH2Ph Me Me MePBB– 21.63 (s, 3 H) a 20.81 (s, 3 H) 0.41 (s, 6 H) 3.77 (dd, 8 H) 6.01 (s, 4 H) 6.09 (s, 4 H) 6.10 (s, 4 H) 6.35 (s, 4 H) 7.19 (d, 8 H, JHH = 7.00) 7.26 (t, 4 H, JHH = 6.50) 7.36 (t, 8 H, JHH = 6.50) CH3B m-CH3 ZrCH3 CH2Ph Cp H1 Cp H2 Cp H3 Cp H4 o-H Ph p-H Ph m-H Ph ª12 b 23.07 (JCH = 120) 41.53 (JCH = 134) 35.91 110.20 112.23 112.82 115.07 126.39 128.33 128.88 131.71 139.14 CH3B m-CH3 ZrCH3 CH2 Cp C3 Cp C1 Cp C2 Cp C4 p-C Ph o-C Ph m-C Ph i-C Cp i-C Ph 2130.87 c (s, 3 F) 2145.56 (d, 3 F, JFF = 23.04) 2145.67 (d, 3 F, JFF = 21.63) 2146.85 (d, 3 F, JFF = 15.05) 2162.92 (t, 3 F, JFF = 21.16) 2165.12 (t, 3 F, JFF = 21.63) 2169.36 (t, 3 F, JFF = 21.16) 2169.87 (t, 3 F, JFF = 21.16) 2170.72 (t, 3 F, JFF = 21.16) 5d in CD2Cl2, 260 8C Zr PhCH2 PhCH2 Me Zr CH2Ph CH2Ph Me Me MeB(C6F5)3 – 20.81 (s, 3 H) d 0.42 (s, 6 H) 3.72 (d, 4 H) 3.90 (d, 4 H) 6.00 (d, 4 H, JHH = 1.68) 6.07 (s, 8 H) 6.37 (d, 4 H, JHH = 1.68) ª7.26 f ª7.30 f ª7.36 f m-CH3 ZrCH3 CH2Ph CH2Ph Cp H1 Cp H2,3 Cp H4 o-H Ph p-H Ph m-H Ph 23.62 e (JCH = 134.6) 35.54 (JCH = 148.4) 41.19 (JCH = 118.9) 109.56 111.63 112.54 114.70 126.70 128.11 128.64 131.54 139.06 m-CH3 CH2Ph ZrCH3 Cp C3 Cp C1 Cp C2 Cp C4 p-C Ph o-C Ph m-C Ph i-C Cp i-C Ph 6a in CD2Cl2, 260 8C Zr CH2Ph CH2Ph Me Me B(C6F5)3 0.20 (br, 3 H) g 0.75 (s, 3 H) 3.87 (d, d, 4 H) 6.07 (s, 2 H) 6.11 (s, 2 H) 6.28 (s, 2 H) 6.39 (s, 2 H) 7.15 (d, 4 H) ª7.30 j ª7.35 j CH3B ZrMe CH2Ph Cp H1 Cp H2 Cp H3 Cp H4 o-H Ph p-H Ph m-H Ph ª24 h 35.47 43.108 (JCH = 123) 110.89 112.90 113.30 115.91 126.68 128.01 128.95 133.00 139.03 CH3B CH2Ph ZrMe Cp C3 Cp C2 Cp C1 Cp C4 p-C Ph o-C Ph m-C Ph i-C Cp i-C Ph 2140.24 i 2165.44 2170.29 o-F p-F m-F 6b in CD2Cl2, 260 8C Zr CH2 CH2Ph Me MeB(C6F5)3 – 0.48 (br, 1.5 H) g ª3.8 (2 H) j 5.38 6.23 6.38 6.55 ª7.35 ª7.27 7.48 CH3B CH2Ph Cp H1 ª Cp H2 Cp H3 Cp H4 p-H Ph o-H Ph m-H Ph ª10 h 35.02 113.4 114.79 115.57 117.62 127.37 128.95 129.06 129.81 138.49 CH3B CH2Ph Cp C1 Cp C3 Cp C2 Cp C4 p-C Ph o-C Ph m-C Ph i-C Ph i-C Ph 2139.57 i 2169.48 2172.29 o-F p-F m-FJ.Chem. Soc., Dalton Trans., 1999, 43–49 47 Table 3 (Contd.) Compound 1H NMR (d, J/Hz) 13C NMR (d) 19F NMR (d) 7a in CD2Cl2, 260 8C Zr CHPh2 CHPh2 Me Me B(C6F5)3 0.16 (br, 1.5 H) g 0.75 (s, 1.5 H) CH3B ZrMe 41.99 h (JCH = 121) ZrMe 7b in CD2Cl2, 260 8C Zr CHPh CHPh2 Me MeB(C6F5)3 – 0.53 (br, 5 H) g 0.84 (q, 3 H) 5.03 (s, 2 H) 5.81 (s, 2 H) 5.92 (s, 2 H) 6.04 (s, 2 H) 6.20 (s, 2 H) 6.8–7.6 (m, 20 H) CH3B ZrMe CHPh Cp H1 Cp H2 Cp H3 Cp H4 Ph ª10 h 49.19 (JCH = 118) 53.39 112.13 113.13 115.15 116.57 127.1 127.9 128.7 134.2 CH3B ZrMe CHPh2 Cp C2 Cp C1 Cp C3 Cp C4 p-C Ph o-C Ph m-C Ph i-C Cp 8 in CD2Cl2, 260 8C Zr Si(SiMe3)3 Si(SiMe3)3 Me Me B(C6F5)3 0.10 (m, 54 H) g 0.68 (s, 3 H) 5.80 (s, 2 H) 6.19 (s, 2 H) 6.43 (s, 2 H) 6.60 (s, 2 H) SiCH3 ZrMe Cp H1 Cp H4 Cp H3 Cp H 0.84 h ª23 45.97 (JCH = 122) 115.47 119.80 120.29 120.75 127.38 SiCH3 CH3B ZrMe Cp C2 Cp C1 Cp C3 Cp C4 i-C Cp Cp Hn (n = 1–4) denotes hydrogens of the C5 ring, coupled to Cp Cn (connectivity determined by CH correlation). a 500 MHz.b 125.7 MHz. c 470 MHz. d 300 MHz. e 75.5 MHz. f Obscured by triphenylethane. g 270 MHz. h 67.8 MHz. i 470 MHz, 290 8C. j Peaks not resolved. CCDC reference number 186/1237. See http://www.rsc.org/suppdata/dt/1999/43/ for crystallographic files in .cif format. Preparations Diphenylmethylcyclopentadiene. Chlorodiphenylmethane (100 g, 500 mmol) in 200 cm3 thf at 278 8C was treated with a solution of NaCp (500 mmol) in 200 cm3 thf by slow addition.The reaction mixture was warmed to room temperature and stirred for 30 min before pouring onto ice–water/NH4Cl. The organic phase was separated and the aqueous phase was washed with light petroleum. The combined organic phases were dried over MgSO4 and the volatiles removed under reduced pressure. The oily residue was vacuum distilled [bp (0.01 mmHg) 125–145 8C] giving an orange liquid.Yield: 30 g, 129 mmol, 25.8%. Bromotris(trimethylsilyl)silane. The compound (Me3Si)4- Si 19221 (19.0 g, 59.2 mmol) was dissolved in 250 cm3 thf. Methyllithium in diethyl ether (42 cm3, 59.2 mmol, 1.4 mol l21) was slowly added. The solution changed to red. The reaction mixture was stirred overnight, changing to orange, and some white, fluVy material was observed. This mixture was added slowly to a solution of 3.0 cm3 (59.2 mmol) Br2 in 50 cm3 thf at 278 8C.The red colour of the bromine slowly changed to yellow, indicating consumption of the bromine. A precipitate was formed, which later dissolved again. After complete addition the reaction mixture was warmed to room temperature and stirred for 1 h. The volatiles were removed under vacuum at room temperature to yield an orange oil, which was extracted with hot light petroleum and the solvent was removed under vacuum to yield a yellow oil.This was extracted again with 100 cm3 CH2Cl2 and the solvent removed under vacuum to yield a yellow oil. Yield: 19.4 g, 59.2 mmol, 100%. Tris(trimethylsilyl)silylcyclopentadiene. Freshly distilled C5H6 (3.91 g, 59.2 mmol) was diluted with 150 cm3 thf and 23.7 cm3 (59.2 mmol) butyllithium (2.5 mol l21) were added at 278 8C. After complete addition the reaction mixture was warmed to room temperature and stirred for 30 min. This mixture was added to a solution of 19.4 g (59.2 ol) of SiBr(SiMe3)3 in 100 cm3 thf at 278 8C.After complete addition the reaction mixture was allowed to warm up and stirred at room temperature for 1 h. The volatiles were removed under vacuum to yield a yellow oil which was extracted with 200 cm3 light petroleum. The filtrate was concentrated to dryness to yield a yellow, waxy material. The 1H NMR spectrum indicated a strong (ª27 H) signal for the SiMe3 protons at around d 0 and a weaker signal (ª4 H) for the Cp protons.Yield 18.3 g, 58 mmol, 98%. [Zr(Á-C5H4CH2Ph)2Cl2] 1a. Benzylcyclopentadiene (7.2 g, 46 mmol) in thf (100 cm3) at 278 8C was treated with butyllithium (18.4 cm3, 46 mmol) at 278 8C. The mixture changed to red. It was slowly warmed to room temperature and stirred for 1 h. The compound [ZrCl4(thf)2] [8.86 g (23 mmol)] was slowly added. The mixture changed to a darker red and finally to a rust brown. After stirring overnight, the solvent was removed under reduced pressure and the red-brown solid residue extracted with 150 cm3 toluene.The extract was stored at 278 8C for 2 d and an oV-white solid formed which was collected by filtration, washed and dried. Yield: 4.6 g, 9.7 mmol, 48.7%. A satisfactory elemental analysis could not be obtained.48 J. Chem. Soc., Dalton Trans., 1999, 43–49 Table 4 Data for ethylene polymerisation with zirconocenes 1b, 2b and 4b a Run 123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Zr/CGA 1b/Ph34C1 1b/Ph3C1 1b/Ph3C1 1b/Ph3C1 1b/Ph3C1 1b/Ph3C1 1b/B(C6F5)3 1b/B(C6F5)3 1b/B(C6F5)3 1b/B(C6F5)3 1b/B(C6F5)3 1b/B(C6F5)3 2b/Ph3C1 2b/Ph3C1 2b/Ph3C1 2b/Ph3C1 2b/Ph3C1 2b/Ph3C1 2b/B(C6F5)3 2b/B(C6F5)3 2b/B(C6F5)3 2b/B(C6F5)3 2b/B(C6F5)3 2b/B(C6F5)3 4b/MAO 4b/MAO 4b/MAO 4b/MAO 4b/MAO 4b/MAO T/8C 60 60 20 20 00 60 60 20 20 00 60 60 20 20 00 60 60 20 20 00 60 60 23 20 00 t/min 111111111111111111111111221111 PE yield/mg 68.3 88.8 123.0 112.5 163.5 154.5 60.5 75.2 133.1 137.2 96.6 137.8 75.0 12.7 57.8 56.7 78.8 64.5 27.4 97.0 52.9 53.6 80.3 69.5 59.1 42.3 96.1 112.1 71.2 58.8 Productivity b 409.8 532.8 738.0 675.0 981.0 927.0 363.0 451.2 798.6 823.2 579.6 826.8 450.0 76.2 346.8 340.2 472.8 387.0 164.4 582.0 317.4 321.6 481.8 417.0 177.3 126.9 576.6 672.6 427.2 532.8 1023 Mw 26.7 36.4 54.3 25.8 22.4 47.5 3.3 6.4 11.2 4.4 6.3 15.4 12.6 13.1 40.2 44.8 122.5 126.0 1023 Mn 7.4 9.1 18.9 8.3 8.8 21.8 1.1 2.5 4.6 1.0 2.8 6.7 3.2 3.4 8.8 10.4 40.4 51.8 Mw/Mn 3.6 4.1 2.9 3.2 2.5 2.2 3 2.5 2.4 4.4 2.2 2.2 3.9 3.7 4.5 4.3 3.0 2.4 a Conditions: 20 cm3 toluene, 1 bar ethylene, 10 mmol [Zr], 10 mmol CGA {cationic generating agent, i.e.[Ph3C]1[B(C6F5)4]2 or B(C6F5)3}; 2 cm3 MAO (10% in toluene, WITCO). b In 103 g PE (mol Zr h bar)21. [Zr( Á-C5H4CHPh2)2Cl2] 2a. Diphenylmethylcyclopentadiene (9.3 g, 40 mmol) in 250 cm3 thf at 278 8C was treated with butyllithium (16 cm3, 40 mmol) at 278 8C in a dropwise manner. The mixture changed to red.It was warmed to room temperature and stirred for 1.5 h. To this was added 6 g (20 mmol) [ZrCl4(thf)2] and stirred overnight. The solvent was removed under vacuum and the residue extracted with 200 cm3 toluene. The filtrate was stored in a freezer for 3 d to obtain white crystals, which were collected by filtration. Yield: 4.9 g, 7.9 mmol, 35%. A satisfactory elemental analysis could not be obtained. [Zr( Á-C5H4CH2Ph)2Me2] 1b. To a suspension of compound 1a (4.60 g, 9.7 mmol) in 200 cm3 diethyl ether at 278 8C was added slowly methyllithium (13.8 cm3, 19.4 mmol).The reaction mixture was stirred for 10 min at 278 8C, slowly warmed to room temperature and stirred for 1 h. The volatiles were removed under reduced pressure and the residue was extracted with 300 cm3 of warm light petroleum. The extract was filtered and the filtrate stored at 5 8C for 3 d which yielded white crystals suitable for X-ray crystallography. A second crop of crystals was obtained from the concentrated mother-liquor after 6 d at 5 8C.Yield: 2.50 g, 5.8 mmol 59.8%. (Found: C, 72.3; H,6.5. Calc.: C, 72.32; H, 6.54%). [Zr(Á-C5H4CHPh2)2Me2] 2b. Compound 2a (4.9 g, 7.9 mmol) was suspended in 250 cm3 diethyl ether at 278 8C. Methyllithium (11.3 cm3, 16.8 mmol) was slowly added. The mixture was stirred for 4 min at 278 8C, slowly warmed to room temperature, and stirred for 1 h. The volatiles were removed under vacuum to yield a white solid which was extracted with 200 cm3 light petroleum.The extract was stored at 230 8C for 1 day to yield white crystals. The residue was repeatedly extracted with warm light petroleum and the combined extracts were stored at 278 8C to yield further product. The fractions were combined and recrystallised from 400 cm3 warm light petroleum and slowly cooled to 230 8C to yield a white solid. (2.1 g, 3.6 mmol 45.6%) (Found: C, 77.87; H, 6.45. Calc.: C, 78.16; H, 6.21%). Attempted synthesis of [Zr(Á-C5H5)(Á-C5H4CH2Ph)Me2] 3b.A suspension of the compound [Zr(h-C5H5)(h-C5H4CH2Ph)- Cl2] (573.5 mg, 1.7 mmol) in diethyl ether at 278 8C was treated with LiMe?LiBr (2.0 cm3, 3 mmol) in Et2O. The reaction mixture was slowly warmed to room temperature and stirred for 1 h. A white precipitate formed. The volatiles were removed under reduced pressure and the residue was extracted with light petroleum. The extract was filtered, concentrated and stored at 278 8C for 1 d to yield a white solid.After filtration, the solid turned into a brown liquid at room temperature and gas evolution was observed. The NMR data indicate the formation of the product 3b as well as intractable side products. [Zr{Á-C5H4Si(SiMe3)3}2Cl2] 4a. The compound C5H5Si- (SiMe3)3 (18.3 g, 50 mmol) in 200 cm3 thf at 278 8C was treated with n-butyllithium (20 cm3, 50 mmol). The reaction mixture was warmed to room temperature and stirred for 2 h. To this was slowly added [ZrCl4(thf)2] (9.6 g, 25 mmol), changing to dark red. The reaction mixture was stirred overnight.The volatiles were removed under vacuum and the residue was extracted with 100 cm3 toluene and stored at 278 8C for 2 d. A light yellow solid was obtained which was collected by filtration. Yield: 3.5 g, 4.5 mmol 18%. (Found: C, 42.04, H, 7.76. Calc.: C, 42.81; H, 7.95%). [Zr{Á-C5H4Si(SiMe3)3}2Me2] 4b. To a suspension of compound 4a (3.5 g, 4.5 mmol) suspended in 250 cm3 diethyl ether at 278 8C was added methyllithium (6.3 cm3, 9 mmol), and the reaction mixture was stirred at low temperature for 30 min before being allowed to warm slowly to room temperature.Stirring was continued for 1 h. The mixture changed from pale yellow to colourless, became clear, and a new precipitate was formed. The volatiles were removed under reduced pressure and the residue was extracted with 200 cm3 light petroleum, andJ. Chem. Soc., Dalton Trans., 1999, 43–49 49 stored at 240 8C for 3 d.The resulting crystals were collected by filtration and the mother-liquor was cooled to 278 8C to yield another crop. Yield: 1.5 g, 2.0 mmol, 45% (Found: C, 45.99; H, 9.36. Calc.: C, 48.38; H, 9.20%). The low carbon content may be due to the formation of SiC. Low temperature NMR studies on cationic compounds Owing to the sensitivity of the cationic complexes, they were generated in situ, and isolation was not attempted. The general procedure was as follows.The zirconocene compound (ca. 0.1 mmol) was dissolved in 0.25 cm3 CD2Cl2 and the solution transferred to a precooled (278 8C) NMR tube. The cation generating agent, such as B(C6F5)3, [Ph3C]1[B(C6F5)4]2 or PBB (ca. 0.11 mmol), was dissolved in 0.28 cm3 CD2Cl2 and transferred to the zirconocene solution in the NMR tube. The tube was sealed and shaken vigorously to ensure complete mixing. The reaction mixture became yellow. The sample was placed in the precooled (260 8C) probe of the NMR spectrometer and 1H, 13C, H-COSY and CH-COSY spectra were recorded at 260 8C.The sample was warmed to ambient temperature at increments of 20 K and at each step a 1H NMR spectrum was recorded. Acknowledgements This work was supported by the Engineering and Physical Science Research Council. J. S. wishes to thank BASF, Germany, for a Ph.D. studentship. M. U. T. would like to thank the Studienstiftung des deutschen Volkes for financial support. We would like to thank Drs.L. H. Doerrer and D. Häußinger for assistance with NMR spectroscopy. We are grateful to RAPRA Technology, Shawbury, UK, for GPC measurements. References 1 M. Bochmann, J. Chem. Soc., Dalton Trans., 1996, 225; W. Kaminsky, Adv. Polym. Sci., 1997, 127, 144; J. Chem. Soc., Dalton Trans., 1998, 1413. 2 H.-H. Brintzinger, D. Fischer, R. Mühlhaupt, B. Rieger and R. Waymouth, Angew. Chem., 1995, 107, 1255; Y. van der Leek, K. Angermund, M. ReVke, R. Kleinschmidt, R. Goretzki and G. Fink, Chem. Eur. J., 1997, 4, 585; R. Kravchenko and R. M. Waymouth, Macromolecules, 1998, 31, 1; M. K. Leclerc and R. M. Waymouth, Angew. Chem., Int. Ed. Engl., 1998, 31, 922; M. Toto, L. Cavallo, P. Corradini, G. Moscardi, L. Resconi and G. Guerra, Macromolecules, 1998, 31, 3431; L. Resconi, F. Piemontesi, I. Camurati, O. Sudmeijer, I. E. Nifant’ev, P. V. Ivchenko and L. G. Kuzmina, J. Am. Chem. Soc., 1998, 120, 2308. 3 J. J. W. Eshuis, Y. Y. Tan, A. Meetsma and J. H. Teuben, J. Mol. Catal., 1990, 62, 277. 4 J. J. W. Eshuis, Y. Y. Tan, A. Meetsma and J. H. Teuben, Organometallics, 1992, 11, 362. 5 J. Tian and B. Huang, Macromol. Rapid Commun., 1994, 15, 923. 6 A. Kucht, H. Kucht, W. Song, M. D. Rausch and J. C. W. Chien, Appl. Organomet. Chem., 1994, 8, 437. 7 H. G. Alt, M. Jung and W. Milius, J. Organomet. Chem., 1998, 558, 111. 8 P. Renaut, G. Taintutier and B. Gauthéron, J. Organomet. Chem., 1978, 148, 35. 9 Y. Dusausoy, J. Protas, P. Renaut, B. Gautheron and G. Tainturier, J. Organomet. Chem., 1978, 157, 167. 10 P. C. Möhring, N. Vlachakis, N. E. Grimmer and N. J. Coville, J. Organomet. Chem., 1994, 483, 159. 11 Y. X. Chen, C. L. Stern, S. Yang and T. J. Marks, J. Am. Chem. Soc., 1996, 118, 12451. 12 Y.-X. E. Chen, M. V. Metz, L. Li, C. L. Stern and T. J. Marks, J. Am. Chem. Soc., 1998, 120, 6287. 13 M. Bochmann and S. J. Lancaster, Angew. Chem., Int. Ed. Engl., 1994, 33, 1634. 14 S. Beck, M.-H. Prosenc, H.-H. Brintzinger, R. Goretzki, N. Herfert and G. Fink, J. Mol. Catal. A, 1996, 111, 67. 15 A. D. Horton, J. de With, J. v. d. Linden and H. v. d. Weg, Organometallics, 1996, 15, 2672. 16 M. Bochmann and S. J. Lancaster, Organometallics, 1993, 12, 633. 17 B. F. Hallam and P. L. Pauson, J. Chem. Soc., 1956, 3030. 18 K. Alder and H. Holzrichter, Liebigs Ann. Chem., 1936, 524, 145. 19 P. J. Bonasia and J. Arnold, Inorg. Synth., 1997, 31, 162. 20 G. Gutekunst and A. G. Brook, J. Organomet. Chem., 1982, 225, 1. 21 H. Gilman and C. L. Smith, J. Organomet. Chem., 1968, 14, 91. Paper 8/07418F

ISSN:1477-9226    

DOI:10.1039/a807418f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1998, Pages 45–49 45 Oxo-centred trinuclear acetate complexes containing mixed-metal clusters. Crystal structure of a chromium(III)iron(III)nickel(II) complex and magnetic properties of a dichromium(III)magnesium(II) complex‡ Antony B. Blake,*,†,a Ekk Sinn,a Ahmad Yavari,a Keith S. Murray b and Boujemaa Moubaraki b a School of Chemistry, University of Hull, Hull, UK HU6 7RX b Department of Chemistry, Monash University, Clayton, Victoria 3168, Australia The crystal structure of [CrIII xFeIII 22xNiIIO(O2CCH3)6(py)3]?py 1 (x = ca. 0.7, py = pyridine), has been determined at 25 8C. Although the molecule has no crystallographically imposed symmetry, the metal atoms form an almost equilateral triangle, M ? ? ? M 3.274(3), 3.284(4), 3.281(3) Å, with the O atom less than 0.006 Å from the M3 plane, and the Cr, Fe, and Ni atoms extensively disordered over the three metal sites. The corresponding Cr2Mg, Cr2Ni, and Fe2Ni complexes are isomorphous with 1.The structure differs from that of the rhombohedral FeIII 2MII (MII = Mn, Fe, Co or Zn) and CrIII 2MII (MII = Mn, Fe or Co) analogues in the orientations of the pyridine rings. The planes of the pyridine ligands in 1 are approximately perpendicular to the M3 plane, and the non-coordinated pyridine molecule lies approximately in this plane. (This contrasts with the rhombohedral complexes, where the ligand pyridine rings are approximately parallel with the M3 plane and the non-co-ordinated pyridine lies with its plane on the three-fold axis of the M3O system.) The magnetic susceptibility of [CrIII 2MgO(O2CCH3)6(py)3]?py has been measured between 5 and 300 K, and fitted by use of an exchange hamiltonian 22JS1·S2, giving JCrCr = 217 cm21, a value considerably smaller than that (ca. 226 cm21) of the Cr2Ni and Cr2Co analogues.In a study some years ago of the magnetic properties of the mixed-metal trinuclear complexes [CrIII 2MIIO(O2CCH3)6(py)3]? py and the analogous FeIII 2MII and CrIIIFeIIIMII complexes (MII = Mg, Mn, Fe, Co, Ni, or Zn; py = pyridine) 1 we noted that they fall crystallographically into two groups.The majority crystallise in the rhombohedral (trigonal) space group R32, with the central O atom necessarily on a site of three-fold symmetry. They are isomorphous with the mixed-valence MnIII 2- MnII and FeIII 2FeII complexes, the crystal structures of which are known,2,3a,b and the crystal structures of three of these mixed-metal rhombohedral complexes, [M2CoO(O2CCH3)6- (py)3]?py (M = Mn,4 Fe 3b or Ru3c), have been reported.However, the Cr2Mg, Cr2Ni, CrFeNi, and Fe2Ni complexes, which are also isomorphous, do not conform to this trigonal class, but crystallise with a monoclinic unit cell, space group Cc or C2/c. We were intrigued by this difference, arising from a seemingly minor change at the atomic level. Further, our interpretation of the magnetic properties, in particular the unexpected effect of the divalent metal on the exchange interaction between the trivalent metal ions, was hampered by the absence of detailed information about the metal–oxygen bond lengths and angles.In the rhombohedral compounds it is obvious that the metal ions must be randomly distributed over the three equivalent metal sites in the molecule, and hence the individual coordination environments of the tri- and di-valent metals cannot be revealed by X-ray crystallography.Recently, extended X-ray absorption-edge fine-structure spectroscopy (EXAFS), an element-specific technique that does not depend on crystallographic order, has been successfully used to obtain limited information about the individual metal co-ordination environments in the rhombohedral CrIII 2CoII complex, throwing some light on the magnetic properties, but the precision of the distances and angles obtained is considerably lower than that achievable by crystallography.5 It seemed possible that in the † Present address: P.O.Box 89A, Kangaroo Valley, 2577, Australia. ‡ Non-SI unit employed: mB ª 9.27 × 10224 J T21. monoclinic compounds the lower molecular symmetry (at most a mirror plane or two-fold axis) might reflect an ordered arrangement of the MIII 2MII triangles, which would permit a full crystallographic determination of the environments of the individual metal ions in these molecules. Our belief in this possibility was encouraged by the case of some monoclinic FeIII 2FeII complexes with substituted pyridine ligands, where the Fe]O bond lengths indicated a substantial degree of order in the di- and tri-valent ion positions.6 We have now determined the crystal structure of the monoclinic complex [CrIII xFeIII 22xNiIIO(O2CCH3)6(py)3]?py 1 (x = ca. 0.7). Unfortunately, we find no evidence for an ordered distribution of the different metal ions among the three metal sites in this molecule. The crucial difference between the monoclinic and the rhombohedral compounds turns out to be in the orientations of the pyridine ligands.The structure is reported below. Our magnetic investigation of the mixed-metal FeIII 2MII (M = Mg, Mn, Co, or Ni) and CrIII 2MII (M = Co or Ni) complexes revealed a remarkable increase in the strength of the superexchange interaction between the two trivalent ions compared with that in the FeIII 3 or CrIII 3 complex.1 Data for the complex [CrIII 2MgO(O2CCH3)6(py)3]?py 2, where the interpretation is particularly simple because only one exchange parameter is involved, were unfortunately not available at that time.We have now measured the magnetic susceptibility of this compound between 5 and 300 K, and the results, also reported below, turn out to deviate significantly from the above generalisation. Experimental The compounds were obtained as well formed single crystals in the manner described previously.1 Chemical analysis Carbon, H, and N were determined by commercial microanalysis, Cr by oxidation to dichromate(VI) and titration against46 J.Chem. Soc., Dalton Trans., 1998, Pages 45–49 iron(II), and Mg by atomic absorption spectrometry {Found: C, 45.1; H, 4.52; Cr, 4.3; N, 6.6. [Cr0.7Fe1.3NiO(O2CCH3)6(py)3]?py 1 requires C, 45.0; H, 4.48; Cr, 4.3; N, 6.6. Found: C, 47.4; H, 4.74; Cr, 12.9; Mg, 3.2; N, 6.95. [Cr2MgO(O2CCH3)6(py)3]?py 2 requires C, 47.2; H, 4.70; Cr, 12.8; Mg, 2.6; N, 6.9%}.Crystallography All measurements on complex 1 were made as previously described,7 using a Rigaku AFC6S diffractometer with graphite-monochromated Mo-Ka radiation (l = 0.71069 Å) on a prismatic crystal having approximate dimensions 0.40 × 0.30 × 0.25 mm. Cell constants and an orientation matrix were obtained from a least-squares refinement based on the setting angles of 25 carefully centred reflections in the range 18.4 < 2q < 28.98. Reflection intensity data were collected at 22 8C by the w–2q scan technique.Scans of angular width (1.63 1 0.30tan q)8 were made at 8.08 min21 (in w), together with stationary background counts on each side of the reflection. Of the 5571 reflections collected, 5141 were unique (Rint = 0.040), and of these, 3242 had Fo 2 > 3s(Fo 2), where s was estimated from counting statistics.8 Lorentz-polarisation corrections were applied. An empirical absorption correction, based on azimuthal scans of several reflections, was applied. The resulting transmission factors ranged from 0.82 to 1.18.The intensities of three standard reflections measured at 150 reflection intervals showed no greater fluctuations than those expected from Poisson statistics. Crystal data. C32H38CrFeN4NiO13, Mr = 853.2, purple-black crystals from pyridine, monoclinic, a = 21.951(4), b = 12.466(3), c = 15.923(5) Å, b = 117.02(2)8, U = 3881(3) Å3, Z = 4, Dc = 1.46 g cm23, space group Cc or C2/c from systematic absences. The non-centric space group Cc (no. 9) was established from the Patterson synthesis, packing considerations, and intensity statistics, and confirmed by successful refinement. (The extra vectors expected in the Patterson synthesis if the space group was C2/c rather than Cc were absent. Moreover, the automatic structure solution packages we tried failed to find a solution in C2/c, and we were unable to find a model that would refine in the latter space group.) Structure solution and refinement.The metal positions were determined from a three-dimensional Patterson synthesis based on all data, and all other atoms from Fourier-difference maps. Full-matrix least-squares refinement, carried out using the TEXRAY program set,9 converged to give agreement indices R = 0.037, R9 = 0.037 for the enantiomorph shown, where R = o||Fo| 2 |Fc||/o|Fo|, R9 = [ow(|Fo| 2 |Fc|)2/owFo 2]� �� . The function minimised was ow(|Fo| 2 |Fc|)2, where w = 4Fo 2/s2(Fo 2) and s2(Fo 2) = [S2(C 1 R2B) 1 (pFo 2)2]/Lp 2, S, C, R, and B are, respectively, the scan rate, total integrated peak count, ratio of scan time to background counting time, and total background count; Lp is the Lorentz-polarisation factor and the factor p = 0.02 is introduced to reduce the weighting of the most intense reflections.For the other enantiomorph, agreement was slightly poorer, R = 0.040. The standard deviation of an observation of unit weight, [ow(|Fo| 2 |Fc|)2/(Nobs 2 Nvar)]� �� , where Nobs = number of observations, Nvar = number of variables, was 1.68.The non-hydrogen atoms were refined anisotropically, and in the absence of information about the distribution of the Cr, Fe, and Ni atoms among the metal sites all three were treated as Fe31. The N atom of the solvate pyridine molecule was apparently disordered over the six ring positions, and this molecule was modelled as benzene with an occupancy factor of 5/6 for the H atoms. Plots of ow(|Fo| 2 |Fc|)2 against |Fo|, reflection order in data collection, (sin q)/l, and various classes of indices, showed no unusual trends.Neutral atom scattering factors were taken from Cromer and Waber.10 Anomalous dispersion effects 11 were included in Fc, the values for Df 9 and Df 0 being those of Cromer.12 CCDC reference number 186/727. Magnetic measurements on complex 2 The sample was ground in air and placed in a gelatin capsule, and its susceptibility measured between 5 and 300 K in a field of 1 T with a SQUID magnetometer (Quantum Design model MPMS).The data were corrected for sample diamagnetism by use of Pascal constants, and for diamagnetism of the sample container. Results and Discussion Crystal structure of [CrIII xFeIII 22xNiIIO(O2CCH3)6(py)3]?py 1 (x = ca. 0.7) The method of preparation of complex 1, reaction of 1 mol [CrII 2(O2CCH3)4(H2O)2] with 1 mol [FeIII 2NiIIO(O2CCH3)6(py)3] in pyridine in the presence of air,1 does not precisely determine the ratio of CrIII to FeIII in the product.It is likely that a continuous range of compositions is possible between the limits CrIII 2NiII and FeIII 2NiII, and that any intermediate product of composition CrIII xFeIII 22xNiII will contain proportions of molecules of the limiting types as well as of the mixed type CrIIIFeIIINiII. The presence of the latter is manifested unequivocally by the appearance of a purple colour, quite distinct from the green and olive green of the respective limiting forms, as first suggested by Weinland and Gussmann,13 and discussed more fully elsewhere.1 Chromium analysis showed that the Cr :Fe ratio was actually 0.7 : 1.3 in the purple-black crystalline product 1.In a statistical distribution this would correspond to a sample containing the Cr2Ni, CrFeNi, and Fe2Ni complexes in the molar proportions 1 : 3.7 : 3.5. Fig. 1 is an ORTEP14 diagram of the trinuclear complex with the atom numbering scheme, and Fig. 2 shows the arrangement of the molecules, including the unco-ordinated pyridine, in the unit cell. Selected bond lengths and angles are given in Table 1. In space group Cc the molecule has no crystallographically imposed symmetry. A careful inspection of the electron-density distribution, bond lengths, and atomic vibrational parameters at the three metal sites does not show any definite evidence for Fig. 1 View of the molecule [CrFeNiO(O2CCH3)6(py)3] 1. In practice, the Cr, Fe, and Ni atoms should be regarded as distributed randomly over the positions shown, with disordered Cr2Ni and Fe2Ni species also present, as discussed in the textJ.Chem. Soc., Dalton Trans., 1998, Pages 45–49 47 localisation of the Ni21 ion in a particular metal site. Although atom M(1) has slightly longer M]O bonds (mean 2.008 Å) than those of M(2) and M(3) (means 1.990, 1.993 Å), which may indicate some preference for this site to be occupied by the larger Ni21 ion, the difference of ca. 0.017 Å is considerably less than would be expected in a fully ordered structure: NiII]O bonds are typically ca. 0.1 Å longer than CrIII]O or FeIII]O bonds, and in the CrIII 2CoII complex studied by EXAFS a difference of ca. 0.1 Å between the CrIII]O and CoII]O distances was found.5 In the largely valence-localised monoclinic complex [FeIII 2FeIIO(O2CCH3)6L3]L (L = 4-ethylpyridine) a difference of 0.074 Å between the longer and shorter Fe]m3-O bond lengths was found.6 Moreover, the three M]N bonds in 1 are essentially equal.Differences between the Cr31 and Fe31 ions would be even smaller, and in the absence of any indication to the contrary we assume that these are also distributed randomly. On the balance of evidence, we feel that any metal ion localisation in compound 1 is too small for the observed bond lengths and angles to be used in a meaningful discussion of the individual metal co-ordination geometries. Unpublished determinations of the crystal structures of the CrIII 2Mg and CrIII 2NiII complexes, which have the same crystal structure as that of 1, indicate that in these also there is extensive disorder in the distribution of the metal atoms over the three sites.15 The overall structure of the complex is similar to that in other oxo-centred trinuclear carboxylates, with the m3-O atom at the centre of an almost equilateral triangle of metal atoms, and we shall not discuss the details except to note that the M]m3-O bond lengths are towards the low end of the range (1.89–1.95 Å) observed in compounds of this general type.2–6,15–20 The main feature of the structure that distinguishes the monoclinic compound 1 and the isomorphous Cr2Mg, Cr2Ni, and Fe2Ni complexes from the other complexes concerns the orientation of the pyridine molecules.The planes of the pyridine ligands in 1 are approximately perpendicular to the M3O plane (dihedral angles 82.0, 83.3 and 90.08), while the non-coordinated pyridine molecule has its plane approximately parallel and nearly coincident with the M3O plane (dihedral angle 13.68).The orientations of the ligand and unco-ordinated pyridine molecules in the Cr2Mg and Cr2Ni complexes are closely similar to those in 1.15 This is in direct contrast with the structure found in all those complexes which crystallise in the trigonal space group R32 (mixed-valence VIII 2VII,16 CrIII 2CrII,17 MnIII 2MnII,2 and FeIII 2FeII complexes,2,3a the FeIII 2CoII,3b RuIII 2- CoII,3c and MnIII 2CoII complexes,4 and presumably all the other mixed-metal complexes that are isomorphous with these,1 including FeIII 2Mg).In those compounds the ligand pyridine rings lie very nearly parallel with the M3O plane, and the nonco- ordinated pyridine molecule lies with its plane perpendicular to the M3O plane, along the three-fold axis (and disordered about this axis). An approximately perpendicular orientation of the terminal Fig. 2 Unit cell of complex 1 ligands with respect to the M3O plane has also been observed in [CrIII 3O(O2CC6H5)6(py)3]ClO4,18 and in two complexes containing a substituted pyridine ligand, [FeIII 2FeIIO(O2CCH3)6- (NC5H4CH3-3)3]?solv (where solv = 3-methylpyridine or toluene). 6 In the 4-ethylpyridine FeIII 2FeII complex,19 and in the valence-localised complex [MnIII 2MnIIO(O2CCH3)6(NC5H4- Cl-3)3],20 two of the ligands are approximately perpendicular to the M3O plane and the third is parallel with it.Apart from the complexes containing a substituted pyridine ligand, where molecular packing effects may be expected tthe ligand orientations, the tendency for a complex of the type [M3O(O2CCH3)6(py)3]?py to adopt the rhombohedral (R32) or monoclinic (Cc) structure, each with a different orientation of the pyridine molecules, must be determined by rather subtle forces arising within the M3O cluster. Dimorphism has not been observed: it seems that the complex always adopts either the one structure or the other, depending on the nature of the metals.Only the presence of nickel (and, to a lesser extent, magnesium) in the cluster appears to favour the perpendicular orientation of the pyridine ligands, all other metals giving the parallel orientation. The effect may possibly be related to the fact that Ni21 has the smallest radius of the divalent metal ions studied (Mg21 having the second smallest), although all of the divalent ions are, of course, larger than Cr31 or Fe31.21 Magnetic properties of [CrIII 2MgO(O2CCH3)6(py)3]?py 2 The molar magnetic susceptibility cm(Cr2Mg) of the complex was calculated by the matrix method 22 from the Hamiltonian (1), where S1 = S2 = ��� .Allowance was made in equation (2) for H = 22JS1 ·S2 1 gmB(S1 1 S2) ·H (1) cm = (1 2 r)cm(Cr2Mg) 1 r(2NAg2mB 2/3kT)(15/4) (2) the presence of a fraction r of the Cr31 in monomeric form (assumed to behave as a simple paramagnet).The calculated susceptibility was then fitted by least squares to the experimental data by the Simplex method,23 minimising the function R2, equation (3). Fig. 3 shows the agreement between the cal- R2 = o[(cm)obs 2 (cm)calc]2/o[(cm)obs]2 (3) culated and observed values of the molar magnetic susceptibility and the effective magnetic moment per molecule, m/mB = [(3k/NAmB 2)(cmT)]� �� . The magnetic moment of a pair of exchange-coupled Cr31 Table 1 Selected interatomic distances (Å) and angles (8) in compound 1 M(1) ? ? ? M(2) M(2) ? ? ? M(3) M(3) ? ? ? M(1) M(1)]O(1) M(1)]O(11) M(1)]O(21) M(1)]O(31) M(1)]O(41) M(1)]N(11) M(2)]O(1) M(2)]O(12) M(1)]O(1)]M(2) M(2)]O(1)]M(3) M(3)]O(1)]M(1) O(1)]M(1)]O(11) O(1)]M(1)]O(21) O(1)]M(1)]O(31) O(1)]M(1)]O(41) O(1)]M(1)]N(11) O(1)]M(2)]O(12) 3.274(3) 3.281(3) 3.284(4) 1.908(4) 2.024(5) 2.034(5) 2.040(5) 2.036(5) 2.184(6) 1.880(4) 2.011(5) 119.7(2) 120.8(2) 119.5(2) 96.8(2) 92.8(2) 95.9(2) 96.2(2) 177.4(2) 98.0(2) M(2)]O(22) M(2)]O(51) M(2)]O(61) M(2)]N(21) M(3)]O(1) M(3)]O(32) M(3)]O(42) M(3)]O(52) M(3)]O(62) M(3)]N(31) O(1)]M(2)]O(22) O(1)]M(2)]O(51) O(1)]M(2)]O(61) O(1)]M(2)]N(21) O(1)]M(3)]O(32) O(1)]M(3)]O(42) O(1)]M(3)]O(52) O(1)]M(3)]O(62) O(1)]M(3)]N(31) 2.011(5) 2.026(5) 2.021(5) 2.189(6) 1.893(4) 2.019(5) 2.009(5) 2.027(5) 2.018(5) 2.180(6) 95.7(2) 95.8(2) 94.2(2) 177.9(2) 94.7(2) 95.9(2) 96.4(2) 94.8(2) 178.8(2)48 J.Chem. Soc., Dalton Trans., 1998, Pages 45–49 ions with negative J would be expected to vanish as the temperature approached zero.Fitting only the data above 50 K and making no allowance for paramagnetic impurity, we obtained a calculated curve of the expected type, with g = 1.94 and J = 214 cm21 (215 cm21 if the value of g is fixed at 1.98). However, it is evident that the experimental susceptibility below about 20 K is dominated by the impurity contribution, and use of the whole data set with allowance for this factor gave a greatly improved fit, with g = 2.00, J = 217 cm21, and r = 0.06.As shown in Fig. 3 (inset), when the impurity is assumed to be paramagnetic the calculated moment levels off to a value of 1.4mB at low temperature. An improved fit to the data below 10 K can be obtained by introducing a small Weiss constant (i.e. replacing T by T 2 q) in the impurity contribution to equation (2), suggesting that the impurity is weakly antiferromagnetic. The best fit had g = 1.988, J = 217.1 cm21, r = 0.073, and q = 21.6 K, with R = 0.011. The rather large value of r is somewhat perplexing, since the sample was of good crystallinity and gave satisfactory chemical analysis.It is conceivable that the impurity might be the CrIII 2- CrII analogue, which could have a ground-state spin of either 2 or 5, depending on the exchange parameters, and would be scarcely detectable by chemical analysis. Whatever the significance of r might be, however, it does not greatly affect the estimation of J.In previous work we have drawn attention to the increase in magnitude of the coupling constants JFeFe and JCrCr in the FeIII 3 and CrIII 3 oxoacetate complexes by a factor of ca. 2.3 when one of the trivalent ions is replaced by a divalent ion, as in the Fe2Mn, Fe2Ni, Fe2Mg, Cr2Co, and Cr2Ni, complexes, Table 2.1 We regard the magnitude of this effect as evidence that the central O atom provides the main pathway for superexchange, and have argued that the increase in J must be due at least Fig. 3 Magnetic moment (solid circles and left-hand axis) and molar magnetic susceptibility (open circles and right-hand axis) of compound 2. Curves: (a) fitted to data above 50 K with J = 214 cm21, (b) fitted to the full data set with J = 217 cm21, r = 0.073, and q = 21.6 K, calculated as described in the text. Inset: the effect of including (lower curve) the Weiss constant q in the paramagnetic impurity correction Table 2 Values of the exchange parameter J(MIII]MIII) in trinuclear oxoacetate complexes [MIII 3O(O2CCH3)6(H2O)3]1 and [MIII 2MIIO- (O2CCH3)6(py)3] Exchange cluster FeIII 2O]FeIII FeIII 2O]MnII FeIII 2O]NiII FeIII 2O]MgII CrIII 2O]CrIII CrIII 2O]CoII CrIII 2O]NiII CrIII 2O]MgII 2J/cm21 30 ± 3 64 ± 3 73 ± 3 62 ± 3 11 ± 1.5 27 ± 2 26 ± 1 17 ± 1 Ref. 24, 25 111 24, 26 11 This work partly to polarisation of the electron cloud at the central O atom due to the asymmetric charge distribution in the trinuclear cluster.1,5 The value of 217 cm21 for J in the Cr2Mg complex is, however, much lower than that of ca. 226 cm21 that would be expected by comparison with the other compounds, if the main factor at work is simply the charge on the Mg21 ion. It is possible that differences in bond lengths or angles play a particularly important role in determining the value of J in this case, but we can at present see no reason why the Cr]O bond lengths and angles in the Cr2Mg complex should be signifi- cantly different from those in the Cr2Co and Cr2Ni complexes (bearing in mind that the radius 21 of Mg21 is considered to lie between those of Co21 and Ni21), especially in view of the similarity in J values shown by the Fe2Mg and Fe2Ni complexes.Further investigation, perhaps by EXAFS measurements, may be necessary to resolve this question. Acknowledgements We thank the SERC (UK) and the ARC (Australia) for financial support, and the computer centres of the University of Hull and the University of Wollongong, NSW, Australia for computing facilities.References 1 A. B. Blake, A. Yavari, W. E. Hatfield and C. N. Sethulekshmi, J. Chem. Soc., Dalton Trans., 1985, 2509. 2 S. E. Woehler, R. J. Wittebort, S. M. Oh, T. Kimbara, D. N. Hendrickson, D. Iniss and C. E. Strouse, J. Am. Chem. Soc., 1987, 109, 1063; A. R. E. Baikie, M. B. Hursthouse, D. B. New and P. Thornton, J. Chem. Soc., Chem. Commun., 1978, 62; J. B. Vincent, H.-R. Chang, K. Folting, J.C. Huffman, G. Christou and D. N. Hendrickson, J. Am.Chem. Soc., 1987, 109, 5703. 3 (a) S. E. Woehrler, R. J. Wittebort, S. M. Oh, D. N. Hendrickson, D. Iniss and C. E. Strouse, J. Am. Chem. Soc., 1986, 108, 2938; (b) H. G. Jang, K. Kaji, M. Sorai, R. J. Wittebort, S. J. Gleib, A. L. Rheingold and D. N. Hendrickson, Inorg. Chem., 1990, 29, 3547; (c) A. Ohto, Y. Sasaki and T. Ito, Inorg. Chem., 1994, 33, 1245. 4 R. D. Cannon, U. A. Jayasooriya, L. Montri, A. K. Saad, E. Karu, S.K. Bollen, W. R. Sanderson, A. K. Powell and A. B. Blake, J. Chem. Soc., Dalton Trans., 1993, 2005. 5 A. B. Edwards, J. M. Charnock, C. D. Garner and A. B. Blake, J. Chem. Soc., Dalton Trans., 1995, 2515. 6 S. M. Oh, S. R. Wilson, D. N. Hendrickson, S. E. Woehler, R. J. Wittebort, D. Iniss and C. E. Strouse, J. Am. Chem. Soc., 1987, 109, 1073. 7 J. R. Backhouse, H. M. Lowe, E. Sinn, S. Suzuki and S. Woodward, J. Chem. Soc., Dalton Trans., 1995, 1489. 8 P. W. P. Corfield, R. J.Doedens and J. A. Ibers, Inorg. Chem., 1967, 6, 197. 9 TEXSAN-TEXRAY, Struis Package, Molecular Structure Corporation, Houston, TX, 1985. 10 D. T. Cromer and J. T. Waber, International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.2A. 11 J. A. Ibers and W. C. Hamilton, Acta Crystallogr., 1964, 17, 781. 12 D. T. Cromer, International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4, Table 2.3.1. 13 R.Weinland and E. Gussmann, Chem. Ber., 1909, 42, 3881. 14 C. K. Johnson, ORTEP II, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. 15 A. B. Blake, M. B. Hursthouse, M. Motevalli and R. L. Short, unpublished work. 16 F. A. Cotton, M. W. Extine, L. R. Falvello, D. B. Lewis, G. E. Lewis, C. A. Murillo, W. Schwatzer, M. Tomas and J. M. Troup, Inorg. Chem., 1986, 25, 3505. 17 F. A. Cotton and W. Wang, Inorg. Chem., 1982, 21, 2675. 18 A. Harton, M. K. Nag, M. M. Glass, P. C. Junk, J. L. Atwood and J. B. Vincent, Inorg. Chim. Acta, 1994, 217, 171. 19 S. M. Oh, D. N. Hendrickson, K. L. Hassett and R. E. Davies, J. Am. Chem. Soc., 1985, 107, 8009. 20 A. R. E. Baikie, M. B. Hursthouse, L. New, P. Thornton and R. G. White, J. Chem. Soc., Chem. Commun., 1980, 684. 21 R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B, 1970, 26, 1046; Sect. A, 1976, 32, 751. 22 A. B. Blake, J. Chem. Soc., Dalton Trans., 1981, 1041.J. Chem. Soc., Dalton Trans., 1998, Pages 45–49 49 23 J. A. Nelder and R. Mead, Comput. J., 1965, 7, 108; implementation in Nottingham Algorithms Group (NAG) Fortran Library. 24 A. Earnshaw, B. N. Figgis and J. Lewis, J. Chem. Soc. A, 1966, 1656; Y. V. Rakitin, T. A. Zhemchuzhnikova and V. V. Zelentsov, Inorg. Chim. Acta, 1977, 23, 145. 25 J. F. Duncan, C. R. Kanekar and K. F. Mok, J. Chem. Soc. A, 1969, 480; G. J. Long, W. T. Robinson, W. P. Tappmeyer and D. L. Bridges, J. Chem. Soc., Dalton Trans.,1973, 573; C. T. Dziobkowski, J. T. Wrobleski and D. B. Brown, Inorg. Chem., 1981, 20, 671. 26 M. Sorai, M. Tachiki, H. Suga and S. Seki, J. Phys. Soc. Jpn., 1971, 30, 750; J. Ferguson and H. U. Güdel, Chem. Phys. Lett., 1972, 17, 547. Received 7th August 1997; Paper 7/05778D

ISSN:1477-9226    

DOI:10.1039/a705778d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

DALTON J. Chem. Soc., Dalton Trans., 1996, Pages 47.53 47 EVect of sterically inhibited axial azaferrocene ligands on the physical properties of iron(III) porphyrins. Crystal structures of bis(azaferrocene) complexes of iron(III) and cobalt(III) porphyrinates Michele Cesario,a Charles Giannotti,*,a Jean Guilhem,a Jack Silver *,¢Ó,b and Janusz Zakrzewski *,c a Institut de Chimie des Substances Naturelles, CNRS, F-91198 Gif-sur-Yvette, France b Department of Biological and Chemical Sciences, Central Campus, University of Essex, Wivenhoe Park, Colchester CO4 3SQ, UK c Department of Organic Chemistry, University of �©odz¢¥, PL-90-136 �©odz¢¥, Narutowicza 68, Poland The crystal structures of [Fe(oep)L2][O3SCF3] and [Co(oep)L2][ClO4] [H2oep = 2,3,7,8,12,13,17,18- octaethylporphyrin, L = [Fe(C5H5)(C4H4N)] have been determined.The azaferrocene ligands L in both compounds are seen to be parallel. The axial M]N (azaferrocene) bonds are longer than those found in analogous imidazole and pyridine complexes.Rhombic EPR spectra for bis(azaferrocene) complexes of iron(III) porphyrins are postulated to be a result of steric interactions of the azaferrocene H 2,5 atoms splitting the degeneracy of the dxz and dyz orbitals. The Mossbauer spectra of these iron complexes were recorded and interpreted as evidence for s bonding of the azaferrocene nitrogen atom to the porphyrin iron core with little or no p back bonding from the porphyrinic iron to the N atom.This is in agreement with the long bond lengths found in the structures. The rhombic EPR spectra observed are in keeping with a parallel arrangement of the ligand planes, as are the magnitudes of the Mossbauer quadrupole splittings. It is shown that the large gz values of the complexes are not unusual and must relate to long axial bonds and to steric effects splitting the dxz and dyz orbitals. There has been considerable interest in the stereochemistry and electronic properties of iron(III) and cobalt(III) porphyrins containing imidazole and pyridine ligands of the general formula [M(por)B2]+ (M = FeIII or CoIII; por = porphyrin dianion, B = imidazole or pyridine).1.7 In part this interest has been triggered by a need to understand the chemistry of cytochromes b.8.15 It is now well established that the rotational orientation of the planes of the axially co-ordinated B ligands with respect to the equatorial M]N (porphyrin) bond vectors strongly influences the physicochemical properties of iron(III) porphyrin complexes such as redox potentials, spin states, NMR, EPR and 57Fe Mossbauer spectra.1,4,5,7 It has been postulated and become widely accepted that many biological functions of haem proteins may be dependent on fine tuning of the orientation of the co-ordinated imidazole ring of the histidine residue.It has been suggested that some of the physicochemical properties correlate and are diagnostic of particular types of orientation.1,4,5,7,12,16.23 A value of 2.43 mm s21 for the Mossbauer quadrupole splitting (q.s.) for the imidazolate complex [Fe(por)(Him)(Im)] was assigned to a parallel orientation of the planar axial ligands, whereas a value of 1.87 mm s21 for [Fe(por)(mim)2]+ (mim = 2- methylimidazole) was assigned to a perpendicular orientation of the imidazole planes.4,5 The latter Mossbauer parameters have been associated with EPR spectra described as highly anisotropic low-spin systems (HALS).1,4,5,7,24 From extensive studies on model complexes it has been established that, for nonsterically hindered porphyrins, non-hindered imidazoles and highly basic pyridines (such as 4-aminopyridine) favour parallel orientation of their axial ligand planes.Such complexes display high q.s. values and normal rhombic EPR spectra. In previous work we have demonstrated the ligating properties of azaferrocene [Fe(C5H5)(C4H4N)] to a variety of acceptor species including metal porphyrin complexes.25.29 We have reported electronic absorption and EPR spectra for [Fe- (oep){Fe(C5H5)(C4H4N)}2][O3SCF3] (H2oep = 2,3,7,8,12,13,- 17,18-octaethylporphyrin) and [Fe(tpp){Fe(C5H5)(C4H4N)}2]- ¢Ó Present address: School of Chemical and Life Sciences, University of Greenwich, Wellington Street, London SE18 6PT, UK.[O3SCF3] (H2tpp = 5,10,15,20-tetraphenylporphyrin).27 These complexes gave rhombic EPR spectra which allowed us to suggest that they were low spin and had a parallel orientation of the azaferrocene ligands.27 Such an orientation was also established by X-ray studies on [Fe(tfpp){Fe(C5H5)- (C4H4N)}2] [H2tfpp = 5,10,15,20-tetrakis(pentafluorophenyl)- porphyrin] which is a low-spin iron(II) complex.28 The EPR spectra had gz parameters of 3.24 for the oep derivative and 3.16 for the tpp derivative.These are larger than those previously reported for normal rhombic spectra (gz �£ 2.9), but smaller than HALS gz values (3.4 for tpp).7 These larger values merited further investigation and we now report Mossbauer spectroscopic studies on [Fe(oep)L2][O3SCF3] 1, [Fe- (tpp)L2][O3SCF3] 2, [Co(tpp)L2][BF4] 3, [Fe(tfpp)L2] 4 and brief crystallographic details of 1 and [Co(oep)L2][ClO4] 5 [L = Fe(C5H5)(C4H4N)].Experimental Azaferrocene and its porphyrin complexes were synthesized and purified as previously described.27,28,30 Synthesis of [Co(oep){Fe(C5H5)(C4H4N)}2][ClO4] 5 The complex [Co(oep)(H2O)2][ClO4] 31 (36.4 mg, 0.05 mmol) and azaferrocene (40 mg, 0.21 mmol) were dissolved in CH2Cl2 (1 cm3).After 5 min of stirring at room temperature (r.t.) heptane (2 cm3) was added and the resulting solution concentrated in vacuo to give purple crystals which were filtered off, washed with pentane, and dried overnight at 0.2 Torr (26.6644 Pa) at r.t. (Found: C, 60.65; H, 5.85; N, 7.90. Calc. for C54H62ClCoFe2- N6O4: C, 60.9; H, 5.85; N, 7.90%). Positive-ion FAB mass spectrum (m-nirobenzyl alcohol matrix): m/z 964 ([Co(oep)- {Fe(C5H5)(C4H4N)}2]+), 777 ([Co(oep){Fe(C5H5)(C4H4N)}]+) and 591 {[Co(oep)]+}.Crystallography Single crystals were obtained by crystallisation from dichloromethane .heptane: dark purple pyramidal crystals for 1, bright purple prismatic crystals for 5.48 J. Chem. Soc., Dalton Trans., 1997, Pages 47–53 Data collection. Graphite-monochromated Mo-Ka radiation (l 0.710 73 Å), four-circle Philips PW1100 diffractometer, 293(2) K, w–2q scan technique, unit-cell dimensions refined from setting angles of 25 reflections (10 < q < 148).Three standard reflections measured every 3 h to monitor instrument and crystal stability; s(I) derived from counting statistics. Lorentz-polarisation effects and empirical absorption corrections applied.32 Structure resolution. Direct methods using SHELXS 86,33 refinement by least-squares method using SHELXL 93 34 for compound 1 and by SHELX 7635 for 5. Compound 1.Crystal structure solved in the space group P1� , taking the centre of inversion at the Fe of the porphyrinate ligand. Disorder of the trifluoromethane sulfonate anion on an inversion centre could not be resolved by simple use of this space group (more than the two expected anion positions were observed). Only three peaks were clearly observed in the Fourier-difference synthesis, so the refinement was concluded in the centrosymmetric space group with the coordinates of the observed cationic complex and the three identified anion peaks, with anisotropic thermal parameters.Two of the eight ethyl chains are also disordered (occupation factors 0.67 and 0.33 respectively). Constraints were applied to chemically equivalent bonds, and H atoms were introduced at theoretical positions. Compound 5. Owing to the low number of data versus the number of parameters, only the Co, Fe and Cl atoms were refined anisotropically. Furthermore, refinement with constraints 36 was applied to chemically equivalent bonds (17 equations); H atoms were introduced in the refinement at theoretical positions (C]H 1.00 Å) and assigned an isotropic thermal parameter equivalent to 1.1 times that of the bonded atom.Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre (CCDC). See Instructions for Authors, J. Ch, Dalton Trans., 1997, Issue 1. Any request to the CCDC for this material should quote the full literature citation and the reference number 186/277.The list of observed and calculated structure factors is available from the authors at the Institut de Chimie des Substances Naturelles. Mössbauer spectroscopy The Mössbauer spectra were recorded for solids using an instrument and techniques previously described.37 Results and Discussion Crystal structures for [Fe(oep){Fe(C5H5)(C4H4N)}2][O3SCF3] 1 and [Co(oep){Fe(C5H5)(C4H4N)}2][ClO4] 5 The molecular structures of these complexes are shown in Figs. 1 (1) and 2 (5) with selected bond lengths. Crystallographic data are reported in Table 1. In both structures the X-ray analysis has revealed an identical orientation of azaferrocene ligands (they are axially linked to the central metal ion in the porphyrinato core, tilted in a parallel configuration). The geometry of these compounds may be compared with that of [Fe(tfpp){Fe(C5H5)(C4H4N)}2] 4.28 For complex 4 the average length of the axial Fe]N bonds is 2.05(2) Å, which is similar to the distance 2.057(5) Å of 1.This latter distance is long in comparison to those of known structures of formula [Fe(por)L2]+ where L is an imidazole type ring [range 1.965(3)–2.015(4) Å].2,38–44 For the six-membered ring pyridine ligands 2,3,24,45 the Fe]Naxial range is 1.989(4)–2.031(2) Å, all but one significantly shorter than those found in compound 1. The larger value for Fe]Naxial distance for 1 may be a consequence of the steric hindrance the azaferrocene suffers when binding to iron porphyrins, though it also signifies a weaker Fe]N bond than in the pyridine complexes.The Fe]Npor bond distances of 1 are 2.004(4) and 2.023(4) Å, in agreement with those in other lowspin Fe(oep) structures.2,3,45 The two azaferrocene molecules related by a crystallographic centre of symmetry are obviously parallel. The dihedral angle between the pyrrolyl plane and the mean plane of the porphyrin core is 70.6(2)8 presumably as a result of the steric hindrance experienced by the C5H5 ring and the haem plane.This compares well to angles of 65 and 678 found in the two molecules of 4.28 The bond lengths and angles in the azaferrocene molecules are within acceptable agreement with those in 4 and the angle between the cyclopentadienyl and pyrrolyl ligand plane is 4.3(5)8 also similar to those reported (3.5 and 5.58) for 4.28 The orientation of the axial ligands with respect to the porphyrin core is usually characterised by the dihedral angle a between the axial ligand plane and the co-ordination plane [defined by a porphyrinato nitrogen atom N(1), the metal atom and the N atom of the axial ligand].In compounds 1, 4 and 5 the tilting of the axial ligand planes to the porphyrin plane means that reporting the angle in this way is not precise, though for 1 it is 22.9(4)8, similar to the value found in the iron(II) compound 4 (208). One can define the orientation of the azaferrocene moiety by the angle b between the same co-ordination plane and the plane Fe]Naxial]Fe(1); b = 72.3(3)8.This latter description of the orientation is much more satisfying for axial ligands where tilting is encountered. The structure of the cobalt(III) compound 5 may be com- Fig. 1 (a) Structure of complex 1. The Fe]Npor bond distances are 2.004(4) and 2.023(4) Å, the axial Fe]N bond distances 2.057(5) Å. The positions of the disordered ethyl chains are indicated by broken circles (see details of structure resolution in the text).(b) View showing the orientation of the two azaferrocene ligands on the porphyrin coreJ. Chem. Soc., Dalton Trans., 1997, Pages 47–53 49 Table 1 Crystal data and structure refinement parameters for complexes 1 and 5 [Fe(oep){Fe(C5H5)(C4H4N)}2][O3SCF3] [Co(oep){Fe(C5H5)(C4H4N)}2][ClO4] Empirical 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/mm21 F(000) Crystal size/mm q/8 hkl Ranges Collected reflections Observed reflections [I > 2s(I)] Least-squares refinement Data, parameters Goodness of fit Final R [I > 2s(I)] (all data) Weighted R Weighting scheme, w Largest difference peaks/e Å23 C55H62F3Fe3N6O3S 1115.77 Triclinic P1� 10.616(5) 11.193(5) 12.882(6) 95.28(4) 110.49(5) 109.40(5) 1314.3(11) 1 1.41 0.92 583 0.15 × 0.35 × 0.40 2–28 214 to 12, 214 to 14, 0–16 5668 4605 Full matrix on F2 5659, 323 0.763 0.094 0.111 0.284 [s2(Fo 2) + (0.3P)2 + 2.8P]21 where P = (Fo 2 + 2Fc 2)/3 1.9, 20.7 C54H62ClCoFe2N6O4 1069.24 Monoclinic P21/n 12.987(5) 33.065(13) 12.215(5) 110.21(5) 4922(7) 4 1.36 1.02 2240 0.05 × 0.10 × 0.15 2–25 215 to 14, 238 to 39, 0–12 5465 1026 Blocked matrix on F 1026, 328 1.10 0.056 0.057 [s2(Fo) + 0.05Fo 2]21 0.8, 20.5 pared with those of the low-spin iron(III) structures from the point of view of the charge on the central metal but also to the Fig. 2 (a) Structure of complex 5.The Co]Npor bond distances range from 1.96(2) to 2.00(2) Å, the axial Co]N bond lengths 1.94(2) and 2.01(2) Å. (b) View of the structure showing the orientation of the two azaferrocene ligands on the porphyrin core iron(II) structures 28,42 from the point of view of the 3d6 electron count on the central metal. The precision of this structure is not sufficient (the crystal was too small) to facilitate discussion on the bond lengths and angles. The two azaferrocene ligands which are not related by symmetry elements (in contrast to compounds 1 and 4) are in parallel configuration.The dihedral angles a between the axial ligands planes and co-ordinate plane Naxial]Co]N(4)por vary from 38(1) to 39(1)8. The rotational orientation of the azaferrocene moieties, which may be defined by the angle b between the co-ordinate plane Co, Naxial, N(4)por and the plane (Co, Naxial, Fe), is 56(1)8. The dihedral angles between the pyrrolyl ligand planes and the mean plane of the porphyrin core are 66(1) and 67(1)8 again similar to those in structures 1 and 4.28 So once again the steric interaction between the C5H5 ring and the porphyrin plane is manifest.The bond lengths and angles are in acceptable agreement with those found in 1 and 4 and also those reported previously in alkyl(h1-azaferrocene)cobaloximes.26,28 Mössbauer spectroscopy The data from the Mössbauer spectra of compounds 1–5 and those of azaferrocene and other relevant literature data appear in Table 2.Compound 4 shows a change of 0.09 mm s21 in the q.s. of co-ordinated azaferrocene compared to azaferrocene itself suggesting a small reduction in the electron density of the donation from the e1 orbital on the ring to the iron centre. This change in q.s. is small compared to changes we have previously found for substituted ferrocenes compared to ferrocene itself,46–48 however, protonation of azaferrocene only causes a change of 0.15 mm s21.25 Thus the first question is does the q.s.of azaferrocene tell us anything about the bonding properties to the FeII in the porphyrin centre? In principle the bonding of the azaferrocene to Fe(tfpp) could be caused in three ways: (a) due to the nitrogen lone-pair s bonding to the Fe of the porphyrin ring (donation of its electrons) and causing a slight polarisation of the ring e1 electrons toward the now electron-deficient N atom; (b) by a combination of (a) and pyrrolyl ring p bonding to the iron porphyrin (by p donation to the metal); (c) by a combination of (a) and pyrrolyl ring50 J.Chem. Soc., Dalton Trans., 1997, Pages 47–53 Table 2 Iron-57 Mössbauer data (mm s21) for compounds 1–5 and azaferrocene a Porphyrinic site Azaferrocene site Complex d D G d D G 1 2345 Azaferrocene b 0.30(6) 0.27(2) 0.44(1) 2.24(6) 1.92(4) 1.25(2) 0.29(2) (0.55)(6) 0.58(4) 0.13(1) 0.52(1) 0.54(1) 0.55(1) 0.53(1) 0.56(1) 0.54(1) 2.45(1) 2.42(1) 2.47(1) 2.42(1) 2.42(1) 2.51(1) 0.12(1) 0.14(4) 0.12(2) 0.11(1) 0.11(1) 0.14(1) a d = Isomer shift, D = quadrupole splitting, G = half-width at half-height.b Taken from ref. 25. p bonding to the iron porphyrin (by metal to ring back bonding). The reduction in the q.s. othe azaferrocene on bonding to Fe(tfpp) provides evidence for which of the above mechanisms is correct. Consider how the q.s. arises in azaferrocene: q.s. µ 2p2 2 p1, where p2 and p1 are the electron populations of the azaferrocene e2 and e1 orbitals.46 The e2 orbitals have twice the effect of the e1 orbitals on the q.s.47,48 Thus removal of electron density from the e1 orbitals [due to bonding to Fe- (tfpp)] would increase the q.s.of the azaferrocene if there was no concomitant back bonding from the iron in the azaferrocene to the pyrrolyl ring via the e2 orbitals. A decrease in q.s. as in this case is consistent with greater back bonding from the iron in the azaferrocene to the pyrrolyl ring showing that electron density has been removed from the pyrrolyl ring on bonding to Fe(tfpp).The q.s. decrease signifies that the e1 orbital is more electron deficient after the azaferrocene binds to the Fe(tfpp). Thus this bonding removes electron density from the ring e1 orbitals. The question is how? Mechanism (c) would involve increase in electron population of the ring e1 orbitals and thus less need for e2 back bonding and hence an increase in the q.s. and so can be ruled out as a possibility.However, both mechanisms (a) and (b) would fit the reduction seen in the q.s. of azaferrocene on bonding to Fe(tfpp). To gain further insight into which is correct we now turn to the Mössbauer data for the iron in the Fe(tfpp) site. The isomer shift and q.s. values for the iron(II) in the porphyrin plane are typical of low-spin iron(II) porphyrins. For such compounds where the axial ligand is imidazole the q.s. is found to be between 0.9 and 1.0 mm s21.49–51 For pyridine ligands higher q.s.values are observed between 1.11 and 1.20 mm s21 at 77 K; the value of 1.25(2) mm s21 for compound 4 is larger than the end of the latter range suggesting bonding similar to but weaker than that of pyridines.49,51–56 Indeed for [Fe(tpp)(py)2] (py = pyridine) the q.s. is 1.15 mm s21.56 This suggests that in complex 4 the axial Fe]N bonds are purely s with little or no p back bonding. The axial Fe]N bond distances support this interpretation.28 Thus mechanism (b) can be eliminated and (a) is verified.It should be noted that in compound 4 the iron t2g orbitals on the iron in the porphyrin are filled and thus further donation to them from the e1 orbitals is impossible again eliminating mechanism (b). Compound 1 (Table 2) is a low-spin iron(III) complex and the Mössbauer parameters for the iron(III) site are typical of a parallel configuration of the axial ligands (Table 2 also shows comparable compounds). Indeed this is found in the structure (see above).This is not surprising in view of our earlier report 28 of the EPR spectrum which is rhombic, however it has a large gz value which we will discuss in the next section more fully. The q.s. of the co-ordinated azaferrocene in compound 1 is only 0.06(1) mm s21 less than that of azaferrocene itself. Again a long axial bond in this structure suggests little back bonding (if any). Yet the dihedral angles of the ligand pyrrolyl planes with the plane defined by one Npor, the central iron and the azaferrocene N was 238, which although not ideal is close to values found by other workers where back bonding was said to occur.2,3,7 We note that in compound 4 similar angles were found28 and as shown (see above) little evidence of back bonding is apparent.In compound 3 the azaferrocene only shows a slight change in q.s., 0.04(1) mm s21 from the parent, and so again this may be explained by azaferrocene being a weak s donor.The Mössbauer parameters for the low-spin iron(III) site in compound 2 are close to others that have been interpreted as belonging to a site in which the axial ligands are nearly perpendicular. 2,4 We fitted the Mössbauer spectrum using a simple doublet for this site rather than two singlets. The fact that the spectrum seems to suggest a perpendicular arrangement of the axial ligands is worthy of note; first the three known structures (compounds 1, 2 and 5) all show parallel configurations, and secondly the EPR data recorded at 3.5 K were rhombic with gz = 3.24, which is consistent with a parallel configuration (but again high).We will return to this below. The azaferrocene q.s. value for both compounds 2 and 4 is 0.09 mm s21 smaller than that of azaferrocene itself. This value is in keeping with no back bonding from the iron(III), especially as there is no change on the azaferrocenyl iron when the oxidation state of the porphyrin iron changes.As both compounds 3 and 4 contain a 3d6 electron configuration at the porphyrinic metal centre then these two centres might have been expected to be good p donors to azaferrocene. No Mössbauer evidence was found in favour of this suggestion, and the fact that the planes of the pyrrolyl rings were found to make an angle of around 708 to the porphyrin plane would also discourage p back bonding. Further reinforcement of this argument comes from the q.s. for compound 5; here again a lowering of 0.09(1) mm s21 relative to azaferrocene is observed, again a 3d6 electron configuration is present and a similar structural pattern occurs and thus no evidence for back bonding is found in the Mössbauer data.Compounds 1 and 2 contain 3d5 electronic configurations and have a hole in the t2g orbitals into which pyrrolyl p electrons could donate; again no evidence is found for this in the Mössbauer parameters or in the structure of 1. Electron paramagnetic resonance The EPR spectra of complexes 1 and 2 have been previously reported.27 The data appear in Table 3 along with relevant literature data.57–62 The spectra of 1 and 2 are of rhombic type but with relatively high gz values approaching those found in HALS spectra. Previously rhombic type spectra of iron porphyrins have only been found for strongly co-ordinating unhindered imidazoles and highly basic pyridines.It is therefore somewhat surprising that compounds 1 and 2 manifest such spectra, although the crystal structure of 1 shows that the pyrrolyl planes are parallel. The Mössbauer spectrum of 1 is also in keeping with such an axial arrangement of the ligands, whereas that of 2 does not fit such an arrangement very well.To explain these observations itJ. Chem. Soc., Dalton Trans., 1997, Pages 47–53 51 Table 3 The EPR data of bis-ligated imidazole, azaferrocene and pyridine porphyrinatoiron(III) complexes in solution at 77 K Complex Solvent gz gy gx Ref.Imidazole derivatives [KL2][Fe(tpp)(4-mim)2] [FeL1(1-mim)2(OH)] [FeL1(Him)(Im)] [Fe(tpp)(4-Hmim)2]I [Fe(tpp)(bim)2]I [Fe(tmp)(1-mim)2]ClO4 [Fe(tpp)(1-mim)2]I [Fe(tpp)(1-mim)2]ClO4 [FeL1(1-mim)2]+ [FeL1(Him)2]+ [Fe(tpp)(dmim)2]I [Fe(tpp)(2-Hmim)2]I [Fe(tpp)(2-Hmim)2]Cl [Fe(tpp)(Hdmbim)2]I [FeL1(2-Hmim)2]+ CH2Cl2 Water–ethanol (1 : 1) Me2SO + OH CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 Me2SO Me2SO CH2Cl2 CH2Cl2 HCONMe2 CH2Cl2 HCONMe2 2.60 2.74 2.76 2.847 2.860 2.886 2.886 2.890 2.97 3.02 3.40 3.40 3.41 3.43 3.48 2.24 2.27 2.28 2.288 2.306 2.325 2.294 2.291 2.27 2.24 2.36 1.82 1.72 1.68 1.590 1.561 1.571 1.549 1.554 1.51 1.51 1.05 57 11 58, 59 60 60 2 60 61 16 59 60 60 15 60 62 Azaferrocene derivatives [Fe(tpp){Fe(C5H5)(C4H4N)}2]+ [Fe(oep){Fe(C5H5)(C4H4N)}2]+ CHCl3 CHCl3 3.16 3.24 2.20 2.05 1.45 1.31 27 27 Pyridine derivatives [Fe(tpp)(dmadmpy)2]I [Fe(tpp)(dmapy)2]I [Fe(oep)(dmpapy)2]ClO4 [Fe(tpp)(apy)2]I [Fe(tpp)(dapy)2]I [Fe(tpp)(dmampy)2]I [Fe(tdcpp)(apy)2]ClO4 [Fe(tmp)(dmapy)2]ClO4 [Fe(tpp)(dmpy)2]ClO4 [Fe(tdcpp)(dmapy)2]ClO4 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 2.785 2.786 2.828 2.830 2.864 2.865 3.24 3.33 3.40 3.54 2.281 2.284 2.278 2.289 2.280 2.286 1.675 1.657 1.642 1.603 1.597 1.591 60 60 2 60 60 60 22 60 2 L2 = 4,7,13,16,21,24-Hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane; 4-mim = 4-methylimidazolate; 1-mim = 1-methylimidazole; Him = imidazole; bim = 1-benzylimidazole; H2tmp = 5,10,15,20-tetramesitylporphyrin; dmim = 1,2-dimethylimidazole; 2Hmim = 2-methylimidazole; Hdmbim = 5,6- dimethylbenzylimidazole; dmadmpy = 4-dimethylamino-3,5-dimethylpyridine; dmapy = 4-dimethylaminopyridine; apy = 4-aminopyridine; dapy = 3,4-diaminopyridine; dmampy = 4-dimethylamino-3-methylpyridine; H2tdcpp = 5,10,15,20-tetrakis(2,6-dichlorophenyl)porphyrin; dmpy = 3,4- dimethylpyridine.is interesting to plot the known EPR spectra in such a way that the gz values for the series are linear (Fig. 3).57–62 Such a plot is known as a Symons plot.63 When this is done it can be appreciated how close to the HALS positions compounds 1 and 2 are.To consider the orbital arrangements for HALS and rhombic splittings three possible cases must be considered. The first (a) is when V @ l and D @ l (V is the rhombic distortion parameter, D is the tetragonal distortion parameter and l is the spin–orbit coupling constant) the ground state is an orbital singlet (2B) and the unpaired electron is localised in dyz (see Fig. 2, ref. 4). This then generates an unequal distribution of charge density in the x and y directions and gives rise to a large crystal field. This is the situation that arises when the two Fig. 3 A Symons diagram showing the range of g values for various low-spin iron(III) porphyrin complexes based on literature data for solid-state imidazole pyridines and our results for the azaferrocene derivatives given in Table 3.The g1 values were placed on a 458 line for display purposes only; the vertical axis has no specific significance axial ligands are parallel and interact in some way with the iron dyz orbital making it the highest-energy orbital (as shown by analysis of EPR data).62,64 Scheidt and co-workers 1,65,66 have shown that such a parallel arrangement of axial ligand planes is thermodynamically the more stable form. Such compounds display rhombic EPR spectra.2,11,16,59–61 The second case (b) is when V < l and D @ l the unpaired electron is delocalised over the dxz and dyz orbital giving rise to an orbital-doublet (2E) ground state for the complex.This situation is brought about by an effective electronic axial symmetry; the crystal field is obviously smaller than that of case (a). This case arises when one of the axial ligands interacts with the iron dxy orbital and the other with the iron dxz orbital. For greatest interaction the ligands should be orientated perpendicular to each other.An example of this is in the structure and EPR spectrum of [Fe(tpp)(py)2]+ where two pyridine molecules adopt a perpendicular geometry and the dxz and dyz orbitals are nearly degenerate.24 Thus this situation gives rise to highly anisotropic low-spin (HALS) spectra.2,15,60 The third case (c) arises when V < l and D < l and the energy states are close together. Electrostatic interactions or spin–orbit coupling may mix the 2E and 2B states or an interaction may arise where there is only a slight preference to split the energies of the dxz and dyz orbitals. It is noted that very small q.s.values for low-spin iron(III) porphyrin complexes have been observed, for example 0.53 mm s21 at 80 K in Na[FeL1(CN)2] (H2L1 = protoporphyrin IX = 3,7,12,17-tetramethyl-8,13-divinylporphyrin-2,18-dipropanoic acid) 67 and in [Fe(tpp)(cpy)2]ClO4 (cpy = 4-cyanopyridine); 68 in these compounds the ground state has been explained as either ‘2B2g’ or as a mixed ‘2B2 and 2Eg’ state.68 As there should be little or no p–p interactions between the52 J.Chem. Soc., Dalton Trans., 1997, Pages 47–53 porphyrin iron and the pyrrolyl ring of the azaferrocene ligand (as seen from the Mössbauer data discussed above), we had therefore expected near degeneracy in the dxy and dyz orbitals thus giving gz values of ª3.5 or higher, despite the fact that in compound 1 the pyrrolyl rings are parallel. Thus we would have expected a HALS spectrum.Significantly the a angle of 22.9(4) is not large and by comparison to the work of Quinn et al.44 we would not expect a large gz value (smaller or equal to 2.9 which is contrary to what we observe). This relatively large gz value is therefore more surprising. The long porphyrin iron to axial pyrrolyl nitrogen bond length of 2.06(1) Å (which suggests little or no p bonding) must be a factor in generating the large gz value. The steric interaction of the H atoms on the pyrrolyl rings with the dxz or dyz orbitals will be the main reason for splitting their degeneracy to produce a rhombic-type spectrum. Clearly here we demonstrate that a p interaction is not needed to produce this splitting, a steric effect is sufficient.Such a long axial bond was also found in [Fe(tfpp){Fe(C5H5)- (C4H4N)}2] 28 and a similar a angle for the eclipsed pyrrolyl rings [in the iron(II) compound no ligand p donation could occur as the metal t2g orbitals are filled].The long axial bonds therefore are the result of weak binding and the steric interactions of the H atoms with the dxz or dyz orbital containing the single electron, and also to the steric interaction of the other end of the azaferrocene molecule with the porphyrin p electrons trying to attain a minimum-energy position. Thus our finding is that the axial bond strength will affect the magnitude of gz. The fact that for compound 1 the q.s. is 2.24(6) mm s21 yet the pyrrolyl rings are totally parallel means that for long axial bonds of this type this q.s.is a maximum, 0.16 mm s21 lower than that found for imidazole ligands.4 Thus a similar lower value for a perpendicular arrangement where the dxz and dyz orbitals are degenerate might also be expected. This means that the q.s. for compound 2 may in fact also represent a parallel arrangement with the q.s. value for perpendicular being below 1.8 mm s21. Indeed for 2 the gz is 3.16, which though still higher than 2.9 is lower than that for 1 as is the q.s.Conclusion We have found for compounds 1 and 2 that rhombic spectra with unusually large g1 values can arise from cases where the axial ligands are parallel, but are sterically hindered. In these cases p bonding from the metal to the ligand is non-existent or very weak. Under such circumstances the Mössbauer q.s. values are consistent with a parallel arrangement of the azaferrocene pyrrolyl planes.Recently Nakamura et al.69 have reported sterically hindered porphyrin complexes which although manifesting perpendicular arrangements of the ligands gave gz values that are more typical of rhombic spectra. It thus appears from this work and that of Nakamura et al.69 that what is important is not the absolute value of gz but the overall shape of the EPR spectrum. Thus there is a region of overlap in gz between values of 3.0 and 3.3 where both rhombic and HALS spectra can occur depending on the nature of the axial ligands and/or steric effects on the porphyrin.We note that these azaferrocene-ligated iron(III) porphyrins give EPR spectra quite different to those we reported for aliphatic ligands.70 Acknowledgements We are grateful to the NATO Scientific Affairs Division for financial support (Grant HTECH CRG 931301) and to the British–Polish joint research collaboration programme [Grant KBN (WAR/992/046)]. References 1 F. A. Walker, B. H. Hyunh, W. R.Scheidt and S. R. Osvath, J. Am. Chem. Soc., 1986, 108, 5288. 2 M. K. Safo, G. P. Gupta, F. A. Walker and W. R. Scheidt, J. Am. Chem. Soc., 1991, 113, 5497. 3 M. K. Safo, F. A. Walker, A. M. Raitsimring, W. P. Walters, D. P. Dolata, P. G. Debrunner and W. R. Scheidt, J. Am. Chem. Soc., 1994, 116, 7760. 4 O. K. Mehdi and J. Silver, J. Chem. Soc., Dalton Trans., 1990, 263. 5 O. K. Mehdi and J. Silver, J. Chem. Soc., Dalton Trans., 1990, 555. 6 H. Bang, J. O. Edwards, J. Kim, R.G. Lawler, K. Reynolds, W. R. Ryan and D. A. Sweigart, J. Am. Chem. Soc., 1992, 114, 2843. 7 M. K. Safo, G. P. Gupta, C. T. Watson, U. Simonis, F. A. Walker and W. R. Scheidt, J. Am. Chem. Soc., 1992, 114, 7066. 8 D. Dolphin (Editor), The Porphyrins, Academic Press, New York, 1979, vols. 4 and 7. 9 A. B. P. Lever and H. B. Gray (Editors), Iron Porphyrins, Part 2, Addison-Wesley, Reading, MA, 1983, parts 1 and 2. 10 G. N. La Mar, P. D. Burns, T. J. Jackson, K. M. Smith, K.C. Langry and P. Strittmatter, J. Biol. Chem., 1981, 256, 6075. 11 G. T. Babcock, W. R. Widger, W. A. Cramer, W. A. Oertling and J. Metz, Biochemistry, 1985, 24, 3638. 12 M. Pierrot, R. Haser, M. Frey, F. Payan and J. P. Astier, J. Biol. Chem., 1982, 257, 14341. 13 J. C. Salerno, J. Biol. Chem., 1984, 249, 2331. 14 A.-H. Tsai and G. Palmer, Biochim. Biophys. Acta, 1983, 722, 349. 15 B. Guiard and F. Lederer, J. Mol. Biol., 1979, 135, 639. 16 K. R. Carter, A. Tsai and G. Palmer, FEBS Lett., 1981, 132, 243. 17 D. Brautigan, B. A. Feinberg, B. M. Hoffman, E. Margoliash, J. Peisach and W. E. Blumberg, J. Biol. Chem., 1977, 252, 574. 18 J. S. de Ropp, V. Thanbal and G. N. La Mar, J. Am. Chem. Soc., 1985, 107, 8268. 19 M. F. Perutz and L. F. Ten Eyck, Cold Spring Harbor Symp. Quant. Biol., 1971, 36, 295. 20 R. F. Pasternack, B. S. Gilles and J. R. Stahlbush, J. Am. Chem. Soc., 1978, 100, 2613. 21 B. B. Hasinoff, H. B. Dunford and D. G. Horne, Can. J. Chem., 1969, 47, 3225. 22 M. Quing-jin, C. A. Tondreau, J. O. Edwards and D. A. Sweigart, J. Chem. Soc., Dalton Trans., 1985, 2269. 23 T. Yoshimura and T. Ozaki, Arch. Biochem. Biophys., 1984, 230, 466. 24 D. Inniss, S. M. Soltis and C. E. Strouse, J. Am. Chem. Soc., 1988, 110, 5644. 25 A. Houlton, R. M. G. Roberts, J. Silver and J. Zakrzewski, J. Organomet. Chem., 1993, 456, 107. 26 M. Cesario, C. Giannotti, J. Guilhem and J. Zakrzewski, Acta Crystallogr., Sect. C, 1992, 48, 798. 27 J. Zakrzewski and C.Giannotti, J. Chem. Soc., Dalton Trans., 1993, 1629. 28 J. Zakrzewski, M. Cesario, J. Guilhem and C. Giannotti, J. Chem. Soc., Dalton Trans., 1992, 3059. 29 J. Zakrzewski and C. Giannotti, Coord. Chem. Rev., 1995, 140, 169. 30 J. Zakrzewski and C. Giannotti, J. Organomet. Chem., 1990, 388, 175. 31 H. Sugimoto, N. Ueda and M. Mori, Bull. Chem. Soc. Jpn., 1981, 54, 3425. 32 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158. 33 G. M. Sheldrick, SHELXS 86, Program for the solution of crystal structures, University of Göttingen, 1986. 34 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993. 35 G. M. Sheldrick, SHELX 76, Program for crystal structure determination, University of Cambridge, 1976. 36 J. Waser, Acta Crystallogr., 1963, 16, 1091. 37 M. Y. Hamed, R. C. Hider and J. Silver, Inorg. Chim. Acta, 1982, 66, 13. 38 W. R. Scheidt, J. L. Kirner, J. L. Hoard and C. A. Reed, J. Am. Chem. Soc., 1987, 109, 1963. 39 D. M. Collins, R.Countryman and J. L. Hoard, J. Am. Chem. Soc., 1972, 94, 2066. 40 K. Hatano, M. K. Safo, F. A. Walker and W. R. Scheidt, Inorg. Chem., 1991, 30, 1643. 41 T. Mashiko, J.-C. Marchon, D. T. Musser, C. A. Reed, M. E. Kastner and W. R. Scheidt, J. Am. Chem. Soc., 1979, 101, 3653. 42 R. G. Little, K. R. Dymock and J. A. Ibers, J. Am. Chem. Soc., 1975, 97, 4532. 43 W. R. Scheidt, R. S. Osvath and Y. J. Lee, J. Am. Chem. Soc., 1987, 109, 1958. 44 R. Quinn, J. S. Valentine, M.P. Byrn and C. E. Strouse, J. Am. Chem. Soc., 1987, 109, 3301. 45 W. R. Scheidt, D. K. Geiger and K. J. Haller, J. Am. Chem. Soc., 1982, 104, 495. 46 A. Houlton, J. R. Miller, R. M. G. Roberts and J. Silver, J. Chem. Soc., Dalton Trans., 1990, 2181. 47 A. Houlton, J. R. Miller, R. M. G. Roberts and J. Silver, J. Chem. Soc., Dalton Trans., 1991, 467.J. Chem. Soc., Dalton Trans., 1997, Pages 47–53 53 48 R. A. Brown, A. Houlton, S. D. Howe, R. M. G Roberts and J. Silver, J. Chem. Soc., Dalton Trans., 1993, 3329. 49 L. M. Epstein, D. K. Straub and C. Maricondi, Inorg. Chem., 1967, 6, 1720. 50 J. P. Collman, J. L. Hoard, N. Kim, G. Lang and C. A. Reed, J. Am. Chem. Soc., 1975, 97, 2676. 51 G. Al-Jaff, J. Silver and M. T. Wilson, Inorg. Chim. Acta, 1990, 176, 307. 52 A. J. Bearden, T. H. Moss, W. S. Caughey and C. A. Beaudreau , Proc. Natl. Acad. Sci. USA, 1965, 53, 1246. 53 T. H. Moss, A. J. Bearden and W. S. Caughey, J. Chem. Phys., 1969, 51, 2624. 54 D. Dolphin, J. R. Sams, T. B. Tsin and K. L. Wong, J. Am. Chem. Soc., 1976, 98, 6970. 55 H. Kobayashi, Y. Maeda and Y. Yanagawa, Bull. Chem. Soc. Jpn., 1970, 43, 2342. 56 D. K. Straub and W. M. Connor, Ann. N. Y. Acad. Sci., 1973, 206, 383. 57 R. Quinn, C. E. Strouse and J. S. Valentine, Inorg. Chem., 1983, 22, 3934. 58 T. Yoshimura and T. Ozaki, Arch. Biochem. Biophys., 1984, 230, 466. 59 J. Peisach and W. B. Mims, Biochemistry, 1977, 16, 2765. 60 F. A. Walker, D. Reis and V. L. Balke, J. Am. Chem. Soc., 1984, 106, 6888. 61 T. Higgins, M. K. Safo and W. R. Scheidt, Inorg. Chim. Acta, 1991, 178, 261. 62 C. T. Migita and M. I. Iwazumi, J. Am. Chem. Soc., 1981, 103, 4378. 63 M. C. R. Symons and R. L. Petersen, Proc. R. Soc. London, Ser. B, 1978, 201, 285. 64 G. Palmer, in The Porphyrins, ed. D. Dolphin, Academic Press, New York, 1979, vol. 4, ch. 6, pp. 313–353; in Iron Porphyrins, eds. A. B. P. Lever and H. B. Gray, Addison-Wesley, Reading, MA, 1983, part 2, ch. 2, pp. 43–88. 65 W. R. Scheidt and D. M. Chapman, J. Am. Chem. Soc., 1988, 110, 5644. 66 W. R. Scheidt and Y. S. Lee, Struct. Bonding (Berlin), 1987, 64, 1. 67 B. Lucas and J. Silver, Inorg. Chim. Acta, 1986, 124, 97. 68 W. R. Scheidt, S. R. Osvath, Y. J. Lee, C. A. Reed, B. Schaevitz and G. P. Gupta, Inorg. Chem., 1989, 28, 1591. 69 M. Nakamura, K. Tajima, K. Tada, K. Ishizu and N. Nakamura, Inorg. Chim. Acta, 1994, 224, 113. 70 P. J. Marsh, J. Silver, M. C. R. Symons and F. A. Taiwo, J. Chem. Soc., Dalton Trans., 1996, 2361. Received 4th April 1996; Paper 6/02366E

ISSN:1477-9226    

DOI:10.1039/a602366e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 51–55 51 Revisiting the synthesis of [Mo6(Á5-C5Me5)O18]2. X-Ray structural analysis, UV-visible, electrochemical and multinuclear NMR characterization Anna Proust,* René Thouvenot and Patrick Herson Laboratoire de Chimie Inorganique et Matériaux Moléculaires, ESA 7071, Université Pierre et Marie Curie, 4 Place Jussieu, Case 42, 75252 Paris Cedex 05, France. E-Mail: proust@ccr.jussieu.fr Received 27th July 1998, Accepted 2nd November 1998 In methanol, [n-Bu4N][MoCp*O3] reacted with [n-Bu4N]2[Mo4O10(OMe)4Cl2] to yield [n-Bu4N][Mo6Cp*O18] in reasonable yield, thus opening a new route in the synthesis of organometallic derivatives of polyoxometalates. The crystal and molecular structure of [n-Bu4N][Mo6Cp*O18]?Me2CO has been determined by single-crystal X-ray diVraction.The ion [Mo6Cp*O18]2 has been thoroughly characterized by multinuclear NMR in solution. The 95Mo and 17O NMR data support the conclusion that the Cp* ligand is a better s 1 p donor than the oxo ligand.The complex is electrochemically active and displays three successive one-electron reduction processes. As it provides an understanding of oxide surface reactions at the molecular level, the chemistry of organometallic derivatives of soluble oxide analogues is rapidly expanding. A significant class of compounds comprises organometallic derivatives of polyoxometalates 1 with major contributions from Klemperer, Finke and Isobe’s groups.Oligomerization of oxometalates in the presence of [(RhCp*Cl2)2] allowed Isobe 2 to obtain molecular cubane-type clusters, which mimic the structure as well as the reactivity of bulk oxides. In the case of [(RhCp*)4V6O19] the Cp*Rh21 fragments stabilize the otherwise unstable {V6O19}82 core.3,4 On the other hand, the polyanion-supported organometallic complexes described by Day and Klemperer 5 and Finke and co-workers 6–11 are rather relevant to the modeling of organometallic catalyst oxide–support interactions.Determination of their molecular and electronic structures either by single-crystal X-ray diVraction or by spectroscopic methods aVords accurate structural and spectroscopic models for interfacial organometallic chemistry. The relative basicities of the oxo ligands, an essential factor in oxide surface reactivity, can also be assessed from studies of the fixation of organometallic units onto the polyoxometallic framework.Most interestingly, the Dawson-type [P2W15Nb3O62]92 polyoxoanion-supported Re(CO)3 1 and Ir(CO)2 1 complexes provide models for the mobility of M(CO)n 1 cations on an oxide surface.12 On the other hand, [(1,5-COD)IrP2W15Nb3O62]82 displays its own reactivity either as a catalyst 13 or as a precursor for polyoxoanion- stabilized Irª300 nanoclusters.14 Moreover, organometallic derivatives of polyoxoanions could also display synergistic eVects or even bifunctional activity.15 In the course of their extensive investigation of cyclopentadienyl –oxo complexes,16 Bottomley and co-workers described the crystal structure of [C5Me5O][Mo6Cp*O18], obtained by air oxidation of [{MoCp*(CO)2}2] in CHCl3.17 This prompted us to benefit from our experience in the functionalization of polyoxometalates 18–23 for developing a more straightforward synthesis for [Mo6Cp*O18]2 and similar complexes.This paper addresses an alternative preparation of [Mo6Cp*O18]2 by selfassembly of [MoCp*O3]2 and oxomolybdenum fragments generated from the [Mo4O10(OMe)4Cl2]22 anion.The singlecrystal structure analysis of [n-Bu4N][Mo6Cp*O18]?Me2CO and the characterization of [Mo6Cp*O18]2 by multinuclear NMR (1H, 13C, 17O and 95Mo), UV-visible spectroscopy and electrochemistry are reported. Experimental Dimethylhydroxylamine hydrochloride, tetrabutylammonium hydroxide 1 M in methanol and HPLC grade acetonitrile were purchased from Aldrich and used as received.Reagent grade methanol was distilled over magnesium methoxide. The salt [n-Bu4N]4[a-Mo8O26] was synthesized according to the published procedure and recrystallized in acetonitrile,24 [n-Bu4N]- [MoCp*O3] by addition of 1 equivalent of n-Bu4NOH to a solution of [t-BuNH3][MoCp*O3] in methylene chloride, prepared as described.25 The salt [n-Bu4N][BF4] was prepared from [n-Bu4N][HSO4] and NaBF4 and dried overnight at 80 8C under vacuum. Infrared spectra were recorded from KBr pellets on a Bio- Rad FT 165 spectrophotometer, electronic absorption spectra on a Shimadzu model UV-2101 spectrophotometer. Elemental analyses were performed at Pierre et Marie Curie University or at the Service Central d’Analyse of the Centre National de la Recherche Scientifique (Vernaison, France).Proton and 13C NMR spectra were recorded in 5 mm o.d. tubes at room temperature on a Bruker AC300 spectrometer, natural abundance 17O and 95Mo NMR spectra from CH3CN–CD3CN (2 cm3–0.2 cm3) solutions in conventional 10 mm o.d.tubes at 343 K on a Bruker AM500 spectrometer. Preparation of [n-Bu4N]2[Mo4O10(OMe)4Cl2] To a suspension of [n-Bu4N]4[a-Mo8O26] (2.15 g, 1 mmol) in methanol (15 cm3) were added 1.91 g (20 mmol) of Me2NOH? HCl. The mixture was heated to reflux until a clear green solution was obtained, which was then allowed to cool slowly to room temperature. Within a few hours, colourless crystals of [n-Bu4N]2[Mo4O10(OMe)4Cl2] deposited (1.35 g, 55%); n& max/ cm21 2962s, 2932m, 2876m, 2822w, 1474m, 1382w, 1025m, 993m, 941s, 917vs, 900s, 738w, 705s, 598w, 548m, 380w, 344w and 311m (Found: C, 35.54; H, 7.02; N, 2.25.C18H42ClMo2NO7 requires C, 35.34; H, 6.92; N, 2.29%). Preparation of [n-Bu4N][Mo6Cp*O18] In a typical experiment, a solution of 0.39 g (0.75 mmol) of [n-Bu4N][MoCp*O3] in 25 cm3 of methanol was added to a suspension of 0.979 g (0.8 mmol) of [n-Bu4N]2[Mo4O10(OMe)4- Cl2] in 25 cm3 of methanol. After a 1.5 h reflux, the solution was52 J.Chem. Soc., Dalton Trans., 1999, 51–55 concentrated to about 40 cm3 and set aside at room temperature. After a few days, 0.2 g of [n-Bu4N][Mo6Cp*O18] was collected as orange-red crystals, which proved too small for an X-ray analysis. A further crop of compound was isolated from the filtrate later on (whole yield 40%, based on [n-Bu4N]- [MoCp*O3]); n& max/cm21 2963m, 2929m, 2874w, 1472w, 1442w, 1372w, 982m, 955s, 798s and 778s (Found: C, 25.66; H, 4.22; Mo, 44.32; N, 1.08.C26H51Mo6NO18 requires C, 25.16; H, 4.14; Mo, 46.37; N, 1.12%); lmax/nm (log e/dm3 mol21 cm21) (CH3CN) 258 (4.5), 283 (sh) (4.4) and 494 (sh) (2.35); dH(300.13 MHz; solvent acetone-d6; standard SiMe4) 2.23 [15 H, s, C5(CH3)5], 3.33 [8 H, m, N(CH2CH2CH2CH3)4], 1.65 [8 H, m, N(CH2- CH2CH2CH3)4], 1.43 [8 H, m, N(CH2CH2CH2CH3)4] and 0.99 [12 H, t, N(CH2CH2CH2CH3)4]; dC(75.47 MHz; solvent CD3- CN; standard SiMe4) 140.0 [5 C, C5(CH3)5] and 12.6 [5 C, C5- (CH3)5], 59.9 [4 C, N(CH2CH2CH2CH3)4], 24.9 [4 C, N(CH2- CH2CH2CH3)4], 20.9 [4 C, N(CH2CH2CH2CH3)4] and 14.4 [4 C, N(CH2CH2CH2CH3)4]; d95Mo(32.6 MHz; solvent CH3CN–CD3- CN, 343 K; standard aqueous alkaline Na2MoO4) 175 (1 Mo, Moaxial), 164 (4 Mo, Moequatorial) and 235 (1 Mo, Cp*Mo); d17O(67.8 MHz; solvent CH3CN–CD3CN, 343 K; standard distilled water) 931 (4 O, Oterminal), 890 (1 O, Oterminal), 613 (4 O, m-O), 580 (4 O, m-O), 559 (4 O, m-O) and 22.7 (1 O, m6-O).Suitable crystals for an X-ray diVraction study were grown from an acetone solution of the crude compound.Their IR spectrum was very similar to that of the crystals obtained in methanol except for two additional bands at 1708 and 1224 cm21, attributed to acetone molecules. Cell parameters have also been determined for crystals obtained from a solution of the compound in acetonitrile : tetragonal, a = b = 18.349(8), c = 51.02(3) Å, U = 17179(23) Å3. Electrochemistry All measurements were carried out in acetonitrile, under nitrogen, at room temperature, by using a standard three-electrode cell, which consisted of the working electrode, an auxiliary platinum electrode, and an aqueous saturated calomel electrode (SCE) equipped with a double junction.Solution concentrations were ca. 1 mM for the compound and 0.1 M for the supporting electrolyte, [n-Bu4N][BF4]. Polarograms were recorded at a dropping mercury electrode on a Tacussel PRG3 device, at the rate of 0.15 V min21.Cyclic voltammograms were recorded at a carbon electrode on a PAR 273 instrument, at the rate of 0.1 V s21. Crystal structure determination for [n-Bu4N][Mo6Cp*O18]? Me2CO Crystal data. C29H57Mo6NO19, M = 1299.4, orthorhombic, space group P212121 (no. 19), a = 12.406(16), b = 16.065(9), c = 21.851(8) Å, U = 4355 (9) Å3, T = 298 K, Z = 4, m(Mo-Ka) = 1.7 mm21, Dc = 1.982 g cm23. The intensity data were collected at room temperature on a CAD4 Enraf-Nonius diVractometer, using graphite-monochromated Mo-Ka radiation.Two reference reflections were periodically monitored for intensity and orientation control. An overall decay of 27% was observed, despite the use of a crystal coated with oil in a Lindeman tube. A total of 4272 reflections were measured (2 < 2q < 508) using the 2q–w method. Intensities were corrected for Lorentz-polarization eVects and a absorption correction was carried out using DIFABS.26 The structure was solved by direct methods,27 neutral-atom scattering factors were used, and anomalous dispersion corrections were included.28 Hydrogen atoms were not included in the model.Distances and angles within the tetrabutylammonium cation were constrained to 1.55 ± 0.05 Å and 109 ± 18, respectively. All atoms were given anisotropic thermal parameters. Refinements were carried out by full matrix least-square procedures. The refinement of the 497 parameters converged at R1 = 0.054 and wR(F2) = 0.060 (w = 1) for 2794 reflections [I > 3s(I)]. Largest peak and hole in the final diVerence map were 11.36, 21.65 e Å23.All computations were performed using the CRYSTALS version for PC.29 Selected bond lengths and angles are listed in Table 1. A CAMERON view30 of the anion is depicted on Fig. 1. CCDC reference number 186/1232. See http://www.rsc.org/suppdata/dt/1999/51/ for crystallographic files in .cif format. Results and discussion Syntheses The synthesis and the X-ray characterization of [C5Me5O]- [Mo6Cp*O18] were reported by Bottomley and Chen 17 in 1992.This compound formed from a mixture of [(MoCp*O2)2(m-O)] and [MoCp*ClO2] resulting from air oxidation of [{MoCp*- (CO)2}2] in CHCl3. Harper and Rheingold 31 had previously described the sealed-tube preparation of [W6Cp*2O17] by reaction of [{WCp*(CO)2}2] with methylarsaoxanes (MeAsO)n. Both [Mo6Cp*O18]2 and [W6Cp*2O17] molecular structures derive from that of the Lindqvist-type polyoxoanion [M6- O19]2232 by formal replacement of one or two oxo ligands by Cp* ligand(s) (see below).As we are currently involved in a systematic study of functionalized polyoxoanions,18–23 we decided to explore the condensation reactions of the [MCp*- O3]2 (M = Mo or W) anions,25 both to improve the syntheses of [Mo6Cp*O18]2 and [W6Cp*2O17] and eventually to characterize novel cyclopentadienyl–oxo polyanions. Indeed it was anticipated that acid-driven condensation of [MCp*O3]2 could aVord Cp*–oxo polyanions, just as that of the parent [MO4]22 anions results in the formation of polyoxoanions.In parallel to acidification reactions, which are under investigation, we looked at the self-assembly reactions of polyoxometalate precursors and [MoCp*O3]2. Up to now, the only characterized product of the reaction between [MoCp*O3]2 anion and [MoO2(acac)2], as a potential source of MoO2 21, is [Mo6O19]22. On the other hand, [n-Bu4N]2[Mo4O10(OMe)4Cl2], which is known to transform into [n-Bu4N]2[Mo6O19] in refluxing methanol, reacts with an equivalent of [n-Bu4N][MoCp*O3] in methanol to aVord red-orange crystals of [n-Bu4N][Mo6- Cp*O18] in reasonable yield.The synthesis of [n-Bu4N]2- [Mo4O10(OMe)4Cl2] was first described by Zubieta and coworkers, 33 and was subsequently improved in our group by treating Me2NOH?HCl with [n-Bu4N]4[a-Mo8O26] in methanol. Although the exact role of the dimethylhydroxylamine is not clear, similar attempts with hydrochloric acid in place of Me2NOH?HCl have failed up to now.The [Mo4O10(OMe)4- Cl2]22 anion consists of a compact arrangement of edgesharing octahedra, with two doubly and two triply bridging methoxo ligands. The chloro ligands occupy terminal sites. The IR spectrum of [n-Bu4N][Mo6Cp*O18] in the 1000–250 cm21 range is typical of a Lindqvist-type species and is in accordance with that previously reported by Bottomley and Chen17 for [C5Me5O][Mo6Cp*O18]. The IR spectrum of [n-Bu4N]2[Mo6O19] displays two main bands at 954 and 800 cm21, attributed to the nasym(Mo]] Ot) (F1u) and nasym(Mo–O– Mo) vibrational modes respectively.34 A splitting of these two characteristic bands is actually observed in the spectrum of [n-Bu4N][Mo6Cp*O18] which shows peaks at 982, 955, 798 and 778 cm21.This is typical of monofunctionalized Lindqvist-type hexamolybdates and reflects the anion symmetry lowering from Oh to C4v on substitution.21,22,35,36 The electronic spectrum of [n-Bu4N][Mo6Cp*O18] was recorded in acetonitrile solution.A broad ligand-to-metal charge-transfer absorption is observed at 258 nm (log e = 4.5) with two shoulders at 283 (log e = 4.4) and 494 nm (log e = 2.35). By comparison, the spectrum of [n-Bu4N]2[Mo6O19] displays a peak at 260 (log e = 4.2) and a shoulder at 325 nm (log e = 4.0). These features account for the change from yellow [n-Bu4N]2[Mo6O19] to orange-red [n-Bu4N][Mo6Cp*O18]. If theJ. Chem. Soc., Dalton Trans., 1999, 51–55 53 absorption at 494 nm is assumed to involve charge transfer from the Cp* ligand, then this supports the conclusion that the latter is a stronger donor than the oxo ligand (see below).This is in keeping with prevailing views that the Cp ligand is a stronger donor than the t-BuN ligand,37 which is itself a stronger donor than the oxo ligand.38 Structure analysis The molecular structure of the anion of [n-Bu4N][Mo6Cp*- O18]?Me2CO (Fig. 1) is closely related to that of [Mo6O19]2239 by formal replacement of a terminal oxo ligand by a h5-Cp* ligand, both ligands being able to form s,2p triple bonds.The distance of Mo(1) to the mean plane through the Cp* ligand is 2.083 Å, while the average of the Mo(1)–C(i) (i = 1–5) distances is 2.40 Å. Previous structural studies of substituted Lindqvist-type hexamolybdates including the C5Me5O1 salt of [Mo6Cp*O18]2 reported by Bottomley and Chen,17 [CpTiMo5- O18]32,35 the imido [Mo6O18(NC6H4Me-p)]22,36 diazenido [Mo6- O18(NNC6F5)]32,40 and hydrazido [Mo6O18(NNMePh)]2241 derivatives have emphasized the geometrical distortions related to the substitution.Owing to the relatively low precision of the structural data for [n-Bu4N][Mo6Cp*O18]?Me2CO, the alteration, in those of the M4(m-O)4 rings involving the substituted centre, of the regular trans-alternation pattern of long and short bond lengths observed for [Mo6O19]22 is unclear. However, the balance of Mo(i)–O(10) distances indicates a signifi- cant displacement of the central oxygen atom O(10) towards Mo(1) [Mo(1)–O(10) 2.14 (1), Mo(6)–O(10) 2.48 (1), average Moeq–O(10) 2.33 Å].This is in close agreement with the reported data for [C5Me5O][Mo6Cp*O18].17 This is the largest deviation observed in a Lindqvist derivative and reflects the weak trans influence of the h5-Cp* ligand. Indeed, structural data for complexes of the type [ReCp*XCl2] (X = O42 or t-BuN43) clearly show that the trans influence decreases along the series O > NR @ Cp*.Electrochemical studies As expected, the [Mo6Cp*O18]2 anion is electrochemically active. The voltammogram of [n-Bu4N][Mo6Cp*O18] in acetonitrile at a carbon electrode displays a reversible (Epa 2 Epc = 0.068 V) reduction process at 20.033 V vs. SCE, as depicted in Fig. 2. No further reduction process could be clearly observed, due to adsorption eVects. On the other hand, the polarogram of [n-Bu4N][Mo6Cp*O18] in acetonitrile at a dropping mercury electrode displays three well defined waves, at 20.09, 21.04 and Fig. 1 Molecular structure of the anion [Mo6Cp*O18]2. 21.44 V vs. SCE respectively, each of which involves one electron by comparison with [n-Bu4N]2[Mo6O19]. The positive shift of the first reduction process of [Mo6Cp*O18]2, compared to [Mo6O19]22 (E =20.37 V), reflects the eVect of the overall charge lowering on substitution. It should be even larger if it were not for the strong donation from the Cp* ligand. NMR studies An exhaustive characterization of [Mo6Cp*O18]2 by multinuclear NMR has been performed.The 95Mo and 17O spectra are presented in Figs. 3 and 4 respectively. On the basis of C4v symmetry for the [Mo6Cp*O18]2 anion in solution, three signals are expected in the 95Mo NMR spectrum. They are indeed observed at d 175, 164 and 235, with respective intensities of 1:4:1. Let us observe that the signal from the four equivalent equatorial Mo atoms is quite narrow (Dn2� 1 ª 60 Hz) with respect to the two other ones (Dn2� 1 ª 200 Hz).This is consistent with a short quadrupolar relaxation rate (larger T2) for the equatorial sites, according to a smaller electric field gradient for a less distorted site symmetry (see above). Fig. 2 Cyclic voltammogram of [n-Bu4N][Mo6Cp*O18] in CH3CN, at a carbon electrode vs. SCE at a scan rate of 0.1 V s21. Table 1 Selected distances (Å) and bond angles (8) for [n-Bu4N]- [Mo6Cp*O18]?Me2CO Mo(1)–O(10) Mo(1)–O(13) Mo(1)–O(15) Mo(1)–C(2) Mo(1)–C(4) Mo(2)–O(2) Mo(2)–O(12) Mo(2)–O(25) Mo(3)–O(3) Mo(3)–O(13) Mo(3)–O(34) Mo(4)–O(4) Mo(4)–O(14) Mo(4)–O(45) Mo(5)–O(5) Mo(5)–O(15) Mo(5)–O(45) Mo(6)–O(6) Mo(6)–O(26) Mo(6)–O(46) Mo(1)–O(12)–Mo(2) Mo(1)–O(14)–Mo(4) Mo(2)–O(23)–Mo(3) Mo(2)–O(26)–Mo(6) Mo(3)–O(36)–Mo(6) Mo(4)–O(46)–Mo(6) 2.14(1) 1.91(2) 1.87(2) 2.34(2) 2.45(3) 1.67(2) 1.90(2) 1.93(2) 1.68(2) 1.92(2) 1.96(2) 1.68(2) 1.95(2) 1.94(2) 1.65(2) 1.93(2) 1.94(2) 1.62(2) 1.85(2) 1.87(2) 116.2(7) 115.9(8) 115.2(6) 119.8(9) 120.4(10) 119.6(9) Mo(1)–O(12) Mo(1)–O(14) Mo(1)–C(1) Mo(1)–C(3) Mo(1)–C(5) Mo(2)–O(10) Mo(2)–O(23) Mo(2)–O(26) Mo(3)–O(10) Mo(3)–O(23) Mo(3)–O(36) Mo(4)–O(10) Mo(4)–O(34) Mo(4)–O(46) Mo(5)–O(10) Mo(5)–O(25) Mo(5)–O(56) Mo(6)–O(10) Mo(6)–O(36) Mo(6)–O(56) Mo(1)–O(13)–Mo(3) Mo(1)–O(15)–Mo(5) Mo(2)–O(25)–Mo(5) Mo(3)–O(34)–Mo(4) Mo(4)–O(45)–Mo(5) Mo(5)–O(56)–Mo(6) 1.94(2) 1.88(2) 2.38(3) 2.40(3) 2.44(3) 2.33(1) 1.92(2) 1.97(2) 2.33(1) 1.96(2) 1.97(2) 2.33(1) 1.90(2) 1.95(2) 2.34(2) 1.95(2) 1.94(2) 2.48(1) 1.84(2) 1.86(2) 116.5(8) 116.8(8) 116.4(7) 117.4(8) 115.8(6) 120.7(8)54 J.Chem. Soc., Dalton Trans., 1999, 51–55 Table 2 Selected NMR data for [Mo6O19]22 derivatives d(17O)a Complex [Mo6O19]2241 [CpTiMo5O18]3235 [VMo5O19]3246 [Mo6O18(NC6H4NO2)]2244 [Mo6Cp*O18]2 d(95Mo)a 125 (6) 143 (5), 51 (1) 175 (1), 235 (1) 164 (4) Ot 934 (6) 863 (4), 834 (1) 1200 (1), 885 (5) 923 (4), 914 (1) 931 (4), 890 (1) m-O 565 (12) 641 (4), 535 (4), 516 (4) 665 (4), 541 (4), 531 (4) 581 (4), 567 (4), 555 (4) 613 (4), 580 (4), 559 (4) a d in ppm; multiplicity in parentheses.Our parallel studies of the imido series [Mo6O18(NC6- H4Z-p)]22 (Z = OMe, Me, H, F, Cl, Br, CF3 or NO2) have shown that the expected signals for the Mo atoms bearing a terminal oxo ligand are lightly deshielded (less than 30 ppm) with respect to [Mo6O19]22 and that they are hardly resolved. On the contrary, the Mo bearing the imido ligand is significantly shielded, to an extent which depends on Z.The corresponding signal is indeed unambiguously assigned since it appears as a 1:1:1 triplet due to scalar 1J(95Mo–14N) coupling.22,44 Accordingly, the most shielded signal at d 235 for [Mo6Cp*O18]2 should be attributed to the Mo atom bearing the Cp* ligand and the two other signals at d 175 and 164 to the two other types of Mo. If we assume a predominant diamagnetic contribution to the 95Mo shielding, at least if the paramagnetic contribution is supposed to be nearly constant in these high-valent hexamolybdates, which seems reasonable since none of them has low-lying vacant electronic levels, then 95Mo NMR data further suggest that donation decreases along the series Cp* > NR (R = aryl) > O, which also simply correlates with the softness of the attached ligand.† The natural-abundance 17O NMR spectrum of [n-Bu4N]- [Mo6Cp*O18] has been measured in CH3CN–CD3CN at 343 K in order to reduce the signal linewidth.It displays a pattern of six lines at d 931, 890, 613, 580, 559 and 22.7. As 17O chemical shifts of polyoxometalates correlate with the metal–oxygen p-bond order,45 assignment of the resonances is straightforward: the two most deshielded lines (4 : 1) are assigned to the two types of terminal oxo ligands (Ot), the three close lines (4:4:4) around d 600 to the three types of symmetry nonequivalent m-O ligand (Ob) and finally the most shielded one at d 22.7 to the central oxygen atom (Oc). Owing to the large distortion of the Mo6Oc octahedron with respect to Oh symmetry, Oc should experience a relatively large electric field gradient, which favours the quadrupolar relaxation of this nucleus; this explains why the Oc resonance could be observed for [Mo6Cp*O18]2, Fig. 3 The 95Mo NMR spectrum of [n-Bu4N][Mo6Cp*O18] in CH3CN– CD3CN, recorded at 343 K. † The correlation of the metal chemical shift with the softness of the ligand has been suggested by a referee.whereas this line was hardly (or not) detected for nitrosyl derivatives under the same experimental conditions of quick pulse repetition (20 Hz).21 When considering the weighted average chemical shift for each set of oxygen atoms, it appears that m-O are deshielded, while terminal oxygen atoms are shielded, with respect to the parent anion [Mo6O19]22 (Table 2). The decrease of the overall charge from [Mo6O19]22 to [Mo6Cp*O18]2 should lead to a deshielding of all oxygen resonances.Therefore the net shielding of the terminal oxygen atoms suggests that some other eVect of the Cp* ligand prevails over the charge eVect. This further supports the conclusion that Cp* is a better donor than the oxo ligand. Moreover the weighted average d(Ot) is nearly the same for the [Mo6Cp*O18]2 monoanion (923) and for the [Mo6O18(NC6H4NO2-p)]22 dianion (921), which suggests that the Cp* ligand is a better donor than the imido ligands, in agreement with the conclusions from 95Mo NMR data.Concluding remarks The Lindqvist-type derivative [n-Bu4N][Mo6Cp*O18] has been successfully prepared by reaction of [n-Bu4N][MoCp*O3] with [n-Bu4N]2[Mo4O10(OMe)4Cl2] in methanol, and has been characterized by single-crystal X-ray diVraction and multinuclear NMR in solution. We are currently exploiting the potential of [MoCp*O3]2 and of its tungsten analogue in the syntheses of Cp*–oxo polyanions by triggering the condensation with diVerent Brönsted and Lewis acids.References 1 P. Gouzerh and A. Proust, Chem. Rev., 1998, 98, 77. 2 K. Isobe, Acc. Chem. Res., 1993, 26, 524. 3 Y. Hayashi, Y. Ozawa and K. Isobe, Chem. Lett., 1989, 425. 4 H. K. Chae, W. G. Klemperer and V. W. Day, Inorg. Chem., 1989, 28, 1423. 5 V. W. Day and W. G. Klemperer, Polyoxometalates: from Platonic Fig. 4 Natural abundance 17O NMR spectrum of [n- Bu4N][Mo6Cp*O18] in CH3CN–CD3CN, recorded at 343 K. * Represents the [Mo6O19]22 impurity.J.Chem. Soc., Dalton Trans., 1999, 51–55 55 Solids to Anti-retroviral Activity, eds. M. T. Pope and A. Müller, Kluwer, Dordrecht, 1994, p. 87. 6 R. G. Finke and M. W. Droege, J. Am. Chem. Soc., 1984, 106, 7274. 7 R. G. Finke, B. Rapko and P. J. Domaille, Organometallics, 1986, 5, 175. 8 D. J. Edlung, R. J. Saxton, D. K. Lyon and R. G. Finke, Organometallics, 1988, 7, 1692. 9 Y. Lin, K. Nomiya and R. G. Finke, Inorg. Chem., 1993, 32, 6040. 10 M. Pohl and R. G. Finke, Organometallics, 1993, 12, 1453. 11 B. Rapko, M. Pohl and R. G. Finke, Inorg. Chem., 1994, 33, 3625. 12 T. Nagata, M. Pohl, H. Weiner and R. G. Finke, Inorg. Chem., 1997, 36, 1366. 13 N. Mizuno, H. Weiner and R. G. Finke, J. Mol. Catal., 1996, 114, 15. 14 J. D. Aiken III, Y. Lin and R. G. Finke, J.. Catal., 1996, 114, 29. 15 A. R. Siedle, C. G. Markell, P. A. Lyon, K. O. Hodgson and A. L. Roe, Inorg. Chem., 1987, 26, 219. 16 F. Bottomley and L. Sutin, Adv.Organomet. Chem., 1988, 28, 339. 17 F. Bottomley and J. Chen, Organometallics, 1992, 11, 3404. 18 P. Gouzerh, Y. Jeannin, A. Proust and F. Robert, Angew. Chem., Int. Ed. Engl., 1989, 28, 1363. 19 A. Proust, P. Gouzerh and F. Robert, Inorg. Chem., 1993, 32, 5291. 20 A. Proust, R. Thouvenot, F. Robert and P. Gouzerh, Inorg. Chem., 1993, 32, 5299. 21 A. Proust, R. Thouvenot, S.-G. Roh, J.-K. Yoo and P. Gouzerh, Inorg. Chem., 1995, 34, 4106. 22 A. Proust, R. Thouvenot, M.Chaussade, F. Robert and P. Gouzerh, Inorg. Chim. Acta, 1994, 224, 81. 23 A. Proust, S. Taunier, V. Artero, F. Robert, R. Thouvenot and P. Gouzerh, Chem. Commun., 1996, 2195. 24 N. H. Hur, W. G. Klemperer and R.-C. Wang, Inorg. Synth., 1990, 27, 78. 25 M. S. Rau, C. M. Kretz, L. A. Mercando, G. L. GeoVroy and A. L. Rheingold, Organometallics, 1993, 12, 3447. 26 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158. 27 G. M. Sheldrick, SHELXS 86, Program for Crystal Structure Determinations, University of Göttingen, 1986. 28 International Tables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. IV, p. 149. 29 D. J. Watkin, C. K. Prout, J. R. Carruthers and P. W. Betteridge, CRYSTALS, Chemical Crystallography Laboratory, University of Oxford, Oxford, 1996. 30 D. J. Watkin, C. K. Prout and L. J. Pearce, CAMERON, Chemical Crystallography Laboratory, University of Oxford, 1996. 31 J. R. Harper and A. L. Rheingold, J. Am. Chem. Soc., 1990, 112, 4037. 32 M. T. Pope, Heteropoly and Isopoly Oxometalates, Springer, Berlin, 1983. 33 H. Kang, S. Liu, S.-N. Shaikh, T. Nicholson and J. Zubieta, Inorg. Chem., 1989, 28, 920. 34 C. Rocchiccioli-DeltcheV, R. Thouvenot and M. Fouassier, Inorg. Chem., 1982, 21, 30. 35 T. M. Che, V. W. Day, L. C. Francesconi, M. F. Fredrich, W. G. Klemperer and W. Shum, Inorg. Chem., 1985, 24, 4055. 36 Y. H. Du, A. L. Rheingold and E. A. Maatta, J. Am. Chem. Soc., 1992, 114, 345. 37 J. Sundermeyer and D. Runge, Angew. Chem., Int. Ed. Engl., 1994, 33, 1255. 38 W. A. Nugent and J. M. Mayer, Metal–Ligand Multiple Bonds, Wiley, New York, 1988. 39 H. R. Allcock, E. C. Bissel and E. T. Shawl, Inorg. Chem., 1973, 12, 2963. 40 S. Bank, S. Liu, S.-N. Shaikh, X. Sun, J. Zubieta and P. D. Ellis, Inorg. Chem., 1988, 27, 3535. 41 H. Kang and J. Zubieta, J. Chem. Soc., Chem. Commun., 1988, 1192. 42 W. A. Herrmann, E. Herdtweck, M. Flöel, J. Kulpe, U. Küsthardt and J. Okuda, Polyhedron, 1987, 6, 155. 43 W. A. Herrmann, G. Weichselbaumer, R. A. Paciello, R. A. Fischer, D. Herdtweck, J. Okuda and D. W. Marz, Organometallics, 1990, 9, 489. 44 S. Taunier, A. Proust, R. Thouvenot and P. Gouzerh, unpublished work. 45 W. G. Klemperer, Angew. Chem., Int. Ed. Engl., 1978, 17, 246. 46 M. Filowitz, R. K. C. Ho, W. G. Klemperer and W. Shum, Inorg. Chem., 1979, 18, 93. Paper 8/05832F

ISSN:1477-9226    

DOI:10.1039/a805832f

出版商:RSC    

年代:1999

数据来源: RSC