High-valent corrole metal complexes

J Biol Inorg Chem (2001) 6: 733±738 DOI 10.1007/s007750100273 COMMENTARY Zeev Gross High-valent corrole metal complexes Received: 26 April 2001 / Accepted: 29 May 2001 / Published online: 1 August 2001 Ó SBIC 2001 Abstract This commentary concentrates on corrole complexes with the three metal ions that are most relevant to oxidation catalysis: chromium, manganese, and iron. Particular emphasis is devoted to the only recently introduced meso-triarylcorroles and a comparison with the traditionally investigated b-pyrrole-substituted corroles. Based on a combination of spectroscopic methods, electrochemistry, and X-ray crystallography, it is concluded that in most high-valent metallocorroles the corrole is not oxidized. Both experimental (for (oxo)chromium(V) corrole) and computational (for (oxo)manganese(V) corrole) evidence indicate that the stabilization of high-valent metal ions by corroles originates from a combination of short metal-nitrogen bonds and large metal out-of-plane displacements in the corrole, which lead to quite unexpected interactions of the oxo-metal p* orbitals with the in-plane orbitals of the corrole. Keywords Corrole á Iron á Manganese á Chromium á Cation radicals Corroles may be considered as the non-natural analogs of the cobalt-coordinating corrin ring in Vitamin B12 with which they share an identical skeleton, or as one meso-carbon short porphyrins with whom they share aromaticity (Scheme 1) [1]. Corrole chemistry, which started in 1964 [2], also led to the ®rst structural characterization of a free-base corrole by Hodgkin and coworkers in 1971 in course of their Vitamin B12 project [3]. The main research e€orts during the years to come were devoted to the synthesis of corroles [4] and to investigation of the coordination chemistry of their metal complexes [5, 6]. Still, compared to porphyrins, the re- Z. Gross Department of Chemistry and Institute of Catalysis Science and Technology, Technion ± Israel Institute of Technology, Haifa 32000, Israel E-mail: chr10zg@tx.technion.ac.il search activity in corrole chemistry remained very low. For example, a computer search of ``article titles, keywords, or abstract'' in the ISI database for 1998 produced 1143 hits for porphyrin*, but only 10 hits for corrol*. There was also no reported application of corroles or their metal complexes in any scienti®c journal or in patent databases up to 1999, and it took 28 years to obtain the second ever X-ray crystal structure of a metalfree corrole [7]. This situation is however to experience a major change, as much more ecient and simple methodologies for the preparation of corroles are constantly being introduced (for the most recent developments that are not covered by the review [4], see [7, 8, 9]). One main di€erence between the corrin, porphyrin, and corrole macrocycles (Scheme 1) is the number of ionizable protons in their N4 coordination core, which is 1, 2, and 3, respectively. Accordingly, in their coordination complexes they act as mono-, di-, and trianionic ligands, respectively. This rather naive distinction is actually of importance, as may be appreciated by the following facts. The catalytic action of Vitamin B12 relies on stabilization of cobalt(I) by the monoanionic corrinate ligand, the most common oxidation states of porphyrin-supported metals are +2 and +3, and the trianionic corrolate ligand supports unusually high metal oxidation states. Within this commentary we will examine the coordination chemistry of corroles with chromium, iron, and manganese, not least because of the major role of these metals in biological systems. The information about the electron-rich b-pyrrole alkylsubstituted corroles has recently been reviewed [5], while electron-poor corroles were only most recently introduced [7, 8]. Accordingly, we will compare the metal complexes of H3(oec), the prototype of electron-rich corroles, with that of H3(tpfc) ± the most electron-poor corrole known to date (for the structures and abbreviations of all discussed corroles, see Scheme 2). The large interest in H3(tpfc) and its metal complexes is that they are the ®rst corroles that were utilized in catalysis and other applications [10, 11], and H3(tpfc) is the ®rst commercially available corrole [12]. 734 One important factor in high valent metallocorroles and metalloporphyrins is the oxidation site dilemma, i.e., con®rming whether or not the macrocycle is oxidized. In porphyrins, resolution of this dilemma is largely assisted by investigations of non-transition metal complexes, especially with zinc and magnesium [13, 14]. Their one-electron oxidation leads to the corresponding porphyrin p-cation radicals, which were extensively examined by spectroscopic methods for elucidation of their characteristic properties. Additional information is achieved from cyclic voltammetry (CV), in which the di€erences between their ®rst oxidation and reduction potentials (Eox±red) provides the so-called electrochemical HOMO-LUMO gap of the porphyrin macrocycle [15, 16]. Until most recently such investigations were not carried out with corroles, which, due to their action as Scheme 1 The skeleton-only structures of porphyrin, corrin and corrole trianionic ligands, do not form stable zinc(II) or magnesium(II) complexes. Although oec complexes with indium(III) [17], germanium(IV) [17], tin(IV) [17, 18], and phosphorus(V) [19] have been prepared, only the tin corroles were oxidized to their p-cation radicals [18]. However, the electrochemistry of a quite large series of main group oec complexes is now described in the review of Erben et al. [5], from which an electrochemical HOMO-LUMO gap of 2.20 V was elucidated. Even more recently, the germanium(IV), tin(IV), and phosphorus(V) tpfc complexes were examined by similar means and a HOMO-LUMO gap of 2.12‹0.02 V was determined [20]. This brings the HOMO-LUMO gap of both type of corroles into the range of 2.20‹0.15 V (see entries 1±3 and 10±12 of Table 1), obtained for a large range of metalloporphyrins [21]. The corrole and porphyrin cation radicals are also similar in terms of their green color and in their reduced intensity Soret band relative to the non-oxidized precursors. However, the characteristic appearances of new bands at above 600 nm in porphyrin radicals is much less pronounced in corrole radicals [20, 22]. Also, while the EPR spectra of porphyrin radicals depend on the type of orbital the electron was removed from ± singlets for oep complexes (2A1u) and resolved nitrogen hyper®ne splittings for most meso-aryl porphyrins (2A2u) ± for all corrole radicals examined to date only singlets were obtained. This di€erence between porphyrin and corrole Scheme 2 Formal drawing and abbreviations of the discussed corroles and their metal complexes M R2=R3=R7=R8=R12=R13=R17=R18=C2H5 H3 R5=R10=R15=H Sn Mn Mn H3 R2=R3=R7=R8=R12=R13=R17=R18=H R5=R10=R15=C6F5 Ga Sn Cr Mn Mn R2=R3=R7=R8=R12=R13=R17=R18=CH3 H3 Cr R5=R10=R15=H Mn Fe R2=R3=R17=R18=CH3 H3 R7=R8=R12=R13=C2H5 Cr Cu R5=R10=R15=H L L¢ Abbreviation M L L¢ Abbreviation ± Cl Cl phenyl ± pyridine Cl O OPPh3 Cl ± ± ± Cl ± O ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± H3(oec) (oec)Sn(Cl) (oec)Mn(Cl) (oec)Mn(C6H5) H3(tpfc) (tpfc)Ga(pyr) (tpfc)Sn(Cl) (tpfc)Cr(O) (tpfc)Mn(OPPh3) (tpfc)Mn(Cl) H3(omc) (omc)Cr (omc)Mn (omc)Fe(Cl) H3(7,8-temc) (7,8-temc)Cr(O) (7,8-temc)Cu Fe Fe Fe Cl phenyl O-Fe ± ± ± (oec)Fe(Cl) (oec)Fe(C6H5) [(oec)Fe]2O Mn Mn Fe Fe Fe Fe Br O Cl O-Fe pyridine ether ± ± ± ± pyridine ether (tpfc)Mn(Br) (tpfc)Mn(O) (tpfc)Fe(Cl) [(tpfc)Fe]2O (tpfc)Fe(pyr)2 (tpfc)Fe(OEt2)2 735 Table 1 Half wave potentials (in Volts vs SCE, in CH2Cl2 unless di€erently indicated) for oxidation and reduction of corrole metal complexes Entry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Metal (axial ligand) IV Sn (Ph) PV(CH3)2 PV(Ph2)2 SnIV(Cl) CrV(O) FeIV(Cl) FeIV(Ph) FeIII(pyr) FeIII(NO) SnIV(Cl) GeIV(OH) PV(OH)2 CrV(O) FeIV(Cl) FeIV(O)FeIV FeIII(pyr)2 FeIII(NO) Corrole Cor/Cor+ Cor/Cor± DE (V)a oec oec oec oec 7,8-temc oec oec oec oec tpfc tpfc tpfc tpfc tpfc tpfc tpfc tpfc 0.47 0.44 0.56 0.67 0.63d 0.76 0.43 1.03e 0.61 1.20 1.13 1.05 1.24 1.24 1.25, 0.88 ±1.74 ±1.75 ±1.71 2.21 2.19 2.27 ±1.92e ±0.94 ±0.99 ±1.05 Mn/Mn+1 0.21 2.53 2.14 2.13 2.10 0.74 1.07 Mn/Mn±1 ±1.46b ±0.33 ±0.08 ±0.62 ±1.04 ±0.41 0.11 0.44 0.13 ±0.78f 0.00 Reference 18 5c 5c 18 25 37 37 37 37 20 20 20 28 20 20 20 20 a The di€erence between the ®rst oxidation and reduction potentials of the corrole Sn(IV)/Sn(II) c In benzonitrile d Originally interpreted as a CrV/CrVI couple, see text e Not assigned with full con®dence f Irreversible process (Epc at 0.1 V/s) b Fig. 1 Perspective drawing of the DFT calculated spin densities of the two highest occupied molecular orbitals of Ga(tpfc) [22] p-cation radicals was recently addressed by computation of the relevant orbitals for both the hypothetical nonsubstituted corrole and the experimentally examined (tpfc)Ga [22, 23]. As shown in Fig. 1, the two highest occupied orbitals of [(tpfc+.)Ga] are very close in energy and the spin density on the nitrogens in both of them is very small. Accordingly, the hyper®ne splittings (aN) are indeed expected to be very small (or unobservable), regardless of the ground state of the radical (2B1 or 2A2). It may thus be concluded that the spectral properties of corrole radicals are signi®cantly less distinctive than those of porphyrin radicals. However, the HOMOLUMO gap of corroles remains a very useful tool for distinguishing between metal and corrole centered oxidation, as demonstrated below. The ®rst reported high valent corrole metal complexes are of chromium. Initially, the reaction product of H3(omc) and Cr(CO)6 was reported as (omc)CrIII [24], but this was recently claimed to be erroneous [5]. On the other hand, the reaction product of H3(7,8-temc) with CrCl2 was con®dently characterized by a combination of various spectroscopic methods as a very stable (oxo)chromium(V) corrole, (7,8-temc)Cr(O) [25]. Based on spectroelectrochemistry, the ®rst oxidation potential at 0.63 V was assigned as involving the metal, i.e., oxidation of (7,8-temc)CrV(O) to [(7,8-temc)CrVI(O)]+. These very interesting results ± (oxo)chromium(V) porphyrins are much less stable [26] ± were not further explored until most recently when chromium was inserted into H3(tpfc) (with Cr(CO)6 as the metal source) [27]. Aerobic workup a€orded the very stable (oxo)chromium(V) complex, (tpfc)Cr(O), that was characterized by spectroscopic methods and X-ray crystallography (Fig. 2). The main structural parameters are of short Cr-N bonds, a long Cr-O bond, and large out-of-plane Cr (0.6 AÊ), while EPR spectroscopy shows that the spin densities are large on the nitrogens and small on the chromium. These results were analyzed as indicating that the experimentally observed stabilization of the [Cr(O)+3] moiety is due to a strong r(N)-donation 736 by the corrolato trianion to the metal. Whether or not this is also a major factor in the stabilization of other high oxidation states by corroles remains to be con®rmed. The di€erence between the oxidation and reduction potentials is 1.13 V in (tpfc)Cr(O) and 0.96 V in (7,8temc)Cr(O) (Table 1, entries 5, 13) [25, 27]. As the HOMO-LUMO gap for the corroles is 2.1±2.2 V, the possibility that both electrochemical processes are corrole-centered may immediately be ruled out. However, the almost identical oxidation potentials of (tpfc)Cr(O) and (tpfc)Sn(Cl) on the one hand and that of (7,8temc)Cr(O) and (oec)Sn(Cl) on the other (the values are 1.24, 1.20, 0.63, and 0.67 V (Table 1, entries 13, 10, 5, 4) [18, 20, 25, 27], respectively) suggest that the oxidation of both (oxo)chromium(V) complexes is corrolecentered. Support for this conclusion comes from the spectroelectrochemistry of (tpfc)Cr(O), which shows that its Soret band shifts to shorter wavelengths and its intensity is heavily reduced upon oxidation (Fig. 3a). A similar, although less pronounced, behavior may also be noticed for (7,8-temc)Cr(O) [25]. Accordingly, the suggestion of Murakami et al about a Cr(V)/Cr(VI) process is probably erroneous, and their improper interpretation of the spectral results might well be the result of an only Fig. 2 Partial views (hydrogen atoms and meso-aryl substituents omitted) of the X-ray structures of the two conformers of (tpfc)Cr(O) [27] Cr-O (AÊ) Cr-N1 (AÊ) Cr-N2 (AÊ) Cr-N3 (AÊ) Cr-N4 (AÊ) N1-Cr-N3 (°) N2-Cr-N4 (°) 1.5700(17) 1.936(2) 1.9425(19) 1.927(2) 1.936(2) 145.53(9) 145.35(9) Fig. 3a, b Spectroelectrochemistry of: a the oxidation processes; b the reduction processes of (tpfc)Cr(O) [28] 1.5317(17) 1.943(2) 1.943(2) 1.929(2) 1.927(2) 144.84(8) 145.03(8) partial oxidation [25]. Full conformation for the oxidation site dilemma in the one-electron oxidation products of the (oxo)chromium(V) corroles may be achieved by their isolation and examination by EPR and NMR (the d0 (oxo)chromium(VI) corrole must be diamagnetic). Most recent results for the oxidation product of (tpfc)Cr(O) are supportive of corrole-centered process, i.e., a triplet EPR is obtained [28]. As for the reduction process, the spectroelectrochemistry of (tpfc)Cr(O) (Fig. 3b) and (7,8-temc)Cr(O) are very similar, and the large red shift in the Soret band and its increased intensity are clearly supportive of a CrV/CrIV process. Consistent with this suggestion is the fact that in situ reduction of (tpfc)Cr(O) by cobaltocene results in an NMR spectrum with sharp and non-shifted resonances (indicative for diamagnetism), which is expected for a low-spin d2 metal coordinated by a porphyrin-like macrocycle [28]. Considering the fact that the stable oxidation states of chromium and iron corroles are +5 and +4, respectively, it is actually surprising that only manganese(III) corroles were reported until most recently [29]. Furthermore, the authors suggested that (oec)Mn and related tetracoordinated complexes are actually better described as manganese(II) corrole cation radicals rather than as manganese(III) corroles. This conclusion is based on 1H NMR spectral analysis, reminiscent of a proposal for manganese(III) porphyrins with very weakly bound axial ligands [30]. However, alternative interpretations of the NMR spectra have been put forward for both the corrole and the porphyrin complexes [5, 31]. In any case, the recent examination of the pentacoordinated (tpfc)Mn(OPh3) by X-ray crystallography, magnetic susceptibility, and high ®eld (high frequency) EPR reveals the characteristics of an isolated manganese(III) complex [32]. In addition, several reports about high valent manganese(IV) and manganese(V) corroles start to appear. In the review of Erben et al. [5], (oec)Mn(Cl) and (oec)Mn(C6H5) are mentioned, which according to the authors are best described as Mn(III) corrole radical and Mn(IV) corrole, respectively. For (tpfc)Mn, its surprisingly low oxidation potential (E1/2=0.75 V) was taken advantage of for 737 Fig. 4 ORTEP views of (tpfc)Mn(Br) (top, Mn-Br bond length: 2.428 AÊ) and (tpfc)Mn(Cl) (bottom, Mn-Cl bond length: 2.312 AÊ) [33] the easy preparation and full characterization of (tpfc)Mn(Br) and (tpfc)Mn(Cl) [33]. Both complexes display the characteristic features of authentic Mn(IV) corroles, a conclusion based on UV-vis (single Soret band) and EPR (broad g=4.3 signal and a g=1.99 signal with resolved Mn hyper®ne structure, aMn=85 G) spectroscopies, X-ray crystallography (Fig. 4), and magnetic susceptibility (l=3.8 lB). Full conformation of these intriguing results and possible di€erences between the tpfc and oec complexes will have to await the full reports for the latter. Finally, when the green (tpfc)Mn was used as catalyst for the epoxidation of styrene by iodosylbenzene, the steady state color of the reaction mixture was red [34]. Independent oxidation of (tpfc)Mn by either ozone, iodosylbenzene, or MCPBA was shown to lead to the same red complex, which was identi®ed as (tpfc)Mn(O). The assignment of this quite stable complex as an (oxo)manganese(V) corrole was mainly based on its diamagnetic 1H and 19F NMR spectra, an achievement that took 20 years for the much less stable (oxo)manganese(V) porphyrins [35]. Starting in 1994, Vogel and coworkers have demonstrated the stability of iron(IV) corroles via the isolation and structural characterization of the mononuclear (oec)Fe(Cl) and (oec)Fe(C6H5) complexes and the dinuclear complex [(oec)Fe]2O [36]. The employment of all relevant spectroscopic methods leaves little doubt about the iron(IV) oxidation state in all three complexes (for a full summary, see also [5]). The oxidation and reduction of these complexes (Table 1, entries 6, 7) were shown to be corrole- and metal-centered, respectively, which also led to isolation of an authentic iron(IV) corrole cation radical [37]. Still, the interpretations regarding the iron(IV) oxidation state were recently challenged via Fig. 5 The X-ray structures of (tpfc)Fe(Cl) and [(tpfc)Fe]2O [20, 39] NMR analysis of (omc)Fe(Cl) and these authors propose that the alternative formulation of iron(III) corrole cation radical cannot be ruled out [38]. The intriguing demonstration that (tpfc)Fe(Cl) is a good oxidation catalyst [10] initiated the full characterization of this and other related iron complexes [20, 39]. The X-ray structures of (tpfc)Fe(Cl) and [(tpfc)Fe]2O (Fig. 5), in combination with several spectroscopic methods showed that both complexes are iron(IV) corroles (see also Table 1, entries 14±17). Electrochemistry and spectroelectrochemistry of (tpfc)Fe(Cl) was also used for assignment of its oxidation and reduction processes as corrole- and metal-centered, respectively [20]. In addition, two iron(III) complexes, (tdcp)Fe(pyr)2 and (tdcp)Fe(OEt2)2, were prepared and used as catalysts. The full characterization of these iron(III) corroles (a preliminary X-ray structure of (tpfc)Fe(pyr)2 can be found in [20]) and the comparison with (tpfc)Fe(Cl) will certainly throw additional light on the electronic structure of the latter complex. We conclude this commentary by predicting that many novel aspects of the coordination chemistry of corrole metal complexes will be revealed very soon, especially with respect to high metal oxidation states. This will eventually lead to a much better understanding of the puzzling situation where an easily oxidized ligand stabilizes metals in high oxidation states. Acknowledgments I wish to acknowledge my grateful thanks to the highly fruitful cooperations with Professor I Goldberg from the Tel 738 Aviv University (Israel) and Professor HB Gray from the California Institute of Technology (USA), and the talented and highly devoted students and co-workers from my laboratory at the Technion (N Galili, I Saltsman, L Simkhovich, A Mahammed, G Golubkov) and from that of Professor HB Gray in CalTech (AE Meier-Callahan, J Bendix), whose enthusiasm made the breakthrough in corrole chemistry possible. References 1. Sessler JL, Weghorn SJ (1997) In: Expanded, contracted, and isomeric porphyrins. Pergamon, Oxford, pp 11±125 2. Johnson AW, Kay IT (1964) Proc Chem Soc 89 3. Harrison HR, Hodder OJR, Hodgkin DC (1971) J Chem Soc B:640 4. Paolesse R (2000) In: Kadish KM, Smith KM, Guilard R (eds) The porphyrin handbook. Academic Press, vol 2, pp 201±232 5. Erben C, Will S, Kadish KM (2000) In: Kadish KM, Smith KM, Guilard R (eds) The porphyrin handbook. Academic Press, vol 2, pp 233±300 6. Licoccia S, Paolesse R (1995) Struct Bonding 84:71±133 7. Gross Z, Galili N, Simkhovich L, Saltsman I, Botoshansky M, BlaÈser D, Boese K, Goldberg I (1999) Org Lett 1:599 8. Gross Z, Galili N, Saltsman I (1999) Angew Chem Int Ed Engl 38:1427 9. Paolesse R, Jaquinod L, Nurco DJ, Mini S, Sagone F, Boschi T, Smith KM (1999) J Chem Soc Chem Commun 1307 10. Gross Z, Simkhovich L, Galili N (1999) J Chem Soc Chem Commun 599 11. Aviezer D, Cotton S, David M, Segev A, Khaselev N, Galili N, Gross Z, Yayon A (2000) Cancer Res 60:2973 12. Strem Chemicals (Illinois) and patent pending application 126,426 (Gross Z, Technion) 13. Barzilay CM, Sibilia SA, Spiro TG, Gross Z (1995) Chem Eur J 1:222 14. Ehlinger N, Scheidt WR (1999) Inorg Chem 38:1316 15. Felton RH (1978) In: Dolphin D (ed) The porphyrins. Academic Press, vol 5, chap 3 16. Fuhrhop JH, Kadish KM, Davis DG (1973) J Am Chem Soc 95:5140 17. Paolesse R, Licoccia S, Boschi T (1990) Inorg Chim Acta 178:9 18. Kadish KM, Will S, Adamian VA, Walther B, Erben C, Ou Z, Guo N, Vogel E (1998) Inorg Chem 37:4573 Note added in proof. In a most recent computational investigation, the formally (oxo)manganese(V) complexes of porphyrin and corrole were examined by DFT [40]. The ground state of the porphyrin complex was found to be a triplet, i.e., (oxo)manganese(IV) porphyrin radical, while the corrole revealed a singlet, i.e., (oxo)manganese(V) corrole. Intriguingly, the singlet states of both the porphyrin (6±10 kcal/mol above the triplet state) and the corrole (3-6 kcal/mol below the triplet states) display short metal-nitrogen bonds and large metal out-of-plane displacement. It was concluded that these two factors are crucial for stabilization of the [Mn(O)]+3 moiety, as they lead to quite unexpected interactions of the Mn=O p* orbitals with the in-plane orbitals of the corrole. This conclusion is very much in line with the earlier mentioned results for the (oxo)chromium(V) corrole. 19. Paolesse R, Licoccia S, Boschi T, Khoury RG, Smith KM (1998) Chem Commun 1119 20. Simkhovich L, Mahammed A, Goldberg I, Gross Z (2001) Chem Eur J 7:1041±1055 21. Kadish KM (1986) Prog Inorg Chem 34:435 22. Bendix J, Dmochowski IJ, Gray HB, Mahammed A, Simkhovich L, Gross L (2000) Angew Chem Int Ed Eng 39:4048± 4051 23. Ghosh A, Wondimagegn T, Parusel ABJ (2000) J Am Chem Soc 122:5100 24. Boschi T, Licoccia S, Paolesse R, Tagliatesta P, Tehran MA, Pelizzi G, Vitali F (1990) J Chem Soc Chem Commun 501 25. Murakami Y, Matsuda Y, Yamada S (1981) J Chem Soc Dalton Trans 855 26. Groves JT, Kruper WJ, Haushalter RC, Butler WM (1982) Inorg Chem 21:1363 27. Meier-Callahan EA, Gray HB, Gross Z (2000) Inorg Chem 39:3605 28. Meier-Callahan EA, Simkhovich L, DiBilio AJ, Gray HB, Gross Z (submitted) 29. Licoccia S, Morgante E, Paolesse R, Polizio F, Senge MO, Tondello E, Boschi T (1997) Inorg Chem 36:1564 30. Turner P, Gunter MJ (1994) Inorg Chem 33:1406 31. Gross Z, Mahammed A, Barzilay CM (1998) J Chem Soc Chem Commun 1505 32. Bendix J, Golubkov G, Gray HB, Gross Z (2000) J Chem Soc Chem Commun 1957 33. Golubkov G, Bendix J, Gray HB, Mahammed A, Goldberg I, DiBilio AJ, Gross Z (2001) Angew Chem Int Ed Eng 2132± 2134 34. Gross Z, Golubkov G, Simkhovich L (2000) Angew Chem Int Ed Eng 39:4045±4047 35. Jin N, Groves JT (1999) J Am Chem Soc 121:2923 36. Vogel E, Will S, Tilling AS, Neumann L, Lex J, Bill E, Trautwein AX, Wieghardt K (1994) Angew Chem Int Ed Eng 33:731 37. VanCaemelbecke E, Will S, Autret M, Adamian VA, Lex J, Gisselbrecht JP, Gross M, Vogel E, Kadish KM (1996) Inorg Chem 35:184 38. Cai S, Walker FA, Licoccia S (2000) Inorg Chem 39:3466 39. Simkhovich L, Galili N, Saltsman I, Goldberg I, Gross Z (2000) Inorg Chem 39:2704 40. de Visser SP, Ogliaro F, Gross Z, Shaik S (2001) Chem Eur J 7 (in press)