摘要:

DALTON PERSPECTIVE J. Chem. Soc., Dalton Trans., 1999, 1353–1361 1353 A mechanistic and structural model for the formation and reactivity of a MnV] O species in photosynthetic water oxidation Julian Limburg,a Veronika A. Szalai b and Gary W. Brudvig *a a Department of Chemistry, Yale University, PO Box 208107, New Haven, CT 06520-8107, USA. E-mail: gary.brudvig@yale.edu b Department of Chemistry, University of North Carolina at Chapel Hill, CB# 3290, Chapel Hill, NC 27599-3290, USA Received 29th September 1998, Accepted 15th January 1999 Photosynthetic water oxidation is carried out by a tetranuclear Mn cluster contained in the membrane-bound protein complex photosystem II (PSII).The mechanism of PSII catalysed water oxidation is unknown; however, several current models invoke a high-valent Mn]] O species as a key intermediate in O–O bond formation. In part, these proposals are based on biophysical studies of the protein which suggest that the redox-active tyrosine residue, YZ, abstracts hydrogen atoms directly from substrate water molecules bound to the Mn4 cluster.In this paper, we consider organic oxidation and O–O bond-forming reactions catalysed by biomimetic Mn and Ru model complexes that are believed to proceed via M]] O intermediates. We also interpret biophysical data concerning the roles of Ca21 and Cl2 in photosynthetic water oxidation, proposing that they are involved in a hydrogen-bonded network between the Mn4 cluster and YZ.Connecting the observed reactivities of model complexes containing M]] O groups to spectroscopic information on the environment of the Mn4 cluster in the protein leads us to favour an O–O bond-forming step in photosynthetic water oxidation that occurs through nucleophilic attack of a calcium-bound hydroxide ligand on the electrophilic oxygen atom of a Mn]] O intermediate. In addition, a new role for Cl2 is proposed in which Cl2 tunes the nucleophilicity of the calcium-bound hydroxide. 1.0 Introduction The mechanism of photosynthetic water oxidation by photosystem II (PSII) remains, for the most part, an unresolved problem.In this review, we discuss recent results on the reactivity patterns of high-valent metal-oxo species and consider them in the context of current biophysical studies of PSII in order to propose a mechanistic model of the O–O bond-forming step of photosynthetic O2 evolution. The model extends previous proposals and includes a new mechanistic role for Ca21 and Cl2 which are required cofactors for water oxidation.The oxygen-evolving complex (OEC) of PSII, Fig. 1, contains a tetranuclear manganese cluster and a redox active tyrosine, YZ.1–4 While the precise structure of the Mn4 cluster is unknown, interpretations of EPR and X-ray absorption data have allowed for some fairly detailed proposals,3,5–9 some of which are shown in Fig. 2. A key structural element indicated from EXAFS studies is the di-m-oxo Mn dimeric unit, and there are numerous examples of oxo-Mn cluster model complexes (reviewed in refs. 6 and 7). To date, no manganese cluster has been synthesised that closely matches the spectroscopic properties of the OEC. Water oxidation is a four-electron process, and the OEC has been shown to cycle through five intermediate oxidation states from S0 to S4 (Fig. 3).10 Each S-state advance is associated with Gary Brudvig received his bachelor’s degree in chemistry from the University of Minnesota in 1976 and then went on to gain his Ph.D.in chemistry from the California Institute of Technology in 1980. His graduate work at Caltech concerned cytochrome c oxidase and was performed under the direction of Sunney Chan. From 1980–82, he was a Miller Fellow at the University of California, Berkeley studying with Kenneth Sauer. He joined the faculty at Yale University in 1982, where he is now a Professor of Chemistry, and where he applies both biophysical and bioinorganic techniques to the study of photosystem II.Julian Limburg was born in London in 1972. He earned his bachelor’s degree in chemistry from the University of East Anglia in 1993. His undergraduate studies included an exchange year at the University of Massachusetts, Amherst (1991–1992). Since 1993 he has been studying for a Ph.D. at Yale University under the direction of Gary Brudvig and Robert Crabtree working on bioinorganic model systems for photosynthetic water oxidation. Veronika Szalai received her bachelor’s degree in chemistry from Bryn Mawr College in 1988.Before entering graduate school, she was employed as a research assistant at EniChem America, Inc. She recently completed her Ph.D. in chemistry at Yale University using electron paramagnetic resonance to study the active site for water oxidation in photosystem II. Currently a Lineberger Comprehensive Cancer Center post-doctoral researcher in Dr. H. Holden Thorp’s laboratory at the University of North Carolina at Chapel Hill, she is interested in biophysical and bioinorganic chemistry.Gary W. Brudvig Julian Limburg Veronika A. Szalai1354 J. Chem. Soc., Dalton Trans., 1999, 1353–1361 light-induced charge separation at the chlorophyll-containing pigment P680 to form the strong oxidant P6801. The tyrosyl radical, YZ ?, formed upon reduction of P6801 oxidises the manganese cluster which, in turn, is reduced by electrons stripped from water.Flash-induced UV 11 and XANES12–14 data have been interpreted as showing that the Mn4 cluster is oxidised on each S-state transition. However, there is a current controversy over whether Mn or an associated ligand is oxidized in the S2 to S3 step.15 The manganese tetramer in PSII exhibits EPR signals from the odd-electron S0 and S2 states.16–19 A combination of EPR 8,20 Fig. 1 Model of photosystem II in the thylakoid membrane. The arrows show the direction of electron-transfer reactions.After initial photoexcitation, the special chlorophyll called P680 transfers an electron to a pheophytin (Pheo) molecule. The electron is rapidly passed to QA, a tightly-bound plastoquinone, and ultimately reduces an exchangeable plastoquinone in the QB-binding site. P6801 is reduced by tyrosine Z (YZ). Oxidized YZ is reduced by a tetranuclear manganese cluster, Mn4. Calcium and chloride are required and expected to be in close proximity to the Mn4 cluster.Fig. 2 Structural models of the OEC based on X-ray absorption spectroscopic data of PSII. Adapted from ref. 3. and XANES3,14,21,22 data has led to the assignment of S2 as MnIIIMnIV 3 although MnIII 3MnIV cannot be ruled out.23,24 Assignment of a MnIIIMnIV 3 configuration for the S2 state of the OEC has recently been supported by Blondin et al.25 who reported a mixed-valence manganese tetramer (Fig. 4) with an S = 1/2 ground state, formed by gamma irradiation of the corresponding EPR-silent MnIV tetramer.The EPR spectrum so obtained is the closest match from a model complex to the S2-state multiline EPR signal of the OEC seen thus far. Dioxygen is released during the transition from S3 to S0, and there have been a number of proposals on the nature of the active species involved in O–O bond formation.3,5,26–41 Many of these mechanisms have not been developed with the chemistry of Mn]] O groups in mind. The most current mechanistic models take into account the point-dipole distance between YZ ? and the Mn4 cluster, estimated to be 7–10 Å.42–45 Spectroscopic studies of YZ indicate that its proton is released to a nearby histidine residue 46–49 upon oxidation to generate the neutral tyrosyl radical, YZ ?,50 which suggests that proton movement is an important component of YZ reactivity. The O–H bonddissociation energy of water terminally ligated to Mn(III) or Mn(IV) has been estimated to lie between 78 and 89 kcal mol21.27,51,52 This is comparable to the O–H bond-dissociation energy of tyrosine, estimated at 86.5 kcal mol21,53,54 allowing for H-atom abstraction from a bound water to be exothermic and so provide extra driving force relative to a purely outer-sphere oxidation of the Mn4 cluster by YZ ?.It has been argued on the basis of studies of inhibited PSII that YZ ? cannot remain a competent oxidant of the Mn4 cluster above the S2 state unless proton-coupled electron-transfer from the Mn4 cluster to YZ ? is invoked.55 As a result, an attractive possibility is that the S4 state contains a Mn]] O species, formed by YZ ? abstracting hydrogen atoms from a water bound to manganese.26,27,30,32–34,41 The proposal for a high-valent Mn]] O-containing intermediate assumes that the manganese cluster is oxidized on each S-state advance.Both Ca21 and Cl2 are required for maximal rates of O2 production by PSII 2,56,57 and, primarily, these cofactors have been assigned structural or electrostatic charge-balance roles.12,26,40,57–61 Most mechanistic proposals of water oxidation attempt to include all four relevant cofactors—Ca21, Cl2, the Mn4 cluster and YZ—because they are believed to be in close proximity to one another in the protein matrix.Removal of calcium or chloride produces a state of the OEC that exhibits a “split” radical EPR signal 62 that has been interpreted as arising from a tyrosyl radical interacting with the Mn4-cluster in its S2 state, and has been formulated as S2YZ ?.43,45,55,63–67 Chloride can be competitively replaced by acetate,63,68 to produce an S2YZ ? state, although the eVects of acetate inhibition do not mimic all the eVects of chloride depletion. 63,69,70 Importantly, acetate-inhibited samples display characteristics of both chloride-depleted and calcium-depleted PSII Fig. 3 The S-state cycle proposed by Kok et al.10 The solid arrows represent light-driven reactions and the dashed arrows represent dark reactions.The box around S1 denotes that this is the dark-stable state. The Mn oxidation states are based on UV and X-ray absorption studies. Fig. 4 Structural core of the complex of Blondin et al.25J. Chem. Soc., Dalton Trans., 1999, 1353–1361 1355 preparations,63 suggesting that the roles of chloride and calcium are related. Spectroscopic studies have demonstrated that bound acetate is close to both YZ and the Mn4 cluster,68,71 and could be bridging between them in a similar manner to that observed in Concanavalin A.63,72 The body of biophysical data on the OEC provides evidence for a mechanism in which YZ ? is involved in H-atom abstraction from Mn-bound substrate water molecules ultimately leading to the formation of a Mn]] O species.In section 3, we propose a model that explains these eVects by assigning chloride as a bridging ligand between calcium and the Mn4 cluster, with calcium being involved in an H-bonded network between the Mn4 cluster and YZ as well as being a site for substrate binding.The participation of these cofactors in hydrogen bonds and proton movement is supported by the observation that Mn-, Ca21- or Cl2-depleted PSII samples cannot undergo S-state transitions and/or YZ oxidation at the same temperatures as intact PSII.1,70,73–80 In this way, Ca21 and Cl2 may be intimately involved in the formation of a Mn]] O species in PSII. This review begins with a discussion of oxygen-atom transfer reactions involving M]] O (M = Ru or Mn) species, emphasizing the factors which aVect the reactivity of the oxo group, especially toward nucleophilic attack.Although there is a vast amount of literature on the reactivity of Fe-based oxidation catalysts as cytochrome P450 models (for leading references see 81–83), none of the Fe-based catalysts has been implicated in O–O bond formation leading to O2 production so they are not discussed in any detail here. The review concludes with a model for photosynthetic water oxidation that combines the biophysical and model chemistry results.We present our proposal of the S-state cycle which involves a MnV]] O as a key intermediate in O–O bond formation. In addition, we suggest roles for both Ca21 and Cl2 and consider why these cofactors are speci- fically required by the OEC. Finally, we support the proposal that the key step in photosynthetic water oxidation may be nucleophilic attack of a Ca21-bound hydroxide/water on an electrophilic MnV]] O.26,41 2.0 Inorganic oxidation catalysts The role of activated metal-oxo species in oxidation chemistry is well established in both biological and inorganic model systems.Such species are capable of oxidizing a wide variety of substrates including phosphines, amines, sulfides, alkenes and alkanes.84–91 Scheme 1 shows a simplified view of how a metal-oxo species may react with a potential nucleophile, using a sulfide as an example.The first example, with a one-electron ratedetermining step, requires a radical intermediate and the second example, with a two-electron rate-determining step, can be thought of as an SN2 reaction involving nucleophilic attack Scheme 1 Reactions of M]] O with nucleophiles highlighting the diVerences between one- and two-electron processes. on an electrophilic oxo. The mechanism of a particular oxidation is strongly dependent on the nature of both the oxidant and the substrate.In this review, we argue that in the case of water oxidation by a Mn]] O, a two-electron, or SN2-like, mechanism would predominate. In the first part of this section, we discuss oxidations involving Ru]] O complexes. These are very well characterized and so serve as excellent models for M]] O reactivity. In the second part, we look at reactions that are thought to proceed via Mn]] O intermediates and discuss their mechanisms by analogy to the better understood Ru chemistry. 2.1 Ruthenium Despite their reactivity, a number of Ru]] O complexes with a variety of coordination environments have been isolated and structurally characterized.92 Ru]] O complexes (Ru = 14 to 17) can oxidize organic substrates 87,88,92 and, most importantly for our interests, have been identified as key intermediates in homogeneous catalysis of water oxidation.93 Studies on electronic eVects governing the reactivity of Ru oxidation catalysts have shown that as the basicity of the ligands to the Ru decrease, the oxo group becomes more reactive.94,95 A simplified way of considering this is that the positive charge residing on the metal increases and so the Ru–O bond becomes more polarized.92 Acquaye et al.95 studied the oxidation of thioanisoles and methyl phenyl sulfoxides by (O)(P(p-C6H5R)3)RuIV complexes, varying the basicities of both the phosphines on the Ru and of the substrates by changing the para substituents on the phenyl rings.When the substituents on the phosphines were varied, the order of the rates of reactivity was CF3 > F > H > Me > OMe, with an approximately ten-fold rate diVerence across the entire range for both sulfide and sulfoxide oxidation.These results clearly indicate that better donor ligands decrease reactivity, presumably by stabilizing the M]] O complex. The authors also showed that the reaction rates increased linearly with an increase in the E2� 1 of the RuIV/III couple.Similar results were observed for substrate reactivity, i.e. the more basic the substrate, the faster it reacted. A comparison of sulfide and sulfoxide reactivity led to some interesting conclusions. Thioanisoles reacted with the Ru]] O complexes approximately 100 times faster than the corresponding sulfoxides. For instance, the second-order rate constant for thioanisole oxidation was 2.30 M21 s21 compared to 0.06 M21 s21 for methyl phenyl sulfoxide, and this is expected as sulfides are more basic than sulfoxides. However, detailed kinetic analyses of the reactions suggested that the two substrates react via diVerent mechanisms.A deuterium-isotope eVect of kH/ kD = 1.14 was measured comparing thioanisole and methyl-d3 phenyl sulfide vs. a kH/kD = 0.64 for the equivalent sulfoxides. This result is consistent with a single-electron transfer mechanism for sulfide oxidation and an SN2-like mechanism for sulfoxide oxidation. The assignment was further supported by Hammett plots on the rates of oxidations when the basicities of the phosphine ligands and substrate were varied.One interpretation of these results is that the rate-determining step in sulfoxide oxidation involves nucleophilic attack on the oxo group on the metal. This change in mechanism r oxidation of sulfides vs. sulfoxides would then be explained by the unfavourable energetics of the one-electron oxidation of sulfoxides (for the one-electron oxidation vs.SCE: E0 thioanisole = 1.45 V,95 E0 methylphenylsulfoxide = 2.30 V96). An important conclusion is that the mechanism—nucleophilic attack vs. radical formation—is decided by the redox properties of the substrate. 2.1.1 O–O bond formation: O2 evolution. The most characterized system for catalytic homogeneous water oxidation is [(bpy)2(H2O)RuORu(H2O)(bpy)2]41 (bpy = 2,29-bipyridine), which was first reported by Gersten et al.97 in 1982, and which1356 J. Chem. Soc., Dalton Trans., 1999, 1353–1361 Fig. 5 Mechanisms to explain 18O incorporation into [(bpy)2(H2O)RuORu(H2O)(bpy)2]41. Adapted from ref. 5. can oxidize water both electrochemically,97 and chemically, using Co31 or Ce41 as primary oxidants.93,98 Subsequent studies have shown that the key intermediate involved in O–O bond formation is a RuV]] O species.93 Single-turnover labeling studies using H2 18O-Ru and unlabeled solvent water, where the O2 formed was analysed using mass spectrometry, were inconclusive as to the precise mechanism of O–O bond formation. Geselowitz and Meyer 98 reported a 36O2 : 34O2 : 32O2 ratio of 13 : 64 : 23, whereas a similar study by Hurst et al.93 gave a 34O2 : 32O2 ratio of approximately 50 : 50, and no 36O2, and also showed that the aqua ligands on the Ru(III) complex exchanged slowly on the timescale of the catalysis.It should be noted that the experiments were run under diVerent conditions. Fig. 5 shows possible mechanisms that are consistent with the results of the labeling studies,5,37,93,98 and multiple pathways must be invoked to account for all the products.Importantly, 34O2 formation requires that one oxygen atom originate from the solvent and the other from a ligand to Ru, and this is consistent with a pathway involving an attack of solvent water on a terminal oxo ligand. 2.2 Manganese In contrast to the Ru chemistry, the detection of Mn]] O catalytic intermediates has been a challenge. Manganates excepted, there are only four structurally characterized Mn]] O complexes,99–102 only one of which has shown reactivity.102 The reactive system was developed by engineering a ligand that could coordinate cations close to the Mn(V)]] O, thereby tuning the electrophilicity of the terminal oxo.102 For instance, adding 5 equivalents of Sc31 to the complex increased the rate of Ph3P oxidation by a remarkable three orders of magnitude compared to the system in the absence of an added cation.There have been a few reports of spectroscopic characterization of Mn]] O intermediates. The first was published by Groves and Stern who synthesized three porphyrin MnIV]] O complexes, (TMP)Mn(O), [(TMP)Mn(O)(OH)] and (TPP)- (Mn)(O) (TMP = 5,10,15,20-tetramesitylporphyrin, TPP = tetraphenylporphyrin),103,104 and spectroscopically characterized them. The Mn]] O group is generated by reacting the appropriate MnIII complex with m-CPBA (meta-chloroperoxybenzoic acid). The presence of Mn]] O was confirmed with 18O labeling and vibrational spectroscopy. A transient species with a strong visible absorbance at approximately 420 nm which cannot be attributed to either MnIV, MnIII or a MnIV]] O-porphyrin radical cation has been seen in the m-CPBA systems, as well as in reaction mixtures using potassium peroxomonosulfate as the oxidant.105 Accordingly, this has been assigned as a MnV]] O.The presence of Mn]] O’s in manganese-porphyrin oxidation chemistry has also been inferred from incorporation of 18O into the product from labeled water in olefin epoxidation,105–108 proposed to proceed by the mechanism shown in Fig. 6. [(salen)MnIII]1 (salen = N,N9-bis(salicylidene)ethylenediamine di-anion) Catalysed epoxidations are thought to require a MnV]] O species analogous to that seen in the porphyrin systems. 109,110 Recently, Feichtinger and Plattner 111 used ES–MS to study the reaction between [(salen)MnIII]1 and iodosobenzene, and saw a m/z peak that was attributable to a MnV]] O complex.This is the first direct evidence for this intermediate in the salen system. Both the salen and porphyrin systems have been well studied, and their reactivities follow broadly the trends seen for RuJ. Chem. Soc., Dalton Trans., 1999, 1353–1361 1357 Fig. 6 Mechanism of incorporation of 18O from solvent water into epoxides formed from the reaction of MnV]] O(porphyrin) species with olefins. Based on refs. 105 and 107.catalysts, that is to say, less stabilizing ligands lead to more reactive intermediates. Chellamani et al.110,112 looked at the electronic eVects on the oxidation of thioanisoles by [(salen)- MnIII]1, analogous to that published for (O)(P(p-C6H5R)3)RuIV complexes.95 As with Ru, it was found that the less basic the ligand, the more reactive the catalyst. Further, Hammett plots similar to those obtained for (O)(P(p-C6H5R)3)RuIV complexes were reported in which both the salen ligand and substrate basicities were varied, suggesting that [(salen)MnIII]1 and (O)(P(p-C6H5R)3)RuIV both oxidize sulfides by analogous mechanisms, i.e.single electron transfer from a M]] O intermediate. Unfortunately, the manganese system did not oxidize the sulfoxides to sulfones under the reported conditions so no information is available on whether a two-electron reaction would predominate here. These electronic eVects manifest themselves dramatically in Jacobsen’s epoxidation catalyst (an enantioselective version of the [(salen)MnIII]1 complex) which shows a direct correlation between electronic eVects and selectivity, i.e.the more reactive the catalyst, the less selective the oxidation.113 In both of these cases, an analogy to the Ru chemistry can be drawn and the oxo on manganese can be thought of as becoming increasingly electrophilic as the other ligands to manganese become less basic. A further, interesting, aspect of Mn-oxidation chemistry is the role of the ligand trans to the active site on the complex.It was found early on that adding a base such as pyridine or imidazole as a co-catalyst increases greatly the reactivity of Mn porphyrins.108,114 Subsequent proposals suggest that these bases act as p-donors to the Mn]] O LUMO, an antibonding orbital (Fig. 7).86,115 This has the eVect of weakening the Mn]] O bond and so increasing the reactivity of the oxo group by, again, making it more electrophilic.115 As a result, a basic group cis to the active site stabilizes a terminal oxo, whereas a p-donor trans to the active site increases its reactivity. 2.2.1 O–O bond formation: O2 evolution.There have been only a few reports of manganese complexes able to catalyse homogeneous O–O bond-forming reactions that lead to dioxygen evolution (reviewed in references 5 and 37). Naruta et al. have reported homogeneous catalytic water oxidation from a face-to-face Mn(porphyrin) dimer (Fig. 8).116 The active species was suggested to be a MnV]] O dimer, and this would be consistent with the other Mn(porphyrin) catalysts. This mechanism has since been supported by studies involving olefin epoxidation. 117 We have developed a Mn di-m-oxo dimer capable of catalysing O2 evolution from oxone5,118 or sodium hypochlorite.119 The catalyst is ultimately lost by formation of permanganate which is unreactive under these conditions. This is the first example of a di-m-oxo Mn dimer, which is a structural element present in the OEC, that has the relevant functional chemistry.When the reaction with hypochlorite is run in H2 18O the label is incorporated into the dioxygen.119 The ratio 36O2 : 34O2 is consistent with water being the source of the oxygen in the O2 evolved, and we interpret the mechanism to involve exchange of oxygen from water with a Mn]] O species. In analogy to the porphyrin dimer, Fig. 7 Molecular orbitals of a M]] O species showing p-donation of bases into the antibonding orbital.Adapted from Jørgensen and Swanstrøm.115 Fig. 8 Structure of the face-to-face [Mn(porphyrin)]2 complex reported by Naruta et al.1161358 J. Chem. Soc., Dalton Trans., 1999, 1353–1361 Fig. 9 Possible mechanism of O–O bond formation from the reaction between [(Terpy)(H2O)Mn(O)2Mn(OH2)(Terpy)](NO3)3 (Terpy = 2,29:69,20- terpyridine) and sodium hypochlorite.118 we have proposed a mechanism with a MnV]] O intermediate (Fig. 9); one possibility for O–O bond formation is attack of a hydroxide on the oxo ligand. 3.0 A mechanism for photosynthetic water oxidation We have presented a basis for an O–O bond-forming reaction that proceeds via nucleophilic attack on a MnV]] O. In this section, we consider how such a reaction could be promoted by the OEC and consider the implications that the chemistry of M]] O species has on the suggestion that a Mn]] O species is formed during photosynthetic water oxidation. In addition, we tie together observations about Ca21 and Cl2 and propose that their role is to align and tune the reactivity of hydroxide as a nucleophile. This is incorporated into a new structural and mechanistic model that extends the H-atom abstraction model presented by Hoganson et al.32,33 Assembling all of this information, we present a detailed picture of the O–O bond-forming reaction which occurs between the S3 and S0 states of the S-state cycle of the OEC.Fig. 10 shows our model of the intermediates in the S-state cycle of photosynthetic water oxidation focusing on the Mn4 cluster.The structure of the manganese cluster shown is based on the structure proposed by Yachandra et al. from X-ray absorption data.9 However, in Fig. 10, a single terminal Mn ion is proposed to be directly involved in the O–O bond-forming step, with the other three Mn ions serving as a source of oxidizing equivalents and/or as structural elements. Therefore, other structures for the Mn4 cluster are readily accommodated by this model.As suggested by recent EPR results,17–19,120 the oxidation states of the manganese atoms in the S0 state are assigned as MnIIMnIIIMnIV 2 in which one dimer consists of MnIV atoms and the other dimer is a MnIIMnIII pair. As there are no examples of di-m-oxo MnIIMnIII complexes, the MnIIMnIII dimer in the cluster is suggested to be bridged by one m-hydroxo and one m-oxo ligand. Advance from the S0 to the S1 state would be followed by deprotonation of the MnIIMnIII m-hydroxo to form a di-m-oxo MnIIIMnIII dimer; there are three such dimers known from model chemistry.121,122 The role of such protoncoupled steps involving interconversion of m-hydoxo and m-oxo bridges in mediating oxidation potentials have been studied in Mn model complexes,123,124 and analogous reactions involving the OEC may allow for approximately equipotential sequential oxidations of the Mn4 cluster in the early S states.Oxidation of the Mn4 cluster from S1 to the S2 state would involve electron transfer from the manganese cluster to YZ ? together with protonation of YZ ? by a non-substrate species.In these steps, the role of BDE’s in providing suYcient driving force for the reactions may be unimportant because a variety of inhibited states of the OEC, including calcium- and chloride-depleted samples, can advance to the S2 state. Advancement from the S2 to S3 and S3 to S4 states would involve oxidation of the Mn4 cluster together with a proton abstraction from a water ligand, ultimately converting the water ligand to a terminal oxo in the S4 state.Importantly, in this proposal these steps cannot proceed if the proton transfer from the substrate water to YZ ? is inhibited. Fig. 11 shows a scheme of OEC oxidation from the S3 to the S4 state which incorporates interpretations of current biophysical data on Ca21, Cl2, the Mn4 cluster and YZ. In the final step, the S4 state would collapse to the S0 state through nucleophilic attack of hydroxide on a MnV]] O group with the concomitant release of O2.Because advance to S states higher than S2 has been proposed to proceed through proton-coupled electron transfer and because Ca21 and Cl2 have been found to be required for the S2-to-S3-state transition, 69,74 we propose detailed structural models of only the higher S-state transitions. Fig. 12 shows a structural model for the OEC incorporating the proposed roles of Ca21 and Cl2.The geometric center-tocenter distance between YZ ? and the Mn4 cluster is ª9 Å in this model, and the point-dipole distance 125 is within the currently accepted 7–10 Å distance estimate limits.42–45,125 Another feature of this model is that it includes a well-ordered hydrogen-bonded network between Ca21, YZ, the Mn4 cluster, and a histidine residue (D1-His 190, the proton acceptor for YZ upon its oxidation). In this hydrogen-bonded network, calcium serves as an anchor between a hydroxide ligated to the manganese cluster and YZ.Part of the chloride’s role would be to position calciumJ. Chem. Soc., Dalton Trans., 1999, 1353–1361 1359 Fig. 10 Proposed S-state cycle of the OEC in PSII. The Mn4 cluster structure shown is the structure proposed by Yachandra et al.9 Fig. 11 Hypothetical O–O bond-forming steps (S3 to S0 states) of the OEC in PSII showing the positions of Ca21, Cl2, YZ and the Mn4 cluster.The Mn4 cluster is based on the model complex of Blondin et al.25 between YZ and the Mn4 cluster. Because YZ ? abstracts a hydrogen atom from the S2 state to convert it to the S3 state, Ca21 and Cl2 together could create a local structure uniquely designed to make advance to the S3 state energetically and kinetically accessible. Removal of Ca21 or Cl2 would disrupt the H-bonded structure between the Mn4 cluster and YZ, thereby preventing the S2 to S3-state oxidation step. Fig. 12 also shows a histidine ligated to the manganese trans to where the oxo group forms. This is consistent with ESEEM studies which have shown that there is at least one histidine ligated to the Mn4 cluster.126 An aromatic amine trans to the oxo group could play an activating role as shown in Fig. 7. The final step, advance to the S4 state, involves H-atom abstraction to form a MnV]] O. We support the hypothesis that O–O bond formation is achieved by nucleophilic attack on the MnV]] O by a hydroxide ligated to calcium.26,41 This type of mechanism is consistent with the calculations of Siegbahn and Crabtree which suggest that Mn]] O groups are electron deficient.127 It is also consistent with the idea that M]] O species react with substrates that are diYcult to oxidize by one1360 J. Chem.Soc., Dalton Trans., 1999, 1353–1361 electron, such as hydroxide, by a two-electron SN2-like process (see preceding sections). In order for the O–O bond-forming reaction to occur, the nucleophile must be delivered at the same step in which the MnV]] O is created.The proposed mechanism also includes a new functional role for chloride as a bridge that communicates the formation of the MnV]] O to Ca21 in order to activate the hydroxide for nucleophilic attack. We envision that by binding between the Mn4 cluster and Ca, chloride tunes the nucleophilicity of the calcium-hydroxide by responding to changes in the oxidation state of the Mn4 cluster.As the oxidation level of the Mn4 cluster is increased, chloride binds more tightly to it. As a result of this, the chloride is pulled away from calcium, thereby causing calcium to become a stronger Lewis acid. In eVect, advance to the S4 state would change the interaction between calcium and chloride so that a nucleophilic hydroxide would be created in concert with an electrophilic MnV]] O species, to facilitate formation of the O–O bond. 4.0 Summary We have presented a new model for photosynthetic water oxidation that combines inorganic and biophysical studies to support the idea that the O–O bond-forming step involves nucleophilic attack of a calcium-bound hydroxide on an electrophilic MnV]] O species.In our proposal, the reactivity of the terminal oxo is enhanced by a histidine ligated trans to it. Another feature of the model assigns roles for calcium and chloride as mediating proton-coupled electron transfer between the Mn4 cluster and YZ ? by way of a hydrogen-bonded network.Further, having chloride as a bridging ligand between calcium and the Mn4 cluster allows it to control the Lewis acidity of calcium, thereby tuning the reactivity of the substrate hydroxide. Acknowledgements We would like to thank Drs. G. T. Babcock, R. H. Crabtree and P. E. M. Siegbahn for copies of submitted manuscripts. This work was supported by grants from the National Institutes of Health. Fig. 12 A structural model of the OEC showing the proposed hydrogen- bonded network between Ca21, Cl2, the Mn4 cluster and YZ. The model was generated by using Chem 3D and metrical parameters from the Cambridge Structural Database, as described in ref. 125. The blue atom trans to OH on the terminal Mn represents a nitrogen atom of a histidine ligand. References 1 B. A. Diner and G. T. Babcock, in Oxygenic Photosynthesis: The Light Reactions, eds. D. R. Ort and C. F. Yocum, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1996, p. 213. 2 R. J. Debus, Biochim. Biophys. Acta, 1992, 1102, 269. 3 V. K. Yachandra, K. Sauer and M. P. Klein, Chem. Rev., 1996, 96, 2927. 4 R. D. Britt, in Advances in Photosynthesis, Oxygenic Photosynthesis: The Light Reactions, eds. D. R. Ort and C. F. Yocum, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1996, p. 137. 5 J. Limburg, G. W. Brudvig and R. H. Crabtree, in Biomimetic Oxidation Chemistry, ed. B. Meunier, Imperial College Press, Singapore, 1999, in press. 6 K. Wieghardt, Angew. Chem., Int. Ed. Engl., 1994, 33, 725. 7 R. Manchanda, G. W. Brudvig and R. H. Crabtree, Coord. Chem. Rev., 1995, 144, 1. 8 G. W. Brudvig, Adv. Chem. Ser., 1995, 246, 249. 9 V. K. Yachandra, V. J. DeRose, M. J. Latimer, I. Mukerji, K. Sauer and M. P. Klein, Science, 1993, 260, 675. 10 B. Kok, B. Forbush and M. McGloin, Photochem. Photobiol., 1970, 11, 457. 11 J. P. Dekker and H. J. van Gorkom, J. Bioenerg. Biomembr., 1987, 19, 125. 12 L. Iuzzolini, J.Dittmer, W. Doerner, W. Meyer-Klauke and H. Dau, Proceedings of the XIth International Congress on Photosynthesis, ed. G. Garab, Kluwer Academic Publishers, Dordrecht, The Netherlands, in press. 13 T. Ono, T. Noguchi, Y. Inoue, M. Kusunoki, T. Matsushita and H. Oyanagi, Science, 1992, 258, 1335. 14 T. A. Ono, T. Noguchi, Y. Inoue, M. Kusunoki, T. Matsushita and H. Oyanagi, Jpn. J. Appl. Phys. Part 1, 1993, 32, 530. 15 T. A. Roelofs, W. Liang, M. J. Latimer, R. M. Cinco, A. Rompel, J.C. Andrews, K. Sauer, V. K. Yachandra and M. P. Klein, Proc. Natl. Acad. Sci. U.S.A., 1996, 93, 3335. 16 G. C. Dismukes and Y. Siderer, FEBS Lett., 1980, 121, 78. 17 K. A. Åhrling, S. Peterson and S. Styring, Biochemistry, 1997, 36, 13148. 18 J. Messinger, J. H. A. Nugent and M. C. W. Evans, Biochemistry, 1997, 36, 11055. 19 C. Goussias, N. Ioannidis and V. Petrouleas, Biochemistry, 1997, 36, 9261. 20 J. Bonvoisin, G. Blondin, J.-J. Girerd and J.-L. Zimmermann, Biophys.J., 1992, 61, 1076. 21 J. E. Penner-Hahn, P. J. Riggs-Gelasco, E. Yu, P. DeMarois and C. F. Yocum, in Photosynthesis: From Light to Biosphere, ed. P. Mathis, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1995, p. 241. 22 J. E. Penner-Hahn, Structure and Bonding, 1998, 90, 1. 23 M. Zheng and G. C. Dismukes, Inorg. Chem., 1996, 35, 3307. 24 M. Zheng and G. C. Dismukes, in Research in Photosynthesis, ed. N. Murata, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1992, vol. 2, p. 305. 25 G. Blondin, R. Davydov, C. Philouze, M.-F. Charlot, S. Styring, B. Åkermark, J.-J. Girerd and A. Boussac, J. Chem. Soc., Dalton Trans., 1997, 4069. 26 V. L. Pecoraro, M. J. Baldwin, M. T. Caudle, W. Hsieh and N. A. Law, Pure Appl. Chem., 1998, 70, 925. 27 M. R. A. Blomberg, P. E. M. Siegbahn, S. Styring, G. T. Babcock, B. Åkermark and P. Korall, J. Am. Chem. Soc., 1997, 119, 8285. 28 G. W. Brudvig and R. H. Crabtree, Proc. Natl. Acad. Sci. U.S.A., 1986, 83, 4586. 29 G. W. Brudvig and R. H. Crabtree, Prog. Inorg. Chem., 1989, 37, 99. 30 M. L. Gilchrist, G. A. Lorigan and R. D. Britt, Proc. Natl. Acad. Sci. U.S.A., 1995, 92, 9545. 31 J. A. Halfen, S. Mahapatra, E. C. Wilkinson, S. Kaderli, V. G. Young, L. Que, A. D. Zuberbuhler and W. B. Tolman, Science, 1996, 271, 1397. 32 C. W. Hoganson, N. Lydakis-Simantiris, X.-S. Tang, C. Tommos, K. Warncke, G. T. Babcock, B. A. Diner, J. McCracken and S. Styring, Photosynth. Res., 1995, 46, 177. 33 C. W. Hoganson and G. T. Babcock, Science, 1997, 277, 1953. 34 J. Messinger, M. Badger and T. Wydrzynski, Proc. Natl. Acad. Sci. U.S.A., 1995, 92, 3209. 35 D. M. Proserpio, R. HoVmann and G. C. Dismukes, J. Am. Chem. Soc., 1992, 114, 4374. 36 D. M. Proserpio, A. K. Rappe and S. M. Gorun, Inorg. Chim. Acta, 1993, 213, 319. 37 W. Rüttinger and G. C. Dismukes, Chem. Rev., 1997, 97, 1. 38 C. Tommos and G. T. Babcock, Acc. Chem. Res., 1998, 31, 18. 39 J. B. Vincent and G. Christou, Inorg.Chim. Acta, 1987, 136, L41.J. Chem. Soc., Dalton Trans., 1999, 1353–1361 1361 40 T. Wydrzynski, W. Hillier and J. Messinger, Physiol. Plant., 1996, 96, 342. 41 V. A. Szalai and G. W. Brudvig, Am. Sci., 1998, 86, 542. 42 P. Dorlet, M. Di Valentin, G. T. Babcock and J. L. McCracken, J. Phys. Chem. B, 1998, 102, 8239. 43 K. V. Lakshmi, S. S. Eaton, G. R. Eaton, H. A. Frank and G. W. Brudvig, J. Phys. Chem. B, 1998, 8327. 44 J. M. Peloquin, K. A. Campbell and R. D. Britt, J.Am. Chem. Soc., 1998, 120, 6840. 45 V. A. Szalai, H. Kühne, K. V. Lakshmi, G. R. Eaton, S. S. Eaton and G. W. Brudvig, Biophys. J., 1998, 74, A75. 46 A.-M. A. Hays, I. R. Vassiliev, J. H. Golbeck and R. J. Debus, Biochemistry, 1998, 37, 11352. 47 B. A. Diner, P. J. Nixon and J. W. Farchaus, Curr. Opin. Struct. Biol., 1991, 1, 546. 48 P. J. Nixon and B. A. Diner, Biochem. Soc. Trans., 1994, 22, 338. 49 R. A. RoVey, D. M. Kramer, Govindjee and R. T. Sayre, Biochim. Biophys.Acta, 1994, 1185, 257. 50 C. Tommos, X.-S. Tang, K. Warncke, C. W. Hoganson, S. Styring, J. McCracken, B. A. Diner and G. T. Babcock, J. Am. Chem. Soc., 1995, 117, 10325. 51 M. J. Baldwin and V. L. Pecoraro, J. Am. Chem. Soc., 1996, 118, 11325. 52 M. T. Caudle and V. L. Pecoraro, J. Am. Chem. Soc., 1997, 119, 3415. 53 J. Lind, X. Shen, T. E. Eriksen and G. Merényi, J. Am. Chem. Soc., 1990, 112, 479. 54 F. G. Bordwell and J.-P. Cheng, J. Am. Chem. Soc., 1991, 113, 1736. 55 V.A. Szalai and G. W. Brudvig, Biochemistry, 1996, 35, 15080. 56 C. F. Yocum, Biochim. Biophys. Acta, 1991, 1059, 1. 57 P. H. Homann, Biochim. Biophys. Acta, 1988, 934, 1. 58 F. Rappaport, M. Blanchard-Desce and J. Lavergne, Biochim. Biophys. Acta, 1994, 1184, 178. 59 P. O. Sandusky and C. F. Yocum, Biochim. Biophys. Acta, 1986, 849, 85. 60 A. W. Rutherford, Trends Biochem. Sci., 1989, 14, 227. 61 J. Tso, M. Sivaraja and G. C. Dismukes, Biochemistry, 1991, 30, 4734. 62 A. Boussac and A.W. Rutherford, Biochemistry, 1988, 27, 3476. 63 V. A. Szalai, H. Kühne, K. V. Lakshmi and G. W. Brudvig, Biochemistry, 1998, 37, 13549. 64 V. A. Szalai and G. W. Brudvig, Biochemistry, 1996, 35, 1946. 65 D. J. MacLachlan and J. H. A. Nugent, Biochemistry, 1993, 32, 9772. 66 X.-S. Tang, D. W. Randall, D. A. Force, B. A. Diner and R. D. Britt, J. Am. Chem. Soc., 1996, 118, 7638. 67 B. J. Hallahan, J. H. A. Nugent, J. T. Warden and M. C. W. Evans, Biochemistry, 1992, 31, 4562. 68 H. Kühne, V. A. Szalai and G. W. Brudvig, Biochemistry, submitted for publication. 69 H. Wincencjusz, H. J. van Gorkom and C. F. Yocum, Biochemistry, 1997, 36, 3663. 70 K. Lindberg and L.-E. Andréasson, Biochemistry, 1996, 35, 14259. 71 D. A. Force, D. W. Randall and R. D. Britt, Biochemistry, 1997, 36, 12062. 72 J. E. Penner-Hahn, R. M. Fronko, V. L. Pecoraro, C. F. Yocum, S. D. Betts and N. R. Bowlby, J. Am. Chem. Soc., 1990, 112, 2549. 73 M. Baumgarten, J. S. Philo and G.C. Dismukes, Biochemistry, 1990, 29, 10814. 74 A. Boussac, J.-L. Zimmermann and A. W. Rutherford, FEBS Lett., 1990, 277, 69. 75 A. Boussac and A. W. Rutherford, J. Biol. Chem., 1994, 269, 12462. 76 R. Damoder, V. V. Klimov and G. C. Dismukes, Biochim. Biophys. Acta, 1986, 848, 378. 77 K. Kalosaka, W. F. Beck, G. W. Brudvig and G. Cheniae, in Current Research in Photosynthesis, ed. M. BaltscheVsky, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1990, vol. 1, p. 721. 78 M. Sivaraja, J. Tso and G. C. Dismukes, Biochemistry, 1989, 28, 9459. 79 P. van Vliet, A. Boussac and A. W. Rutherford, Biochemistry, 1994, 33, 12998. 80 P. van Vliet and A. W. Rutherford, Biochemistry, 1996, 35, 1829. 81 S. Ferguson-Miller and G. T. Babcock, Chem. Rev., 1996, 96, 2889. 82 M. Sono, M. P. Roach, E. D. Coulter and J. H. Dawson, Chem. Rev., 1996, 96, 2841. 83 Cytochrome P450. Structure, Mechanism and Biochemistry, ed. P. R. Ortiz de Montellano, Plenum, New York, 1996. 84 R. H. Holm, Coord. Chem. Rev., 1990, 100, 183. 85 R. S. Drago, Coord. Chem. Rev., 1992, 117, 185. 86 K. A. Jørgensen, Chem. Rev., 1989, 89, 431. 87 F. Montanari and L. Casella, in Metalloporphyrins Catalyzed Oxidations, eds. R. Ugo and B. R. James, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1994. 88 W. P. GriYth, Chem. Soc. Rev., 1992, 21, 179. 89 Selective Hydrocarbon Activation: Principle and Progress, ed. J. A. Davies, VCH, New York, 1994. 90 R. A. Sheldon and J.K. Kochi, Metal Catalyzed Oxidation of Organic Compounds, Academic Press, New York, 1991. 91 Metalloporphyrins in Catalytic Oxidations, ed. R. A. Sheldon, Marcel Dekker, New York, 1994. 92 C.-M. Che, Pure Appl. Chem., 1995, 67, 225. 93 J. K. Hurst, J. Zhou and Y. Lei, Inorg. Chem., 1992, 31, 1010. 94 J. G. Muller, J. H. Acquaye and K. J. Takeuchi, Inorg. Chem., 1992, 31, 4552. 95 J. H. Acquaye, J. G. Muller and K. J. Takeuchi, Inorg. Chem., 1993, 32, 160. 96 M. Chanon and S.Andre, in Chemistry of Sulfones and Sulfoxides, eds. S. Patai, Z. Rappoport and C. J. M. Stirling, John Wiley and Sons Ltd, 1988, p. 1048. 97 S. W. Gersten, G. J. Samuels and T. J. Meyer, J. Am. Chem. Soc., 1982, 104, 4029. 98 D. Geselowitz and T. J. Meyer, Inorg. Chem., 1990, 29, 3894. 99 T. J. Collins and S. W. Gordon-Wylie, J. Am. Chem. Soc., 1989, 111, 4511. 100 T. J. Collins, R. D. Powell, C. Slebodnick and E. S. UVelman, J. Am. Chem. Soc., 1990, 112, 899. 101 F. M.MacDonnell, N. L. P. Fackler, C. Stern and T. V. O’Halloran, J. Am. Chem. Soc., 1994, 116, 7431. 102 C. G. Miller, S. W. Gordon-Wylie, C. P. Horwitz, S. A. Strazisar, D. K. Peraino, G. R. Clark, S. T. Weintraub and T. J. Collins, J. Am. Chem. Soc., 1998, 120, 11540. 103 J. T. Groves and M. K. Stern, J. Am. Chem. Soc, 1987, 109, 3812. 104 J. T. Groves and M. K. Stern, J. Am. Chem. Soc., 1988, 110, 8628. 105 J. T. Groves, J. Lee and S. S. Marla, J. Am. Chem. Soc., 1997, 119, 6269. 106 R. Song, A. Sorokin, J. Bernadou and B. Meunier, J. Org. Chem., 1997, 62, 673. 107 J. Bernadou, A.-S. Fabiano, A. Robert and B. Meunier, J. Am. Chem Soc., 1994, 116, 9375. 108 B. Meunier, E. Guilmet, M.-E. De Carvahlo and R. Poilblanc, J. Am. Chem. Soc., 1984, 106, 6668. 109 D. L. Hughes, G. B. Smith, J. Liu, G. C. Dezeny, C. H. Senanayake, R. D. Larsen, T. R. Verhoeven and P. J. Reider, J. Org. Chem., 1997, 62, 2222. 110 A. Chellamani, N. M. I. Alhaji and S. Rajagopal, J. Chem. Soc., Perkin Trans. 2, 1997, 299. 111 D. Feichtinger and D. A. Plattner, Angew. Chem., Int. Ed. Engl., 1997, 36, 1718. 112 A. Chellamani, N. I. Alhaji, S. Rajagopal, R. Sevvel and C. Srinivasan, Tetrahedron, 1995, 51, 12677. 113 M. Paluki, N. S. Finney, P. J. Pospisil, M. L. Gueler, T. Ishida and E. N. Jacobsen, J. Am. Chem. Soc., 1998, 120, 948. 114 O. Bortolini and B. Meunier, J. Chem. Soc., Perkin Trans. 2, 1984, 1967. 115 K. A. Jørgensen and P. Swanstrøm, Acta Chem. Scand., 1988, 43, 822. 116 Y. Naruta, M. Sasayama and T. Sasaki, Angew. Chem., Int. Ed. Engl., 1994, 33, 1839. 117 K. Ichihara and Y. Naruta, Chem. Lett., 1998, 185. 118 J. Limburg, G. W. Brudvig and R. H. Crabtree, J. Am. Chem. Soc., 1997, 119, 2761. 119 J. Limburg, J. S. Vrettos, L. M. Liable-Sands, A. L. Rheingold, R. H. Crabtree and G. W. Brudvig, Science, 1999, 283, 1524. 120 J. Sarrou, N. Ioannidis, Y. Deligiannakis and V. Petrouleas, Biochemistry, 1998, 37, 3581. 121 P. A. Goodson, A. R. Oki, J. Glerup and D. J. Hodgson, J. Am. Chem. Soc., 1990, 112, 6248. 122 N. Kitajima, U. P. Singh, H. Amagai, M. Osawa and Y. Moro-oka, J. Am. Chem. Soc., 1991, 113, 7757. 123 H. H. Thorp, J. E. Sarneski, G. W. Brudvig and R. H. Crabtree, J. Am. Chem. Soc., 1989, 111, 9249. 124 R. Manchanda, H. H. Thorp, G. W. Brudvig and R. H. Crabtree, Inorg. Chem., 1992, 31, 4040. 125 V. A. Szalai, D. A. Stone and G. W. Brudvig, Proceedings of the XIth International Congress on Photosynthesis, ed. G. Garab, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, in press. 126 X.-S. Tang, B. A. Diner, B. S. Larsen, M. L. Gilchrist, G. A. Lorigan and R. D. Britt, Proc. Natl. Acad. Sci. U.S.A., 1994, 91, 704. 127 P. E. M. Siegbahn and R. H. Crabtree, J. Am. Chem. Soc., 1999, 121, 117. Paper 8/07583B

ISSN:1477-9226    

DOI:10.1039/a807583b

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1999, 1363–1364 1363 High-nuclearity cobaltadendrimers Edwin C. Constable,* Oliver Eich and Catherine E. Housecroft* Institut für Anorganische Chemie, Universität Basel, Spitalstrasse 51, CH-4056 Basel, Switzerland. E-mail: housecroft@ubaclu.unibas.ch Received 27th January 1999, Accepted 16th March 1999 New dendritic polyalkynes have been prepared and reacted with [Co2(CO)8] to give cobaltadendrimers containing up to 40 cobalt atoms.Metalladendrimers are of intrinsic structural and synthetic interest 1–3 and oVer potential applications as light-collecting devices, information storage devices and polyfunctional catalysts. We have prepared compounds in which the metals are part of the backbone dendritic connectivity,1 or in which they decorate the dendrimer.4 The generation of C2Co2(CO)6 clusters from alkynes is a facile method for the introduction of multiple metal centres which has not been widely used in metalladendrimer chemistry.5 We now report an extension of our previous work (which led to a starburst decorated with C2Co2- (CO)6 through to the third generation) 6 to genuinely dendritic systems. Our synthetic strategy involves the preparation of a starburst or dendritic polyalkyne and post-functionalization with [Co2- (CO)8].6 The doubly-protected compound 1 was prepared in 53% yield † by the sequential reaction of 1,3,5-tribromobenzene with two equivalents of (TIPS)C]] ] CH (TIPS = triisopropylsilyl) and a three-fold excess of (TMS)C]] ] CH (TMS = trimethylsilyl) in each case in the presence of [(Ph3P)2PdCl2], CuI and NEt3, followed by alkaline cleavage of the TMS group.The use of the base-stable TIPS protecting group is critical to the success of this strategy and allows the preparation of asymmetrically substituted derivatives. The reaction of C(p-C6H4I)4 7 with an excess of 1 under similar palladium(II)-catalysed coupling conditions followed by deprotection of 2 with [nBu4N]F in THF yielded the intensely luminescent (lmax 352 nm) dendritic dodecaalkyne 3 as white crystals.‡ Upon stirring 3 with [Co2(CO)8] in CH2Cl2, a dark coloured solution was obtained from which the dendritic tetracosacobalt complex 4 was isolated as deep red crystals in 28% yield.§ Linear extension of these systems proved to be facile.The reaction of the polyalkyne 3 with p-IC6H4C]] ] C(TMS) under standard Pd-coupling conditions yielded the protected dendritic icosaalkyne 5 as yellow crystals.Subsequent basic deprotection gave icosaalkyne 6, the reaction of which with an excess of [Co2(CO)8] produced the deep red crystalline tetracontacobalt compound 7 ¶ (Fig. 1). This latter compound and all others described were characterized by the normal spectroscopic and analytical methods adopted for ‘small molecules’ even though modelling indicates that 7 has a diameter of ª35 nm. We are currently investigating the chemical and structural aspects of these novel high-nuclearity species.Fig. 1 Proposed structure of compound 7.1364 J. Chem. Soc., Dalton Trans., 1999, 1363–1364 Acknowledgements This work was supported by the Schweizerischer Nationalfonds zur Förderung der wissenschaftlichen Forschung and the University of Basel. We thank Dr G. Scherer for assistance with the NMR spectroscopic experiments. Notes and references † 1: 1,3,5-Br3C6H3 (1.35 g, 4.28 mmol), (TIPS)C]] ] CH (1.64 g, 8.99 mmol), CuI (81.5 mg, 0.43 mmol) and [(PPh3)2PdCl2] (300 mg, 0.43 mmol) were stirred in NEt3 (50 ml) for 4 h at 35 8C.Treatment with (TMS)C]] ] CH (1.26 g, 12.8 mmol) under analogous conditions to those described above (reaction time 12 h), followed by chromatographic work-up gave the protected intermediate; this was dissolved in THF (120 ml) and 1 M NaOH (150 ml) added; the solution stirred for 3.5 h. After extraction, the residue was purified by column chromatography to give a colourless oil (1.06 g; 53%). 1H NMR (250 MHz, CDCl3) d 7.52– 7.50 (m, 3H, Ar), 3.08 (s, 1H, CCH), 1.12 (s, 42H, TIPS); 13C NMR (75 MHz, CDCl3) d 135.2, 124.2, 122.6, 105.1, 92.4, 82.0, 78.3, 18.7 (TIPS), 11.3 (TIPS); MS (MALDI-TOF) m/z 502 [M 1 K]1, 486 [M 1 Na]1. ‡ 3: compound 1 (0.37 g, 0.80 mmol), C(p-C6H4I)4 (0.16 g, 0.19 mmol), CuI (10.8 mg, 0.06 mmol) and [(Ph3P)2PdCl2] (40.0 mg, 0.06 mmol) were stirred in NEt3 (10 ml) and DMF (20 ml) for 70 h at 35 8C. Chromatographic work-up gave 2 as yellow crystals (0.35 g, 84.5%).Deprotection using [nBu4N]F (1.60 mmol) in THF (50 ml, room temperature, 4 h) yielded 3 as white crystals (62 mg, 42%). 1H NMR (300 MHz, CDCl3) d 7.61 (d, J 1.4 Hz, 8H), 7.56 (t, J 1.5 Hz, 4H), 7.45 (d, J 8.6 Hz, 8H), 7.21 (d, J 8.5 Hz, 8H), 3.11 (s, 8H, CCH); 13C NMR (75 MHz, CDCl3) d 146.2, 135.1, 131.3, 130.9, 124.0, 122.9, 120.8, 90.3, 87.9, 81.8, 78.5, 48.4; MS (MALDI-TOF) m/z 1828 [2M]1, 914 [M]1. § 4: alkyne 3 (54.3 mg, 0.06 mmol) and Co2(CO)8 (0.49 g, 1.43 mmol) were stirred in CH2Cl2 (10 ml) for 1.5 h at room temperature, the solvent removed, and the residue purified by column chromatography to give a deep red crystalline solid (71.9 mg, 28%).IR (KBr disc, cm21) nCO 2093 s, 2055 vs, 2020 vs; 1H NMR (250 MHz, CDCl3) d 7.67 (d, J 2.0 Hz, 8H), 7.57 (d, J 8.3 Hz, 8H), 7.50 (t, J 1.7 Hz, 4H), 7.18 (d, J 8.8 Hz, 8H), 6.37 (s, 8H, Hcluster); 13C NMR (101 MHz, CDCl3) d 199.0 (CO), 145.7, 140.4, 139.3, 135.9, 131.7, 130.4, 129.2, 128.3, 91.4, 90.3, 88.5, 72.8, 64.8; MS (MALDI-TOF) m/z 4317 [M 2 CO]1.¶ 6 and 7: alkyne 3 (62 mg, 68 mmol), p-IC6H4C]] ] C(TMS) (0.24 g, 0.82 mmol), CuI (10.4 mg, 0.05 mmol) and [(PPh3)2PdCl2] (38.1 mg, 0.05 mmol) were stirred in dry NEt3 (5 ml) for 100 h at 39 8C. Chromatographic work-up gave compound 5 as yellow crystals (67 mg, 43%); it was dissolved in THF (30 ml), and 1 M NaOH (30 ml) added; the solution was stirred for 4 h. Water was added and after extraction with CH2Cl2, the residue was purified by column chromatography to give 6 as white crystals (21.3 mg, 43%).Reaction of 6 (15.4 mg, 8.98 mmol) with [Co2(CO)8] under the same conditions as for 4 gave 7 as deep red crystals (27 mg, 41%). 6: 1H NMR (400 MHz, CDCl3) d 7.65 (br s, 12H), 7.48–7.46 (m, 40H), 7.24 (d, J 8.6 Hz, 8H), 3.19 (s, 8H, CCH); 13C NMR (75 MHz, CDCl3) d 146.2, 134.3, 132.1, 131.5, 131.3, 130.9, 124.0, 123.8, 123.1, 122.3, 120.9, 90.3, 90.1, 89.6, 88.1, 83.1, 79.2; MS (MALDI-TOF) m/z 1713 [M]1. 7: IR (KBr disc, cm21) nCO 2091 s, 2055 vs, 2020 vs; 1H NMR (400 MHz, CD2Cl2) d 7.79 (d, J 1.6 Hz, 8H), 7.72 (t, J 1.7 Hz, 4H), 7.62 (d, J 8.6 Hz, 8H), 7.56 (d, J 8.4 Hz, 16H), 7.48 (d, J 8.5 Hz, 16H), 7.24 (d, J 8.6 Hz, 8H), 6.46 (s, 8H, CCH); 13C NMR (101 MHz, CD2Cl2) d 199.8 (CO), 199.4 (CO), 199.2 (CO), 146.2, 140.9, 140.6, 138.4, 137.9, 136.3, 131.9, 131.2, 129.9, 129.4, 128.8, 91.8, 91.7, 90.9, 90.8, 89.6, 73.4, 69.5; MS (MALDI-TOF) m/z 7298 [M 2 5CO]1, 7144 [M 2 10 CO]1. 1 E. C. Constable, Chem. Commun., 1997, 1073. 2 G. R. Newkome, C. N. Moorefield and F. Vögtle, Dendritic Molecules, Wiley-VCH, Weinheim, 1996. 3 V. Balzani, A. Juris, M. Venturi, S. Campagna and S. Serroni, Chem. Rev., 1996, 96, 759. 4 E. C. Constable, C. E. Housecroft and L. A. Johnston, Inorg. Chem. Commun., 1998, 1, 68. 5 D. Seyferth, T. Kugita, A. L. Rheingold and G. A. P. Yap, Organometallics, 1995, 14, 5362. 6 E. C. Constable, O. Eich, C. E. Housecroft and L. A. Johnston, Chem. Commun., 1998, 2661. 7 M. Simard, D. Su and J. D. Wuest, J. Am. Chem. Soc., 1991, 113, 4696. Communication 9/02093D

ISSN:1477-9226    

DOI:10.1039/a902093d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1999, 1365–1368 1365 [Me2Si] Ansa bridged complexes of permethyltitanocene: synthesis and structural characterization of fulvene derivatives with trialkylidenemethane character † Hyosun Lee, JeVrey B. Bonanno, Tony Hascall, Joseph Cordaro, Juliet M. Hahn and Gerard Parkin * Department of Chemistry, Columbia University, New York, New York 10027, USA Received 6th January 1999, Accepted 1st March 1999 A series of permethylated [Me2Si] ansa bridged titanocene complexes has been synthesized and structurally characterized by X-ray diffraction; the dialkyl complexes [Me2Si(C5Me4)]TiR2 are thermally unstable towards elimination of alkane (RH), thereby yielding fulvene derivatives [Me2Si(C5Me4)(C5Me3CH2)]TiR.We have recently reported how incorporation of the [Me2Si] ansa bridge may have a profound eVect on the chemistry of the permethylzirconocene system by increasing the electrophilicity of the metal center.1 In this paper, we describe chemistry of the corresponding titanium system, [Me2Si(C5Me4)2]TiX2, which includes (i) an unusual coupling reaction to form a biphenyl derivative and (ii) the synthesis of fulvene derivatives that possess trialkylidenemethane character.The dichloride [Me2Si(C5Me4)2]TiCl2 2 provides a convenient entry to a series of ansa titanocene complexes (Scheme 1).3 For example, reduction of [Me2Si(C5Me4)2]TiCl2 with Mg(Hg) under an atmosphere of CO gives the dicarbonyl, [Me2Si- (C5Me4)2]Ti(CO)2, while reactions with RLi (R = Me, Ph, CH2SiMe3) and LiNC4H4 yield [Me2Si(C5Me4)2]TiR2 4 and [Me2Si(C5Me4)2]Ti(NC4H4)2, respectively.Most interestingly, however, the reaction of [Me2Si(C5Me4)2]TiCl2 with excess PhLi results in C–C coupling and the formation of the biphenyl-2,29- diyl complex, [Me2Si(C5Me4)2]Ti[(C6H4)2] (Scheme 1).5 The dialkyl complexes [Me2Si(C5Me4)2]TiR2 may be used as † Supplementary data available: tables of analytical and spectroscopic data, and preparative details.For direct electronic access see http:// www.rsc.org/suppdata/dt/1999/1365/, otherwise available from BLDSC (No. SUP 57522, 19 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton). precursors to other derivatives via reaction at the Ti–C bond; for example, [Me2Si(C5Me4)2]Ti(OAc)2 and [Me2Si(C5Me4)2]Ti- (NC4H4)Me, accompanied by elimination of methane, are obtained upon treatment of [Me2Si(C5Me4)2]TiMe2 with AcOH and C4H4NH, respectively.6 In the absence of a substrate, the dialkyls [Me2Si(C5Me4)2]TiR2 (R = Me, Ph, CH2SiMe3) eliminate alkane (RH) to give the corresponding fulvene derivatives [Me2Si(C5Me4)(C5Me3CH2)]TiR (Scheme 2).7,8 The molecular structure of the phenyl derivative [Me2Si(C5Me4)- (C5Me3CH2)]TiPh has been determined by X-ray diVraction Scheme 2 Scheme 11366 J.Chem. Soc., Dalton Trans., 1999, 1365–1368 (Fig. 1),‡ thereby confirming that metallation occurs adjacent to the ansa bridge.Furthermore, the diVraction study indicates that the fulvene moiety is coordinated in a most asymmetric manner, with individual Ti–C bond lengths ranging from 2.16 Å to 2.57 Å, with the shortest being that adjacent to the methylene group, and the longest being those most distant from the methylene group.9 For comparison, the longest Ti–C bond length in the diphenylfulvene complex (C5H4CPh2)2Ti is only 2.40 Å.10,11 The fulvene moiety of [Me2Si(C5Me4)(C5Me3CH2)]- TiPh is thus slipped, such that the principal interaction is with four carbon atoms, namely C11, C12, C13 and C32 (Fig. 1). In this regard, the complex is perhaps better represented as a trialkylidenemethane, i.e. C(CR2)3, derivative,12 with the bonding supplemented by a weak olefin (C14 and C15) interaction (see Fig. 2). Supporting this notion, trialkylidenemethane character is indicated by the observation that the Ti–C12 bond is eVectively perpendicular to the C11–C13–C32 plane, passing through its centroid.13 It is important to note that this view of the metal–fulvene interaction as possessing trialkylidenemethane character has not previously been oVered as a description for fulvene complexes; specifically, bonding in transition metal fulvene complexes is typically only discussed Fig. 1 Selected bond lengths (Å): Ti–C11 2.250(4), Ti–C12 2.156(4), Ti–C13 2.395(4), Ti–C14 2.572(4), Ti–C15 2.491(4), Ti–C21 2.294(4), Ti–C22 2.370(4), Ti–C23 2.459(4), Ti–C24 2.447(4), Ti–C25 2.336(4), Ti–C32 2.328(5), C12–C32 1.408(6).in terms of whether the interaction is better described as coordination of a neutral h6-fulvene (i.e. a “diene-olefin”) or as that of a dianionic h5,h1-ligand (Fig. 2).10,11,14 The trialkylidenemethane- olefin view of the interaction, therefore, provides an alternative description that may more appropriately describe the bonding in certain fulvene complexes.15 Deuterium kinetic isotope and labeling studies 16 indicate that elimination of methane from [Me2Si(C5Me4)2]TiMe2 occurs via rate determining a-H abstraction of one of the titanium methyl groups giving {[Me2Si(C5Me4)2]Ti]] CH2}, followed by rapid transfer of a ring methyl hydrogen to the methylene group (Scheme 2).Such a mechanism is precedented by Bercaw’s detailed study of methane elimination from Cp*2TiMe2 to give Cp*(C5Me4CH2)TiMe.17 In view of the structural changes introduced by incorporating a [Me2Si] ansa bridge (see below), it is perhaps surprising that the rate of methane elimination is virtually unaVected by this modification;18 the primary kinetic isotope for elimination from [Ti(CD3)2] is, however, considerably greater for the ansa system [kH/kD = 5.16 at 100 8C] than that for the permethyltitanocene system [kH/kD = 2.92 at 98.3 8C].19 Elimination of benzene from [Me2Si(C5Me4)2]TiPh2 presumably occurs via a similar mechanism involving a benzyne intermediate,20,21 although the reaction is significantly more facile than elimination of methane from [Me2Si(C5Me4)2]- TiMe2.22 Evidence for the presence of a benzyne intermediate is provided by the observation that {[Me2Si(C5Me4)2]Ti(h2- C6H4)} may be trapped by C2H4 or C2H2 to give [Me2Si- (C5Me4)2]Ti(h2-C6H4CH2CH2) and [Me2Si(C5Me4)2]Ti(h2- C6H4CH]] CH), respectively (Scheme 3).Furthermore, the fulvene complex [Me2Si(C5Me4)(C5Me3CH2)]TiPh also reacts with C2H4 to yield [Me2Si(C5Me4)2]Ti(h2-C6H4CH2CH2), suggesting that isomerization of [Me2Si(C5Me4)(C5Me3CH2)]- TiPh to {[Me2Si(C5Me4)2]Ti(h2-C6H4)} is kinetically facile.The molecular structures of the majority of the above [Me2Si(C5Me4)2]TiXX9 complexes have been determined by Fig. 2 Coordination modes of fulvene ligands. Scheme 3J. Chem. Soc., Dalton Trans., 1999, 1365–1368 1367 Table 1 Geometrical data for [Me2Si(C5Me4)2]TiXX9 derivatives Cp*2TiCl2 b [Me2Si(C5Me4)2]TiCl2 c [Me2Si(C5Me4)2]Ti(CO)2 [Me2Si(C5Me4)2]Ti(CH2SiMe3)2 [Me2Si(C5Me4)2]Ti(NC4H4)2 [Me2Si(C5Me4)2]TiMe2 [Me2Si(C5Me4)2]Ti(NC4H4)Me [Me2Si(C5Me4)2]Ti(C6H4C2H4) [Me2Si(C5Me4)2]TiPh2 [Me2Si(C5Me4)2]Ti(C6H4C2H2) [Me2Si(C5Me4)2]Ti[(C6H4)2] d(M–Cpcent)/Å 2.128 2.136 2.046 2.167 2.154 2.138 2.133 2.129 2.143 2.090 2.095 d(M–C)/Å 2.404–2.484 2.365–2.552 2.281–2.465 2.392–2.574 2.386–2.557 2.371–2.530 2.375–2.517 2.383–2.521 2.396–2.521 2.363–2.470 2.366–2.468 d(M–C) range/Å 0.080 0.187 0.184 0.182 0.171 0.159 0.142 0.138 0.125 0.107 0.102 a/8 137.4 132.2 139.7 130.4 130.6 132.8 132.3 132.5 131.1 134.5 134.4 b/8 135.4 120.6 128.1 119.9 120.6 123.3 123.1 124.5 124.2 128.1 128.1 g a/8 1.0 5.8 5.8 5.3 5.0 4.8 4.6 4.0 3.5 3.2 3.2 a g = (a 2 b)/2.b From ref. 27. c From ref. 3. X-ray diVraction ‡ and details of the coordination of the ansa ligands are summarized in Table 1. By comparison with the non-bridged Cp*2TiCl2 derivative, in which the Cp* ligands are coordinated in a symmetric h5-fashion, with individual Ti–C bond lengths diVering by less than 0.08 Å, the cyclopentadienyl groups in [Me2Si(C5Me4)]TiXX9 derivatives are coordinated much less symmetrically. Thus, individual Ti–C bond lengths in [Me2Si(C5Me4)]TiX2 diVer by up to 0.19 Å for each complex, increasing in the sequence Ti–C1 < Ti–C2,5 < Ti–C3,4,23 such that the [Me2Si(C5Me4)] ligand adopts a modified geometry which approaches an h3,h3-coordination mode.This modifi- cation is accompanied by a small tilting (g) of the cyclopentadienyl groups away from the normal of the M–Cpcent vector towards the ansa-bridge, i.e.the Cpcent–M–Cpcent angle (a) is greater than the angle between the Cp ring normals (b). As with the corresponding zirconium system, the modified h3,h3-coordination geometry creates a more electrophilic metal center, as judged by the greater n(CO) stretching frequencies of the dicarbonyl complex [Me2Si(C5Me4)2]Ti(CO)2 (1879 and 1955 cm21) compared to those for Cp*2Ti(CO)2 (1858 and 1940 cm21).24–26 In summary, a variety of permethylated ansa-titanocene complexes has been synthesized.The dialkyl complexes [Me2- Si(C5Me4)]TiR2 are thermally unstable towards elimination of alkane (RH), thereby yielding fulvene derivatives [Me2Si- (C5Me4)(C5Me3CH2)]TiR. Interestingly, the titanium–fulvene interaction in these complexes may be considered to possess character analogous to that of a metal–trialkylidenemethane derivative. Acknowledgements We thank the U.S.Department of Energy, OYce of Basic Energy Sciences (#DE-FG02-93ER14339) for support of this research. Notes and references ‡ Crystal data. [Me2Si(C5Me4)2]TiMe2, C22H36SiTi, M 376.50, orthorhombic, Pnma, a 10.1151(9), b 14.966(2), c 13.7309(12) Å, U 2078.6(3) Å3, Z 4, m 0.471 mm21, T 293(2) K, R1 0.0878, wR2 0.1427 for 2494 reflections collected. [Me2Si(C5Me4)2]Ti(CH2SiMe3)2, C28H52Si3Ti, M 520.87, monoclinic, P21/n, a 9.322(4), b 15.914(8), c 20.647(9) Å, b 98.715(11)8, U 3027(2) Å3, Z 4, m 0.415 mm21, T 203(2) K, R1 0.1029, wR2 0.1152 for 20564 reflections collected. [Me2Si(C5Me4)2]TiPh2, C32H40SiTi, M 500.63, monoclinic, P21/n, a 11.518(5), b 15.820(7), c 15.041(7) Å, b 98.815(13)8, U 2708(2) Å3, Z 4, m 0.379 mm21, T 203(2) K, R1 0.0583, wR2 0.0992 for 20007 reflections collected.[Me2Si- (C5Me4)2]Ti[(C6H4)2], C32H38SiTi, M 498.61, triclinic, P1� , a 9.5788(10), b 10.0927(12), c 14.3197(16) Å, a 81.039(2), b 77.790(2), g 74.042(2)8, U 1293.9(3) Å3, Z 2, m 0.396 mm21, T 203(2) K, R1 0.0862, wR2 0.1302 for 9737 reflections collected. [Me2Si(C5Me4)2]Ti(C6H4C2H2), C28H36- SiTi, M 448.56, monoclinic, P21/n, a 10.2119(7), b 13.8023(8), c 17.0937(11) Å, b 100.3010(10)8, U 2370.5(3) Å3, Z 4, m 0.424 mm21, T 203(2) K, R1 0.0624, wR2 0.1175 for 17424 reflections collected.[Me2Si(C5Me4)2]Ti(C6H4C2H4), C28H38SiTi, M 450.57, monoclinic, P21/n, a 10.3168(5), b 13.9236(6), c 16.7155(8) Å, b 100.6990(10)8, U 2359.39(19) Å3, Z 4, m 0.427 mm21, T 203(2) K, R1 0.0499, wR2 0.1039 for 17101 reflections collected.[Me2Si(C5Me4)2]Ti(CO)2, C22H30O2SiTi, M 402.45, monoclinic, P21/c, a 9.7826(10), b 23.523(3), c 10.0518(11) Å, b 115.319(6)8, U 2090.8(4) Å3, Z 4, m 0.480 mm21, T 298(2) K, R1 0.1258, wR2 0.1518 for 3024 reflections collected. [Me2Si(C5Me4)2]TiMe(NC4H4), C25H37NSiTi, M 427.55, monoclinic, C2/m, a 17.078(7), b 10.981(5), c 13.495(5) Å, b 114.27(2)8, U 2307(2) Å3, Z 4, m 0.433 mm21, T 293(2) K, R1 0.0546, wR2 0.1202 for 2288 reflections collected.[Me2Si(C5Me4)2]Ti(NC4H5)2, C28H38N2SiTi, M 478.59, tetragonal, P43212, a 11.2367(6), b 11.2367(6), c 19.998(2) Å, U 2525.0(3) Å3, Z 4, m 0.405 mm21, T 293(2) K, R1 0.0814, wR2 0.1105 for 2791 reflections collected. [Me2Si(C5Me4)(C5Me3CH2)]TiPh, C26H34SiTi, M 422.52, triclinic, P1� , a 8.7542(8), b 9.0925(8), c 15.4375(14) Å, a 76.673(2), b 83.731(2), g 69.141(2)8, U 1116.83(17) Å3, Z 2, m 0.446 mm21, T 213(2) K, R1 0.1364, wR2 0.1747 for 8338 reflections collected.CCDC reference number 186/1397. 1 H. Lee, P. J. Desrosiers, I. Guzei, A. L. Rheingold and G. Parkin, J. Am. Chem. Soc., 1998, 120, 3255. 2 P. Jutzi and R. Dickbreder, Chem. Ber., 1986, 119, 1750. 3 Other [Me2Si(C5Me4)2]TiX2 derivatives include [Me2Si(C5Me4)2]TiCl and [Me2Si(C5Me4)2]Ti[h2-C2(SiMe3)2]. See: V. Varga, J. Hiller, R. Gyepes, M. Polasek, P. Sedmera, U. Thewalt and K. Mach, J. Organomet. Chem., 1997, 538, 63. 4 For unsubstituted analogs, [Me2Si(C5H4)2]TiR2, see: (a) R. Gómez, T. Cuenca, P. Royo, W. A. Herrmann and E. Herdtweck, J. Organomet. Chem., 1990, 382, 103; (b) R. Gómez, T. Cuenca, P. Royo and E. Hovestreydt, Organometallics, 1991, 10, 2516. 5 Alternatively, [Me2Si(C5Me4)2]Ti[(C6H4)2] may also be obtained by reaction of [Me2Si(C5Me4)2]TiPh2 with PhLi. 6 H. Lee, J. Cordaro and J. B. Bonanno, unpublished work. 7 Non-bridged fulvene analogs Cp*(C5Me4CH2)TiR (R = H, Me, CH2CMe3, CH2SiMe3, Ph, CH]] CH2) have also been reported.See, for example: (a) C. McDade, J. C. Green and J. E. Bercaw, Organometallics, 1982, 1, 1629; (b) G. A. Luinstra, P. H. P. Brinkmann and J. H. Teuben, J. Organomet. Chem., 1997, 532, 125; (c) G. A. Luinstra and J. H. Teuben, Organometallics, 1992, 11, 1793; (d ) J. L. Polse, R. A. Andersen and R. G. Bergman, J. Am. Chem. Soc., 1996, 118, 8737; (e) J. L. Polse, A. W. Kaplan, R. A. Andersen and R. G. Bergman, J. Am. Chem. Soc., 1998, 120, 6316; ( f ) R.Beckhaus, J. Oster and T. Wagner, Chem. Ber., 1994, 127, 1003; ( g) R. Beckhaus, J. Oster, I. Sang, I. Strauß and M. Wagner, Synlett, 1997, 241. 8 The CH2 groups in [Me2Si(C5Me4)(C5Me3CH2)]TiR are characterized by 13C NMR signals at ca. d 80, with 1JC–H coupling constants of ca. 150 Hz: Me (d 78.3, 149 and 154 Hz), Ph (d 82.5, 150 and 153 Hz) and CH2SiMe3 (d 80.2 ppm, 149 and 151 Hz). These values are comparable to those for Cp*(C5Me4CH2)TiR derivatives (see, for example, ref. 7). 9 The Ti–C bond lengths for the unmetallated cyclopentadienyl group range from 2.29 Å to 2.46 Å. 10 For a review of bonding in fulvene complexes, see: J. A. Bandy, V. S. B. Mtetwa, K. Prout, J. C. Green, C. E. Davies, M. L. H. Green, N. J. Hazel, A. Izquierdo and J. J. Martin-Polo, J. Chem. Soc., Dalton Trans., 1985, 2037. 11 For further comparison, the longest Ti–C bond length in the Ti(III) complex Cp*(C5Me4CH2)Ti is 2.47 Å. See: J. M. Fischer, W. E.Piers and V. G. Young, Jr., Organometallics, 1996, 15, 2410. 12 For trialkylidene methane derivatives of zirconium, see: G. Rodriguez and G. C. Bazan, J. Am. Chem. Soc., 1997, 119, 343. 13 Thus, the Ti–C12 bond deviates by only 2.58 from the Ti–centroid vector. 14 (a) L. E. Schock, C. P. Brock and T. J. Marks, Organometallics, 1987, 6, 232; (b) A. R. Bulls, W. P. Schaefer, M. Serfas and J. E. Bercaw, Organometallics, 1987, 6, 1219.1368 J. Chem. Soc., Dalton Trans., 1999, 1365–1368 15 For example, the wide range of M–C bond lengths in the zirconium and hafnium complexes, Cp*(C5Me4CH2)ZrPh (2.28–2.62 Å) 14a and Cp*(C5Me4CH2)HfCH2Ph (2.25–2.60 Å),14b suggests that they may also be considered to possess trialkylidenemethane character. 16 Specifically, [Me2Si(C5Me4)2]Ti(CD3)2 yields principally the isotopomers [Me2Si(C5Me4)(C5Me3CH2)]Ti(CD2H) and CD4, rather than [Me2Si(C5Me4)(C5Me3CH2)]Ti(CD3) and CD3H. 17 C. McDade, J. C. Green and J. E. Bercaw, Organometallics, 1982, 1, 1629. 18 For [Me2Si(C5Me4)2]TiMe2: DH‡ = 27.8(8) kcal mol21, DS‡ = 25(2) e.u. (1 e.u. = 4.184 J K21 mol21). For Cp*2TiMe2: DH‡ = 27.6(3) kcal mol21, DS‡ = 22.9(7) e.u. (ref. 17). 19 Interestingly, values of kH/kD ª 5 have also been observed for elimination of methane from Cp*2Ti(CH3)(C6D5) (5.7 at 33 8C) and Cp*2Ti(CH3)(CD=CD2) (5.1 at 80 8C). See ref. 7(c). 20 A benzyne intermediate has also been proposed in the formation of Cp*(C5Me4CH2)ZrPh by thermal elimination of PhH and H2 from Cp*2ZrPh2 14a and Cp*2Zr(Ph)H (F.D. Miller and R. D. Sanner, Organometallics, 1998, 7, 818), respectively. 21 Cp2TiPh2 has been proposed to decompose via ortho-hydrogen abstraction by the other phenyl group, generating a benzyne intermediate, [Cp2Ti(h2-C6H4)]. See: (a) J. Dvorak, R. J. O’Brien and W. Santo, Cheommun., 1970, 411; (b) C. P. Boekel, J. H. Teuben and H. J. de Liefde Meijer, J. Organomet. Chem., 1975, 102, 161. 22 At 40 8C, the rate constants for elimination of RH from [Me2- Si(C5Me4)2]TiR2 are: Ph [1.30(1) × 1024 s21], CH2SiMe3 [9.85(9) × 1025 s21] and Me [2.0 × 1028 s21]. The value for the methyl derivative is that determined by the activation parameters listed in ref. 18. 23 The numbering system is such that C1 is the ring carbon attached to silicon. 24 D. J. Sikora, M. D. Rausch, R. D. Rogers and J. L. Atwood, J. Am. Chem. Soc., 1981, 103, 1265. 25 ESR spectroscopic studies suggest that 2-methyltetrahydrofuran binds more strongly to [Me2Si(C5Me4)2]TiCl than to Cp*2TiCl, providing further evidence for enhanced electrophilicity of the ansa titanocene system. See ref. 3. 26 It should, however, be noted that computational studies suggest that ring slippage and tilt introduced by incorporation of bulky substituents are not the principal factors responsible for modifi- cation of the electron density at a metal center in a series of non-bridged titanocene complexes; rather the changes in electron density merely reflect the inductive eVects of the various substituents. See: B. E. Bursten, M. R. Callstrom, C. A. Jolly, L. A. Paquette, M. R. Sivik, R. S. Tucker and C. A. Wartchow, Organometallics, 1994, 13, 127. 27 T. C. McKenzie, R. D. Sanner and J. E. Bercaw, J. Organomet. Chem., 1975, 102, 457. Communication 9/00173E

ISSN:1477-9226    

DOI:10.1039/a900173e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1999, 1369–1371 1369 The first evidence for activation of exogenous O2 on a vanadium(IV) center: synthesis and characterization of a peroxo vanadium(V) complex with hydrotris(3,5-diisopropylpyrazol-1-yl)- borate Masahiro Kosugi, Shiro Hikichi,* Munetaka Akita and Yoshihiko Moro-oka * Research Laboratory of Resources Utilization, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan. E-mail: shikichi@res.titech.ac.jp Received 8th February 1999, Accepted 19th March 1999 Reaction of a vanadium(IV) hydroxo complex bearing hydrotris(3,5-diisopropylpyrazol-1-yl)borate (TpPri 2) with O2 results in reductive O2 activation to give TpPri 2VV(O)(Á2- O2)(L); the first example of the formation of a peroxo vanadium complex from molecular oxygen.Reductive O2 activation (O2 æÆ O2 2 æÆ O2 22 æÆ 2 × O22) via oxidative addition of molecular oxygen to a transition metal center is a fundamental process in various synthetic O2 oxidation reactions and physiological O2 metabolism.Therefore, research into the reactivity of transition metal complexes toward O2 activation and the characterization of the resulting peroxo complexes is essential to the understanding of the synthetic and metabolic O2 activation mechanisms.1 It is known that vanadium–peroxo species take part in various catalytic oxidations ranging from industrial processes to enzymatic reactions. 2 However, the oxidant used in most of these processes (including halide oxidation by haloperoxidases) is not O2 but ROOH (R = H, alkyl); only a limited number of aerobic oxidation reactions by vanadium catalysts have been reported.Moreover, the role of the vanadium center in aerobic oxidation processes is proposed to assist the autoxidation reaction (i.e. degradation of the alkylhydroperoxides to induce radical chain reaction).3 In addition, previously reported vanadium(V)– peroxo complexes were prepared by the reaction of appropriate V(V)–oxo or –hydroxo precursors with ROOH,2 and no examples of the formation of a peroxo complex via incorporation of an external O2 molecule was known to date.4 In this communication, we report the first evidence for the activation of exogenous O2 on V(IV) centers resulting in formation of V(V)–peroxo species.† We have been investigating the peroxo and related complexes of various first- and second-row, late transition metals (Mn, Fe, Co, Ni, Cu, Ru, Rh, Pd) with the hindered hydrotris- (pyrazolyl)borate ligands (TpR) in order to obtain comprehensive insights into the role of metal ions in various oxidation processes.5 Recently, our research target has been extended to the early transition metals, and a V(IV)–hydroxo complex with hydrotris(3,5-diisopropylpyrazol-1-yl)borate (TpPri 2), TpPri 2- VIV(O)(OH)(OH2) 1, has been synthesized and characterized successfully.6 When a toluene solution of 1 was exposed to O2 (1 atm) in the presence of 1 equiv.of 3,5-diisopropylpyrazole (PzPri 2H)‡ at room temperature, the purple solution changed to a dark red one within 2 h. NMR (51V, 13C and 1H) and IR spectra of the resulting products clearly indicated the existence of two sets of TpPri 2VV(PzPri 2H) moieties in a 1:1:molar ratio and, in addition, two sets of n(V]] O) and n(O–O) vibration bands appeared in the 960–890 cm21 region (in the IR region). Finally, these two products were identified as mononuclear V(V)-cis-dioxo and –oxo–peroxo complexes, TpPri 2VV(O)2- (PzPri 2H) 2 and TpPri 2VV(O)(h2-O2)(PzPri 2H) 3, respectively, by comparison of the spectral data with those of the independently prepared authentic samples whose molecular structures were successfully determined by X-ray crystallography (see below).§ An authentic sample of 2 was synthesized by chemical oxidation of 1 with ButOOH or KMnO4 in the presence of 1 equiv. of the pyrazole ligand (PzPri 2H), and subsequent dehydrative condensation of the isolated 2 with H2O2 (in the presence of Na2SO4 as dehydrating reagent) yielded the corresponding V(V)–oxo–peroxo complex 3 (Scheme 1).¶ The pale yellow dioxo complex 2 involves a slightly distorted octahedral vanadium center coordinated by two terminal oxo ligands in cis configuration [Fig. 1(a)]. The structural and spectral features of 2 [(i) somewhat long V]] O lengths: V–O(1) 1.621(3), V–O(2) 1.636(3) Å, (ii) elongation of two of the three V–NTp bond lengths trans to the oxo ligands are due to the strong trans eVect from the terminal oxo ligands, and (iii) relatively low n(V]] O) vibration frequencies (921 and 893 cm21)] are similar to those found for the previously reported V(V)–cis-dioxo complexes.7 Scheme 11370 J.Chem. Soc., Dalton Trans., 1999, 1369–1371 The red peroxo complex 3 has a seven-coordinated pentagonal- bipyramidal metal center containing an h2-peroxo ligand [V–O(11) 1.862(5), V–O(12) 1.887(5) Å; O(11)–V–O(12) 43.2(2)8], and the overall structures of 2 and 3 are very similar when it is assumed that the peroxo ligand occupies a single coordination site [Fig. 1(b)]. The distance from the vanadium center to the terminal oxo ligand [V–O(1) 1.603(4) Å, n(V]] O) 947 cm21] is shorter than those found for the dioxo complex 2. The relatively short O–O length [O(11)–O(12) 1.379(6) Å] is almost at the shortest end of the typical range for the O–O distances of peroxide ligands,8 and is indicated by the relatively high n(O–O) value of 960 cm21.Retention of the side-on bound peroxo ligand in solution is supported by a UV-vis spectrum of a toluene solution involving the peroxo-to-vanadium chargetransfer band around 495 nm (e = 280 M21 cm21).2 The O2 activation on the V(IV) center of the hydroxo complex 1 was evidenced by a labeling experiment with 18O2 and an external substrate oxidation ability. Of the four characteristic vibrations [n(O–O) and n(V]] O), see above] of a reaction mixture of 1 and O2, only the n(O–O) band at 960 cm21 disappeared upon treatment with 18O2 [overlapped with the tail of the n(V]] O) peak at 893 cm21 (2); observed n(18O–18O) value = 900 cm21 in the sample prepared by the reaction of 2 with H2 18O2], and the remaining three n(V]] O) bands were not shifted.We thus conclude that the peroxo ligand in 3 arises from the Fig. 1 Molecular structures of TpPri 2V(O)2(PzPri 2H) 2 (a) and TpPri 2V(O)(h2–O2)(PzPri 2H)?THF 3?THF (b) drawn at the 50% probability level.All hydrogen atoms except those attached to the nitrogen atoms [ N(42)] of the coordinating pyrazole ligands (2 and 3), the disordered carbon atoms of one of the three 5-Pri groups in the TpPri 2 ligand and the THF solvate (3) are omitted for clarity. Selected bond lengths (Å) and angles (8): (a) cis-dioxo complex 2 V–O(1) 1.621(3), V– O(2) 1.636(3), V–N(11) 2.324(5), V–N(21) 2.203(5), V–N(31) 2.122(5), V–N(41) 2.161(5), O(1)–V–O(2), 103.8(1); (b) oxo–h2–peroxo complex 3 O(11)–O(12) 1.379(6), V–O(1), 1.603(4), V–O(11), 1.862(5), V–O(12), 1.887(5), V–N(11) 2.324(5), V–N(21) 2.203(5), V–N(31) 2.122(5), V– N(41) 2.161(5); O(11)–V–O(12), 43.2(2).external dioxygen molecule and the origin of the terminal oxo ligands in both 2 and 3 is the oxygen atoms of the oxo and the hydroxo ligands in 1 (not O2). Trapping of the external O2 molecule as the peroxide (]] O2 22) ligand on the vanadium center (i.e. formation of 3) indicates that the present aerobic oxidation of 1 is clearly diVerent from the usual 4e2 oxidation process of [VIV(]] O)]21 compounds yielding the vanadium(V)–dioxo compounds [eqn.(1)].2 4[VIV(]] O)]21 1 2H2O 1 O2 æÆ 4[VV(]] O)2]1 1 4H1 (1) Remarkably, the V(IV)–hydroxo complex 1 showed aerobic PPh3 oxygenation activity [under 1 atm O2, r.t., reaction time: 30 min, yield of O]] PPh3: 150% (based on 1 in the presence of 30 equiv. of PPh3)], whereas the isolated mononuclear V(V)–h2- peroxo complex 3 exhibited relatively low oxo-transfer activity under the same condition (yield 31% based on 3) and the V(V)– dioxo complex 2 could not oxidize PPh3 under any conditions.These observations supported the fact that the reductive O2 activation was mediated by another vanadium–O2 species formed at an initial stage of the oxygenation of 1. In the present vanadium system, the electron donating ability of TpPri 2 might make the reductive O2 activation on the vanadium(IV) center possible.9 In conclusion, the reductive O2 activation took place on the V(IV) center of the hydroxo complex containing TpPri 2.The resulting monomeric V(V)–h2-peroxo complex is the first example of the V(V)–peroxo complex derived from the direct oxygenation of the V(IV) precursor. Detailed investigation of the reactivities of the vanadium–peroxo species is now underway. Acknowledgements We are grateful to the Ministry of Education, Science, Sports and Culture of the Japanese government for financial support of the research (Grant-in-Aid for Specially Promoted Scientific Research: No. 08102006). Notes and references † Abbreviations used in this paper: TpR, hydrotris(3,5-substitutedpyrazol- 1-yl)borate; TpPri 2, hydrotris(3,5-diisopropylpyrazol-1-yl)- borate; PzPri 2H, 3,5-diisopropylpyrazole. ‡ Reaction of 1 with O2 in the absence of the additional PzPri 2H also resulted in the formation of the PzPri 2H containing complexes 2 and 3, although the yields of them were quite low (>30% based on 1) due to partial decomposition of the TpPri 2 ligand providing the PzPri 2H ligands.§ Crystal data. For 2: C36H62N8O2BV, M = 700.69, monoclinic, space group P21/n (no. 14), a = 9.698(10), b = 20.54(2), c = 22.84(1) Å, b = 92.51(4)8, V = 4018(2) Å3, Z = 4, Dc = 1.16 g cm21, m(Mo-Ka) = 2.87 cm21, T = 260 8C, R(Rw) = 0.043 (0.046) (based on F) for 3613 [I > 3s(I)] reflections with 442 parameters.For 3?THF: C40H70N8O4BV, M = 788.80, monoclinic, space group P21/n (no. 14), a = 9.171(2), b = 17.458(4), c = 25.12(1) Å, b = 98.97(8)8 V = 4493(6) Å3, Z = 4, Dc = 1.17 g cm23, m(Mo-Ka) = 2.67 cm21, T = 260 8C, R(Rw) = 0.077, (0.068) (based on F) for 4472 [I > 3s(I)] reflections with 509 parameters. CCDC reference number 186/1394. ¶ Selected spectroscopic data. For 2: IR (KBr pellet, n/cm21): 3433(N– H), 2541 (B–H), 921, 893 (V]] O). FD-MS: m/z 702 (M 1 H1). UV-vis (toluene, l/nm, e/M21 cm21): 749 (7.3). 51V NMR (C6D6, reference; VOCl3): d 541. 1H NMR (C6D6): d 9.03 (1H, pyrazole N-H). For 3: IR (KBr pellet, n/cm21): 3436 (N–H), 2543 (B–H), 960 (16O–16O), 900 (18O– 18O; prepared by rection of 2 with H2 18O2), 947 (V]] O). FD-MS: m/z 718 (M 1 H1). UV-vis (toluene, l/nm, e/M21 cm21): 495 (280). 51V NMR (C6D6, reference; VOCl3): d 552. 1H NMR (C6D6): d 8.43 (1H, pyrazole N–H). 1 Metal-Catalyzed Oxidations of Organic Compounds, eds. R. A. Sheldon and J.K. Kochi, Academic Press, New York, 1981; special thematic issue of Chem. Rev. for “Metal-Dioxygen Complexes”, Chem. Rev., 1994, 94, 567–856. 2 A. Butler, M. J. Clague and G. E. Meister, Chem. Rev., 1994, 94, 625; C. Slebodnick, B. J. Hamastra and V. L. Pecoraro, Struct. Bonding (Berlin), 1997, 89, 51. 3 C. J. Chang, J. A. Labinger and H. B. Gray, Inorg. Chem., 1997, 36, 5927.J. Chem. Soc., Dalton Trans., 1999, 1369–1371 1371 4 One example of the formation of the V(V)–peroxo complex without using H2O2 has been reported. However, the origin of the oxygen atoms of the peroxide ligand has not been revealed: M.Shao, X. Dong and Y. Tang, Sci. Sinica, Ser. B (Engl. Ed.), 1988, 31, 789. 5 Co–OOR: S. Hikichi, H. Komatsuzaki, M. Akita and Y. Moro-oka, J. Am. Chem. Soc., 1998, 120, 4699; Co–, Ni–oxo: S. Hikichi, M. Yoshizawa, Y. Sasakura, M. Akita and Y. Moro-oka, J. Am. Chem. Soc., 1998, 120, 10567; Mn–OO(R): H. Komatsuzaki, Y. Nagasu, K.Suzuki, T. Shibasaki, M. Satoh, F. Ebina, S. Hikichi, M. Akita and Y. Moro-oka. J. Chem. Soc., Dalton Trans., 1998, 511; H. Komatsuzaki, M. Satoh, N. Sakamoto, S. Hikichi, M. Akita and Y. Moro-oka, Inorg. Chem., 1998, 37, 6554; Fe–catecholate oxygenation: T. Ogihara, S. Hikichi, M. Akita and Y. Moro-oka, Inorg. Chem., 1998, 37, 2614; Pd–O2: M. Akita, T. Miyaji, S. Hikichi and Y. Moro-oka, Chem. Commun., 1998, 1005; Rh–O2: Y. Takahashi, M. Hashimoto, S. Hikichi, M. Akita and Y. Moro-oka, submitted. 6 M. Kosugi, S. Hikichi, M. Akita and Y. Moro-oka, Inorg. Chem., submitted. 7 See for examples of six-coordinate cis-dioxo–V(V) complexes: D. C. Crans, A. D. Keramidas, S. S. Amin, O. P. Anderson and S. M. Miller, J. Chem. Soc., Dalton Trans., 1997, 2799; D. C. Crans, A. D. Keramidas, M. Mahroof-Tahir, O. P. Anderson and M. M. Miller, Inorg. Chem., 1996, 35, 3599; D. Schulz, T. Weyhermüller, K. Wieghardt and B. Nuber, Inorg. Chim. Acta, 1995, 240, 217. 8 M. H. Gubelmann and A. F. Williams, Struct. Bonding (Berlin), 1983, 55, 1. 9 Similar O2 activation was observed for the oxygenation of the dinuclear manganese(II)–hydroxo complex bearing the same TpPri 2 ligand reported by us: N. Kitajima, M. Osawa, M. Tanaka and Y. Moro-oka, J. Am. Chem. Soc., 1991, 113, 8952. Communication 9/01054H

ISSN:1477-9226    

DOI:10.1039/a901054h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON COMMUNICATION J. Chem. Soc., Dalton Trans., 1999, 1373–1374 1373 A singly stranded, helical di-ruthenium(II) complex of a novel 6,6-ethynyl-linked bis(terpyridine) ligand. Distortion of the ethyne linkage and inversion of helicity Charles M. Chamchoumis and Pierre G. Potvin * Department of Chemistry, York University, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3. E-mail: pgpotvin@yorku.ca Received 2nd March 1999, Accepted 12th March 1999 The new, ethynyl-bridged ditopic ligand, 1,2-bis[49-(4- methylphenyl)-2,29:6920-terpyridin-6-yl]ethyne (L) was prepared and used in the formation of the helical dinclear complex [L{Ru(ttpy)}2][PF6]4 (ttpy = 49-(4-methylphenyl)- 2,29:69,20-terpyridine), whose crystal structure exhibits pronounced bending of the central ethynyl bridge and whose solution NMR spectra indicate a rapid inversion of helicity at 335 K.Multinuclear complexes of Ru(II) have been built with a number of bridging ligands 1 but helical complexes based on terpyridine (tpy) units are few.2–5 Helicating ditopic ligands can be built by coupling monotopic, unsymmetrically substituted terpyridines, which are available by Kröhnke synthesis.6 Linkages through the 6 position can lead to strong inter-ligand congestion but alkynyl substituents are relatively sterically undemanding.We report herein the successful preparation and structural characterization of a helical, singly stranded dinuclear Ru(II) complex of the first ditopic ligand assembled from two tpy units joined at the 6 position through an ethynyl bridge.This is the first structurally characterized complex of this type. The electronic character of the bridge was also of interest as ethyne substituents are electron-withdrawing groups and, like these,7 can dramatically enhance triplet lifetimes and luminescence yields over those of Ru(tpy)2 21.8 The starting material I was obtained in near quantitative yield from the condensation of 2-acetyl-6-bromopyridine 9 with p-tolualdehyde (Scheme 1).† The reaction of I with pyridinium salt II 10 produced the 6-bromoterpyridine III in ª90% yield.This was coupled with trimethylsilylacetylene to aVord the protected 6-ethynylterpyridine IV, which was then deprotected to give the terminal alkyne V in ª75% overall yield. The ditopic ligand L was assembled by Sonogashira coupling 11 of III and V in 98% yield (ª66% overall from 2-acetyl-6-bromopyridine). The direct reaction of L with Ru(ttpy)Cl3 12 failed to proceed.However, the dinuclear complex [L{Ru(ttpy)}2][PF6]4 1 was obtained after activation of Ru(ttpy)Cl3 (AgBF4, acetone, reflux, 0.5 h) and heating with 0.5 equivalent of L (dmf, reflux, 8 h), followed by precipitation with aqueous NH4PF6, chromatography (silica, 14:2:1 CH3CN–saturated KNO3–H2O), reprecipitation (NH4PF6) and recrystallization, in 65% isolated yield.† There are only two possible conformations for 1: the flat, fully conjugated (and achiral) form, or the non-conjugated, helical (and chiral) form in which both metals have the same absolute configurations.The 1H NMR spectrum of 1 showed considerable complexity but, with the help of COSY spectra, revealed one set of signals for two symmetrical monotopic ttpy units. The outer pyridines of these units gave broad signals which sharpened upon heating to 335 K. This symmetrization implies an interconversion in solution between the enantiomers of the helical form.In contrast, NMR suggested that the only other comparable, singly-stranded Ru(II) helicate, bearing a 6,6-linked m-phenylene bridge, was rigid at room temperature.2 The crystal and molecular structure of 1 at 150 K is shown in Fig. 1.‡ This clearly revealed the cation to be helical, with a 123.88 dihedral angle about the triple bond. There appears to be considerable distortion at the ethynyl linkage, which is bent out of linearity, and in the linked ttpy units, which are bowed such that the least-squares planes of the inner pyridines (ring 1) are 168 out of planarity with the outer pyridines (ring 3) and form the longest bonds (2.107 Å) to the Ru.The other Ru–N distances are normal and the two N3 binding planes are nearly orthogonal to one another (88.38). Another notable feature is the twisting by 33.48 of the tolyl ring on the ttpy ligand (ring 8) to become almost parallel (4.48) with the nearest pyridine (ring 1A) of the ethynyl-linked ttpy unit at the neighbouring metal, thus enabling p–p interactions at a 3.6–3.8 Å inter-planar distance, whereas the tolyl groups on L (ring 4) are more coplanar (inter-planar angle 7.88) with the attached pyridines (ring 2).The cyclic voltammogram§ of 1 showed one reversible Ru31/21 couple (1.42 V vs. SCE) and several reduction waves. The two least negative reduction waves (21.07 and 21.16 V) were both more positive than the first reductions of Ru(ttpy)2 21 (21.24 V).13 Reductions of both L and ttpy ligands in 1 are expected to be easier than in Ru(ttpy)2 21, on the one hand because of the electron-withdrawing eVect of the ethyne group combined with the presence of a second metal and, on the other hand, because the loss of conjugation between the tolyl and terpyridine portions of the ttpy ligand in 1 (see above) amounts Scheme 1 i, CH3C6H4CHO-4, KOH in 80% MeOH; ii, excess NH4OAc in AcOH, reflux, 4 h; iii, 6 mol% Pd(PPh3)4, 12 mol% CuI, iPr2NH, room temp., 8 h; iv, KF, MeOH, room temp., 1 h; v, 10 mol% Pd(PPh3)4, iPr2NH, room temp., 8 h.N N N X Ar N Br O Ar N H3C Br O N N N Ar Ar N N N N+ O N L I Ar = C6H4CH3-4 i ii iii iv v II I– III X = Br V X = C CH IV X = C CSiMe31374 J. Chem. Soc., Dalton Trans., 1999, 1373–1374 to a loss of an electron-donating group, relative to the conjugated ttpy in Ru(ttpy)2 21. In any case, the potentials here indicate significant stabilizations of both metal-centred HOMO and ligand-centred LUMO to equal extents (170 mV), when compared with Ru(ttpy)2 21,13 unlike the eVects of other electron-withdrawing groups on terpyridine complexes.7 There resulted the same HOMO–LUMO gap (E2� 1 31/21 2 E2� 1 21/1) and similar MLCT band lmax values [494 nm, e 35 800 M21 cm21, with a shoulder at 454 nm, vs. 490 nm, e 15 500 M21 cm21 for Ru(ttpy)2 21]. This contrasts with the 49,49-ethyne-bridged analogue,14 where the bridge and the second metal centre act as more typical electron-withdrawing groups, i.e.aVecting mostly the LUMO and red-shifting the MLCT lmax. Our case is reminiscent of the 6-vinylterpyridine complex 15 in which similar eVects were observed and attributed to a weaker binding typical of 6-substituted terpyridines. The long Ru–N bond measured here would support such an explanation in the present case. On the other hand, data from the singly stranded, 6,6-linked m-phenylene-bridged 2 and quinquepyridine 3 helicates indicate that these linkages behaved as e2-donating groups.A doubly stranded, 5,5-ethanyl linked helicate showed unsurprisingly weak eVects.4 Whether linked at positions 6, 5 or 49, all cases except the unsymmetrical quinquepyridine helicate 3 showed single Ru31/21 couples, implying weak electronic communication between the metals. Fig. 1 ORTEP16 diagram of the complex cation of 1?10.5CH3COCH3 with H atoms omitted for clarity. The unlabelled portion is the symmetry equivalent of the labelled portion and is drawn with 50% probability thermal ellipsoids.The first digit of a ring atom label is the ring number. Selected bond lengths (Å) and angles (8): Ru1–N11 2.107(3), Ru1–N21, 1.986(3), Ru1–N31 2.060(3), Ru1–N51 2.069(3), Ru1–N61 1.976(3), Ru1–N71 2.066(3), C17–C17A 1.215(8), C16–C17– C17A 167.8(5), C15–C16–C17 118.7(3), N11–C16–C17 119.5(3), C16– C17–C17A–C16A 123.8. Notes and references † All new compounds were fully characterized by 1H NMR, 13C-NMR, and EI- or FAB-mass spectra, and by elemental analysis.‡ Crystal data: C121.5H129F24O10.5P4Ru2, M = 2707.39, monoclinic, space group C2/c (no. 15), a = 25.6950(7), b = 14.0790(4), c = 35.4180(9) Å, b = 100.370(2)8, U = 12603.5(6) Å3, Z = 4, Dc = 1.427m23, m(Mo-Ka) = 0.388 mm21, F(000) = 5560, T = 150.0(1) K. Of 41,430 reflections collected, the final R indices were, for all data, R = 0.0834 and Rw = 0.1850 and for the 9463 reflections where I > 2s(I), R = 0.0603 and Rw = 0.1652. CCDC reference number 186/1381.See http://www.rsc.org/suppdata/dt/1999/1373 for crystallographic files in .cif format. § In CH3CN with 0.1 M nBu4NPF6 at 298 K; values were ±0.02 V and peak-to-peak separations were all 50–75 mV. 1 K. Kalyanasundaram and Md. K. Nazeeruddin, Inorg. Chim. Acta, 1994, 226, 213. 2 P. K.-K. Ho, S.-M. Peng, J.-Y. Wong and C.-M. Che, J. Chem. Soc., Dalton Trans., 1996, 1829. 3 C. J. Cathey, E. C. Constable, M. J. Hannon, D. A. Tocher and M.D. Ward, J. Chem. Soc., Chem. Commun., 1990, 621. 4 J. D. Crane and J.-P. Sauvage, New. J. Chem., 1992, 16, 649. 5 P. K.-K. Ho, K.-K. Cheung and C.-M. Che, Chem. Commun., 1996, 1197. 6 F. Kröhnke, Synthesis, 1976, 1. 7 M. Maestri, N. Armaroli, V. Balzani, E. C. Constable and A. M. W. Cargill Thompson, Inorg. Chem., 1995, 34, 2759. 8 A. C. Beniston, V. Grosshenny, A. C. Harriman and R. Ziessel, Angew. Chem., Int. Ed. Engl., 1994, 33, 1884. 9 J. E. Parks, B. E. Wagner and R. H. Holm, J. Organomet. Chem., 1973, 56, 53. 10 F. Kröhnke, Angew. Chem., Int. Ed. Engl., 1963, 2, 386. 11 K. Shonogashira, Y. Tohda and N. Hagihara, Tetrahedron Lett., 1975, 4467; K. Sonogashira, in Comprehensive Organic Synthesis, ed. B. M. Trost and I. Fleming, Pergamon Press, Oxford, 1991, vol. 3, p. 521. 12 J.-P. Collin, S. Guillerez and J.-P. Sauvage, J. Chem. Soc., Chem. Commun., 1989, 776. 13 J.-P. Collin, S. Guillerez, J.-P. Sauvage, F. Barigelletti, L. De Cola, L. Flamigni and V. Balzani, Inorg. Chem., 1991, 30, 4230. 14 V. Grosshenny, A. Harriman, J.-P. Gisselbrecht and R. Ziessel, J. Am. Chem. Soc., 1996, 118, 10315; A. Harriman and R. Ziessel, Chem. Commun., 1996, 1707. 15 K. T. Potts, D. A. Usifer, A. Guadalupe and H. D. Abruna, J. Am. Chem. Soc., 1987, 109, 3961. 16 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. Communication 9/01696A

ISSN:1477-9226    

DOI:10.1039/a901696a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 1381–1385 1381 A [RuII(bipy)3]-[1,9-diamino-3,7-diazanonane-4,6-dione] twocomponent system, as an eYcient ON–OFF luminescent chemosensor for Ni21 and Cu21 in water, based on an ET (energy transfer) mechanism Fabrizio Bolletta,a Ilaria Costa,b Luigi Fabbrizzi,*b Maurizio Licchelli,b Marco Montalti,a Piersandro Pallavicini,b Luca Prodi *a and Nelsi Zaccheroni a a Dipartimento di Chimica “G. Ciamician”, Università degli Studi di Bologna, via Selmi 2, 40126 Bologna, Italy b Dipartimento di Chimica Generale, Università di Pavia, v.Taramelli, 12, 27100 Pavia, Italy Received 2nd November 1998, Accepted 1st March 1999 A dioxo-tetramine ligand (1,9-diamino-3,7-diazanonane-4,6-dione º dioxo-2,3,2-tet) has been appended to a RuII(bipy)3 unit. This new system, 2, is water-soluble and capable of sensing Cu21 and Ni21 cations thanks to the strong quenching of the Ru(bipy)3 fluorescence, which takes place when a metal cation is coordinated by the dioxo-2,3,2-tet binding unit.Coordination requires the energetically expensive deprotonation of the amide nitrogens, so that only Cu21 and Ni21 are able to promote it among the series of divalent first-row transition metal cations. Moreover, the complexation reaction is pH-dependent and one can distinguish between the two metal cations on working at the proper pH. The quenching mechanism has been examined by measuring the lifetime of the excited state of the ruthenium luminophore both on the metal-free and metal-complexed system and by flash photolysis experiments carried out on the complexed systems.The results clearly indicate that an energy transfer mechanism holds both for the Cu21 and Ni21 complex. The characterization of 2 as a water soluble ON–OFF sensor for copper and nickel has also been checked for its lowest detection limit, finding that these two metals can be detected down to a 1027 M concentration.Moreover, also system 3, containing a dioxo-2,3,2-tet ligand and the ReI(CO)3bipy(Cl) luminophore, has been examined as another possible water-soluble ON–OFF fluorescent sensor for the same transition metal cations. Again, only Cu21 and Ni21 are bound with a pH-dependent equilibrium, but incomplete luminescence quenching was observed, which prevented the determination of the quenching mechanism. Introduction Luminescence quenching or enhancement is currently receiving a lot of attention as regards its application in sensing cationic, anionic or neutral species in solution.1 A multi-component approach is commonly employed: a binding unit is covalently connected to a luminophore in a single molecular system.The interaction of a substrate with the binding unit, usually with a 1 : 1 stoichiometry, induces significant changes in the emitting properties of the luminophore, thus allowing it to signal the presence of the substrate.2 Of particular interest is the possibility of building multi-component systems in which a wanted species is selectively bound by the binding unit, and/or, once bound, it causes a selective variation of the luminescence among a series of similar substrates. If the target species is a metal cation, sensors may be synthesized in which the choice of the binding unit can be based on the wide knowledge coming from classical coordination chemistry.We have recently found3 that a well known category of ligands, dioxo-tetraamines (and in particular dioxo-2,3,2-tet º 1,9-diamino-3,7-diazanonane- 4,6-dione) can be successfully employed for signalling the presence of Ni21 and Cu21 in aqueous solution, if incorporated in a two-component fluorescent molecule, where the fluorophore is an anthracene fragment (see ligand 1): in the absence of metal cations the system displays its full fluorescence, but when the transition metal cation is bound by the dioxo-2,3,2-tet unit, according to equilibrium (1), the anthracene emission is quenched through an electron-transfer (eT) mechanism.2a According to this, the system behaves as an ON–OFF sensor.Dioxo-tetramine fragments appear as ligands of choice, as they (i) do not influence significantly the fluorescence of the connected fluorophore; 3 (ii) can bind only Ni21 and Cu21 among the series of divalent transition metal cations belonging to the first transition row; (iii) allow us to select between Ni21 and Cu21, as the complexation process (equilibrium (1)) is pHdependent and takes place at remarkably diVerent pH values for the two cations.One major drawback of system 1, however, is the need for organic/aqueous solvent mixtures as the working media due to the pronounced lipophilicity of the anthracene fragment. In view of the possible analytical use of systems of this kind in natural and biologically relevant environments (i.e. water as solvent, at a neutral or slightly basic pH) we switched to more hydrophilic luminescent fragments, with the aim of obtaining a water-soluble sensor, selective for Cu21 and Ni21.Thus, the two-component system 2 has been prepared, in which the dioxo-2,3,2-tet binding unit has been appended to the highly fluorescent, water soluble [RuII(bipy)3]21 unit (bipy = 2,29- bipyridine),4 a well studied luminophore whose emitting properties have already been found to vary when transition metal cations are bound in an appended tetraaza macrocyclic unit.5 In this work, the synthesis, coordinative properties and use of 2 as an ON–OFF sensor for Ni21 and Cu21 in water have been studied and described, with particular attention to the photophysical properties of its metal complexes and to the quenching mechanism.A [ReI(CO)3Cl(bipy)] derivative of the dioxo-2,3,2- tet ligand (ligand 3) has also been prepared with the same aim and an overview of the chemical and photophysical studies on its binding properties towards Ni21 and Cu21 in water is also presented here.Experimental Instrumentation Absorption spectra were recorded with a Perkin-Elmer Lambda 16 spectrophotometer. Uncorrected emission, corrected excit-1382 J. Chem. Soc., Dalton Trans., 1999, 1381–1385 ation spectra and phosphorescence lifetimes were obtained with a Perkin-Elmer LS 50 spectrofluorimeter. The fluorescence lifetimes (uncertainty ±5%) were obtained with an Edinburgh single-photon counting apparatus, in which the flash lamp was filled with D2.Luminescence quantum yields (uncertainty ±15%) were determined using [Ru(bpy)3]21 in aqueous solution (F = 0.028) 6 as a reference. In order to allow comparison of emission intensities, corrections for instrumental response, inner filter eVects, and phototube sensitivity were performed.7 A correction for diVerences in the refractive index was introduced when necessary. Degassed solutions were obtained with the freeze-thaw-pump method. Emission spectra at 77 K were obtained using quartz tubes immersed in a quartz Dewar filled with liquid nitrogen.Transient absorption experiments were performed as described in reference 8. Fluorimetric titrations For the fluorimetric titrations, concentrations of 1 × 1024 M were used for both 2 and 3. Metal ions were added as chloride or perchlorate salts. Standard HCl, HClO4 and NaOH were used for changing the pH conditions for the luminescence vs. pH experiments. BuVers at pH 7.0 and 9.0 were obtained using the lutidine/lutidinium and borate/boric acid couples, respectively.Syntheses 5-(Bromomethyl)-2,29-bipyridine 9 was prepared according to literature methods. All the other reagents are commercially available and were used as such, except where indicated in the experimental procedure. Diethyl (2,29-bipyridin-5-yl)methylmalonate 4. 1.75 g (7.04 mmol) 5-(bromomethyl)-2,29-bipyridine were dissolved in 10 ml of dry toluene and added dropwise under a nitrogen atmosphere, over a 20 minute period, to 20 ml of absolute ethanol, in which 1.34 g (8.4 mmol) diethyl malonate and 0.192 g (8.4 mmol) Na had been dissolved.When the addition was complete, the reaction mixture was heated to reflux for 1 hour and then kept well stirred, at room temperature, for 24 hours, after which time the abundant white NaBr precipitate was separated with a paper filter and thoroughly washed with several 2 ml portions of CH2Cl2. The gathered organic solutions were then evaporated to dryness to give a pale yellow oil, which was a mixture of 4 and of the bis[5-(2,29-bipyridyl)]methyl substituted diethyl malonate derivative 5 (as identified by NMR spectroscopy).The mixture was separated on a basic Al2O3 column (using 10 :1 v :v n-hexane/ethyl acetate as eluent), to give 0.72 g (yield = 35%) of 4 as a colorless oil. NMR (CD3OD) d 8.7 (dd, 1H), 8.55 (ds, 1H), 8.35 (m, 2H), 7.8 (dt, 1H), 7.6 (dd, 1H), 7.3 (m, 1H), signals relative to the bipyridine protons; 4.2 (q, 4H, O–CH2–CH3); 3.65 (t, 1H, CH2-CH(COOEt)2); 3.3 (d, 2H, CH2–CH(COOEt)2); 1.3 (t, 6H, O–CH2–CH3).IR (NaCl cells, pure sample film) 1732 cm21 (C]] O stretch of the ester group). 1,9-Diamino-5-(2,29-bipyridin-5-ylmethyl)-3,7-diazanonane- 4,6-dione 6. 0.35 g (1.2 mmol) of compound 4 were dissolved in 15 ml of dry ethylenediamine (freshly distilled from KOH, under a nitrogen atmosphere), the solution flushed with nitrogen and maintained well stirred at room temperature for 7 days.Excess ethylenediamine was then removed on a rotary evaporator to give a yellowish solid that, on washing with diethyl ether and drying under vacuum, gave a sample of 0.3 g (yield = 70%) of compound 6 as a white, hygroscopic solid. NMR (CD3OD) d 8.6 (dd, 1H), 8.5 (ds, 1H), 8.2 (m, 2H), 7.9 (dt, 1H), 7.8 (dd, 1H), 7.4 (m, 1H), signals relative to the bipyridine protons; 3.3 (d, 2H, CH2–CH–(COONHR)2); 3.2 (t, 4H, NH–CH2–CH2); 2.8 (t, 1H, CH2–CH–(COONHR)2); 2.7 (t, 4H, CH2–CH2–NH2).[Ru(bipy)2(6)](PF6)2?NH4PF6 (2?(PF6)2?NH4PF6). 0.122 g (0.342 mmol) of ligand 6 and 0.111 g (0.213 mmol) Ru(bipy)2- Cl2?2H2O were added to 12 ml absolute ethanol, under a nitrogen atmosphere. The resulting deep violet mixture was heated to reflux for 3 hours, after which time a bright orange solution was obtained, from which the solvent was then removed on a rotary evaporator. The residue was dissolved in 5 ml H2O and treated with 0.200 g of solid NH4PF6, obtaining the precipitation of an orange-brown solid.This crude product was then purified on a column of basic Al2O3, using CH3CN as eluent and then a gradient of mixtures of CH3CN with 1 M aqueous NH4PF6. From the fractions eluted with 5% aqueous NH4PF6 a bright orange solid was obtained after slow evaporation, which was identified as 2?(PF6)2?NH4PF6?H2O. Yield: 0.065 g (24%). Anal. calc. for C38H46N11O3P3F18Ru: C 33.70, H 3.39, N 11.37%. Found: C 33.40, H 3.51, N 11.12%. Mass spec-J.Chem. Soc., Dalton Trans., 1999, 1381–1385 1383 trum (ESI) 915 ([M 1 PF6]1), 385 ([M]21/2). NMR (CD3OD): d 8.6 (m, 6H), 8.2 (m, 5H), 7.9 (m, 2H), 7.7 (m, 4H), 7.6 (m, 2H), 7.4 (m, 4H), signals relative to plain and substituted bipy protons; 2.8–3.4 (m, 11H), signals relative to the CH2– CH(CONHCH2CH2NH2)2 system. [Re(CO)3Cl(6)] 3. 0.190 g (0.53 mmol) of ligand 6 and 0.193 g (0.53 mmol) of Re(CO)5Cl were dissolved in 20 ml CH3OH and the colorless solution obtained was heated to reflux under a nitrogen atmosphere for 2.5 hours, after which time it became deep yellow.The volume of the solvent was then reduced to 5 ml on a rotary evaporator and 10 ml of diethyl ether added, to give a yellow oil which precipitated as a thin film on the flask wall. The solvent mixture was mechanically removed and the oil kept under high vacuum for 2 hours, after which time it became a well treatable solid which was collected mechanically.Yield 0.210 g (58%). Anal. calc. for 3?H2O, C21H24N6O5ClRe: C 37.10, H 3.82, N 12.35%. Found: C 37.01, H 3.90, N 12.11%. Mass spectrum (ESI): 627 ([M 2 Cl2]1). Results and discussion System 2 as a sensor for Ni21 and Cu21 in water The two-component system 2 has been prepared with the aim of obtaining a system capable of selectively signalling the presence of nickel and copper cations in water at low concentrations and in biologically relevant conditions (e.g.at pH values close Scheme 1 to the intracellular H2PO4 2/HPO4 22 or blood H2CO3/HCO3 2 buVers, i.e. 7.2 and 7.4, respectively). The solubility in water of the chosen components, [Ru(bipy)3]21 and the diamine-diamide ligand dioxo-2,3,2-tet, is fair, and the same has been observed for 2, which can be dissolved in pure water with concentrations ranging up to 1022 M. Moreover, selectivity towards Ni21 and Cu21 should be imparted by the peculiar coordinating properties of the dioxo-2,3,2-tet fragment: it is well established 10 that plain or substituted ligands of this type can bind a metal ion, in a square planar fashion, with the simultaneous release of the two protons of the amide nitrogens, with a pH-dependent complexation equilibrium (see Scheme 1).However, it should be noted that the very endergonic deprotonation of the amide groups can take place only in the presence of metal ions which profit from a large ligand field stabilization, e.g. divalent metal cations late in the 3d series.As a matter of fact only Ni21 and Cu21 within the series of M21 cations of the first transition series can promote the deprotonation of the amide nitrogens and form the neutral complex sketched in Scheme 1, with both plain 11 and substituted 3 dioxo-2,3,2-tet ligands. This makes ligands of this kind selective for these two cations and the diVerent ligand field stabilization (higher in the case of Cu21 with respect to Ni21) allows the ligand to distinguish between them: complexation with Cu21 takes place at lower pH values than in the case of Ni21 (~2 pH units: e.g.with plain dioxo- 2,3,2-tet, R = H in Fig. 1, complexation with Ni21 and Cu21 is complete at pH 7 and 9, respectively), and working with solutions buVered at the appropriate pH allows these types of ligands to bind only Cu21 even in the presence of Ni21.3,12 Complexation properties of 2 towards transition metal cations have been examined by observing the variation of its luminescence intensity (If) and lifetime (t) as a function of pH, in the presence or in the absence of transition metal cations.When no metal cations are added to solutions containing 2, If remains constant over the 2 < pH < 12 range, as has already been observed for the related system 1 (in acetonitrile/water 4 : 1 v/v mixtures). In this pH range, the luminescence quantum yield (0.030) and lifetime (440 ns) of 2 in aerated water solutions are very similar to those observed for the [Ru(bpy)3]21 chromophore under the same conditions,13 indicating that the dioxo-2,3,2-tet fragment does not substantially perturb the excited state properties of the Ru core, in agreement with what is generally found for aliphatic amine-functionalized Rupolypyridine complexes.14 On the other hand, when Ni21 or Cu21 (as their perchlorate or chloride salts) are added in 1 : 1 molar ratio with respect to system 2, the If vs.pH plot shows a typical sigmoidal profile (Fig. 1), which indicates that binding of the metal ion by the dioxo-2,3,2-tet fragment, according to equilibrium (1), takes place in the narrow pH range of the steeply descending portion of the figure.3 Fig. 1 Luminescence intensity (arbitrary units) vs. pH for solutions containing ligand 2 and a metal cation in 1 : 1 stoichiometry. The metal species referring to each curve is indicated in the plot.1384 J. Chem. Soc., Dalton Trans., 1999, 1381–1385 Thus, from Fig. 1 it can be said that complexation by 2, under the titration conditions (see the Experimental), begins at pH 5.8 and is complete at pH 6.8 for Cu21, while it begins at pH 7.5 and is complete at pH 8.5 for Ni21. In the descending If vs. pH portion can be observed, together with the intensity decrease, the appearance in the excited state decay profile of a second component with a much shorter lifetime (11 and 15 ns for Ni21 and Cu21, respectively), clearly indicating an intramolecular quenching process; this component becomes the only one present at higher pH values.The observed behaviour is not surprising, since transition metal cations are usually capable of quenching fluorescence emission of a series of fluorophores, including [Ru(bipy)3]21, through electron transfer (eT) or energy transfer (ET) mechanisms, when bound to appended and adjacent ligands.1–3 From eqn. (2), where t8 and t are the excited state lifetimes of kq = 1/t 2 1/t8 (2) free and complexed 2, respectively, intramolecular quenching rate constant values, kq, of 6.4 × 107 s21 for Cu21 and 8.9 × 107 for Ni21 can be calculated.In order to better discriminate between eT and ET processes, flash photolysis experiments were carried out on a 5 × 1025 M solution of 2 and Cu21 at pH 7 and on a 5 × 1025 M solution of 2 and Ni21 at pH 9. In both cases, after the 532 nm excitation pulse, no evidence for the presence of RuI species could be detected, while the only transient absorbing species were due to the excited states of the RuII chromophore. This result indicates that a charge separated species is not formed, or that the back electron transfer reaction is much faster than the forward one, so that the charge separated state cannot accumulate nor, as a consequence, be detected.Furthermore, steady-state quenching experiments performed at 77 K showed that the quenching process was very fast (1.5 × 107 s21 and 2.1 × 107 s21 for Cu21 and Ni21, respectively) also in frozen medium at low temperature, where the charge separated state is strongly destabilized.15 All these findings suggest that the ET transfer is the more appropriate candidate for explaining the luminescence quenching of 2.In this paper, system 2 behaves as an ON–OFF sensor for these two metal cations based on an ET mechanism, as If is reduced to less then 5% of its maximum value when complexation processes are complete, and selectivity is proven by the negligible luminescence intensity variation in the If vs.pH plots in the presence of other metal centres (1 : 1 molar ratio), such as Mn21, Fe21, Zn21 and Co21(see Fig. 1, circles); this suggests that with these metal cations no complexation takes place at the dioxo-2,3,2-tet fragment,16 in agreement with literature data for both plain and substituted (in the same position as 2) dioxo- 2,3,2-tet ligands.3,4,10 Furthermore, incorporation of Ni21 and Cu21 into the binding part of 2 was checked by titrations of solutions containing 2 and buVered at pH 8.5 and 7.0, respectively: a linear decrease of luminescence is observed, which is completely quenched after the addition of 1 equivalent (i.e.ligand/metal molar ratio 1 : 1) of metal cation. In addition, working on a solution of 2, buVered at pH 7.0, no variation in If is observed by addition of up to 2 eq. of Ni21 (or other divalent first row transition metal cations), while subsequent addition of Cu21 causes the expected luminescence quenching, demonstrating that selectivity of Cu21 on Ni21 can be obtained by choosing the correct pH value (see Fig. 2). Finally, solutions buVered at pH 7.0 and containing system 2 at concentrations as low as 1027 M revealed an easily detectable variation of If on addition of 1 eq. of Cu21, indicating that 2 is a suitable sensor for copper cations under analytically relevant conditions. Finally, it is worth mentioning that the ReI derivative, [Re(CO)3Cl(6)], ligand 3, was synthesized through a straightforward route,18,19 in an eVort to prepare another easily attainable, water soluble sensor, selective towards Ni21 and Cu21, similar to 2 and based, in this case, on the well established properties of the ReI(CO)3(bipy)Cl luminophore.18 However, although having the same binding unit as 2, the behaviour of 3 as a sensor towards Ni21 and Cu21 was less eVective.In the presence of 1 eq. of Cu21 and Ni21, in the If vs.pH curve a sigmoid curve was observed, which begins and ends, in both cases, at pH values similar to those found for 2, but If is quenched only to 30% and 60% of the full emission value, for Cu21 and Ni21, respectively. Moreover, in buVer solutions containing 3 (pH 7.2 and 8.5, for Cu21 and Ni21, respectively) additions of sub-stoichiometric quantities of aqueous metal perchlorates made If decrease, reaching the minimum value (30% and 60% of the starting value for Cu21and Ni21, respectively) at a metal cation/ligand molar ratio of 0.7 for Cu21 and 0.4 for Ni21, respectively.No changes were observed in the excited state lifetime of the chromophore 3 (51 ns). These data seem to indicate that ligand 3 is not completely available for complexation with metal cations, i.e. there is a concurrent equilibrium which involves the dioxo-2,3,2-tet binding unit. However, in water as the working medium, in the absence of any added species and in the examined concentration range (1023–1025 M), only the hypothesis of an intramolecular interaction can be put forward, possibly due to the folding of the dioxo-2,3,2-tet unit towards the Re centre, e.g.due to the dipole–dipole interaction between the amine fragment(s) and the bound Cl2 anion. Although quite exotic, this hypothesis could be supported by what has already been found for a related Re(CO)3(bipyR)Cl system (R = alkyl chains) 20 and by the fact that, on changing the medium (e.g.in methanol–water mixtures), If is quenched to a diVerent extent and, in buVered solutions, it reaches its minimum value at a diVerent cation/ligand molar ratio than in water (this, in particular, excludes the possibility that incomplete quenching and complexation could be due to some luminescent, Re-containing impurity). The observation of the excited state decay of the complexed species is in this case prevented by the quite intense residual luminescence coming from uncomplexed 3.Thus, although displaying behaviour similar to that found for 2, the residual fluorescence of system 3, even in the presence of excess Ni21 and Cu21, makes it less feasible as a sensor for these two cations. Acknowledgements Thanks are due to the Italian Ministry of University Research and Technology (MURST) for funding within the Dispositivi Supramolecolari project and to the University of Bologna (Funds for Selected Topics). Thanks are also due to Dr. Francesca Benevelli, Dr.Luca Gianelli and Centro Grandi Strumenti (Università di Pavia) for NMR and mass spectra and to Dr. Lucia Flamigni and Dr. Nicola Armaroli, Istituto Fig. 2 Luminescence intensity (arbitrary units) vs. equivalents of added Ni21 (j) and Cu21 (d) for ligand 2, in a solution buVered at pH 7.0.J. Chem. Soc., Dalton Trans., 1999, 1381–1385 1385 FRAE-CNR Bologna, for having performed the transient absorption experiments. References and notes 1 A. P. da Silva, H.Q. N. Gunaratne, T. Gunnlaugsson, A. J. M. Huxley, C. P. McCoy, J. T. Rademacher and T. E. Rice, Chem. Rev., 1997, 97, 1515. 2 (a) L. Fabbrizzi, M. Licchelli, P. Pallavicini, A. Perotti, D. Sacchi and A. Taglietti, Chem. Eur. J., 1996, 2, 75; (b) L. Fabbrizzi, M. Licchelli, P. Pallavicini, D. Sacchi and A. Taglietti, Analyst, 1996, 121, 1763. 3 L. Fabbrizzi, M. Licchelli, P. Pallavicini, A. Perotti and D. Sacchi, Angew. Chem., Int. Ed. Engl., 1994, 33, 1975. 4 V. Balzani, F.Barigelletti and L. De Cola, Top. Curr. Chem., 1989, 158, 31. 5 (a) E. Kimura, S. Wada, M. Shionoya, T. Takahashi and Y. Iitaca, J. Chem. Soc., Chem. Commun., 1990, 397; (b) E. Kimura, X. Bu, M. Shionoya, S. Wada and S. Maruyama, Inorg. Chem., 1992, 31, 4542; (c) S. Rawle, P. Moore and N. Alcock, J. Chem. Soc., Chem. Commun., 1992, 684. 6 K. Nakamaru, Bull. Chem. Soc. Jpn., 1982, 55, 2697. 7 A. Credi and L. Prodi, Spectrochim. Acta, Part A, 1998, 54, 159. 8 (a) L. Flamigni, J.Phys. Chem., 1992, 96, 3331; (b) L. Flamigni, J. Chem. Soc., Faraday Trans., 1994, 90, 2331. 9 P. Moore, S. C. Rawle and N. W. Alcock, J. Chem. Soc., Chem. Commun., 1992, 684. 10 M. Kodama and E. Kimura, J. Chem. Soc., Dalton Trans., 1979, 325. 11 H. A. O. Hill and K. A. Raspin, J. Chem. Soc. (A), 1968, 3036. 12 G. De Santis, M. Di Casa, L. Fabbrizzi, M. Licchelli, C. Mangano, P. Pallavicini, A. Perotti, A. Poggi, D. Sacchi and A. Taglietti, in Transition Metals in Supramolecular Chemistry, NATO-ASI Series, ed. L. Fabbrizzi and A. Poggi, Kluwer Academic Publishers, Dordrecht, 1996. 13 A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser and A. von Zelewski, Coord. Chem. Rev., 1988, 84, 85. 14 J. V. Caspar and T. J. Mejer, Inorg. Chem., 1983, 22, 2444. For the lack of electron transfer processes between aliphatic amines and Ru polypyridine complexes, see R. Ballardini, G. Varani, M. T. Indelli, F. Scandola and V. Balzani, J. Am. Chem. Soc., 1978, 100, 7219. 15 G. L. Gaines, III, M. P. O’Neil, W. A. Svec, M. P. Niemczyk and M. R. Wasielewski, J. Am. Chem. Soc., 1990, 113, 719. 16 DiVerently from transition metal cations, the d10 Zn21 cation is not usually capable of luminescence quenching 17 and the invariability of If vs. pH in the presence of 1 eq. of Zn21 cannot in principle exclude its complexation; however, competition experiments carried out with 1 eq. of 2, one eq. of Ni21 or Cu21 and a 50 eq. excess of Zn21 rule out this possibility (If vs. pH profiles are found to be identical to those found in the absence of Zn21), in agreement with literature data 10 that exclude complexation of Zn21 by related ligands. 17 E. U. Akkaya, M. E. Huston and A. W. Czarnik, J. Am. Chem. Soc., 1990, 112, 3590. 18 (a) I. Costa, L. Fabbrizzi, P. Pallavicini, A. Poggi and A. Zani, Inorg. Chim. Acta, 1998, 275–276, 117; (b) Y. Shen and P. Sullivan, Inorg. Chem., 1995, 34, 6235. 19 K. C. Schanze, D. B. MacQueen, T. A. Perkins and L. Cabana, Coord. Chem. Rev., 1993, 122, 63. 20 G. A. Reitz, J. N. Demas, B. A. DeGraV and E. M. Stephens, J. Am. Chem. Soc., 1988, 110, 5051. Paper 8/08476I

ISSN:1477-9226    

DOI:10.1039/a808476i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 1387–1391 1387 Compounds in which the Mo2 41 unit is embraced by one, two or three formamidinate ligands together with acetonitrile ligands Malcolm H. Chisholm,*a F. Albert Cotton,*b Lee M. Daniels,b Kirsten Folting,a John C. HuVman,a Suri S. Iyer,a Chun Lin,b Ann M. Macintosh a and Carlos A. Murillo *b a Department of Chemistry and Molecular Structure Center, Indiana University, Bloomington, IN 47405, USA b Department of Chemistry and Laboratory for Molecular Structure and Bonding, Texas A&M University, PO Box 30012, College Station, TX 77842-3012, USA Received 13th January 1999, Accepted 1st March 1999 The cationic complexes [Mo2(DPhF)(MeCN)6]31[BF4 2]3?MeCN 1a [throughout this paper the formamidinate anions, RNC(H)NR2, will be abbreviated as DRF2, with specific aryl groups represented by Ph for phenyl, and Ani for p-anisyl], [Mo2(DAniF)(MeCN)6]31[BF4 2]3?1.59MeCN 1b, cis-[Mo2(DPhF)2(MeCN)4]21[BF4 2]2?MeCN 2a, cis- [Mo2(DAniF)2(MeCN)4]21[BF4 2]2?2MeCN 2b, and [Mo2(DPhF)3(MeCN)2]1[BF4 2] 3, have been prepared from the reactions between Mo2(DArF)4 and HBF4?Et2O or (Me3O)BF4 in an appropriate stoichiometry in acetonitrile or CH2Cl2–MeCN mixtures.A better procedure for the preparation of 3 involves the reaction between 2a and Li1(DPhF2) in acetonitrile. Compound 2a undergoes a reaction in pyridine to give the compound trans-[Mo2(DPhF)2- (py)4]21[BF4 2]2 4. The new compounds 1a, 1b, 2a, 2b and 4 have been structurally characterized and shown to contain Mo–Mo quadruple bonds.Introduction In previous work1 attempts have been made to link M–M multiply bonded complexes, bridged by carboxylates, together to form parallel or perpendicular polymers of the type shown below in I and II, respectively. Unfortunately for such oligomers of molybdenum and tungsten supported by carboxylate ligands, ligand scrambling occurred and rendered the polymers and smaller oligomers kinetically labile.Indeed, even a linked dimer of “dimers” was susceptible to disproportionation of the type shown in eqn. (1). [M2(O2CR)3]2(bridge) M2(O2CR)4 1 1/n[M2(O2CR)2(bridge)]n (1) Reactions of type (1) arise because the M2 41 centers, where M = Mo and W, are kinetically labile to substitution reactions. Carboxylate scrambling can be catalyzed by adventitious carboxylate anions or acid present in solution.2 In order to minimize such facile ligand scrambling we reasoned that the use of ligands lacking the active lone pairs of carboxylates might lead to more kinetically persistent linked di- and poly-nuclear species. As the first part of the continuing project in developing such systems, we report here the synthesis and characterization of cationic formamidinate bridged dimolybdenum complexes which, unlike their Mo2–carboxylate counterparts, are much less labile to ligand scrambling.A few examples of how they might be linked have been reported,3 but many more will be described in the near future.4 Results and discussion Synthesis This investigation was carried out independently by the groups at Indiana University and at Texas A&M University; one worked with the ligand DPhF [PhNC(H)NPh2], while the other worked with the anisyl (Ani) analogue, DAniF [AniNC(H)- NAni2].Though the synthetic procedures employed by both groups are fairly similar, there are important diVerences which make some comparison worthwhile.The general synthetic strategy for the formation of phenyl derivatives of Mo2- (DPhF)4 2 n n1 cationic complexes is shown in eqn. (2). Mo2(DPhF)4 1 2nHBF4?Et2O 25 8C MeCN [Mo2(DPhF)4 2 n(MeCN)2n]n1[BF4 2]n 1 n[H2DPhF]1[BF4 2]n (2) Reaction (2) is complicated by the fact that Mo2(DPhF)4 is very sparingly soluble in MeCN as a solvent. Thus, the reaction proceeds by the addition of HBF4?Et2O to a slurry of Mo2(DPhF)4. The resultant cationic Mo2 species are more soluble and thus go into solution.As a consequence the addition of HBF4?Et2O must proceed slowly if any control of product is desired beyond formation of the Mo2(MeCN)8 41 and Mo2- (DPhF)(MeCN)6 31 salts. To this end it is desirable to dilute the HBF4?Et2O in MeCN solution and to add the acid slowly. All the species have somewhat diVerent solubilities and colors. The Mo2(MeCN)8 41 cation is blue; Mo2(DPhF)- (MeCN)6 31 1a is purple; Mo2(DPhF)2(MeCN)4 21, 2a is red and Mo2(DPhF)3(MeCN)2 1 3 is pumpkin orange while Mo2- (DPhF)4 is yellow.The most diYcult compound in the series to obtain is [Mo2(DPhF)3(MeCN)2]1[BF4 2] 3, which in our hands was best prepared by the reaction shown in eqn. (3). [Mo2(DPhF)2(MeCN)4]21[BF4 2]2 1 LiDPhF 25 8C MeCN [Mo2(DPhF)3(MeCN)2]1[BF4 2] 1 Li1BF4 2 (3)1388 J. Chem. Soc., Dalton Trans., 1999, 1387–1391 Dissolving [Mo2(DPhF)2(MeCN)4]21[BF4 2]2 in pyridine leads to a facile replacement of the MeCN ligands by pyridine and a slow isomerization to give trans-[Mo2(DPhF)2(py)4]21[BF4 2]2 4.The anisyl analogues were prepared by slightly modified procedures which oVered more control over the reaction, especially for the preparation of 2b. Compound 1b was prepared in a 1 : 4 mixture of acetonitrile and dichloromethane. Mo2(DAniF)4 1 HBF4?Et2O(excess) 25 8C MeCN–CH2Cl2 [Mo2(DAniF)(MeCN)6][BF4 2]3 1 3 [H2DAniF][BF4] (4) Under these conditions it is not necessary to control the rate of addition of HBF4?Et2O. Furthermore, Mo2(DAniF)4 and [H2DAniF][BF4] are soluble in this mixture, and therefore the only species that precipitates is the purplish 1b.For the preparation of 2b we found that the use of wet (Me3O)BF4 is more advantageous than neat HBF4?Et2O. Mo2(DAniF)4 1 (Me3O)BF4(excess) 1 4H2O 25 8C MeCN [Mo2(DAniF)2(MeCN)4][BF4 2]2 1 2 [H2DAniF][BF4] 1 Me2O 1 4 MeOH (5) An acetonitrile slurry of (Me3O)BF4 and Mo2(DAniF)4 does not react but addition of a small amount of deoxygenated water slowly dissolves all of the Mo2(DAniF)4 producing a clear red solution.After removal of the solvent, the solid is washed with Et2O, then recrystallized from a mixture of CH2Cl2– MeCN which is layered with Et2O to produce an essentially quantitative yield of 2b. Structural characterizations [Mo2(DPhF)(MeCN)6]31[BF4 2]3?MeCN 1a. In the unit cell of space group P21/n there are two [Mo2(DPhF)(MeCN)6]31 cations and six BF4 2 anions. The Mo–Mo distances, 2.15(1) Å, are essentially identical for both molecules and the gross structural features are very similar.The Mo–N (formamidinate) distances 2.08(1) Å (average) are shorter by 0.05 to 0.10 Å than the Mo–N distances to the acetonitrile ligands. Each Mo atom is coordinated to four N atoms that lie roughly in a plane and the central Mo2N8 skeleton is virtually eclipsed as expected for Mo–Mo quadruply bonded complexes. In addition, there is a weak, 2.58 Å, interaction between Mo(1) and an axial MeCN.Selected bond distances are given in Table 1. [Mo2(DAniF)(MeCN)6]31[BF4 2]3?1.59MeCN 1b. Compound 1b crystallizes in space group P1� with the three BF4 2 groups and the weakly-interacting axial MeCN groups all disordered to some extent. One axial position is fully occupied although the MeCN group has two orientations in the approximate ratio of 53 :47. The MeCN group at the other axial position has a refined occupancy of 59%. Two of the BF4 2 sites were modeled as two interpenetrating tetrahedra. The third anion did not Table 1 Selected bond distances (Å) for [Mo2(DPhF)(MeCN)6]31- [BF4 2]3?MeCN 1a Mo(1)–Mo(2) Mo(1)–N(9) Mo(1)–N(18) Mo(1)–N(21) Mo(1)–N(24) Mo(1)–N(27) Mo(2)–N(11) Mo(2)–N(30) Mo(2)–N(33) Mo(2)–N(36) 2.149(1) 2.086(8) 2.58(1) 2.150(9) 2.19(1) 2.124(9) 2.078(8) 2.14(1) 2.168(9) 2.11(1) Mo(39)–Mo(40) Mo(39)–N(47) Mo(39)–N(56) Mo(39)–N(59) Mo(39)–N(62) Mo(39)–N(65) Mo(40)–N(49) Mo(40)–N(68) Mo(40)–N(71) Mo(40)–N(74) 2.151(1) 2.098(8) 2.567(9) 2.129(9) 2.165(9) 2.15(1) 2.076(8) 2.139(9) 2.173(9) 2.13(1) behave as well and the ensemble of electron density was simply modeled as a group of 8 F atoms of variable occupancy, with the total occupancy constrained to equal 4.Again, the Mo2N8 core is eentially eclipsed (torsion angles all less than 28) as expected. The axial interactions are very weak, with Mo–N distances of 2.70 and 2.82 Å. A view of the molecule is given in Fig. 1. Selected bond distances are given in Table 2.cis-[Mo2(DPhF)2(MeCN)4]21[BF4 2]2?MeCN 2a. Compound 2a crystallizes in the space group P1� with two formula units in the unit cell. One MeCN molecule is weakly ligated in an axial position to one of the Mo atoms, namely Mo(2) via a long, 2.595(3) Å, interaction. Selected bond distances are given in Table 3. The two m-formamidinate ligands are mutually cis and the shortest Mo–N distances involve formamidinate nitrogen atoms that are bound to Mo(1) which has only two attendant MeCN ligands.The Mo–Mo distance of 2.1457(7) Å is comparable to that in 1 and only somewhat longer than that in Mo2(DPhF)4, 2.0944(8) Å.5 cis-[Mo2(DAniF)2(MeCN)4]21[BF4 2]2?2MeCN 2b. Compound 2b also crystallizes in the space group P1� ; the structure is very similar to that of its phenyl analogue. The Mo–Mo distance of 2.1439(6) Å is equivalent to that of 2a. A view of the dication, shown in Fig. 2, reveals that a coordinated acetonitrile molecule is located at a distance of 2.590(4) Å from Mo(1).All other chemically equivalent dimensions are comparable for 2a and 2b. Selected bond distances are given in Table 4. Fig. 1 A view of the [Mo2(DAniF)(MeCN)6]31 cation in 1b, with ellipsoids drawn at the 40% probability level. Table 2 Selected bond distances (Å) for [Mo2(DAniF)(MeCN)6]31- [BF4 2]3?1.59MeCN 1b Mo(1)–Mo(2) Mo(1)–N(1) Mo(1)–N(3) Mo(1)–N(5) Mo(1)–N(7) 2.152(1) 2.09(1) 2.14(1) 2.17(1) 2.13(1) Mo(2)–N(2) Mo(2)–N(4) Mo(2)–N(6) Mo(2)–N(8) 2.09(1) 2.15(1) 2.14(1) 2.13(1) Table 3 Selected bond distances (Å) for cis-[Mo2(DPhF)2- (MeCN)4]21[BF4 2]2?MeCN 2a Mo(1)–Mo(2) Mo(1)–N(3) Mo(1)–N(18) Mo(1)–N(33) Mo(1)–N(36) 2.1457(7) 2.086(3) 2.100(3) 2.181(3) 2.183(3) Mo(2)–N(5) Mo(2)–N(20) Mo(2)–N(39) Mo(2)–N(42) Mo(2)–N(45) 2.118(3) 2.105(3) 2.180(3) 2.185(3) 2.595(3)J.Chem. Soc., Dalton Trans., 1999, 1387–1391 1389 Fig. 2 A drawing of the cation cis-[Mo2(DAniF)2(MeCN)4]21 in 2b, with ellipsoids drawn at the 40% probability level.[Mo2(DPhF)2(py)4]21[BF4 2]2?4py 4. The asymmetric unit contained one half of the dinuclear unit together with one BF4 2 anion and two molecules of interstitial pyridine. One of the solvent molecules C(40)–C(51) was severely disordered. However, the remainder of the structure was well behaved. A drawing of the centrosymmetric dication is given in Fig. 3 which shows the trans arrangement of the formamidinate ligands. Also in Fig. 3 we show a view looking down the Mo–Mo bond which reveals the eclipsed geometry of the ligands and the favorable p–p stacking of the pyridine ligands.Selected bond distances and angles are given in Table 5. Of note is the fact that the Mo–N distances are longer by ca 0.1 Å to the pyridine nitrogen atoms than those to the formamidinate nitrogen atoms. Also the Mo–Mo–N angles to the pyridine ligands are larger to accommodate a more favorable p–p stacking. Electronic spectra The color of the various anisyl derivatives changes from blue in the unsubstituted Mo2(MeCN)8 41 cation, to purple for Mo2- (DAniF)(MeCN)6 31 1b, to red for Mo2(DAniF)2(MeCN)4 21 2b, and to yellow for Mo2(DAniF)4.Their lowest transition ener- Table 4 Selected bond distances (Å) for cis-[Mo2(DAniF)2- (MeCN)4]21[BF4 2]2?2MeCN 2b Mo(1)–Mo(2) Mo(1)–N(1) Mo(1)–N(3) Mo(1)–N(7) Mo(1)–N(5) 2.1439(6) 2.115(3) 2.117(3) 2.185(3) 2.188(3) Mo(2)–N(2) Mo(2)–N(4) Mo(2)–N(6) Mo(2)–N(8) 2.085(3) 2.087(3) 2.172(3) 2.180(3) Table 5 Selected bond distances (Å) and angles (8) for trans-[Mo2- (DPhF)2(py)4]21[BF4 2]2 4 Mo(1)–Mo(19) Mo(1)–N(2) Mo(1)–N(4) Mo(19)–Mo(1)–N(2) Mo(19)–Mo(1)–N(17) 2.107(2) 2.151(6) 2.132(6) 92.6(2) 99.6(2) Mo(1)–N(17) Mo(1)–N(23) Mo(19)–Mo(1)–N(4) Mo(19)–Mo(1)–N(23) 2.226(6) 2.243(5) 92.3(2) 103.8(2) gies, which were assigned to dÆd* transitions, are 597, 570, 516 and 430 nm, respectively.This spectroscopic blue shift supports the idea that by increasing the number of the formamidinate ligands, the Mo–Mo bond strength increases.This correlates with the decrease of their Mo–Mo bond distances, which are 2.180,6 2.152, 2.144 and 2.096 Å,5 respectively. Reactivity studies of the phenyl substituted formamidinates In CD3CN as solvent the cationic complexes 1a and 2a exhibit facile CH3CN for CD3CN exchange. Similarly, for 4 exchange of coordinated py ligands with py-d5 is rapid. However, our key observations involve the lack of exchange of the formamidinate ligands.Thus, compound 2a and Mo2(DPhF)4 failed to yield the compound [Mo2(DPhF)3(MeCN)2]1[BF4 2] as judged by NMR spectroscopy in acetonitrile solvent. In part, this could have been attributed to the insolubility of Mo2(DPhF)4. However, a reaction between compound 2a and Li(DPhF) in MeCN did yield 3 in conjunction with some formation of Mo2(DPhF)4 (which could be removed by filtration). Because the solubilities of complex 2a and complex 3 are very similar, the optimum preparation of 3 involves the reaction between 2a and Li- (DPhF) (1.5 equivalents) in MeCN which results in 90% of 3 and 10% of 2a after filtration to remove Mo2(DPhF)4. We also attempted to study the comproportionation of the dicationic complex 2a with [Mo2(MeCN)10]41[BF4 2]4 7 in CD3CN.Here both species are fully dissolved. However, we observed no ligand exchange leading to the formation of the tricationic complex 1 as judged by the 1H NMR spectroscopy.Thus, in contrast to carboxylate ligands, we find the formamidinate ligands to be relatively inert to ligand scrambling. The reaction between compound 3 and the oxalate dianion was studied as shown in eqn. (6). 2[Mo2(DPhF)3(MeCN)2]1[BF4 2] 1 [Bun 4N1]2C2O4 25 8C MeCN [Mo2(DPhF)2]2(m-O2CCO2) 1 2[Bun 4N1][BF4 2] (6) The product of the reaction was a red-orange powder essentially insoluble in CD3CN, CD2Cl2 and py-d5. Given the recent1390 J. Chem. Soc., Dalton Trans., 1999, 1387–1391 report by some of us,3 we presume that reaction 6 provides an alternate synthesis of the m-oxalate type complex.Concluding remarks In this work we have prepared dimolybdenum cationic complexes supported by one, two and three formamidinate ligands. These have been shown to be notably less labile to ligand scrambling than their related carboxylate counterparts. The success in the preparation of the m-oxalate linked dimer of “dimers” in reaction (6) suggests that formamidinate ligands may lead to kinetically more persistent quartets (dimers of “dimers”) and higher oligomers.Experimental All manipulations were carried out under an inert atmosphere Fig. 3 An ORTEP10 drawing of [Mo2(DPhF)2(py)4][BF4]2 4, front-on view (top) and looking down the metal–metal bond axis (bottom), with the hydrogen atoms and the BF4 counter anions omitted for clarity. Thermal ellipsoids are drawn at the 50% probability level. by using standard Schlenk and glove-box techniques.All solvents were dried and degassed by standard methods and distilled prior to use. 1H NMR spectra were recorded on Varian Gemini-2000 and Varian XL-200E NMR spectrometers and referenced to residual protio impurities in the deuteriated solvent. Infrared spectra were obtained as KBr pellets on Nicolet 510 P FT-IR or Perkin-Elmer 16PC FTIR spectrometers. Elemental analyses were performed by Atlantic Microlabs, Norcross, GA or Canadian Microanalytical Service, Delta, BC.Tetrafluoroboric acid (54 wt.% or 85 wt.% in diethyl ether), N,N9-diphenylformamidine and n-butyllithium (1.6 M in hexanes) were purchased from Aldrich Chemical Company and used as received. Mo2(DPhF)4 was prepared according to the method described previously for Mo2[(p-tol)- NCHN(p-tol)]4 (p-tol = p-tolyl);8 the anisyl analogue Mo2- (DAniF)4 prepared by following a published procedure.5 Li(DPhF) was obtained by neutralizing a THF solution of N,N9-diphenylformamidine with 1 equivalent of n-butyllithium at 278 8C.After warming to room temperature, the volatile components were removed under vacuum and the resulting oV- white solid was washed with hexane. Syntheses [Mo2(DPhF)(MeCN)6][BF4]3?MeCN 1a. A solution of 0.84 mL (6.4 mmol) of HBF4?Et2O in 25 mL of MeCN was added dropwise over a 45 min period to a slurry containing 1.0 g (1.02 mmol) of Mo2(DPhF)4 in 25 mL of acetonitrile. Over this time, the color of the solution changed from pale yellow to plum.The solution was stirred for an additional 2 h to ensure completion and then filtered to remove unreacted Mo2(DPhF)4. The reaction mixture was concentrated to 10 mL after which 25 mL of diethyl ether was added resulting in the formation of a purple precipitate. The solid was isolated by filtration through a medium frit and dried under vacuum to yield 0.73 g (76%). Deep red crystals suitable for X-ray analysis were obtained by cooling a concentrated solution of 1, in MeCN, to 215 8C for several days. 1H NMR (CD3CN): d 9.05 (s, 1H), 7.3 (m, 4H), 7.2 (m, 6H), 1.95 (s, 11H). IR (cm21): 1599w, 1520m, 1450m, 1322m, 1217m, 1063vs (br), 761s, 699m, 692m, 520w, 461m. [Mo2(DAniF)(MeCN)6][BF4]3?1.59MeCN 1b. To a stirred solution of Mo2(DAniF)4 (150 mg, 0.124 mmol) in 20 mL of CH2Cl2 and 5 mL of MeCN, was added 0.3 mL of HBF4?Et2O (85% in Et2O). The yellow color quickly changed to red then purple. The reaction mixture was stirred at room temperature for 30 min.Diethyl ether (20 mL) was added. Then the solvent was decanted oV and the solid residue was extracted into MeCN (3 mL) and then filtered. The crystalline product was obtained by addition of CH2Cl2 (30 mL) to the filtrate. Yield, 100 mg (85%). Purple single crystals suitable for X-ray analysis were grown by diVusion of Et2O into a MeCN solution. 1H NMR (CD3CN): d 8.97 (s, 1H, NCHN), 7.15 (d, 4H, aromatic, 3J = 9.0), 6.88 (d, 4H, aromatic, 3J = 9.0 Hz), 3.75 (s, 6H, OCH3).IR (KBr, cm21): 2374w, 2338w, 2283w, 2050w, 1654w, 1637w, 1608m, 1528m, 1505s, 1461m, 1442m, 1401m, 1298m, 1246s, 1214s, 1175s, 1082s, 1035s, 832m, 805w, 767w, 593w, 533w, 522w, 464w. UV-vis, lmax/nm (e/M21 cm21): 570 (3130), 460 (2470), 290 (sh). [Mo2(DPhF)2(MeCN)4][BF4]2?MeCN 2a. To a slurry of Mo2(DPhF)4 (0.51 g, 0.51 mmol) in 25 mL of MeCN, a solution of HBF4?Et2O (0.28 mL, 2.1 mmol dissolved in 25 mL of MeCN) was added dropwise over a period of several hours.The addition of HBF4 resulted in a color change from pale yellow to deep red. The mixture was then stirred for an additional 2 h, filtered to remove any unreacted Mo2(DPhF)4, concentrated and cooled to 215 8C. Over a period of 2 days a red precipitate formed and was isolated by filtration through a medium frit and dried under vacuum to yield 0.343 g (72%). Deep red crystals ofJ. Chem. Soc., Dalton Trans., 1999, 1387–1391 1391 Table 6 Summary of crystal data Empirical formula M Crystal system Space group a/Å b/Å c/Å a/8 b/8 g/8 U/Å3 ZT /K Dc/g cm23 Crystal size/mm m(Mo-Ka)/cm21 2q range/8 No.unique data R(F) a or R(F 2, all data) b Rw(F) a or Rw(F 2, all data) b Goodness of fit 1a C27H32Mo2B3N9F12 934.90 Monoclinic P21/n 23.831(3) 12.218(1) 26.018(3) 94.13(0) 7555.76 8 104 1.644 0.25 × 0.25 × 0.40 7.377 6–45 9840 0.0539 a 0.0473 a 1.123 1b C30.18H37.77B3F12Mo2N9.59O2 1019.20 Monoclinic P1� 11.017(8) 11.065(8) 18.77(2) 87.28(3) 85.49(3) 82.67(5) 2260(4) 2 273 1.498 0.10 × 0.20 × 0.30 6.43 4.4–45 5562 0.114 b 0.255 b 1.256 2a C36H37B2F8Mo2N9 961.24 Triclinic P1� 12.027(3) 17.335(5) 10.341(3) 91.76(1) 109.71(2) 80.43(1) 2000.52 2 103 1.596 0.10 × 0.17 × 0.20 7.027 6–50 7101 0.0360 a 0.0352 a 0.955 2b C42H48B2F8Mo2N10O4 1122.40 Triclinic P1� 12.323(3) 13.155(2) 16.817(4) 94.04(1) 94.78(1) 111.90(1) 2505.4(9) 2 213 1.488 0.25 × 0.25 × 0.20 5.800 4–46 6490 0.058 b 0.134 b 1.071 4 C66H62B2F8Mo2N12 1388.79 Monoclinic C2/c 27.753(6) 12.462(2) 18.715(4) 106.51(1) 6205.83 4 98 1.486 0.15 × 0.15 × 0.30 4.800 6–45 4080 0.0632 a 0.0575 a 1,359 a R(F) = S Fo| 2 |Fc /S|Fo|, Rw(F) = [Sw(|Fo| 2 |Fc|)2/SwFo 2]� �� .b R(F 2) = [Sw(Fo 2 Fc)2/SwFo 2]� �� , Rw(F 2) = {S[w(Fo 2 2 Fc 2)2]/Sw(Fo 2)2}� �� . 2 suitable for X-ray analysis were obtained by slow evaporation of an acetonitrile solution. 1H NMR (CD3CN): d 8.8 (s, 1H), 7.1 (m, 6H), 6.7 (d, 4H, J = 8.7 Hz), 1.95 (s, 9H). IR (cm21): 2849w, 2775vw, 2363w, 1577s, 1491s, 1233m, 1094s (br), 841w, 776m.(Calc. for Mo2C34H34N8B2F8: C, 44.38; H, 3.73; N, 12.18. Found: C, 44.48; H, 3.88; N, 11.85%). [Mo2(DAniF)2(MeCN)4][BF4]2?2MeCN 2b. A 100 mL, threenecked, round-bottomed flask was charged with 242 mg (0.20 mmol) of Mo2(DAniF)4, 178 mg (1.20 mmol) of (Me3O)BF4, and 60 mL of MeCN. To the stirred suspension was added a small amount of deoxygenated H2O (ca. 2–3 drops). The resulting mixture was stirred at room temperature for ca. 5 h, after which time all of the yellow starting material had reacted to give a clear red solution. After the removal of solvent, the red residue was washed with Et2O (2 × 10 mL), then extracted with CH2Cl2–MeCN (19: 1, 2 × 5 mL). Diethyl ether (75 mL) was then carefully added to the extract, and the solution was stored for 24 h at room temperature. Large orange-red block-shaped crystals formed; they were collected by filtration and dried for 4 h in vacuo. The yield was essentially quantitative.Single crystals suitable for X-ray analysis were grown by diVusion of Et2O into an MeCN solution. 1H NMR (CD3CN): d 8.68 (s, 2H, NCHN), 6.69 (m, 16H, aromatic), 3.69 (s, 12H, OCH3). IR (KBr, cm21): 2373w, 2314w, 2283w, 2045w, 1695m, 1610m, 1532s, 1506s, 1464m, 1443m, 1342w, 1291m, 1249s, 1215s, 1178m, 1083s, 1030s, 830m, 766w, 723w, 592w, 526w. UV-vis, lmax/nm, (e/M21 cm21): 516 (2320), 410 (3430), 287 (sh). (Calc. for C38H42B2F8Mo2N8O4: C, 43.87; H, 4.07; N, 10.77.Found: C, 43.23; H, 4.07; N, 10.10%). [Mo2(DPhF)3(MeCN)2][BF4] 3. A 100 mL Schlenk flask was charged with 0.30 g (0.33 mmol) of 2 and 0.096 g (0.49 mmol) of Li(DPhF). To this mixture was added 35 mL of MeCN. The reaction was stirred at room temperature for 12 h, after which time the solution was filtered to remove the Mo2(DPhF)4, which formed as a result of 3 undergoing additional reaction with Li(PhNCHPh). Removal of the volatile components under dynamic vacuum yielded a pumpkin colored powder that contained 90% of [Mo2(DPhF)3(MeCN)2][BF4] and 10% of [Mo2(DPhF)2(NCMe)4][BF4]2 by 1H NMR. 1H NMR (CD3CN): d 9.1 (s, 2H), 8.4 (s, 1H), 7.1 (m, 8H), 7.0 (m, 4H), 6.8 (m, 6H), 6.7 (d, 8H), 6.2 (d, 4H), 1.95 (s, 5H). IR (cm21): 2197vw, 1696m, 1595m, 1534vs, 1489vs, 1319s, 1215s, 1063s (br), 758s, 696m, 517w, 434w. [Mo2(DPhF)2(py)4][BF4]2 4. To a flask containing 0.2 g (0.19 mmol) of 2 was added 25 mL of pyridine. The reaction was stirred at room temperature for 1 h.After 15 min the color of the solution changed from red to orange. The reaction mixture was filtered through a fine frit, concentrated and left undisturbed at ambient temperature. Hexagonal crystals of 4 suitable for X-ray analysis were obtained after 1 week. The yield is essentially quantitative. 1H NMR (pyridine-d5): d 9.9 (s, 1H), 7.6 (d, 4H), 7.5 (m, 6H), 7.2 (m, 25H), 6.7 (d, 4H). IR (cm21): 2900w, 1600s, 1520s, 1310s, 1094 vs (br), 841w, 776m (Calc. for Mo2C66H62N12B2F8: C, 57.08; H, 4.50; N, 12.10. Found: C, 57.10; H, 4.58; N, 12.11%). Crystallographic study General operating procedures and listings of programs have been described.4,9 A summary of crystal data is given in Table 6. CCDC reference number 186/1368. See http://www.rsc.org/suppdata/dt/1999/1387/ for crystallographic files in .cif format. Acknowledgements We thank the National Science Foundation for support. References 1 R. H. Cayton, M. H. Chisholm, J. C. HuVman and E. B. Lobkovsky, J. Am. Chem. Soc., 1991, 113, 8709. 2 M. H. Chisholm and A. M. Macintosh, J. Chem. Soc., Dalton Trans., 1999, 1205. 3 F. A. Cotton, C. Lin and Coc., Dalton Trans., 1998, 3151. 4 F. A. Cotton, L. M. Daniels, C. Lin and C. A. Murillo, J. Am. Chem. Soc., in press. 5 C. Lin, J. D. Protasiewicz, E. T. Smith and T. Ren, Inorg. Chem., 1996, 35, 6422. 6 F. A. Cotton, L. M. Daniels, C. A. Murillo and X. Wang, Polyhedron, 1998, 2781. 7 F. A. Cotton and K. J. Wiesinger, Inorg. Synth., 1992, 29, 134. 8 F. A. Cotton, X. Feng and M. Matusz, Inorg. Chem., 1989, 28, 594. 9 M. H. Chisholm, K. Folting, J. C. HuVman and C. C. Kirkpatrick, Inorg. Chem., 1984, 23, 1021. 10 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. Paper 9/00389D

ISSN:1477-9226    

DOI:10.1039/a900389d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 1393–1399 1393 Periodic trends in metal–metal bonding in cubane clusters, (C5H5)4M4E4 [M 5 Cr, Mo, E 5 O, S] John E. McGrady Department of Chemistry, The University of York, Heslington, York, UK YO10 5DD. E-mail: jem15@york.ac.uk Received 25th January 1999, Accepted 17th March 1999 Trends in structure and bonding in a series of metal cubane clusters are examined using broken-symmetry density functional theory. For the metal–sulfur clusters, (C5H5)4Mo4S4 and (C5H5)4Cr4S4, the twelve cluster valence electrons are delocalised in six metal–metal single bonds, giving an approximately tetrahedral metal core.In (C5H5)4Cr4O4, however, no strong Cr–Cr bonds are present, and three cluster valence electrons remain localised on each of the chromium centres. Antiferromagnetic coupling across four of the six edges of the tetrahedron, and ferromagnetic coupling across the remaining two give rise to a spin-singlet ground state and a distinct rhombic distortion. The driving force for the distortion is only 12 kJ mol21, and consequently inter- and intra-molecular steric eVects may play a major role in determining the structure of the cluster in the solid state.Both chromium clusters have low-lying excited states in which the bonding pattern is completely reversed, with six Cr–Cr bonds present in (C5H5)4Cr4O4 but none in (C5H5)4Cr4S4. In each case the excited state lies less than 45 kJ mol21 above the ground state, despite the fact that a substantial structural rearrangement is involved.Changes in metal–metal bond strength and spin polarisation energy are found to contribute approximately equally to the periodic trend towards electron localisation in the chromium clusters. Introduction The role of iron–sulfur clusters in biological electron transfer is well documented,1 and as a result their physical properties have been studied using a wide variety of spectroscopic 2–4 and theoretical 5 techniques.Until very recently, all available data indicated that there was no direct metal–metal bonding present in these clusters under physiological conditions, but this view has been challenged by recent EXAFS studies on the Fe protein of the enzyme nitrogenase.6 Reduction of the single Fe4S4 cluster to the all-ferrous oxidation state causes substantial structural changes, consistent with the formation of Fe–Fe bonds. This observation presents the intriguing possibility that reversible redox induced formation of metal–metal bonds may play a previously unsuspected role in the control of biological electron transfer.A wide variety of model cubane clusters have been synthesised over the years, the most extensive series being the cyclopentadienyl-capped systems, (C5R5)4M4E4 (M = Mo, Cr, Ru, Ti, V, Fe, Ir, Co; E = O, S) illustrated in Chart 1 and Table 1. Within this closely related series, the cubane unit shows a high degree of structural flexibility, from the highly distorted arrangement in (C5H5)4Ru4S4 21,15 to the almost perfect tetra- Chart 1 Structure of the (C5H5)4M4E4 unit.hedral cores found in (C5H4 iPr)4Mo4S4 and (C5H5)4Cr4S4.8,12,13 On the basis of isoelectronic relationships, a tetrahedral core might also be anticipated for the chromium–oxygen clusters, (C5H5 2 xMex)4Cr4O4, x = 0,1,5,9–11 but in fact, only the pentamethylated system, (C5Me5)4Cr4O4, exhibits six approximately equivalent Cr–Cr distances.11 In (C5H4Me)4Cr4O4, a rhombic distortion compresses the tetrahedron along one two-fold axis,9 giving four short and two long Cr–Cr separations, while in (C5H5)4Cr4O4 an additional twist about the principal axis further reduces the symmetry, giving three distinct pairs of Cr–Cr distances.10 The contrast between (C5H5 2 xMex)4Cr4O4 and (C5H4 iPr)4Mo4S4 is further illustrated by their magnetic properties: the chromium clusters are all antiferromagnetic, with substantial room temperature magnetic moments, while the molybdenum species is diamagnetic.A clue to the origin of the diVerences between the two classes of cluster comes from a comparison of the photoelectron spectra of (C5H5)4Cr4O4 and (C5H4 iPr)4Mo4S4.22 The absence of resolved structure in the metal ionisation band in the former prompted the authors 22 to suggest that the metal-based electrons in the chromium system are very weakly coupled, in contrast to the strong bonding present in the molybdenum analogue.If this is indeed the case, then the comparison of the molybdenum–sulfur systems with their chromium–oxygen analogues provides an ideal opportunity to examine the factors which influence the balance between electron localisation and delocalisation in cubane clusters. As a result, it may shed light on the possible role of redox induced bond formation in the biological systems. Several attempts to rationalise the structural properties of the metal–sulfur systems have been made using bonding models based on the interaction of four equivalent (C5H5)M fragments.23–26 Each fragment has three orbitals, dz2, dxy and dx2 2 y2, which are approximately non-bonding with respect to the ligands, and therefore available for metal–metal bonding.The dz2 orbitals are directed radially towards the centre of the cluster, whilst dxy and dx2 2 y2 form a degenerate pair oriented tangentially around the circumference of the cluster (Fig. 1). In the limit of perfect tetrahedral symmetry [assuming eVective1394 J. Chem. Soc., Dalton Trans., 1999, 1393–1399 Table 1 Crystallographically determined structural parameters of metal–sulfur and metal–oxygen cubane clusters Complex (C5Me5)4Ti4S4 (C5H4Me)4V4S4 1 (C5H4Me)4V4S4 (C5H4 iPr)4Mo4S4 21 (C5H4 iPr)4Mo4S4 1 (C5H4Me)4Cr4O4 (C5H5)4Cr4O4 (C5Me5)4Cr4O4 (C5H5)4Cr4S4 (C5H4Me)4Cr4S4 (C5H4 iPr)4Mo4S4 (C5H5)4Fe4S4 21 (C5H4Me)4Ru4S4 21 (C5H5)4Fe4S4 1 (C5H5)4Fe4S4 (C5H4Me)4Ru4S4 (C5Me5)4Ir4S4 21 (C5H5)4Co4S4 1 (C5Me5)4Ir4S4 (C5H5)4Co4S4 Cluster valence electron count 4 78 10 11 12 12 12 12 12 12 18 18 19 20 20 22 23 24 24 r(M–M)/Å 2 × 2.930 4 × 3.008 2.852–2.855 2.868–2.884 2 × 2.805 4 × 2.894 2.860–2.923 2 × 2.896 4 × 2.759 2 × 2.896 2 × 2.823 2 × 2.706 2.828–2.840 2.818–2.891 2.822–2.848 2.892–2.912 2 × 3.254 4 × 2.834 2.794 2 × 2.784 2 × 3.474 3.564 2 × 2.652 2 × 3.188 2 × 3.319 2 × 2.650 4 × 3.363 2 × 2.753 4 × 3.601 2.764 4 × 3.565 3.683 2 × 3.330 4 × 3.172 3.584–3.602 3.236–3.343 r(M–E)a/Å 2.36 2.28 2.30 2.34 2.34 1.95 1.94 1.95 2.26 2.25 2.34 2.19 2.31 2.21 2.22 2.33 2.36 2.22 2.37 2.23 r(M–C)a/Å 2.37 2.25 2.28 2.33 2.35 2.28 2.26 2.27 2.24 2.24 2.36 2.13 2.22 2.11 2.14 2.22 2.26 2.08 2.19 2.11 Ref. 7 778 89 10 11 12 13 8 14 15 16 17 18 19 20 21 20 a Averaged values. infinite rotational symmetry for the (C5H5)M group], the four dz2 orbitals transform as a1 1 t2, while the eight dxy and dx2 2 y2 orbitals form a basis for e 1 t2 1 t1 representations. Dahl and co-workers 14,23 proposed that the metal–metal bonding a1, e and t2 orbitals lie below their antibonding counterparts, t1 and t2, leading to a net order of 1.0 for each metal–metal bond in clusters such as (C5H5)4Mo4S4 with twelve valence electrons.In (C5H5)4Co4S4, the additional valence electrons occupy all components of the antibonding t1 and t2 orbitals, again giving a symmetric structure, but this time with much longer distances between the metal centres,20 consistent with a Co–Co bond order of zero.In between these closed-shell limits, distortions from tetrahedral symmetry occur. For example, the Fe4 cores in (C5H5)4Fe4S4 21 (eighteen valence electrons) and (C5H5)4Fe4S4 (twenty valence electrons) are elongated and compressed respectively along one two-fold axis.14,17 Both types of rhombic distortion can be rationalised in terms of Jahn–Teller instability arising from the partial occupancy of the degenerate t1 and t2 orbitals.While the Dahl bonding model has been remarkably successful in rationalising the properties of the metal–sulfur cubane clusters, the apparently anomalous properties of their chromium–oxygen analogues cannot be simply explained within the same framework. On the basis of extended Hückel calculations, Bottomley and Grein 24 proposed an alternative scheme where strong interactions with the bridging oxide ligands destabilise the metal–metal bonding a1 orbital.The resulting ground-state configuration, e4t2 6t1 2, would then be Jahn–Teller unstable, giving rise to a distortion from tetrahedral symmetry. More recent calculations, however, suggest that their result was an artifact of the chosen parameter set, and favour the original orbital ordering shown in Fig. 1.10c,26 The distorted structures and non-zero room temperature magnetic moments of the chromium–oxygen clusters could also, in principle, arise from a dynamic Jahn–Teller distortion due to the thermal population of the antibonding t1 and t2 orbitals.This is, however, inconsistent with variable temperature crystallographic studies, which show a marginal increase, rather than decrease, in the magnitude of the rhombic distortion as the temperature is reduced.10c In summary, simple molecular orbital arguments based on the interactions of four equivalent (C5H5)M units have so far failed to provide a convincing rationale for the distortions observed in (C5H5)4Cr4O4.Green and co-workers 22 noted that if the electrons in (C5H5)4Cr4O4 are weakly coupled, single configuration molecular orbital methods such as those used by Dahl, Bottomley and others 23–26 are likely to be inadequate, because the eVects of electron correlation are neglected. In a series of recent papers reporting density functional studies on bimetallic clusters,27 it has been shown that this can result in the underestimation of metal–metal bond lengths by as much as 1.0 Å.The eVects of electron correlation can, however, be modelled within spin-unrestricted density functional theory by removing all symmetry elements connecting the metal ions (brokensymmetry density functional theory, see methodology section), thereby permitting the electrons to localise. The result is a much improved description of antiferromagnetic coupling, and optimised metal–metal bond lengths in excellent agreement with experiment. Noodleman and Norman have shown that this broken-symmetry methodology28 is essential for an accurate description of metal–metal interactions in tetrametallic clusters such as Fe4S4(SR)4,5a–c as well as numerous other polymetallic systems.29 A key feature of the broken-symmetry technique isJ.Chem. Soc., Dalton Trans., 1999, 1393–1399 1395 that it permits, but does not force, the electrons to localise. Thus, the fully delocalised electron spin density distribution characteristic of strong metal–metal bonding can be recovered if it represents a more stable situation than the localised alternative.The broken-symmetry technique is therefore capable of modelling both weakly coupled and strongly bonded limits without bias, and is ideal for a study of periodic trends. In this paper, broken-symmetry density functional theory is used to investigate periodic trends in the structure and bonding of (C5H5)4M4E4, M = Cr, Mo, E = O, S. In particular, the aim is to establish whether there is an abrupt change in electronic structure between the chromium–oxygen clusters and their molybdenum–sulfur counterparts, and whether this gives rise to the unusual structural properties of the former.The long term goal is to establish a theoretical technique capable of analysing redox-induced metal–metal bond formation in the iron–sulfur and iron–molybdenum–sulfur clusters, and the potential implications of such processes for electron transfer pathways in redox enzymes.Methodology All calculations described in this paper are based on approximate density functional theory, which has been used to probe structural, energetic and mechanistic problems in numerous transition metal-based systems.30 Calculations were performed using the Amsterdam Density Functional (ADF) program Version 2.3, developed by Baerends and co-workers.31 A double-z Slater-type basis set, extended with a single polaris- Fig. 1 Schematic molecular orbital diagram for cubanes (after refs. 23 and 26). ation function, was used to describe the hydrogen, carbon, oxygen and sulfur atoms, while molybdenum and chromium were modelled with a triple-z basis set. Electrons in orbitals up to and including 1s {C,O}, 2p {S}, 3p {Cr} and 4p {Mo} were considered part of the core and treated in accordance with the frozen core approximation. The local density approximation was employed in all cases,32 along with the local exchangecorrelation potential of Vosko, Wilk and Nusair 33 and gradient corrections to exchange and correlation proposed by Perdew and Wang.34 All structures were optimised using the gradient algorithm of Versluis and Ziegler.35 The cyclopentadienyl rings were constrained to local D5h symmetry, and were aligned with one carbon atom of each ring eclipsing an edge of the M4 tetrahedron, giving a nuclear framework with D2d symmetry (see Chart 1). All four clusters in the current study have spin-singlet ground states, a situation that can be achieved in one of two ways depending on the distribution of spin density throughout the cluster.In the limit of full electron delocalisation, each of the twelve metal-based electrons, six with spin-a, six with spin-b, have equal amplitude on all four metal centres, and the electron spin density has the full symmetry of the nuclear framework [D2d, Scheme 1, structure (a)]. At the opposite extreme of complete electron localisation, the spin-a electrons are concentrated on the upper half of the cluster, the spin-b electrons on the lower half, resulting in an electron spin density distribution of lower symmetry than the nuclear framework [C2v, Scheme 1, structure (b)].This localised limit can be modelled by removing all symmetry elements connecting the two halves of the cluster, and imposing an initial excess of spin-a and spin-b electron density on the upper and lower M2 units respectively. To aid comparison between the localised and delocalised limits, orbitals will be labelled according to the representations of the C2v subgroup throughout this manuscript. Results Ground-state electronic structure of (C5H5)4Cr4O4 Optimised structural parameters for the ground states of all four clusters are summarised in Table 2, along with total energies and the net spin densities per metal centre.The calculated parameters for (C5H5)4Cr4O4 are in excellent accord with crystallographic data, with Cr–Cr, Cr–O and Cr–C separations lying within 0.06 Å of the experimentally determined values.Most significantly, a distinct ground-state Scheme 1 Electron spin density distributions in M4E4 cubanes. Bold lines indicate strong metal–metal bonds, broken lines indicate weak magnetic coupling. M M M M M M M M M M M M (c) – EM-M + ESP + EM-M – ESP –1/3EM-M +1/3EM-M –2/3EM-M + ESP +2/3EM-M – ESP (a) (b)1396 J. Chem. Soc., Dalton Trans., 1999, 1393–1399 Table 2 Optimised energies, spin distributions and structural parameters for ground and excited states of (C5H5)4M4E4 (M = Cr, Mo, E = O, S) Ground-state Excited-state (C5H5)4Cr4O4 (C5H5)4Cr4S4 (C5H5)4Mo4O4 (C5H5)4Mo4S4 (C5H5)4Cr4O4 (C5H5)4Cr4S4 (C5H5)4Mo4O4 (C5H5)4Mo4S4 Configuration, Ga,b a1 2,2a2 1,1b1 2,1b2 1,2 a1 3,3a2 1,1b1 1,1b2 1,1 a1 3,3a2 1,1b1 1,1b2 1,1 a1 3,3a2 1,1b1 1,1b2 1,1 a1 3,3a2 1,1b1 1,1b2 1,1 a1 2,2a2 1,1b1 2,1b2 1,2 a1 2,2a2 1,1b1 2,1b2 1,2 a1 2,2a2 1,1b1 2,1b2 1,2 r(M–M)/Å 2 × 2.886 4 × 2.753 2 × 2.800 4 × 2.802 2 × 2.674 4 × 2.682 2 × 2.930 4 × 2.950 2 × 2.466 4 × 2.471 2 × 3.250 4 × 2.973 2 × 3.068 4 × 2.605 2 × 3.358 4 × 2.870 r(M–E)/Å 1.950 2.250 2.088 2.402 1.919 2.326 2.096 2.436 r(M–C)/Å 2.306 2.253 2.454 2.411 2.276 2.278 2.456 2.407 Relative energy/kJ mol21 0.00 0.00 0.00 0.00 146 143 1110 1215 Net spin ±2.78 0.00 0.00 0.00 0.00 ±2.53 ±0.76 ±0.76 rhombic distortion emerges, with two long and four short Cr–Cr distances.The net spin densities of ±2.78 per metal centre indicate that the electrons are localized, with a spin density distribution similar to that shown in Scheme 1, structure (b).The most logical starting point for the discussion of the molecular orbital structure is therefore not the delocalised view implicit in the Dahl model, but rather that of four isolated d3 single ions in their spin-quartet ground states. The molecular orbital diagram for the broken-symmetry ground state of (C5H5)Cr4O4 is summarised in Fig. 2, with orbitals localised on the upper and lower halves of the cluster shown on the left and right sides respectively. On the upper half, the in-phase and out-of-phase combinations of the radially oriented dz2 orbitals transform as a1 1 b1, while their tangential counterparts, dxy and dx2 2 y2, transform as a1 1 b1 and a2 1 b2 respectively, giving a total of 2a1 1 a2 1 2b1 1 b2 symmetry orbitals. The orbitals on the lower half transform in a similar manner, except that the rotation by 908 about the principal axis interconverts the b1 and b2 symmetry labels, yielding 2a1 1 a2 1 b1 1 2b2 symmetry orbitals.Where metal–metal interactions are weak, each of these twelve symmetry orbitals remains singly occupied, those on the upper half of the cluster by spin-a electrons, those on the lower half by spin-b, leading to Fig. 2 Molecular orbital diagram for the broken-symmetry ground state of (C5H5)4Cr4O4: spin-a orbitals are shown as full lines, spin-b orbitals as dashed lines.Orbitals which are predominantly involved in metal–ligand, rather than metal–metal, interactions are not shown. Sketches are shown only for the spin-a orbitals, their spin-b counterparts are identical, but localised on the opposite half of the cluster. –1.0 –2.0 –3.0 –4.0 –5.0 1a1 2a1 3a1 4a1 1a2 2a2 1b1 2b1 1b2 2b2 3b2 4b2 1a1 2a1 3a1 4a1 1a2 2a2 1b2 2b1 1b1 2b1 3b1 4b1 E /eV a ground-state electronic configuration Ga,b = a1 2,2a2 1,1b1 2,1b2 1,2.In terms of magnetic coupling, the chromium centres within each dimeric unit are ferromagnetically coupled (spins parallel), while the two dimeric units are coupled antiferromagnetically (spins antiparallel). The extent of the rhombic distortion described in the introduction is defined by the angle q between the centroid and two chromium centres in either the upper or lower half of the cluster. Where q = 109.58, the M4 core is perfectly tetrahedral, while larger values correspond to an increased separation between the ferromagnetically coupled chromium centres.The dependence on q of the energies of the occupied spin-a orbitals of (C5H5)4Cr4O4 is summarised in Fig. 3 (the corresponding spin-b orbitals, localised on the opposite half of the cluster, behave in identical fashion). As q increases, the major eVect is a stabilisation of the 2b1 orbital, an antibonding combination of the dx2 2 y2 orbitals on the two centres. Thus, it is the reduction in antibonding character between the two ferromagnetically coupled chromium centres that provides the driving force for the rhombic distortion in (C5H5)4Cr4O4.The poor overlap between the small chromium 3d orbitals makes this driving force relatively weak, and the optimised structure (q = 114.78) is only 12 kJ mol21 more stable than the perfectly tetrahedral one (q = 109.58). The calculations described so far do not address the question of the origin of the twist observed in the crystal structure of (C5H5)4Cr4O4, because this type of distortion does not lie on the C2v-symmetric potential energy surface.The structure was therefore re-optimised using C2 point symmetry, with one Cr2 unit twisted relative to the other. The potential energy curve describing the twist is found to be extremely flat (<10 kJ mol21 above the minimum for rotations of up to 108 about the principal axis), but no evidence was found for additional minima, optimisation always resulting in a C2v-symmetric structure.In conclusion, the electronically-favoured structure for the (C5H5)4Cr4O4 system is the rhombically distorted one, with both the tetrahedral and twisted structures lying approximately 10–12 kJ mol21 higher in energy. The almost perfectly tetrahedral structure observed for pentamethylated species, (C5Me5)4Cr4O4, along with the low-symmetry distortion seen in (C5H5)4Cr4O4, must therefore arise because the weak electronic driving force favouring the distortion is of similar magnitude to intra- or inter-molecular steric interactions.Ground-state electronic structure of (C5H5)4Cr4S4, (C5H5)4- Mo4O4 and (C5H5)4Mo4S4 In the metal–sulfur clusters, the optimised ground-state structural parameters are again in excellent agreement with crystallographic data. In particular, no rhombic distortion isJ. Chem. Soc., Dalton Trans., 1999, 1393–1399 1397 Fig. 3 Walsh diagram showing the dependence on q of the energy of the occupied orbitals of (C5H5)4Cr4O4. predicted, consistent with the presence of six equivalent metal– metal single bonds lying along each of the edges of the tetrahedron.A molecular orbital diagram for the ground state of (C5H5)4Mo4S4 is illustrated in Fig. 4 [the qualitative features of the molecular orbital diagrams for (C5H5)4Cr4S4 and (C5H5)4Mo4O4 are identical]. In contrast to Fig. 2, no distinction is made on the basis of either spin or spatial distribution, because the net spin densities of zero indicate that all metalbased electrons are fully delocalised over the whole cluster [Scheme 1, structure (a)].The occupied manifold contains three narrow bands, made up, in ascending order, of 1a1, 2a1 1 1a2 and 3a1 1 1b1 1 1b2, giving a ground-state electronic con- figuration of Ga,b = a1 3,3a2 1,1b1 1,1b2 1,1, which correlates directly with the G = a1 2e4t2 6 configuration shown in Fig. 1 (see Table 3). Thus, when the electrons are delocalised over the whole cluster, Fig. 4 Molecular orbital diagram for the ground state of (C5H5)4- Mo4S4: no distinction is made on the basis of either spin or spatial distribution, as all cluster valence electrons are delocalised. Orbitals which are predominantly involved in metal–ligand, rather than metal– metal, interactions are not shown. 1a1 2a1 + 1a2 4a1 1b1 + 1b2 –5.0 E / eV –1.0 –2.0 –3.0 –4.0 3a1 3b1 + 3b2 2b1 + 2b2 2a2 Table 3 Descent in symmetry from Td to C2v Td a1 et 1 t2 C2v a1 a1 1 a2 a2 1 b1 1 b2 a1 1 b1 1 b2 the bonding model obtained from the broken-symmetry technique converges with that proposed by Dahl and others.14,23 In contrast, the Ga,b = a1 2,2a2 1,1b1 2,1b2 1,2 ground-state configuration of the chromium–oxygen analogue correlates with an excited state, G = a1 2e4t2 4t1 2, in tetrahedral symmetry, confirming that the Dahl model is inappropriate in the limit of weak metal– metal bonding.Excited-state properties Despite the fact that the electron spin-density distributions are very diVerent for (C5H5)4Cr4O4 and (C5H5)4Cr4S4, their groundstate configurations diVer only in the location of two electrons, which are either in 2b1a/2b2b (localised) or 3a1a/b (delocalised).Thus population of a doubly excited state may cause substantial changes in the electron distribution, and hence in the structure of the cluster. In (C5H5)4Cr4O4, promotion of two electrons from 2b1a/2b2b to 3a1a/b causes the metal-based electrons to delocalise completely, resulting in a net spin density of zero.The Cr–Cr distances contract by 0.3–0.4 Å, consistent with the formation of six Cr–Cr single bonds, and the rhombic distortion disappears. In contrast, precisely the opposite is observed in the excited state of (C5H5)4Cr4S4 (Ga,b= a1 2,2a2 1,1b1 2,1b2 1,2), where the metal-based electrons localise (net spin densities = ±2.53), the Cr–Cr separations lengthen by 0.35– 0.45 Å, and a strong rhombic distortion emerges.Perhaps the most surprising aspect of this observation is that in each case the excited state lies less than 45 kJ mol21 above the ground state, indicating that gross structural rearrangements, involving changes of almost 0.5 Å in metal–metal bond lengths, can be energetically facile. Moreover, not only do the two clusters lie close to the localised/delocalised borderline, they are also symmetrically placed on either side of it. Accordingly, the structure of the mixed oxide–sulfide cluster, (C5H5)4Cr4S2O2 has been optimised in both electronic configurations, and the localised and delocalised states are found to be separated by only 2 kJ mol21, despite having Cr–Cr separations which diVer by over 0.3 Å.This observation suggests that, given a judicious choice of ligands, dramatic changes in cluster structure and bonding might be induced by very minor electronic or steric perturbations. Despite the very similar ground-state properties of (C5H5)4- Cr4S4, (C5H5)4Mo4O4 and (C5H5)4Mo4S4, the excited-state properties of the two molybdenum clusters diVer significantly from those of the chromium–sulfur system described above.In both molybdenum clusters, two of the six Mo–Mo separations increase by approximately 0.4 Å, while the remaining1398 J. Chem. Soc., Dalton Trans., 1999, 1393–1399 four contract slightly, giving a very strong rhombic distortion [Scheme 1, structure (c)]. The net spin densities of 10.76 con- firm that only two of the six Mo–Mo bonds have been broken, which is precisely the result that might have been anticipated on the basis of the bonding/antibonding character of the 2b1/2b2 and 3a1 orbitals.The former are antibonding with respect to the metal–metal bonds, while the latter are bonding (see Fig. 2), and so the total bond order for the clusters is reduced from six to four in the excited state. This then raises the question of why promotion of the same two electrons in (C5H5)4Cr4S4 causes the cleavage of all six, rather than just two, of the metal–metal bonds.In the following section, the diVerent properties of the chromium and molybdenum systems are rationalised in terms of the competing eVects of metal–metal bond strength and spin polarisation energy. The balance between localised and delocalised electron distributions In previous papers 27 it has been established that the relative stability of localised and delocalised electron density distributions is determined by the balance between the strength of the metal–metal bonds, favouring delocalisation, and the spin polarisation energy, favouring localisation.The absolute magnitude of this spin polarisation energy is related to n[n 2 1]/2, where n is the number of unpaired electrons per metal centre. A quantitative estimate of the magnitude of these two competing terms can be obtained by defining a reference configuration in which neither metal–metal bonding nor spin polarisation are present.In this case, the appropriate configuration is spin restricted G = 1a1 12a1 11a2 13a1 11b1 11b2 12a2 12b1 12b2 14a1 13b1 13b2 1 (see Fig. 4), where single occupation of all bonding and antibonding combinations ensures a net metal–metal bond order of zero. The metal–metal bond energy (denoted EM–M) is then taken as the diVerence in energy between the reference configuration and that where all electrons are involved in metal–metal bonds (Ga,b = a1 3,3a2 1,1b1 1,1b2 1,1). Similarly, the spin polarisation energy, ESP, is taken as the diVerence between the reference configuration and one in which there are three unpaired electrons per metal centre.Unfortunately the energy of the Ga,b = a1 2,2a2 1,1b1 2,1b2 1,2 broken-symmetry state described in the previous section is not appropriate, because in the molybdenum clusters, the electrons are only partially localised. Instead, a high-spin configuration, Ga,b = a1 4,0a2 2,0b1 3,0b2 3,0, with twelve unpaired electrons, is defined corresponding to single occupation of all metal-based bonding and antibonding orbitals.In this configuration, which diVers from the localised broken-symmetry state only in the orientations of the spins on the two halves of the molecule, maximum spin polarisation is assured, while metal–metal bonding is completely eliminated. A detailed discussion of this form of analysis is given in ref. 27(c). Values of EM–M and ESP for the four clusters are summarised in Table 4.The spin polarisation energy falls by over 260 kJ mol21 on replacement of chromium by molybdenum because the more diVuse 4d orbitals increase the average separation between the electrons. Replacement of oxygen with sulfur has a similar, although smaller, eVect due to the greater covalence of the metal–sulfur bond, which also increases the average interelectronic separation. Trends in EM–M are exactly opposite to those in ESP, because strong metal–metal bonds are favoured by strong orbital overlap, and therefore by large metal d orbitals.Thus EM–M increases when chromium is replaced by molybdenum, and when oxygen is replaced by sulfur. In both molybdenum systems, EM–M @ ESP, indicating that the strength of six metal–metal bonds outweighs the spin polarisation, and so the delocalised electron spin density distribution prevails. For the corresponding chromium clusters, the increase in ESP and concomitant decrease in EM–M combine to cause a shift towards the localised limit.The greater tendency of heavier members of the triad to form metal–metal bonds is well established, and is usually interpreted solely in terms of the more eVective overlap aVorded by the larger 4d and 5d orbitals. The analysis presented here shows that changes in orbital overlap are only one factor, and in fact for the oxide clusters, the decrease in spin polarisation energy is more important. Thus when considering the tendency of a given complex to form metal–metal bonds, it is important to consider not only the potential gain in energy associated with orbital overlap, but also the loss of spin polarisation energy associated with the delocalisation of the electrons.For (C5H5)4Cr4S4, EM–M remains signifi- cantly larger than ESP, and accordingly the electrons remain delocalised, but for (C5H5)4Cr4O4, the two terms are almost identical. The additional stability associated with the antiferromagnetic coupling in the broken-symmetry state (rather than ferromagnetic in the high-spin state on which this analysis is based) is then suYcient to tip the balance in favour of localisation.The diVerent structural changes associated with the promotion of two electrons in the molybdenum and chromium clusters can also be rationalised in terms of the balance between spin polarisation energy and metal–metal bond strength. As noted in the previous section, the bonding/antibonding properties of the orbitals would suggest that only two of the six bonds should be aVected in the excited state, changing the total energy associated with the metal–metal bonds from EM–M to approximately 2/3EM–M [Scheme 1, structures (a)–(c)].If polarisation of the core- and ligand-based electrons is neglected, there is no increase in spin polarisation energy associated with this transition, because only a single unpaired electron is generated per metal centre (n[n 2 1]/2 = 0).Cleavage of the remaining four bonds completely eliminates the residual metal–metal bonding (2/3EM–M), but this is now oVset by the emergence of the spin polarisation energy, ESP, associated with the generation of three unpaired electrons per metal centre [Scheme 1, structures (a)–(c)]. The balance between partial and complete bond cleavage in the excited state is therefore determined by the relative magnitudes of 2/3EM–M (favouring retention of four bonds) and ESP (favouring complete cleavage).For the molybdenum clusters, 2/3EM–M @ ESP, and the metal– metal bonds are therefore strong enough to resist complete electron localisation. In contrast, ESP > 2/3EM–M for (C5H5)4Cr4S4, and the large gain in spin polarisation energy outweighs the combined strength of the four residual Cr–Cr bonds, causing complete electron localisation. Conclusions In this paper, the delicate balance between strong metal–metal bonding and weak antiferromagnetic coupling in metal cubane clusters has been illustrated.Only relatively minor structural and electronic changes are necessary to cause an abrupt transition from one regime to the other. The degree of electron localisation is determined by the competition between orbital overlap (favouring delocalisation) and spin polarisation energy (favouring localisation). For the molybdenum clusters, metal– metal bonding is strong, and the electrons are delocalised in the ground state, giving an almost perfectly tetrahedral M4 core.For the chromium systems, the two terms are more delicately balanced, and in (C5H5)4Cr4O4, the spin polarisation term dominates, causing the electrons to localise. In this case, the metalbased electrons are antiferromagnetically coupled across four Table 4 Metal–metal bond (EM–M) and spin polarisation (ESP) energies (C5H5)4Cr4O4 (C5H5)4Cr4S4 (C5H5)4Mo4O4 (C5H5)4Mo4S4 EM–M/kJ mol21 798 915 992 1222 ESP/kJ mol21 788 767 514 502J.Chem. Soc., Dalton Trans., 1999, 1393–1399 1399 edges of the tetrahedron, and ferromagnetically coupled across the other two, giving rise to a rhombic distortion. Both chromium clusters have low-lying excited states (<45 kJ mol21 above the ground state) in which the nature of the metal–metal bonding is completely reversed relative to the ground state (weak coupling for the sulfur system, strong bonding for the oxygen analogue). The transition from ground to excited state is therefore associated with major structural rearrangements, despite the relatively low energy involved.In future work, the influence of redox changes on metal–metal bonding in related clusters will be analysed, with the aim of determining whether electron transfer at biologically attainable potentials can also induce abrupt transitions from one bonding regime to another. References 1 R. Cammack, Adv. Inorg. Chem., 1992, 38, 281; H. Beinert, R. H. Holm and E. Münck, Science, 1997, 277, 653. 2 W. R. Hagen, Adv. Inorg. Chem., 1992, 38, 165; I. Bertini, S. Ciurli and C. Luchinat, Struct. Bonding (Berlin), 1995, 83, 1. 3 S. J. Yoo, Z. Hu, C. Goh, E. L. Bominaar, R. H. Holm and E. Münck, J. Am. Chem. Soc., 1997, 119, 8732; E. P. Day, J. Peterson, J. J. Bonvoisin, I. Moura and J. J. G. Moura, J. Biol. Chem., 1988, 263, 3684. 4 H. C. Angove, S. J. Yoo, B. K. Burgess and E. Münck, J. Am. Chem. Soc., 1997, 119, 8730; P. A. Lindahl, E. P. Day, T. A. Kent, W.H. Orme-Johnson and E. Münck, J. Biol. Chem., 185, 260, 11160. 5 (a) L. Noodleman and D. A. Case, Adv. Inorg. Chem., 1992, 38, 423; (b) A. Aizman and D. A. Case, J. Am. Chem. Soc., 1982, 104, 3269; (c) L. Noodleman, C. Y. Peng, D. A. Case and J.-M. Mouesca, Coord. Chem. Rev., 1995, 144, 199; (d ) A. J. Thomson, J. Chem. Soc., Dalton Trans., 1981, 1180; (e) A. Ceulemans and P. W. Fowler, Inorg. Chim. Acta, 1985, 105, 75. 6 K. B. Musgrave, H. C. Angove, B. K. Burgess, B. Hedman and K.O. Hodgson, J. Am. Chem. Soc., 1998, 120, 5325. 7 J. Darkwa, J. R. Lockemeyer, P. D. W. Boyd, T. B. Rauchfuss and A. L. Rheingold, J. Am. Chem. Soc., 1988, 110, 141. 8 J. A. Bandy, C. E. Davies, J. C. Green, M. L. H. Green, K. Prout and D. P. S. Rodgers, J. Chem. Soc., Chem. Commun., 1983, 1395. 9 I. L. Eremenko, S. E. Nefedov, A. A. Pasynskii, B. Orazsakhatov, O. G. Ellert, Yu. T. Struchkov, A. I. Yanovsky and D. V. Zagorevsky, J. Organomet. Chem., 1989, 368, 185. 10 (a) F. Bottomley, D. E. Paez and P. S. White, J. Am. Chem. Soc., 1981, 103, 5581; (b) F. Bottomley, D. E. Paez and P. S. White, J. Am. Chem. Soc., 1982, 104, 5651; (c) F. Bottomley, D. E. Paez, L. Sutin, P. S. White, F. H. Köhler, R. C. Thompson and N. P. C. Westwood, Organometallics, 1990, 9, 2443. 11 F. Bottomley, J. Chen, S. M. MacIntosh and R. C. Thompson, Organometallics, 1991, 10, 906. 12 C. Wei, L.-Y. Goh, R. F. Bryan and E. Sinn, Acta Crystallogr., Sect. C, 1986, 42, 796. 13 A. A. Pasynskii, I. L. Eremenko, Yu. V. Rakitin, V. M. Novotortsev, O. G. Ellert, V. T. Kalinnikov, V. E. Shklover, Yu. T. Struchkov, S. V. Lindeman, T. Kh. Kurbanov and G. Sh. Gasanov, J. Organomet. Chem., 1983, 248, 309. 14 Trinh-Toan, B. K. Teo, J. A. Ferguson, T. J. Meyer and L. F. Dahl, J. Am. Chem. Soc., 1977, 99, 408. 15 E. J. Houser, J. Amarasekera, T. B. Rauchfuss and S. R. Wilson, J. Am. Chem. Soc., 1991, 113, 7440; E. J. Houser, T. B. Rauchfuss and S. R. Wilson, Inorg.Chem., 1993, 32, 4069. 16 Trinh-Toan, W. P. Fehlhammer and L. F. Dahl, J. Am. Chem. Soc., 1977, 99, 402. 17 R. A. Schunn, C. J. Fritchie and C. T. Prewitt, Inorg. Chem., 1966, 5, 892; C. H. Wei, G. R. Wilkes, P. M. Treichel and L. F. Dahl, Inorg. Chem., 1966, 5, 900. 18 J. Amarasekera, T. B. Rauchfuss and S. R. Wilson, J. Chem. Soc., Chem. Commun., 1989, 14. 19 A. Venturelli and T. B. Rauchfuss, J. Am. Chem. Soc., 1994, 116, 4824. 20 G. L. Simon and L. F. Dahl, J. Am. Chem.Soc., 1973, 95, 2164. 21 D. A. Dobbs and R. G. Bergman, J. Am. Chem. Soc., 1992, 114, 6908. 22 C. E. Davies, J. C. Green, N. Kaltsoyannis, M. A. MacDonald, J. Qin, T. B. Rauchfuss, C. M. Redfern, G. H. Stringer and M. G. Woolhouse, Inorg. Chem., 1992, 31, 3779. 23 A. S. Faust and L. F. Dahl, J. Am. Chem. Soc., 1970, 92, 7337; G. L. Simon and L. F. Dahl, J. Am. Chem. Soc., 1973, 95, 2175. 24 F. Bottomley and F. Grein, Inorg. Chem., 1982, 21, 4170. 25 S. Harris, Inorg. Chem., 1987, 26, 4278; S. Harris, Polyhedron, 1989, 8, 2843; C. S. Bahn, A. Tan and S. Harris, Inorg. Chem., 1998, 37, 2770. 26 P. D. Williams and M. D. Curtis, Inorg. Chem., 1986, 25, 4562. 27 (a) J. E. McGrady, R. Stranger and T. Lovell, Inorg. Chem., 1998, 37, 3802; (b) J. E. McGrady, R. Stranger and T. Lovell, J. Phys. Chem. A, 1997, 101, 6265; (c) J. E. McGrady, T. Lovell and R. Stranger, Inorg. Chem., 1997, 36, 3242; (d) T. Lovell, J. E. McGrady, R. Stranger and S. A. Macgregor, Inorg. Chem., 1996, 35, 3079; (e) J. E. McGrady and R. Stranger, J. Am. Chem. Soc., 1997, 119, 8512. 28 L. Noodleman and J. G. Norman, Jr., J. Chem. Phys., 1979, 70, 4903; L. Noodleman, J. Chem. Phys., 1981, 74, 5737. 29 C. A. Brown, G. J. Remar, R. L. Musselman and E. I. Solomon, Inorg. Chem., 1995, 34, 688; P. K. Ross and E. I. Solomon, J. Am. Chem. Soc., 1991, 113, 3246; H. Jacobsen, H.-B. Kraatz, T. Ziegler and P. M. Boorman, J. Am. Chem. Soc., 1992, 114, 7851; A. Bencini and D. Gatteschi, J. Am. Chem. Soc., 1986, 108, 5763; J. Andzelm and E. Wimmer, J. Chem. Phys., 1992, 96, 1280; K. E. Edgecombe and A. D. Becke, Chem. Phys. Lett., 1995, 244, 427; N. A. Baykara, B. N. McMaster and D. R. Salahub, Mol. Phys., 1984, 52, 891. 30 T. Ziegler, Chem. Rev., 1991, 91, 651. 31 ADF 2.3.0, Theoretical Chemistry, Vrije Univeriteit, Amsterdam, E. J. Baerends, D. E. Ellis and P. Ros, Chem. Phys., 1973, 2, 42; G. te Velde and E. J. Baerends, J. Comput. Phys., 1992, 99, 84. 32 R. G. Parr and W. Yang, Density Functional Theory of Atoms and Molecules, Oxford University Press, New York, 1989. 33 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200. 34 J. P. Perdew and Y. Wang, Phys. Rev. B, 1991, 44, 13298; J. P. Perdew and Y. Wang, Phys. Rev. B, 1992, 45, 13244. 35 L. Versluis and T. Ziegler, J. Chem. Phys., 1988, 88, 322. Paper 9/00660E

ISSN:1477-9226    

DOI:10.1039/a900660e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 1401–1406 1401 Structural analysis and magnetic properties of the 1-D compounds [M(NCS)2bpa2] [M 5 Fe, Co, Ni and bpa 5 1,2-bis(4-pyridyl)- ethane] Margarita L. Hernández,a M. Gotzone Barandika,a,b M. Karmele Urtiaga,c Roberto Cortés,b Luis Lezama,a M. Isabel Arriortua c and Teófilo Rojo *a a Departamento de Química Inorgánica, Facultad de Ciencias, Universidad del País Vasco, Apdo. 644, Bilbao 48080, Spain. E-mail: qiproapt@lg.ehu.es b Departamento de Química Inorgánica, Facultad de Farmacia, Universidad del País Vasco, Apdo. 450, Vitoria 01080, Spain c Departamento de Mineralogía-Petrología, Facultad de Ciencias, Universidad del País Vasco, Apdo. 644, Bilbao 48080, Spain Received 5th January 1999, Accepted 11th March 1999 The increasing interest in designing polymeric compounds has focused this work on the synthesis and magnetostructural characterization of three 1-D compounds of general formula [M(NCS)2bpa2] [M = Fe, Co, Ni and bpa = 1,2-bis(4-pyridyl)ethane] which exhibit double bpa bridges between metallic cations.X-Ray single crystal diVraction analysis was carried out on 1 [Fe(NCS)2bpa2] while for 2 [Co(NCS)2bpa2] and 3 [Ni(NCS)2bpa2] the structural characterization was performed on powdered samples. UV-VIS results were consistent with high-spin metallic cations in tetragonally distorted octahedral fields. Although in the three compounds weak antiferromagnetic coupling has been observed taking place through the –M–(bpa)2–M– bridges, for the theoretical analysis of the global magnetic behaviour of all of them some other eVects have been considered, i.e. zero field splitting and spin–orbit coupling.Introduction Over the last few years, the design of polymeric compounds has been an active area of research in coordination chemistry. So far, extended systems of a variety of metals and ligands have been characterized providing very interesting information about magnetostructural co-relations in these compounds.1,2 The generation of this type of molecular architecture rests on the combination of several factors like the coordination geometry of the metals, the performance of the ligands and the reaction conditions.In relation to the ligands, the use of pseudo-halides and voluminous organic molecules has been well established and results in covalent high-dimensional arrangements. While the role of ambidentate pseudo-halides is clearly related to their ability to form interaction bridges, the function of the voluminous ligands has been mainly linked to their coordination-site-blocking behaviour.3,4 However, the bridging capability exhibited by several polydentate organic molecules enhances the possibility of generating highdimensional systems.Compounds obtained by using N,N9 bidentate spacers 5–11 like 4,49-bipyridine, pyrazine and other related ligands are illustrative of this point. For a self-assembly strategy of the synthesis of these compounds, the rigidity of the aforementioned ligands is clearly a limiting factor.Therefore, the use of flexible linkers represents an excellent alternative for further research. Among them, 1,2- bis(4-pyridyl)ethane has been used to this purpose in several recent works.12,13 This ligand has been observed to exhibit two diVerent conformations, anti and gauche (Scheme 1), being remarkably versatile as a spacer in extended systems.14 Thus, isomerism and its manifestation in the structure have become issues which demand much work in this area.Taking into account the aforementioned aspects, three compounds of formula [M(NCS)2bpa2] [M = Fe 1, Co 2 and Ni 3 and bpa = 1,2-bis(4-pyridyl)ethane] were prepared in order to study the influence of both ligands in the structure and magnetic properties of the 1-D systems obtained. Experimental Synthesis Synthesis of compound 1 (yield 63%) was carried out by mixing an ethanolic solution of FeCl2?4H2O (0.5 mmol, 25 ml) also containing an excess of NaNCS with a methanolic solution of bpa (1 mmol, 25 ml).This solution was left to stand at room temperature. After several days, prismatic, yellow, X-ray quality single crystals were obtained. Synthesis of compound 2 (yield 82%) was performed by mixing an ethanolic solution of Co(CH3CO2)2?4H2O (0.5 mmol, 25 ml) with a methanolic solution (25 ml) containing bpa (1.0 mmol) and NaNCS (2.5 mmol) resulting in a red precipitate. Compound 3 (yield 80%) was also obtained as a precipitate (light blue coloured) after following the procedure for compound 2 with NiCl2?6H2O.Further attempts to recrystallize these precipitates in several mixtures of ethanol, methanol, acetone and water gave rise to poor quality single crystals. Elemental analysis and atomic absorption results were in good agreement with the MC26N6H24S2 (M = Fe, Co, Ni) stoichiometry for the three compounds.Found (calc.)%: Fe, Scheme 1 N N N N anti gauche1402 J. Chem. Soc., Dalton Trans., 1999, 1401–1406 Fig. 1 ORTEP33 view (50% probability) of the linear chains extending along the [010] direction for [Fe(NCS)2bpa2] 1. 10.43 (10.34); S, 11.53 (11.87); N, 16.01 (15.56); C, 58.07 (57.76); H, 4.27 (4.47) for 1: Co, 11.10 (10.85); S, 11.61 (11.77); N, 15.56 (15.47), C, 57.36 (57.45); H, 4.24 (4.45) for 2: Ni, 11.10 (10.80); S, 11.68 (11.80); N, 15.58 (15.47), C, 57.74 (57.48); H, 4.21 (4.45) for 3.TG curves obtained for compounds 1, 2 and 3 (25–500 8C) showed that, as reported for similar compounds,15 the three of them undergo pyrolysis of the ligands taking place in two steps. The first of them, being centred at 275 8C, can be attributed to the loss of a molecule of bpa per formula unit while the second one takes place at around 425 8C. The identification of the final residues through X-ray diVraction was not possible due to the low crystallinity of the samples.Physical measurements Microanalyses were performed with a Perkin-Elmer 2400 analyser. Analytical measurements were carried out in an ARL 3410 1 ICP with Minitorch equipment. IR spectroscopy was performed on a Nicolet 520 FTIR spectrophotometer in the 400–4000 cm21 region. DiVuse reflectance spectra were registered at room temperature on a CARY 2415 spectrometer in the range 5000–45000 cm21. TG curves were obtained using a Perkin-Elmer System-7 DSC-TGA unit at a heating rate of 5 8C min21 in nitrogen.ESR spectroscopy was performed on powdered samples at X frequency, with a Bruker ESR 300 spectrometer, equipped with a standard Oxford low-temperature device, calibrated by the NMR probe for the magnetic field, the frequency being measured by using a Hewlett-Packard 5352B microwave frequency computer. Magnetic susceptibilities of powdered samples were carried out in the temperature range 1.8–300 K using a Quantum Design Squid magnetometer, equipped with a helium continuous-flow cryostat.The magnetic field was approximately 1000 G, a field at which the magnetisation versus magnetic field curve is linear even at 1.8 K. The experimental susceptibilities were corrected for diamagnetism of the constituent atoms (Pascal tables). Table 1 Crystal data and structure refinement for compound 1 Formula M Space group a/Å b/Å c/Å b/8 U/Å3 Z Fe0.5C13N3SH12 270.24 C2/c 19.077(4) 9.995(3) 14.588(4) 110.23(3) 2610(1) 8 T/8C l/Åa robs/g cm23 rcalc/g cm23 m/cm21 Unique data Observed data R(R9) b 20 0.71070 1.33(2) 1.375 7.64 3940 3168 0.0355(0.0927) a Mo-Ka radiation, graphite monochromator.b R = [S |Fo| 2 |Fc| /S|Fo|], R9 = {S[w(Fo 2 2 Fc 2)2]/S[w(Fo 2)2]}� �� where w = 1/s2(|Fo|). Crystal structure determination X-Ray measurements for compound 1 were taken at room temperature on an Enraf-Nonius CAD-4 diVractometer, operating in w–2q scanning mode using suitable crystals for data collection. Accurate lattice parameters were determined from least-squares refinement of 25 well-centred reflections. Intensity data were collected in the q range 2.28–31.078.During data collection, two standard reflections periodically observed showed no signi variation. Corrections for Lorentz and polarization factors were applied to the intensity values. The structure was solved by heavy-atom Patterson methods using the program SHELXS8616 and refined by a full-matrix least-squares procedure on F2 using SHELXL93.17 Nonhydrogen atomic scattering factors were taken from ref. 18. Crystallographic data and processing parameters for compound 1 are shown in Table 1. CCDC reference number 186/1380. See http://www.rsc.org/suppdata/dt/1999/1401/ for crystallographic files in .cif format. X-Ray powder diVraction data for compounds 2 and 3 were collected on a PHILIPS X’PERT powder diVractometer with Cu-Ka radiation in steps of 0.028 (2q) over the 5–608 (2q) angular range and a fixed-time counting of 4 s at 25 8C.The powder diVraction pattern was indexed with the FULLPROF19 program based on the Rietveld method 20–21 using the Profile Matching option. Crystallographic data and processing parameters for compounds 2 and 3 are given in Table 2. Results and discussion Structural analysis The structure of [Fe(NCS)2bpa2] complex 1 consists of linear Table 2 Structural data and refinement for compounds 2 and 3 Compound Formula M Space group a/Å b/Å c/Å b/8 U/Å3 T/8C l/Å Rb a Rp b Rwp c 2 Co0.5C13N3SH12 271.66 C2/c 19.762(1) 9.9430(7) 14.581(1) 110.388(4) 2685.6(1) 25 1.5418 2.88 9.19 12.7 3 Ni0.5C13N3SH12 271.55 C2/c 19.7841(7) 9.8630(6) 14.5239(7) 110.050(5) 2662.3(7) 25 1.5418 1.35 6.27 8.48 a Rb = 100[S|Io 2 Ic|]/S|Io|.b Rp = 100[S|yo 2 yc|]/S|yo|. c Rwp = (S[w|y 2 yc|2]/S[w|yo|2])� �� .J. Chem. Soc., Dalton Trans., 1999, 1401–1406 1403 chains extending along the [010] crystallographic direction (Fig. 1). Metallic Fe(II) ions are bridged through two N,N9 bidentate bpa ligands.The coordination sphere around Fe(II) is a FeN6 octahedron in which the equatorial positions are occupied by four Nbpa atoms while the axial positions are occupied by terminal N-bonded isothiocyanate ligands which are nearly linear. The Fe–NNCS distances [2.069(2) Å] are shorter than the Fe–Nbpa ones (2.221 Å, on average) giving rise to a compressed octahedral coordination sphere in which all the angles are close to the ideal ones. Table 3 summarizes the selected bond distances and angles for compound 1.The intrachain distance between metallic cations [9.995(3) Å] is shorter than the one found in a similar 1-D compound of Fe(II) where the metallic ions are bridged by 4,49-bipyridine 5 as a result of the gauche conformation of the bpa in compound 1 [torsion angle C3–C6–C7–C10 is 68.2(3)8]. On the other hand, the Fe–Fe interchain distance along the z axis in compound 1 [7.292(2) Å] is shorter than the intrachain distance as the chains are connected through hydrogen bonds between SNCS atoms and the pyridyl rings (Table 3).The resulting 2-D network of non-planar layers can be seen in Fig. 2, the Fe–Fe interchain distance being 10.768(4) Å. X-Ray diVraction patterns for compounds 2 and 3 are very similar to the theoretical pattern generated for compound 1. Therefore, the values corresponding to the cell parameters and space group of compound 1 were used as initial data for the refinement of the experimental patterns for compounds 2 and 3.The experimental, calculated (according to the best fit parameters shown in Table 2) and diVerence patterns are shown Fig. 2 View of the non-planar packing for [Fe(NCS)2bpa2] 1. The discontinuous lines represent interchain H-bonds. Table 3 Selected bond distances (Å) and angles (8) for compound 1 Fe–N1 Fe–N2(ii) Fe–N3 N3–C13 C13–S 2.220(2) 2.222(2) 2.069(2) 1.118(2) 1.562(2) N3(i)–Fe–N3 N1(i)–Fe–N1 N1(i)–Fe–N2(iii) N2(ii)–Fe–N2(iii) N3(i)–Fe–N2(ii) N3–C13–S 175.84(9) 84.84(8) 91.99(6) 86.59(8) 94.78(7) 179.3(2) Hydrogen bonds D–H C9–H9 0.86(2) D? ? ?A C9 ? ? ? S2(v) 3.518(2) H? ? ?A H9 ? ? ? S2(v) 2.79(2) D–H? ? ?A C9–H9 ? ? ? S2(v) 143.1(2) Symmetry codes: (i) 2x, y, 2z 1 1/2; (ii) x, y 1 1, z; (iii) 2x, y 1 1, 2z 1 1/2; (iv) x, y 2 1, z; (v) x 2 1/2, y 2 1/2, z.in Fig. 3 for compounds 2 and 3. These results clearly indicate that compounds 1, 2 and 3 are isomorphous.IR and UV-VIS spectroscopies It is worth mentioning that the major application of IR spectroscopy in pseudo-halide containing compounds is usually focused on the bands corresponding to these ambidentate ligands as they can provide very useful information about the pseudo-halide coordination mode. In particular, for N-bonded isothiocyanate, the most intense band appears at about 2050 cm21 being related to the asymmetric streching mode of the C]] N bond, while for S-bonded thiocyanate, this band can be found at about 2100 cm21.Thus, for N,S-bonded thiocyanate, this absorption appears as a split band in the vicinity of these values. A summary of the most important IR bands corresponding to compounds 1, 2 and 3 together with their tentative assignments is given in Table 4. As can be observed, the three IR spectra show an intense single band at about 2065 cm21 which is associated with the nasym streching mode of the C]] N bond of N-coordinated isothiocyanate.On the other hand, the frequencies of the bands related to the bpa ligand in the three compounds are very close to their positions in the free ligand (which are also displayed in Table 4) indicating that the pyridyl rings are nearly planar in the complexes. These results are indicative of a similar coordination of the ligands to the metallic ions, as expected for isomorphous compounds. The diVuse reflectance spectrum for compound 1 is typical of octahedral high spin Fe(II) complexes with Jahn–Teller distortion which specially aVect the 5eg orbitals as a result of the shorter axial distances.22 Thus, the spectrum shows the only spin-allowed transition 5T2gÆ5Eg split into two bands at 10810 and 19646 cm21. The corresponding value of (8/3)ds = 8836 cm21 is comparable to those calculted for Fe(II) octahedral complexes with tetradentate, conjugated, cyclic N4-ligands.23 Fig. 3 Experimental, calculated and diVerence powder X-ray diVraction patterns for (top) [Co(NCS)2bpa2] 2 and (bottom) [Ni(NCS)2- bpa2] 3.1404 J.Chem. Soc., Dalton Trans., 1999, 1401–1406 Table 4 Selected IR bands for bpa and [Fe(NCS)2bpa2] 1, [Co(NCS)2bpa2] 2 and [Ni(NCS)2bpa2] 3 together with their assignments Compound Thiocyanate, n(CN) Pyridyl ring stretching, n(C]] C) n(ArC–C, C]] N) Pyridyl ring breathing, d(ArC–H) Pyridyl out of plane bending, n(ArC–H) Pyridyl ring in plane vibration bpa 1594i 1413i 982m 817i 539i,sp 1 2056i 1614m 1415m 1016m 827m 525m,sp 2 2062i 1614m 1418m 1016m 812m 521m,sp 3 2079i 1614m 1420m 1016m 827m 546m,sp i = intense, m = medium, sp = split.Additionally, the spectrum exhibits a charge-transfer band above 21000 cm21 which is in accordance with the fact that the metallic cation is surrounded by four pyridine rings. For compound 2, the diVuse reflectance spectrum exhibits two absorptions which have been attributed to the spin allowed transitions from 4T1g to 4T2g and 4T1g(P), respectively.The first absorption is split into two at 9435 and 12075 cm21, respectively, while the second is located at 20000 cm21. The split of the first transition must also be interpreted in terms of the Jahn– Teller distortion aVecting the 5eg orbitals. The expected transition to 4A2g appears as a shoulder on the second transition, its position not having been measured. The values of Dq = 1059 cm21 and B = 795 cm21 which have been estimated from these transitions are typical for high spin Co(II) complexes.22 The value of B is indicative of an 80% covalency of the Co–N bonds in compound 2.This spectrum also shows a charge-transfer band above 25000 cm21. Compound 3 exhibits two reflectance spectrum bands at 10566 and 17452 cm21 which have been assigned to the spin allowed transitions from 3A2g to 3T2g and 4T1g. According to these bands, a va Dq = 1057 cm21 has been calculated which is typical for Ni(II) compounds.22 This spectrum also shows a charge-transfer band above 29000 cm21 which hides the expected third transition from 3A2g to 3T1g(P). In summary, diVuse reflectance data clearly indicate that the bpa ligand generates a low perturbation on the octahedral coordination spheres of the title compounds.Additionally, the shorter M–NNCS distances in the axial positions give rise to a tetragonal Jahn–Teller distortion according to which dx2 2 y2 orbitals becomes more stable than dz2 orbitals. ESR and magnetic properties ESR measurements were carried out at several temperatures in the range 2–300 K for compounds 1, 2 and 3.As expected for compounds 1 and 3, having non-Kramer metallic ions, ESR spectra did not show any signals over the whole temperature range. For compound 2, even if X-band isotropic spectra were Fig. 4 X-Band ESR spectra for [Co(NCS)2bpa2] 2 at 80 and 4 K. recorded below 100 K, just those corresponding to temperatures lower than 40 K acquired rhombic resolution. This is illustrated in Fig. 4, where the spectra recorded at 80 and 4 K can be seen. The spectrum at 4 K can be described in terms of an eVective spin doublet S = 1/2 arising from the splitting of the 4T1 term as a result of the spin–orbit coupling in octahedral Co(II). The sum of the three observed g-values, g1 = 5.20, g2 = 4.15 and g3 = 3.00, are close to the theoretical value of 13 proposed by Abragam and Pryce 24 which is consistent with the slightly distorted octahedral sphere around Co(II) in compound 2.Measurement of the thermal variation of the magnetic susceptibility for compounds 1, 2 and 3 showed that cm values continuously increase upon cooling for all three compounds, not exhibiting any maxima in the entire temperature range studied. The experimental data for compound 1, plotted as the thermal variation of the reciprocal susceptibility and the cmT product are shown in Fig. 5. The variation of cm 21 is well described by the Curie–Weiss law over the whole temperature range, with values of Cm and q of 3.88 cm3 K mol21 and 23.7 K, respectively.The calculated value of g, 2.27, lies among the usual ones for octahedral Fe(II) (2.08–2.33).25 The cmT term is practically constant down to 100 K, rapidly decreasing upon further cooling. The thermal variation of cmT together with the sign of the Weiss constant could be consistent with the occurrence of antiferromagnetic coupling. Eqn. (1) can be proposed as a theoretical approach to the magnetic behaviour of compound 1.In this expression, cm is a function of the J parameter due to exchange couplings along an infinite-spin, linear chain 26 with S = 2. cm = 6Ng2b2 3kT F1 2 u 1 1 uG (1) where u = T T0 2 coth S T T0D and T0 = 12 J k N and k are the Avogadro and Boltzmann constants, respectively, and b is the Bohr magneton. According to eqn. (1), the best fit parameters for compound 1 have been determined to be g = 2.27 and J = 20.5 K. The fact that the theoretical curve shows a slight discrepancy with the experimental data at very low temperatures could be indicative Fig. 5 Thermal evolution of cm 21 and cmT for [Fe(NCS)2bpa2] 1 and their corresponding theoretical curves.J. Chem. Soc., Dalton Trans., 1999, 1401.1406 1405 of the occurrence of a second eVect in addition to the antiferromagnetic interactions. Even if the zero field splitting could be proposed to be this second eVect, the occurrence of a spin transition might also be considered.As reported elsewhere,27,28 the N-coordinated bpa could be expected to cause a spin transition in Fe(II) systems. Obviously, the temperature required for the spin transition to occur is related to the perturbation originated by bpa on the ligand field which has been observed to be very weak. Thus, assuming that the spin transition does exist for compound 1, it should be concluded that it takes place at around 10 K finishing at a temperature below 2 K (below the studied range).Comparison between this low temperature and the T1/2 value of 138 K observed for [Fe(DPEA)(NCS)2] 29 [DPEA = (2-aminoethyl)bis- (2-pyridylmethyl)amine] as well as the temperature range of 100.250 K attributed to the spin transition for [Fe(tvp)2(NCS)2]? CH3OH30 [tvp = 1,2-bis(4-pyridyl)ethylene] is illustrative of the weaker ligand field force generated by bpa. Plots of cm 21 and cmT vs. temperature are shown in Fig. 6 for compound 2. As can be seen, the Curie.Weiss law is followed at temperatures down to 50 K, with values of Cm and q of 3.40 cm3 K mol21 and 222.6 K, respectively.The decreasing value of cmT upon cooling could indicate the existence of antiferromagnetic coupling. However, the long pathway through the organic ligand as well as the high value of the Weiss constant preclude that the decrease of meff should be mainly attributed to spin.orbit coupling. In order to theoretically study the magnetic behaviour of compound 2, eqn.(2) has been used.25 This expression proposes meff to be a function of l (related to the spin.orbit coupling) and a (corresponding to the ligand field). The minimum experimental value of meff, 3.82 mB, coincides with the theoretical low field value for octahedral Co(II) which is due to the low perturbation generated by the bpa ligand (as concluded from diVuse reflectance spectra). Therefore, a = 1.5 has been considered for the fitting of eqn. (2). In this way, the best fit value for l has been found to be 2144.8 cm21.This model quite accurately reproduces the experimental data at low temperatures. The slight deviation observed at high temperatures could be caused by the anisotropy of the octahedral sphere. Additionally, the decrease in l in relation to the value corresponding to the free ion (85%) can be attributed to the covalency of the Co.N bonds. The thermal evolution of cm 21 for compound 3, shown in Fig. 7, can be described by the Curie.Weiss law down to 20 K with Cm = 1.19 cm3 K mol21 and q = 20.4 K.The calculated value of g (2.18) is typical for octahedral Ni(II). The thermal meff 2 = 7(3 2 a)2x 5 1 12(2 1 a)2 25a 1 I I O 2(11 2 2a)2x 45 1 176(2 1 a)2 675a ¢� ¢© ¢­ x expS25ax 2 D1 I I O (5 1 a)2x 9 2 20(2 1 a)2 27a ¢� ¢© ¢­ exp(24ax) x 3 I I O 3 1 2expS25ax 2 D1 exp(24ax) ¢� ¢© ¢­ (2) where x = l(kT)21 Fig. 6 Thermal evolution of cm 21 and cmT for [Co(NCS)2bpa2] 2 and their corresponding theoretical curves.trend exhibited by cmT, also shown in Fig. 7, indicates a sharp decrease of meff at temperatures below 8 K which could be associated with the occurrence of antiferromagnetic interactions and a zero field splitting eVect, or a zero field splitting eVect, only. With the aim of analysing the magnetic behaviour of compound 3, eqn. (3) and (4) have been considered. In eqn. (3), cm = 2Ng2b2 3kT I I I I E 2 x 2 2 exp(2x) x 2 1exp(2x) 1 1 2 exp(2x) ¡Æ ¢« ¢« ¢« ¢� (3) where x = D kT cm = Ng2b2 kT F 2 1 0.0194X 1 0.777X2 3 1 4.346X 1 3.232X2 1 5.834X3G (4) where X = J(kT)21 which corresponds to the Van Vleck equation 31 for S = 1, cm is given as a function of the zero field splitting term D.On the other hand, by means of eqn. (4), cm is supposed to depend on the J parameter related to the antiferromagnetic coupling in a 1-D system 32 with S = 1, by considering the ground term 3A2 and the interactions between nearest neighbours. According to both suppositions, the best fit parameters have been found to be g = 2.183 and D = 1.9 cm21 for eqn.(3) and g = 2.182 and J = 0.1 cm21 for eqn. (4). In Fig. 7 the theoretical curves obtained from eqn. (3) and (4) are shown up to 50 K. Theoretical values above this temperature have been omitted as they do not show any significant discrepancies. As can be seen, the curve corresponding to eqn. (3) [cm = f(D)] fits the experimental data more eYciently than the curve corresponding to eqn.(4) [ cm = f(J)]. Additionally, the low value of J obtained from eqn. (4) indicates that, even if the existence of antiferromagnetic coupling should not be neglected, the eVect of the zero field splitting term is more relevant for the magoperties of compound 3. Conclusions Three polynuclear compounds were prepared using the transition metal cations Fe(II), Co(II) and Ni(II) in combination with the pseudo-halide NCS2 and the bidentate organic ligand bpa. X-Ray single crystal diVraction characterization carried Fig. 7 Thermal evolution of cm 21 and cmT for [Ni(NCS)2bpa2] 3 and their corresponding theoretical curves [continuous and discontinuous lines for cmT correspond to cm = f(D) and cm = f(J), respectively].1406 J. Chem. Soc., Dalton Trans., 1999, 1401–1406 out on the compound with Fe(II) revealed that this compound consists of linear chains in which the metallic cations are connected through two bpa ligands in gauche disposition. For compounds with Co(II) and Ni(II), X-ray powder diVraction analysis indicated that the three compounds are isomorphous.IR, UV-VIS and TG data were also consistent with this latter point. Thus, while the isothiocyanate acts as a terminal ligand in these compounds, the bpa ligand has been observed to perform as an eYcient spacer giving rise to 1-D complexes. The thermal variation of the magnetic susceptibility for the three compounds has been interpretated in terms of the occurrence of antiferromagnetic coupling between metal ions along with some other eVects.However, the theoretical treatment of the experimental data revealed that these interactions are weak, in accordance with the long pathway through the voluminous bpa ligand. Acknowledgements This work has been carried out with the financial support of the Universidad del País Vasco/Euskal Herriko Unibertsitatea (Project UPV 130.310.EB234/95) and the Gobierno Vasco/ Eusko Jaurlaritza (Project PI96/39).M. L. H. also thanks the Universidad del País Vasco/Euskal Herriko Unibertsitatea for the grant UPV 130.310.EB234/95. References 1 M. M. Turnbull, T. Sugimoto and L. K. Thompson, Molecule-Based Magnetic Materials, American Chemical Society, Washington, DC, 1996. 2 O. Kahn, Molecular Magnetism, VCH, Weinheim, 1993. 3 S. M. Kuang, Z. Z. Zhang, Q. G. Wang and T. C. W. Mak, J. Chem. Soc., Dalton Trans., 1997, 4457. 4 R. Cortés, M. Drillon, X. Solans, L. Lezama and T. Rojo, Inorg.Chem., 1997, 36, 677. 5 M. L. Tong, X. M. Chen, X. L. Yu and T. C. W. Mak, J. Chem. Soc., Dalton Trans., 1998, 5. 6 A. J. Blake, S. J. Hill, P. Hubberstey and W. S. Li, J. Chem. Soc., Dalton Trans., 1998, 909. 7 T. O. S. Jung, S. H. Park, D. C. Kim and K. M. Lim, Inorg. Chem., 1998, 37, 610. 8 L. Carlucci, G. Ciani, D. M. Proserpio and A. J. Sironi, J. Chem. Soc., Dalton Trans., 1997, 1801. 9 M. Kondo, T. Yoshitomi, K. Seki, H. Matzusaka and S. Kitagawa, Angew. Chem., Int.Ed. Engl., 1997, 36, 1725. 10 A. J. Blake, N. R. Champness, A. Khlobystov, D. A. Lemenovskii, W. S. Li and M. Schröder, Chem. Commun., 1997, 2027. 11 R. Cortés, M. K. Urtiaga, L. Lezama, J. L. Pizarro, M. I. Arriortua and T. Rojo, Inorg. Chem., 1997, 36, 5016. 12 T. L. Hennigar, D. C. MacQuarrie, P. Losier, R. D. Rogers and M. J. Zaworotko, Angew. Chem., Int. Ed. Engl., 1997, 36, 972. 13 K. N. Power, T. L. Hennigar and M. J. Zaworotko, Chem. Commun., 1998, 595. 14 M. Fujita, Y. K. Kwon, M. Miyazawa and K. Ogura, J. Chem. Soc., Chem. Commun., 1994, 1977. 15 J. Lu, T. Paliwala. S. C. Lim, C. Yu, T. Niu and A. J. Jacobson, Inorg. Chem., 1997, 36, 923. 16 G. M. Sheldrick, SHELXS86, Program for the Solution of Crystal Structures, University of Göttingen, 1985. 17 G. M. Sheldrick, SHELXL93, Program for the Refinement of Crystal Structures, University of Göttingen, 1993. 18 International Tables for X-ray Crystallography, Kynoch Press, Birmingham, 1974, vol. IV. 19 J. Rodriguez Carvajal, FULLPROF, Program Rietveld Pattern Matching Analysis of Powder Patterns, 1997. 20 H. M. Rietveld, Acta Crystallogr., 1967, 12, 151. 21 H. M. Rietveld, J. Appl. Crystallogr., 1969, 6, 65. 22 A. B. P. Lever, Inorganic Electronic Spectroscopy, Elsevier Science B.V., Amsterdam, 1984. 23 V. L. Goedken and D. H. Busch, J. Am. Chem. Soc., 1972, 94, 7355. 24 A. Abragam and M. H. L. Pryce, Proc. R. Soc. (London) A, 1951, 206, 173. 25 F. E. Mabbs and D. J. Machin, Magnetism in Transition Metal Complexes, Chapman and Hall, London, 1973. 26 M. E. Fisher, Am. J. Phys., 1964, 32, 343. 27 P. Gütlich, A. Hauser and H. Spiering, Angew. Chem., Int. Ed. Engl., 1994, 33, 2024. 28 H. Toftlund, Coord. Chem. Rev., 1989, 94, 67. 29 G. S. Matouzenko, A. Bousseksou, S. Lecocq, P. J. van Koningsbruggen, M. Perrin, O. Kahn and A. Collet, Inorg. Chem., 1997, 36, 2975. 30 J. A. Real, E. Andrés, M. C. Muñoz, M. Julve, J. Granier, A. Bousseksou and F. Verret, Science, 1995, 268, 265. 31 J. H. Van Vleck, The Theory of Electrical and Magnetic Susceptibilities, Oxford University Press, Oxford, 1932. 32 A. Meyer, A. Gleizes, J. J. Girerd, M. Verdaguer and O. Kahn, Inorg. Chem., 1982, 21, 1729. 33 C. K. Johnson, ORTEP, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976. Paper 9/00096H

ISSN:1477-9226    

DOI:10.1039/a900096h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

DALTON PERSPECTIVE J. Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 1413 Highly active metallocene catalysts for olefin polymerization Walter Kaminsky Institute for Technical and Macromolecular Chemistry, University of Hamburg, Bundesstrasse 45, D-20146 Hamburg, Germany The fascinating story of the discovery of the metallocene– methylaluminoxane catalysts for olefin polymerization is reviewed from its conception up until the first commercial production of polymers. A great number of diVerent titanocenes and zirconocenes have been synthesized that give tailored polymers of totally diVerent structures, and allows control of polymer tacticity, molecular weight and molecular weight distribution to be more eYcient.New kinds of copolymers and elastomers can be synthesized. 1 Introduction The discovery of metallocene catalysts for the polymerization of olefins has opened a frontier in the area of organometallic chemistry, polymer synthesis, and processing.What is the cause for this surge in interest? The classical polymers like polyethylene, polypropylene, and polystyrene are not only the most extensively used polymers, but the production of diVerent polyolefin types also records above-average growth rates. New tailor-made polymers can be synthesized by metallocene catalysts. Some recent reviews give detailed information on metallocene catalysis.1–4 Professor Dr. Walter Kaminsky was born in 1941 in Hamburg. Since 1979 he has been a Full Professor of Technical and Macromolecular Chemistry, was Director of the Institute of Technical and Macromolecular Chemistry and Dean of the Department of Chemistry at the University of Hamburg.He has published more than 200 papers/books, and holds 20 patents. His research interests include metallocene catalysis for olefin polymerization and the pyrolysis of plastic wastes, scrap tyres, and oil shale/oil sand for recycling. He received several awards for his investigations and in 1997 was awarded an Honorary Fellowship from the Royal Society of Chemistry.In 1995, 53.6 × 106 tons of polyolefins were produced worldwide. This amount makes up 47% of the entire production of plastics (Table 1). Furthermore, in the past the extent of the production of individual types such as LLDPE (linear low density polyethylene) or PP doubled in a period of about 5 to 7 years; which is an outstanding growth rate when compared to that of other materials. Estimates show that this development will continue.In the year 2005, the proportion of polyolefines will climb to 55%, taking into account a simultaneous increase in the entire production of plastics. In turn, this means that polyolefins will displace some of the commercial plastics of today that are less easy to manufacture or pose more problems for recycling or waste disposal. Polyolefins are composed solely of carbon and hydrogen (hence the expression ‘sliceable mineral oil’).Being thermoplastics they can be easily processed; used polyolefin materials can be recycled or combusted with a gain in energy, the only products being merely carbon dioxide and water. The basic units ethene and propene are easily obtained from the crack process of mineral oil. Apart from LDPE discovered by ICI, which has a highly branched structure and is produced radically at ethene pressures of 1000–3000 bar (1 bar = 105 Pa), polyolefins are synthesized at far lower pressures using catalysts.The discovery of the catalyst based on titanium tetrachloride and diethylaluminium chloride as cocatalyst was made by Karl Ziegler, who succeeded in polymerizing ethene into HDPE (high density polyethylene) at standard pressure and room temperature in 1953 at the Max-Planck Institute in Mülheim. A little later Giulio Natta, at the Polytechnical Institute of Milan, was able to demonstrate that an appropriate catalyst system was capable of polymerizing propene into semi-crystalline polypropylene.Both Zeigler and Natta were awarded the Nobel Prize for chemistry in 1963 for these discoveries. Natta perceived that propene and long-chain olefins can be assembled in a stereoregular manner, the building blocks of the resulting chain having a defined and recurring arrangement. This alignment has a considerable influence on the functional properties of the material. A statistical arrangement leads to amorphous polypropylene which flows at room temperature, whereas stereoregular polypropylene is crystalline having a melting point of 165 8C.Currently, the Ziegler–Natta catalysts most often applied are Table 1 Worldwide production of polyolefins * (source: Parpinelli Tecnon) (×106 tons) 1983 1990 1995 2005 Polyethylene LDPE 11.3 14.0 14.4 15.8 Polyethylene HDPE/LLDPE 7.9 16.1 22.1 36.1 Polypropylene (PP) 6.4 12.6 17.1 27.7 Fraction of plastics production (%) — 43 47 55 * LDPE = low density polyethylene, HDPE = high density polyethylene, LLDPE = linear low density polyethylene1414 J.Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 heterogeneous. They are comprised of titanium tetrachloride supported on magnesium chloride with triethylaluminium as cocatalyst.1 Lewis bases, such as ethylbenzoate or silanes, are added in the polymerization of propene in order to improve the stereocontrol of the polymerization. As these heterogeneous catalysts are complex systems with diVerent active sites, the polymer structure can be influenced only to a limited degree. 2 Metallocene Catalysts In comparison, metallocene catalysts represent a great development: they are soluble in hydrocarbons, show only one type of active site and their chemical structure can be easily changed. These properties allow one to predict accurately the properties of the resulting polyolefins by knowing the structure of the catalyst used during their manufacture and to control the resulting molecular weight and distribution, comonomer content and tacticity by careful selection of the appropriate reactor conditions.In addition, their catalytic activity is 10–100 times higher than that of the classic Zeigler–Natta systems. The structure of metallocenes, so called ‘sandwich compounds’ in which a p-bonded metal atom is situated between two aromatic ring systems, was uncovered by Ernst O. Fischer and GeoVrey Wilkinson in 1952.5,6 They were both awarded the Nobel Prize in 1973 for this achievement.This compound class initiated a more resourceful organometallic chemistry that did not play a role in larger industrial processes in the past. Metallocenes, in combination with the conventional aluminium alkyl cocatalysts used in Ziegler systems, are indeed capable of polymerizing ethane, but only at a very low activity. Only with the discovery and application of methylaluminoxane (MAO) in our institute in Hamburg in 1977 was it possible to enhance the activity, surprisingly, by a factor of 10 000.7,8 Therefore, MAO plays a crucial part in catalysis with metallocenes. Methylaluminoxane is a compound in which aluminium and oxygen atoms are arranged alternately and free valences are saturated by methyl substituents. It is gained by careful partial hydrolysis of trimethylaluminium and, according to investigations by Sinn 9 and Barron,10 it consists mainly of units of the basic structure [Al4O3Me6], which contains four aluminium, three oxygen atoms and six methyl groups.As the aluminium atoms in this structure are co-ordinatively unsaturated, the basic units (mostly four) join together forming clusters and cages (Fig. 1). These have molecular weights from 1200 to 1600 and are soluble in hydrocarbons. If metallocenes, especially zirconocenes (Fig. 2), are treated with MAO, then catalysts are acquired that allow the polymerization of up to 100 tons of ethene per g of zirconium. At such high activities the catalyst can remain in the product.The insertion time (for the insertion of one molecule of ethene into the growing chain) amounts to some 1025 s only. A comparison with enzymes is not far-fetched. It is generally assumed that the function of MAO is firstly to undergo a fast ligand exchange reaction with the metallocene dichloride, thus rendering the metallocene methyl and dimethyl aluminium compounds (Fig. 3). In a further step, either Cl2 or CH3 2 is abstracted from the metallocene compound by an Fig. 1 Suggestion of a structure for methylaluminoxane clusters by Sinn 9 Al O Al Al O Al O Al O O Al Al Al O O O Al O Al Al O Al Al Al O Al O Al Fig. 2 Structures of metallocenes that are used in the polymerization of olefins Zr X Cl Cl 7 X = C2H4 10 X = Me2Si M X Cl Cl 5 M = Zr, X = C2H4 6 M = Hf, X = C2H4 9 M = Zr, X = Me2Si Zr X Cl Cl R1 R2 R3 R2 R3 R1 R X M Cl Cl 8 X = C2H4, R1 = R2 = R3 = Me 11 X = Me2Si, R1 = R2 = R3 = Me 12 X = Me2Si, R1 = Me, R2 = Ph, R3 = H 15 M = Zr, X = Me2C, R = H 16 M = Hf, X = Me2C, R = H 17 M = Zr, X = Ph2C, R = H 19 M = Zr, X = Me2C, R = But X Zr Cl Cl 14 X = Me2C 18 X = Ph2C Zr R Cl Cl R 1 R = H 2 R = neomenthyl ZrCl2 Me2Si O Me2Si 4 Me ZrCl2 Me2Si Me 13 Fig. 3 Mechanism of the polymerization of olefins by zirconocenes. Step 1: the cocatalyst MAO converts the zirconocene after complexation into the active species which has a free co-ordination position for the monomer and stabilizes the latter.Step 2: the monomer (olefin) is allocated to the complex. Step 3: insertion of the olefin into the zirconium] alkyl bond and provision of a new free co-ordination position. Step 4: repetition of step 3, in a very short period of time (about 2000 propene molecules per catalyst molecule per s), thus rendering a polymer chain ZrCl2 + MAO Zr Me + R Step 1 Step 2 Zr Me + R R n Steps 3 + 4 Zr + Me nJ. Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 1415 Al-centre in MAO, thus forming a metallocene cation and a MAO anion.11–13 The alkylated metallocene cation represents the active centre.Meanwhile, other weakly co-ordinating cocatalysts, such as tetra(perfluorophenyl)borate anions [(C6F5)4B]2, have been successfully applied to the activation of metallocenes.14–17 A further milestone was reached when Brintzinger 18 synthesized chiral bridged metallocenes in 1982 at the University of Konstanz and in 1984, when Ewen,19 at the Exxon Company (USA), was able to demonstrate that appropriate titanocenes render partially isotactic polypropylene. A little later, highly isotactic material was obtained with analogous zirconocenes in our institute.20 After this discovery, a fervent development of industrial and scientific research in the metallocene sector commenced and, up until today, it has not been concluded.Polyolefins, with diVerent microstructures and characteristics, can be custom-made just by varying the ligands on the metallocene (Fig. 2).21–27 By combining diVerent olefins and cycloolefins with one another, the range of characteristics can be further broadened. The production of polyolefins with narrow molecular weight distributions (Mw/Mn = 2), of syndiotactic polymers and of chemically uniform copolymers has not yet been achieved by conventional heterogeneous catalysts. Using metallocene catalysts, it was possible for the first time to produce polyethylenes, polypropylenes and copolymers with narrow molecular weight distributions,28 syndiotactic polypropylene (in technical scale amounts),29 syndiotactic polystyrene, 30 cyclopolymerisates of 1,5-hexadiene,31 cycloolefin copolymers (COC) with high catalytic activity,32 optically active oligomers 33 and composite materials of biomass, powdered metals with polyolefines.34 Organic or inorganic particles (starch, cellulose, quartz sand or powdered metal) can be coated with a hydrocarbon soluble metallocene catalyst and in turn, after polymerization, with a polyolefin film of variable thickness.35 3 Ethene and Propene Polymerization The catalysts allow polymerizations to be controlled precisely.The building blocks are joined together in a linear fashion (see Fig. 3) at only one type of active centre (‘single site’). The polymerization of ethene and propene with selected metallocenes under the same conditions is summarized in Table 2.36 It is evident that the unsubstituted zirconocene Cp2ZrCl2 has a remarkable activity in the polymerization of ethene.Hafnocenes are less active than the analogous zirconocenes. Metallocenes containing tetrahydroindenyl rings which are bridged by C2H4 or Me2Si groups aVord polymers with especially high molecular weights. Surprisingly, the sterically congested and highly substituted metallocene 11 displays the highest activity in the polymerization of ethene amongst the compounds listed in Table 2. This leads to the conclusion that electronic eVects are particularly important for the insertion reaction.37 The molecular weight of polyethylene as well as of polypropylene can be adjusted to lower values by raising the polymerization temperature, the addition of small amounts of hydrogen or decreasing the monomer concentration.Using blends of metallocenes that render polymers with diVerent molecular weights makes the tuning of molecular weight distributions to values of Mw/Mn = 5–10 possible. Apart from isotactic and atactic polypropylene, the syndiotactic as well as the isoblock and stereoblock materials can be achieved for the first time in large quantities and in high purity (Fig. 4). Syndiotactic polypropylene can be synthesized by complexes containing fluorenyl and cyclopentadienyl ligands bridged by X (compounds 15–17, Table 2).29 The block lengths of equal tacticity in isoblock PP (in which the position of the methyl groups are in one direction) can be varied in a broad range (4–100) on the nanoscale (4–100 units).This leads to microcrystalline PP which is suitable for making transparent foils. The development of applications for elastic stereoblock PP has just begun.38 By inserting a silyl bridge and substituting the indenyl ligands in zirconocene 1, the metallocene work-group at Hoechst AG in Frankfurt was able to enhance the activity and stereoselectivity of the isotactic functioning catalyst considerably (compounds 9–13). These catalysts aVord isotactic PP of high molecular weights with a melting point of 161 8C.39 Polypropylenes made by metallocenes exhibit distinct diVerences to conventionally produced polypropylenes, such as narrow molecular weight distributions, higher stiVness and greater tensile strength (Table 3).This is caused not only by the more uniform structure but also by the extremely low fractions of oligomeric products of low molecular weight. These fractions amount to less than 0.1%, compared to 2–4% in Ziegler–Natta PP.Copolymers of ethene and oct-1-ene produced by analogous ‘single site’ catalysts 20 of the Exxon and Dow companies Fig. 4 Microstructures of polypropylene Table 2 Homopolymerization of ethene and propene at 30 8C, 2.5 bar monomer pressure, 6.25 × 1026 mol l21 metallocene concentration, molar MAO–metallocene ratio = 250:1 Ethene Propene Catalyst 123 (C5Me4Et)2ZrCl2 456789 10 11 12 13 14 15 16 17 Activity a 60 900 12 200 18 800 57 800 41 100 2 900 22 200 78 000 36 900 30 200 111 900 16 600 7 600 1 550 2 000 890 2 890 103 M 620 1000 800 930 140 480 1000 190 260 900 250 730 450 25 500 560 630 Activity a 140 170 290 230 1 690 610 1 220 750 1 940 7 700 3 800 15 000 6 100 180 1 550 130 1 980 103 M 23 0.2 0.3 323 446 24 418 > 79 44 192 650 380 3 159 750 729 Isotacticity b (%) 7 59 7 24 95 94 98 99 97 95 94 99 98 49 0.6 0.7 0.4 a Measured in kg polyolefin per mol metallocene per h per concentration of metallocene.b mmmm (meso) Pentads as determined by 13C NMR spectroscopy.Table 3 Comparison of some properties of polypropylene made by Ziegler–Natta and metallocene catalysts 18 Melting point/8C Mw/Mn Stress tensile/N mm22 Hardness/N mm22 Extractable fraction (%) PP from metallocene catalyst 161 2.5 1620 86 0.1 PP from Ziegler–Natta catalyst 162 5.8 1190 76 2–41416 J. Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 in the USA have already captured a market in the packaging industry and in the field of medicinal applications.40 High contents of octene of more than 20 mol % aVord polyolefin elastomers (POE). 4 Elastomers One of the biggest impacts of metallocene catalysts will be in the manufacture of elastomers with elastic properties similar to rubber. The copolymers of ethene and propene with a molar ratio of 1 : 0.5 up to 1 : 2 are of great industrial interest. These ethene–propene (EP) polymers show elastic properties and, together with 2–5 wt. % of dienes as third monomers they are used as elastomers (EPDM).41,42 Since there are no double bonds in the backbone of the polymer, it is less sensitive to oxidation and degradation reaction by daylight.The dienes 5- ethylidene-2-norbornene (ENB) (norbornene = bicyclo[2.2.1]- hept-2-ene), hexa-1,4-diene and dicyclopentadiene are used. Up until today, in most technical processes for the production of EP and EPDM rubber, soluble or highly dispersed vanadium components are used. Similar elastomers which are less coloured, can be obtained with the same metallocene–MAO catalysts used for the homopolymerization of ethene or propene at a much higher activity.The regiospecificity of the metallocene catalysts towards propene leads exclusively to the formation of head-to-tail enchainments. 5-Ethylidene-2-norbornene polymerizes via vinyl polymerization of the cyclic double bond and the tendency to branch is low. The molecular weight distribution of about 2 is narrow. At low temperatures the polymerization time to form one polymer chain is long enough to consume one monomer and then to add another one.So, it becomes possible to synthesize block copolymers if the polymerization, catalyzed specifically by hafnocenes, starts with propene and, after the propene is nearly consumed, continues with ethene. High branching, which is caused by the incorporation of long chain olefins into the growing polymer chain, is obtained with a new class of silyl bridged amido(cyclopentadienyl)- titanium compound 20.43,44 This catalyst, used by Dow and Exxon in combination with MAO or borates, incorporates oligomers with vinyl end-groups which are formed during polymerization by b-hydrogen transfer resulting in long chain branched polyolefins.In contrast, structurally linear polymers are obtained when catalyzed by other metallocenes. Long branched copolymers of ethene with oct-1-ene show elastic properties as long as the comonomer content is more than 20%.Other elastomers with diVerent microstructures can be synthesized from dienes.45 5 Cycloolefin Copolymers (COC) Metallocene catalysts are particularly important for the polymerization of cycloolefins (cyclopentene, norbornene and their substituted compounds) (Fig. 5). In this process, only the double bond is opened and not the ring. Crystalline polycycloolefins are rendered, that have extremely high melting points of at least 380 8C, sometimes being higher than the decomposition temperature.46 While homopolymerization of cyclopentene results in 1,3- enchainment of the monomer units, norbornene is inserted in 1,2-enchainment as usual for olefin polymerization.The prob- Si N Ti Cl Cl 20 lems of processing that arise from the high melting temperatures of the homopolymers can be solved by copolymerizing cycloolefins with ethene, for example (Fig. 6).47–49 The insertion of norbornene units into the growing polymer chain is very easy.As seen by the copolymerization parameter r1, which is between 2.0 and 3.4 and shows how much faster ethene is inserted than norbornene when the previous insertion was ethene, it is easy to incorporate this huge monomer. For the copolymerisation of ethene and propene r1 is between 3 and 6. Table 4 compares activities and incorporation of norbornene for diVerent catalysts. The metallocene 19 shows not only high activities for the copolymerization of ethene with norbornene, but gives an alternating structure, too.50 Most metallocenes produce polymers with a statistical structure.It is impossible to achieve copolymers with more than 50 mol % of norbornene. The melting point of the alternating copolymer depends on the molar ratio of norbornene units in the polymer while the glass transition temperature is nearly independent of this. A maximum melting point of 320 8C was reached. Such materials characteristically have an excellent transparency and a very high continuous service temperature. From cycloolefin insertion rates of 10 mol % upwards, these cycloole- fin copolymers (COC) are no longer crystalline but amorphous.They are very resistant towards solvents and chemicals, they exhibit high softening temperatures (glass temperatures of up to 200 8C) and can be processed on a thermoplastic basis. A further peculiarity of these materials is their tendency to absorb little light, which makes them suitable for optoelectronic applications.Norbornene–ethene copolymers are most interesting for technical uses because of their readily available monomers. Currently, such COC polymers are already being used for the production of compact discs in Japan within the framework of a joint venture between the Hoechst and Mitsui companies. Applications for optoelectronic data transfer and storage as well as for other areas of high technology have been provided for. Fig. 5 Cycloolefins used for polymerization by metallocenes Cyclopentene Norbornene DMON Fig. 6 Structures of norbornene (N)–ethene (E) copolymers. Alternating blocks can also occur NENEN EENEE Table 4 Copolymerization of norbornene (N) and ethene by diVerent metallocene–MAO catalysts at 30 8C* Catalyst 1597 15 17 18 Time/min 30 10 15 40 10 10 15 Activity/kg mol21 h21 1200 9120 2320 480 7200 6000 2950 Incorporation of N (%) 21.4 26.1 28.4 28.1 28.9 27.3 33.3 * Conditions: MAO–Zr = 200, [metallocene] = 5 × 1026 mol l21; ethene pressure = 2 bar, [N] = 0.05 mol l21.J.Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 1417 Fig. 7 Contour diagrams of (a) zirconocene 5 and (b) zirconocene 8. Complete conformational analysis of the angle of torsion in 58 steps. ce is the centroid of the five-membered ring of the indenyl and ce9 that of the second indenyl. C(bridge)]ce]Zr]ce9 is a measurement of the orientation of the indenyl group relative to the ZrCl2 fragment, while C9(bridge)]ce9]ce]C(bridge) describes the orientation of the indenyl rings with respect to one another.The lines represent isoenergetic areas. Energies are scaled to the energy minimum 11.3 kcal mol21 (Emin = 0 at l for 5) and 15.8 kcal mol21 (at l for 8) 6 Supporting of Metallocene Catalysts Metallocene catalysts in dissolved forms are unsuitable for the production of polyethylene or isotactic polypropylene on an industrial scale. In order to use them in existing technical processes (drop-in technology) by exchanging them for the conventional Ziegler–Natta catalysts, the metallocenes have to be applied to a powdery, insoluble substrate. One way to do so is to support them on silica, alumina, magnesium dichloride or other supports.DiVerent methods are possible.51 Two of them are: (1) initial absorption of MAO on the support with subsequent addition of metallocenes in a second step which is mostly used. These washed catalysts are used in combination with additional MAO or other aluminium alkyls in the polymerization. (2) Another way is either absorption and immobilization of the metallocene first or direct bonding by a spacer to the support surface.Then after addition of MAO, this catalytic system is used in the polymerization process. Both procedures aVord diVerent catalysts and these in turn produce polyolefins with diVerent properties.52,53 The polymers obtained by method (1) are very similar to those obtained by the homogeneous system. Each metallocene on the support forms an active center and the starting point for the growth of a polymer chain. As the active sites on the surface of each catalyst grain are identical, all chains grow uniformly resulting in polymers with narrow molecular weight distributions.If the metallocene is linked to the support first, diVerent absorptions occur. A large part of the metallocene is also destroyed by acid centers. This diVerent bonding leads to diVerent active sites. Therefore, the activity is much lower than in the case of the homogeneous system and the molecular weight distribution of the produced polymer is much broader. 7 Syndiotactic Polystyrene Idemitsu was able to demonstrate that titanium compounds combined with MAO are capable of polymerizing styrene in a syndiotactical manner.30 Moreover, trichloro(cyclopentadienyl) titanium (CpTiCl3) has been proved to be remarkably active.54 Syndiotactic polystyrene is crystalline and shows a melting point of 275 8C, which nearly makes it a high performance plastic (Table 5). Previously, it was already possible to produce isotactic polystyrene with classical Ziegler–Natta catalysts with very low polymerization activities.However, it crystallized so slowly that technical usages were unthinkable. Furthermore, the polymerization activity of CpTiCl3–MAO catalysts was also unsatisfactory for technical usage. If fluorinated complexes are employed such as trifluoro(pentamethylcyclopentadienyl)titanium, the activity can then be improved by a factor of 30 (Table 6).55 At the same time the molecular weight rises from 169 000 to 660 000.The copolymerization of styrene with ethene, as examined by Mülhaupt, expands the property domains and employment areas beyond that.56 Syndiotactic polystyrene has already been produced in technical amounts by Idemitsu. 8 Modeling of Metallocene Catalysis Explanations for the strong temperature dependency of particular zirconocenes that function isotactically were sought using force-field calculations in order to enable the optimization of metallocenes that are as stereorigid as possible.Ethylidene bridged zirconocenes are possibly less rigid than assumed so far. Potential flexibility of the molecular structure could coincide with loss of stereospecifity for polymerisations at ‘high’ temperatures. Substituents in the 2- and 7-positions of the indenyl structure should restrict the mobility of the molecule, especially the bridge ‘twist’, because of their interaction with the ethylidene bridge.For this reason, the mobility of ethylidenebis[(2,4,7-trimethyl)indenyl]zirconium dichloride 8 was compared to that of ethylidenebis(indenyl)zirconium dichloride 5 (Fig. 7).57 The contour diagram of the energy hypersurface proves that, Table 5 Properties of atactic, isotactic and syndiotactic polystyrene (PS) Structure Crystallization rate Glass temperature/8C Melting point/8C Atactic PS Amorphous — 100 — Isotactic PS Crystalline Slow 99 240 Syndiotactic PS Crystalline Fast 100 275 Table 6 Synthesis of syndiotactic polystyrene Catalyst CpTiCl3 CpTiF3 Cp*ZrCl3 Cp*TiCl3 Cp*TiF3 Temperature/8C 50 50 30 50 50 Activity * 1 100 3 000 0.01 15 690 M.p./8C 258 265 249 275 275 Mn 140 000 100 000 20 000 169 000 660 000 Mw/Mn 1.9 2.0 2.2 3.6 2.0 * Measured in kg PS per mol metallocene per h.1418 J.Chem. Soc., Dalton Trans., 1998, Pages 1413–1418 considering the same energy, 5 in comparison to 8 can occupy an area on the energy hypersurface that is four times larger.Furthermore, the modeling shows that 5 exists in two conformers, the d- and l-form, that are only separated by a small barrier of 1.5 kcal mol21 (cal = 4.184 J). This is supplemented by NMR spectroscopic investigations which show that a conversion of the d-form into the l-form is feasible at room temperature. In contrast, substituted 8 exists only in the l-form. Silyl bridged zirconocenes also exhibit only one energy minimum, by which its higher stereospecifity can be explained.Force-field calculations demonstrate, thereby, why some metallocenes exhibit such a high temperature dependency of the isotacticity.58 The reason can be found, for example, in the coexistence of two conformers, that are only separated by a low barrier and that are, therefore, more unstable aVording polypropylene with a lower isotacticity. 9 Outlook Meanwhile, the first products synthesized by metallocene catalysts, such as LLDPE, POE, EPDM elastomers or certain types of isotactic polypropylene, are already commercially available.The COC polymers and syndiotactic polystyrene are to follow. The application of metallocenes beyond the synthesis of polymers is discernable as in the enantioselective preparation of fine chemicals of low molecular weight. A racemic mixture of chiral metallocenes can be separated into its enantiomers. Using optically active catalysts, oligomerization leads to optically active branched olefins.The development of metallocene catalysts has not yet been concluded and for certain processes, such as the isotactic polymerization of propene, only temporary optimum values have been reached. The future will certainly yield further custommade structures and a whole string of new polymers and oligomers; the catalysts involved will have metallocenes as their main component. 10 References 1 K.Soga and T. Shiono, Prog. Polym. Sci., 1997, 22, 1503. 2 W. Kaminsky and M. Arndt, Adv. Polym. Sci., 1997, 127, 143. 3 M. Bochmann, J. Chem. Soc., Dalton Trans., 1996, 255. 4 G. Fink, R. Mülhaupt and H. H. Brintzinger (Editors), Ziegler Catalysts, Springer-Verlag, Berlin, 1995; H. H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger and R. Waymouth, Angew. Chem., 1995, 107, 1255. 5 G. Wilkinson and I. M. Birmingham, J. Am. Chem. Soc., 1954, 76, 4281. 6 E. O. Fischer, Angew. Chem., 1952, 22, 620. 7 H. Sinn, W. Kaminsky, H. J. Vollmer and R. Woldt, Angew. Chem., 1980, 92, 396. 8 W. Kaminsky, Macromol. Chem. Phys., 1996, 197, 3907. 9 H. Sinn, Macromol. Symp., 1995, 97, 27. 10 Y. Koide, S. G. Bott and A. R. Barron, Organometallics, 1996, 15, 2213. 11 J. J. Eisch, S. I. Bombrick and G. X. Zheng, Organometallics, 1993, 12, 3856. 12 P. Tait, in Transition Metals and Organometallics as Catalysts for Olefin Polymerization, eds. W. Kaminsky and H. Sinn, Springer Press, Berlin, 1988. 13 R. F. Jordan, W. E. Dasher and S. F. Echols, J. Am. Chem. Soc., 1986, 108, 1718. 14 C. Sishta, R. M. Hathorn and T. J. Marks, J. Am. Chem. Soc., 1992, 114, 114. 15 J. Powell, A. Lough and T. Saeed, J. Chem. Soc., Dalton Trans., 1997, 4137. 16 R. Duchateau, S. J. Lancaster, M. Thornton-Pett and M. Bochmann, Organometallics, 1997, 16, 4995. 17 J. K. G. Erker and R. Fröhlich, J. Am. Chem. Soc., 1997, 119, 11 165. 18 F. R. W. P. Wild, L. Zsolnai, G. Huttner and H.H. Brintzinger, J. Organomet. Chem., 1982, 232, 233. 19 J. A. Ewen, J. Am. Chem. Soc., 1984, 106, 6355. 20 W. Kaminsky, K. Külper, H. H. Brintzinger and F. R. W. P. Wild, Angew. Chem., Int. Ed. Engl., 1985, 24, 507. 21 W. Spaleck, F. Küber, A. Winter, J. Rohrmann, B. Bachmann, M. Antberg, V. Dolle and E. F. Raulus, Organometallics, 1994, 13, 954. 22 R. Mülhaupt, T. Ruschek and B. Rieger, Makromol. Chem., Macromol. Symp., 1991, 48/49, 317. 23 P. Schwarz, E. Siebel, R. D. Fischer, D.C. Apperly, N. A. Davies and R. K. Harris, Angew. Chem., 1995, 107, 3813. 24 H. Schumann, M. Glanz, E. C. E. Rosenthal and H. Hemling, Z. Anorg. Allg. Chem., 1996, 622, 1865. 25 H. G. Alt, J. S. Han and U. Thewalt, J. Organomet. Chem., 1993, 456, 89. 26 V. Busico, R. Cipullo, G. Monaco and M. Vacatello, Macromolecules, 1997, 30, 6251. 27 A. Carvill, J. Tritto, P. Locatelli and M. C. Sacchi, Macromolecules, 1997, 30, 7056. 28 H. Sinn and W. Kaminsky, Adv. Organomet. Chem., 1980, 18, 99. 29 J. A. Ewen, R. L. Jones, A. Razavi and J. P. Ferrara, J. Am. Chem. Soc., 1988, 110, 6255. 30 N. Ishihara, M. Kuromoto and M. Uol, Macromolecules, 1988, 21, 3356. 31 G. W. Coates and R. M. Waymouth, J. Am. Chem. Soc., 1991, 113, 6270. 32 W. Kaminsky and R. Spiehl, Makromol. Chem., 1989, 190, 515. 33 W. Kaminsky, A. Ahlers and N. Möller-Lindenhof, Angew. Chem., 1989, 101, 1304; Angew. Chem., Int. Ed. Engl., 1989, 28, 1216. 34 J. Dutschke, W. Kaminsky and H. Lüker, in Polymer Reaction Engineering, eds.K. H. Reichert and W. Geiseler, Hanser Publications, München, 1983, p. 207. 35 W. Kaminsky and H. Zielonka, Polym. Adv. Technol., 1993, 4, 415. 36 W. Kaminsky, R. Engehausen, K. Zoumis, W. Spaleck and J. Rohrmann, Makromol. Chem., 1992, 193, 1643. 37 B. Peifer, M. B. Welch and H. G. Alt, J. Organomet. Chem., 1997, 544, 115. 38 M. D. Bruce, G. W. Coates, E. Hauptmann, R. W. Waymouth and J. W. Ziller, J. Am. Chem. Soc., 1997, 119, 11 174. 39 W. Spaleck, M. Aulbach, B. Bachmann, F. Küber and A. Winter, Macromol. Symp., 1995, 89, 237. 40 J. P. Stevens, Proceedings MetCon Houston, May 26–28, 1993, no. 15. 41 W. Kaminsky and M. Miri, J. Polym. Sci., Polym. Chem. Ed., 1985, 23, 2151. 42 J. C. W. Chien and D. He, J. Polym. Sci., Part A: Polym. Chem., 1991, 29, 1585. 43 P. J. Shapiro, E. Bunel, W. P. Schaefer and J. E. Bercaw, Organometallics, 1990, 9, 867. 44 A. K. Hughes and A. J. Kingsley, J. Chem. Soc., Dalton Trans., 1997, 4139. 45 G. Ricci and L. Porri, Macromol. Chem. Phys., 1997, 198, 3647. 46 W. Kaminsky, A. Bark and M. Arndt, Makromol. Chem., Macromol. Symp., 1991, 47, 83. 47 H. Cherdron, M.-J. Brekner and F. Osan, Angew. Makromol. Chem., 1994, 223, 121. 48 N. Herfert, P. Montag and G. Fink, Makromol. Chem., 1993, 194, 3167. 49 S. Collins and W. M. Kelly, Macromolecules, 1992, 25, 233. 50 M. Arndt and I. Beulich, Catalysis Symposium Friedrichroda, 20–23 March, 1996, poster no. 7. 51 W. Kaminsky and F. Renner, Makromol. Chem., Rapid Commun., 1993, 14, 239. 52 K. Soga and M. Kaminaka, Makromol. Chem., Rapid Commun., 1992, 13, 221. 53 S. Collins, W. M. Kelly and D. A. Holden, Macromolecules, 1992, 25, 1780. 54 A. Zambelli, L. Olivia and C. Pellecchia, Macromolecules, 1989, 22, 2129. 55 W. Kaminsky, S. Lenk, V. Scholz, H. W. Roesky and A. Herzog, Macromolecules, 1997, 30, 7647. 56 S. Jüngling, R. Mülhaupt, D. Fischer and F. Langhauser, Angew. Makromol. Chem., 1995, 229, 93. 57 G. Schupfner and W. Kaminsky, J. Catal., 1995, 102, 59. 58 G. Guerra, L. Cavallo, V. Venditto, M. Vacatello and P. Corradini, Makromol. Chem., Macromol. Symp., 1993, 69, 237. Received 2nd January 1998; Paper 8/00056E

ISSN:1477-9226    

DOI:10.1039/a800056e

出版商:RSC    

年代:1998

数据来源: RSC