DALTON FULL PAPER J. Chem. Soc., Dalton Trans., 1999, 119–125 119 Combined versus individual labilising eVects of H1, Na1 and nucleophile on catalysed substitution reactions: studies on [Fe4S4X4]22 (X 5 Cl or PhS)† Richard A. Henderson John Innes Centre, Nitrogen Fixation Laboratory, Norwich Research Park, Colney, Norwich, UK NR4 7UH. E-mail: richard.henderson@bbsrc.ac.uk Received 6th October 1998, Accepted 27th November 1998 The reactions between [Fe4S4X4]22 (X = PhS or Cl) and Et2NCS2 2 to form [Fe4S4X2(S2CNEt2)2]22 have been studied in MeCN.The kinetics are consistent with a dissociative mechanism under all conditions. The addition of Na1 led to an increase in rate for [Fe4S4(SPh)4]22 and analysis of the kinetics indicates that a single Na1 binds and labilises the cluster. Comparison is drawn with the established eVect of H1 on the lability of this cluster. The presence of a thiolate ligand is necessary to bind Na1 since the reaction between [Fe4S4Cl4]22 and Et2NCS2 2 is unaVected by the addition of Na1.The addition of acid to [{Fe4S4(SPh)4}Na]2 further accelerates the rate of substitution. Quantitative analysis shows that the combined labilising eVect of Na1 and H1 is no more than that expected from the individual labilisation aVorded by each cation. Similar analyses show the same is true for H1 and nucleophile in acid-catalysed associative substitution mechanisms, and two H1 in the acid-catalysed dissociative mechanisms of Fe–S-based clusters.The generality of these observations is discussed. Introduction There are many substitution reactions which are catalysed or inhibited in the presence of other species; for example, substitution reactions 1 in the presence of H1. Elaborations of this chemistry could involve catalysis by several components [A, B, C, etc. . . ., as illustrated in equation (1)]. For example, A and B could both be H1. M–X 1 Y A, B, C, etc. . . . M–Y 1 X (1) An important mechanistic question in such systems is, “Is the combined labilising eVect of all these species diVerent to that expected from each contributor or is there a cooperative eVect when all the components A, B, C, etc.are present?” Specifically, for the acid-catalysed reactions, “Is the eVect of two H1 different from that expected from compounding the eVect from one H1 with another H1?” Although these are fundamental mechanistic questions which relate to the reactions of many compounds, we are unaware of any study which addresses this problem.This is because such a study requires that the elementary rate constants for the dissociation of M–X to be determined in the presence of A, in the presence of B, as well as in the presence of A and B together. In most systems it is not possible to ‘dissect’ kinetically the dissociation rate constants from the binding constants of A and B. However, our studies on the acid-catalysed substitution reactions of synthetic Fe–S-based clusters 2–7 have shown that the binding of H1 or nucleophile are rapid equilibrium reactions which are followed by the slow dissociation of the leaving group.Analysis of the kinetic data invariably allows us to determine the dissociation rate constants. Herein, we report kinetic studies on the substitution reaction shown in equation (2) (X = Cl or PhS) and: (i) compare the eVects that Na1 and H1 have on the lability † Supplementary data available: kinetic data.For direct electronic access see http://www.rsc.org/suppdata/dt/1999/119/, otherwise available from BLDSC (No. SUP 57466, 5 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton). [Fe4S4X4]22 1 2Et2NCS2 2 H1, Na1, etc. [Fe4S4(S2CNEt2)2X2]22 1 2X2 (2) of the clusters; (ii) quantify the combined eVect of binding Na1 and H1 on the lability of the cluster in terms of the individual labilising eVects of these two cations, and (iii) a quantitative analysis of the relative labilising eVects of binding H1 and PhSH in the associative substitution mechanisms of [Fe4S4Cl4]22. In order to investigate the eVect of Na1 it has been necessary to use Et2NCS2 2 as the nucleophile.Previously similar studies used RS2 or ArS2 as the nucleophile.2–7 However, both NaSR and NaSAr are very poorly soluble in MeCN, and precipitation of these compounds precludes studying the reactions. NaS2CNEt2 is suYciently soluble in MeCN to avoid this complication.Clearly, in the reactions of [Fe4S4(SPh)4]22 in the presence of Na1, some NaSPh will be produced. However, the low amounts of NaSPh formed ([NaSPh] £ 0.2 mmol dm23) are suYciently soluble in MeCN for the reaction to remain homogeneous. Results and discussion EVects of H1 on the reactivity of [Fe4S4(SPh)4]22 In a series of kinetic studies we have been studying the acidcatalysed substitution reactions of a variety of synthetic Fe–Sbased clusters: reactions essential in understanding the multiproton, multi-electron, substrate transformation chemistry of these compounds.8 The generalised picture which has emerged from these studies is exemplified by that of [Fe4S4(SPh)4]22 shown in Fig. 1. Initial protonation occurs at the thiolato-S and subsequently at two m3-S atoms. It is protonation of these m3-S which labilises the cluster towards substitution. Protonation of the thiolate ligand is, apparently, not appreciably labilising. The reasons for this have been discussed in detail earlier 5,8 but, briefly, are a consequence of protonation at this site decreasing the s-donor but increasing the p-acceptor abilities of the ligand.The nett120 J. Chem. Soc., Dalton Trans., 1999, 119–125 Fig. 1 EVect of successive addition of H1 to [Fe4S4(SPh)4]22 on the lability of the cluster in the dissociative substitution reactions with EtS2. For simplicity only one PhS ligand is shown; d = Fe, s = S. eVect is that the Fe–thiolate and Fe–thiol bond strengths are very similar, and consequently the lability is unchanged.This proposal is consistent with structural studies on mononuclear thiolate complexes.9 These reactions of [Fe4S4(SPh)4]22 operate by an acidcatalysed dissociative substitution mechanism and analysis of the kinetics gives the values of k0 = 1.0 ± 0.2 × 1022 s21, kH = 8.0 ± 0.2 × 1022 s21 and kHH = 0.39 ± 0.02 s21. The ratio, kH/ k0 = 8.4 ± 1.6, describes the labilising eVect a single protonation has on the dissociation of the leaving group.Similarly kHH/ k0 = 41 ± 8 describes the eVect of diprotonation. It is evident, that to a reasonable approximation, kHH/k0 = (kH/k0)2. That is, the labilising eVect of each successive H1 is compounded. Clearly, we are not looking for an exact relationship here. Merely a guide as to whether there are orders of magnitude diVerence between the combined and the individual labilising eVects of each contributor to the activated complex.In the remainder of this paper we will see that similar equations describe the labilisation of the leaving group in Fe–S clusters by the combined eVects of: (i) H1 and Na1 and (ii) H1 and nucleophile. The clusters studied in this work10,11 and the products of the reactions [equation (2)] have already been structurally well characterised. Earlier synthetic studies showed that the addition of at least two mole equivalents of Et2NCS2 2 to [Fe4S4X4]22 (X = Cl or PhS) results in the formation of [Fe4S4- (S2CNEt2)2X2]22 and X-ray crystallography has established that the Et2NCS2-ligands are bound in a bidentate fashion to the Fe atoms.12 EVect of Na1 on the reactivity of [Fe4S4(SPh)4]22 and [Fe4S4Cl4]22 When studied on a stopped-flow apparatus, the reaction between [Fe4S4(SPh)4]22 and an excess of [NBun 4]S2CNEt2 is associated with a biphasic absorbance–time curve, provided [Et2NCS2 2] < 20 mmol dm23.The initial absorbance is that of [Fe4S4(SPh)4]22 and the final absorbance corresponds to [Fe4S4(SPh)2(S2CNEt2)2]22.At higher concentrations of Et2- NCS2 2 the absorbance–time curve becomes more complicated, with an increasing absorbance over protracted times (>20 s). For simplicity we have: (i) studied the kinetics only when [Et2NCS2 2] < 20 mmol dm23 and (ii) restricted the discussion to the first substitution reaction, corresponding to the initial phase of the absorbance–time curve. In order to get accurate rate constants for this first phase a method has been adopted, used in earlier studies, involving fitting the entire curve to two exponentials, and from this analysis obtaining the rate constant for the faster phase.The rate of the initial substitution reaction exhibits a first order dependence on the concentration of [Fe4S4(SPh)4]22 but is independent of the concentration of Et2NCS2 2 (kobs = 2.0 ± 0.5 × 1022 s21). This value is in good agreement with the rate constant measured using EtS2 or ButS2 in earlier studies.The first-order dependence on the concentration of cluster is indicated by the exponential shape of the absorbance–time curve, and is confirmed by experiments in which the concentration of the cluster was varied ([Fe4S4(SPh)4 22] = 0.02–0.2 mmol dm23) whilst keeping the concentration of Et2NCS2 2 constant (5.0 mmol dm23). Under these conditions the value of kobs did not vary. These kinetics are consistent with the uncatalysed dissociative mechanism shown in the centre of Fig. 2, in which dissociation of the Fe–SPh bond has to occur before Et2NCS2 2 binds to the cluster. The addition of Na[BPh4] results in an increase in the rate of the reaction as shown in Fig. 3. The dependence on the concentration of Na1 is complicated. At low concentrations of Na1 the rate exhibits a first order dependence on the concentration of Na1, but at high concentrations the rate becomes independent of the concentration of Na1.Experiments in which the concentration of Et2NCS2 2 was varied (maintaining [Na1] = 10.0 mmol dm23), showed that the rate of the reac-J. Chem. Soc., Dalton Trans., 1999, 119–125 121 Fig. 2 Dissociative substitution pathways in the reactions of [Fe4S4(SPh)4]22 with Et2NCS2 2 in MeCN at 25.0 8C. Shown (from left to right) are: (i) acid-catalysed pathway; (ii) uncatalysed pathway; (iii) Na1-catalysed pathway and (iv) Na1- with acid-catalysed pathway. Only one PhS ligand is shown for simplicity; d = Fe, s = S.tion is independent of thiolate. Analysis of these data by the usual “double reciprocal” graph13 gives the rate law shown in equation (3). 2d[Fe4S4(SPh)4 22] dt = {(2.0 ± 0.5) × 1022 1 (1.5 ± 0.1) × 102[Na1]} 1 1 (5.1 ± 0.3) × 102[Na1] [Fe4S4(SPh)4 22] (3) Fig. 3 Kinetics for the reaction between [Fe4S4(SPh)4]22 (0.1 mmol dm23) and Et2NCS2 2 in MeCN at 25.0 8C. The bottom curve illustrates the eVect of Na1 on the rate of the reaction (d); the curve is that defined by equation (3).The top curve shows the combined eVect of Na1 and H1; data points correspond to: [NHEt3 1] = 10 mmol dm23 (s), [NHEt3 1] = 20 mmol dm23 (g); [Et2NCS2 2] = 2.0 mmol dm23, [NEt3] = 1–20 mmol dm23. The curve is that defined by equation (6). This behaviour is consistent with the pathway shown in Fig. 2 in which Na1 rapidly binds to the cluster and this labilises the Fe–SPh bond to dissociation. This behaviour is directly analogous to that observed with H1, and the eVects of Na1 and H1 will be compared below. First, the way Na1 binds to [Fe4S4- (SPh)4]22 will be considered.The binding of Na1 to other Fe–S clusters has been observed crystallographically. Thus, in [Na2{Fe6S9(SMe)2}2]62, each Na1 is bound to three m3-S;14 in [a-Na2Fe18S30]82 and [b-Na2Fe18- S30]82, each Na1 is bound to four m-S 15,16 and in [Na9Fe20Se38]92 each Na1 is bound to four m-Se.16 Finally, there is evidence that Na1 interacts with the “double-cubane” [{MoFe3S4(SEt)2- (Cl4cat)}2(m-SEt)2]42 (Cl4cat = C6Cl4O2 22).17 Molecular modelling studies indicate that a Na1 could bind to two m3-S and two m-SEt residues.With these precedents in mind, it seems likely that Na1 binds to [Fe4S4(SPh)4]22 using one SPh and two m3-S as shown in Fig. 2. The rate law for the reactions in the presence of Na1 is readily derived by assuming that binding Na1 is a rapid equilibrium reaction (complete within the dead-time of the stopped-flow apparatus, 2 ms), and that subsequent dissociation of the Fe–SPh bond is rate-limiting.The result is shown in equation (4), and comparison of equations 2d[Fe4S4(SPh)4 22] dt = {k0 1 kNaKNa[Na1]} 1 1 KNa[Na1] [Fe4S4(SPh)4 22] (4) (3) and (4) gives k0 = 1.5 ± 0.5 × 1022 s21, kNa = 0.30 ± 0.04 s21 and KNa = 5.1 ± 0.3 × 102 dm3 mol21. A quantitative measure of the labilising eVect of Na1 is given by kNa/k0 = 31 ± 6. Equation (4) is directly comparable to the rate law describing the eVect of [NHEt3]1 on the substitution reactions 2 of [Fe4- S4(SPh)4]22.In this case the total ‘proton’ concentration is expressed as [NHEt3 1]/[NEt3], and the rate law is that shown in equation (5), with k0 = 1.0 ± 0.2 × 1022 s21, kH = 8.0 ± 0.1 × 1022 s21 and KH = 1.2 ± 0.1.122 J. Chem. Soc., Dalton Trans., 1999, 119–125 2d[Fe4S4(SPh)4 22] dt = {k0 1 kHKH[NHEt3 1]/[NEt3]} 1 1 KH[NHEt3 1]/[NEt3] [Fe4S4(SPh)4 22] (5) By comparing kNa and kH derived from equations (4) and (5) respectively, a quantitative measure of the relative labilising eVects of H1 and Na1 is obtained, kNa/kH = 4.1 ± 0.3.Although this is not a large diVerence it is, at first sight a surprising result. Intuitively, it might be expected that H1 would be more labilising than Na1 since H1 is a more polarising cation. In addition, in the reactions with acid, H1 is labilising a thiol ligand (Fig. 1) whereas Na1 is labilising a thiolate ligand (Fig. 2), making the greater labilising power of Na1 even more unexpected.The reasons for this are not entirely clear but we suggest that (at least) part of the reason is because a single Na1 is suYciently large to interact with both the thiolate ligand and two m3-S simultaneously. Consequently, the labilising interactions of Na1 with the leaving group and m3-S are always in concert. In contrast, with the smaller H1, such a concerted interaction is not possible and multiple protonations must occur to attain the same eVect.The importance of the thiolate ligand in facilitating the binding of Na1 to [Fe4S4(SPh)4]22 is emphasised in studies with [Fe4S4Cl4]22. The kinetics of the reactions of [Fe4S4Cl4]22 with [NBun 4]S2CNEt2 are independent of the concentration of Et2NCS2 2, with k0 C = 3.0 ± 0.5 s21. This rate constant is in good agreement with that observed earlier for the dissociative substitution pathway using PhS2 as the nucleophile (k0 C = 2.0 ± 0.3 s21).4 The rate of the reaction of [Fe4S4Cl4]22 with Et2NCS2 2 is unaVected by the presence of Na1.Since the geometries of the cluster cores of [Fe4S4(SPh)4]22 and [Fe4S4Cl4]22 are essentially identical, this indicates that Cl is a poorer ligand than PhS for Na1. Assuming that the rate law shown in equation (4) operates in the reactions of [Fe4S4Cl4]22, a limit for the value of KNa C (the binding constant of Na1 to [Fe4S4Cl4]22) can be calculated. Since, there is no evidence for the binding of Na1 even when [Na1] = 20.0 mmol dm23, KNa C < 5 dm3 mol21 (i.e.Na1 is bound to [Fe4S4Cl4]22 at least 100 times more weakly than it is to [Fe4S4(SPh)4]22). Combined eVect of Na1 and H1 on the lability of [Fe4S4(SPh)4]22 When [Na1] � 10 mmol dm23, all of the [Fe4S4(SPh)4]22 in solution has a Na1 bound to it, i.e. [{Fe4S4(SPh)4}Na]2. Under these conditions, the addition of [NHEt3]1 results in a further increase in the rate as shown in Fig. 3. Analysis of the kinetics shows that the reaction exhibits a non-linear dependence on, [NHEt3 1]/[NEt3], such that at low values of this ratio the rate exhibits a first order dependence on [NHEt3 1]/[NEt3], but is independent of the ratio at high values of [NHEt3 1]/ [NEt3].In additional experiments, [NHEt3 1]/[NEt3] was kept constant and the concentration of PhSH varied. Under these conditions the rate of the reaction does not depend on the concentration of PhSH. The rate law which fits these data is shown in equation (6). 2d[Fe4S4(SPh)4 22] dt = {0.30 ± 0.03 1 (0.38 ± 0.04)[NHEt3 1]/[NEt3]} 1 1 (0. ± 0.02)[NHEt3 1]/[NEt3] × [Fe4S4(SPh)4 22] (6) This rate law is identical to that observed with essentially every Fe–S-based cluster we have studied to date,2–8 the only diVerence is that in this case Na1 is additionally bound to the cluster. The mechanism is shown in Fig. 2. Rapid protonation of [{Fe4S4(SPh)4}Na]2 labilises the cluster towards dissociation of the Fe–SPh bond. The kinetics clearly demonstrate that H1 does not displace the bound Na1, otherwise kobs at high [NHEt3 1]/[NEt3] would correspond to the value observed in the presence of [NHEt3]1 and defined by equation (5).By considering all the pathways shown in Fig. 2, the general rate law shown in equation (7) can be derived, assuming that 2d[Fe4S4(SPh)4 22] dt = {k0 1 kNaKNa[Na1] 1 kNaHKNaKNaH[Na1][NHEt3 1]/[NEt3]} 1 1 KNa[Na1] 1 KNaKNaH[Na1][NHEt3 1]/[NEt3] × [Fe4S4(SPh)4 22] (7) binding of Na1 and H1 are rapidly established equilibria complete within the dead-time of the stopped-flow apparatus, and dissociation of the Fe–SPh bonds are the rate-limiting steps.Under conditions where [Na1] > 10 mmol dm23, KNa[Na1] @ 1 and equation (7) simplifies to equation (8). 2d[Fe4S4(SPh)4 22] dt = {k0 1 kNa 1 kNaHKNaH[NHEt3 1]/[NEt3]} 1 1 KNaH[NHEt3 1]/[NEt3] × [Fe4S4(SPh)4 22] (8) Comparison of equations (6) and (8) gives (k0 1 kNa) = 0.30 ± 0.03 s21, and using k0 = 1.5 ± 0.5 × 1022 s21 (the mean value of k0 derived from this and earlier studies), kNa = 0.28 ± 0.03 s21.This value is in good agreement with that derived from studies in the presence of only Na1 [equation (4)]; in addition, kNaH = 1.5 ± 0.2 s21 and KNaH = 0.25 ± 0.02. The question which must now be addressed is, “Where does this proton bind?” Using the value of KNaH = 0.25 ± 0.02 derived from these studies together with the pKa of [NHEt3]1 in MeCN (18.46),18 the pKa = 17.9 of [{Fe4S4(SPh)4}Na]2 can be calculated.This value is slightly smaller than for the parent [Fe4S4(SPh)4]22 (pKa = 18.6),5 consistent with the presence of the electron-withdrawing Na1 bound to the cluster. Most important the pKa associated with [{Fe4S4(SPh)4}Na]2 falls in the range 17.9 £ pKa £ 18.9, observed for all Fe–S-based clusters in MeCN.5 This is consistent with protonation occuring at the cluster core; most probably a m3-S. Since, the above analyses have yielded the values of kNaH and kNa, and earlier work2 gave kH we are in a position to discuss quantitatively the relative labilising eVects of Na1, H1 and the combined eVect of both Na1 and H1 on the cluster. From the studies with [NHEt3]1 alone [equation (5)], kH/ k0 = 8.4 ± 1.6, and in studies where only Na1 is added [equation (4)], kNa/k0 = 31 ± 6.The addition of both Na1 and H1 results in an increase in the rate [equation (8)], kNaH/k0 = 170 ± 20. This is close to the value which can be calculated using the simple equation, kNaH/k0 = (kNa/k0)(kH/k0) = 260 ± 50.That is, the labilising eVect of Na1 and H1 together is not appreciably diVerent from the product of the individual labilising components. We will see in the next section that similar behaviour is observed in the eVects of H1 and nucleophile on the lability of the cluster in an associative mechanism. EVect of H1 on the dissociative lability of [Fe4S4Cl4]22 Previous studies showed that the substitution reaction between [Fe4S4Cl4]22 and PhS2 occurs predominantly by an associativeJ.Chem. Soc., Dalton Trans., 1999, 119–125 123 mechanism, and protonation (by [NHEt3]1) accelerates the rate.4 Previously, the relative contributions to the labilisation of the cluster from binding H1 and PhSH could not be assessed. However, because Et2NCS2 2 is a poor nucleophile the substitution reaction with [Fe4S4Cl4]22 occurs exclusively by a dis- Fig. 4 Kinetics for the reaction between [Fe4S4Cl4]22 (0.1 mmol dm23) and Et2NCS2 2 in the presence of [NHEt3]1 in MeCN at 25.0 8C.Data shown: [NHEt3 1] = 10.0 mmol dm23 (d), [NHEt3 1] = 20.0 mmol dm23 (g); [Et2NCS2 2] = 2.0 mmol dm23, [NEt3] = 0.7–20 mmol dm23. Curve drawn is that defined by equation (9). sociative pathway. This permits a quantification of the eVect H1 alone has on the rate of dissociation of the chloro-group. Comparison with the earlier studies allows us to estimate the individual eVects that binding H1 and PhSH have on the labilisation of the chloro-group in the associative pathway.The addition of [NHEt3]1 to the reaction between [Fe4S4- Cl4]22 and Et2NCS2 2 leads to an increase in the rate of reaction as shown in Fig. 4. The rate of reaction exhibits a first order dependence on the concentration of [Fe4S4Cl4]22 (as indicated by the exponential shape of the absorbance–time curve) and the usual non-linear dependence on the ratio, [NHEt3 1]/[NEt3]. The rate law consistent with these data is shown in equation (9). 2d[Fe4S4Cl4 22] dt = {3.0 ± 0.5 1 (30.2 ± 0.2)[NHEt3 1]/[NEt3]} 1 1 (2.0 ± 0.2)[NHEt3 1]/[NEt3] × [Fe4S4Cl4 22] (9) This is consistent with the dissociative mechanism shown in Fig. 5. The rate law associated with this mechanism is shown in equation (10), assuming that protonation is a rapidly estab- 2d[Fe4S4Cl4 22] dt = {k0 C 1 kH CKH C[NHEt3 1]/[NEt3]} 1 1 KH C[NHEt3 1]/[NEt3] [Fe4S4Cl4 22] (10) lished equilibrium and dissociation of the chloro-group is ratelimiting. Comparison of equations (9) and (10) gives: k0 C = 3.0 ± 0.5 s21; kH C = 15.0 ± 1.0 s21 and KH C = 2.0 ± 0.2.The value of Fig. 5 Summary of the uncatalysed dissociative, and the acid-catalysed dissociative and associative substitution pathways for the reaction between [Fe4S4Cl4]22 and RSH (R = Et2NCS or Ph). Only one Cl ligand is shown for simplicity; d = Fe, s = S.124 J. Chem. Soc., Dalton Trans., 1999, 119–125 Table 1 Comparison of the individual and combined eVects of H1, Na1 and PhSH on the dissociation of the leaving group in the substitution reactions of [Fe4S4X4]22 (X = Cl or PhS) kAB/k0 Cluster [Fe4S4(SPh)4]22 A HH kA/k0 a 8.4 ± 1.6 8.4 ± 1.6 B H Na kB/k0 8.4 ± 1.6 31 ± 6 Obs. 41 ± 8 170 ± 20 Calc. 71 ± 13 260 ± 50 kAB/k0 C kAk0 C b kBk0 C Obs. Calc. [Fe4S4Cl4]22 H 5.0 ± 0.3 PhSH 83 ± 4 �250 415 ± 20 a k0 = (1.0 ± 0.2) × 1022 s21, for studies with [Fe4S4(SPh)4]22. b k0 C = 3.0 ± 0.5 s21, for studies with [Fe4S4Cl4]22. KH C is in good agreement with that determined in the earlier studies with PhSH (KH C = 2.2 ± 0.1).4 Earlier studies on the reaction between [Fe4S4Cl4]22 and PhSH in the presence of [NHEt3]1 showed that the mechanism involved rapid protonation of the cluster, followed by the binding of PhSH (KT), then rate-limiting cleavage of Fe–Cl (kTH)4 (Fig. 5). Analysis of the kinetics gave KTkTH = 1.5 × 104 dm3 mol21 s21. A limit to the value of KT can be estimated since, even at the highest concentration of PhSH used ([PhSH] = 5.0 mmol dm23), there is no kinetic evidence for the accumulation of appreciable amounts of the cluster with PhSH bound; hence (5.0 × 1023)KT £ 0.1, and KT £ 20 mol dm23; consequently kTH � 7.5 × 102 s21.The labilisation aVorded by binding H1 and PhSH is kTH/k0 C � 250. We are now in a position to estimate the individual labilising eVects of H1 (kH C/k0 C) and nucleophile (kT/k0 C). The studies with Et2NCS2 2, reported herein, show that protonation of the cluster core labilises the chloro-group to dissociation, kH C/ k0 C = 5.0 ± 0.3.Earlier studies 4 showed that the dissociation of the chloro-group after binding of PhS2 was associated with a rate constant, kT = 2.5 ± 0.1 × 102 s21. Consequently, labilisation aVorded by binding of PhS2 is kT/k0 C = 83 ± 4. Although, strictly, this is the labilisation aVorded by PhS2 rather than PhSH, currently it is the best we can do, and does at least give an estimate of the eVect of PhSH.Using these values we can calculate (kT/k0 C)(kH/k0 C) = 415 ± 20, consistent with the simple relationship, kTH/k0 C = (kT/k0 C)(kH/k0 C). Previously, in a study 5 on id-catalysed substitution reactions of the linear trinuclear cluster, [Cl2FeS2VS2FeCl2]32 we came to the conclusion that H1 alone was not particularly labilising but for maximum labilisation both protonation and binding a thiol was necessary. Herein, a detailed quantitative analysis confirms our earlier proposal.Labilisation by multiple components Throughout this paper we have emphasised that the labilising eVect of adding more than one reactant (H1, Na1 or nucleophile) to a cluster is not appreciably more labilising than that expected from the individual eVects of each contributor. This is born out by the summary of the results shown in Table 1, where we return to the generalised designations (A, B, C, etc. . . .) introduced in equation (1). Thus, two species A and B will aVect the lability of the leaving group by an amount described by the simple relationship shown in equation (11), where k0 is the rate constant associated with the uncatalysed reaction.kAB/k0 = (kA/k0)(kB/k0) (11) Close inspection of Table 1 reveals that our (perhaps oversimplistic) equations consistently over-emphasise the labilising power of the combination of several components. This is probably not too surprising considering the electronic origins of the eVects we are discussing.The strength, and hence lability, of Fe–Cl, Fe–SPh or Fe–SHPh bonds are defined by the s- and p-orbital overlap between ligand and Fe. The electron density distribution within these s- and p-orbital components is perturbed by the presence of H1, Na1 or nucleophile. It seems likely that in the presence of several components the electron distribution is distorted predominantly by one component such that the others do not have the eVect (when acting in concert) that they do when acting individually.However, it is clear that this is a rather minor eVect and that (at least in these systems) there is no cooperative labilising eVect from having several components present. Experimental All manipulations were routinely performed under an atmosphere of dinitrogen using Schlenk or syringe techniques as appropriate. [NBun 4]2[Fe4S4(SPh)4] 19 and [NBun 4]2[Fe4S4Cl4] 11 were prepared by the literature methods and characterised as described earlier.MeCN was dried by distillation from CaH2 under an atmosphere of dinitrogen. Na[BPh4] was purchased from Aldrich and used as received. NaS2CNEt2?3H2O (Aldrich) was recrystallised from methanol– diethyl ether and dried in vacuo. [NHEt3]BPh4 was prepared by the literature method.20 Preparation of [NBun 4]S2CNEt2 [NBun 4]Br (2.9 g, 8.9 mmol) was added to a solution of NaS2CNEt2?3H2O (2.0 g, 8.9 mmol) in methanol (ca. 50 mL), and the solution stirred for 30 min.The solvent was then removed in vacuo to leave a pale yellow solid. MeCN (ca. 20 mL) was added to the solid and after stirring for 30 min the mixture was filtered through Celite to remove NaBr. Diethyl ether (ca. 60 mL) was added to the clear filtrate which went cloudy (a further small amount of NaBr). The mixture was again filtered through Celite, then addition of a large excess of diethyl ether (ca. 200 mL) to the clear solution produced no further cloudiness. The solution was cooled to 220 8C overnight to produce fine, pale yellow needles of the product, which was removed by filtration, washed with diethyl ether and then dried in vacuo.Kinetic studies The reactions were studied on a Hi-Tech Stopped-Flow apparatus modified to handle air-sensitive solutions.21 The temperature was maintained at 25.0 8C using a Grant LE8 thermostat tank. The spectrophotometer is interfaced to a Viglen computer via an analogue-to-digital convertor. All solutions were prepared immediately prior to study and used within 1 h.J.Chem. Soc., Dalton Trans., 1999, 119–125 125 Under all conditions the reactions exhibited exponential absorbance–time curves which were fitted using a computer program. The dependence on the concentration of other reagents was established using conventional graphical methods as presented in Results and discussion. Acknowledgements The BBSRC is acknowledged for supporting this work. References 1 R. G. Wilkins, Kinetics and Mechanism of Reactions of Transition Metal Complexes, VCH, Weinheim, 2nd edn., 1991, ch. 1 and 4. 2 R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Dalton Trans., 1993, 1467. 3 R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Dalton Trans., 1993, 1473. 4 R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Chem. Commun., 1994, 377. 5 K. L. C. Grönberg and R. A. Henderson, J. Chem. Soc., Dalton Trans., 1996, 3667. 6 R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Dalton Trans., 1998, 1731. 7 V. R. Almeida, C. A. Gormal, K. L. C. Grönberg, R. A. Henderson, K. E. Oglieve and B. E. Smith, Inorg. Chim. Acta, submitted. 8 K. L. C. Grönberg, R. A. Henderson and K. E. Oglieve, J. Chem. Soc., Dalton Trans., 1998, 3093. 9 D. Sellmann and J. Sutter, Acc. Chem. Res., 1997, 30, 460, and refs. therein. 10 B. A. Averill, T. Herskovitz, R. H. Holm and J. A. Ibers, J. Am. Chem. Soc., 1973, 95, 3523. 11 G. B. Wong, M. A. Bobrik and R. H. Holm, Inorg. Chem., 1978, 17, 578. 12 M. G. Kanatzidis, D. Coucouvanis, A. Simopoulos, A. Kostikas and V. Papaefthymiou, J. Am. Chem. Soc., 1985, 107, 4925. 13 Ref. 1, p. 24. 14 H. Strasdeit, B. Krebs and G. Henkel, Inorg. Chem., 1984, 23, 1816. 15 J.-F. You, B. S. Snyder, G. C. Papefthymiou and R. H. Holm, J. Am. Chem. Soc., 1990, 112, 1067. 16 J.-F. You, G. C. Papefthymiou and R. H. Holm, J. Am. Chem. Soc., 1992, 114, 2697. 17 J. Huang, C. Goh and R. H. Holm, Inorg. Chem., 1997, 36, 356. 18 K. Izutsu, Acid–Base Dissociation Constants in Dipolar Aprotic Solvents, Blackwell Scientific, Oxford, 1990, p. 17. 19 B. V. Pamphilis, B. A. Averill, T. Herskovitz, L. Que, jun. and R. H. Holm, J. Am. Chem. Soc., 1974, 96, 4159. 20 J. R. Dilworth, R. A. Henderson, P. Dahlstrom, T. Nicholson and J. A. Zubieta, J. Chem. Soc., Dalton Trans., 1987, 529. 21 R. A. Henderson, J. Chem. Soc., Dalton Trans., 1982, 917. Paper 8/07769J