摘要:
1973Redistribution Reactions of Some Transition-metal Chelates. Part 1.Thermodynamics of Bidentate Ligand Exchange between Nickel(ii)C he1 a tesBy J. C. Lockhart and W. J. Mossop, Department of Inorganic Chemistry, The University, Newcastle uponThe redistribution reaction between chelates NiLl, and NiL2, to give NiL1L2 where L1 and L2 are bidentate salicyl-aldimine or P-keto-imine residues has been monitored by use of contact shifts in the l H or 19F n.m.r. spectra of thechelates. Equilibrium constants, enthalpies, and entropies for the redistribution are presented for several systems.The mixed compound is in general formed in less than statistical amount, enthalpies of redistribution then usuallybeing positive. This unusual result i s discussed in relation to previous work.In a number of systems no reactionwas observed. The redistribution between NiLl, and ML2, (where M is Cu, Zn, or Co) has also been examined.These reactions have shed light on the reason for non-reactivity of the bis- (N-n-propylsalicyaldiminato)nickel(li)-bis- (N-t- butylsalicylaldiminato) nickel (11) system.Tyne NE1 7RUTHIS series will be concerned mainly with studies ofequilibria such as (1), where L1 # L2 are bidentateligands. Most of the compounds used were of types(1) and (2). Structural information for these compoundsin solution is ~lentiful.l-~ A most useful property is theparamagnetism which gives rise to contact shifts in then.m.r. spectra of complexes (1) and (2) (where these haveany tetrahedral character), Whereas the n.m.r.spectrumof the diamagnetic ligand covers a range of 10 p.p.m.the spectrum of the co-ordinated ligand may cover arange of 250 p.p.m., effectively providing a non-linearlyexpanded spectrum of the ligand. Thus quantitativeanalysis of the reactants and products in equation (1)NiL1, + NiL2, =+ 2NiL1L2 (1)[(NiLl,, NiL2, = (1) or (2)] which would be difficult indiamagnetic systems can be achieved by use of n.m.r.signal areas as a measure of c~ncentration.~ We haveused this method of analysis to determine the kineticsand thermodynamics of a number of redistribution re-actions6 such as (1) and some thermodynamic resultsare discussed.The redistribution shown in equation (2) has also beenexamined where M is Co, Cu, or Zn. A third equilibrium(3) together with equations (1) and (2) is necessary toKZ NiLl, + ML2, + NiL2, + MLl,describe the mixed-metal reaction completely.Al-though there are six components in the mixture we wereML1, + ML2, === 2ML1L2able to monitor only the nickel compounds by contactshift spectra. The presence of other components wasR. H. Holm, G. W. Everett, jun., and A. Chakravorty,Progr. Inorg. Chem., 1966, 7, 83.L. Sacconi, P. Paoletti, and M. Ciampolini, J . Amer. Chem.SOC., 1963, 85, 411.A preliminary account has a~peared.~(2)(3)K ,inferred from mass-spectral analyses and materialbalance or (for diamagnetic zinc compounds) unshiftedn.m.r. signals.EXPERIMENTALPrepavation of Complexes.--Xll the complexes wereprepared by previously published procedures, the P-keto-imine derivatives by a non-aqueous chelation pr~cedure,~and the salicylaldimines by the reaction of preformedmetal salicylaldehydate with amine.1,Redistribution Studies.-To a weighed mixture of twosolid reactants was added a known volume of solvent froman Agla micrometer syringe. After dissolution the re-action mixture was transferred to an n.m.r.tube andcooled to -78 K. The tube was sealed and transferred tothe n.m.r. spectrometer, adjusted t o the appropriatetemperature. Spectra were recorded as soon as possible.For slow reactions (which took up to 30 h to reach equili-brium) spectra were recorded at intervals until equilibriumwas reached. For fast reactions, which reached equilibriumbefore the first spectrum was recorded, spectra weremeasured at a minimum of five temperatures in the range290-380 K.Normally two or three ' equilibrium ' spectrawere measured at each temperature. All n.m.r. spectrawere recorded in tetrachloroethylene or deuteriochloroform,relative to Me,Si or CFCl,, with a Bruker SpectrospinHFK-6 instrument. N.ni.r. signal areas were integratedwith the aid of a Dupont 310 Curve Analyser. Preliminarystudies showed that signal area could be used as a measureof concentration for the NiLl, and NiL2, complexes and thesame is assumed for NiL1L2 complexes which cannot ingeneral be isolated. Reactions were normally run a tseveral different initial concentrations and with equimolarratios of reagent. The concentration range (ca.10-1-10-2~)was limited on the one hand by the solubility of the reagentsand on the other by the resolution possible with the n.m.r.instrument.RESULTSThree categories of equation (1) were studied, arbitrarilydistinguished thus: (i) fast reactions (complete in less than5 min), (ii) slow reactions (for which equilibrium wasreached in longer periods), and (iii) systems for which noreaction could be detected by the n.m.r. technique described.3 G. W. Everett, jun., and R. H. Holm, J . Amer. Chem. Soc.,1965, 88, 2117. * R. H. Holm, Accounts Chem. Res., 1969, 2, 307.5 J. C. Lockhart and W. J. Mossop, Chem. Fomm., 1971, 63.J. C. Lockhart, ' Redistribution Reactions, Academic Press,New York, 197020 J.C.S. DaltonThe thermodynamic parameters for equilibrium (1) in bothfast and slow systems (i) and (ii) are in Table 1.Values of the equilibrium quotient K , = [NiLILa] "I-("iL1.J [NiL2,]) were determined at several differenttemperatures and the values quoted (all a t 298 K) wereobtained graphically from a ' least-squares ' treatment ofthe plot of log K against the reciprocal of the temperature.From the slope of this plot values of the enthalpy (Table 1)were obtained.The values of AS were obtained from theused to monitor the reaction which took ca. 20 min to cometo completion. The systems can be characterised by threeindependent equilibrium constants K,, K,, and K,. It ispossible to determine K , for the all-nickel equilibrium[equation (l)] and the values of K a ee quite well withvalues for the system (1) in isol$ioE For the nickel-copper and nickel-cobalt systems it was not possible todetermine the areas of the signals due to copper or cobaltcompounds since the signals were very broad.The zincTABLE 1Thermodynamic parameters for equation (1) in C&1, solution (unless otherwise noted)NiL1,' R1 x or ~ 2 '(1) But(1) But(2) Pri(1) But(1) Et,CH(1) But(1) Pri(1) Pri(1) But(1) But(1) But(1) PhHH3!HHHHHHHHHHNiLZ2r c\ Kl *R1 X or R2 (298 K)(2) Pri(2) Pri(2) Pri(2) Pri(2) Pri(1) Pri(1) Prn(1) Pri(2) Pri(2) Pri(1) Et,CH(1) Pri(2) Pr'(1) PhMePhMeMePhHH4MeMeMeH3Me0HMe138.243.63.61.441.321.20.820.800.620.160-130.087AH A SqTzF J mol-l K-1-2.8 3-2.2 f 2-0.1 f 0.5l f 3-1 f 35-45 f 218 & 95.2 f 211 f 1.613 f 35.5 f 115 f 228 f 43 8 f 112 & 9 l o & l o b11.2 f 2 =,=13 f.921 f 762 f 2519 f 1535 f 1542 & 914.5 f 1038 f 1777 & 1667 f . 9 :107 f 3S dSSSSFFFSSFFFF* These values are calculated from total concentrations of a particular compound in all its stereochemical forms present in solution(Le., square planar, tetrahedral, and associated). Only for the first four reactions in the Table does this represent equilibriumbetween three compounds of the same (tetrahedral) geometry. In principle two values of K , (K, tetrahedral and K, planar)could be obtained for each of the other systems but in practice i t is not possible to establish the percentage of tetrahedral geometrypresent in the mixed compounds since these cannot be isolated directly.Data obtained at two temperaturesonly.CDC1, solution.S, slow, F, fast reaction.second law. The errors in peak measurement are of theorder &tax and lead to large errors in the value of K whichlie within the limits f20%. The errors in AH and ASquoted in Table 1 were obtained from the least-squaresanalysis.Table 2 lists systems which gave no detectable reaction.TABLE 2Mixtures in which no reaction according to equation (1)was observed after several days at room temperature aR1 X o r R 2 R1 X or R2NiL21 NiL2aPri MePr* CF, b(1) Prn H (2)Pr* Ph(1) Prn H (2)Pri CF, C(2) Pr* Me (2)But H(1) But H (2)(1) Bun H (1)(1) But H (1) MeO*[CH,],* HPrn HPrn HH (3 dl ;: Me (3 d,0 No reaction after heating to 360 K for several hours.Thesereactions however occurred in the mass spectrometer, sinceparent ions corresponding to NiLlL2 were observed for everymixture. b Addition of pyridine (which changes the geometryof the nickel complexes) caused this reaction to proceed rapidly.c Addition of free ligand causes reaction. d (3) = Squareplanar nickel chelate of N- (n-propyl) -o-hydroxyacetophenone-imine.Mixed-metal Reactions.-The reactions were investigatedin CDCl, solution by the method outlined above and weretoo fast for kinetic studies. In some cases an approach toequilibrium could be followed (e.g., see Figure).In thisreaction the t-butyl group of N-t-butylsalicylaldiminato-nickel compounds gave suitable lH n.m.r. signals and wascomplexes are diamagnetic and byconcentrations of each ligand on zincresolution was insufficient for K3 to90 -c.5). 3II wm.-50-v-0s\CC 0 umaterial balance thecould be found. Thebe determined, and aI8 161010Time/ minPlot of concentration against time for NiL, = (1; R = But) insystem A , bis-(N-t-butylsalicylaldiminato)nickel(II) and bis-(N-n-propylsalicylaldiminato)zinc(II) in CDC1, a t 303 Kvalue of 4 (statistical value) was assumed in order tocalculate K,, the intersystem equilibrium constant. Table 3contains the equilibrium constants1973 21A qualitative order of reaction rates was observed andthis was D and E > A > C FZ B (see Table 3 for labelling).TABLE 3Equilibrium constants obtained from studies of mixed-metal systems [equation (2)]; NIL, = (1; R = But,X = H)(1) = ML2,M R X Kl K2 TIKA Zn Prn H 6.9 x 10-3 0.6" 3032.6 x 332B Zn Pri H 1.9C c u Pri H 1.5 303D c o Pri H 2.0 303E Ni Pr' H 1.5 303This was estimated assuming the zinc exchange to berandom with K , = 4.Mass Spectra.-The presence of all components inequations (1)-( 3) was demonstrated in mass spectraof mixtures.The mixture of bis- (N-n-propylsalicyl-aldiminato) zinc and bis- (N-t-but ylsalicylaldiminato) nickelhad as major parent ions the products of equation (2) , butthe mixed species [equations (1) and (3)] and startingmaterials of equation (2) were in very small abundance.Although ion abundance cannot be taken as a directmeasure of concentration, these observations are in accordwith a highly non-random reaction, suggested also by thesolution n.m.r.studies.U.V. Spectra.-The cobalt-nickel mixed metal systemwas examined by U.V. spectroscopy. Although there wasa change in the spectrum, overlap of peaks preventedquantitative analysis.DISCUSSIONTable 1 shows a variation of lo2 in the values of KB8for individual reactions. All but three of the K298values are less than the statistical value K, = 4. Wediscuss these low values first. The correspondingenthalpies are in general positive and unfavourable tothe mixed NiL1L2 compound.The entropy gain forequation (1) from statistical considerations should be+11.6 J K-l but the experimental results are in generalmore positive than this value, which tends to stabiliseNiL1L2 opposing the effect of the positive enthalpy. Itis exceptional to find a system in which the mixedcompound is not favoured. Only a few instances havebeen re~orded.~jVery few data have been available concerning ex-change of bidentate ligands. Holm and his co-workerssupposed the exchange of bidentate salicylaldimines onnickel to be randomJ8 while the exchange of dithiolateson nickel gave more product than expected on statisticalconsideration^.^ Exchange of p-diketonate ligands hasbeen studied on aluminium lo and gallium l1 (six-co-ordinate) and zirconium 12913 and hafnium (eight-co-tmhd = tetramethylheptane-dionate; hfac = hexafluoroacetylacetonate; and tfac = tri-fluoroacetylacetonate.K. Moedritzer, Adv.Organometallic Chem., 1968, 6, 171.A. Chakravorty and R. H. Holm, J . Amer. Chem. SOL,A. Davison, J. A. McCleverty, E. T. Shawl, and E. J.* acac = Acetylacetonate ;1964, 86, 3999.Wharton, J . Amer. Chem. Soc., 1967, 89, 830.ordinate) .13 In the six-co-ordinate systems reactionsinvolving two p-diketonates with simple alkyl sub-stituents (e.g., acac,* tmhd *) were approximatelyrandom but if one ligand contained fluorinated alkylgroups (hfac) exchange was very strongly in favour ofthe mixed compound. The enthalpyll in the reaction[equation (4)] was -17.2 & 3 kJ mol-l and the entropy0.4 & 10 J K-1.In contrast the eight-co-ordinatezirconium and hafnium compounds M(acac), andQGa(acac), + tGa(hfac), =+ Ga(hfac),(acac) (4)M(tfac), gave slightly more than the statistical amountof mixed compounds: the enthalpies for the Zr systemswere found to be nearly zero, but the entropies weregreater than expected.12 The co-ordination number ofthe central atom seems to be important in determiningthe enthalpy of these processes. Exothermic reactionsare the usual pattern for non-random redistributionsand can be interpreted in terms of increased bondenergies in the mixed compounds (a synergic effect).In the six-co-ordinate compounds this reinforced(synergic) bonding occurs when two different ligands(acac and hfac) are bound to the same central atom, butsynergism is not evident in the eight-co-ordinate com-pounds.There has been no comparable study ofasymmetric bidentate ligands such as the systemsexamined here. Our finding of unf avourable enthalpiessuggests that far from being reinforced, bonds arerelatively weaker in the mixed nickel salicylaldiminecomplexes. Since the ligands are asymmetric O,Ndonors, the reinforced bonding associated with thepresence of ligands of different bonding capabilities mayalready operate in the starting materials (1) (NiL12),and be destroyed in the mixed compound NiL1L2. Thisfinding is extremely important, since the constructionof mixed-ligand catalyst systems of enhanced reactivityis a current synthetic aim.l4.l5 Extensive redistributionstudies for a variety of multidentate ligands and centralatoms of different stereochemistry are needed.Atpresent, trends in redistribution chemistry are docu-mented mostly for monofunctional substituents on MainGroup There has been a somewhat differentapproach to the problem of mixed-ligand complexes,which can however yield information similar to thatfrom redistribution studies.14 Extensive studies offormation constants for mixed-ligand complexes ofcopper indicate reinforced bonding.14J5E$ect of Szcbstitzttiort.-Reactions within the salicyl-aldimine series [equation (l), NiLl,, NiL1, = (l)] showonly a tenfold variation in Kl [200-fold if the reaction inTable 3 of (1; R = Prn) is included].Substituents Xon the aryl ring of (1) and substituents R on nitrogenhave significant effects. If NiL, = (1; R = But) islo J . J. Fortman and R. E. Sievers, Inorg. Chem., 1967, 6, 2022.l1 T. J. Pinnavaia and S. 0. Nweke, Inorg. Chem., 1969, 8,l2 T. J. Pinnavaia and R. C. Fay, Inorg. Chew., 1966, 5, 233.l3 A. C. Adams and E. M. Larsen, Inorg. Chem., 1966, 5, 228.l4 R. Griesser and H. Sigel, Inorg. Chem., 1971, 10, 2229.l5 H . Sigel and D. B. McCormick, Accounts Chem. Res., 1970,639.3, 20122 J.C.S. Daltontaken as reference compound, exchange deviates mostfrom random for NiL2, = (1 ; R = Ph or Prn). We canalso compare the reactions of a set of complexes withNIL, = (2; R1 = Me, R2 = Pri) as reference. Fullytetrahedral complexes such as (1; R = But) and (2;R2 = CF,, R1 = Pri) gave random or greater thanrandom values of K and reaction (1) is exothermic.For other salicylaldimines, the reaction is endothermic,most particularly for R = Ph.The variation of substituents R is known to have acritical effect on the geometry a t the nickel centre.l**Substituents X affect the geometry most when in the3-position of the aryl ring.It is clear that the Rsubstituents will to a first approximation determine thegeometry of the mixed complex NiL1L2. Where thegeometries of the starting compounds NiL1, and NiL1L2differ, the largest deviations from random behaviourare to be expected. The reactions which were found togive the greatest deviations were those (i) of squareplanar (1 ; R = Prn) with fully tetrahedral (1 ; R = But)to give a mixed complex with partial tetrahedralcharacter, and (ii) of associated (octahedral?) (1; R =Ph) with fully tetrahedral (1; R = But) or (2; R1 =Pri, R2 = CF,).The first three reactions in Table 1with more f avourable enthalpies are between compoundswhich are fully tetrahedral.Nickel(I1) in four-co-ordination has Lewis acidtendencies which are satisfied in two ways: (i) self-association via bridging ligands to increase the co-ordination number and (ii) x-bonding internally in thefour-co-ordinate monomer. Series (1) has a strongertendency for association, (2) for x-bonding. Contactshift measurements suggest that series (2; Rz = Me,R1 = Pri) are more strongly x-bonding by a factor ofcn.1.5 than series (1; R = Pri).3 We suggest tenta-tively that ligands giving rise to rc-bonding tetrahedralmonomers can take part more effectively in synergicbonding. However when a ligand with associatingtendencies is exchanged with a x-bonding ligand themixed complex is less able either to associate or tox-bond and is thus unstable with respect to the startingcomplexes.Mixed-metal Reactions.-One of the most obviousfeatures of the nickel redistributions noted by Holm andhis co-workers is the failure of certain square-planarnickel complexes to react. We have noted (Table 2) anumber of such systems which did not react even a t hightemperatures. The reason for Holm’s observation(which we have confirmed) that N-n-propylsalicyl-aldiminatonickel(I1) and N-t-butylsalicylaldiminato-nickel(I1) do not react became immediately clear fromthe reaction in the zinc-nickel system A (Table 3).Tetrahedral (1; R = But) does not appear to react withsquare planar (1; R = Prn) because the system isalready in its most stable form. When the ligandN-n-propylsalicylaldimine is introduced [equation (3)]as its tetrahedral zinc chelate, the most stable nickelchelate (1; R = Prn) is formed. It appears that thetwo ligands (R = But, R = Prn) are about equallystable on zinc. The equilibrium constant for equation(1) has been obtained from the zinc-nickel system. Theintersystem constant could be obtained from materialbalance if the assumption was made that K3 = 4 for thezinc system [equation (3)].We thank the S.R.C. for a research studentship (toW. J. M.), Dr. A. K. Covington for the use of a DupontCurve Resolver, Mr. P. Kelly for mass spectra, and Mr.J. A. W. Akitt for discussion of n.m.r. spectra.[2/1048 Received, 10th May, 1972
ISSN:1477-9226
DOI:10.1039/DT9730000019
出版商:RSC
年代:1973
数据来源: RSC