1972 2617Thermodynamics of the Interactions of Catechol with Transition Metals.Part 1 1 P The Effect of 4-Chloro- and 4-Nitro-substitution on Proton andMetal Catechol Complex FormationBy R. F. Jameson and M. F. Wilson, The Chemistry Department, The University, Dundee DD1 4HNThe effect on proton and metal complex formation of substitution of chloro- and nitro-groups in the catechol4-position has been investigated by potentiometric, calorimetric, and spectrophotometric methods with 0.1 OOM-KNOs as background electrolyte at 25 'C. Thermodynamic quantities for complex formation have been deter-mined and some interpretation of the effect of substitution on bonding has been made.PREVIOUS investigations in this laboratory of protonand metal complex formations with 4-substitutedcatechols have been concerned mainly with catechol-amine ligands having side-chains which are essentially' electron-donating ' sub~tituents.~?~ The ligandsadrenaline and L-Dopa are characteristic.We nowreport a study of the effect on complex formation ofelectron-withdrawing substituents in the catechol4-position.EXPERIMENTALApparatus and Techniques.-The potentiometric as-sembly, treatment and calibration of glass electrodes,and preparaion of silver-silver chloride electrodes werede~cribed.~ The LKB 8700 Precision Calorimeter wastested by means of the THAM reaction5 and heats ofneutralisation of the ligands were measured as in Part I.6The method for the determination of heats of neutralisationof metal-ligand mixtures was described in Part II 1 and abatch method was used since titration introduced a timelag during which chlorocatechol was oxidised.Nitro-catechol was stable to oxidation under these conditionsprovided that pcH was not above 9.5. Copper(I1) andnickel(r1) ions were titrated with the ligands with aligand : metal ratio of 2.5 : 1 as a precaution against theformation of hydroxo-complexes. Spectrophotometric de-terminations were made with a Unicam SP 700 spectro-photometer.4-Nitrocstechol (Puriss ; Ralph Emmanuel Ltd.) was98.27 yo pure by potentiometric titration with KOH.4-Chlorocatechol (Aldrich) was twice purified [charcoal inlight petroleum (b.p. 60-90 "C)} and recrystallised, and wasthen triply sublimed ; potentiometric titrations indicatedbetter than 99.90/, purity.The preparation of solutionsof copper and nickel and potassium hydroxide titrant wasdescribed in Part 1I.l AnalaR potassium nitrate wasused.RE S U LTSDissociation Constants of the Ligands.-For each liganda serious of titrations was carried out and the averagenumber of protons, g, abstracted from the ligand wascalculated by use of the previously determined value of theionic product of water.4 Dissociation constants werecalculated by the computer programme LETA-GROPVRID.7.8 Dissociation constants of the first andPart 11, R. F. Jameson and M. F. Wilson, preceding paper.2 R. F. Jameson and W. F. S. Neillie, J . Inorg. Nuclear Chem.,R. F. Jameson and J. E. Gorton, J . Chem. SOC. ( A ) , 1968,1965, 27, 2623.26 15.second proton ionisations of the ligands are in Table 1.The results are compared with corresponding values forcatechol determined in previous work.Stability Constants of C O ~ ~ M ( I I ) and Nickel (11) Comfilexes.-These were obtained from potentiometric titration dataand formation curves were also plotted revealing steps a tvalues of fi = 1 and 2 corresponding to the formation ofcomplexes of the type ML and ML,.By using a range ofmetal concentrations it was established that no polynuclearcomplex was formed. Stability constants were calculatedagain by use of the computer programme LETA-GROPVRID, the input data being the master variables ii(the average number of ligands bound per metal), pL(-log[free ligand]), and [MT] the total metal concentration.The output data were values of the overall stoicheiometricstability constants PIM and pzM and from these the corre-sponding values of log KIM and log K2M were found; stand-ard deviations were also computed and these are givenwith stability constant values in Table 2.TABLE 1Acid dissociation constants of the catechol ligands, to-gether with standard deviations, measured in anionic background of 0.100~-KNO, a t 25 "CLigand -log Kla -log K2aCatechol 9.195 f 0.002 12.98 f 0.01 *4-Chlorocatechol 8.522 f 0.002 11.974 & 0.0074-Nitrocatechol 6.701 f 0.002 10.853 f 0.004* Part I.Calorimetric Results for the Ionisation of the Ligands.-As in the case of catecho1,s a t pcH values as low as 8.8oxidation of 4-chlorocatechol occurred and interferedwith the determination of the enthalpy change for thesecond ionisation; hence only the first ionisation of theligand could be investigated by calorimetrv. Similarlysome decomposition of 4-nitrocatechol was detected, butonly a t high pcH values.Heats of neutralisation of the ligands, Qn, were determinedcalorimetrically by neutralisation (1) with potassiumhydroxide.Corrections were made for the heat of forma-tion oE water,6 Qw (which was dependent upon the concen-tration of the excess of acid in the calorimeter solution)and also for the heat of dilution of the added base Qdil.R. I;. Jameson and M. F. Wilson, J.C.S. Dalton, 1972, 2607.R. J. Irving and I. Wadso, Acta Chem. Scand., 1964,18, 195. * Part I, R.F. Jameson and M. F. Wilson, J.C.S. Dalton, 1972,L. G. SillBn, Acta Chem. Scand., 1964, 18, 1085.N. Ingri and L. G. SillCn, Arkiu Kemi, 1964, 23, 97.2610J.C.S. DaltonHence the uncorrected measured heat of reaction Qr, wasgiven by equation (2).The concentration of ionised ligand after reaction wasdetermined from the known pKIB value and the finalhydrogen-ion concentration ; hence the enthalpy change,TABLE 2Stoicheiometric stability constants of copper(i1) andnickel(ir) complexes of catechols a t 25 "C and ionicstrength 0- 1 0 0 ~ (KNO,) together with standarddeviationsA =Ligand log KIM log I q f log KIM - log K 2 MCopper( 11)Catechol 13.827 10.921 2.906 *4-Chlorocatechol 12.894 10.163 2.7314-Nitrocatechol 11.666 9.282 2.384& 0.002 & 0.003& 0.002 f 0.003& 0.004 5 0.005Nickel (11)Catechol 8.927 5.561 3.37 *4-Chlorocatechol 8.375 5.467 2.194-Nitrocatechol 7.887 5.504 2.38i 0 .0 1 2 &0*020i 0 . 0 1 4 i 0 . 0 1 80.005 f 0-008* Part 11.AH,, for the first proton neutralisation of the ligandswas calculated.the first protonResults for theAHioo are givenThe corresponding enthalpy change forionisation was given by expression (3).AHion = AH, - AH, ( 3)calorimetric determination of AH,, andin Table 3 together with standard de-TABLE 3Calorimetric results for the enthalpy changes of the firstproton neutralisation and ionisation of 4-chloro- and4-nitro-catechol a t 25 "C and ionic strength 0 . 1 0 0 ~(KNO,), showing standard deviationsConcentration and pcH ranges AH, AHton102[LT]/~~ 1O3[HL]/~ pcH kJ mol-l kJ mol-I4-Chlorocatechol3.000- 1~8i2- 7.319- -27.83 28-683.250 2.091 7.374 f0-40 & 0-404-Nitrocatechol0.65% 1.782- 6.228- -32.51 23.980.720 1.889 6.324 C0.18 &O*lSviations.Thermodynamic quantities for the first protonionisation of the ligands are summarised in Table 4.Calorimetric Results for Cofiper Complex Formation.-Calorimetric results were obtained for the formation of1 : 1 copper complexes alone since i t was not possible tomake similar determinations in the case of nickel becauseof interference by catalytic oxidation of the ligands.Heats of neutralisation of metal-ligand mixtures withpotassium hydroxide were measured with use of an excessof ligand to prevent metal-ion hydrolysis, and i t wasarranged that the final ionic strength was 0 .1 0 0 ~ withrespect to potassium nitrate; the proportions of metal,ligand, and added base were also adjusted so that only the1 : 1 complex was formed [reaction (4)]. CorrectionsH,L + Cu2+ + 20H-+ CuL + 2H,Oto the observed heats of reaction were made to allow forthe heat of formation of water and heat of dilution of theadded base. (The enthalpy change for the formation ofwater in O-~OOM-KNO, a t 25 "C has been determined.6) Thecorrection for the heat of formation of water was calculated(4)TABLE 4Thermodynamic quantities for the first proton ionisationof catechol ligands in 0.100~-KNO, at 25 "C, showingstandard deviationLigand -log Kla k J mol-l k J mol-l J K-l mol-lAG AH A S --Catechol 9.195 52.48 34.46 -60.264-Chlorocatechol 8.522 48.66 88.68 - 66.9410.002 f0.04 kO.09 f0.42 *,t0*002 f0.04 &O-40 j 1 - 2 6-C0-002 -- 40.04 k0.18 kO.814-Nitrocatechol 6.701 38-23 23.98 - 48.1 1* Part I.6by determining the change in hydrogen-ion concentrationduring the reaction, and the amount of complex formedwas determined from a knowledge of the metal-ligandstability constant, the ligand dissociation constant, andthe final pcH value.The method of calculation is givenin Part 11.1The enthalpy change for the neutralisation reaction,AH,, was calculated from the heat of neutralisation and theconcentration of 1 : 1 complex formed, (ML]. Resultsfor a series of determinations are in Table 5.TABLE 5Calorimetric results for copper(I1) complex formation with4-chloro- and 4-nitro-catechol a t 25 "C and ionicstrength 0- 1 0 0 ~ (KNO,), showing standard deviationsConcentration ranges AHn AH*1O3[RIT]/~ 103[LT]/hl 104[ML]/hr k J nlol-' k v4-Chlorocatechol2.480- 6.190- 4.165- -101.1 f 0.3 11.93 j 0.366.442 16.11 9.2954-Nitrocatechol1.982- 4.870- 3.822- -105.9 & 0.2 7.15 f 0.212.973 7.305 9.282Using previously determined thermodynamic datatogether with the values of AHn, we calculated enthalpychanges for other equilibria.First, using the value ofthe enthalpy change for the formation of water in the samemedium, from equation (4) we calculated the enthalpychange, AH*, for the equilibrium ( 5 ) , together with theCu2+ + H2L ---t CuL + 2H+ ( 5 )corresponding values of AG* and AS*.Thermodynamic quantities for the ionisation of theligands (Table 4) were also used to calculate correspondingthermodynamic values for equilibrium (6).Table 6Cu2+ + HL- + CuL + H+ (61972 2619summarises the thermodynamic quantities for these equili-bria and they are compared with corresponding valuesfor the 1 : 1 copper-catechol complex reported in Part 11.'TABLE 6Thermodynamic quantities for the formation of copper(I1)complexes with catechols in O-~OOM-KNO, a t 25 "Cshowing standard deviationsLigand log K kJ mol-l kJ mol-l J K-' mol-lAG AH A S --(a) Cu2+ + H,T, + CuL + 2HfCatechol -8.345 47-66 13.39 -115 It 2 *4-Chlorocatechol - 7.602 43.40 11.93 -105 f 14-Nitrocatechol - 5.888 33.60 7.15 -88.8 f 0.4kO.009 f0.04 -J=0*80f O * O l l f0.04 f0.36jO.010 f0.04 f0.21(b) Cu2+ + HL-+ CuL + H+Catechol 0.85 -4.85 -20.9 -54 f 2 *4-Chlorocatechol 0.92 -6.27 -16.7 -38 & 24-Xitrocatechol 0-81 -4.64 -16.7 -42 :k 2f0-01 f0.04 5 0 .8&O*Ol f0.04 +Om8+O.Ol 10.04 f0.4* Part 11.'Optical Stzcdzes.-In acidic aqueous solution the ligandscatechol and 4-chlorocatechol are found to have high-intensity U.V. spectra with well defined peaks at 34 kK;no absorption occurs for these ligands in the region below30 kK and hence the solutions are colourless. The ligand4-nitrocatechol absorbs in the visible region and has a welldefined peak at 28 kK.The visible spectra of some 1 : 1 copper(r1) complexes withsubstituted catechols are shown in Figure 1.In mostcases the spectra have two important characteristics :(a) a typical broad asymmetric d-d band a t 13.5 k ~ , oflow intensity (spin-forbidden), characteristic of the spectraof copper(I1) co-ordinated to four oxygen atoms, and(b) for most of the ligands shown, a high-intensity charge-transfer band whose peak varies between 22-5 kK forI I : \ Is / k KFIGURE 1 Spectra of copper(I1) complexes of catechols:(-), catechol; (...........), 4-chlorocatechol; (- - -),4-formylcatechol; (- . - - -), DopaL-Dopa to 26 kK for catechol, and whose position is clearlydependent upon the nature of the substituent. Thecopper(11)-4-nitrocatechol complex shows no charge-transferband, but with 4-formylcatechol a shoulder at 22 kKsuggests the presence of a charge-transfer band beneaththe ligand absorption.Similar charge-transfer spectrawere not observed in optical studies of the nickel complexes,which makes their instability in the presence of oxygenless readily explicable.DISCUSSIONThe effect of substituent on ionisation of the ligandscan be seen by considering the dissociation constants40 80 120-ASIJ K-' mol"FIGURE 2 Plots of AG against A S for phenolic-type ionisations:circles and full line, orlho-substituted phenols with valuesa t infinite dilution ; triangles and broken line, 4-substitutedcatechols measured in O-~OOM-KNO,; A,A' = catechol;B = 4-chlorocatechol; C = 4-nitrocatechol; D = o-propyl-phenol; E = o-methylphenol; F = o-methoxyphenol; G =phenol; H = o-formylphenol; I = o-chlorophenol; J =o-nitrophenolin Table 1.The pK (and hence AGion) decreases withincrease in electron-withdrawing power of the sub-stituent for both the first and second proton ionisationand therefore the acid strength decreases in the orderof substituent NO, > C1> H. The Born model ofelectrostatic solute-solvent interaction for weak acidsin solution predicts a linear relationship betweenAGione and ASione for acids of a similar structure; theproportionality between these thermodynamic quantitiesASione/AGione = ( 6 In D/GT)p (7)may be shown to be (7), where D is the dielectric con-stant, T the absolute temperature, and p the press~re.~Figure 2 shows a plot of AGione against ASione for theionisation of ortlzo-substituted phenols and a goodlinear relationship between these parameters is observed.The corresponding values of AGione against ASione forcatechol (first proton ionisation, obtained by extra-polation to infinite dilution 6) are plotted on the samediagram and it is seen that catechol, an ortho-substitutedphenol, does not correlate with the other ortho-compounds. One explanation is that the structure!a E.J. King, ' Acid-Base Equilibria,' Pergarnon Press,London, 19652620 J.C.S. Daltonvalues of AH indicate that for process (5) less energy isrequired as the ligand acidity increases. One possibleexplanation is that the bonding in the complexes is notsimply due to coulombic attractions between species ;however, since extensive hydrogen bonding is almostcertainly present in these species? solvation effectsare expected to have a marked influence and this isthe most likely explanation for the above effect; inother words it cannot be assumed that hydrationalenthalpies are constant from ligand to ligand.For equilibrium (6) values of AH are -20-9, -16.7,and -16-7 kJ mol-l for substituents H, C1, and NO,respectively and overall these results suggest that theeffects of removing the second ligand proton and formingthe 1: 1 complex compensate each other more thanin equilibrium (5).In this respect the second protonof the catechols appears to behave more like a sim2lephenolic proton, as suggested by the approximate valueof the enthalpy change (AH = 21 kJ mol-l) for thesecond proton ionisation of catechol determined inPart I.6Values of AH for the first proton ionisation of thecatechols (Table 4) indicate that the strength of bindingof the proton decreases with increasing acidity of theligand as measured by the pKla value (i.e., AH = 34.46,28.68, and 23.98 kJ mol-l for substituents H, C1, and NO,respectively).The strength of binding of the secondcatechol proton (as measured by AH values for theionisation) would also be expected to follow this order.Further, although it was not possible t o determinesuch AH values, provided that the binding of thesecond proton does follow this order, then by use of theAH values for equilibrium (6) given in Table 6, togetherwith arbitrary values of AH for the second protonionisation, it can be shown that for the equilibrium (8)cu2+ + L2- + CUL (8)the strength of bonding in the 1 : 1 copper complexes,as measured by the AH value for this reaction, decreasesin the order catechol > 4-chlorocatechol > 4-nitro-catechol. This indicates that the bonding present inthese complexes is simple o-bonding and that x-bondingby back-donation of &electrons from copper to ligandis of little significance.For oxidising metal ions and oxidisable ligandscharge-transfer bands can be observed correspondingto transitions in which the metal ion is reduced.Ingeneral for charge transfer from ligand to metal thefrequency is smaller the more oxidising the cation andthe more reducing the ligand.In the case of thecopper(I1) catechol complexes the charge transfer isalmost certainly from the ligand to the metal; thisfollows from the dependence of the frequency of thetransition upon the nature of the substituent in the4-position. Thus the energy of the band decreases inthe order catechol, 4-chlorocatecholJ L-Dopa and thisof catechol in aqueous solution may be considerablydifferent because of strong hydrogen bonding betweenadjacent hydroxyl group^.^ There is, however, somecorrelation between the values of AG and AS obtainedfor the first proton ionisation of the catechol ligands inin 0.100~-KNO,; this seems more plausible if onepostulates a different structure for catechols in solution.Table 2 shows that the stabilities of the copper(I1)complexes for both the 1 : 1 and 2 : 1 species decreasewith substituent in the order H > C1 > NO,, and forthe formation of the 1 : 1 nickel(11) complexes a similartrend is found.This is the order expected for complexesformed by simple o-bonding between ligand and metal,i.e., the acidity of the ligand increases with increasein electron-withdrawing power of the substituent andhence the base strength of the deprotonated ligandspecies is diminished. The Table also shows that forthe addition of a second ligand anion to the 1 : 1copper(I1) complex of that anion, the log formationconstant, log K2M, is generally between 2 and 3 unitslower than log KIM. This may be understood by con-sidering structures (I) and (11). The considerableincrease in stability of formation of complex (I) overthat of (11) is best explained in terms of the chargeeffects of the complex ions formed.lO Thus in the form-ation of complex (11) , coulombic repulsions betweenthe negative oxygen atoms would tend to lower thestability, and an increase in the relative stability of thisspecies with increase in electron-withdrawing powerof the substituent would also be expected if the effectof the substituent is transmitted through the ligandsuch that the electron density on the catechol oxygensis lowered.This effect can be seen more clearly byexamining the values of A (= log KIM - log givenin Table 2. For formation of the nickel(I1) complexesvalues of A do not fall uniformly, although the overalltrend is similar to that of copper.An examination of thermodynamic quantities for theformation of 1 : 1 copper(1r) complexes (Table 6) showsthat AH values for equilibrium (5) are positive anddecrease with increasing acidity of the ligand (i.e.,AH = 13.39, 11-93, and 7-15 kJ mol-1 for substituentsH, C1, and NO, respectively).In general, for a seriesof substituted bidentate ligands of similar structure,the effects of removing two ligand protons and formingcopper-ligand bonds would normally be expected tocompensate each other as we move along the series(provided that enthalpies of hydration are constantand that the strength of bonding is determined by thebasicity of the deprotonated ligand species alone) andhence AH for this process should remain relativelyunaffected by changing substituent. In this case the10 G. A. L‘Heureux and X. E. Martell, J . Inorg. Nucleav Chem.,1966, 28, 481; this paper reports an inexplicably low value ofpK,a for catechol (1 1.93)1972 2621is the same order as the increase in x-electron donatingpower of the 4-substituent. In the case of 4-nitro-catechol no charge-transfer band was detected and thisis to be expected since the nitro-group is a strongelectron-withdrawing substituent. Thus these resultsenable the catalytic oxidation of L-Dopa by molecularoxygen in the presence of iron(II1) to be rationalised,although it has been found necessary to postulate atwo-electron transfer in this case (probably as a rapidsuccession of two one electron-transfer steps) from theligand to metal to oxygen.ll This is probably partlydue to the role of the 0, molecule, however, since in theoxidation of ascorbic acid with copper(11) as catalyst, asimilar overall two-electron change seems to demandthe participation of two copper(I1) ions on the basis ofkinetic evidence.12[2/1000 Received, 4th May, 19721l3 R. F. Jameson, A. E. Martell, and (in pat) J. E. Gorton,l2 N. J. Blackburn and R. F. Jameson, unpublished work.unpublished work