Rate constants and equilibria of monochelate formation of iron(III) with 2-acetylcyclohexanone Carlos A. Blanco* and Antonio Rojas Departamento de Facultad de Ciencias, Universidad de V alladolid, 47005 Quïç mica Fïç sica, V alladolid, Spain Kinetics and equilibria in aqueous 1 : 1 chelation of iron(III) by 2-acetylcyclohexanone have been investigated spectrophotometrically in aqueous solution at 25 °C and 0.5 mol dm~3 In conjunction with earlier work, the results suggest NaClO4 .that the deprotonation rate of the keto tautomer of 2-acetylcyclohexanone has a minimum value when compared with 1,3- dicarbonylic ligands of similar structure. The mechanism proposed to account for the kinetic data involves pathways in which both Fe3` and Fe(OH)2` react with the enol tautomer of the ligand. The relative rates of complex formation are shown to depend not only on the metal species involved but also on ligand structure factors such as steric hindrances, ring strain and intramolecular hydrogen bonding.Metal chelates of 1,3-dicarbonylic ligands are important in several areas of applied chemistry.1 Although kinetic studies into the formation of coordination complexes of metal ions with acyclic 1,3-diketones have been thoroughly investigated, 2h4 very few kinetic studies of alicyclic systems have been carried out. The kinetic behaviour of 2-acetylcyclokanones towards metal ions depends on factors such as the reaction media (solvent, pH, etc.), the nature of the metal ion (hard»soft properties, lability, hydrolytic behaviour, .. .) and of course the nature of the ligand (keto»enol ratios, substituents, ring size, steric eÜects). It is well known that 2- acetylcycloalkanones form enol isomers and that the extent of enolization is controlled in the main by the strong intramolecular hydrogen bond formed, as well as the concentration, solvents and temperature.5 Moreover, tautomeric equilibria of these ligands depend upon the nature of the ring considered, i.e.replacement of a –ve- by a six-membered ring gives a more stable endocyclic enolic form. Furthermore, intramolecular proton transfer reactions in these ligands are particularly attractive, even from a theoretical point of view,6 which may allow quantitative comparisons with experimental data. Thus, a major focus of our studies has been an attempt to understand the factors that aÜect the rate constants of these chelation reactions.A kinetic study for the reaction of 2- acetylcyclohexanone (Hachx) with iron(III) to form metalenolate complexes is now reported. Apart from the intrinsic interest of the ligand, it has been discovered that the corrosion performance of single-phase glycol»water compositions containing the usual inhibitors can be improved by the addition of an eÜective amount of 2-acetylcyclohexanone.7 Furthermore, the fact that cast iron is frequently involved in these corrosive processes adds greater interest to this iron complexation study.Experimental Reagent-grade 2-acetylcyclohexanone (Aldrich) was freshly distilled under reduced pressure. Stock solutions were standardized by titration with standard sodium hydroxide. Solutions of iron(III) were prepared from reagent grade (Merck) and were standardized by gravimet- Fe(NO3)3 … 9H2O ric analysis, by weighing Fe2O3 . Perchloric acid (Merck) was used as a source of hydrogen ions. The ionic strength of the reactant solutions was adjusted to 0.5 mol dm~3 using (Merck). NaClO4 …H2O Unless otherwise indicated, all materials were of reagent grade (Merck) and were used without further puri–cation.pH measurements were made using a CRISON 2002 pHmeter calibrated to read hydrogen ion concentration directly by titrating solutions of perchloric acid with standard sodium hydroxide solutions. End-points were determined using the method of Johansson.8 UV»VIS spectra were recorded on a Spectronic 1201 UV»VIS spectrophotometer coupled to a BBC SE 790 chart recorder.Kinetic measurements were made either on the said spectrophotometer, which includes a resident kinetics program, or using a home-made stopped-—ow apparatus incorporating a tungsten lamp as the light source. Apart from some temperature-dependence measurements, the mixing chamber and reactions reservoirs were thermostatted at 25^0.1 °C. Results and Discussion In contrast to acyclic 1,3-diketones, alicyclic 1,3-diketones may exist simultaneously in endo- and exo-cyclic enol form, thus, the ligand can exist as the keto form (HK), the enolate ion (E~), and two enol forms (HE).Under our reaction conditions the dissociation equilibria are shown in Scheme 1. The interconversion of enolic structures does not require the dissociation of the intramolecular hydrogen bond and reduces to the migration of a proton between two oxygen atoms. Therefore it is very rapid. The lifetime of the tautomer is less than 10~4 s.All other tautomeric transformations in this system are, as a rule, slow. The ligand dissociation constant was evaluated potentiometrically by titrating solutions of the ligand having concentrations in the range 1.0]10~3»1.0]10~2 mol dm~3 with standard sodium hydroxide solutions in a double-walled titration cell under an atmosphere of nitrogen gas. Data treatment was carried out using the SUPERQUAD9 program. With this program the dissociation constant of the ligand was determined by minimization of an error-square sum based on measured electrode potentials.Since compounds of this class readily undergo cleavage in alkaline solution, it was thought desirable to check these results by the method of parallel J. Chem. Soc., Faraday T rans., 1998, 94(10), 1447»1450 1447Scheme 1 straights10 through spectrophotometric measurements. The ligand concentration used was 5.05]10~5 mol dm~3. Standard buÜer solutions were prepared in a pH range close to the pK of the ligand.The wavelengths were selected at regular intervals around the absorption maximum of the ligand (300 to 315 nm). The average dissociation constant at 25 °C KHL was found to be 5.49^0.05]10~11 mol dm~3. The composition of the complex that iron(III) forms with 2-acetylcyclohexanone in aqueous solution has been investigated spectrophotometrically by the methods of Job,11 and Yoe and Jones.12 Results obtained indicate a 1 : 1 stoichiometry.The monochelated complex formation may be structurally regarded as a hexacoordinated sphere where two molecules of water are replaced by two carbonyl oxygens from the 2- acetylcyclohexanone, simultaneous with loss of the proton from the tertiary ring carbon. The values of the equilibrium constant, KML\[ML][H`]/ [M][HL], were obtained spectrophotometrically. The absorbance change at 495 nm, where the complex absorbs strongly, is a function of the equilibrium constant, and the metal, ligand and proton concentrations : [HL]/A\[H`]/([Fe3`]eKML)]1/e (1) From a plot of the left-hand side of eqn.(1) vs. [H`]/ [Fe3`], may be readily determined. Several series of solu- KML tions were prepared, where the concentration of iron(III) ranged from (2.0»9.04)]10~3 mol dm~3, and [Fe3`][10[HL] with [HL]\4.0]10~4 mol dm~3. The value obtained for was 24.6^0.1, which agrees with the KML results obtained by using Schwarzenbach coefficients.13 The stability constant of the monocomplex, b\[ML]/[M][L], may be calculated from the relationship and b\KML/KHL , was found to be 4.5]1011. Kinetic measurements of complex formation have been carried out with the metal ion concentration greater than 20-fold the ligand concentration to ensure pseudo-–rst-order conditions.The range of hydrogen ion concentration (2.0]10~3 to 1.0]10~2 mol dm~3) was high enough to prevent the formation of hydrolysis products other than On the other hand the pH was kept high [Fe(H2O)5(OH)]2`. enough to ensure that complex formation was appreciable.Depending on the pH of the reactant solutions, the reactive ligand species can be either or both of the protonated tautomers and/or the enolate ion. Since these reactions have been carried out in an acidic medium where the pH is considerably less than the pK of the ligand, the fraction of the ligand present as the enolate ion is very small indeed and need not be considered. Pseudo-–rst-order constants were obtained by –tting ln data against time, where A and are the complex (Aa[A) Aa absorptions at time t and after completion of the reaction, respectively. Data for at least three half lives were utilised in these calculations.Each observed rate constant, subse- kobs , quently used for further calculation, is the average of at least three determinations. The standard deviation in individual runs was usually less than 2%. Several series of experiments were carried out to study the in—uence of the metal and proton concentrations on the observed rate constants, while the ligand concentration, ionic strength and temperature were kept constant.The dependence of on metal ion and hydrogen ion con- kobs centration has been studied. Fig. 1 shows a typical plot of kobs against iron(III) concentration at 25 °C. Results of complexation reactions for both ligands are consistent with a mechanism which can be represented by Scheme 2, considering that both and react with the enol Fe(H2O)63` Fe(H2O)5(OH)2` tautomer of the ligand, HE, to form the monocomplex (coordinated water is omitted for clarity).Fe3`]H2O A8B Kh Fe(OH)2`]H` HE]Fe3` A8B k~1 k1 FeE2`]H` HE]Fe(OH)2` A8B k~2 k2 FeE2`]H2O Scheme 2 Once the rate of formation of FeE2` is developed, it may be easily integrated and, considering the conditions used in this investigation : [HL]0\[FeE2`] (2) [Fe3`]\[FeIII][[Fe(OH)2`] \[FeIII][H`]/([H`]]Kh) (3) k1/k~1\KML/y; where y\[HE]/([HE]][HK]) (4) the pseudo-–rst-order rate constant derived, may be kobs , expressed as : kobs\ k1(H`]]k2Kh KML/y ] k1[H`]]k2Kh [H`]]Kh y[FeIII] (5) The hydrolysis constant of Fe3`, was Kh\1.86]10~3, taken from the literature.14 The ligand enolic fraction, y, was determined by a kinetic procedure.Since interconversion between endo- and exocyclic enol tautomers is very fast (Scheme 1), Hachx exists as Fig. 1 Observed rate constants as a function of the total concentration of iron(III) at 25 °C and ionic strength 0.5 mol dm~3, [Hachx]\1.0]10~4 mol dm~3 1448 J.Chem. Soc., Faraday T rans., 1998, V ol. 94equilibrium mixtures of the keto and enol forms, and the reaction scheme for proton transfer can be described by Scheme 3. HK A8B kb ka H`]E~ A8B kd kc HE Scheme 3 For a more useful kinetic treatment, Scheme 3 may be approximated without serious error to : HK A8B kf ke HE (I) A bromination procedure15 has been used to determine the value of [HE]/[HK], the ratio of the enolization rate and the ketonization rate.The enol tautomer reacts instantly with the bromine. To maintain equilibrium some of the keto tautomer converts to the enol tautomer, which then reacts with bromine. Hence, the rate of enolization parallels the rate of bromination, which can be easily determined. The ligand concentrations in the cell were in the range (1.5» 2.5)]10~3 mol dm~3, whereas bromine concentrations were (1.4»2.4)]10~3 mol dm~3.Changes in adsorption at 452 nm were recorded at 12 s intervals. Each of the bromine solutions contained 10~3 mol dm~3 NaBr in order to suppress hydrolysis of bromine. The ketonization rate constant, was determined graphi- kf , cally from the –rst-order bromination equation,15 and from acid hydrolysis of a solution of the ligand. In this latter case, the value of was determined by reacting a solution of the kf ligand, which had been adjusted to pH ca. 11 with sodium hydroxide, and which contained appreciable quantities of enolate ion, with excess perchloric acid.The enolate ion is rapidly protonated and the subsequent readjustment of the equilibrium to the equilibrium concentrations of keto and enol tautomers was monitored spectrophotometrically at 290 nm. A –rst-order plot of these absorbance changes gives (ke from which can be readily evaluated. ]kf), kf Thus, at 25 °C, values of and obtained were 5.9]10~4 ke kf s~1 and 1.37]10~3 s~1, from which the value of y was found to be 0.29.Eqn. (5) now only contains two unknown parameters, k1 and The kinetic data were –tted by a curve-–tting routine k2 . (NAG Fortran Library) to the derived rate equation of the mechanism. The rate constants obtained were k1\0.313 dm3 mol~1 s~1 and dm3 mol~1 ^0.006 k2\1.09^0.02 s~1. The goodness of –t estimated by the corrected residual sum of squares (less than 5%), clearly demonstrates that the data are well described by the former mechanism.Kinetic experiments have been carried out at several temperatures and, by using the corresponding thermodynamic and kinetic results, second-order rate constants may be similarly determined (Table 1). From the and values, which k1 k2 satisfy the Arrhenius law and Eyring equation, Gibbs energies of activation for both processes have been determined. The results obtained, kJ mol~1 and *G1º\78.5^0.1 *G2º\75.6 ^0.1 kJ mol~1, show that the reaction pathway through the Table 1 Rate constants (dm3 mol~1 s~1) for the formation of iron(III) monochelate of 2-acetylcyclohexanone T /K k1/10~2 k2/10~1 RSSa 283.15 7.43 3.05 4.31 285.15 8.66 3.65 3.23 288.15 11.6 4.63 4.61 293.15 16.6 6.47 3.75 298.15 31.3 10.9 4.26 a RSS\100 C; Wi(kobs[kcalc)2 ; W1(kcalc)2 D1@2 .monohydrolysed metal ion is kinetically favoured. In Scheme 3, and represent the KHK\ka/kb KHE\kd/kc dissociation constants for the respective reactions. Noting that [E~] represents the total concentration of enolate ion, the following relationships can be derived : 1/KHL\1/KHE]1/KHK (6) KHK/KHE\[HE]/[HK] (7) From these, and have been determined to be KHK KHE 1.89^0.05]10~10 and 7.73^0.05]10~11, respectively.The rate of proton loss from the keto tautomer, has ka\ke , been determined by the bromination procedure described above, and the value of was calculated from the kd\KHEkc known values of and assuming that the pathway is KHE kc , kc diÜusion controlled (ca. 3]1010 dm3 mol~1 s~1).16 The slow rate of ionization of the keto tautomer of Hachx compared with other 1,3-dicarbonylic compounds (Table 2) suggests that some special factor contributes to the formation of the enolate ion.The –rst two compounds in Table 2 are structurally similar alkyl, alkyl substituted b-diketones, the third and fourth compounds are tri—uoro derivatives in which the groups are electron withdrawing through strong wCF3 inductive eÜects. Finally, Hachx, which contains the 1,3- dicarbonylic system as part of a ring, does in fact have an abnormally low rate constant.This slow rate of proton loss from the keto tautomer may be ascribed to the fact that the 2-carbon on Hachx has only one proton attached compared to two on the methine carbon in linear b-diketones. In addition, the acidic proton is attached to a tertiary ring carbon and this may hinder the process of proton removal. According to Scheme 2, the Gibbs energy changes of both processes have been determined, kJ *G10\[11.0^0.1 mol~1 and kJ mol~1, at 25 °C, indicat- *G20\[26.5^0.1 ing that the pathway through the monohydrolysed metal ion is also energetically advantageous.In fact, hydroxide complexes of FeIII usually give larger rate constants than the aquometal ions for many substitution reactions ;3 in this particular case, the reactivity of Fe(OH)2` towards Hachx seems to be limited by some factors. Our results have established that the iron (III)»2-acetylcyclohexanonate complex has, in acidic solutions, an abnormally slow rate of formation compared with the rate constant of complexation of iron(III) with 2- aceylcyclopentanone, which is at least ten times higher.17 The slow rate of complexation of Hachx dm3 (k2\1.09 mol~1 s~1) may be justi–ed, taking into account the relatively slow rate of proton release from Hachx enol tautomer s~1) compared to s~1 in 2- (kd\5.67 kd\3150 acetylcyclopentanone (Hacpt).17 The factors which control the exo- and endo-enol ratios of 2-acetylcycloalkanones would in—uence the diÜerential rates of proton release. The relative stabilities of the two enol forms (endo- and exo-cyclic) are strongly in—uenced by the ring size and it is usually accepted that the exo-enol is more stable than the endo-enol for –ve-membered ring systems and vice versa for the six-membered ring systems.18 The slow rate of loss of proton from the predominate endo-enol tautomer in Hachx Table 2 Rates of ionization of the keto tautomer of 1,3-dicarbonylic compounds at 25 °C in aqueous solution ke/s ref. pentane-2,4-dione 1.52]10~2 27 heptane-3,5-dione 5.78]10~3 4 1,1,1-tri—uoro- 3.46]10~3 28 pentane-2,4-dione 4,4,4-tri—uoro-1-(2-thienyl)- 8.40]10~3 29 butane-1,3-dione 2-acetylcyclohexanone 5.90]10~4 this work J.Chem. Soc., Faraday T rans., 1998, V ol. 94 1449can be ascribed to steric hindrance compared to the reduced steric eÜects in the predominant exo-enol form in Hacpt.It is interesting to note that the extent of enol contribution to the structure of 2-acetylcyclohexanone is nearly double that of the cyclopentyl derivative,17 as could be expected.19 A second factor that is known to aÜect signi–cantly the rates of proton transfer may be attributed to an imbalanced transition state in which the development of resonance, solvation etc. lags behind proton transfer.20 This eÜect can be seen as a manifestation of the principle of non-perfect synchronization (PNS) which states that a product stabilizing factor whose development lags behind bond formation at the transition state depresses the subsequent rate constant.21 Another factor that may aÜect the rate constant of proton transfers is the capability of the transition state to become stabilized. The comparison of ligand-substitution rate constants seems consistent with dissociative activation, which means a bulky transition state, for substitution on and associative activation for Fe(H2O)5(OH)2`, This drastic mechanistic diÜerence between Fe(H2O)63`.22h26 both exchange paths may be due to the strong electrondonating capability of OH~.The strong bonding between the metal centre and this group will weaken the remaining metal» water bonds, most probably the trans solvent molecules. The complex thus becomes more labile and dissociative activation is favoured. There is also a clear relation between the acidity of a particular 2-acetylcycloalkanone (Hachx or Hacpt) and the properties of its iron(III) chelates.The more acidic ligand, Hacpt, gives a less stable complex (b\1.39]1010), whereas the less acidic, Hachx, gives a more stable complex (b\4.47]1011). This suggests that the major factor determining the properties of the iron complexes is electronic. As a general rule, it could be expected that a strongly acid diketone would give an anion which is a poor donor and transfers little charge to the metal atom; conversely a diketone with a high pK transfers considerable charge, with the result that the iron chelate is more stable.authors acknowledge the –nancial support of the Junta The de Castilla y Leon. References 1 M. A. Ribeiro Da Silva, NAT O ASI Ser., Ser. C., 1984, 119, 317. 2 R. C. Mehrota, R. Bohra and D. P. Gaur, Metal b-Diketonates and Allied Derivatives, Academic Press, London, 1978. 3 M. J. Hynes, Rev. Inorg. Chem., 1991, 11, 21. 4 C. Blanco and M. J. Hynes, Inorg.Chim. Acta, 1990, 173, 115. 5 N. Cir and L. W. Reeves, Can. J. Chem., 1965, 43, 3057. 6 D. Lee, C. K. Kim, B. S. Lee, I. Lee and B. C. Lee, J. Comput. Chem., 1997, 18, 56. 7 H. German (Dow Chemical Co.) US Pat., 4, 324, 676, 1982. 8 A. Johansson, Analyst (L ondon), 1970, 95, 535. 9 P. Gans, A. Sabatini and A. Vacca, J. Chem. Soc., Dalton T rans., 1985, 1195. 10 P. Maroni and J. P. Calmon, Bull. Soc. Chim., 1964, 519. 11 P. Job, Ann. Chim., 1928, 9, 13. 12 J. H. Yoe and A. L. Jones, Ind. Eng. Chem. Anal. Ed., 1944, 16, 11. 13 A. Ringbom, de complejos en Formacioç n Quïçmica Analïçtica, Alhambra, Madrid, 1979. 14 G. H. Khoe, P. L. Brown, R. N. Sylva and R. G. Robins, J. Chem. Soc., Dalton T rans., 1986, 1901. 15 M. L. Moyaç , A. Rodriguez, F. Sanchez, C. Blanco and M. J. Hynes, Int. J. Chem. Kinet., 1992, 24, 359. 16 M. Eigen and L. C. Maeyer, in T echniques of Organic Chemistry, ed. S. L. Friess, E. S. Lewis and A. Weissberger, Interscience, New York, 1963, vol. VIII, part II. 17 C. A. Blanco and J. Sumillera, New J. Chem., 1994, 18, 223. 18 K. A. K. Ebraheem, S. T. Hamdi and M. N. Khalaf, Can. J. Spectrosc., 1981, 26, 225. 19 W. M. Coleman and B. M. Gordon, Appl. Spectrosc., 1988, 42, 666. 20 C. F. Bernasconi, D. A. A. Ohlberg and M. W. Stronach, J. Org. Chem., 1991, 56, 3016. 21 C. F. Bernasconi, T etrahedron, 1985, 41, 3219; Acc. Chem. Res., 1987, 20, 301. 22 M. Grant and R. B. Jordan, Inorg. Chem., 1980, 20, 55. 23 A. E. Merbach, Pure Appl. Chem., 1982, 54, 1479. 24 F. K. Meyer, A. R. Monnerat, K. E. Newman and A. E. Merbach, Inorg. Chem., 1982, 21, 774. 25 A. E. Merbach, Pure Appl. Chem., 1987, 59, 161. 26 J. Burgess, Metal Ions in Solution, Wiley, Chichester, 1990. 27 D. P. Fay, A. R. Nichols and N. Sutin, Inorg. Chem., 1971, 10, 2096. 28 M. J. Hynes and B. D. OœRegan, J. Chem. Soc., Dalton T rans., 1980, 7. 29 M. R. JaÜe, D. P. Fay, M. Cefola and N. Sutin, J. Am. Chem. Soc., 1971, 93, 2878. Paper 8/00521D; Received 19th January, 1998 1450 J. Chem. Soc., Faraday T rans., 1998, V ol. 94