King Khalid Military Academy, P.O. Box 22140, Riyadh 11490, Saudi Arabia. John Burgess* and Ali Shaker**. Department of Chemistry, University of Leicester, ...
Low-spin FeII Schi base complexes
Transition Met. Chem., 23, 689±691 (1998)
689
Attenuation of substituent eects on reactivities of low-spin iron(II) complexes of Schi base ligands Saad Alshehri King Khalid Military Academy, P.O. Box 22140, Riyadh 11490, Saudi Arabia John Burgess* and Ali Shaker** Department of Chemistry, University of Leicester, LE1 7RH, UK Summary Substituent eects on reactivity are compared for base hydrolysis of tris-ligand-iron(II) complexes of Schi base ligands derived from 2-acetylpyridine and substituted benzylamines and their aniline analogues. The methylene spacer in the former eectively isolates the iron from substituent eects in the phenyl ring, as indicated by the almost equal rate constants for base hydrolysis of the complexes derived from benzylamine and its 4-methyl, 4-¯uoro, and 4-tri¯uoromethyl derivatives. Rate constants for base hydrolysis of the (substituted) benzylamine derivatives are also reported for some Me2CO-H2O and DMSO-H2O solvent mixtures. Reaction is faster in all these binary aqueous mixtures, mainly due to the higher chemical potentials of hydroxide ion. Introduction Considerable eort has been expended over the past ®ve decades in establishing substituent eects on reactivity for substitution reactions ± dissociation (aquation), hydroxide attack (base hydrolysis), and reaction with cyanide ± of low-spin tris-ligand iron(II) complexes. The ®rst such reports related to substituted 1,10-phenanthroline ligands(1), with subsequent papers dealing with further work on a range of 1,10-phenanthroline complexes(2,3,4), on complexes of substituted 2,2¢-bipyridyls(5), on complexes of Schi base ligands derived from pyridine 2-carboxaldehyde(6) or 2-benzoyl pyridine(7), and other derivatives(8). Such substituent eects can be quite large, as for example in the case of aquation of [Fe(X-phen)3]2+, where going from the 5-nitro- to the 5,6-dimethyl derivative results in an almost ®fty-fold reduction in rate constant(2). There is a more dramatic substituent eect on stabilities, with log b3 for the [Fe(Xphen)3]2+ cations increasing from 15.6 to 23.0 on going from the 5-nitro(8) to the 5,6- or 4,7-dimethyl(9) derivatives. We have recently become interested in the eects of replacing H by F in inorganic complexes on reactivities(10), on solvation(10,11), and on solvatochromism(12). In respect to solvation, our preferred method of assessing solvation changes through transfer chemical potentials derived from solubility measurements requires that the complexes being studied are both stable and inert. As ¯uorine-containing substituents tend to reduce stability constants but increase substitutional lability, it seemed worthwhile to try to prevent such * Author to whom correspondence should be directed. ** Permanent address: Department of Chemistry, Faculty of Science, South Valley University, Sohag 84215, Egypt. 0340±4285
Ó 1998 Kluwer Academic Publishers
substituents from aecting stability and lability while leaving them on the periphery of complexes where they could have their full eect on solvation. We have started this line of investigation by preparing complexes of ligands in which ¯uorine-containing substituents on a phenyl ring are insulated by a methylene spacer from the central part of the complex. The availability of suitably substituted benzylamines for the synthesis of substitution-inert low-spin iron(II)-Schi base complexes makes this approach feasible. We report on substituent eects on base hydrolysis rate constants for tris-ligand iron(II) complexes of the ligands shown in Scheme 1, which provide direct comparisons between series of benzylamine-based and aniline based ligands. We have also carried out an initial investigation of medium eects on reactivity for base hydrolysis of four of the benzylaminederived complexes, in aqueous acetone and in aqueous dimethyl sulphoxide.
Scheme 1. The Schi base ligands and their abbreviations.
The reaction occurring in these kinetic experiments is nucleophilic attack of hydroxide at the complex, which gives free ligand and, in the presence of the dissolved oxygen in the solutions, colloidal iron(III) hydroxide. The cell contents are colourless and optically clear at the end of each run, but a very faint yellowish-brown turbidity can sometimes be detected if the cell contents are retained for several days. Experimental The complexes were prepared by Krumholz's method(13) from iron(II) ammonium sulphate, 2-acetylpyridine, and the appropriate (substituted) aniline or benzylamine, in EtOH-H2O media. The intensely coloured complex cations thus produced were precipitated as their iodide
690
Burgess et al.
Transition Met. Chem., 23, 689±691 (1998)
salts [Fe(LL)3]I2 by the addition of KI. Krumholz prepared and characterised [Fe(apai)3]2+; the other complexes are reported here for the ®rst time. All gave satisfactory CHN microanalyses for anhydrous [Fe(LL)3]I2 after drying for some days in a vacuum desiccator. The choice of iodide as counterion was determined by our wish to avoid over-rapid combustion of perchlorate salts in the CHN microanalysis apparatus. Kinetic experiments were carried out in 1 cm cells in thermostatted cells in a Pye-Unicam Sp1800 or a Shimadzu UV160 spectrophotometer. First-order rate constants for single-stage runs were computed by means of a least-mean-squares program from data collected over the ®rst two and a half half-lives for each run(14). Rate constants for two-stage runs were calculated by appropriate combinations of graphical and computer methods.
Table 1. Observed ®rst-order rate constants, kobs, for base hydrolysis of tris-ligand iron(II) complexes of Schi bases derived from 2-acetyl pyridine and (substituted) anilines, in aqueous solution, 0.106 mol dm)3 NaOH, at 298.2 K. Initial concentration of complex 10)4 mol dm)3 Ligand
Substituenta
104 kobs (s)1)
apai apmai apmxai aptfmai apfai
H Me OMe CF3 F
7.6 3.0 2.7 4.8 15.0
a
i.e. X in formula 1.
2 ÿ ÿdFe
LL2 3 =dt fk1 k2 OH gFe
LL3
over the normally-studied range of hydroxide concentrations. In practice the k1 term is generally very small compared with the k2[OH)] term at a hydroxide concentration of 0.106 mol dm)3 as used in the present investigation. We have carried out a series of runs, in weakly acid solution and in the presence of edta(19), which have shown that the k1 term is indeed negligible (
