Trans effect of the good .sigma. donor ligands dimethyl phosphate

of SN1lim reactions of dimethylphosphonatocobaloxime and methylcobaloxime complexes. William C. Trogler, Robert Charles Stewart, and Luigi G. Marz...
0 downloads 0 Views 409KB Size
3697 R

R

0

0 4

3 R = ribitol side chains

for 3, A,,, 472 nm ( E = 4.15 X 10' M-' cm-' ) for 4, Amax 496 nm ( E = 3.9 X lo3 M - ' cm-') for 3, pK, = 4.8

the basic form of 2. The analytical, nmr, spectral, and titrimetric data as well as the inability of the product to form a Cu(I1) complex* allows an assured assignment of the structure of the product as being that of the 8hydroxy isomer (2). The kineticsg$lo of the oxidation of 1,5-dihydroisoalloxazines (reduced flavines) have been logically interpreted t o require the formation of an 02-dihydroflavine adduct. Massey, Palmer, and Ballou'O have suggested O2addition to the 1,5-dihydroisoalloxazinesat the 4a and 10a positions while Hemmerich and Miiller'l have provided arguments in support of addition of 0, at the 4a and 8 or preferably 6 positions. The 10a position has been ruled out as an obligatory position of 0, addition in this l a b o r a t ~ r y . ~ If~ ~the ' ~ dihydro form of 1

the case for other dihydroisoalloxazines, l 4 then O2 addition to the 8 position would provide the same intermediate as HOO- addition to oxidized 1. Thus, 0, oxidation of reduced 1 should provide 2 just as does HOO- treatment of oxidized 1 (eq 2). Experiments carried out at pH 5.0 and 9.5 established that 20 cycles of alternate dithionite reduction and 0, oxidation of 1 gave a ca. 9 4 x yield of unaltered 1. No absorbance at 480 nm could be detected at the end of either experiment. Therefore, the dihydro form of 1 does not undergo oxidation by addition of O2 at either the 6 or 8 position. It should be noted that the overall conversion of 1 to 2 represents a two-step synthesis of an 8-hydroxyisoalloxazine. Compound 2 is prepared by treatment of the corresponding isoalloxazine with sulfite in aqueous solution at ambient temperature.4 Acknowledgment. This work was supported by a grant from the National Science Foundation. (14) The structure of dihydroisoalloxazines is being investigated in this laboratory.

Stephen B. Smith, Martin Briistlein, Thomas C. Bruice* Department of Chemistry, University of CaLfornia Santa Barbara, California 93106 Received February 25, 1974

The Trans Effect of the Good +Donor Ligands ( C H 3 0 ) ~ PO, (CH30)(CsH5)PO,and (C6Hs),P0. Evidence for Complete Bond Breaking in the Transition State of S N l l i , Reactions of Dimethylphosphonatocobaloxime and Methylcobaloxime Complexes 80,-

0 1

Sir : The interesting class of uninegative phosphorus donor ligands of the type 0

R'

2

exists totally or in part as the 1,5-dihydro isomer, as is (8) A property of 6-hydroxyisoalloxazines: G. Schollnhammer and P. Hemmerich, Z . Nururforsch. B, 27, 1030 (1972). (9) Q. H. Gibson and J. W. Hastings, Biochem. J., 8 3 , 368 (1962). (lo) V. Massey, G. Palmer, and D. Ballou in "Flavins and Flavoproteins," H. Kamin, Ed., University Park Press, Baltimore, Md., 1971, p 349. (11) P. Hemmerich and F. Muller, Ann. N . Y . Acad. Sci., 212, 1 3 (1973). (12) G. J. Kasperek, L. Main, and T. C. Bruice, J . Chem. Soc., Chem. Commun., 847 (1972). (13) At p H 6.4 (phosphate buffer), = 0.1 with KC1 and with -10-7 M superoxide dismutase and equimolar ratio of oxidized flavine and re2 duced flavine (5 X 10-'M), the stopped flow-rate constants for the 0 oxidation of 1,5-dihydrolumiflavine and the loa-blocked 1,5-dihydro-3methyl-10-(2'-6 '-dimethylphenyl)isoalloxazine are comparable (Robert F. Williams and T. C. Bruice, unpublished).

have received little attention. 1-6 Since the existence of such ligands is not widely recognized, it is not surprising that no rate or mechanistic studies have been reported on metal complexes of these ligands. We felt such a study might prove valuable because dialkylphosphonates (R = alkoxy) are known to be good Q donors and to have a substantial trans influence.6 We have therefore prepared some new cobaloxime complexes of the type RR'P(O)Co(DH),L, where L = neutral N- or P-donor ligand, D H = monoanion of dimethylglyoxime, and R and R ' = CH,O or C6Hs. Cobaloximes (complexes containing the Co(DH)z moiety) are particularly suitable for a comparative mechanistic study because kinetic information is avail(1) A. Pidcock and C.R. Waterhouse, J . Chem. Sac. A , 2080 (1970). (2) F. H. Allen, A. Pidcock, and C. R . Waterhouse, J . Chem. SOC.A, 2087 - (1970). .-.,. . ~

(3) J. Bennet, A. Pidcock, C. R. Waterhouse, P. Coggan, and A. T. McPhail, J . Chem. SOC.A , 2094 (1970). (4) R. J. Haines, I. L. Marais, and C. R. Nolte, Chem. Commrm., 547 (1970). ( 5 ) R. J. Haines and C . R. Nolte, J . Organometa[. Chem., 24, 725 (1970). (6) G. G . Mather and A. Pidcock, J. Chem. SOC.,Dalton Trans., 560 (1973). The term trans influence is used to describe a ground state, not a kinetic, effect.

Communications to the Editor

3698

able on substitution reactions of cobaloxime complexes sentially independent of both the concentration and the containing the effective trans activating alkyl and sulnature of L’ when (1) went to completion. The data fit0 ligand^.^^^ Although the substitution reactions of are therefore consistent with the reaction scheme of eq cobalt(II1) complexes have received great a t t e n t i ~ n , ~ - ’ ~ 2 and 3. Formation of the five-coordinate interin only a few cases have unassisted dissociation reackl tions been well documented.10s13 Substitution reRR’P(O)Co(DH)*L RR’P(O)Co(DH)2 + L (2) k actions of alkylcobaloximes have been extensively kz studied as models for Blz coenzyme. 11-13 However, RR’P(O)Co(DH)2 + L’ +RR’P(O)Co(DH)*L’ (3) most studies have revealed second-order kinetic beI;-* havior (except for the nmr studies of Brown’s groupi3). mediate is rate limiting and kobsd = k1. The specific complexes prepared contain the three In order to compare the kinetic parameters of the ligands PI, PII, and P I I ~ . The substitution reactions cobaloxime complexes over a temperature range, benzene was employed as a solvent. The reaction of the tBu(py) complexes (0.01 M ) of PI, PII, PIII, and CH3 with Bu3P (0.2 M ) gave the following results for kobsd PI PI I PI11 (25O, sec-I), AH* (kcal/mol), and A S * (eu), respec26.7 f 0.3, 12.0 =k 1.0; PII, tively: PI, 8.3 X studied involve replacement of neutral ligands L by 1.0 X 26.4 i: 0.4, 16.0 f 1.3; PIII, 1.1 X other ligands L’, where L and L’ are phosphines, 24.9 f 0.6, 11.0 f 2.0; and CH3, 3.9 X lo-*, 23.7 =t phosphites, or heterocyclic N donors, according to (1). 0.6, 5.0 =t 2.0. Activation parameters for methylcobaloximes (in nonaqueous solutions) have been reRR’P(O)Co(DH)zL + L’ +RR’P(O)Co(DH)zL’ + L (1) ported.i3,16 It is clear that the RR’P(0)- ligands are Evidence for P-bonding derives from both the large very good trans labilizers. Combining our data with trans labilizing effect of the RR’P(0)- ligands and the other comparisons in the l i t e r a t ~ r e , ~ ~ 8 ~wei 0 ,find 1 ~ the characteristic v(P=O) stretch. l 4 kinetic trans effect order to be PIII PII > CH3 > PI Rates of reaction 1 in the forward direction were SO3 >> NOz. The higher trans effect of PI1l and PII us. determined for the substitution of PICO(DH)~L(where PI is most likely due to the greater electron donating inL = CNpy and Ph3P) by a number of L’ in CH2Clz ductive effect of the phenyl group than the methoxy solvent and for the substitution of t-Bu(py) in XCogroup. 18,19 (DH),(t-Bu(py)), (X = PI, PII, PIII, CH3) by B U R in Extensive theoretical and experimental effort has been C6H6 solvent. The reactions wer: monitored by devoted toward explaining the kinetic trans effect. 2o A changes in the visible region (-5000 A, under Nzfor X clear understanding of the effect depends on an accurate = CH,) and found to be first order in complex concenknowledge of the nature of the activated complex. tration for up to five half-lives in some cases.’; Some Competition studies allow insight into the degree of reactions were also followed using pmr spectroscopy. bond breaking in the transition state of dissociative reThere was good agreement between the two techniques, actions in suitable cases. This is because competition but only the more precise spectrophotometric results are ratios close to unity imply a very reactive interreported here. At sufficiently high L’ concentrations mediate.9*10 It follows from Hammond’s postulatez1 and in the absence of added L, reaction rates were esthat, in such a case, almost complete bond rupture has occurred in the transition state of the dissociative re(7) A. L. Crumbliss and W. I