Table 1 Substituent effect on methyl transfer from an ammonium ion to
arylthiolatocobaloximes
(a)
+
+
SET
Me NR2 Ar• –
ArS•
+
ArS[CoIII]
+
+
Me NR2 Ar
Starting
material
Ammonium
ion
Yield
(%)
X
Product
+
ArS Me
R2NAr
+
[CoIII]+
3b
3b
3b
3a
3b
3c
3d
3a
3b
3c
3d
9a
9b
9c
11
11
11
11
14
14
14
14
OMe
H
NO2
OMe
H
10
10
10
19
30
38
77
72
70
65
61
51
48
44
(b)
+
+
ArS[CoIII]
ArS• + R3N–CHnPh3–n
[CoIII]+
• –
+
R3N CHnPh3–n
12a
12b
12c
12d
15a
15b
15c
15d
ArS–CHnPh3–n + NR3 + [CoIII]+
Scheme 4
ArS• + •CHnPh3–n
III +
Cl
+ NR3
+
[Co ]
CN
OMe
H
Cl
CN
[Scheme 4, (b)] because these migrations occur in preference to
methyl group migration in spite of greater steric hindrance. In
accordance with the SET mechanism, the reactions proceed
more efficiently in polar MeCN than in less polar CHCl3 or
protic MeOH.
80
Direct methyl transfer to a thiolate anion8 via an SN2
mechanism cannot explain the involvement of coenzyme-B12 in
enzymatic processes. The present experimental findings ac-
count for assistance by a cobalt complex and suggest a possible
scheme for the enzymatic process, in which N5-protonated-
N5-methyltetrahydrofolic acid and homocysteinylthiolato–co-
enzyme B12 complex are reaction partners.
40
We thank Waseda University and the Ministry of Education,
Sports, and Culture of Japan for financial support via the Annual
Project Program (95A-252) and a Grant-in-aid for Scientific
Research (09640649), respectively.
10
t / h
20
Footnotes and References
Fig. 1 Time course of the reaction of arylthiolatocobaloximes 3 with
benzyldimethylammonium salt 11: (8) 3a, (5) 3b, («) 3c and (2) 3d
* E-mail: mtada@mn.waseda.ac.jp
† Reaction conditions were the same for all reactions of 3. A mixture of 3
(0.2 mmol) and an ammonium salt (0.25 mmol) in MeCN (10 cm3) was
heated to 80 °C for 22 h and then concentrated to 1 cm3. The concentrated
mixture was subjected to Florisil chromatography (hexane–CH2Cl2) to
remove polar non-volatile materials. The yields of 5 were obtained by GC
analyses of the eluate using an internal standard, and those of 8, 10, 12 and
15 refer to the isolated products. In most cases the starting materials
persisted but the reactions were stopped after 22 h to assess the relative
reactivities.
of 3 must be inversely proportional to the bond strength and
hence the bond distance of the sulfur–cobalt bond. X-Ray
crystal analyses11 determined the Co–S distance of 3 as shown
in Table 2. However, this correlation between the oxidation
potential (Eox) and the bond distance (dCo–S) breaks down upon
moving from 3c and 3d. The X-ray analysis unexpectedly
showed the Co–S bond in 4-chlorophenylthiolatocobalox-
ime(iii) 3c to be shorter than the others. As we pointed out
earlier,12 the strength of the cobalt–sulfur bond is a consequence
of the donation and back-donation from sulfur to cobalt. Cyclic
voltammetry of arylthiolatocobaloximes 3a–d showed a revers-
ible oxidation wave as shown in Table 2, and the reactivity of
the present methyl and arylmethyl group transfer has a trend
parallel with the oxidation potential (Eox) of 3a–d; a lower
oxidation potential results in higher reactivity.
The substituent effects on both sides of the reactants suggest
a single electron transfer (SET) mechanism. The intermediacy
of the arylthiyl radical is supported by the formation of diaryl
disulfide in all the reactions. As one of the plausible mecha-
nisms for the methyl transfer, we propose a homolytic
substitution by an arylthiyl radical on the methyl group of the
zwitterionic species formed by SET [Scheme 4, (a)].‡ The
migrations of the benzyl and diphenylmethyl groups are
considered to proceed via a radical coupling mechanism
‡ The direct attack of the arylthiyl radical on ammonium salts is ruled out
by the lack of reactivity of the arylthiyl radical generated by PhSH–AIBN
or photolysis of (PhS)2.
1 R. G. Matthews and J. T. Drummond, Chem. Rev., 1990 90, 1275.
2 A. A. DiMarco, T. A. Babik and S. S. Wolfe, Annu. Rev. Biochem.,
1990, 59, 355; K. M. Noll, Method in Enzymology, ed, L. Lacker,
Academic, San Diego, 1995, vol. 251, p. 470.
3 R. T. Tayor, B12, ed D. Dolphin, Wiley, New York, 1982, vol. 2, p. 307;
C. Temple, Jr. and J. A. Montogomery, Folates and Pterins, ed, R. L.
Blackley and S. J. Benkovic, Wiley, New York, 1984, vol. 1, p. 62.
4 P. Van Beelen, A. P. M. Stassen, W. G. Bosch, G. D. Vogels, W. Guijt
and C. A. G. Haasnoot, Eur. J. Biochem., 1984, 138, 563; J. T. Keltjens,
P. C. Raemakers- Franken and G. D. Vogles, Microbial Growth on
C1-Compounds, ed, J. C. Murrell and D. P. Kelly, Intercept, Andover,
1993, p. 135.
5 R. G. Matthews, R. V. Banerjee and S. W. Ragsdale, Bio Factors, 1990,
3, 147; B. Krautler, FEMS Microbiol. Rev., 1990, 87, 349.
6 S. M. Polsen, H. Hansen and L. G. Marzilli, Inorg. Chem., 1997, 36,
307.
7 H. P. C. Hogenkamp, G. T. Bratt and S.-Z. Sun, Biochemistry, 1985, 24,
6428; G. N. Schrauzer and R. J. Windgassen, J. Am. Chem. Soc., 1967,
89, 3607.
8 E. Hilhorst. T. B. R. A. Chen, A. S. Iskander and U. K. Pandit,
Tetrahedron, 1994, 50, 7837.
9 R. G. Matthews, Folates and Pterins, ed, R. L. Blackley and S.
J. Benkovic, Wiley, New York, 1984, vol. 1, p. 497.
10 C. D. Taylor and R. S. Wolfe, J. Biol. Chem., 1974, 249, 4879.
11 Y. Inouye, T. Kambe and M. Tada, unpublished work.
12 M. Tada and R. Shino, J. Inorg. Biochem., 1991, 44, 89; M. Tada,
T. Yoshihara and K. Sugano, J. Chem. Soc., Perkin Trans. 1, 1995,
1941.
Table 2 Oxidation potentials and bond lengths of arylthiolatocobal-
oxime(iii) 3
Compound
X
Eox/Va
d(Co–S)/Å
3a
3b
3c
3d
OMe
H
Cl
+0.483
+0.542
+0.563
+0.777
2.291
2.280
2.261
2.274b
CN
a Pt electrodes; voltage vs. Ag/AgNO3; 3a–d (0.2 mmol dm23), Bu4NClO4
(0.1 mol dm23) in MeCN; scan rate, 0.100 V s21. The voltages were
referenced to ferrocene (Eox = 0.083 V). b Mean value of two independent
molecules in the asymmetric unit.
Received in Cambridge, UK, 22nd September 1997; 7/06823I
42
Chem. Commun., 1998