Definitive evidence of diffusing radicals in Grignard reagent formation

Full text of DOI:10.1021/ja963208m

J. Am. Chem. Soc. 1997, 119, 253-254
253
Definitive Evidence of Diffusing Radicals in
Grignard Reagent Formation
Since ks can be decreased by appropriately deuterating the
solvent (kinetic isotope effect), studies of solvent isotope effects
can distinguish definitively between the KRW mechanism and
the D model.
John F. Garst,* Ferenc Ungv a´ ry,*,1a and James T. Baxter1b
Cyclopropyl bromide (CpBr) is an appropriate substrate for
such studies because in its reactions, unlike those of typical
Department of Chemistry
The UniVersity of Georgia
Athens, Georgia 30602
6
,13,14
alkyl bromides, s is significant.
Accordingly, we have
investigated the effects of solvent perdeuteration on reactions
of cyclopropyl bromide (CpBr) in diethyl ether (DEE), MgBr2/
DEE, THF, and MgCl2/THF.
ReceiVed September 12, 1996
In every case (Table 1), the yield of r (CpMgBr) or c (CpCp),
or both, increases on solvent deuteration. This disproves the
KWR mechanism.
In reactions of magnesium with alkyl halides (RX) in ether
solvents (SH), intermediate alkyl radicals (R ) reduce to Grignard
•
reagent (r), couple and disproportionate (c), and attack the
solvent (s).2-14
Since it is generally agreed that s is a reaction of diffusing
radicals, the observed increases of 12-14% in the yields of
CpMgBr in MgBr2/DEE, THF, and MgCl2/THF demonstrate
that at least 12-14% of the CpBr reacts to give diffusing
Possible mechanisms can be distinguished by
the roles assigned to adsorption and diffusion for these steps.
In an “A” (adsorption) step, the radicals remain adsorbed at
the magnesium surface until they react. Otherwise, the step is
•
radicals Cp that are reduced to CpMgBr in the deuterated media.
“
D” (diffusion), a reaction of radicals that diffuse in solution.
This proves that radicals that leave the magnesium surface can
be converted to Grignard reagent, an essential element of the
Specifying the nature of the steps in the order rcs, three of the
possible mechanisms are AAD [Kharasch-Reinmuth-Walbor-
7
sky (KRW) mechanism],2
-8
9-14
D model that has been disputed.
ADD, and DDD (D model).
D-model calculations can provide quantitative predictions of
the effects of solvent deuteration as a function of the value of
10
the isotope effect, ks(SH)/ks(SD). Consider MgBr2/DEE. The
viscosity is higher than that of DEE, so the reaction is slower
-
6
-2 -1
and c is negligible. For V ) 2.0 × 10 mol cm
s
(V )
flux of radical formation at the magnesium surface) and the
values of all other rate parameters except δ the same as those
•
13
used (and justified) previously for Cp , the experimental value
of the yield of CpMgBr (71%) is matched by setting δ to 0.010
-
1
Å
(δ ) κ/D; κ ) heterogeneous rate constant for r; D )
•
diffusion coefficient of Cp ). Decreasing ks by factors from 2
to 10 leads to calculated yields of CpMgBr of 77-87%. The
observed value in MgBr2/DEE-d10, 84%, is matched with an
isotope effect of 6, a plausible value. Reactions in THF and
MgCl2/THF are quantitatively similar.
The effect of diffusion can be evaluated by comparing the
D-model calculation with one based on a (pseudo-)first-order
kinetic model for competing r and s. For a decrease in ks by a
factor of 6, the first-order model predicts an
Although a considerable body of data verifies quantitative
9-13
predictions of the D model,
the KRW mechanism has not
been disproved. It cannot be tested against such data because
it does not support quantitative predictions. It does, however,
provide a definite prediction of the effect of decreasing the rate
constant (ks) for s; the product distribution will remain un-
changed (because r, c, and desorption, and not s, are product-
determining). In contrast, the D model predicts that decreasing
ks will increase the amount of r, c, or both at the expense of s.
increase in the yield of CpMgBr from 71% to 94%, a significant
overestimate. Relative to the first-order model, diffusion
attenuates the rate of r in the D model, reducing the calculated
yield of CpMgBr from 94% to 84%.
(1) (a) Permanent address: Institute of Organic Chemistry, University
of Veszpr e´ m, 8201 Veszpr e´ m, Hungary. (b) Permanent address: Chemistry
Department, Valdosta State University, Valdosta, GA 31698.
Among the media studied, pure DEE is unique in that its
deuteration increases c by a factor of nearly 4 but r only
marginally. Here s is coupled strongly to c but weakly (if at
all) to r, a result that is predicted by the ADD mechanism. There
are several plausible explanations. (1) The mechanism in DEE
is ADD. It is not clear why it should be ADD in one medium
and DDD in the others. (2) The mechanism is DDD in all
media, but in pure DEE the buildup of polar solutes (MgBr2,
RMgBr) adjacent to the magnesium surface creates a zone that
radicals do not often re-enter, once they have left it. If most of
the s and c occur outside that zone, then the observed behavior
results. Indeed, we have found two liquid phases in the product
(
2) Kharasch, M. S.; Reinmuth, O. Grignard Reactions of Nonmetallic
Substances; Prentice-Hall: New York, 1954.
(
3) Walborsky, H. M.; Young, A. E. J. Am. Chem. Soc. 1964, 86, 3288.
4) Walborsky, H. M.; Aronoff, M. S. J. Organomet. Chem. 1973, 51,
(
3
4
1.
(
5) Walborsky, H. M. Acc. Chem. Res. 1990, 23, 286.
(
6) Walborsky, H. M.; Zimmermann, C. J. Am. Chem. Soc. 1992, 114,
996.
(
7) Walborsky, H. M.; Topolski, M.; Hamdouchi, C.; Pankowski, J. J.
Org. Chem. 1992, 57, 6188.
8) Hamdouchi, C.; Walborsky, H. M. In Handbook of Grignard Reagent;
Silverman, G. S., Rakita, P. E., Eds.; Dekker: New York, 1996; Chapter
(
1
1
1
0, pp 145-218.
9) Garst, J. F.; Deutch, J. M.; Whitesides, G. M. J. Am. Chem. Soc.
986, 108, 2490.
10) Garst, J. F.; Swift, B. L.; Smith, D. W. J. Am. Chem. Soc. 1989,
11, 234.
(
1
5
mixture from some reactions in DEE. One of these could lie
adjacent to the magnesium and constitute a viscous, polar zone
that could have these effects. (3) The early part of the reaction
in pure DEE involves mostly s and c, consistent with the
proposition that r requires the presence of MgBr2, which builds
(
(
(
(
11) Garst, J. F.; Swift, B. L. J. Am. Chem. Soc. 1989, 111, 241.
12) Garst, J. F. Acc. Chem. Res. 1991, 113, 95.
13) Garst, J. F.; Ungv a´ ry, F.; Batlaw, R.; Lawrence, K. E. J. Am. Chem.
Soc. 1991, 113, 5392.
(14) Garst, J. F.; Lawrence, K. E.; Batlaw, R.; Boone, J. R.; Ungv a´ ry,
F. Inorg. Chim. Acta 1994, 222, 365.
(15) Unpublished observations of R. Batlaw in our laboratories.
S0002-7863(96)03208-8 CCC: $14.00 © 1997 American Chemical Society

254 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Communications to the Editor
Table 1. Effects of Solvent Deuteration on Products of Reactions of Cyclopropyl Bromide with Magnesiuma
n
[MgX
2
]
0
CpMgBr
CpCp
CpS
SS
DEE
THF
SH
SD
SH
SD
6
0.0
0.0
2.6
2.6
52 (49-57)
54 (52-55)
71 (70-72)
84 (84-85)
3 (1-4)
14 (13-14)
2 (1-3)
4
3 (3-4)
7 (4-8)
3
3
2
3
1.2 (1.0-1.6)
5 (2-8)
2
0.3
0.04 (0.03-0.05)
SH
SD
SH
SD
7
2
4
2
0.0
0.0
0.5
0.5
58 (55-61)
70 (69-71)
68 (65-74)
80 (78-82)
b
b
b
b
10 (6-18)
3
6 (3-9)
6 (3-9)
16 (7-25)
0.3
4 (2-6)
0.07 (tr-0.14)
c
a
[CpBr]
0
2 0 2 2
) 0.40 M; temperature ) 37 °C. [MgX ] is the initial molar concentration of MgBr (DEE) or MgCl (THF). n is the number of
replicate experiments. For CpMgBr, CpCp, and CpS, the yield is the percentage of Cp groups (of the CpBr consumed) accounted for in the
•
product. For CpS and SS, the yield is the percentage of Cp groups accounted for as residues (S) of solvent attack by Cp . Grignard reagent was
determined by titration with 2-butanol; CpCp, CpS, and SS were determined by GC.1
3,14
The tabulated yields are averages. The limits of observed
b
variations are given in parentheses. Where no limit is given, the range was zero (for the tabulated number of significant figures). Solvent interference
prevented reliable GC analyses for CpCp; the yields appear to be low. c For one of these experiments, [CpBr]
) 0.57 M and [MgCl ) 0.7 M.
0
2 0
]
up as the reaction proceeds.14 Deuterating DEE would divert
some s to c, but not r, during the early part of the reaction. The
effects of deuteration during the later part, when r is more
important, would be attenuated by the effects involving s and c
only in the early part. These and perhaps other possible
diphenylcyclopropane (*CpBr).3-8 This is not necessary at
present because there are several viable alternatives, including
•
a minor retention pathway along which *Cp is not an
3-8
intermediate, as is suggested by Walborsky.
explanations deserve further exploration.
As noted previously,11 quantitative predictions of the D model
Acknowledgment. We are grateful for support provided by a grant
from the National Science Foundation. Acknowledgment is made to
the donors of the Petroleum Research Fund, administered by the
American Chemical Society, for partial support of this work. We thank
Dr. Susan D. Richardson and Alfred D. Thruston, Jr., for GC-MS
assistance.
do not depend on whether or not there is transient adsorption
of intermediate alkyl radicals at the magnesium surface, provided
that r is their only surface reaction. The D model will
accommodate such adsorption if it becomes necessary to invoke
it to account for the partial retention of configuration that is
found for reactions of optically active 1-bromo-1-methyl-2,2-
JA963208M