¨
E. Erdik, O.O. Pekel / Journal of Organometallic Chemistry 693 (2008) 338–342
¨
339
Our interest in the reactivity of mixed organozinc
reagents prompted us to carry out a systematic study to
probe the origin of transfer selectivity and also to find
the relative transfer ability of organyl groups in the reac-
tions of the mixed diorganozincs. To the best of our knowl-
edge, this subject has not been investigated in detail [6–11].
A theoretical study was published by Bolm and co-workers
for phenyl versus ethyl transfer in the addition of diorgano-
zincs to aldehydes [29].
For evaluating the effect of preparation method, reac-
tion temperature and time on the relative transfer ability
of organyl groups in the CuI catalyzed benzoylation of n-
BuPhZn 1ab, the ratio of product yields, i.e. 4a/4b, was
found. The yield of uncatalyzed reaction did not exceed
20% with a 4a:4b ratio of 50:50. The background yields
for benzoylation of mixed diorganozinc 1ab were found
to be 90% for din-butylzinc 1a and 66% for diphenylzinc
1b.
Here we report the dependence of the relative transfer
ability of n-butyl and phenyl groups on the reaction condi-
tions in the Cu(I) catalyzed benzoylation of n-butyl phenyl-
zinc and our success in the control of organyl group
transfer by changing the solvent or using an additive.
The amount of CuI catalyst 3a was optimized to be
10 mol% and the reaction time to be 20 min at the room
temperature. Using 1:1 or 2:1 as molar ratio of 1ab:2 did
not make any appreciable change on the benzoylation yield
and on the 4a:4b ratio. The reaction carried out at lower
temperatures than room temperature gave lower yield
and took a longer time to reach the yield obtained at room
temperature. The total benzoylation yield and relative
transfer ability of n-Bu and Ph groups of 1ab prepared
by Methods A1 and A2 are the same in the error limit of
GC analysis, since we found that the optimized yield is
78% and 86% and the ratio of 4a:4b is 58:42 and 60:40,
respectively. This result shows that the organic group,
which is originally bound to Mg, i.e. Ph and n-Bu in Meth-
ods A1 and A2, respectively, has no effect on the reaction
outcome as expected.
(ii) We carried out the benzoylation of n-BuPhZn 1ab in
THF using coordinating solvents and also Lewis base or
Lewis acid additives. We used N-methyl-2-pyrrolidinone
(NMP), N,N,N0,N0-tetramethylethylenediamine (TME-
DA), N,N-dimethyl propyleneurea (DMPU), hexamethyl-
phosphoric triamide (HMPA) and diethylene glycol
dimethyl ether (diglyme) as a coordinating cosolvent and
n-Bu3P (tri n-butyl phosphine) as a Lewis base additive.
The yields and relative transfer ability of n-Bu and Ph
groups in THF and in THF:cosolvent (and/or additive)
are given in Table 1. We were surprised to see that the
transfer ability of n-Bu group increases in THF:NMP
(3:1), in THF:diglyme (2:1), in THF:DMPU (2:1) and in
THF:HMPA (1:1). Since we found that the ratio of 4a:4b
is 87:13, 89:11, 81:19 and 79:21 (entries 6, 9, 10 and 11),
respectively, compared to 60:40 ratio in THF alone (entry
3). We were also delighted to see that transfer of only n-
Bu group takes place in the presence of n-Bu3P (1 equiv.)
with a 4a:4b ratio of 100:0 (entry 14) and transfer of almost
only Ph group takes place in THF:TMEDA (2:1) with
4a:4b ratio of 8:92 (entry 17). The benzoylation yield is
lowered to 59–68% in the presence of N-donor solvents
and does not change in the presence of n-Bu3P, i.e. 82%
compared to 86% in THF alone, but increases to 94% in
the presence of O-donor solvent, i.e. diglyme. Before look-
ing more closely at these results, we also determined the
background yields, i.e. benzoylation yields of n-Bu2Zn
and Ph2Zn in THF:NMP, in THF:diglyme, in THF:TME-
DA and in the presence of n-Bu3P (entries 4, 7, 12 and 15
for n-Bu2Zn; entries 5, 8, 13 and 16 for Ph2Zn).
2. Results and discussion
For the influence of reaction conditions on the group
transfer selectivity of n-butylphenylzinc in the Cu(I) cata-
lyzed acylation with benzoyl chloride,
CuðIÞcatalyst
3
n-BuPhZn þ PhCOCl
PhCOn–Bu þ PhCOPh ð1Þ
!
1ab
2
THF:cosolvent
4a
4b
we focused on the following parameters:
(i) Preparation method of the mixed diorganozinc, reac-
tion temperature and time and (ii) cosolvents and additives.
We carried out the coupling reaction by addaing ben-
zoyl chloride 2 to n-butylphenylzinc 1ab in the presence
of a Cu(I) catalyst 3 in THF or THF:cosolvent (or addi-
tive). We determined the relative transfer ability of n-Bu
and Ph groups by finding the GC yields of ketones 4a
and 4b using authentic samples.
(i) In this study we used magnesium-based organozinc
reagents, i.e. n-butyl- and phenylmagnesium bromides to
be transmetallated. We prepared n-butylphenylzinc by
three different methods. In the in situ preparation meth-
ods [18–20], n-butylzinc bromide prepared by transmetal-
lation of n-butylmagnesium bromide was allowed to
react with phenylmagnesium bromide (Method A1) or
the same procedure was applied by reacting phenylzinc
bromide with n-butylmagnesium bromide (Method A2).
We also mixed equimolar amounts of din-butylzinc and
diphenylzinc (Method B). This method allows the use
of commercially available diorganozincs, and it was
already shown that the equilibrium constants favored
the formation of R1R2Zn type reagents from R2 Zn
1
2
and R2 Zn [19,30]. However we found one pot successive
Mg to Zn transmetallation reactions (Methods A1 and
A2) more practical for the preparation of Grignard
reagent based mixed diorganozincs. In addition, we also
wanted to avoid the effect of equilibrium on the forma-
tion of mixed diorganozinc prepared by Method B.
Methods A1 and A2 are also expected to give the same
reagent; however, we also wanted to check if the group
originally attached to Zn, which is n-Bu or Ph, respec-
tively, could make a change in the relative transfer abil-
ity of these groups.
The data in THF:NMP (3:1) show that the transfer abil-
ity of n-Bu group in n-Bu2Zn and n-BuPhZn is 83% (entry