thus possible to reduce the necessary amount of diphenylzinc
to 0.65 equiv and still transfer both aryl groups to the
aldehyde. Mechanistic studies suggested that the active zinc
species is a mixed phenylzincethyl, which is less active than
diphenylzinc and thus more selective.8,1b However, the inter-
change of phenyl and ethyl groups in this reaction is con-
sistently difficult to monitor under relevant reaction condi-
tions. Three further improvements were recently made in this
field. The use of boronic acids as an aryl source enabled the
transfer of functionalized aryl rings to aldehydes.1 Triph-
Figure 1. Boron compounds as phenylzinc precursors.
c,9
1
0
enylborane was demonstrated to be a viable phenyl source
and finally, additives enabled the use of simple, commercially
mixture was quenched by addition of iodine in absolute
dichloromethane and analyzed by GC, using decane as
internal standard (see the Supporting Information).
1
1
available ligands for the triphenylborane protocol.
A serious drawback of all these protocols is in our eyes
still to be seen in the price and availability of the aryl sources.
While the expensive and sensitive diphenylzinc is only
available and reasonably employable for research purposes,
the arylboronic acid protocol calls for a huge excess of
diethylzinc (6 to 7 equiv) and drastic conditions for the
transmetalation step. Triphenylborane (1) on the other hand
is a reasonably priced but still a rather air sensitive and
flammable compound.
We therefore searched for other aryl sources for the
transmetalation step yielding active arylzinc reagents. Our
attention was first drawn to complexes of triphenylborane.
Triphenylborane sodium hydroxide complex is a com-
mercially available compound that is sold by DuPont as a
flame retardant. Various other complexes of triphenylborane
are known, many of which can easily be prepared by addition
of, e.g., a nitrogen nucleophile to the triphenylborane
solution. We chose to prepare the well-known ammonia
complex 2 via a simple protocol. Initial experiments showed
that this ammonia complex is an outstandingly stable,
versatile, and economic precursor for arylzinc reagents in
asymmetric catalysis. We here describe a detailed study on
its transmetalation behavior, as well as its application in
asymmetric catalysis.
To gain some understanding of the two occurring pro-
cesses, namely transmetalation and (enantioselective) 1,2-
addition of the intermediary formed active zinc reagent to
the aldehyde, we undertook kinetic studies of the ammonia
complex (2). Additionally, we studied the transmetala-
tion of diphenyl borinate (3) and triphenylborane (1) to
compare reactivity patterns and enantioselectivity of the
intermediary formed arylzinc reagents in addition reactions
to aldehydes.
The transmetalation reactions of compounds 1 to 3 were
studied at room temperature (20 °C) in toluene, the solvent
of choice for most reported enantioselective addition reac-
tions. After the given reaction time, a sample of the reaction
12
Figure 2. Transmetalation of boron compounds in toluene.
As it can be seen from Figure 2, the rate of the
transmetalation reaction of the three compounds is different.
While triphenylborane exchanges about 50% of its three
phenyl groups practically immediately (first sample was
taken after 1 min of reaction time), the two complexes take
longer. Borinate 3 shows a rather fast transmetalation
reaction, exchanging one of its two phenyl groups within
60 min. The rate of transmetalation for the ammonia complex
2
is decisively slower. However, after 60 min all three borane
compounds have exchanged about half of their available
phenyl groups. For triphenylborane and the borinate, the
amount of exchanged phenyl groups remains constant over
13
several hours. A slight decrease of active phenylzinc reagent
generated from ammonia complex 2 was observed in several
kinetic experiments and an experimental artifact can therefore
1
4
be ruled out.
The dissimilar transmetalation patterns should result in a
changed behavior in catalysis. To examine the catalytic
reaction, we performed the addition of phenylzinc generated
from the corresponding borane (1-3) to benzaldehyde in
(
8) Hermanns, N. Ph.D. Thesis, RWTH Aachen, Germany, 2002.
(9) (a) Bolm, C.; Rudolph, J. J. Am. Chem. Soc. 2002, 124, 14850. (b)
Rudolph, J.; Schmidt, F.; Bolm, C. Synthesis 2005, 840.
10) (a) Rudolph, J.; Hermanns, N.; Bolm, C. J. Org. Chem. 2004, 69,
(
3
8
997. (b) Rudolph, J.; Schmidt, F.; Bolm, C. AdV. Synth. Catal. 2004, 346,
67.
(13) Even after 16 h the amount of “active“ phenyl groups remains
unchanged.
(14) The formation of an equilibrium could be discussed giving rise to
an ammonia stabilized Ph2BEt + PhZnEt. It is unclear, however, which
process could lead to an intermediary higher level of detectable phenylzinc.
(
11) Rudolph, J.; Lormann, M.; Bolm, C.; Dahmen, S. AdV. Synth. Catal.
005, 347, 1361.
12) Layton, W. J.; Niedenzu, K.; Niedenzu, P. M.; Trofimenko, S. Inorg.
Chem. 1985, 24, 1454.
2
(
4598
Org. Lett., Vol. 7, No. 21, 2005