ORDER
REPRINTS
2
940
WU AND SHEN
6
is an excellent reagent for asymmetric aldol condensation. However, it is
rather expensive and commercially available only as 1.0 M solutions in low
boiling organic solvents. Its high sensitivity to air also causes additional
difficulty in storage. For these reasons, we chose to prepare n-Bu BOTf
2
from triethylborane ourselves. In literature searching, we found, to our
surprise, that despite its broad utility and the long history, essentially all
the tributylborane used in organic synthesis over the last 60 years was pre-
8
pared by Johnson’s procedure using butyl Grignard reagent (Eq. 1), and
no-one seemed to have tried using n-butyllithium in preparation of tribu-
tylborane, although since the late 1970s n-butyllithium has been the most
widely and most frequently used organometallic compound, and can be
found practically in every synthetic laboratory.
n-BuMgX þ F BEOEt ! n-Bu B þ MgXF
ðEq: 1Þ
3
2
3
It appears to be rather rare in organic chemistry that the preparation
of a given compound essentially relies on one single procedure over some
6
1
0 years. In further literature studies we found that Brown did mention in
9
986 that lithium was not suitable for preparing tributylborane, when he
reported a one-pot protocol. However, judging from the reaction conditions
employed there, this failure to obtain tributylborane from lithium does not
necessarily suggest that butyllithium could not react with boron trifluoride
to form the desired alkylborane in synthetically useful yields. This consid-
eration and the potential convenience of preparing tributylboron from
n-butyllithium encouraged us to examine the reaction of commercially avail-
able n-butyllithium with boron trifluoride (Eq. 2).
3
n-BuLi þ F BEOEt ! n-Bu B þ 3LiF:
ðEq: 2Þ
3
2
3
Similarly to the Grignard reagent, n-butyllithium reacts violently with
boron trifluoride. However, the reaction phenomenon was somewhat differ-
ent from that with the Grignard reagent, where the inorganic salt formed
during the reaction precipitates from solution. In some runs we obtained
substantial amounts of ‘‘ate’’ compound, which mixed with the inorganic
salt to give a sticky paste. On contact with water in the work-up, a violent
exothermic reaction took place, giving off large amounts of gas (butane).
This led to drastically reduced yields of tributylborane. By changing the
amount of ether and the method of mixing the reactants, we obtained the
most satisfactory results by adding an ethereal solution of freshly distilled
1
0
boron trifluoride to n-butyllithium. In this way, tributylborane can be
obtained on a 30 mmol scale in ca. 64% distilled yield with rather narrow
boiling point range.
Although using butyllithium to replace butyl Grignard reagent to react
with boron trifluoride is unremarkable in theory, this work may clear the
Wu
