Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
10.1002/ejoc.201801173
European Journal of Organic Chemistry
FULL PAPER
very slow in such media.[17] Also, chelating ligands as well as
lithium salts are commonly employed to modify the properties of
organolithiums, since they demonstrated high influence on the
nature and reactivity of organometallics aggregates.[16,17,18]
Therefore we envisioned that such additives could allow the
reaction to occur or could improve it when it furnished improvable
yields.
Interestingly, whereas only low conversion rates were observed
when hexane or toluene was used as solvent (15 and 18 %
respectively, entries 6 and 7), the use of dimethoxyethane (DME,
1.0 equiv.) as co-solvent proved to be effective, since it allowed
On the other hand, it has been reported previously[18h,19] that the
use of 2.0 equiv. of t-BuLi in the reaction conditions could
potentially “remove” a full equivalent of ligand by preferential
complexation of the latter with the lithium bromide salt generated
by elimination reaction starting from the t-BuBr formed and the
excess of base (t-BuLi). To address this problem, one solution is
to increase the [ligand:t-BuLi] ratio, either by increasing the
amount of ligand used or by decreasing the stoichiometry of the
base. Attempts to use a [ligand:t-BuLi] ratio of [1:1] in the reaction
performed with DME had a deleterious effect on the yield (29%,
entry 4, Table 2). This result naturally led us to question on the
recovering of 3a in 45 % yield (almost quantitative conversion rate, actual effect of LiBr on the coupling reaction.
entry 8, Table 1). A similar trend was observed for the carbocyclic
analog bromobenzene 1b, although lower yields were generally
obtained.
Subsequently, to acquire further insight on the effect of the
chelating DME on the coupling reaction, we decided to examine
the influence of the ligand stoichiometry on the reaction
developed in toluene (Table 2).
Subsequently, to acquire further insight regarding the influence of
lithium bromide, inevitably generated along with benzyne, as well
as in the first stage of the reaction when using 2.0 equiv. of t-BuLi,
we investigated the effect of this salt when added to the reaction
in toluene during the metalation step, with and without DME
(Table 3). Importantly, in addition to the classical metalation route
employed until now with t-BuLi for 1a (Metalation A), a second
metalation protocol (Metalation B) using n-BuLi (1.0 equiv.) as the
base was developed to avoid the initial formation of LiBr in the
reaction medium, making this metalation pathway the “salt-free
reference method”.
Table 2. Influence of the number of DME equivalents on the coupling
reaction in toluene.
Considering first attempts performed with DME, it is clear that the
salt positively affected the reaction: up to 69% yield was indeed
achieved when an excess of LiBr was added to the reaction using
the usual metalation path A (entry 7, metalation A) overtaking the
previous results obtained without further introduction of salt by
approximately 10 points (58%, entry 3, metalation A). Interestingly,
whether metalation path A or B was used for the metalation of 3-
entry
DME
(m eq)
t-BuLi
(p eq)
Yield[a] (%)
bromothiophene, similar results were obtained when
a
3a
comparable amount of salt was added to the reaction (entries 3-
7). Especially an appreciable increase was reported when more
than 3.0 equiv. of salt were added to the reaction. The LiBr effect
is actually demonstrated by the considerable drop in yield
observed without any salt in the reaction mixture apart from the
one formed concomitantly with benzyne (40%, entry 2 –salt-free
metalation B without LiBr addition). Nevertheless, addition of the
ligand remains essential to the reaction since all experiments
failed to occur when performed without DME, whatever the
amount of LiBr used or the metalation path followed (entry 1). It
appeared overall that the previous hypothesis on the possible
trapping of the ligand by the salt generated in the metalation step
with t-BuLi —that could have explained the considerable
improvement of the yield observed when using an increased
amount of ligand— is not consistent with these observations, and
that the combination of both salt and ligand seemed favorable to
the reaction.
1
2
3
4
5
6
7
0
0.5
1
1
2
4
10
2
2
2
1
2
2
2
0
26
45
29
54
58
56
[a] Isolated yields after centrifugal thin-layer chromatography purification.
The obtained results clearly showed that the use of an increased
amount of ligand had a beneficial effect on the yield. While the
use of
a substoichiometric amount of ligand seemed to
dramatically affect the reaction (entry 2), better yields were not
achieved when increasing the amount of ligand to 10.0 equiv.
(entry 7) and it appeared overall that 4.0 equiv. of ligand proved
to be optimal (entry 6), results remaining pretty close to those
initially obtained in THF (see Table 1, entry 2).
This article is protected by copyright. All rights reserved.