Pd–PEPPSI Complexes
FULL PAPER
Table 1. Evaluation of catalyst turnover for consecutive KTC cross-cou-
substrates whether oxidative addition did in fact occur to a
reasonable degree with complexes 12, 13 and 14; a more de-
tailed investigation on this aspect is given in the sp2–sp2 and
sp3–sp3 Negishi cross-coupling section (vide infra). The
anionic ligands on palladium (ClÀ vs. BrÀ) showed no nota-
ble effects on catalytic activity (1–3, 6), whereas the elec-
tronics of pyridine throw-away ligands revealed some de-
pendence. Complex 4 consistently provided a slightly lower
yield, which is probably the result of the bulky methyl
groups slowing down the halide–aryl exchange during cata-
lyst generation/reduction. To explain the lower reactivity of
lutidine derivative 4, we decided to conduct a rate study on
the formation of 15 in the hope of gaining some insight into
the details of this reaction (Figure 2).
pling reactions to form 15.[a,b]
Catalyst Conversion after 24 h [%]
(Yield [%])
Conversion after 48 h [%]
(Yield [%])
1
4
5
100 (93)
85 (81)
100 (91)
90 (87)
78 (75)
91 (87)
[a] Conditions: PEPPSI complex 1, 4 or 5 (2 mol%), p-chloroanisole
(0.5 mmol), p-tolylmagnesium bromide (0.65 mmol), total solvent
volume=2.1 mL. After 24 h and complete conversion to 15, additional p-
chloroanisole (0.5 mmol) and p-tolylmagnesium bromide (0.65 mmol)
were added and the reaction continued for an additional 24 h. Control
experiments with no catalyst showed no conversion in all cases. [b] The
procedure was repeated three times and the isolated yields were obtained
with <5% variation.
modest loss of activity after one reaction cycle. Note that, as
observed in the previous study, complex 4 was comparative-
ly less active despite the theoretical generation of an identi-
cal NHC–Pd0 species in each reaction vial. This lagging ac-
tivity could suggest that it is not just activation that is prob-
lematic with the hindered pyridine derivative, but that per-
haps the pyridine ligands attach to and detach from the
NHC–Pd0 complex in solution. Although lutidine is more
bulky, it is also more electron rich, which means that it
could be a stronger s-donating ligand than simple pyridine,
and most certainly a better coordinator than electron-poor
3-chloropyridine. The crystal structures of complexes 1 and
4 (Figure 3) provide further evidence to support both the
steric congestion argument during reduction and the better
~
&
*
Figure 2. Rate of formation of 15 by using catalysts 1 ( ), 4 ( ), 5 (
)
À
coordination of lutidine; the Pd N bond of complex 4
!
and PdCl2 ( ; control). The relative rates (initial) are as follows: 1: 2.11ꢄ
10À4 molLÀ1 sÀ1; 4: 5.15ꢄ10À5 molLÀ1 sÀ1; 5: 7.89ꢄ10À4 molLÀ1 sÀ1. Con-
version was determined by using GC/MS against a calibrated internal
standard (undecane). All experiments were performed in triplicate. Con-
trol experiments with no catalyst showed no conversion.
(2.086 ꢃ) is comparatively shorter than that of 1 (2.137 ꢃ),
despite unfavourable sterics that direct the methyl groups in
toward the metal centre. Additionally, the fact that no
blacking-out of Pd occurs supports the notion that the pyri-
dines reattach to the Pd centre, which stabilises it and
lengthens the catalyst lifetime.
At a 4% loading, our results indicate that complex 5
forms the same active catalyst as 1 in solution, albeit at
about one quarter of the rate, which indicates that there is
some difference between the dissociation of pyridine and 3-
chloropyridine. Alternatively, complex 4 displayed a more
obvious reduction in activity that led ultimately to a lower
conversion percentage after 24 h. Because 4 did successfully
form product 15, lutidine dissociation undoubtedly did take
place. To test whether catalyst death was occurring at any
time during the 24 h study, we next addressed the lifetime of
the three catalysts (Table 1). After an initial period of 24 h,
which gave quantitative conversion to 15 for 1 and 5, we in-
jected a second aliquot of the coupling partners into the
same reaction vial, allowed the contents to stir for a further
24 h with the remaining catalyst of the previous reaction.
After the initial complete conversion to the product, cata-
lysts 1, 4 and 5 remained highly active and suffer only a
In the next KTC study, we subjected the more active IPr-
NHC catalysts (1–8) to a more challenging KTC cross-cou-
pling reaction to see if any further differences in perfor-
mance could be detected with different substrates
(Figure 4). Complex 5 showed a 20% yield enhancement
over 1 for the formation of 16, and although the previous
two studies suggested that the same active species forms in
solution with similar ease, this may not be the case with this
particular group of substrates. One of our initial theories
was that pyridine, being less electron-withdrawing than 3-
chloropyridine, would have a higher dissociation energy
from PdII/Pd0 and thus allow more precatalyst to be con-
served in 5 than in 1. Assuming that conversion in this reac-
tion is slower than in Figure 1, catalyst lifetime would
become an important issue. The dissociated pyridine ligand
could now re-coordinate with the NHC–Pd0 complex after a
Chem. Eur. J. 2010, 16, 10844 – 10853
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10847