.
Angewandte
Communications
Table 2: Substrate scope.
Taking into account the importance of steric bulk in
promoting carbon–halogen reductive elimination, carbamoyl
chloride 1a was chosen for the reaction optimization
(Table 1). Initial screening revealed that both Pd(PtBu3)2
and Pd(Q-Phos)2 were inefficient catalysts for the chlorocar-
bamoylation reaction, giving low yields of 2a (entries 1 and
2). Upon switching to Pd(PtBu2Ph)2, a significant increase in
reactivity was observed (entry 3). Attempts to push the
Table 1: Reaction optimization.
Entry Pd catalyst
Ligand
(y mol%) [h]
Time Conv.[a] Yield[a] E/Z[a]
(x mol%)
[%]
[%]
[%]
1
2
3
4
Pd(PtBu3)2 (5)
Pd(Q-Phos)2 (5)
Pd(PtBu2Ph)2 (5)
Pd2dba3 (2.5)
Pd2dba3 (1.25)
Pd2dba3 (1.25)
–
–
–
–
18
18
18
3
19
28
92
100
100
26
12
21
86
99
99
8
N.D.[b]
N.D.[b]
>99:1
PA-Ph (10)
PA-Ph (5)
PA-Ph (5) 21
PA-Ph (5)
dppf (30) 24
>99:1
>99:1
>99:1
–
5[c]
6[c]
7[c]
8[d]
5
[a] Pd2(dba)3 (2.5 mol%), PA-Ph (10 mol%). [b] Pd2(dba)3 (2.5 mol%),
PA-Ph (10 mol%), 1008C. [c] Yield determined by 1H NMR analysis of
the crude reaction mixture using 1,3,5-trimethoxybenzene as the internal
standard.
5
7
24
0
<2
Pd(OAc)2 (10)
N.D.
[a] Conversions, NMR yields, and E/Z ratios were determined by
1H NMR analysis of the crude reaction mixture using 1,3,5-trimethoxy-
benzene as the internal standard. NMR yields of the major isomer are
given. [b] Formation of other side products prevented clear determina-
tion of the E/Z ratios by 1H NMR analysis of the crude reaction mixture.
[c] Concentration: 0.2m. [d] With NaI (4.0 equiv), 0.067m, reflux.
N.D.=not determined.
carbamoyl chlorides (1g and 1r) were significantly less
reactive under the standard conditions. Having additional Cl
atoms on either the aromatic backbone (1d, 1i) or the benzyl
group (1p) did not compromise the reaction. However, the
À
presence of a more reactive Csp2 Br bond (1q) did have
reaction to full conversion by switching to higher temper-
atures and/or catalyst loadings were met with failure. It should
be noted that in all cases, the E isomer, formally resulting
from a trans addition process, was observed as the major
product. In search of a more efficient catalyst, we were
prompted to investigate ligands based on the phosphaada-
mantane core, as its steric bulk and electronic properties are
comparable to those of a PtBu2 group and a P(OR)2 group,
respectively.[13] Furthermore, these ligands are stable to air,
moisture, and column chromatography, making them partic-
ularly attractive for reaction development. After screening
a variety of PA-Ar ligands (Supporting Information, Scheme
S1), PA-Ph proved to be the best, providing 2a in 99% yield
(E/Z > 99:1) after 3 h (entry 4).[14] At higher concentration
and longer reaction times, the catalyst loading could be
decreased to 1.25 mol% (entry 5). Entries 6 and 7 demon-
strate that maintaining a 1:2 Pd/L ratio is crucial for the
reactivity, and that the reaction does not proceed in the
absence of Pd. Notably, applying Tongꢀs conditions[2f] to
substrate 1a led to inferior results (entry 8).
a negative impact on the yield. The reaction efficiency was
minimally impacted upon changing the nature of the
N-protecting group (1l–1o), but having a bulky silyl alkynyl
substituent proved to be necessary (1s vs. 1t). Aryl substitu-
ents were incompatible with the standard reactions conditions
(1u and 1v). In all examples, exclusive selectivity for the
E isomer (> 99:1) was observed. To confirm our stereochem-
ical assignment, an X-ray crystal structure of (E)-2b was
obtained, and the E/Z ratios for all other examples were
determined by comparing analogous chemical shifts in the
1H NMR spectra.[15]
To account for the high trans selectivity, we propose two
possible mechanistic pathways (A or B) and present exper-
imental and computational data that support pathway B
(Scheme 2a).[16] Both catalytic cycles begin with oxidative
addition of Pd0 into the carbamoyl chloride bond. In pathway
A, an in situ released chloride ion (from another molecule of
1a) could trap intermediate I by a trans chloropalladation
À
process, which, upon C C reductive elimination, would give
(E)-2a. To test the feasibility of an ionic pathway, we
conducted the reaction in the presence of exogenous halide
additives (KI or NaI, Scheme 2b).[17] In both cases, 1a was
fully converted into 2a (X = Cl) with no halide exchange
products observed in the 1H NMR spectrum of the crude
reaction mixture, thus disfavoring pathway A. Alternatively,
With our optimized conditions in hand, we then inves-
tigated the substrate scope (Table 2). Substrates with elec-
tron-withdrawing substituents at the 5- (1b, 1c, 1e) or 6-
position (1h, 1j, 1k) were well tolerated. Substrate 1 f
demonstrated excellent reactivity, but more electron-rich
ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 15897 –15900