.
Angewandte
Communications
entries 5 and 6). However, with Pd(OAc)2 and [Pd(acac)2],
only traces of product could be observed, with low conversion
of 1a (Table S1, entries 7 and 8). By using PdI2 as the
palladium source, the reaction even proceeded at 808C,
although giving a lower yield (Table S1, entry 9). In general,
the catalyst system was shown to be robust and almost the
same yield and selectivity (92%, 89%; Table S1, entry 11)
was achieved at a lower catalyst loading (0.25 mol%). To our
surprise, the addition of 0.75 mol% p-toluenesulfonic acid led
to a significant decrease in the yield of the product (64%
yield; Table S1, entry 13). Monitoring the gas consumption
showed that the reaction reached over 90% conversion after
10 h (see the Supporting Information for details).
With the optimized conditions in hand, a range of easily
available and structurally diverse olefins were tested
(Table 1). Primarily, 1-decene (1b) gave a very similar result
to 1-octene (1a, Table 1, entries 1 and 2; 87% and 88% yield
of isolated 3, 89% n-selectivity). To our delight, different
cyclic olefins (1 equiv) also reacted smoothly. For example,
the reactions with cyclopentene (1c) and cyclohexene (1d)
resulted in complete conversion and good yields of the
isolated products (Table 1, entries 3 and 4; 83% and 90%
yield, respectively). With norbonene (1e) as substrate, the
reaction was shown to be completely exo-selective, with the
imide moiety in the equatorial position (58% yield; Table 1,
entry 5). Interestingly, the basic industrial feedstock ethylene
(1 f) gave 3 fa in 91% yield (Table 1, entry 6). Excellent linear
selectivities were observed in the case of sterically hindered 1-
alkenes. For example, 4-methyl-1-pentene (1g) and allylcy-
clohexane (1h) led to yields of 85% and 86% of isolated
product with an n-selectivity of 89% for 3ga and 3ha
(Table 1, entries 7 and 8). With 4-vinylcyclohexene (1i) as
substrate, 97% linear selectivity was obtained with a 90%
yield of the isolated product (Table 1, entry 9). It is note-
Scheme 2. Hydroamidocarbonylation of 1-octene. Reaction conditions:
Ligand variation. 2a (1.0 mmol), 1a (1.0 mmol), PdCl2 (1 mol%),
monodentate ligand (4 mol%) or bidentate ligand (2 mol%), toluene
(2 mL), CO (50 bar) heating at 1058C for 20 h. Yield (%) of the
mixture of 3aa and 3aa’ determined by GC using isooctane as an
internal standard, the number in parenthesis indicates the 3/3’
selectivity determined by GC. N.R. indicates no observable product by
GC.
respectively). Notably, the best result (61% yield, 84%
selectivity) was observed using DPEPhos [(oxydi-2,1-phenyl-
ene)bis(diphenylphosphine)] (L11). Thus, L11 was chosen for
further studies. Similar to the Xantphos ligand, the sterically
hindered cyclohexyl and tert-butyl analogues of DPEPhos
(L12 and L13) were shown to be less reactive (14% yield,
75% selectivity and no conversion, respectively). Interest-
ingly, the so-called dtbpx ligand a,a’-bis(di-tert-butylphos-
phino)-o-xylene (L14) , which is known to be highly active for
palladium-catalyzed carbonylation reactions of alkenes and
for the aminocarbonylation of butadiene was shown to be
completely inactive in this case.[13,7c]
In contrast to most of the recently developed palladium-
catalyzed carbonylation reactions of olefins, the presented
imide formation takes place under “acid-free” conditions.
Under such conditions, slow isomerization of 1-octene leads
to an accumulation of internal octenes, which are not in fast
equilibrium with the terminal olefin. Hence, the use of an
increased amount of olefin (2 equiv) resulted in the yield of
the desired terminal imide increasing to 85% with similar n-/
iso selectivity (84%). It is noteworthy that carbonylation
products derived from internal olefins are not observed under
the “acid-free” conditions, which shows that the reaction rates
of the terminal alkenes are much faster than the internal ones.
Next, the effect of the counterion was investigated by
testing different palladium sources (see Table S1 in the
Supporting Information). [Pd(cod)Cl2] and [Pd(CH3CN)2Cl2]
resulted in similar linear selectivity, however with slightly
decreased yields (Table S1, entries 3 and 4). Surprisingly,
almost quantitative yields and an increased linear selectivity
of the reactions were observed with both PdBr2 and PdI2 as
the palladium source (85% and 88%, respectively; Table S1,
=
worthy that in this case the internal C C bond remained
intact after reaction, thus showing that the terminal alkene is
much more reactive in this reaction than the cyclic internal
alkene. Natural oil derived substrate citronellene (1j) gave
a 76% yield of 3ja with an n-selectivity of more than 99%.
The trisubstituted internal double bond also stayed intact in
this case (Table 1, entry 10). To our delight, the optimized
conditions are also compatible with more bulky substrates,
such as 1k and 1l, which resulted in 55% and 60% yields,
respectively, with complete linear selectivity (Table 1,
entries 11 and 12). The electronic effect of substituted
allylbenzenes was also studied. The reaction between allyl-
benzene and N-methylacetamide went to complete conver-
sion at a slightly elevated temperature (1208C), and the
corresponding linear imide 3ma was isolated in 88% yield
with an n-selectivity of 90% (Table 1, entry 13). With 1-allyl-
4-fluorobenzene (1n) as the substrate, the reaction was
complete at 1058C and afforded 3na in 85% yield and 88%
selectivity (Table 1, entry 14). Finally, styrene was successfully
employed as an example of an aromatic olefin (Table 1,
entry 15; 63% yield and 70% n-selectivity). This system was
further demonstrated to tolerate Br and Cl substituents on
styrene derivatives. Under the optimized reaction conditions,
low conversions were observed (28% and 31% yield of
isolated product for 3pa and 3qa, respectively). By increasing
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 10239 –10243