R
SYNTHETIC COMMUNICATIONSV
3
organic synthesis. Recent years have witnessed an increasing interest in decarbonylative
coupling reactions using various carbonyl functionalities (acyl chlorides, anhydrides,
esters, amides, aldehydes, and ketones) as counterparts.[33] With our continuous efforts
in Pd-catalyzed C–N bond-forming reactions for the assembly of functional mole-
cules,[34] we were interested in decarbonylative coupling reactions of readily accessible
carbamoyl chlorides for the C–N formation.[35]
Results and discussion
Methyl(phenyl)carbamic chloride 1a was selected as a model substrate for a palladium-
catalyzed decarbonylative C–N coupling reaction to construct the symmetrical tetrasub-
stituted urea 2a. The blank experiment (without the catalyst) of 1a was examined in
1,4-dioxane at 100 ꢁC for 48 h using Na2CO3 or KHCO3 as the base, and no desired
product 2a was obtained (entry 1, Table 1). Fortunately, we found that 1a can be trans-
formed into the desired product 2a in 86% yield in the presence of 5 mol% Pd(PPh3)4
(entry 2, Table 1). Thus, to optimize reaction conditions, different catalysts, bases, sol-
vents, reaction time, and temperature were examined, as shown in Table 1. First, the
effect of bases (such as Na2CO3, Cs2CO3, CsF, K2CO3, NaHCO3, KOAc, K3PO4,
KHCO3, NaOH, KOH, CH3ONa, KOt-Bu, and NEt3) on the reaction was investigated
using 5 mol% Pd(PPh3)4 as the catalyst (entries 2–14, Table 1). It was found that the
nature of base greatly influenced the outcome of the reaction, weak inorganic bases pro-
vided better yields (86–95%, entries 2–9) than strong inorganic bases and organic base
(73–79%, entries 11–15). Among all the weak bases examined, KHCO3 seemed to be
more effective and gave 2a in 95% yield (entry 9). It was also worth noting that the
reaction provided the worse results in the absence of KHCO3 or N2, 2a was obtained in
45% and 10% yields, respectively (entry 9, Table 1). In addition, a higher ratio of base
to substrate is preferable to afford a better result [80% (1:1), and 95% (2:1); entry 9].
When the amount of catalyst was reduced to 2 mol% and 1 mol%, the desired product
2a was obtained in 82% and 73% yield, respectively (entry 10). Second, a survey of reac-
tion media showed that weak polar solvents generally provided better results (95%, 1,4-
dioxane, entry 9; 80%, acetonitrile, entry 21) than strong polar solvents (trace, DMSO;
39%, DMF; 67%, DMA; 72%, NMP; entries 16–19) and non-polar solvent (64%, toluene,
entry 20). The effect of palladium sources on the reaction was next investigated,
Pd(PPh3)4 was found to give the best result (95% yield, entry 9), and 1a can also be
converted into 2a in high yields using a phosphine-free palladium catalyst (entries
22–24). Finally, a series of experiments were conducted to reveal the role of reaction
time and temperature on the reaction (entries 25–29). Generally, higher yields were
obtained with the reaction time increasing; however, when the reaction temperature is
over 100 ꢁC, the yield gradually decreased with the increase of temperature.
Having established the feasibility of symmetrical urea synthesis via palladium-cata-
lyzed decarbonylative C–N coupling reaction, we then explored the scope and limitation
of carbamoyl chlorides under the standard conditions (Table 2). As shown in Table 2, a
variety of diversely substituted groups (such as alkyl, -OMe, -OPh, -F, -Cl, -CN, -CF3,
-COOMe, -SMe, -NMe2, morpholino) on the aromatic moiety of anilines were applic-
able, and the corresponding ureas 2fꢀ2z can be obtained in moderate to high yields