2
Table 2. values of oxocyclobutenyl ligands of 2a–2e.*
Complexes 2
(13C-NMR)/ppm
Ar
p-MeOC6H4− 182.9 153.3 207.6 53.6
reaction mixture, and the simply coupled compounds 5d and
5e were obtained as the major products (Entries 4 and 5,
respectively). Furthermore, it is worth mentioning that alco-
holic products resulting from the attack of organozincs on
the carbonyl carbon of the cyclobutenone ligand were not
observed under these reaction conditions,.
C1
2
L
C2
C3
C4
2a PPh3
2b PPh3
2c PPh3
p-ClC6H4−
183.0 153.8 211.2 53.8
183.1 153.8 211.2 53.8
C6H5−
The satisfactory results obtained from the stoichio-
metric reactions prompted us to explore the feasibility of the
catalytic version of the reaction of propargyl chlorides 1, CO,
and terminal alkynes 3. A plausible catalytic cycle is illus-
trated in Scheme 1. First, oxocyclobutenyl Pd 2 is generated
from 1, CO, and Pd(0) species via allenyl/propargyl Pd spe-
cies (6/7).6 Second, the in situ-generated copper(I) acetylide
would substitute the chloride ligand of 2. Finally, reductive
elimination would afford 2,3-disubstituted cyclobutenones 4
as the product with the concomitant release of the Pd(0) spe-
cies.
2d PMePh2 p-MeOC6H5− 183.3 153.7 204.1 54.8
2e PPh3
p-MeC6H4− 183.0 153.7 209.6 53.6
* Measured in CDCl3 at 125 MHz.
Table 1 summarizes the results obtained from the So-
nogashira-type coupling reactions between complexes 2a–2d
and terminal alkynes 3a and 3b. It is evident from the table
that the isolated, non-optimized yield of 4a–4d varied from
29% to 64%. Products 4a–4d were sufficiently stable to be
handled under air, even in solution state at room temperature
for several days. Table 2 summarizes the 13C-NMR values
of the oxocyclobutenyl ligands in complexes 2a–2e. These
values are scarcely influenced by the aryl substituents or the
phosphine ligands, suggesting that there are little differences
in the electronic states of the complexes. Therefore, the rela-
tively good yield of complex 2d (64%; Entry 5) may arise
not from its electronic state but from its sterically less-
crowded environment.
A similar coupling was found to occur between com-
plexes 2 and organozincs [Eq. (4)]. When complex 2c (Ar =
C6H5−) was treated with 1.1 equivalent of phenylzinc bro-
mide in THF at room temperature for
3 h, 2,3-
diphenylcyclobutenone (5c) was obtained in 19% isolated
yield based on complex 2c (Table 2, Entry 3).7
Scheme 1: Possible catalytic cycle for the conversion of 1
into cyclobutenone 4.
(4)
A revision on the CO pressure required for the genera-
tion of complex 2 was performed before attempting the cata-
lytic reaction [Eq. (2)]. Complex 2c was prepared in 75%
and 89% isolated yields from [Pd(PPh3)4] and 1a (Ar =
C6H5−) under 4 and 20 atm of CO, respectively. Based on
this result, we set 4 atm of CO as the reaction pressure.
Table 3. Coupling between complexes 2 and organozincs.
Entry 2
Ar
Organozinc
R
5
Y/%*
33
1
2
3
4
5
2a p-MeOC6H4− Ph-ZnBr C6H5− 5a
2b p-ClC6H4− Ph-ZnBr C6H5− 5b
2c
2c
31
19
74
40
C6H5−
C6H5−
Ph-ZnBr C6H5− 5c
Et2Zn
Et2Zn
C2H5− 5d
C2H5− 5e
2e p-MeC6H4−
(5)
* Isolated yield of 5 based on 2. Not optimized.
Table 3 summarizes the results of the Negishi-type
coupling reactions. The results show that the isolated yield of
5a–5e varied from 19% to 74%. The reaction mixture
changed its color from light yellow to black immediately
after the addition of the organozinc solution. Based on this
appearance change, the reaction appeared to proceed quite
smoothly, with negligible amount of 2 left. Because cyclobu-
tenones 5a–5e were thermally stable, as mentioned for 4a–
4d, the low-to-moderate yield of 5a–5e may be attributed to
their low chemical stability in the presence of coexisting
chemical species.
We chose propargyl chloride 1a and p-ethynyltoluene
(3b; R = p-CH3C6H4−) as the model reactants to synthesize
cyclobutenone 4d using the catalytic system [Eq. (5)]. In the
currently optimized conditions, 1a (2 mmol), NEt3 (2 mmol),
[Pd(PPh3)4] (0.1 mmol), CuI (0.2 mmol), and acetonitrile (20
mL) as a solvent were stirred in an autoclave with pressur-
ized CO (4 atm) at room temperature. Subsequently, a solu-
tion of 3b (2.5 mmol) in acetonitrile (10 mL) was gradually
added into the pressurized autoclave using an HPLC pump
for 5 h, and the mixture was stirred for an additional 20 h.
After washing and chromatographic separation, 4d was iso-
lated in 20% yield from 1a (Table 4, Entry 2). This result
clearly shows that the reaction was catalytic on the Pd atom
(TON value is approximately 4).
We first anticipated that the reaction of 2 with diethyl-
zinc might afford product 5, with the R group being hydro-
gen as a result of the possible -hydrogen elimination of the
ethyl group. However, no such products were found in the