Full Paper
portant yield enhancement, especially in the benzyl series, for
which the yield of 26a rose from 38 to 77%. Thus, although
the reaction time was lengthened, fewer byproducts were gen-
erated in the course of the reaction.
Finally, replacement of the RhI catalyst [Rh(Cl)(CO)2]2 with
[Rh(cod)Cl]2, always in the presence of 50 mol% of dppp (the
conditions of Mukai and co-workers),[23] brought about signifi-
cant improvement in the PKR and resulted in the formation of
25b in quantitative yield (Cb series; Table 1, entry 7) and 26a
in 90% yield (Bn series; Table 2, entry 8). These results suggest
that the nature of the catalyst appears to be an underlying
factor of the reaction.[14,26]
Our study then focused on the influence of solvent and
ligand nature on the reaction. Replacement of toluene by THF
or acetonitrile did not improve the results, with the catalyst
complex being insoluble in these solvents (Table 1, entries 5, 6,
and 8; Table 2, entry 3). If a cationic rhodium complex was
used, that is, [Rh(Cl)(CO)2]2/dppp or [Rh(cod)Cl]2/dppp in the
presence of AgOTf, inconclusive results were observed with
substrate 13a (R1: OBn; Table 2, entries 7 and 9).[27] In the
benzyl series, the reaction was next explored in the presence
of various phosphorus ligands (phosphines, hydroxyphos-
phines, or phosphites: Xantphos, PO, and P(OEt)3) but with
inconclusive outcomes (Table 2, entries 5, 6, and 10).[28]
Scheme 8. Preparation of alkoxyallene-ynes in oxygen series. a) NaH, prop-
argyl bromide, THF, 08C to RT, then reflux, 12 h, 95%. b) nBuLi, TMSCl, THF,
À788C, 45 min, quant. c) DDQ, CH2Cl2/H2O, 08C to RT, 12 h, 87%. d) (COCl)2,
DMSO, Et3N, CH2Cl2, À788C to RT, 92%. e) nBuLi, THF, À788C, 30 min, then
aldehyde 22, THF, À788C, 30 min. f) TBSOTf, 2,6-lutidine, CH2Cl2, 08C, 3 h.
g) K2CO3, MeOH. DDQ: 2,3-dichloro-5,6-dicyano-1,4-benzoquinone.
It should be noted that bicyclic [5,6] products resulting from
the PKR between the vicinal double bond of the alkoxy-allene
and the alkyne function have never been observed.
With the feasibility of the alkoxyallene-yne Pauson–Khand
cyclocarbonylation established and the optimized reaction
conditions in hand (that is, [Rh(cod)Cl]2 (10 mol%) modified
with dppp (50 mol%), in toluene heated to reflux under 1 atm
of CO), we set out to explore the scope of this reaction for the
preparation of a number of perhydroazulene systems. Interest-
ingly, in the carbamate and benzyl series (R1: Cb or Bn), the
treatment of alkynes 12a, 12c–e, and 13b (R2: H, Et, TMS, or
CH2OBn) under optimal conditions provided the unsaturated
bicyclo[5.3.0] ring systems 25a, 25c–e, and 26b in similar
yields, which illustrates the high tolerance of this cycloaddition
reaction (Scheme 9 and Scheme 10). Moreover, extension of
the repertoire of allenes to the PMB series (R1: PMB) also per-
Table 1. Optimization of the alkoxyallene-yne Pauson–Khand reaction
with 12b (R1: Cb; R2: Me).
Catalyst/ligand
Solvent
T
t
Conversion Yield of
[8C] [h] [%]
25b [%]
1
2
3
4
5
6
7
8
[Rh(Cl)(CO)2]2/–
[Rh(Cl)(CO)2]2/–
[Rh(Cl)(CO)2]2/–
toluene
toluene
THF
90
RT
110
110
1.5 100
70
70
70
88
–
8
2
3
100
100
100
0
[Rh(Cl)(CO)2]2/dppp[a] toluene
[Rh(Cl)(CO)2]2/dppp
[Rh(Cl)(CO)2]2/dppp
[Rh(cod)Cl]2/dppp
[Rh(cod)Cl]2/dppp
acetonitrile 110 20
THF
toluene
THF
110 20
110 20 100
110 20
0
–
quant.
–
0
[a] dppp: 1,3-bis(diphenylphosphino)propane.
the desired products 25b and 26a with full conver-
sion within 1.5 h. Notably, 25b was isolated in good
yield (70%; Table 1, entry 1), whereas 26a was ob-
tained in moderate yield (38%), along with byprod-
ucts, which indicated stability problems (Table 2,
entry 1). If the temperature was lowered to 258C, the
reaction of 12b (Cb series) interestingly reached
completion in 8 h (Table 1, entry 2), whereas only
10% conversion was observed after 20 h in the
benzyl series (Table 2, entry 2). Alternatively, if the re-
action was performed in THF, there was no improve-
ment whatever in the series (Tables 1 and 2, entry 3).
Exposure to 10 mol% of [Rh(Cl)(CO)2]2 modified
with dppp (50 mol%) in toluene heated to reflux
under CO (1 atm; Table 1 and 2, entry 4) led to an im-
Table 2. Optimization of the alkoxyallene-yne Pauson–Khand reaction with 13a (R1:
Bn; R2: Me).
Catalyst/ligand
Solvent
T
t
Conversion Yield of
[8C] [h]
[%]
26a [%]
1
2
3
4
5
6
7
8
9
10
[Rh(Cl)(CO)2]2/–
[Rh(Cl)(CO)2]2/–
[Rh(Cl)(CO)2]2/–
[Rh(Cl)(CO)2]2/dppp
[Rh(Cl)(CO)2]2/P(OEt)3
[Rh(Cl)(CO)2]2/PO[a]
[Rh(Cl)(CO)2]2/dppp, AgOTf toluene
[Rh(cod)Cl]2/dppp
toluene
toluene
THF
toluene
acetonitrile 110
90
RT
110
110
1.5 100
38
–
20
2
6
4
4
10
100
100
100
100
20
77
24
52
–
90
–
–
THF
110
110
110
110
110
1.5 20
toluene
toluene
THF
28
20
20
100
0
0
[Rh(cod)Cl]2/dppp, AgOTf
[Rh(cod)Cl]2/Xantphos
[a] PO: 1-(2-(diphenylphosphanyl)phenyl)ethan-1-ol.
Chem. Eur. J. 2016, 22, 4938 – 4944
4941
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim