propargylic displacements with silicon nucleophiles invol-
ving the transmetalation of interelement linkages. A semi-
was also high for phosphate (γ:R = 90:10). The pro-
pargylic bromide reacted, however, with poor selectivity
(γ:R = 67:33) (Table 1, entries 1ꢀ3). The remaining
common oxygen leaving groups all favor R substitution,
and γ:R ratios are in fact good for carbamate and benzoate
(Table 1, entries 4 and 6). It is noteworthy that, compared
to Sawamura’s investigations,9,12a,13 the carbonate leaving
group yielded a poor γ:R ratio (Table 1, entry 5). Chemical
yields were generally lower for oxygen leaving groups than
those for chloride and bromide (quantitative yield).
ꢀ
nal paper by Szabo et al. showed that palladium(II) pincer
complexes indeed catalyze the heterolytic cleavage of a
SiꢀSn bond (left, Scheme 1), and the thus-formed reactive
PdꢀSi intermediate participates in the γ-selective substitu-
tion of propargylic chlorides.8 Later, Sawamura et al.
adopted our rhodium(I)-catalyzed SiꢀB transmetalation
(yet without added water) and elaborated a practical
functional-group-tolerant allenylic silane synthesis with a
carbonate leaving group (middle, Scheme 1).9 In this
Letter, we demonstrate that our straightforward reaction
setup for copper(I)-catalyzed SiꢀB bond activation6 is
applicable to the γ-selective substitution of propargylic
chlorides as well as phosphates (right, Scheme 1).10ꢀ13
Table 1. Copper-Catalyzed Propargylic Substitution: Survey of
Leaving Groups
Scheme 1. Transition-Metal-Catalyzed Interelement Activation
in Propargylic Substitution with Nucleophilic Silicon
propargylic
precursor
leaving group
X
γ:R
yield
(%)b
entry
ratioa
1
2
3
4
5
6
1a
2a
3a
4a
5a
6a
Cl
100:0c
67:33
90:10
6:94
94
98
66
42
49
47
Br
OP(O)(OEt)2
OC(O)NHPh
OC(O)OMe
OC(O)Ph
24:76
5:95
a Ratio of regioisomers determined by GLC analysis prior to pur-
ification. b Combined isolated yield after flash chromatography on silica
gel. c No linear regioisomer detected by GLC analysis.
The leaving group-dependent propargylic substitution
provides an access to both allenylic (γ-selectivity) and
propargylic (R-selectivity) silanes in synthetically useful
As we continued using the protocol for the allylic
substitution,6 we immediately began with a survey of
leaving groups (Table 1). Again, CuCN (5.0 mol %) and
NaOMe (2.0 equiv) in THF at ꢀ78 °C were routinely used,
but this time, only a slight excess of Suginome’s Me2Ph-
SiꢀBpin reagent14 (1.2 equiv as opposed to 1.5 equiv) was
necessary. We were delighted to find that the γ:R ratios in
the propargylic substitution largely parallel those obtained
in the allylic transposition. The chloride leaving group
secured perfect regiocontrol (γ:R= 100:0), and γ-selectivity
0
γ:R ratios. With our focus on SN -type substitution, we
extended the substrate scope for propargylic chlorides
(1aꢀ1h, Table 2). All aryl- and alkyl-substituted precursors
were cleanly converted into allenes (Table 2, entries 1ꢀ6).
The parent compound, propargylic chloride, also yielded
the allene exclusively (Table 2, entry 7). In agreement with
our previous findings,6 the γ:R ratio was completely eroded
by a terminal Me3Si group (Table 2, entry 8).
While the chloride leaving group emerged as superior, it
would not be useful in enantioselective displacements as
enantioenriched R-chiral propargylic chlorides are not
available. Instead, R-chiral phosphates are easy to make,
and the γ:R ratio was also promising (Table 1, entry 3).
We therefore prepared the R-chiral propargylic phos-
phates (S)-3i (R = Ph) and (S)-3j (R = n-Bu) from the
known corresponding enantiopure alcohols, obtained by
enzymatic kinetic resolution. Subjecting those to our stan-
dard protocols afforded the chiral allenes with superb γ:R
ratios [(S)-3i/(S)-3jf(aR)-γ-7i/(aR)-γ-7j, Scheme 2].15
Gratifyingly, the central-to-axial chirality transfer wasalso
ꢀ
ꢀ
(8) Kjellgren, J.; Sunden, H.; Szabo, K. J. J. Am. Chem. Soc. 2005,
127, 1787–1796.
(9) Ohmiya, H.; Ito, H.; Sawamura, M. Org. Lett. 2009, 11, 5618–
5620.
(10) For an uncatalyzed reaction of terminally metalated propargyl
substrates with Me2PhSiꢀBpin, see: Shimizu, M.; Kurahashi, T.;
Kitagawa, H.; Hiyama, T. Org. Lett. 2003, 5, 225–227.
(11) We think that the current investigation is closely connected to
the recent progress in CuꢀO-mediated transmetalation of SiꢀH and
BꢀB bonds. The resultant CuꢀH and CuꢀB nucleophiles undergo
γ-selective propargylic reduction12 (with carbonate12a or acetate12b leaving
groups) and borylation13 (with carbonate leaving group).
(12) (a) Zhong, C.; Sasaki, Y.; Ito, H.; Sawamura, M. Chem. Com-
mun. 2009, 5850–5852. (b) Deutsch, C.; Lipshutz, B. H.; Krause, N. Org.
Lett. 2009, 11, 5010–5012.
(13) Ito, H.; Sasaki, Y.; Sawamura, M. J. Am. Chem. Soc. 2008, 130,
15774–15775.
(14) Suginome, M.; Matsuda, T.; Ito, Y. Organometallics 2000, 19,
4647–4649.
(15) It is important to note that we had also tested the cognate (R = Ph)
carbonate 5i and benzoate 6i but both showed no conversion.
Org. Lett., Vol. 13, No. 16, 2011
4463