COMMUNICATION
Table 1. Optimization of the coupling reaction of 1 with allyl carbonate 2a.[a]
hydro-dehalogenation (ꢀ10%), b-hydride elimina-
tion (ꢀ10%), and homocoupling (ꢀ10%). The
side reactions, however, could be significantly sup-
pressed by employing a stoichiometric amount of
MgCl2 as the sole additive (Table 1, entries 16–19).
A
combination of 1 equivalent of MgCl2 and
15 mol% of 6 gave the highest yield (91%). The
amount of 6 could be further reduced to 10 mol%
without diminishing the yield by utilizing 1.5 equiv-
alents of MgCl2 (Table 1, entry 19).
By using the optimized conditions for 2a
(Table 1, entries 4 and 19), a diverse set of substitut-
ed allylic carbonates 2b–2m were examined for the
coupling reaction with 1a and 1b (Table 2). For the
3-alkyl-substituted allyl substrates 2b (trans) and 2c
(cis), moderate to good yields were obtained with
1a (Table 2, entries 1 and 3), whereas excellent
yields were observed with 1b (Table 2, entries 2 and
4). It should be noted that only the trans-product
3c was isolated for the reaction of cis-3-ethyl car-
bonate 2c (Table 2, entries 3 and 4). For trans-3-
phenyl carbonate 2d, both 1a and 1b delivered the
allylated product 3d in high yields, despite the fact
that 1b appears to be more efficient (Table 2, en-
tries 5 and 6). The allylic carbonates bearing 2-
methyl groups, as in 2e and 2 f, generally displayed
much lower reactivity, giving poor results for both
1a and 1b (Table 2, entries 7–10). The 2-phenyl car-
bonate 2g, however, behaved differently from the
2-alkyl-substituted analogues, providing moderate
and excellent yields for 1a and 1b, respectively
(Table 2, entries 11 and 12). The coupling reactions
of 1a and 1b with 1-alkyl-substituted allyl sub-
strates 2h–2k were generally highly efficient
X
Ligand
([mol%])
R
Y
Additive
([mol%])
T
[8C]
t
Yield
[%][b]
G
E
[h]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
I
I
I
I
4a (15)
4a (15)
4a (15)
4a (15)
4b (15)
4a (15)
4a (15)
4a (15)
4b (15)
4c (15)
4d (15)
4e (15)
5 (15)
iPr
iPr
iPr
iPr
H
iPr
iPr
iPr
H
Me
sBu
tBu
–
iPr
–
H
H
H
H
H
H
H
H
H
H
H
H
–
none
CuI (20)
CuI (50)
CuI (30)
CuI (20)
none
none
CuI (20)
CuI (20)
CuI (20)
CuI (20)
CuI (20)
25
25
25
35
25
25
80
80
80
80
80
80
80
80
80
80
80
80
80
12
12
6
67
82
82
82
5
6
I
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
n.d.[c]
7
45
10
20
40
n.d.[c]
35
CuI (20)
4a (15)
6 (15)
4a (15)
4 f (15)
6 (15)
H
–
H
OMe
–
CuI (20)/MgCl2 (1:1)
CuI (20)/MgCl2 (1:1)
MgCl2 (100)
MgCl2 (100)
MgCl2 (100)
MgCl2 (150)
65
65
76
68
91
91
iPr
iPr
–
6 (10)
–
–
[a] Reaction conditions: 1 (100 mol%, 0.15m in DMA), 2 (200 mol%), [NiACTHNUTRGNE(UNG cod)2]
(10 mol%), Zn (300 mol%), DMA (1 mL). Ligand, additive, temperature, and reac-
tion time are variables. [b] Isolated yields. [c] n.d.=not detected.
substrates bearing leaving groups, such as acetate and tBu-
carbonates, resulted in low yields; no products were ob-
tained for allyl bromide and phenyl sulfonate reagents.
When 4-bromo-1-tosylpiperidine (1b) was subjected to
(Table 2, entries 13–20). Poor results were delivered for 1-
phenyl allylic carbonate 2l due to extensive homocoupling
(Table 2, entries 21–22). Finally, the sterically more congest-
ed 1,3-diethyl compound 2m provided 3m in poor yields
(Table 2, entries 23 and 24).
the [NiACHTUNGTRENNUNG(cod)2]/4a/Zn/DMA conditions at 258C, only hydro-
dehalogenation and b-hydride elimination of 1b were ob-
served (Table 1, entry 6). Raising the temperature produced
3a in a highest yield of only 7% at 808C (Table 1, entry 7
and Table S1 in the Supporting Information).[17] The addi-
tion of CuI (20 mol%) boosted the yield to 45% (Table 1,
entry 8 and Table S2 in the Supporting Information).[17] Ex-
tensive examination of ligands (Table 1, entries 9–13) and
solvents (Table S3 in the Supporting Information) did not
result in any further improvements.[17] However, the addition
of anhydrous MgCl2 (20 mol%) promoted the yield to 65%
by using 4a as the ligand (Table 1, entry 14),[18] which is
comparable to the yield found when using tBu-Terpy (6;
Table 1, entry 15) as the ligand. Further screening of differ-
ent Ni sources and reductants did not provide better results
(Tables S4 and S5 in the Supporting Information). Under
the improved reaction conditions (Table 1, entry 14), 1b was
consumed after 8 h, whereas a substantial amount of 2a was
recovered; the byproducts derived from 1b were from
Interestingly, extensive screening of other ligands for the
example reactions in Table 2 indicated that the coupling effi-
ciency for 1a with 3-alkyl and 2-aryl allylic carbonates, and
1b with 2-alkyl and 3,3’-dimethyl allylic carbonates could be
drastically improved by using ligand 4g (Table 3). The gen-
erality of these scenarios was evident for 2b, 2c, and 2n
(Table 3, entries 1–3), 2e, 2 f, and 2o (Table 3, entries 4–6),
and 2g and 2p (Table 3, entries 7 and 8).
Notably, all reactions examined in Tables 2 and 3 gave E-
alkenes only. The excellent regioselectivities (>20:1) for un-
symmetrical allylic substrates were derived from the addi-
tion of the alkyl group to the sterically less hindered allylic
terminus. In general, sterically more encumbered allylic sub-
strates, such as 2m, led to lower yields. When the allylation
was less efficient, hydro-dehalogenation and homocoupling
of 1a or 1b were usually the major factors accounting for
poor results, with the latter being the main side reaction.
Chem. Eur. J. 2012, 18, 808 – 812
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
809