Angewandte
Communications
Chemie
Table 1: Optimization of the tethered aminoalkynylation.[a]
C).[9b–f] While elegant, both works remain limited to a single
À
C N bond formation.
Recently, our group has demonstrated that vicinal amino
alcohols could be synthesized from allyl amines using an
acetal tether and a palladium catalyst to form simultaneously
[10a]
À
À
new C O and C C bonds.
We envisioned that an allyl
Entry
Changes from standard conditions[b]
Yield [%][c]
amine and an imine could in situ form an aminal, which would
undergo a Pd-catalyzed carboamination reaction to generate
1
2
3
4
5
6
7
8
none
2b instead of 2a
2c instead of 2a
2d instead of 2a
2e instead of 2a
2 f instead of 2a
2g instead of 2a
2h instead of 2a
99 (95)[d]
<3
15
36
5
À
À
an imidazolidine by formation of both a C N and a C C bond
followed by hydrolysis to give the free diamine (Scheme 1,
D).[11] All the reported mixed aminals employed for olefin
functionalization bear electron-withdrawing groups on both
nitrogen atoms, which greatly stabilize them. However, the
poor nucleophilicity of the amine precursor prevents their fast
synthesis.[7] We thought to use nucleophilic allyl amines to
solve this issue. Three challenging conditions had to be met to
ensure the success of the envisioned transformation:
1) formation of the aminal must be complete, fast, and
selective to avoid side reactions or catalyst deactivation;
2) the mixed aminal formed needs to be both stable and
suitable for the envisioned Pd-catalyzed carboamination;
3) the imidazolidine obtained must be stable under the
reaction conditions to avoid catalyst poisoning, but still be
amenable to a straightforward deprotection to free the
diamine.
35
45
5
9
2i instead of 2a
7
10
11
12
13
14
15
16
17
18
19
PPh3 (12 mol%) instead of P(2-furyl)3
PhDavePhos (11 mol%) instead of P(2-furyl)3
DavePhos (11 mol%) instead of P(2-furyl)3
BINAP (6 mol%) instead of P(2-furyl)3
DPEPhos (6 mol%) instead of P(2-furyl)3
XANTPhos (6 mol%) instead of P(2-furyl)3
CsHCO3 instead of Cs2CO3
K2CO3 instead of Cs2CO3
Benzyl allylcarbamate instead of 1
without [Pd2dba3] or P(2-furyl)3 or base
39
81
11
3
90
58
44
89
8
<3
[a] Reactions conditions: 0.10 mmol 1, 0.11 mmol 2a, 0.13 mmol 3,
0.33m in PhMe, 20 h. [b] DavePhos: 2-dicyclohexylphosphino-2’-(N,N-
dimethylamino)biphenyl; PhDavePhos: 2-diphenylphosphino-2’-(N,N-
dimethylamino)biphenyl; BINAP: (1,1’-binaphthalene-2,2’-diyl)bis(di-
phenylphosphine); DPEPhos: bis[(2-diphenylphosphino)phenyl] ether;
XANTPhos: 4,5-bis(diphenylphosphino)-9,9-dimethylxanthene; PMP: p-
methoxyphenyl. [c] Yield determined through NMR spectroscopy using
3,4,5-trichloropyridine as internal standard. [d] Yield of isolated product
on 0.30 mmol scale.
The efficient implementation of this approach is reported
herein, employing electron-rich allyl amines, carbamate-
protected trifluoroaldimines in their stable hemiaminal
form, and a bromide, in the presence of a commercially
available palladium source and phosphine ligands. The
developed conditions allowed for a broad scope in both
allyl amines and alkynyl, aryl, or vinyl bromides, as well as for
high stereo-, regio-, and diastereoselectivity.
We started our studies with allylbenzylamine (1) and silyl
acetylene 3 as efficient electrophilic partner in Pd0/PdII-
catalyzed olefin functionalization (Table 1).[10,12] We identi-
fied aldimine precursor 2a as efficient tethering reagent in
combination with [Pd2dba3] as palladium source, P(2-furyl)3
as ligand, and cesium carbonate as base. Under these
conditions, imidazolidine 4 was isolated in excellent yield
and diastereoselectivity (95%, d.r. > 20:1) (Table 1, entry 1).
Importantly, 2a is easily available on multigram scale,[13]
bench stable for months, and only a slight excess is required
to achieve complete conversion into 4. As anticipated, the
tether was of great importance for the reaction outcome.
Electron-neutral aldimine 2b was ineffective (entry 2), and
even more-activated analogues still suffered from poor
reactivity (entries 3 and 4). Various N-protected trifluorome-
thylaldimines or their precursors did provide some improve-
ment, but furnished inferior results to 2a (entries 5–9).
Changing the ligand had also a profound effect. While using
simple triphenylphosphine decreased the yield (entry 10),
bulky PhDavePhos afforded 4 in high yield (entry 11).
Electron-rich DavePhos was not efficient (entry 12). Regard-
ing bidentate phosphines, BINAP was not competent, DPE-
Phos yielded 90% of 4, and closely related XANTPhos only
58% (entries 13–15). Switching to cesium hydrogen carbon-
ate as base reduced the yield significantly but still afforded
a reasonable amount of 4, while potassium carbonate
provided 89% of the desired imidazolidine (entries 16 and
17). The use of such mild bases is very rare for Pd0/PdII-
catalyzed carboamination reactions, which often require
stronger alkoxides. Deactivating the allyl amine with a carba-
mate protecting group led to very low conversion (entry 18).
Finally, without palladium or ligand or base, 4 was not
observed (entry 19).
The scope of the reaction was then examined (Scheme 2).
We performed an assessment of functional-group tolerance
and electronic effects by varying the substitution on the
nitrogen atoms (Scheme 2, A). Use of a methyl carbamate
protected tether provided product 5 in excellent yield.
Modifying the benzyl group with electron-donating or -with-
drawing groups had only minor influence (products 5–10).
Both the useful aryl bromide (product 8) and chloride
(product 9) were preserved under these conditions. When
the reaction was performed on gram-scale, compound 10 was
obtained in quantitative yield. A simple allyl group was also
tolerated, with no Heck side reaction observed (product 11).
Imidazolidines bearing a furan heterocycle or a ferrocene
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 6
These are not the final page numbers!