Angewandte
Communications
Chemie
Table 1: Screening of catalysts and additives, and the effect of temper-
Table 2: Scope with respect to the malononitrile.[a]
ature.[a]
Entry Catalyst
Additive (2 equiv) Yield 1 [%][b] Yield 2 [%][b]
1
2
3
4
[RhCl(cod)]2
Cs2CO3; 1008C
K3PO4
ZnCl2
93 (89)
25
8
7
8
3
5
5
2
5
71
36
30
55
28
30
[RhCl(cod)]2
[RhCl(cod)]2
[RhCl(cod)]2
[RhCl(cod)]2
Rh(nor)2BF4
Rh(cod)2OTf
[Rh(OH)(cod)]2 K3PO4; 1 equiv H2O
[Rh(OH)(cod)]2 i-Pr2NEt
[Rh(OH)(cod)]2 Proton sponge
CuI
5
K3PO4; 1 equiv H2O
K3PO4; 1 equiv H2O
K3PO4; 1 equiv H2O
6[c]
7[c]
8
68 (55)
90
9
10
11
0
0
98 (95)[d]
95[d]
[RhCl(cod)]2
Proton sponge
[a] Reaction conditions: 0.75 mmol PhB(OH)2, 0.50 mmol DMMN,
1 mol% catalyst and 1 mmol additive in 1 mL dioxane at 808C, 20 h.
[b] HPLC assay yields; values in parentheses are isolated yields.
[c] 2 mol% catalyst. [d] Reaction time of 4 h. cod=1,5-cyclooctadiene,
nor=norbornene, Proton sponge=1,8-bis(dimethylamino)naphtha-
lene.
[a] Reaction conditions: 1.5 equiv PhB(OH)2, 1.0 equiv malononitrile,
1 mol% [Rh(OH)cod]2 and 2.0 equiv proton sponge in 2 mL dioxane at
808C, 6 h; yields of isolated product are given. [b] Reaction was done at
1008C. [c] Reaction time of 24 h.
dioxane, 1008C). Under these conditions, the transnitrilation
pathway is predominant, and 1 and 2 are formed in 93% and
5% assay yields, respectively.[11] It was noted that the use of
K3PO4 as the base resulted in a shift towards favoring the b-
ketonitrile product 2 (71% assay yield, entry 2). We thought
the addition of water to the reaction mixture would facilitate
hydrolysis of the putative iminonitrile adduct and potentially
further increase the amount of b-ketonitrile formation.
However, the inclusion of 1 equiv of water in reactions with
different Rh catalysts resulted in reduced assay yields of 2 and
1 (entries 5–8). The use of [RhOH(cod)]2 as the catalyst
provided the best ratio of 2 to 1 of 68:5 (entry 8). A
breakthrough in both yield and selectivity occurred when
organic amine bases were employed instead of inorganic
bases. When Hꢀnigꢁs base was used (entry 9), the b-ketoni-
trile product 2 was formed in 90% assay yield, while only 2%
assay yield of 1 was detected. The best results were obtained
with a proton sponge. Under these conditions, the reaction
was completely selective for 2. With either [RhOH(cod)]2 or
[RhCl(cod)]2 as the catalyst, the reaction was complete after
just 4 h, and 2 was formed in 98% and 95% assay yields,
respectively (entries 10 and 11). Due to the slightly higher
yield obtained with [RhOH(cod)]2, the conditions of entry 10
were chosen as optimal.
With the optimized reaction conditions in hand, the scope
of b-ketonitrile formation was examined with respect to the
dialkylmalononitrile (Table 2). The reaction was compatible
with both cyclic (3–6) and acyclic (7–10) malononitriles,
giving b-ketonitrile products in yields ranging from 68–93%.
The reaction was sensitive to steric hindrance, since reduced
yields were obtained with acyclic malononitriles relative to
the tied-back cyclic malononitriles. The lowest yield of 68%
was observed for the bulkiest malononitrile 9 (product 17), for
which the reaction time was extended from 6 to 24 h. Notably,
the use of mono-substituted malononitrile 10, which bears an
acidic a-proton, was possible and gave b-ketonitrile 18 in
82% yield.
The reaction scope was next explored with respect to
variation of the aryl boronic acid structure (Table 3). The
presence of p-methyl, p-methoxy, p-phenoxy, or m-methoxy
substituents on the aryl boronic acid resulted in excellent
yields of products 19–23. 2-Methoxy substitution of the
boronic acid was tolerated (product 24), though the reaction
was run at increased temperature. The sterically hindered 2,6-
disubstituted product 25 could also be formed in good yield at
1008C. Halogens such as fluoride and chloride were amenable
(products 26 and 27), as were functional groups such as amine
(28), acetamide (29), ester (30), and ketone (31) groups.
Importantly, the presence of the ester and ketone functional
groups would likely not be tolerated by the nitrile anion
carbonylation method of Skrydstrup and Beller.[10] Finally, the
reaction was possible with a heterocyclic boronic acid, giving
indolyl product 32 in 77% yield.
The utility of the b-ketonitrile products was demonstrated
by diastereoselective additions to the derived tert-butanesul-
finyl ketimines to give b-aminonitriles. The requisite keti-
mines were prepared through condensation of (S)-tert-buta-
nesulfinamide with the b-ketonitrile in the presence of neat
Ti(OEt)4 (Scheme 1).[14] Imines 33–35 were obtained as single
geometrical isomers in good yields.
With the b-ketiminonitriles in hand, the diastereoselective
addition of nucleophiles was explored (Table 4). Reduction
with NaBH4 gave b-aminonitrile 36 in > 97:3 diastereoselec-
tivity and 94% yield. The addition of MeLi proceeded in high
yield and diastereoselectivity, giving the a-quaternary-b-
quaternary aminonitrile 37. The production of such a sterically
hindered system in a diastereoselective manner is, to the best
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!