Organic Letters
Letter
byproducts.8 In this context, alkyl sulfonates are more
attractive than alkyl halides for cross-coupling reactions
because these compounds can be readily prepared from the
easily available alcohols in generally high yields. Inspired by
our previous work, we expected that alkyl sulfonates might also
serve as efficient electrophiles for cyanation reactions. It was
noted that alkyl sulfonates have rarely been used in nickel-
catalyzed cross-coupling reactions,9 possibly due to their low
tendency to engage in a SET process that usually operate with
alkyl halides. In this paper, we report the first example of
nickel-catalyzed cyanation of alkyl sulfonates using Zn(CN)2 as
the cyanide source owing to its relatively lower toxicity
(intraperitoneal, LD50 = 100 mg kg−1,10a compared with
NaCN, KCN,10b intraperitoneal, LD50 = 4.72−5.55 mg kg−1)10
and low solubility in organic solvents which was beneficial to
mitigate catalyst poisoning by cyanide ions (Scheme 1c). The
method also avoids the use of highly toxic reagents such as
MCN (M = Na, K) as the cyanide source.11 Although these
reagents could be used for cyanation of alkyl sulfonates, the
systematic studies on these reactions are quite rare. In some
cases, the reactions suffer from an excess of cyanide,11a,b
serious side reactions,11c and low yields.11d,e
Table 1. Optimization of the Reaction Conditions
Our investigation commenced with the nickel-catalyzed
cyanation of 1-tosylpiperidin-4-yl methanesulfonate 1a as the
model substrate. Initially, the reactions were examined under
the conditions for cyanation of alkyl chlorides or bromides
(NiCl2·6H2O/Xantphos/Zn/DMAP/n-Bu4NCl or Ni(acac)2/-
Xantphos/DMAP/Zn) reported in our previous paper.7a
However, trace or no desired cyanation products were
observed. The effects of nickel catalysts, ligands, reductants,
additives, solvents, substrate concentration, etc. on the reaction
were then studied, and the detailed results are shown in the
the desired reaction took place smoothly to give the alkyl
nitrile 2a in 88% yield at 100 °C using NiCl2·6H2O/Xantphos
as the catalyst system in the presence of 40 mol % zinc powder
(Adamas, 325 mesh), 1.0 equiv DMAP, and 3.0 equiv n-Bu4NI
(Table 1, entry 1). Zinc acts as a reductant to reduce Ni(II). It
was found that the additive of n-Bu4NI played a key role in this
reaction because, without n-Bu4NI, no desired product was
observed (entry 2). The use of n-Bu4NBr as the additive
afforded 2a in low yield, along with the formation of 4-bromo-
1-tosylpiperidine, alkene 4a, and dehalogenated product 6a as
the byproducts (entry 3). The results indicated that alkyl
bromide might act as the active intermediate under this
condition. When n-Bu4NCl was added, the corresponding alkyl
chloride was formed in 84% yield (entry 4). No product was
found in this case, possibly due to the lower reactivity of the
alkyl chloride toward cyanation compared with alkyl bromide.
Based on these results, we assumed that alkyl iodide might be
produced in the presence of n-Bu4NI,9d,12 which should be
more reactive than alkyl bromides or alkyl chlorides. n-Bu4NI
might also act as a phase transfer reagent to promote the
cyanide ion entering into the organic phase. When KI was
employed instead of n-Bu4NI, the yield of 2a decreased to 41%
(entry 5). Reducing the amount of n-Bu4NI to 2.0 equiv
caused an erosion in yield (entry 6). Without DMAP, 2a was
formed in 67% yield, along with the formation of alkene 5a and
dehalogenated product 6a, indicating that DMAP could
effectively reduce the side reactions (entry 7 vs entry 1).
DMAP may also promote cyanide ion dissociation through the
formation of a DMAP-Zn(CN)2 complex and acts as a
coligand as well.7b The use of 4-aminopyridine or Cs2CO3
a
Determined by 1H NMR using 1,3,5-trimethoxybenzene as an
b
c
internal standard. Isolated yield. 4-Bromo-1-tosylpiperidine was
formed in 28% yield. 4-Chloro-1-tosylpiperidine was formed in 84%
yield.
d
instead of DMAP afforded 2a in similar yields (82−85%,
entries 8−9). NiCl2(DME) or NiCl2 could also catalyze this
reaction, albeit with lower yields of 60−72% (entries 10 and
11). Other bidentate phosphine ligands such as dppf gave 2a in
74% yield, while the use of PCy3 or 2,2′-bpy was not effective
(entries 12−14). Control experiments indicated that NiCl2·
6H2O, Xantphos, and Zn were indispensable (entries 17−19).
Alkyl tosylate was also compatible for this reaction, while no
desired products were observed using alkyl acetate or benzoate
as the substrates (entries 20 and 21).
We next studied the substrate scope of this new cyanation
reaction under the best reaction conditions (Table 1, entry 1).
The results are shown in Scheme 2. Boc- or CO2Et-protected
piperidin-4-yl methanesulfonates could be cyanated efficiently
in 88−90% yields (2b−2c). Cyclic substrates bearing benzene-
fused five- or six-membered rings coupled smoothly with
Zn(CN)2 (2d−2e). Substrate 1f containing a ketal moiety
could be successfully converted to 2f in 82% yield. Cyanation
of 1g bearing a large-sized ring afforded 2g in low yield (24%).
When the optically pure alkyl sulfonate 1h derived from the
nature product was used as the substrate, a mixture of two
diastereomers (2h) was produced. The results revealed that
the radical species might be generated during the reaction.
Next, the reactivity of acyclic alkyl mesylates was investigated.
B
Org. Lett. XXXX, XXX, XXX−XXX