A robust palladium-catalyzed cyanation procedure: Beneficial effect of zinc acetate

Article abstract of DOI:10.1016/j.tetlet.2003.12.040

Conditions that ensure a more reproducible Pd(0)-mediated procedure to convert aryl bromides to aryl cyanides have been developed that entail the use of zinc dust to keep the Pd in the zero oxidation state and zinc acetate to ensure high catalytic activity. This procedure is applicable to a wide range of substrates and obviates the need for stringent removal of oxygen from the reaction medium.

Full text of DOI:10.1016/j.tetlet.2003.12.040

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 1441–1444
A robust palladium-catalyzed cyanation procedure:
beneficial effect of zinc acetate
Ramakrishnan Chidambaram*
Process Research and Development, Bristol–Myers Squibb, New Brunswick, NJ 08908, USA
Received 1 December 2003; revised 9 December 2003; accepted 9 December 2003
Abstract—Conditions that ensure a more reproducible Pd(0)-mediated procedure to convert aryl bromides to aryl cyanides have
been developed that entail the use of zinc dust to keep the Pd in the zero oxidation state and zinc acetate to ensure high catalytic
activity. This procedure is applicable to a wide range of substrates and obviates the need for stringent removal of oxygen from the
reaction medium.
Ó 2003 Elsevier Ltd. All rights reserved.
A very important and useful reaction in organic syn-
thesis is the conversion of an aryl bromide or chloride to
an aryl cyanide;1 several procedures involving the use of
transition metals are available.2 Unfortunately, most of
these procedures require high temperatures (ꢀ120–
140 °C), and more importantly, the reactions tend to be
capricious.
As expected, when the DMF was saturated with oxygen,
there was absolutely no conversion of I into II.
This implicated oxygen as the main source in deactiva-
tion of the catalyst, resulting in lower conversion.4 Since
oxygen most likely oxidizes the active palladium(0)
species to a palladium(II) species, the in situ use of a
reductant might reduce the palladium(II) back to the
palladium(0) species and reintroduce it into the catalytic
cycle. Since zinc has been used in cyanation reactions to
reduce palladium(II) species,3b;5 we added 3–4 mol %
zinc dust to our cyanation reaction and obtained
excellent conversion (>99 area percent) of I into II
without the need to deoxygenate the DMF. The cya-
nation procedure using zinc dust was then successfully
scaled up to produce multi-kilogram quantities of II.
However, attempted cyanation of I purchased from a
vendor was totally unsuccessful (<5 AP conversion).
The difference between the two lots of I was that acetic
acid was used in our procedure to prepare I, whereas the
vendor material was prepared in an acetic acid-free
environment. Addition of 5 wt % of acetic acid to the
cyanation reaction using vendor supplied I gave a >99
AP conversion to II in ꢀ5 h.
Our efforts were directed towards finding a general and
scalable procedure to convert substituted bromoacet-
anilide I to the corresponding proprietary cyano com-
pound II. The two most recently reported systems for
such a conversion are:
(a) Pd(OAc)2 as the catalyst, dppe as a ligand and acet-
one cyanohydrin as the cyanide source2 and
(b) Pd2(dba)3 in combination with dppf [1,10-bis(diph-
enylphosphino)ferrocene] as a ligand and Zn(CN)2
as the cyanide source.3
Due to safety concerns as well as availability issues
associated with acetone cyanohydrin, we decided to
focus on the use of the Pd2(dba)3/dppf system. We found
that the cyanation under the reported conditions was
not reproducible with conversions varying from 50 to
>98 AP (area percent, HPLC) even after careful deoxy-
genation of the DMF by bubbling in nitrogen at 90 °C.
Since acetic acid could potentially generate hydrogen
cyanide upon reacting with zinc cyanide, we started
looking at safer alternates that would produce the same
catalytic effect. Our investigations led to the observation
that the use of 3–4 mol % zinc acetate6 was just as
effective as acetic acid in catalyzing the cyanation. In
order to ascertain whether the effect of zinc acetate
had wide applicability, we examined a wide range of
Keywords: Cyanation; Aryl bromides; Palladium; Zinc acetate; Zinc
dust.
* Tel.: +1-732-227-6941; fax: +1-732-227-3944; e-mail: ramakrishna.
chidambaram@bms.com
0040-4039/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.12.040

1442
R. Chidambaram / Tetrahedron Letters 45 (2004) 1441–1444
Table 1. Cyanation of aryl bromides in the presence of zinc acetate1
substrates with varying electronic and steric effects.7 In
all cases reagent grade DMF was used without any
additional purification other than performing the reac-
tion under nitrogen.
Ar
Time
(h)
Conversion
(HPLC area
percent)
Isolated
yielda
As seen in Table 1, with the exception of nitro compound
1g (entry 7) and the free amine 1i (entry 9), sub-
strates with different electronic and steric effects reacted
with excellent conversion and yields. In order to confirm
the beneficial effect of zinc acetate, the aryl bromides
were dissolved in an appropriate solvent, dried over
potassium carbonate to remove any traces of acetic acid
that may have been used for their preparation and then
cyanated in the presence of zinc dust and in the absence
of zinc acetate.8 The conversion rates were then com-
pared.
12
14
2
>98
>99
>99
>99
91
94
92
89
The results are summarized in Table 2.
As can be seen in Table 2, with the exception of electron-
deficient substrate 1c, which did not require Zn(OAc)2,
the conversions were sharply lower than those in which
zinc acetate was used. This pointed to a definite beneficial
effect of zinc acetate. In order to understand whether this
effect represented a rate enhancement due to the presence
of zinc or whether it was an effect of the counterion, the
cyanation of 1e was carried out in the presence of other
additives.9 The results are summarized in Table 3.
6
9
>99
90
7
3
>99
<1
87
––
Additives such as magnesium acetate and zinc trifluoro-
acetate also had a beneficial effect, whereas zinc oxide
and zinc chloride did not, pointing to a counterion effect
rather than a metal effect.
Efforts were then directed towards understanding the
role of Zn(OAc)2 in the catalytic cycle. Cyanation of I
(Fig. 1) was carried out in the absence of Zn(OAc)2 for
about 3.5 h at which point there was <5 area percent
conversion as expected, 4 mol % of Zn(OAc)2 and
3 mol % of zinc dust were then added to the reaction
mixture. The conversion increased to 50 area percent
after 4 h and then stalled. However, in a separate
experiment, when 3 mol % of zinc dust and 4 mol % of
Zn(OAc)2 were introduced 24 h after the cyanation was
carried out in the absence of Zn(OAc)2, there was no
additional conversion beyond that observed early in the
experiment.10
3
>99
80
24
5
ꢀ20
Not
isolatedb
Not detected
––
a Unless otherwise stated, all reactions were carried out on a 5-g scale.
Isolated yields have not been optimized. The materials isolated were
compared with authentic samples (1H and 13C NMR). The reactions
were run with 0.1 mol % Pd2(dba)3, 0.25 mol % dppf, 3–4 mol % Zn
dust and 3–4 mol % Zn(OAc)2 in DMF containing 1% by volume
water.
The fact that the cyanation reaction slowed down sig-
nificantly and finally stalled suggests an irreversible
competitive deactivation of the catalyst and that
Zn(OAc)2 presumably keeps the catalyst ÔaliveÕ in the
catalytic cycle long enough to complete the cyanation.
b The formation of product was checked by comparing the retention
time on the HPLC with an authentic sample.
In the presence of cyanide ions, palladium is known to
generate species such as [Pd0(CN)n]nꢁ, which can no
longer participate in the catalytic cycle.2 These com-
pounds have strongly bound cyanide groups on the
palladium11 that interfere with the necessary oxidative
addition and thus slow down the conversion.
4 mol % Zn dust and 3–4 mol % Zn(OAc)2, there was
only about a 50% conversion after 9 h (compared to
>99% conversion using normal conditions). Using
0.1 mol % of a less tightly bound, easily reducible pal-
11;12
Additionally, when cyanation of 1e was carried out with
0.1 mol % of K2Pd(CN)4 or Pd(CN)2 along with 3–
ladium compound like Pd(OAc)2
conversion was >95 AP after 9 h.
to cyanate 1e, the

R. Chidambaram / Tetrahedron Letters 45 (2004) 1441–1444
1443
Table 2. HPLC area percent conversion of substrates in the absence of
zinc acetate but in the presence of zinc dust
DMF is used without stringent deoxygenation controls.
Mechanistic implications of this observation as well as
other applications of this system are currently being
investigated.
Substratea
AP Conversion
[without Zn(OAc)2]b;
(time, h)
AP Conversion with
Zn(OAc)2 (from Table 1);
(time, h)
1a
1b
1c
1d
1ec
1f
75 (12)
11 (24)
>99 (2)
14 (9)
>98 (12)
>99 (14)
>99 (2)
>99 (6)
>99 (9)
>99 (6)
>99 (3)
Acknowledgements
The author thanks Drs. Percy Manchand, William
Nugent, Edward Delaney, Richard Mueller, David
Kronenthal and Wendel Doubleday for fruitful discus-
sions. The author also thanks Profs. Barry Trost and
Donna Blackmond for their valuable insights.
60 (12)
49 (6)
80 (3)
1h
a The substrates were dissolved in MTBE or ethyl acetate, dried over
potassium carbonate, filtered and then the solvents removed under
reduced pressure.
b The reactions were run with 0.1 mol % Pd2(dba)3, 0.25 mol % dppf
and 3–4 mol % Zn dust in DMF containing 1% by volume water.
c This potassium carbonate washed substrate was also cyanated in the
presence of Zn(OAc)2. The conversion was >99 area percent (ꢀ90–
92% isolated yields).
References and notes
1. (a) Ellis, G. P.; Ronmey-Alexander, T. M. Chem. Rev.
1987, 87, 779; (b) Grushin, V. V.; Alper, H. Chem. Rev.
1994, 94, 1074; (c) House, H. O.; Fisher, W. F. J. Org.
Chem. 1969, 34, 3626; (d) Kleeman, A.; Engel, J.;
Kutscher, B.; Reichart, D. Pharmaceutical Substances:
Syntheses, Patents, Applications. Fourth ed.; Georg Thi-
eme: Stuttgart, 2001.
Table 3. Effect of other additives on the cyanation reaction with 1e
2. Sundermeier, M.; Zapf, A.; Beller, M. Angew. Chem., Int.
Ed. 2003, 42, 1661–1664, and references cited therein.
3. (a) Maligres, P. E.; Waters, M. S.; Fleitz, F.; Askin, D.
Tetrahedron Lett. 1999, 40, 8193–8195; (b) Jin, F.;
Confalone, P. N. Tetrahedron Lett. 2000, 40, 3271–3273.
4. Attempts to quantitate the dissolved oxygen in DMF were
unsuccessful since DMF dissolved the available probe
sensors.
Additive
Conversion (AP)a
NaOAc (9)
Mg(OAc)2 (9)
ZnCl2 (9)
85 (stalled)
>99
55
Zn(OCOCF3)2 (9)
ZnO (9)
No additive
>99
40
60
5. Okano, T.; Iwahara, M.; Kiji, J. Synlett 1998, 243.
6. When the cyanation of 1e was run with 50 mol % of
Zn(OAc)2, the reaction was very sluggish (only 50 HPLC
area percent completion after 9 h).
a The reactions were run for 9 h with 0.1 mol % Pd2(dba)3, 0.25 mol %
dppf, 3–4 mol % Zn dust and 4 mol % additive in DMF containing 1%
by volume water.
7. Representative procedure for cyanation. To a 100-mL,
3-necked flask connected to a nitrogen-inlet adapter and
reflux condenser was charged the substrate 1e (5 g;
21.9 mmol), DMF (10 mL), water (100 lL), dppf (0.03 g;
0.06 mmol), Pd2(dba)3 (0.1 mol %; 0.02 g; 0.02 mmol),
Zn(CN)2 (1.39 g; 11.8 mmol), zinc dust (4 mol %; 0.06 g;
0.88 mmol) and Zn(OAc)2 (4 mol %; 0.16 g; 0.88 mmol).
The mixture was heated and the temperature maintained
between 90 and 100 °C. The reaction was monitored by
HPLC and was worked up after the reaction was deemed
complete (<2 AP of starting material). Workup of liquid
products were done according to the procedure described
in Ref. 4. Workup of solid products were done according
to the procedure described in Ref. 3. Commercially
available starting materials were used for the cyanations
and the isolated products were compared with authentic
samples that were purchased from various vendors.
8. The substrates were dissolved in either methyl t-butyl ether
or ethyl acetate, dried over K2CO3 for 30 min, filtered and
the solvent removed under reduced pressure. Cyanations
were then carried out as described above without the
Zn(OAc)2. The reaction rates were compared by HPLC
area percent.
NHAc
NHAc
R
R
Br
CN
I
II
Figure 1.
These data are again consistent with a scenario in which
Zn(OAc)2 keeps the catalyst active in the catalytic cycle.
In the absence of Zn(OAc)2, a palladium species that
can no longer be activated by the Zn/Zn(OAc)2 combi-
nation appears to form irreversibly. At this juncture it is
not clear as to whether the role of Zn(OAc)2 is to just
activate the palladium catalyst.
In summary, a much more robust and reliable cyanation
of aromatic bromo compounds has been developed in
which the introduction of zinc acetate in combination
with a reducing agent such as zinc dust keeps the pal-
ladium from being deactivated even when reagent grade
9. The substrates were cyanated as described in Ref. 7 with
the appropriate additive being added instead of Zn(OAc)2.
10. No additional conversion was observed with 1a, 1b and 1e
when Zn(OAc)2 and additional Zn dust were added 24 h
after cyanation was carried out without Zn(OAc)2.

1444
R. Chidambaram / Tetrahedron Letters 45 (2004) 1441–1444
11. Nakamura, M.; Fujiwara, S. J. Coord. Chem. 1972, 1,
221–227.
12. When Pd(OAc)2 or K2Pd(CN)4, which are palladium(II)
species, were used in the absence of zinc dust, the
conversion was <5 HPLC area percent. The cyanation of
1e using 5 mol % of Pd(OAc)2, Zn dust and dppf without
any Zn(OAc)2 resulted in a reaction that was 80 HPLC
area percent complete after 9 h. However, the reaction
profile indicated the formation of several unidentified
impurities (totaling ꢀ15 HPLC area percent) that were
not seen during the course of a normal cyanation
reaction.