methylene groups using either hypophosphorous acid or a
combination of hypophosphorous and phosphorus acids.
Furthermore, we report a successful application of this
novel method to the synthesis of an arylhalogen-contain-
ing anticancer agent, Lonafarnib.
studies to optimize the reaction conditions to minimize the
byproduct and improve the operability on a large scale. A
numberofexperimentswere carried out, and the results are
listed in Table 1.
As indicated by entry 2 of Table 1, reducing the amount
of H PO and replacing HI with NaI did not reduce the
3
2
level of the byproduct. Use of less potent reducing agents
such as H PO (entry 3) or NaH PO (entry 4) in the
3
3
2
3
Scheme 1. Proposed Synthesis of Lonafarnib
presence of 2 equiv of NaI and aqueous HBr completed
thereduction witha longerreaction time. However, neither
of the conditions prevented the formation of the iodo
byproduct. We thought that the level of the iodo displace-
ment byproduct should be proportional to the concentra-
tionofiodide and thatthe less the iodide concentration, the
lower the byproduct. To our delight, reducing the amount
of NaI from 2 to 0.1 equiv suppressed the formation of the
byproduct to an undetectable level. However, use of a
catalytic amount of NaI (0.1 equiv) also slowed the desired
reductions. For example, use of 3.5 equiv of H PO and
3
2
0.1 equiv of NaI (entry 5) completely reduced the diaryl
ketone, but only partially reduced the nitro group. On the
other hand, a mixture of 8 equiv of H PO and 0.1 equiv of
3
3
NaI (entry 6) reduced the nitro group quickly, but the
carbonyl group very slowly. Concentrated HCl was not
tested in this study.
Recently, Fry reported a chemoselective reduction of
aryl ketones in the presence of bromo substituents with
6
a combination of H PO and I in refluxing HOAc. This
3
2
2
method is safer than a previously reported procedure
7
using red phosphorus and HI under refluxing conditions.
However, similar conditions are not known to reduce
9
nitroarenes. Thus, we initially planned a two-stage reduc-
a
Table 1. A Novel Reduction of Nitroarenes and Aryl Ketones
tion: first reducing the aryl ketone with the H PO -based
3
2
method and then the nitro group with our ZnX -modu-
2
5
lated Pd/C catalyst. In light of safety considerations, we
focused our effort on H PO -based reducing agents, rather
3
2
than the red phosphorus. To our suprise, the first reaction
with 11 equiv of H PO and 4 equiv of HI in refluxing
3
2
water not only reduced the ketone group to a methylene
group but also reduced the nitro group to the correspond-
ing aniline; both the bromo and chloro substituents were
largly preserved. To the best of our knowledge, this is the
first example of nitroarene reduction with the HIꢀH PO
3
2
system. This method can be complementary to the current
methods such as hydrogenation, hydride, and metal-based
procedures. However, there were two issues associated
with this procedure: (1) a major byproduct was observed
at about the 5% level; and (2) a severe foaming at the
beginning of the reduction. The byproduct was identified
as the 3-iodo analogue of 2, apparently derived from an
iodo replacement of the 3-bromo substituent on the pyr-
idine ring, and was very difficult to remove to a pharma-
ceutically acceptable level. Thus, we undertook additional
reducing
agents
(equiv)
catalyst
entry
(equiv) solvents 2:2-iodo comments
1
2
3
4
5
6
7
H
3
PO
2
(11)
HI (4)
H
2
O
95:5
Complete
Complete
Complete
Complete
H PO (3.5) Nal (2) HBr/H O 95:5
3
2
2
2
2
2
2
2
H
3
PO
3
(8)
Nal (2) HBr/H
O
O
O
O
O
95:5
95:5
NaH
2
PO
2
(11) Nal (2) HBr/H
H
H
H
H
3
PO
3
PO
3
PO
3
PO
2
3
2
3
(3.5) Nal (0.1) HBr/H
100:0 Nitro remain
100:0 Ketone remain
100:0 Complete
(8)
Nal (0.1) HBr/H
Nal (0.1) HBr/H
(2)/
(3.5)
a
All reactions were carried out at 115 °C for 1 to 16 h.
(
6) (a) Hicks, L.; Han, J. K.; Fry, A. J. Tetrahedron Lett. 2000, 41,
817. (b) Gordon, P. E.; Fry, A. L. Tetrahedron Lett. 2001, 42, 831.
7) (a) Bradsher, C. K.; Vingiello, F. A. J. Org. Chem. 1948, 13, 786.
b) Lee, W.; Park, C. J. Org. Chem. 1993, 59, 7149.
7
(
Although the foaming was manageable on laboratory
scales byadjusting both the agitation and the addition rate,
itwould not be easytooperate inplantproduction. Careful
observations indicated that the severe foaming takes place
only at the beginning of the reaction. We speculated that a
(
(8) Blatt, A. H.; Gross, N. J. Org. Chem. 1957, 22, 1046.
(9) (a) Wu, G. G.; Wong, Y.; Poirier, M. Org. Lett. 1999, 1, 745.
(
b) Bernard, C. F.; Casey, M.; Chen, F. X.; Grogan, D. C.; Poirier, M.;
Williams, R. P.; Wong, Y.; Wu, G. U.S. Patent 6,495,689, December 17,
002.
2
Org. Lett., Vol. 13, No. 19, 2011
5221