Expanding the scope of silane-mediated hydrodehalogenation reactions

Article abstract of DOI:10.1016/j.tetlet.2013.06.063

A palladium-catalysed, silane-mediated hydrodehalogenation (HDH) reaction with increased substrate scope has been developed. Whereas previous attempts to reduce carboxylic acid or phenol-containing aryl halides using silane-based HDH conditions failed, the current protocol allows efficient access to the reduced products. Chemoselective HDH in the presence of sensitive functional groups is also presented.

Full text of DOI:10.1016/j.tetlet.2013.06.063

Tetrahedron Letters 54 (2013) 4518–4521
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Expanding the scope of silane-mediated hydrodehalogenation
reactions
⇑
Gary M. Noonan , Barry R. Hayter, Andrew D. Campbell, Timothy W. Gorman,
Benjamin E. Partridge, Gill M. Lamont
AstraZeneca Mereside, Alderley Park, Cheshire SK10 4TG, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
A palladium-catalysed, silane-mediated hydrodehalogenation (HDH) reaction with increased substrate
scope has been developed. Whereas previous attempts to reduce carboxylic acid or phenol-containing
aryl halides using silane-based HDH conditions failed, the current protocol allows efficient access to
the reduced products. Chemoselective HDH in the presence of sensitive functional groups is also
presented.
Received 12 February 2013
Revised 31 May 2013
Accepted 14 June 2013
Available online 24 June 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Hydrodehalogenation
Palladium catalysed
Silane
Halogenated arenes and heteroarenes are widely utilised build-
ing blocks in organic chemistry. The carbon–halogen bond usually
serves as a functional handle in carbon–carbon or carbon–hetero-
atom bond forming reactions, however, the removal of ‘redundant
halides’ from reaction products is sometimes necessary. These
‘redundant halides’ derive from the use of halides as activating/
blocking groups1,2 for other transformations, or due to the fact that
it is sometimes easier to access, via chemical synthesis or commer-
cial sources, a more extensively halogenated compound.
Many methodologies have been developed for the selective re-
moval of halogens from arenes and heteroarenes.2–10 While the
methodologies cited above have proven effective in many situa-
tions, they all suffer limitations. HDH reactions employing harsh
reagents such as Grignard reagents or other strong bases3–7 pro-
vide limited functional group tolerance, while other methodologies
are of limited substrate scope,7,10 or employ undesirably high reac-
tion temperatures.9
Of the hydrodehalogenation (HDH) methods that have been
developed to-date, those which utilise silanes as cheap and readily
available reducing agents are particularly attractive. Silane-based
HDH methodologies avoid the use of hydrogen gas (which is an
explosion risk and is not always accessible), and provide chemose-
lective conditions for HDH in the presence of sensitive
functionalities. Although a number of reports of silane-mediated
palladium-catalysed HDH exist,11–13 these published methodolo-
gies are of limited substrate scope. Conditions employing
polymethylhydrosilane with potassium fluoride as the base were
completely ineffective for phenolic and carboxylic acid containing
substrates, while protocols employing triethylsilane as a reducing
agent have only demonstrated the reduction of a single aryl
chloride (chlorobenzene).11–13 Herein we report a chemoselective
silane-mediated HDH protocol which addresses these limitations.
To facilitate a drug discovery campaign we wished to access the
polyfunctional pyridine 3 (Scheme 1).
Compound 3 was not commercially available so we utilised 2,6-
dichloropyridine 1 as the starting material. Access to the desired
compound would then be achieved via selective displacement of
the 6-chloro substituent with ammonia, followed by reduction
(HDH) of the unwanted carbon–chlorine bond.
We initially attempted the HDH of 2 under catalytic hydrogena-
tion conditions [H2 (1 atm), Pd/C], but this led to the simultaneous
reduction of the nitrile functionality, as well as the formation of
secondary amine side-products. Dissolving metal type reductions
(using Zn/AcOH or Zn/TFA) were then trialed, however, despite
being relatively successful in providing the desired product, issues
during the reaction work-up, that is, nitrile hydrolysis, led us to
F
CN
Cl
F
CN
Cl
F
CN
H
NH3
HDH
Cl
N
H2N
N
H2N
N
1
2
3
commercially
available
⇑
Corresponding author. Tel.: +44 1625 233264.
E-mail address: gary.noonan@astrazeneca.com (G.M. Noonan).
Scheme 1. Synthetic route to polyfunctional pyridine 3.
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.06.063

G. M. Noonan et al. / Tetrahedron Letters 54 (2013) 4518–4521
4519
Table 1
Screening of the hydrodehalogenation conditions
F
CN
HN
N
CN
F
CN
H
N
Pd (5 mol%), ligand
F
CN
Cl
H2N
N
3
Et3N (2 eq.), 1,4-dioxane
H2N
F
5
F
H2N
N
2
Et3SiH (5 eq.)
100 °C, 3 h
CN
CN
F
CN
H2N
N
H2N
N
4
SiEt3
N
F
6
NH2
Entrya
Catalyst
2b
3b
4b
5/6b,c
1
2
3
4
5
6
Pd2(dba)3/t-Bu-X-Phosd
Pd(PPh3)2Cl2
62
96
85
—
—
—
24
4
15
88
85
91
—
—
—
—
15
9
14
—
—
12
—
—
Pd(dppf)2Cl2
Pd2(dba)3/X-Phosd
Pd(A-Phos)2Cl2
Pd(dtbpf)Cl2
a
b
c
Reactions carried out using 0.6 mmol substrate in 3 mL of 1,4-dioxane.
Percentage product is based on LC–MS (uncalibrated) of the crude reaction mixtures.
Structures 5 and 6 are both possible based on [M+H]+ observed in crude LC–MS data.
10 mol % ligand.
d
Table 2
seek an alternative approach. Thus we explored palladium-silane
based hydrodehalogenation conditions and Table 1 outlines the re-
sults of our initial catalyst screen.
Substrate scope of the hydrodehalogenation conditions17,18
[Pd(dtbpf)Cl2] (5 mol%)
Et3N (2 eq.), 1,4-dioxane
Cl/Br
H
Analysis of the crude reaction mixtures (Table 1) was carried
out by LC–MS and the percentage value shown for each compound
2–6 was that observed in the uncalibrated chromatogram (UV
detection). The identities of the side-products 4–6 are proposed
based on mass ions observed in the crude LC–MS data.14
The most successful catalysts were those derived from bulky,
electron-rich phosphines, with both monodentate (entry 5) and
bidentate ligands (entry 6) proving successful. The reaction using
[Pd(dtbpf)Cl2] (entry 6) was particularly effective, producing
mainly the desired product, together with a small amount of the
silylated side-product 4.
R
R
Et3SiH (5 eq.)
100 °C
Entry
Substrate
Time (h)
Product
Yielda
1
16
75b
Br
This reaction (entry 6) was then repeated using a 1H NMR inter-
nal standard. After 1 h clean conversion to the required product
was observed, with an internal standard based yield of 68% (see
Supplementary data for details). The reaction mixture was allowed
to stir for a further 30 min, after which a small impurity started to
become apparent in the 1H NMR spectrum. The reaction was
stopped after 2 h and following work-up and purification, the re-
quired product was isolated in a 56% yield (average of two runs).15
Although the isolated yields were moderate, this methodology pro-
vided the required synthetic intermediate in a quick and efficient
manner.
As hydrodehalogenation is a reaction we utilise on a regular ba-
sis, we decided to explore the scope of the current HDH conditions.
We were particularly interested in testing these reaction condi-
tions in the hydrodehalogenation of substrates which have previ-
ously failed to produce the desired product using related
protocols. The results of this study are presented in Table 2.
We carried out initial test reactions on 4-bromo/iodo tert-butyl-
benzene. The reaction of 4-bromo-tert-butylbenzene produced the
desired product in a moderate yield (entry 1), whereas reduction of
the iodo analogue was less efficient (entry 2). Significant unknown
impurities were observed in both reactions, with 33% of an un-
known side-product observed for the 4-iodo substrate. These
2
3
16
16
47c
I
O
OH
O
OH
98
(65)
Cl
OH
OH
4
5
20
16
87d
Cl
Cl
(70)d
N
OH
N
OH
(continued on next page)

4520
G. M. Noonan et al. / Tetrahedron Letters 54 (2013) 4518–4521
both phenols and carboxylic acids are very common functional-
Table 2 (continued)
ities. Reduction of a chloro-indole was also successful (entry 6),
but again required the use of the more active [Pd(dba)2]-S-Phos
catalyst system for complete conversion. The purification of the
resulting indole proved challenging due to co-elution with silane
based impurities, however, when more polar/elaborated substrates
were employed (see below) purification was straight-forward
using flash silica chromatography.
Entry
Substrate
Time (h)
2
Product
Yielda
92d
Cl
6
N
H
N
H
Bn
Bn
O
O
97d
In order to test the chemoselectivity of these reaction condi-
tions we carried out HDH of substrates containing other function-
ality prone to reduction. HDH of chloro- and bromo-arenes
containing benzyl-protected alcohols provided the reduced prod-
ucts with no observable alcohol deprotection (entries 7 and 8).
For the reasons stated above, small amounts of silane-based impu-
rities were persistent in purified samples of benzyloxybenzene.
However, high conversion into the desired product was observed
in the crude reaction mixture based on the proton NMR internal
standard (see Supplementary data). It is noteworthy that the
HDH of 3-(benzyloxy)-5-bromopyridine, using the conditions from
Table 2 (Pd(dba)2-S-phos, t = 1.5 h), produced a significant amount
of a side product (ꢀ23% in the crude LC–MS), which appeared to be
3-hydroxypyridine. However, it is possible that this unwanted
side-reaction may be avoided upon reduction in the number of
equivalents of silane reducing agent and/or lowering the reaction
temperature.
7
2
(53)e
Cl
O
Bn
Bn
O
8
9
2
1
(62)f
Br
NO2
NO2
NO2
73g
Cl
NO2
It also proved possible to reduce aryl halides in the presence of
nitro or alkenyl functionalities (entries 9–11). The present HDH
conditions were amenable to selective reduction of the carbon–
halogen bonds in these substrates, but the reactions required close
monitoring to the point where a maximum yield of the desired
product was obtained while minimising over-reduction (entries
9–11). For example, reduction of both the halogen and the nitro
groups was observed over prolonged reaction times with the 4-
halo-nitrobenzene substrates.
Notably, we failed to reduce 3-chloro-2-nitropyridine selec-
tively, with the crude 1H NMR spectrum showing multiple species
present. An isoxazole-containing substrate, namely, 5-(4-chloro-
phenyl)isoxazole also failed to give the desired HDH product selec-
tively, but instead produced what appeared to be a ring-opened
product (due to cleavage of the isoxazole N–O bond).
10
11
0.5
57h
Br
0.25
90
Br
Yields based on 1H NMR internal standard with yields of isolated products in
a
parentheses; confirmation of the identity of the products was provided by 1H NMR
spectroscopy of authentic product samples under conditions identical to the reac-
tion conditions. All isolated products are >95% pure unless otherwise stated.
b
12% of an unknown side-product.
30% of an unknown side-product.
5% [Pd(dba)2], 10% S-Phos.
c
d
e
Silane impurities are present (<10%).
90% pure based on 1H NMR using maleic acid as internal standard (Silane
f
Due to the fact that the isolated yields obtained in Table 2 were
somewhat lower than those based on the internal standard, we
carried out two larger scale HDH reactions (ꢀ5 mmol scale) on
substrates which would produce more easily purified non-volatile
products. We also wanted to test the robustness of the present pro-
impurities present).
g
16% starting material remaining.
5% aniline formed.
h
impurities may be the result of silylation at the 4-position versus
hydrodehalogenation, but in the absence of isolation of these
impurities, this is purely speculative.
Cl
Pleasingly, we found that para-chlorobenzoic acid could be re-
duced effectively under our reaction conditions, with near quanti-
tative conversion into benzoic acid observed in the crude reaction
mixture (entry 3). It was noted previously that para-chlorophenol
was also completely unreactive under silane-based HDH condi-
tions.11 Attempted reduction of this substrate under our standard
reaction conditions (using [Pd(dtbpf)Cl2] as catalyst) was only
moderately successful, providing the desired product in 24% yield
based on an internal standard. However, upon moving to the bulky
and highly active S-Phos ligand,16 with [Pd(dba)2] as the palladium
source, the desired product was formed in a high yield (entry 4).
Interestingly, for this reaction, we initially observed the formation
of what appeared to be a silylated product, which then went on to
form the reduced ‘free phenol’ (see Supplementary data).
Reduction of a chloro-pyridone substrate was successful, produc-
ing the desired product in 70% isolated yield (entry 5). Successful
HDH of these phenolic and carboxylic substrates under silane-
mediated HDH conditions expands the scope of this reaction, as
[Pd(dba)2] (5 mol%)
S-Phos (10 mol%)
O
S
N
O
O
S
N
O
Et3N (2 eq.), 1,4-dioxane
Et3SiH (5 eq.)
100 °C, 16 h
N
H
N
H
~4 mmol
Br
77% Isolated
[Pd(dba)2] (5 mol%)
S-Phos (10 mol%)
Et3N (2 eq.), 1,4-dioxane
Et3SiH (5 eq.)
H2N
O
H2N
O
100 °C, 16 h
87% Isolated
~5 mmol
Scheme 2. Larger scale HDH reactions.

G. M. Noonan et al. / Tetrahedron Letters 54 (2013) 4518–4521
11. Maleczka, R. E., Jr.; Rahaim, R. J., Jr. Tetrahedron Lett 2002, 43, 8823.
4521
tocol and so, in carrying out these reactions, commercial-grade
anhydrous solvent was used directly and none of the reagents were
degassed prior to the reaction. Pleasingly, high isolated yields of
the desired products were obtained for both of these reactions
(Scheme 2). Reduction of these amide and sulfonamide substrates
further exemplifies the substrate scope of the current
methodology.
In summary, this HDH protocol expands significantly the sub-
strate scope of the silane-mediated hydrodehalogenation reaction,
with silane-mediated reduction of hydroxyl and carboxylic acid
containing substrates now possible. The current conditions have
been shown to be effective for the reduction of a variety of func-
tionalised haloarenes and, most notably, deactivated aryl chlorides.
This methodology provides an operationally simple and relatively
functional group tolerant hydrodehalogenation of arenes and het-
eroarenes using readily available reagents and should find applica-
tion, for example, in discovery-scale medicinal chemistry
campaigns.
12. Manuel, G.; Chatgilialoglu, C.; Boukherroub, R. Organometallics 1996, 15, 1508.
13. Azadi, R.; Firouzabadi, H.; Iranpoor, N. J. Organomet. Chem. 2010, 695, 887.
14. We propose that structure 5 is the most likely structure of the impurity
labelled 5/6 in Table 1 and is formed via Buchwald–Hartwig coupling between
the hydrodehalogenated product and the chloroarene starting material,
although the isomeric structure 6 is also plausible.
15. Further optimisation of the reaction conditions, that is, stopping the reaction at
the point of maximum conversion into the desired product with minimum
side-product formation should prove possible, but was beyond the scope of the
current work.
16. Barder, T. E.; Walker, S. D.; Martinelli, J. R.; Buchwald, S. L. J. Am. Chem. Soc.
2005, 127, 4685.
17. General procedure: To a sparge-degassed solution of Et3N (0.405 g, 0.554 mL,
4 mmol) and 1,4-dimethoxybenzene (internal standard) (0.138 g, 1 mmol) in
anhydrous 1,4-dioxane (12 mL) under a nitrogen atmosphere was added 6-
amino-2-chloro-5-fluoronicotinonitrile (343 mg, 2 mmol), [(1,10-bis-(di-tert-
butyl-phosphino)ferrocene)PdCl2] (64 mg, 0.1 mmol) and Et3SiH (1.160 g,
1.592 mL, 10 mmol). The mixture was then heated to 100 °C with stirring.
The reaction was monitored using 1H NMR spectroscopy, taking aliquots of the
crude reaction mixture. Upon completion, the mixture was diluted with CH2Cl2
(20 mL) and washed with H2O (1 ꢁ 20 mL). The aqueous layer was then
extracted with CH2Cl2 (2 ꢁ 20 mL). The combined organics were dried over
anhydrous MgSO4, filtered, and the solvent was removed under reduced
pressure to leave the crude reaction product. The crude product was purified
by flash silica chromatography with 0–5% methanolic NH3 (7 M) in CH2Cl2 as
the eluent. The solvent was then removed from fractions containing the
desired product to leave 6-amino-5-fluoronicotinonitrile (3) (148 mg,
1.076 mmol).
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2013.06.
063.
Mp = 181–182 °C; 1H NMR (400 MHz, DMSO-d6, 22 °C): 7.38 (2H, s), 7.84 (1H,
dd, J = 11.43, 1.78 Hz), 8.16–8.26 (1H, m); 13C NMR (101 MHz, DMSO-d6,
22 °C): 94.30, 117.64, 122.76, 144.14 (J = 254.4 Hz), 148.87, 152.35; HRMS:
calcd [M+] = 137.0389, found: m/z (EI) [M+] = 137.0389.
Note: All other isolated samples were purified by flash chromatography, with
the exception of 2-pyridone, which was purified on an SCX ion-exchange
column.
An identical procedure was followed using the [Pd(dba)2]-S-Phos catalyst
system, substituting [Pd(dba)2] (5 mol %) and S-Phos (10 mol %) for the [(1,10-
bis-(di-tert-butyl-phosphino)ferrocene)PdCl2] catalyst.
References and notes
1. Liegault, B.; Petrov, I.; Gorelsky, S. I.; Fagnou, K. J. Org. Chem. 2010, 1047, 75.
2. Jimenez, L. S.; Ramanathan, A. Synthesis 2010, 217.
3. Nolan, S. P.; Grasa, G. A.; Viciu, M. S. Organometallics 2001, 20, 3607.
4. von Wangelin, A. J.; Mayer, M.; Grupe, S.; Czaplik, W. M. Chem. Commun. 2010,
6350.
5. Cazin, C. S. J.; Bisaro, F.; Akzinnay, S. Chem. Commun. 2009, 5752.
6. Cucullu, M. E.; Nolan, S. P.; Belderain, T. R.; Grubbs, R. H. Organometallics 1999,
18, 1299.
18. General procedure for reactions monitored by 1H NMR: Prior to the beginning of
the reaction (i.e., before the reaction mixture was heated), a sample of the
reaction mixture (0.1 mL) was removed via syringe, placed in an NMR tube and
diluted with CDCl3 (0.5 mL). The 1H NMR of the sample was then measured
(t = 0, 1H NMR). Throughout the reaction samples were removed (0.1 mL),
diluted with CDCl3 (0.5 mL) and the 1H NMR spectrum was measured. Changes
in the amount of the starting material remaining versus internal standard and
product formed were then measured by integration of the resulting spectra
(see Supplementary data for more details).
7. Chelucci, G. Tetrahedron Lett. 2010, 51, 1562.
8. Koutentis, P. A.; Ioannidou, H. A. Tetrahedron 2011, 67, 3348.
9. Roy-MacArthur, A. H.; Lin, S.; Kelly, C. K.; Geuther, M. E.; Cannon, K. A.
Organometallics 2011, 30, 4067.
10. Bei, X.; Hagemeyer, A.; Volpe, A.; Saxton, R.; Turner, H.; Guram, A. S. J. Org.
Chem. 2004, 69, 8626.