Development of a Stepwise Reductive Deoxygenation Process by Ru-Catalysed Homogeneous Ketone Reduction and Pd-Catalysed Hydrogenolysis in the Presence of Cu Salts

Article abstract of DOI:10.1002/cctc.201200526

A stepwise catalytic reduction of ketone 1 to alcohol 2 and subsequently to aryl(imidazo[1,2-b]pyridazinyl)methane 3 is described, which provides synthetically useful chemoselectivity at acceptably low catalyst loadings. Undesired reactive sites include a

Full text of DOI:10.1002/cctc.201200526

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201200526
Development of a Stepwise Reductive Deoxygenation
Process by Ru-Catalysed Homogeneous Ketone Reduction
and Pd-Catalysed Hydrogenolysis in the Presence of Cu
Salts
Damian M. Grainger,*[a] Antonio Zanotti-Gerosa,[a] Kevin P. Cole,*[b] David Mitchell,[b]
Scott A. May,[b] Patrick M. Pollock,[b] and Joel R. Calvin[b]
A stepwise catalytic reduction of ketone 1 to alcohol 2 and
subsequently to aryl(imidazo[1,2-b]pyridazinyl)methane 3 is de-
scribed, which provides synthetically useful chemoselectivity at
acceptably low catalyst loadings. Undesired reactive sites in-
clude an aryl chloride, heteroarylchloride and benzylic amine
group. The presence of these functional groups presents a sig-
nificant challenge to chemoselectivity for both reduction steps.
For selective C=O reduction of highly functionalised 1, high
chemoselectivity was observed at low catalyst loading by
using Wills’ tethered Ru transfer-hydrogenation catalyst 13.
The selective hydrogenolysis of 2 was then accomplished
under acidic hydrogenation conditions by using a Pd/C catalyst
in the presence of Cu salts. This procedure has been demon-
strated on a multi-gram scale, which makes this approach
a viable method to use a combination of homogeneous and
heterogeneous catalysis.
Introduction
This work describes the development of an efficient two-step
catalytic method for the reductive deoxygenation of ketone
1 to aryl(imidazo[1,2-b]pyridazinyl)methane 3, which is a key
building block in the preparation of LY2784544 (Scheme 1).
LY2784544 is a JAK2 inhibitor currently undergoing clinical in-
vestigations for the treatment of myeloproliferative disor-
ders.[1,2] To date, this deoxygenation has been accomplished in
a single step by the treatment of 1 with six equivalents of trie-
thylsilane in the presence of twelve equivalents of trifluoroace-
tic acid as a promoter and solvent.[2] Although the isolated
yields and purity for this transformation are high, the desire to
avoid large amounts of fluoride- and silicon-containing waste
prompted us to investigate alternatives. We identified two al-
ternative reductions for the single-step deoxygenation, but
each had drawbacks. Trichlorosilane with triethylamine worked
well,[3] but the volatile nature of trichlorosilane and a difficult
reaction workup eliminated this method from contention. Hy-
pophosphorus acid/iodine reductions were also effective,[4] but
iodide-induced catalyst poisoning was encountered in the
downstream chemistry,[2] which caused us to abandon this ap-
proach. Wolff–Kishner reduction and numerous other methods
were found to be ineffective, which highlights the surprising
difficulty of the desired transformation.
Experimental Results
The most direct catalytic approach to the target molecule
would be the hydrogenation of 1 to 2 and the one-pot hydro-
genolysis to 3 in the presence of heterogeneous catalysts.[5]
A
5% Pd/C catalyst (Johnson Matthey (JM), Type 5R39) was
chosen to test a broad range of reaction variables, which in-
clude choice of solvent (MeOH, THF, toluene, AcOH, THF/water
and THF/AcOH), temperature (30–708C) and H2 pressure (6–
30 bar).[6] The consumption of starting material 1 was observed
to varying extents but it invariably produced none of the de-
sired product 3, only traces of 2 and significant amounts of
side-products that arise from morpholine cleavage along with
dechlorination. The two main side-products were tentatively
identified by using LC–MS as 5 and 6.[6] It is known that addi-
tives that act as chloride sources help to suppress dechlorina-
tion side-reactions.[5] Hence several acids and salts (including,
among others, HCl, NaCl, ZnCl2,[7] CuCl2 and CuSO4) were
tested without any noticeable improvement in conversion. In-
terestingly, CuCl2 and CuSO4 in THF or THF/water were found
to reduce the formation of dechlorinated and morpholine-
cleaved side-products, although their addition resulted in no
conversion of 1. When the screen was extended to other pre-
[a] Dr. D. M. Grainger, Dr. A. Zanotti-Gerosa
Johnson Matthey, Catalysis and Chiral Technologies
Cambridge Science Park, Unit 28
Cambridge, CB4 0FP (United Kingdom)
Fax: (+44)01223-438037
E-mail: damian.grainger@matthey.com
[b] Dr. K. P. Cole, Dr. D. Mitchell, Dr. S. A. May, P. M. Pollock, J. R. Calvin
Chemical Product Research and Development
Lilly Research Laboratories
Eli Lilly and Company
Lilly Corporate Center
Indianapolis, IN 46285 (USA)
Fax: (+1)317-276-4507
E-mail: k_cole@lilly.com
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201200526.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1205

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Scheme 1. LY2784544, key intermediates 1 and 3 and side-products identified by using LC–MS.
cious-metal catalysts on different supports and in different sol-
Table 2. Homogeneous hydrogenation of 1.[a]
vents, only Ir catalysts gave somewhat encouraging results
with good selectivity for the reduction of 1 to 2. Ir/CaCO3 (JM,
Type 30) gave a clean conversion to 2 in MeOH (808C, 5 bar
H2, 8 h, 5% dry weight catalyst loading) but the conversion re-
mained moderate at best (up to 67% 2) and the reaction
could not be optimised into a preparative process.[6]
Entry
Catalyst
S/C
T [8C]
Solvent
Base
(5%)
2
[b] [%]
1
2
3
4
5
6
7
15
16
17
17
17
18
18
100:1
50
50
50
50
60
50
60
MeOH
iPrOH
MeOH
iPrOH
iPrOH
MeOH
MeOH
–
39
4
3
71
36
100
99
1000:1
1000:1
1000:1
2000:1
1000:1
2000:1
tBuOK
tBuOK
tBuOK
tBuOK
tBuOK
tBuOK
As a result of the chemoselectivity problems associated with
the direct hydrogenation/hydrogenolysis of 1, it was decided
to switch the research focus to a two-step approach. Initially, 2
was easily prepared by the reduction of 1 with NaBH4. As an al-
ternative to NaBH4, catalytic reductions of 1 with homogene-
ous transfer-hydrogenation catalysts (Table 1) and hydrogena-
tion catalysts (Table 2) were examined. We exploited the fact
that homogeneous catalysts, which operate under an entirely
different mechanistic pathway from heterogeneous catalysts,
can display much higher chemoselectivity towards carbonyl re-
duction versus dechlorination and hydrogenolysis.
[a] Reactions were carried out under H2 (27 bar) for 16 h on a scale be-
tween 0.25 and 0.5 mmol (0.1–0.2m). [b] By HPLC analysis, XBridge C18,
4.6ꢁ150 mm, 228 nm.
Table 1. Homogeneous transfer hydrogenation of 1.[a]
Entry Catalyst Hydride source Amount
S/C
t [h]
2
[b] [%]
1
13
13
13
13
14
NH4OOCH
NH4OOCH
NH4OOCH
NaOOCH
10 equiv. 1000:1
16
20
100
100
99.5
41
2[c]
3
4 equiv.
4 equiv.
4 equiv.
5000:1
10000:1 16
5000:1
4
5
16
16
NH4OOCH
10 equiv. 1000:1
60
Figure 1. Homogeneous transfer-hydrogenation catalysts (13 and 14) and
hydrogenation catalysts (15–18). COD=1,5-cyclooctadiene.
[a] Reactions were performed at 808C in AcOEt/H2O 4:1 for 16 h on
a scale between 0.5 and 4 mmol (0.1–0.8m). [b] By HPLC analysis, XBridge
C18, 4.6ꢁ150 mm, 228 nm. [c] Reaction on a 14.2 mmol (6 g) scale.
mixtures) as well as sodium formate and ammonium formate
in a biphasic system of EtOAc/water. The biphasic system was
found to be particularly convenient to facilitate workup and
product isolation and rapidly allowed the reduction of the cat-
alyst loading from S/C (equivalents S=substrate/C=catalyst)
1000:1 (Table 1, entry 1) to S/C 5000:1 (entry 2) and S/C
10000:1 (entry 3). Ammonium formate was a more efficient re-
Following the original work by Wills on tethered Ru chiral
catalysts for asymmetric transfer hydrogenation,[8] we have re-
cently developed an achiral version of the catalyst, 13
(Figure 1).[9] Initial tests indicated that different reducing
agents were effective: formic acid/triethylamine (5:2 and 1:1
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1206

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ducing agent than sodium formate (entry 4), and the reaction
at S/C 5000:1 (w/w 4280:1, entry 2) was demonstrated on
a multi-gram scale. After 20 h, full conversion was achieved to
give 2 with a high isolated yield (97%). Although ammonium
formate has been previously used with homogeneous transfer
hydrogenation for the reductive amination of ketones,[10] selec-
tive conversion to 2 was obtained in this case. The non-teth-
ered Noyori catalyst 14[11] displayed a much lower activity
under partially optimised reaction conditions (entry 5). These
results confirmed the advantage of the tethered catalyst
design already reported in the area of chiral catalysis.[8,9] The
improved results are probably a consequence of increased cat-
alyst robustness in the presence of poly-functionalised sub-
strates.
Table 3. Small-scale hydrogenolysis of 2 with 5% Pd/C.[a]
Additive [mol%] Conv.[b] [%] 3[b] [%] Side products
Entry Solvent
1
THF/H2O
THF/HCl
80:20
THF/HCl
80:20
THF/HCl
80:20
THF/HCl
80:20
Toluene/HCl CuSO4 (10)
80:20
AcOEt/HCl CuSO4 (10)
80:20
iPrOAc/HCl CuSO4 (10)
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
NaCl
–
3
–
–
–
–
2[c]
53% 8 and 9,[d]
21% 10
3[e]
4
–
–
–
96% 8 and 9,[d]
CuSO4 (10)
CuCl2 (10)
43
41
27
52
76
76
75
76
78
87
86
81
50
–
–
–
–
–
–
5
31
6
94
7
97
8
>99
>99.5
>99.5
>99.5
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
As an alternative catalytic method for the ketone reduction
to alcohol, homogeneous achiral hydrogenation catalysts were
tested (Table 2, Figure 1). The Rh catalyst 15,[12] known to
reduce ketones, gave partial conversion at S/C 100:1 (Table 2,
entry 1). Low conversion was obtained with Noyori’s catalyst
16 (entry 2), which was used in iPrOH/tBuOK as generally re-
quired by this class of catalyst.[13] Baratta’s catalyst 17[14] provid-
ed up to 71% conversion to 2 at S/C 1000:1 in iPrOH/tBuOK
(entry 4) and 36% conversion at S/C 2,000:1 (entry 5). Catalyst
18[15] provided increased activity with full conversion at S/C
2000:1 in MeOH with 5% tBuOK (608C and 27 bar H2). The
higher activity associated with the use of a tridentate amine
ligand may be again associated with increased catalyst stability
in the presence of substrates that, such as 1, are capable of
various metal coordination modes.
9
CuSO4 (10)
CuCl2 (10)
Cu(OAc)2(10)
CuSO4 (1)
–
[f]
10
11
12
13
14
15
16
17
18
19
20
21
[f]
[f]
[f]
CuSO4 (0.5)
CuSO4 (0.25)
CuSO4 (0.125)
FeCl2 (1)
11% 11
37% 11
11% 8,79% 11
8% 8, 65% 11
12% 8,78% 11
8% 8, 69% 11
9% 8, 82% 11
NiCl2 (1)
–
Having in hand some options for a clean and efficient reduc-
tion of 1 to 2, we set out to study the hydrogenolysis step (2
to 3), choosing again a 5% Pd/C catalyst (JM, Type 5R39) as
the starting point (Table 3). No conversion was obtained in
THF/water in the presence of excess NaCl (entry 1). When the
reactions were conducted in a 4:1 mixture of THF and 2n
aqueous HCl (four equivalents of HCl to substrate), the main
reaction products that could be identified by direct LC–MS
analysis of these reactions were 8 and 9, derived from dech-
lorination at the pyridazine ring and hydrogenolysis of the
benzylic morpholine substituent, and 10, from further hydro-
genolysis of the benzylic alcohol (entry 2). Reduced reaction
temperatures only led to increased selectivity towards 8 and 9
(entry 3).
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
AcOH/HCl
80:20
CeCl3 (1)
–
CoCl2 (1)
–
MgBr2 (1)
–
Zn(OAc)2 (1)
Cu(OAc)2(1)
–
7% 8, 72% 11
22[g] AcOH/HCl
70:30
95
[f]
[a] All reactions were carried out under H2 (5 bar) at 708C in a Biotage En-
deavour reactor on a scale of 0.2–0.25 mmol of 2 (0.1m), with 5% Pd/C
5R39 (5 wt% on a dry basis) for 8–16 h. [b] By HPLC analysis, XBridge
C18, 4.6ꢁ150 mm, 228 nm. [c] 20 bar H2 and 708C. [d] The HPLC method
initially used (entries 1–3) did not separate products 8 and 9. [e] 20 bar
H2 and 308C. The same result was obtained under 5 bar H2. [f] Dimers
were detected in variable amounts. [g] 5% Pd/C A405038 (5 wt% on
a dry basis), HCl 1.33n, [2]=0.1m, 808C, 27.5 bar H2.
The breakthrough came when it was observed that the addi-
tion of Cu salts, CuCl2 and CuSO4, (10% to 2) had an extraordi-
nary effect on the reaction selectivity (Table 3, entries 4 and 5)
to produce for the first time significant amounts of 3. Follow-
ing these encouraging results, several solvents were tested in
combination with 2n aqueous HCl, and AcOH was chosen as
the preferred solvent for further optimisation (entries 6–9). Al-
ternative Cu salts were also tested, and in each case compara-
ble results were obtained (entries 9–11). The amount of Cu salt
additive was optimised in AcOH/aqueous HCl (entries 12–15)
and it was found that between 0.5% and 1% CuSO4 gave the
highest reaction purity (entries 12 and 13). However, below the
0.5% threshold, an increasing amount of dechlorinated prod-
uct 11 was formed (entries 14 and 15). Interestingly, when salt
additives that did not contain Cu were tested in AcOH/aque-
ous HCl (FeCl2, NiCl2, CeCl3, CoCl2, MgBr2, Zn(OAc)2), 3 was not
formed, and high amounts of dehalogenated product 11 (65–
82%) together with minor amounts of 8 were observed (en-
tries 16–21).
Further small-scale experimentation[6] in the presence of 1%
Cu(OAc)2 led to the adjustment of the amount of water from
20 to 30% of the total solvent volume (without taking into ac-
count that some water is introduced into the reaction from
the catalyst; 5% Pd/C is a paste that contains ca. 50% water
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1207

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
by weight) and to the identification of an alternative catalyst
(5% Pd/C, JM, Type A405038). The reaction temperature was
increased to 808C, the H2 pressure was increased to 27.5 bar
and, on a small scale, the reaction reproducibly gave 3 in 95–
96% HPLC purity (from direct HPLC analysis of the crude reac-
tion mixture) (Table 3, entry 22). The only significant side-prod-
ucts detected at this stage corresponded to late-eluting HPLC
peaks. Structural assignments made by using a combination of
The optimised hydrogenolysis conditions were 5% Pd/C
(5 wt% loading on a dry basis, JM, Type 5R39) with 1.4%
CuSO4 in 2.75 volumes of H3PO4, 2.75 volumes of 5n HCl (six
equivalents to substrate) in an autoclave under 34.5 bar
(500 psi) of H2 at 608C for 24 h (Table 4, entry 6). Although a re-
action performed at 508C gave a slightly lower conversion
(entry 7), the other reactions in H3PO4/aqueous HCl (entries 5
and 8) reliably resulted in>99.5% consumption of 2 and <6%
area impurities by HPLC. The reaction was estimated to be
completed in 14–15 h; little to no increase in impurity levels
were observed upon prolonged exposure of the product to
the hydrogenolysis conditions. The reaction workup involved
catalyst filtration, adjustment of the pH of the aqueous phase
with 50% NaOH to approximately 7 and the addition of tolu-
ene. Compound 3 was observed to partition into the organic
layer. Partial distillation of the toluene layer followed by the
addition of heptane resulted in the crystallisation of 3. The iso-
lated yields were typically 70–80%, with an additional 10–15%
lost in the mother liquor.
1
LC–MS and H NMR spectroscopy suggested the presence of
dimers.[17]
The reaction required the presence of Cu in molar amounts
similar to that of Pd (therefore, catalytic with respect to 2), and
there appeared to be an induction time for the formation of
a more chemoselective catalyst with >50% of the total impuri-
ties formed in the first 2–3 h of the reaction.[6] Further analysis
of the crude reaction solution after catalyst separation, indicat-
ed a significant reduction in the amount of solubilised Cu, ap-
proximately 75% reduction, calculated by comparison to a con-
trol reaction with no added Pd/C or substrate. In conjunction,
analysis of the separated Pd/C catalyst indicated that a signifi-
cant amount of Cu was present.[6] Additional experiments con-
firmed that no reaction occurred in the presence of Cu salts
alone (without Pd/C) or by replacing Pd/C with PdCl2
(1 mol%).[6] In the latter case only small amounts of O-acetylat-
ed derivative 12 (4%) were detected.[20]
Conclusions
A stepwise reduction of 1 to 2 and 2 to 3 has been demon-
strated, which provides synthetically useful chemoselectivity at
acceptably low catalyst loadings. In the presence of supported
metal catalysts, 1 was mostly unreactive towards C=O reduc-
tion, although other undesired reactions took place more
easily. However, we took advantage of the inherently higher
chemoselectivity of homogeneous catalysts to overcome the
otherwise intractable problem of the reduction of 1 to 2. In
particular, Wills’ tethered Ru transfer-hydrogenation catalyst 13
and Baratta’s pincer Ru hydrogenation catalyst 18 showed su-
perior reactivity in the presence of highly functionalised 1.
Substrate 2 became amenable to hydrogenolysis to 3 in the
presence of Cu salts, an effect that, to the best of our knowl-
edge, has never been reported. Several salts were tested as ad-
ditives but only Cu salts prevented both dechlorination side-re-
actions and hydrogenolysis of the benzylic morpholine. The
use of an excess of acids (e.g., HCl) is well established to accel-
erate the hydrogenolysis of benzylic alcohols as well as to pre-
vent aromatic dechlorination.[5] An additional, well-established
effect of the acidic environment is to protonate the basic het-
erocyclic sites of both substrate and products and to prevent
catalyst deactivation.[5] On the contrary, the exact role of the
Cu additives in such a complex catalytic system is only
a matter of hypothesis. Cu salts completely inhibited the re-
duction of 1 but were necessary to achieve the chemoselective
hydrogenolysis of 2.
Unfortunately, the reactions in AcOH/aqueous HCl were diffi-
cult to reproduce on a multi-gram scale (10–12 g). The best re-
action conditions given in Table 3 were repeated in 25 and
50 mL stainless-steel autoclaves but gave reduced conversion
to 3 (Table 4, entries 1 and 2). A combination of increased
Table 4. Multi-gram-scale hydrogenolysis of 2.[a]
Entry Solvent
t [h] T [8C] Conv.[b] [%]
3
[b] [%] Yield
3 [%][c]
1[d]
2[d]
3
4
5
6
7
8
AcOH/HCl 70:30 17
AcOH/HCl 70:30
80
70
60
50
60
60
50
60
>99
98
99.4
98
99.7
99.8
97
60
70
–
–
8
AcOH/HCl 50:50 17
AcOH/HCl 50:50 46
H3PO4/HCl 50:50 18
H3PO4/HCl 50:50 24
H3PO4/HCl 50:50 24
H3PO4/HCl 50:50 24
91.5
92
94.2
94.9
92.7
94.3
63 (98.4)
72 (98.2)
67 (99.3)
76 (98.9)
80 (97)
80 (99)
99.7
[a] Reactions were carried out on a 10–12 g scale with 5% Pd/C JM 5R39
(5 wt% on a dry basis) and 1.4% CuSO4 in an autoclave under 34.5 bar
H2. [b] By HPLC analysis, XBridge C18, 4.6ꢁ150 mm, 228 nm. [c] HPLC
purity of isolated 3 in brackets. [d] Catalyst: 5% Pd/C A405038 (5 wt% on
a dry basis), 1% Cu(OAc)2, 20 bar H2.
The use of Cu modifiers on supported Pd catalysts under hy-
drogenation conditions has some precedent in areas as differ-
ent as selective dechlorination in the presence of C=C
bonds,[21] denitration of water[22] and diastereoselective imine
reduction.[23] Literature precedents usually employ pre-formed
bimetallic Pd-Cu catalysts.[22,23] Depending on the application,
it has been suggested that higher selectivity is associated with
the presence of CuII or, more specifically, that a catalytic cycle
occurs in which Cu(0) is oxidised to CuO (e.g., in the NO2 to
water content (AcOH/aqueous HCl 6n), reduced reaction tem-
peratures, and a move to Hastelloy Parr autoclaves increased
the conversion to 3 (entries 3 and 4).[18] Finally, phosphoric acid
was identified as an additional suitable reaction medium (en-
tries 5–8), which appeared to have a slight advantage over
acetic acid in terms of reproducibility and overall impurity pro-
file, to provide a small but reproducible increase in the conver-
sion to 3.[19]
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1208

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
removed, and the aqueous layer extracted with additional toluene
(50 mL). The combined organic layers were washed with aqueous
NaHCO3 (0.5m, 60 mL) and water (2ꢁ25 mL). Occasional heat was
applied to the solutions to avoid haziness/product precipitation
during the extractions. The toluene layer was then concentrated to
a volume of approximately 36 mL in a 250 mL flask at 50–608C.
The product solution was held at 608C while heptane (144 mL)
was added dropwise over 45 min; 20 mL into the heptane addi-
tion, a small amount of seed crystals of 3 was added, which in-
duced product crystallisation. When the heptane addition was
complete, the slurry was cooled from 60 to 08C over 6 h, and
stirred overnight. The solids were isolated by vacuum filtration and
were washed with 20% toluene in heptane (36 mL). The solid was
dried in vacuo to afford 8.95 g (81.1%). Quantitative HPLC analysis
of the filtrate revealed a loss of 1.45 g (3.54 mmol, 13.2%).
NO reduction step), which is then reduced by activated hydro-
gen from the neighbouring Pd atom.[22] Based on literature
data and our own analysis of the reaction mixture after separa-
tion of the catalyst, it can be envisaged that under the reaction
conditions Cu precipitation occurs to form a metal layer, which
acts as modifier of the Pd catalyst.
The current combination of homogenous achiral catalysts
and Pd/C in the presence of Cu additives provides the basis for
a viable process. Our results highlight the benefits of open-
minded experimentation with both homogenous and hetero-
geneous hydrogenation technology for achiral transformations
of synthetic importance.
Experimental Section
Reagents and catalysts: Heterogeneous catalysts are commercially
available from Johnson Matthey.[24] Homogeneous catalysts[9,11–15]
and 1[1,2] were prepared according to literature procedures.
Acknowledgements
We thank Prof. Walter Baratta for supplying a sample of catalyst
HPLC analysis: Waters XBridge C18 column, 4.6ꢁ150 mm, 3.5 mm
particle size; flow rate=1.5 mLminÀ1
; T=308C; detection at
18.
228 nm. Solvent A: NH4OH in water (0.1 mLLÀ1); Solvent B: NH4OH
in CH3CN (0.1 mLLÀ1). Gradient elution: 70% A at t=0 min to 15%
A at t=8 min, 15% A at t=15 min to 70% A at t=16 min, 18 min
total run time. Retention times: 2: 7.4 min; 1: 8.5 min, 3: 9.7 min.
Synthesis of 2: A 100 mL round-bottomed flask with a magnetic
stirrer bar was charged with 1 (6.0 g, 14.2 mmol), ammonium for-
mate (3.57 g, 56.7 mmol) and 13 (1.4 mg, S/C 5000:1). The flask
was purged with N2, and H2O (7.1 mL) and EtOAc (28 mL) were
added. The slurry was heated to 808C for 20 h and then cooled to
room temperature. The reaction mixture was diluted with EtOAc
(30 mL), and the aqueous phase was separated. The organics were
washed with H2O (3ꢁ10 mL) and brine (10 mL), dried (MgSO4) and
concentrated under reduced pressure. The crude product (5.83 g,
97%) was obtained in >99% HPLC purity. Pale yellow powder;
Keywords: hydrogenation
catalysis · ruthenium · homogeneous catalysis
·
palladium
·
heterogeneous
[1] T. P. Burkholder, J. R. Clayton, L. Ma, Amino pyrazole compound. US Pat.
Appl. Publ. US 20100152181A1 20100617; CAN 153:97762, AN
2010:753991.2010.
[2] D. Mitchell, K. P. Cole, P. M. Pollock, D. M. Coppert, T. P. Burkholder, J. R.
Clayton, Org. Process Res. Dev. 2012, 16, 70–81.
[3] R. A. Benkeser, Acc. Chem. Res. 1971, 4, 94–100.
[4] a) L. D. Hicks, J. K. Han, A. J. Fry, Tetrahedron Lett. 2000, 41, 7817–7820;
b) G. G. Wu, F. X. Chen, D. LaFrance, Z. Liu, S. G. Greene, Y.-S. Wong, J.
Xie, Org. Lett. 2011, 13, 5220–5223; c) J. E. Milne, T. Storz, J. T. Coyler,
O. R. Thiel, M. D. Seran, R. D. Larsen, J. A. Murry, J. Org. Chem. 2011, 76,
9519–9524.
[5] a) G. V. Smith, F. Notheisz, Heterogeneous Catalysis in Organic Chemistry,
Academic Press, San Diego, USA, 1999; b) R. L. Augustine, Heterogene-
ous Catalysis for the Synthetic Chemist, Marcel Dekker, New York, 1996;
c) P. N. Rylander, Catalytic Hydrogenation in Organic Synthesis, Academic
Press, New York, USA, 1979.
[6] Detailed experimental results are reported in the Supporting Informa-
tion.
[7] G. Wu, M. Huang, M. Richards, M. Poirier, X. Wen, R. W. Draper, Synthesis
2003, 11, 1657–1660.
[8] a) A. M. Hayes, D. M. Morris, G. J. Clarkson, M. Wills, J. Am. Chem. Soc.
2005, 127, 7318; b) D. J. Morris, A. M. Hayes, M. Wills, J. Org. Chem.
2006, 71, 7035; c) F. K. Cheung, A. J. Clarke, G. J. Clarkson, D. J. Fox,
M. A. Graham, C. Lin, A. Lorente Crivillꢃ, M. Wills, Dalton Trans. 2010, 39,
1395–1402.
1
m.p. (toluene/heptane)=133.0–134.08C; H NMR (400 MHz, CDCl3):
d=7.73 (t, 3J(H,H)=8.4 Hz, 1H; CH), 7.23 (s, 1H; CH), 7.21 (dd,
3J(H,H)=1.6, 8.4 Hz, 1H; CH), 7.04 (dd, 3J(H,H)=1.6, 10.4 Hz, 1H;
3
CH), 6.56 (brs, 1H; CH), 4.01 (brd, J(H,H)=4.4 Hz, 1H; OH), 3.94 (s,
3
3
2H; CH2), 3.77 (t, J(H,H)=4.4 Hz, 4H; CH2), 2.58 (t, J(H,H)=4.4 Hz,
4H; CH2), 2.31 ppm (s, 3H; CH3); 13C NMR (100 MHz, CDCl3): d=
159.4 (d, J=249 Hz), 146.5, 141.5, 138.5, 136.3, 134.4 (d, J=
10.3 Hz), 128.9 (d, J=4.6 Hz), 126.4 (d, J=12.0 Hz), 124.4 (d, J=
12.2 Hz), 116.2, 116.0, 115.7, 66.9 (2C), 61.5 (d, J=3.0 Hz), 55.9, 53.8
(2C), 14.2 ppm; 19F NMR (376 MHz, CDCl3): d=À114.9 ppm (t, J=
9.4 Hz); IR (neat): n˜ =3268 (br, OH), 2959 (w), 2863 (w), 1610 (m),
1579 (m), 1544 (s), 1486 (m), 1441 cmÀ1 (s); HRMS m/z: calcd for
C19H20Cl2FN4O2: 425.0942 [M+H]; found: 425.0940.
Synthesis of 3:[2] To a 160 mL Hastelloy Parr reactor were charged
2 (12.05 g, 26.97 mmol), CuSO4 (61 mg, 0.38 mmol), wet Pd/C
(1.44 g, JM 5R39, 5 wt% on a dry basis), phosphoric acid (32 mL)
and HCl (32 mL, 5n). The reactor was purged twice with N2 and
three times with H2. The reaction was placed under 34.5 bar H2
with stirring at 400 rpm and heated to 608C. After 24 h, the reac-
tion mixture was cooled and purged with N2. HPLC analysis
showed 99.7% conversion of 2 and 5.4% area impurities. Toluene
(50 mL) was added, and the slurry was stirred for 30 min. The slurry
was then filtered through a bed of Hyfloꢂ Super Celꢂ (filter aid, flux
calcined, treated with Na2CO3), which was washed with water
(36 mL) and then toluene (50 mL). The combined filtrates were
added to water (20 mL) and toluene (20 mL). NaOH (35 mL, 50%
solution) was then slowly added to the biphasic mixture (exother-
mic!) to adjust the pH to approximately 7. The organic layer was
[9] K. E. Jolley, A. Zanotti-Gerosa, F. Hancock, A. Dyke, D. M. Grainger, J. A.
Medlock, H. G. Nedden, J. M. Le Paih, S. J. Roseblade, A. Seger, V. Sivaku-
mar, I. Prokes, D. J. Morris, M. Wills, Adv. Synth. Catal. 2012, 354, 2545–
2555.
[10] For example: J. S. D. Boggs, J. D. Cobb, K. J. Gudmundsson, L. A. Jones,
R. T. Matsuoko, A. Millar, D. E. Patterson, V. Samano, M. D. Trone, S. Xie,
X.-M. Zhou, Org. Process Res. Dev. 2007, 11, 539–545.
[11] a) Y. Crameri, K. Puentener, M. Scalone (F. Hoffman–La Roche AG),
EP0915076 B1; b) The chiral version of the catalyst was first described
by Noyori: S. Hashiguchi, A. Fujii, J. Takeara, T. Ikariya, R. Noyori, J. Am.
Chem. Soc. 1995, 117, 7562–7563; c) K.-J. Haack, S. Hashiguchi, A. Fujii,
T. Ikariya, R. Noyori, Angew. Chem. 1997, 109, 297–300; Angew. Chem.
Int. Ed. Engl. 1997, 36, 285–288.
[12] a) M. J. Burk, T. G. P. Harper, J. R. Lee, C. Kalberg, Tetrahedron Lett. 1994,
35, 4963–4966; b) I. R. Butler, W. R. Cullen, T. J. Kim, Synth. React. Inorg.
Met.-Org. Chem. 1986, 75, 109–116.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1209

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[13] H. Doucet, T. Ohkuma, K. Murata, T. Yokozawa, M. Kozawa, E. Katayama,
A. F. England, T. Ikariya, R. Noyori, Angew. Chem. 1998, 110, 1792–1796;
Angew. Chem. Int. Ed. Engl. 1998, 37, 1703–1707.
[14] W. Baratta, E. Herdtweck, K. Siega, M. Toniutti, P. Rigo, Organometallics
2005, 24, 1660–1669.
which resulted in large quantities of inorganic solids unless a copious
amount of water was employed.
[20] Independent preparation of the acetylated product 12 and its submis-
sion to standard hydrogenolysis conditions showed no higher conver-
sion than the reaction that started from 2. Although 12 may be re-
duced to 3, there is no evidence that the formation of 12 is necessary
for the catalytic cycle.
[15] W. Baratta, E. Herddtweck, K. Siega, M. Toniutti, P. Rigo, Chem. Eur. J.
2008, 14, 9148–9160.
[16] Optimised reaction conditions shown in Table 3 were re-tested with 1,
but resulted in no conversion.
[21] T. Mallꢄt, J. Petrꢅ, Appl. Catal. 1990, 57, 71–81. The improved selectivity
of Pd/C catalysts in dehalogenation versus C=C hydrogenation in pyri-
dine and in the presence of Et3N and Cu salts was attributed to metal
deposition on Pd, which partly modifies the active sites.
[22] J. Sꢄ, N. Barrabꢃ, E. Kleymenov, C. Lin, K. Fçttinger, J. A. van Bokhoven,
M. Nachtegaal, A. Urakawa, G. A. Crespo, G. Rupprechter, Catal. Sci.
Technol. 2012, 2, 794–799 and references therein. Bimetallic Pt-Cu and
Pd-Cu catalysts are used in water denitration.
[23] J. Mꢆslehiddinog˘lu, J. Li, S. Tummala, R. Deshpande, Org. Process Res.
Dev. 2010, 14, 809–894. A bimetallic Pd-Cu/C catalyst (Johnson Matthey
A701023-4) was used for the highly diastereoselective imine hydroge-
nation. The Pd-Cu/C catalyst gave a higher selectivity than Pd-Cu/Al2O3,
which was attributed to the higher amount of CuII in the Pd-Cu/C cata-
lyst.
[17] The mechanism by which these dimers are formed is unclear. Their for-
mation was strongly impacted by the amount of Cu and H2 pressure,
and low pressures (<6.5 bar) resulted in higher levels of impurities. In
the author’s experience with other chemical transformations of the aryl-
(imidazo[1,2-b]pyridazinyl)methane moiety, this is prone to oxidation
especially under radical conditions, possibly because of the existence of
a highly stabilised and delocalised radical or ionic reactive intermediate.
[18] Small-scale screening reactions were performed in glass-lined Biotage
Endeavour reactors. However, upon performing the reactions in 25 mL
316 stainless-steel Parr reactors, etching and corrosion were observed
on the surface of the metal reactor, which were associated with a green-
ish colour of the reaction mixture and to degradation of the reaction
performance with higher levels of dimeric impurities being formed.
Upon switching to Hastelloy Parr reactors, no further corrosion was
noted.
[24] Handbook of Pharmaceutical Catalysis, 2009, available at www.jmcata-
lysts.com/pharma.
[19] Acetic acid was felt to be potentially advantageous in the workup be-
cause of the possibility for distillative removal. Extraction of the desired
free-base product from the phosphoric acid mixture was complicated
by the necessary partial quenching of the phosphoric acid with base,
Received: August 1, 2012
Published online on February 1, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 1205 – 1210 1210

Copyright of ChemCatChem is the property of Wiley-Blackwell and its content may not be copied or emailed to
multiple sites or posted to a listserv without the copyright holder's express written permission. However, users
may print, download, or email articles for individual use.