Selective Hydrogenation of Halogenated Nitroaromatics to Haloanilines in Batch and Flow

Article abstract of DOI:10.1021/acs.oprd.5b00170

The selective hydrogenation of functionalized nitroaromatics poses a major challenge from both academic as well as industrial viewpoints. As part of the CHEM21 initiative (www.chem21.eu), we are interested in highly selective, catalytic hydrogenations of halogenated nitroaromatics. Initially, the catalytic reduction of 1-iodo-4-nitrobenzene to 4-iodoaniline served as a model system to investigate commercial heterogeneous catalysts. After determining optimal hydrogenation conditions and profiling performances of the best catalysts, hydrogenations were transferred from batch to continuous flow. Finally, the optimized flow conditions were applied to transformations which represent important steps in the syntheses of the active pharmaceutical ingredients clofazimine and vismodegib.

Full text of DOI:10.1021/acs.oprd.5b00170

Article
pubs.acs.org/OPRD
Selective Hydrogenation of Halogenated Nitroaromatics to
Haloanilines in Batch and Flow
,†,‡
‡
,†
†
†
‡
Patrick Loos,* Hannes Alex, Jorma Hassfeld,* Kai Lovis, Johannes Platzek, Norbert Steinfeldt,
‡
and Sandra Hu
†
̈
bner
Bayer Pharma AG, Friedrich-Ebert-Str. 217-333, D-42117 Wuppertal, Germany
‡
Leibniz-Institut fu
̈
r Katalyse e.V. (LIKAT), Albert-Einstein-Str. 29a, D-18059 Rostock, Germany
*
S Supporting Information
ABSTRACT: The selective hydrogenation of functionalized nitroaromatics poses a major challenge from both academic as well
as industrial viewpoints. As part of the CHEM21 initiative (www.chem21.eu), we are interested in highly selective, catalytic
hydrogenations of halogenated nitroaromatics. Initially, the catalytic reduction of 1-iodo-4-nitrobenzene to 4-iodoaniline served
as a model system to investigate commercial heterogeneous catalysts. After determining optimal hydrogenation conditions and
profiling performances of the best catalysts, hydrogenations were transferred from batch to continuous flow. Finally, the
optimized flow conditions were applied to transformations which represent important steps in the syntheses of the active
pharmaceutical ingredients clofazimine and vismodegib.
INTRODUCTION
by far the most widespread support for hydrogenation catalysts,
very likely due to its competitive price.
■
Nitration is one of the most versatile ways of introducing
nitrogen into aromatic compounds. The resulting nitro-
aromatics can easily undergo further functionalization, e.g. by
nucleophilic aromatic substitution or hydrogenation. This is
one reason for their widespread use en-route to highly
functionalized molecules. For example, 2-chloroaniline, 4-
chloroaniline, and 2,4-dichloroaniline can be converted into a
variety of functional molecules like active pharmaceutical
ingredients, agrochemicals, or pigments. These anilines are
produced by nitration of chlorobenzene and dichlorobenzene
followed by selective reduction of the nitro group. However, it
can be challenging to hydrogenate the nitro group in densely
functionalized aromatics selectively. Many catalysts will also
reduce functionalities such as multiple bonds, cyano groups,
heterocycles, benzyl protecting groups, or halogens in addition
to the nitro group. Therefore, the selective reduction of
nitroaromatics to anilines has been subject of substantial
High yield and selectivity are of major importance in order to
develop efficient processes, but aspects of sustainability have to
be considered to promote the continuing improvement in this
field. Hence, low-cost, abundant metals, and solvents, which
can be produced in a sustainable fashion, should be taken into
account.
With these considerations in mind we set out to investigate
selective reductions of nitro aromatics, initially focusing on
retaining halogens in the substrate. In order to facilitate
separation, we confined this study to heterogeneous catalysts
and the most atom-efficient reducing agent hydrogen.
Continuous processes in microstructured reactors can offer
excellent control over residence time, reaction temperature, and
1
e
2
12,13
mixing efficiency.
They can also be advantageous for highly
exothermic reactions and for the synthesis of unstable or
14
explosive compounds. Providing large interfaces can be
especially important for multiphasic hydrogenation reactions
1
,3
research efforts over the last decades.
While well-defined homogeneous catalysts allow fine-tuning
of the catalytic activity, we decided to focus this study on
15
involving a gas phase, a liquid phase and a solid catalyst bed.
Until today, the continuous hydrogenation of nitroaromatics in
4
16,17
heterogeneous catalysts to facilitate product separation. Apart
from the abundantly used heterogeneous Pt- and Pd-catalysts,
other metals like Ru, Rh, Ag, Ir, and Au, have also been used for
hydrogenations, with Au-based catalysts showing exceptionally
liquid phase has been investigated only to a limited extent.
Therefore, we compared promising catalysts for different
halogen-substituted nitroaromatics, so as to provide an
informed and balanced comparison of batch and flow processes.
Initially, the selective hydrogenation of 1-iodo-4-nitro-
benzene 1a to 4-iodoaniline 4a served as a model system
5
,6
high chemoselectivity. Considering the limited supply of
conventional resources, catalysts based on the nonprecious
7
8
8b,d,9
transition metals iron, cobalt, or nickel
should be
(Figure 1), which is already well-known in the litera-
1
0
1d,11b,d,17a
preferred over platinum group metals. However, catalyst
activity, space-time yield, catalyst recycling (or long-term
performance for continuous reactions), and ease of metal
recycling are also important factors for the evaluation of the
sustainability of a process. A variety of supports have been used
to immobilize catalytically active metals. Activity and selectivity
of the resulting catalysts can be influenced by both the metal
ture.
Mechanistically, reduction of 1-iodo-4-nitro-
benzene 1a can proceed via nitroso compound 2 and
Special Issue: Continuous Processing, Microreactors and Flow
Chemistry
Received: May 29, 2015
11
and the respective support. Nevertheless, activated carbon is
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
1
e
Figure 1. Mechanistic pathways for the reduction of 1-iodo-4-nitrobenzene.
Figure 2. Product composition in the batch hydrogenation of 1-iodo-4-nitrobenzene 1a for different catalysts after 2/4 h reaction time (data point
with higher amount of 4-iodoaniline 4a shown). Dark gray: 4-iodoaniline 4a, white: nitrobenzene 5a, black: aniline 6a, light gray-striped:
unmonitored/adsorbed compounds based on detected starting material. The absolute height of the bars reflects conversion of the starting material;
Reaction conditions: 2 mmol 1-iodo-4-nitrobenzene 1a in 10 mL of THF/H O (95:5), 1 mol % catalyst (*Raney Co: 15 mol %, 110 °C, 20 bar H ),
2
2
5
bar H , 80 °C, 2 or 4 h. Internal analytical GC standard: diethylene glycol dibutyl ether.
2
hydroxylamine 3 to iodoaniline 4a. In this sequence, reduction
of hydroxylamine 3 is usually the rate limiting step.
of hydrodehalogenation can affect solubility of the products and
the pH of the reaction mixture.
1
e
Dehalogenation of the nitro starting material (1a → 5a) is
The seemingly simple hydrogenation proceeds through a
reaction network with many intermediates, which can be
influenced by multiple parameters. At the same time,
iodoaniline 4a is highly prone for hydrodehalogenation.
Trapping of this intermediate is therefore a challenging task.
possible, but the dehalogenation of the more electron-rich
1e
aniline product 4a is usually faster (4a → 6a). Another
possible pathway can accrue from condensation of these
intermediates. For example, nitroso (2) and hydroxylamine (3)
intermediate can condense to give azoxy compound 7.
Similarly, nitroso intermediate 2 can react with aniline 4a to
form diazo compound 8. These compounds can be reduced to
iodoaniline 4a via a hydrazo intermediate 9. Additionally, the
formation of water during in the hydrogenation and HI in case
BATCH REACTION OPTIMIZATION
■
Catalyst Screening. Using the 1-iodo-4-nitrobenzene
model system, suitable catalysts were identified by a broad
screening of commercially available catalysts (Figure 2). In
B
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 3. (a−b) Hydrogenation of 1-iodo-4-nitrobenzene 1a in different solvents; conversions/yields represent mean values of two experiments;
reaction conditions: 1 mol % Pt−V/C, 25 bar H , 95 °C, 0.05 M 1-iodo-4-nitrobenzene, 10 mL of the corresponding solvent, analytical standard:
2
diethylene glycol dibutyl ether; (a) conversion of 1-iodo-4-nitrobenzene 1a in squares: THF; circles: MeOC H ; diamonds: EtOAc; triangles:
5
9
anisole; (b) yield of aniline 6a (det. by GC): squares: THF; circles: MeOC H ; diamonds: EtOAc; triangles: anisole; all data points below the dotted
5
9
line were measured at <100% conversion of 1a.
these batch experiments, samples were taken after 2 and 4 h in
order to accommodate for varying hydrogenation rates.
Relative and absolute amounts of compounds 1a, 4a, 5a, and
15 mol % Raney Co at 5 bar H , 80 °C resulted in only 2%
conversion after 4 h. Pt−V/C and Raney Co were chosen as
representative precious and base metals for further studies. The
first step was a solvent and additive screening using Pt−V/C as
catalyst.
2
6
a in the reaction mixture (i.e., the supernatant) were
determined by GC against diethylene glycol dibutyl ether as
nonreactive internal analytical standard. Each bar in Figure 2
shows the composition of the reaction mixtures at the specified
reaction time. The absolute height of the bar represents the
conversion of the starting material. Therefore, the deviation
between conversion and the sum of the monitored products
can be attributed to unmonitored compounds (i.e., compounds
of the diazo pathway) or compounds, which remain adsorbed
to the catalyst surface. Further optimization might suppress the
formation of the unknown intermediates in favor of 4-
iodoaniline 4a.
Solvent and Additive Screening. Solvents can represent
a substantial fraction of the material used for the production of
19
chemicals. The development of more sustainable processes
should therefore include the screening of solvents which can be
20
produced in a sustainable fashion.
Toward this end we screened the solvents 2-methyl-THF,
methoxycyclopentane (MeOC H ), ethyl acetate, and anisole
5
9
to compare reaction rates, selectivities as well as solubility of
substrate 1a in these solvents in analogy to THF. The more
polar solvents DMF and N-methyl-2-pyrrolidone have been
omitted from this study due to their high toxicity. Alcohols
have not been investigated due to very low solubility of the
substrates in these solvents (see below).
Solubility is especially important for continuous reactions in
order to prevent potential precipitation issues and subsequent
clogging in the reactor. At ambient temperature, the model
substrate 1-iodo-4-nitrobenzene 1a was soluble in THF in
concentrations of approximately 0.6 M (149 g/L; in the
presence of the analytical standard diethylene glycol dibutyl
ether solubility decreased drastically). In EtOAc the solubility
was between 0.2 and 0.25 M, while in EtOH solubility it was
less than 0.05 M.
18
In line with earlier findings, low conversion was observed
for Pd(S)/C, which shows less than 10% conversion of 1-iodo-
4-nitrobenzene 1a after 4 h (a control reaction without catalyst
did not give any conversion after 4 h, not shown). For many
other catalysts hydrogenation of the nitro group was
accompanied by significant hydrodehalogenation. It has to be
taken into account that the product 4-iodoaniline 4a is
significantly more prone to dehalogenation than the starting
1e
material 1-iodo-4-nitrobenzene 1a. At the same time, nitro
reduction is usually faster than hydrodehalogenation. There-
fore, examples where 1-iodo-4-nitrobenzene 1a is fully
converted show comparatively large amounts of aniline 6a. In
order to achieve precise control over the reaction time on
technical scale, the use of a flow reactor might prove
advantageous.
In the next step, we compared the conversion of 1a in the
chosen solvents. To prevent compromising solubility issues, a
substrate concentration of 0.05 M was used. Figure 3a shows
that almost full conversion (99.5%) was observed in THF after
Particularly Pt−Fe and Pt−V based catalysts on activated
carbon gave promising results. The most selective catalyst in
this screening was Raney Co with less than 1% overall
dehalogenation, albeit 15 mol % had to be used at elevated H2
pressure and temperature to adjust for its lower activity. Using
5 min. In MeOC H , which was comparable to 2-methyl-THF
5 9
(Supporting Information Figure S1), full conversion was
observed after 15 min. EtOAc also showed complete
C
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 4. (a) Product composition in the batch hydrogenation of 1-iodo-4-nitrobenzene 1a with Pt−V/C for different concentrations; conversions/
yields represent mean values of two experiments; reaction conditions: 1 mol % Pt−V/C, 25 bar H , 95 °C, 10 mL of THF, analytical standard:
2
diethylene glycol dibutyl ether; (b) product composition in the batch hydrogenation of 1-iodo-4-nitrobenzene 1a with Raney Co for different
concentrations; conversions/yields represent mean values of two experiments; reaction conditions: 15 mol % Raney Co, 20 bar H , 110 °C, 10 mL
2
THF, analytical standard: diethylene glycol dibutyl ether; dark gray: 4-iodoaniline 4a, black: aniline 6a, light gray-striped: unmonitored/adsorbed
compounds based on detected starting material. The absolute height of the bars reflects conversion of the starting material.
2
5
consumption of the starting material after 15 min, while in
anisole full conversion was observed after 40 min.
Below 100% conversion, similar levels of dehalogenation
were observed for these solvents (Figure 3b, values below the
dotted line). Maximal 2.6% dehalogenation were observed
tion. Slower nitro hydrogenation might be a side effect of this
phenomenon. However, in contrast to the results reported for
Pt/C, using ZnI in combination with Pt−V/C did not lead
2
5
2
to significantly lower dehalogenation at high conversion.
Concentration Screening. Initially, the concentration
screening for the hydrogenation was performed in batch with
Pt−V/C as a catalyst. It was observed that substrate
concentration plays a significant role for the hydrogenation
rate and selectivity of the reaction (Figure 4a). At 0.2 M
concentration in THF, 94% of the substrate 1-iodo-4-
nitrobenzene 1a were converted after 240 min. With a 0.1 M
solution 93% conversion was observed after 21 min. At 0.05 M
(EtOAc, anisole) when substrate 1a was still present in the
reaction mixture. Upon full conversion, the dehalogenation rate
increased significantly in all investigated solvents.
In summary, using alkyl ethers like THF or 2-methyl-THF
(15 min: full conversion, 2.5% aniline 6a, Figure S1) resulted in
high reaction rates together with good substrate solubility. In
anisole (and toluene), much lower reaction rates were
observed. EtOAc showed a high reaction rate, but lower
substrate solubility compared to THF. For further optimization,
THF was chosen to allow for maximum concentration, fast
turnover, and maintaining high selectivity.
9
9.5% were consumed after 5 min. At the same time,
dehalogenation decreased with lower substrate concentration
from 27% over 8% to 1%.
On the other hand, in order to achieve high conversions
together with high selectivity, precise control over the reaction
time was found to be crucial. For example, at 0.05 M
concentration 1.4% dehalogenation were observed after 5 min
As shown above, Pt-based catalysts usually offer high catalytic
activity but often suffer from decreasing selectivity at
conversions close to 100%. Hence, numerous modifiers for
21
Pt-based catalysts have been reported. Using modifiers can
not only increase selectivity but can also suppress the formation
of hydroxylamine intermediates, whose accumulation can lead
to run-away reactions. Vanadium seems to be useful in this
(
at 99.5% conversion). After 10 min, full conversion and 6.8%
aniline were observed (Figure 3). Residence time can be
controlled very precisely in continuous flow reactors, which
might prove advantageous in order to achieve conversions close
to 100%, while separating the dehalogenation prone product
from the catalyst as quickly as possible.
The same concentrations were employed for hydrogenations
with Raney Co as catalyst (Figure 4b). In contrast to reactions
with Pt−V/C, conversions between 77 and 82% were observed
after very similar reaction times of 80−100 min for the different
concentrations. For Raney Co, dehalogenation did not vary
significantly for different concentrations, less than 2% aniline
were detected. Hence, hydrogenations using Raney Co could
allow to maximize substrate concentration and reduce solvent
consumption without significant decrease in product selectivity.
Even at full conversion, dehalogenation increased only
2
2
23
regard for Pt- as well as for Ni-based catalysts. Thiol
21
modifiers have been added in order to retain vinyl groups.
The presence of heteroaryl halides was tolerated well when a
Pt(S)/C catalyst was used. Remarkably, unmodified Pd/C
and Pt/C catalysts have been used for the selective hydro-
genation of 1-iodo-2-methyl-4-nitrobenzene by simply adding
24
25
0
.4% zinc iodide to the reaction mixture. It was therefore
applied to the hydrogenation of our model substrate 1-iodo-4-
nitrobenzene 1a with Pt−V/C (Supporting Information, Figure
S2). Interestingly, the use of ZnI slowed down hydrogenation
2
of starting material 1a as well as dehalogenation of 4-
iodoaniline 4a. It has been suggested that ZnI might block
2
highly active catalytic centers and thus suppress dehalogena-
D
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 5. Product composition over time in 200 mL batch hydrogenations of 1-iodo-4-nitrobenzene 1a with Pt−V/C and Raney Co; conversions/
yields represent mean values of two experiments; (a) reaction conditions: 1 mol % Pt−V/C, 25 bar H , 95 °C, 200 mL of THF, analytical standard:
2
diethylene glycol dibutyl ether; (b) reaction conditions: 15 mol % Raney Co, 20 bar H , 110 °C, 10 mL THF, analytical standard: diethylene glycol
2
dibutyl ether; squares: conversion of 1-iodo-4-nitrobenzene 1a; circles: 4-iodoaniline 4a; diamonds: aniline 6a.
Figure 6. Hydrogenation with Raney Co on different scales; conversions/yields represent mean values of two experiments; reaction conditions: 15
mol % Raney Co, 20 bar H , 110 °C, 0.05 M 1-iodo-4-nitrobenzene 1a, THF:H O = 95:5, analytical standard: diethylene glycol dibutyl ether; (a)
2
2
conversion of 1-iodo-4-nitrobenzene 1a; (b) formation of 4-iodoaniline 4a; circles: 10 mL of solvent, 127 mg of 1-iodo-4-nitrobenzene 1a; squares:
00 mL of solvent, 2.5 g of 1-iodo-4-nitrobenzene 1a; triangles: 1000 mL of solvent, 12.7 g of 1-iodo-4-nitrobenzene 1a.
2
moderately when Raney Co was used (1.3% in 50 min, Figure
S3).
Raney Ni, which is highly pyrophoric when freshly
26
prepared, has also been used for hydrogenations in order to
compare its performance to Raney Co. Raney Ni showed lower
catalytic performance compared to cobalt and slightly higher
dehalogenation. We observed that Raney Ni lost its catalytic
activity in hydrogenation reactions successively, when it was
stored for several months in water at ambient atmosphere. In
contrast, Raney Co showed stable catalytic activity after up to
1.5 years of storage time under the same conditions. The Raney
Co used in this study contained up to 2.5% nickel as specified
by the manufacturer; varying amounts of nickel in Raney Co
might lead to differences in catalytic performance. In our hands
Not unexpectedly, Raney Co showed significantly slower
reaction rates compared to Pt−V/C, necessitating higher metal
loadings (15 mol % vs 1 mol %) and longer reaction times.
However, it has to be taken into account, that in the case of
Pt−V/C, the metal is finely dispersed on the carbon support. In
contrast, a comparably large number of cobalt atoms are buried
deeply in the sponge-like Raney Co and thus not available for
catalysis. Heterogeneous catalysts featuring cobalt which is
immobilized on carrier materials might improve the reaction
rates.
E
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 7. Flow scheme of the reaction setup for continuous flow reactions with a solid catalyst bed in microfluidic devices.
Raney Co was less pyrophoric than freshly prepared Raney Ni.
However, when water was removed by washing of the Raney
cobalt catalyst with EtOH and drying of the catalyst was
attempted under a constant stream of air, the material
eventually ignited. In general, using a pyrophoric material will
hamper the scale-up of a reaction.
To summarize, Pt−V/C is by far more active than Raney Co,
but less selective, especially at higher concentrations, while Co
is selective over a broad range of concentrations. After these
screening experiments with reaction volumes of 10 mL, we
wished to evaluate hydrogenations with Raney Co on larger
scale using the previously optimized conditions.
Scale-Up Experiments. The scale-up experiments were
performed with low substrate concentration (0.05 M) in order
to minimize effects of increasing dehalogenation with a higher
substrate concentration (Figure 4). First, the 10 mL reaction
volume used in the screening experiments was increased to 200
mL in a 300 mL autoclave, all other parameters staying
constant (Figure 5).
With Pt−V/C as well as with Raney Co, it was observed that
after full conversion, additional reaction time was necessary
before maximal yields of 4-iodoaniline 4a were obtained, an
observation which can be attributed to the reduction of
hydroxylamine 3 and diazo compounds 7−9. Reactions with
Pt−V/C showed a strong increase of the dehalogenation rate
immediately after reaching full conversion (Figure 5a) as
opposed to reactions with Raney Co (Figure 5b) where
dehalogenation increased from 0.5% (260 min) to 0.8% (450
min). 4-Iodoaniline 4a was formed in 81% yield.
After the promising results with 200 mL reaction volume, we
focused on scale-up experiments with Raney Co (Figure 6).
Upon scaling up from 10 to 200 mL reaction volume, longer
reaction times were required for full conversion (Figure 6a) and
maximum product formation (Figure 6b). Heating was slower
for larger reaction volumes. However, this can only partially
account for the longer reaction time (at 10 mL scale, autoclaves
were heated with a preheated block, heating the reaction
solution to 110 °C was completed after 1−2 min; on 200 mL
scale heating required 40 min). For this reason it can be
assumed that the decreased performance might be due to a
hydrogen mass transfer limitation. Scaling up from 200 to 1000
mL necessitated using a multipurpose 1.8 L stainless steel
autoclave. This autoclave had been used with a variety of
rhodium-, palladium-, and platinum-based catalysts before.
After dismantling, manual scrubbing, and using clean-in-place
cleaners, hydrogenation experiments with 1-iodo-4-nitro-
benzene 1a, but without Raney Co were performed in order
to determine the influence of residual catalysts. After 6 h at 110
27
°C and 20 bar H , 10% of substrate 1a were converted to a
2
mixture of nitrobenzene, 4-iodoaniline, and aniline. Further
suppression of the background reactivity could not be achieved
by cleaning. Accordingly, the yields for 4a which had been
observed on a 10 and 200 mL scale could not be attained on a
1000 mL scale. After full conversion of 1a (Figure 6a),
comparatively stronger dehalogenation of 4-iodoaniline 4a to
aniline 6a was observed (not shown). Concomitantly, the yield
of 4-iodoaniline 1a decreased (Figure 6b).
Therefore, using highly selective Raney Co on sensitive
substrates in batch processes might necessitate the use of
designated equipment. At this point, we investigated the use of
Pt−V/C and Raney Co in continuous flow hydrogenations.
Flow Experimental Setup. Continuous reactions were
performed using a customized microreactor, based on a
modular microreaction system from Ehrfeld Mikrotechnik
28
BTS (Figure 7).
29
A Knauer K-501 HPLC pump (0.001−9.999 mL/min) was
used to pump the reaction solution through the microreactor.
Valve 1 allowed switching between a rinsing solution used for
conditioning and washing of the catalyst and the reaction
solution. A check valve behind the pump ensured that no gas
could pass from the reaction system to the HPLC pump. After
passing a pressure sensor and a heat exchanger, which allowed
to preheat the reaction solution, liquid and gaseous phase were
combined in a micro mixer. The gaseous phase was introduced
30
into the system with a mass flow controller. Valve 2 allowed
switching between nitrogen and hydrogen. In order to prevent
uncontrolled pressure build-up in case of clogging, a pressure
release valve (95 bar) was included into the system. After
passing a temperature sensor, the gas−liquid mixture was
introduced into an Ehrfeld cartridge reactor 240. For safety
reasons, the housing of this cartridge reactor was continuously
flushed with argon during the hydrogenations. The total
F
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 8. Product composition in the continuous hydrogenation of 1-iodo-4-nitrobenzene 1a with Pt−V/C; (a) first run, and (b) second run after
cartridge recycling; reaction conditions: 547 mg of Pt−V/C (11 μmol Pt), 20 bar of H , 95 °C, 0.125 M in THF, 1.25 mL/min, cartridge recycling
2
after 6 h: washing with THF (1 mL/min, 30 min), then second run of hydrogenation under otherwise identical conditions, analytical standard:
diethylene glycol dibutyl ether; dark gray: 4-iodoaniline 4a, black: aniline 6a, light gray-striped: unmonitored compounds based on detected starting
material. The absolute height of the bars reflects conversion of the starting material 1a.
volume of the reactor was 5 mL; the catalyst bed in the
cartridge had a constant volume of 2.5 mL. After leaving the
reactor, the mixture passed a second temperature sensor and an
31
Equilibar back pressure valve, which assured precise back-
pressure control. Valve 5 allowed to switch between discarding
and collecting of the reaction mixture. The Ehrfeld cartridge
reactor 240 has a pressure limit of 100 bar at 25 °C and 90 bar
at 150 °C. In this study we limited the pressure to 85 bar and
the temperature to 150 °C.
Flow Experiments with Pt−V/C. When this system was
used with Pt−V/C, the conversion of 1-iodo-4-nitrobenzene 1a
usually decreased rapidly over time. Figure 8a shows a typical
experiment, with conversion dropping from 88 to 51% over 5 h.
Flushing reactor and catalyst with THF could only partially
restore the catalytic activity (Figure 8b).
This effect was more pronounced at temperatures below 100
°C, but even at temperatures as high as 150 °C, conversion
dropped from 100 to 80% over 6 h. Decreasing the
concentration to 0.05 M or reducing the flow rate to 0.5
mL/min did not alter this effect. Pt−V/C did not meet our
expectation to show stable performance over time, which is
perceived to be an important prerequisite for continuous flow
reactions. For this reason, comparison to the reactions in batch
was not possible and the continuous hydrogenation of other
nitroaromatics was not further investigated with Pt−V/C as
catalyst. Continuous hydrogenation with Raney Co was
investigated next.
Continuous Hydrogenation with Raney Co. Under
continuous flow conditions, Raney Co showed higher long-
term stability in comparison to Pt−V/C. Compared to
hydrogenations in batch (Figure 5b) at 20 bar and 110 °C, it
was found that increasing the pressure to 85 bar and lowering
the temperature to 80 °C, increased the yield of 4-iodoaniline
Figure 9. Continuous hydrogenation of 1-iodo-4-nitrobenzene 1a with
Raney Co; reaction conditions: 1.32 g of Raney Co (16.9 mmol Co),
8
5 bar H , 1.8 mL of H /min, 80 °C, 0.05 M in THF/H O (95:5), 0.5
2
2
2
mL/min; dark gray: 4-iodoaniline 4a, black: aniline 6a, light gray,
striped: unmonitored compounds based on detected starting material.
to 9 equiv with respect to the substrate did not change the
catalytic performance significantly. After 7 h, the system was
flushed with THF and could be reused for several different
substrates before a drop in performance became apparent.
Application to Pharmaceutically Relevant Substrates.
While 1-iodo-4-nitrobenzene 1a represents a suitable model
compound for the profiling of different catalysts and the
optimization of general reaction conditions, we also inves-
tigated transformations, which are relevant examples for the
synthesis of active pharmaceutical ingredients.
4
a from 80 to 90% together with slightly higher overall
dehalogenation to aniline 6a between 1.5 and 2% at full
conversion (Figure 9). 1−2 h of equilibration were required to
reach stable catalytic performance. Three equiv of hydrogen
were used for these experiments. Increasing the hydrogen flow
The reduction of N-(4-chlorophenyl)-2-nitroaniline 1b, an
intermediate in the synthesis of clofazimine, which is used for
3
2
the treatment of leprosy,
has been reported using
33a 33b
stoichiometric amounts of zinc,
iron powder,
or
G
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
3
3c
SnCl2
(Scheme 1). Interestingly, unlike 1-iodo-4-nitro-
Scheme 3. Reduction of 4-Chloro-3-(pyridine-2-
yl)nitrobenzene 1d in the Synthesis of Vismodegib and
Reduction to Dehalogenated Aniline 6d
3
5
benzene 1a, nitro compound 1b features a chlorine atom
Scheme 1. Reduction of N-(4-Chlorophenyl)-2-nitroaniline
33b
1
b in the Synthesis of Clofazimine, Reduction of
Iodinated Analogue 1c, and Reduction to Dehalogenated
Aniline 6b
at slightly higher temperature (80 → 100 °C) for hydro-
genation of nitro compounds 1b−1c on days 2 and 3.
Clofazimine precursor 1b was hydrogenated on day 2,
yielding the corresponding chloroaniline 4b almost quantita-
tively (Figure 10a). Dehalogenation occurred below the
detection limit of the HPLC analysis (ca. 0.5%). On day 3,
iodinated analog 1c showed 2.5−3.5% dehalogenation at
conversions of 98−100% (Figure 10b).
which is attached to an electron-rich aniline. Therefore, we also
investigated iodinated model compound 1c to benchmark our
reaction setup, assuming that this substrate would be even more
prone to dehalogenation to aniline 6b than electron-poor
nitroaromatic 1a.
Subjecting the same catalyst cartridge to vismodegib
precursor 1d on day 4, almost quantitative conversion and
The synthesis of N-(4-chlorophenyl)-2-nitroaniline 1b from
-fluoro-2-nitrobenzene 10 and 4-chloro-aniline 11 (Scheme 2)
1
1
−2% dehalogenation were observed by HPLC. However,
upon HPLC monitoring of this hydrogenation, an additional
side product was observed, eluting together with chloroaniline
Scheme 2. Synthesis of Halogenated Nitro Compounds 1b
and 1c; LiHMDS: Lithium Bis(Trimethylsilyl)amide
4d. The two peaks could not be separated completely by
HPLC, but after evaporation of the solvent mixture this
3
6
compound was converted into the azoxy compound
approximately 10% in the crude product), suggesting that
(
the unidentified side product could have been the hydroxyl-
amine intermediate 4-chloro-N-hydroxy-3-(pyridine-2-yl)-
aniline.
The experiment with vismodegib precursor 1d of day 4 was
repeated with fresh catalyst (Figure 10c). No formation of
hydroxylamine or azoxy compound was observed in this
experiment, which suggests beginning deactivation of the
catalyst in the previous experiment. Higher dehalogenation
tendency compared to the aromatic chloride 1b (<0.5% vs 2%)
might be explained by the adjacent coordination site provided
by the pyridine nitrogen.
After equilibration times of 1−2 h, the reaction mixtures of
4b−4d were collected for the purpose of product isolation
(Table 1). Entry 1 shows conversion and product composition
of crude model compound 4a for comparison. After collecting
for 5−6 h (total hydrogenation time: 7 h) acid−base extraction
was used in order to separate the crude products 4b − 4d from
the internal analytical standards. Dispersion in acidic aqueous
solution as aqueous ammonium salt was followed by washing
with organic solvents. For clofazimine related anilines 4b and
4c, this procedure also allowed to remove traces of unreacted
starting materials (Table 1, entries 2−5). Aniline 4b was
isolated by this procedure with a yield of 98% (entry 3).
Iodinated analog 4c was isolated with a yield of 90% after
washing, but contained 2% of the dehalogenated derivative 6c
(entry 5). This impurity could be further depleted by
crystallization to furnish iodinated aniline in an overall yield
3
4
has been described using KF at 180 °C for 48 h. However,
strong decomposition was observed when these conditions
were used for the synthesis of iodoaniline 1c. Using the
stronger base LiHMDS at −35 °C, complete and clean
conversion was achieved after 45 min. After inverse quench
into 10% citric acid, washing of the precipitate and
recrystallization, chloroaniline 1b and iodoaniline 1c were
obtained in 88 and 78% yield, respectively (Scheme 2). This
procedure was easily scalable to produce 40−50 g batches of
diphenylamines 1b and 1c.
The selective transformation of chlorinated nitro compound
d to aniline 4d en-route to the Hedgehog signaling pathway
inhibitor vismodegib (Scheme 3) has been described to be
challenging and is only reported to be successful by using
overstoichiometric amounts of iron or SnCl2. Nitroaromatic
d also features a pyridyl substituent, which is potentially prone
to reduction. Hence, we felt it would be worthwhile to
investigate the catalytic hydrogenation of 1d. This substrate was
synthesized according to previously described procedures.
Figure 9, Figure 10a, and Figure 10b show a three-day
hydrogenation experiment using the same Raney Co cartridge
for nitro compounds 1a−1c. After hydrogenation of 1-iodo-4-
nitrobenzene 1a on day 1 (Figure 9), the cartridge was reused
1
35
1
35
H
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 10. Continuous hydrogenation of other substrates with Raney Co; (a) nitro compound 1b (day 2); conditions and cartridge from day 1
(
Figure 9) but at 100 °C; dark gray: chloroaniline 4b, light gray, striped: unmonitored compounds based on detected starting material; (b) nitro
compound 1c (day 3); conditions and cartridge from day 2; dark gray: iodoaniline 4c, black: aniline 6c; (c) nitro compound 1d; reaction conditions
as for day 3 but in THF/H O (9:1), fresh catalyst used; dark gray: chloroaniline 4d, black: aniline 6d, light gray, striped: unmonitored compounds
2
based on detected starting material. The absolute height of the bars reflects conversion of the starting material.
of 78% (entry 6). In case of the vismodegib intermediate, the
previously described washing procedure delivered the chlori-
nated aniline 4d together with 2% of dehalogenated aniline 6d
significantly affect the selectivity. Upon scale-up of batch
hydrogenations, it was found that undesired background
reactions can play a role when using multipurpose equipment.
Moving to continuous flow conditions using Pt−V/C led to
a fall in catalyst performance under various conditions, which is
in agreement with previous work. This prevented use of Pt−
V/C for long-term continuous hydrogenation at high
conversion. On the other hand, Raney Co showed stable
catalytic performances in continuous flow for up to 20 h
reaction time. Chlorinated intermediates of the active
pharmaceutical ingredients clofazimine and vismodegib as
well as iodinated model compounds were hydrogenated with
Raney Co in continuous flow and less than 2% dehalogenation
at nearly quantitative conversions. Depending on the substrate,
aqueous extraction or crystallization afforded the pure anilines
in overall yields of 78−98%.
(
entries 7−8). Again, removal of the dehalogenated aniline 6d
was possible by crystallization (entry 9), giving chlorinated
17f
aniline 4d with an overall yield of 88%.
CONCLUSION
■
The catalysts Pt−V/C and Raney Co were found to be high
performance catalysts in the selective hydrogenation of
halogenated nitroaromatics. Dialkylethers like THF, 2-methyl-
THF, or MeOC H are suitable solvents for this trans-
5
9
formation, because they show both, high substrate solubilities
and high hydrogenation rates. Alternatively, anisole and toluene
are also suitable as solvents but showed lower substrate
solubilities and hydrogenation rates in our investigation. In the
hydrogenation of 1-iodo-4-nitrobenzene 1a using Pt−V/C and
THF in batch, selective reaction to 4-iodo-aniline 4a was highly
dependent on the substrate concentration. In contrast, using
Raney Co with different substrate concentrations did not
It has to be kept in mind that cobalt gives space−time yields
which are almost 1 order of magnitude lower than space−time
yields for Pt−V/C. However, at the same time the price per
mol cobalt is more than 1 order of magnitude lower than the
I
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 1. Data on Hydrogenation of 4-Iodoaniline 4a and Isolation of Halogenated Anilines 4b−4d by Extraction and
Crystallization
a
b
Product compositions were determined by HPLC with internal analytical standard. Product compositions were determined by comparing peak
areas of HPLC separations at 230 nm.
price per mol platinum. Nevertheless, considering scale-up and
process intensification, optimization toward higher substrate
concentration and flow rate will be crucial for further
development.
EXPERIMENTAL SECTION
■
General Procedure for the Synthesis of Substrates 1b
and 1c. The corresponding haloaniline 11 or 4a (178.9 mmol)
was dissolved in THF (90 mL) and cooled to −50 °C. After
addition of 1-fluoro-2-nitrobenzene 10 (25 g, 177.2 mmol) and
rinsing with THF (35 mL), LiHMDS (1 M solution in THF,
54 mL, 354 mmol) was added dropwise over 30 min. The
purple reaction mixture was stirred at −40 to −30 °C until the
-fluoro-2-nitrobenzene was completely converted (30−45
min, monitored by GCMS). The reaction mixture was added
over approximately 20 min to 10% aqueous citric acid (2000
mL) at 0 °C. The dark red precipitate was washed with water
(1000 mL) and dried in vacuo. The crude product was
recrystallized as indicated to give bright red needles of the
corresponding N-(4-halophenyl)-2-nitroaniline.
OUTLOOK
■
A rational choice between batch and flow might be alleviated by
the following approach. In a first step, a catalyst system suitable
for the given transformation should be identified. A quick
evaluation of a suitable parameter window in batch will allow to
judge whether development of a flow process might be useful.
This can be the case for very fast and exothermic reactions or if
the reaction has to be stopped within a short time frame, e.g., to
isolate an intermediate. It might also be useful if very high
pressures or temperatures are necessary or if mass transport
limitations play an important role in the process. Telescoping
into a second reaction or continuous workup can streamline a
process, but it can also be inevitable if the intermediate is
3
1
34
N-(4-Chlorophenyl)-2-nitroaniline 1b. Recrystallized from
EtOH (2200 mL) and water (900 mL). Yield: 39.1 g (88%) N-
4-chlorophenyl)-2-nitroaniline; purity: 98%, determined by
14
unstable or highly toxic.
(
At the moment, we are also investigating catalysts with
higher hydrogenation rates like Pt and Pd. For these catalysts,
advantages like precise control over residence time and
increased availability of hydrogen on the catalyst surface
might favor selectivity in continuous processing. However, a
sustainable process should involve recycle/reuse of the precious
metals from used catalysts.
For cobalt-based systems, tolerance of other functional
groups and the reduction of solvent amounts and catalyst
loadings will be vital in order to exploit the surpassing
chemoselectivity for technical applications.
quantitative NMR using dimethyl sulfone (Sigma-Aldrich
standard for quantitative NMR, catalog no. 41867) as internal
1
analytical standard; H NMR (300 MHz, DMSO-d ): δ = 9.30
6
(
s, 1H), 8.11 (dd, J = 8.5, 1.6 Hz, 1H), 7.52 (ddd, J = 8.6, 7.0,
1
.6 Hz, 1H), 7.47−7.39 (m, 2H), 7.37−7.29 (m, 2H), 7.21 (dd,
13
J = 8.6, 1.2 Hz, 1H), 6.92 (ddd, J = 8.4, 7.0, 1.2 Hz, 1H);
C
NMR (75 MHz, DMSO-d ): δ = 141.2, 138.7, 135.9, 134.3,
6
1
C
29.3, 128.2, 126.2, 124.8, 118.7, 117.3; HRMS (ESI): for
+
H
ClN
NaO
([M + Na] ) calc.: 271.02448, found:
12
9
2
2
271.0245.
J
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
N-(4-Iodophenyl)-2-nitroaniline 1c. Recrystallized twice
from EtOAc (600 mL), EtOH (2275 mL), and water (420
mL). Yield: 48.3 g (79%) N-(4-iodophenyl)-2-nitroaniline.
Purity: 98%, determined by quantitative NMR using dimethyl
sulfone (Sigma-Aldrich standard for quantitative NMR, catalog
and H (3 equiv corresponding to the substrate) at 100 °C and
2
85 bar for 10 min before switching to the substrate solution.
Samples were collected every 30 min, diluted with MeOH (20
volumes), and analyzed by HPLC. The reaction solution was
collected in 30 min fractions. The fractions collected between
120 and 420 min were combined and concentrated in vacuo.
The crude product was dissolved in a mixture of EtOH (60
mL) and THF (52 mL); then aq. HCl (0.45 M, 145 mL) was
added. The mixture was extracted with petroleum ether (2 × 40
mL). The aqueous phase was filtered, and the filter was washed
with a mixture of EtOH (20 mL), THF (17 mL), and aq. HCl
(48 mL). After adding pH 10 carbonate buffer (105 mL) to the
combined aqueous phases, the mixture was extracted with
EtOAc (3 × 150 mL). The organic phases were combined,
1
no. 41867) as internal analytical standard; H NMR (300 MHz,
DMSO-d ): δ = 9.26 (s, 1H), 8.10 (dd, J = 8.5, 1.6 Hz, 1H),
6
7
.76−7.65 (m, 2H), 7.52 (ddd, J = 8.6, 7.0, 1.6 Hz, 1H), 7.25
(
dd, J = 8.6, 1.2 Hz, 1H), 7.19−7.07 (m, 2H), 6.93 (ddd, J =
1
3
8
1
8
3
.4, 7.0, 1.3 Hz, 1H); C NMR (75 MHz, DMSO-d ): δ =
6
40.8, 139.7, 138.1, 135.9, 134.6, 126.2, 125.0, 118.9, 117.5,
−
8.1; HRMS (ESI): for C H IN O ([M − H] ) calc.:
1
2
8
2
2
38.96359, found: 338.9641.
Three-Day Hydrogenation Experiment in the Ehrfeld
Microreaction System with Halogenated Nitro Com-
dried (MgSO ), and concentrated in vacuo to give N-(4-
4
pounds 1a−c. Reactor Cartridge Preparation. Raney Co
iodophenyl)benzene-1,2-diamine 4c as a purple solid (2.100 g,
90%), which contained around 2.5% N-phenylbenzene-1,2-
diamine 6b (det. by comparing HPLC peak areas at 230 nm).
The crude product was recrystallized from a mixture of EtOAc
(34 mL) and n-heptane (165 mL, 80 °C to −30 °C). The
mother liquor was concentrated in vacuo and recrystallized
from a mixture of EtOAc (4 mL) and n-heptane (10 mL, 80 °C
to −30 °C). The beige-colored crystals were washed with cold
n-pentane (0 °C, 3 × 5 mL) and dried in vacuo to give N-(4-
iodophenyl)benzene-1,2-diamine 4c in a combined yield of
78% (containing approximately 0.4% N-phenylbenzene-1,2-
diamine 6b and 0.3% of the diiodinated azoxycompound N-(4-
iodophenyl)-2-[{2-[(4-iodophenyl)amino]phenyl}-NNO-
(1.322 g, ca. 16.8 mmol) was mixed with glass beads (20−200
μm) to give a total volume of 2.5 mL. The cartridge (Ehrfeld
cartridge reactor 240) was consecutively filled with layers of
glass wool (150 mg), Celite (150 mg), glass wool (150 mg), the
mixture of catalyst with glass beads, glass wool (150 mg), Celite
(
150 mg), and glass wool (150 mg).
Day 1: Hydrogenation of 1-Iodo-4-nitrobenzene 1a. 1-
Iodo-4-nitrobenzene (6.358 g, 12.5 mmol) and diethylene
glycol dibutyl ether (2.781 g, 6.3 mmol, analytical standard)
were dissolved in 500 mL of stabilized THF and water (95:5).
The Ehrfeld system was equilibrated with THF (0.5 mL/min)
and H (3 equiv corresponding to the substrate) at 80 °C and
2
8
5 bar for 30 min before switching to the substrate solution.
azoxy]aniline, det. by HPLC comparing peak areas at 230 nm).
1
Samples were collected hourly, diluted with stabilized THF
1:5) and analyzed by GC. After 6 h the system was rinsed with
H NMR (400 MHz, DMSO-d ): δ = 7.42−7.35 (m, 2H), 7.33
6
(
(s, 1H), 6.97 (dd, J = 7.8, 1.5 Hz, 1H), 6.88 (ddd, J = 7.9, 7.3,
1.5 Hz, 1H), 6.74 (dd, J = 8.0, 1.5 Hz, 1H), 6.58−6.47 (m, 3H),
pure THF for 45 min to be reused on day 2.
4.75 (s, 2H); 13C NMR (75 MHz, DMSO-d ): δ = 146.1,
Day 2: Hydrogenation of N-(4-Chlorophenyl)-2-nitroani-
line 1b. N-(4-chlorophenyl)-2-nitroaniline 1b (3.177 g, 12.5
mmol) and 2-methoxynaphtalene (333 mg, 2.1 mmol, analytical
standard) were dissolved in 250 mL of stabilized THF and
water (95:5). The Ehrfeld system was primed with THF (0.5
6
143.0, 137.1, 126.4, 125.0, 124.3, 116.6, 116.4, 115.3, 78.1;
+
HRMS (ESI): for C H IN ([M + H] ) calc.: 311.00397,
12
12
2
found: 311.0041.
Hydrogenation of 2-(2-Chloro-5-nitrophenyl)pyridine
1d with Fresh Catalyst. The reactor cartridge was filled with
fresh Raney Co as described before. 2-(2-Chloro-5-
nitrophenyl)pyridine 1d (3.087 g, 12.5 mmol) and 2-
methoxynaphtalene (332 mg, 2.1 mmol) were dissolved in
250 mL of stabilized THF and water (9:1). The Ehrfeld system
was primed with THF (0.5 mL/min) and H2 (3 equiv
corresponding to the substrate) at 100 °C and 85 bar for 45
min before switching to the substrate solution. Samples were
collected every 30 min, diluted with MeOH (20 volumes) and
analyzed by HPLC. The reaction solution was collected in 30
min fractions. The fractions collected between 90 and 420 min
were combined and concentrated in vacuo. The crude product
was suspended in EtOH (35 mL), and aq. HCl (0.45 M, 120
mL) was added. The mixture was extracted with petroleum
ether (3 × 25 mL) and EtOAc (1 × 15 mL). pH 10 carbonate
buffer (65 mL) was added, and the aqueous phase was
extracted with EtOAc (4 × 50 mL). The organic phases were
mL/min) and H (3 equiv corresponding to the substrate) at
2
100 °C and 85 bar for 10 min before switching to the substrate
solution. Samples were collected every 30 min, diluted with
MeOH (20 volumes), and analyzed by HPLC. The reaction
solution was collected in 30 min fractions. After 7 h the system
was rinsed with THF for 45 min and reused on day 3. The
fractions collected between 60 and 420 min were combined and
concentrated in vacuo. The crude product was dissolved in
EtOH (45 mL) and aq. HCl (0.45 M, 150 mL) was added. The
mixture was extracted with petroleum ether (3 × 30 mL). pH
10 carbonate buffer (79 mL, cf. Supporting Information) was
added to the aqueous solution, and the precipitate was filtered
and washed with water until the filtrate reached pH 7. The
precipitate was dried in vacuo to give N-(4-chlorophenyl)-
33
benzene-1,2-diamine 4b (1.936 g, 98%) as beige-colored solid
1
(
>99% area by HPLC, 230 nm). H NMR (400 MHz, CDCl ):
3
δ = 7.19−7.12 (m, 2H), 7.12−7.00 (m, 2H), 6.84−6.73 (m,
1
3
2
H), 6.69−6.60 (m, 2H), 5.19 (bs, 1H), 3.39 (bs, 2H);
C
combined, dried (MgSO ), and concentrated in vacuo to give
4
NMR (100 MHz, CDCl ): δ = 144.2, 142.1, 129.3, 128.1,
4-chloro-3-(pyridin-2-yl)aniline 4d as a yellow solid (1.622 g,
96%, containing 2% 3-(pyridin-2-yl)aniline 6d, det. by HPLC
comparing peak areas at 230 nm). The crude product was
recrystallized from toluene (27 mL, 80 °C to −30 °C) and n-
pentane (8 mL). The yellow-colored crystals were washed with
n-pentane (3 × 1 mL) and dried in vacuo to give 4-chloro-3-
3
1
26.3, 125.2, 124.0, 119.4, 116.38, 116.37; HRMS (ESI): for
+
C H ClN ([M + H] ) calc.: 219.06835, found: 219.0686.
1
2
12
2
Day 3: Hydrogenation of N-(4-Iodophenyl)-2-nitroaniline
c. N-(4-Iodophenyl)-2-nitroaniline 1c (4.386 g, 12.5 mmol)
1
and 4-methylanisole (1.541 g, 2.0 mmol, analytical standard)
were dissolved in 250 mL of stabilized THF and water (95:5).
The Ehrfeld system was equilibrated with THF (0.5 mL/min)
35
(pyridin-2-yl)aniline 4d (1.485 g, 88% yield, >99% purity, det.
1
by HPLC, 230 nm). H NMR (400 MHz, benzene-d ): δ =
6
K
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
8
7
1
.57 (ddd, J = 4.8, 1.9, 1.0 Hz, 1H), 7.53−7.50 (m, 1H), 7.17−
.15 (m, 1H), 7.10 (d, J = 8.5 Hz, 1H), 6.88 (d, J = 2.8 Hz,
H), 6.70 (ddd, J = 7.5, 4.8, 1.2 Hz, 1H), 6.09 (dd, J = 8.5, 2.9
(b) Jagadeesh, R. V.; Surkus, A.-E.; Junge, H.; Pohl, M.-M.; Radnik, J.;
Rabeah, J.; Huan, H.; Schu
Washington, DC, U. S.) 2013, 342, 1073−1076.
(8) (a) Braden, R.; Knupfer, H.; Hartung, S. Process for the
̈ ̈
nemann, V.; Bruckner, A.; Beller, M. Science
(
1
3
Hz, 1H), 2.88 (bs, 2H); C NMR (100 MHz, benzene-d ): δ =
6
preparation of unsaturated amino compounds. U.S. Patent 4,002,673,
Jan 11, 1977. (b) Twigg, M. V. Catalytic reduction process for the
production of aromatic amino compounds. Eur. Pat. Appl. 0211545,
Feb 25, 1987, EP 211545. (c) Chen, J. P.; Lee, K. M.; Sorensen, C. M.;
Klabunde, K. J.; Hadjipanayis, G. C. J. Appl. Phys. (Melville, NY, U. S.)
1994, 75, 5876. (d) Raja, R.; Golovko, V. B.; Thomas, J. M.;
Berenguer-Murcia, A.; Zhou, W.; Xie, S.; Johnson, B. F. G. Chem.
Commun. (Cambridge, U. K.) 2005, 2026−2028. (e) Westerhaus, F. A.;
1
1
2
57.8, 149.8, 146.1, 140.4, 135.3, 130.9, 125.0, 122.1, 121.0,
+
18.5, 116.2; HRMS (ESI): for C H ClN ([M + H] ) calc.:
1
1
10
2
05.05270, found: 205.0528.
ASSOCIATED CONTENT
■
* Supporting Information
S
Additional experiments and data on catalysts and analytical
methods. The Supporting Information is available free of
charge on the ACS Publications website at DOI: 10.1021/
acs.oprd.5b00170.
Jagadeesh, R. V.; Wienho
̈
fer, G.; Pohl, M.-M.; Radnik, J.; Surkus, A.-E.;
ckner, A.; Beller, M.
Rabeah, J.; Junge, K.; Junge, H.; Nielsen, M.; Bru
̈
Nat. Chem. 2013, 5, 537−543. (f) Jagadeesh, R. V.; Stemmler, T.;
Surkus, A.-E.; Bauer, M.; Pohl, M.-M.; Radnik, J.; Junge, K.; Junge, H.;
AUTHOR INFORMATION
Corresponding Authors
E-mail: patrick.loos@catalysis.de.
E-mail: jorma.hassfeld@bayer.com.
Author Contributions
This work was done as part of the European Union CHEM21
project.
Notes
̈
Bruckner, A.; Beller, M. Nat. Protoc. 2015, 10, 916−926. (g) Stemmler,
■
T.; Westerhaus, F. A.; Surkus, A.-E.; Pohl, M.-M.; Junge, K.; Beller, M.
Green Chem. 2014, 16, 4535−4540.
*
(9) Harrad, M. A.; Boualy, B.; El Firdoussi, L.; Mehdi, A.; Santi, C.;
*
Giovagnoli, S.; Nocchetti, M.; Ali, M. A. Catal. Commun. 2013, 32,
92−100.
(10) Hunt, A. J. in Element Recovery and Sustainability, RSC Green
Chem. Ser. No. 22; The Royal Society of Chemistry: Cambridge,
2
(
013; pp 1−24.
11) (a) Takasaki, M.; Motoyama, Y.; Higashi, K.; Yoon, S.-H.;
Mochida, I.; Nagashima, H. Org. Lett. 2008, 10, 1601−1604.
b) Motoyama, Y.; Lee, Y.; Tsuji, K.; Yoon, S.-H.; Mochida, I.;
The authors declare no competing financial interest.
(
ACKNOWLEDGMENTS
■
Nagashima, H. ChemCatChem 2011, 3, 1578−1581. (c) Ma, H.; Sun,
K.; Li, Y.; Xu, X. Catal. Commun. 2009, 10, 1363−1366.
(d) Evangelisti, C.; Aronica, L. A.; Botavina, M.; Martra, G.;
Battocchio, C.; Polzonetti, G. J. Mol. Catal. A: Chem. 2013, 366,
288−293. (e) Xu, K.; Zhang, Y.; Chen, X.; Huang, L.; Zhang, R.;
Huang, J. Adv. Synth. Catal. 2011, 353, 1260−1264. (f) Witte, P. T.;
Berben, P. H.; Boland, S.; Boymans, E. H.; Vogt, D.; Geus, J. W.;
Donkervoort, J. G. Top. Catal. 2012, 55, 505−511.
The research for this work has received funding from the
Innovative Medicines Initiative (IMI) joint undertaking project
CHEM21 under grant agreement no. 115360, resources of
which are composed of financial contribution from the
European Union’s Seventh Framework Programme (FP7/
007-2013) and EFPIA companies in kind contribution. We
wish to thank Manuela Pritzkow (LIKAT) for excellent
technical assistance.
2
(12) Reviews: (a) Newman, S. G.; Jensen, K. F. Green Chem. 2013,
1
5, 1456. (b) Frost, C. G.; Mutton, L. Green Chem. 2010, 12, 1687−
1
703.
REFERENCES
■
(
13) (a) Dudas, J.; Hanika, J. Chem. Eng. Res. Des. 2009, 87, 83−90.
(
1) Reviews: (a) Downing, R. S.; Kunkeler, P. J.; van Bekkum, H.
Catal. Today 1997, 37, 121−136. (b) Blaser, H.-U.; Malan, C.; Pugin,
B.; Spindler, F.; Steiner, H.; Studer, M. Adv. Synth. Catal. 2003, 345,
(
b) Johnson, M. D.; May, S. A.; Calvin, J. R.; Remacle, J.; Stout, J. R.;
Diseroad, W. D.; Zaborenko, N.; Haeberle, B. D.; Sun, W.-M.; Miller,
M. T.; Brennan, J. Org. Process Res. Dev. 2012, 16, 1017−1038.
1
03−151. (c) Chen, B.; Dingerdissen, U.; Krauter, J. G. E.; Lansink
Rotgerink, H. G. J.; Mobus, K.; Ostgard, D. J.; Panster, P.; Riermeier,
T. H.; Seebald, S.; Tacke, T.; Trauthwein, H. Appl. Catal., A 2005, 280,
(14) Cantillo, D.; Damm, M.; Dallinger, D.; Bauser, M.; Berger, M.;
̈
Kappe, C. O. Org. Process Res. Dev. 2014, 18, 1360−1366.
15) Acres, G. J. K.; Cooper, B. J. J. Appl. Chem. Biotechnol. 1972, 22,
69−785.
16) Review: Irfan, M.; Glasnov, T. N.; Kappe, C. O. ChemSusChem
011, 4, 300−316.
17) (a) Cantillo, D.; Baghbanzadeh, M.; Kappe, C. O. Angew. Chem.
(
7
(
2
1
1
(
7−46. (d) Blaser, H.-U.; Steiner, H.; Studer, M. ChemCatChem 2009,
, 210−221. (e) Pietrowski, M. Curr. Org. Synth. 2012, 9, 470−487.
2) Booth, G. Nitro Compounds, Aromatic. Ullmann’s Encyclopedia of
Industrial Chemistry; Wiley-VCH: Weinheim, 2005.
3) Hoogenraad, M.; van der Linden, J. B.; Smith, A. A. Org. Process
Res. Dev. 2004, 8, 469−476.
4) (a) Wienhofer, G.; Baseda-Kru
F. A.; Baumann, W.; Jackstell, R.; Junge, K.; Beller, M. Chem. Commun.
Cambridge, U. K.) 2013, 49, 9089−9091. (b) Corma, A.; Gonzalez-
Arellano, C.; Iglesias, M.; Sanchez, F. Appl. Catal., A 2009, 356, 99−
02. (c) Chepaikin, E. G.; Khidekel, M. L.; Ivanova, V. V.; Zakhariev,
(
(
2
1
4
012, 124, 10337−10340; Angew. Chem., Int. Ed. 2012, 51, 10190−
0193. (b) Chen, M.; Buchwald, S. L. Angew. Chem. 2013, 125, 4341−
344; Angew. Chem., Int. Ed. 2013, 52, 4247−4250. (c) Kreutzer, M.
(
̈
̈
ger, M.; Ziebart, C.; Westerhaus,
T.; Kapteijn, F.; Moulijn, J. A. Catal. Today 2005, 105, 421−428.
(
́
(
(
d) Tadepalli, S.; Lawal, A. Int. J. Chem. React. Eng. 2008, 6, A112.
e) Hatziantoniou, V.; Andersson, B.; Scho n, N.-H. Ind. Eng. Chem.
́
̈ ̈
o
1
Process Des. Dev. 1986, 25, 964−970. (f) Yeong, K. K.; Gavriilidis, A.;
Zapf, R.; Hessel, V. Catal. Today 2003, 81, 641−651. (g) Yeong, K. K.;
Gavriilidis, A.; Zapf, R.; Hessel, V. Chem. Eng. Sci. 2004, 59, 3491−
A. I.; Shopov, D. M. J. Mol. Catal. 1980, 10, 115−119. (d) Xu, S.; Xi,
X.; Shi, J.; Cao, S. J. Mol. Catal. A: Chem. 2000, 160, 287−292.
(
e) Knifton, J. F. J. Org. Chem. 1976, 41, 1200−1206. (f) Toti, A.;
3
494. (h) Kataoka, S.; Takeuchi, Y.; Harada, A.; Takagi, T.; Takenaka,
Y.; Fukaya, N.; Yasuda, H.; Ohmori, T.; Endo, A. Appl. Catal., A 2012,
427−428, 119−124. (i) Jones, R.; Godorhazy, L.; Szalay, D.;
, G.; Urge, L.; Darvas, F. QSAR Comb. Sci.
Frediani, P.; Salvini, A.; Rosi, L.; Giolli, C. J. Organomet. Chem. 2005,
90, 3641−3651. (g) Deshmukh, A. A.; Prashar, A. K.; Kinage, A. K.;
Kumar, R.; Meijboom, R. Ind. Eng. Chem. Res. 2010, 49, 12180−12184.
5) Review: Cardenas-Lizana, F.; Keane, M. A. J. Mater. Sci. 2013, 48,
43−564.
6) Makosch, M.; Sa,
A.; Hungerbuhler, K. ChemCatChem 2012, 4, 59−63.
7) (a) Deshpande, R. M.; Mahajan, A. N.; Diwakar, M. M.; Ozarde,
P. S.; Chaudhari, R. V. J. Org. Chem. 2004, 69, 4835−4838R.
6
̈
̈
́
̈
(
5
(
́
Gerencser
́ ́
, J.; Dorman
2005, 24, 722−727.
́
J.; Kartusch, C.; Richner, G.; van Bokhoven, J.
(18) Kosak, J. R. Ann. N. Y. Acad. Sci. 1970, 172, 175−185.
(19) Jimenez-Gonzalez, C.; Ponder, C. S.; Broxterman, Q. B.;
Manley, J. B. Org. Process Res. Dev. 2011, 15, 912−917.
̈
(
(20) Prat, D.; Hayler, J.; Wells, A. Green Chem. 2014, 16, 4546−4551.
L
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(
21) Example: Makosch, M.; Lin, W.-I.; Bumbal
J. W.; Hungerbuhler, K.; van Bokhoven, J. A. ACS Catal. 2012, 2,
079−2081.
22) Baumeister, P.; Blaser, H.-U.; Studer, M. Catal. Lett. 1997, 49,
19−222.
23) Studer, M.; Neto, S.; Blaser, H.-U. Top. Catal. 2000, 13, 205−
12.
24) Kasparian, A. J.; Savarin, C.; Allgeier, A. M.; Walker, S. D. J. Org.
Chem. 2011, 76, 9841−9844.
25) Wu, G.; Huang, M.; Richards, M.; Poirier, M.; Wen, X.; Draper,
R. W. Synthesis 2003, 11, 1657−1660.
26) Billica, H. R.; Adkins, H. Org. Synth. 1955, 7, 176 and references
therein.
́ ́
ek, V.; Sa, J.; Medlin,
̈
2
(
2
(
2
(
(
(
(
27) Clean-in-place cleaners have been obtained from Ecolab
(http://www.ecolab.com).
(
(
(
(
(
(
28) http://www.ehrfeld.com.
29) http://www.knauer.net.
30) http://www.bronkhorst-maettig.de.
31) http://www.equilibar.com.
32) Hooper, M. Chem. Soc. Rev. 1987, 16, 437−465.
̈
33) (a) Kirsch, P.; Schonleben-Janas, A.; Schirmer, R. H. Liebigs
Ann. 1995, 1995, 1275−1281. (b) Albrecht, B.; Gehling, V. S.; Hewitt,
M. C.; Taylor, A. M.; Harmange, J.-C. Bromodomain inhibitors and
uses thereof. WO2012151512, Nov 8, 2012. (c) Wilberg, A. Ber. Dtsch.
Chem. Ges. 1902, 35, 954−959.
(
34) Kulagowski, J. J.; Rees, C. W. Synthesis 1980, 1980, 215.
(
35) (a) Robarge, K. D.; Brunton, S. A.; Castanedo, G. M.; Cui, Y.;
Dina, M. S.; Goldsmith, R.; Gould, S. E.; Guichert, O.; Gunzner, J. L.;
Halladay, J.; Jia, W.; Khojasteh, C.; Koehler, M. F. T.; Kotkow, K.; La,
H.; LaLonde, R. L.; Lau, K.; Lee, L.; Marshall, D.; Marsters, J. C., Jr.;
Murray, L. J.; Qian, C.; Rubin, L. L.; Salphati, L.; Stanley, M. S.;
Stibbard, J. H. A.; Sutherlin, D. P.; Ubhayaker, S.; Wang, S.; Wong, S.;
Xie, M. Bioorg. Med. Chem. Lett. 2009, 19, 5576−5581. (b) Gunzner, J.
L.; Sutherlin, D. P.; Stanley, M. S.; Bao, L.; Castanedo, G.; Lalonde, R.;
Wang, S.; Reynolds, M. E.; Savage, S. J.; Malesky, K.; Dina, M. S.;
Koehler, M. F. T. Pyridyl inhibitors of hedgehog signaling. U.S. Pat.
Appl. 20120094980 A1, Apr 19, 2012.
(
36) The dichlorinated azoxy compound 2-(2-chloro-5-{[4-chloro-3-
(
pyridine-2-yl)phenyl]-NNO-azoxy}phenyl)pyridine was identified
using HRMS; it might be formed by condensation of nitroso and
hydroxylamine intermediate (compare 2 + 3 → 7 + H O).
2
M
DOI: 10.1021/acs.oprd.5b00170
Org. Process Res. Dev. XXXX, XXX, XXX−XXX