Fine-Bubble-Based Strategy for the Palladium-Catalyzed Hydrogenation of Nitro Groups: Measurement of Ultrafine Bubbles in Organic Solvents

Article abstract of DOI:10.1055/s-0036-1588869

Fine bubbles of hydrogen were employed as a new reaction medium for the autoclave-free gas-liquid-solid multiphase hydrogenation of nitro groups on a multigram scale. Furthermore, ultrafine bubbles were examined by nanoparticle-tracking analysis in organic solvents.

Full text of DOI:10.1055/s-0036-1588869

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2017, 49, A–E
letter
A
en
N. Mase et al.
Letter
Synlett
Fine-Bubble-Based Strategy for the Palladium-Catalyzed Hydroge-
nation of Nitro Groups: Measurement of Ultrafine Bubbles in Or-
ganic Solvents
Nobuyuki Mase*a,b,c
Yuki Nishinaa
Shogo Isomuraa
Kohei Satoa
Tetsuo Narumia,b,c
Naoharu Watanabeb
a Applied Chemistry and Biochemical Engineering Course, Department
of Engineering, Graduate School of Integrated Science and Technology,
Shizuoka University, 3-5-1 Johoku, Hamamatsu, Shizuoka 432-8561,
Japan
b Graduate School of Science and Technology, Shizuoka University, 432-
8561, Japan
c Green Energy Research Division, Research Institute of Green Science
and Technology, Shizuoka University, 432-8561, Japan
mase.nobuyuki@shizuoka.ac.jp
Received: 21.04.2017
Accepted after revision: 11.05.2017
Published online: 06.07.2017
Aromatic amines are important intermediates in the in-
dustrial production of materials such as urethane mono-
mers, pharmaceuticals, herbicides, dyes, and rubber-pro-
cessing chemicals. Consequently, many studies on the gas–
liquid–solid multiphase heterogeneous catalytic hydroge-
nation of nitroarenes to anilines have been reported.9 How-
ever, existing problems need to be addressed to improve the
reactivity and chemical yield. The process requires a pres-
sure-resistant reaction vessel operating under high-pres-
sure conditions and/or vigorous bubbling and mechanical
stirring, and finely powdered Pd/C is required to obtain a
high yield.10 However, the removal of the Pd/C by filtration
often necessitates the adoption of a batch system and en-
tails long operating times; careful monitoring to prevent
clogging of the filter is required in continuous-flow sys-
tems. These problems can be solved by increasing the con-
centration of the dissolved gas at atmospheric pressure,
without vigorous bubbling and stirring, by using a nonpow-
dery supported Pd catalyst. We therefore investigated the
hydrogenation of nitroarenes under H2-FB conditions using
DOI: 10.1055/s-0036-1588869; Art ID: st-2017-u0280-l
Abstract Fine bubbles of hydrogen were employed as a new reaction
medium for the autoclave-free gas–liquid–solid multiphase hydrogena-
tion of nitro groups on a multigram scale. Furthermore, ultrafine bub-
bles were examined by nanoparticle-tracking analysis in organic sol-
vents.
Key words microbubbles, ultrafine bubbles, hydrogenation, hetero-
geneous catalysis, green chemistry, nitro compounds
The development of new chemical reactor technologies
that permit practical improvements in the efficiencies of
gas–liquid multiphase reactions for industrial production
or academic research has been a longstanding require-
ment.1 In recent years, fine bubbles (FBs)2,3 have been used
for gas–liquid reactions in the Cu/TEMPO-catalyzed aerobic
oxidation of primary alcohols to aldehydes, providing a
simple, safe, and user-friendly protocol.4 This method has
been further applied to gas–liquid–solid multiphase Pd-cat-
alyzed hydrogenations of C–C unsaturated bonds in an au-
toclave-free environment under atmospheric pressure
(Scheme 1, Equation 1).5 Although the behavior of FBs in or-
ganic solvents still remains unclear,6 in principle the FB-
mediated method can generally be applied to gas–related
multiphase reactions.7 Here, we report a Pd-catalyzed gas–
liquid–solid multiphase hydrogenation of nitroarenes
(Scheme 1, Equation 2) and we also report that the H2-FB
method overcomes the drawbacks of the conventional bub-
bling method. Furthermore, the existence of ultrafine bub-
bles of H2 (H2-UFB) has been confirmed by nanoparticle-
tracking analysis (NTA)8 in organic solvents.
previous work
Pd/Al2O3 (0.01–3 mol%)
H2-FB
R1
R3
R2
R4
R1
R3
R2
R4
(1)
MeOH, 30 °C, 1–6 h
yield up to >99.9%
up to 50 times higher yield than
obtained by conventional bubbling
this work
NO2
NH2
Pd/Al2O3 (3 mol%)
H2-FB
(2)
R1
R1
AcOEt, 30 °C, 5–12 h
yield up to >99.9%
up to 91 times higher yield than
obtained by conventional bubbling
Scheme 1 FB-mediated Pd-catalyzed hydrogenations
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 49, A–E

B
N. Mase et al.
Letter
Synlett
Pd on alumina spheres (Pd/Al2O3, 0.5% Pd, approx. 2–4 mm),
despite the lower reactivity of this catalyst compared with
that of Pd/C powder.
ences the reaction rate. In fact, hydrogenation was consid-
erably promoted under FB conditions, because the solution
rapidly became saturated with H2.5
Hydrogenation of nitrobenzene (1a; 10 mmol, 1.2 g) on
a gram scale was chosen as a model reaction for the evalua-
tion of the efficiency of the reaction (Table 1). To identify an
appropriate solvent for the FB-mediated hydrogenation, we
carried out solvent screening. Water and DMF were less ef-
fective solvents (Table 1, entries 1 and 2), whereas N-meth-
ylaniline was obtained as a byproduct in methanol, which is
often used as a solvent for hydrogenation reactions (entry
3).11 In frequently used solvents such as ethanol, THF, acetic
acid, or ethyl acetate, the desired aniline (2a) was obtained
in yields of over 90% (entries 4–7). Ethyl acetate was chosen
as the optimal solvent, due to its low miscibility with water
and the lack of byproduct formation during hydrogenation.
The superiority of the FB method was examined by com-
parison with conventional methods. The use of a balloon (1
atm) or bubbling of H2 through the solution from a conven-
tional gas-dispersion tube, fitted with a porous fritted-glass
tip, at the same H2 flow rate as used in the FB method re-
sulted in low yields (entries 8 and 9). The yield was not im-
proved even when a pressure-resistant reaction vessel at
high pressure (0.3 MPa) was used (entry 10). Although it is
possible to forcibly improve the yield by vigorous stirring, it
is not possible to reuse the catalyst under these conditions
due to the resulting damage. Since three hydrogen mole-
cules are involved in the hydrogenation of the nitro group,12
the concentration of dissolved hydrogen markedly influ-
Next, we examined the scope of the substrate in the FB
method13 in comparison with the conventional bubbling
method (Table 2). In the case of electron-donating substitu-
ents, such as methyl or methoxy, the corresponding aniline
derivatives 2 were produced in quantitative yields (entries
2 and 3). Similarly, hydrogenation of unprotected hydroxy-
or amino-substituted nitroarenes proceeded smoothly (en-
tries 4 and 5). In the case of (4-nitrophenyl)amine (1e), the
difference in reactivity between conventional bubbling and
FB conditions was remarkable (91 times faster; entry 5).
During the hydrogenation of the nitro group, the aniline
product is adsorbed onto the catalyst and might decrease
its catalytic activity.14 However, the exchange of the prod-
uct with hydrogen at the catalytically active site is consid-
ered to be rapid under FB conditions in which hydrogen is
dissolved at high concentrations. As a result, deactivation of
the catalyst is suppressed, even in the case of benzene-1,4-
diamine (2e), which is easily adsorbed onto the catalyst. Al-
though nitroarenes substituted with electron-withdrawing
groups were readily hydrogenated in high yields, simulta-
neous partial dehalogenation of 1-chloro-4-nitrobenzene
(1g) proceeded in 13% yield (entries 6–8). The heterocyclic
5-nitroquinoline (1i) was efficiently hydrogenated under FB
conditions (entry 9).
Because FB-mediated hydrogenations can efficiently
and easily be performed at atmospheric pressure, this
method is expected to be suitable for the industrial produc-
tion of 3-(4-aminophenoxy)aniline (2j), a monomer used in
the preparation of a manmade high-performance aramid fi-
ber. Monomer 2j has been prepared on an industrial scale
by the hydrogenation of [3-(4-aminophenoxy)phe-
nyl]amine (1j) in DMF under high-temperature and high-
pressure conditions.15 When the amount of catalyst was in-
creased to 4 mol% to compensate for the low hydrogenation
reactivity in DMF (Table 1, entry 2), the desired diamine
monomer 2j was obtained twice as efficiently when com-
pared to conventional bubbling conditions (entry 10).
Because N-methylaniline was obtained as a byproduct
during nitro-group reduction in methanol (Table 1, entry
3),11 reductive amination reactions were expected to occur
in the presence of ketones or aldehydes. Indeed, the second-
ary and tertiary amines 3 and 4, in the presence of acetone
and butyraldehyde, respectively, were produced in one-pot
operations in yields of over 90% (Scheme 2, Equations 1 and
2). Not only are nitro and imino groups smoothly reduced
in the FB method, but also carbonyl groups are efficiently
reduced in a synthetically useful manner; for example
benzaldehyde was quantitatively reduced by using 3 mol%
Pd/Al2O3 and H2-FB (5 mL/min) to give benzyl alcohol (see
Supporting Information).
Table 1 Comparison of Hydrogenation Conditionsa
NO2
NH2
Pd/Al2O3 (3 mol%), H2
30 °C, 5 h
1a
2a
Entry
Conditions
H2 (mL/min) Solvent
Yieldb (%)
1
2
FB
5
5
5
5
5
5
5
–
5
–
H2O
69.0
41.0
68.6c
93.0
96.1
>99.9
>99.9
1.6
FB
DMF
3
FB
MeOH
EtOH
THF
4
FB
5
FB
6
FB
AcOH
AcOEt
AcOEt
AcOEt
AcOEt
7
FB
8
balloon
bubbling
autoclaved
9
5.2
10
1.0
a Reaction conditions: aniline (1a, 10 mmol), Pd/Al2O3 (3 mol%), solvent
(80 mL, 0.125 M), H2-FB or bubbling (5 mL/min) or balloon, 30 °C, 5 h.
b Determined by GC analyses (column: GL Sciences TC-17).
c N-Methylaniline was formed as a byproduct in 20% yield.
d This reaction was carried out at 0.3 MPa.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 49, A–E

C
N. Mase et al.
Letter
Synlett
Table 2 Substrate Scope in FB-Mediated Hydrogenation of Nitroarenesa
Pd/Al2O3 (3 mol%)
Conditions
A: H2-FB
NO2
NH2
B: H2-bubbling
R
R
AcOEt, 30 °C
1
2
Entry
Substrate
R
Time (h)
Product
Yield Ab (%)
Yield Bc (%)
Ratiod
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1f
H
5
12
5
2a
2b
2c
2d
2e
2f
>99.9
>99.9
>99.9
91.6
5.2
8.7
19
11
24
7
4-Me
4-OMe
4-OH
4-NH2
4-F
4.2
9
14.0
<1.0
6.1
6
91.0
91
16
28
24
7
>99.9
86.7
1g
1h
1i
4-Cl
7
2g
2h
3.1e
3.7
4-CO2Me
12
90.4
NO2
9
10
2i
97.0
3.4
29
N
1j
H2N
O
10f
7
2j
>99.9
48.0
2
NO2
a Reaction conditions: nitroarene 1 (10 mmol), Pd/Al2O3 (3 mol%), solvent (80 mL, 0.125 M), H2-FB (A) or bubbling (B) (5 mL/min), 30 °C (see also Reference 13).
b Yield under FB conditions, determined by GC (column: GL Sciences TC-17).
c Yield under bubbling conditions, determined by GC (column: GL Sciences TC-17)
c Ratio = Yield A/Yield B.
d The dehalogenation product 2b was obtained in 13% yield.
e The reaction was carried out in DMF.
been studied. We reported preliminary analytical results
H
Pd/Al2O3 (3 mol%)
H2-FB (10 mL/min)
NO2
N
for UFBs in methanol. The USBs were produced by using an
FB generator (MA3-FS), and the H2-UFB number (2.7 × 107
particles/mL) and the average size of the bubbles (158 nm)
were determined by using a NanoSight LM10-HS nanopar-
ticle-tracking analysis (NTA) instrument, fitted with a red
(638 nm) laser.5 To understand the behavior of UFBs in
more detail, we chose to use a NanoSight LM10-VHST in-
strument, which has a purple (405 nm) laser that increases
the intensity of the light scattered by the UFBs. H2-UFBs
were observed in the 50–200 nm size range in water, with
an average of 148.5 ± 6.4 nm (Figure 1, left). The number of
UFBs per milliliter was determined to be (1.35 ± 0.17) × 108;
therefore, purple lasers can visualize more UFBs than can
red lasers. Despite the small numbers of visually observable
microscale bubbles in the ethyl acetate used during the hy-
drogenation of nitroarenes, the number and sizes of the H2-
UFBs in this solvent were confirmed to be similar to those
in H2O [average size: 134.1 ± 10.8 nm, number of H2-
UFBs/mL: (2.17 ± 0.38) × 107, Figure 1, right].16
(1)
(2)
AcOEt/acetone = 1:1
30 °C, 12 h
O
O
O
1c
1c
3 (97.6%)
Bu
Pd/Al2O3 (3 mol%)
H2-FB (10 mL/min)
PrCHO (10 equiv)
NO2
N
Bu
AcOEt, 30 °C, 12 h
O
4 (94.1%)
Scheme 2 FB-mediated reductive aminations
UFBs are too small to be observed by visual inspection
or optical microscopy; consequently the solutions appear to
be clear because of the extremely small size of the dis-
persed bubbles. However, it has recently become possible to
examine UFBs through technological innovations in
nanoparticle-analysis equipment. Consequently, the behav-
ior of UFBs in water is gradually being revealed. On the oth-
er hand, the analysis of UFBs in organic solvents has hardly
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 49, A–E

D
N. Mase et al.
Letter
Synlett
In principle, FBs can be generated from various gases;
indeed, UFBs of various gases were observed not only in
water, but also in ethyl acetate (Figure 3). The UFB number
in water ranged from 1 × 108 to 1.6 × 108 and that in AcOEt
ranged from 2 × 107 to 2.6× 107. Interestingly, the UFB num-
ber depends on the solvent, but is largely independent of
the gas.
Figure 1 Intensity as a function H2-UFB size in H2O (left) and AcOEt
(right)
Next, air-UFBs were examined in ten conventional or-
ganic solvents (Figure 2); UFBs were observed in all the or-
ganic solvents examined except hexane. The microbubble
(MB) rising speed depends on the viscosity (η) of the liquid,
in accordance with the Stokes equation, and the surface
zeta potential of the MB depends on the permittivity (ε) of
the liquid, in accordance with the Smoluchowski equation.2
Consequently, the viscosity and permittivity of the liquid
should not only affect the stability of MBs, but also that of
UFBs. We observed no clear correlation between the viscos-
ity or permittivity and the UFB number (the number of
UFBs per milliliter). In contrast, the calculated value (CV)
obtained by multiplying the normalized viscosity and per-
mittivity, based on the values for water, correlates linearly
with the UFB number (Figure 2). In hexane, which has a low
viscosity (0.3 mPa·s) and low permittivity (1.9 ε), UFBs are
hardly observed under our measurement conditions. When
the calculated value is larger than 0.01, the UFB number can
be determined. As the viscosity and/or permittivity in-
creases, more UFBs are detected. For example, propan-2-ol
(η = 2.0, ε = 19.9, CV = 0.58) produces an air-UFB number of
6.2 × 107. Water (η = 0.9, ε = 78.5, CV = 1.00) and DMSO (η =
2.0, ε = 46.5, CV = 1.32) contain more than 108 UFBs per mil-
liliter of liquid.
Figure 3 UFB numbers of various gases in H2O and AcOEt
For example, when less water-soluble nitrogen gas was
used, 1.08 × 106 UFBs per milliliter were observed, whereas
for highly soluble ammonia gas in water, 1.13 × 106 UFBs
per milliliter were observed. It is apparent that the average
UFB size does not depend on the kind of gas or the solvent
under any conditions. The average UFB sizes of various gas-
es in H2O or AcOEt lie in the 100–200 nm range (Figure S1
in Supporting Information). With our current level of un-
derstanding, the reason why the numbers of UFBs and their
average sizes are largely independent of the gas or solvent
remains unclear. However, the rate of gas dissolution into
the liquid and the saturation depend on the type of gas and
liquid, but UFB numbers and sizes are about the same in
their gas–liquid-saturated equilibrium states.
In conclusion, H2-FBs were used as a new reaction medi-
um for gas–liquid–solid multiphase reactions in autoclave-
free multigram-scale hydrogenations of nitro groups. In ad-
dition, several kinds of UFB were examined by NTA in or-
ganic solvents. FBs were generated by using various gases
and liquids; therefore, the FB method is widely applicable.
This environmentally friendly FB-mediated gas-related re-
action system has the potential to contribute to the synthe-
sis of various fine chemicals, as well as bulk chemicals.
Funding Information
This work was supported in part by a Grant-in-Aid for Scientific Re-
search (B) (No. 15H03844) for scientific research from the Japan Soci-
Figure 2 Plot of the air-UFB number as a function of the solvent
permittivity × viscosity (normalization based on H2O)
ety for the Promotion of Science.
a
Jp
a
n
S
o
ecity
o
f
r
hte
P
or
m
o
itn
of
Secince
1(
5
H
0
3
8
4
4)
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 49, A–E

E
N. Mase et al.
Letter
Synlett
Supporting Information
(9) (a) Blaser, H.-U.; Steiner, H.; Studer, M. ChemCatChem 2009, 1,
210. (b) Blaser, H.-U.; Malan, C.; Pugin, B.; Spindler, F.; Steiner,
H.; Studer, M. Adv. Synth. Catal. 2003, 345, 103. (c) Downing, R.
S.; Kunkeler, P. J.; van Bekkum, H. Catal. Today 1997, 37, 121.
(10) Pd/C can ignite on exposure to air; see: Committee on Prudent
Practices for Handling, Storage, and Disposal of Chemicals in Lab-
oratories, Prudent Practices in the Laboratory, Handling and Dis-
posal of Chemicals; National Academy Press: Washington, 1995.
(11) Although the formation mechanism is not clear, the reaction is
considered to proceed through the reductive amination of N-
methyleneaniline, formed in situ from aniline and formalde-
hyde; the latter is in turn generated by dehydrogenation of
methanol on the catalyst surface.
Supporting information for this article is available online at
https://doi.org/10.1055/s-0036-1588869.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
References and Notes
(1) (a) Kashid, M. N.; Renken, A.; Kiwi-Minsker, L. Chem. Eng. Sci.
2011, 66, 3876. (b) Kiwi-Minsker, L.; Renken, A. Catal. Today
2005, 110, 2. (c) Jakobsen, H. A.; Lindborg, H.; Dorao, C. A. Ind.
Eng. Chem. Res. 2005, 44, 5107; and references cited therein.
(2) (a) Parmar, R.; Majumder, S. K. Chem. Eng. Process. 2013, 64, 79.
(b) Craig, V. S. J. Soft Matter 2011, 7, 40. (c) Agarwal, A.; Ng, W.
J.; Liu, Y. Chemosphere 2011, 84, 1175; and references cited
therein.
(12) Möbus, K.; Wolf, D.; Benischke, H.; Dittmeier, U.; Simon, K.;
Packruhn, U.; Jantke, R.; Weidlich, S.; Weber, C.; Chen, B. Top.
Catal. 2010, 53, 1126.
(3) Fine bubbles (FBs) include microbubbles (MBs) and nanobub-
bles (NBs). The term NB is used to describe gas-filled spherical
bubbles that have diameters of less than 1000 nm. An alterna-
tive and equivalent term, also used in the literature, is ‘ultrafine
bubbles’ (UFBs). Since the International Standards Organization
is currently evaluating standards for UFBs (ISO/TC281), we use
the term ‘UFB’ in this communication; see: Alheshibri, M.; Qian,
J.; Jehannin, M.; Craig, V. S. J. Langmuir 2016, 32, 11086.
(4) Mase, N.; Mizumori, T.; Tatemoto, Y. Chem. Commun. (Cam-
bridge) 2011, 47, 2086.
(13) Hydrogenation of Nitroarenes 1 by a H2-FB-Based Strategy;
General Procedure
The hydrogenation was carried out in a 100 mL vial equipped
with an FB generator, without additional stirring. Nitroarene 1
(10 mmol) was dissolved in AcOEt (80 mL) and the solution was
warmed to 30 °C. By using the FB generator (MA3-FS), H2-FB
were introduced into the reactor in the presence of Pd on
alumina spheres (0.5% Pd, 2–4 mm, 0.3 mmol, 3 mol%) at a H2
flow rate of 5 mL/min. Samples of the reaction mixture were
removed periodically to permit monitoring of the progress of
the reaction by GC analysis. When the hydrogenation reaction
was complete, the AcOEt was evaporated in vacuo to afford the
desired aniline 2 with good to excellent purity.
(5) (a) Mase, N.; Isomura, S.; Toda, M.; Watanabe, N. Synlett 2013,
24, 2225. (b) Nagano, T.; Kanemitsu, M.; Motoyama, T.; Mase, N.
JP 2013023460, 2013.
(6) Although UFBs appear to be specific to water and aqueous solu-
tions according to a first report,6a ‘surface UFBs’ have been
observed in solvents such as formamide, ethylammonium
nitrate, and propylammonium nitrate, but not in propylene car-
bonate or dimethyl sulfoxide.6b To date it is thought that
bubbles generated in nonaqueous solutions are not stable and
disappear rapidly: (a) Seddon, J. R. T.; Lohse, D. J. Phys.: Condens.
Matter 2011, 23, 133001. (b) An, H.; Liu, G.; Atkin, R.; Craig, V. S.
J. ACS Nano 2015, 9, 7596.
(7) The synthesis of a five-membered cyclic carbonate derived from
an epoxide in the presence of carbon dioxide micro- and
nanobubbles has recently been disclosed; see: Uruno, M.;
Takahashi, K.; Kimura, K.; Muto, K.; Tanigawa, M.; Hanada, K. JP
2016190799, 2016.
GC analyses: SHIMADZU GC-2010, capillary column: GL Sci-
ences TC-17; He = 0.80 MPa, H2 = 0.50 MPa, air = 0.50 MPa, flow
rate: 1.4 mL/min, Tinj = 250 °C, Tdet = 250 °C, Ti = 100 °C, Tf =
250 °C, rate = 10 °C/min; Nitrobenzene (1a, CAS Reg. No.: 98-
95-3; tR = 5.3 min), aniline (2a, CAS Reg. No: 62-53-3; tR = 5.5
min) (see Supporting Information).
(14) Takasaki, M.; Motoyama, Y.; Higashi, K.; Yoon, S.-H.; Mochida, I.;
Nagashima, H. Org. Lett. 2008, 10, 1601.
(15) Shimada, K.; Yoshisato, E.; Yoshitomi, T.; Matsumura, S. JP
1995242606, 1995.
(16) The reactivity in H2O was lower than that in AcOEt despite the
larger number of UFBs observed in H2O (Table 1, entry 1 vs. 7).
Because the reaction rate of the hydrogenation was propor-
tional to the concentration of dissolved hydrogen, the reactivity
decreased in H2O, in which H2 gas has a low solubility. In addi-
tion, the solubility of the hydrophobic substrate should also
affect the reaction rate and, consequently, the reactivity in H2O
would be reduced.
(8) (a) Tian, X.; Nejadnik, M. R.; Baunsgaard, D.; Henriksen, A.;
Rischel, C.; Jiskoot, W. J. Pharm. Sci. 2016, 105, 3366. (b) Filipe,
V.; Hawe, A.; Jiskoot, W. Pharm. Res. 2010, 27, 796.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 49, A–E