A green approach to ethyl acetate: Quantitative conversion of ethanol through direct dehydrogenation in a Pd-Ag membrane reactor

Article abstract of DOI:10.1002/chem.201202005

Pincers do the trick: The conversion of ethanol to ethyl acetate and hydrogen was achieved using a pincer-Ru catalyst in a Pd-Ag membrane reactor. Near quantitative conversions and yields could be achieved without the need for acid or base promoters or hydrogen acceptors (see scheme).

Full text of DOI:10.1002/chem.201202005

DOI: 10.1002/chem.201202005
A Green Approach to Ethyl Acetate: Quantitative Conversion of Ethanol
through Direct Dehydrogenation in a Pd–Ag Membrane Reactor
[
a, b]
[a, c]
[a, c]
[a, b]
[a, b]
Gaofeng Zeng,
Tao Chen,
Lipeng He,
Ingo Pinnau,
Zhiping Lai,*
and
[
a, c]
Kuo-Wei Huang*
II
Ethyl acetate (EA) is an important industrial product em-
ployed in a wide range of applications from solvent to inter-
We have recently demonstrated that our PNN–Ru com-
plexes 1 and 2 (PNN = phosphorus-nitrogen-nitrogen pincer
ligands) show comparable catalytic activities towards dehy-
drogenation of primary alcohols to the corresponding esters
[1]
mediates. In current industrial processes, EA is mainly
produced by Fischer esterification of ethanol (EtOH) and
acetic acid, the Tishchenko reaction of acetaldehyde, or ad-
dition of acetic acid to ethylene (BPꢀs Avada process) on a
smaller scale. These processes, in addition to the use of cor-
rosive and/or toxic reagents/substrates, require energy-inten-
sive distillation operations. Direct conversion of EtOH to
EA through dehydrogenative coupling with the liberation of
[9]
to that of Milsteinꢀs catalyst 3 (Entries 1 and 2, Table 1).
II
[a]
Table 1. Esterification of primary alcohols with PNN–Ru complexes.
H is an attractive alternative and represents a simple, non-
2
corrosive, relatively nontoxic, and economical process. The
reaction only consumes EtOH, an inexpensive, renewable
raw material, as the feedstock, and provides EA and valua-
ble hydrogen as the products. Many heterogeneous catalysts
have been examined for this goal, yet most of them suffer
[
2–6]
from low conversion rates and/or selectivity.
As EtOH
R
Cat.
Cat. loading
[mol%]
T
[8C]
Conv.
[%]
[b]
A
H
U
G
E
N
N
forms an azeotrope with EA, separation of EA from the re-
sulting mixtures is challenging and energy consuming. On
the other hand, homogeneous catalysts have the potential to
achieve high conversion and selectivity, and in fact, a series
of Ru–pincer complexes have shown promising catalytic ac-
tivities in the dehydrogenative coupling of primary alcohols
1
2
3
4
5
6
CH
CH
Me
Me
Me
Me
3
3
A
H
G
R
N
N
2
2
)
)
4
ꢀ
1
3
1
1
3
2
0.1
0.1
0.5
0.5
0.5
0.5
160
157
78
160
160
160
98
91.5
0
[
c]
A
H
U
G
R
N
U
G
4
ꢀ
[d]
[d]
75
75
69
[
d]
[7,8]
to esters in high yields at 110–1608C.
However, extension
[
a] Alcohol (5 mmol) and a suitable amount of ruthenium complex in a
of these reactions to alcohols with lower boiling points, such
as ethanol and 1-propanol, still represents a major challenge
due to the high reaction temperature needed.
Schlenk flask equipped with a cooling finger were heated at the stated
temperature under argon flow for 24 h. [b] Determined by H NMR spec-
troscopy (>99% selectivity for EA formation). [c] Ref. [7]]. [d] The reac-
tions were performed in 25 mL sealed tubes under argon atmosphere.
1
Consistent with previous literature results, when EtOH was
employed no reaction was observed after heating at reflux
+
+
[
a] Dr. G. Zeng, Dr. T. Chen, Dr. L. He, Prof. Dr. I. Pinnau,
Prof. Dr. Z. Lai, Prof. Dr. K.-W. Huang
[8]
in the presence of 1 for 24 h (Entry 3, Table 1). However,
9–75% conversion was achieved by carrying out the reac-
Division of Chemical and Life Sciences and Engineering
King Abdullah University of Science and Technology (KAUST)
Thuwal 23955-6900 (Saudi Arabia)
6
tions in a pressure tube with 1–3 (Entries 4–6, Table 1), indi-
II
cating that the pincer–Ru catalysts also work effectively for
Fax : (+966)2-8082407
E-mail: zhiping.lai@kaust.edu.sa
small alcohols when the reaction can be conducted at elevat-
ed temperatures. Because a closed system is required to ach-
ieve such conditions, we rationalized that the moderate con-
versions were due to the reached equilibrium limit. As the
hkw@kaust.edu.sa
+
[
b] Dr. G. Zeng, Prof. Dr. I. Pinnau, Prof. Dr. Z. Lai
Advanced Membranes and Porous Material Center
Thuwal 23955-6900 (Saudi Arabia)
total pressure builds up, because of the formation of H , the
2
+
[
c] Dr. T. Chen, Dr. L. He, Prof. Dr. K.-W. Huang
KAUST Catalysis Center
chemical equilibrium is backshifted [Eq. (1)].
Thuwal 23955-6900 (Saudi Arabia)
+
[
] These authors made equal contributions.
2
EtOH ! EA þ 2H
2
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202005.
ð1Þ
o
ꢀ1
o
¼ 52:2 J molꢀ1
D H
¼ 26 kJ mol ; D S
r
298 K
r
298 K
15940
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15940 – 15943

COMMUNICATION
From its thermodynamics, the formation of EA and H is
preferred at high temperature and under low pressure.
brane was swept by argon with a typical volume flow rate of
2
ꢀ
1
200 mLmin to purge all the permeate hydrogen to a GC
(Agilent 7890 A) for the online-composition analysis which
enables the hydrogen release rate to be calculated.
Therefore, if the generated H can be selectively separated
2
in situ from the reaction mixtures, one will be able to drive
the reaction to completion. In this regard, supported Pd-
based composite membranes are the most applicable candi-
dates for selective hydrogen removal owing to their high
and exclusive hydrogen permeability, high thermal stability,
II
In a typical reaction, EtOH (1 mL) and PNN–Ru com-
plex 1 were loaded into the reaction chamber. The reaction
was then monitored until completion. The H permeability
2
of the membrane in our design ensures complete removal of
the produced hydrogen during the course of the reaction.
[10–15]
moderate chemical resistance, and mechanical strength.
Such a membrane-reactor concept has been explored previ-
ously for ethanol dehydrogenation with limited suc-
Figure 2 shows the H release rate and the total pressure of
2
[13,16,17]
cess.
Beller and co-workers recently reported that up
to 81% yield could be achieved for the dehydrogenation of
[18]
EtOH to EA with a more reactive Ru catalyst at 908C.
Considering the extreme difficulty in the separation of
EtOH and EA, higher conversions and yields are practically
desirable. In this communication, we report our design to
employ a defect-free ultrathin Pd–Ag/ceramic membrane re-
actor to overcome the above-mentioned issues to achieve
quantitative conversion of EtOH to EA and H without the
2
use of solvent, acid or base promoters or additional H ac-
2
ceptors.
The 5 mm thick Pd Ag membrane supported on an
7
5
25
asymmetric porous ceramic tube was prepared by a modi-
fied electroless plating process (see details in the Supporting
Information). The H fluxes of the fresh membrane under
2
ꢀ
2
ꢀ1
Figure 2. H2 liberation rate and the total pressure of the reaction at dif-
ferent reaction times at 1608C.
DP=100 kPa were 0.28 and 0.47 molm
s
at 150 and
3
508C, respectively. The H /N selectivity was almost infinite
2
2
because no N leakage was detected (see details in the Sup-
2
porting Information). The tubular alloy membrane was
mounted in a custom-designed reaction cell as shown in
Figure 1 and the cell was placed vertically inside a tubular
furnace. Both ends of the membrane were sealed with
graphite O-rings to create a reaction chamber with an effec-
the chamber at time intervals of 20 s during the reaction
when carried out at 1608C. The final conversion of EtOH
1
and the yield of EA were analyzed by H NMR spectrosco-
py and/or GC-MS. Moreover, this was also confirmed with
the known total pressure inside the reaction chamber as
well as the total volume of the hydrogen released by inte-
grating the hydrogen release curve in Figure 2. If we assume
that there is no hydrogen accumulation in the system, the
total pressure inside the reaction chamber will be deter-
mined only by the partial pres-
2
tive membrane surface area of approximately 6 cm and a
volume of about 2 mL between the membranes and the
shell of the cell, in which the solution of EtOH and catalysts
are contained. The pressure of the chamber was monitored
by a pressure sensor. The lumen side of the tubular mem-
sures of EtOH and EA accord-
ing to Raoultꢀs law in which x is
the molar fraction of EtOH in
the
reaction
mixture
[
19]
[
Eq. (2)].
Sat
EtOH
Sat
EA
P ¼ P
x þ P ð1ꢀxÞ
ð2Þ
For example at 1608C, based
on NMR and GC-MS analysis,
the EtOH conversion was
roughly 99% and the EA yield
was roughly 99% after 24 h.
The EtOH conversion calculat-
ed from the hydrogen release
curve was 96%, while the total
pressure indicated that the
Figure 1. Schematic of the membrane reactor for the dehydrogenation of EtOH.
Chem. Eur. J. 2012, 18, 15940 – 15943
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15941

Z. Lai, K.-W. Huang et al.
EtOH conversion was 98%. The influences of reaction tem-
perature and catalyst loading on the conversion of EtOH
were examined. The results are summarized in Table 2. The
conversions of EtOH with 0.02 mol% of catalyst were
found to increase with elevated temperatures, reaching 36,
57, and 65% at 160, 180, and 2008C, respectively (Figure 3).
Increasing the catalyst loading at 1608C improved the con-
Table 2. Dehydrogenative coupling of EtOH to EA and H
Ru complex 1 in a Pd–Ag membrane reactor.
2
with PNN–
II
1
T
[8C]
Conv.
[%]
Yield
[%]
Conv. [%]
V 2
H
[
a]
[b]
[b]
[c]
[d]
A[ mol%]
C
H
T
U
N
G
T
R
E
N
N
U
N
G
DP
1
2
3
4
5
6
0.01
0.02
0.02
0.02
0.10
0.50
200
160
180
200
160
160
25
36
57
65
96
99
98
98
99
99
99
99
1
22
33
59
66
94
96
27
40
60
68
93
98
Figure 4. H
catalytic loading of complex 1.
2
release rates of EtOH dehydrogenation at 1608C versus the
H and N fluxes of the membrane were measured be-
2
2
II
[
a] PNN–Ru complex 1. [b] Determined by H NMR spectroscopy and
GC-MS. [c] Determined by H produced. [d] Determined by the total
pressure inside the reaction chamber.
tween 150–3508C after all reactions were finished. No N2
leakage of the membrane was detected at room temperature
with DP=800 kPa. However, the H2 flux of the Pd–Ag
membrane at 3508C with DP=100 kPa dropped to
2
ꢀ
2
ꢀ1
0
.17 molm s , approximately 37% of the pristine mem-
brane. At the same time, the activation energy of H perme-
2
ꢀ
1
ation increased from 5.8 to 17.4 kJmol . However, after
washing with methanol followed by annealing at 5008C in
H2 for 48 h, the H2 flux can be recovered to 77% of the
fresh membrane and the activation energy can be dropped
ꢀ
1
to 8.3 kJmol (Figure S2 in the Supporting Information).
This indicates that the decline of H flux and the decrease
2
in permeability were most likely caused by the physical ad-
sorption of reagents and catalysts on the surface of the
membrane. Inspection of the Pd–Ag membrane after reac-
tion by SEM showed no coke formation on the surface (Fig-
ure S3 in the Supporting Information). It was reported that
EtOH may decompose to CH , CO, and H in the presence
4
2
[5]
of Pd catalysts, but such reactions were not observed in
our system, since the pressure of the reaction chamber was
synchronously dropped to zero when it was cooled to room
temperature, indicating no formation of other gases during
our process.
2
Figure 3. H release rates of EtOH dehydrogenation with 0.02 mol%
complex 1 versus the reaction temperature.
In summary, we have demonstrated that quantitative pro-
duction of ethyl acetate and hydrogen directly from EtOH
can be achieved in an ultrathin Pd–Ag membrane reactor.
This environmentally benign reaction is homogeneously cat-
alyzed by dearomatized pincer Ru complexes and no acid or
base promoters or hydrogen acceptors are required. A
defect-free Pd–Ag membrane was employed for the in situ
removal of the liberated hydrogen, allowing the reaction to
be operated at elevated temperatures and reach almost
100% conversion and EA production yields. Furthermore,
version from 36 (0.02) to 96% (0.10 mol%) and eventually
to 99% (0.50 mol%) (Figure 4). As can be seen, conversions
calculated by different methods agree very well with one an-
other, indicating that the Pd–Ag membrane can efficiently
remove hydrogen from the system, and more importantly,
suggests that in our membrane reactor system, the progress
of the reaction can be monitored accurately. This may po-
tentially serve as a powerful tool to study the reactionꢀs ki-
netics.
The mechanical stability of the Pd–Ag membrane was
proven by the tolerability to high pressure of 12–30 bar in
the reaction temperature range of 160–2008C. The mem-
brane also showed good thermal stability and chemical re-
sistance over more than 500 h of service for this reaction
system.
the liberated H fluxes and the inner pressure of the reactor
2
can both be used to accurately monitor the reaction progress
in real time. Further investigations to extend the scope of
our design in other dehydrogenative reactions are currently
ongoing in our laboratories and will be reported in due
course.
15942
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15940 – 15943

Green Approach to Ethyl Acetate
COMMUNICATION
Acknowledgements
[7] J. Zhang, G. Leitus, Y. Ben-David, D. Milstein, J. Am. Chem. Soc.
2
005, 127, 10840–10841.
8] C. Gunanathan, L. Shimon, D. Milstein, J. Am. Chem. Soc. 2009,
31, 3146–3147.
9] L. He, T. Chen, D. Gong, Z. Lai, K.-W. Huang, Organometallics
012, 31, 5208–5211.
[
[
We are grateful for the generous financial support from King Abdullah
University of Science and Technology.
1
2
[
[
[
10] Y. Yamada, T. Watanabe, T. Beppu, N. Fukuyama, K. Torii, Y.
Keywords: dehydrogenation · ethanol · ethyl acetate ·
membrane reactor · ruthenium
Uozumi, Chem. Eur. J. 2010, 16, 11311–11319.
11] S. Niwa, M. Eswaramoorthy, J. Nair, A. Raj, N. Itoh, H. Shoji, T.
Namba, F. Mizukami, Science 2002, 295, 105–107.
12] L. Shi, A. Goldbach, G. Zeng, H. Xu, J. Membr. Sci. 2010, 348, 160–
1
66.
[
1] K. Weissermel, H. Arpe, in Industrial Organic Chemistry, 4th ed.,
Wiley-VCH, Weinheim, 2003, pp. 188–189.
[13] J. Keuler, L. Lorenzen, Ind. Eng. Chem. Res. 2002, 41, 1960–1966.
[14] G. Zeng, A. Goldbach, H. Xu, J. Membr. Sci. 2009, 326, 681–687.
[15] G. Zeng, A. Goldbach, L. Shi, H. Xu, Int. J. Hydrogen Energy 2012,
37, 6012–6019.
[
[
2] K. Inui, T. Kurabayashi, S. Sato, J. Catal. 2002, 212, 207–215.
3] E. Santacesaria, G. Carotenuto, R. Tesser, M. Serio, Chem. Eng. J.
2
012, 179, 209–220.
4] S. Colley, J. Tabatabaei, K. Waughb, M. A. Wood, J. Catal. 2005,
36, 21–33.
[16] B. A. Raich, H. C. Foley, Ind. Eng. Chem. Res. 1998, 37, 3888–3895.
[17] W. H. Lin, H. F. Chang, Catal. Today 2004, 97, 181–188.
[18] M. Nielsen, H. Junge, A. Kammer, M. Beller, Angew. Chem. 2012,
124, 5809–5811; Angew. Chem. Int. Ed. 2012, 51, 5711–5713.
[19] G. Thomson, Chem. Rev. 1946, 38, 1–39.
[
[
[
2
5] A. Sꢂnchez, N. Homs, S. Miachon, J. Dalmon, J. Fierro, P. Piscina,
Green Chem. 2011, 13, 2569–2575.
6] A. Johansson, E. Zuidema, C. Bolm, Chem. Eur. J. 2010, 16, 13487–
Received: June 7, 2012
1
3499.
Published online: November 7, 2012
Chem. Eur. J. 2012, 18, 15940 – 15943
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15943