Novel anion exchange resin-based catalyst for liquid-phase methanol synthesis at 373-393 K

Article abstract of DOI:10.1246/cl.2004.1252

A thermo-stable anion exchange resin-Raney Cu system was found as the most effective solid catalyst for low-temperature liquid-phase methanol synthesis at 373 to 393 K under 5.0 MPa of syngas (2H₂/CO). With the catalyst (20 mL of the resin and 2.0 g of Cu) suspended in methanol solution 72% of CO was converted to methanol (70%) and methyl formate (HCOOCH₃) (30%) in 4 h.

Full text of DOI:10.1246/cl.2004.1252

1252
Chemistry Letters Vol.33, No.10 (2004)
Novel Anion Exchange Resin-based Catalyst for Liquid-phase Methanol Synthesis at 373–393 K
Ken-ichi Aika,ꢀ Hidenobu Kobayashi, Kenji Harada, and Koji Inazu
Department of Environmental Chemistry and Engineering, Interdisciplinary Graduate School of Science and Engineering,
Tokyo Institute of Technology, G1-13-4259 Nagatsuta, Midori-ku, Yokohama 226-8502
(Received July 22, 2004; CL-040863)
A thermo-stable anion exchange resin–Raney Cu system
was found as the most effective solid catalyst for low-tempera-
ture liquid-phase methanol synthesis at 373 to 393 K under
5.0 MPa of syngas (2H2/CO). With the catalyst (20 mL of the
resin and 2.0 g of Cu) suspended in methanol solution 72% of
CO was converted to methanol (70%) and methyl formate
(HCOOCH3) (30%) in 4 h.
Table 1. Percentage CO conversion of methanol synthesis
(2H2 þ CO ¼ CH3OH) under the equilibrium as functions of
temperature and pressure
Total Pressure/MPa
CH3OH Vapor
Temperature/K
0.1
0.5
2.0
5.0
10.0
Pressurea/MPa
473
453
423
403
373
1.1
3.3
27
83
—
82
92
—
80
89
93
97
84
89
94
96
98
90
93
96
98
99
3.8
2.6
21
1.3
—
—
0.60
0.35
Methanol is produced through gas phase reaction using cop-
per–zinc based catalyst at 523–573 K under 5–20 MPa. Under
these conditions, equilibrium limit of methanol yield is 20 to
40%. The separation and recycling process of the unreacted
syngas requires large energy consumption.
Equilibrium conversion of methanol synthesis is high at low
temperatures and high pressures as shown in Table 1. The group
of italic number shows the condition under which methanol va-
por pressure is lower than the equilibrium methanol pressure, so
that methanol synthesized in the gas phase condenses automati-
cally to liquid.1 If any solid catalyst is developed and is active at
such low temperatures, neither separation nor recycle process is
necessary. The solved syngas and intermediate methyl formate
(HCOOCH3) are the possible impurities of the methanol
product.
aAutomatic product separation occurs if methanol vapor pressure is lower
than the equilibrium pressure of the synthesized methanol.
The aim of our project is to replace sodium methoxide with
solid bases or anion exchange resins to find an active solid cata-
lytic system for methanol synthesis. Solid bases such as MgO or
CaO did not catalyze the carbonylation reaction described by
Eq 1, while anion-exchange resins such as AmberlystÒ A26
(Rohm and Haas Co. Ltd., –CH2N(CH3)3þClꢁ, maximum oper-
ating temperature of 333 K) is known to be active to exchanging
alkoxides (Eq 3) and to catalyze Eq 1.6–8 Around 83% of CO was
converted to methyl formate at 333 K under 5.0 MPa of CO using
A26 exchanged with methoxide.7
ResinþClꢁ þ NaOCH3 ! ResinþCH3Oꢁ þ NaCl
ð3Þ
Several catalyst systems such as alkali alkoxide mixed with
Ni compounds or Cu–Cr2,3 and sodium methoxide–Ni4 have
been reported as the first success for the low temperature liquid
phase methanol synthesis. The reaction proceeds through the
following two steps:
Carbonylation (Eq 1) is proved to be such an easy reaction
over resin–alkoxide system working at low temperature. In this
instance is needed any heat-resistant anion-exchanging resin
working at high temperatures (373 to 423 K). Because the hydro-
genation of methyl formate (Eq 2) is a slow reaction over metals,
we intended to operate the two reactions above 373 K at the
same time. Meanwhile, a heat-resistant anion-exchanging resin
with –(CH2)4–N(CH3)3þClꢁ functional group (DIAIONÒ
TSA1200, Mitsubishi Chemical Co. Ltd., maximum operating
temperature of 373 K) was newly developed,9 and tested for
carbonylation (Eq 1) at 373–393 K.
CH3OH þ CO ! HCOOCH3
ð1Þ
ð1-aÞ
CH3Oꢁ(þNaþ) þ CO ! CH3OCOꢁ(þNaþ)
CH3OCOꢁ(þNaþ) þ CH3OH
! HCOOCH3 þ CH3Oꢁ(þNaþ)
HCOOCH3 þ 2H2 ! 2CH3OH
ð1-bÞ
ð2Þ
20 mL of the resin10 was added to a solvent media composed
of methanol (0.40 mol) and tri(ethylene glycol) dimethyl ether
(triglyme) then subjected to the reaction. The reaction was car-
ried out in a magnetically stirred autoclave (234 mL) with a
batch operation. The reactant gas (CO/Ar = 4/1) was pressur-
ized to 5.0 MPa. Then, the temperature was increased to 373–
393 K. The extent of methanol carbonylation over TSA1200
reached about 80% of the equilibrium in 15 min (before reaching
the set temperature) and then the equilibrium was attained in
40 min at 373–393 K. The CO was added again to the same
batch. The initial reaction (CO decrease) rates were much the
same for the 2 runs for TSA1200, suggesting that the resin
was not deactivated at 373–393 K. On the other hand, the activity
of A26 declined in several runs even at 373 K.
Sodium methoxide catalyzes the first step (Eqs 1-a and 1-b are
the mechanistic details), and metals (Ni or Cu) catalyzes the sec-
ond. CH3ONa–Raney Ni dissolved and suspended in methanol
solution work even at 423 K under 5 MPa of syngas with the
CO conversion of 95% and methanol selectivity of 99% in
4 h.4 Excepting small amount of methyl formate, no other by-
product was formed.4 These systems are being to be used for
low-temperature liquid phase methanol synthesis in the future,
too. However, the level of sodium methoxide in the reactor un-
dergoes a continuous decrease because it is soluble in methanol.
Thus, the separation process of catalyst (CH3ONa–metal) from
methanol is inevitable. Conventional Cu-Zn catalysts can work
according to Eqs 1 and 2 (but not Eqs 1-a and 1-b) using an al-
cohol-based co-catalyst under temperatures as low as 443 K.
However, the activity is not high enough.5
Next, we studied the hydrogenation of HCOOCH3 (Eq 2) by
Copyright Ó 2004 The Chemical Society of Japan

Chemistry Letters Vol.33, No.10 (2004)
1253
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
120
100
80
60
40
20
0
5.5
5.0
4.5
4.0
3.5
3.0
0
50
100
150
200
250
0
200
400
600
800
1000 1200
Time /min
Time /min
Figure 1. Time course of pressure of syngas (H2/CO/Ar = 8/4/1)
during methanol synthesis using Raney Cu (1.0 g ( ), 2.0 g ( ), or
4.0 g ( )) and TSA1200 (20 mL) in triglyme (70 mL) and methanol
(406 mmol). Temperature was raised at 5 Kꢃminꢁ1 and kept constant
at 393 K.
Figure 2. Time course of temperature (dotted line) and pressure (sol-
id line) of syngas (H2/CO/Ar = 8/4/1) using Raney Cu + TSA1200
(2.0 g + 20 mL) in triglyme and methanol (406 mmol). After 4-h run,
the syngas was added to 5.0 MPa and the next run started. Consecutive
three runs were recorded.
using Ni and Cu (both in Raney form11). Hydrogenation of meth-
yl formate (13 mL, 97.5%) was carried out over 2.0 g of Ni (or
Cu) in triglyme (87 mL) solution under 5.0 MPa of H2/Ar
(=8/1) in the same autoclave for 15 h. A comparable amount
(24 mmol) of CH4 was produced in addition to methanol
(29 mmol) for the case of Ni at 423 K. On the other hand, hydro-
genation of HCOOCH3 over Cu gave CH3OH exclusively;
106 mmol at 423 K for 3 h, 39 mmol at 393 K for 5 h. indicating
Raney Cu was active and selective for Eq 2.
Now the resin and Raney Cu were added together and meth-
anol synthesis under 5.0 MPa of syngas (CO/H2/Ar = 4/8/1)
was tested using the same autoclave. The results obtained at
393 K for different amount of Cu are shown in Figure 1. The
gas pressure drops rapidly for the first 15 min before reaching
the set temperature of 393 K, independent from the Cu amount.
In this period, the methanol carbonylation (Eq 1) proceeds over
the resin. The slow pressure drop during the initial period de-
pended on the Cu amount. Here, the hydrogenation of
HCOOCH3 (Eq 2) occurs. After the 4 h-run with 2.0 g of Cu,
72% of CO was reacted to produce 33 mmol of methanol and
6.9 mmol of methyl formate.
15%). Under the same reaction condition Raney Ni-TSA1200
exchanged with CH3Oꢁ gave no methanol but only methyl
formate from the syngas. Probably Cu (and Cuþ) activates H2
heterolytically (to Hþ + Hꢁ), while Ni does it in homolytic
ꢂ
ꢂ
way (to H + H ).
From these experiments the followings were concluded. 1)
Methanol was catalytically produced with TSA1200/Cu at
373–393 K with an appreciable rate. 2) The second step (Eq 2)
was slower (rate-determining step) than the first (Eq 1). Howev-
er, the hydrogenation rate can be accelerated by increasing the
amount of Cu or by increasing the reaction temperature. (At
present the temperature limit of the resin is 373–393 K.) Thus,
the amount of intermediate HCOOCH3 can be controlled. 3)
This is the first example of active solid catalyst for low-temper-
ature liquid-phase methanol synthesis from syngas at 373–
393 K.
This work was supported by Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports,
Science and Technology (11450305 and 16656247) and CREST,
Japan Science and Technology Agency.
The same type of experiments were conducted for 3 times at
373 K. The results are shown in Figure 2. Here also, the first
pressure drop is due to CO consumption (Eq 1) and the second
due to H2 reaction (Eq 2). Since the second reaction is slow,
some methyl formate is left unconverted and thus the H2/CO ra-
tio is higher than 2 at the end of every run. However, syngas with
the ratio of 2 (H2/CO) was again added at the beginning of 2nd
and 3rd runs. Thus, the consumption of CO was smaller
(41 mmol) in the 2nd run and much smaller (26 mmol) in the
3rd run than in the 1st run (57 mmol), due to the equilibrium lim-
itations (methyl formate is left unconverted).
The total hydrogen consumption (126 mmol) after 3 runs
was estimated from the pressure drop and gas analysis. The half
of this amount (63 mmol) must be equal to produced methanol
(Eqs 1 and 2). The measured value of 59 mmol corresponded
to the above mentioned value. The left methyl formate after third
run was 71 mmol. The sum of the two products (130 mmol) cor-
responded to CO consumption (124 mmol). About 5% discrep-
ancy of carbon mass balance may be brought by the vague meth-
anol analysis due to low methanol increment (59/406 mmol =
References and Notes
1
Z. Liu, J. W. Tierney, Y. T. Shah, and I. Wender, Fuel Process. Technol.,
23, 149 (1989).
2
V. M. Palekar, H. Jung, J. W. Tierney, and I. Wender, Appl. Catal., A,
102, 13 (1993); V. M. Palekar, J. W. Tierney, and I. Wender, Appl. Catal.,
A, 103, 105 (1993).
3
G. H. Graaf, J. G. M. Winkelman, E. J. Stamhuis, and A. A. C. M.
Beenackers, Chem. Eng. Sci., 43, 2161 (1988).
4
5
E. S. Lee and K. Aika, J. Mol. Catal. A: Chem., 141, 241 (1999).
P. Reubroycharoen, T. Vitidsant, Y. Yoneyama, and N. Tsubaki, Catal.
Today, 89, 447 (2004).
6
7
8
9
D. L. Smathers, U. S. Patent, 4100360 (1978); Chem. Abstr., 89, 196987g
(1978).
M. DiGirolamo, M. Lami, M. Marchionna, D. Sanfilippo, M. Andreoni,
A. M. R. Galletti, and G. Sbrana, Catal. Lett., 38, 127 (1996).
C. Carlin, M. DiGirolamo, M. Marchionna, A. M. R. Galletti, and G.
Sbrana, Stud. Surf. Sci. Catal., 119, 491 (1998).
H. Kubota, Petrotech, 27, 507 (2004).
10 The resin was treated with HCl aq solution, NaOH aq solution, and
CH3ONa (1 molꢃLꢁ1) methanol solution. Clꢁ in the resin was confirmed
to be exchanged almost completely with methoxide by SEM-EDX.
11 Raney Ni or Cu were prepared by leaching 4.0 g of an Al–Ni (50/50 wt %)
or Al–Cu (50/50 wt %) alloy in 50 mL of 5 N NaOH at 353 K. The surface
area of the Raney Ni (Cu) was 84 (92) m2ꢃgꢁ1
.
Published on the web (Advance View) August 28, 2004; DOI 10.1246/cl.2004.1252