Catalysis Letters Vol. 79, Nos.

1±4, April 2002 (# 2002) 129

Low-temperature methanol synthesis (LTMS) in liquid phase on

novel copper-based catalysts
Wei Chu a;b *, Tao Zhang c , Chuanhua He a , and Yutang Wu a
a
Chengdu Institute of Organic Chemistry, P.O. Box 415, Chinese Academy of Sciences, Sichuan 610041, China
b
Department of Chemical Engineering, Sichuan University, Chengdu 610065, Sichuan, China
c
Dalian Institute of Chemical Physics, P.O. Box 110, Chinese Academy of Sciences, Liaoning 116023, China

Received 27 August 2001; accepted 14 December 2001

Novel Cu-Cr based catalysts with alumina promotion have been prepared and studied in low-temperature methanol synthesis (LTMS) in
a slurry reactor. The new method of co-precipitation from the sol state gave better results than those of three other preparation methods.
From the results of temperature programmed desorption and XRD, there were more active sites in the presence of alumina.
KEY WORDS: methanol synthesis; low temperature; LTMS; catalyst; Cu-Cr-Al; co-precipitation from sol state; slurry; TPD characterization.

1. Introduction 2. Experimental

In the chemical industries of coal and natural gas, 2.1. Preparation of catalysts
methanol is a very useful and important product and
raw material [1,2]; the world's annual production of Three methods have been used for catalyst preparation:
methanol is 30  106 tons/year. In the ICI industrial co-precipitation, impregnation, and co-precipitation from
process, the pressure was 40±200 atm and temperature the sol state [12,15]. For the impregnation method, using
220±280 8C. In recent years, a new process has been the alumina powder, a solution of nitrate of copper
reported with signi®cant improvements in the low-tem- (C ˆ 0.501 mole/l) and nitrate of chromium (C ˆ
perature methanol synthesis in the slurry phase (LTMS) 0.496 mole/l) was impregnated; then the precursor was
[3±14]. There were several advantages: higher conversion calcined by thermal decomposition. The copper content
at single-pass (more than 90%, while conversion in the was 36% in the catalyst precursor. For the co-precipi-
ICI process was only 20%), higher selectivity, and lower tation, a mixed solution of nitrate of copper, nitrate of
reaction temperature (100±150 8C against 230±280 8C), chromium and nitrate of aluminium was reacted with the
lower operation cost, etc. Typical results [4±12] have precipitation agent NH3 :H2 O; then it was ®ltered, dried
been presented on the nickel catalysts, on the copper and calcined. There were the same components of equal
catalysts, and also by Wu et al. using the Cu-Cr based molar content in the catalyst. For the co-precipitation
catalysts. The reaction conditions are: T ˆ 90±150 8C, from the sol state [12], two initial liquids were prepared
P ˆ 20±60 atm, the addition of CH3 ONa is imperative (A, B); using an aqueous solution of Al(NO3 )3 :9H2 O
and promoting. The syngas conversion was more than (AR), the Al(OH)3 sol (A) was prepared by adding the
90%, the CH3 OH ‡ MF selectivity was higher than ammonia (NH3 :H2 O). A solution of Cu(NO3 )2 was
95%. The catalysts developed by Wu et al. were operated mixed with the concentrated ammonia water solution,
at 110 8C and 50 atm; the yield was 97.8 g lÿ1 hÿ1 . then a Cu(NH3 )2‡ 4 complex solution of dark-blue colour
The purposes of the present research consist of the was obtained; then a solution of (NH4 )2 Cr2 O7 was added
preparation and characterization of novel supported with good agitation, and the liquid (B) was prepared. The
Cu-Cr based catalysts in order to improve the catalytic two liquids (A, B) were mixed together and kept at 60 8C
performances in low-temperature methanol synthesis. for 1 h; the pH value was adjusted to 6.0 with the HNO3
The alumina presence promoted signi®cantly the activity solution; the precipitation was obtained under continuous
and selectivity. agitation. Then it was aged, ®ltered, washed, dried, ground
and calcined to obtain the catalyst samples.

2.2. Test of catalytic reaction of LTMS in liquid phase

* To whom correspondence should be addressed. The activity of the catalyst was evaluated in a 500 ml
E-mail: [email protected]; [email protected] stainless steel autoclave reactor in slurry phase with a

1011-372X/02/0400-0129/0 # 2002 Plenum Publishing Corporation

130 W. Chu et al. / New copper catalysts for low-temperature methanol synthesis

magnetic stirrer [12±14]. The reaction proceeded at

115 8C and 5.0±6.0 MPa. The solvent xylene (170 ml)
was dehydrated with a fresh 5 AÊ molecular sieve before
reaction, and a catalyst sample of 6.0 g was used; 30 ml
CH3 ONa solution (C ˆ 4.995 mole/l) in methanol was
added as the co-catalyst (the molar concentration of
CH3 ONa in liquid reactant was 0.749 mole/l), the feed
gas composition was H2 : CO ˆ 2.3 :1; the volume of the
total reactant-mixture system was 200 ml. At 115 8C,
the system was introduced to a pressure of 6.0 MPa
and the initial time was recorded; when the pressure of
the system decreased to 5.0 MPa, the time was also
noted, and the feed gas was re-introduced quickly to a Figure 1. E€ect of the preparation method on catalytic performances:
T ˆ 115 8C, P ˆ 5.5 MPa, H2 : CO ˆ 2.3 : 1, Cat: Cu : Cr : Al ˆ 1 : 1 : 0.5 (in
pressure of 6.0 MPa. The above processes were repeated
mole). 1A: impregnation; 2B: co-precipitation; 3C: co-precipitation from
many times, and the reaction was continued for several the sol state.
hours. The products in liquid phase of this batch reactor
were analyzed by gas chromatograph (GC103) with a
TCD detector. The catalytic activity was represented higher temperature of 750 8C; high catalytic performance
by the space time yield (STY) of methanol. was obtained. In this work, three preparation techniques
have been used, and the novel method of co-precipitation
2.3. Catalyst characterization (TPR, TPD) [13±16] from the sol state is presented here. The performances are
shown in ®gure 1; the activity and methanol selectivity of
The apparatus for TPR experiments was homemade. the catalyst from the co-precipitation method were
The catalyst sample (50 mg) was placed in a quartz slightly higher than those of the catalyst from the impreg-
tube, pretreated in situ under pure Ar ¯ow, then satu- nation method; results of the catalyst issued from the
rated with H2 ‡ N2 (4.3%) mixture. The temperature novel method (co-precipitation from the sol state) are
was raised to 600 8C at a rate of 12 8C/min, while the signi®cantly better. The reason could be the di€erent
TPR diagram was recorded on line by a thermal con- surface dispersion of active centers and the di€erence of
ductivity detector. their particle sizes.
For the temperature programmed desorption (TPD) During the preparation procedure, the temperature
test, a 200 mg sample was charged; pre-reduced using and the pH value of precipitation, the ageing, and the
H2 ‡ N2 mixture (4.3%); then it was ¯ushed by pure calcinations were also important for preparing a good
Ar gas and cooled to room temperature. The probe gas catalyst. In ®gure 2, the in¯uence of pH value on activity
(H2 or CO) was introduced for adsorption for 45 min; is shown. When pH value was changed from 5.4 to 7.0,
then ¯ushed by Ar gas (more than 30 min) until the base- there was a signi®cant evolution of activity; the pH
line was stabilized; the temperature increased to 600 8C value should be controlled at 6.0 for the best activity.
at 12 8C/min, and the signal was measured. From the color of the catalyst precursor at di€erent
pH value, some general in¯uence could be predicted;
when the catalyst precursor was dark red, the activity
3. Results and discussion was quite good.

3.1. E€ect of preparation method and parameters

The catalyst activity and structure could be very

di€erent while using di€erent preparation methods and
processes. The sol-gel method and plasma techniques
were very useful for preparing e€ective catalytic materials
[17,18]. In our previously reported work [18], the plasma
technique was used for the preparation of the highly-
active Ni/Al2 O3 catalysts for the direct oxidation of
methanol to form synthesis gas. The precursors were pre-
pared by the conventional impregnation method. The
plasma is a fourth state of material, other than in its
solid, liquid, or gas state. The precursor was charged
and treated in a plasma ®eld activated from the discharge
under high voltage for 30 min; the structure of this sample Figure 2. In¯uence of pH value of precipitation on STY of methanol.
is very di€erent from that treated conventionally at the Conditions: as in ®gure 1 legend.
 W. Chu et al. / New copper catalysts for low-temperature methanol synthesis 131

3.2. E€ect of alumina promotion Table 1

Results of temperature programmed desorption (TPD-H2 )

In the same reaction conditions, the activity of Cu-Cr

Catalyst sample Peak Peak Height Width Peak
catalyst (Cu :Cr ˆ 1 :1, in mole) was compared to that of number temperature (cm) (cm) at area
Cu-Cr-Al catalyst (Cu : Cr :Al ˆ 1 :1 :0.5, in mole). It (8C) mid-height (a.u.)
was observed that the activity of Cu-Cr-Al was signif-
icantly better than that of Cu-Cr catalyst. The STY of Cu : Cr ˆ 1 : 1 1 76 1.9 1.9 3.6
2 170 1.2 1.6 1.9
the Cu-Cr-Al catalyst increased to 172.5 g lÿ1 hÿ1 and
the methanol selectivity was 96%, while the STY of the Cu : Cr : Al ˆ 1 : 1 : 0.5 1 105 5.1 3.3 16.8
Cu-Cr catalyst was just 117.5 g lÿ1 hÿ1 and the methanol
selectivity was only 79.5%. This indicated that the
addition of alumina support could enhance the e€ective however, much better stability was obtained in the con-
dispersion of active components in the catalyst, and the tinuously-¯owing slurry reactor.
number of active sites on the surface increased; the
hydrolysis activity of the catalyst was thus improved 3.3. Temperature programmed characterization (TPR,
and resulted in the augmentation of catalytic activity TPD)
and methanol selectivity. Compared with commercial
methanol catalysts (e.g. Cu-Zn-Al), the turnover rates For understanding the bene®cial e€ect of alumina,
of the new catalysts were higher. several techniques such as XRD, TPR, and TPD-H2
The evolution of activity versus reaction cycle-times is were used for the catalyst characterization. From XRD
illustrated in ®gure 3. In the initial stage, the primary diagrams, there were both CuO and CuCrO2 peaks for
activity of the Cu-Cr catalyst was very close to that of the Cu-Cr catalyst; however there were only CuCrO2
Cu-Cr-Al. With the time going on or while the reaction for the Cu-Cr-Al catalyst, no longer the CuO peak.
cycle-times increased, the activity of the Cu-Cr catalyst These data suggest that the alumina addition promoted
decreased quite rapidly; however, the activity of the the copper dispersion and stabilized the CuCrO2 active
Cu-Cr-Al catalyst decreased quite smoothly and slowly. phase.
This indicated that the addition of alumina stabilized The results of TPR were obtained and discussed. With
the catalyst active components [19] and improved the the un-promoted Cu-Cr catalyst, there was only one
catalyst stability. From refs. [4,5], initial activities were reduction peak at 285 8C. For the Cu-Cr-Al catalyst
usually reported; the evolution with time on stream with promotion, there was also one peak, shifted to a
was hardly discussed for the new process of low- lower temperature at 240 8C; the peak area was only
temperature methanol synthesis (LTMS). In our pre- 46.2 a.u., compared with 108.8 a.u. for that of the Cu-
vious work, we studied the deactivation of the catalyst Cr catalyst. Besides a lesser amount of copper in the
system and the regeneration; the change of CH3 ONa catalyst, an important reason could be the interaction
and the coverage of copper catalyst were the principal and better dispersion of the copper component in the
factors. From the modi®cation of catalyst design and Cu-Cr-Al catalyst (in contrast, there was a large
the change of reactor design, the stability could be amount of CuO crystal phase in the Cu-Cr catalyst).
enhanced. Fast scanning of the activities of new catalysts The results of TPD-H2 are reported in table 1. There
could be obtained from the results in a batch reactor; were two peaks for the Cu-Cr catalyst at 76 8C and
170 8C; there was only one bigger peak for the Cu-Cr-
Al catalyst at 105 8C; the area of the latter (Cu-Cr-Al)
was three times the sum of the two peaks of the former
(Cu-Cr); the spillover e€ect of H2 [20±22] could be an
explanation.

4. Conclusions

For the Cu-Cr catalyst, the alumina promotion was

observed, both in the initial activity and the selectivity.
The co-precipitation from the sol state was an ecient
method for preparing highly active Cu-Cr-Al novel
catalysts for LTMS. A good yield of 172 g lÿ1 hÿ1
CH3 OH has been obtained at 115 8C, 5.5 MPa.
TPR, XRD, and TPD-H2 were used for catalyst
Figure 3. E€ect of reaction cycle-times on catalytic performance. Condi- characterization, and promoting e€ects were illustrated.
tions: as in ®gure 1 legend. In the presence of alumina, the copper component was
132 W. Chu et al. / New copper catalysts for low-temperature methanol synthesis

better dispersed; there were more active sites on the [6] V.M. Palekar, H. Jung, J.W. Tierney et al., Appl. Catal. A. 102 (1993)
surface. 13.
[7] S. Oyama, Appl. Catal. A. 180 (1999) 217.
[8] W. Tierney, I. Wender and V.M. Palekar, US patent 5221652 (1993);
5385949 (1995).
Acknowledgment [9] Yutang Wu, S.Z. Luo et al., Petroleum Chem. Eng. (Chin.) 22(1)
(1993) 10.
This work was supported by NNSFC (No. 29903011) [10] Yutang Wu, W.K. Chen, S.Z. Luo et al., CN 1136 979A, 1996.
[11] X.Q. Liu, Yutang Wu, Z.L. Yu et al., Stud. Surf. Sci. Catal. 119 (1998)
and by the Ministry of Sciences and Technology of 557.
China (a subdivision of the ``973'' key-project). Useful [12] Wei Chu, C.H. He, Yutang Wu et al., Appl. CN 001 127 27.6, 2000.
discussions with Dr. R. Ohnishi, Prof. Dr. Can Li, Mr. [13] Wei Chu, Yutang Wu, Liwu Lin et al., Chemical Progress 13(2), (2001)
Kuiling Wang, Mr. Shizhong Luo, and other colleagues 128.
are also very gratefully acknowledged. [14] X. Q. Liu, Yutang Wu et al., Chin. J. Catal. (Cuihua Xuebao) 20
(1999) 81.
[15] W. Chu, Q.G. Yan, Z.L. Yu et al., Stud. Surf. Sci. Catal. 119 (1998)
849.
References [16] W. Chu, R. Kie€er, A. Kiennemann and J.P. Hindermann, Appl.
Catal. A. 121 (1995) 95.
[1] K.C. Waugh, Catal. Today, 15 (1992) 51. [17] Yuhong Zhang, Guoxing Xiong, Yuan Kou et al., React. Kinet. Catal.
[2] K. Klier, Adv. Catal. 31 (1983) 243. Lett. 69(2) (2000) 325.
[3] (a) M. Marchionna, M. Lami and F. Ancillotti, European patent [18] Yong Zhang, W. Chu, C.R. Luo, K. Zhou et al., Plasma Chem.
0375071 (1990). (b) A.T. Bell et al., Catalysis Looks to the Future Plasma Process, 20(1) (2000), 137.
(US National Academy Press, Washington, 1992). [19] M. Saito, T. Fujitani et al., Appl. Catal. A. 138 (1996) 311.
[4] M. Marchionna, M. Lami and A.M. Galletti, Chemtech, April (1997) [20] K.D. Jung and A.T. Bell, J. Catal. 193(2) (2000) 207.
27. [21] T. Shishido and H. Hattori, Appl. Catal. A 146 (1996) 157.
[5] V.M. Palekar, J.W. Tierney and I. Wender, Appl. Catal. A. 103 (1993) [22] T. Do, M. Che and L. Bonneviot, Stud. Surf. Sci. Catal. 77 (1993) 125.
105.