Topics in Catalysis

, Volume 56, Issue 9, pp 725–729

Liquid-Phase Low-Temperature and Low-Pressure Methanol Synthesis Catalyzed by a Raney Copper-Alkoxide System

Authors

  • Bo Li
    • Faculty of TechnologyTelemark University College
    • Faculty of TechnologyTelemark University College
Original Paper

DOI: 10.1007/s11244-013-0031-4

Cite this article as:
Li, B. & Jens, K.J. Top Catal (2013) 56: 725. doi:10.1007/s11244-013-0031-4

Abstract

Current CuZnO methanol synthesis technology may be improved by development of a catalyst allowing a “once through” methanol synthesis step. A Raney copper/alkoxide catalyst has been reported in the patent literature operating at 50 bar and 120 °C. We report here the effect of temperature, pressure, Raney copper concentration and CH3OK concentration by orthogonal array design on this catalyst system. A synergistic effect on catalyst activity and methanol formation has been observed between CH3OK and Raney copper.

Keywords

Very low-temperature methanol synthesisRaney copperCH3OKOrthogonal array design

1 Introduction

Methanol has been considered as a potential fuel, as a convenient energy-storage molecule or as a feedstock to synthesize hydrocarbons [1]. Current CuZnO methanol synthesis technology is highly optimized but capital intensive. Reduction of capital cost is thus one of the important items for improvement. If a very low temperature catalyst system could be identified allowing a “once-through” methanol synthesis, an adiabatic reforming step utilizing air could be part of the methanol synthesis process [2].

A route for synthesis gas conversion to methanol under mild conditions via a methyl formate (MF) intermediate was first proposed in 1919 by Christiansen [3]. This route has received continuous attention over the years and catalyst systems can be classified into two types: copper based catalysts (typically CuCr2O4)/alkali alkoxide [4] and nickel based catalysts (typically Ni2+–NaH–alcohol)/alkali alkoxide) [5]. The reaction route can be divided into two steps:
$${\text{ROH }} + {\text{ CO}}\, \leftrightarrows \,{\text{HCOOR}} $$
(1)
$${\text{HCOOR }} + {\text{ 2H}}_{ 2} \, \leftrightarrows \,{\text{CH}}_{ 3} {\text{OH }} + {\text{ ROH}} $$
(2)
$$2 {\text{H}}_{ 2} + {\text{ CO}}\; \leftrightarrows \;{\text{CH}}_{ 3} {\text{OH}} $$
(3)

For nickel based catalysts, the mechanism of the reaction is thought to involve carbonylation of methanol to MF [6] catalysed by KOCH3, while [HNi(CO)3] catalyses the rate determining step of MF reduction to two molecules of methanol [7]. Apart from being a highly toxic catalyst system, the catalyst deactivates in the presence of CO2 and H2O. However, a comparison of the two catalyst types shows the nickel system to give particularly high methanol productivity [8].

Raney copper based conventional methanol catalysts are known [9] and have shown methanol synthesis productivities close to that of industrial catalysts. A patent [10] reports a composite Raney-Cu/CH3OK catalyst to be active for methanol synthesis at 50 bar and 120 °C. This contribution investigates the effect of different reaction variables on this catalyst system.

2 Experimental

2.1 Catalyst and Solvent Preparation

All the chemicals including the alloy were purchased from Sigma-Aldrich and used as received unless otherwise noted. The syn-gas (CO: 33.3 mol % ± 2 %, H2: 66.7 mol % ± 2 %) was purchased from Yara Praxair AS. Raney copper was prepared by the following procedure: 20 g of Devarda’s alloy (Cu: 50 %, Al: 45 %, Zn: 5 %) was added into 100 ml of water under a blanket of N2 gas followed by slow addition of NaOH pellets until the solution stopped boiling. The slurry was stirred for over 1 h at 55 °C followed by an additional 1 h at 45 °C. Subsequently the solid was washed with water until the pH of the supernatant solution reached 7. The water was filtered off and the solid was washed 3 times with dry THF followed by drying under vacuum for 30 min at room temperature. The dried Raney copper was stored under N2 at room temperature and was used without further purification. All catalytic experiments reported used the same Raney copper catalyst batch. The THF solvent was dried by distillation over potassium metal before use.

2.2 Catalyst Characterization

BET surface area of catalysts was measured by single-point measurement under atmospheric pressure of 30 % N2 in helium at liquid nitrogen temperature (77 K) on a Micromeritics Flowsorb III-2310 Surface Area Analyzer. The Raney copper sample used for the BET determination was from the same catalyst batch as was used for the catalytic reactions. The BET sample (0.163 g) was weighed in air. The BET surface area was determined to be 30.5 m2/g.

2.3 Reaction Procedures

The reaction was carried out in batch operation in a 70 ml stainless steel autoclave (PREMEX Reactor AG, Switzerland) equipped with a dip tube for sampling, an internal thermocouple, a stirrer operating at a fixed speed of 230 rpm, a spring operated safety valve and a pressure transmitter connected to a PC. The electric stirrer was connected into the reactor by a magnetic coupling and had oblique impeller blades extending into the reaction liquid reaching near the bottom of the reactor. The reactor was placed into an oil heated block; the temperature of the oil was controlled by a thermostat (Huber, ministat 230). The temperature inside the reactor was independently logged by a PC using the internal thermocouple.

For all experiments the reaction suspension was stirred at a fixed speed of 230 rpm. A test run not containing catalyst was performed as a “blank-test” in order to determine the background pressure drop due to synthesis gas solubility in the reaction solvent and also to identify possible mass transfer limitations in the experimental set-up. The initial pressure drop of a typical blank test curve as shown in Fig. 1 indicates that the system does not seem to be mass transfer limited.
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-013-0031-4/MediaObjects/11244_2013_31_Fig1_HTML.gif
Fig. 1

Typical temperature and pressure change during the course of a catalytic run. (a Blank test pressure. b Reaction pressure. c Temperature)

Predetermined amounts of CH3OK and Raney copper catalyst together with 30 g dry THF were successively filled into the reactor under N2 protection. The syn-gas was introduced into the autoclave at room temperature. The system was purged with syn-gas and pressurized to the designated pressure. Thereafter the temperature was increased to the desired reaction level. In all runs the reaction was stopped after 4 h. The reaction time was chosen to cover all the phases (see Fig. 1) of the catalytic reaction. The range of operating conditions is given in Table 1.
Table 1

Range of operating conditions

Temperature

80–120 °C

Total pressure

10–20 bar

Reactant gas ratio (H2/CO)

2.0

Stirrer speed

230 rpm

Since potassium alkoxide was employed, trace amounts of CO2 and H2O could deactivate the catalyst. Although the Raney copper catalyst has been dried in the synthesis step, there could be trace amounts of residual water left in the pores of the catalyst. For this reason, we carried out a background test to identify possible methanol formation due to reaction between residual water and potassium alkoxide. The background test was carried out in the same way as the catalytic reaction (see above), but replacing the syn-gas with 1 atm N2. Such background test was performed for each catalytic reaction and used as correction factor in the methanol formation calculation.

Solubility of the syn-gas in the solvent under reaction conditions will influence the pressure drop observed during reaction. Therefore we performed a blank test before each catalytic experiment. The blank test was performed in the same manner as the catalytic reaction without adding catalyst. The result of the blank test was used as a correction factor in the syn-gas conversion calculation.

After each experiment a 1 g sample of the liquid phase was taken using the dip tube. 0.05 g heptane was added into each sample as an internal GC standard. The samples were manually (not using an auto-sampler) injected into a gas chromatograph (Thermo Electron S.p. A, GC Focus Series) equipped with a FID detector at 250 °C. A 60 m, 0.32 mm I.D., 1.2 μ film thickness, 007 series CARBOWAX 20 M column and a programmed temperature profile were employed for product analysis. The initial temperature of the oven was 40 °C with a hold time of 0.5 min, thereafter the oven was ramped 10 °C/min up to the final temperature of 200 °C.

The formation of methanol was calculated by subtracting the quantity of methanol of the background test from the quantity of methanol measured after the catalytic reaction. The pressure drop of syn-gas was defined as the difference between the pressure before and after the catalytic reaction corrected for the blank test at room temperature. A typical example of the recorded temperature and pressure curves during the course of a catalytic run is shown in Fig. 1.

2.4 Orthogonal Array Design

A L9(34) orthogonal array design was used to optimize the experimental conditions for obtaining high methanol formation and syn-gas conversion. The design involved four factors at three levels as shown in Table 2. The order of the experiments in Table 3 was obtained by inserting the parameters shown in Table 2 into a L9(34) orthogonal array table [11]. The data analysis includes identification of the individual influence of each factor.
Table 2

Experimental parameters and their levels in L9(34)

Parameters

Levels

1

2

3

Temperature (°C)

80

100

120

Syngas pressure (bar)

10

15

20

Dosage of CH3OK (g)

0.5

1.0

1.5

Dosage of Raney copper (g)

0.2

0.5

1.0

Table 3

Experiment order for the L9(34) orthogonal array design

No.

T (°C)

P (bar)

CH3OK (g)

Raney copper (g)

1

80

10

0.5

0.2

2

80

15

1.0

0.5

3

80

20

1.5

1.0

4

100

10

1.0

1.0

5

100

15

1.5

0.2

6

100

20

0.5

0.5

7

120

10

1.5

0.5

8

120

15

0.5

1.0

9

120

20

1.0

0.2

3 Results and Discussion

3.1 Catalytic Test Run and Product Identification

Figure 1 shows the temperature and pressure curve of a typical test run. Curve b shows the pressure change during the catalytic reaction while curve a shows the pressure change during the blank test not containing catalyst. Assuming the solubility of the synthesis gas in the reaction mixture not to be affected by addition of the catalyst, comparing curve a and b it appears that the course of the catalytic run can be divided into 3 distinct periods. In the first period (0–25 min), it is unexpected to observe the lower gas absorption rate in the catalytic run (curve b) as compared to the blank test (curve a). It appears that a gaseous reaction product may be formed. In the second period (25–50 min), the relatively constant pressure of curve b indicates that the actual catalyst could be formed in situ during this period. In the final period (after 50 min), the continuous pressure drop in curve b indicates that the catalytic reaction proper is proceeding.

Based on thermodynamic considerations [12], CH3OCH3 and water could be possible reaction side products in the liquid phase. In our experiments CH3OCH3 was not detected in the liquid phase. In addition to THF solvent and the heptane internal standard two compounds were identified in the liquid sample, MF and methanol (see Fig. 2). MF is expected to be an intermediate in the synthesis reaction. Table 4 shows how product and MF formation varies according to the different reaction conditions.
https://static-content.springer.com/image/art%3A10.1007%2Fs11244-013-0031-4/MediaObjects/11244_2013_31_Fig2_HTML.gif
Fig. 2

Typical GC chromatogram of a reaction product sample containing heptane (internal standard), THF reaction solvent, methyl formate (MF)- and methanol product

Table 4

Experimental results of the L9(34) orthogonal array design

No.

T (°C)

P (bar)

CH3OK (g)

Raney copper (g)

Methanol (mg/g)

Methyl formate (mg/g)

Syn-gas conversion (%)

1

80

10

0.5

0.2

1.97

1.38

8.09

2

80

15

1.0

0.5

4.74

2.04

26.92

3

80

20

1.5

1.0

9.20

0.84

49.67

4

100

10

1.0

1.0

7.33

ND

48.91

5

100

15

1.5

0.2

2.70

3.03

22.63

6

100

20

0.5

0.5

0.51

5.55

16.40

7

120

10

1.5

0.5

5.52

ND

37.27

8

120

15

0.5

1.0

2.34

5.19

22.47

9

120

20

1.0

0.2

6.00

2.07

34.14

ND signifies that the material was not detected

3.2 Direct Observation Analysis of the Orthogonal Array Design

The impact of parameters on the formation of each product can be statistically treated using direct observation analysis to evaluate the effect or the importance of a given factor. In the direct observation analysis of the orthogonal array design, the average formation (Y) of each product at each level of a factor was obtained from the data listed in Table 4. The average product formation and syn-gas conversion (Y) at each level of a factor and range is shown in Table 5.
Table 5

The average product formation and syn-gas conversion (Y) at each level of a factor and their ranges

No.

T (°C)

P (bar)

CH3OK (g)

Raney copper (g)

Methanol

 Y1

5.30 (80 °C)

4.94 (10 bar)

1.61 (0.5 g)

3.56 (0.2 g)

 Y2

3.51 (100 °C)

3.26 (15 bar)

6.02 (1.0 g)

3.59 (0.5 g)

 Y3

4.62 (120 °C)

5.24 (20 bar)

5.81 (1.5 g)

6.29 (1.0 g)

Range

1.79

1.98

4.41

2.73

Methyl formate

 Y1

1.42 (80 °C)

0.46 (10 bar)

4.04 (0.5 g)

2.16 (0.2 g)

 Y2

2.86 (100 °C)

3.42 (15 bar)

1.37 (1.0 g)

2.53 (0.5 g)

 Y3

2.42 (120 °C)

2.82 (20 bar)

1.29 (1.5 g)

2.01 (1.0 g)

Range

1.44

2.96

2.75

0.52

Syn-gas conversion

 Y1

28.22 (80 °C)

31.42 (10 bar)

15.65 (0.5 g)

21.62 (0.2 g)

 Y2

29.31 (100 °C)

24.00 (15 bar)

36.66 (1.0 g)

26.86 (0.5 g)

 Y3

31.29 (120 °C)

33.40 (20 bar)

36.52 (1.5 g)

40.35 (1.0 g)

Range

3.07

9.40

21.00

18.73

The value of each factor is calculated as the mean value of the amount of each product formed or conversion of syn-gas at each level. For example, the value of Y1 (5.30) for methanol formation is thus the average of the value obtained from the experiments 1, 2 and 3 at 80 °C reaction temperature. The mean value of each level of each factor shows how the product formation or syn-gas conversion will change when changing the levels of that factor. In all cases each product formation or syn-gas conversion reported is the average of three experiments in which the parameter of interest (temperature, pressure…) was kept constant while the other parameters were changed (see Table 3).

3.3 Effect of Different Reaction Parameters

In this orthogonal array design, we focus on three components of the reaction output (methanol formation, MF formation, syn-gas conversion). The value of the range of each factor on each component is defined by the maximum difference of the Y value among the three levels. For each individual component, the factor with a larger range value gives a stronger effect on this specific component. Based on the range values shown in Table 5 the following conclusions can be drawn:
  1. (1).

    For methanol formation, CH3OK concentration is the most important factor, while Raney copper concentration also shows a relatively strong effect.

     
  2. (2).

    For MF formation, the reaction pressure and CH3OK concentration are important.

     
  3. (3).

    For syn-gas conversion or catalyst activity, CH3OK concentration and Raney copper concentration are important.

     

From the reactions (1)–(3), one would expect MF formation to be dependent on CH3OK concentration and CO pressure, which was also shown by our experiments. For methanol formation, the strong influence of CH3OK dosage is unexpected while the strong influence of Raney copper concentration is to be expected.

The individual influence of each factor on each product formation and syn-gas conversion can be summarized as follows:

3.3.1 Reaction Temperature

The effect of the reaction temperature on the catalyst activity was investigated by varying temperature from 80 to 120 °C as shown in the temperature column of Table 5. The range values of the temperature factor for each component of experiment output are all relatively small as compared to the other factors, which means that under our experimental conditions the influence of the temperature change on catalytic activity is not significant as compared with the change due to the other parameters.

3.3.2 Reaction Pressure

The effect of the reaction pressure on catalyst activity was investigated by varying the pressure from 10 to 20 bar as shown in the pressure column of Table 5. As discussed above, the effect of the reaction pressure is significant for MF formation. With an increase of syn-gas pressure, formation of MF shows an increasing trend.

3.3.3 CH3OK Concentration

The effect of CH3OK dosage on the catalyst activity is shown in the CH3OK column of Table 5. The relatively large range value of the CH3OK factor for each component of experiment output shows that this factor is very important for all of them. Increased concentration of CH3OK induces an increase of the methanol formation and the catalyst activity, However, it is at the same time retarding MF formation.

3.3.4 Raney Copper Concentration

The effect of Raney copper dosage on the catalyst activity is shown in the Raney copper column of Table 5. The Raney copper concentration factor is important for methanol formation and the syn-gas conversion. As the concentration of Raney copper increases, methanol formation and the syn-gas conversion increase. This result is expected since Raney copper is known to be a catalyst for hydrogenation of MF to methanol [4].

4 Conclusions

The result of the orthogonal array design shows that the activity of the Raney copper/CH3OK system is, at our reaction conditions, influenced most by CH3OK and Raney copper concentration. However, while the CH3OK concentration gives a positive effect on the methanol formation it is at the same time retarding the MF formation. This implies a synergetic effect between CH3OK and Raney copper for the hydrogenation step which is stronger than the effect of CH3OK concentration on MF formation. This synergetic effect between Raney copper and CH3OK may be connected to the observation that the actual catalyst may be formed in situ in the starting period of the catalytic reaction.

Copyright information

© Springer Science+Business Media New York 2013