processes

Article
Methanol Synthesis: Optimal Solution for a Better
Efficiency of the Process
Grazia Leonzio ID

Department of Industrial and Information Engineering and Economics, University of L’Aquila,

Via Giovanni Gronchi 18, 67100 L’Aquila, Italy; grazia.leonzio@graduate.univaq.it; Tel.: +39-3401537804

Received: 31 January 2018; Accepted: 16 February 2018; Published: 25 February 2018

Abstract: In this research, an ANOVA analysis and a response surface methodology are applied
to analyze the equilibrium of methanol reaction from pure carbon dioxide and hydrogen. In the
ANOVA analysis, carbon monoxide composition in the feed, reaction temperature, recycle and
water removal through a zeolite membrane are the analyzed factors. Carbon conversion, methanol
yield, methanol productivity and methanol selectivity are the analyzed responses. Results show that
main factors have the same effect on responses and a common significant interaction is not present.
Carbon monoxide composition and water removal have a positive effect, while temperature and recycle
have a negative effect on the system. From central composite design, an optimal solution is found in
order to overcome thermodynamic limit: the reactor works with a membrane at lower temperature with
carbon monoxide composition in the feed equal to 10 mol % and without recycle. In these conditions,
carbon conversion, methanol yield, methanol selectivity, and methanol production are, respectively,
higher than 60%, higher than 60%, between 90% and 95% and higher than 0.15 mol/h when considering
a feed flow rate of 1 mol/h. A comparison with a traditional reactor is also developed: the membrane
reactor ensures to have a carbon conversion higher of the 29% and a methanol yield higher of the 34%.
Future researches should evaluate an economic analysis about the optimal solution.

Keywords: methanol reaction; equilibrium reactions; ANOVA analysis; response surface

methodology; optimization

1. Introduction
Methanol production via CO2 hydrogenation is an important representative among the chemical
conversions of CO2 and offers challenging opportunities for sustainable development.
As a raw material, methanol constitutes the basis for the production of hundreds of
chemicals, such as formaldehyde, methyl tertbutyl ether (MTBE), acetic acid, methyl methacrylate,
dimethyl terephthalate, and olefins (ethylene and propylene), all basic chemical building blocks
for a number of common products [1]. Methanol can also be used as transportation fuel, as a fuel
cell hydrogen carrier, as well as in wastewater treatment or in electricity production. It is then an
excellent fuel and a key starting material of important industrial reactions [2]. In the recent years, also,
methanol is suggested as alternative for chemical energy carrier [3,4].
Generally, in industrial applications, methanol can be produced from syngas, a mixture of CO,
CO2 and H2 at 50–100 bar and 250 ◦ C–300 ◦ C, using copper and zinc-based catalyst [5,6]. These catalysts
are already active at 200 ◦ C and selective towards the formation of H2 and CO2 . Alternative routes
and its use as fuel and chemical are the core of methanol economy [4,7]. Pontzen et al. [8]
experimentally show that methanol can be produced from CO2 and H2 using conventional catalysts as
CuO/Al2 O3 /ZnO. However, studies show that the equilibrium yield of methanol from CO2 at 200 ◦ C
and 50 bar is slightly less than 40%, whereas the yield from a mixture of CO2 and CO at the same
conditions is greater than 80% [9].

Processes 2018, 6, 20; doi:10.3390/pr6030020 www.mdpi.com/journal/processes

Processes 2018, 6, 20 2 of 20

The production of methanol from pure CO2 and H2 has also some thermodynamic limits, as shown
by Zachopoulos and Heracleous [10], suggesting the removal of water through sorbents. In fact,
water inhibits the reaction rate of methanol formation via CO2 hydrogenation [11]. In particular,
water produced from reverse water gas shift greatly reduces the methanol synthesis rate by suppressing
the reaction of methanol production.
Works related to these considerations are few in literature. In fact, works about the production
of methanol by syngas are mainly present and different strategies are proposed to improve the
efficiency of the process. Mainly, studies involve the in situ methanol removal. Examples are methanol
adsorption on fine alumina powder or the use of a solvent, such as tetraethylene glycol, n-butanol, or
n-hexane [12]. Westerterp et al. [13] suggest the selective adsorption of water and methanol on a solid,
in a trickle bed reactor. An alternative method is based on the in situ separation of reaction products
by condensing them on the surface of a condenser inside the reactor, close to the catalyst bed [14].
These solutions have some disadvantages due to the introduction of other chemicals, complicated
operations and low space time yields. An alternative to overcome chemical equilibrium is the in-situ
condensation of methanol without additional adsorbents or coolers. Van Bennekom et al. [15] suggest
the condensation of methanol in situ at high pressure (200 bar) and low temperature (200 ◦ C). It is
evident that thermodynamic barrier can be eliminated by removal of reaction products based on the
principle of Le Chatelier. Gallucci and Basile [16] suggest the recycle of unconverted synthesis gas
after product separation by condensation or the water removal using selective, permeable membranes.
The first membrane reactor for methanol production is used by Struis et al. [17] using lithiated Nafion
membrane at 200 ◦ C and 4.3 bar. In another work of Struis and Stucki [18] a modeling of membrane
reactor for methanol reaction is developed, using a Nafion membrane: a single pass yield is improved
by 40%. However, this application is limited because the operative temperature of Nafion membrane
is lower than 200 ◦ C. Chen and Yuan [19] analyze a one-dimensional isothermal pseudo-homogeneous
model for membrane reactor producing methanol. A silicone rubber/ceramic composite membrane
is used. Results show that conversion in membrane reactor is increased by 22% when compared to
traditional fixed bed reactor. Also, Barbieri et al. [20] use a ceramic membrane for a reactor producing
methanol. Gallucci et al. [21] analyze a zeolite membrane reactor: at the same conditions, conversions
are higher in membrane reactor respect to traditional reactor. In other words, the same conversions can
be obtained in membrane reactor with lower temperature and pressure compared to the traditional
reactor. This aspect should notably reduce the energy demand. A mathematical model for zeolite
membrane reactor permeable to water is developed by Gallucci and Basile [16]. Results shows that it
is possible to have s higher conversion and methanol selectivity with respect to a traditional reactor at
the same operating experimental conditions.
When considering the analyzed works reported in literature where syngas is used in the feed,
the feasibility of methanol synthesis from CO2 can be achieved by circumventing the thermodynamic
limits through innovative solutions: the presence of CO, the operation at lower temperature, the recycle
of unconverted gas, the presence of membrane permeable to water inside the reactor. In fact,
a mixture of CO–CO2 allows for having a higher conversion and yield, as shown by Kunkes and
Behrens [22]. Several studies show that a maximum in methanol production from a mixture of CO–CO2
occurs at CO2 concentration of 2–5 mol % of total carbon [23–25]. Lower temperatures are preferred
by thermodynamic even if kinetic is not favored and a more active catalyst should be developed.
The recycle of unconverted gas to ensure a higher conversion is also suggested by Montebelli et al. [26].
However, this results in high investment and operating costs and large pressure drops. The extraction
of water, an inhibitor for reaction, through membrane shifts the equilibrium toward to product,
but requires higher capital investment [27].
In this research, an analysis of chemical equilibrium for methanol production by pure CO2 and
H2 is carried out: the above proposed solutions, as the presence of CO in the feed considering that
the maximum methanol production is obtained with 5 mol % of CO2 , lower operating temperature,
the recycle of unconverted gas, and the presence of zeolite membrane permeable only to water,
Processes 2018, 6, 20 3 of 20

are analyzed. An ANOVA analysis and a response surface methodology (RSM) through a central
composite design (CCD) are developed to find the optimal solution that overcome thermodynamic
limits ensuring a higher carbon conversion, methanol yield, methanol production and methanol
selectivity. Data for these analyses are obtained by simulations of methanol reactor in Aspen Plus.

2. Materials and Method

2.1. ANOVA Modeling

For the methanol reactor, the estimation of main and interaction effects is developed by ANOVA
analysis (analysis of variance); in this analysis, it is determined if effects and interactions among
investigated factors are significant respect to experimental error (σε ). Main factors are evaluated
by Yates’s algorithm through Excel 2016 software. Statistical significance is checked by F-value
(Fischer variation ratio) and p-value (significant probability value). Model terms are selected or rejected
based on probability value within 95% of confidence interval (or 5% significance level). In this research,
σε is evaluated by means of the mean square (MS) of interactions that are not significant. A 24 full
factorial design with 16 simulation tests is performed for this research [28]. A mathematical model
could be obtained with significant factors and the quality of the model is assessed by the coefficient of
determination, R2 . R2 represents a pure correlation between measured and predicted values, and it is
indicative of response variation explained by model. Then, in this statistic method all factor levels are
fully changed so that it is possible to measure any variation in response. One of the most important
advantages of this method is the limited number of experiments necessary to identify the best solution.
CO composition in the feed, reaction temperature, the recycle of unconverted gases, and the removal
of water through a zeolite membrane are considered as factors in this research. Carbon conversion,
methanol yield, methanol production and methanol selectivity are considered as responses, for which
mathematical models are developed.

2.2. Response Surface Methodology

The aim of RSM is to find the optimal operating conditions of the process or to determine a region
that satisfies the operating specifications, maintaining a reduced number of experiments [29]. RMS is
then used in process design and optimization.
The RSM methodology is developed using a central composite design for fitting a second order
model. Generally, the CCD consists of 2k factorial tests, 2k stars tests and nc center point tests, where k
is the number of studied factors in the experiment. Values at center points are used to estimate the
curvature of surface plot. Star points are located at distance α from center point and are used to
estimate the coefficient of quadratic terms. Factorial points are used to estimate the coefficient of linear
terms and two-way interactions.
Experimental or simulation data are used to develop a second-order polynomial model as in the
following correlation (see Equation (1)) [30]:

k k k k
Y = β o + ∑ β i · Ai + ∑ ∑ β ij · Ai · A j + ∑ β ii · A2i + ε (1)
i =1 i =1 j =1 i =1

where βo is intercept, βi , βij , βii are first-order, interactive, and second-order effects, respectively, i and
j represent the number of k factors, while ε is residual error. This method is then able to evaluate
interaction effects, pure quadratic effects, or third- and fourth-order effects and so on; it is the most
efficient evaluation method, and it improves the quality of data. To determine the significance of
each term in the equation and to estimate the goodness or fitting quality, the polynomial equation is
validated by analysis of variance. A response surface plot is constructed by using the fitted model.
In this research, with four factors, star points are set on the centers of each face of factorial face:
the value of α is equal to 1 and the design is denominated as face-centered central composite design.
Processes 2018, 6, 20 4 of 20

Processes 2018, 6, 21  4 of 20 
It requires only three levels of each factor, and in practice it is frequently difficult to change factor
simulations in Aspen Plus are developed: 16 factorial tests, eight star tests, and six central tests, and 
levels. 31 simulations in Aspen Plus are developed: 16 factorial tests, eight star tests, and six central
one replication test. Software Minitab 18 is used to carry out the CCD analysis for response surface 
tests, and one replication test. Software Minitab 18 is used to carry out the CCD analysis for response
methodology. 
surface methodology.

2.3. Reactor Modeling 
2.3. Reactor Modeling
In order to carry out ANOVA and RSM analysis a modeling of methanol reactor is developed in 
In order to carry out ANOVA and RSM analysis a modeling of methanol reactor is developed in
Aspen Plus
Aspen Plus software,
software,  as  shown 
as shown in  Figure 
in Figure 1. It is1. 
anIt  is  an  equilibrium 
equilibrium reactor 
reactor where where 
reactions reactions 
and and 
equilibrium
equilibrium constants are defined and it is fed with CO  and H , at flow rate equal to 1 mol/h, as 
constants are defined and it is fed with CO2 and H2 , at flow rate equal to 1 mol/h, as shown in Table 1.
2 2

shown in Table 1. Carbon dioxide and hydrogen are in stochiometric conditions. 
Carbon dioxide and hydrogen are in stochiometric conditions.  

 
Figure 1. Equilibrium reactor for methanol production in Aspen Plus environmental. 
Figure 1. Equilibrium reactor for methanol production in Aspen Plus environmental.

Table 1. Material and energy balances of methanol reactor obtained in Aspen Plus environmental. 
Table 1. Material and energy balances of methanol reactor obtained in Aspen Plus environmental.

Factor Feed Product

Factor Feed Product
Temperature (K)  493  493 
Temperature (K) 493 493
Pressure (bar)  55  55 
Pressure (bar) 55 55
Vapor Vapor fraction 
fraction 1 1  1  1
Total flow rate (mol/h) 
Total flow rate (mol/h) 1 1  0.766 0.766
CO2 flowCOrate (mol/h)
2 flow rate (mol/h)  0.250.25  0.165 0.165
H2 flow rate (mol/h) 0.75
H2 flow rate (mol/h)  0.75  0.51 0.51
H2 O flow rate (mol/h) 0 0.084
H
CO flow2O flow rate (mol/h) 
rate (mol/h) 0 0  0.084 
0.0069
CO flow rate (mol/h)  0  0.0069 
The reactor is characterized according the following reactions (see Equations (2) and (3)):
The reactor is characterized according the following reactions (see Equations (2) and (3)): 
CO2 + 3H2 ↔ CH3 OH + H2 O∆H298K = −49.43 kJ/mol (2)
CO 3H ↔ CH OH H O ∆ 49.43 kJ/mol  (2) 

CO H 2 +↔H CO
CO 2 ↔ COH
+HO 2 O ∆H298K
∆ = +41.12 41.12
kJ/molkJ/mol  (3)
(3) 
The equilibrium constants of two reactions are respectively the following
The equilibrium constants of two reactions are respectively the following (see Equations (4) and 
(see Equations
(5)) [24]:  (4) and (5)) [24]:
7060
ln(Keq1 ) = 7060− 24.389 (4)
ln T 24.389  (4) 
4773
ln(1/Keq2 ) = − + 4.672 (5)
4773
T
ln 1/ 4.672 
First reaction is exothermic so equilibrium constant decreases with temperature; second reaction
(5) 

is endothermic so equilibrium constant increases with temperature. As shown in [24], initially CO2
First reaction is exothermic so equilibrium constant decreases with temperature; second reaction 
reacts to give CO as well as methanol. Since the reverse water gas shift is endothermic, and because
is endothermic so equilibrium constant increases with temperature. As shown in [24], initially CO2 
of decreasing CO2 concentration, this slows down the hydrogenation of CO2 at the reactor inlet.
reacts to give CO as well as methanol. Since the reverse water gas shift is endothermic, and because 
The reverse water gas shift reaches its equilibrium value and switches direction. This results in an
of decreasing CO2 concentration, this slows down the hydrogenation of CO2 at the reactor inlet. The 
inflection point in temperature profile and a concentration evolution of water and methanol. From this
reverse  water  gas  shift  reaches  its  equilibrium  value  and  switches  direction.  This  results  in  an 
point onward, reactions are strictly in series.
inflection point in temperature profile and a concentration evolution of water and methanol. From 
this point onward, reactions are strictly in series.   
A  membrane  reactor  is  modelled  using  six  stages  in  series  of  reactors  and  water  separators, 
because the aim of this research is to analyze the equilibrium of methanol reaction by removing water. 
Processes 2018, 6, 20 5 of 20

A membrane reactor is modelled using six stages in series of reactors and water separators,
because the aim of this research is to analyze the equilibrium of methanol reaction by removing water.
In fact, membrane is considered to be both not permeable and poorly permeable to other compounds,
due to the condensation of water in membrane pores. With these considerations, the presence of
membrane inside the reactor is evaluated as percentage of water removal by reaction site to permeation
site, as is also suggested by Gallucci and Basile [16].
SRK thermodynamic model is used to carry out simulations in Aspen Plus, because it is the better
thermodynamic model for methanol system as found by Cheng et al. [31].

3. Results and Discussion

3.1. Results of ANOVA Analysis

Table 2 shows factors (CO concentration in the feed, reaction temperature, the recycle of
unconverted gas, water removal through membrane) and the values of their levels chosen for the
factorial plant.

Table 2. Factors and values of their levels chosen in the factorial design of ANOVA analysis.

Level
Code Factor
(−) (+)
A CO concentration in the feed (mol %) 0 20
B Temperature (◦ C) 200 280
C Recycle (%) 0 90
D Water removal (%) 0 80

Carbon conversion to methanol, methanol yield, methanol production, and methanol selectivity
are the analyzed responses. Carbon conversion, methanol yield and methanol selectivity are defined,
respectively, as in the following relations (see Equations (6)–(8))

(CO2,in + COin ) − (CO2,out + COout )

Conversion (%) = (6)
(CO2,in + COin )

MeOHout
MeOH yield (%) = (7)
CO2,in + COin
MeOHout
MeOH selectivity (%) = (8)
(CO2,in + COin ) − (CO2,out + COout )
Figure 2 shows the results of ANOVA analysis with significant factors and interactions,
while Table 3 reports the general result of ANOVA analysis. As shown in Figure 2 carbon conversion
in %, methanol production in mol/h, methanol yield in %, methanol selectivity in % are the
chosen responses.
Processes 2018, 6, 20 6 of 20
Processes 2018, 6, 21 6 of 20

40

20

0
A B C D AC BC BD ABC ABD
-20

Effects
-40

-60

-80

-100
Carbon conversion (%) Methanol yield (%)
Methanol production (mol/h)·[1000] Methanol selectivity (%)
Factors

Figure 2. Results of ANOVA analysis considering only significant factors and interactions (A = CO concentration in the feed in mol %, B = Temperature in °C, C =
Figure 2. Results of ANOVA analysis considering only significant factors and interactions (A = CO concentration in the feed in mol %, B = Temperature in ◦ C,
Recycle in %, D = Water removal in %) with values of their effects.
C = Recycle in %, D = Water removal in %) with values of their effects.
Table 3. Results of ANOVA analysis (A = CO concentration in the feed in mol %, B = Temperature in °C, C = Recycle in %, D = Water removal in %).
Table 3. Results of ANOVA analysis (A = CO concentration in the feed in mol %, B = Temperature in ◦ C, C = Recycle in %, D = Water removal in %).
Carbon Conversion (%) Methanol Yield (%) Methanol Production (mol/h) Methanol Selectivity (%)
Factors
Effects F-Ratio p-Value Remarks Effects F-Ratio p-Value Remarks Effects F-Ratio p-Value Remarks Effects F-Ratio p-Value Remarks
Carbon Conversion (%) Methanol Yield (%) Methanol Production (mol/h) Methanol Selectivity (%)
A 12.84
Factors 10.02 0.051 95% 10.40 15.92 0.016 98% 1.99 17.12 0.014 99% 18.12 71.14 0.001 100%
B −32.54 Effects
64.32 0.001 p-Value
F-Ratio 100% Remarks
−30.07 F-Ratio p-Value
Effects 133.03 0.000 Remarks
100% Effects 190.85p-Value0.000
16.63 F-Ratio Remarks100% F-Ratio p-Value
Effects −18.01 70.33 Remarks
0.001 100%
AB −4.71
A 1.35
12.84 0.310
10.02 69% 95% 1.14 10.40
0.051 0.19
15.92 0.684
0.016 32%
98% 1.990.02 17.120.32 0.014 0.60399% 40%
18.12 3.56
71.14 2.74
0.001 0.173
100% 83%
C −21.13
B −27.12
32.54 0.006
64.32 99% 100%−26.85
0.001 −30.07 106.08
133.03 0.001
0.000 100% 16.63
100% 13.26 190.85
54.08 0.000 0.002100% 100%
−18.01 −29.71
70.33 191.31
0.001 0.000
100% 100%
AC −12.80
AB −
9.95
4.71 0.034
1.35 97% 69% −15.251.14
0.310 34.22
0.19 0.004
0.684 100%
32% 0.024.28 0.3238.28 0.603 0.00340% 100% 3.56 −21.17
2.74 97.12
0.173 0.001
83% 100%
BC 9.34C −5.30
21.13 27.12
0.083 0.006
92% 99% 11.78−26.85 20.44
106.08 0.001
0.011 100%
99% 13.262.55 54.089.19 0.002 0.039100% 96% −29.71 191.31
5.74 0.000
7.14 100%
0.056 94%
ABC AC
4.77 −1.38
12.80 9.95
0.305 0.034
69% 97% 10.79−15.25 17.1434.22 0.004
0.014 100%
99% 4.282.14 38.2819.78 0.003 0.011100% 99%−21.17 97.12
6.03 0.001
7.88 100%
0.048 95%
D BC
21.83 9.34
28.96 5.30
0.006 0.083
99% 92% 15.1011.78 33.55 20.44 0.011
0.004 99%
100% 2.554.19 9.19 9.43 0.039 0.03796% 5.74
96% 7.14
14.89 0.056
48.06 94%
0.002 100%
AD ABC
−3.93 4.77
0.94 1.38
0.387 0.305
61% 69% −6.3210.79 17.14
5.87 0.014
0.072 99%
93% 2.140.73 19.786.37 0.011 0.06599% 93%6.03 7.88
−3.90 0.048
3.30 95%
0.144 86%
D 21.83 28.96 0.006 99% 15.10 33.55 0.004 100% 4.19 9.43 0.037 96% 14.89 48.06 0.002 100%
BD −11.87 8.56 0.043 96% −8.99 11.89 0.026 97% 1.49 4.30 0.107 89% −6.62 9.51 0.037 96%
AD −3.93 0.94 0.387 61% −6.32 5.87 0.072 93% 0.73 6.37 0.065 93% −3.90 3.30 0.144 86%
ABD 1.89
BD −0.22
11.87
0.665
8.56
33%
0.043 96%
7.94
−8.99 9.28 11.89
0.038
0.026
96%
97% 1.49
1.16
4.30
11.24
0.107
0.028
89%
97%
−6.62 6.70
9.51
9.73
0.037
0.036
96%
96%
CD 10.67
ABD 6.92
1.89 0.058
0.22 94% 33% 3.95 7.94
0.665 2.30
9.28 0.204
0.038 80%
96% 1.160.29 11.242.01 0.028 0.22997% 77%6.70 1.15
9.73 0.29
0.036 0.621
96% 38%
ACD 3.75
CD 0.85
10.67 0.408
6.92 59% 94% 1.53 3.95
0.058 0.35
2.30 0.588
0.204 41%
80% 0.290.04 2.01 0.40 0.229 0.55977% 44%1.15 0.90
0.29 0.17
0.621 0.698
38% 30%
BCD −6.19
ACD 2.33
3.75 0.202
0.85 80% 59% −3.44 1.53
0.408 1.74
0.35 0.258
0.588 74%
41% 0.040.22 0.40 0.03 0.559 0.88044% 12%0.90 −2.75
0.17 1.64
0.698 0.269
30% 73%
ABCD −1.91
BCD −
0.22
6.19 0.663
2.33 34% 80% 4.09−3.44 2.46
0.202 1.74 0.192
0.258 81%
74% 0.220.31 0.03 3.25 0.880 0.14612% −2.75
85% 2.96
1.64 1.90
0.269 0.241
73% 76%
ABCD −1.91 0.22 0.663 34% 4.09 2.46 0.192 81% 0.31 3.25 0.146 85% 2.96 1.90 0.241 76%
Processes 2018, 6, 20 7 of 20

Results show that significant factors and interactions have the same effect on the
analyzed responses.
The effect indicates how the analyzed responses vary by changing the level of factor. In other
words, the effect is the effect of an independent variable on a dependent variable averaging across the
levels of any other independent variables. If by increasing the level of factor, the analyzed response
increases the effect is positive; if by increasing the level of factor the analyzed response decreases the
effect is negative.
In this study, all the main factors are significant. CO concentration and water removal through a
membrane have a positive effect on all responses, while temperature and the recycle of unconverted
gases have a negative effect on the studied responses. In order to improve the efficiency of the process
with a higher methanol production, it is better to operate in presence of CO or with a membrane
permeable to water or with a lower temperature or without the recycle of unconverted gas, allowing
to have a reactor with lower sizes and more compact.
The positive effect of CO in the feed is reported by Van den Bussche and Froment [24], funding an
optimal CO2 composition in the feed. The positive effect of water removal is also found by Zachopoulos
and Heracleous [10], using water sorption system. The negative effect of temperature is also reported
by Skrzypek et al. [32]. The negative effect of recycle is due to the decreasing of kinetic. At fixed
number of tubes, recycle has a negative effect: with a higher flow rate, superficial velocity increases
and kinetic decreases producing a lower methanol [33]. Manenti et al. [34] suggest that conversion
increases when the inlet flowrate is lowered, thanks to the increase in residence time. In fact, a lower
volume of gas causes high residence times in the reactor and reaction is near to chemical equilibrium.
For these reasons, stages of adiabatic reactors with intermedia refrigeration are also used.
Reaction temperature has a higher effect on carbon conversion, methanol yield and methanol
production as compared to other significant factors and interactions. Recycling, instead, has a higher
effect on methanol selectivity respect to other significant factors and interactions. These results suggest
that temperature has an important role on equilibrium reaction. A higher efficiency of reaction is
obtained operating at lower temperature, so new kind of catalysts need to be developed to operate at
lower temperature.
Interaction AC has a negative effect: the effect of factor C is lower negative with factor A.
The presence of CO and the recycle of produced gas determines a lower efficiency. Interaction BC
has a positive effect on methanol production and methanol yield. It is possible to operate at a
higher temperature using the recycle of produced gas in order to have a good efficiency of the
process. However, the effect of this interaction is lower compared to the effect of main factors.
Also, this interaction is not significant for carbon conversion and methanol selectivity.
Interaction BD has a negative effect on carbon conversion, methanol yield and methanol selectivity:
the positive effect of membrane, factor D, becomes negative at a higher temperature, factor B.
Also, in the presence of membrane, temperature must to be lower. Interactions of third order ABC and
ABD have a positive effect on methanol yield, production and selectivity.
Generally, these results suggest that the use of membrane permeable to water or the presence of CO
in the feed could improve the efficiency of the process with a higher carbon conversion, methanol yield,
selectivity and production avoiding the use of recycle that increases the reactor volume and costs or
avoiding the operation at a lower temperature that decreases kinetic.
In particular, membrane has a higher effect on carbon conversion and methanol yield while the
presence of CO has a higher effect on methanol production and selectivity.
A mathematical model is developed for the analyzed responses as function of significant factors
and interactions, as the following relations (see Equations (9)–(12)):

Conversion (%) = 31.09 + 6.42· A − 16.26· B − 10.56·C − 6.39· AC + 10.91· D

(9)
−5.93· BD R2 = 0.9


Processes 2018, 6, 20 8 of 20

MeOH yield (%)

Processes 2018, 6, x FOR PEER REVIEW 8 of 20
= 27.9 + 5.2· A − 7.62· AC + 5.89· BC + 5.39· ABC + 7.5·4D (10)
−4.49· BD + 3.97· ABD − 15.03· B − 13.42·C R2 = 0.96


(%)
=
 27.9 + 5.2 ∙ − 7.62 ∙ + 5.89 ∙ + 5.39 ∙ + 7.5 ∙ 4 (10)
MeOH production − mol
4.49
h ∙ + 3.97 ∙ − 15.03 ∙ − 13.42 ∙ ( = 0.96)
= 0.08 + 0.012· A − 0.042· B − 0.022·C − 0.019· AC + 0.009· BC (11)
( ) 2
ℎ · ABC + 0.009· D + 0.010· ABD R = 0.95


+0.013 (11)
= 0.08 + 0.012 ∙ − 0.042 ∙ − 0.022 ∙ − 0.019 ∙ + 0.009 ∙
MeOH selectivity+ (0.013
%) ∙ + 0.009 ∙ + 0.010 ∙ ( = 0.95)
= 25.61
(%) + 9.05 · A − 9.00 · B − 14.85 · C − 10.58· AC + 3.01· ABC (12)
+
=7.44 · D+−9.05
25.61 3.31∙ · BD
− 9.00 ∙ ·−ABD
+ 3.34 14.85+∙ 7.44 · D ( R∙ 2 =+0.9
− 10.58 ) ∙
3.01 (12)
+ 7.44 ∙ − 3.31 ∙ + 3.34 ∙ + 7.44 ∙ ( = 0.9)
where A, B, C, D are main factors respectively CO composition in the feed, temperature, recycle of
where A, gases,
unconverted B, C, Dand are main
waterfactors respectively
removal, while BD,CO AC,
composition
BC are in
thethe feed, temperature,
interactions recycle
of second of and
order
unconverted gases, and water removal,
ABD and ABC are the interactions of third order. while BD, AC, BC are the interactions of second order and
ABD and ABC are the interactions of third order.
The trend of residues in Figure 3 shows that the obtained mathematical models are correct: a good
The trend of residues in Figure 3 shows that the obtained mathematical models are correct: a
agreement between calculated and simulation data is present. This is also confirmed by the values of
good agreement between calculated and simulation data is present. This is also confirmed by the
R2 : itvalues
is closer
of Rto unit
2: it for alltomathematical
is closer models. models.
unit for all mathematical

30
a
25

20

15
Residues (%)

10

0
0 20 40 60 80 100
-5

-10

-15

-20
Calculated carbon conversion (%)

10b
8
6
4
Residues (%)

2
0
0 20 40 60 80 100
-2
-4
-6
-8
-10
Calculated MeOH yield (%)

Figure 3. Cont.
Processes 2018, 6, 20 9 of 20
Processes 2018, 6, x FOR PEER REVIEW 9 of 20

c
0.02
Processes 2018, 6, x FOR PEER REVIEW 9 of 20
0.015
0.01c
Residues (mol/h)
0.02
0.005
0.015
0
0.010 0.05 0.1 0.15 0.2 0.25
-0.005
Residues (mol/h)

0.005
-0.01
0
-0.015 0 0.05 0.1 0.15 0.2 0.25
-0.005
-0.02
-0.01
-0.025
-0.015 Calculated MeOH production (mol/h)
-0.02
d
15 -0.025
Calculated MeOH production (mol/h)

10 d
15
5
Residues (%)

10
0
50 20 40 60 80 100
Residues (%)

-5
0
0 20 40 60 80 100
-10
-5

-15-10
Calculated MeOH selectivity (%)

-15
Figure 3. Trend of residues for (a) Calculated
carbon conversion, (b) methanol
MeOH selectivity (%) yield, (c) methanol production,
Figure 3. Trend of residues for (a) carbon conversion, (b) methanol yield, (c) methanol production,
and (d) methanol selectivity.
and (d) Figure
methanol selectivity.
3. Trend of residues for (a) carbon conversion, (b) methanol yield, (c) methanol production,
Anand
accurate analysis
(d) methanol is also developed for interaction factors with the test of two levels. Figure
selectivity.
An accurate
4 shows analysis
the analysis is also developed
of interaction AC and BD for interaction
for carbon factors
conversion. withshow
Results the that
test atofa two
higherlevels.
Figure 4 shows
level ofAnfactorthe
A, itanalysis
accurate analysis
is possibleofalso
is interaction
to developed
improve AC and BD for
for interaction
carbon conversion tocarbon
factors conversion.
with the
methanol testoperating
by Results
of two levels.
withFigureshow
factor C that
at a4 lower
at a higher shows
level the
ofanalysis
level,factor of it
A,
then without interaction
is possibleAC
recycle. and
Ato BD value
improve
higher for carbon
carbon conversion.
conversion
of conversion Results
can show that
to obtained
be methanol at afactor
by
with higher
operating
D at with
factora C level
at aoflevel
higher factor
lower A,factor
it is possible
level,
and then to improve
B at without
a lower carbonAconversion
recycle.
level. higher valueto methanol by operating
of conversion canwith
be factor C with
obtained
at a lower level, then without recycle. A higher value of conversion can be obtained with factor D at
factor D at a higher level and factor B at a lower level.
a higher level and factor B at a lower level.
a 60

a 60
50
Carbon conversion (%)

50
Carbon conversion (%)

40
40 C lower level
30
30 CClower
higher level
level
20
C higher level
20
10
10
0
-1.5 -1 -0.5 0 0 0.5 1 1.5
-1.5 -1 -0.5 0 A
Factor 0.5 1 1.5
Factor A

Figure 4. Cont.
Processes 2018, 6, 20 10 of 20

Processes 2018, 6, x FOR PEER REVIEW 10 of 20

Processes 2018, 6, x FOR PEER REVIEW 10 of 20

70
b 70
b 60
60

Carbon conversion (%)

50

Carbon conversion (%)

50
40
40 B lower level
30
BBhigher
lowerlevel
level
30
20 B higher level
20
10
100
-1.5 -1 -0.5 0 0.5 1 1.5
0
-1.5 -1 -0.5
Factor
0
D 0.5 1 1.5
FactorofDinteraction AC in carbon conversion; (b) Test of two
Figure 4. (a) Test of two levels for the analysis
Figure 4. (a) for
levels Testtheofanalysis
two levels for the BD
of interaction analysis of interaction
in carbon conversion. (AAC in carbon
= CO conversion;
composition in mol%, B (b)= Test of
Figure
two levels 4. (a) Test
for the in of two
analysis levels for the
of interactionanalysis of interaction
BD inremoval AC in
carboninconversion.carbon conversion; (b) Test
(A = CO composition in of twomol%,
temperature °C, C = recycle in %, D = water %).
levels for the analysis
◦ of interaction BD in carbon
B = temperature in C, C = recycle in %, D = water removal in %). conversion. (A = CO composition in mol%, B=
temperature in °C, the
Figure 5 shows C = test
recycle in %,
of two D = water
levels removal inAC,
for interactions %).BC, BD, that are significant for methanol
yield.
Figure 5 shows the test of two levels for interactions AC, BC, BD, that are significant for
Figure 5 shows the test of two levels for interactions AC, BC, BD, that are significant for methanol
methanol yield.
yield. a
30
a
25
30

20
MeOH yield (%)

25

15 C lower level
20
MeOH yield (%)

C higher level
10
15 C lower level
5 C higher level
10
0
-1.5 -1 -0.5 5 0 0.5 1 1.5
Factor A
0
-1.5 -1 -0.5 70 0 0.5 1 1.5
b Factor A
60
70
b 50
MeOH yield (%)

60
40
C lower level
50
30
C higher level
MeOH yield (%)

40
20
C lower level
30
10
C higher level
200
-1.5 -1 -0.5 0 0.5 1 1.5
Factor
10 B

0
-1.5 -1 -0.5 0 0.5 1 1.5
Factor B

Figure 5. Cont.
Processes 2018, 6, x FOR PEER REVIEW 11 of 20
Processes 2018, 6, 20 11 of 20
Processes 2018, 6, x FOR PEER REVIEW 60 11 of 20
c
60
50
c
50
MeOH yield (%) 40

40
30
MeOH yield (%)

B lower level
30
B higher
B lower level level
20
B higher level
20
10
10

0
-1.5 -1 -0.5 00 0.5 1 1.5
-1.5 -1 -0.5 0 0.5 1 1.5
Factor D
Factor D

Figure 5. (a)5. Test

Figure of two
(a) Test of twolevels for
levels forthe
theanalysis
analysis of interactionAC
of interaction ACinin methanol
methanol yield;
yield; (b) Test
(b) Test of twoof two
Figure 5. (a)
levelslevels
for for the analysis of interaction BC in methanol yield (c) Test of two levels for the analysis of two
the Test of
analysis twoof levels for
interaction theBCanalysis
in of
methanol interaction
yield (c)AC
Testinofmethanol
two yield;
levels for (b)
the Test of
analysis of
levels for
interaction the
BD analysis
in of
methanol interaction
yield. (A =BC
CO in methanol
composition yield
in (c)
mol%, Test
B = of two levels
temperature
interaction BD in methanol yield. (A = CO composition in mol%, B = temperature in °C, C = recycle in for
in °C,the
C =analysis
recycle of
in
interaction BD in methanol yield. (A = CO composition in mol%, B = temperature in ◦ C, C = recycle in
%, D = water removal
%, D = water removal in %). in %).
%, D = water removal in %).
WithWithfactorfactor
A atAaatlower
a lower level,a ahigher
level, highermethanol
methanol yield
yieldisisobtained
obtained with factor
with factorC atCa at
lower level.level.
a lower
With factor
However, in A at
these a lower
conditions, level, a
a lower higher
carbon methanol
conversionyield is obtained
is obtained.
However, in these conditions, a lower carbon conversion is obtained. Factor B and C at a lower with
Factor B andfactor
C atCa at a
lower lower
level level.
level
in
However, interaction BC allows to have a higher methanol yield. In interaction BD, factor B and D must to
in interaction BC allows to have a higher methanol yield. In interaction BD, factor B and D must in
in these conditions, a lower carbon conversion is obtained. Factor B and C at a lower level to
be at aBC
interaction lower and to
allows higher
have level, respectively,
a higher methanol toyield.
have aInhigher methanol
interaction BD,yield:
factor carbon
andconversion
B carbon D must to is
be at is
a
be atalso
a lower
higher. and higher level, respectively, to have a higher methanol yield: conversion
lower and
also higher. higher level, respectively, to have a higher methanol yield: carbon conversion is also higher.
Figure 6 shows the test of two levels for interactions AC and BC in methanol production.
Figure
Figure 66 shows
shows the theistest
test of two
of at
two levels for
levels for interactions
interactions AC AC andand BC BC in in methanol production.
production.
Methanol production higher lower level of factor A and C in interaction AC. methanol
Factor B and C at a
Methanol
Methanol production is higher at lower level of factor A and C in interaction AC. Factor B and
C atCa
lower production
level ensure is to higher at lower
have a higher level ofproduction.
methanol factor A and C inconditions
These interaction AC. FactorBC
in interaction B allow
and
at a lower
lowerforlevellevel
also ensure ensure to
to have atohigher have
have methanol a higher
a higher yield. methanol production. These conditions
methanol production. These conditions in interaction BC allow in interaction BC allow
for
for also
also to
to have
have aa higher
higher methanol
methanolyield.
yield.
a
0.08
a 0.07
0.08
MeoH production (mol/h)

0.06
0.07
MeoH production (mol/h)

0.05
0.06
0.04 C lower level
0.05
0.03 C higher level
0.04 C lower level
0.02
0.03 C higher level
0.01
0.02
0
-1.5 -1 -0.5 0 0.5 1 1.5
0.01
Factor A
0
-1.5 -1 -0.5 0 0.5 1 1.5
Figure 6. Cont.
Factor A

Figure 6. Cont.
Figure 6. Cont.
Processes 2018, 6, 20 12 of 20
Processes 2018, 6, x FOR PEER REVIEW 12 of 20
Processes 2018, 6, x FOR PEER REVIEW 12 of 20

0.18
b 0.18
b 0.16
(mol/h) 0.16
0.14
(mol/h)
0.14
0.12
0.12
production

0.1
production

0.1
C lower level
0.08 C lower level
0.08 C lower level
0.06 C lower level
MeOH

0.06
MeOH

0.04
0.04
0.02
0.02
0
-1.5 -1 -0.5 0 0 0.5 1 1.5
-1.5 -1 -0.5 0 B
Factor 0.5 1 1.5
Factor B
Figure 6. (a) Test of two levels for the analysis of interaction AC in methanol production; (b) Test of
Figure 6.
Figure (a) Test
6. (a) Test of two
two levels for
for the
the analysis of
of interaction AC in
in methanol
methanol production;
production; (b)
(b) Test
Test of
two levels for the ofanalysislevels
of interactionanalysis interaction
BC in methanol AC
production (A = CO composition in mol%,ofB
two
two levels for the
levels for the analysis
analysis of interaction
of interaction BC in methanol production (A = CO composition in mol%,
= temperature in °C,◦ C = recycle in %, D =BC in methanol
water removalproduction
in %). (A = CO composition in mol%, B
=B temperature
= temperature in in
°C, C,
C =Crecycle
= recycle
in in
%,%,D =Dwater
= water removal
removal in in
%).%).
Figure
Figure 77 shows
shows an an analysis
analysis of
of significant
significant interaction
interaction AC
AC and
and BDBD for
for methanol
methanol selectivity.
selectivity. As
As for
for
Figure
carbon 7 showsfor
conversion, an interaction
analysis of AC
significant
a higherinteraction
selectivityAC
is and BD for
obtained methanol
with factor A selectivity.
at a higher As for
level
carbon conversion, for interaction AC a higher selectivity is obtained with factor A at a higher level
carbon
and conversion, for interaction AC a higher selectivity is obtained with factor A at a higher level
and factor
factor CC at
at aa lower
lower level. In
level. In interaction
interaction DB,
DB, factor
factor D
D and
and B B at
at aa lower
lower level
level ensures
ensures aa higher
higher
and factor
methanol C at a lower
selectivity. level. In interaction DB, factor D and B at a lower level ensures a higher
methanol selectivity.
methanol
The selectivity.two levels suggests that a common interaction that ensures a higher efficiency
The analysis
analysis of of two levels suggests that a common interaction that ensures a higher efficiency
for The analysis of two levels suggests that a common interaction that ensures a higher efficiency
for all analyzed responses is
all analyzed responses is not
not present.
present. In
In order
order to
to optimize
optimize the
the process,
process, it
it is
is preferable
preferable to
to consider
consider
for
the all
mainanalyzed
factors,responses
in is
particularnot present.
the In
addition order
of CO to optimize
to pure CO the process, it is preferable to consider
2 in the feed or the use of membrane
the main factors, in particular the addition of CO to pure CO2 in the feed or the use of membrane
the main factors,
permeable to water.
water.in particular the addition of CO to pure CO2 in the feed or the use of membrane
permeable to
permeable to water.

a 70
a 70
60
60
(%)

50
(%)

50
selectivity

40
selectivity

40
30 C lower level
30 C lower level
C higher level
MeOH

20 C higher level
MeOH

20
10
10
0
-1.5 -1 -0.5 0 0 0.5 1 1.5
-1.5 -1 -0.5 0 0.5 1 1.5
Factor A
Factor A

Figure 7. Cont.
Figure 7. Cont.
Figure 7. Cont.
Processes 2018, 6, 20 13 of 20
Processes 2018, 6, x FOR PEER REVIEW 13 of 20

30
b
MeOH selectivity (%) 25

20

15
B lower level
B higher level
10

0
-1.5 -1 -0.5 0 0.5 1 1.5
Factor D

Figure 7. (a) Test of two levels for the analysis of interaction AC in methanol selectivity; (b) Test of
Figure 7. (a) Test of two levels for the analysis of interaction AC in methanol selectivity; (b) Test of
two levels for the analysis of interaction BD in methanol selectivity (A = CO composition in mol %, B
two levels for the analysis of interaction BD in methanol selectivity (A = CO composition in mol %,
= temperature in °C, C = recycle in %, D = water removal in %).
B = temperature in ◦ C, C = recycle in %, D = water removal in %).

3.2. Results of Response Surface Methodology

3.2. Results of Response Surface Methodology
Table 4 shows the chosen factors and the values of their levels set in a face centered composite
Table 4 shows
central design, usedthe chosen
to find thefactors and
response the values
surface of the
plot of their levels setanalyzed
responses in a faceincentered
ANOVA composite
analysis.
central design, used to find the response surface plot of the responses analyzed in ANOVA
Factors are the same used in previous ANOVA analysis: CO concentration in the feed, temperature, analysis.
Factors are the same used in previous ANOVA analysis: CO concentration in the feed,
the recycle of unconverted gas and the removal of water through a membrane. An analysis of three temperature,
the recycle
levels of unconverted
is developed for eachgas and the removal of water through a membrane. An analysis of three
factors.
levels is developed for each factors.
Table 4. Factors and values of their levels chosen in CCD analysis.
Table 4. Factors and values of their levels chosen in CCD analysis.
Level
Code Factor
(−)Level(0) (+)
Code Factor
A CO concentration in the feed (mol %) (−) 0 (0) 10 (+) 20
B A Temperature
CO concentration (°C) (mol %)
in the feed 0 20010 240 20 280
C B Recycle (%)
Temperature (◦ C) 200 0 240 45 280 90
D C Recycle
Water (%) (%)
removal 0 0 45 40 90 80
D Water removal (%) 0 40 80
Figure 8 shows the response surface plot of carbon conversion: it is possible to have a conversion
Figure 8 shows the response surface plot of carbon conversion: it is possible to have a conversion
higher than 60% using a membrane reactor at single pass with a removal of water equal to 40% and
higher than 60% using a membrane reactor at single pass with a removal of water equal to 40% and a
a CO composition in the feed equal to 10% (Figure 8D). In this case, temperature can be lower,
CO composition in the feed equal to 10% (Figure 8D). In this case, temperature can be lower, improving
improving kinetic. The same conversion can be obtained also in other cases with membrane reactor
kinetic. The same conversion can be obtained also in other cases with membrane reactor and recycle of
and recycle of gases. However, this determines a reactor with greater sizes, increasing its costs, so
gases. However, this determines a reactor with greater sizes, increasing its costs, so they are not the
they are not the better solutions. Then, among all better solutions a membrane reactor at single pass
better solutions. Then, among all better solutions a membrane reactor at single pass is chosen because
is chosen because it allows to decrease the costs. A carbon conversion higher than 60% can be
it allows to decrease the costs. A carbon conversion higher than 60% can be obtained also at lower
obtained also at lower temperature, with CO composition in the feed equal to 20%, with recycle and
temperature, with CO composition in the feed equal to 20%, with recycle and water removal equal
water removal equal to 40% in the first case (Figure 8A). In the second case, carbon conversion higher
to 40% in the first case (Figure 8A). In the second case, carbon conversion higher than 60% can be
than 60% can be obtained also in membrane reactor (water removal equal to 80%), with recycle, at
obtained also in membrane reactor (water removal equal to 80%), with recycle, at lower temperature
lower temperature and with CO composition in the feed equal to 10% (Figure 8E). Figure 8 shows
and with CO composition in the feed equal to 10% (Figure 8E). Figure 8 shows also that in many
also that in many cases a temperature◦higher than 240 °C decreases carbon conversion because kinetic
cases a temperature higher than 240 C decreases carbon conversion because kinetic is not favored.
is not favored. In particular, carbon conversions are between 10% and 30%.
In particular, carbon conversions are between 10% and 30%.
Processes 2018, 6, 20 14 of 20
Processes 2018, 6, x FOR PEER REVIEW 14 of 20

Processes 2018, 6, x FOR PEER REVIEW 14 of 20

Figure 8. Surface plots of carbon conversion obtained by CCD analysis: a) carbon conversion as
Figure 8. Surface plots of carbon conversion obtained by CCD analysis: (a) carbon conversion
function of temperature and CO composition in the feed; b) carbon conversion as function of recycle
as function of temperature and CO composition in the feed; (b) carbon conversion as function
and CO composition in the feed; c) carbon conversion as function of water removal and CO
of recycle and CO
composition
composition
in the feed;
in the
d) carbon
feed; (c)ascarbon
carbonconversion
conversionobtainedfunction
conversion
of recycle
as
and
function of water removal
Figure 8. Surface plots of by CCD analysis: a) temperature; e) carbon
carbon conversion as
and CO composition
conversion in the
as temperature
function feed; (d) carbon
of water conversion as function of recycle and temperature;
function of and COremoval and temperature;
composition in the feed; b)f)carbon
carbon conversion
conversion as function
as function of water
of recycle
(e) carbon
removalconversion
and as
recycle. function of water removal and temperature; (f) carbon
and CO composition in the feed; c) carbon conversion as function of water removal and CO
conversion as function
of watercomposition
removal and recycle.
in the feed; d) carbon conversion as function of recycle and temperature; e) carbon
Using a membrane
conversion as functionreactor
of wateratremoval
single andpass as in Figure
temperature; 8D, conversion
f) carbon methanolasyield, production,
function of water and
Using a membrane
selectivity are also
removal higher,
and recycle. reactor
as shown at in
single
Figures pass
9–11,as in FigureA 8D,
respectively. methanol
methanol yield,than
yield higher production,
60%,
a methanol production
and selectivity higher than
are also higher, 0.15 mol/h
as shown and a methanol
in Figures selectivity between
9–11, respectively. 90% and yield
A methanol 95% can higher
be Using a membrane reactor at single pass as in Figure 8D, methanol yield, production, and
obtained.
than 60%, a methanol production higher than 0.15 mol/h and a methanol selectivity between 90% and
selectivity
Figure 9are also higher,
shows as shown
that a quite highinmethanol
Figures 9–11,
yieldrespectively.
between 50% A methanol
and 60%yield can behigher than 60%,
obtained also in
95% cana be obtained.
methanol production higher than 0.15 mol/h and a methanol selectivity between 90% and 95% can is
membrane reactor with recycle (but increasing costs) in two cases. In the first case, temperature
Figure 9 shows that a quite high methanol yield between 50% and 60% can be obtained also in
be obtained.
equal to 200 °C, CO composition in the feed is equal to 20 mol %, recycle is equal to 45%, and water
membrane reactor with recycle
Figure 9 shows (buthigh
that a quite increasing
methanol costs) in two50%
yield between cases.
and In60%thecanfirst case, temperature
be obtained also in is
removal ◦is equal to 40% (Figure 9A). In the second case, temperature is 200 °C, water removal is 80%,
equalCOtomembrane
200 C, reactor
CO with recyclein(but
composition the increasing
feed is costs)to
equal in20two cases.
mol %, In
composition in the feed is equal to 10 mol % and recycle is equal to 45% (Figure 9E).
the first
recycle is case,
equal temperature
to 45%, is water
and
removalequal to 200to°C,
is equal 40%CO(Figure
composition
9A).inInthe
thefeed is equal
second to 20temperature
case, mol %, recycle
isis200
equal
◦ C,towater
45%, and water is 80%,
removal
removal is equal to 40% (Figure 9A). In the second case, temperature is 200 °C, water removal is 80%,
CO composition in the feed is equal to 10 mol % and recycle is equal to 45% (Figure 9E).
CO composition in the feed is equal to 10 mol % and recycle is equal to 45% (Figure 9E).

Figure 9. Surface plots of methanol yield obtained by CCD analysis: a) methanol yield as function of
temperature and CO composition in the feed; b) methanol yield as function of recycle and CO
Figure 9. Surface plots of methanol yield obtained by CCD analysis: a) methanol yield as function of
composition
Figure 9. Surfaceinplots
the feed; c) methanol
of methanol yield
yield as function
obtained by of
CCDwater removal(a)
analysis: and CO composition
methanol yield asinfunction
the
temperature and CO composition in the feed; b) methanol yield as function of recycle and CO
of temperature and
composition in CO composition
the feed; inyield
c) methanol the as
feed; (b) methanol
function yield and
of water removal as function of recycle
CO composition in theand CO
composition in the feed; (c) methanol yield as function of water removal and CO composition in the
feed; (d) methanol yield as function of recycle and temperature; (e) methanol yield as function of water
removal and temperature; (f) methanol yield as function of water removal and recycle.
 Processes 2018, 6, x FOR PEER REVIEW 15 of 20

Processes 2018, 6, 20
feed; d) methanol yield as function of recycle and temperature; e) methanol yield as function of water15 of 20
removal and temperature; f) methanol yield as function of water removal and recycle.

It is possible to have a methanol yield between 50% and 60% also in membrane reactor at
It is possible to have a methanol yield between 50% and 60% also in membrane reactor at single
single pass with CO composition in the feed equal to 20 mol %, temperature equal to 240 ◦ C and
pass with CO composition in the feed equal to 20 mol %, temperature equal to 240 °C and water
water removal equal to 40% (Figure 9B). However, carbon conversion is between 40% and 50%.
removal equal to 40% (Figure 9B). However, carbon conversion is between 40% and 50%. Generally,
Generally, lower methanol yields are present at lower carbon conversions where higher temperatures
lower methanol yields are present at lower carbon conversions where higher temperatures are
are present.
present.
Figure 10 shows that methanol production higher than 0.15 mol/h can also be obtained in other
Figure 10 shows that methanol production higher than 0.15 mol/h can also be obtained in other
conditions that however do not ensure a higher carbon conversion and methanol yield (Figure 10A,E,F)
conditions that however do not ensure a higher carbon conversion and methanol yield (Figure
or10A,E,F)
utilize aorreactor
utilizewith higher with
a reactor costs higher
(Figurecosts
10B,F). Results10B,F).
(Figure show that lower
Results methanol
show productions
that lower methanolare
clearly and rightly present at lower carbon conversions.
productions are clearly and rightly present at lower carbon conversions.

Figure 10. Surface plots of methanol production obtained by CCD analysis: a) methanol production
Figure 10. Surface plots of methanol production obtained by CCD analysis: (a) methanol production
as function of temperature and CO composition in the feed; b) methanol production as function of
as function of temperature and CO composition in the feed; (b) methanol production as function of
recycle and CO composition in the feed; c) methanol production as function of water removal and CO
recycle and CO composition in the feed; (c) methanol production as function of water removal and CO
composition in the feed; d) methanol production as function of recycle and temperature; e) methanol
composition in the feed; (d) methanol production as function of recycle and temperature; (e) methanol
production as function of water removal and temperature; f) methanol production as function of
production as function of water removal and temperature; (f) methanol production as function of water
water removal and recycle.
removal and recycle.

Figure11
Figure 11shows
showsthat
that aa methanol
methanol selectivity
selectivity higher
higher than
than 95%
95% can
can be
be also
also obtained
obtainedin intwo
twocases
cases
(Figure 11 C,E), but carbon conversion varies between 10% and 50%. The first case, Figure 11C, 11C,
(Figure 11 C,E), but carbon conversion varies between 10% and 50%. The first case, Figure is
is when
when CO composition in the feed is 20 mol %, water removal is 80%, recycle is 45% and temperature
CO composition in the feed is 20 mol %, water removal is 80%, recycle is 45% and temperature is 240 ◦ C.
is 240 °C. In this condition, carbon conversion is between 40% and 50%. The second case, Figure 11E,
In this condition, carbon conversion is between 40% and 50%. The second case, Figure 11E, is when,
is when, temperature is 280 °C, recycle 45%, water removal 80%, and CO composition in the feed is
temperature is 280 ◦ C, recycle 45%, water removal 80%, and CO composition in the feed is 10 mol %:
10 mol %: carbon conversion is lower and between 10% and 40%.
carbon conversion is lower and between 10% and 40%.
Processes 2018, 6, 20 16 of 20
Processes 2018, 6, x FOR PEER REVIEW 16 of 20
Processes 2018, 6, x FOR PEER REVIEW 16 of 20

Figure 11. Surface plots of methanol selectivity obtained by CCD analysis: a) methanol selectivity as
Figure
Figure 11.
11. Surface plots of methanol
methanol selectivity
selectivity obtainedby byCCD
CCDanalysis:
analysis:
a) (a) methanol selectivity
function ofSurface plots and
temperature of CO composition in obtained
the feed; b) methanol methanol
selectivity selectivity
as function as
of recycle
as function
function of of temperature
temperature and and
CO CO composition
composition in in
the the
feed; feed;
b) (b) methanol
methanol selectivity
selectivity as as
function function
of of
recycle
and CO composition in the feed; c) methanol selectivity as function of water removal and CO
recycle
and and CO
CO compositioncomposition
in d)
thein the feed; (c) methanol
feed; c) selectivity
methanol as selectivity
selectivity as function
as recycle
function of water
oftemperature;removal
water removal and
and COCO
composition in the feed; methanol function of and e) methanol
composition
composition ininthe
thefeed;
feed;(d)
d) methanol
methanol selectivity
selectivity as function
as function of recycle
of recycle and
and temperature;
temperature; e) (e)methanol
methanol
selectivity as function of water removal and temperature; f) methanol selectivity as function of water
selectivity
selectivityas
asfunction
function of water removal and temperature; (f) methanol selectivity as function of water
removal and recycle. of water removal and temperature; f) methanol selectivity as function of water
removal and recycle.
removal and recycle.
Optimalconditions
Optimal conditionsandandreactor
reactor configurations
configurations are
are then
then found
found to
to produce
produce methanol
methanol byby CO
CO22and
and
H2. These results are in agreement with results found in ANOVA analysis where main factors2 and
Optimal conditions and reactor configurations are then found to produce methanol by CO can
HH22.. These
These results are in agreement
resultsefficiency
are in agreement with results found in ANOVA analysis where main
with results found in ANOVA analysis where main factors canfactors can
ensure a higher of the process.
ensure
ensure a higher
a higher efficiency
efficiency of the
of the process.
process.
Figures 12–15 show the trend of surface plot of carbon conversion, methanol yield, methanol
Figures 12–15 show the trend of surface plot of carbon conversion, methanol yield,
yield,methanol
methanol
production and methanol selectivity insurface
Figures 12–15 show the trend of plot
the space asof carbon of
function conversion, methanol
the analyzed factors.
production and methanol selectivity in the space as function of the analyzed
production and methanol selectivity in the space as function of the analyzed factors. factors.

Figure 12. Carbon conversion plots as function of analyzed factors (hold values: recycle 45%,
Figure 12. Carbon
temperature conversion
240 °C, conversion
water removal plots
40%,as CO
function of analyzed
composition factors
in the feed (hold values:
10 mol%): recycle 45%,
Figure 12. Carbon plots as function of analyzed factors (hold a) carbon
values: conversion
recycle 45%,
temperature
as function of 240 °C,
recycle water removal
and water 40%,
removal; CO composition
b) carbon in the
conversion feed 10 mol%): a) carbon conversion
temperature 240 ◦ C, water removal 40%, CO composition in theas function
feed of temperature
10 mol%): and water
(a) carbon conversion
as functionc)ofcarbon
recycle and water removal; b) carbon conversionand as function
recycle; of temperature and water
asremoval;
function of recycle conversion as function
and water removal; of temperature
(b) carbon conversion as function d) carbon conversion
of temperature as
and water
removal;
function c) CO
of carbon conversion
composition in as feed
the function
and of temperature
water removal; and
e) recycle;
carbon d) carbon
conversion as conversion
function of as
CO
removal; (c) carbon conversion as function of temperature and recycle; (d) carbon conversion as
function
compositionof CO composition
incomposition
the feed andinin the feed
recycle; and water
f) carbon removal;
conversion e)
as(e) carbonof
function conversion as function
CO composition offeed
in the CO
function of CO the feed and water removal; carbon conversion as function of CO
composition
and temperature.in the feed and recycle; f) carbon conversion as function of CO composition in the feed
composition in the feed and recycle; (f) carbon conversion as function of CO composition in the feed
and temperature.
and temperature.
Processes 2018, 6, 20 17 of 20
Processes 2018, 6, x FOR PEER REVIEW 17 of 20
Processes 2018, 6, x FOR PEER REVIEW 17 of 20

Figure 13. Methanol yield plots as function of analyzed factors (hold values: recycle 45%, temperature
Figure 13. Methanol yield plots as function of analyzed factors (hold values: recycle 45%, temperature
Figure
240 °C,13. Methanol
water removalyield plots
40%, COascomposition
function of analyzed factors
in the feed 10 mol(hold
%): values: recycle
a) methanol 45%,astemperature
yield function of
240 ◦ C, water removal 40%, CO composition in the feed 10 mol %): (a) methanol yield as function
240 °C, water
recycle removal
and water 40%, CO
removal; composition
b) methanol in the
yield feed 10 mol
as function %): a) methanol
of temperature andyield
wateras removal;
function of
c)
of recycle and water removal; (b) methanol yield as function of temperature and water removal;
recycle
methanolandyield
wateras removal;
function b)
of methanol yieldand
temperature as function of methanol
recycle; c) temperatureyieldandaswater removal;
function of COc)
(c) methanol yield as function of temperature and recycle; (d) methanol yield as function of CO
methanol
compositionyield as feed
in the function of temperature
and water removal; e) and recycle;
methanol c) as
yield methanol
function yield
of COas function ofin CO
composition the
composition in the feed and water removal; (e) methanol yield as function of CO composition in the
composition in the
feed and recycle; f) feed and water
methanol removal;
yield as functione) of
methanol yieldand
temperature as function of CO composition
CO composition in the feed.in the
feed and recycle; (f) methanol yield as function of temperature and CO composition in the feed.
feed and recycle; f) methanol yield as function of temperature and CO composition in the feed.

Figure 14. Methanol production surface plots as function of analyzed factors (hold values: recycle
Figure 14. Methanol
45%, temperature 240production surface plots
°C, water removal 40%,as
COfunction of analyzed
composition in the factors
feed 10(hold values:
mol%): recycle
a) methanol
Figure 14. Methanol production surface plots as function of analyzed factors (hold values: recycle 45%,
45%, temperature
production 240 °C,ofwater
as ◦function removal
recycle 40%, CO
and water composition
removal; in the feed
b) methanol 10 mol%):
production as a) methanol
function of
temperature 240 C, water removal 40%, CO composition in the feed 10 mol%): (a) methanol production
production
temperatureas andfunction of recycle
water removal; and water
c) methanol removal; as
production b)function
methanol production as
of temperature andfunction
recycle; of
c)
as function of recycle and water removal; (b) methanol production as function of temperature and water
temperature and water
methanol production asremoval;
function c)
of methanol production
CO composition as function
and water of temperature
removal; e) methanoland recycle; as
production c)
removal; (c) methanol production as function of temperature and recycle; (d) methanol production
methanol
function ofproduction as function
CO composition of CO
in the composition
feed andf) water
and recycle; removal;
methanol e) methanol
production production
as function of COas
as function of CO composition and water removal; (e) methanol production as function of CO
function of CO
composition composition
in the in the feed and recycle; f) methanol production as function of CO
feed and temperature.
composition in the feed and recycle; (f) methanol production as function of CO composition in the feed
composition in the feed and temperature.
and temperature.
 45%, temperature 240 °C, water removal 40%, CO composition in the feed 10 mol%): a) methanol
production as function of recycle and water removal; b) methanol production as function of
temperature and water removal; c) methanol production as function of temperature and recycle; c)
methanol production as function of CO composition and water removal; e) methanol production as
function
Processes of CO composition in the feed and recycle; f) methanol production as function of CO18 of 20
2018, 6, 20
composition in the feed and temperature.

Figure 15. Methanol selectivity surface plots as function of analyzed factors (hold values: recycle 45%,
temperature 240 ◦ C, water removal 40%, CO composition in the feed 10 mol%): (a) methanol selectivity
as function of recycle and water removal; (b) methanol selectivity as function of temperature and water
removal; (c) methanol selectivity as function of temperature and recycle; (d) methanol selectivity as
function of CO composition in the feed and water removal; (e) methanol selectivity as function of
CO composition in the feed and recycle; (f) methanol selectivity as function of temperature and CO
composition in the feed.

The quadratic expressions of these plots are the following (see Equations (13)–(16)):

Conversion (%) = 32.55 + 5.37· A − 14.82· B − 9.26·C − 6.22· AC + 10.86· D

(13)
+8.65· BC + 4.36·CD

MeOH yield (%) = 32.97 + 7.63· A − 3.7· AC + 11.71· BC + 8.36· D − 3.7· AC

(14)
−5.2· AB − 12.12· B − 11.98·C
 
MeOH production mol h
= 0.114 + 0.017· A − 0.023· B − 0.019· AC + 0.022· D (15)
+0.021· BD + 0.038· BC
MeOH selectivity (%) = 77 − 0.8·C + 8.42· BD (16)

where A, B, C, D are main factors respectively CO composition in the feed, temperature, recycle and
water removal, while BD, AC, BC are the interactions of second order.
A comparison between a membrane reactor with the optimal operating conditions found through
the previous response surface methodology and a traditional reactor is developed. The traditional
reactor is at 200 ◦ C, 40 bar and the feed in amount of 1 mol/h contains CO2 and H2 in stochiometric
conditions. Results are reported in Table 5. The advantages of membrane reactor are evident,
allowing for a higher conversion, methanol production, and yield. In particular, carbon conversion
and methanol yield are respectively increased by 29% and 34%. In addition, methanol production and
hydrogen conversion are higher in membrane reactor, respectively, equal to 0.15 mol/h and 49%.

Table 5. Comparison between membrane reactor and traditional reactor for methanol production.

Factor Membrane Reactor Traditional Reactor

Carbon conversion (%) 62 33
Hydrogen conversion (%) 49 32
Methanol yield (%) 61 27
Methanol production
0.15 0.076
(mol/h)
Processes 2018, 6, 20 19 of 20

4. Conclusions
The main obstacle to methanol synthesis from CO2 rich streams is thermodynamic. For pure CO,
a one pass methanol yield of nearly 55% can be obtained at 252 ◦ C, while pure CO2 would only yield 18%.
In this research, different strategies to overcome thermodynamic limits in methanol production
from pure CO2 and H2 are suggested and analyzed. The addition of CO to the feed, the operating
at lower temperature, the recycle of unconverted gas, and the use of membrane removing water
inside the reactor can be applied to have higher efficiencies in methanol production. An ANOVA
analysis and a response surface methodology are carried out with the aim to find the best solution.
In these analyses, chosen factors are: CO composition in the feed, operating temperature, the recycle of
unconverted gases, and the presence of zeolite membrane permeable to water. The analyzed responses
are: carbon conversion, methanol production, methanol yield and methanol selectivity. Results suggest
that a higher carbon conversion, methanol yield, selectivity and production are obtained from a
mixture of CO2 and H2 , adding 10 mol % of CO in the feed, operating at 200 ◦ C, and using a membrane
reactor without recycle. In these conditions, when considering a feed flow rate of 1 mol/h carbon
conversion, methanol yield, selectivity, and production are respectively higher than 60%, higher than
60%, 0.15 mol/h, and between 90% and 95%. A tradition reactor, operating at 200 ◦ C, with hydrogen
and carbon dioxide in stochiometric conditions allows for having a conversion only equal to 33% and
a methanol yield of 27%: the advantages of optimized systems are evident. In addition, the traditional
system produces only 0.076 mol/h of methanol, while membrane reactor can produce 0.15 mol/h of
methanol with a feed flow rate of 1 mol/h. The feasibility of methanol synthesis from carbon dioxide
can be achieved by circumventing the thermodynamic limitations, through innovative reactor design.
Future research should evaluate economic considerations about the optimal solution.

Acknowledgments: The author of the study would like to thank the University of L’Aquila for funding this work.
Conflicts of Interest: The author declares no conflict of interest.

References
1. Kung, H.H.; Cheng, W.H. Methanol Production and Use; M. Dekker: New York, NY, USA, 1994.
2. Bansode, A.; Urakawa, A. Towards full one-pass conversion of carbon dioxide to methanol and
methanol-derived products. J. Catal. 2014, 309, 66–70. [CrossRef]
3. Olah, G.A. After Oil and gas: Methanol economy. Catal. Lett. 2004, 93, 1–2. [CrossRef]
4. Olah, G.A.; Goeppert, A.; Prakash, G.K.S. Chemical Recycling of Carbon Dioxide to Methanol and
Dimethyl Ether: From Greenhouse Gas to Renewable, Environmentally Carbon Neutral Fuels and Synthetic
Hydrocarbons. J. Org. Chem. 2009, 74, 487–498. [CrossRef] [PubMed]
5. Sunggyu, L. Handbook of Alternative Fuel Technologies; CRC Press: New York, NY, USA, 2007; pp. 297–321.
6. Jadhav, S.G.; Vaidya, P.D.; Bhanage, B.M.; Joshi, J.B. Catalytic carbon dioxide hydrogenation to methanol:
A review of recent studies. Chem. Eng. Res. Design 2014, 92, 2557–2567. [CrossRef]
7. Olah, G.A.; Prakash, G.K.S.; Goeppert, A. Anthropogenic Chemical Carbon Cycle for a Sustainable Future.
J. Am. Chem. Soc. 2011, 133, 12881–12898. [CrossRef] [PubMed]
8. Pontzen, F.; Liebner, W.; Gronemann, V.; Rothemel, M.; Ahlers, B. CO2 -Based Methanol and DME—Efficient
Technologies for Industrial Scale Production. Catal. Today 2011, 171, 242–250. [CrossRef]
9. Arakawa, H. Research and development of on new synthetic routes for basic chemicals by catalytic
hydrogenation of CO2 . Stud. Surf. Sci. Catal. 1998, 114, 19–30.
10. Zachopoulos, A.; Heracleous, E. Overcoming the equilibrium barriers of CO2 hydrogenation to methanol
via water sorption: A thermodynamic analysis. J. CO2 Util. 2017, 21, 360–367. [CrossRef]
11. Saito, M.; Fujitani, T.; Takeuchi, M.; Watanabe, T. Development of copper/zinc oxide based multicomponent
catalyst for methanol synthesis from carbon dioxide and hydrogen. Appl. Catal. A 1996, 138, 311–318.
[CrossRef]
12. Reubroycharoen, P.; Vitidsant, T.; Asamic, K.; Yoneyama, Y.; Tsubaki, N. Accelerated methanol synthesis in
catalytically active supercritical fluid. Catal. Commun. 2003, 4, 461–464. [CrossRef]
Processes 2018, 6, 20 20 of 20

13. Westerterp, K.R.; Kuczynski, M.; Bodewes, T.N.; Vrijland, M.S.A. Neue Konvertersysteme für die
Methanol-Synthese. Chemie Ingenieur Technik 1989, 61, 193–199. [CrossRef]
14. Haut, B.; Halloin, V.; Amor, H.B. Development and analysis of a multi- functional reactor for equilibrium
reactions: Benzene hydrogenation and methanol synthesis. Chem. Eng. Process. 2004, 43, 979–986. [CrossRef]
15. Van Bennekom, J.G.; Venderbosch, R.H.; Winkelman, J.G.M.; Wilbers, E.; Assink, D.; Lemmens, K.P.J.;
Heeres, H.J. Methanol synthesis beyond chemical equilibrium. Chem. Eng. Sci. 2013, 87, 204–208. [CrossRef]
16. Gallucci, F.; Basile, A. A theoretical analysis of methanol synthesis from CO2 and H2 in a ceramic membrane
reactor. Int. J. Hydrog. Energy 2007, 32, 5050–5058. [CrossRef]
17. Struis, R.P.W.J.; Stucki, S.; Wiedorn, M. A membrane reactor for methanol synthesis. J. Membr. Sci. 1996, 113,
93–100. [CrossRef]
18. Struis, R.P.W.J.; Stucki, S. Verification of the membrane reactor for methanol synthesis. Appl. Catal. A Gen.
2001, 216, 117–129. [CrossRef]
19. Chen, G.; Yuan, Q. Methanol synthesis from CO2 using a silicone rubber/ceramic composite membrane
reactor. Sep. Purif. Technol. 2004, 34, 227–237. [CrossRef]
20. Barbieri, G.; Marigliano, G.; Golemme, G.; Drioli, E. Simulation of CO2 hydrogenation with CH3 OH removal
in zeolite membrane reactor. Chem. Eng. J. 2002, 85, 53–59. [CrossRef]
21. Gallucci, F.; Paturzo, L.; Basile, A. An experimental study of CO2 hydrogenation into methanol involving a
zeolite membrane reactor. Chem. Eng. Proc. 2004, 43, 1029–1033. [CrossRef]
22. Kunkes, E.; Behrens, M. Methanol Chemistry. Available online: http://pubman.mpdl.mpg.de/pubman/
item/escidoc:1587235/component/escidoc:1719633/Methanol%20Chemistry%20final.pdf (accessed on
2 February 2018).
23. Sahibzada, M.; Metcalfe, I.S.; Chadwick, D. Methanol synthesis from CO/CO2 /H2 over Cu/ZnO/Al2 O3 at
differential and finite conversions. J. Catal. 1998, 174, 111–118. [CrossRef]
24. Van den Bussche, K.M.; Froment, G.F. A steady state kinetic model for methanol synthesis and the water gas
shith reaction on commercial Cu/ZnO/Al2 O3 catalyst. J. Catal. 1996, 161, 1–10. [CrossRef]
25. Lee, J.S.; Lee, K.H.; Lee, S.Y.; Kim, Y.G. A Comparative-Study of Methanol Synthesis from CO2 /H2 and
CO/H2 over a Cu/ZnO/Al2 O3 Catalyst. J. Catal. 1993, 144, 414–424. [CrossRef]
26. Montebelli, A.; Visconti, C.G.; Groppi, G.; Tronconi, E.; Ferreira, C.; Kohler, S. Enabling small-scale methanol
synthesis reactor through the adoption of high conductive structured catalyst. Catal. Today 2013, 215, 176–185.
[CrossRef]
27. Hansen, J.B.; Nielsen, P.E.H. Methanol Synthesis. In Handbook of Heterogenous Catalysis; Ert, G., Knötzinger, H.,
Weitkamp, J., Eds.; Weily-VCH: New York, NY, USA, 2008.
28. Montgomery, D.C. Design and Analysis of Experiments; John Wiley & Sons: New York, NY, USA, 2005.
29. Dutka, B.M.; Ditaranto, M.; Løvås, T. Application of a Central Composite Design for the Study of NOx
Emission Performance of a Low NOx Burner. Energies 2015, 8, 3606–3627. [CrossRef]
30. Box, G.E.P.; Hunter, J.S.; Hunter, W.G. Statistics for Experimenters: Design, Innovation, and Discovery;
Wiley-Interscience: Hoboken, NJ, USA, 2005; pp. 447–455.
31. Cheng, T.; Rousseau, R.W.; Kliopatric, P.K. Methanol synthesis reactions: Calculations of equilibrium
conversions using equations of state. Ind. Eng. Chem. Process Des. Dev. 1986, 25, 477–481. [CrossRef]
32. Skrzypek, J.; Lachowska, M.; Grzesik, M.; Słoczyński, J.; Nowak, P. Thermodynamics and kinetics of low
pressure methanol synthesis. Chem. Eng. J. 1995, 58, 101–108. [CrossRef]
33. An, X.; Zuo, Y.; Zhang, Q.; Wang, J. Methanol synthesis from CO2 hydrogenation with a Cu/Zn/Al/Zr
fibrous catalyst. Chin. J. Chem. Eng. 2009, 17, 88–94. [CrossRef]
34. Manenti, F.; Cieri, S.; Restelli, M. Considerations on the steady state modeling of methanol synthesis fixed
bed reactor. Chem. Eng. Sci. 2011, 86, 152–162. [CrossRef]

© 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).