Impact of Membrane Thickness on Characteristics of Biogas Dry Reforming Membrane Reactor Using Pd/Cu Membrane and Ni/Cr/Ru Catalyst

1. Introduction

Global warming is a serious problem in the world. There is a lot of promising procedures to solve this problem. One of the procedures is green H₂, e.g. H₂ production from a renewable energy. This study focuses on a biogas dry reforming (BDR) as a procedure to produce green H₂. A biogas is a fuel consisting of CH₄ (55-75 vol%) and CO₂ (25-45 vol% [1], generally. A biogas is produced from fermentation by the action of anaerobic microorganisms on raw materials, e.g., garbage, livestock excretion, and sewage sludge. In 2020, 1.46 EJ of biogas was produced in the world [2], which was five times as large as that in 2000 [2]. Therefore, this study excepts that the amount of produced biogas increases more and more in the near future.

Generally speaking, a biogas is utilized as a gases fuel for gas engines and micro gas turbines [3]. However, the power generation efficiency becomes low compared to a natural gas due to lower heating value caused by including CO₂. Considering to solve this problem, this study proposes the combination system consisting of a BDR reactor with a solid oxide fuel cell (SOFC) [4,5,6]. CO which is a by-product from BDR can be utilized as a fuel for SOFC. Since SOFC can be a co-generation system, the total energy conversion efficiency is higher compared to existing power generation systems, i.e., gas engines and micro gas turbines.

There are many reported researches on BDR [7,8,9,10,11]. This study focuses on the selection of catalyst since it influences the performance of a BDR reactor. According to the literature survey, Ni-based catalysts have been investigated for BDR [7,8,9,10,11]. Ni/Al₂O₃ catalyst which was synthesized by a conventional wet impregnation performed the CH₄ conversion of approximately 100 % and the CO₂ conversion of approximately 95 % at 850 °C [7]. The CH₄ conversion and the CO₂ conversion increased with the increase in the reaction temperature from 700 °C to 850 °C. The CO₂ rich condition (CO₂/CH₄ = 1.2) was slightly better in terms of conversion and it induced a significant reverse water gas shift reaction (RWGS), which provided a low H₂/CO ratio and H₂ selectivity. Ni₃Co supported on Al₂O₃ performed CH₄ conversion of 59.8 % and 96.2 % at 600 °C and 750 °C, respectively [8]. In addition, this catalyst also performed CO₂ conversions of 24.4 % and 45.4 % at 600 °C and 750 °C, respectively. Though the significant carbon formation caused the catalyst deactivation, the carbon formation was less at higher reaction temperature. NiCeZrO₂ which was prepared by sol-gel method performed the CH₄ conversion of 50 % and the CO₂ conversion of 50 % at 800 °C [9]. The increase in the molar ratio of CH₄:CO₂ to 1.5 caused the formation of coke, which was suppressed by the addition of water. Ni/MgO which was prepared by a wet impregnation method performed the CH₄ conversion of approximately 70 % and the CO₂ conversion of approximately 85 % at 800 °C [10]. According to the investigation on the impact of reaction temperature from 300 °C to 800 °C, the highest CH₄ conversion as well as the highest CO₂ conversion were obtained at 800 °C. Ni-La/SBA-16 catalyst which was prepared by a hydrothermal process performed the CH₄ conversion of approximately 95 % and the CO₂ conversion of approximately 93 % at 750 °C [11]. The CH₄ conversion and the CO₂ conversion increased with the increase in the reaction temperature from 600 °C to 750 °C where the CH₄ conversion as well as the CO₂ conversion were approximately 70 % at 600 °C.

On the other hand, Ru-based catalysts have been also investigated [12,13,14]. Ru/Ni catalyst which was prepared by a wet impregnation performed the CH₄ conversion of approximately 78 % and the CO₂ conversion of approximately 72 % at 750 °C [12]. When changing the ratio of Ru to Ni from 0 to 2, the ratio of 0.8 showed the best performance. Ru supported on a binary La₂O₃-SiO₂ catalyst which was prepared by incipient wetness impregnation performed the CH₄ conversion of 5.3 % and the CO₂ conversion of 5.5 % at 550 °C [13]. Ru-Ni/MgAl₂O₄ which was prepared by a wet impregnation method performed the CH₄ conversion of 93 % and the CO₂ conversion of 96 % at 750 °C [14]. When decreasing in the reaction temperature from 750 °C to 600 °C, the CH₄ conversion and the CO₂ conversion decreased to 65 % and 73 %, respectively.

Although several Ni-based catalysts have been investigated, the Ni/Cr catalyst has not been examined well without the authors’ previous study [5]. In addition, the Ni/Cr/Ru catalyst has not been investigated well without the authors’ previous study [15].

Additionally, it is important to operate at a lower temperature to improve the thermal energy efficiency of BDR since BDR is an endothermic reaction. This study thinks a membrane reactor is an effective procedure to meet this purpose since the H₂ production is promoted by means of providing the non-equilibrium state with H₂ separation from the reaction site [5]. Though several Pd alloy membranes such as Pd/Ag and Pd/Au are commercialized as well as used for the research, the cost of Pd/Cu is relatively smaller than that of Pd/Ag, Pd/Au and pure Pd. This study thinks the cost is important to apply the system proposed by this study to the industry in the near future. Consequently, this study selected Pd/Cu as H₂ separation membrane. The authors’ previous study using Ni/Cr/Ru catalyst [15] used the Pd/Cu membrane whose thickness was 20 μm only. However, there is no study to investigate the impact of thickness of Pd/Cu membrane on the characteristics of BDR using Ni/Cr/Ru catalyst. This study thinks there would be the optimum thickness of Pd/Cu membrane to obtain the higher performance of membrane reactor using Ni/Cr/Ru catalyst proposed by this study, resulting from that the H₂ separation rate of H₂ separation membrane and the H₂ production rate of catalyst are thought to be important to realize the non-equilibrium state of BDR.

Consequently, the aim of this study is to clarify the impact of thickness of Pd/Cu membrane on the characteristics of BDR using Ni/Cr/Ru catalyst. The impact of reaction temperature, the molar ratio of CH₄:CO₂, the differential pressure between the reaction chamber and the sweep chamber, and the introduction of a sweep gas on the characteristics of BDR reactor using Pd/Cu membrane as well as Ni/Cr/Ru catalyst is investigated. The molar ratio of CH₄:CO₂ = 1.5:1 simulates a biogas in this study.

The reaction scheme of CH₄ dry reforming (DR) is described as follows:

CH₄ + CO₂ ↔ 2CO + 2H₂ + 247 kJ/mol

(1)

Additionally, the following reaction schemes can be considered in this study:

CO₂ + H₂ ↔ CO + H₂O + 41 kJ/mol

(2)

CO₂ + 4H₂ ↔ CH₄ + 2H₂O – 164 kJ/mol

(3)

CH₄ + H₂O ↔ CO + 3H₂ + 206 kJ/mol

(4)

where Equation (2) indicates a reverse water gas shift reaction (RWGS), Equation (3) indicates a methanation reaction (MR), and Equation (4) indicates a steam reforming of CH₄ (SR). As to a carbon deposition, the following reaction schemes are considered in this study:

CH₄ ↔ C + 2H₂ + 75 kJ/mol

(5)

2CO ↔ C + CO₂ – 173 kJ/mol

(6)

CO₂ + 2H₂ ↔ C + 2H₂O – 90 kJ/mol

(7)

CO + H₂ ↔ C + H₂O – 131 kJ/mol

(8)

2. Experiment

2.1. Experimental Set-Up

Figure 1 shows the schematic drawing of the experimental apparatus of this study. The experimental apparatus consists of a gas cylinder, mass flow controllers (S48-32; HORIBA METRON INC. Beijing, China), pressure sensors (KM31), valves, a vacuum pump, a reactor composed of a reaction chamber and a sweep chamber, and gas sampling taps. The reactor is placed in an electric furnace. This study controls the temperature in the electric furnace by far-infrared heaters (MCHNS1; MISUMI, Tokyo, Japan). CH₄ gas whose purity is over 99.4 vol% and CO₂ gas whose purity is over 99.9 vol% are controlled by the mass flow controllers and mixed before flowing into the reaction chamber. We measure the pressure of the mixed gas at the inlet of the reaction chamber by means of pressure sensors. We control Ar gas whose purity is over 99.99 vol% by a mass flow controller as a sweep gas and measure the pressure of Ar gas by a pressure sensor. We suction the exhausted gases at the outlet of the reaction chamber and sweep chamber by means of gas syringe via gas sampling taps. We measure the concentration of sampled gas by a TCD gas chromatograph (GL Science, Tokyo, Japan). The minimum resolution of the TCD gas chromatograph and the methanizer is 1 ppmV. We measure the gas pressure at the outlet of the reactor by means of a pressure sensor. We measure the gas concentration and pressure at the outlet of the reaction chamber and the sweep chamber, respectively. We install the valves at the outlet of the reaction chamber and close them when we flow the reaction gas consisting of CH₄ and CO₂ into the reaction chamber. As a result, H₂ in the reaction chamber after contacting the Ni/Cr/Ru catalyst penetrates through the H₂ separation membrane, i.e. Pd/Cu membrane, and flows into the sweep chamber. In other words, no H₂ flows out of the reaction chamber unless it penetrates through the H₂ separation membrane.

Figure 2 shows the details of the reactor in this study. The reactor is composed of a reaction chamber, a sweep chamber, and a H₂ separation membrane, i.e. Pd/Cu membrane. The reaction chamber and the sweep chamber are made of stainless steel with the size of 40 mm×100 mm×40 mm. The volume of reaction space is 16×10^-5 m³. We charge a porous Ni/Cr/Ru (Ni: 69.2 wt%, Cr: 29.6 wt%, Ru: 1.2 wt%) in the reaction chamber. The mean hole size of the Ni/Cr/Ru catalyst used in this study is 1.95 mm. According to the manufacture’s brochure, the porosity of the Ni/Cr/Ru catalyst used in this study is 0.93. The weight of Ni/Cr/Ru catalyst is 66.3 g.

Figure 3 shows a photograph charged in the reaction chamber in this study [15]. We selected a Pd/Cu membrane (Cu: 40 wt%; Tanaka Kikinzoku, Tokyo, Japan) as a H₂ separation membrane. The thickness of the Pd/Cu membrane was changed by 20 μm, 40 μm and 60 μm. We measured the temperatures at the inlet, the middle, and the outlet of the reaction chamber and the sweep chamber by means of K-type thermocouples. We controlled the initial operation temperature and set by a far-infrared heater and thermocouples. We saved the data of measured temperature and pressure by means of a data logger (GL240; Graphic Corporation, Tokyo, Japan).

Table 1 lists the experimental parameters applied in this study. The molar ratio of provided CH₄:CO₂ was changed by 1.5:1, 1:1 and 1:1.5. The molar ratio of CH₄:CO₂ simulated biogas in this study. According to the authors’ previous study [16], the feed ratio of sweep gas to supply gas, which is defined as the flow rate of sweep gas divided by the flow rate of supply gas composed of CH₄ and CO₂, was set at 1.0. This is the optimum feed ratio of sweep gas to supply gas according to the authors’ previous study [16]. We investigate the impact of the installation of sweep gas. We think it is necessary to separate H₂ from the membrane reactor with a sweep gas and extract the H₂ from the sweep gas for the purpose of applying the system proposed by the authors. We have to consider that the additional energy and equipment are necessary to install the system proposed by this study in the actual industry. On the other hand, the additional energy and equipment are not necessary if we can separate H₂ without a sweep gas from the membrane reactor. Consequently, this study has conducted a comparison of the presence vs. absence of sweep gas. We changed the differential pressure between the reaction chamber and the sweep chamber by 0 MPa, 0.010 MPa and 0.020 MPa. This study measures and confirms the differential pressure by the pressure sensors installed at the outlet of the reaction chamber and the sweep chamber. As to the differential pressure between the reaction chamber and the sweep chamber, we had tried to carry out the experiment at the differential pressure of 0.030 MPa and the temperature of 600 °C. As a result, the hole was made within the Pd/Cu membrane. Consequently, this study reports the results below 0.020 MPa. We changed the initial reaction temperature by 400 °C, 500 °C and 600 °C. We measure the initial reaction temperature by thermocouples before flowing the mixed gas of CH₄ and CO₂ as well as the sweep gas into the reactor. The temperature of the reaction chamber decreased by approximately 3 °C during the experiment due to the endothermic reactions, which is shown above. The gas concentrations at the outlet of the reaction chamber and the sweep chamber were detected by the FID gas chromatograph (GC320; GL Science) and methanizer (MT221; GL Science). We show the average data of five trials for each experimental conditions investigated in this study in the following figures. The distribution of each gas concentration is below 10 %.

2.2. Assessment Factor to Evaluate the Performance of BDR Mmbrane Ractor in This Study

This study conducts the evaluation on the performance of the proposed BDR membrane reactor by examining gas concentration at the outlet of the reaction chamber and the sweep chamber. This study evaluates CH₄ conversion (X_CH4), CO₂ conversion (X_CO2), H₂ yield (Y_H2), H₂ selectivity (S_H2), and CO selectivity (S_CO) by using these data. This study defines these assessment factors as follows:

X_CH4 = (C_{CH4, in} – C_{CH4, out})/(C_{CH4, in})×100

(9)

X_CO2 = (C_{CO2, in} – C_{CO2, out})/(C_{CO2, in})×100

(10)

Y_H2 = (1/2)(C_{H2, out})/(C_{H4, in})×100

(11)

S_H2 = (C_{H2, out})/(C_{H2, out} + C_{CO, out})×100

(12)

S_CO = (C_{CO, out})/(C_{H2, out} + C_{CO, out})×100

(13)

where C_{CH4, in} indicates the concentration of CH₄ at the inlet of the reaction chamber [mol],

C_{CH4, out} indicates the concentration of CH₄ at the outlet of the reaction chamber [mol],

C_{CO2, in} indicates the concentration of CO₂ at the inlet of the reaction chamber [mol],

C_{CO2, out} indicates the concentration of CO₂ at the outlet of the reaction chamber [mol],

C_{H2, out} indicates the concentration of H₂ at the outlet of the reaction chamber and the sweep

chamber [mol], and C_{CO, out} indicates the concentration of CO at the outlet of the reaction

chamber [mol].

Additionally, this study evaluates H₂ recovery (H) and permeation flux (F) as follows:

H = (C_{H2, out, sweep})/(C_{H2, out, sweep} + C_{H2, out, react})×100

(14)

$$F={\displaystyle \frac{P\left(\sqrt{P_{react,ave}}-\sqrt{P_{sweep,ave}}\right)}{\delta }}\times 100$$

(15)

where C_{H2, out, sweep} indicates the concentration of H₂ at the outlet of the sweep chamber [mol],

C_{H2, out, react} indicates the concentration of H₂ at the outlet of the reaction chamber [mol], P

indicates the permeation factor [mol/(m∙s∙Pa^0.5)], P_{react, ave} indicates the average pressure of

the reaction chamber [MPa], P_{sweep, ave} indicates the average pressure of the sweep chamber

[MPa], and  indicates the thickness of the Pd/Cu membrane [m].

Furthermore, this study also evaluates the thermal efficiency of the membrane reactor (), which is defined as follows:

$$\eta ={\displaystyle \frac{Q_{H2}}{\left(W_{R.C.}+W_{S.C.}+W_P\right)}}\times 100$$

(16)

where Q_H2 indicates the heating value of produced H₂ based on the lower heating value

[W], W_R.C. indicates the amount of pre-heat of supply gas for the reaction chamber [W],

W_S.C. indicates the amount of pre-heat of the sweep gas for the sweep chamber [W], and

W_p indicates the pump power to provide the differential pressure between the reaction

chamber and the sweep chamber [W].

3. Results and Discussion

3.1. Impact of Thickness of Pd/Cu Membrane on Each Gas Concentration in the Reaction Chamber and the Concentration of H₂ in the Sweep Chamber When Changing the Initial Reaction Temperature and the Differential Pressure Between the Reaction Chamber and the Sweep Chamber with and Without a Sweep Gas

Figure 4 and Figure 5 show the impact of thickness of Pd/Cu membrane on each gas concentration in the reaction chamber and the concentration of H₂ in the sweep chamber changing the initial reaction temperature (it is described as reaction temperature later) and the differential pressure between the reaction chamber and the sweep chamber without a sweep gas, respectively. In addition, Figure 6 and Figure 7 show the impact of thickness of Pd/Cu membrane on each gas concentration in the reaction chamber and the concentration of H₂ in the sweep chamber changing the reaction temperature and the differential pressure between the reaction chamber and the sweep chamber with a sweep gas, respectively. In these figures, the reaction temperature was changed by 400 °C, 500 °C and 600 °C and the differential pressure between the reaction chamber and the sweep chamber was changed by 0 MPa, 0.010 MPa and 0.020 MPa. Moreover, the molar ratio of CH₄:CO₂ is 1.5:1 in these figures. This study shows the results with the gas concentration in the unit of ppmV in these figures. However, it can be converted into the unit of mol. Regarding the gas concentration, 10,000 ppmV is 1 vol%. The gas volume can be calculated by multiplying the gas concentration by the volume of the reaction chamber, i.e. 40 mm×100 mm×40 mm or the volume of the sweep chamber, i.e., 40 mm×100 mm×40 mm. After that, the unit of gas volume, e.g., m³ can be converted into mol by multiplying 22.4 kmol/m³. Since the volume of reaction chamber and sweep chamber as well as the conversion factor of 22.4 kmol/m³ are constant, this study thinks that the gas concentration in the unit of ppmV can be used as an evaluation value for the performance of the reaction.

According to Figure 4 and Figure 6, we can see that the concentration of H₂ in the reaction chamber increases with the increase in the reaction temperature irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber and existing a sweep gas or not. Since DR, RWGS and SR are endothermic reactions as shown in Equations (1), (2) and (4), the reaction progresses well with the increase in the reaction temperature according to the theoretical kinetic study [17].

On the other hand, we can see from Figure 5 and Figure 7 that the concentration in the sweep chamber increases with the increase in the reaction temperature irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber and existing a sweep gas or not. Since the concentration of H₂ in the reaction chamber is higher at higher reaction temperature, the driving force to penetrate through the Pd/Cu membrane becomes higher because of the high H₂ partial differential pressure between the reaction chamber and the sweep chamber, i.e., a high concentration difference of H₂ between the reaction chamber and the sweep chamber. Consequently, a higher concentration of H₂ in the sweep chamber is obtained.

Considering the impact of thickness of Pd/Cu membrane on the concentrations of H₂ in the reaction chamber and the sweep chamber, the highest concentrations of H₂ in the reaction chamber and the sweep chamber are obtained for the thickness of 40 μm mainly among the investigated conditions, i.e. the reaction temperature, the differential pressure between the reaction chamber and the sweep chamber, and existing a sweep gas or not. It can be thought that H₂ might be penetrated through H₂ separation membrane when the thickness of H₂ separation membrane is thinner. However, the optimum thickness of Pd/Cu membrane is 40 μm in this study. If the H₂ separation rate of the Pd/Cu membrane is higher than the reaction rate of Ni/Cr/Ru catalyst used in this study, the effect of H₂ separation, which provides a non-equilibrium state in DR, is not effective sufficiently [15]. This study thinks that it is necessary to match the H₂ production rate of the Ni/Cr/Ru catalyst and the H₂ separation rate of the Pd/Cu separation membrane in order to obtain the non-equilibrium state of DR [15]. When using the thicker membrane, i.e., the thickness of 60 μm, the H₂ separation rate is too low due to big resistance to separate H₂. On the other hand, the H₂ separation rate is too fast when using the thinner membrane, i.e., the thickness of 20 μm due to small resistance to separate H₂. Consequently, this study thinks that the optimum thickness of Pd/Cu membrane is 40 μm in this study.

3.2. Impact of Thickness of Pd/Cu Membrane on Each Gas Concentration in the Reaction Chamber and the Concentration of H₂ in the Sweep Chamber When Changing the Molar Ratio and the Differential Pressure Between the Reaction Chamber and the Sweep Chamber with and Without a Sweep Gas

Figure 8 and Figure 9 show the impact of thickness of Pd/Cu membrane on each gas concentration in the reaction chamber and the concentration of H₂ in the sweep chamber changing the molar ratio and the differential pressure between the reaction chamber and the sweep chamber without a sweep gas, respectively. In addition, Figure 10 and Figure 11 show the impact of thickness of Pd/Cu membrane on each gas concentration in the reaction chamber and the concentration of H₂ in the sweep chamber changing the molar ratio and the differential pressure between the reaction chamber and the sweep chamber with a sweep gas, respectively. In these figures, the molar ratio of CH₄:CO₂ was changed by 1.5:1, 1:1 and 1:1.5 and the differential pressure between the reaction chamber and the sweep chamber was changed by 0 MPa, 0.010 MPa and 0.020 MPa. Moreover, the reaction temperature is 600 °C in these figures.

We can see from Figure 8 and Figure 10 that the highest concentration of H₂ in the reaction chamber is obtained for the molar ratio of CH₄:CO₂ = 1.5:1 among the investigated molar ratios irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber as well as existing a sweep gas or not. We can claim that this tendency matches with the authors’ previous studies investigating Ni, Ni/Cr and Ni/Cr/Ru catalysts [5,15]. The reaction mechanism to explain why the highest concentration of H₂ is obtained in the case of the molar ratio of CH₄:CO₂ = 1.5:1 can be explained as follows [5]: since the amount of CH₄ in this case, (i) H₂ is produced by the reactions shown in Equations (1) and (5), (ii) the produced H₂ is consumed by the reaction shown in Equation (2), resulting that CO is produced, (iii) a part of CO produced by the reactions shown in Equations (1) and (2) is consumed by Equation (6), and (iv) H₂O produced by the reactions shown in Equations (2) and (3) is consumed during Equation (4).

In addition, it is seen from Figure 9 and Figure 11 that the concentration of H₂ in the sweep chamber is the highest for the molar ratio of CH₄:CO₂ = 1.5:1 among the investigated molar ratios irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber as well as existing a sweep gas or not. Since the concentration of H₂ is the reaction chamber in the case of the molar ratio of CH₄:CO₂ = 1.5:1 is higher compared to the other molar ratios, the driving force to penetrate H₂ through the Pd/Cu membrane becomes larger because of the high H₂ partial pressure between the reaction chamber and the sweep chamber, i.e., a large concentration difference of H₂ between the reaction chamber and the sweep chamber. Consequently, a higher concentration of H₂ in the sweep chamber is obtained in the case of the molar ratio of CH₄:CO₂ = 1.5:1.

Moreover, it can be seen from Figure 8, Figure 9 and Figure 10 that the concentrations of CO, CH₄ and CO₂ are higher than those of H₂ in the reaction chamber and the sweep chamber. This study considers the reactions shown by Equations (1) – (8). The following steps of reactions might have occurred [15]:

(i)

H₂ is produced by Equations (1) and (5).

(ii)

The produced H₂ is consumed by Equations (2) and (3), resulting that CO, CH₄ and H₂O are produced.

(iii)

The produced CO is consumed by Equations (6) and (8), resulting that C, CO₂ and H₂O are produced.

Consequently, this study thinks that the concentrations of CO, CH₄ and CO₂ are higher than those of H₂ in the reaction chamber and the sweep chamber. After all experiments including 54 experimental conditions were conducted for each thickness of Pd/Cu membrane, the weights of the Ni/Cr/Ru catalyst for the thickness of 20 μm, 40 μm and 60 μm have increased by 0.108 g, 0.017 g and 0.315 g, respectively, indicating that a carbon is produced. Though the amount of produced carbon was small, a carbon deposition was confirmed in this study. Additionally, regarding H₂O production which is explained by Equations (2) and (3), it was confirmed by naked eye observation using a gas bag as shown in Figure 12. The methanizer and TCD gas chromatograph used in this study cannot detect H₂O since they are destroyed by absorbing H₂O on the surface of the column. Consequently, this study adopted a gas bag to capture H₂O exhausted from the reactor installed at the outlet of the reactor. We can observe that the existence of H₂O by the naked eye since the phase of H₂O is changed from a gas into a liquid [15]. In Figure 12, the left figure shows the gas bag used to capture H₂O, remaining in the red circle shown in Figure 12. The charged color parts on the white bar to capture H₂O shown in the right figure in Figure 12 indicates the production of liquid H₂O. In Figure 12, the white bar is used to show the existence of liquid H₂O. Though H₂O exists as a vapor during the experiment due to the reaction temperature over 400 °C, the phase of H₂O changes into a liquid due to a temperature drop after capturing H₂O by the gas bag.Considering the impact of thickness of Pd/Cu membrane, the highest concentration of H₂ is obtained for the thickness of 40 μm mainly among the investigated conditions, i.e. the molar ratio of CH₄:CO₂, the differential pressure between the reaction chamber and the sweep chamber, and existing a sweep gas or not. Since the concentration of H₂ is higher in the case of 40 μm, the above described reaction steps (ii) and (iii) might not occur strongly. As discussed above, it can be thought that H₂ might be penetrated through H₂ separation membrane when the thickness of H₂ separation membrane is thinner. However, the optimum thickness of Pd/Cu membrane is 40 μm in this study. If the H₂ separation rate of the Pd/Cu membrane is higher than the reaction rate of Ni/Cr/Ru catalyst used in this study, the effect of H₂ separation, which provides a non-equilibrium state in DR, is not effective sufficiently [15]. This study thinks that it is necessary to match the H₂ production rate of the Ni/Cr/Ru catalyst and the H₂ separation rate of the Pd/Cu separation membrane in order to obtain the non-equilibrium state of DR [15]. Therefore, this study claims that the optimum thickness of Pd/Cu membrane is 40 μm.

3.3. Comparison of Assessment Factors Among the Investigated Experimental Conditions

To investigate the performance of the proposed membrane reactor using the Ni/Cr/Ru catalyst and Pd/Cu membrane, Table 2, Table 3 and Table 4 show the comparison of CH₄ conversion, CO₂ conversion, H₂ yield, H₂ selectivity, CO₂ selectivity, H₂ recovery, permeation flux and thermal efficiency for the different thicknesses of Pd/Cu membrane, the molar ratios of CH₄:CO₂, and the differential pressures between the reaction chamber and the sweep chamber. The reaction temperature is 600 °C in this study.

According to Table 2, Table 3 and Table 4, we can see that the CO₂ conversion indicates a negative value from -46.0 % to -113 % irrespective of investigated conditions. It can be thought that a reaction consuming CH₄ and a producing CO₂ occur [5,15]. Additionally, we can see from Table 2, Table 3 and Table 4 that the CO selectivity is much higher than the H₂ selectivity irrespective of investigated conditions, which is similar to the authors’ previous studies [5,15]. On the other hand, we can see from Table 2, Table 3 and Table 4 that the CH₄ conversion indicates a positive value from 67.1 % to 88.2 %. It can be thought that CH₄ is consumed by DR and Equation (5). It is seen from Figure 5, Figure 7, Figure 9 and Figure 11 that H₂ is moved to the sweep chamber. However, it is also seen from Figure 6, 8, Figure 10 and Figure 12 that some H₂ produced remains in the reaction chamber, indicating that all H₂ does not move to the sweep chamber by penetrating through Pd/Cu membrane. Consequently, this study claims that the following reaction mechanism can be explained [5,15]: (i) H₂ is produced by the reactions shown by Equations (1) and (5); (ii) the produced H₂ is consumed by the reaction shown by Equation (2), resulting that CO is produced; (iii) a part of CO produced by the reactions shown by Equations (1) and (2) is consumed during the reaction shown by Equation (6); and (vi) H₂O produced during the reactions shown by Equations (2) and (3) is consumed by Equation (4). As discussed before, the production of carbon as well as H₂O was confirmed.

According to the investigation in this study, the highest concentration of H₂ was obtained for the thickness of 40 μm, the molar ratio of CH₄:CO₂ = 1.5:1 and the differential pressure between the reaction chamber and the sweep chamber of 0 MPa without a sweep gas, which is 4890 ppmV in the reaction chamber and 38 ppmV in the sweep chamber, respectively. It is revealed from Table 2, Table 3 and Table 4 that CH₄ conversion, H₂ yield and thermal efficiency are 75.0 %, 0.214 % and 2.92 %, respectively under this condition. Though the CH₄ conversions obtained in this study are competitive to the previous studies [7,8,9,10,11], the H₂ yield and the thermal efficiency are still low. Therefore, this study suggests the following approaches to improve the performance of proposed membrane reactor using Ni/Cr/Ru catalyst and Pd/Cu membrane: (i) to optimize the catalyst shape and composition (the pore size and the weight ratio of Ni, Cr and Ru, (ii) to optimize the thickness and the composition of Pd/Cu membrane, (iii) to match the H₂ separation rate of Pd/Cu membrane and the H₂ production rate of Ni/Cr/Ru catalyst, which can decide the optimum operation condition. This study would like to investigate these subjects in the near future.

4. Conclusions

This study has conducted the investigation to reveal the impact of thickness of Pd/Cu membrane on the characteristics of BDR using Ni/Cr/Ru catalyst. This study has also investigated the impact of reaction temperature, the molar ratio of CH₄:CO₂, the differential pressure between the reaction chamber and the sweep chamber, and the introduction of a sweep gas on the characteristics of BDR reactor using Pd/Cu membrane as well as Ni/Cr/Ru catalyst. As a result, this study can draw the following conclusions:

(i)

It is revealed that the concentration of H₂ in the reaction chamber increases with the increase in the reaction temperature irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber and existing a sweep gas or not. The reaction progresses well with the increase in the reaction temperature since DR, RWGS and SR are endothermic reactions.

(ii)

It is revealed that the concentration in the sweep chamber increases with the increase in the reaction temperature irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber and existing a sweep gas or not. When the concentration of H₂ in the reaction chamber is higher at higher reaction temperature, the driving force to penetrate through the Pd/Cu membrane is higher due to the high H₂ partial differential pressure between the reaction chamber and the sweep chamber.

(iii)

It is revealed that the highest concentration of H₂ in the reaction chamber as well as that in the sweep chamber are obtained for the molar ratio of CH₄:CO₂ = 1.5:1 among the investigated molar ratios irrespective of the thickness of Pd/Cu membrane, the differential pressure between the reaction chamber and the sweep chamber as well as existing a sweep gas or not.

(iv)

The highest concentration of H₂ is obtained for the thickness of 40 μm mainly among the investigated conditions, i.e. the reaction temperature, the molar ratio of CH₄:CO₂, the differential pressure between the reaction chamber and the sweep chamber, and existing a sweep gas or not. This study claims that the optimum thickness is 40 μm.

(v)

It is clarified that the highest concentration of H₂ is obtained for the thickness of 40 μm, the molar ratio of CH₄:CO₂ = 1.5:1 and the differential pressure between the reaction chamber and the sweep chamber of 0 MPa without a sweep gas, which is 4890 ppmV in the reaction chamber and 38 ppmV in the sweep chamber, respectively. Under this condition, CH₄ conversion, H₂ yield and thermal efficiency are 75.0 %, 0.214 % and 2.92 %, respectively.

(vi)

In the near future, the following subjects can be considered: to optimize the catalyst shape and composition (the pore size and the weight ratio of Ni, Cr and Ru); to optimize the thickness and the composition of Pd/Cu membrane; to match the H₂ separation rate of Pd/Cu membrane and the H₂ production rate of Ni/Cr/Ru catalyst, deciding the optimum operation condition.