How-to, Tips and Articles: The Effect of Graphite and Sub-bituminous Coal Characteristics on the Resistance of the Inner Direct Carbon Fuel Cell (DCFC) System

ABSTRACT

A direct carbon fuel cell (DCFC) system directly converts the chemical energy of solid carbonaceous fuel into electrical energy. The electrochemical reaction of this system has an influence on the properties of solid fuel, such as crystal structure, element composition, and surface properties. In addition, when using raw coals as DCFC fuel, the volatile gases released from coal at a high temperature aﬀect cell performance. The purpose of this study is to investigate the eﬀect of fuel characteristics on the resistance of the inner DCFC system by electrochemical impedance spectroscopy (EIS) and equivalent circuits. Two kinds of solid carbon, graphite and sub-bituminous coal, were prepared to compare electrochemical characteristics by EIS measurement. The equivalent circuits were applied to the constant phase element (CPE) in a Randle circuit to explain the correlation between the fuel characteristic and electrochemical reaction resistance.

1. INTRODUCTION

The direct carbon fuel cell (DCFC) generates electricity from the oxidation reaction of carbon, and it is rather diﬀerent from conventional fuel cells that convert gaseous or liquid fuels into electricity. The DCFC system has many advantages; it is comprised of eco-friendly equipment based on using no harmful substances and can employ a variety of solid-phase carbonaceous fuels.1 In addition, its reaction products are easily collected because it is theoretically pure carbon dioxide.2−4 The DCFC system also has high eﬃciency when compared to other power systems, owing to independence of the Carnot cycle, near 100% fuel utilization, and 100% theoretical conversion eﬃciency. The solid carbon fuels are reacted with carbonate ions at the anode, and CO2 and O2 produced carbonate ions followed by the electrochemical reaction.5,6

Anode reaction

C(s) + 2CO₃²⁻ = 3CO2 + 4e⁻ (1)

Cathode reaction

2CO2 + O2 + 4e⁻ = 2CO₃²⁻ (2)

The performance of DCFC is dependent upon the supplied fuels and the working electrode condition.7−9 The eﬀects of fuel properties have been studied as DCFC fuels with various solid carbonaceous fuels.9 However, when using coal as fuel, it is diﬃcult to predict the correlation between the fuel properties and the electrochemical reaction, mainly because the characteristic of coal is heterogeneous. Coal has various components besides carbon, such as oxygen, nitrogen, sulfur, and metallic oxides, and these properties are therefore altered with increasing temperature.10 The electrolytes used in DCFC, K2CO3 and Li2CO3, cause the gasification reaction of coal.11 In addition, MgO and Fe2O3, ash components of inner coal, function as catalysts of gasification of solid carbon within molten carbonate.8 Each of these factors contributing to uncertainty of the cell is interesting from the perspective of DCFC research.

Chien et al. show the eﬀect of the flow rate of carrier gas (He) and the Boudouard reaction on DCFC performance.10 The cell performance was increase with decreasing gas flow rates. The low flow rates caused an increasing CO concentration and improved the electrochemical oxidation of CO gas. In addition, CO2 produced by the electrochemical reaction reacted with solid carbon through Boudouard reaction and regenerated CO that has an eﬀect on the fuel cell performance.

Li et al. has investigated the eﬀect of gasification at temperatures ranging from 700 to 1000 °C, reporting that gases released from coal aﬀect the electrochemical reaction using K, Ca, and Ni as gasification catalysts of carbon black.11 The gasification eﬀects of catalysts with CO2 are K to be the best catalyst, Ni to occupy an intermediate eﬀect, and Ca to be the worst through decreasing the gasification temperature about 200, 150, and 130 °C with K, Ni, and Ca, respectively. These results showed that the power density increases with CO concentrations and the operating temperature of DCFC is reduced by the catalytic gasification process.

Hao et al. studied the eﬀect of impurities present in solid carbonaceous fuel in a molten carbonate DCFC.12 They analyzed the correlation between the cell performance and impurities using impedance spectroscopy. Some ash components, such as K, Ca, and Mg, involved fuel gasification as a catalyst on DCFC. The reactions from inherent ash components in the anode led to high performance compared to the expected performance of renewable biomass and wastes.

In our previous studies,13 through comparison of three coals and their chars, the gas eﬀects on the DCFC system were investigated and the three raw coals showed higher open circuit voltage (OCV) and maximum power density than those of their chars. From the results of measuring the gas component, the released gases from coals that were H2, CH4, CO, and CO2 had an influence on the DCFC performance. CO predominated over 600 °C as a result of the reverse Boudouard reaction and the reaction between the carbon and carbonate ions, which produced carbon monoxide.

Table 1. Proximate and Ultimate Analyses of Fuels

In the present study, the objective is to perform experiments with various operating conditions and to analyze inner resistance diﬀerences through electrochemical impedance spectroscopy (EIS) and equivalent circuit. To investigate electrochemical reaction parameters in various conditions, two types of solid carbonaceous fuels, graphite particle and Adaro coal, which have significantly diﬀerent characteristics of volatile matter, were used and two temperature conditions were used. The electrochemical reactions were analyzed using EIS, and impedance spectra were presented as equivalent circuits containing electrolyte resistance, double-layer capacitance, and charge transfer resistance. Fuel properties were examined by four forms of analysis, being thermogravimetric analysis (TGA), X-ray diﬀraction (XRD), gas adsorption, and X-ray photo-electron spectroscopy (XPS), to compare the diﬀerences between the graphite particle and Adaro coal. The correlation between fuel properties and the characteristics of electro-chemical reactions were studied by i−V, EIS, and equivalent circuits.

2. EXPERIMENTAL SECTION

2.1. Sample Preparation

Two types of solid carbonaceous fuels were used to draw comparisons between physical characteristics and electrochemical reactions. The Adaro coal used is considered sub-bituminous coal, and the graphite particle is almost entirely composed of carbon. These samples were supplied with lithium and potassium carbonate powder at the working electrode. All samples used in each experiment were analyzed by both the proximate and ultimate analyses, and these results are listed in Table 1. Proximate analysis was conducted by TGA701 (LECO, St. Joseph, MI) according to the standard set forth by ASTM D5142. The ultimate analysis was submitted to the Vario Micro Cube at the Korea Basic Science Institute (KBSI, Busan Center, Korea).

2.2. Analysis of Fuel Properties

To compare the fuel properties of Adaro coal and graphite particle, measurements of thermal reactivity, crystal structure, specific surface area, pore volume, and functional groups on the fuel surface were carried out.

TGA was conducted using SDT Q600 (TA Instrument, New Castle, DE) to analyze thermal reactivity of fuels with increasing temperature. Carbonaceous samples of 10 mg were used in each experiment and heated to 900 °C at a 10 °C/min heating rate under an Ar atmosphere. The atmospheric gas was supplied at a 100 mL/min flow rate.

The surface area and pore volume were identified using Brunauer− Emmett−Teller (BET, ASAP 2020, Micromeritics Co., Norcross, GA). Nitrogen gas was used as the adsorbent to measure the specific surface area and total pore volume. A total of 0.5−1 g of the sample was degassed at 50 °C and then kept constant at approximately −200 °C using liquid nitrogen. Nitrogen gas was adsorbed from 0.05 to 0.95 of the P/P0 range and desorbed from 0.95 to 0.1 of the P/P0 range.

XRD was employed to investigate the physical structure of each fuel. The results of XRD were obtained from 10° to 90° in 2θ using Cu Kα radiation (0.154 nm). The Bragg equation (eq 3) and the Scherrer equation (eq 4) were used to determine the lattice parameter14:

λ = 2d sin θ (3)

t = 0.9λ /(β cos θB) (4)

where λ is the wavelength, d is the distance between planes of the crystal lattice, θ is the angle of diﬀraction between the incident ray and the lattice planes, θB is the Bragg angle, and β is the full width at half maximum (FWHM).

Through XPS, the elemental composition of the fuel surface was investigated using a Theta Probe AR-XPS system (Thermo Fisher Scientific, U.K.). XPS is a quantitative spectroscopic technique that measures the elemental composition, chemical state, and electronic state of the elements that exist within 10 nm from the material surface. This measurement used monochromatic Al Kα lines and power supply (15 kV and 150 W) to emit X-rays. The resultant data were obtained in units of 0.1 eV.

2.3. DCFC System and Experimental Method

The DCFC experimental system is described in detail in previous literature.15,16 Figure 1 shows a schematic diagram of a half cell of DCFC.

Figure 1. Schematic diagram of the DCFC experimental system

The electrolyte was dried at 110 °C and mixed with a 62:38 ratios of Li2CO3 and K2CO3. These mixed molten carbonates are usually used as electrolytes in molten carbonate fuel cell (MCFC) and have advantages in the DCFC system because of high conductivity, appropriate melting temperatures, and stability with CO2 generated by electrochemical reactions.17 For each experiment, the electrolyte used was 460 g and the fuel used was 10 wt % (46 g) of the electrolyte. Three electrodes were made of platinum, with one being the working electrode, the next being the counter electrode, and the last acting as the reference electrode. Through previous studies, Li2PtO3 is formed on the Pt electrode surface under a CO2 atmosphere when heated in a molten carbonate electrolyte over 650 °C.18,19 On the basis of this, each experiment was performed within 6 h. The current density and power density were calculated by considering the eﬀective area of the working electrode (1.34 cm2 in the standard cell). Using an alumina tube, the counter electrode and reference electrode were isolated from the fuels. The alumina tubes have four holes with a diameter of 0.1mm for ion diﬀusion. The performance of the fuels was measured by the potentiostatic method using Versastat3 (Princeton Applied Research, Oak Ridge, TN). Prior to reaching the operating temperature, the counter electrode and reference electrode were filled with CO2 gas (50 mL/min). After attainment of the operating temperature, CO2 changed into a mixed gas of O2 (33.3 mL/min) and CO2 (66.7 mL/min).

2.3.1. Linear Sweep Potentiometry

The performance of DCFC was indicated by current density and power density, with the current being estimated from OCV to 0 V with a scan rate of 0.001 V/s using linear sweep potentiometry. The current density and power density can be obtained by a potentiostat. The results were divided on the basis of the contact area of the working electrode and then calculated to the current density and power density.

2.3.2. EIS

EIS is a method that takes alternating current (AC) signals and measures the response of a system by frequency.20 The frequency range for measuring EIS was from 0.1 Hz to 100 kHz. The amplitude of frequencies was 10 mV. From the results of the impedance spectrum, which responded to the frequencies within range, the acquired data were related to the characteristics of the electrochemical reaction system, such as electrolyte resistance, double-layer capacitance, and charge transfer resistance. For analysis of the eﬀect of fuel properties on system, the equivalent circuit was used. The equivalent circuits used for fitting impedance data implied the electrical double layer between the anode and molten electrolyte interface.21

3. RESULTS AND DISCUSSION

3.1. Characteristics of Adaro Coal and Graphite Particle

Figure 2. Thermal behaviors in TGA under Ar condition.

Figure 2 shows the results of thermal reactivity of the graphite particle and Adaro coal under an Ar atmosphere using TGA.15,16 In the case of the graphite particle, its weight was kept constant until 900 °C, because of bare moisture and volatile matter, while Adaro coal was decomposed actively. Weight variation of Adaro coal occurred around 100 °C, most likely from moisture evaporation. After the evaporation step, the volatile matter of inner coal particles was released from 300 to 500 °C by the devolatilization process. At 650 and 700 °C, which were the operating temperatures of this research, the fixed carbon of coal was decomposed and light gases, such as hydrogen, methane, carbon monoxide, and carbon dioxide, were produced. These gases influenced the electrochemical reactions at the working electrode, as demonstrated previously.13 In reviewing the results with graphite particle and Adaro coal, the eﬀect of thermal decomposition on each resistance value for each electrochemical reaction was clarified.

Table 2. Characteristics of Gas Adsorption

Table 2 exhibits the results from the BET method, including specific surface area, total pore volume, and mean pore diameter of each fuel. These surface characteristics are important physical properties because they are correlated to the triple-phase boundary between the electrolyte and carbon fuel. The specific surface area and total pore volume of the graphite particle are similar to those of Adaro coal in the raw state. However, thermal decomposition was confirmed from thermal reactivity of Adaro coal. Volatile matter inner raw coal particles, when they are emitted, generate many small-sized pores and change the surface properties. The char state of Adaro coal, which was carbonized until 700 °C, becomes much larger than in the raw coal state. In consideration of the cell-operating temperature, Adaro coal has a much higher surface area and total pore volume than those of the graphite particle, larger by 200 and 49 times, respectively.

The crystal structures of the fuels were analyzed by X-ray, and the results are shown in Figure 3.

Figure 3. XRD patterns of the graphite particle and Adaro coal.

In comparison to the XRD pattern of the two fuels, both Adaro coal and graphite particle have a peak around 26°. This means that the two solid carbonaceous fuels have a similar carbon crystal structure, being the (002) peak. The intensity of (002) peaks of the graphite particle, however, was much higher than that of Adaro coal because of the degree of crystallization and high carbon content. The graphite particle is mainly composed of crystal grains of (002) and has small sizes of the crystal grain compared to Adaro coal. For this reason, the carbon of Adaro coal is more easily reacted and an enhancement in the performance of the DCFC system was expected.7,8,13

Table 3 shows the quantitative elemental analysis of each fuel surface by XPS. The ratio of oxygen to carbon (O/C) means the degree of oxidation on the fuel surface, and the ratio of Si + Al to carbon [(Si + Al)/C] means the results of the ash component on the fuel surface. Adaro coal has a lower surface carbon composition compared to the graphite particle, which has 97.9 atomic %. However, Adaro coal has a high oxygen/ carbon ratio on the fuel surface, which indicates surface oxygen functional groups. The presence of surface functional groups permit reactivity with carbonate ions more readily and accelerate electrochemical reactions.7,13 Therefore, Adaro coal may have a better performance, owing to reaction kinetics and wettability between fuels and electrolyte.

Table 3. Quantitative Elemental Analysis of the Fuel Surface

3.2. Characteristics of i−V and i−P Results

The evaluation of the DCFC performance was conducted at 650 and 700 °C, and the results were revealed by the potential of the power density with the current density to analyze features of electrochemical reactions.

Figure 4. i−V and i−P curves of the (a) eﬀect of fuel types on the performance at 650 °C and (b) eﬀect of the operating temperature on the performance at 650 and 700 °C.

As presented in Figure 4, the OCVs of Adaro coal and graphite particle are −1.05 and −0.81 V, respectively. The potential of Adaro coal is higher than the graphite particle at the same current density level. In addition, the maximum power densities of Adaro coal and graphite particle were presented as 22.1 and 15.3 mW/cm2, respectively. The maximum power densities of the graphite particle and Adaro coal were significantly improved at approximately 45%. This is probably because the gases produced through the decomposition process have an influence on the DCFC performance.13 According to analysis of exhaust gas at the anode, the gases H2, CH4, CO, and CO2 were detected in the working electrode. Unlike the Adaro coal results, the graphite particle has a diﬀerent behavior because it has almost no volatile matter content. Therefore, the potential of the graphite particle is sharply diminished at the low current density region. An improved performance of raw coal may be caused by the volatile content, specific surface area, crystal structure, and surface functional groups.

Another reason for the performance boost in raw cool is that increasing the specific surface area and pore volume of fuels results in the enhancement of the DCFC performance, as indicated previously.2,7,15 The specific surface area and pore volume are closely related to the contact area of the molten electrolyte. Therefore, the DCFC performance is aﬀected through changing the reactive area. In addition, as mentioned in section 3.1, the crystal structure and oxygen/carbon ratio have a bearing on low performance. The high crystallization and low oxygen of the graphite particle create a higher bond energy compared to that of Adaro coal, so that a higher energy is required for carbon oxidation of the fuel surface.2

3.3. Results of EIS

To compare more details of the electrochemical reactions of Adaro coal and graphite particle, EIS was adopted. EIS was measured at OCV, and the impedance data are shown in Figure 5 as a Nyquist plot. In Figure 5a, the two fuels show diﬀerent trends of an impedance spectrum. At 650 °C, the real part of the impedance of the graphite particle is larger than that of Adaro coal within the whole frequency area. The electrolyte resistances (Rs) of the graphite particle and Adaro coal, with the value of the imaginary part being 0 in the high-frequency region, were 0.3137 and 0.5547 Ω cm2, respectively.20 These electrolyte resistances are higher values compared to previous studies, which have a similar DCFC system and experimental conditions.22 This discrepancy of electrolyte resistance may be responsible for the small size holes of alumina tubes preventing contact with solid fuels. In addition, the diﬀerence between two electrolyte resistances demonstrates that tar and other impurities emitted from Adaro coal have eﬀects on the resistance of electrolytes (Rs). However, charge transfer resistance (Rct) is dominant in this system because diﬀerences between Rct of the graphite particle, which is 26.18 Ω cm2, and that of Adaro coal, which is 5.16 Ω cm2, are detected at about 21 Ω cm2, while electrolyte resistance diﬀerences are shown at 0.241 Ω cm2.23

The measured impedance spectra were analyzed by equivalent circuits and fitted using ZSimpWin (EChem Software, Ann Arbor, MI). Panels a and b of Figure 5 show the fitted arcs obtained by using each equivalent circuit model, with the models themselves presented in Figure 6. The fitted EIS data acquired from the equivalent circuits were in agreement with the measured EIS data. The parameters of the equivalent circuit model and all values are represented in Table 4. The equivalent circuit is composed of Rs, Qs, Rct, and Cdl, and these parameters refer to electrolyte resistance, diﬀusion resistance, charge transfer resistance, and electrical double layer of capacitance.24 Reviewing these data are necessary to understand the behavior of the electrodes under various conditions.

Figure 5. Comparison of experimental and fitted impedance spectra for (a) Adaro coal and graphite particle at 650 °C and (b) Adaro coal at 650 and 700 °C.

Figure 6. Equivalent circuits by diﬀerent conditions for (a) graphite particle at 650 °C, (b) Adaro coal at 650 °C, and (c) Adaro coal at 700 °C.

In the case of the graphite particle, the equivalent circuit consisted of electrolyte resistance and one Q−R parallel circuit, which is associated with charge transfer and diﬀusion, with this model being displayed in Figure 6a. This concept based on volatile gases does not aﬀect the graphite particle at operating temperatures.25 The equivalent circuit model of Adaro coal is shown in Figure 6b, made up of two Q−R parallel circuits in series. This is similar to the equivalent circuit used with high-temperature fuel cells, where the gases were used as fuels.26 Through comparison of these results and the characteristics of solid fuels, the equivalent circuit models of the graphite particle and Adaro coal were revealed to be diﬀerent based on the eﬀect of gases released as fuels.

First, Q−R of Figure 6b is presented as the characteristic between the electrode and fuel surface.27 Rct, the charge transfer resistance at the interface between the electrode and fuels, is the diﬀerence between two points on the graphs of the semicircle that cross the real axis. The results of Adaro coal in Figure 5 appear as the first small semicircle in the high-frequency region. A second semicircle, which has a larger radius than the first semicircle, is shown in the low-frequency region. However, in the case of the graphite particle, the impedance values are much greater than those of Adaro coal in the whole area. Table 4 shows that the Rct of graphite particle is 4 times as large as Rct1 of Adaro coal. Tubilla et al. investigated the correlation between electrochemical reactivity and charge transfer resistance because of the increased contact area of the solid fuel and electrode.28 The Q was used to show an electrical double layer, a behavior of an incomplete capacitor.29 The modified Randles circuit imposed the “constant phase element (Q)” component, which is the circuit component used for equivalent circuit analysis.

The power of Q (Qn) having a range of 0−1 represents the behavior of the capacitor at interfaces close to 1 and the behavior of resistances close to 0.20,23 The Qn values of Adaro coal and graphite particle of 0.5−0.6, except for the Qn2 value of Adaro coal at 650 °C. Zoltowski reported the CPE, which has 0.5 because the power of CPE represents the Warburg element.30 Rhie et al. also investigated that the CPE has roles in Warburg impedance when the Qn value is between 0.5 and 0.6 on the DCFC system.23 Therefore, it implies that the end of the electrochemical reaction of all solid fuels is related with unrestricted diﬀusion of the flat electrode surface. The power of Q also means that the interfacial reaction between the electrode and the fuel is well-distributed when Adaro coal is used as fuel, most likely because of gases released by the thermal decomposition reaction, such as H2, CH4, CO, and CO2. In addition to eq 1, which is the electrochemical oxidation of solid carbon, H2 and CO gases can be oxidized electrochemically at the anode, such as the following equations:

H2 + CO₃²⁻ → CO2 + H2O + 2e⁻ (5)

CO + CO₃²⁻ → 2CO2 + 2e⁻ (6)

Therefore, the Qn2 value of Adaro coal is closer to 1 than other n values based on the extra electrochemical reactions of released gases within the interface region.

When using the same fuel, each value of the equivalent circuit elements was diﬀerent as a result of the operating temperature rising. When the operating temperature is 700 °C, the value of Rs is decreased to about 0.19. This is shown by an improvement in ion conductivity in the electrolytes with an increasing temperature.

Table 4. Parameters of the Equivalent Circuit of Fuels

At 700 °C conditions, the impedance data at the low frequency seem to be flat. The reasons for this are ascribed by the decrease of Qn1 and the change of the second Q−R. The decrease of Qn1 is caused by an inhomogeneous distribution of the reaction rate and non-uniform current distribution. It is described as a thermal decomposition phenomenon, which results in the surface characteristic changes of solid carbon fuels. The Q of the second Q−R, referring to the contact interface between fuels, was also changed to Cdl. This result signified the behavior of a constant phase element being replaced by the behavior of the capacitor with an increasing temperature. Both charge transfer resistances Rct1 and Rct2 also decreased with the rising temperature. As shown in Figure 5b, this can be confirmed by the small semicircle in the low-frequency region. This is because the produced gases can be used as fuel. CO2 of the reaction product and carbon in the fuel are converted to CO according to the Boudouard reaction at 700 °C, and the reaction distribution at the interface between the fuels is more widely spread. In addition, as described before, H2 also can be reacted with carbonate ion and produce electrons.13,23,31

4. CONCLUSION

(1) The maximum power density of Adaro coal is 22.1 mW/ cm2, a 45% performance improvement over the graphite particle at 15.3 mW/cm2. This is most likely because of the fuel characteristics of the released gases, such as specific surface area, total pore volume, and amorphous structure.

(2) At the same operating temperature, in the case of Adaro coal, the second parallel circuit, Q−R, is presented in comparison to the equivalent circuit of the graphite particle based on the influence of the gases released by thermal decomposition reactions. The charge transfer resistance (Rct1 + Rct2) of the Adaro coal graphite particle is also 4 times lower than Rct of the graphite particle.

(3) As the operating temperature increases, the CPE of the second Q−R changes to reflect the behavior of the capacitor. This is because CO is produced by carbon and CO2 through the Boudouard reaction at 700 °C and H2 is released from raw coal at the operating temperature. These produced gases function to improve non-uniform current distribution.

NOMENCLATURE

λ= wavelength of X-ray (nm)
θ = Bragg angle (deg)

β = full width at half maximum (fwhm)
d = distance of parallel planes (nm)

t = bulk of crystal lattice (nm)
L = inductance (H)

Rs = solution (electrolyte) resistance (Ω)
Q = constant phase element (CPE) (S sn)
Rct = charge transfer resistance (Ω)

Qn = power of Q (0 < n < 1)

Cdl = capacitance of electric double layer (F)

Subscripts

s = solution

ct = charge transfer

dl = electric double layer

Source: Seongyong Eom,† Jaemin Cho,† Seongyool Ahn,‡ Yonmo Sung,† Gyungmin Choi,*,† and Duckjool Kim†

† School of Mechanical Engineering, Pusan National University, Busan 609-735, Republic of Korea

‡ Central Research Institute of Electric Power Industry (CRIEPI), 2-6-1 Nagasaka, Yokosuka-shi

The 10 largest coal producers and exporters in Indonesia:

Click Here for Top World Clean Coal Technologies Contractors

Monday, March 16, 2020

The Effect of Graphite and Sub-bituminous Coal Characteristics on the Resistance of the Inner Direct Carbon Fuel Cell (DCFC) System