How-to, Tips and Articles: Investigation of Performance of Raw and Ash-free Coal in the Operation of a Direct Carbon Fuel Cell (DCFC)

Abstract

We have investigated the comparable performance of raw and ash-free coal in the operation of a direct carbon fuel cell (DCFC). The various structural and morphological analyses using SEM, TEM, EDX, XPS, XRD, and TGA are carried out to study the distinct physicochemical properties of coals. Due to contained volatile organic compounds, raw coal generates about a two-fold higher fuel cell performance compare to ash-free coal below a reaction temperature of 750 °C. However, over a cell temperature of 900 °C, both of them reach a similar power density of 170 mW cm⁻². In the long-term operation of a DCFC, we observe a distinctly more durable power performance using ash-free coal than that of raw coal.

1. Introduction

Coal is the second largest primary energy source in the world after oil, and it is a major source of electricity generation by thermal power plants [1–3]. The average thermal efﬁciency of conventional coal-ﬁred power plants, however, is around 40%, and at the same time they produce copious amounts of carbon dioxide (CO2) [2]. Since it is not enough to make electricity without coal, signiﬁcant improvements in the coal fuel conversion and utilization efﬁciencies of power generation systems are needed to solve the problems of a potential energy crisis and global warming, including the devel- opment of more efﬁcient carbon capture and storage technology.

For the direct use of solid coal fuel without additional gasiﬁcation and reforming processes, a direct carbon fuel cell (DCFC) has been developed as a promising novel concept of a high-temperature energy conversion system. The motivation for this development is owing to (i) an ideal energy efﬁciency of 100%, (ii) very pure CO2 generation through applying clean coal, and (iii) the possible usage of various types of carbon sources, such as coal, coke, char, graphite, and even biomass [4–6]. Raw coal (parent coal) is composed of complex chemical and physical heterogeneous structures, which contain high-value aromatic and hydroaromatic units linked together and up to 10% inorganic components [7–9]. These impurities might cause serious energy dissipation in the long-term operation of a DCFC.

The underlying progress in the development of a coal-based DCFC has been categorized mainly according to the electrolyte materials used, such as molten carbonate [5,10–13] and solid oxide [14–22]. In particular, the solid oxide electrolyte based DCFCs has been intensively researched due to a relatively stable electrolyte and easy to scale-up [17,23]. Nakagawa et al. [14] ﬁrst examined the performance of a porous Pt-YSZ anode by the direct internal gasiﬁcation of solid charcoal. Recently, Gür et al. [16] integrated a ﬂuidized bed DCFC with a Boudouard-type dry gasiﬁer and showed a maximum power density of 450 mW cm⁻² by applying CO gas generated from coalchar. Of late, Chuang and co-workers [17–19] reported that an Ag pasted Ni-YSZ anode-supported cell using coconut coke resulted in a maximum power density of 146 mW cm⁻² [19]. Despite a long history of research, only a few feasibility studies with regards to using real coal fuels have been reported on the DCFC development. In an another attempt to produce electricity in DCFCs, Li et al. [11,12] proposed the various pretreated coal to enhance the electrochemical reactivity in a molten carbonate slurry by increasing high amount of surface oxygen functional groups, as well as to protect the electrodes and the electrolytes through removal of the mineral impurities from coal. 4 M HNO3 treated coal showed the highest power density of 129 mW cm⁻² at 800 °C.

Fig. 1. Detailed description of the modiﬁed cell structure in a direct carbon fuel cell (DCFC).

However, for the molten carbonate electrolyte, the corrosion of cell components and unstable electrolyte conditions of leaking and consumption are a serious challenge for long-term stable operation [4,19,23].

In this study, we try to investigate the comparative performance in a DCFC cell using raw and ash-free coals in order to consider for problems of further feasible demonstration. We have focused on the comprehensive characterization of the physicochemical properties of coals, and subsequently a single cell test is performed under various operating conditions of the cell.

2. Experimental

In order to study the effect of impurities on the DCFC performance, three different solid carbon fuels were applied. A commercial carbon black (CB; ENSACO 350G, Timcal, Switzerland) was selected to serve as a reference substance. Ash-free coal (AFC) was obtained by the thermal extraction of sub-bituminous raw coal (RC) using 1-methylnaphthalene solvent, as described in detail elsewhere [24]. Each of coal substances was separately crushed and sieved to 50 mesh (<300 µm) and then the samples were dried in an oven at 70 °C for 6 h. Prior to the fuel cell test, 0.5 g of each prepared fuel was mixed with 0.5 ml of ethylene glycol (Junsei Chemical Co., Ltd.) in the gel state and the fuel was then directly placed at the anode to provide initial direct physical contact to the porous anode interface.

An optimal design of the conﬁguration of DCFC reactor was developed to exactly evaluate the activity and durability of the fuels under different conditions, as detailed in Fig. 1. The experimental DCFC is composed of an alumina ceramic reactor placed inside a furnace and the operating workstation (NARA Cell-Tech, Korea). Anode-supported SOFC button cells (Ceramic Fuel Cell Power, Korea) were used and the cell consisted of a porous Ni-YSZ anode (660 µm), a dense 8 mol% YSZ electrolyte (25 µm), and a porous lanthanum strontium manganate (LSM) cathode layer (15 µm). The diameter of the whole cell was 30 mm and the active area corresponding to the masked anode surface was 1.0 cm2. Pt mesh (99.9%, 52 mesh woven, Alfa Aesar) and Pt wires (99.99%, 0.5 mm, Alfa Aesar), attached to the anode surface by applying Ag paste (Dotite D-500, Fujikura Kasei, Japan), were used together as a current collector. For use as a ﬂexible compression seal in the prevention of air leakage, two gaskets (Thermiculite 866, USA) in the form of rings were sealed by a sandwich arrangement of the cell. The cell was then mounted onto the end of cathode alumina tube and was pressed between two alumina tubes by adjusting the mechanical spring and micrometer screw of the anode part of the device. A high-temperature ceramic adhesive (Aremco 668, USA) was pasted around the alumina tubes to ensure constant experimental conditions.

Prior to heating up, inert Ar gas (purity 99.999%, 50 ml min⁻¹) was ﬂowed through the anode chamber to remove residual gas. Pure O2 gas (purity 99.99%, 50 ml min⁻¹) was then fed into the cathode. Once the desired operating temperature was achieved, power generation experiments were conducted using the electrochemical fuel cell workstation. To determine the real-time response of the variation in the ohmic contact resistance, the internal resistance of the cell was measured at a high frequency of 1 kHz using an AC-impedance meter (3560 AC m𝞨 HiTester, Hioki, Japan) [25,26].

The surface morphology and the elemental qualitative analysis of the fuel substances were carried out via high-resolution transmission electron microscopy (HR-TEM; JEOL, JEM-2100, Japan), ﬁeld emission scanning electron microscopy (FE-SEM; S-4700, Hitachi, Japan), and energy dispersive X-ray spectroscopy coupled (EDX) with the TEM–EDX (Oxford, INCAx, England) and SEM–EDX (EMAX 7200-H, Horiba, England). The composition of the coal substances was subsequently analyzed via X-ray photoelectron spectroscopy (XPS; Multilab 2000, Thermo VG Scientiﬁc, UK) with a Mg Ka X-ray source (1253.6 eV) at a base pressure of 2x10⁻⁹ Torr and all binding energies were referenced to the carbon signal at 284.6 eV. The pyrolysis and gasiﬁcation of substances were conducted under thermogravimetric analysis (TGA; TGA-50 A, Shimadzu, Japan) under N₂ and air (21.2% O2, balance N₂) at a ﬂow rate of 80 ml min⁻¹. The crystallinity and elemental analysis of the fuels were investigated using X-ray diffraction (XRD; MiniFlex II, Rigaku, Japan).

3. Results and discussion

The surface morphology and grain size of the fuel substances were observed by TEM images, as shown in Fig. 2. Unlike the particle surface of the AFC and CB substances, dispersed black spots on the RC surface were observed. In Fig. 2(b), the inset TEM–EDX quantitative data and the selected area electron diffraction pattern indicated that the surface spots contain the crystalline structured metal oxides of Al2O3 and SiO2.

Table 1 lists the TEM– and SEM–EDX results for different structural compositions of the fuel substances. RC contains a variety of mineral components such as Mg, Al, Si, K, Ca, and Fe, as well as it was determined a low carbon ratio due to the low-rank coal (sub-bituminous coal) with relatively low ﬁxed-carbon yield. On the contrary, the thermal extraction treated coal (AC) shows an increase in the amount of carbon ratio and it is attributable to the removal of inorganic ash matters from the parent coal (RC). In addition, three substances contain an appreciable amount of sulfur elements. It is expected that sulfur poisoning on the DCFC anode can readily be deactivated and degraded [5,15,16]. In addition, the two analyses show different sensitivity of the weight percentage contents of samples because TEM–EDX is more accurate microanalysis than SEM–EDX for the analysis of individual particles [27].

Fig. 2. TEM and HR-TEM images of various substances at different scales: (a) and (b) are raw coal, (c) and (d) are ash-free coal, and (e) and (f) are carbon black. The inset in (b) displays the corresponding selected area electron diffraction (SAED) patterns with only the crystalline structure and EDX analysis around the spots.

Table 1 Structural composition of different substances using TEM– and SEM–EDX analyses.

Table 2 Surface chemical composition analysis of the substances using XPS wide-scan spectra.

Fig. 3. Characterization of the thermal decomposition and oxidation activity using TGA under (a) N₂ and (b) air for raw coal (solid line), ash-free coal (dashed line), and carbon black (dotted line) substances. Insert is the formation of coals as a function of temperature [9,27].

XPS wide-scan spectra in a binding energy range from 0 to 800 eV were measured to identify the overall surface element compositions of the fuel substances and carry out a quantitative analysis. As summarized in Table 2, the main signals are attributed to C 1s peaks (at ca. 285 eV) and O 1s peaks (at ca. 532 eV). In contrast to AFC and CB substances, several inorganic peaks such as N 1s (at ca. 400 eV), Al 2p (at ca. 75 eV), and Si 2p (at ca. 103 eV) were detected in the RC. In addition, it was found that RC had the highest oxygen atomic ratio (26.16 at.%) due to the presence of oxygen- containing surface functional groups such as C--O (at ca. 286.3 eV), C==O (at ca. 287.5 eV), and O--C==O (at ca. 289 eV) forms on the coal, as well as its relatively high content of oxygenated impurities such as Al and Si, which are in good agreement with previous reports [11,12,28] and our previous TEM– and SEM–EDX observations. On the other hand, the results for AFC showed a signiﬁcant decrease in the amount of O 1s peaks on the coal surface. This indicates that surface oxygen functional groups are decreased via the solvent extraction treatment of coals in comparison with other treatment methods [12,28,29]. The order of the magnitude of the O/C ratio is RC > AFC > CB. No other elements were detected from wide-scan spectra at the surface of the fuel substances.

Fig. 4. Comparison of the electrochemical performance of the DCFC for (a) raw coal, (b)ash-free coal, and (c) carbon black fuels at different temperatures.

In order to advance the understanding of the thermal oxidation behavior of fuels, the thermal decomposition and air oxidation behaviors of three fuel substances were investigated by TGA under N₂ and Air. Fig. 3 illustrates the weight loss curves and common thermal decomposition behaviors of the coal in terms of the change in form. As part of TGA under N₂ for RC and AFC, the ﬁrst weight loss was observed below 120 °C, corresponding to the removal of surface moisture. Two distinct stages of weight loss occurred at 300 °C and 800 °C, which were caused by thermal decomposition of the contained volatile matters, such as H2, CO, CO2, and CH4.

Fig. 5. Comparison of the electrochemical data for raw coal (square), ash-free coal (circle), and carbon black (triangle) substances for (a) the maximum power density and (b) high frequency resistance (closed symbol) and series resistance (open symbol) as a function of temperature.

The pyrolytic decomposition behavior of coal is comprised of the following steps: (i) the decomposition of aliphatics and aromatics begins at 250 °C, (ii) the light hydrocarbon gases are released around 400 °C, and (iii) H2 and CO are produced with the formation of char above 700 °C [9,30]. RC was observed to have 15 wt.% higher amounts of volatile matters than that found in AFC.

Fig. 6. Time dependence of the polarization curves for DCFC at a constant current density of 50 mA cm⁻² . The cell was operated at 900 °C and supplied with 0.2 g of (a) raw coal (dotted line), (b) ash-free coal (solid line), and (c) carbon black (dashed line) fuels.

Fig. 7. Comparison of the XRD patterns for the (i) as-prepared fuels and (ii) after- tested residual fuels on the Ni-YSZ anode by using (a) raw coal, (b) ash-free coal, and (c) carbon black fuels.

The difference in the weight of the residual coal substances at 1000 °C is ascribed to the amount of ﬁxed carbon in the coal and it shows a 2-fold difference between RC and AFC, similar to the result of the previous proximate analysis data of Cho et al. [24]. Unlike for the coals, a very slight weight loss for CB over 800 °C was observed due to less contamination of volatile matters and moisture.

In the TGA under oxygen atmosphere, the oxidation of three substances was started at ca. 300 °C, 450 °C, and 550 °C and the order of the onset oxidation temperature was RC < AFC < CB. It was found to considerably enrich the existence of surface functional groups in the coals, which can affect the coal oxidation mechanism due to enlargement of the reactive sites and cleavage of functional groups with weaker bonds [30–32]. In addition, the onset oxidation temperature of CB was much higher than that of the two coals because the enhanced thermal oxidation stability under air could be attributed to its high degree of carbon crystallinity [32]. Despite the thermal oxidation conditions, the remaining coal content of 4.3 wt.% was observed after 550 °C and it seemed to be comprised of residual ashes.

Fig. 8. After long-term stability test by using raw coal fuel, (a) SEM images of Ni-YSZ anode surface, (b) Photo images of residual ashes on the anode surface, and (c) EDX analysis of the anode surface.

To provide a direct comparison of the electrochemical performance for three substances in the DCFC, the polarization curves for the Ni-YSZ anode-supported cell were measured from 750 °C to 925 °C. In Fig. 4, the measured open-circuit voltage (OCV) increases with an increase in the operating temperature and the cell resistance linearly decreases. RC shows the highest OCV value with the highest electrochemical reactivity among the prepared fuels in all temperature ranges, while the polarization curves of AFC and CB become very similar in shape with increasing temperature. In particular, the measured OCVs of three fuels are somewhat less than that of the theoretical OCV of 1.02 V. This result implies that the overall cell potential could be decreased due to the partial electrochemical oxidation of the by-products via the decomposition of coal and the mixed multi-step reactions. Similar phenomena occur in the direct partial oxidation of CH4 in SOFC [33] and in methanol crossover in a direct methanol fuel cell, a so-called mixed potential [34]. Moreover, the thermal decomposition of coal might also be derived in the DCFC anode due to its high operating temperature above 700 °C and the by-products also can partially affect th electrocatalytic performance for DCFC reactions. In the low current density region, i.e. due to activation polarization, the secondary fuels of the CO gas produced by a 2-electron reaction of the carbon electrochemical oxidation and the by-product gases by the decomposition of surface functional groups can probably occupy a large proportion of the DCFC performance compared with the sluggish kinetics of the solid carbon reaction.

In Fig. 5(a), the maximum power density (Pmax) of raw coal linearly increased with increasing operating temperature and the Pmax gap of three substances became smaller because of the similar activation energy of all fuels. To make a clear comparison of three substances in terms of the resistance of the cell at OCV, we measured both the internal resistance and series resistance, as shown in Fig. 5(b). The resistance of a high frequency in AC mode results from both the different ionic conductivity of the solid electrolyte and various current collectors. The resistances of high frequency resistance (HFR) and series resistance (Rs) show similar values at temperatures over 850 °C. The observed HFR values were reduced by up to six-times due to improved current collection through the optimized DCFC reactor system, compared to the results in our previous report [25]. This observation indicates that the conversion of the coal from the solid to the liquid state occurred during the heating procedure. The porous anode might be permeated with the liquid coal, which could enhance the physicochemical contact between the solid substances and the porous anode interface.

However, the predicted ohmic resistance value of 25 µm YSZ is 0.13 𝞨cm2 at 700 °C and 0.06 𝞨cm2 at 800 °C, based on the ionic conductivity of YSZ of 0.0188 S cm⁻¹ at 700 °C and 0.0428 S cm⁻¹ at 800 °C [35]. Therefore, most ohmic losses of the DCFC might still occur in the electrodes/current collector interface. In addition, we observed a signiﬁcantly high value of Rs for CB at 750 °C and the HFR was not detected in the temperature range between 750 °C and 850 °C.

In order to more intensively observe the DCFC durable performance, we investigated long-term stability tests by feeding 0.2 g of three fuels in galvanostatic mode at 50 mA cm⁻² at 900 °C, as shown in Fig. 6. Interestingly, an initial cell voltage was decreased signiﬁcantly from OCV to ca. 0.8 V when the current density was applied on the cell. This initial voltage degradation might be attributed to the deactivation of the anode catalyst due to the poisoning by sulfur impurities and the formation of carbon deposit under open-circuit conditions [16,36–38]. In contrast to the results shown by the polarization curves (activity), the cell performances for AFC and CB, which have relatively small amounts of surface functional groups, were maintained at a high enough level with a small degradation slope until 200 min. On the other hand, RC showed a large degradation in the DCFC performance. Such results indicate that the ash in the coal fuels were not affected by short-term stability, while the ash for the durability test caused signiﬁcant inhibition of the anode reaction surface and pore structures, as a poisoning species. Moreover, the coal located near Si and Al compounds tended to suppress the gasiﬁcation, since the reactant gas was unable to have contact with the carbon surface [39].

Fuel utilization was achieved for the supplied 0.2 g of RC, AFC, and CB fuels, corresponding to 12%, 45%, and 42%, respectively.

More than half of the fuels were therefore not utilized via electrochemical reaction. This may be due to carbon fuel losses caused by secondary gases formation via thermal decomposition and the reverse Boudouard reaction, in reasonable agreement with the previous work of Nürnberger et al. [23].

Fig. 7 shows typical XRD patterns for the as-prepared fuels and after-tested residual fuels on the Ni-YSZ anode surface. The broad amorphous carbon peak at 25° is present in the as-tested RC and AFC fuels, while the carbon peak for the as-tested CB is observed at 26.5°. Another weak broadened peak at around 44° is attributed to the graphite crystal faces reﬂection [32]. In contrast with as-tested AFC and CB fuels, various mineral diffraction peaks corresponding to SiO2 (JCPDS No. 87-0703), Al2O3 (JCPDS No. 81-2266), Fe2O3 (JCPDS No. 87-1165), and CaO (JCPDS No. 82-1691) are appeared by using RC fuel. This XRD result is in good agreement with previous observations of RC substrates.

In addition, Fig. 8 presents SEM and EDX analysis for after-tested Ni-YSZ anode using RC fuel and photo image of the residual fuels on the anode. The EDX data conﬁrmed that the presence of the impurities of Al and Si was deposited on the anode surface and a respectable amount of ash substrates was observed (see Fig. 8b). These results suggest that the surface active sites of anode could be physically blocked by the presence of the ashes. Therefore, the most important factors affecting the long-term stability of direct coal-DCFC could be the amount of ash.

4. Conclusion

We have investigated the inﬂuence of the physicochemical characteristics of raw and ash-free coals on the electrochemical activity and durability performance in a direct carbon fuel cell (DCFC) and compared them with commercial carbon black. In the raw coal, the signiﬁcant ash contamination and abundant oxygen-containing surface functional groups were detected by various analyses. In ash- free coal, in contrast, the inorganic contaminants were successfully removed by the thermal extraction method. During polarization curve measurements, the electrocatalytic oxidation of coal fuels in the DCFC was not compromised by the existence of ashes. However, based on the long-term durability test it has been proposed that the DCFC performance is more signiﬁcantly affected by the presence of the ash due to poisoning of the anode catalyst. The enhanced durability of the DCFC performance was accomplished using ash-free coal due to the eliminate ash effect and it was operated for 25 h g⁻¹ at 900 °C with a corresponding fuel utilization of 45%. Finally, it is anticipated that an advanced treatment of coal for sulfur removal technology is quite necessary to realize a practical coal-DCFC system.

Source: HyungKuk Jua,1, Jiyoung Eoma,1, Jae Kwang Leeb, Hokyung Choic, Tak-Hyoung Limd,Rak-Hyun Songd, Jaeyoung Leea,b,∗
a School of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST)
b Ertl Center for Electrochemistry and Catalysis, Research Institute for Solar and Sustainable Energies (RISE), South Korea
c Clean Fuel Center, Korea Institute of Energy Research (KIER)

The 10 largest coal producers and exporters in Indonesia:

Click Here for Top World Clean Coal Technologies Contractors

Tuesday, March 17, 2020

Investigation of Performance of Raw and Ash-free Coal in the Operation of a Direct Carbon Fuel Cell (DCFC)