Review of Fuels for Direct Carbon Fuel Cells - Energy & Fuels (ACS

Review pubs.acs.org/EF

Review of Fuels for Direct Carbon Fuel Cells Adam C. Rady,† Sarbjit Giddey,‡ Sukhvinder P. S. Badwal,‡ Bradley P. Ladewig,† and Sankar Bhattacharya*,† †

Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia Commonwealth Scientific and Industrial Research Organisation (CSIRO) Energy Technology and Advanced Coal Technology Portfolio, Private Bag 33, Clayton South, Victoria 3169, Australia

‡

ABSTRACT: In this paper, the current status of direct carbon fuel cell (DCFC) technology has been reviewed. Recent promising advances in the design of fuel cells has resulted in a reprisal of research into the DCFC technology. As a result, more is understood about the roles of species and mechanisms that govern the performance of DCFC systems. Of particular interest to industry and researchers are the direct carbon molten carbonate fuel cell (MCFC) and solid oxide fuel cell (SOFC) arrangements, with the bulk of research articles and large-scale investment focused on these DCFC types. However, the variety of fuels used and trialled within these fuel cells is limited. This is especially true for the SOFC arrangement, with the overwhelming fuel of choice for researchers being carbon black and light gases for industry. The application of more readily available and cheaper fuels in this type of DCFC is unassessed. This review addresses this gap in the literature by reviewing all fuels tested in direct carbon MCFC and SOFC systems, in particular critically evaluating low-rank coals and biomass, among other alternative fuels.

1. INTRODUCTION Fuel cells are widely considered as a clean, alternative means of electricity generation, because of their high efficiencies and predictable, low emissions. Direct carbon fuel cells (DCFCs) operate on the same electrochemical principles as conventional fuel cells but rely on a solid carbonaceous material as the fuel source. Their theoretical energy efficiency is marginally greater than 100% at 600 °C for eq 3,1,2 which can be calculated as theoretical efficiency = [ΔG°(T )/ΔH °std ] × 100

electrolyte and reported high efficiency, but his system was later found to have achieved an efficiency of only 8%. A number of less successful electrolyte mixtures were proposed before Baur,5 in 1910, investigated a molten sodium hydroxide fuel cell, which included an air electrode and ran on various fuels, including coal and sugar. Baur and co-workers continued to develop and refine their fuel cells and, in 1921, arrived on a design incorporating molten carbonate electrolyte adsorbed in a porous ceramic. The fuel cell operated at 800 °C and used an iron rod anode and an iron oxide cathode. Early work on molten salt electrolytes paved the way for the modern molten carbonate fuel cell (MCFC). During and after the 1980s, industry involvement in MCFC development expanded rapidly and companies, such as Mitsubishi, Hitachi, Toshiba, Fuji, IHI, and Siemens, all made notable contributions to large-scale and long-life operation of MCFCs. However, for MCFCs to be a competitive source of electricity, the cost must reach a limit of around €1.500 kW−1.4 The solid oxide fuel cell (SOFC) was developed shortly after the MCFC. The yttrium-stabilized zirconium (YSZ) material used as the electrolyte in SOFCs today was first developed by Nernst in 1897 for light generation. A student of Nernst, Schottky, realized the potential of ceramic as an electrolyte for a fuel cell (the SOFC) in 1935. Baur and Pries conducted much of the pioneering work of the SOFC in the 3 years between 1937 and 1939, using Nernst’s YSZ electrolyte as well as other solid electrolytes, such as clay. Their cells ran on gaseous fuels of hydrogen, town gas, and carbon monoxide. Industry interest in the SOFC has been greater and more sustained than that of the MCFC, beginning with Westinghouse Electric Corp., who developed the tubular SOFC in 1958.

(1) 1

The actual efficiency of DCFCs lie between 80 and 95% and is a product of the theoretical efficiency, use efficiency, and voltage efficiency. The actual efficiency is equivalent to the electrical energy output of a cell divided by the combustion higher heating value (HHV) of the fuel used. DCFCs are unique in that they are the only means by which electrical energy can be attained directly from solid carbon fuel or the oxidation of solid carbonaceous material without a reforming process. Potential sources of carbon fuel include coal, coke, graphite, municipal solid waste (MSW), natural gas, and other carbonaceous material, including biomass, such as rice hulls, nut shells, corn husks, grass, and woods. These fuels are cheap, abundant, and readily available, and unlike other fuel cells, widespread implementation of DCFCs would not require a radical change to infrastructure to accommodate a hydrogenor gas-based economy. The first attempted use of carbon directly in a fuel cell was in 1855 by Antoine César Becquerel, who used platinum and carbon electrodes immersed in a molten sodium nitrate solution.3,4 In the early stages of development of the DCFC, the main variable was the molten electrolyte, and several patents were put forward when a new successful candidate returned positive results. One such early candidate was that of molten hydroxide, which was patented in 1896 by William W. Jacques.4 Jacques employed coal as the fuel with air injected into the © 2012 American Chemical Society

Received: October 31, 2011 Revised: January 30, 2012 Published: January 31, 2012 1471

dx.doi.org/10.1021/ef201694y | Energy Fuels 2012, 26, 1471−1488

Energy & Fuels

Review

Section 5 will introduce a new DCFC in the HDCFC. There are a large variety of direct carbon fuel cell types and designs. The focus of this paper is on the effect of the fuel type and fuel characteristics on cell performance, for a given cell arrangement. It should be noted that there is an emphasis on molten carbonate fuel cells and solid oxide fuel cells (in particular the latter). Sections 6 and 7 explore properties of coals used in DCFC systems and the effect of impurities, physical nature of the coals, and treatment options on the performance of DCFC systems. Section 8 investigates analytical techniques used by researchers on fuels used in direct carbon MCFCs and SOFCs, and section 9 summarizes the performance of these and other fuels in DCFCs.

In 1987, Osaka Gas and Tokyo Gas operated a 3 kW stack of cells for 15 000 h on natural gas. Since realizing the industrial applications of SOFCs on gaseous fuels, many other companies have invested in SOFC development, including Fuji Electric Corp., Siemens, Doriner, Mitsubishi Heavy Industries, RollsRoyce, and Ceramic Fuel Cells Limited. Thanks largely to industry investment, the modern DCFC can be either mobile or a stationary conglomerate of units close to the fuel source. As of 2008, a new DCFC, the hybrid direct carbon fuel cell (HDCFC), has entered the scene.6 This is just one of many recent advances in DCFC technology that has enjoyed a reprisal in interest and has seen many new researchers enter the field. Much of this renewed interest is due to recent advances is DCFC design, such as the work of Cherepy and co-workers at the Lawrence Livermore National Laboratory.7 The authors dispersed carbon particulates in molten carbonate, improving reaction rates, and shifted research away from use of carbon rods in direct carbon MCFC systems. Also, for direct carbon SOFC systems, the two-compartment high-temperature fuel cell from Gur and Huggins8 paved the way for future detached-type direct carbon SOFC cells with independent temperature control of both the anodic fuel chamber and the electrolyte. The focus of this review will be on fuels for direct carbon fuel cells, in particular those for direct carbon MCFCs and SOFCs. 1.1. DCFC Reaction Basics. DCFC reactions occur at temperatures below that of traditional coal combustion or gasification. However, the operating temperatures are still high, typically in the range of 600−900 °C and vary depending upon electrolyte and cell type/design. In all DCFCs, a side reaction known as the reverse Boudouard reaction occurs, where the operating conditions allow. This reaction is particularly prevalent in high-temperature fuel cells (such as direct carbon MCFCs and SOFCs) because the reverse Boudouard reaction is thermodynamically favored at temperatures above 700 °C.8,9 The Boudouard reaction has great and unique implications for the performance of both the direct carbon MCFC and SOFC and, hence, is an extremely important consideration to the fuel cells of interest in this paper. The reverse Boudouard reaction is also known as the gasification reaction and is given below as eq 2. C + CO2 ↔ 2CO

2. DIRECT CARBON MCFC The MCFC and direct carbon MCFC operate using molten carbonate salt electrolytes, which are composed of K2CO3, Li2CO3, and Na2CO3. These carbonates are present in the electrolyte as either binary eutectics of (Li/K)2CO3 or (Li/Na)2CO3 or ternary eutectics of (Li/K/Na)2CO3. The celloperating temperature required to achieve high conductivity of molten salt is 600−800 °C. Because of this high operating temperature, cell components may be exposed to a highly corrosive environment,10 limiting material options,2,10,11 and high pretreatment costs are associated with fuels that contain sulfur impurities. Because of the corrosive operating environment, rather exotic materials must be employed, such as Inconel, Hastelloy, and Kanthal-A.1 The high corrosivity of the electrolyte plagued the development of commercial MCFC stacks. In the decade spanning 1996−2006, however, cell life was improved from just a few months to up to 2 years.12 This advancement was largely thanks to incremental improvements to corrosivity resistance of critical cell components. There have been various designs of the MCFC throughout its history, notably focused on the application of coal within the cell itself, ranging from solid anodes to the more modern coal/ electrolyte slurry. The early design of a solid carbon-based anode was plagued by various shortcomings, such as the leakage of impurities (e.g., ash) into the electrolyte and mechanical instability of coal-based electrodes. Other key drawbacks of the early cell designs as identified by Cherepy and co-workers13 include low reaction rates as well as the economical and logistical impracticalities of carbon electrodes. The majority of these issues were overcome to a large degree in a novel design by Vutetakis and co-workers.14 This new design involved suspending coal particulates within the molten carbonate electrolyte and realized increased reaction rates. This work itself was developed by combining the work on coal slurry electrolysis by Coughlin and Farooque15−18 and coal-powered molten salt fuel cells. Cherepy and co-workers further researched the effect of various solid fuel sources and their respective physical nature (particle size, surface area, and crystallinity) on the performance of this arrangement. The authors believe that a specifically designed cell, which operated under constant polarization, should achieve a fuel use at or about 100%. Li et al.19 believe that this fuel cell arrangement by Cherepy and co-workers showed potential for a continuous process and, hence, commercialization. Several authors13,20 also realized a relatively pure product of CO2 at operating temperatures of 700−900 °C, despite some earlier researchers predicting a high concentration of CO via reaction 2 because of the high temperatures.

(2)

The overall cell reaction for all DCFC arrangements simplifies to the familiar oxidation of carbon in eq 3. C + O2 → CO2

(3)

However, the oxidation is electrochemical in nature and involves several reaction steps involving the electrolyte for non-solid oxide arrangements. Understanding the reaction mechanisms involved in all conversion steps from solid carbonaceous matter to carbon dioxide is crucial to pairing the most appropriate fuel with a given DCFC system. It also allows us to determine which fuel preparation techniques will benefit fuel use, cell performance, and cell life. This paper has been written to address recent advances in the design of and knowledge surrounding DCFCs because there are no reviews on the current status of DCFC technology. This paper also focuses on appropriate fuel sources, their composition, and preparation of these fuels for use in DCFCs, as well as what has been reported in the literature. In the following three sections (sections 2−4), three prominent DCFC types and current arrangements will be introduced along with the current understanding of their reaction mechanisms. 1472


Energy & Fuels

Review

3. DIRECT CARBON MOLTEN HYDROXIDE FUEL CELL (MHFC) The MHFC has a low operating temperature of around 650 °C. The direct carbon MHFC has a similar arrangement to the MCFC, with the main advantage of the molten sodium hydroxide electrolyte having a greater ionic conductivity than that of molten sodium carbonate.1 Molten NaOH also has lower overpotentials and a greater oxidation rate than molten Na2CO3. These factors allow for the MHFC to operate at lower temperatures, allowing for cheaper materials to be used for cell casing and cathodes, such as 300 series stainless steel and ultralow carbon steel.1 However, carbonate formation has been an issue preventing the industrial application of molten NaOH as a DCFC electrolyte because this design was first conceived in 1896 by William W. Jacques. The reaction mechanism for carbonate formation as proposed by Goret and Tremillon22 involved two reaction steps given below as eqs 7 and 8.

The simple half-cell reactions are not however disputed and are given as eqs 4 and 5. Anode reaction: C + 2CO32 − → 3CO2 + 4e−

(4)

Cathode reaction: C + 2CO2 + 4e− → 2CO32 −

(5)

A more detailed mechanism of the anodic electrochemical oxidation of coal as proposed by Cooper and colleagues,13 adapted from Vetter’s original mechanism as cited in ref 1, is illustrated in Figure 1. The mechanism is similar to the Hall

(7)

C + 3O2 − → CO32 − + 4e−

(8)

Reaction 7 is a fast chemical reaction, whereas reaction 8 is a rate-limiting electrochemical reaction. A mitigating technique for the prevention of carbonate formation has since been proposed by Zecevic et al.23 This technique involves increasing the concentration of water in the electrolyte solution to drive the equilibrium of eq 7 to the left, reducing the amount of O2− in solution, which is an intermediate product in the formation of carbonate. Increasing the concentration of water in the electrolyte solution had several additional benefits. Namely, an increase in the ionic conductivity of the electrolyte/fuel solution and reduced corrosion rates of iron, nickel, and chromium (key metals used in the cathode and cell casings). High corrosivity, as for MCFCs, is still an issue for MHFCs at operating temperatures of 400−800 °C.10 While Ni and Ni foams are prominent cathodic materials, an alternative in Fe2Ti shows good resistance to corrosion as a cathodic and cell casing applications.24 The half-cell reactions are given as eqs 9 and 10. Anode reaction:

Figure 1. Reaction mechanism for electrochemical oxidation of carbon in molten carbonate.

C + 4OH− → CO2 + 2H2O + 4e−

process and takes the form of a five-step reaction. Minor variations of this mechanism exist in the literature, such as that by Li and co-workers,19 which includes the necessary decomposition of a carbonate ion to form an oxygen ion as a preliminary step. Reaction 2 (the reverse Boudouard reaction) also occurs either when the carbon in the anode compartment is not polarized (for example, during cell standby, open circuit operation, or remote regions of the anode that experience low current flow) or when the carbon is not in physical contact with the anode.13 Carbon consumed via the reverse Boudouard reaction has been described as “carbon fuel loss” because it prevents a carbon atom from taking part in a reaction with the molten carbonate electrolyte. It is however a necessary reaction in the direct carbon SOFC. An alternative electrochemical mechanism has been suggested for CO production at high temperatures (800 °C) in molten carbonate.21 C(s) + CO32 − → CO2(g) + CO(g) + 2e−

6OH− ↔ 3O2 − + 3H2O

(9)

Cathode reaction: O2 + 2H2O + 4e− → 4OH−

(10)

4. DIRECT CARBON SOFC There are three main classes of direct carbon solid oxide fuel cells (direct carbon SOFCs), and their classification depends upon the method of contacting between the fuel and the anode. The three classes are detached type, physical contact type, and carbon-deposited type [also referred to as the rechargeable direct carbon fuel cell (RDCFC)] SOFC. Other modern arrangements of direct carbon SOFCs also exist, such as gasificationdriven SOFCs, fluidized-bed SOFCs (FB-DCFC),25,26 and the hybrid (solid oxide/molten carbonate) DCFC27 (HDCFC; see section 5). This section will cover only the three main classes, along with an emerging variation of the detached type in a gasification-driven SOFC. The SOFC is a high-temperature fuel cell, and therefore, normal operating conditions involve temperatures around

(6) 1473


Energy & Fuels

Review

800−1000 °C required to achieve desirable ionic conductivity of the electrolyte. The cell performance is also heavily dependent upon the kinetics of the gas-phase reactions, in particular the gasification of coal or the Boudouard reaction when gasified with CO2. Therefore, an ideal coal preparation would be that of a carbon fuel that is easily gasified. This is opposed to the direct carbon MCFC, where surface oxygen functional groups play an important role (see section 7.2). However, the SOFC is flexible in terms of fuel sources and can operate on CO, H2, or light hydrocarbon gases. 4.1. Common Design Features of the Direct Carbon SOFC. While there are numerous types and arrangements of the SOFC, many design features, such as materials used for cell components, are rather universal. This section will outline a few well-established design characteristics of the modern direct carbon SOFC. The YSZ electrolyte allows for the direct carbon SOFC to operate at high temperatures of 850−1000 °C, higher than other DCFC types. The high temperatures are required to maintain high ionic conductivity of YSZ, which is very sensitive to temperature change. Current research efforts in materials engineering are focused on reducing the temperature requirements for high ionic conductivity to allow for lower temperature cell operation,28−31 as well as improved resistances to catalyst poisoning.32 The high-temperature operation also negates the need for expensive platinum catalysts, which also have issues with CO poisoning. There is also no consumption of or reaction with the electrolyte or risk of electrolyte leakage. The ceramic-based electrolyte and electrodes means that there is little risk of corrosion of crucial cell components at the high temperatures. However, the high operating temperatures may result in mechanical instability of the cell, enhanced corrosion issues with cell casing and longer start-up times. Sulfur and other impurities must first be removed from the fuel prior to entering the SOFC to maximize the lifetime of the cell. The electrolyte material for the SOFC is a denser ceramic than the anode, which is porous. Common choices for ceramic include 8% mol Yttrium oxide in YSZ (Y8SZ), scandia-stabilized zirconia (ScSZ), 9% mol Sc2O3 (9ScSZ), and gadolinium-doped ceria (GDC). The two foremost desirable properties of the ceramic electrolyte are low electrical conductivity, to minimize current leakage away from the anode, and high ionic conductivity. High temperatures allow for acceptable O2− ion transport kinetics. Currently, the lower limit for operating temperature is around 600 °C, below which the resistance for ion transport is too high. Achieving current research targets of 500 °C operating temperature would result in a marked increase in available materials for use in the SOFC. The physical design consideration of the electrolyte is its thickness. Sufficient electrolyte thickness is necessary to prevent current leakage, which would effectively short-circuit the cell. However, electrolyte thickness dictates the distance that O2− ions have to travel. This coupled with the inherent electrolyte resistance results in ion transport resistance. The microstructure of the electrolyte grains also plays a key role in resistivity, with some structure patterns more resistive to ion transport than others. It is as though a columnar grain structure may provide lower resistance by eliminating grain boundary contributions,33,34 whereas a microtubular design may enhance thermal mechanical properties of the cell.28 The YSZ electrolyte may react with modern cathode materials, such as lanthanum strontium cobalt ferrite (LSCF). This can be

negated by separating the electrolyte and cathode by a thin GDC film.35 The cathodic material in a SOFC is a thin porous ceramic layer, which, for electrode purposes, must be conductive of electrons and preferentially conductive of ions. The common material choice for the cathode is lanthanum strontium manganite (LSM). LSM is a popular candidate because of its mechanical suitability with YSZ in that both materials have similar thermal expansion properties, minimizing stresses at the junction of the high-temperature fuel cell. The two materials are also chemically stable. These factors extend the lifetime of the cell. However, LSM has a low ionic conductivity below 800 °C. For cathodic materials with poor ionic conductivity (such as LSM in intermediate−low-temperature applications), the reaction zone is confined to the “triple-phase boundary” (TPB). This “reaction zone” is where the electrochemical conversion of oxygen in air to O2− takes place and can occur beyond the TPB and into the cathode, where there is good ionic conductivity. The anodic material in a SOFC is a highly porous ceramic, which, similar to the cathode, must be conductive of electrons. However, it is also necessary for the anode to be conductive of ions. The common arrangement is a “cermet” (a composite of a ceramic, “cer”, and a metal, “met”) composed of the electrolyte ceramic material and Ni. YSZ is a popular candidate for the ceramic component because grain growth of the metal component (Ni) is retarded by YSZ. Oxidation occurs at the anode, which results in generation of electrons, with the loss of electrons from the oxygen ion, as well as generation of heat from the production of CO2. As such, the anode may require cooling in some cases. This cooling may be satisfied in situ, where coal or other carbonaceous material is used as the fuel source via the gasification process. Gasification of the carbon with either H2O or CO2 is an endothermic process and, therefore, acts to cool the anode by supplying it with cooler gas. The anode itself acts as a catalyst for steam reforming. 4.2. Deposited-Type Direct Carbon SOFC (RDCFC). The carbon deposited-type direct carbon SOFC (also known as RDCFC) is a unique DCFC in that it requires the supply of a gaseous hydrocarbon (usually light hydrocarbons) to the anode as the fuel. Therefore, the application of coal-based fuels to this cell type is minimal. It is possible that light gases from coal and biomass pyrolysis may be used in this cell; however, they have to be free of tar, which is detrimental to the longevity of cell operation, and some key cell components, such as the anode. The deposited carbon is supplied via thermal decomposition of the gaseous fuel onto the porous anode surface. Initially, the deposited carbon is electrochemically oxidized by oxygen ions at the anode/electrolyte TPB as per eq 11. Once CO and CO2 are formed, the kinetics of reactions 12 and 2 (reproduced here for clarity) dominate and become the focus of the fuel cell operation.36 C + O2 − → CO/CO2 + 2/4e−

(11)

CO + O2 − → CO2 + 2e−

(12)

C + CO2 ↔ 2CO

(2)

37

Li and co-workers analyzed the performance of a button cell using CH4 deposition. The results of the authors show significant activation polarization at high voltages (above 1 V; see Figure 2). These results support the mechanism above because activation 1474


Energy & Fuels

Review

performance. It has also been suggested that previous research into SOFC/liquid electrolyte hybrid cells was an attempt to increase the contact between the carbon and electrolyte based on the belief that reaction 13 is the reaction mechanism.38 However, the trends of the later results by Li and coworkers37 (summarized in Tables 1 and 2) do not demonstrate Table 1. Summary of Cell Performance Trends for Physical Contact-Type Direct Carbon SOFC37

Figure 2. Cell voltage and current density for a deposited-type direct carbon SOFC with different times of open circuit operation and significant activation polarization at high voltages (circled), adapted from Li and co-authors.37

(13)

CO + O2 − → CO2 + 2e−

(12)

C + CO2 ↔ 2CO

Ar (case II) Ar flow rate ↑ cell performance ↓ CO2 (case II)

CO2 flow rate ↑ cell performance ↑

CO2 flow rate ↑ cell performance ↑

Table 2. Summary of Cell Performance Trends for Detached-Type Direct Contact SOFC37

polarization is linked to the direct electrochemical oxidation of carbon, which requires an overpotential.37 4.3. Physical Contact-Type Direct Carbon SOFC. As the name suggests, the fuel is in physical contact with the anode. In research-based systems, this is commonly achieved by sandwiching the carbon fuel to the cell, often with a thin, porous plate or felt9,37 or between the anode and the electrolyte.10 In other cases, the fuel is housed in a tubular SOFC, where it is allowed to rest against the inner anode surface of the tube.38 There is increasing cause for uncertainty over whether or not physical contact is beneficial, because of a shift in thinking over fundamental and controlling reaction mechanisms. Whether or not direct oxidation of carbon occurs at the anode is in dispute. What is agreed upon recently however by most researchers is that, if it does occur, it is not the dominant mechanism for consumption of carbon.36,39 The reaction mechanisms proposed by authors are presented as case-specific and are dependent upon not only the physical cell arrangement but also the atmosphere in which the carbon fuel is consumed. Most notably, the comparison between mechanisms for reactive atmospheres of CO/CO2 and inert atmospheres are proposed. What reactions occur in the presence of sufficient residual active gases (CO/CO2) or recycled anode gas is a hot topic. It is thought by some that the direct electrochemical oxidation of carbon in solid carbonaceous fuels, as per eq 13, is possible,9,39,40 where carbon is in intimate contact with the anode.39 It is largely accepted however that, as with RDCFCs, irrespective of whether or not reaction 13 takes place, eqs 12 and 2 are the dominant reactions.39 C + 2O2 − → CO2 + 4e−

Ar (case I) Ar flow rate ↑ cell performance ↓ CO2 (case I)

Ar (case I)

Ar (case II)

Ar flow rate ↑ cell performance ↓ CO2 (case I)

Ar flow rate ↑ cell performance ↑ CO2 (case II)

CO2 flow rate ↑ cell performance ↓

CO2 flow rate ↑ cell performance ↓

that the exact same reaction mechanisms are taking place as those in the detached-type direct carbon SOFC, contrary to the interpretation of the authors. The trends in Table 1 appear to be case-independent (for a given atmosphere), as opposed to those observed for the detached type (Table 2). The polarization curves were produced after case I, cell attached to an open circuit for 30 min, and case II, cell discharged at 0.5 V for 5 min. Figure 3 shows how CO acts as the fuel in the physical contacttype direct carbon SOFC, where reaction eqs 12 and 2 are

(2)

Figure 3. Depiction of the fuel oxidation in a physical contact-type direct carbon SOFC, as proposed by Nakagawa and Ishida.41

Some authors believe that reaction 13 can only take place at the anode/electrolyte TPB, and for solid carbonaceous fuels, this is an unfeasible mechanism, irrespective of intimate contact of the fuel with the electrolyte.36 Li et al. also claim that electrochemical oxidation of carbon at the anode does not occur, and therefore, the anodic reaction mechanisms are the same as those for detached-type direct carbon SOFCs. If this statement is correct, then there are associated implications for the cell design, in that improving contact between fuel and anode will not enhance cell

dominant, as accepted by most authors. The dashed arrows indicate two options for the CO2 and CO products: to participate in further reactions or to escape the anode as product gas. Li and co-authors25 conducted tests in a fluidized-bed reactor, where carbon particles were fluidized, allowing for contact with the anode. The experiments were conducted in a relatively inert helium fluidizing environment; however, the exact nature of the 1475


Energy & Fuels

Review

fluidizing environment was unknown because of reported leakage problems. According to the authors, there was no direct electrochemical oxidization of the carbon. The reactions they reported were

2O2 − ↔ O2 + 4e−

(14)

C + O2 → CO2

(3)

2C + O2 → 2CO

(15)

CO + O2 − → CO2 + 2e−

(12)

C + CO2 ↔ 2CO

(2) Figure 4. Depiction of the fuel oxidation in a detached-type direct carbon SOFC, as proposed by Nakagawa and Ishida.41

However, the authors could only speculate on the extent of reaction 2 and named the first four reactions as the key reactions. 4.4. Detached-Type Direct Carbon SOFC. The detachedtype direct carbon SOFC involves the fuel being physically removed from the anode. This system offers the greatest flexibility in cell design, and there are many variations, most notably the gasification-driven SOFC. As with physical contact-type direct carbon SOFCs, research into the effect of the anodic cell atmosphere on reaction mechanisms has been conducted to help clarify the process. Tang and Liu38 reported that catalyzing the Boudouard reaction in the presence of sufficient residual active gases (CO/ CO2) markedly improved the performance of the cell. They concluded that the formation of CO from CO2 through the Boudouard reaction was a pivotal step in the function of the cell. The authors used their findings to support the mechanism proposed by Nakagawa and Ishida,41 which incorporated the following dual-reaction mechanism for the oxidation of carbon in the solid oxide fuel cell: CO + O2 − → CO2 + 2e−

feasible industrial-scale practice. Under these conditions, Li and co-workers36 proposed the following mechanism:

2O2 − ↔ O2 + 4e− C + O2 → 2CO

(3)

2C + O2 ↔ 2CO

(15)

The above reaction mechanism, supported by Gur and Huggins,8 is most likely associated with very poor kinetics because the reactant gases are not present in high concentrations. However, this mechanism does not require the intimate contact between the carbon and the anode and, therefore, occurs in the “detached” reactor types, as employed by Gur and Huggins.8 The mechanism is similar to that proposed by Li and co-authors25 in that O2− reacts with itself to form O2 at the anode when the concentration of other active gases is low. However, it is unlikely that the mechanism proposed by Gur and Huggins is complete because it is widely accepted now that CO can be electrochemically oxidized to CO2. Li and co-workers37 investigated the effect of changing the flow rates of both an inert gas in Ar and a reactive gas in CO2 through the fuel sample and anode. The results are summarized in Table 2. Of interest is the relationship between the cell performance and the supply and removal rate of CO2 from the fuel sample and the use of this to justify the proposed reaction mechanisms. Adjusting the flow rates influenced the concentration of CO2 in and around the fuel, which has implications for the Boudouard reaction. The fact that there is a switch in cell performance between high Ar flow rates and low Ar flow rates makes it hard to pinpoint the exact reaction mechanism. The authors argue that, because of the varying results between cases for the Ar-rich anode gas and the CO2-rich anode gas, the dominant reaction mechanisms differ depending upon the subjected atmosphere of the fuel and anode. This is not necessarily the case. Because low CO2 flow rates for both cases resulted in the greatest cell performance, it is difficult to justify the Boudouard reaction as the dominant reaction mechanism or at least the limiting reaction. One possibility is that the CO2 flow is diluting CO around the anode, necessary for the electrochemical reaction, more than the CO2 flow enhances the generation of CO. There would therefore be an ideal flow rate and recycle of CO2 to the gasification chamber to maximize cell performance and fuel use.

(12)

C + CO2 ↔ 2CO

(14)

(2) 37

Equation 12 occurs at the anode three-phase boundary. The interfacial, polarization and activation resistances govern the kinetics of this reaction. Equation 2 occurs at the carbonaceous fuel source, with concentration resistance governing the kinetics of the reaction. Because both reactions depend upon each other for reactants, the kinetics of both reactions are critical to the overall reaction rate. The cell performance will be greatly compromised if one of the reactions is retarded. Therefore, it is necessary to catalyze both reactions. Figure 4 shows fuel oxidation in a detached-type direct carbon SOFC and how CO acts as the fuel in the cell where reactions 12 and 2 are dominant. There is an option to recycle a fraction of the anodic gas back to the detached carbon fuel source. Note however that this is not necessarily a pure stream of CO2 and will likely contain varying concentrations of unconverted CO. This relative concentration depends upon factors such as anodic gas residence time in the anode chamber, temperature, and reactivity of the carbonaceous fuel. In the absence of sufficient residual active gases (CO/CO2), product gases are not present in a significant concentration in the anode compartment. This is the case when the anode chamber is flushed with an inert gas and the anodic gas is not recycled back to the anode. Such arrangements are for the benefit of understanding reaction mechanisms and are not a 1476


Energy & Fuels

Review

4.5. Gasification-Driven SOFC. While it is largely accepted that the gasification of carbonaceous fuel is a necessary mechanism for the operation of the detached-type direct carbon SOFC, the gasification-driven SOFC is so named because it often incorporates an external gasifier. There are two classes of these gasification-driven fuel cells: (1) single-step direct approach (unconstrained, in situ gasification) and (2) multi-step indirect approach (constrained, separate gasification) with (a) indirect gasification (IDG) or (b) autothermal gasification (ATG). Lee and co-workers42 investigated the effect of different fuel sources and different gasification feeds (H2O and CO2) on the thermal efficiencies of ATG and IDG systems. They reported (for their simulations) that the work output differed by less than 1%, despite changes to gas composition. Their model, as the authors acknowledge, did however neglect certain inherent fuel cell irreversibilities contributing to cell resistances, as well as the concentration and thermochemical driving forces. The authors noted that gas flow to the gasifier also affords some operational benefits. The solid fuel may be delivered to the gasifier entrained flow within CO2/H2O/O2 feed. Agitation via fluidization of carbonaceous matter from the gaseous feed also improves mixing as well as enhancing segregation or removal of ash. In the case of single-step gasification, the gasification occurs in the same chamber as the anode. This is a common arrangement for experimental research and is likened to the detachedtype direct carbon SOFC. An example of a single-step gasification system is given as Figure 5, which is only constrained by

The fuel cell runs on the dominant species supplied by the gasification process (CO if CO2-fed gasification and H2 if H2Ofed gasification). An example of an IDG setup is depicted in Figure 6.

Figure 6. IDG.

The gasifier requires heat input to sustain the endothermic gasification reaction. There are several potential sources of heat: (1) addition of oxygen into the gasifier (see the ATG system), (2) recycling of flue gas back to the gasifier (in a CO2-fed gasification system), (3) heat exchange of flue gas with steam (in a H2O-fed gasification system), and (4) combustion of syngas (a side stream) and transfer of heat through heat coils or pipes within the gasifier or upstream of pyrolysis syngas. Considerations for the gasifier-operating temperature include the enhancement of the gasification reaction rate with an increased temperature or by catalytic means, as well as the impact on the downstream SOFC operating temperature. Considerations for the SOFC operating temperature include enhancement of electrolyte ionic conductivity with an increasing temperature and temperature dependence of the Boudouard reaction. High operating temperatures are ideal for reaction kinetics; however, low temperatures are desired for material selection. Unlike single-step gasification systems, the indirect approach allows for the possibility of hot gas cleanup after gasification, before reaching the anodic compartment. Single-step gasification systems must employ anodic barriers or sulfur-resistant anodes if sulfur levels are too high. Also, ash is removed in a prior stage and physically removed from the fuel cell. ATG systems are similar arrangement to IDG systems but with the addition of an air separation unit (ASU) and introduction of another reactive species, oxygen, into the gasification step. The oxygen is supplied to the fuel to generate heat necessary for the endothermic gasification reaction, through an exothermic reaction in the partial combustion of coal. The coal exposed to O2 is mostly only partially oxidized to CO because of the low concentration of O2 in the gasification atmosphere. An example of an ATG setup is depicted in Figure 7. An ASU is often employed to avoid nitrogen diluting the gasification and anodic gas, as well as to reduce additional heat required to heat feed air.

Figure 5. Single-step (in situ) gasification.

the additional heat exchanger (a non-essential process, used only to preheat the incoming air, minimizing heat loss from the system). This single-step gasification arrangement is limited by several design constraints. First, ash produced in the same chamber as the fuel cell may not be simple to remove, and some will undoubtedly interact with the anode. Second, independent control of the exothermic (electrochemical oxidation of CO) and endothermic (gasification) reactions is impossible with in situ gasification. This makes temperature control of the fuel cell difficult. This arrangement may be appropriate for batch processes and laboratory-based exploratory work but would not be ideal for industrial application. For IDG, the gasification step occurs in a unit separate from the fuel cell. The gasification unit is located upstream of the fuel cell and supplies the syngas to the anode compartment. 1477


Energy & Fuels

Review

only an issue at low temperatures (below that of 700 °C) for cells that are not gasification-promoted (without the addition of a CO2 or H2O feed into the anode chamber). At higher temperatures, there is minimal difference between physical contact-type and detached-type direct carbon SOFC systems where gasification is promoted, as discussed in the previous section of this review. The benefit of the HDCFC cell over that of direct carbon SOFC cells, with respect to the temperature, was evaluated by Nabae and co-workers,44 and their findings are reproduced in Figure 8. The “cell” in Figure 8 is essentially a

Figure 7. ATG.

5. HDCFC The HDCFC is a “hybrid” of key elements of both the direct carbon MCFC and SOFC designs. Essentially, both electrolytes from the MCFC and SOFC are used within the one cell with the fuel dispersed within the molten carbonate electrolyte. The physical cell resembles that of a typical SOFC cell, with the molten electrolyte residing in the anode chamber, so that it may wet the anode. The electrolyte consists of a binary eutectic mixture of Li2CO3 and K2CO3.43 There is scope however to try alternative electrolyte mixtures, such as the range used in direct carbon MCFCs. The HDCFC is a new technology and was conceived by Irvine and co-workers at the University of St Andrews, U.K.6 The cell concept was created with the vision to combine advantages of the direct carbon MCFC and SOFC cells and, more so, to alleviate some issues with each cell type. Nabae and co-workers44 address some of the “shortcomings” of traditional stand-alone direct carbon MCFC and SOFC systems and how the combination of the two technologies solves some of these key issues. These include separation of the cathode and the molten carbonate via a solid oxide electrolyte, reducing the possibility of cathode corrosion, as observed in MCFC systems.44,45 Also, direct carbon MCFCs require management of CO2 in the electrolyte mix,44 and having a separate cathode chamber negates the need for CO2 circulation.12,44,45 The primary motivation for incorporating molten carbonate in the SOFC anode chamber, however, is to improve the kinetics of electrochemical carbon oxidation in direct carbon SOFCs at low temperatures.44 This is because, in a direct carbon, it is thought that contact between the anode and carbon fuel is necessary for the direct electrochemical oxidation of carbon (as per eq 13). This is the case at low temperatures (temperatures below where the Boudouard reaction is favored). The suspension of carbon in a molten carbonate solution effectively extends the anodic electrochemical reaction zone away from the TPB of SOFC systems. The electrochemical oxidation of carbon may occur in the slurry mixture because of the ionic conductivity of the carbonate and current carrying capacity of the carbon.46 However, the requirement to enhance the direct electrochemical oxidation of carbon in direct carbon SOFC systems is

Figure 8. Comparison between direct carbon SOFC cell performance with and without molten carbonate, adapted from Nabae and coworkers.44

tubular direct carbon SOFC cell, which upon the addition of molten carbonate into the anode compartment becomes a makeshift HDCFC unit. The fuel used by the authors was carbon black (CB), and the “cell (dry)” run was carried out as a physical contact-type direct carbon SOFC experiment, with no CO2 feed. It can be seen from Figure 8 that, at temperatures above 820 °C, the direct carbon SOFC system becomes favorable because CO2 produced in the anode compartment promotes Boudouard gasification of the fuel, improving the kinetics of reactions contributing to the overall conversion of C to CO2 (see sections 4.3 and 4.4). It is also worth noting that the ionic conductivity of the YSZ electrolyte is highly dependent upon the temperature. There is an emphasis upon minimizing the solid oxide electrolyte thickness in HDCFC cells, because resistance to O2− transfer through the electrolyte is high under the low-temperature operation of these cells. At higher temperatures, the HDCFC cell experiences significant decomposition of the molten carbonate, producing O2− ions in the melt via eq 16. CO32 − → CO2 + O2 −

(16)

2−

The presence of O ions contributes to the high open circuit voltage (OCV) values seen in HDCFC systems. This is due to the recombination of O2− with CO2, following electrochemical oxidation of CO (produced by the reverse Boudouard reaction) to CO2. The presence of O2− ions in the carbon/electrolyte slurry therefore reduces the activity of CO2, increasing the Nernst potential of eq 3.46 The same authors in an earlier study44 investigated the effect of the thermal history of the cell on OCV. They noted that the OCV could be increased to 1.5 V at 550 and 700 °C after high-temperature operation (900 °C stepped down to 550 and 700 °C). In their subsequent study, the authors reported that the OCV could be improved, without 1478


Energy & Fuels

Review

Less is known about the effects of species on molten carbonate systems than for solid oxide systems; however, species may be tentatively placed into categories of inhibitive and catalytic thanks largely to the work by Li and co-workers.48 The authors studied the effect of metallic oxides common to coal ash on the electrochemical performance of direct carbon MCFCs. These included a fixed amount (8%) of oxides of calcium, magnesium, iron, aluminum, and silicon impregnated in activated carbon (AC). Their results are summarized in Table 3.

compromising the carbonate electrolyte through high-temperature treatment, by inclusion of a Ni catalyst into the carbon fuel.46 The authors investigated the effect of loading CB with 10, 30, and 50 wt % Ni on the cell performance and OCV at various temperatures between 550 and 900 °C. They found the most significant improvement to peak power occurred at lower temperatures and by a factor of 7.6 at 550 °C for a Ni loading of 50 wt %. Jiang and Irvine45 ran their HDCFC with a CB/carbonate mixture at varying carbonate loadings (0, 20, 50, and 80 mol % carbonate) and temperatures to determine the effect on OCV. The authors chose an 80:20 mol % carbon/carbonate for their subsequent cell runs. The low carbonate concentration was chosen as a compromise between anodic polarization at high carbon loadings and sufficient carbonate to wet the carbon while minimizing corrosion issues. After optimization for minimum activation polarization, the authors then produced impedance spectra to assess cell performance and achieved a peak power density of around 18 and 53 mW cm−2 at 700 and 800 °C, respectively. Jain and co-workers43 investigated the performance of pyrolyzed medium-density fiberboard (pMDF). In an attempt to improve the poor performance of the pMDF in the fuel cell, the authors employed various fuel preparation techniques: immersion of pMDF sticks in the eutectic molten carbonate mixture, immersion of pMDF sticks in a saturated aqueous solution of the carbonates before drying (a similar method to that employed by Cao and co-workers47), and milling of pMDF with carbonate powder. The best performance was achieved by immersion of pMDF in the molten carbonate solution at an operating temperature of 800 °C. However, at temperatures below 700 °C, the pMDF samples prepared by immersion in aqueous solution and by milling show greater reactivity (with milled pMDF being the standout performer). As the authors note, there are also handling issues associated with the use of molten carbonate mixtures outside of the cell. As a new technology, there is much scope for further research into the benefits of low-temperature operation of the HDCFC. However, advances in the ionic conductivity of the YSZ electrolyte at lower temperatures would significantly improve the output capability of these cells.

Table 3. Direct Carbon MCFC Performance Using AC Impregnated with Various Metallic Oxides21,48 fuel AC AC AC AC AC AC

(pure) (8% MgO) (8% CaO) (8% Fe2O3) (8% Al2O3) (8% SiO2)

OCV (V)

current density (mA cm−2 at −0.9 V)

Pmax (mW cm−2)

−1.21 −1.22 −1.22 −1.20 −1.19 −1.18

16 22 18 17 15 14

46.3 51.8 50.9 47.3 43.3 40.8

From Table 3, it appears that, for direct carbon MCFCs, MgO, CaO, and Fe2O3 act as catalysts for the electrochemical process, whereas Al2O3 and SiO2 act as inhibitors. However, systematic studies into the effect of these species at various concentrations are required before they can be accurately categorized as either inhibitive or catalytic and the extent of their inhibitive or catalytic effects. More is known about the effect of species on direct carbon SOFCs because the performance of this cell is directly dependent upon the performance of the gasification process, a well-studied field. The mechanism proposed by Furimsky and co-workers49 for the catalyzed gasification of coal is given as follows. In the presence of CO2: xFe + CO2 = FexO + CO

(17)

FexO + C = x Fe + CO

(18)

In the presence of steam:

6. CELL PERFORMANCE: EFFECT OF COAL COMPOSITION Each coal sample has its own unique composition represented by ultimate analysis (carbon, hydrogen, nitrogen, sulfur, oxygen, and ash) and ash analysis (composition expressed as oxides of silicon, aluminum, calcium, magnesium, sodium, iron, titanium, and sulfur, among others). The outcomes of both analyses hold important information regarding the makeup of the coal. This section will summarize our current understanding of how some of these analyses, in particular sulfur and metallic oxides in ash, influence the performance of both direct carbon MCFCs and SOFCs. From this, an attempt to predict the performance of Victorian brown coals within these DCFC types may be made. 6.1. Role of Metals in Coal. Every species present in coal has an impact on the DCFC performance and operation. Some elements play an inhibitive role, whereas others exhibit catalytic effects. Which category these species fall into for both direct carbon MCFCs and SOFCs is very similar; however, the mechanisms by which they perform their catalytic or inhibitive effects may differ and are largely unknown.

xFe + H2O = FexO + H2

(19)

FexO + C = x Fe + CO

(18)

The effect of catalyzing the Boudouard reaction on direct carbon SOFC performance has been investigated by Tang and Liu,38 who used pure AC and Fe-loaded AC as their fuel source. The presence of Fe is known to enhance the generation of CO, as also seen in Figure 9. The enhancement of cell performance because of the presence of Fe is also illustrated in Figure 10 and Table 5 (by comparison of cells II and III). The maximum power output of the already GDC-catalyzed cell was almost doubled upon the addition of Fe. Other species were also tested for their impact on the direct carbon SOFC system. Li and co-workers36 investigated the catalysis of CO2-fed gasification in a tubular reactor where the button cell and fuel were noncontacting and maintained at different temperatures (750 and 700−1000 °C, respectively). They concluded that K, Ni, and Ca all showed catalytic effects in order of decreasing effect of K > Ni > Ca. Their results are summarized in Table 4. The authors reported that the decreases in the fuel gasification temperature to achieve the same power 1479


Energy & Fuels

Review

In terms of enhancing the electrochemical performance of the direct carbon SOFC, however, because of the cell operating on a gaseous fuel and the desire to keep cell components isolated from impurities, catalysis via elements within the coal is limited. More work has been performed to embed catalytic species into the anodes and cathodes themselves and. therefore, is a material selection issue. This is opposed to the potential for catalytic species to exist in the electrolyte solution in MCFCs. One such material was investigated by Tang and Liu.38 The authors used a pure Ag-coated anode and a GDC−Ag-coated anode (55:45 wt % GDC/Ag), in a tubular cell design, to analyze the effect of GDC on the cell performance. The results of their experiments are shown in Figure 10, where the presence of GDC has a large positive impact on the cell performance (cells I and II). The results are summarized in Table 5. Equally important as catalyzing elements and reactions is monitoring and understanding the roles of species that have an inhibitive effect on cell performance. Cherepy and co-workers13 experienced degradation of direct carbon MCFC performance via sulfidation corrosion when using cokes with 2.5−6 wt % sulfur. The authors found that Ni-based electrodes experienced densification of the porous electrodes through the formation of nickel sulfides, reducing contact with the carbon particulates. Indeed, it has been found that many species react detrimentally with Ni, a common additive to DCFC cell components. A list of known catalysts and inhibitors to direct carbon MCFC and SOFC cells is provided in Table 6. It is important to note that not all of the impurities are derived from the fuel source. Some are also derived from metallic components of the cell itself (as in the case with Cr2O3 evaporation). 6.2. Composition of Victorian Brown Coals. Victorian brown coals are high-moisture (around 60% as received) and low-ash (around 2% when dried) coals. Ultimate analysis and ash analysis23 of three major Victorian brown coals are presented in Tables 7 and 8, respectively. As is evident, these coals have very low sulfur and ash contents; however, in the ash, there are significant levels of metals that potentially may have both catalytic and inhibitive effects during their use in a DCFC. Given that no research has been conducted in the application of Victorian brown coals within DCFCs, predicting their performance in a DCFC is difficult. This is especially so because of the limited knowledge of mechanisms surrounding the catalytic and inhibitive effects of individual species. However, Victorian brown coals show promise based on their low sulfur and ash contents and relative abundance of catalytic elements. Coals apparently suited to gasification-driven SOFCs (for a given ash content, high proportions of Fe, Ca, Mg, Na, and K) are those from the Morwell, Hazelwood, Yallourn, and Maryvale mines. Coals apparently suited to direct carbon MCFC application (low in ash but, for a given ash content, high in Fe, Mg, and Ca and low concentrations of Si and Al) are those from the Yallourn, Morwell and Maryvale mines. Volatile matter is also undesirable in direct carbon MCFCs. All Victorian brown coals are high in volatile content; therefore, these would benefit from pyrolysis (devolatilization) prior to their use in the direct carbon MCFCs. For continuous operation of fuel cells, in particular the direct carbon MCFC, a low ash content is desirable. Also, for good fuel use in a gasification-driven SOFC, a high fixed carbon

Figure 9. CO molar percentage as a function of the temperature and presence of Fe, adapted from Tang and Liu.38

Figure 10. Cell performance for different anodic materials:38 cell I (pure Ag anode), cell II (Ag−GDC anode), and cell III (Ag−GDC anode with Fe).

Table 4. Catalytic Effects of Selective Coal Inorganics on the Performance of CO2-Fed Detached-Type Direct Carbon SOFC36 power density (mW cm−2) fuel

700 °C

750 °C

800 °C

CB−K CB−Ni CB−Ca

95.6

147.7 112.3 103.4

185.3 153.3 149.1

57.7

Table 5. Effect of Different Anodic Materials on the Cell Performance: Summary of Figure 10 Results38

cell I (pure Ag anode) cell II (Ag−GDC anode) cell III (Ag−GDC anode with Fe)

Pmax (mW cm−2)

polarization resistance (2.3 Ω cm2)

4 24 45

38 2.3 1

density output from CB for K, Ni, and Ca were 200, 150, and 130 °C, respectively. As described earlier, it is necessary to catalyze both reactions 12 and 2, which are coupled and widely accepted as the dominant reactions for most direct carbon SOFC arrangements. Therefore, catalysis of both the chemical and electrochemical oxidation of CO is required. The electrochemical oxidation of CO occurs at the anode three-phase boundary of the anode, electrolyte, and fuel.37 1480


Energy & Fuels

Review

Table 6. Summary of Catalysts and Inhibitors for Direct Carbon SOFCs and MCFCs MCFC catalysts MgO48 CaO48 Fe2O348

MCFC inhibitors

SOFC catalysts

SiO248 Al2O348 Cl > 0.01 ppm (vol)67 F > 0.01 ppm (vol)67 S (as COS, H2S, mercaptanes, and thiophenes)13,67 dust > 5 μm (from gas streams)67 Si (as silanes and siloxanes)67

SOFC inhibitors

Group I and II metals38 (K > Ni > Ca, decreasing catalytic effect on the CO2 gasification reaction)36

S (as H2S), reacts with Ni65,66 Se (as H2Se), reacts with Ni65 As, reacts with Ni65 Cr2O3 vapor68 P (as PH3 or HPOx), reacts with Ni65

Group VIII metals (Fe)4938

Table 7. Ash Analysis of Victorian Coal (Atomic wt %) Conducted by Brown and Co-workers69 Morwell Hazelwood PS Hazelwood Yallourn Maryvale Loy Yang Vale Point

SiO2

AlO1.5

FeO1.5

CaO

MgO

NaO0.5

KO0.5

TiO2

SO3

0.19 4.38 5.20 0.85 3.18 7.52 59.44

1.21 1.49 3.57 1.89 1.09 26.02 27.48

11.57 6.50 10.14 24.73 41.68 9.67 3.66

29.95 36.74 15.97 11.78 11.56 4.67 2.46

24.12 29.95 33.42 36.52 22.81 20.11 1.58

14.29 7.35 17.78 13.57 10.74 19.83 1.11

0.60 0.51 0.84 0.30 0.58 0.25 2.80

0.22 0.12 0.21 0.11 0.12 0.41 1.03

16.76 11.47 10.08 10.23 5.47 8.20 0.43

reactor, wire-mesh reactor, drop-tube reactor (also called entrained flow reactor), and Curie-point pyrolyzer. While a variety of pyrolyzer reactors have been employed by researchers, no discernible preference is noted for any given coal type. The preparation of chars via pyrolysis for use in DCFCs, in particular direct carbon SOFCs, is a promising and apparently necessary concept; however, little is known about the performance of these chars in DCFC systems. It is possible to subject the raw coal to a single heat treatment of pyrolysis followed by gasification within the direct carbon SOFC, but this would have undesirable consequences. Namely, the direct carbon SOFC would be subject to several impurities released during the pyrolysis of the coal in tars and other gaseous and condensable species, such as compounds of S and Cl. Pyrolysis would likely be of the greatest benefit to a gasification-driven SOFC system as a means of generating char as the fuel. The volatiles produced may be used for an alternative means, such as the generation of heat for the gasifier, or to satisfy another market, such as benzene, toluene, and xylene (BTX) and other nonfuel chemical production. See the work by Hayashi and Miura50 for a concise overview of chemical production from pyrolysis of coals. The authors also outline effects on soot formation, with the introduction of steam to the pyrolyzing atmosphere greatly reducing soot formation as well as substantially increasing COx evolution. Chars prepared via pyrolysis experience an increase in surface area but have been found to be a poor fuel for direct carbon MCFCs51,52 (see Table 9). This is due to the reduction in surface oxygen-containing compounds on the char surface, which are largely responsible for the reactivity of the coal in a molten carbonate environment. There are also implications for the application of chars in direct carbon SOFCs based on the pyrolysis conditions. Many species are volatilized during pyrolysis, and some of these leave with the volatile components, whereas some reattach themselves to the char. Which species remain in the char is an important consideration for the downstream gasification of the char in a direct carbon SOFC and is largely dependent upon the gasification reactor configuration. An overview of alkali and alkaline earth metallic (AAEM) species volatilization based on the peak temperature is shown in Figure 11.

Table 8. Typical Ultimate Analysis Ranges of Victorian Brown Coals on a Percent Mass Dry Basis70 C (%, db) H (%, db) N (%, db) S (%, db) O (%, db) ash (%, db) Na (%, db) Cl (%, db)

Yallourn

Morwell

Loy Yang

64.9−66.6 4.00−4.60 0.43−0.51 0.21−0.25 24 1.90−2.40 0.08−0.09 0.06−0.10

64.9−69.3 4.40−5.10 0.44−0.64 0.23−0.39 22.2 1.70−3.00 0.05−0.09 0.04−0.07

66.1−68.7 4.50−4.90 0.53−0.68 0.27−0.37 22.7 1.09−2.79 0.05−0.15 0.04−0.10

content is desired for high char yields. Table 8 gives the ultimate analysis ranges of some Victorian brown coals and highlights their low ash and sulfur contents. It is important to note however that the analyses given in Tables 7 and 8 are those of raw coals. Any form of fuel pretreatment, such as heat treatment or acid washing, will have a marked effect on the fuel composition as well as, in some cases, its physical nature. The effects of pretreatment for fuel preparation are discussed in the following section.

7. FUEL PREPARATION OPTIONS AND COAL CHARACTERISTICS Two major classes of pretreatment are heat treatment and acid washing. This section will address the effects of these two methods of pretreatment as well as the potential for application of treated coals as fuels for DCFCs. 7.1. Heat Treatment. There is an abundance of literature available on the heat treatment of coals and other carbonaceous matter in the form of pyrolysis and gasification. The focus of this literature review, in terms of heat treatment as a means of preparation of fuels for DCFC, was pyrolysis. The relevance of gasification to DCFC technology has been discussed in the previous section on direct carbon SOFCs, where it acts as not a pretreatment method but a necessary mechanism for fuel consumption. Various papers on the pyrolysis of low-rank coals, in particular Victorian brown coals, were reviewed. The major types of reactors mostly used for coal pyrolysis are the fluidized-bed 1481


Energy & Fuels

Review

Table 9. Summary of Heat Treatment Results on Raw Coal Properties and Effect on the Direct Carbon MCFC Performance36 heat treatment

surface functional groups

SBETa

Vtotala

Dporea

oxidation in 6.6% increase in surface O and 10.1% 23.1% negligible air (500 °C) 66.7% decrease in surface S increase decrease change pyrolysis 16.5% decrease in surface O and 58.2% 53.8% 12.9% (750 °C) 83.3% decrease in surface S decrease decrease increase pyrolysis 48.4% decrease in surface O and 10.1% 69.2% 15.2% (950 °C) 83.3% decrease in surface S decrease decrease increase a

crystallinity

OCV

Pmax

48% increase 0.06 V penalty (at 800 °C 53 mW cm−2 (52.3%) penalty in La cell operation) (at 800 °C cell operation) 76% increase 0.12 V penalty (at 800 °C 71 mW cm−2 (64.0%) penalty in La cell operation) (at 800 °C cell operation) 108% increase 0.16 V (73.9%) penalty (at 82 mW cm−2 penalty (at 800 in La 800 °C cell operation) °C cell operation)

SBET, Vtotal, and Dpore values were achieved via N2 adsorption.

Figure 11. Temperature map of AAEM species volatilization during pyrolysis of Victorian brown coal.

Figure 11 is not universal for all pyrolysis conditions. There are many variables for pyrolysis, such as reactor type, heating rate, peak temperature, and holding time.53 Figure 11 is supplied as a guide for typical results of pyrolysis. The peak temperature has a greater effect on volatilization of AAEM species than the heating rate.50,54 However, the peak temperature may still be misleading because Manzoori and Agarwal55 demonstrated that at least 80% of Mg, Ca, and Na can be volatilized at 700 °C with a holding time of 40 s and a complete volatilization of Cl. Therefore, it is necessary to ensure that holding times are sufficient to allow for complete volatilization, especially for fast pyrolysis. 7.2. Acid Washing. As opposed to heat treatment, the use of acid-washed coals in DCFCs is well-documented. Researchers have implemented acid washing for various purposes but primarily to remove metallic impurities to allow for selective impregnation of AAEM species, to determine their individual catalytic/inhibitive properties. The effect of acid washing on the porosity, surface area, and surface functional groups and the associated impact of these factors on the cell performance have also been examined. An overview of the effects of acid washing on coal and other commonly studied carbon sources is given in Table 10. With respect to the values stated in Table 10, the percentages are case-specific and not universal. They are provided as a means for quick comparison. It is accepted that HF is very

effective at removing Al and Si from coal. However, because of the extreme hazards surrounding this acid, the use of HF in acid digestion is being phased out. Unfortunately, the use of more benign acids, such as HNO3 and HCl, do not achieve the desired Al/Si removal. Microwave acid digestion is one method of increasing the capacity of an acid to adsorb metals from the coal. This method is particularly beneficial for weaker acids and is often a precursor to inductively coupled plasma (ICP) spectrometry analysis.56−58 Wang and co-workers 56 used microwave digestion of coal with HNO3 at 250 °C and 7.5 MPa and reported good recovery of trace elements in the coal samples. The absence of HF from the digestion allowed for the use of a quartz vessel. Popular reagents for use in microwave acid digestion include HCl, HNO3, and H2O2. The weight loss resulting from the decomposition of surface functional groups becomes more prominent with acid treatment, as shown in Figure 12.17 This is particularly evident with HNO3 treatment because this acid is known to generate high amounts of surface oxygen functional groups.21 Such acid treatment increases the reactivity to oxygen of the material, reducing its thermal stability. In some cases of acid washing, the acid decreases the thermal stability of the surface functional groups, which lowers the onset temperature of decomposition. This is the case for CO-containing functional groups when washed with HCl.21 1482


Energy & Fuels

Review

Table 10. Effect of Acid Washing on Various Coal Properties and Performance in Direct Carbon MCFCs fuel type species removal

raw coal52

surface functional groups

raw coal52

SBET Vtotal Dpore crystallinity (by La value) Pmax (molten carbonate) a

AC21 CB21 raw coal52 AC47 raw coal52 AC47 raw coal52 raw coal52 AC21 CB21 raw coal52

HNO3

HCl

modestly effective for Al and S removal and increase in surface N 125.2% increase in surface O and 176.9% increase in surface N 32.5% increase in surface O 390% increase in surface O 92.4% increase 1.9% increase 100.0% increase 2.4% increase 2.0% increase negligible change minor increase (6.7%) notable increase (31.0%) 18 mW cm−2 improvement (at 800 °C)

HF

modestly effective for Al removal 30.8% increase in surface O

extremely effective for Si and Al removal 27.5% increase in surface O

13.8% increase in surface O 190% increase in surface O 72.2% increase a 84.6% increase a 4.6% increase negligible change negligible increase notable increase (27.6%) 4 mW cm−2 improvement (at 800 °C)

a a 112.7% increase52 6.1% increase 107.7% increase 4.9% increase 2.6% increase negligible change a a 6 mW cm−2 improvement (at 800 °C)

Not reported.

Table 11. Analytical Techniques Used for DCFC Fuel Analysis analytical technique

Figure 12. Typical TGA curve for AC and response to acid washing.

Acid washing holds benefits for research by allowing for examination into the extent of pore blockage by metal ions as well as being a necessary precursor to ion-exchange analysis. Acid washing with HNO3 has even been shown to significantly enhance the performance of direct carbon MCFCs by increasing the number of surface oxygen groups. However, acid washing on an industrial scale is not feasible, and the primary benefit of laboratory-based acid washing of coals for use in DCFCs is insight gained into the catalytic/inhibitive effect of individual AAEM species. Therefore, the most suitable coal for use in a commercial DCFC would be one low in ash, Al, and Si contents, to negate the need for a large-scale acidwashing process. Minimal examination into the effect of acid washing on fuels for direct carbon SOFCs has been performed because the fuel of choice for researchers has been CB (see Table 12). Investigations into the effects of metal ions on the reaction mechanisms for direct carbon SOFCs has been investigated through impregnation of CB (Fe-based catalysis38) and AC (K-, Ni-, and Ca-based catalysis36). An overview of the analysis of these fuels is supplied in the following section (section 8).

MCFC

SOFC

used for

SEM

13

59 and 60

XPS

48 and 52

59

ICP−OES TDS/TPD

48 21, 48, 51, and 52

TGA XRD

21, 48, 51, and 52 13, 21, 48, 51, and 52

gas adsorption of N2 gas adsorption of CO2 frequency response analyzer

21, 47, 48, 51, and 52

particle size distribution surface functional complex ash composition surface oxygenated complex proximate analysis degree of crystallinity and crystallite size surface area

48 and 52

surface area

21 and 51

electrical conductivity

60 9

respective DCFCs than it does with the type of fuels used, as will be discussed in section 9. 8.1. X-ray Photoelectron Spectroscopy (XPS). The surface characteristics of fuel are accepted as one of the most important factors in the fuel reactivity and, hence, the achievable current density from the DCFC. As one method of determining surface and sub-surface (to 12 nm depth) composition, XPS was a popular method for Li and co-workers.48,52 Of particular interest to Li and co-workers was the surface oxygen/carbon ratio because surface oxygen functional groups have a significant positive effect on the performance of DCFCs, in particular direct carbon MCFCs. It should be noted that Li examined the surface compositions of several carbon fuels using various techniques, including inductively coupled plasma−optical emission spectroscopy (ICP−OES), temperature-programmed desorption (TPD), and thermogravimetric analysis (TGA). For direct carbon SOFCs, there was no fuel analysis via XPS. The only found application of XPS was that of an elemental analysis of a post-CH4-deposited anode of a deposited-type direct carbon SOFC. 8.2. ICP−OES. The ICP−OES method was used by Li and co-workers48 to determine the ash composition in coal samples after the oxidation of the coal in an air-fed tube furnace at 700 °C for 2 h. This technique was not used in any papers investigating direct carbon SOFCs, again because of the predominant use of CB in this type of fuel cell and the ash-free nature of CB.

8. ANALYTICAL TECHNIQUES The types of analytical techniques used by authors to discern fuel properties are also indirectly dependent upon the DCFC type. There are a disproportionate number of analytical techniques and tests conducted into fuels used in direct carbon MCFCs over those used in direct carbon SOFCs (see Table 11). This has less to do with the importance of certain fuel properties on the performance of the 1483


CB graphite

AC

miscellaneous fuels

coal class

0.7 1.34 1.34 1.35 1.33 a a a a 1.26 1.26 1.31 1.28 1.29 1.25 a

molten carbonate (Li2CO3−K2CO3−Al2O3 as 1.05:1.2:1 by mass) molten carbonate molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−Na−K) molten carbonate (eutectic Li−Na−K) molten carbonate solid oxide (detached type, CO2 fed, button cell)

1484

TIM-CAL M-292 Desulco graphite particles flake graphite micronized graphite (IEA International Trading Qingdao Co., Ltd., China)

a

a 0.77

solid oxide (detached type, CO2 fed, button cell) molten carbonate (Li−K−Al as 1.05:1.2:1 by mass)

0.90 1.05 a

solid oxide solid oxide solid oxide (tube, ScSZ electrolyte film, no introduction of anodic gas, self-contained) molten carbonate molten carbonate (eutectic Li−K) molten carbonate (Li−K−Al as 1.05:1.2:1 by mass) solid oxide (YSZ−GDC anode, button cell, He environment)

1.05 a 0.65 0.70

52

a 0.65

solid oxide (YSZ−GDC anode, button cell, CO2 environment)

14 (0.8 V) 58 (0.8 V) a a

17 46 63 6.5

24 45 75

42

a 0.65

17 (0.8 V) 46 (0.8 V) a

49 (at 0.4 V)

a

a

85 83 a a a 84 56 51 a 32 a a a a 46 185.3 (at 0.7 V, not peak power) 153.3 (at 0.7 V, not peak power) 149.1 (at 0.7 V, not peak power) 116

61 39 46 61 36 141

peak power density (mW cm−2)

solid oxide (direct contact type, carbon pellet pressed between anode and electrolyte, N2 anodic gas) solid oxide (YSZ−GDC anode, button cell, He environment)

a

a

a

solid oxide (detached type, CO2 fed, button cell)

vapor-grown carbon fiber (Showa Denko K.K. Co., Japan) granular AC (Calgon BPL) granular AC (Calgon BPL) granular AC, HNO3 washed (Calgon BPL) granular AC, HCl washed (Calgon BPL) peach pit AC acid washed coconut AC acid washed coal-derived AC acid washed wood-derived AC (HF washed) Koppers Continex N220 Koppers Continex N220 Koppers Continex N220, HNO3 washed Koppers Continex N220, HCl washed Koppers Continex N220, air plasma treated Koppers Continex N660 Black Pearls 2000 (K catalyzed), GP-3848, Cabot Corporation, Boston, MA Black Pearls 2000 (Ni catalyzed), GP-3848, Cabot Corporation, Boston, MA Black Pearls 2000 (Ca catalyzed), GP-3848, Cabot Corporation, Boston, MA conductive CB (Vulcan XC72, Cabot Corporation, Boston, MA) pressed conductive CB (Vulcan XC72, Cabot Corporation, Boston, MA) ball-milled CB (Vulcan XC-72 R, Cabot Corporation, Boston, MA) ball-milled CB (Vulcan XC-72 R, Cabot Corporation, Boston, MA) pure AC, Taishan Yueqiao Reagent Plastic Co., Ltd. AC with Fe, Taishan Yueqiao Reagent Plastic Co., Ltd. ShaoXing RenFei Carbon Blacks Co.

77 (0.8 V) 110 (0.8 V) 47 (0.8 V) 87 (0.8 V) 58 (0.8 V) 251.56 (at peak power) a 50 (0.8 V) 36.5 (0.9 V) 86 (0.9 V) 53.5 (0.9 V) 124 (0.8 V) 102 (0.8 V) 65 (0.8 V) 45 (0.8 V) 22 (0.8 V) 16 (0.9 V) 46 (0.9 V) 32 (0.9 V) 22 (0.9 V) 37 (0.8 V) a

a a a a a 0.82

molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (eutectic Li−K) molten carbonate (Li2CO3−K2CO3−Al2O3 as 1.05:1.2:1 by mass)

acetylene black (from acetylene pyrolysis) Arosperse 3 (from furnace oil) Arosperse 15 (from methane pyrolysis) carbon aerogel microspheres (pyrolyzed, glassy carbon) needle petroleum coke (calcined, milled) green needle coke, crushed (Jinzhou Petrochemical Co.)

current density (mA cm−2)

OCV (V)

electrolyte/cell arrangement

fuel name/origin

Table 12. DCFC Performances Using Different Fuels

800 800 650 700

800 800 800

800

800

900

650

800

800

650 800 800 800 800 800 800 800 a 800 a a a a 800 800

800 800 800 800 800 650

temperature (°C)

reference

51 13 71 9

38 38 39

9

9

10

71

36

36

71 51 21 21 21 13 13 13 47 51 21 21 21 21 51 36

13 13 13 13 13 71

Energy & Fuels Review


52 800 129

Data not reported. a

1.33 molten carbonate

bituminous

8.3. X-ray Diffraction (XRD). Aside from surface characteristics, crystallinity, often measured by XRD, has been identified as the other main factor in influencing fuel reactivity in DCFCs. This method has been employed by authors to compare the crystallinity of fuel samples and its effect on the DCFC performance, as well as the effect of preparation techniques on crystallinity, in particular heat treatment.52 Kulkarni and co-workers9 employed XRD to investigate the stability of the LSCF anode in contact with CB. This was achieved via XRD analysis of a 50:50 (wt %) mixture of LSCF powder and CB. All of the XRD measurements encountered were conducted at room temperature. Many of the fuel samples that have been analyzed are those of chars produced by heat treatment. As the fuel sample cools to room temperature, the morphology of the char undergoes certain changes, including a change in crystallite properties. Therefore, crystallite size estimates from XRD measurements are limited to the conditions of the XRD analysis and the properties of the char at those temperatures. They do not necessarily reflect the coal properties under the conditions experienced by the fuel during preparation (or more importantly those that the fuel undergoes during reforming in the fuel cell). Therefore, XRD conducted with samples in situ and at high temperature [hightemperature XRD (HTXRD)] will be a useful analytical technique for characterizing fuel samples for DCFC application. 8.4. Scanning Electron Microscopy (SEM). In terms of fuel analysis, this was not a widely used method. Cherepy and co-workers13 used SEM imaging to estimate coal particle size and observe particle size distribution, with further image processing. The only found use of SEM imaging to discern fuel properties for use in direct carbon SOFCs was for the carbon deposit fuel cell types.59,60 The SEM was not used to analyze the raw fuel as such but rather the deposited carbon on the anode, because the fuel sources analyzed were methane. 8.5. Surface Area Measurement by Gas Adsorption. There are two main gas adsorption techniques used for the analysis of surface characteristics of coal and other carbon-based fuels. These are N2 adsorption and CO2 adsorption. N2 adsorption is a popular method for analysis of the porosity of a fuel sample, in particular a coal-based fuel sample. Most authors calculated the overall surface area using the Brunauer−Emmett−Teller (BET) method. Some authors estimated the mesoporous surface area as the total surface area minus the microporous surface area.47,51 Accurate mesoporous volume calculations are crucial because the mesoporous volume has been identified as an important factor for direct carbon MCFC performance,51 because they are much more accessible to the molten carbonate electrolyte than micropores. For this reason, the total surface area is not a reliable base for predicting the reactivity of a fuel in a direct carbon MCFC. Li and co-workers21 report that a mesoporous structure in the fuel can be detected by hysteresis loops in the N2 adsorption isotherms. Li and co-workers48,52 used CO2 adsorption to determine Vmicro, Smicro, and Dmicro via the Dubinin−Raduskevich (DR) method. It is thought by many that differences between N2 and CO2 specific surface area results (in favor of larger CO2 area readings) are due to an activated diffusion effect rendering the microporous structure inaccessible to N2 at 77 K.52,61,62 However, it may have just as much to do with CO2 chemisorption in some cases. CO2 chemisorbs onto alkaline earth metal oxides. This is consistent with CO2 chemisorption onto Ca-loaded lignite chars.63 Linares-Solano and co-workers64 identified that this could result in misleading surface-area approximations based on CO2 adsorption when comparing their analytical results for pore volume and surface area for various Ca loadings. 8.6. Thermal Desorption Spectroscopy (TDS/TPD). TPD is another useful method21,48,51,52 for analyzing the surface chemistry of direct carbon MCFC fuel sources. As was the case with XPS, the authors used TPD to investigate surface oxygen complexes. This was used to compare the reactivity of the fuels in the direct carbon MCFC based on their surface oxygen group concentrations, as well as the effect of preparation techniques on the formation/destruction of these surface oxygen groups. 8.7. TGA. Li and co-workers21,48,51,52 used TGA to measure the moisture, volatile matter, fixed carbon, and ash content of various raw coals, as well as acid- and heat-treated coals, for use in a direct carbon

47 (0.9 V)

52 800 115 1.29 molten carbonate

32 (0.9 V)


35 (0.9 V)

52 800 58 26 (0.9 V) 1.24 molten carbonate


19 (0.9 V)

800 40 22 (0.9 V) 1.19 molten carbonate

5.9 63 85 111

graphite powder Newlands (Queensland, Australia) coking-grade raw coal Blackwater (Queensland, Australia) coking-grade raw coal Germancreek (Queensland, Australia) coking-grade raw coal Germancreek coking-grade coal (Queensland, Australia) pyrolyzed (750 °C) Germancreek coking-grade coal (Queensland, Australia) pyrolyzed (950 °C) Germancreek coking-grade coal (Queensland, Australia) heat treated in air (500 °C) Germancreek coking-grade coal (Queensland, Australia) acid washed (HF) Germancreek coking-grade coal (Queensland, Australia) acid washed (HCl) Germancreek coking-grade coal (Queensland, Australia) acid washed (HNO3)

1.05 1.25 1.32 1.31

a 25 (0.9 V) 31 (0.9 V) 29 (0.9 V)

800 800 800 800

72 48 48 48 and 52 52

Review

solid oxide (YSZ), Ar gas feed to fuel side molten carbonate molten carbonate molten carbonate

peak power density (mW cm−2) coal class

Table 12. continued

fuel name/origin

electrolyte/cell arrangement

OCV (V)

current density (mA cm−2)

temperature (°C)

reference

Energy & Fuels

1485


Energy & Fuels

Review

MCFC. They also compared the thermogravimetric results to assess changes to thermal stability of the coal after acid washing (see Figure 12). Zhao and co-workers60 also employed TGA but in a direct carbon SOFC. However, this was again for a deposited-type direct carbon SOFC, and the TGA was employed to analyze the deposited carbon on the anode. This test was performed in support of their SEM imaging, to gain insight into the reaction mechanism for CH4-deposited direct carbon SOFCs. 8.8. Frequency Response Analyzer. Electrical conductivity is an important factor for coal materials in DCFCs because they are the anodic material and sometimes used as the electrode. They play a major role in contributing to the global ohmic resistance of the cell. The electrical conductivity of a coal is directly related to its structure, in particular its graphitic nature. Therefore, frequency response analysers may be employed by researchers who are concerned with carbon fuels, which are also used as the anode electrode, or what impact preparation techniques may have on the potential for certain fuels to be used as the anode in a DCFC. However, this arrangement has been somewhat superseded by the coal dispersed in liquid electrolyte design. Also, it is widely accepted that the reactivity of a coal is more important than a high electrical conductivity, because an improvement in one factor often leads to a decline in the other, because of crystallinity. 8.9. Fuel Analysis Overview. Table 11 shows the authors who have used various analytical techniques to discern certain properties of carbonaceous fuels used in both direct carbon SOFCs and MCFCs. There were numerous other analytical techniques employed by many authors to evaluate the performance and output of cells, such as online gas chromatography and soap-film burets; however, the focus of this table is purely on fuel analysis. It is clear from Table 11 that there is a greater emphasis placed on analyzing fuels for direct carbon MCFCs as opposed to direct carbon SOFCs. This is due to a greater variety of fuels used in direct carbon MCFCs (see Table 12), as well as greater variety of preparation techniques. The overwhelming majority of direct carbon SOFC fuels used in the papers reviewed were CB. Because the CBs used were commercial CBs, most of the fuel properties are supplied by the supplier and, therefore, additional analysis of the CB was only required when the fuel was altered, such as when the CB underwent ion impregnation. Also, most of the direct carbon SOFC papers that included fuel analysis were deposited types, with the analysis focused on the anode post-deposition. Table 11 highlights the lack in both the variety of fuels used within direct carbon SOFCs as well as investigation of preparation techniques used to enhance the fuel performance within the DCFC.

application of coals with limited pretreatment. This is particularly important because much remains unknown about the mechanisms of individual species as well as uncertainty about some fundamental reaction mechanisms. These factors all contribute to a lack of predictability of DCFC performance for a given fuel, much less one that has not been trialed before in Victorian brown coal.

10. CONCLUDING COMMENTS DCFCs can operate at high temperatures and therefore have significant thermodynamic advantage of almost zero entropy change, resulting in high theoretical electrochemical conversion efficiency. The current status of DCFC technology and its different configurations and associated analysis requirements have been reviewed in this paper. In particular, emphasis has been given on the review of all fuels tested in direct carbon MCFC and SOFC systems and critical evaluation of low-rank coals and biomass, among other alternative fuels. Most of the research to date has focused on MCFC- and SOFC-type DCFC systems. Only limited types of fuels have been used, with major fuels being CB, AC, and light gases. However, neither of the two major solid fuels can potentially be used in larger scale systems. Only a limited number of studies have used coal, assessing the catalytic and inhibitive effects of some of the metallic elements present in ash on cell performance. Compounds of Ni, Ca, Mg, and Fe have been demonstrated to exert catalytic effects, while the presence of Si and Al are shown to exhibit inhibitive effects. However, a systematic study assessing the effects of their concentration on cell performance is warranted. There is also a limited understanding of the mechanisms by which these compounds interact with the fuels or fuel cell components (in particular the anode and electrolyte) and how they perform their catalytic and inhibitive roles. Low-rank fuels have not been assessed in general for DCFC applications. Some of the low-rank fuels, such as dried Victorian brown coal and dried biomass, and their heat-treated products can potentially be good fuel sources for DCFCs because of the low ash content and presence of a significant amount of the catalytic elements. However, the effect of the levels of the inhibitive elements and the solid trace metals in these fuels on their electrochemical performance in a DCFC is unknown. The review also identifies the typical fuel preparation methods for use of these fuels in DCFC systems. The associated analytical methods for characterizing the prepared fuels and assessing their performance in DCFC systems have also been identified and discussed.

9. FUEL TYPES USED IN DCFCS Table 12 is a summary of the performance of modern DCFC arrangements based on the fuels used. As Table 12 highlights, the vast majority of research into direct carbon SOFCs has been conducted using CB samples. This is because CB is a highly pure source of carbon with low ash content and results in minimal degradation of the fuel cell. Because of the absence of impurities, CB also allows for accurate investigation into the effect of metal ions on the performance of the cell, after ion exchange or ion impregnation. Despite the obvious advantages for research that CB has, there is a need to expand fuel sources to coal-derived carbon for industrial application of the DCFC. Minimal research has been conducted into the use of coal-derived fuels in SOFCs, in particular the use of lignite-based fuels. Investigation into application of raw coals and their products in DCFCs is paramount in assessing the feasibility of DCFCs as a large-scale provider of energy. A greater understanding of achievable outputs from raw coals as well as treated coals is therefore required. The use of these fuels would also help validate whether recent knowledge gained into the effect of individual metal ions on cell performance is translatable to the potential real-world

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge Brown Coal Innovation Australia (BCIA) for this research.

■

REFERENCES

(1) Muthuvel, M.; Jin, X.; Botte, G. G. Fuel cellsExploratory fuel cells | Direct carbon fuel cells. In Encyclopedia of Electrochemical Power

1486


Energy & Fuels

Review

Sources; Jürgen, G., Ed.; Elsevier: Amsterdam, The Netherlands, 2009; pp 158−171. (2) Cao, D.; Sun, Y.; Wang, G. Direct carbon fuel cell: Fundamentals and recent developments. J. Power Sources 2007, 167 (2), 250−257. (3) Williams, K. R.; Koslow, R. An introduction to fuel cells. J. Electrochem. Soc. 1966, 113 (6), 139C−140C. (4) Kurzweil, P. History | Fuel cells. In Encyclopedia of Electrochemical Power Sources; Jürgen, G., Ed.; Elsevier: Amsterdam, The Netherlands, 2009; pp 579−595. (5) Baur, E. Z. Elektrochem. Angew. Phys. Chem 1910, 16, 300. (6) Elsevier. Direct carbon SOFC/MCFC hybrid developed at St Andrews. Fuel Cells Bull. 2008, 2008 (8), 10. (7) Cherepy, N.; Krueger, R.; Cooper, J. F. In Power Sources for the New Millenium; Jain, M., Ryan, M. A., Surampudi, S., Marsh, R. A., Nagarajan, G., Eds.; The Electrochemical Society: Pennington, NJ, 2000; The Electrochemical Society Proceedings Series, p 64. (8) Gur, T. M.; Huggins, R. A. Direct electrochemical conversion of carbon to electrical energy in a high temperature fuel cell. J. Electrochem. Soc. 1992, 139 (10), L95−L97. (9) Kulkarni, A.; Giddey, S.; Badwal, S. P. S Electrochemical performance of ceria-gadolinia electrolyte based direct carbon fuel cells. Solid State Ionics 2011, 194 (1), 46−52. (10) Desclaux, P.; Nürnberger, S.; Rzepka, M.; Stimming, U. Investigation of direct carbon conversion at the surface of a YSZ electrolyte in a SOFC. Int. J. Hydrogen Energy 2010, 36 (16), 10278− 10281. (11) Dicks, A. L. The role of carbon in fuel cells. J. Power Sources 2006, 156 (2), 128−141. (12) Selman, J. R. Molten-salt fuel cellsTechnical and economic challenges. J. Power Sources 2006, 160 (2), 852−857. (13) Nerine, J.; Cherepy, R. K.; Fiet, K. J.; Jankowski, A. F.; Cooper, J. F. Direct conversion of carbon fuels in a molten carbonate fuel cell. J. Electrochem. Soc. 2005, 152 (1), A80−A87. (14) Vutetakis, D. G.; Skidmore, D. R.; Byker, H. J. Electrochemical oxidation of molten carbonate-coal slurries. J. Electrochem. Soc. 1987, 134 (12), 3027−3035. (15) Coughlin, R. W.; Farooque, M. Hydrogen production from coal, water and electrons. Nature 1979, 279 (5711), 301−303. (16) Coughlin, R. W.; Farooque, M. Consideration of electrodes and electrolytes for electrochemical gasification of coal by anodic oxidation. J. Appl. Electrochem. 1980, 10 (6), 729−740. (17) Coughlin, R. W.; Farooque, M. Electrochemical gasification of coal-simultaneous production of hydrogen and carbon dioxide by a single reaction involving coal, water, and electrons. Ind. Eng. Chem. Process Des. Dev. 1980, 19 (2), 211−219. (18) Farooque, M.; Coughlin, R. W. Electrochemical gasification of coal (investigation of operating conditions and variables). Fuel 1979, 58 (10), 705−712. (19) Li, H.; Liu, Q.; Li, Y. A carbon in molten carbonate anode model for a direct carbon fuel cell. Electrochim. Acta 2010, 55 (6), 1958−1965. (20) Liebhafsky, H. A.; Cairns, E. J. Fuel Cells and Fuel Batteries. Guide to Their Research and Development; John Wiley and Sons: New York, 1968; p 692. (21) Li, X.; Zhu, Z.; Chen, J.; De Marco, R.; Dicks, A.; Bradley, J.; Lu, G. Surface modification of carbon fuels for direct carbon fuel cells. J. Power Sources 2009, 186 (1), 1−9. (22) Goret, J.; Tremillon, B. Propriétés chimiques et électrochimiques en solution dans les hydroxydes alcalins fondusIV. Comportement électrochimique de quelques métaux utilisés comme électrodes indicatrices. Electrochim. Acta 1967, 12 (8), 1065−1083. (23) Zecevic, S.; Pattona, E. M.; Parhamia, P. Direct electrochemical power generation from carbon in fuel cells with molten hydroxide electrolyte. Chem. Eng. Commun. 2005, 192 (12), 1655−1670. (24) Zecevic, S.; Patton, E. M.; Parhami, P. Carbon−air fuel cell without a reforming process. Carbon 2004, 42 (10), 1983−1993. (25) Li, S.; Lee, A. C.; Mitchell, R. E.; Gur, T. M. Direct carbon conversion in a helium fluidized bed fuel cell. Solid State Ionics 2008, 179 (27−32), 1549−1552.

(26) Lee, A. C.; Li, S.; Mitchell, R. E.; Gur, T. M. Conversion of solid carbonaceous fuels in a fluidized bed fuel cell. Electrochem. Solid-State Lett. 2008, 11 (2), B20−B23. (27) Pointon, K.; Lakeman, B.; Irvine, J.; Bradley, J.; Jain, S. The development of a carbon−air semi fuel cell. J. Power Sources 2006, 162 (2), 750−756. (28) Suzuki, T.; Hasan, Z.; Funahashi, Y.; Yamaguchi, T.; Fujishiro, Y.; Awano, M. Impact of anode microstructure on solid oxide fuel cells. Science 2009, 325 (5942), 852−855. (29) Hong Huang, M. N.; Su, P.; Fasching, R.; Saito, Y.; Prinz, F. B. High-performance ultrathin solid oxide fuel cells for low-temperature operation. J. Electrochem. Soc. 2007, 154 (1), B20−B24. (30) Zongping, S.; Haile, S. M. A high-performance cathode for the next generation of solid-oxide fuel cells. Nature 2004, 431 (7005), 170−173. (31) Steele, B. C. H.; Heinzel, A. Materials for fuel-cell technologies. Nature 2001, 414 (6861), 345−352. (32) Yang, L.; Wang, S.; Blinn, K.; Liu, M.; Liu, Z.; Cheng, Z.; Liu, M. Enhanced sulfur and coking tolerance of a mixed ion conductor for SOFCs: BaZr0.1Ce0.7Y0.2−xYbxO3−δ. Science 2009, 326 (5949), 126− 129. (33) Xing, Y.-Z.; Li, C.-J.; Li, C.-X.; Yang, G.-J. Influence of throughlamella grain growth on ionic conductivity of plasma-sprayed yttriastabilized zirconia as an electrolyte in solid oxide fuel cells. J. Power Sources 2008, 176 (1), 31−38. (34) Garbayo, I.; Dezanneau, G.; Bogicevic, C.; Santiso, J.; Gracia, I.; Sabate, N.; Tarancon, A. Pinhole-free YSZ self-supported membranes for micro solid oxide fuel cell applications. Solid State Ionics 2011, DOI: 10.1016/j.ssi.2011.09.011. (35) Rose, L.; Menon, M.; Kammer, K.; Kesler, O.; Larsen, P. H. Processing of Ce1−xGdxO2−δ (GDC) thin films from precursors for application in solid oxide fuel cells. Adv. Mater. Res. 2006, 15−17, 293−298. (36) Li, C.; Shi, Y.; Cai, N. Performance improvement of direct carbon fuel cell by introducing catalytic gasification process. J. Power Sources 2010, 195 (15), 4660−4666. (37) Li, C.; Shi, Y.; Cai, N. Effect of contact type between anode and carbonaceous fuels on direct carbon fuel cell reaction characteristics. J. Power Sources 2011, 196 (10), 4588−4593. (38) Tang, Y.; Liu, J. Effect of anode and Boudouard reaction catalysts on the performance of direct carbon solid oxide fuel cells. Int. J. Hydrogen Energy 2010, 35 (20), 11188−11193. (39) Liu, R.; Zhao, C.; Li, J.; Zeng, F.; Wang, S.; Wen, T.; Wen, Z. A novel direct carbon fuel cell by approach of tubular solid oxide fuel cells. J. Power Sources 2010, 195 (2), 480−482. (40) Badwal, S. P. S; Giddey, S. The holy grail of carbon combustionThe direct carbon fuel cell technology. Mater. Forum 2010, 34, 181−185. (41) Nakagawa, N.; Ishida, M. Performance of an internal directoxidation carbon fuel cell and its evaluation by graphic exergy analysis. Ind. Eng. Chem. Res. 1988, 27 (7), 1181−1185. (42) Lee, A. C.; Mitchell, R. E.; Gur, T. M. Thermodynamic analysis of gasification-driven direct carbon fuel cells. J. Power Sources 2009, 194 (2), 774−785. (43) Jain, S. L.; Barry Lakeman, J.; Pointon, K. D.; Marshall, R.; Irvine, J. T. S. Electrochemical performance of a hybrid direct carbon fuel cell powered by pyrolysed MDF. Energy Environ. Sci. 2009, 2 (6), 687−693. (44) Nabae, Y.; Pointon, K. D.; Irvine, J. T. S. Electrochemical oxidation of solid carbon in hybrid DCFC with solid oxide and molten carbonate binary electrolyte. Energy Environ. Sci. 2008, 1 (1), 148−155. (45) Jiang, C.; Irvine, J. T. S. Catalysis and oxidation of carbon in a hybrid direct carbon fuel cell. J. Power Sources 2011, 196 (17), 7318− 7322. (46) Nabae, Y.; Pointon, K. D.; Irvine, J. T. S. Ni/C slurries based on molten carbonates as a fuel for hybrid direct carbon fuel cells. J. Electrochem. Soc. 2009, 156 (6), B716−B720. 1487


Energy & Fuels

Review

(67) Huppmann, G. Fuel cellsMolten carbonate fuel cells | Systems. In Encyclopedia of Electrochemical Power Sources; Jürgen, G., Ed.; Elsevier: Amsterdam, The Netherlands, 2009; pp 479−496. (68) Liu, Y.; Chen, D. Y. Protective coatings for Cr2O3-forming interconnects of solid oxide fuel cells. Int. J. Hydrogen Energy 2009, 34 (22), 9220−9226. (69) Brown, L.; Cashion, J.; Ledger, R. Coal ash composition of Australian low rank coals. Hyperfine Interact. 1992, 71 (1), 1411−1414. (70) Bhattacharya, S.; Beaupeurt, I.; Topping, B. Gasification Tests in a Pressurised Fluidised Bed Gasifier-Process Development Unit Using Air/ Oxygen and Steam as Reactants; Cooperative Research Centre for Clean Power from Lignite: Melbourne, Australia, 2006. (71) Chen, M.; Wang, C.; Niu, X.; Zhao, S.; Tang, J.; Zhu, B. Carbon anode in direct carbon fuel cell. Int. J. Hydrogen Energy 2010, 35 (7), 2732−2736. (72) Kim, J.-P.; Lim, H.; Jeon, C.-H.; Chang, Y.-J.; Koh, K.-N.; Choi, S.-M.; Song, J.-H. Performance evaluation of tubular fuel cells fuelled by pulverized graphite. J. Power Sources 2010, 195 (22), 7568−7573.

(47) Cao, D.; Wang, G.; Wang, C.; Wang, J.; Lu, T. Enhancement of electrooxidation activity of activated carbon for direct carbon fuel cell. Int. J. Hydrogen Energy 2010, 35 (4), 1778−1782. (48) Li, X.; Zhu, Z.; De Marco, R.; Bradley, J.; Dicks, A. Evaluation of raw coals as fuels for direct carbon fuel cells. J. Power Sources 2010, 195 (13), 4051−4058. (49) Furimsky, E.; Sears, P.; Suzuki, T. Iron-catalyzed gasification of char in carbon dioxide. Energy Fuels 1988, 2 (5), 634−639. (50) Hayashi, J.-i.; Miura, K. Pyrolysis of Victorian brown coal. In Advances in the Science of Victorian Brown Coal; Chun-Zhu, L., Ed.; Elsevier Science: Amsterdam, The Netherlands, 2004; pp 134−222. (51) Li, X.; Zhu, Z. H.; Marco, R. D.; Dicks, A.; Bradley, J.; Liu, S.; Lu, G. Q. Factors that determine the performance of carbon fuels in the direct carbon fuel cell. Ind. Eng. Chem. Res. 2008, 47 (23), 9670− 9677. (52) Li, X.; Zhu, Z.; De Marco, R.; Bradley, J.; Dicks, A. Modification of coal as a fuel for the direct carbon fuel cell. J. Phys. Chem. A 2010, 114 (11), 3855−3862. (53) Quyn, D. M.; Wu, H.; Li, C.-Z. Volatilisation and catalytic effects of alkali and alkaline earth metallic species during the pyrolysis and gasification of Victorian brown coal. Part I. Volatilisation of Na and Cl from a set of NaCl-loaded samples. Fuel 2002, 81 (2), 143− 149. (54) Li, C. Z.; Sathe, C.; Kershaw, J. R.; Pang, Y. Fates and roles of alkali and alkaline earth metals during the pyrolysis of a Victorian brown coal. Fuel 2000, 79 (3−4), 427−438. (55) Manzoori, A. R.; Agarwal, P. K. The fate of organically bound inorganic elements and sodium chloride during fluidized bed combustion of high sodium, high sulphur low rank coals. Fuel 1992, 71 (5), 513−522. (56) Wang, J.; Nakazato, T.; Sakanishi, K.; Yamada, O.; Tao, H.; Saito, I. Single-step microwave digestion with HNO3 alone for determination of trace elements in coal by ICP spectrometry. Talanta 2006, 68 (5), 1584−1590. (57) Wang, J.; Nakazato, T.; Sakanishi, K.; Yamada, O.; Tao, H.; Saito, I. Microwave digestion with HNO3/H2O2 mixture at high temperatures for determination of trace elements in coal by ICP−OES and ICP−MS. Anal. Chim. Acta 2004, 514 (1), 115−124. (58) Chang, C.-Y.; Wang, C.-F.; Mui, D. T.; Chiang, H.-L. Application of methods (sequential extraction procedures and highpressure digestion method) to fly ash particles to determine the element constituents: A case study for BCR 176. J. Hazard. Mater. 2009, 163 (2−3), 578−587. (59) Li, C.; Shi, Y.; Cai, N. Mechanism for carbon direct electrochemical reactions in a solid oxide electrolyte direct carbon fuel cell. J. Power Sources 2011, 196 (2), 754−763. (60) Zhao, X. Y.; Yao, Q.; Li, S. Q.; Cai, N. S. Studies on the carbon reactions in the anode of deposited carbon fuel cells. J. Power Sources 2008, 185 (1), 104−111. (61) Şenel, I. G.; Gürüz, A. G.; Yücel, H.; Kandas, A. W.; Sarofim, A. F. Characterization of pore structure of Turkish coals. Energy Fuels 2001, 15 (2), 331−338. (62) Cazorla-Amoros, D.; Alcaniz-Monge, J.; de la Casa-Lillo, M. A.; Linares-Solano, A. CO2 as an adsorptive to characterize carbon molecular sieves and activated carbons. Langmuir 1998, 14 (16), 4589−4596. (63) Radović, L. R.; Walker, P. L. Jr.; Jenkins, R. G. Importance of catalyst dispersion in the gasification of lignite chars. J. Catal. 1983, 82 (2), 382−394. (64) Linares-Solano, A.; Hippo, E. J.; Walker, P. L. Jr. Catalytic activity of calcium for lignite char gasification in various atmospheres. Fuel 1986, 65 (6), 776−779. (65) Yokokawa, H.; Yamaji, K.; Brito, M. E.; Kishimoto, H.; Horita, T. General considerations on degradation of solid oxide fuel cell anodes and cathodes due to impurities in gases. J. Power Sources 2011, 196 (17), 7070−7075. (66) Cheng, Z.; Liu, M. Characterization of sulfur poisoning of Ni− YSZ anodes for solid oxide fuel cells using in situ Raman microspectroscopy. Solid State Ionics 2007, 178 (13−14), 925−935. 1488


Review of Fuels for Direct Carbon Fuel Cells - Energy & Fuels (ACS

Recommend Documents