One-Step Fabrication of Protonic Ceramic Fuel Cells Using a Convenient Tape Calendering Method

: The present paper reports the preparation of multilayer protonic ceramic fuel cells (PCFCs) using a single sintering step. The success of this fabrication approach is due to two main factors: the rational choice of chemically and mechanically compatible components, as well as the selection of a convenient preparation (tape calendering) method. The PCFCs prepared in this manner consisted of a 30 µ m BaCe 0.5 Zr 0.3 Dy 0.2 O 3– δ (BCZD) electrolyte layer, a 500 µ m Ni–BCZD supporting electrode layer and a 20 µ m functional Pr 1.9 Ba 0.1 NiO 4 + δ (PBN)–BCZD cathode layer. These layers were jointly co-ﬁred at 1350 ◦ C for 5 h to reach excellent gas-tightness of the electrolyte and porous structures for the supported and functional electrodes. The adequate fuel cell performance of this PCFC design (400 mW cm − 2 at 600 ◦ C) demonstrates that the tape calendering method compares well with such conventional laboratory PCFC preparation techniques such as co-pressing and tape-casting.

Conventional FCs, including SOFCs and PCFCs, are comprised of a minimum of three functional components (layers): electrolyte, cathode and anode. Each layer requires its own unique processing regime [17][18][19]. In terms of the electrode-supported configuration of FCs, a thin electrolyte layer is formed onto a Ni-based anode and then these two layers are jointly fired at high temperatures to form a half-cell, combining the porous anode substrate and dense electrolyte. The next step consists of cathode formation onto the electrolyte surface followed by its sintering at a lower temperature, which produces the final FC [20][21][22][23]. As can be seen, the provided scheme includes several temperature regimes along with their various parameters (temperatures, heating/cooling rates, soaking times). Since the multi-stage preparation process of conventional FCs is time-, energy-and resource-intensive, the possibility of transition to a single-stage preparation process is of high interest.
Appl. Sci. 2020, 10, 2481 2 of 9 The one-step sintering of FCs represents an important technological strategy, allowing the entire FC to be produced using a single temperature regime. This strategy has recently been successfully used in the fabrication of both SOFCs [24,25] and PCFCs [26,27]. However, it should be noted that the FCs in the referenced studies were prepared by means of a co-pressing method requiring high pressing pressures. Although such a method can be quite easily used for fabricating the single cells with a small size, its adaptation for large-scale production presents severe challenges. Therefore, as an alternative approach, we herein propose a tape calendering method as a simple, scalable and efficient means for one-step fabrication of PCFCs. The possibility of one-step fabrication is based on our recent achievements in materials science (the selection of chemically and thermally compatible materials [28,29]) and technological (the rational parameter tuning of the tape calendering method [30,31]) fields.

Materials and Methods
In the present work, the tape calendering method was utilised to prepare the entire PCFC, which was then sintered in a single step. In order to achieve successful results, it was necessary to take several factors into consideration: (a) the electrolyte must be fully densified, while the electrodes should have a sufficient porosity; (b) all the functional materials need to exhibit similar thermomechanical properties; (c) no significant chemical interactions should occur at the anode/electrolyte and electrolyte/cathode interfaces. Taking these limitations into account, BaCe 0.5 Zr 0.3 Dy 0.2 O 3-δ + 0.5 wt% CuO (BCZD) was selected as a proton-conducting electrolyte [28,32], while Pr 1.9 Ba 0.1 NiO 4+δ (PBN) was used to form a cathode [29]. In detail, BCZD is characterised by good protonic conductivity levels, coupled with a high chemical stability in highly moisturised atmospheres. A small amount of copper oxide was added as an aid [33,34] to promote BCZD sinterability. PBN was selected due to its good chemical and thermal compatibilities with BCZD, even at 1350 • C [29].
The PCFC was fabricated according to the principal scheme presented in Figure 1. First, four individual powders were prepared for the corresponding functional layers: 1.
An anode supporting layer (ASL) was prepared by thorough the mixing of the NiO, BCZD and pore former (starch) powders in a weight ratio of 60:40:20; 2.
The anode functional layer (AFL) consisted of NiO and BCZD (without starch) powders mixed in a weight ratio of 55:45; 3.
An electrolyte layer (EL) was formed using the BCZD powder only; 4.
A cathode functional layer (CFL) was prepared by thorough mixing of the PBN, BCZD and starch in a weight ratio of 70:30:10.
The second stage included the preparation of mixtures of the corresponding powders in an organic binder (components and their concentrations are provided elsewhere [30]). These mixtures were dried overnight to remove evaporating solvents (acetone, benzene). The dried residues were rolled using a Durtson rolling mill to fabricate the corresponding film layers with the required thicknesses. Since the minimal gap between the rolls was~100 µm, the PCFC with a 30 µm-thick electrolyte was prepared by rolling the 100 µm-thick films several times: first EL with CFL, and then the obtained film with AFL to produce the 100 µm-thick structure AFL|EL|CFL. The green ASL|AFL|EL|CFL cell was prepared by co-rolling the latter film with a 500 µm-thick ASL film using a larger roll gap. Finally, the green cell was co-sintered according to the following regime: heating up to 600 • C with a heating rate of 1 • C min −1 , heating up to 1350 • C with a heating rate of 3 • C min −1 , isothermal soaking at 1350 • C for 5 h, and cooling down to room temperature with a cooling rate of 3 • C min −1 .
Scanning electron microscopy (SEM, TESCAN MIRA 3 LMU microscope, TESCAN, Czech Republic) coupled with energy-dispersive X-ray spectroscopy (EDX, INCA Energy 350 microanalysis system, Oxford Instruments Analytical, UK) was used to analyse the microstructural parameters of the PCFC.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 3 of 9 system, Oxford Instruments Analytical, UK) was used to analyse the microstructural parameters of the PCFC. The fabricated PCFC was electrochemically characterised utilising a complex of Amel 2550 potentiostat/galvanostat (Amel, Italy) and a Materials M520 frequency material response analyser (Amel, Italy) to obtain data on volt-ampere and impedance dependences. The impedance spectra were collected in a frequency range of 3·10 4 -5·10 −2 Hz with an acceleration voltage of 30 mV and a cell bias corresponding to open circuit voltage (OCV) or maximal power density (Pmax) conditions. Humid air with water vapour partial pressure (pH2O) of 0.30 atm was used as an oxidant, while wet hydrogen (pH2O = 0.03 atm) was used as a fuel. The required pH2O values were achieved by feeding the corresponding gases through the water bubbler thermostated at certain temperatures (69 and 25 °C, respectively). The electrochemical characterisation of the PCFC was carried out at temperatures varying from 550 to 700 °C ; the cell was equilibrated for 1 h after each temperature change.

Results and Discussion
The electrode-supported PCFC fabricated via the tape calendering method followed by one-step sintering and was characterised in terms of its microstructure. According to Figure 2, the thicknesses of AFL, EL and CFL were estimated to be 40, 30 and 20 μm, respectively; while the thickness of the entire PCFC was ~600 μm. The layers exhibit essential properties, including a dense state for the electrolyte and a porous structure for the functional cathode and anode; moreover, there is a good interface contact (adhesion) for both AFL/EL and EL/CFL couples. These results imply that the combination of the used methods, temperature regimes and material compositions was quite well optimised. Here we should stress that the present example of an asymmetric PCFC, co-sintered during a single stage, represents one of the initial reports, achieving successful results comparable to The fabricated PCFC was electrochemically characterised utilising a complex of Amel 2550 potentiostat/galvanostat (Amel, Italy) and a Materials M520 frequency material response analyser (Amel, Italy) to obtain data on volt-ampere and impedance dependences. The impedance spectra were collected in a frequency range of 3·10 4 -5·10 −2 Hz with an acceleration voltage of 30 mV and a cell bias corresponding to open circuit voltage (OCV) or maximal power density (P max ) conditions. Humid air with water vapour partial pressure (pH 2 O) of 0.30 atm was used as an oxidant, while wet hydrogen (pH 2 O = 0.03 atm) was used as a fuel. The required pH 2 O values were achieved by feeding the corresponding gases through the water bubbler thermostated at certain temperatures (69 and 25 • C, respectively). The electrochemical characterisation of the PCFC was carried out at temperatures varying from 550 to 700 • C; the cell was equilibrated for 1 h after each temperature change.

Results and Discussion
The electrode-supported PCFC fabricated via the tape calendering method followed by one-step sintering and was characterised in terms of its microstructure. According to Figure 2, the thicknesses of AFL, EL and CFL were estimated to be 40, 30 and 20 µm, respectively; while the thickness of the entire PCFC was~600 µm. The layers exhibit essential properties, including a dense state for the electrolyte and a porous structure for the functional cathode and anode; moreover, there is a good interface contact (adhesion) for both AFL/EL and EL/CFL couples. These results imply that the combination of the used methods, temperature regimes and material compositions was quite well optimised. Here we should stress that the present example of an asymmetric PCFC, co-sintered during a single stage, represents one of the initial reports, achieving successful results comparable to those obtained via conventional sintering at such high temperatures as 1350 • C. In other cases [26,27], asymmetric-type PCFCs were fabricated at lower sintering temperatures (below 1250 • C), which, as a rule, are insufficient for the required densification of Zr-containing (especially, Zr-enriched) BaCeO 3 materials.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 4 of 9 those obtained via conventional sintering at such high temperatures as 1350 °C . In other cases [26,27], asymmetric-type PCFCs were fabricated at lower sintering temperatures (below 1250 °C), which, as a rule, are insufficient for the required densification of Zr-containing (especially, Zr-enriched) BaCeO3 materials.

Figure 2.
Cross-section of the fabricated PCFC with the 30 μm-thick BaCe0.5Zr0.3Dy0.2O3-δ (BCZD) electrolyte and corresponding energy-dispersive X-ray spectroscopy (EDX) analysis. Figure 3 shows the volt-ampere (U-j) characteristics of the PCFC recorded under OCV conditions (when j = 0) to ~100 mV with a voltage step of 25 mV. The high current density conditions, corresponding to a voltage range of 0-100 mV, were used in order to prevent significant heating of the cell. Analysing the free-current conditions, the cell generates OCVs varying from 1.040 V at 550 °C to 0.935 V at 700 °C. At low temperatures, these OCVs are very close to the theoretical values (1.056 V at 550 °C), but start to deviate from them (by ~80 mV at 700 °C) at higher temperatures, indicating the electron transport of the BCZD membrane [35]. Here, the theoretical levels were estimated on the basis of a Nernstian equation for a pure protonic conductor (EH = RT/2F·ln[p'H2/p''H2]). The PCFC load results in the generation of Pmax, which increases from 290 mW cm −2 at 550 °C to 470 mW cm −2 at 700 °C. The EIS analysis was further utilised to reveal the main factors affecting the overall fuel cell performance. For this purpose, the impedance spectra were recorded in two conditions. One of these corresponds to the conventional measurements at OCV, while the other is related to realising Pmax.  Figure 3 shows the volt-ampere (U-j) characteristics of the PCFC recorded under OCV conditions (when j = 0) to~100 mV with a voltage step of 25 mV. The high current density conditions, corresponding to a voltage range of 0-100 mV, were used in order to prevent significant heating of the cell. Analysing the free-current conditions, the cell generates OCVs varying from 1.040 V at 550 • C to 0.935 V at 700 • C. At low temperatures, these OCVs are very close to the theoretical values (1.056 V at 550 • C), but start to deviate from them (by~80 mV at 700 • C) at higher temperatures, indicating the electron transport of the BCZD membrane [35]. Here, the theoretical levels were estimated on the basis of a Nernstian equation for a pure protonic conductor (E H = RT/2F·ln[p H 2 /p"H 2 ]). The PCFC load results in the generation of P max , which increases from 290 mW cm −2 at 550 • C to 470 mW cm −2 at 700 • C.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 4 of 9 those obtained via conventional sintering at such high temperatures as 1350 °C . In other cases [26,27], asymmetric-type PCFCs were fabricated at lower sintering temperatures (below 1250 °C), which, as a rule, are insufficient for the required densification of Zr-containing (especially, Zr-enriched) BaCeO3 materials.

Figure 2.
Cross-section of the fabricated PCFC with the 30 μm-thick BaCe0.5Zr0.3Dy0.2O3-δ (BCZD) electrolyte and corresponding energy-dispersive X-ray spectroscopy (EDX) analysis. Figure 3 shows the volt-ampere (U-j) characteristics of the PCFC recorded under OCV conditions (when j = 0) to ~100 mV with a voltage step of 25 mV. The high current density conditions, corresponding to a voltage range of 0-100 mV, were used in order to prevent significant heating of the cell. Analysing the free-current conditions, the cell generates OCVs varying from 1.040 V at 550 °C to 0.935 V at 700 °C. At low temperatures, these OCVs are very close to the theoretical values (1.056 V at 550 °C), but start to deviate from them (by ~80 mV at 700 °C) at higher temperatures, indicating the electron transport of the BCZD membrane [35]. Here, the theoretical levels were estimated on the basis of a Nernstian equation for a pure protonic conductor (EH = RT/2F·ln[p'H2/p''H2]). The PCFC load results in the generation of Pmax, which increases from 290 mW cm −2 at 550 °C to 470 mW cm −2 at 700 °C. The EIS analysis was further utilised to reveal the main factors affecting the overall fuel cell performance. For this purpose, the impedance spectra were recorded in two conditions. One of these corresponds to the conventional measurements at OCV, while the other is related to realising Pmax. The EIS analysis was further utilised to reveal the main factors affecting the overall fuel cell performance. For this purpose, the impedance spectra were recorded in two conditions. One of these corresponds to the conventional measurements at OCV, while the other is related to realising P max . Despite the P max conditions describing the real operation of a fuel cell, this function has yet to be analysed in detail in the literature. The EIS results presented in Figure 4 indicate that the spectra have different shapes depending on the measurement temperatures and/or cell bias. In detail, at least three parts (described by RQ elements, where R is the resistance, Q is the constant phase element) can be clearly distinguished at low temperatures; these merge to enable the detection of at least two (low-and high-frequency) processes at higher temperatures. Due to the revealed complexity, only the overall polarisation resistance (R p ) was calculated in order to analyse the electrochemical behaviour of the PCFC electrodes. Considering the temperature effect (Figure 4a), a typical decrease in both ohmic resistance (R o ) and R p is observed, which is in line with the thermo-activated nature of ongoing processes. However, when the bias is changed from OCV conditions to PCFC load, the following interesting tendencies occur: R o attains almost constant values, but the R p level grows more noticeably, implying a deterioration in the electrochemical behaviour. The latter agrees with a number of studies [36][37][38], which report similar R p growth with decreasing cell bias. The observed R p deterioration might either be associated with a decrease in the p-type electronic conductivity of the proton-conducting electrolyte, thus providing a lower current distribution at the EL/CFL interface [39,40], or with an increase in the concentration type of R p taking place at high current densities [41,42].
Appl. Sci. 2020, 10, x FOR PEER REVIEW 5 of 9 Despite the Pmax conditions describing the real operation of a fuel cell, this function has yet to be analysed in detail in the literature. The EIS results presented in Figure 4 indicate that the spectra have different shapes depending on the measurement temperatures and/or cell bias. In detail, at least three parts (described by RQ elements, where R is the resistance, Q is the constant phase element) can be clearly distinguished at low temperatures; these merge to enable the detection of at least two (low-and high-frequency) processes at higher temperatures. Due to the revealed complexity, only the overall polarisation resistance (Rp) was calculated in order to analyse the electrochemical behaviour of the PCFC electrodes. Considering the temperature effect (Figure 4a), a typical decrease in both ohmic resistance (Ro) and Rp is observed, which is in line with the thermo-activated nature of ongoing processes. However, when the bias is changed from OCV conditions to PCFC load, the following interesting tendencies occur: Ro attains almost constant values, but the Rp level grows more noticeably, implying a deterioration in the electrochemical behaviour. The latter agrees with a number of studies [36][37][38], which report similar Rp growth with decreasing cell bias. The observed Rp deterioration might either be associated with a decrease in the p-type electronic conductivity of the proton-conducting electrolyte, thus providing a lower current distribution at the EL/CFL interface [39,40], or with an increase in the concentration type of Rp taking place at high current densities [41,42].  Figure 5 depicts the absolute values of Ro, Rp, Rt along with their ratio. As can be seen, the overall PCFC resistance is equal to 0.86 Ω cm 2 at 550 °C and 0.43 Ω cm 2 at 700 °C for the OCV conditions. At the same time, the Rp contribution is lower than 50% for all measured temperatures; this indicates that the parameters of the electrolyte (resistance and, especially, thickness) are dominant factors affecting the performance of the PCFC. When the cell bias corresponds to the Pmax conditions, the overall PCFC resistance rises to 0.97 Ω cm 2 at 550 °C and 0.48 Ω cm 2 at 700 °C due to an increase in Rp (its contribution attains ~51% at 550 °C and 16% at 700 °C). According to the obtained data, further research efforts should be aimed at obtaining an improvement in electrode behaviour under load rather than the conventional OCV conditions.  Figure 5 depicts the absolute values of R o , R p , R t along with their ratio. As can be seen, the overall PCFC resistance is equal to 0.86 Ω cm 2 at 550 • C and 0.43 Ω cm 2 at 700 • C for the OCV conditions. At the same time, the R p contribution is lower than 50% for all measured temperatures; this indicates that the parameters of the electrolyte (resistance and, especially, thickness) are dominant factors affecting the performance of the PCFC. When the cell bias corresponds to the P max conditions, the overall PCFC resistance rises to 0.97 Ω cm 2 at 550 • C and 0.48 Ω cm 2 at 700 • C due to an increase in R p (its contribution attains~51% at 550 • C and 16% at 700 • C). According to the obtained data, further research efforts should be aimed at obtaining an improvement in electrode behaviour under load rather than the conventional OCV conditions. that the parameters of the electrolyte (resistance and, especially, thickness) are dominant factors affecting the performance of the PCFC. When the cell bias corresponds to the Pmax conditions, the overall PCFC resistance rises to 0.97 Ω cm 2 at 550 °C and 0.48 Ω cm 2 at 700 °C due to an increase in Rp (its contribution attains ~51% at 550 °C and 16% at 700 °C). According to the obtained data, further research efforts should be aimed at obtaining an improvement in electrode behaviour under load rather than the conventional OCV conditions. To illustrate the prospect of the tape calendering method as a means of fabricating PCFCs via a one-step procedure, a comparative analysis was performed ( Table 1). As can be seen, the rational selection of the functional materials, as well as the utilisation of convenient laboratory technology, allow PCFC fabrication with promising performance across the intermediate temperature range. It should be noted that the effect of ohmic resistance is excluded from consideration in order to provide a clear relationship between the polarisation resistance and power density. From this viewpoint, the achieved electrode activity competes with the electrochemical activity of electrodes as prepared by other methods, although the number of temperature regimes is reduced to only one. 1 Abbreviations: temperature regimes (TR), tape calendaring (TC), dry-pressing (DP), painting (P), dip-coating (DC), screen-printing (SP), phase inversion (PI). 2 These data are provided for 700 • C.

Conclusions
The present work summarises technological advances in the fabrication of electrode-supported protonic ceramic fuel cells (PCFCs) using a single temperature regime. This achievement was realised due to the utilisation of the tape calendering (co-rolling) method recently adopted for solid oxide fuel cells based on proton-conducting electrolyte membranes. The PCFC fabricated in this way demonstrated both gas-tightness (1.03 V at 600 • C in the current-free mode) and high performance (~400 mW cm −2 at 600 • C) as a result of achieving the required morphological properties-full densification of the electrolyte and sufficient electrode porosity.