Comparison of literature studies of total conductivity of BCY10 in nominally dry and wet H2 at 400°C  (reproduced by permission of The Royal Society of Chemistry).
Proton-conducting perovskites are among the most promising electrolytes for Proton Ceramic Fuel Cells (PCFCs), electrolysers and separation membranes. Particularly, yttrium-doped barium cerate, BaCe1-xYxO3-δ (BCY), shows one of the highest protonic conductivities at intermediate temperatures (σ ∼ 10−3 S cm−1 at 400°C); values that are typically achieved under humidified atmospheres (p H2O ∼ 10−2 atm). However, BCY has commonly been discarded for such applications due to its instability in the presence of water vapour and carbonaceous atmospheres. A recent discovery has shown that BCY10 exhibits pure protonic conductivity under very low humidity contents (∼10−5–10−4 atm), owing to its very high equilibrium constant for hydration. This peculiar characteristic allows this material to retain its functionally as a proton conductor in such conditions, while preventing its decomposition. Hence, this chapter explores the electrochemical properties of the BaCe0.9Y0.1O3-δ (BCY10) composition, comprehensively establishing its limiting operation conditions through defect chemistry and thermodynamic analyses. Moreover, the importance of such conditions is highlighted with respect to potential industrially relevant hydrogenation/de-hydrogenation reactions at low temperatures under low humidity.
- barium cerate
- protonic conductivity
- transport number
- nominally dry conditions
Ceramic proton conductors have been highlighted for electrochemical synthesis, as potential membranes in hydrogenation and dehydrogenation reactions . One of the best compositions for this role is that of the doped barium cerate, e.g. BaCe1-xMxO3-δ (M = Y3+, In3+, Gd3+, etc.), which can show very high levels of proton conductivity at intermediate temperatures (i.e. σ ∼ 10−3 S cm −1 at 400°C) [2, 3, 4, 5, 6, 7, 8, 9]. This material belongs to the perovskite family with ABO3 ceramic oxide structure, including a divalent alkaline earth element, such as Ba2+ (also, Sr2+ or Ca2+), in the A-cation site, while a tetravalent rare-earth element, Ce4+, is present in the B-cation site. The introduction of dopants in the B-site with suitable acceptor elements, such as Y3+, In+3 or Gd3+ trivalent cations, leads to the formation of charge compensating oxygen vacancies :
In addition to potential oxide-ion conductivity, these acceptor-substituted materials are also capable of offering both protonic and electronic conductivity, depending on the temperature and atmospheric conditions. The protonic conductivity is the most significant characteristic of these materials that is usually associated with the existence of protonic defects (), upon filling of these oxygen vacancies in the presence of water vapour, as expressed by Eq. (2) [10, 11, 12, 13]:
Accordingly, the equilibrium constant for hydration, K w, is given by the following equation:
Due to the significant importance of humidity to promote protonic conductivity, most of the reported studies of barium cerate based materials have focused on highly wetted atmospheres with typical water vapour partial pressure p H2O ∼ 3 × 10−2 atm [14, 15, 16, 17, 18]. Unfortunately, these works also underline the tendency of this material for reacting with acidic gases, viz. carbon dioxide (CO2) and water vapour (H2O), leading to the formation of insulating carbonate or hydroxide phases, respectively, on the surface of the material. This complication impedes the ability of this material to be used in highly humidified and carbon-based fuels, thus, limiting its potential application range [3, 14, 15, 16, 17, 18, 19, 20]. The typical degradation reactions in such atmospheres include:
The chemical stability of doped barium cerates is well documented in the literature and huge efforts have been made to explore the reasons behind its chemical instability, using both conventional and non-conventional techniques [21, 22, 23, 24, 25]. For instance, Matsumoto et al.  studied the effect of dopant M in BaCe0.9M0.1O3-δ (M = Y, Tm, Yb, Lu, In, or Sc) on the electrical conductivity in the temperature range 400–900°C and on the chemical stability with respect to CO2 by thermogravimetry (TG). Both the electrical conductivity (moistened H2 or O2, p H2O = 1.9 × 10−2 atm) and the stability against carbonate formation were shown to decrease with increasing ionic radius (Figure 1), corresponding to an increase in basicity. Nonetheless, all compounds were found to interact with pure CO2 at temperatures below 900°C, failing to succeed in the mitigation of the chemical instability in the doped barium cerate.
Against this scenario, one common alternative is the use of the barium zirconates or compounds containing both Ce and Zr elements, where the introduction of Zr can significantly increase their chemical stability. Nonetheless, it has also been demonstrated that increased amounts of Zr negatively impact the total conductivity of these materials, due to an increase in their refractive nature and in their grain growth, which aggravate the problem of resistive grain boundaries. As such, much lower values of total conductivity are, typically, reported for the zirconate materials than for their cerate analogues, even though their bulk protonic conductivities are actually greater [9, 26, 27, 28, 29, 30, 31, 32, 33].
More recently, the work of Kim et al.  reported that the chemical instability of the barium cerates is due to the presence of a nanometre-thick amorphous phase found at the grain boundaries in proton-conducting BaCeO3 polycrystals, which not only leads to a reduced proton mobility, but also can act as a penetration path for H2O and CO2 gas molecules, facilitating chemical decomposition and collapse of the microstructure (Figure 2a). Furthermore, this effect could be minimised by controlling the composition to obtain Ba-deficient samples in which the intergranular amorphous layer could be minimised, leading to a mitigation of the reactivity with such gases (Figure 2b). The presence of an amorphous layer on the interfaces between grains has also been documented in barium zirconate-based compositions [26, 35], where this feature can exert significant complications during fabrication of complete electrochemical cells [19, 36].
In summary, the high electrical conductivity and the facile processing of the doped barium cerates demands further investigation to succeed to overcome their limited stabilities. In fact, it is only very recently that research in these materials has moved towards a more fundamental and, yet, critical aspect, concerning a deeper understanding of the limiting atmospheric conditions that are necessary to retain their functionality. Taking this into account, Loureiro et al.  reanalysed the barium cerate stability limits by thermodynamic calculations, considering its decomposition products in the presence of water vapour and CO2 (Figure 3). According to this theoretical study, no degradation would be expected for humidity values of ∼3 × 10−2 atm and temperatures higher than ∼500°C. However, when considering the formation of barium carbonate (Figure 3), the thermodynamics predict that much stricter conditions need to be applied, where only very low partial pressures of CO2 (e.g. p CO2 < ∼10−8 atm at 400°C) are able to avoid barium cerate degradation.
For this reason, only very few reports can be found on successful applications of BCY membranes for chemical reactions. Most of these have concerned, ammonia synthesis [39, 40, 41], or the conversion of propane to propylene . In these cases, no chemical instability has been reported and the survival of the BCY material is likely to be related to the effective absence of CO2 or significant water vapour in these operations. To understand this further, Figure 4 presents the maximum water vapour partial pressure (p H2O) that could be tolerated in different carbonaceous atmospheres to provide an equilibrium partial pressure of CO2 that remains below that of the BCY stability limit. These results demonstrate that, for example, at 400°C, these values should range between the values of 10−3 < p H2O < ∼ 10−4 atm in order to avoid decomposition of the perovskite phase, for the potential hydrocarbon atmospheres of CH4, C3H8 or C6H6 .
Nonetheless, one of the requirements for operating in such low water vapour partial pressures is that the protonic conductivity must be maintained in order to ensure the functionality of the electrolyte membrane in these applications. In this respect, protonic conductors are complex materials as they are capable to offer mixed conductivity (protonic, oxide-ion and electronic), depending on the temperature and on the nature of the surrounding atmosphere [37, 38]. One of the most promising compositions for this type of application is that of the yttrium-doped barium cerate, BaCe1-xYxO3-δ (BCY), which has very high protonic conductivity at lower temperatures under humidified atmospheres (e.g. ∼10−3 S cm−1 at 400°C, p H2O ∼ 10−2 atm) [1, 38].
Therefore, the current chapter will be focus on the electrochemical transport properties of the BaCe0.9Y0.1O3-d (BCY10) in reducing and oxidising conditions when operating in very low humidity levels. The aim of this chapter is to comprehensively explain the working limits of BCY10 and to assess its applicability as an electrolyte membrane for fuel cell, electrolysers and other electrochemical-based applications, with special focus on operation under low water vapour partial pressures.
2. Electrochemical properties of BCY10 in nominally dry reducing conditions
Figure 5 depicts the total conductivity of BCY10 analysed by impedance spectroscopy between 100 and 500°C in H2, 10%H2-N2 and N2, highlighting that no significant differences can be observed in the conductivity measured under these atmospheres. In addition, at the higher temperature range, a notable decrease of the activation energy is observed in all cases, as a result of the exsolution of protons from the structure of BCY10, and the concomitant decrease of the protonic contribution to the electrical transport . Interestingly, and also surprisingly at first sight, current results of total conductivity in nominally dry H2 are close to those corresponding to the available data in literature for humidified H2 (Table 1).
To analyse the contribution of electronic conductivity to this material in nominally dry conditions, the total conductivity was also analysed as a function of hydrogen partial pressure (p H2) , as shown in Figure 5b. A slight increase in total conductivity can be observed towards higher p H2 values.
To be able to understand this behaviour, firstly the potential for an electronic component to conductivity must be assessed. In reducing conditions (e.g. H2-containing atmospheres), the cerium cations from the B-site of the perovskite structure of BCY10 can reduce from a higher oxidation state, Ce4+, to a lower one, Ce3+, altering the contribution of the concentration of the electronic charge carriers. This phenomenon is well documented in the literature for various cerium-based compositions [37, 38, 44, 45, 46, 47], being described as small-polaron electronic conductivity (i.e., a localised, mobile electron, ). Due to the high mobility of electronic conductors, such electronic contribution can exceed that of the ionic, under very reducing conditions and high temperatures [37, 45, 46, 47]. However, in the case of BCY10, the extent of cerium reduction has been assessed by Loureiro et al. , who performed coulometric titration measurements to study the potential role of electronic contribution in BaCe0.9Y0.1O3-δ in reducing conditions as a function of temperature. This technique has been widely adopted to quantify the changes in the oxygen non-stoichiometry (Δδ), which can be associated with the reduction of Ce4+ to Ce3+, following the equation:
with the equilibrium constant for reduction reaction given by:
The results of coulometric titration (Figure 6)  show considerable variations of Δδ with oxygen partial pressure only at very high temperature, with a lower impact as temperature decreases. Thus, Figure 6 demonstrates that very extreme reducing conditions and very high temperatures are required to produce appreciable increase in the oxygen-vacancy and electronic concentrations in BCY [47, 48]. These results contrast with those of fluorite-ceria-based materials which usually show high reducibility under milder conditions [46, 49].
Thus, to take the possibility of reduction into account, the methodology applied by Loureiro et al.  for the determination of reduction equilibrium follows the method reported elsewhere , as described below.
The corresponding mass action constant (Eq. (7)) can be combined with the electroneutrality condition:
and other mass and lattice position restrictions, on neglecting defect interactions and assuming nearly ideal behaviour, with the following relations between the concentrations of relevant species, stoichiometric changes (Δδ), and fraction of trivalent additive (x):
where Z is the number of atoms per unit cell and n0, the unit cell volume. Substitution in Eq. (7) leads to the values of the equilibrium constant for reduction (KR) from the entire range of values of Δδ versus p O2 at a given temperature T:
The following equation was then determined to describe the temperature dependence of K R, from the results of oxygen-nonstoichiometry shown in Figure 6:
with an enthalpy for reduction, ΔH R = 804.99 kJ mol−1. This value is significantly higher than those obtained by other authors for fluorite ceria-based materials (Table 2) [46, 49], underscoring the low reducibility of BCY10 in such conditions from intermediate to low temperatures.
On the basis of these results, the potential rehydration of the BCY10 material was then assessed by thermogravimetric experiments . Figure 7 depicts the concentration of protonic charge carriers as a function of temperature, calculated from the following methodology.
By expressing the equilibrium constant for water incorporation reaction (Eq. (3)) in terms of entropy, ΔS w, and enthalpy, ΔH w:
where T and R have usual meanings. Given Eq. (1) and knowing that the number of oxygen sites per formula unit of barium cerate is restricted to 3, implying the site restriction relationship:
and then, the concentration of protonic defects is given by
where K′ = K w.p H2O and S = [YCe’]. Because the formation of protonic defects is usually accompanied by a significant weight increase, the concentration of protonic defects as a function of temperature and water vapour partial pressure is generally measured by thermogravimetric analysis (TG). From Figure 7, one can observe an increase in the concentration of protonic species as a function of decreasing temperature, even in nominally dry 10%H2/N2. This factor is most likely related to the intrinsic formation of water vapour under the presence of hydrogen and oxygen impurities in the feed stream:
This result emphasises the existence of protonic conductivity in nominally dry hydrogen-containing atmospheres, as even trace amounts of oxygen can form water vapour, potentially contributing to the hydration of the BCY10 material. Hence, the partial conductivities can be obtained by combining the results from both coulombic titration and TG experiments using a defect chemistry methodology .
Figure 8 shows the partial conductivities of all species (protons, oxide-ions and electrons) obtained at the temperature range (350–600°C) in nominally dry H2 (p H2O ∼ 10−4 atm). One can observe a dominance of the ionic charge carriers over the electronic carriers in the whole temperature range, corroborating the negligible reducibility of cerium cations measured by coulombic titration (Figure 6).
Furthermore, at the low temperature range (350–400°C), the dominance of protonic conductivity is related to the high equilibrium constant for water incorporation in BCY10, allowing a significant hydration even at p H2O values as low as ∼10−4 atm , as confirmed by TG (Figure 7). This behaviour also explains the slight p H2 dependence of conductivity shown in Figure 5b that is due, not to electronic behaviour, but to changes in the effective water vapour partial pressure arising from Eq. (19) and subsequent slight increase in ionic conductivity due to a higher level of hydration Eq. (18). In contrast at higher temperatures in the (550–600°C) range, oxide-ion conductivity starts to become dominant at due to the loss of protons from the structure (Figure 7).
3. Electrochemical properties of BCY10 in low humidity oxidising conditions
The transport numbers of BCY10 in oxidising atmospheres were firstly studied by Oishi et al.  and by Grimaud et al. . Later, Lim et al.  determined the concentration of charge carriers in BCY10 by thermogravimetric analysis (TGA) under two different humidity conditions (dry and wet, p H2O ∼ 10−5 and 10−3 atm, respectively). More recently, Loureiro et al.  focused on the determination of the transport properties of this composition at temperatures below 600°C and under very low humidity levels (p H2O ≤ 10−4 atm).
In oxidising conditions, the absence of hydrogen species, shifts the water formation reaction, Eq. (19), away from the water product, leading to a lower intrinsic water vapour partial pressure that can, in turn, decrease the protonic transport number . Therefore, at the intermediate temperature range, 350–600°C, it is necessary to externally add humidity to guarantee a sufficient level of protonic conductivity. Moreover, BCY10 is known to possess p-type electronic conductivity in oxidising atmospheres, which can importantly impact the total conductivity in these conditions , as expressed by
with the following mass action constant
Figure 9 shows the total conductivity of BCY10 measured in the temperature range 350–600°C in wet and low humidity O2 and N2. From Figure 9, this expected decrease in the concentration of protonic species is corroborated, as in both, N2 and O2, total conductivity is shown to be higher in wet conditions (p H2O ∼ 10−3 atm) than in low humidity conditions (p H2O ∼ 10−7 atm). It is also possible to observe that low humidity N2 (p H2O ∼ 10−7 atm) the total conductivity is lower in the whole measured temperature range in comparison to wet N2 (p H2O ∼ 10−3 atm), as a result of dehydration of the sample according to Eq. (22). In contrast, in O2, the total conductivity in low humidity and wet conditions are similar, particularly at higher temperatures, a factor that can be explained due to the presence and dominance of p-type electronic conductivity [57, 58] (see Eq. (20)):
In agreement, the presence of p-type electronic conductivity can explain the slightly higher activation energy registered in low humidity O2, 0.49 eV, in comparison to the other studied atmospheres.
Figure 10 illustrates the partial conductivities obtained in wet (p H2O ∼ 10−3 atm) and low humidity (p H2O ∼ 10−7 atm) conditions in N2 and O2. Figure 10a and b show that in moderate wet conditions (p H2O ∼ 10−3 atm) the protonic conductivity is dominating in both atmospheres with activation energies similar to that obtained for the protonic conduction (∼0.4–0.5 eV) [16, 17]. In contrast, in low humidity conditions (Figure 10c and d) a drop on protonic conductivity with increasing temperature is observed, due to predominant oxide-ion conductivity in both atmospheres. In the case of hole conductivity, the activation energies obtained were found to be lower at low humidity conditions (0.61–1.03 eV, T = 350–500°C) in comparison with those obtained in wet conditions (1.29–1.75 eV, T = 350–500°C). This can be explained by the creation of electronic defects (Eq. (20)), upon filling the oxygen vacancies.
4. Comparison between reducing and oxidising conditions under low humidity
As discussed previously, to maximise the protonic conductivity is necessary to maintain a minimum level of humidity in the order of 10−4–10−5 atm. It is also important to emphasise that, while this level of humidity is intrinsically formed in nominally dry hydrogen-containing atmospheres, in the case of oxidising atmospheres this level must be externally supplied. A comparison of the partial conductivities in all cases is shown in Figure 11, for p H2O ∼ 10−4 atm. At temperatures below 450°C, the total conductivity is dominated by protonic conductivity, with the oxide-ion conductivity taking a negligible role. In contrast, at higher temperatures (T > 450°C), the oxide- ion conductivity dominates the total conductivity with a simultaneous decrease of protonic conductivity. With respect to the electronic conductivity, this term increases as p O2 increases, being only relevant in oxidising conditions and/or high temperatures. This can be explained due to the creation of electronic holes, which become more relevant with increasing p O2 and temperature (Eq. (20)).
Overall, BCY10 is shown to be a predominant protonic conductor in both reducing and oxidising atmospheres at sufficiently low temperatures ≤500°C, even under relatively low water vapour partial pressures (p H2O ∼ 10−4–10−5 atm). Moreover, the level of conductivity measured at 400°C in these conditions is high, e.g. ∼10−3 S cm−1. The origin of protonic conductivity is due to a high equilibrium constant for water absorption that allows this material to offer high bulk protonic conductivity at intermediate temperatures in these very low humidity conditions. From Figure 12, one can immediately envisage that this is a particular behaviour of BCY10 that cannot be obtained in other competing proton-conducting perovskites, due to their much lower values of K w.
This is a very exciting result since it opens a wide range of possibilities for using the BCY material, in particular, in different applications that involve very low humidity levels and low temperatures of operation. The most well-known is that of ammonia electrochemical synthesis [39, 40, 41], although many other processes concerning hydrogenation and de-hydrogenation reactions can also be considered that agree with these operating conditions (Table 3).
The current chapter highlights that the transport properties of BaCe0.9Y0.1O3-δ (BCY10) in very low humidity conditions are dependent on the nature of the surrounding atmosphere and on the temperature, being significantly different in reducing and oxidising conditions and at high and low temperatures. In reducing conditions, BCY10 shows a very high protonic conductivity (e.g. ∼ 10−3 S cm−1) at low temperatures i.e. < 400°C, even in nominally dry atmospheres with negligible oxide-ion/electronic influence.
In the other hand, in oxidising conditions, the same behaviour can only be obtained by externally supplying humidity in the range (p H2O ∼ 10−4–10−5 atm) at low temperatures ≤500°C. At higher temperatures, at this low humidity, the onset of hole conductivity can be noted at higher oxygen partial pressures due partial material dehydration.
The present discussion shows the importance of controlling the humidity levels in order to maximise the protonic conductivity of BCY under operation. The very low levels of humidity required (p H2O ∼ 10−4–10−5 atm), to ensure predominant proton conductivity in both reducing and oxidising atmospheres at low temperatures ≤500°C, are highly interesting as they highlight the possibility of using this composition in applications where low humidity levels and temperatures are required, such as the suggested de-hydrogenation/hydrogenation chemical reactions, while maintaining its stability against decomposition.
The authors acknowledge Fundação para a Ciência e Tecnologia (FCT) for the PhD grants – PD/BDE/142837/2018, SFRH/BD/130218/2017, and PD/BDE/114353/2016. The authors also acknowledge the projects UID/EMS/00481/2019-FCT and CENTRO-01-0145-FEDER-022083 - Centro Portugal Regional Operational Programme (Centro2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF).