Proton-Conducting Ceramic Membranes for the Production of Hydrogen via Decarbonized Heat: Overview and Prospects

Abstract

Proton-conducting ceramic membranes show high hydrogen ion conductivity in the temperature range of 300–700 °C. They are attracting significant attention due to their relevant characteristics compared to both higher-temperature oxygen ion-conducting ceramic membranes and lower-temperature proton-conducting polymers. The aim of this review is to integrate the fundamentals of proton-conducting ceramic membranes with two of their relevant applications, i.e., membrane reactors (PCMRs) for methane steam reforming (SMR) and electrolysis (PCEC). Both applications facilitate the production of pure H₂ in the logic of process intensification via decarbonized heat. Firstly, an overview of various types of hydrogen production is given. The fundamentals of proton-conducting ceramic membranes and their applications in PCMRs for SMR and reversible PCEC (RePCEC), respectively, are given. In particular, RePCECs are of particular interest when renewable power generation exceeds demand because the excess electrical energy is converted to chemical energy in the electrolysis cell mode, therefore representing an appealing solution for energy conversion and grid-scale storage.

1. Introduction

The world faces fundamental choices that will determine whether in the framework of global warming the 1.5 °C path, as defined at COP26, will remain within reach. Electrification and efficiency are key resulting in a cut of nearly 37 gigatonnes in annual CO₂ emissions by 2050 [1]. These reductions can be achieved through (1) increases in production and uses of renewables-based electricity; (2) improvements in energy efficiency [2]; (3) electrification of end-use sectors such as electric vehicles and heat pumps; (4) clean hydrogen and its derivatives [3,4]; (5) bioenergy coupled with carbon capture and storage [5]; and (6) the use of carbon capture and storage (Figure 1).

In this context, the decarbonization of heat is a cross-cutting opportunity. Many recent studies have emphasized its crucial role [6,7,8]. Indeed, a large portion of industrial sector emissions, estimated at about 7.5 Gt of CO₂ [9], about 21% of global greenhouse gas (GHG) emissions [10], derive from the generation of over 100 EJ_th of thermal energy. The combustion of three fuels, i.e., coal, natural gas and oil, generates the vast majority of this heat and it’s associated with CO₂ emissions.

Thiel and Stark [11] discussed four strategies for decarbonizing heat production, i.e., zero-carbon fuels, zero-carbon heat sources, the electrification of heat and better heat management, as schematized in Figure 2.

Among these four pathways, there is electrification of heat, i.e., the generation of thermal energy from electricity. It must be emphasized that the greenhouse gas benefits of the electrification of heat are strictly related to grid decarbonization; in other words, electrified heat is decarbonized heat only when electricity is generated without carbon emissions. As the grid moves toward 100% low-carbon electricity, the GHG benefits of electrification will increase, but until then, it is important to recognize the proportion of low-carbon energy in the grid to understand the magnitude of electrification benefits in the near future [1]. Even though a rapid increase in power generation from solar and wind was observed covering in summer up to 30–40% of the total energy demand at the expense of traditional energy sources from fossil fuels [12], the intermittent nature of these renewable energies has hastened the need for low-cost storage over a wide range of time scales, from seconds to days and even seasonal storage. Energy storage systems as well as decarbonizing industrial processes are two current grand challenges to be faced [7]. In addition to direct electrical energy storage such as pumped hydroelectricity, batteries, flow batteries [13], compressed air and emerging options with direct ties to the grid [14], there is the indirect approach of converting variable energy like wind and solar to energy carriers (chemical energy) such as hydrogen (H₂), ammonia, and ethanol, storing and transporting those energy carriers, and then using them in industry. In this context, the use of electrolysis fed by renewable energy to produce low-carbon energy carriers such as H₂, and the use of those chemicals as fuels, reductants, feedstock, and for upgrading petroleum products in refining, represents a low-carbon pathway to net-zero GHG emissions for a number of industrial processes [8].

In this paper, an overview of recent advances in the production of hydrogen within the framework of decarbonizing heat by means of proton-conducting ceramic membranes is given. In particular, after a section on hydrogen production, some general considerations on proton-conducting ceramic membranes as well as types of proton-conducting ceramic membranes and their applications in membrane reactors (MRs) and reversible protonic ceramic electrochemical cells (PCECs) are reported.

2. Hydrogen Production

The energy industry uses color codes to differentiate hydrogen on the basis of the type production.

Across a spectrum of carbon intensity (Figure 3), there is black and brown hydrogen using black coal or lignite (brown coal) in the hydrogen-making process and green hydrogen, which is made by using clean electricity from surplus renewable energy sources, such as solar or wind power, in water electrolysis [15,16,17] (discussed below). The black and brown hydrogen, which are the opposite of green hydrogen, are the most environmentally damaging.

Currently, gray hydrogen is the most common form of hydrogen production: it is obtained from natural gas, or methane, using steam methane reforming (SMR) (discussed below), but without capturing the greenhouse gases made in the process. Blue hydrogen is essentially the same as gray hydrogen, but includes the use of carbon capture and storage (CCS) to trap and store the CO₂ produced in the process. Moreover, there are relatively new hydrogen codes such as yellow hydrogen, turquoise hydrogen and pink hydrogen. Yellow hydrogen indicates hydrogen made through electrolysis using solar power. Turquoise hydrogen is made using a methane pyrolysis to produce hydrogen and solid carbon. In the future, such type of hydrogen may be valued as a low-emission hydrogen, if the thermal process is powered with renewable energy and the carbon being permanently stored. Pink hydrogen is generated through electrolysis powered by nuclear energy. In addition, the very high temperatures from nuclear reactors could be used in other hydrogen productions by producing steam for more efficient electrolysis or fossil-gas-based steam methane reforming [18].

The current cost for green H₂ is five to seven times that of gray H₂ and three times that of blue H₂, but such cost difference is expected to narrow with increased development and scale [19]. The CO₂ reduction between an electrolytic H₂ via grid-supplied electricity and gray H₂ is around 20–30%. For industry, H₂ via electrolysis using low-carbon sources could drive a much larger drop in CO₂ emissions [8].

2.1. SMR

Today, industrial-scale H₂ is produced via SMR, which supplies a range of chemical industries including synthesis of ammonia via the Haber-Bosch process, fuels and methanol. In SMR, natural gas (methane) reacts with steam according to the following equations:

$${\mathrm{CH}}_4+{\mathrm{H}}_2\mathrm{O} \leftrightarrows \mathrm{CO}+3{\mathrm{H}}_2 {\mathsf{\Delta }}_{SMR}^0=206 \mathrm{kJ}/\mathrm{mol}$$

(1)

$$\mathrm{CO}+{\mathrm{H}}_2\mathrm{O} \leftrightarrows {\mathrm{CO}}_2+{\mathrm{H}}_2 {\mathsf{\Delta }}_{WGS}^0=-41 \mathrm{kJ}/\mathrm{mol}$$

(2)

The synthesis gas known as syngas (CO + H₂) composition is governed by the methane reforming reaction (1) and the water–gas shift (2). Syngas conversion processes occur at high pressure and therefore, process economy favors operating the SMR at high pressure to decrease compression cost in the plant design. However, looking at the stoichiometry of the reaction (1), SMR is adversely impacted by high pressure due to Le Chatelier principle, shifting the equilibrium to left. Consequently, high temperatures are required to drive the reaction. Typical conditions are 30 bar and 900 °C [20]. A large-scale industrial reformer can contain up to several hundred tubular reactors in a large furnace, favored for uniform distribution of heat, which is produced by combustion of fossil fuels [20]. For most large-scale industrial reformers, radiation is the primary heat-transfer mechanism. Therefore, the combustion process must occur several hundred degrees above the reaction temperature to supply the necessary heat flux for the catalyst [20,21,22].

A typical large-scale industrial syngas plant for pure H₂ production has several reaction units, being steam-reforming equilibrium-limited and even, in the case of complete fuel conversion, a hydrogen-rich gas mixture containing carbon oxides and other by-products is produced. Typically, a high-temperature reformer, high- and low-temperature shift reactors with subsequent separation and compression downstream via pressure swing adsorption to separate H₂ from CO₂ feature an industrial steam reforming plant (Figure 4).

A typical fired SMR process based on natural gas emits 7–11 kg CO₂/kg H₂ [24]. The global production of syngas accounts for nearly 3% of global CO₂ emissions [25,26]. The heat source can be a variety of hydrocarbons with a high to low carbon content, i.e., from coal or coke to natural gas to renewables. In the context of decarbonized heat, i.e., to replace the fossil fuel combustion, excess electric power electricity from renewables can be transformed into heat via microwaves, induction, plasma, or joule heating (Figure 5).

Several studies have investigated relevant endothermic processes proposing electrified heat as listed in

Table 1. Selected studies on strategies of electrification of endothermic processes to produce H₂.

Entry #	Process	Electrified Heat Transformation Route	Refs.
1	SMR	Induction	[27,28,29]
2	SMR	Joule heating	[24,26,30,31,32,33]
3	SMR	Plasma	[34]
4	Dry reforming of methane	Plasma	[34,35]
5	Biogas reforming	Induction	[36]
6	Dry reforming of methane	Microwave	[37]
7	Catalytic decomposition of methane	Plasma	[34]
8	Methane conversion to higher hydrocarbons	Microwave	[38]
9	Methane conversion to higher hydrocarbons	Plasma	[39]
10	CO2 reforming	Microwave	[38]
11	CO2 reforming	Joule heating	[40,41]
12	Decomposition of ammonia	Joule heating	[42]
13	Decomposition of ammonia	Plasma	[43,44]
14	Decomposition of ammonia	Microwave	[45,46]

.

In particular, joule heating instead of heating from fossil fuel combustion has been proposed for highly endothermic chemical synthesis such as SMR (Table 1) as discussed below. Noteably, electrified heat is only decarbonized heat when the electricity is generated without carbon emissions [11]. Therefore, hydrogen synthesis from methane should include carbon capture converting emitted carbon dioxide into valuable commodity chemicals (Figure 6).

2.2. Water Electrolysis

Another relevant strategy to produce hydrogen via decarbonized heat is by water electrolysis fed by renewable electric energy, which in other words is transforming renewable electric energy into chemical energy by producing feedstocks such as H₂ (Figure 6).

By sourcing hydrogen exclusively by electrolysis instead of by SMR would avoid the generation of CO₂ altogether.

Hydrogen produced by electrolysis can be used for several purposes ranging from low-carbon heating to the synthesis of chemicals, representing a relatively new opportunity to reduce our dependence on carbon-based sources of heat and feedstock. Water electrolysis is a well-established mature technology used in special applications [47]. The overall electrolysis reaction is the electrochemical splitting of water molecules, owing to an electric current between cathode and anode, separated by electrolyte.

H₂O → H₂ + ½ O₂

(3)

Electrolytic hydrogen production technology can be classified on the basis of the employed ionic conductor, i.e., alkaline electrolysis cells (AECs), polymer electrolyte membrane electrolysis cells (PEMECs), protonic ceramic electrolysis cells (PCECs), and solid oxide electrolysis cells (SOECs). An overview of the four types of electrolytic hydrogen production technology is given in Table 2. In particular, AECs conventionally operate at lower temperatures in the range of 25−100 °C and utilize a concentrated alkaline solution as a medium for the conduction of hydroxide ions (OH⁻), which is fed to the cathode and anode for the corresponding hydrogen evolution reaction (HER) (Equation (4)) and oxygen evolution reaction (OER) (Equation (5))

2H₂O + 2e⁻ → H₂ + 2OH⁻

(4)

2OH⁻ → ½ O₂ + H₂O + 2e⁻

(5)

To avoid the mixing of the produced H₂ and O₂ (Equations (4) and (5)), a separator is typically placed between the electrodes, generally a polyphenylene sulfide (PPS)-based diaphragms [48] for which the main drawbacks are a limited current density (normally below 400 mA·cm⁻²) and an insufficient energy efficiency [49].

In the case of PEMECs, the conducted ion is proton (H⁺) transported through a polymer electrolyte membrane, while the deposited electrocatalysts promote the HER and OER on both sides as reported in Equations (6) and (7):

2H⁺ + 2e⁻ → H₂

(6)

H₂O → ½ O₂ + 2H⁺ +2e⁻

(7)

Additionally, PCECs use H⁺ as the conducted ion in electrolyte materials (with identical HER and OER expressions to those shown in Equations (6) and (7)), whilst the operating temperatures are in the range of 550−700 °C, requiring very different cell structures and material selections. The electrolyte materials mainly comprise proton conducting ceramic membranes as discussed below in this review, whereas cermet materials incorporating metals are utilized as electrocatalysts for remarkable catalytic activity.

SOECS have similar structures and materials to PCECs with a further increase in the working temperature to >700 °C, whereas the ion conduction in the electrolyte is dominated by the migration of oxygen ions (O²⁻), which requires a water supply to the hydrogen electrode (cathode) and involves distinct HER and OER expressions, as reported in Equations (8) and (9):

H₂O + 2e⁻ → H₂ + O²⁻

(8)

O²⁻ → ½ O₂ + 2e⁻

(9)

Table 2. Specification of AECs, PEMECs, PCECs and SOECs (adapted from [49] with permission of American Chemical Society).

	AECs	PEMECs	PCECs	SOECs
HER	2H2O +2e− → H2 + 2OH−	2H+ + 2e− → H2	2H+ + 2e− → H2	H2O + 2e− → H2 + O2−
OER	2OH− → ½ O2 + H2O + 2e−	H2O → ½ O2 + 2H+ + 2e−	H2O → ½ O2 + 2H+ + 2e−	O 2− → ½ O2 + 2e−
Charge carrier	OH−	H+	H+	O2−
Electrolyte	NaOH/KOH solution	Polymers	Ceramics	Ceramics
Electrocatalyst	Ni	Pt/C IrO2	Ceramics and cermets	Ceramics and cermets
Temperature (°C)	25–100 (low range) 100–250 (high range)	25–100 (low range) 100–250 (high range)	550–700	>700

Table 2. Specification of AECs, PEMECs, PCECs and SOECs (adapted from [49] with permission of American Chemical Society).

	AECs	PEMECs	PCECs	SOECs
HER	2H2O +2e− → H2 + 2OH−	2H+ + 2e− → H2	2H+ + 2e− → H2	H2O + 2e− → H2 + O2−
OER	2OH− → ½ O2 + H2O + 2e−	H2O → ½ O2 + 2H+ + 2e−	H2O → ½ O2 + 2H+ + 2e−	O 2− → ½ O2 + 2e−
Charge carrier	OH−	H+	H+	O2−
Electrolyte	NaOH/KOH solution	Polymers	Ceramics	Ceramics
Electrocatalyst	Ni	Pt/C IrO2	Ceramics and cermets	Ceramics and cermets
Temperature (°C)	25–100 (low range) 100–250 (high range)	25–100 (low range) 100–250 (high range)	550–700	>700

3. Proton Conducting Ceramic Membranes: General Considerations

Membrane technologies have all the features of process intensification strategy, i.e., low energy consumption, easy scalability, modularity [50]. In particular, H₂ separation using membranes can operate continuously using potentially less than half the energy required for the PSA process at the downstream of SMR (Figure 4) [51]. Among the four kinds of hydrogen separation membranes (polymeric, metallic, carbon, and ceramic [52]), only polymeric membranes have been in commercial use to any considerable extent at the current state of technology. Among the three inorganic categories, i.e., metallic, carbon and ceramic, proton–electron conducting ceramic membranes are the most promising due to their two main advantages, i.e., high selectivity and low cost [53]. Indeed, proton–electron-conducting ceramic membranes are permeable only to hydrogen ions, i.e., protons allowing the extraction of hydrogen from gas mixtures with a theoretical selectivity of 100%.

Overall, the process of H₂ transport through a proton–electron-conducting ceramic membrane mainly involves two steps, as schematized in Figure 7: (i) H₂ is first adsorbed onto the surface of the membrane, then dissociates into protons and electrons (surface process); (ii) protons and electrons diffuse together to the other side of the membrane surface where they reassociate to form molecular H₂ again (bulk diffusion) [54].

The permeation of hydrogen through the membrane is primarily controlled by the surface process and bulk diffusion.

Regarding bulk diffusion, the transport of protons through the membrane is the rate-determining step if the electron conductivity is much higher than the proton conductivity. Taking into account the chemical/electrochemical potentials and conductivities of both protons and electrons within the membrane, the H₂ flux (J_H2) can be described by the Wagner equation as follows [55]:

$${\mathrm{J}}_{\mathrm{H}2}=\frac{RT}{4F^{2}l} \frac{{\sigma }_{H+}{\sigma }_{el}}{{\sigma }_{H+}{+\sigma }_{el}}\mathrm{l}\mathrm{n}\left(\frac{P_{H2}^{\prime }}{P_{H2}^{″}}\right)$$

(10)

where R is the universal gas constant, T is the temperature, F is the Faraday constant, l is the thickness of the ceramic membrane, σ_H+ is the proton conductivity within the membrane, σ_el is the sum of the electron and hole conductivities, $P_{H2}^{\prime }$ is the H₂ pressure at upstream membrane and $P_{H2}^{″}$ is the H₂ pressure at downstream membrane (Figure 7). Proton and electron conductivities as well as the temperature and pressure at the membrane upstream and downstream as well as membrane thickness control the H₂ flux.

Two opposite cases can occur.

The first is when the proton conductivity σ_H+ within the membrane is much smaller than the electron conductivity σ_el; that is σ_H+ << σ_el, Equation (10) can be rewritten as:

$${\mathrm{J}}_{\mathrm{H}2}=\frac{RT}{4F^{2}l} {\sigma }_{H+}\mathrm{l}\mathrm{n}\left(\frac{P_{H2}^{\prime }}{P_{H2}^{″}}\right)$$

(11)

In this case, the development of highly conductive proton-conducting ceramic membranes is very important for achieving high H₂ flux as discussed below.

The second case is if the proton conductivity within the membrane is much larger than the electron conductivity, i.e., σ_H+ >> σ_el. So, Equation (10) can be rewritten as:

$${\mathrm{J}}_{\mathrm{H}2}=\frac{RT}{4F^{2}l} {\sigma }_{el}\mathrm{l}\mathrm{n}\left(\frac{P_{H2}^{\prime }}{P_{H2}^{″}}\right)$$

(12)

In this case, ceramic membranes with high electron conductivity become crucial in achieving high H₂ flux during the separation process.

Normally, it is difficult to obtain both high proton and high electron conductivities. This depends on the membrane materials.

Different types of ceramic H₂ separation membranes are reported in the literature:

(1)

Cermet membranes, consisting of a combination of a ceramic phase and a metallic phase. The metal is used as the electron-conducting phase and the ceramic oxide serves as the proton-conducting phase. Combining these two phases together may lead to high h₂ permeation because both σ_h+ and σ_el are high;

(2)

Single-phase ceramic oxides membranes, which simultaneously transport protons and electrons.

(3)

Cer-cer membranes, i.e., dual-phase ceramic oxides membranes, which combine a protonic conducting perovskite phase and electronic (non-perovskite) phase.

Single-phase ceramic membranes in turn may be classified into two subcategories: perovskite-type and non-perovskite-type membranes.

A scheme of the various types of proton conducting ceramic membranes is shown in Figure 8. In Table 3 a comparison of thickness normalized H₂ permeation flux of ceramic membranes from the literature is given.

3.1. Perovskite-Type Ceramic Membranes

The term perovskite originally refers to a mineral calcium titanium oxide, CaTiO₃. Today perovskite-type metal oxides is the name given to a group of materials with general formula ABO₃ (Figure 9) having the same structure as the mineral calcium titanate (CaTiO₃) [56]. In this general formula, A = Ca, Ba, Sr; B = Ce, Zr. The distance between the oxygen ions, which plays an important role in proton conductivity, can be changed by doping a trivalent M element into the structure leading to a significant improvement [57]. The formula of these high-temperature, proton-conducting perovskite oxides can be written as AB_1−xM_xO_3−δ. Doping a trivalent M cation into the B site can increase the vibration distance between the oxygen ions and also create more oxygen vacancies, leading to higher proton conductivity [58,59,60]. The most studied high temperature perovskite-type oxides are SrCeO₃ and BaCeO₃, which have high protonic conductivity but poor electronic conductivity. Electron conductivity has been improved by doping the B site in SrCeO₃ and BaCeO₃ with a multivalent cation [61,62,63]. Moreover, BaCeO₃ and SrCeO₃ are chemically unstable, easily reacting with CO₂ and H₂O [64,65,66]. In particular, formation of undesired phases such as carbonates would be formed because of the reaction between alkali earth ions, i.e., Ba²⁺ and Sr²⁺ and CO₂ at elevated temperature. This is the weak point of ceramic-based membranes, while the drawback of Pd-based membranes (not discussed in this paper) is hydrogen-induced embrittlement and sulfur poisoning.

In contrast, zirconate-based proton-conducting perovskite oxides are stable, but their conductivity is limited by highly resistive grain boundaries [68,69]. Indeed, the total conductivity of Y-doped BaZrO₃, the most promising zirconate in H₂O and CO₂-containing atmospheres, decreases sharply, owing to the blocking effect of grain boundaries combined with the small grain sizes in these materials [69]. In order to develop advanced membrane materials, solid solutions of cerate and doped zirconate (Y-doped BaZrO₃-BaCeO₃) have been investigated. Yttrium-doped barium zirconate/cerate materials (BaCe_xZr_1−x−y Y_yO_3−d, BCZY) represent the best compromise between the high proton conductivity of cerate-based and high stability of zirconate-based proton conducting perovskite oxides. They are stable over a wide range of temperatures (300–800 °C) and chemical environments and are among the most studied materials for hydrogen separation membranes in membrane reactors (MRs) [70,71,72,73,74] and reversible protonic ceramic electrochemical cells (PCECs) [75,76] as discussed below.

3.2. Non-Perovskite-Type Ceramic Membranes

The non-perovskite type membranes for hydrogen separation are mainly doped rare earth metal oxides and fluorite-structured metal oxide. Alkali earth ion-free oxides such as LaNbO₄, Ln₆WO₁₂, Y₂O₃, CeO₂ were developed in order to overcome the drawback of other phase formation such as carbonates under a CO₂ rich atmosphere. Ln₆WO₁₂ presents relatively high mixed protonic and electronic conductivities and reasonable stability under CO₂ and H₂O [77]. Moreover, among the Ln₆WO₁₂-based membrane materials, Mo-doped, Nd_5.5WO_11.25−d, and Re-doped materials showed almost the highest ambipolar conductivity and hydrogen permeation flux.

3.3. Cer-Cer Dual-Phase Ceramic Membranes

As reported by Elangovan et al. [78], H₂ permeating material with high performance can be obtained by combining (i) a protonic conducting perovskite phase BaCe_1−xM_xO_3−d and (ii) an electronic conducting fluorite phase Ce_1−yM_yO_2−d (M is the metal dopant). Such combination is named as cer-cer composite. It has certain degree of percolation, which provides efficient pathways for proton and electronic transport through the membrane. In such systems (Table 3), the addition of the doped ceria phase enhances the stability of BaCeO₃ phase, which otherwise, in the presence of CO₂ and H₂O, would decompose to BaCO₃, Ba(OH)₂ and CeO₂.

3.4. Cermet Membranes

In cermet membranes, a ceramic phase is combined with a metallic phase in order to enhance H₂ permeation because the ceramic phase is a proton conductor and the metallic phase is a highly electron conductor. Depending on the degree of hydrogen conductivity in the metal phase, mainly three different combinations can be realized [23]: (i) a metal with low hydrogen conductivity in combination with a highly proton-conductive oxide; (ii) a metal or an alloy with high hydrogen permeability such as Pd, Pd/Ag, Pd/Cu, Pd/Cu, combined with a ceramic of low hydrogen permeability; (iii) a combination where both the metallic and ceramic phases conduct hydrogen.

Table 3. Thickness-normalized values of H₂ permeation rates (J_H2) through ceramic membranes (adapted from [53] under CC-BY 4.0).

Material	Gas Atmosphere Feed–Sweep	T (°C)	JH2, Norm (mL·min−1 ·cm−1)	Ref.
Perovskite ceramic membranes
BaCe0.80Y0.10Ru0.10O3−δ	wet H2 in Ar–Ar	800	4.3 × 10−3	[79]
BaCe0.95Nd0.05O3−δ	wet 80% H2 in He–dry Ar + Ne	900	1.3 × 10−3	[80]
BaZr0.80Y0.15Mn0.05O3−δ	wet 50% H2 in He–wet Ar	900	1.4 × 10−3	[81]
SrCe0.95Tm0.05O3−δ	10% H2 in He–air	700–900	1.2 × 10−3/1.9 × 10−3	[82]
SrCe0.95Tm0.05O3−δ	10% H2 in He–air	700–900	2.6 × 10−3/6.4 × 10−3	[83]
SrCe0.95Tm0.05O3−δ	10% H2 dry in He–20% O2 in Ar	750/900	4.3 × 10−3/6.8 × 10−3	[84]
SrCe0.95Tb0.05O3−δ	20% H2 in He–CO in Ar	750/900	4.0 × 10−4/1.6 × 10−3	[85]
SrCe0.95Y0.05O3−δ	80% H2 in He–Ar	900	3.2 × 10−4	[86]
SrCe0.95Y0.05O3−δ	10% H2 in He–air	677	2.7 × 10−3	[87]
SrCe0.95Eu0.05O3−δ	100% H2–He	700–850	2.3 × 10−4/7.4 × 10−4	[88]
SrCe0.95Sm0.05O3−δ	100% H2–He	850	5.4 × 10−4	[88]
SrCe0.75Zr0.20Tm0.05O3−δ	H2 in He–wet Ar	750/900	5.0 × 10−3	[89]
SrCe0.75Zr0.20Tm0.05O3−δ	10% H2 dry in He–20% O2 in Ar	900	8.0 × 10−4/2.4 × 10−3	[84]
SrCe0.70Zr0.25Ln0.05O3−δ (Ln = Tm, Yb)	wet 20% H2–wet sweep (not specified)	900	2.3 × 10−4	[90]
SrCe0.65Zr0.20Eu0.15O3−δ	100% H2–He	900	8.5 × 10−4	[91]
SrZr0.95Y0.05O3−δ	20% H2–air in He	700	<2.3 × 10−5	[87]
Non-perovskite ceramic membranes
La5.5WO11.25−δ	wet 50% H2 in He–wet Ar	750/900	1.5 × 10−3/4.7 × 10−3	[92]
La5.5W0.8Mo0.2O11.25−δ	wet 50% H2 in He–wet Ar	700	2.7 × 10−3	[92]
La5.5W0.8Re0.2O11.25−δ	wet 50% H2 in He–wet Ar	700	5.9 × 10−3	[92]
Nd5.5WO11.25−δ	wet 20% H2 in He–wet Ar	1000	1.2 × 10−3	[93]
(Nd5/6La1/6)5.5WO12−δ	wet 50% H2 in He–wet Ar	900	1.2 × 10−3	[94]
(La5/6Nd1/6)5.5WO12−δ	wet 50% H2 in He–wet Ar	900	1.4 × 10−3	[95]
Nd5.5W0.5Mo0.5O11.25−δ	wet 50% H2 in He–wet Ar	900	7.5 × 10−3	[96]
Nd5.5W0.5Re0.5O11.25−δ	wet 50% H2 in He–wet Ar	900	4.1 × 10−3	[97]
Cer-Cer dual phase membranes
La5.5WO11.25−δ:La0.87Sr0.13CrO3−δ (50:50 vol.%)	wet 50% H2 in He–wet Ar	700	5.5 × 10−3	[98]
BaCe0.80Eu0.20O3−δ:Ce0.80Eu0.20O2−δ (50:50 vol.%)	H2, CH4, H2O, CO, CO2–He	900	<7.0 × 10−2	[78]
BaCe0.80Y0.20O3−δ:Ce0.80Y0.20O2−δ (50:50 wt.%)	wet 50% H2 in He–wet Ar	900	1.1 × 10−2	[99]
BaCe0.65Zr0.20Y0.15O3−δ:Ce0.85Gd0.15O2−δ (50:50 vol.%)	wet 50% H2 in He–wet Ar	755	1.76 × 10−2	[100]
BaCe0.20Zr0.70Y0.10O3−δ:Sr0.95Ti0.90Ni0.10O3−δ (50:50 vol.%)	9% H2 in He–dry Ar	800	1.1 × 10−3	[101]
SrZrO3:SrFeO3 (80:20 vol.%)	H2 in He–wet Ar	900	4.8 × 10−3	[102]
Cermet membranes
BaCe0.95Tb0.05O3−δ:Ni (50:50 wt.%)	50% H2 in N2–He	850	8.3 × 10−3	[103]
BaCe0.90Y0.10O3−δ:Ni (60:40 vol.%)	4% H2 in He–100 ppm H2 in N2	800	1.7 × 10−2	[104]
BaCe0.80Y0.20O3−δ:Ni (60:40 vol.%)	3.8% H2 in N2–100 ppm H2 in N2	900	2.0 × 10−3	[105]
BaCe0.85Zr0.10Tb0.05O3−δ:Ni (50:50 wt.%)	50% H2 in 50% He–Ar	800	8.5 × 10−3	[106]
BaCe0.70Zr0.10Y0.20O3−δ:Ni (60:40 vol.%)	4% H2 in He–100 ppm H2 in N2	900	5.6 × 10−3	[107]
BaCe0.70Zr0.10Y0.10Yb0.10O3−δ:Ni (60:40 vol.%)	20% H2–wet, 60% CO2, 20% He–N2	900	3.5 × 10−3	[108]
Ce0.50La0.4875Ca0.0125O2−δ:Ni (60:40 vol.%)	wet 20% H2, 77% N2–Ar	900	1.5 × 10−3	[109]
YSZ:Pd (40:60 vol.%)	90% H2 in He–N2	400/900	4.7 × 10−2/9.4 × 10−2	[110]

Table 3. Thickness-normalized values of H₂ permeation rates (J_H2) through ceramic membranes (adapted from [53] under CC-BY 4.0).

Material	Gas Atmosphere Feed–Sweep	T (°C)	JH2, Norm (mL·min−1 ·cm−1)	Ref.
Perovskite ceramic membranes
BaCe0.80Y0.10Ru0.10O3−δ	wet H2 in Ar–Ar	800	4.3 × 10−3	[79]
BaCe0.95Nd0.05O3−δ	wet 80% H2 in He–dry Ar + Ne	900	1.3 × 10−3	[80]
BaZr0.80Y0.15Mn0.05O3−δ	wet 50% H2 in He–wet Ar	900	1.4 × 10−3	[81]
SrCe0.95Tm0.05O3−δ	10% H2 in He–air	700–900	1.2 × 10−3/1.9 × 10−3	[82]
SrCe0.95Tm0.05O3−δ	10% H2 in He–air	700–900	2.6 × 10−3/6.4 × 10−3	[83]
SrCe0.95Tm0.05O3−δ	10% H2 dry in He–20% O2 in Ar	750/900	4.3 × 10−3/6.8 × 10−3	[84]
SrCe0.95Tb0.05O3−δ	20% H2 in He–CO in Ar	750/900	4.0 × 10−4/1.6 × 10−3	[85]
SrCe0.95Y0.05O3−δ	80% H2 in He–Ar	900	3.2 × 10−4	[86]
SrCe0.95Y0.05O3−δ	10% H2 in He–air	677	2.7 × 10−3	[87]
SrCe0.95Eu0.05O3−δ	100% H2–He	700–850	2.3 × 10−4/7.4 × 10−4	[88]
SrCe0.95Sm0.05O3−δ	100% H2–He	850	5.4 × 10−4	[88]
SrCe0.75Zr0.20Tm0.05O3−δ	H2 in He–wet Ar	750/900	5.0 × 10−3	[89]
SrCe0.75Zr0.20Tm0.05O3−δ	10% H2 dry in He–20% O2 in Ar	900	8.0 × 10−4/2.4 × 10−3	[84]
SrCe0.70Zr0.25Ln0.05O3−δ (Ln = Tm, Yb)	wet 20% H2–wet sweep (not specified)	900	2.3 × 10−4	[90]
SrCe0.65Zr0.20Eu0.15O3−δ	100% H2–He	900	8.5 × 10−4	[91]
SrZr0.95Y0.05O3−δ	20% H2–air in He	700	<2.3 × 10−5	[87]
Non-perovskite ceramic membranes
La5.5WO11.25−δ	wet 50% H2 in He–wet Ar	750/900	1.5 × 10−3/4.7 × 10−3	[92]
La5.5W0.8Mo0.2O11.25−δ	wet 50% H2 in He–wet Ar	700	2.7 × 10−3	[92]
La5.5W0.8Re0.2O11.25−δ	wet 50% H2 in He–wet Ar	700	5.9 × 10−3	[92]
Nd5.5WO11.25−δ	wet 20% H2 in He–wet Ar	1000	1.2 × 10−3	[93]
(Nd5/6La1/6)5.5WO12−δ	wet 50% H2 in He–wet Ar	900	1.2 × 10−3	[94]
(La5/6Nd1/6)5.5WO12−δ	wet 50% H2 in He–wet Ar	900	1.4 × 10−3	[95]
Nd5.5W0.5Mo0.5O11.25−δ	wet 50% H2 in He–wet Ar	900	7.5 × 10−3	[96]
Nd5.5W0.5Re0.5O11.25−δ	wet 50% H2 in He–wet Ar	900	4.1 × 10−3	[97]
Cer-Cer dual phase membranes
La5.5WO11.25−δ:La0.87Sr0.13CrO3−δ (50:50 vol.%)	wet 50% H2 in He–wet Ar	700	5.5 × 10−3	[98]
BaCe0.80Eu0.20O3−δ:Ce0.80Eu0.20O2−δ (50:50 vol.%)	H2, CH4, H2O, CO, CO2–He	900	<7.0 × 10−2	[78]
BaCe0.80Y0.20O3−δ:Ce0.80Y0.20O2−δ (50:50 wt.%)	wet 50% H2 in He–wet Ar	900	1.1 × 10−2	[99]
BaCe0.65Zr0.20Y0.15O3−δ:Ce0.85Gd0.15O2−δ (50:50 vol.%)	wet 50% H2 in He–wet Ar	755	1.76 × 10−2	[100]
BaCe0.20Zr0.70Y0.10O3−δ:Sr0.95Ti0.90Ni0.10O3−δ (50:50 vol.%)	9% H2 in He–dry Ar	800	1.1 × 10−3	[101]
SrZrO3:SrFeO3 (80:20 vol.%)	H2 in He–wet Ar	900	4.8 × 10−3	[102]
Cermet membranes
BaCe0.95Tb0.05O3−δ:Ni (50:50 wt.%)	50% H2 in N2–He	850	8.3 × 10−3	[103]
BaCe0.90Y0.10O3−δ:Ni (60:40 vol.%)	4% H2 in He–100 ppm H2 in N2	800	1.7 × 10−2	[104]
BaCe0.80Y0.20O3−δ:Ni (60:40 vol.%)	3.8% H2 in N2–100 ppm H2 in N2	900	2.0 × 10−3	[105]
BaCe0.85Zr0.10Tb0.05O3−δ:Ni (50:50 wt.%)	50% H2 in 50% He–Ar	800	8.5 × 10−3	[106]
BaCe0.70Zr0.10Y0.20O3−δ:Ni (60:40 vol.%)	4% H2 in He–100 ppm H2 in N2	900	5.6 × 10−3	[107]
BaCe0.70Zr0.10Y0.10Yb0.10O3−δ:Ni (60:40 vol.%)	20% H2–wet, 60% CO2, 20% He–N2	900	3.5 × 10−3	[108]
Ce0.50La0.4875Ca0.0125O2−δ:Ni (60:40 vol.%)	wet 20% H2, 77% N2–Ar	900	1.5 × 10−3	[109]
YSZ:Pd (40:60 vol.%)	90% H2 in He–N2	400/900	4.7 × 10−2/9.4 × 10−2	[110]

4. Applications of Proton-Conducting Ceramic Membranes to Produce Decarbonized H₂

Regarding hydrogen production, proton-conducting ceramic membranes are used in reversible fuel cells, i.e., Reversible Protonic Ceramic Electrochemical Cells (RePCECs), to produce both H₂ and electricity [111,112], in Proton Ceramic Membrane Reactors (PCMRs) for dehydrogenation of ethane to ethylene [71] and SMR [72,73,74]. Below some case studies are discussed.

4.1. PCMRs

Membrane reactors can be classified on the basis of the membrane role [113,114]. The membrane can have three different functions (Figure 10):

(i).

It can be used to extract a reaction product from the reaction zone in the so-called extractor.

(ii).

It can control the introduction of one of the reactants into the reaction zone in the so-called distributor.

(iii).

It can facilitate the contact between reactants and catalyst in the so-called contactor.

All the case studies discussed below involve decarbonized hydrogen production concern extractor type MRs for SMR (

Table 4. Selected case studies of PCMRs.

Entry #	Reaction	Ceramic Membrane	Anodic Electrode	Cathodic Electrode	Temperature (°C)	Ref.
1	SMR	BaZr0.7Ce0.2Y0.1O2.9	NiBZCY72 cermet	Cu	450–650	[74]
2	SMR	BaZr0.7Ce0.2Y0.1O3−δ (BZCY72)/BaZr0.7Ce0.1Y0.2O3−δ (BZCY71)/BaZr0.8Ce0.1Y0.1O3−δ (BZCY81)	BCZY	Ni	800	[72]
3	SMR	BaZr0.7Ce0.2Y0.1O3−δ (BZCY72)/BaZr0.8Ce0.1Y0.1O3−δ (BZCY81)	BCZY	Ni	750	[115]

, entries 1–3). Indeed, the reactions involved are equilibrium reactions and the membrane, which allow continuous removal of the H₂ product, i.e., its extraction; as a result increases the reaction conversion according to the Le Chatelier principle.

In the specific case of SMR, with respect to a classical configuration consisting of a reaction unit in series with a separation unit, i.e., PSA (Figure 4), in an MR, the integration of a reaction to produce H₂ with its concomitant separation by the membrane reduces capital costs, improves yields and selectivities, and drastically reduces downstream separation costs [116]. In most studies, metallic membranes, predominantly based on Pd or Pd-Ag alloy are used [23,116]. The difference in hydrogen’s chemical potential constitutes the driving force for hydrogen transport through the Pd membrane, from which it follows that even though complete methane conversion has been obtained, high pressures upstream are necessary and the pressure of the produced H₂ is low, and further compression requires multistage compressors, increasing energy consumption. In PCMR, protons are driven through the protonic ceramic membrane through the application of a voltage (or current) (Figure 11), which indirectly drives the flux of hydrogen gas (Figure 7) avoiding higher pressure. A gas mixture of methane saturated with steam is fed over the electrocatalyst (anode) where catalytic reactions (1)–(2) occur. With the use of an external power source, the produced hydrogen is extracted thanks its conversion at the anode in protons which are transported through the proton conducting ceramic membrane to the cathode according to the mechanism schematized above (Figure 7). At the cathode, protons recombine with electrons, forming molecular H₂.

Kyriacou et al. [74] (Table 4, entry 1) used BaZr_0.7Ce_0.2Y_0.1O_2.9 perovskite as a proton-conducting membrane, a tubular NiBZCY72 cermet as anodic electrode, and a film of Cu as the cathode. The results showed that hydrogen migration was beneficial for methane conversion and hydrogen yield. Figure 12 shows the effect of proton removal from the anode on the increase in methane consumption. Initially, reacting methane correlates linearly with removed protons (under closed circuit operation). The experiments carried out to investigate the stability indicate suitable activity thanks to the outstanding BaZr_xCe_1−xY_0.1O_3−δ perovskite membrane stability for x > 0.3 without the of formation barium carbonate as reported in the literature [117,118,119].

Malerød-Fjeld et al. [72] proposed a PCMR operating at 800 °C and using a BaZrO₃-based proton conducting membrane (Table 4, entry 2). They simultaneously realized the extraction and shift of a thermodynamically limited reaction towards full conversion of methane and the production of high-purity H₂ as in MRs based on Pd membranes, but differently; on the latter, PCMR delivers heat to the strongly endothermic reaction (1) through the electrical operation of the membrane, which acts as separator and compressor, compressing hydrogen directly at the sweep side of the membrane. In

Table 5. Comparison between Pd-based membrane reformers and PCMR investigated by Malerød-Fjeld et al. [72].

Reformer Type	JH2 (mL min−1 cm−2)	XCH4 %	SCO2 (%)	Temperature (°C)	HR (%)	Ref.
Pd-based membrane	4–9	≤95	-	495–550	~95	[120]
Pd-based membrane	19	87	95	550	92	[121]
PCMR	25	99.9	98	800	99	[72]

comparison of hydrogen production rate (J_H2), methane conversion (X_CH4), CO₂ selectivity (S_CO2), and hydrogen recovery (HR) of PCMR with Pd-based membrane reformers is given.

Malerød-Fjeld et al. [72] showed that the reaction heat required to form H₂ from CH₄ and H₂O scales linearly with the production rate. The heat evolved from the galvanic operation of the membrane distributes along the length of the membrane and originates from compression and Joule contributions [72]. The complete reformer plant was simulated as a function of H₂ production rate by changing the applied current in order to evaluate system heat integration (Figure 13).

Noteably, in order to assess the practical implications of PCMR technology, Malerød-Fjeld et al. made a techno-economic analysis, comparing the operating energy costs of PCMR with water electrolysis and SMR (Figure 14). The analysis reveals that PCMR and SMR are the less sensitive technologies to energy prices. The PCMR technology allows significant process intensification. High energy efficiency together with its hybrid nature, which uses 1/3 of electricity and 2/3 of natural gas, decreases the carbon footprint especially when the sources of electricity are renewable energies. Moreover, PCMR can benefits from producing a nearly pure stream of CO₂, enabling carbon capture, use, and storage for industrial scale H₂ production at locations with access to infrastructure for CO₂ storage and use.

Recently, Clark et al. [115] (Table 4, entry 3) proposed an optimized PCMR which retains the energy efficiency and H₂ recovery of single cells while achieving a 36-fold increase in H₂ production capacity. The PCMR stacks showed high-pressure H₂ at high purity and a CO₂-rich effluent at a hydrogen recovery and methane conversion > 99%. These results are very promising compared to Pd based membrane reformers.

4.2. RePCECs

As reported above, water electrolysis or water splitting via electricity enables energy storage in the H₂ molecule, which can be converted back into electricity through fuel cells or used in chemical processing (Figure 6, above). Among the different types of electrolyzers discussed above, PCECs in which electron flow through the device is accompanied by a flow of protons through a ceramic electrolyte are promising because of the potential for high energy efficiency and the possibility to produce dry pressurized H₂ [122]. RePCECs allow the seasonal energy storage challenge of intermittent renewable electricity generation, favoring the paradigm based on decarbonized heat, to be addressed. Indeed, conventional batteries are less suited for seasonal energy storage due to self-discharge and economic constraints than they are for short-term energy storage [119,123,124,125,126]. A. RePCEC produces H₂ from water when there is an excess of renewable electricity on the grid that might otherwise go to waste; otherwise, it can work in fuel cell mode using the stored H₂ to produce electricity when there is not enough electricity available on the grid [127]. In particular, in an RePCEC, in the electrolysis (EC) mode, the applied external voltage drives H⁺ from the air electrode to which steam, i.e., H₂O, is fed toward the electrode where H₂ is produced; in the fuel cell mode (FC), electric power is generated as a result of the oxidation of H₂, which is fed to the fuel side of the cell (Figure 15).

Dailly et al. [76] used BaCe_0.8Zr_0.1Y_0.1O_3−d (BCZY81) as a proton conducting ceramic membrane. Several reversible cycles into fuel cell/steam electrolysis were performed with an accelerated electrical degradation observed (+5%/kh under EC and −8%/kh under FC mode over 800 h). However, the authors of this study reported that deeper investigations in terms of operation protocols and the optimization of cell architecture (in particular the air electrode coating) are needed in order to improve electrical Faradaic efficiency (FE).

In a subsequent study, Duan et al. [75] correlate the poor FE observed in electrolysis mode with the electronic leakage due to minority p-type electronic conduction in common proton-conducting ceramic membranes. Indeed, electron holes are produced via the incorporation of molecular oxygen into oxygen vacancies as follows:

$$½{\mathrm{O}}_2+V_O^{..}\leftrightarrow 2h^{.}+O_{O^{.}}^x$$

(13)

Instead, the desired hydroxide defect, ${OH}_O^{.}$, is formed via the incorporation of water into oxygen vacancies via the hydration reaction:

$${\mathrm{H}}_2\mathrm{O}\left(\mathrm{g}\right)+O_O^x+V_O^{..}\rightleftarrows 2{OH}_{O^{.}}^{.}$$

(14)

So, a competition between oxidation (Equation (13)) and hydration (Equation (14)) for oxygen vacancies exists [111]. Increasing the favorability of reaction (14) suppresses reaction (13) by simultaneously increasing the proton transference number and decreasing the hole transference number, which in turn decreases electronic leakage. These relationships present the opportunity to tune the composition of proton conducting ceramic membrane and cell operating conditions to achieve higher FE. The equilibrium constant of the hydration reaction (14), i.e., K_hydration, can be expressed in terms of thermodynamic parameters [111]:

$${\mathrm{K}}_{\mathrm{hydration}}=\frac{{{[OH}_{O^{.}}^{.}]}^2}{\left[V_O^{..}\right]\left[O_O^x\right]p_{H_{2}O}}=\exp \left(\frac{{\mathsf{\Delta }S}_{hydration}^{°}}{R}\right)\mathrm{e}\mathrm{x}\mathrm{p}(-\frac{{\mathsf{\Delta }H}_{hydration}^{°}}{RT})$$

(15)

where ${\mathsf{\Delta }S}_{hydration}^{°}$ and ${\mathsf{\Delta }H}_{hydration}^{°}$ are the standard hydration entropy and enthalpy, respectively. Equation 15 correlates proton concentration with hydration thermodynamics and provides a tool for the design of materials with high proton concentration at relatively high temperatures. For ABO₃-based simple perovskites, prior investigation [129,130,131] has suggested that ${\mathsf{\Delta }H}_{hydration}^{°}$ is correlated with the electronegativities of the cations, which relate to the basicity of the oxide. At 600 °C, cerates such as barium cerates, i.e., BCY exhibits the largest hydration equilibrium constant, which both favors high proton concentration and suppresses the parasitic oxidation reaction. At the opposite end, the lower K_hydration value observed in zirconates (such as BZY) is the cause of low proton concentration and conductivity, and also of more oxygen vacancies available for oxidation resulting in higher hole concentrations and greater electronic leakage.

The most common proton-conducting ceramic membranes used in RePCECs employ yittrium-doped barium zirconate (BZY), barium cerates (BCY), and solid solutions thereof (BCZY). These ceramic membranes are mixed ionic-electronic conductors and not pure proton conductors. Even though electronic conduction does not greatly affect device performance in fuel cell mode, because transport across the membrane is dominated by protons and oxygen vacancies, in electrolysis mode, electronic conduction transport of delocalized small polarons $O_O^{.}$ is not negligible. BZY20 and BCZYYb show significantly higher electronic transport with severe electronic leakage in BZY20 more than in BCZYYb, which has a high Ce content at the B site, and therefore lower FE in electrolysis mode. So, the findings by Duan et al. [75] suggest that the preferred ceramic membrane composition for RePCECs and proton-conducting fuel cells (PCFCs) should be different. Indeed, one of the main challenges in the development of RePCECs is to enhance their performance by tailoring the composition of the ceramic membrane acting as electrolyte [111,132]. Putilov and Tsidilkovski [128] investigated theoretically the impact of acceptor-bound states of ionic defects in a proton-conducting oxide with proton and hole conductivity on the performance of protonic ceramic fuel and electrolysis cells operating on humidified hydrogen. The results are presented for the electrochemical cell based on BaZr_0.8M_0.2O_3−δ. Acceptor doping of a proton conducting oxide is necessary to create oxygen vacancies required for the oxide hydration. To analyze the effect of acceptor doping, Putilov and Tsidilkovski [128] have considered proton hopping over bound and free oxygen sites for two types of potential energy landscape (Q) (Figure 16): (i) ΔQ = 0, which corresponds to a weak effect of acceptor impurities; (ii) ΔQ = ΔE_H where ΔE_H is the trapping energy of protons, which represents the situation where impurities considerable reduce both the proton bound-state energy and the saddle point energy for transition between neighboring bound sites.

Fuel cells generate less electric power at a given current, and, conversely, higher applied electric power is required for electrolyzers to maintain the same current when impurities that provide deeper traps for protons and oxygen vacancies are present. Such an effect of acceptor impurities on the power density (P) of a PCEC is shown in Figure 17a, where the values of the power density calculated for Y-, Sc-, Gd- and In-doped BaZrO₃ with the corresponding trapping energies of protons (ΔE_H) and vacancies (ΔE_V) are given. It can be observed that the reduction in the P for the considered dopants is minimal for Y-doped BaZrO₃ (ΔE_H = 0.11 eV, ΔE_H = 0.31 eV), while for In-doped BaZrO₃, the largest power loss (ΔE_H = 0.17 eV, ΔE_H = 0.52 eV) can be obtained. Figure 17b shows U–j and P–j curves at different trapping energies.

5. Conclusions and Outlook

Proton-conducting ceramic membranes play a relevant role in energy conversion and storage applications including fuel cells for power generation, electrolyzers for hydrogen production, reversible electrochemical cells for long-term energy storage, electrochemical cells/membrane reactors for natural gas conversion. In particular, among the types of reactors for extracting hydrogen from natural gas, electrochemical membrane reactors based on proton ceramic membranes offer a promising opportunity for sustainable industrial growth, possessing unique characteristics such as 100% selectivity gas separation avoiding high pressures at the membrane upper stream, facilitating the use of joule heating instead of thermal heating based on fossil fuels. However, despite their potential in various fields, commercialization of proton conducting ceramic membranes remains a challenge: their performance and stability should be improved and large-scale membrane fabrication strategies are needed. To address such challenges, some key points should be considered:

(i).

The defect chemistry and transport properties of proton-conducting ceramic membranes and positive electrode materials as well as the relationship of defect chemistry and transport properties of these materials with composition, stoichiometry, microstructure, and operating conditions.

(ii).

The design of positive electrodes for PCECs and RePCECs to enhance bulk proton conductivity.

(iii).

The development of appropriate catalysts with specific activities for natural gas reforming and compatibility with other components of protonic ceramic devices.

(iv).

The engineering of the interface between the catalyst and the negative electrode to tune the electrochemical reactions.