English

• 1. Introduction

  The fast development of the global economy heavily relies on the exploration and utilization of fossil fuels such as crude oil and natural gas [1]. The excessive consumption of carbon-based energy carriers results in the raising level of CO₂ emission in atmosphere over 410 ppm in 2018, 40% higher than that before "industry revolution" [2]. The "biological carbon cycle" in nature with the participation of plants could only make limited contribution to the alleviation of this challenge via the natural photosynthesis, and severe environmental and ecological issues have been brought due to the CO₂ emission induced climate change [3]. Therefore, the revolution of our current energy system: i.e., the technologies for artificial capture/conversion of CO₂ and promoted share of low/zero carbon energy (such as H₂) are urgently required [4]. With the consideration of low weight energy storage density and high cost of lithium battery (< 260 Wh/kg, > 100 USD/kWh) [5], an ideal renewable system calls for advanced resolutions transforming fluctuating renewable electricity (wind or solar) into clean, value-added and transportable chemicals/fuels for continuous demands and multipurpose. Recent studies have illustrated that approximately half of the renewable electricity system requires further reasonable energy storage section. And an estimated value could be over 7500 terawatt-hours in Euro to realize completely decarbonization [6]. Thereby, the high-efficient electrolysis is undoubtedly the core step toward the realization of 100% renewable earth, which essentially consumes "exhausted carbon" to produce "valuable carbon" or convert H₂O into H₂, completing the gap in carbon neutral cycle with near-zero emission [7].

  So far, the electrolysis cells with different configurations and working principles are reported, including alkaline electrolysis cells (AECs), polymer electrolyte membrane electrolysis cells (PEMECs), and solid oxide electrolysis cells (SOECs). The AECs cell typically contains two electrodes isolated by a separator immersed in an aqueous alkaline solution (e.g., KOH). But the corrosive liquid electrolyte restricts its durable application at higher temperature (e.g., > 80 ℃). PEMECs utilize a solid polymer electrolyte with tri-functionality for proton conduction, electrolyzed gases separation, and electronic conduction suppression. Although it can migrate the challenge of low current density of AECs, the high price and vulnerability of proton exchanged membrane still hinder its large-scale industrialization. Compared to the former two counterparts, SOECs have unique advantages of cell robustness, low manufacture cost and unrivaled electrolysis efficiency [8].

  SOECs were originally proposed in the mid-1960s as a technique of electrolysis of CO₂ for oxygen regeneration in the US manned spaceflight programs. In the mid-1990s, Sridhar et al. [9] proposed to use CO and O₂ produced by solid oxide cell in electrolytic mode during the day to generate electricity in fuel cell mode at night. Since the beginning of the 21^st century, research groups around the world have been committed to developing novel materials to improve the performance and durability of stacks and reduce the production cost of syngas [10].

  The ideal green energy scenarios incorporating SOEC is demonstrated by Song et al. and shown in Fig. 1 [11]. SOECs can not only split steam (H₂O) into H₂, but also electrolyze CO₂ to CO. More appealingly, the co-electrolysis of H₂O/CO₂ toward synthetic gas (mixture of H₂/CO) enables the integration of SOEC with a wide range of downstream chemical processes such as Fischer-Tropsch synthesis, enhancing the total energy efficiency. A representative SOEC device displays an all-solid lamellar-structure layout with three main components: a porous cathode (fuel electrode) where the reduction reaction (e.g., CO₂ to CO and H₂O to H₂) happens with electrons flowing in, a porous anode (air electrode) where the oxidative reaction (O²⁻ to O₂) happens, and a dense electrolyte in the middle for ion transportation. From the perspective of configuration, SOEC could be classified into electrode (anode or cathode)-support, electrolyte-support and metal-supported ones. And the appearance of the electrolyzer could be either tubular, planar or flat-tubular shape (Fig. 1b) [12]. Furthermore, the operation of device in a reversed mode, named SOFC mode, is able to take advantage of the produced fuel (synthesized in SOEC mode) to generate electricity, while the power grid is on heavy demand [13].

  Figure 1

  

  Figure 1.  (a) The schematic of carbon neutral cycle based on renewable energy and solid oxide electrolysis cell. Copied with permission [11]. Copyright 2019, Wiley. (b) The SOEC devices with different configuration and appearances. Copied with permission [12]. Copyright 2017, Springer Nature. (c) Thermodynamics of CO₂ and H₂O reduction reactions. Copied with permission [14]. Copyright 2011, Elsevier. (d) Schematic representation of technology development potential of different electrolyzers. Copied with permission [15]. Copyright 2015, International Energy Agency.

  DownLoad: Full-Size Img PowerPoint

  SOEC can only properly work at elevated temperature (e.g., > 500 ℃), which brings the incomparable merits from both thermodynamic and kinetic aspects. As the electrolysis temperature increases, the theoretical electrolysis voltage (electrical energy demand) gradually decreases for either CO₂ (blue line) or H₂O (red line) splitting processes (Fig. 1c), following the Eq. 1 shown below [14]. This feature enables the electrical energy saving over 20%-30%, while compared to AECs and PEMECs operated below 200 ℃ [15].

  (1)

  Furthermore, the higher working temperature signifies the accelerated reaction kinetics, endowing the enhanced electrolysis current and conversion rate. For example, an SOEC for water splitting can generate a current density of ~1.5 A/cm² at a voltage less than 1.3 V, while the PEM electrolyzer required a much higher voltage of 1.6 V to generate a smaller current density of 1 A/cm² [16]. Such kinetics preponderance of SOEC is even more straightforward on sluggish CO₂ electrolysis. The SOEC for CO₂ reduction can easily reach a current density over 1 A/cm² with the overpotential less than 0.45 V and faradaic efficiency over 95%. While it is extremely difficult for the low temperature electrolyzer to reach over 500 mA/cm² with the maintenance of Faradic efficiency over 60% [17]. Moreover, the all-solid nature of SOEC has granted the device superior retention for long-term operation. The CO₂-to-CO electrolysis technology developed by Haldor Topsøe, Demark, demonstrated a prolonged measurement for almost 8000 h with an acceptable degradation rate of 16 mV per 1000 h [18], which is more than one order of magnitude higher than that of the state-of-the-art low temperature analogue [19]. In Sep. 2022, the Board of Topsoe has decided to start the construction of the world's largest SOEC electrolyzer with the capacity of 500 MW/year in Denmark. The superiorities in activity and durability further strengthens the industrialization potential of SOEC in regard to reduced operating and capital costs per electrolyzer unit.

  2. Single cell level

  2.1 Electrolyte

  The exploration of highly performed and stable single SOEC cell is the first milestone of commercialization. According to the different ionic conducting mechanisms, two types of SOECs, using oxygen ion conducting electrolyte (O-SOEC) and proton-conducting electrolyte (P-SOEC), were comprehensively studied, respectively [20]. For O-SOEC, the CO₂ or H₂O molecules receive electrons to produce CO or H₂, respectively. And the stripped oxygen ions are driven by the electric field and transported into anode chamber via electrolyte. The accumulated O²⁻ in anode are catalyzed to form O₂ and spontaneously release electrons (Fig. 2a) [21]. While in P-SOEC, the splitting of H₂O can only occur in anode to directly generate O₂ and the protons are transported into cathode to form H₂. The CO₂ electrolysis in P-SOEC is slightly more complicated, which requires the assistance of H₂O or H₂ splitting in anode [22,23]. However, the activation energy of temperature-dependent proton conducting is usually lower than that of ionic conducting, enabling the possibility to realize high efficiency electrolysis at intermediate temperature region [24].

  Figure 2

  

  Figure 2.  (a) The configuration schematic of solid oxide electrolysis cell (SOEC) with supported cathode layer, fuel cathode layer, electrolyte, buffer layer and anode layer. Copied with permission [21]. Copyright 2020, Springer Nature. (b) The temperature dependent ionic conductivity plot of doped ZrO₂. Copied with permission [27]. Copyright 2022, The American Ceramic Society. (c) The temperature dependent ionic conductivity plot of doped LSG samples. Copied with permission [30]. Copyright 1994, American Chemical Society. (d, e) The SEM image of Co-doped Sr₂Fe_1.5Mo_0.5O_6–δ (SFMC) reduced for 2 h and the EIS plot for SFMC before and after exsolution for CO₂ electrolysis at 1.6 V and 800 ℃. Copied with permission [66]. Copyright 2020, Wiley. (f) polarization curve and SEM image comparison of cell with straight pore and tortuous cathode decorated SOECs. Copied with permission [86]. Copyright 2022, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  The ideal electrolyte should possess good sinterability, high ionic conductivity and negligible electronic conductivity. Currently, two types of oxide with fluorite structure and perovskite structure were extensively employed as potential oxygen ion electrolytes in SOEC, respectively. The aliovalent cation doped ZrO₂ is the most commonly used electrolyte material in reversible solid oxide cell. The pure ZrO₂ demonstrates cubic structure only above 2370 ℃, which showed highest conductivity among the derivates with other phases [25]. In order to stabilize such highly desired structure, low valence elements of Y³⁺ and Sc³⁺ are doped into ZrO₂, which further promoted the ionic conductivity through vacancy hopping mechanism. The 8 mol% Y doped ZrO₂ shows a high ionic conductivity of ~0.03 S/cm at 800 ℃ [26]. However, its high activation energy for ionic transportation significantly restricted the application potential at lower temperature (Fig. 2b) [27]. The Sc doped ZrO₂ demonstrates an enhanced conductivity over > 0.08 S/cm at 800 ℃. But the high price of Sc is still the bottleneck for practical application [28]. The doped ceria (such as Gd or Sm doped CeO₂, (GDC, SDC)) presents higher ionic conductivity than doped ZrO₂ while the operation temperature below 600 ℃. However, while exposed to reducing atmosphere in cathode of SOEC, the emergence of Ce³⁺/Ce⁴⁺ redox pair in doped ceria leads to the electronic current leakage, the short-circuit of cell and lower faradic efficiency [29]. The Sr and Mg co-doped LaGaO₃ perovskite (La_0.9Sr_0.1Ga_0.8Mg_0.2O_3-x, LSGM) showed an outstanding ionic conductivity of 0.17 S/cm at 800 ℃ (Fig. 2c) [30]. However, the LSGM still suffers from the challenges of possible side-reaction with conventional Ni electrode to form less-conducting phase, and La diffusion into electrode to form catalytically inert phase [31].

  Lowering the operating temperature of proton conductor can not only broaden the selection of sealant and interconnect components, but also simplify the downstream purification process of electrolysis products (e.g., dry H₂ for steam electrolysis) [32]. Usually, one of the most predominant factor affecting proton conduction is the hydration capability and quantity of oxygen vacancy. While exposed to vapor for example, the water molecules are incorporated into surface oxygen vacancies to create two ^•OH. The barium cerates of Ba(Zr_0.1Ce_0.7Y_0.2)O₃ (BZCY) [33] and BaCe_0.5Zr_0.3Y_0.16Zn_0.04O_3-d (BCZYZ) [34] exhibited high proton conductivity. The additive of Zn, Cu, Ni and Fe were reported to enhance the sinterability of this series of proton conductor. However, they all suffers from the decomposition in H₂O- or CO₂-containing atmospheres. The common SOEC reactants, shorten their lifetimes. Alternatively, a hybrid ion conductor of BaZr_0.1Ce_0.7Y_0.2-xYb_xO_3-δ (BCZYYb) has been developed and regarded as most promising electrolyte candidate for H-SOECs because of its high conductivity and stability even treated the critical condition of 50% H₂-50% H₂O or 50% H₂-50% CO₂ for over 300 h [24,35]. Duan et al. demonstrated the desirable stability of BCZYYb-based H-SOECs for 1200 h at 600 ℃ with the degradation rate of 30 mV per 1000 h [32]. Yet, the extremely high concentration of H₂O and CO₂ still lead to the gradual deposition of BCZYYb [36], indicating the huge room for future exploration of better proton conductor.

  2.2 Fuel electrode (CO₂, H₂O)

  The cathode in SOEC is the place where CO₂ and H₂O were reduced to target products, regardless of the type of electrolyte utilized. The material candidates for cathode must have high electrical conductivity, electrocatalytic activity, chemical stability/combability with the electrolyte, as well as economic/manufacture feasibility [37]. The Ni-cermet (mixture of Ni and electrolyte (e.g., Y₂O₃ stabilized ZrO₂ (YSZ)) is currently the commercial cathode in SOEC, in which Ni phase provides electronic conducting pathway and reactive sites, and the electrolyte supplies ions transported from anode sides [38]. The electrochemical reduction will only happen at the triple phase boundary where Ni, electrolyte and reactant gases meet each other. During CO₂ electrolysis, some researchers thought it is necessary to add the reductive gas such as CO or H₂ to prevent the oxidation of Ni-YSZ by weak oxidants of CO₂ [39]. However, these protective gases will dilute the reactants concentration and decrease the electrolysis performance. Some other studies claimed the superfluousness of protective gas since the negative cathodic potential guarantees the electrons enriched environment on cathode to avoid the deep oxidation [40]. Unlike H₂ without carbon atom, the usage of CO as protective gas will inevitably bring another challenges of carbon deposition, stemming from the electroreduction of carbon species at the high electrolysis voltage and Boudouard reaction while the CO concentration is above the reaction threshold. The as-form carbon deposition could exist in the form of amorphous, graphite or carbon tube, blocking the active sites and degrading the devices [41]. Compared on Ni-YSZ, despite the Ni-BCZY cermet in P-SOEC exhibits better performance at intermediate temperature region, the poor redox stability and mechanical strengthen in highly humidified and reductive atmosphere are still obstacles needed to be resolved [42].

  Alternatively, perovskite oxide cathode candidates with regular formula of ABO₃ have attracted much attention because of its earth abundance, mixed ionic and electronic conductivity and excellent redox stability. However, the superficial surface of perovskite is usually dominated by catalytically inactive A-O terminal with limited amount of oxygen vacancies, and B-cation enriched region is buried beneath the surface, restricting the performance of the cell [43]. Therefore, diverse intrinsic/extrinsic material engineering methodologies are developed to improve the electrocatalytic activity of the perovskite-related oxides.

  Aliovalent cation doping is a very effective way to tune the structural properties and electronic configurations of perovskites, and further influence their electrocatalytic activities. The perovskite SrFeO_3–δ has a narrowed bandgap of ~2.0 eV, expressing good electrical conductivity but poor activity toward CO₂ splitting. The infiltration of Ni nanoparticles (NPs) significantly enhanced the CO₂ electrolysis performance with long-term stability for 100 h and 8 redox cycles [44]. Another example of electronic conducting La_0.2Sr_0.8TiO₃ (LST) also stressed the important role of doping effect. The doping of Mn into B-site of LST ((La_0.2Sr_0.8)_0.9(Ti_0.9Mn_0.1)O_3–δ (LSTM)) can create sufficient oxygen vacancies and enhance ionic conductivity. The material demonstrated desirable redox stability between SOFC mode (800 ℃ in a 5% H₂/Ar) and SOEC mode (1.4 V, CO₂, 800 ℃) [45]. Vanadium (V) substituting La_0.5Sr_0.5FeO_3-δ was also shown to have enhanced content of oxygen vacancies which facilitates the CO₂ absorption. The 0.05 mol% of V doped into La_0.5Sr_0.5Fe_1-xV_xO_3-δ enhanced the CO₂ electrolysis current density by 51.2% at 1.6 V and 800 ℃ [46].

  The wet impregnation is another widely used method to load active components onto oxide scaffold [47]. Loading of 2 wt% Fe₂O₃ on LSCM electrode was showed to drastically enhance the CO₂ electrolysis current by 75% at 800 ℃, which could deliver a stable current of 0.36 A/cm² at 2 V [48]. Carneiro et al. reported that metal oxophilicity (metal-oxygen bond strength) was strongly correlated to the CO₂ activation and electrode deactivation processes, which can explain why the electrode with infiltrated Fe has better performance than those with Pd and Ni [49]. Ni, Cu and Ni Cu metals are successfully incorporated into LSCM porous framwork, and the comparison of electrochemical measurements showed that the NiCu-LSCM composite cathode had best activity of 0.782 A/cm² at 1.6 V and 800 ℃) with high robustness for 100 h and ~10 redox cycles [50]. Besides metal, the MnCo₂O₄ spinel oxide were also reported to be impregnated into GDC skeleton with uniform distribution and served as electrode for SOEC. The current density for pure CO₂ electrolysis was able to reach 0.75 A/cm² with applied voltage of 1.5 V at 800 ℃ [51].

  Within the last decade, an alternative electrode design principle, called in-situ exsolution, emerged and was capable to resolve the weaknesses of poor metal (or alloy)/oxide adhesion, uneven distribution and fabrication complexity possessed by impregnation. Briefly, the catalytically active elements were uniformly incorporated into the lattice of perovskite scaffold during first-step annealing, and then reduced, segregated and anchored onto the surface of oxide support [52]. The anomalous catalytic interface of emergent metallic/alloy nanoparticles embedded in oxide parent displays uniform dispersion, enhanced stability. Intensive researchers have discovered both internal and external manipulation parameters to boost thermodynamic and kinetic of exsolution process. For instance, the point non-stoichiometry in lattice [53], topotactic ion exchange [54] and lattice strain [55] were applied to enhance the quality of exsolved particles. And the electrochemical poling method also facilitated the exsolution by creating transit ultra-low oxygen partial pressure (P_O2) [56]. Using La_0.75Sr_0.25Cr_0.5Mn_0.5O_3-δ (LSCM) scaffold as an example again, the exsolved Ni nanoparticles decorated LSCM electrode demonstrated an current density rise from 0.27 A/cm² to 0.38 A/cm² at 2.0 V and Faradaic efficiency enhancement by 60% [57]. On the case of La_0.4Sr_0.4Fe(or Ni)_0.06Ti_0.94O_2.97, Tsekouras et al. evidenced that the presence of electrocatalytically active exsolved Fe or Ni nanoparticles and co-generated oxygen vacancies dramatically lowered the activation barrier to steam electrolysis (900 ℃, 47% H₂O and 53% N₂) compared to pristine perovskite [58]. Moreover, the Ni-Fe, Fe-Co and Ru-Fe alloy were also reported to exsolve from various perovskite scaffolds [59-65], which not only have superb electrolysis activity, but also deliver enhance coking resistance. Also, the reduction temperature and duration are also key parameters influencing morphology of exsolved particle, further the electrolysis performance (Figs. 2d and e) [66]. The plenty of theoretical simulation results proved the preferable absorption and activation of reactant at exsolved metal-support interface. And the as-formed oxygen vacancies adjacent to these interfaces can also contributed to the enhanced charge carrier transfer kinetics, collectively facilitating the overall electrolysis process.

  2.3 Air electrode

  The anodic reaction for typical SOEC is often the sluggish 4-electron involving oxygen evolution reaction (OER), which has a redox potential of 1.229 V (vs. SHE) and occupies the majority energy consumption of SOEC. Therefore, the development of novel anode materials with minimized overpotential is crucial for highly efficient SOEC [67,68]. The first generation of anode electrode is Sr doped LaMnO₃ perovskite (LSM) which had also been applied for SOFC for decades [69]. The electronic conducting characteristic of LSM (over 100 S/cm at 800 ℃) [70] determines that it has to be well mixed with ionic conducting electrolyte materials (YSZ, GDC) to form mixed ionic and electronic conducting composite [71], and the reaction can only happen at the triple phase boundary (TPB) where LSM, electrolyte and gas meet.

  The poor intrinsic activity of LSM motivates the development of second-generation air electrode of La_0.6Sr_0.4Co_0.2Fe_0.8O₃ (LSCF) perovskite which has a reasonable electronic conductivity of 200-300 S/cm, but an orders of magnitude higher ionic conductivity of 0.01-0.1 S/cm. Therefore, the TPB length can be substantially extended to the whole air electrode region, which enhances the reaction kinetics even without the mixture of electrolyte [72]. Recently, other types of oxide were also developed to broaden the family of air electrode. The Ruddlesden-Popper (RP) structure oxide such as rare earth nickelate of Nd₂NiO₄ [73], usually demonstrate a higher oxygen ionic conductivity than that of simple perovskite by several orders of magnitudes, owing to the highly anisotropic interstitial diffusion mechanism of oxygen ion on the a-b plane, confirmed by Molecular dynamics (MD) simulation on La₂NiO₄ [74]. Beyond simple perovskite, the double perovskite of AA′BB′O_6-δ with the AO-BO₂-AO-BO₂ stacking along c-axis have shown outstanding oxygen surface exchange coefficient and ionic diffusion rates. MD results showed that oxygen ions migrated in PrBaCo₂O_5.48 [75] and GdBaCo₂O_5+x [76] were also highly anisotropic. The widely studied fuel electrode of Sr₂Fe_1.5Mo_0.5O_6-δ (SFM) was also found to be efficient as air electrode with the polarization resistance is only 0.26 Ω cm² at open-circuit condition, and the electrolysis current density at 1.3 V reaches 0.88 A/cm² At 900 ℃, which was better than that of the LSM-YSZ anode [77]. The cell with PrBa_0.5Sr_0.5Co_1.5Fe_0.5O_5+δ (PBSCF) double perovskite anode can deliver the maximum current density of 1.3 A/cm² is obtained at 1.3 V and humidity of 50% at 800 ℃. And the cell exhibited excellent durability of 120 h and good reversibility [78,79].

  2.4 Buffer layer

  Upon prolonged operation for thousands of hours, the significant ion inter-diffusion between different components will lead to the formation of insulating phase, resulting in the deactivation of electrolysis stack. The unwanted reactions between electrode and electrolyte can be alleviated by adding the buffer layer between them. For example, LSCF is one of the most promising air electrodes for O-SOEC, which however, will react with YSZ electrolyte to form SrZrO₃ and La₂Zr₂O₇ insulating phases [80]. Thereby, doped ceria with comparable ionic conductivity as YSZ was usually used as a buffer layer to avoid the undesired reaction between the LSCF-based electrode and YSZ electrolyte. Lyu et al. used one-pot hydrothermal method to prepared LSCF@GDC twined hybrid which showed a 40-fold promoted performance under oxygen starving conditions, compared to bare LSCF at 750 ℃, suggesting that the additive of buffer layer component can also facilitate the oxygen activation kinetics. Nevertheless, the additive of GDC will unavoidably decrease the total conductivity of LSCF and enhance the total ohmic resistance of the cell [81]. Besides YSZ, another commonly used electrolyte of LSGM will react with Ni cermet to form lanthanum nicklates which is catalytically inert for fuel activation. Consequently, again, a buffer layer of doped ceria has to be incorporated between the Ni cermet and LSGM electrolyte to resolve this problem [82].

  2.5 Microstructure engineering

  The microstructure engineering of the specific component, especially electrode, of SOEC is equally vital as the material selection, which is essentially correlated to the mass/heat transportation and the lengthen of TPB. Typically, the fuel electrode of the mixture of Ni and YSZ has high porosity over 40% which is created by the additive of pore maker during fabrication. The high active surface area due to porous structure of electrodes could guarantee the sufficient exposed active sites on which the reactants meet. Unlike graphite and potato starch which generates coarser and uneven pores, respectively, the PMMA can generate much finer pores, leading to enhanced electrochemical performance. The SOEC using PMMA as pore former shows a highest electrolysis at −713 mA/cm² at 1.3 V and 800 ℃ with a stable operation for over 400 h [83]. Liu et al. disclosed the optimum PMMA content of 10 wt% in cathode fabrication, resulting in a high porosity of 45% and electronic conductivity of 6726 S/cm [84]. Alternatively, the porous framework manufacture and deposition of catalytic site can be done by two separated steps. For example, Fan et al. reported the construction of ionic conductive YSZ matrix concomitantly with the dense electrode. And then the LSCF catalytic air electrode compound could be introduced by impregnation to form TPB. However, the agglomeration of LSCF was observed after 8 charged/discharged cycles, responding for its unsatisfied redox stability [85]. The manipulation of morphology of pore can not only boost the electrolysis efficiency, but also migrate the challenges confronted in extreme condition. Li et al. reported that the modification of tortuous pore to finger-like straight pore can enhance the carbon deposition resistance of the SOEC for CO₂ conversion. The critical current density at which coking starts to occur was found the be enhanced from 0.5 and 1.0 A/cm² to 1.0 and 1.5 A/cm², respectively. The electrochemical impedance spectroscopy (EIS) results disclosed that straight pore paths accelerate the removal of as-produced CO to suppress the Boudouard reaction (Fig. 2f) [86]. The microstructure engineering was also employed to alleviate the delamination of electrode/electrolyte interface during SOEC/SOFC cycling [87]. Khan et al. added a SSZ porous layer between the dense SSZ electrolyte and the porous LSM air electrode and found that electrolysis current density increased from 294 mA/cm² to 407 mA/cm² after 5 cycles at 1.3 V in 10% H₂O-90% H₂, at 800 ℃, owing to the raised quantity of TPB and migrated high oxygen partial pressure. In comparison, The cell without such porous buffer layer suffered from the formation of insulating zirconate phase, LSM aggregation and undesirable silver deposition [88].

  2.6 Operation optimization

  2.6.1   Fuel-assisted O-SOEC

  Now that the electricity accounts for a majority of total energy consumption of SOEC, the cheap low-carbon molecules were added into anode chamber to modify the elementary reaction happening on electrode, aiming to lowering the overpotential for electrolysis. The CH₄-assisted SOEC by coupling CH₄ in air electrode was systematically modeled by Ni's groups on 2016 (Fig. 3a) [89]. It was shown that the introduction of CH₄ in anode of CO₂/H₂O co-electrolysis device effectively lower the equilibrium potential of SOEC (Fig. 3b), and thus decrease the electricity consumption. The reaction can be triggered even at the low potential of 0.2 V at 750 ℃, and the device could realize the co-generation of electricity and syngas at the current density less than 600 A/m² and at 850 ℃.

  Figure 3

  

  Figure 3.  (a) The model of CH₄ assist-SOEC. Copied with permission [89]. Copyright 2016, Elsevier. (b) The calculated energy consumption for conventional H₂O/CO₂ co-electrolysis and CH₄ assisted electrolysis. Copied with permission [89]. Copyright 2016, Elsevier. (c) The schematic of CH₄ oxidative coupling assisted CO₂ electrolysis. Copied with permission [65]. Copyright 2018, Wiley. (d) The polarization curves of conventional SOEC and carbon gasification assisted SOEC at 850 ℃. Copied with permission [93]. Copyright 2016, Elsevier. (e) The calculated thermodynamic equilibrium results for the reaction of 2 mol O₂ with 98 mol carbon at different temperatures. Copied with permission [93]. Copyright 2016, Elsevier. (f) The amount of H₂ depends on the current in the cathode in C₂H₆ dehydrogenation assisted P-SOEC. Copied with permission [94]. Copyright 2020, American Chemical Society. (g) CH₄ partial current density and partial Faradaic efficiency in C₂H₆ dehydrogenation assisted P-SOEC. Copied with permission [95]. Copyright 2020, Springer Nature.

  DownLoad: Full-Size Img PowerPoint

  To enhance the total productivity and efficiency of fuel-assisted SOEC, the highly selective catalyst were facilely prepared to convert CH₄ to valuable syngas via partial oxidation of methane (POM) reaction [90]. The total electrical energy consumed of the POM-assisted SOEC for the CO₂/H₂O co-electrolysis is decreased compared with that of the convenient SOEC. An electrolysis current of 350 mA/cm² can be obtained at a cell voltage of only 0.4 V in the POM assisted mode. Following this line of thinking, Lu et al. developed the Ni Cu alloy exsolved LSCM electrode and assembly the CH₄-assist SOEC to efficiently reform of CH₄/CO₂ to produce syngas by an electrochemical way. It was found that the cell can reach the current density to 1.2 A/cm² at 1.5 V and the productivity of CO is highly dependent on the Ni: Cu ratio in alloy (Fig. 3c). The cell demonstrated no obvious degradation after 300 h of high-temperature operation and 10 redox cycles [65].

  By optimizing the catalyst in electrode, the methane can be electrochemically converted to ethylene in SOEC via an oxidative coupling reaction. On a Fe exsolved Sr₂Fe_1.5+xMo_0.5O_6-δ electrode, the concentration of C₂ in anode outlet can reach 16.7% with total selectivity over 80% and CH₄ conversion rate over 40%. This process realizes a stable output for 100 h and 10 redox cycles, suggesting its potential for practical application [91]. While replacing CH₄ with C₂H₆, more value C₂₊ products can be expected via the advanced electrochemical-thermal coupling catalytic process. This design was realized by Song et al. by the impregnation of γ-Al₂O₃ onto the surface of La_0.6Sr_0.4Co_0.2Fe_0.8O_3−δ-Sm_0.2Ce_0.8O_2-δ (LSCF-SDC) anode in an O-SOEC. Usually, the thermal oxidative dehydrogenation of ethane is challenged by the deep oxidation, resulting in low selectivity. In comparison, in Song's work, a high ethylene selectivity of 92.5% and ethane conversion of 29.1% at 600 ℃ can be observed [92]. The infiltrated Al³⁺ was bonded to Fe via an Al-O-Fe structure and enriched the state density around Fermi level of Fe, facilitating the conversion of C₂H₆. Besides gaseous fuel, the solid carbon can also be employed to promote the activity of SOEC in which the oxidation of carbon is predominant reaction in anode. The theoretical overpotential of C-assist CO₂/H₂O co-electrolysis can be reduced by about 1 V at 850 ℃, compared to traditional analogues (Figs. 3d and e). The thermodynamic equilibrium calculation further proved that CO is the main product while the operation temperature was over 550 ℃, suggesting its good combability with high temperature electrolysis [93].

  2.6.2   Fuel-assisted P-SOEC

  As mentioned above, the P-SOEC can migrate the biggest weakness of poor intermediate temperature performance of O-PCEC, offering brighter potential for industrialization. Moreover, for the fuel-assist SOEC, the proton conducting mode without existence of oxygen species in anode, providing an alternative way to realize the co-generation of dehydronated and hydrogenated products in separated chambers. Zhang et al. reported an electrolysis of CO₂ in a P-SOEC with the electrochemical conversion of C₂H₆ to C₂H₄ in anode. The Ni_xCu_1-x alloy exsolved Nb_1.33(Ti_0.8Mn_0.2)_0.67O_4-δ anode showed improved stability and high coking resistance for 10 h with the highest ethane conversion of 75.2% and ~100% ethylene selectivity at 0.8 V (Fig. 3f). The CO₂ was both directly reduced to CO and involved into the reversed water gas shift reaction on cathode, leading to a ~6.3% H₂ in the cathode outlet [94]. Using the same cell configuration with SDC supported Ir (SDC/Ir) electrode, Li et al. reported the achievement of selectivity over 95% toward CH₄ of electrochemical hydrogenation of CO₂ at 400 ℃. Further enhancing the bias leads to the promoted CH₄ selectivity to reach almost 100%. The selectivity is strongly correlated to the stabilization of different transition state of carbon species, which is controlled by the Ir-O hybridization (Fig. 3g) [95].

  On the other hand, the CO₂ electrolysis in P-SOEC mode can direct generate CH₄ with the assist of H₂O splitting in anode. Pan reported that at an optimal low temperature of 450 ℃ and electrolysis current of 1 A/cm², a high CH₄ productivity of 34.6% can be achieved, which is much higher than the yield of thermo-catalytic CO₂ methanation. Further enhancing the electrolyte thickness could improve the CH₄ yield to 40.9% by enhancing the faradic efficiency. And recycling the produced H₂ in anode could boosts the CH₄-yield ratio to 71.2%. Such high performance is not only due to low operating temperature, but also owing to the physical separation of the CO₂ and H₂O feedstocks. Such electrochemical CO₂ upgrading with PCECs is proved to be economically benign by the techno-economic analysis projects [96].

  3. Stack level

  Compared to individual cell, the SOEC stack brings more challenges beyond activity. The stacked cells compose single SOEC module which is the basic unit of SOEC system. Taking planar cell as an example, each individual cell is assembled into metallic frames, which is sandwiched by interconnect component with incurved flow channel. A sealant is simultaneously employed to prevent potential gas leakage or mixing. For steam splitting, the pre-purification system as well as gas storage system are usually required to purify the steam reactant and store H₂ product, respectively.

  3.1 Interconnect

  Since the Interconnects not only separate the gas in cathode and anode, but also provide the electrical conduction pathway ways among all stacks, the qualified interconnect should be stable in reductive, oxidative and moisture atmospheres. The chromic based perovskites such as LaCrO₃ could serve as potential interconnect due to its high chemical and thermal stability. However, its high cost for manufacture restricts its commercialization. In comparison, the high-chromium (Cr) containing steels (430, 441, etc.) are the alternatives owing to its economic efficiency, promoted mechanical strength, and better thermal conductivity. Nevertheless, the high chromium content in steel resulted in deteriorate Cr vaporization and further diffusion into the cell, worsening the performance of the cell. To overcome this problem, these stainless steels are usually coated to slow down it degradation rate and minimize Cr poisoning on cell [97].

  For example, the ChromLok^TM MCO-based composition (Mn, Co)₃O₄ is applied to Crofer 22 APU stainless steel for O-SOEC stack for over 700 h. The MCO coating is found to decrease the oxidation rate by about one order of magnitude, and decrease the Cr evaporation rate by fourfold [98]. On the other hand, the (Co, Mn, Fe)₃O₄ spinel oxide coated Crofer 22 APU stainless steel successfully operated for over 40,000 h. Due to the presence of vapor at high temperature, the coating in fuel side degraded faster than one in cathode side [99].

  3.2 Sealant

  Unlike tubular SOEC with great spatial separation of fuel and air, the SOEC with planar configuration takes higher risk of potential gas leakage. the sealants for a SOEC stack should be chemically, mechanically and thermally stable under harsh working conditions. Further, reactivity between sealant and interconnect has to be as low as possible. The noble metals are chemically inert and have good elasticity. However, they are usually extremely expensive. Glass and glass-ceramic composite seals such as borosilicates, or alkaline earth metal oxides containing silicate or borosilicate glasses, exhibit much higher cost efficiency [100]. Nevertheless, upon prolonged measurement, it is universally found that the silica-containing impurities diffused from sealant into cell can block the electrocatalytically active sites, inevitably cause degradation and lead to increased polarization resistance [101].

  3.3 System operation

  To improve the electrolysis production rate and decrease degradation of the cell, the endeavors have to be devoted to optimize the operation conditions of the stacks, which is usually ignored for button cell measurement. Xu et al. reported the effect of the flows of hydrogen electrode (Figs. 4a and b), water pressure, and electrolysis current density on the performance of stack composited by a 30-layer single cell of NiO-YSZ/NiO-YSZ/YSZ/LSM-YSZ with an active area of 63 cm². High flow and low electrolysis current density are beneficial to increase electrolysis efficiency and reduce conversion (Fig. 4c). The speculated reason is that the heat generated from the stack internal resistance is much higher than that required for water decomposition at high current densities, leading to increases in the operating voltage. Under the output voltage 39 V (thermoneutral voltage), the current density increased from 0.185 A/cm² to 0.22 A/cm² with the water pressure increased from 70% to 90%. Besides, the degradation rate of the stack was 11.7% 1000 h⁻¹ under current density 0.15 A/cm² and 80% water pressure at 800 ℃, and the electrolysis efficiency decreased from 113% to 92%. The Post-mortem analysis revealed that both the oxygen and hydrogen electrode at inlet area were delaminated from the electrolyte (Fig. 4d), and the accumulation of Ni particles was obvious at the hydrogen electrode [102]. Xu et al. further improved the performance and long-term stability of the stack with a single layer by replacing the LSM-YSZ by an LSCF-GDC cathode. The LSCF-GDC stack exhibited a high hydrogen production rate with 159.2 Nml cm⁻² h⁻¹, much higher than 99.5 Nml cm⁻² h⁻¹ of the LSM-YSZ stack under electrolysis voltage with 1.1 V at 800 ℃. When it operated under 0.5 A/cm² at 750 ℃ for 640 h, the degradation rate of stack performance is 4.33%, which is better than that of LSM-YSZ at 4.89%. Besides, there was no significant flaking of the LSCF electrode after the test (Fig. 4e) [103].

  Figure 4

  

  Figure 4.  (a, b) 3D view of the stack module mounting. Copied with permission [102]. Copyright 2014, Elsevier. (c) Effect of different flows on the hydrogen electrode on the steam conversion rates and electrolysis efficiencies of the stack module at 800 ℃ and a constant 80% H₂O-20% H₂. Copied with permission [102]. Copyright 2014, Elsevier. (d, e) SEM images of the cross-section (fracture surface) of the Ni-YSZ/YSZ/LSM–YSZ and Ni-YSZ/YSZ/GDC/LSCF-GDC cells after electrolysis at air inlet area. Copied with permission [102,103]. Copyright 2015, Elsevier. (f) Stack with NiO-YSZ/YSZ/LSCF cell test 1000 h at 0.2 A/cm² in electrolysis mode. Copied with permission [104]. Copyright 2013, Elsevier. (g) Gas outlet measurements of the 10-cell stack at 800 ℃ with cathodic inlet: 65% H₂O-25% CO₂-10% H₂ with conversion rate of oxidized species is 32% at 0.5 A/cm². Copied with permission [107]. Copyright 2015, Elsevier. (h) Gas analysis of the CSC and the ESC stack according to the steady state co-electrolysis experiments. Copied with permission [108]. Copyright 2020, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  Regarding interconnector, O'Brien et al. found that Cr poisoning can seriously affect the performance of the cell in contact with the Inconel collector, the total degradation rate of the stack with 10-layer NiO-YSZ/YSZ/LSCF is 8.9% K/h under 0.2 A/cm² and 70% H₂O at 800 ℃, where the degradation rate of the top cell is 26.8% and that of other cells is 3.7% (Fig. 4f) [104]. To prevent the electrode from being corroded, the LSM and MnCo_1.9Fe_0.1O₄ (MCF) materials were used as an electrode protection layer [105,106]. Fang et al. applied the MCF as a protection layer between the LSCF oxygen electrode and the connector, the degradation rate of the stack is only 0.7%/1000 h under 0.5 A/cm² and 50% H₂O at 800 ℃, indicating high stability of stack with protection layer [106].

  The O-SOEC stack could be also applied to the co-electrolysis mode of H₂O and CO₂. Mougin et al. performed co-electrolysis using a stack with 10-layer NiO-YSZ/YSZ/GDC/LSC single cell under 65% H₂O-25% CO₂-10% H₂. A maximum current density with −0.8 A/cm² and gas conversion of 52% were obtained at 1.15 V at 800 ℃. When the electrolysis current with 0.8 A/cm² was applied at 800 ℃, the H₂-CO syngas production was 0.34 Nm³/h, the O₂ production was 0.17 Nm³/h, and no methane production was observed by chromatography (Fig. 4g), which is consistent with the thermodynamic prediction, proving the feasibility of the co-electrolysis at the stack [107]. Riedel et al. investigated the effect of pressure on the electrolysis performance of a stack with a 10-layer single cell, and compared the performance of the electrolyte support cell (ESC) and cathode support cell (CSC), where the structure of former was NiO-GDC/GDC/3YSZ/GDC/LSCF and the latter was NiO-YSZ/YSZ/GDC/LSCF. The co-electrolytic atmosphere was 60% H₂O-30% CO₂-10% H₂ and test pressure was 1.4, 4, and 8 bar, respectively. The current density of CSC was 0.75 A/cm² under electrolyte voltage of 1.3 V and pressure of 1.4 bar at 750 ℃ (Fig. 4h). The CH₄ yield increased from 2% to 7% with the external pressure increased from 1.4 bar to 8 bar under the electrolysis current 0.5 A/cm² at 750 ℃, which is consistent with high pressure, high H₂/CO ratio, and low temperature is favorable for methanation reaction. Besides, Increasing the pressure could decrease the voltage of stack and increase the core temperature of stack for the methanation reaction is exothermic. In addition, the electrolytic current of ESC was only 0.3 A/cm² under 1.3 V at 800 ℃ and the maximum methane yield is only 5% for the higher internal resistance (Fig. 4h) [108].

  4. Perspectives

  In this review, we have briefly summarized the recent developments of SOEC technology mainly at cell and stack levels. Meanwhile, we have stressed that further optimization of various parameters of SOEC is still urgently required to further promote the overall activity and durability.

  Both performance and cell durability have been improved drastically by order of magnitude in the past decade. The lifetime of the SOEC should be at least 50000 h to maintain its competitiveness in the market. Haldor Topsoe in Demark has commercialized the commercially competitive CO₂ electrolysis technology with lifetimes over 7000 h and degradation rate below 0.4% per 1000 h [109]. However, how to completely prevent Ni segregation and stability of Ni-YSZ fuel electrode (e.g., resistance against moisture for steam splitting, or resistance against carbon deposition in CO₂ electrolysis), especially at the condition of high electrocatalysis current/voltage are still the long-lasting issues that should be addressed carefully. Moreover, the mechanical stability of the cell upon redox operation is still critical. Impregnating functional layers or anchoring Ni nanoparticles via exsolution could be the alternative resolutions. But how to transplant the promoting effect on button cell level to stack level is still challenging.

  For O-SOEC, the operation of the cell by fuel assistant at anode can facilitate the thermodynamics of CO₂ electrolysis at cathode and simultaneously take full advantage of the generated oxygen molecules at anode to upgrade the alkanes to valuable products. On the other hand, the basic working principle of the P-SOEC allows the generation of pure hydrogen without extra step for gas separation. The lower activation energy for proton diffusion determines its lower operation temperature than that of O-SOEC. Nevertheless, so far, the technology readiness level (TRL) of the P-SOEC is far lower than that of conventional O-SOEC, and no commercial stacks or systems has been demonstrated yet. The remaining challenges related to P-SOEC arise from both structures and materials employed in different stack/cell designs.

  From the economical perspective, the total electrolysis products amount over the SOEC stack's lifetime directly determine the unit price (such as cost of 1 kg H₂). This core parameter will be compared to that of alkaline or PEM electrolyzers which directly determine its commercialization potential. Therefore, the optimization of operation condition of the stack (temperature, gas inlet, current, etc.) should be considered collectively to maximize its merits in respect of efficiency, and minimize the weakness in poorer reliability.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This research was supported financially by the National Key Research & Development Program of China (No. 2018YFE0124700), the National Natural Science Foundation of China (Nos. 22272136, 22102135, 22202041, 22172129, 52072134, U1910209, 51876181 and 51972128) Science and Technology Projects of Innovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province (IKKEM) (No. HRTP-[2022]-23) and Hubei Province (Nos. 2021CBA149 and 2021CFA072). Y. Sun acknowledges the financial support from Guangdong Basic and Applied Basic Research Foundation (Nos. 2022A1515010069 and 2020A1515110904) and the Natural Science Foundation of Fujian Province (No. 2021J01212759).

  
  
• 1. [1]

     Z. Caineng, Y. Zhi, H. Dongbo, et al., Pet. Explor. Dev. 45 (2018) 604–618. doi: 10.1016/S1876-3804(18)30066-1

  2. [2]

     J.E. Bistline, Joule 5 (2021) 2551–2563. doi: 10.1016/j.joule.2021.09.012

  3. [3]

     K.O. Yoro, M.O. Daramola, CO₂ emission sources, greenhouse gases, and the global warming effect, in: K.O. Yoro (Ed. ) Advances in carbon capture, Elsevier, New York, 2020, pp. 3-28.

  4. [4]

     J.B. Hansen, Faraday Discuss. 182 (2015) 9–48. doi: 10.1039/C5FD90071A

  5. [5]

     M.S. Ziegler, J. Song, J.E. Trancik, Energy Environ. Sci. 14 (2021) 6074–6098. doi: 10.1039/D1EE01313K

  6. [6]

     D. Connolly, H. Lund, B.V. Mathiesen, Renew. Sust. Energ. Rev. 60 (2016) 1634–1653. doi: 10.1016/j.rser.2016.02.025

  7. [7]

     M. Aresta, A. Dibenedetto, A. Angelini, Chem. Rev. 114 (2014) 1709–1742. doi: 10.1021/cr4002758

  8. [8]

     Y. Tian, N. Abhishek, C. Yang, et al., Matter 5 (2022) 482–514. doi: 10.1016/j.matt.2021.11.013

  9. [9]

     K.R. Sridhar, B.T. Vaniman, Solid State Ionics 93 (1997) 321–328. doi: 10.1016/S0167-2738(96)00513-9

  10. [10]

      Y. Zheng, J. Wang, B. Yu, et al., Chem. Soc. Rev 46 (2017) 1427–1463. doi: 10.1039/C6CS00403B

  11. [11]

      Y. Song, X. Zhang, K. Xie, et al., Adv. Mater. 31 (2019) 1902033. doi: 10.1002/adma.201902033

  12. [12]

      Y. Zhou, Study on Fabrication and Performance of Metal-Supported Solid Oxide Fuel Cells, Springer, Shanghai, 2017.

  13. [13]

      D. Gu, G. Zhang, J. Zou, Chin. Chem. Lett. 32 (2021) 3548–3552. doi: 10.1016/j.cclet.2021.01.050

  14. [14]

      C. Graves, S.D. Ebbesen, M. Mogensen, K.S. Lackner, Renew. Sust. Energ. Rev. 15 (2011) 1–23. doi: 10.1016/j.rser.2010.07.014

  15. [15]

      A. Körner, C. Tam, S. Bennett, Gagné, Technology roadmap-hydrogen and fuel cells, International Energy Agency, Paris, 2015.

  16. [16]

      S. Shiva Kumar, V. Himabindu, Mater. Sci. Energy Technol. 2 (2019) 442–454.

  17. [17]

      R. Küngas, J. Electrochem. Soc. 167 (2020) 044508. doi: 10.1149/1945-7111/ab7099

  18. [18]

      R. Küngas, P. Blennow, T. Heiredal-Clausen, et al., Progress in SOEC Development Activities at Haldor Topsøe, 2019.

  19. [19]

      L. Wang, M. Chen, R. Küngas, et al., Renew. Sust. Energ. Rev. 110 (2019) 174–187. doi: 10.1016/j.rser.2019.04.071

  20. [20]

      I. Zvonareva, X.Z. Fu, D. Medvedev, Z. Shao, Energy Environ. Sci. 15 (2022) 439–465. doi: 10.1039/D1EE03109K

  21. [21]

      A. Hauch, R. Kungas, P. Blennow, et al., Science 370 (2020) eaba6118. doi: 10.1126/science.aba6118

  22. [22]

      X. Zhang, L. Ye, K. Xie, Energ. Fuel 36 (2022) 11576–11583. doi: 10.1021/acs.energyfuels.1c03830

  23. [23]

      N. Shi, Y. Xie, D. Huan, et al., J. Mater. Chem. A 7 (2019) 4855–4864. doi: 10.1039/C8TA12458B

  24. [24]

      L. Lei, J. Zhang, Z. Yuan, et al., Adv. Funct. Mater. 29 (2019) 1903805. doi: 10.1002/adfm.201903805

  25. [25]

      M. Laguna-Bercero, J. Power Sources 203 (2012) 4–16. doi: 10.1016/j.jpowsour.2011.12.019

  26. [26]

      M. Ni, M.K. Leung, D.Y. Leung, Int. J. Hydrogen Energy 33 (2008) 2337–2354. doi: 10.1016/j.ijhydene.2008.02.048

  27. [27]

      D.W. Strickler, W.G. Carlson, J. Am. Ceram. Soc. 47 (1964) 122–127. doi: 10.1111/j.1151-2916.1964.tb14368.x

  28. [28]

      K. Yamaji, H. Yokokawa, 8 - Oxygen ionic conductor, in: C.C. Sorrell, S. Sugihara, J. Nowotny (Eds. ), Materials for Energy Conversion Devices, Woodhead Publishing, New York, 2005, pp. 212–234.

  29. [29]

      H. Sumi, E. Suda, M. Mori, Int. J. Hydrogen Energy 42 (2017) 4449–4455. doi: 10.1016/j.ijhydene.2016.09.176

  30. [30]

      T. Ishihara, H. Matsuda, Y. Takita, JACS 116 (1994) 3801–3803. doi: 10.1021/ja00088a016

  31. [31]

      Z. Gao, H. Wang, E. Miller, et al., ACS Appl. Mater. Interfaces 9 (2017) 7115–7124. doi: 10.1021/acsami.6b15224

  32. [32]

      C. Duan, R. Kee, H. Zhu, et al., Nat. Energy 4 (2019) 230–240. doi: 10.1038/s41560-019-0333-2

  33. [33]

      N.L.R.M. Rashid, A.A. Samat, A.A. Jais, et al., Ceram. Int. 45 (2019) 6605–6615. doi: 10.1016/j.ceramint.2019.01.045

  34. [34]

      L. Gan, L. Ye, S. Wang, et al., Int. J. Hydrogen Energy 41 (2016) 1170–1175. doi: 10.1016/j.ijhydene.2015.11.032

  35. [35]

      L. Yang, S. Wang, K. Blinn, et al., Science 326 (2009) 126–129. doi: 10.1126/science.1174811

  36. [36]

      C. Duan, J. Tong, M. Shang, et al., Science 349 (2015) 1321–1326. doi: 10.1126/science.aab3987

  37. [37]

      X. Zhang, Y. Song, G. Wang, X. Bao, J. Energy Chem. 26 (2017) 839–853. doi: 10.1016/j.jechem.2017.07.003

  38. [38]

      Y. Zheng, Z. Chen, J. Zhang, Electrochem. Energy Rev. 4 (2021) 508–517. doi: 10.1007/s41918-021-00097-4

  39. [39]

      M. Chen, Y.L. Liu, J.J. Bentzen, et al., J. Electrochem. Soc. 160 (2013) F883–F891. doi: 10.1149/2.098308jes

  40. [40]

      Y. Song, Z. Zhou, X. Zhang, et al., J. Mater. Chem. A 6 (2018) 13661–13667. doi: 10.1039/C8TA02858C

  41. [41]

      T.L. Skafte, Z. Guan, M.L. Machala, et al., Nat. Energy 4 (2019) 846–855. doi: 10.1038/s41560-019-0457-4

  42. [42]

      M.C. Tucker, Int. J. Hydrogen Energy 45 (2020) 24203–24218. doi: 10.1016/j.ijhydene.2020.06.300

  43. [43]

      J. Druce, H. Téllez, M. Burriel, et al., Energy Environ. Sci. 7 (2014) 3593–3599. doi: 10.1039/C4EE01497A

  44. [44]

      C. Zhu, S. Hou, L. Hou, K. Xie, Int. J. Hydrogen Energy 43 (2018) 17040–17047. doi: 10.1016/j.ijhydene.2018.07.148

  45. [45]

      L. Ye, X. Hu, X. Wang, et al., J. Mater. Chem. A 7 (2019) 2764–2772. doi: 10.1039/C8TA10188D

  46. [46]

      Y. Zhou, Z. Zhou, Y. Song, et al., Nano Energy 50 (2018) 43–51. doi: 10.1016/j.nanoen.2018.04.054

  47. [47]

      Y. Niu, W. Huo, Y. Yu, et al., Chin. Chem. Lett. 33 (2022) 674–682. doi: 10.1016/j.cclet.2021.07.037

  48. [48]

      C. Ruan, K. Xie, L. Yang, et al., Int. J. Hydrogen Energy 39 (2014) 10338–10348. doi: 10.1016/j.ijhydene.2014.04.204

  49. [49]

      J. Carneiro, X.K. Gu, E. Tezel, E. Nikolla, Ind. Eng. Chem. Res. 59 (2020) 15884–15893. doi: 10.1021/acs.iecr.0c02773

  50. [50]

      C. Zhu, L. Hou, S. Li, et al., J. Power Sources 363 (2017) 177–184. doi: 10.1016/j.jpowsour.2017.07.070

  51. [51]

      L. Zhang, Y. Tian, Y. Liu, et al., ChemElectroChem 6 (2019) 1359–1364. doi: 10.1002/celc.201801831

  52. [52]

      O. Kwon, S. Joo, S. Choi, et al., J. Phys. Energy 2 (2020) 032001. doi: 10.1088/2515-7655/ab8c1f

  53. [53]

      D. Neagu, G. Tsekouras, D.N. Miller, et al., Nat. Chem. 5 (2013) 916–923. doi: 10.1038/nchem.1773

  54. [54]

      S. Joo, O. Kwon, K. Kim, et al., Nat. Commun. 10 (2019) 697. doi: 10.1038/s41467-019-08624-0

  55. [55]

      H. Han, J. Park, S.Y. Nam, et al., Nat. Commun. 10 (2019) 1471. doi: 10.1038/s41467-019-09395-4

  56. [56]

      J.H. Myung, D. Neagu, D.N. Miller, J.T. Irvine, Nature 537 (2016) 528–531. doi: 10.1038/nature19090

  57. [57]

      C. Ruan, K. Xie, Catal. Sci. Technol. 5 (2015) 1929–1940. doi: 10.1039/C4CY01254B

  58. [58]

      G. Tsekouras, D. Neagu, J.T.S. Irvine, Energy Environ. Sci. 6 (2013) 256–266. doi: 10.1039/C2EE22547F

  59. [59]

      S. Liu, Q. Liu, J.L. Luo, J. Mater. Chem. A 4 (2016) 17521–17528. doi: 10.1039/C6TA06365A

  60. [60]

      X. Sun, Y. Ye, M. Zhou, et al., J. Mater. Chem. A (2022) 2327–2335.

  61. [61]

      S. Ding, M. Li, W. Pang, et al., Electrochim. Acta 335 (2020) 135683. doi: 10.1016/j.electacta.2020.135683

  62. [62]

      Y. Tian, Y. Liu, A. Naden, et al., J. Mater. Chem. A 8 (2020) 14895–14899. doi: 10.1039/D0TA05518B

  63. [63]

      H. Lv, L. Lin, X. Zhang, et al., J. Mater. Chem. A 7 (2019) 11967–11975. doi: 10.1039/C9TA03065D

  64. [64]

      H. Lv, L. Lin, X. Zhang, et al., Nat. Commun. 12 (2021) 1–11. doi: 10.1038/s41467-020-20314-w

  65. [65]

      J. Lu, C. Zhu, C. Pan, et al., Sci. Adv. 4 (2018) eaar5100. doi: 10.1126/sciadv.aar5100

  66. [66]

      H. Lv, L. Lin, X. Zhang, et al., Adv. Mater. 32 (2020) 1906193. doi: 10.1002/adma.201906193

  67. [67]

      N. Han, R. Ren, M. Ma, et al., Chin. Chem. Lett. 33 (2022) 2658–2662. doi: 10.1016/j.cclet.2021.09.100

  68. [68]

      J. Sun, R. Ren, H. Yue, et al., Chin. Chem. Lett. 33 (2022) 107776.

  69. [69]

      N.P. Brandon, S. Skinner, B.C. Steele, Annu. Rev. Mater. Res. 33 (2003) 183–213. doi: 10.1146/annurev.matsci.33.022802.094122

  70. [70]

      J. Fleig, Annu. Rev. Mater. Res. 33 (2003) 361–382. doi: 10.1146/annurev.matsci.33.022802.093258

  71. [71]

      K. Chen, N. Ai, Int. J. Hydrogen Energy 39 (2014) 10349–10358. doi: 10.1016/j.ijhydene.2014.05.013

  72. [72]

      S.B. Adler, Chem. Rev. 104 (2004) 4791–4844. doi: 10.1021/cr020724o

  73. [73]

      M.A. Laguna-Bercero, A.R. Hanifi, H. Monzon, et al., J. Mater. Chem. A 2 (2014) 9764–9770. doi: 10.1039/C4TA00665H

  74. [74]

      A. Chroneos, D. Parfitt, J.A. Kilner, R.W. Grimes, J. Mater. Chem. 20 (2010) 266–270. doi: 10.1039/B917118E

  75. [75]

      S. Streule, A. Podlesnyak, D. Sheptyakov, et al., Phys. Rev. B 73 (2006) 094203. doi: 10.1103/PhysRevB.73.094203

  76. [76]

      A. Tarancón, D. Marrero-López, J. Peña-Martínez, et al., Solid State Ionics 179 (2008) 611–618. doi: 10.1016/j.ssi.2008.04.028

  77. [77]

      Q. Liu, C. Yang, X. Dong, F. Chen, Int. J. Hydrogen Energy 35 (2010) 10039–10044. doi: 10.1016/j.ijhydene.2010.08.016

  78. [78]

      Y. Tian, J. Li, Y. Liu, et al., Int. J. Hydrogen Energy 43 (2018) 12603–12609. doi: 10.1016/j.ijhydene.2018.03.187

  79. [79]

      Y. Tian, D. Yan, B. Chi, et al., ECS Trans. 78 (2017) 533. doi: 10.1149/07801.0533ecst

  80. [80]

      J. Laurencin, M. Hubert, D.F. Sanchez, et al., Electrochim. Acta 241 (2017) 459–476. doi: 10.1016/j.electacta.2017.05.011

  81. [81]

      Q. Lyu, N. Wang, Z. Li, et al., J. Power Sources 521 (2022) 230933. doi: 10.1016/j.jpowsour.2021.230933

  82. [82]

      X. Zhang, S. Ohara, R. Maric, et al., J. Power Sources 83 (1999) 170–177. doi: 10.1016/S0378-7753(99)00293-1

  83. [83]

      M.A. Laguna-Bercero, A.R. Hanifi, L. Menand, et al., Electrochim. Acta 268 (2018) 195–201. doi: 10.1016/j.electacta.2018.02.055

  84. [84]

      L. Mingyi, Y. Bo, X. Jingming, C. Jing, Int. J. Hydrogen Energy 35 (2010) 2670–2674. doi: 10.1016/j.ijhydene.2009.04.027

  85. [85]

      H. Fan, M. Keane, N. Li, et al., Int. J. Hydrogen Energy 39 (2014) 14071–14078. doi: 10.1016/j.ijhydene.2014.05.149

  86. [86]

      N. Li, L. Wang, M. Wang, et al., J. Power Sources 518 (2022) 230787. doi: 10.1016/j.jpowsour.2021.230787

  87. [87]

      X. Wang, B. Yu, W. Zhang, et al., Int. J. Hydrogen Energy 37 (2012) 12833–12838. doi: 10.1016/j.ijhydene.2012.05.093

  88. [88]

      M.S. Khan, X. Xu, R. Knibbe, Z. Zhu, ACS Appl. Mater. Interfaces 10 (2018) 25295–25302. doi: 10.1021/acsami.8b05504

  89. [89]

      H. Xu, B. Chen, J. Irvine, M. Ni, Int. J. Hydrogen Energy 41 (2016) 21839–21849. doi: 10.1016/j.ijhydene.2016.10.026

  90. [90]

      Y. Wang, T. Liu, S. Fang, et al., J. Power Sources 277 (2015) 261–267. doi: 10.1016/j.jpowsour.2014.11.092

  91. [91]

      C. Zhu, S. Hou, X. Hu, et al., Nat. Commun. 10 (2019) 1173. doi: 10.1038/s41467-019-09083-3

  92. [92]

      Y. Song, L. Lin, W. Feng, et al., Angew. Chem. Int. Ed. 58 (2019) 16043–16046. doi: 10.1002/anie.201908388

  93. [93]

      L. Lei, Y. Wang, S. Fang, et al., Appl. Energy 173 (2016) 52–58. doi: 10.1016/j.apenergy.2016.03.116

  94. [94]

      X. Zhang, L. Ye, H. Li, et al., ACS Catal. 10 (2020) 3505–3513. doi: 10.1021/acscatal.9b05409

  95. [95]

      M. Li, B. Hua, L.C. Wang, et al., Nat. Catal. 4 (2021) 274–283. doi: 10.1038/s41929-021-00590-5

  96. [96]

      Z. Pan, C. Duan, T. Pritchard, et al., Appl. Catal. B 307 (2022) 121196. doi: 10.1016/j.apcatb.2022.121196

  97. [97]

      J.C.W. Mah, A. Muchtar, M.R. Somalu, M.J. Ghazali, Int. J. Hydrogen Energy 42 (2017) 9219–9229. doi: 10.1016/j.ijhydene.2016.03.195

  98. [98]

      E. Dogdibegovic, S. Ibanez, A. Wallace, et al., Int. J. Hydrogen Energy 47 (2022) 24279–24286. doi: 10.1016/j.ijhydene.2022.05.206

  99. [99]

      P. Piccardo, R. Spotorno, C. Geipel, Energies 15 (2022) 3458. doi: 10.3390/en15093458

  100. [100]

       K. Singh, T. Walia, Int. J. Energy Res. 45 (2021) 20559–20582. doi: 10.1002/er.7161

  101. [101]

       A. Hauch, J.R. Bowen, L.T. Kuhn, M. Mogensen, Electrochem. Solid St. 11 (2008) B38–B41. doi: 10.1149/1.2828845

  102. [102]

       Y. Zheng, Q. Li, W. Guan, et al., Ceram. Int. 40 (2014) 5801–5809. doi: 10.1016/j.ceramint.2013.11.020

  103. [103]

       Y. Zheng, Q. Li, T. Chen, et al., Int. J. Hydrogen Energy 40 (2015) 2460–2472. doi: 10.1016/j.ijhydene.2014.12.101

  104. [104]

       X. Zhang, J.E. O'Brien, R.C. O'Brien, et al., Int. J. Hydrogen Energy 38 (2013) 20–28. doi: 10.1016/j.ijhydene.2012.09.176

  105. [105]

       J. Mougin, A. Chatroux, K. Couturier, et al., Energy Procedia 29 (2012) 445–454. doi: 10.1016/j.egypro.2012.09.052

  106. [106]

       Q. Fang, L. Blum, N.H. Menzler, J. Electrochem. Soc. 162 (2015) F907–F912. doi: 10.1149/2.0941508jes

  107. [107]

       M. Reytier, S. Di Iorio, A. Chatroux, et al., Int. J. Hydrogen Energy 40 (2015) 11370–11377. doi: 10.1016/j.ijhydene.2015.04.085

  108. [108]

       M. Riedel, M.P. Heddrich, A. Ansar, et al., J. Power Sources 475 (2020) 228682. doi: 10.1016/j.jpowsour.2020.228682

  109. [109]

       R. Küngas, P. Blennow, T. Heiredal-Clausen, et al., ECS Trans. 91 (2019) 215–223. doi: 10.1149/09101.0215ecst

• 

  Figure 1  (a) The schematic of carbon neutral cycle based on renewable energy and solid oxide electrolysis cell. Copied with permission [11]. Copyright 2019, Wiley. (b) The SOEC devices with different configuration and appearances. Copied with permission [12]. Copyright 2017, Springer Nature. (c) Thermodynamics of CO₂ and H₂O reduction reactions. Copied with permission [14]. Copyright 2011, Elsevier. (d) Schematic representation of technology development potential of different electrolyzers. Copied with permission [15]. Copyright 2015, International Energy Agency.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) The configuration schematic of solid oxide electrolysis cell (SOEC) with supported cathode layer, fuel cathode layer, electrolyte, buffer layer and anode layer. Copied with permission [21]. Copyright 2020, Springer Nature. (b) The temperature dependent ionic conductivity plot of doped ZrO₂. Copied with permission [27]. Copyright 2022, The American Ceramic Society. (c) The temperature dependent ionic conductivity plot of doped LSG samples. Copied with permission [30]. Copyright 1994, American Chemical Society. (d, e) The SEM image of Co-doped Sr₂Fe_1.5Mo_0.5O_6–δ (SFMC) reduced for 2 h and the EIS plot for SFMC before and after exsolution for CO₂ electrolysis at 1.6 V and 800 ℃. Copied with permission [66]. Copyright 2020, Wiley. (f) polarization curve and SEM image comparison of cell with straight pore and tortuous cathode decorated SOECs. Copied with permission [86]. Copyright 2022, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) The model of CH₄ assist-SOEC. Copied with permission [89]. Copyright 2016, Elsevier. (b) The calculated energy consumption for conventional H₂O/CO₂ co-electrolysis and CH₄ assisted electrolysis. Copied with permission [89]. Copyright 2016, Elsevier. (c) The schematic of CH₄ oxidative coupling assisted CO₂ electrolysis. Copied with permission [65]. Copyright 2018, Wiley. (d) The polarization curves of conventional SOEC and carbon gasification assisted SOEC at 850 ℃. Copied with permission [93]. Copyright 2016, Elsevier. (e) The calculated thermodynamic equilibrium results for the reaction of 2 mol O₂ with 98 mol carbon at different temperatures. Copied with permission [93]. Copyright 2016, Elsevier. (f) The amount of H₂ depends on the current in the cathode in C₂H₆ dehydrogenation assisted P-SOEC. Copied with permission [94]. Copyright 2020, American Chemical Society. (g) CH₄ partial current density and partial Faradaic efficiency in C₂H₆ dehydrogenation assisted P-SOEC. Copied with permission [95]. Copyright 2020, Springer Nature.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a, b) 3D view of the stack module mounting. Copied with permission [102]. Copyright 2014, Elsevier. (c) Effect of different flows on the hydrogen electrode on the steam conversion rates and electrolysis efficiencies of the stack module at 800 ℃ and a constant 80% H₂O-20% H₂. Copied with permission [102]. Copyright 2014, Elsevier. (d, e) SEM images of the cross-section (fracture surface) of the Ni-YSZ/YSZ/LSM–YSZ and Ni-YSZ/YSZ/GDC/LSCF-GDC cells after electrolysis at air inlet area. Copied with permission [102,103]. Copyright 2015, Elsevier. (f) Stack with NiO-YSZ/YSZ/LSCF cell test 1000 h at 0.2 A/cm² in electrolysis mode. Copied with permission [104]. Copyright 2013, Elsevier. (g) Gas outlet measurements of the 10-cell stack at 800 ℃ with cathodic inlet: 65% H₂O-25% CO₂-10% H₂ with conversion rate of oxidized species is 32% at 0.5 A/cm². Copied with permission [107]. Copyright 2015, Elsevier. (h) Gas analysis of the CSC and the ESC stack according to the steady state co-electrolysis experiments. Copied with permission [108]. Copyright 2020, Elsevier.

  下载: 全尺寸图片 幻灯片
•