Interfacial Disordering and Heterojunction Enabling Fast Proton Conduction

doi:10.21203/rs.3.rs-2436390/v1

Download PDF

Article

Interfacial Disordering and Heterojunction Enabling Fast Proton Conduction

https://doi.org/10.21203/rs.3.rs-2436390/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Interfacial disorder is a general method to change the metal-oxygen compatibility and carrier density of heterostructure materials for ionic transport modulation. Herein, to enable high proton conduction, a semiconductor heterostructure based on spinel ZnFe₂O₄ (ZFO) and fluorite CeO₂ is developed and investigated in terms of structural characterization, first principle calculation, and electrochemical performance. Particular attentions are paid to the interfacial disordering and heterojunction effects of the material. Results show that the heterostructure induces a disordered oxygen region at the hetero-interface of ZFO-CeO₂ by dislocating oxygen atoms, leading to fast proton transport. As a result, the ZFO-CeO₂ exhibits a high proton conductivity of 0.21 S/cm and promising fuel cell power output of 1070 mW/cm² at 510 ℃. Based upon these findings, a new mechanism is proposed to interpret the diffusion and acceleration of protons in ZFO-CeO₂. Our study provides a new strategy to customize semiconductor heterostructure to enable fast proton conduction.

Physical sciences/Energy science and technology/Fuel cells

Physical sciences/Energy science and technology/Renewable energy

In recent years, research interest in protonic ceramics fuel cells (PCFCs) is growing to develop cost-effective solid oxide fuel cells (SOFCs) by lowering operational temperatures¹. The critical challenges of PCFCs are to search for high proton-conducting electrolytes and suitable cathodes². So far, the most frequently used proton-conducting ceramics are based on perovskite oxides BaCeO₃ and BaZrO₃³. However, their proton conductivities are in the level of 10^-3~10^-2 S/cm at 600 ℃^{4, 5}, which are far below the required value of 10^-1 S/cm for reaching high fuel cell performance. Thus, the performance of current PCFCs still lags behind that of the SOFC based on yttria-stabilized zirconia (YSZ). Facing this challenge, a new scenario has emerged recently to develop semiconductor ionic materials with remarkably high proton conduction to replace those perovskite proton electrolytes, drawing intensive attention from the PCFC field^{6, 7, 8}.

Semiconductor ionic materials hold great promise to address the fundamental issues of PCFC electrolytes^{9, 10}. Recent studies have found that proton conduction can be tuned in SmNiO₃¹¹, ZnO¹², CeO₂¹³, Sm₂O₃¹⁴_,and La₂O₃¹⁵ via surface engineering under fuel cell operating conditions such as forming core-shell structure¹³ and through surface doping,¹⁶ resulting in high proton conductivity of ~0.1 S/cm at 550 °C ¹⁷. Furthermore, by forming heterostructure, the charge carrier transport, thermal stability, and the mechanical strength of the semiconductor ionic materials can be modulated via creating structural disorder in the lattice and built-in-electric field (BIEF) at the interface^{18, 19}. Particularly, high ionic conductivity can be invoked in the semiconductor heterostructure materials with dissimilar structures (such as fluorite and perovskite), as highly disordered oxygen planes are easily produced at the hetero-interface due to lattice mismatch between two different phases, leading to high-mobility pathway to yield fast ionic transport²⁰ ²¹ ²². In addition, it would also be possible to induce high ionic conduction in the semiconductor heterostructure by the junction effect and BIEF, which can suppress the electron flow to avoid fuel cell short circuit. For instance, by virtue of hetero-interface and BIEF, the heterostructure of La_0.6Sr_0.4Co_0.2Fe_0.8O_3-δ (LSCF) and Ca_0.04Ce_0.8Sm_0.16O_2-δ (SCDC) has achieved a high ionic conductivity of 0.188 S cm^-1 at 600 °C with high SOFC performance^{23, 24}.

When it comes to the design of a new alternative PCFC electrolyte, spinel oxide ZnFe₂O₄ (ZFO) and fluorite oxide CeO₂ are taken into account to develop a semiconductor ionic heterostructure. ZFO belongs to the spinel ferrite mineral with the general formula of MFe₂O₄ (M is a metal ion such as Zn²⁺, Ni²⁺, Fe²⁺, Mg²⁺, Ca²⁺, etc.). In the inverse spinel structure, all the Zn²⁺ are replaced by Fe³⁺ via depriving the octahedral sites, which acts as the role of lattice defect phase²⁵. By creating oxygen vacancies and structural disorder in its spinel structure, the ZFO can attain improved electrical conductivity for high-efficient photocatalytic and energy storage applications^{26, 27}. Fluorite CeO₂ is a typical basis for developing electrolyte materials for intermediate-temperature SOFC through ion doping. Recent advances proposed a non-doping approach to explore the ionic conductivity of CeO₂ based on surface engineering. They found that the formation of CeO_2-δ@CeO₂ core-shell structure leads to a space charge region between the surface and core to facilitate the ionic transport. As a result, the CeO₂ sample obtained a high proton conductivity of 0.16 S/cm at 520 °C and 200 h durability when used in SOFCs as the electrolyte layer^{13, 28}. Such spinel ZFO and fluorite CeO₂have highly dissimilar structures, making it possible to create interfacial disorder and break the compatibility between the two phases to develop high ionic conduction^{22, 29}. Thus, the ZFO and CeO₂ hold great promise as the basic materials to construct a heterostructure electrolyte for PCFCs.

Following this line, in the present work, a new semiconductor heterostructure material made of ZFO and CeO₂ is proposed to develop alternative electrolytes for PCFCs via interfacial disordering and heterojunction studies. The ZFO-CeO₂ is investigated in terms of material characterization and first principle calculations, followed by being applied in fuel cells for proton conductivity and electrochemical performance tests. The designed heterostructure material exhibits high proton conductivity, appreciable fuel cell performance, and long-term durability at low operating temperatures of 360-510 ℃. Our findings and analysis suggest the great potential of ZFO-CeO₂ for PCFC uses, and this study provides a new strategy to customize the semiconductor heterostructure to enable fast proton conduction.

Phase structure and morphology. Figure 1(a) exhibits the XRD pattern of the prepared ZFO, CeO₂, and ZFO-CeO₂. The high-intensity peak of ZFO is located at 2\({\theta }^\circ\) angle of (3 1 1) plane, verifying the single-phase cubic spinel structure of ZnFe₂O₄. The cubic crystal structure of the spinel ferrite sample (ZFO) is verified using jade and X\({\prime }\)Pert high score plus software with the JCPD CARDS of #81–0792 and #01-087-1230. The XRD refinement analysis was also carried out to confirm the exact single-phase cubic nature of the synthesized ZFO. From the refinement data, it is verified that ZFO belongs to the Fd-3m space group with lattice parameters a = b = c = 8.4432 Å³⁰. The theoretical parameters of the prepared ZFO sample are in good agreement with the spinel ferrite crystal structure literature ³¹. While CeO₂ phase was also confirmed using X\({\prime }\)Pert high score plus software. It is found that the high intensity peaks are located at 2\({\theta }^\circ\) angle of values of 28.54\(^\circ\) (1 1 1), 33.07\(^\circ\) (2 0 0), 47.47\(^\circ\) (2 2 0), 56.33\(^\circ\) (3 1 1), 59.07\(^\circ\) (2 2 2), 69.40 (4 0 0) with standard JCPDS#01-081-0792. The prepared CeO₂ belongs to the fluorite structure with lattice parameter values of a = b = c = 5.4124^{32, 33}. In the XRD pattern of ZFO-CeO₂ (Supplementary Fig. S1), the diffraction peaks of ZFO and CeO₂ are identified without extra peaks, suggesting the successful formation of the heterostructure of ZFO and CeO₂.

Morphological analysis of pure ZFO, pure CeO₂, and ZFO-CeO₂ heterostructure composite were studied using high-resolution transmission electron microscopy (HR-TEM) and energy dispersive spectroscopy (EDS). Figure 1(b) exhibits the randomly distributed spherical-shaped ZFO nanoparticles with a size of 45–50 nm, which identifies a clear interface between ZFO and CeO₂ grains (See Fig. b3). Two methods (Line intercept method and Digital Micrograph) were employed to measure the lattice planes and d spacing of the ZFO-CeO₂. The d spacing of the ZFO-CeO₂ heterostructure agrees with the XRD measured data. The slight difference between lattice spaces, i.e., 2.67 Å and 2.69 Å, as obtained from Digital Micrograph software, verifies the exact formation of ZFO-CeO₂ heterostructure. The elemental distribution of ZFO and CeO₂ nanoparticles in the heterostructure was further confirmed using Selected Area Electron Diffraction (SAED) and EDS mapping analysis as depicted in Fig. 1c (c1-c5) and 1d (d1 and d2). Zn, Fe, Ce, and O particles are randomly distributed in the sample to indicate the successful heterostructure.

First principle calculations. Density functional theory (DFT) calculation is an appropriate way to theoretically analyze the ions diffusion mechanism and evaluate the semiconducting nature of the prepared samples. The crystal structures and electronic bandgaps of ZFO and CeO₂ have also been considered and presented in Fig. 2 (a-d). The structures, electronic bandgap, and density of state (DOS) are constructed and evaluated using Quantum ATK software with LDA and GGA approximations. Further procedure to calculate electronic band structure and DOS has been explained in detail in the experimental section. The obtained electronic bandgap using DFT calculations for ZFO and CeO₂ samples is 1.85 eV and 3.5 eV, respectively, which is similar to our experimental values, where the optical bandgaps obtained from UV-Vis spectra are presented in Supplementary Fig. S2. The calculated optical bandgaps of the pure ZFO and CeO₂ samples were 1.85 eV and 3.15 eV, respectively. The small error between electronic bandgap and optical bandgap values has been noted and verified with previous literature ³⁴, with the fact that different U values have been used in first principle calculations. The DOS of individual elements and projected DOS indicates the constitution of each element to the valance and conduction zone in the pristine ZFO and CeO₂ samples principally, as shown in Fig. 2 (e). In ZFO, it has been observed that O ions principally exist in the valance zone, and Fe ions are in the conduction zone with the smaller contribution of oxygen ions. Steven Jankov et al.³⁵ reported on the existence of each element in its valance and conduction zones, and our work agrees with their study. For CeO₂, the fully occupied and unoccupied states are composed of O-2p states and Ce-4 f states, which is in accordance with the report of Rong Han et al.³⁶.

Interfacial disordering and heterojunction. To evaluate the interfacial disorder and conduction mechanism of ZFO-CeO₂, DFT was employed with the computational parameters of the constructed structure, as mentioned in the experimental section. Our calculations based on heterostructure showed that the lattice mismatch is less than 5% by indicating valid interface formation between the ZFO and CeO₂ using the QUANTUM ATK builder. The interfacial structure without optimization geometry (OG) is presented in the calculations in Fig. 3(a). It can be observed that the constructed heterostructure has an interfacial connection between ZFO and CeO₂, and each participating atoms of both samples (Zn, Fe, Ce, and O) are well organized with similar M-O distances as pure ZFO and CeO₂ structure. After OG, by applying force tolerance of 1\(\times\)10⁻⁵ eV/Å, the disorder crystal structure with different variations in M-O distances are observed, as displayed in Fig. 3 (b). The M-O distances of the constructed heterostructure (before and after) using OG are calculated and inserted in Supplementary Fig. S3 (a and b). At the same time, before OG, the band structure is displayed as I, and after OG, it is inserted as II in Fig. 2(i and ii). Such a kind of disordering induces a disordered oxygen plane at the interface between ZFO and CeO₂ by dislocating oxygen atoms. It can be found that the electronic states of oxygen (O) exhibit hybridized nature with those of ZFO and CeO₂ species, which play a crucial role in promoting proton migration. The interfacial disorder dominates at the interface grain boundaries of ZFO and CeO₂ as both structures break the compatibility and create disorder at the interface during heterostructure formation. The ZFO-CeO₂ heterostructure represents to 6 space group and P-2y hall symbol as presented in Supplementary Fig. S4. Interestingly, The ZFO-CeO₂ interface atoms are well connected, without crystal structure breakage and presented in Fig. S4 (c). Therefore, we conclude that interface disorder has been successfully created at ZFO/CeO₂ hetero-interface because of weak M-O bonding, which can be beneficial for ionic migrations.

Furthermore, the interfacial diagram and electronic states can be verified by the calculated DOS, as illustrated in Fig. 3(c). The crystal orbital overlap population (COOP) suggests a strong relation between ZFO and CeO₂ lattices. COOP calculations elaborate the projection of DOS based on specific molecular bonds, which can be constructed as a complementary tool to understand the projected DOS aspects³⁷. COOPs can further reveal the oxygen vacancy concentration and additional bonding information of the built structures. In our case, COOP demonstrates that the oxidation between ZFO and CeO₂ is significant in understanding the constructed heterostructure electronic states. The band shifting towards the Fermi energy level is another crucial parameter to understand the DOS of ZFO-CeO₂ heterostructure. Naveed et al.³⁸ reported that the band moving near the Fermi level is more favorable to enhance the DOS of neighboring atoms, and the improved DOS at the Fermi level may cause increased efficiency of ionic transport. Our results show that the total DOSs of simple ZFO and CeO₂ are far from the fermi level compared to the case of ZFO-CeO₂ heterostructure. Moreover, the DOSs of each element in the ZFO-CeO₂ heterostructure present a shift towards the fermi level, indicative of an increased transfer rate of the electrons to the adsorbed oxygen species. The higher DOS in ZFO-CeO₂ heterostructure, especially in Ce neighboring atoms near the fermi level, can be a reason to increase the charge transfer rate.

When using the ZFO-CeO₂ heterostructure to replace conventional electrolytes to fabricate PCFCs, one question may be raised that how to avoid electronic short-circuit problem. Previous studies have proposed to use hetero-junctions to regulate the charge carriers of semiconductors to evade the current leakage risk of fuel cells^{23, 39}. Typically, the p-n bulk-heterojunction based on p-type and n-type semiconductors⁴⁰, and the metal-semiconductor Schottky formed between the reduced anode (metal or alloy) and the semiconductor electrolyte have been applied in design PCFCs⁴¹. In this study, a catalyst Ni_0.8Co_0.15Al_0.05LiO_2−δ (NCAL) with high H⁺/O²⁻/e⁻ triple conductivity and good catalytic activity was employed as the symmetrical electrodes⁴². At the anode side, the NCAL can be reduced into Ni/Co alloys in an H₂ environment, thus establishing interaction with the semiconductor heterostructure to build a Schottky contact. The Schottky junction involves the propagation of space charge region to block electrons from passing through the anode/electrolyte interface to block electrons crossover and avoid the short-circuiting issue in the fuel cell. Figure 3(d) shows the electron density difference (EDD) of ZFO-CeO₂ heterostructure under the equally applied strains to investigate the charger transfer rate. The different positive and negative regions represent the accumulation of electrons in the ZFO-CeO₂. The positive region in the EDD indicates electron accumulation, and the negative region reflects the electron depletion of the fabricated device.

Conducting property and electrochemical performance. To understand the oxidation states of the constructed heterostructure, X-ray photoelectron spectra (XPS) measurement was performed to verify the surface chemical states of these samples. Figure 4(a) exhibits the XPS survey spectra of pure ZFO and CeO₂ compared to the ZFO-CeO₂ heterostructure. The spectra confirm the presence of Zn, Fe, Ce, and O elements on the surface of the ZFO-CeO₂ sample. For CeO₂, it was observed in work reported by Zhang et al. ⁴³ that the ceria oxidation states play a crucial role in enhancing oxygen vacancies at the constructed sample interface. Figure 4 (b and c) presents the O 1s core-level spectra of pure ZFO and the ZFO-CeO₂ heterostructure for comparison. The O 1s peaks of the two samples are deconvoluted into two peaks at 528.7-530.1 eV and 528.4-530.8 eV, respectively, corresponding to adsorbed oxygen specie (O_ads) and lattice oxygen species (O_lattice). It is calculated that the relative ratio value of (O_ads/O_lattice) increases from 0.654 for ZFO to 0.817 for ZFO-CeO_2, indicative of significantly incremented oxygen vacancies on the surface of the heterostructure sample. Cai et al. ⁴⁴ attributed the higher relative ratio of from 1.13 for SDC to 1.21 for SDC–STO heterostructure as a result of improved concentration of hetero-interface oxygen vacancies due to lattice mismatch, which would lead to enhanced ionic conductivity.

Figure 4(d and e) displays the measured results of ultraviolet photoelectron spectroscopy (UPS) for pure ZFO and ZFO-CeO₂ heterostructure. The optical bandgap for all synthesized samples is calculated using the Kubelka − Munk function. The valance band (VB) maxima in the UPS exhibited an energy cutoff of 21.20 eV, calibrated by He-I light energy. The VB maxima of both ZFO and ZFO-CeO₂ heterostructure-based samples are determined by the cutoff of low binding energy and high binding energy level. The UPS measurements show valance band energy levels are 6.17 eV for ZFO and 6.88 eV for ZFO-CeO₂ heterostructure. In parallel, the conduction band edges for all synthesized samples are calculated to be 4.33 eV to 4.47 eV. The detailed bandgap analysis of ZFO-CeO₂ heterostructure has been carried out using DFT calculations.

Moreover, the fuel cells electrochemical impedance spectroscopy (EIS) was measured at 360\(-\)510 °C under open circuit voltage (OCV) conditions. Typical EIS spectra of ZFO, CeO_2, and ZFO-CeO₂ based PCFCs measured at 510°C are presented in Fig. 4(f), and the corresponding fitting parameters based on the EIS measured at 485-360°C are displayed in Supplementary Fig. S7 (a, b and c). An equivalent circuit of LR_o(R₁-Q₁)(R₂-Q₂) was used to fit the experimental data recorded at 510°C by ZSIMPWIN Software, where R represented resistance and Q was the constant phase element representing a non-ideal capacitor. The fitting parameters of ZFO, CeO₂ and ZFO-CeO₂ devices are listed in Supplementary Tables S1, S2, and S3, respectively. According to the calculated capacitance for each resistance, the R₁ can be ascribed to the charge transfer process while the R₂ is associated with the mass transfer process. The smaller values of R_o for ZFO-CeO₂ PCFC compared to the other two cells indicate higher ionic conduction of ZFO-CeO₂ than pure ZFO and CeO₂. The ZFO-CeO₂ PCFC also presents lower values of R₁ and R₂, suggesting that the device has higher electrode reaction activity while using the same NCAL electrodes. As reported, the grain boundaries (interface regions) play a crucial role in ion transportation, leading to promoted grain boundary conductance and fast electrode reactions ^{45, 46, 47}. Thus, we conjecture that the high interfacial ionic conductivity of ZFO-CeO₂ brought about fast grain-boundary conduction and promoted both electrolyte ionic transport and electrode dynamic processes. The ionic conductivity (σ_i) was estimated based on the ohmic linear part of the I-V polarization curve and presented in Fig. 5(g). The obtained ionic conductivity of ZFO-CeO₂ in the cell exhibits high values of 0.11-0.21 S/cm at 460-510°C, which is remarkably higher than that of pure ZFO and CeO₂. The corresponding activation energy (Ea) can be calculated from Arrhenius equation. As shown in the inset of Fig. 5(g), the activation energy values for ZFO-CeO₂, CeO₂ and ZFO are 0.63 eV, 0.80 eV and 0.88 eV, respectively. The small activation energy of ZFO-CeO₂ reflects that proton transport dominates the total conductivity of ZFO-CeO₂. It is also noticed that ZFO-CeO₂ exhibits lower activation energy as compared to CeO₂ and ZFO, which suggests that the ZFO-CeO₂ sample probably gained improved proton conduction by forming a heterostructure.

To verify the proton conduction, the proton-conducting property of ZFO-CeO₂ was investigated by isotopic effect in 5% H₂/95% Ar and 5% D₂ 95% Ar atmospheres in comparison to air at different operational temperatures. The EIS measurements in the D₂, H₂ (5%), and H₂/air atmosphere at 510°C are demonstrated in Fig S8 (Supplementary Information), while the calculated conductivities from EIS data are presented in Fig. 5(h). The temperature-dependent conductivity of the ZFO-CeO₂ sample shows a clear H₂-D₂ isotopic effect, i.e., the conductivity is significantly higher in H₂ (5%) that in D₂. We can conclude that proton migration is more favorable in the ZFO-CeO₂ heterostructure. Figure 4(i) presents the calculated conductivity of ZFO-CeO₂ compared to the reported results of some typical electrolytes ^{48, 49, 50}. As can be seen, the conductivity of ZFO-CeO₂ is not only remarkably higher than those of YSZ and La_0.9Sr_0.1Ga_0.8Mg_0.2O₃ (LSGM) but also superior to that of proton conductors BaZrO₃ (BZY) and SmNiO₃.

The proton conduction of ZFO-CeO₂ heterostructure was further evaluated in PCFCs at 510 − 360°C. The current-voltage (I-V) and current-power (I-P) characteristics of PCFCs were measured in H₂/air. The power outputs of the PCFCs based on ZFO, CeO₂, and ZFO-CeO₂ as electrolytes are presented in Fig. 5(a) for comparison, while I-V & I-P characteristics of the ZFO-CeO₂ device at 510 − 360°C is shown in Fig. 5(b), and those for pure ZFO and CeO₂ cells are shown in Supplementary Fig. S9 and S10. It is observed that the constructed fuel cells attained high OCVs above 1.05 V at 510 − 360°C. The ZFO, CeO₂, and ZFO-CeO₂ based devices achieved peak power densities of 427 mW/cm², 515 mW/cm², and 1070 mW/cm² at 510°C, respectively, indicating noticeable performance increment by forming the heterostructure. The obtained higher power outputs can be ascribed to two aspects: i) the interfacial disordering leads to faster ionic transport at the interface, resulting in reduced ohmic resistance of the ZFO-CeO₂ as compared to individual CeO₂ and ZFO; ii) the hetero-junction provided BIEF at the interface, which accelerated the ions to enhance the mobility thus further reduce the ohmic resistance of ZFO-CeO₂. Moreover, the peak power densities (PPDs) of ZFO-CeO₂ based cell reach 790 mW/cm², 732 mW/cm², 425 mW/cm², 240 mW/cm², 201 mW/cm², 70 mW/cm² at 485°C, 460°C, 435°C, 410°C, 385°C and 360°C, respectively, manifesting a good capability of low-temperature running, whereas the fuel cells with pure ZFO and CeO₂ display much lower outputs and a difficulty to be operated below 460°C. In our measurements, it is also acquired that the synthesized ZFO-CeO₂ with mass ratio of 2:8 is the optimal sample with the highest power density, while the performances of the cells with 3:7, 4:6, and 5:5 ZFO-CeO₂ present lower power outputs as presented in Supplementary Fig. S11. Additionally, the ZFO-CeO₂ fuel cell was assembled with a BaZr_0.8Y_0.2O_3−δ (BZY) filter layer to verify the proton conductivity of ZFO-CeO₂ in a configuration of Ni-NCAL/BZY/ZFO-CeO₂/BZY/Ni-NCAL, as the BZY layer allows only protons to pass through the electrolyte with negligible oxide ions and electron migration ⁵¹. The I-V and I-P characteristics of the ion-filtering cell and the corresponding cross-sectional SEM image are presented in Fig. 5 (c) and Fig. 5(d), respectively. The ion-filtering cell achieved a PPD of 961 mW/cm² at 510°C, which is slightly lower but comparable to the PPD of the ZFO-CeO₂ cell without the filter layer. The slight loss can be attributed to the extra polarization resistance of the device caused by the two additional interfaces of BZY and ZFO-CeO₂. For comparison, the ZFO electrolyte cell was also assembled with a BZY filter layer and achieved a PPD of 415 mW/cm² at 510°C (Supplementary Fig. S12).

Eventually, the durability of the proposed electrolyte and fuel cells was measured. Figure 5(e) exhibits the working voltage of the ZFO-CeO₂ cell under a stationary current density of 110 mA/cm² lasting more than 120 h. The inset figure is an original image directly exported from the I-V data recording software after 72 h. The initial voltage drop is due to the activation process. It subsequently gradually reaches a stable state with a value of around 0.9 V in the following 110 h. Figure 5(f) shows a cross-sectional SEM of the cell after a durability test to assess the compatibility of the ZFO-CeO₂ electrolyte with other components, in which the NCAL-Ni cathode, ZFO-CeO₂ electrolyte, and NCAL-Ni anode can be distinguished. The electrolyte with a thickness of ~ 650 µm is gas-tight and well adhered to the nether porous electrode but shows partial delamination with the upper electrode, which is probably due to thermal incompatibility between the two materials or the squeezing damage when scissoring fuel cell pellet for SEM characterization. These results reflect the potential of ZFO-CeO₂ for practical PCFCs application, while compatibility of the electrode materials with ZFO-CeO₂ requires more consideration.

In comparison with the latest PCFCs literature related to high proton conductivity, the proposed ZFO-CeO₂ electrolyte and the corresponding fuel cells have revealed remarkably higher proton conductivity and lower area-specific ohmic resistance (ASR_o) ^{17, 39, 52, 53}, thus delivering an excellent and promising fuel cell performance of 1070 mW/cm² at 510 ^oC. According to above characterization and analysis, the interfacial disorder is one of the major factors resulting in such high proton conductivity⁵³. Generally, proton conduction in PCFC electrolytes is based on the Grotthuss mechanism and Vehicle mechanism. The Grotthuss mechanism allows H⁺ to diffuse through the hydrogen bond network of water molecules via the formation and cleavage of bonds. For instance, the H⁺ hops along O-O bond and rotates and locates at the site of lattice O to transport. The vehicle mechanism allows H⁺ to migrate bound to another element. A typical example is the formation of OH⁻ and an oxygen vacancy that occurs in perovskite structured oxides. As confirmed by the HR-TEM, XPS and DFT calculation results, the interfacial disorder in our ZFO-CeO₂ heterostructure is mainly caused by the dislocation or rearrangement of O atoms at the hetero-interface, which leads to the formation of oxygen defects in the lattice. Thus, it can be anticipated that the formation of oxygen defects in the ZFO-CeO₂ heterostructure would weaken the bonding of partial M-O and thus reduce the bond length of O-O, which is beneficial to the hopping of H⁺ along O-O bond in the Grotthuss mechanism. This has been studied in BaCeO₃ by K.D. Kreue that the resistance for proton diffusion can be reduced when the length of O-O bond is shortened from 312 pm to 281 pm ⁵⁴. W. Münch also confirmed that the hopping process of proton between O-O bond is a speed-controlled step of the Grotthuss mechanism, while the proton rotation and position process is always very rapid with low resistance ⁵⁵. In this regard, the interfacial disorder can promote the Grotthuss mechanism of proton conduction in the ZFO-CeO₂ heterostructure. Moreover, the formation of oxygen defects in the ZFO-CeO₂ lattice also has a positive impact on the Vehicle mechanism of protons, as these oxygen defects are mostly in the forms of oxygen vacancies, which would bring about higher oxygen ion conduction in the material and thus allow more H⁺ to migrate bound to the mobile oxygen ions. Normally, the Vehicle mechanism tends to take place at higher operating temperatures such as 700°C, so the possibility of such promoting effect is lower than the former one (Grotthuss mechanism) in our case as the operating temperature of PCFCs is 360–510°C. To clarify the above proton conduction mechanisms in ZFO-CeO₂, two schematic diagrams are proposed and illustrated in Fig. 6 (a, b).

In addition, the fast proton diffusion, high fuel cell performance, and good stability are believed to be linked with the heterojunction effect in the ZFO-CeO₂ sample and fuel cell device. The p-n bulk-heterojunction might establish in the heterostructure of p-type ZFO and n-type CeO₂ at hetero-interface, as illustrated in Fig. 6(c). To confirm the existence of junction, the diode rectification characteristic of ZFO-CeO₂ under bias voltage was measured and presented in Supplementary Fig. S5. The nonlinear behavior of the I-V curve represents the formation of junction and the corresponding BIEF in ZFO-CeO₂ electrolyte ⁵⁶. When ZFO and CeO₂ forming a compact contact at the interface, the high concentrated n-type CeO₂ accommodates with p-type ZFO, and electrons transfer toward the conduction band and holes move in the direction of the valance band according to ZFO-CeO₂ concentration. Redistribution of charges constitutes across the ZFO/CeO₂ interface, leading to a depletion layer with BIEF to generate the p-n heterojunction. From theoretical calculations, it can be found that ZFO (p-type) and CeO₂ (n-type) have different tendencies to bend towards the Fermi energy level. Therefore, conduction band edges and valance band edges can be induced during junction formation and adjust itself by creating a potential barrier. This potential barrier prevents the electrons passing through the junction, i.e., confined electrons locally without long-distance aviation. As a consequence, electronic conduction is suppressed in ZFO-CeO₂. Since the ZFO and CeO₂ forms homogeneous hetero-structure, the junctions are three-dimensionally distributed in the material at the grain/particle scale to form spatial p-n heterojunction. From a micro point of view, the BIEF of the junction at grain/particle scale can localize electrons at the hetero-interface to avoid current leakage of fuel cell and facilitate the transport of O²⁻ and H⁺ as another driving force in addition to the chemical concentration gradient. From a macro point of view, the fuel cell operational atmosphere could induce the whole ZFO-CeO₂ layer to be distributed in polarity due to the amphoteric characteristic of oxide semiconductors, where the material close to H₂ with low oxygen partial pressure presents n-type conduction and that exposed in O₂ presents p-type^{10, 57}. This polarity distribution would establish a macro p-n junction throughout the whole electrolyte. In this way, the junction effect of ZFO-CeO₂ can prevent the electron from passing through the electrolyte layer (from anode to cathode), while promoting the transport of O²⁻ and H⁺ as shown in Fig. 6(c). This hetero-junction effect could play an essential role in enabling the fast proton conduction in ZFO-CeO₂ heterostructure and avoiding the current leakage risk of the ZFO-CeO₂ PCFCs.

In summary, a new ZFO-CeO₂ heterostructure has been developed to enable high proton conduction. The microstructure, electronic structure, interface property, proton conductivity, and electrochemical performance of the heterostructure were studied. The ZFO-CeO₂ has achieved a high ionic conductivity of 0.21 S/cm and fuel cell power density of 1070 mW/cm² at 510°C and normal operation at 360°C with fairly good stability. As an important factor, the interfacial disorder in ZFO-CeO₂ heterostructure as verified by theoretical calculations was found to be one the major cause to create the oxygen defects in the heterostructure, which could promote the proton transport of ZFO-CeO₂ on the basis of Grotthuss mechanism and Vehicle mechanism. But the hetero-junction effect of p-type ZFO and n-type CeO₂ was further and discussed and highlighted as a potential mechanism to enhance the proton conductivity while eliminating the electronic short circuit problem. These findings provide a new semiconductor heterostructure material for developing high-performance PCFCs and point out a new insight into designing heterostructure to enable fast proton conduction at low temperatures.

Powder synthesis. Sol-gel auto-combustion approach was applied to synthesize ZnFe₂O₄ nanoparticles. The Stoichiometric amounts of Zn (NO₃)₂. 6H₂O and Fe (NO₃)₂⋅9H₂O were dissolved in the de-ionized (DI) water to make a homogenous solution. Subsequently, the solution was mixed and stirred on magnetic stirrer hot plate with 450 rpm rate for 3 hours. A specific amount of citric acid (C₆H₈O₇⋅H₂O) was added into the prepared homogenous solution as a chelating agent to avoid large crystal formation. The ammonia solution (NH₄OH) was also used to maintain the pH ~ 7. After that, sol-gel technique is used to convert the prepared solution into a dry gel based on xerogel method. Finally, the obtained gel was heated at 110°C temperature for combustion process using magnetic stirrer hot plate, and the resulting gray ash was dried into the drying oven for 24 hours and then calcined at 800°C for 4 hours in N₂ environment to obtain cubic nature spinel phase. The similar procedure has been followed to obtain single-phase (cubic) ceria oxide (CeO₂). For CeO₂, the calcination temperature was kept 750°C for 4 hours. For ZnFe₂O₄-CeO₂ heterostructure, the suitable amounts of both ZnFe₂O₄ and CeO₂ were added into motor and pestle and mixed for 3 hours with different ratios such as 2:8, 3:7 and 4:6, respectively. The chosen amounts were individually mixed and then ball milled (Fritsch Pulverisette 6, Germany) at a speed of 200 rpm for 20 hours.

Cell preparation. For electrodes preparation, the commercially purchased Ni_0.8Co_0.15Al_0.05LiO_2−δ (NCAL) was mixed with terpineol to make slurry gel and then painted on 13 mm diameter nickel foam (Ni-foam) to form a Ni-NCAL. In the next step, Ni-NCAL was put into drying oven for 40 min for dry synthesis process. For the fuel cell fabrication, ZFO-CeO₂ powders were used as an electrolyte and sandwiched between two electrodes with the configuration of Ni-NCAL/ZFO-CeO₂/Ni-NCAL under 250 MPa load into a steel mold. The fabricated cylindrical fuel cell device was sensitively cleaned from its sides to avoid short-circuiting issue. The constructed fuel cell device with the active area of 0.64 cm² and thickness of 1.0 mm was used for electrochemical and fuel cell measurements. A dc-load instrument (ITECH8511, ITECH Electrical Co., Ltd.) was employed to measure the fuel cell performance of fabricated device under H₂ and air environment. The H₂ flow rate was kept around 100–120 ml/min, while air flow rate was used around 150–180 ml/min for fuel cell device activation. The I-V (current density-voltage) and I-P (current density-power density) curves were recorded in the 510°C to 360°C temperature range. While electrochemical impedance spectroscopy (EIS) was measured at 510°C to 360°C under H₂/air conditions using 3000 Gamry Reference workstation in a 0.1Hz to 10⁵ Hz frequency range. ZSIMPWIN software was used to fit the model circuit with obtained EIS data. While, for proton conductivity, commercially purchased BZY was used as a filter layer with the configuration of Ni-NCAL/BZY/ZFO-CeO₂/BZY/Ni-NCAL under air, D₂, H₂ (5%) and H₂/air atmosphere at different elevated temperatures.

Characterizations. The phase identification of synthesized pure ZnFe₂O₄ and CeO₂ and their ZFO-CeO₂ heterostructure composites were determined by X-ray diffractometer (Germany, Bruker Corporation with Cu Kα radiation, λ = 1.5418 Å) with a 2θ scan range of 0–85\(^\circ\) and a scanning step of 0.02\(^\circ\)’. The Rietveld refinement of both ZnFe₂O₄ and CeO₂ was done using X\({\prime }\)Pert HighScore Plus software for single-phase verification. High resolution transmission electron microscopy (HR-TEM- Tecnai G2 F30 S-TWIN 300 kV/FEG) was employed to analyze the morphological analysis of synthesized nanoparticle and their heterostructure composites, while High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM and EDS) were applied to for surface analysis. X-ray photoelectron spectroscopy (XPS, Physical Electronics Quantum 2000, Al Kα X-ray source) is an important technique for chemical analysis of synthesized samples and XPS41 software was applied to analyze the XPS data. Additionally, UV-Vis and ultraviolet photoelectron spectroscopy (UPS) was used to deduce the valence band level under the unfiltered HeeI (21.22 eV) gas discharge lamp.

First principle calculations. The crystal unit cell for ZFO and CeO₂ samples are sketched using the refined parameters and quantum ATK VIRTUAL NANOLAB as well as VESTA software’s for further process. The refined parameters such as volume of the cell, metal ions (M-O) distances as well as Rwp, Rp, GOF values were taken from the literature. The first principle ZFO and CeO₂ calculations were done using the Atomistic Toolkit quantum (ATK) software package. Usually, local density approximations (LDA) can be employed for density functions theory (DFT), but we preferred Kohn-Sham Hamiltonian eigen functions, which were further expanded into the linear combination of atomic orbital (LCAO). The basis sets for CFT were chosen as pseudopotential SG15. Meta-GGA function was identified for exchange-correlation interaction, which was built-in by the function of TB09LDA in ATK software. The occupation method was applied according to Fermi-Dirac functions, while grid mesh cut-off was set as 75 Hartree. The k-point was set to be \(5\times 5\times 100\) with meta-GGA exchange-correlation potential. For ZFO-CeO₂ heterostructure, LBFGS optimizer method was set for geometry optimization with 0.0001 Hartree self-consistent field iteration. The lattice vectors, space group, and Bravais lattice type were freely relaxed during the geometry optimization.

Acknowledgement

This work was supported by the National Natural Science Foundation of China (NSFC) (Grant No. 51772080 and 22109022), the Fundamental Research Funds for the Central Universities (Grant No.3203002105A2, 2242021k30028), Jiangsu Fundamental Research Program No. JSSCRC2021491.

Author Contributions

Muhammad Yousaf conceived and designed the project under Zhu Bin supervision. Muhammad Yousaf conducted experiments such as synthesis, electrochemical testing, and fuel cell fabrication with the help of Muhammad Akbar, Naveed Mushtaq and Enyi Hu. Yuzheng Lu and M.A.K Yousaf shah successfully done DFT. Xia Chen checked whole manuscript grammatical mistakes. Finally, Yousaf Muhammad, Xia Chen, and Zhu Bin wrote the manuscript with all authors contributions.

Competing interests

The authors declare no competing interests.

Wachsman ED, Lee KT. Lowering the temperature of solid oxide fuel cells. Science 334, 935-939 (2011).
Fop S, et al. High oxide ion and proton conductivity in a disordered hexagonal perovskite. Nature Materials 19, 752-757 (2020).
Duan C, et al. Highly durable, coking and sulfur tolerant, fuel-flexible protonic ceramic fuel cells. Nature 557, 217-222 (2018).
Hara S, Miyayama M. Proton conductivity of superacidic sulfated zirconia. Solid State Ionics 168, 111-116 (2004).
Ding H, et al. Self-sustainable protonic ceramic electrochemical cells using a triple conducting electrode for hydrogen and power production. Nature Communications 11, 1907 (2020).
Duan C, et al. Readily processed protonic ceramic fuel cells with high performance at low temperatures. Science 349, 1321-1326 (2015).
Wang X, et al. Charge-transfer modeling and polarization DRT analysis of proton ceramics fuel cells based on mixed conductive electrolyte with the modified anode–electrolyte interface. ACS applied materials & interfaces 10, 35047-35059 (2018).
Norby T, Magrasó A. On the development of proton ceramic fuel cells based on Ca-doped LaNbO4 as electrolyte. Journal of Power Sources 282, 28-33 (2015).
Raza R, Zhu B, Rafique A, Naqvi MR, Lund P. Functional ceria-based nanocomposites for advanced low-temperature (300–600° C) solid oxide fuel cell: a comprehensive review. Materials Today Energy 15, 100373 (2020).
Zhu B, et al. A nanoscale perspective on solid oxide and semiconductor membrane fuel cells: Materials and technology. Energy Mater 1, 100002 (2021).
Zhou Y, et al. Strongly correlated perovskite fuel cells. Nature 534, 231-234 (2016).
Ma G, Wen Z, Han J, Zhang J, Wen Y. Enhanced proton conduction of BaZr0. 9Y0. 1O3-δ by hybrid doping of ZnO and Na3PO4. Solid State Ionics 281, 6-11 (2015).
Xing Y, et al. Proton shuttles in CeO2/CeO2− δ core–shell structure. ACS Energy Letters 4, 2601-2607 (2019).
Wang F, et al. Surface‐Engineered Homostructure for Enhancing Proton Transport. Small methods 6, 2100901 (2022).
Wang F, et al. Tuning La 2 O 3 to high ionic conductivity by Ni-doping. Chemical Communications 58, 4360-4363 (2022).
Liu Y, Fan L, Cai Y, Zhang W, Wang B, Zhu B. Superionic conductivity of Sm3+, Pr3+, and Nd3+ triple-doped ceria through bulk and surface two-step doping approach. ACS applied materials & interfaces 9, 23614-23623 (2017).
Choi S, et al. Exceptional power density and stability at intermediate temperatures in protonic ceramic fuel cells. Nature Energy 3, 202-210 (2018).
Lund TL. Nano-scale view into solid oxide fuel cell and semiconductor 3 membrane fuel cell: material and technology 4. Perspective 1, 2 (2021).
Lu Y, Zhu B, Shi J, Yun S. Advanced low-temperature solid oxide fuel cells based on a built-in electric field. Energy Materials 1, 100007 (2021).
Sultana S, Rafiuddin. Electrical conductivity in TlI–TiO2 composite solid electrolyte. Physica B: Condensed Matter 404, 36-40 (2009).
Pennycook TJ, Beck MJ, Varga K, Varela M, Pennycook SJ, Pantelides ST. Origin of Colossal Ionic Conductivity in Oxide Multilayers: Interface Induced Sublattice Disorder. Physical Review Letters 104, 115901 (2010).
Garcia-Barriocanal J, et al. Colossal ionic conductivity at interfaces of epitaxial ZrO2: Y2O3/SrTiO3 heterostructures. Science 321, 676-680 (2008).
Zhu B, et al. Charge separation and transport in La0. 6Sr0. 4Co0. 2Fe0. 8O3-δ and ion-doping ceria heterostructure material for new generation fuel cell. Nano Energy 37, 195-202 (2017).
Wang B, et al. Preparation and characterization of Sm and Ca co-doped ceria–La 0.6 Sr 0.4 Co 0.2 Fe 0.8 O 3− δ semiconductor–ionic composites for electrolyte-layer-free fuel cells. Journal of Materials Chemistry A 4, 15426-15436 (2016).
Gao Z, et al. Electrostatic self-assembly synthesis of ZnFe2O4 quantum dots (ZnFe2O4@C) and electromagnetic microwave absorption. Composites Part B: Engineering 179, 107417 (2019).
Li Y, et al. Structural disorder controlled oxygen vacancy and photocatalytic activity of spinel-type minerals: A case study of ZnFe2O4. Chemical Geology 504, 276-287 (2019).
Yao L, et al. Green synthesis of mesoporous ZnFe2O4/C composite microspheres as superior anode materials for lithium-ion batteries. Journal of Power Sources 258, 305-313 (2014).
Wang B, et al. Fast ionic conduction in semiconductor CeO2-δ electrolyte fuel cells. NPG Asia Materials 11, 1-12 (2019).
Irvine JT, Neagu D, Verbraeken MC, Chatzichristodoulou C, Graves C, Mogensen MB. Evolution of the electrochemical interface in high-temperature fuel cells and electrolysers. Nature Energy 1, 1-13 (2016).
Jesus C, Mendonça E, Silva L, Folly W, Meneses C, Duque J. Weak ferromagnetic component on the bulk ZnFe2O4 compound. Journal of magnetism and magnetic materials 350, 47-49 (2014).
Boumaza S, Boudjemaa A, Bouguelia A, Bouarab R, Trari M. Visible light induced hydrogen evolution on new hetero-system ZnFe2O4/SrTiO3. Applied Energy 87, 2230-2236 (2010).
Gerward L, Olsen JS, Petit L, Vaitheeswaran G, Kanchana V, Svane A. Bulk modulus of CeO2 and PrO2—An experimental and theoretical study. Journal of Alloys and Compounds 400, 56-61 (2005).
Lee JG, Park JH, Shul YG. Tailoring gadolinium-doped ceria-based solid oxide fuel cells to achieve 2 W cm− 2 at 550 C. Nature communications 5, 1-10 (2014).
Ulpe AC, Bredow T. GW‐BSE Calculations of Electronic Band Gap and Optical Spectrum of ZnFe2O4: Effect of Cation Distribution and Spin Configuration. ChemPhysChem 21, 546 (2020).
Jankov S, Armaković S, Tóth E, Srdic V, Cvejic Z, Skuban S. Electronic structure of yttrium-doped zinc ferrite–Insights from experiment and theory. Journal of Alloys and Compounds 842, 155704 (2020).
Han R, Qi M, Mao Z, Lin X, Wu P. The electronic structure, magnetic and optical properties of B-doped CeO2 (111) surface by first-principles. Physics Letters A 384, 126526 (2020).
Shi Z, Liu S, Zhou Y, Xing X, Ren X, Yang Q. Structure and properties of YAlO3/NbC heterogeneous nucleation interface: First principles calculation and experimental research. Journal of Alloys and Compounds 773, 264-276 (2019).
Mushtaq N, et al. Promoted electrocatalytic activity and ionic transport simultaneously in dual functional Ba0. 5Sr0. 5Fe0. 8Sb0. 2O3-δ-Sm0. 2Ce0. 8O2-δ heterostructure. Applied Catalysis B: Environmental 298, 120503 (2021).
Xia C, Mi Y, Wang B, Lin B, Chen G, Zhu B. Shaping triple-conducting semiconductor BaCo0. 4Fe0. 4Zr0. 1Y0. 1O3-δ into an electrolyte for low-temperature solid oxide fuel cells. Nature communications 10, 1-9 (2019).
Zhu B, et al. A new energy conversion technology based on nano-redox and nano-device processes. Nano Energy 2, 1179-1185 (2013).
Zhu B, et al. Schottky Junction Effect on High Performance Fuel Cells Based on Nanocomposite Materials. Advanced Energy Materials 5, 1401895 (2015).
Fan L, Su P-C. Layer-structured LiNi0. 8Co0. 2O2: a new triple (H+/O2−/e−) conducting cathode for low temperature proton conducting solid oxide fuel cells. Journal of Power Sources 306, 369-377 (2016).
Zhang W, et al. La0. 6Sr0. 4Co0. 2Fe0. 8O3− δ/CeO2 heterostructured composite nanofibers as a highly active and robust cathode catalyst for solid oxide fuel cells. ACS applied materials & interfaces 11, 26830-26841 (2019).
Cai Y, et al. A Bulk-Heterostructure Nanocomposite Electrolyte of Ce 0.8 Sm 0.2 O 2-δ–SrTiO 3 for Low-Temperature Solid Oxide Fuel Cells. Nano-micro letters 13, 1-14 (2021).
Shim JH, et al. Enhanced oxygen exchange and incorporation at surface grain boundaries on an oxide ion conductor. Acta Materialia 60, 1-7 (2012).
Zhu B. Solid oxide fuel cell (SOFC) technical challenges and solutions from nano‐aspects. International Journal of Energy Research 33, 1126-1137 (2009).
Chen G, et al. Advanced fuel cell based on perovskite La–SrTiO3 semiconductor as the electrolyte with superoxide-ion conduction. ACS applied materials & interfaces 10, 33179-33186 (2018).
Sun W, Shi Z, Liu M, Bi L, Liu W. An easily sintered, chemically stable, barium zirconate‐based proton conductor for high‐performance proton‐conducting solid oxide fuel cells. Advanced Functional Materials 24, 5695-5702 (2014).
Bi L, Fabbri E, Sun Z, Traversa E. A novel ionic diffusion strategy to fabricate high-performance anode-supported solid oxide fuel cells (SOFCs) with proton-conducting Y-doped BaZrO 3 films. Energy & Environmental Science 4, 409-412 (2011).
Lan R, Tao S. Novel proton conductors in the layered oxide material LixlAl0. 5Co0. 5O2. Advanced Energy Materials 4, 1301683 (2014).
Wu Y, et al. Proton conduction and fuel cell using the CuFe-oxide mineral composite based on CuFeO2 structure. ACS Applied Energy Materials 1, 580-588 (2018).
An H, et al. A 5× 5 cm2 protonic ceramic fuel cell with a power density of 1.3 W cm–2 at 600° C. Nature Energy 3, 870-875 (2018).
Ni M, Shao Z. Fuel cells that operate at 300 to 500 C. Science 369, 138-139 (2020).
Kreuer K. Aspects of the formation and mobility of protonic charge carriers and the stability of perovskite-type oxides. Solid State Ionics 125, 285-302 (1999).
Münch W, Seifert G, Kreuer K, Maier J. A quantum molecular dynamics study of proton conduction phenomena in BaCeO3. Solid State Ionics 86, 647-652 (1996).
Jiang Z, Wang J, Hu E, Liu J, Yu L, Wang F. Accelerated Proton Transport Based on a p-i-n Heterostructure Membrane for Low-Temperature Solid Oxide Fuel Cells. ACS Applied Energy Materials 4, 13963-13973 (2021).
Singh K, Nowotny J, Thangadurai V. Amphoteric oxide semiconductors for energy conversion devices: a tutorial review. Chemical Society Reviews 42, 1961-1972 (2013).

There is NO Competing Interest.

Supplementarydata.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Interfacial Disordering and Heterojunction Enabling Fast Proton Conduction

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1