Volume 7 Number 45 7 December 2019 Pages 25675–26188 Journal of Materials Chemistry A Materials for energy and sustainability rsc.li/materials-a ISSN 2050-7488 COMMUNICATION Albert Tarancón et al. Thin film oxide-ion conducting electrolyte for near room temperature applications Journal of Materials Chemistry A This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online COMMUNICATION View Journal | View Issue Thin film oxide-ion conducting electrolyte for near room temperature applications† Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. Cite this: J. Mater. Chem. A, 2019, 7, 25772 Iñigo Garbayo, ‡a Francesco Chiabrera, a Nerea Alayo, a José Santiso, b Received 15th July 2019 Alex Morata a and Albert Tarancón *ac Accepted 9th September 2019 DOI: 10.1039/c9ta07632h rsc.li/materials-a Stabilized bismuth vanadate thin films are presented here as superior applications,8–10 micro sensors for integrated sensing oxide ionic conductors, for application in solid state electrochemical networks,11,12 or resistive switching devices for data devices operating near room temperature. Widely studied in the 90s in processing.13–15 bulk form due to their unbeatable ionic conduction, this family of In the case of TF-SOFCs, the use of hundreds of nanometres materials was finally discarded due to poor stability above 500  C. thick YSZ electrolytes set a new minimum operating tempera- Here, we however unveil the possibility of using Bi4V1.8Cu0.2O10.7 at ture at ca. 450  C.16–18 When compared to the typical bulk SOFC reduced temperatures in thin film-based devices, where the material working temperatures, such a reduction in temperature keeps its unmatched conduction properties and at the same time undoubtedly meant a signicant technological breakthrough, shows good stability over a wide oxygen partial pressure range. attaining quick and low power consuming start-up processes. However, the new temperature range still entailed important restrictions to the cell performance of TF-SOFCs. First, an The quest for high oxygen ion ceramic conductors has been operating temperature of 450  C was incompatible with state- a very active eld of research for decades.1 Aiming to reduce the of-the-art SOFC ceramic electrodes,19 limiting the portfolio of operating temperature of different oxide ion-driven solid state possible electrode materials to inherently instable porous devices such as solid oxide fuel cells (SOFC) or oxygen sensors, it metallic lms.8,17,18,20,21 Also, the integration of TF-SOFCs with became necessary to look for alternatives to the state-of-the-art additional components of the nal portable powering device yttria-stabilized zirconia (YSZ) electrolyte material.2–5 The slug- (reformers, post-combustors, or simply electronics)22,23 required gish ion hopping of YSZ typically limited the device operation of a full integration of the system microfabrication in CMOS- bulk systems to e.g. 800  C in the case of SOFCs,1 or 450  C for compatible Si technology, typically restricted to maximum oxygen sensors.6 With the standardization of thin lm deposi- peak temperatures of 450  C.24 This limited the broad imple- tion techniques in the last decades, a signicant lowering of the mentation of TF-SOFC (and, for similar reasons, also gas device operating temperature could be attained, simply derived microsensors) from both a materials and a microfabrication from a substantial reduction of the electrolyte thickness point of view. Further lowering of the operating temperature is (around 3 orders of magnitude, down to the sub-micrometre therefore still required, for which we are entailed to look for range).7 The use of oxide thin lms gave rise to new families alternative high oxygen conducting electrolyte materials able to of integrated solid state MicroElectroMechanical System operate at reduced temperatures, eventually room temperature. (MEMS) devices operating at intermediate temperatures, such Low temperature operating oxide ionic conductors nd as thin lm SOFCs (TF-SOFCs) for portable power important applications as well in the emerging data storage devices based on transition metal oxide electronic systems, via a Department of Advanced Materials for Energy Applications, Catalonia Institute for resistive switching,15 electronic phase transitions25 or voltage Energy Research (IREC), Jardins de les Dones de Negre 1, 08930 Sant Adrià del controlled ion gating.26 In this sense, it is of particular interest Besòs, Barcelona, Spain. E-mail: atarancon@irec.cat b to this work a recent pioneering paper from Yildiz and co- Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC, Barcelona Institute of Science and Technology, Campus UAB, Bellaterra, Barcelona, 08193, Spain workers, who showed that electrochemically driven phase c ICREA, Passeig Lluı́s Companys 23, 08010, Barcelona, Spain changes in oxide materials (in particular, VxOy) can be used to † Electronic supplementary information (ESI) available: Supplementary images trigger metal-insulator transitions and, thus, force large S1–S8 distributed in ve supplementary sections (S1 to S5). See DOI: changes in resistivity that could be used for data storage.27 In 10.1039/c9ta07632h such system, an ionic conductor is used as gating material, ‡ Present address: CIC Energigune, Parque Tecnológico de Álava, 01510 pumping oxygen to (or pulling it out from) the oxide working Vitoria-Gasteiz, Spain. 25772 | J. Mater. Chem. A, 2019, 7, 25772–25778 This journal is © The Royal Society of Chemistry 2019 View Article Online Communication Journal of Materials Chemistry A electrode, forcing the phase change. However, high oxygen doping with different transition metals has been proved to mobility needed in the gate limited the operation with solid successfully stabilize the g-phase down to room temperature. state materials (YSZ) to a minimum temperature of 300  C, Among the different options, BICUVOX (which stands for the while room temperature operation was only achieved using corresponding BIMEVOX doped with Cu and formula Bi4V1.8- This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. ionic liquid gates. Solid state gate materials with superior oxide Cu0.2O10.7, for a typical 10 mol% doping) was shown to present ionic conduction can therefore open interesting opportunities the highest ionic conductivity.28 However, a non-trivial elec- for the fabrication of electrochemically driven, room tempera- tronic conductivity already at intermediate temperatures (T  ture operating compact all-solid state memories. 500  C),30 irreversible phase changes under reducing atmo- Clearly, the number of alternatives to YSZ is very limited, and spheres31 and a high thermal expansion coefficient32 drastically all come at a cost. Materials with higher ionic conductivity than hindered the broad application of this material (and the rest of Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. YSZ at intermediate temperatures usually suffer from fast BIMEVOXes) in nal pellet-based SOFCs, operating in the degradation under reducing conditions, see e.g. ceria-based intermediate-temperature range. Consequence of this, the electrolytes at T > 600  C.1 Among the explored alternatives, it material was practically abandoned by the end of that decade, is to be highlighted the BIMEVOX family, as the electrolyte except for its use in oxygen separation membranes.33,34 material with the highest ionic conductivity at intermediate However, the possibility of fabricating thin lm devices, in temperatures ever reported. Firstly described in 1990 by which the operating temperature can be signicantly lowered, Abraham et al.,28 BIMEVOX presents a layered aurivillius opens the door to explore such an outstanding material in a new structure consisting of alternating {Bi2O2}2+ and transition low temperature range (T  500  C), where the chemical metal-doped perovskite {VO3.5,0.5}2 layers, see Fig. 1a. The instability is expected to be signicantly hindered. Despite the high presence of disordered oxygen vacancies in the perovskite promising possibilities for low temperature applications, three sheets is the responsible for the high oxygen mobility, which is decades aer the discovery of the high ionic conductivity of maximized when stabilizing the tetragonal g-phase. In the BIMEVOX only a few works have dealt with the thin lm undoped parent compound Bi4V2O11, temperatures below deposition of this material (thickness <500 nm and low 550  C lead to the formation of additional monoclinic a and temperature processing),35,36 and a study of its electrochemical orthorhombic b phases with lower conductivity;29 however, performance at this temperature range is still lacking. In this Fig. 1 (a) Atomic representation of the g-Bi4V2O11 aurivillius layered structure (drawings produced by VESTA 3 software).50 (b–g) Top view SEM images of representative films deposited by PLD at different background pressures (p ¼ 0.0067–0.267 mbar) and a constant target-to-substrate distance of 11 cm, on top of Si substrates. (h) Bi/V ratio as a function of deposition pressure. Dashed line is included as guide for the eyes. Four different pressure regions are identified (A–D) (i) X-ray diffractogram obtained for the same BICUVOX film series. Main diffractions of the g- Bi4V2O11 phase are indicated (ICDD reference code 00-048-0273), as well as the appearance of secondary phases (Bi2O3 *, BiVO4 †) (j) Raman spectra of the BICUVOX film series, plus references for g-Bi4V1.8Cu0.2O10.7 (measured on the pellet used as PLD target),42 a-Bi4V2O11 (ref. 43) and BiVO4.41 This journal is © The Royal Society of Chemistry 2019 J. Mater. Chem. A, 2019, 7, 25772–25778 | 25773 View Article Online Journal of Materials Chemistry A Communication work, we optimized the deposition of BICUVOX thin lms by stoichiometry is recovered, and thus in the lms (region D). The means of Pulsed Laser Deposition (PLD) and we studied its presence of a Bi-decient region at the lowest deposition pres- electrochemical performance at T ¼ 50–450  C. First, we extend sure (region A) is probably related to the high volatility of Bi, previous works and show the tremendous capabilities of PLD which would provoke that at low background pressures, where This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. for tuning the stoichiometry in this type of lms, simply no collisions are expected in the plume, Bi simply ies away.40 changing the deposition conditions. Then, we present the In our case, two pressure regions are clearly identied as conduction properties of the optimized BICUVOX lms, by preferred for obtaining stoichiometric BICUVOX lms, regions measuring its oxygen conductivity as a function of temperature. B (p ¼ 0.0267 mbar) and D (p > 0.187 mbar). Noting the layered crystal structure of BICUVOX with oxide Importantly, no clear trend could be detected for Cu with the ionic diffusion taking place mainly along the {VO3.5,0.5}2 background pressure. However, important differences were Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. layers, we investigated the anisotropy in the conduction by detected between the different deposition pressure regions. preparing epitaxial lms and measuring their conductivity in First, in the low pressure regimes A and B, p ¼ 0.0067–0.0267 orthogonal directions (parallel and perpendicular to the mbar, we could identify the presence of homogeneously perovskite sheets). distributed Cu-rich micrometre-size particles (probably CuO), The optimization of PLD deposition of BICUVOX thin lms see top-view SEM image in Fig. S3.† At the same time, in this was carried out by preparing a set of polycrystalline lms range it was impossible to detect Cu out of these particles (i.e. in (thickness ranging from 100 to 300 nm) under varying depos- the lm), suggesting that Cu could not be successfully inte- iting conditions, viz. background oxygen pressure (p) and target- grated in the material and that the lms deposited at low p are to-substrate distance (d). Other parameters such as deposition composed by undoped Bi4V2O11. Opposed to that, for p > 0.0267 temperature and laser uence were previously optimized for mbar (regions C and D), Cu was detected all along the lms in obtaining homogeneous crystalline lms and then intentionally the right proportion and no particles were observed. This set constant, at 600  C and 0.45 J cm2 respectively. Thin lms suggests that in these regimes Cu is indeed successfully incor- were grown from a home-made Bi4V1.8Cu0.2O10.7 target, p was porated in the BICUVOX structure of the lms. varied from 0.0067 to 0.267 mbar and two target-to-substrate X-Ray Diffraction (XRD) and Raman spectroscopy were distances were studied, d ¼ 9 and 11 cm (note the longer carried out in order to analyse the phase evolution through the distances typically used in our large area PLD system compared lm series, and the correlation between Bi/V ratio and phase to “standard” research-scale PLD systems; more details in the stabilization (Fig. 1i and j). Indeed, the appearance of expect- Experimental section).37 Dense lms were obtained within the able secondary phases is clearly observed in the lms with whole experimental window of pressures and distances, with stoichiometry distant from the targeted one. Analysing the XRD considerable variations in the grain size (see representative top pattern in Fig. 1i, one can observe: (i) the appearance of a Bi- view Scanning Electron Microscopy (SEM) images in Fig. 1b–g, decient BiVO4 phase at low deposition pressures (p < 0.0267 for selected background oxygen pressures and d ¼ 11 cm; cor- mbar, region A), probably a mix of clinobisvanite and dreyerite responding images for d ¼ 90 mm can be found in the ESI, phases,41 and (ii) the presence of Bi2O3 in the V-decient region Fig. S1†). A stoichiometric analysis of the deposited lms was C, 0.0267 < p < 0.187 mbar (peaks marked as *). Moreover, even then carried out by Energy-Dispersive X-ray spectroscopy (EDX), though the desired stoichiometry was measured in lms Fig. 1h and S2† (specic lms were also measured by Wave- deposited in region B, XRD revealed the presence of additional length Dispersive Spectroscopy, WDS, conrming the trends). reections (see e.g. the triplet peak at 2q z 28 ), presumably We observed an evolution of the cationic ratio Bi/V as a function due to the aforementioned non-incorporation of Cu in the of the pressure in analogy with similar representations in the aurivillius structure and a subsequent non-stabilization of the literature involving depositions of heavy vs. light elements (H/L) tetragonal g-phase. Contrary, at the highest deposition pres- by PLD.38 We differentiate four different regions; a rst region A sures, p > 0.187 mbar, single g-Bi4V1.8Cu0.2O10.7 phase could for low p below 0.0267 mbar, where Bi decient lms are ob- only be detected. tained; a region B (p z 0.0267 mbar), where a good stoichi- Similar conclusions can be drawn by Raman spectroscopy, ometry is maintained; a region of intermediate pressures C Fig. 1j. In fact, it is to be highlighted the goodness of Raman (0.0267 < p < 0.187 mbar) in which an important V deciency is spectroscopy for identifying phase formation in the Bi4V2O11 observed; and a nal region D at higher pressures (p > 0.187 family, since a clear shi in wavenumber, from 860 cm1 to mbar), where the lm approaches again the target stoichiom- 830 cm1, happens when moving from the tetragonal g-phase etry. The observed trend matches well with the expected plume to monoclinic a-phase (see difference between the g-Bi4V1.8- dynamics as a function of the background pressure.38,39 In short, Cu0.2O10.7, measured in the pellet used as PLD target and at lower background pressures weak scattering is expected in perfectly matching previous reports,42 and the a-Bi4V2O11 from the laser plume for all the species independently on their size, ref. 43). This way, we could clearly identify the presence of the so the stoichiometry can be well maintained. With increasing desired tetragonal phase in the lms deposited at high pres- pressure, scattering is expected to affect greater to the lighter sure, region D (p > 0.187 mbar). Meanwhile, a-Bi4V2O11 was species (V in our case), giving rise to off-stoichiometric lms detected in the intermediate pressures, and BiVO4 was (light element-decient). If we keep increasing the background presumably measured at the lowest pressure tested (note here pressure, heavy elements (Bi) start to scatter as well and the that similar spectra have been reported for both a-Bi4V2O11 and plume ends up in a diffusion front in which the correct BiVO4, making impossible to differentiate them by this 25774 | J. Mater. Chem. A, 2019, 7, 25772–25778 This journal is © The Royal Society of Chemistry 2019 View Article Online Communication Journal of Materials Chemistry A technique). Surprisingly, at p z 0.0267 mbar Raman spectros- plane matching thanks to the very similar in-plane lattice copy also shows the stabilization of the g-phase, even though parameters. Cu is apparently not well incorporated in the structure. Still, we Furthermore, the quality of the lms was analysed by top- can conclude that XRD and Raman results fully support, from view non-contact AFM images, Fig. 2d and e. The images This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. a structural point of view, the stoichiometric evolution previ- showed a clear formation of terraces, typical of layer by layer ously observed, and suggest the need of high background growth,35 and a complete absence of signicant defects or pressures of p > 0.187 mbar for the stabilization of the auri- particulates in the top surface. The height prole obtained villius BICUVOX g-phase. along the top surface (black dashed arrow in Fig. 2e) revealed Based on this, we additionally prepared epitaxial BICUVOX very small steps of 3.5 Å, see Fig. 2f. These steps are clearly lms grown on (001)-oriented SrTiO3 substrates (undoped – smaller than the out-of-plane lattice parameter (cBICUVOX ¼ Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. STO and doped with 1.0 at% Nb – Nb:STO), at the optimum 15.44 Å, as calculated from the out-of-plane XRD pattern and deposition pressure of p ¼ 0.267 mbar. Fig. 2a shows a repre- matching previous literature).28 We speculate that these steps sentative XRD pattern of a 36 nm-thick BICUVOX lm deposited correspond to single {Bi2O2}2+ or {VO3.5,0.5}2 sub-layers, 4 on Nb:STO (see section S2 of the ESI† for the corresponding every unit cell, thus matching the obtained average value. In measurements on undoped STO). The appearance of a set of conclusion, we were able to grow high quality epitaxial BICU- reections corresponding to the (00L) family of planes reveals VOX lms in two functional substrates, one dielectric (STO), one a perfect epitaxial growth of the tetragonal aurivillius g-phase, conductive (Nb:STO), ideal for further electrochemical analysis. (00L)-oriented in the c-axis (more details in section S2†). The 4- Ionic conductivity in the whole set of BICUVOX lms was scan presented in the inset of Fig. 2a conrms a cube-on-cube investigated by means of Electrochemical Impedance Analysis growth of the tetragonal BICUVOX on top of Nb:STO with (EIS). Epitaxial (on STO-based substrates) and polycrystalline perfect match of the (100) planes of both materials, see sche- (on Si3N4/SiO2/Si substrates) lms were measured in a temper- matics in Fig. 2c. Note the almost identical in-plane lattice ature range from 50 to 450  C, under ambient conditions (cor- parameter of BICUVOX, aBICUVOX ¼ 3.907 Å,28 as compared to responding Nyquist plots can be found in the ESI, Fig. S6†). the cubic STO, aSTO ¼ 3.905 Å, meaning a lattice mismatch of First, the effect of lm off-stoichiometry and presence of only +0.05%. The epitaxy was conrmed by performing high secondary phases (see Fig. 1) on the conduction properties of resolution XRD reciprocal space maps (RSM), Fig. 2b and S4.† the polycrystalline lms was conrmed, correlating the stabili- The gure shows the coexisting reections (103) of the zation of the fast conducting and stable g-phase with the Nb:STO and (1011) of the BICUVOX, revealing a perfect in- obtaining of highest (and stable) conductivity (see section S3 in Fig. 2 (a and b) X-ray diffraction patterns of a 36 nm thick BICUVOX film deposited on Nb-doped SrTiO3. (a) Out-of-plane q–2q X-ray diffraction pattern, where the set of peaks corresponding to the 00L reflections is visible. A 4 scan around the (103) reflection of SrTiO3 is also included in the inset, showing a 4-fold symmetry corresponding to a single in-plane orientation (100) BICUVOX/(100) SrTiO3. (b) Shows the in-plane Reciprocal Space Map around the 103 reflection peak of the Nb-doped SrTiO3 substrate. (c) Schematic representations of the epitaxial growth of g-Bi4V2O11 on top of 001-oriented SrTiO3 single crystals, top and cross-sectional views. (d and e) AFM images of an epitaxial BICUVOX film grown on top of Nb-doped SrTiO3, with two different magnifications. (f) Height profile corresponding to the black dashed arrow shown in figure (e). The average step height is shown as well (only single step terraces are considered for the average calculation). This journal is © The Royal Society of Chemistry 2019 J. Mater. Chem. A, 2019, 7, 25772–25778 | 25775 View Article Online Journal of Materials Chemistry A Communication the ESI†). We then focus on the analysis of the polycrystalline vs. conduction vs. state-of-the-art YSZ and CGO, especially when epitaxial lms, all deposited at the optimum pressure range, p ¼ reducing temperature (due to the lower activation energies, Ea). 0.267 mbar. Fig. 3a presents the Arrhenius representation of the The highest conductivity is clearly obtained in the epitaxial conductivity evolution with temperature, for one polycrystalline BICUVOX lm when measuring in-plane, followed by the poly- This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. lm and two epitaxial BICUVOX lms measured at two different crystalline lm and, nally, the epitaxial measured cross-plane. congurations, viz. in-plane (on STO) and cross-plane (on As for the activation energies, one single mechanism could be Nb:STO), using top Au electrodes (further details on the observed for the whole range of measuring temperatures, with different measuring congurations can be found in section slightly higher Ea for the cross-plane than for the in-plane S3†). The obtained results are compared with reference bulk conduction. polycrystalline BICUVOX, as well as single crystal BICUVOX in The measured in-plane conductivity for the epitaxial lm Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. two directions (parallel and perpendicular to the (100)BICUVOX goes in line with literature values for bulk BICUVOX, conrming planes) and, nally, YSZ and Gd-doped CeO2 (CGO) thin lm the outstanding oxide ionic conductivity of this material in thin conductivity.16,47–49 lm form (the slightly higher Ea values could be related to the Although signicant differences are found among the plausible presence of structural defects in the lms or small off- different lms, all present an improvement in terms of ionic stoichiometries). The expected anisotropy in the conduction, depending on the oxygen hopping direction, is also clearly observed in the measured lms, with more than two orders of magnitude higher in-plane conductivity than cross-plane. Similar anisotropy was previously reported in single crystal BICUVOX and matches well with our XRD results, Fig. 2, since enhanced oxygen mobility in BICUVOX is known to happen along the {VO3.5,0.5}2 sub-layers, i.e. in the direction parallel to the substrate surface (in-plane), and hindered when crossing the {Bi2O2}2+ sub-layers, i.e. in the (001)BICUVOX direction, across the lm.48,50 Interestingly, the polycrystalline lm presents both an intermediate conductivity and activation energy, probably derived from the random orientation of the grains and the presence of grain boundaries. Notably, these results position the prepared epitaxial BICU- VOX lms as the thin lm material with highest (in-plane) ionic conductivity ever reported for low temperature applications (T < 300  C), with important implications for their application in advanced thin lm-based devices. However, apart from the outstanding oxide ionic conductivity, low temperature opera- tion with BICUVOX lms also requires high stability in a wide range of pO2, from oxidizing to highly reducing atmospheres (especially for applications in powering devices). Indeed, as commented before, this has been the main drawback of BICU- VOX hindering its use as electrolyte material at higher temperatures in the past. We therefore studied the stability vs. pO2 of the utmost performing BICUVOX epitaxial lm in its new operating temperature range, T < 300  C. Fig. 3b presents the in- plane conductivity measured by EIS at different temperatures, from 150  C to 350  C, and oxygen partial pressures, from pO2 ¼ 0.21 bar (synthetic air) to pO2 ¼ 1019 bar (5 vol% H2 in Ar). For temperatures below 300  C, the gure shows great stability (at Fig. 3 (a) Arrhenius representation of the evolution of ionic conduc- proles) within the whole range of pO2 (from oxidizing to highly tivity for different BICUVOX (polycrystalline, epitaxial) films with reducing conditions), proving the effectiveness of the material temperature, all deposited at p ¼ 0.267 mbar. References for bulk to be used in thin lm-based low temperature operating polycrystalline BICUVOX,28 single crystal BICUVOX (parallel and perpendicular to the (100)BICUVOX planes),48 YSZ thin film16,46 and CGO devices. The maximum operating temperature for epitaxial thin film44,45 are included for comparison. Inset schematics show the BICUVOX is set in this study at T ¼ 350  C, since a clear effect on direction of oxygen mobility in the two measuring configurations, for the conductivity is observed there when reducing pO2. It is the epitaxial films. Activation energies (Ea) are also indicated for the important to note, though, that the observed trend at this measured films. (b) pO2 dependence of BICUVOX in-plane conduc- temperature matches well with an unexpected p-type conduc- tivity, measured in an epitaxial film at different temperatures (T ¼ 150– 350  C) and in a polycrystalline sample at T ¼ 250  C. Oxygen partial tion behaviour. The appearance of this conduction comport- pressure is measured by an oxygen sensor at the outlet of the gases. ment at this temperature requires attention and will be a matter Dashed lines are included as guide for the eyes. of discussion in further works. 25776 | J. Mater. Chem. A, 2019, 7, 25772–25778 This journal is © The Royal Society of Chemistry 2019 View Article Online Communication Journal of Materials Chemistry A Finally, we also analysed the stability vs. pO2 of the BICUVOX Thin lm structural and compositional characterization polycrystalline lms. Although not as outstanding as the in-plane Long range crystalline structure of the deposited lms was conductivity measured in the epitaxial lms (between one and characterized by X-ray Diffraction (XRD, Bruker D8 Advanced), two orders of magnitude lower), the polycrystalline BICUVOX under a typical q/2q Bragg–Brentano geometry. For the epitaxial This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. already represents an important reduction of 150  C and lms, a four-angle diffractometer (X'Pert MRD-Panalytical) was 100  C vs. YSZ and CGO respectively, with a potential operating used with a Cu Ka radiation source. Additionally, the near range temperature below 300  C. As shown in Fig. 3b stability tests order was analysed by Raman spectroscopy (Horiba Jobin Ybon (open symbols), negligible conductivity changes vs. pO2 were iHR320). For the Raman measurements, green laser (l ¼ 532 detected also in the polycrystalline lms at T ¼ 250  C, thus nm) was utilized under no light polarization. Thin lms proving the viability of the material in the low temperature microstructure was studied by Scanning Electron Microscopy Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. operating range. Taking into account that for certain applications top view imaging (SEM, ZEISS AURIGA). SEM-coupled Energy- this microstructure might be signicantly easier to implement Dispersive X-ray spectroscopy (EDX) was used for the than an epitaxial lm (e.g. in the already existing TF-SOFC elemental distribution analysis. Errors in the EDX analysis were systems), polycrystalline BICUVOX are expected to play a crucial estimated by calculating the standard deviation from multiple role in future low temperature devices too. measurements in several samples. Non-contact Atomic Force Microscopy (AFM, XE 100 Park System Corp.) was additionally Conclusions used for top view imaging of the epitaxial BICUVOX lms. Epitaxial and polycrystalline BICUVOX thin lms are presented Thin lm electrochemical measurements here as superior oxygen conducting electrolyte materials, for their application in thin lm-based low temperature operating elec- Electrochemical Impedance Spectroscopy (EIS) was used for trochemical solid state devices. The highest conductivity was measuring the electrical properties of the BICUVOX lms, with observed in the epitaxial BICUVOX lms when measuring parallel the help of a Novocontrol Alpha-A Impedance analyser. 100 nm- to the {VO3.5,0.5}2 planes, matching that previously reported for thick top Au blocking interdigitated (for in-plane measure- bulk BICUVOX and thus surpassing any other thin lm oxide ion ments) and circular (for cross-plane measurements) electrodes conducting material ever reported at the low temperature oper- with varying sizes were rst dened by photolithography and ating range (T < 300  C). At the same time, polycrystalline BICU- then deposited by e-beam evaporation (OERLIKON UNIVEX VOX lms also showed an outstanding performance, with 450B). Impedance was measured in a frequency range of 1 forecasted operating temperature below 300  C. The high ionic MHz–0.1 Hz applying an AC voltage of 50 mV, in a temperature- conductivity, added to the high stability vs. pO2 proved in this low controlled Linkam probe station, from 50 to 450  C. Results temperature range, anticipates the potential use of BICUVOX were analysed with a ZView soware and tting was carried out down to room temperature in a new family of non-heated solid applying equivalent circuit modelling. state electrochemical systems. Importantly, the use of this mate- rial opens up amazing possibilities for the fabrication of solid Conflicts of interest state devices fully integrated in CMOS processing technologies and with possible encapsulation in plastics. There are no conicts to declare. Experimental section Acknowledgements Thin lm deposition by PLD The research was supported by the Generalitat de Catalunya- BICUVOX thin lms were deposited by large-area pulsed laser AGAUR (2017 SGR 1421). This project has received funding deposition (PLD5000, PVD Products) under different back- from the European research Council (ERC) under the European ground O2 pressures (0.0067 to 0.267 mbar) and target-to- Union's Horizon 2020 research and innovation programme substrate distances (9 to 11 cm), from a home-made Bi4V1.8- (ULTRASOFC, Grant Agreement number: 681146). Cu0.2O10.7 pellet target. For the fabrication of the target, Bi4- V1.8Cu0.2O10.7 powder was rst synthesized in-house by solid References state synthesis, from the basic oxide precursors (Bi2O3, V2O5 and CuO) mixed in the right stoichiometric proportion. Then, 1 J. B. Goodenough, Annu. Rev. Mater. Res., 2003, 33, 91–128. a 300 pellet was pressed by uniaxial pressing and sintered under 2 S. C. Singhal and K. Kendall, High Temperature Solid Oxide ambient air at 850  C for 6 h. All BICUVOX lms were deposited Fuel Cells: Fundamentals, Design, and Applications, Elsevier at a substrate temperature of 600  C, and on various types of Advanced Technology, 2003. substrates, including single crystal Si (100-oriented), sapphire, 3 A. Tarancón, Energies, 2009, 2, 1130–1150. fused silica, and Si (100) coated with Si3N4 (300 nm)/SiO2 (100 4 Z. Gao, L. V. Mogni, E. C. Miller, J. G. Railsback, S. A. Barnett, nm) dielectric bilayer. For the growth of epitaxial BICUVOX J. B. Goodenough, K. Huang, A. Tarancon, H. Yokokawa, lms, SrTiO3 (STO) and Nb-doped STO (Nb:STO) substrates J. Janek, Y. Shao-Horn, J. H. Lang, J. H. Yang, T. McColm, (Crystal GmbH, Germany) were used, for in-plane and cross- R. Trezona, A. Selcuk, M. Schmidt and L. Verdugo, Energy plane electrochemical analysis, respectively. Environ. Sci., 2016, 9, 1602–1644. This journal is © The Royal Society of Chemistry 2019 J. Mater. Chem. A, 2019, 7, 25772–25778 | 25777 View Article Online Journal of Materials Chemistry A Communication 5 B. Singh, S. Ghosh, S. Aich and B. Roy, J. Power Sources, 2017, 27 Q. Lu, S. R. Bishop, D. Lee, S. Lee, H. Bluhm, H. L. Tuller, 339, 103–135. H. N. Lee and B. Yildiz, Adv. Funct. Mater., 2018, 28, 1803024. 6 A. M. Azad, S. A. Akbar, S. G. Mhaisalkar, L. D. Birkefeld and 28 F. Abraham, J. C. Boivin, G. Mairesse and G. Nowogrocki, K. S. Goto, J. Electrochem. Soc., 1992, 139, 3690. Solid State Ionics, 1990, 40–41, 934–937. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. 7 D. Beckel, A. Bieberle-Hütter, A. Harvey, A. Infortuna, 29 F. Abraham, M. F. Debreuille-Gresse, G. Mairesse and U. P. Muecke, M. Prestat, J. L. M. Rupp and L. J. Gauckler, G. Nowogrocki, Solid State Ionics, 1988, 28–30, 529–532. J. Power Sources, 2007, 173, 325–345. 30 T. Iharada, A. Hammouche, J. Fouletier, M. Kleitz, 8 M. Tsuchiya, B.-K. Lai and S. Ramanathan, Nat. Nanotechnol., J. C. Boivin and G. Mairesse, Solid State Ionics, 1991, 48, 2011, 6, 282–286. 257–265. 9 A. Evans, A. Bieberle-Hütter, J. L. M. Rupp and L. J. Gauckler, 31 M. Guillodo, J. Fouletier, L. Dessemond and P. Del Gallo, Open Access Article. Published on 13 September 2019. Downloaded on 6/2/2020 1:47:29 PM. J. Power Sources, 2009, 194, 119–129. Electrochim. Acta, 2002, 47, 2809–2815. 10 I. Garbayo, D. Pla, A. Morata, L. Fonseca, N. Sabaté and 32 M. C. Steil, F. Ratajczak, E. Capoen, C. Pirovano, A. Tarancón, Energy Environ. Sci., 2014, 7, 3617–3629. R. N. Vannier and G. Mairesse, Solid State Ionics, 2005, 11 B.-U. Moon, J.-M. Lee, C.-H. Shim, M.-B. Lee, J.-H. Lee, 176, 2305–2312. D.-D. Lee and J.-H. Lee, Sens. Actuators, B, 2005, 108, 271– 33 J. C. Boivin, C. Pirovano, G. Nowogrocki, G. Mairesse, 277. P. Labrune and G. Lagrange, Solid State Ionics, 1998, 113– 12 S. Yu, Q. Wu, M. Tabib-Azar and C.-C. Liu, Sens. Actuators, B, 115, 639–651. 2002, 85, 212–218. 34 M. H. Paydar, A. M. Hadian and G. Falek, J. Mater. Sci., 13 R. Waser and M. Aono, Nat. Mater., 2007, 6, 833–840. 2006, 41, 1953–1957. 14 D. Ielmini, Semicond. Sci. Technol., 2016, 31, 063002. 35 C. Sant and J. P. Contour, J. Cryst. Growth, 1995, 153, 63–67. 15 C. Bäumer and R. Dittmann, in Metal Oxide-Based Thin Film 36 K. Trzciński, J. Gasiorowski, A. Borowska-Centkowska, Structures, Elsevier, 2018, pp. 489–522. M. Szkoda, M. Sawczak, K. Hingerl, D. R. T. Zahn and 16 I. Garbayo, A. Tarancón, J. Santiso, F. Peiró, E. Alarcón- A. Lisowska-Oleksiak, Thin Solid Films, 2017, 638, 251–257. LLadó, A. Cavallaro, I. Gràcia, C. Cané and N. Sabaté, Solid 37 A. Ojeda-G-P, C. W. Schneider, M. Döbeli, T. Lippert and State Ionics, 2010, 181, 322–331. A. Wokaun, J. Appl. Phys., 2017, 121, 135306. 17 K. Kerman, B.-K. Lai and S. Ramanathan, J. Power Sources, 38 S. Amoruso, in Metal Oxide-Based Thin Film Structures, 2011, 196, 2608–2614. Elsevier, 2018, pp. 133–160. 18 J. An, Y.-B. Kim, J. Park, T. M. Gür and F. B. Prinz, Nano Lett., 39 A. Ojeda-G-P, C. W. Schneider, M. Döbeli, T. Lippert and 2013, 13, 4551–4555. A. Wokaun, Appl. Surf. Sci., 2016, 389, 126–134. 19 I. Garbayo, V. Esposito, S. Sanna, A. Morata, D. Pla, 40 S. Havelia, S. Wang, M. Skowronski and P. A. Salvador, J. L. Fonseca, N. Sabaté and A. Tarancón, J. Power Sources, Appl. Phys., 2009, 106, 123509. 2014, 248, 1042–1049. 41 R. L. Frost, D. A. Henry, M. L. Weier and W. Martens, J. 20 K. Kerman, B.-K. Lai and S. Ramanathan, J. Power Sources, Raman Spectrosc., 2006, 37, 722–732. 2012, 202, 120–125. 42 R. Bacewicz and P. Kurek, Solid State Ionics, 2000, 127, 151– 21 Y. Takagi, S. Adam and S. Ramanathan, J. Power Sources, 156. 2012, 217, 543–553. 43 S. Kumar and P. D. Sahare, Nano, 2012, 08, 1350007. 22 D. Pla, A. Sánchez-González, I. Garbayo, M. Salleras, 44 G. Chiodelli, L. Malavasi, V. Massarotti, P. Mustarelli and A. Morata and A. Tarancón, J. Power Sources, 2015, 293, E. Quartarone, Solid State Ionics, 2005, 176, 1505–1512. 264–273. 45 J. L. M. Rupp and L. J. Gauckler, Solid State Ionics, 2006, 177, 23 B. Scherrer, A. Evans, A. J. Santis-Alvarez, B. Jiang, 2513–2518. J. Martynczuk, H. Galinski, M. Nabavi, M. Prestat, R. Tölke, 46 J. H. Joo and G. M. Choi, Solid State Ionics, 2006, 177, 1053– A. Bieberle-Hütter, D. Poulikakos, P. Muralt, 1057. P. Niedermann, A. Dommann, T. Maeder, P. Heeb, 47 P. Kurek, J. R. Dygas and M. W. Breiter, Solid State Phenom., V. Straessle, C. Muller and L. J. Gauckler, J. Power Sources, 1994, 39–40, 247–250. 2014, 258, 434–440. 48 E. Pernot, M. Anne, M. Bacmann, P. Strobel, J. Fouletier, 24 S. Sedky, A. Witvrouw, H. Bender and K. Baert, IEEE Trans. R. N. Vannier, G. Mairesse, F. Abraham and Electron Devices, 2001, 48, 377–385. G. Nowogrocki, Solid State Ionics, 1994, 70–71, 259–263. 25 N. Shukla, A. V. Thathachary, A. Agrawal, H. Paik, A. Aziz, 49 N. P. Brandon, S. Skinner and B. C. H. Steele, Annu. Rev. D. G. Schlom, S. K. Gupta, R. Engel-Herbert and S. Datta, Mater. Res., 2003, 33, 183–213. Nat. Commun., 2015, 6, 7812. 50 K. Momma and F. Izumi, J. Appl. Crystallogr., 2011, 44, 1272– 26 A. L. Krick and S. J. May, APL Mater., 2017, 5, 042504. 1276. 25778 | J. Mater. Chem. A, 2019, 7, 25772–25778 This journal is © The Royal Society of Chemistry 2019