Publication Preprint: Direct determination of the Fe3+/2+ charge transition level in BaTiO3 and isovalently substituted Ba0.82Ca0.18Ti0.92Zr0.08O3 by X-ray photoelectron spectroscopy

Direct determination of the Fe3+/2+ charge transition level in BaTiO3and isovalently substituted Ba0.82Ca0.18Ti0.92Zr0.08O3by X-ray photoelectron spectroscopy Savita Chaoudhary,1Anna M. Paulik,2Niklas Bertelmann,1Katharina N.S. Lohaus,1Lisanne Gossel,1Melissa A. Larsson,1Hebatallah Ali,3Raoul Blume,4Jurij Koruza,2and Andreas Klein1 1Technical University of Darmstadt, Institute of Materials Science, Electronic Structure of Materials, 64287 Darmstadt, Germany 2Graz University of Technology, Institute for Chemistry and Technology of Materials, 8010 Graz, Austria 3Forschungszentrum J¨ulich GmbH, Institute of Energy Technologies, Fundamental Electrochemistry, 52425 J¨ulich, Germany 4Max Planck Institute for Chemical Energy Conversion, Department of Heterogeneous Reactions, 45470 M¨ulheim an der Ruhr, Germany (Dated: November 7, 2025) Charge transition levels of dopants in oxides and other semiconductors are key factors affecting a wide range of material properties. Despite their importance, only very few charge transition levels are known quantitatively. This work aims to validate the direct experimental determination of charge transition levels of dopants in oxides by means of X-ray photoelectron spectroscopy (XPS). The approach is used to derive the energy level associated to the Fe3+/2+ transition, which is determined as 2.45 ±0.05 eV and 2.65 ±0.05 eV above the valence band maximum of BaTiO3and Ba0.82Ca0.18Ti0.92Zr0.08O3, respectively. The former agrees with thermogravimetric and electric measurements. The results consolidate that XPS is a versatile and reliable technique to experimentally determine charge transition levels, which can be used to reveal systematic dependencies on concentration, temperature, and host material. It is further demonstrated that high-temperature near-ambient pressure XPS performed at a synchrotron is ideally suited for the determination of charge transition levels. INTRODUCTION Chemical doping, i.e. heterovalent chemical substitution, is one of the most widely established approaches for the modification of the functional properties of modern electronic materials. Depending on the material system, dopant type, and processing conditions, the material’s reaction to the addition of heterovalent ions can include ionic compensation, electronic compensation, dopant segregation, polaron formation, or the change of dopant’s oxidation state [1]. Charge-transition levels (CTLs) are the energy levels at which the dopant changes its charge state when the Fermi level moves across it. The position of the Fermi level, determined by overall charge neutrality, is therefore also affecting the charge state of the dopant. As a consequence, the charge transition energies are directly affecting defect concentrations [2] and hence properties of a material. BaTiO3is one of the most extensively used materials for dielectric energy storage [3, 4] and lead-free piezoelectric applications [5]. Widely studied defects in BaTiO3 are Fe and Mn, which can be present in various oxidation states from +2 to +5 [2, 6–8]. Iron and manganese are typically acceptor-type defects, which are used to suppress electrical conductivity of BaTiO3dielectrics [9] and for ferroelectric hardening [10]. Compared to singlevalence acceptors such as Mg or Al, multivalent Mn has been shown to be more efficient in suppressing resistance degradation of BaTiO3[11], which extends the operation temperature of multilayer ceramic capacitors [9, 12, 13]. Except for few examples, including Fe and Mn, experimental data about charge transition levels of dopants in BaTiO3are scarce [1]. The available data are derived from defect chemical experiments, such as thermogravimetry [14, 15] or temperature and oxygen partial dependent conductivity measurements [8, 16]. Information about defect charge states or energy levels of defects can also be extracted from other techniques, such as electron paramagnetic resonance, photoand cathodoluminesce, M¨ossbauer spectroscopy, or deep level transient spectroscopy [6–8, 17–20]. However, none of these allow for a direct determination of the charge transition levels, i.e, information from other experiments or specific dedicated preparation is required to either identify the defect species or assign the observed transition level. The presence of more than just one dopant or defect may further complicate or even disable the analysis. It has been demonstrated recently for 0.2 mol% vanadium-doped BaTiO3that X-ray photoelectron spectroscopy (XPS) can directly and selectively identify CTLs [21]. In that particular case, comparison with literature could only be performed with theoretical defect calculations. To consolidate the determination of CTLs by means of XPS, this study reports on the experimental determination of the Fe3+/2+ charge transition level in BaTiO3, which is well established to lie at an energy of 2.4 eV above the valence band maximum [2, 14, 19]. However, determining the charge transition level of Fe is more challenging than that of V, as the most intense Fe 2pemission line is spread over a 2 much wider energy range than that of the V 2pemission [22]. Doping concentrations of 1 % or less will thus not be detectable within a reasonable measurement time. We have selected Fe rather than Mn in the present case, as Mn-doped BaTiO3exhibits a much stronger tendency to crystallize in the hexagonal phase than Fedoped BaTiO3[23]. We further apply the approach to Fe-doped Ba0.82Ca0.18Ti0.92Zr0.08O3(BCZT), which is widely studied as a lead-free piezoelectric, due to its superior performance as compared to BaTiO3[24–28]. EXPERIMENTAL The 5 mol% Fe-doped Ba0.82Ca0.18Ti0.92Zr0.08O3sample was synthesized via conventional solid-state processing using BaCO3, CaCO3, TiO2, ZrO2, and Fe2O3as starting materials. Ba0.82Ca0.18Ti0.92Zr0.08O3, substitutionally doped with 5 mol% of Fe on the perovskite Bsite, was sintered at 1500 ◦C. Details on the solid-state processing parameters can likewise be found in the Supplementary Information. No hexagonal perovskite phase was observed in the sintered Ba0.82Ca0.18Ti0.92Zr0.08O3 sample, which is consistent with previous literature reports on Ba0.85Ca0.15Ti0.9Zr0.1O3[29]. Details of sample processing of BaTiO3are mentioned in the Supplementary Information. A final sintering temperature of 1300 ◦C was selected for BaTiO3, which effectively suppressed the formation of the hexagonal phase (see Fig. S1 in the Supplementary Information). As the sintered samples are electrically highly insulating and not suitable for XPS analysis, the samples are reduced at 1100 ◦C in a Ar/5 %H2mixture for 12 h in a tube furnace. The samples are quickly cooled to room temperature by a fan. In order to remove undesired surface phases, which can emerge after the reduction step [30], the samples are ground with SiC sand paper and then polished using Phoenix 4000 semiautomatic machine using polishing cloths with 6, 3, 1 and 0.25 µm diamond paste. The same procedure for reduction and surface preparation has been applied in a previous experiment [21]. XPS analysis and different surface treatments, which are required to change the Fermi level and the oxidation state of Fe at the surface, are performed in the Darmstadt Integrated System for Materials Research (DAISYMAT) [31], which combines a Physical Electronics PHI 5700 (Physical Electronics, Chanhassan, MN) with several vacuum chambers for surface processing and thin film deposition. Oxygen plasma treatments are performed using a tectra MK II plasma source operated in atomic mode. In the DAISY-MAT system, XPS analysis of the Ba0.82Ca0.18Ti0.92Zr0.08O3and BaTiO3samples is conducted at room temperature utilizing monochromatic Al Kα radiation with a photon energy of 1486.6 eV at a takeoff angle of 45 ◦. Due to their low intensity, Fe 2p FIG. 1. XPS survey spectra of Fe-doped BaTiO3and Fedoped Ba0.82Ca0.18Ti0.92Zr0.08O3after two consecutive oxygen plasma treatments. Carbon contamination (C 1s) are effectively removed by these treatments. spectra are recorded with a pass energy of 23.5 eV, while all other core levels and the valence band spectra are recorded with a pass energy of 11.75 eV for better energy resolution. Binding energies are calibrated on a daily basis with the Fermi edge emission of a sputter-cleaned silver foil. The Gaussian broadening of the Fermi edge is 0.4 eV. Fe-doped BaTiO3samples are also analyzed using near-ambient pressure XPS at the ISISS bending dipole beamline and endstation of the BESSY II synchrotron in Berlin, Germany [32]. The Fe spectra are measured using a photon excitation energy of 970 eV and a pass energy of 20 eV. All other detailed spectra are recorded using a photon excitation energy of 720 eV with a pass energy of 10 eV. The variation of the photon energies is accounted for by recording the O 1sspectra with first and second order diffraction of the beamline monochromator. RESULTS AND DISCUSSION Figure 1 displays the XPS survey spectra of 2 mol% Fe-doped BaTiO3and 5 mol% Fe-doped BCZT after reduction, polishing and oxygen plasma treatment. All expected core level lines (Ba, Ca, Zr, Ti, Fe, and O) are observed in the spectra, and the C 1semissions have almost disappeared. The Fe 2psignal is noticeable for the higher doped BCZT, but barely for the BaTiO3sample. Except for samples measured directly after insertion into the vacuum system, the Ba 3d, Ba 4d, and Ti 2p core levels of BaTiO3and BCZT exhibit no pronounced changes in line shapes but some variation of binding en- 3 ergies. The same is the case for the Ca 2pand Zr 3demissions of BCZT. The respective spectra are displayed in the supplementary information. Fe 2pcore level and valence band spectra of the BaTiO3and the BCZT sample, recorded after insertion into the vacuum chamber (”polished” state) and after two consecutive oxygen plasma treatments, are presented in Fig. 2. The changes in the oxidation state of Fe are clearly visible in the Fe 2pspectra, while the shift of the Fermi level can be discerned from the variation of the valence band onset. The Fe3+ oxidation state is identified by the peak maximum at a binding energy of ≈711 eV and by the satellite peak with a broad maximum at ≈719.5 eV. The separation between the maximum and the satellite is characteristic for F e3+ [33, 34]. In contrast, the F e2+ state is characterized by a broader partially split maximum at ≈710 eV and by a satellite emission at ≈716 eV. The positions of the characteristic features are indicated in the left panel of Fig. 2. According to the characteristic features of two oxidation states of Fe, one can assign the spectra as follows: The Fe in the as polished (reduced) samples is present in the Fe2+ oxidation state, while a pure Fe3+ state is present after the second oxygen plasma treatment for BaTiO3. The first oxygen plasma treatment results in an oxidation of the surface, evident from shift of the peaks to lower binding energies (see Fig. S4 in the Supplementary Information). Nevertheless, the Fe-doped BaTiO3 samples exhibits still some Fe2+, evident not only from the small additional shoulder at low binding energies of the main Fe 2ppeak, but also from the presence of both satellites. The behavior of the BCZT sample is analogous. The Fe2+ oxidation state of the reduced samples is restored after a subsequent annealing of the samples in vacuum at 300 ◦C. However, the Fe 2pintensity is also increased after this annealing. This might be due to some segregation of Fe to the surface. The shift of the Fermi level is evident from the shift of the low binding energy onset of the valence bands. Both samples exhibit ≈0.4 eV higher valence band onsets in the polished and vacuum-annealed state as compared to the oxygen plasma treated samples. The shifts of the valence band maxima induced by the oxygen plasma treatment is in parallel to that observed in the core level spectra. The corresponding shifts are displayed in Fig. 3. The charge transition levels are indicated by the dashed lines. Their position is ≈50 meV higher than the Fermi level after the first oxygen plasma treatment, for which a mixture of a small amount of Fe2+ and a dominant Fe3+ oxidation state is observed. The addition of 50 meV is motivated by the thermal broadening of the Fermi distribution function at room temperature. Accordingly, the Fe3+/2+ charge transition levels are determined as 2.45 ±0.05 eV and 2.65 ±0.05 eV above the valence band maxima of BaTiO3and BCZT, respectively. The determination of the charge transition level according to the procedure outlined above is straightforward to implement and no special experimental equipment is required. A standard XPS spectrometer with a monochromatic X-ray source and proper binding energy calibration [36, 37] is sufficient. Sample processing can, but does not need to be performed in-situ. Important requirements for sample processing are i) that sufficient electrical conductivity is achieved to avoid sample charging, ii) the surface composition is maintained, iii) the treatments lead to shifts of the Fermi level at least at the surface, and iv) that the surface contamination by adsorbates is kept low for sufficient signal intensity. The constraint of the electrical conductivity of the sample is lifted if the measurements are performed at elevated temperatures. For example, temperatures of 500 ◦C or higher are expected to be sufficient for fully oxidized BaTiO3 samples to avoid charging during XPS measurement [38]. Such measurements have to be performed in controlled atmosphere, however, as exchange of oxygen with the environment at such temperatures results in changes of the Fermi level. Near-ambient pressure XPS is ideally suited for this purpose [32, 39, 40]. Fe 2pNAP-XP spectra recorded from two 2 mol% Fedoped BaTiO3pellets at 350 −400 ◦C in different gas atmospheres are displayed in the upper part of Fig. 4. One sample has been reduced in Ar/H2and polished using the same parameters as those used for the BaTiO3sample reported above. The second sample is processed identically, but not reduced before the measurement. This sample is referred to as oxidized. For both samples, changes of the oxidation state of Fe from +2 to +3 are evident. A clear contribution of Fe2+ is observed for the oxidized sample in reducing atmosphere, while the oxidation state of the reduced sample changes to Fe3+ in oxidizing atmosphere. In principle, the binding energies and oxidation states should not depend on sample pre-treatment. The observed differences indicate that the samples are not in equilibrium with the environment. This is not surprising as bulk equilibration of such samples takes many hours even at 600 ◦C. The absence of equilibrium of the bulk does not affect the extraction of the charge transition levels, as the Fermi level and oxidation states are only probed within the measurement depth of a few nanometers. The respective Fermi level positions extracted from the valence band and O 1score level spectra are presented in the lower panel of Fig. 4. Overall, the oxidized sample exhibits lower binding energies corresponding to a lower Fermi level. The difference is to be expected for oxidized and reduced samples. By combining the Fermi levels and the Fe 2pspectra, the Fe3+/2+ charge transition level is derived as ≈2.6 eV for the oxidized and as 2.42±0.05 eV for the reduced sample, respectively. While the latter value is in very good agreement with that determined by the standard room temperature XPS measurement of the reduced sample and with experimental data reported in 4 FIG. 2. Background-subtracted Fe 2pcore level (left) and valence band spectra (right) of Fe-doped BaTiO3and Ba0.82Ca0.18Ti0.92Zr0.08O3after reduction and polishing, and after two consecutive O2plasma treatments. Features in the Fe 2pspectra associated with Fe2+ and Fe3+ are indicated. The Fe 2pspectra of the plasma treated samples are shifted by 0.4 eV to higher binding energies to remove the shift of the Fermi level, which is evident from the shift of the valence band maxima. literature [2, 14, 19], the apparent difference of the oxidized sample is likely an experimental artifact induced by residual charging of the sample. In the experimental setup used, the temperature in reducing (H2) atmosphere was limited to 400 ◦C. At this temperature, the resistivity of the sample may not be low enough to completely avoid charging. The reduced sample does not suffer from low conductivity. The lower temperature is beneficial in the case of the reduced sample, as complete oxidation of the bulk is not possible within the measurement time, leaving most of the sample in a reduced (conducting) state. The Fe3+/2+ charge transition levels in BaTiO3and BCZT differ by about 0.2 eV (see Fig. 3). This difference may partially be related to the sample composition, as a significant difference of the Fe3+/2+ charge transition level in BaTiO3and SrTiO3has also been reported [19]. However, the CTLs in Fig. 3 are taken relative to the valence bands of the two materials, which may also have a different energy [31]. The Zr-states in BCZT are expected to mostly affect the conduction band position [41–43], while the partial substitution of Ba by Ca is likely to lower the valence band maximum [43]. This is in agreement with the valence band maximum of BaTiO3 being about 0.2 eV above that of SrTiO3[43, 44]. In order to examine the relative energies of the valence band maxima (i.e. the band alignment) of BaTiO3and BCZT, the interface formation with RuO2is studied by stepwise deposition of RuO2or Sn-doped In2O3(ITO) onto different undoped and doped BCZT substrates. The band alignment can then be evaluated using the method described by Kraut and coworkers [45]. Unfortunately, the binding energy shifts of the different BCZT core levels induced by RuO2deposition are different and also vary from sample to sample (see Fig. S7 in the Supplementary Information). The lowest Fermi level position of 1.8±0.1 eV above the valence band maximum is observed for RuO2deposition onto a reduced 5 % Fe-doped BCZT sample. This value is 0.1 eV higher than that determined previously at the BaTiO3/RuO2interface (1.7 eV) [30], suggesting the valence band of BCZT to be 0.1eV lower than that of BaTiO3. This difference corresponds well with the difference in energy gap between BaTiO3and 20 % Ca-substituted BaTiO3of 0.1 eV [46] and also corresponds well with a lower valence band maximum of CaTiO3compared to BaTiO3[43]. An interface experiment, in which ITO was deposited at 300 ◦C onto an undoped reduced BCZT sample, reveals a Fermi level of 2.8±0.1 eV above the valence band maximum. The difference to the Fermi level at the BCZT/RuO2inter- 5 FIG. 3. Fermi level position relative to the valence band for Fe-doped BaTiO3(left) and Ba0.82Ca0.18Ti0.92Zr0.08O3 (right) recorded after reducing and polishing, two consecutive oxygen plasma treatments, and vacuum annealing of the samples. Except for the vacuum annealed BaTiO3sample, the shifts of the core levels are in parallel to those of the valence band maximum, which is evident after subtracting the energy differences indicated in the graphs. The numbers used for BaTiO3are identical to those used in previous work [30, 35], which are determined from a large set of different samples. The dashed horizontal lines indicate the Fe3+/2+ charge-transition level, representing the energy at which Fe changes its valence state. face of 1.0 eV is only slightly less than that observed at other oxide materials [31]. This indicates the absence of interface Fermi level pinning, which can strongly modify the band alignment [31]. The Fermi level position at interfaces between high-temperature deposited ITO films and single crystalline BaTiO3is also determined as 2.8 eV above the valence band maximum (see Fig. S8 in the Supplementary Information), in good agreement with the results for BCZT. Using these results and considering the Fe3+/2+ charge transition levels determined above, the energy level diagrams displayed in Fig. 5 were derived. The agreement of the Fe3+/2+ charge transition level in BaTiO3determined by XPS with those derived from thermogravimetric and electrical measurements [2, 14, 19] confirms the reliability of the approach. A main advantage of the determination of the CTL by XPS compared to other techniques is that it is a direct method, revealing the oxidation state of the measured species and the Fermi level from the same set of spectra. The energy band alignment of different materials can also be determined by XPS [45, 47, 48], enabling to study the FIG. 4. (top) Fe 2pspectra recorded on an as-sintered (oxidized) and a reduced 2 mol% Fe-doped BaTiO3bulk ceramic. Spectra are background subtracted, normalized to the peak height for better comparison and shifted on the energy axis to remove the effect of the Fermi level variation. Labels at the left of the spectra indicate the pressure in mbar and gas atmosphere. The measurement temperatures are indicated at the right. (bottom) The Fermi level position relative to the valence band maximum under different measurement conditions for as-sintered (oxidized) and reduced 2 mol% Fedoped BaTiO3samples is determined from O1s core-level and valence band spectra. Horizontal dashed lines indicate the charge transition levels extracted from the Fe 2pspectra. transferability of the CTLs from one material to another. This transferability of CTLs would finally enable to predict the effect of specific dopants without prior knowledge and open the path for Fermi level engineering [1]. The alignment of defect energy levels has already been demonstrated for transition metal impurities in semiconductors [49–52], for hydrogen in semiconductors [53], and for oxygen vacancies in various oxides [54]. Nevertheless, the extent to which charge transition levels of transition metal species in oxides can be transferred remains to be elaborated. In this context, it is also important to analyze how CTLs depend on concentration and on temperature. A comparison of the Fe3+/2+ and Co3+/2+ transitions in (La,Sr)Fe3−δand (La,Sr)Co3−δwith those in 6 FIG. 5. Energy level diagrams of the Fe3+/2+ charge transition levels of BaTiO3and Ba0.82Ca0.18Ti0.92Zr0.08O3. Bi(Fe,Co)O3indicates that the energy levels might not depend significantly on the concentration of the species [55, 56], which would make the transferability rather general. The temperature dependence of the defect energy levels is relevant for understanding high temperature device operation and sample processing. It is well-known that the band gap of materials decreases with temperature [16, 57, 58]. However, knowledge about how much the CTLs change with temperature, which is also required for quantitative defect modelling, is very limited. Available data, for example for Fe-doped SrTiO3[16], are extracted from temperature and p(O2) dependent conductivity measurements. Such measurements are affected by all defects present in the material and thus depend on the defect model used to describe the data. Direct measurements, such as those offered by the XPS approach described in this contribution, can become a major tool for better informed defect models. Near-ambient pressure XPS is ideally suited for this purpose. The difference in CTL in BaTiO3and Ba0.82Ca0.18Ti0.92Zr0.08O3of 0.2 eV is at least partially related to a lower valence band maximum of the latter, which can be attributed to the partial substitution of Ba by Ca. Taking this into account, the Fe3+/2+ transition in BaTiO3remains about 0.1 eV lower in energy compared to that in Ba0.82Ca0.18Ti0.92Zr0.08O3 on an absolute energy scale. While this is within the experimental uncertainty, it may also partially be related to the higher Fe concentration in the studied Ba0.82Ca0.18Ti0.92Zr0.08O3sample. At a defect concentration of 5 %, defect interactions cannot be ignored. This might affect the CTL. Future work has to address the concentration and host dependence of the defect energy levels explicitly. Finally, the information on the exact position of the CTL is of direct relevance for the use of these materials in dielectric and piezoelectric applications. This is of particular importance since the dopants used in industrial compositions are often transition metal ions with one or more CTLs that lie inside the band gap, such as Fe. For example, the co-firing of electroceramics with basemetal electrodes requires the use of specific atmospheres [9], which may shift the Fermi level towards a CTL and consequently result in a change of the dopant’s oxidation state. This could change the conductivity behavior of the material. Moreover, in the case of piezoceramics, the oxidation state of acceptor dopants often defines the amount of charge-compensating oxygen vacancies and the type of defect complexes [59]. Fe has been shown to incorporate into the B-site of the perovskite lattice of BaTiO3-based materials. Therefore, one oxygen vacancy is formed for every two Fe3+ ions, while one oxygen vacancy is created to compensate one Fe2+ ion. Divalent ions were thus shown to have stronger influence on the reduction of the Curie temperature, which depends on the concentration of both the acceptor-dopant-ions and v·· O[60], and could also form different types of defect complexes than trivalent ions. In addition, these defects have an electrostatic and elastic interaction with ferroelectric domain walls [61] and thus have a direct influence on material’s functional properties, such as piezoelectric coefficients, polarization, and aging/fatigue behavior. SUMMARY AND CONCLUSIONS The energy level of the Fe3+/2+ charge transition in BaTiO3has been quantified using XPS and near ambient pressure XPS with in-situ manipulation of the Fermi level at the surface. The extracted CTL of EF−EVB = 2.45±0.05 eV is the same as those extracted from thermogravimetric and electrical measurements [2, 14, 19], confirming the reliability of the XPS approach. The capacity of XPS to directly determine CTLs offers the unique opportunity to study the transferability of defect energy levels and their dependence on concentration and temperature. It is expected that general doping recipes for electroceramic oxides can be established from systematic studies of these dependencies. The comparison of the Fe3+/2+ CTL in 5 % Fe-doped BCZT and in 2 % Fe-doped BaTiO3reveals a difference of only 0.1 eV on an absolute energy scale. The latter has been obtained from energy band alignment. The close agreement suggests that defect energy levels can be transferred from BaTiO3to Ba0.82Ca0.18Ti0.92Zr0.08O3. The detailed knowledge of defect properties of BaTiO3should thus constitute a direct starting point to set up quantitative defect models for BCZT to better understand the material and enhance the piezoelectric properties. 7 SUPPLEMENTARY MATERIAL The Supplementary Material contains additional information on sample processing, additional photoelectron spectra and data on band alignment from interface studies. ACKNOWLEDGEMENT The presented work has been carried out in the framework of the collaborative research centre FLAIR (Fermi level engineering applied to oxide electroceramics), which is supported by the German Research Foundation (DFG), project-ID 463184206 – SFB 1548, and by the Austrian Fonds zur F¨orderung der wissenschaftlichen Forschung (FWF, Austrian Science Fund), project [Grant-DOI: 10.55776/I6450]. Measurements carried out at the ISISS station at the BESSY II operated by the Helmholtz-Zentrum Berlin f¨ur Materialien und Energie were supported within proposal-ID 251-13190-ST. For the purpose of open access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. [1] A. Klein, K. Albe, N. Bein, O. Clemens, K. A. Creutz, P. Erhart, M. Frericks, E. Ghorbani, J. P. Hofmann, B. Huang, B. Kaiser, U. Kolb, J. Koruza, C. K¨ubel, K. N. S. Lohaus, J. R¨odel, J. Rohrer, W. Rheinheimer, R. A. De Souza, V. Streibel, A. Weidenkaff, M. Widenmeyer, B.-X. Xu, and H. Zhang, The Fermi energy as common parameter to describe charge compensation mechanisms: A path to Fermi level engineering of oxide electroceramics, J. Electroceram. 51, 147–177 (2023). [2] B. A. Wechsler and M. B. Klein, Thermodynamic point defect model of barium titanate and application to the photorefractive effect, J. Opt. Soc. Am. B 5, 1711 (1988). [3] C. A. Randall and P. Yousefian, Fundamentals and practical dielectric implications of stoichiometry and chemical design in a high-performance ferroelectric oxide: BaTiO3, J. Eur. Ceram. Soc. 42, 1445 (2022). [4] X. Kong, L. Yang, F. Meng, T. Zhang, H. Zhang, Y.- H. Lin, H. Huang, S. Zhang, J. Guo, and C.-W. Nan, High-entropy engineered BaTiO3-based ceramic capacitors with greatly enhanced high-temperature energy storage performance, Nature Commun. 16, 885 (2025). [5] M. Acosta, N. Novak, V. Rojas, S. Patel, R. Vaish, J. Koruza, G. A. Rossetti, and J. R¨odel, BaTiO3-based piezoelectrics: Fundamentals, current status, and perspectives, Appl. Phys. Rev. 4, 041305 (2017). [6] H.-J. Hagemann, A. Hero, and U. Gonser, The valence change of Fe in BaTiO3studied by M¨ossbauer effect and gravimetry, phys. stat. sol. (a) 61, 63 (1980). [7] R. N. Schwartz and B. A. Wechsler, Electronparamagnetic-resonance study of transition-metal-doped BaTiO3: Effect of material processing on Fermi-level position, Phys. Rev. B 48, 7057 (1993). [8] H. Kr¨ose, R. Scharfschwerdt, O. F. Schirmer, and H. Hesse, Light-induced charge transport in BaTiO3via three charge states of rhodium, Appl. Phys. B 61, 1 (1995). [9] H. Kishi, Y. Mizuno, and H. Chazono, Base-Metal Electrode-Multilayer Ceramic Capacitors: Past, Present and Future Perspectives, Jpn. J. Appl. Phys. 42, 1 (2003). [10] Y. A. Genenko, J. Glaum, M. J. Hoffmann, and K. Albe, Mechanisms of aging and fatigue in ferroelectrics, Mater. Sci. Engin. B 192, 52 (2015). [11] S.-H. Yoon, C. A. Randall, and K.-H. Hur, Difference between resistance degradation of fixed valence acceptor (Mg) and variable valence acceptor (Mn)-doped BaTiO3 ceramics, J. Appl. Phys. 108, 064101 (2010). [12] T. Baiatu, R. Waser, and K. H. H¨ardtl, dc Electrical Degradation of Perovskite-Type Titanates: III, A Model of the Mechanism, J. Am. Ceram.c Soc. 73, 1663 (1990). [13] K. Hong, T. H. Lee, J. M. Suh, S.-H. Yoon, and H. W. Jang, Perspectives and challenges in multilayer ceramic capacitors for next generation electronics, J. Mater. Chem. C 7, 9782 (2019). [14] H.-J. Hagemann, Acceptor-ions in BaTiO3and SrTiO3 and their effect on the properties of titanate ceramics, Ph.D. thesis, RWTH Aachen University (1980). [15] H.-J. Hagemann and D. Hennings, Reversible Weight Change of Acceptor-Doped BaTiO3, J. Am. Ceram. Soc. 64, 590 (1981). [16] I. Denk, W. M¨unch, and J. Maier, Partial conductivities in SrTiO3: Bulk polarization experiments, oxygen concentration cell measurements, and defect-chemical modeling, J. Am. Ceram. Soc. 78, 3265 (1995). [17] E. Possenriede, O. F. Schirmer, H. J. Donnerberg, G. Godefroy, and A. Maillard, ESR identification of Fe containing defects in BaTiO3, Ferroelectrics 92, 245 (1989). [18] S. Zhang, B. Maliˇc, J.-F. Li, and J. R¨odel, Lead-free ferroelectric materials: Prospective applications, J. Mater. Res. 36, 985 (2021). [19] I. Suzuki, L. Gura, and A. Klein, The energy level of the Fe2+/3+ -transition in BaTiO3and SrTiO3single crystals, Phys. Chem. Chem. Phys. 21 (2019). [20] D. K. Schroder, Semiconductor material and device characterization (John Wiley & Sons, Hoboken, 2006). [21] S. Chaoudhary, A. Aumen, E. C. Dickey, and A. Klein, Direct determination of the charge transition level of dopants in oxides by x-ray photoelectron spectroscopy, Phys. Rev. Mater. 9, 075002 (2025). [22] J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of X-ray Photoelectron Spectroscopy (Physical Electronics, Inc., Eden Prairie, 1995). [23] M. Prades, N. Mas´o, H. Beltr´an, E. Cordoncillo, and A. R. West, Polymorphism of BaTiO3Acceptor Doped with Mn3+, Fe3+, and Ti3+, J. Am. Ceram. Soc. 91, 2364 (2008). [24] W. Liu and X. Ren, Large Piezoelectric Effect in Pb-Free Ceramics, Phys. Rev. Lett. 103, 257602 (2009). [25] J. Koruza, A. J. Bell, T. Fr¨omling, K. G. Webber, K. Wang, and J. R¨odel, Requirements for the transfer of lead-free piezoceramics into application, J. Materiomics 4, 13 (2018). [26] R. Hayati, M. A. Bahrevar, T. Ebadzadeh, V. Rojas, N. Novak, and J. Koruza, Effects of Bi2O3additive on 8 sintering process and dielectric, ferroelectric, and piezoelectric properties of (Ba0.85Ca0.15)(Zr0.1Ti0.9)O3leadfree piezoceramics, J. Eur. Ceram. Soc. 36, 3391 (2016). [27] M. Sayagu´es, A. Otero, L. Santiago-Andrades, R. Poyato, M. Monz´on, R. Paz, F. Gotor, and R. Moriche, Finegrained BCZT piezoelectric ceramics by combining highenergy mechanochemical synthesis and hot-press sintering, J. Alloys Comp. (a) 1007, 176453 (2024). [28] A. M. Paulik, S. Kunz, M. Widenmeyer, O. Clemens, A. Weidenkaff, and J. Koruza, Solid-state formation reactions of Ba0.82Ca0.18Zr0.08Ti0.92O3piezoceramics, J. Eur. Ceram. Soc. 46, 117736 (2026). [29] B. Gong, F. Huang, Y. Shao, L. Lei, L. Liu, J. Wang, S. Yan, X. Lu, and J. Zhu, Multiferroic properties and magnetoelectric coupling of fe-doped (ba0.7ca0.3)tio3ba(zr0.2ti0.8)o3 ceramics, physica status solidi (a) 217, 1900826 (2020). [30] K. N. S. Schuldt, H. Ding, J.-C. Jaud, J. Koruza, and A. Klein, Influence of defects on the Schottky barrier height at BaTiO3/RuO2interfaces, phys. stat. sol. (a) 218, 2100143 (2021). [31] A. Klein, Interface Properties of Dielectric Oxides, J. Am. Ceram. Soc. 99, 369 (2016). [32] J. J. Velasco-V´elez, V. Pfeifer, M. H¨avecker, R. Wang, A. Centeno, A. Zurutuza, G. Algara-Siller, E. Stotz, K. Skorupska, D. Teschner, P. Kube, P. BraeuningerWeimer, S. Hofmann, R. Schl¨ogl, and A. Knop-Gericke, Atmospheric pressure X-ray photoelectron spectroscopy apparatus: Bridging the pressure gap, Rev. Sci. Instr. 87, 053121 (2016). [33] C. Brundle, T. Chuang, and K. Wandelt, Core and valence level photoemission studies of iron oxide surfaces and the oxidation of iron, Surface Science 68, 459 (1977). [34] T.-C. Lin, G. Seshadri, and J. A. Kelber, A consistent method for quantitative XPS peak analysis of thin oxide films on clean polycrystalline iron surfaces, Appl. Surf. Sci. 119, 83 (1997). [35] K. N. S. Lohaus, Defect level identification in acceptordoped polycrystalline barium titanate, Ph.D. thesis, Technical University of Darmstadt (2023). [36] G. Greczynski and L. Hultman, Compromising science by ignorant instrument calibration - need to revisit half a century of published XPS data, Angew. Chem. Int. Ed. 59, 5002 (2020). [37] G. Greczynski and L. Hultman, A step-by-step guide to perform x-ray photoelectron spectroscopy, J. Appl. Phys. 132 (2022). [38] H.-I. Yoo, T.-S. Oh, H.-S. Kwon, D.-K. Shin, and J.- S. Lee, Electrical conductivity–defect structure correlation of variable-valence and fixed-valence acceptor-doped BaTiO3in quenched state, Phys. Chem. Chem. Phys. 11, 3115 (2009). [39] M. Salmeron and R. Schl¨ogl, Ambient pressure photoelectron spectroscopy: A new tool for surface science and nanotechnology, Surf. Sci. Rep. 63, 169 (2008). [40] I. C. G. van den Bosch, J. U. Zaman, G. Shterk, M. H. Hamed, M. Schneider, V. Ratovskii, Y. Bu, P. M. Dietrich, G. Koster, and C. Baeumer, Laboratory-based in situ and operando tricolor x-ray photoelectron spectroscopy, Science Advances 11, eadw6673 (2025). [41] H. Lee, Y. S. Kang, S.-J. Cho, B. Xiao, H. Morko¸c, T. D. Kang, G. S. Lee, J. Li, S.-H. Wei, P. G. Snyder, and J. T. Evans, Dielectric functions and electronic band structure of lead zirconate titanate thin films, J. Appl. Phys. 98, 094108 (2005). [42] R. D. King-Smith and D. Vanderbilt, First-principles investigation of ferroelectricity in perovskite compounds, Phys. Rev. B 49, 5828 (1994). [43] L. Bjaalie, B. Himmetoglu, L. Weston, and A. J. C. G. V. de Walle, Oxide interfaces for novel electronic applications, New J. Phys. 16, 025005 (2014). [44] S. Balaz, Z. Zeng, and L. J. Brillson, Heterojunction band offsets and dipole formation at BaTiO3/SrTiO3interfaces, J. Appl. Phys. 114, 183701 (2013). [45] E. A. Kraut, R. W. Grant, J. R. Waldrop, and S. P. Kowalczyk, Precise Determination of the Valence-Band Edge in X-Ray Photoemission Spectra: Application to Measurement of Semiconductor Interface Potentials, Phys. Rev. Lett. 44, 1620 (1980). [46] S. Lee, R. D. Levi, W. Qu, S. C. Lee, and C. A. Randall, Band-gap nonlinearity in perovskite structured solid solutions, J. Appl. Phys. 107, 023523 (2010). [47] A. Klein, Energy band alignment at interfaces of semiconducting oxides: A review of experimental determination using photoelectron spectroscopy and comparison with theoretical predictions by the electron affinity rule, charge neutrality levels, and the common anion rule, Thin Solid Films 520, 3721 (2012). [48] A. Klein, Application of photoelectron spectroscopy to align the energy levels of photocatalysts, in Applications of X-ray photoelectron spectroscopy to catalytic studies, edited by S. Zafeiratos (World Scientific, 2023) pp. 193– 230. [49] A. Zunger, Composition dependence of deep impurity levels in alloys, Phys. Rev. Lett. 54, 849 (1985). [50] A. Zunger, Theory of 3d Transition Atom Impurities in Semiconductors, Ann. Rev. Mater. Sci. 15, 411 (1985). [51] J. M. Langer and H. Heinrich, Deep-Level Impurities: A Possible Guide to Prediction of Band-Edge Discontinuities in Semiconductor Heterojunctions, Phys. Rev. Lett. 55, 1414 (1985). [52] J. M. Langer, C. Delerue, M. Lannoo, and H. Heinrich, Transition-metal impurities in semiconductors and heterojunction band lineups, Phys. Rev. B 38, 7723 (1988). [53] C. G. Van de Walle and J. Neugebauer, Universal alignment of hydrogen levels in semiconductors, insulators and solutions, Nature 423, 626 (2003). [54] C. Linder¨alv, A. Lindman, and P. Erhart, A Unifying Perspective on Oxygen Vacancies in Wide Band Gap Oxides, J. Phys. Chem. Lett. 9, 222 (2018). [55] Y. Liu, S. Frick, K. N. S. Lohaus, and A. Klein, Co2+/3+ and Fe2+/3+ charge transition levels in (La,Sr)CoO3−δ and (La,Sr)FeO3−δ, Phys. Rev. Mater. 9, 075405 (2025). [56] N. S. Bein, P. Machado, M. Coll, F. Chen, M. Makarovic, T. Rojac, and A. Klein, Electrochemical Reduction of Undoped and Cobalt-Doped BiFeO3Induced by Water Exposure: Quantitative Determination of Reduction Potentials and Defect Energy Levels Using Photoelectron Spectroscopy, J. Phys. Chem. Lett. 10, 7071 (2019). [57] S. M. Sze and K. K. Ng, Physics of Semiconductor Devices (John Wiley & Sons, Hoboken, 2007). [58] R. Moos and K. H. H¨ardtl, Defect chemistry of donordoped and undoped strontium titanate ceramics between 1000 degrees and 1400 degrees C, J. Am. Ceram. Soc. 80, 2549 (1997). [59] G. Arlt and H. Neumann, Internal bias in ferroelectric ceramics: Origin and time dependence, Ferroelectrics 87, 9 109 (1988). [60] P. Hansen, D. Hennings, and H. Schreinemacher, Dielectric Properties of Acceptor-Doped (Ba,Ca)(Ti,Zr)O3 Ceramics, Journal of Electroceramics 2, 85 (1998). [61] T. Rojac and D. Damjanovic, Domain walls and defects in ferroelectric materials, The Japan Society of Applied Physics 56, 10PA01 (2017).