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

Di ec de e mina ion o he Fe3+/2+ cha ge ansi ion le el in BaTiO3and iso alen ly subs i u ed Ba0.82Ca0.18Ti0.92Z 0.08O3by X- ay pho oelec on spec oscopy Sa i a Chaoudha y,1Anna M. Paulik,2Niklas Be elmann,1Ka ha ina N.S. Lohaus,1Lisanne Gossel,1Melissa A. La sson,1Heba allah Ali,3Raoul Blume,4Ju ij Ko uza,2and And eas Klein1 1Technical Uni e si y o Da ms ad , Ins i u e o Ma e ials Science, Elec onic S uc u e o Ma e ials, 64287 Da ms ad , Ge many 2G az Uni e si y o Technology, Ins i u e o Chemis y and Technology o Ma e ials, 8010 G az, Aus ia 3Fo schungszen um J¨ulich GmbH, Ins i u e o Ene gy Technologies, Fundamen al Elec ochemis y, 52425 J¨ulich, Ge many 4Max Planck Ins i u e o Chemical Ene gy Con e sion, Depa men o He e ogeneous Reac ions, 45470 M¨ulheim an de Ruh , Ge many (Da ed: No embe 7, 2025) Cha ge ansi ion le els o dopan s in oxides and o he semiconduc o s a e key ac o s a ec ing a wide ange o ma e ial p ope ies. Despi e hei impo ance, only e y ew cha ge ansi ion le - els a e known quan i a i ely. This wo k aims o alida e he di ec expe imen al de e mina ion o cha ge ansi ion le els o dopan s in oxides by means o X- ay pho oelec on spec oscopy (XPS). The app oach is used o de i e he ene gy le el associa ed o he Fe3+/2+ ansi ion, which is de- e mined as 2.45 ±0.05 eV and 2.65 ±0.05 eV abo e he alence band maximum o BaTiO3and Ba0.82Ca0.18Ti0.92Z 0.08O3, espec i ely. The o me ag ees wi h he mog a ime ic and elec ic measu emen s. The esul s consolida e ha XPS is a e sa ile and eliable echnique o expe imen- ally de e mine cha ge ansi ion le els, which can be used o e eal sys ema ic dependencies on concen a ion, empe a u e, and hos ma e ial. I is u he demons a ed ha high- empe a u e nea -ambien p essu e XPS pe o med a a synch o on is ideally sui ed o he de e mina ion o cha ge ansi ion le els. INTRODUCTION Chemical doping, i.e. he e o alen chemical subs i u- ion, is one o he mos widely es ablished app oaches o he modi ica ion o he unc ional p ope ies o mode n elec onic ma e ials. Depending on he ma e ial sys em, dopan ype, and p ocessing condi ions, he ma e ial’s eac ion o he addi ion o he e o alen ions can include ionic compensa ion, elec onic compensa ion, dopan seg- ega ion, pola on o ma ion, o he change o dopan ’s oxida ion s a e [1]. Cha ge- ansi ion le els (CTLs) a e he ene gy le els a which he dopan changes i s cha ge s a e when he Fe mi le el mo es ac oss i . The posi- ion o he Fe mi le el, de e mined by o e all cha ge neu- ali y, is he e o e also a ec ing he cha ge s a e o he dopan . As a consequence, he cha ge ansi ion ene gies a e di ec ly a ec ing de ec concen a ions [2] and hence p ope ies o a ma e ial. BaTiO3is one o he mos ex ensi ely used ma e ials o dielec ic ene gy s o age [3, 4] and lead- ee piezoelec- ic applica ions [5]. Widely s udied de ec s in BaTiO3 a e Fe and Mn, which can be p esen in a ious oxida- ion s a es om +2 o +5 [2, 6–8]. I on and manganese a e ypically accep o - ype de ec s, which a e used o supp ess elec ical conduc i i y o BaTiO3dielec ics [9] and o e oelec ic ha dening [10]. Compa ed o single- alence accep o s such as Mg o Al, mul i alen Mn has been shown o be mo e e icien in supp essing esis ance deg ada ion o BaTiO3[11], which ex ends he ope a ion empe a u e o mul ilaye ce amic capaci o s [9, 12, 13]. Excep o ew examples, including Fe and Mn, expe - imen al da a abou cha ge ansi ion le els o dopan s in BaTiO3a e sca ce [1]. The a ailable da a a e de- i ed om de ec chemical expe imen s, such as he mo- g a ime y [14, 15] o empe a u e and oxygen pa ial dependen conduc i i y measu emen s [8, 16]. In o ma- ion abou de ec cha ge s a es o ene gy le els o de- ec s can also be ex ac ed om o he echniques, such as elec on pa amagne ic esonance, pho o- and ca hodolu- minesce, M¨ossbaue spec oscopy, o deep le el ansien spec oscopy [6–8, 17–20]. Howe e , none o hese allow o a di ec de e mina ion o he cha ge ansi ion le els, i.e, in o ma ion om o he expe imen s o speci ic dedi- ca ed p epa a ion is equi ed o ei he iden i y he de ec species o assign he obse ed ansi ion le el. The p es- ence o mo e han jus one dopan o de ec may u he complica e o e en disable he analysis. I has been demons a ed ecen ly o 0.2 mol% anadium-doped BaTiO3 ha X- ay pho oelec on spec- oscopy (XPS) can di ec ly and selec i ely iden i y CTLs [21]. In ha pa icula case, compa ison wi h li e a u e could only be pe o med wi h heo e ical de- ec calcula ions. To consolida e he de e mina ion o CTLs by means o XPS, his s udy epo s on he ex- pe imen al de e mina ion o he Fe3+/2+ cha ge ansi- ion le el in BaTiO3, which is well es ablished o lie a an ene gy o 2.4 eV abo e he alence band maximum [2, 14, 19]. Howe e , de e mining he cha ge ansi- ion le el o Fe is mo e challenging han ha o V, as he mos in ense Fe 2pemission line is sp ead o e a 2 much wide ene gy ange han ha o he V 2pemis- sion [22]. Doping concen a ions o 1 % o less will hus no be de ec able wi hin a easonable measu emen ime. We ha e selec ed Fe a he han Mn in he p esen case, as Mn-doped BaTiO3exhibi s a much s onge endency o c ys allize in he hexagonal phase han Fe- doped BaTiO3[23]. We u he apply he app oach o Fe-doped Ba0.82Ca0.18Ti0.92Z 0.08O3(BCZT), which is widely s udied as a lead- ee piezoelec ic, due o i s su- pe io pe o mance as compa ed o BaTiO3[24–28]. EXPERIMENTAL The 5 mol% Fe-doped Ba0.82Ca0.18Ti0.92Z 0.08O3sam- ple was syn hesized ia con en ional solid-s a e p ocess- ing using BaCO3, CaCO3, TiO2, Z O2, and Fe2O3as s a ing ma e ials. Ba0.82Ca0.18Ti0.92Z 0.08O3, subs i u- ionally doped wi h 5 mol% o Fe on he pe o ski e B- si e, was sin e ed a 1500 ◦C. De ails on he solid-s a e p ocessing pa ame e s can likewise be ound in he Sup- plemen a y In o ma ion. No hexagonal pe o ski e phase was obse ed in he sin e ed Ba0.82Ca0.18Ti0.92Z 0.08O3 sample, which is consis en wi h p e ious li e a u e e- po s on Ba0.85Ca0.15Ti0.9Z 0.1O3[29]. De ails o sam- ple p ocessing o BaTiO3a e men ioned in he Supple- men a y In o ma ion. A inal sin e ing empe a u e o 1300 ◦C was selec ed o BaTiO3, which e ec i ely sup- p essed he o ma ion o he hexagonal phase (see Fig. S1 in he Supplemen a y In o ma ion). As he sin e ed samples a e elec ically highly insula - ing and no sui able o XPS analysis, he samples a e e- duced a 1100 ◦C in a A /5 %H2mix u e o 12 h in a ube u nace. The samples a e quickly cooled o oom em- pe a u e by a an. In o de o emo e undesi ed su ace phases, which can eme ge a e he educ ion s ep [30], he samples a e g ound wi h SiC sand pape and hen polished using Phoenix 4000 semiau oma ic machine us- ing polishing clo hs wi h 6, 3, 1 and 0.25 µm diamond pas e. The same p ocedu e o educ ion and su ace p epa a ion has been applied in a p e ious expe imen [21]. XPS analysis and di e en su ace ea men s, which a e equi ed o change he Fe mi le el and he oxida ion s a e o Fe a he su ace, a e pe o med in he Da m- s ad In eg a ed Sys em o Ma e ials Resea ch (DAISY- MAT) [31], which combines a Physical Elec onics PHI 5700 (Physical Elec onics, Chanhassan, MN) wi h se - e al acuum chambe s o su ace p ocessing and hin ilm deposi ion. Oxygen plasma ea men s a e pe - o med using a ec a MK II plasma sou ce ope a ed in a omic mode. In he DAISY-MAT sys em, XPS analysis o he Ba0.82Ca0.18Ti0.92Z 0.08O3and BaTiO3samples is conduc ed a oom empe a u e u ilizing monoch oma ic Al Kα adia ion wi h a pho on ene gy o 1486.6 eV a a akeo angle o 45 ◦. Due o hei low in ensi y, Fe 2p FIG. 1. XPS su ey spec a o Fe-doped BaTiO3and Fe- doped Ba0.82Ca0.18Ti0.92Z 0.08O3a e wo consecu i e oxy- gen plasma ea men s. Ca bon con amina ion (C 1s) a e e - ec i ely emo ed by hese ea men s. spec a a e eco ded wi h a pass ene gy o 23.5 eV, while all o he co e le els and he alence band spec a a e eco ded wi h a pass ene gy o 11.75 eV o be e ene gy esolu ion. Binding ene gies a e calib a ed on a daily basis wi h he Fe mi edge emission o a spu e -cleaned sil e oil. The Gaussian b oadening o he Fe mi edge is 0.4 eV. Fe-doped BaTiO3samples a e also analyzed using nea -ambien p essu e XPS a he ISISS bending dipole beamline and ends a ion o he BESSY II synch o on in Be lin, Ge many [32]. The Fe spec a a e measu ed using a pho on exci a ion ene gy o 970 eV and a pass ene gy o 20 eV. All o he de ailed spec a a e eco ded using a pho on exci a ion ene gy o 720 eV wi h a pass ene gy o 10 eV. The a ia ion o he pho on ene gies is accoun ed o by eco ding he O 1sspec a wi h i s and second o de di ac ion o he beamline monoch oma o . RESULTS AND DISCUSSION Figu e 1 displays he XPS su ey spec a o 2 mol% Fe-doped BaTiO3and 5 mol% Fe-doped BCZT a e e- duc ion, polishing and oxygen plasma ea men . All ex- pec ed co e le el lines (Ba, Ca, Z , Ti, Fe, and O) a e obse ed in he spec a, and he C 1semissions ha e al- mos disappea ed. The Fe 2psignal is no iceable o he highe doped BCZT, bu ba ely o he BaTiO3sample. Excep o samples measu ed di ec ly a e inse ion in o he acuum sys em, he Ba 3d, Ba 4d, and Ti 2p co e le els o BaTiO3and BCZT exhibi no p onounced changes in line shapes bu some a ia ion o binding en- 3 e gies. The same is he case o he Ca 2pand Z 3demis- sions o BCZT. The espec i e spec a a e displayed in he supplemen a y in o ma ion. Fe 2pco e le el and a- lence band spec a o he BaTiO3and he BCZT sample, eco ded a e inse ion in o he acuum chambe (”pol- ished” s a e) and a e wo consecu i e oxygen plasma ea men s, a e p esen ed in Fig. 2. The changes in he oxida ion s a e o Fe a e clea ly isible in he Fe 2pspec a, while he shi o he Fe mi le el can be disce ned om he a ia ion o he alence band onse . The Fe3+ oxida ion s a e is iden i ied by he peak maximum a a binding ene gy o ≈711 eV and by he sa elli e peak wi h a b oad maximum a ≈719.5 eV. The sepa a ion be ween he maximum and he sa elli e is cha ac e is ic o F e3+ [33, 34]. In con as , he F e2+ s a e is cha ac e ized by a b oade pa ially spli maxi- mum a ≈710 eV and by a sa elli e emission a ≈716 eV. The posi ions o he cha ac e is ic ea u es a e indica ed in he le panel o Fig. 2. Acco ding o he cha ac e is ic ea u es o wo oxida- ion s a es o Fe, one can assign he spec a as ollows: The Fe in he as polished ( educed) samples is p esen in he Fe2+ oxida ion s a e, while a pu e Fe3+ s a e is p esen a e he second oxygen plasma ea men o BaTiO3. The i s oxygen plasma ea men esul s in an oxida ion o he su ace, e iden om shi o he peaks o lowe binding ene gies (see Fig. S4 in he Supplemen- a y In o ma ion). Ne e heless, he Fe-doped BaTiO3 samples exhibi s s ill some Fe2+, e iden no only om he small addi ional shoulde a low binding ene gies o he main Fe 2ppeak, bu also om he p esence o bo h sa elli es. The beha io o he BCZT sample is analo- gous. The Fe2+ oxida ion s a e o he educed samples is es o ed a e a subsequen annealing o he samples in acuum a 300 ◦C. Howe e , he Fe 2pin ensi y is also inc eased a e his annealing. This migh be due o some seg ega ion o Fe o he su ace. The shi o he Fe mi le el is e iden om he shi o he low binding ene gy onse o he alence bands. Bo h samples exhibi ≈0.4 eV highe alence band onse s in he polished and acuum-annealed s a e as compa ed o he oxygen plasma ea ed samples. The shi s o he a- lence band maxima induced by he oxygen plasma ea - men is in pa allel o ha obse ed in he co e le el spec- a. The co esponding shi s a e displayed in Fig. 3. The cha ge ansi ion le els a e indica ed by he dashed lines. Thei posi ion is ≈50 meV highe han he Fe mi le el a e he i s oxygen plasma ea men , o which a mix- u e o a small amoun o Fe2+ and a dominan Fe3+ oxida ion s a e is obse ed. The addi ion o 50 meV is mo i a ed by he he mal b oadening o he Fe mi dis- ibu ion unc ion a oom empe a u e. Acco dingly, he Fe3+/2+ cha ge ansi ion le els a e de e mined as 2.45 ±0.05 eV and 2.65 ±0.05 eV abo e he alence band maxima o BaTiO3and BCZT, espec i ely. The de e mina ion o he cha ge ansi ion le el ac- co ding o he p ocedu e ou lined abo e is s aigh o - wa d o implemen and no special expe imen al equip- men is equi ed. A s anda d XPS spec ome e wi h a monoch oma ic X- ay sou ce and p ope binding ene gy calib a ion [36, 37] is su icien . Sample p ocessing can, bu does no need o be pe o med in-si u. Impo an equi emen s o sample p ocessing a e i) ha su icien elec ical conduc i i y is achie ed o a oid sample cha g- ing, ii) he su ace composi ion is main ained, iii) he ea men s lead o shi s o he Fe mi le el a leas a he su ace, and i ) ha he su ace con amina ion by adso ba es is kep low o su icien signal in ensi y. The cons ain o he elec ical conduc i i y o he sample is li ed i he measu emen s a e pe o med a ele a ed em- pe a u es. Fo example, empe a u es o 500 ◦C o highe a e expec ed o be su icien o ully oxidized BaTiO3 samples o a oid cha ging du ing XPS measu emen [38]. Such measu emen s ha e o be pe o med in con olled a mosphe e, howe e , as exchange o oxygen wi h he en- i onmen a such empe a u es esul s in changes o he Fe mi le el. Nea -ambien p essu e XPS is ideally sui ed o his pu pose [32, 39, 40]. Fe 2pNAP-XP spec a eco ded om wo 2 mol% Fe- doped BaTiO3pelle s a 350 −400 ◦C in di e en gas a - mosphe es a e displayed in he uppe pa o Fig. 4. One sample has been educed in A /H2and polished using he same pa ame e s as hose used o he BaTiO3sample e- po ed abo e. The second sample is p ocessed iden ically, bu no educed be o e he measu emen . This sample is e e ed o as oxidized. Fo bo h samples, changes o he oxida ion s a e o Fe om +2 o +3 a e e iden . A clea con ibu ion o Fe2+ is obse ed o he oxidized sam- ple in educing a mosphe e, while he oxida ion s a e o he educed sample changes o Fe3+ in oxidizing a mo- sphe e. In p inciple, he binding ene gies and oxida ion s a es should no depend on sample p e- ea men . The obse ed di e ences indica e ha he samples a e no in equilib ium wi h he en i onmen . This is no su p ising as bulk equilib a ion o such samples akes many hou s e en a 600 ◦C. The absence o equilib ium o he bulk does no a ec he ex ac ion o he cha ge ansi ion le els, as he Fe mi le el and oxida ion s a es a e only p obed wi hin he measu emen dep h o a ew nanome- e s. The espec i e Fe mi le el posi ions ex ac ed om he alence band and O 1sco e le el spec a a e p esen ed in he lowe panel o Fig. 4. O e all, he oxidized sample exhibi s lowe binding ene gies co esponding o a lowe Fe mi le el. The di e ence is o be expec ed o oxidized and educed samples. By combining he Fe mi le els and he Fe 2pspec a, he Fe3+/2+ cha ge ansi ion le el is de i ed as ≈2.6 eV o he oxidized and as 2.42±0.05 eV o he educed sample, espec i ely. While he la e alue is in e y good ag eemen wi h ha de e mined by he s anda d oom empe a u e XPS measu emen o he educed sample and wi h expe imen al da a epo ed in 4 FIG. 2. Backg ound-sub ac ed Fe 2pco e le el (le ) and alence band spec a ( igh ) o Fe-doped BaTiO3and Ba0.82Ca0.18Ti0.92Z 0.08O3a e educ ion and polishing, and a e wo consecu i e O2plasma ea men s. Fea u es in he Fe 2pspec a associa ed wi h Fe2+ and Fe3+ a e indica ed. The Fe 2pspec a o he plasma ea ed samples a e shi ed by 0.4 eV o highe binding ene gies o emo e he shi o he Fe mi le el, which is e iden om he shi o he alence band maxima. li e a u e [2, 14, 19], he appa en di e ence o he ox- idized sample is likely an expe imen al a i ac induced by esidual cha ging o he sample. In he expe imen al se up used, he empe a u e in educing (H2) a mosphe e was limi ed o 400 ◦C. A his empe a u e, he esis i - i y o he sample may no be low enough o comple ely a oid cha ging. The educed sample does no su e om low conduc i i y. The lowe empe a u e is bene icial in he case o he educed sample, as comple e oxida ion o he bulk is no possible wi hin he measu emen ime, lea ing mos o he sample in a educed (conduc ing) s a e. The Fe3+/2+ cha ge ansi ion le els in BaTiO3and BCZT di e by abou 0.2 eV (see Fig. 3). This di e - ence may pa ially be ela ed o he sample composi ion, as a signi ican di e ence o he Fe3+/2+ cha ge ansi- ion le el in BaTiO3and S TiO3has also been epo ed [19]. Howe e , he CTLs in Fig. 3 a e aken ela i e o he alence bands o he wo ma e ials, which may also ha e a di e en ene gy [31]. The Z -s a es in BCZT a e expec ed o mos ly a ec he conduc ion band posi ion [41–43], while he pa ial subs i u ion o Ba by Ca is likely o lowe he alence band maximum [43]. This is in ag eemen wi h he alence band maximum o BaTiO3 being abou 0.2 eV abo e ha o S TiO3[43, 44]. In o de o examine he ela i e ene gies o he alence band maxima (i.e. he band alignmen ) o BaTiO3and BCZT, he in e ace o ma ion wi h RuO2is s udied by s epwise deposi ion o RuO2o Sn-doped In2O3(ITO) on o di e en undoped and doped BCZT subs a es. The band alignmen can hen be e alua ed using he me hod desc ibed by K au and cowo ke s [45]. Un o una ely, he binding ene gy shi s o he di e en BCZT co e le - els induced by RuO2deposi ion a e di e en and also a y om sample o sample (see Fig. S7 in he Supple- men a y In o ma ion). The lowes Fe mi le el posi ion o 1.8±0.1 eV abo e he alence band maximum is obse ed o RuO2deposi ion on o a educed 5 % Fe-doped BCZT sample. This alue is 0.1 eV highe han ha de e mined p e iously a he BaTiO3/RuO2in e ace (1.7 eV) [30], sugges ing he alence band o BCZT o be 0.1eV lowe han ha o BaTiO3. This di e ence co esponds well wi h he di e ence in ene gy gap be ween BaTiO3and 20 % Ca-subs i u ed BaTiO3o 0.1 eV [46] and also co - esponds well wi h a lowe alence band maximum o CaTiO3compa ed o BaTiO3[43]. An in e ace expe - imen , in which ITO was deposi ed a 300 ◦C on o an undoped educed BCZT sample, e eals a Fe mi le el o 2.8±0.1 eV abo e he alence band maximum. The di e ence o he Fe mi le el a he BCZT/RuO2in e - 5 FIG. 3. Fe mi le el posi ion ela i e o he alence band o Fe-doped BaTiO3(le ) and Ba0.82Ca0.18Ti0.92Z 0.08O3 ( igh ) eco ded a e educing and polishing, wo consecu- i e oxygen plasma ea men s, and acuum annealing o he samples. Excep o he acuum annealed BaTiO3sample, he shi s o he co e le els a e in pa allel o hose o he alence band maximum, which is e iden a e sub ac ing he ene gy di e ences indica ed in he g aphs. The numbe s used o BaTiO3a e iden ical o hose used in p e ious wo k [30, 35], which a e de e mined om a la ge se o di e en samples. The dashed ho izon al lines indica e he Fe3+/2+ cha ge- ansi ion le el, ep esen ing he ene gy a which Fe changes i s alence s a e. ace o 1.0 eV is only sligh ly less han ha obse ed a o he oxide ma e ials [31]. This indica es he absence o in e ace Fe mi le el pinning, which can s ongly mod- i y he band alignmen [31]. The Fe mi le el posi ion a in e aces be ween high- empe a u e deposi ed ITO ilms and single c ys alline BaTiO3is also de e mined as 2.8 eV abo e he alence band maximum (see Fig. S8 in he Supplemen a y In o ma ion), in good ag eemen wi h he esul s o BCZT. Using hese esul s and con- side ing he Fe3+/2+ cha ge ansi ion le els de e mined abo e, he ene gy le el diag ams displayed in Fig. 5 we e de i ed. The ag eemen o he Fe3+/2+ cha ge ansi ion le el in BaTiO3de e mined by XPS wi h hose de i ed om he mog a ime ic and elec ical measu emen s [2, 14, 19] con i ms he eliabili y o he app oach. A main ad- an age o he de e mina ion o he CTL by XPS com- pa ed o o he echniques is ha i is a di ec me hod, e ealing he oxida ion s a e o he measu ed species and he Fe mi le el om he same se o spec a. The en- e gy band alignmen o di e en ma e ials can also be de e mined by XPS [45, 47, 48], enabling o s udy he FIG. 4. ( op) Fe 2pspec a eco ded on an as-sin e ed (oxi- dized) and a educed 2 mol% Fe-doped BaTiO3bulk ce amic. Spec a a e backg ound sub ac ed, no malized o he peak heigh o be e compa ison and shi ed on he ene gy axis o emo e he e ec o he Fe mi le el a ia ion. Labels a he le o he spec a indica e he p essu e in mba and gas a mosphe e. The measu emen empe a u es a e indi- ca ed a he igh . (bo om) The Fe mi le el posi ion ela i e o he alence band maximum unde di e en measu emen condi ions o as-sin e ed (oxidized) and educed 2 mol% Fe- doped BaTiO3samples is de e mined om O1s co e-le el and alence band spec a. Ho izon al dashed lines indica e he cha ge ansi ion le els ex ac ed om he Fe 2pspec a. ans e abili y o he CTLs om one ma e ial o ano he . This ans e abili y o CTLs would inally enable o p e- dic he e ec o speci ic dopan s wi hou p io knowl- edge and open he pa h o Fe mi le el enginee ing [1]. The alignmen o de ec ene gy le els has al eady been demons a ed o ansi ion me al impu i ies in semicon- duc o s [49–52], o hyd ogen in semiconduc o s [53], and o oxygen acancies in a ious oxides [54]. Ne e heless, he ex en o which cha ge ansi ion le els o ansi ion me al species in oxides can be ans e ed emains o be elabo a ed. In his con ex , i is also impo an o ana- lyze how CTLs depend on concen a ion and on empe - a u e. A compa ison o he Fe3+/2+ and Co3+/2+ an- si ions in (La,S )Fe3−δand (La,S )Co3−δwi h hose in 6 FIG. 5. Ene gy le el diag ams o he Fe3+/2+ cha ge ansi- ion le els o BaTiO3and Ba0.82Ca0.18Ti0.92Z 0.08O3. Bi(Fe,Co)O3indica es ha he ene gy le els migh no depend signi ican ly on he concen a ion o he species [55, 56], which would make he ans e abili y a he gen- e al. The empe a u e dependence o he de ec ene gy le els is ele an o unde s anding high empe a u e de- ice ope a ion and sample p ocessing. I is well-known ha he band gap o ma e ials dec eases wi h empe a- u e [16, 57, 58]. Howe e , knowledge abou how much he CTLs change wi h empe a u e, which is also e- qui ed o quan i a i e de ec modelling, is e y limi ed. A ailable da a, o example o Fe-doped S TiO3[16], a e ex ac ed om empe a u e and p(O2) dependen con- duc i i y measu emen s. Such measu emen s a e a ec ed by all de ec s p esen in he ma e ial and hus depend on he de ec model used o desc ibe he da a. Di ec mea- su emen s, such as hose o e ed by he XPS app oach desc ibed in his con ibu ion, can become a majo ool o be e in o med de ec models. Nea -ambien p es- su e XPS is ideally sui ed o his pu pose. The di e ence in CTL in BaTiO3and Ba0.82Ca0.18Ti0.92Z 0.08O3o 0.2 eV is a leas pa - ially ela ed o a lowe alence band maximum o he la e , which can be a ibu ed o he pa ial subs i u- ion o Ba by Ca. Taking his in o accoun , he Fe3+/2+ ansi ion in BaTiO3 emains abou 0.1 eV lowe in ene gy compa ed o ha in Ba0.82Ca0.18Ti0.92Z 0.08O3 on an absolu e ene gy scale. While his is wi hin he expe imen al unce ain y, i may also pa ially be ela ed o he highe Fe concen a ion in he s udied Ba0.82Ca0.18Ti0.92Z 0.08O3sample. A a de ec concen- a ion o 5 %, de ec in e ac ions canno be igno ed. This migh a ec he CTL. Fu u e wo k has o add ess he concen a ion and hos dependence o he de ec ene gy le els explici ly. Finally, he in o ma ion on he exac posi ion o he CTL is o di ec ele ance o he use o hese ma e i- als in dielec ic and piezoelec ic applica ions. This is o pa icula impo ance since he dopan s used in indus- ial composi ions a e o en ansi ion me al ions wi h one o mo e CTLs ha lie inside he band gap, such as Fe. Fo example, he co- i ing o elec oce amics wi h base- me al elec odes equi es he use o speci ic a mosphe es [9], which may shi he Fe mi le el owa ds a CTL and consequen ly esul in a change o he dopan ’s oxida ion s a e. This could change he conduc i i y beha io o he ma e ial. Mo eo e , in he case o piezoce amics, he oxi- da ion s a e o accep o dopan s o en de ines he amoun o cha ge-compensa ing oxygen acancies and he ype o de ec complexes [59]. Fe has been shown o inco po a e in o he B-si e o he pe o ski e la ice o BaTiO3-based ma e ials. The e o e, one oxygen acancy is o med o e e y wo Fe3+ ions, while one oxygen acancy is c e- a ed o compensa e one Fe2+ ion. Di alen ions we e hus shown o ha e s onge in luence on he educ ion o he Cu ie empe a u e, which depends on he concen- a ion o bo h he accep o -dopan -ions and ·· O[60], and could also o m di e en ypes o de ec complexes han i alen ions. In addi ion, hese de ec s ha e an elec o- s a ic and elas ic in e ac ion wi h e oelec ic domain walls [61] and hus ha e a di ec in luence on ma e ial’s unc ional p ope ies, such as piezoelec ic coe icien s, pola iza ion, and aging/ a igue beha io . SUMMARY AND CONCLUSIONS The ene gy le el o he Fe3+/2+ cha ge ansi ion in BaTiO3has been quan i ied using XPS and nea ambi- en p essu e XPS wi h in-si u manipula ion o he Fe mi le el a he su ace. The ex ac ed CTL o EF−EVB = 2.45±0.05 eV is he same as hose ex ac ed om he mo- g a ime ic and elec ical measu emen s [2, 14, 19], con- i ming he eliabili y o he XPS app oach. The capaci y o XPS o di ec ly de e mine CTLs o e s he unique op- po uni y o s udy he ans e abili y o de ec ene gy le els and hei dependence on concen a ion and em- pe a u e. I is expec ed ha gene al doping ecipes o elec oce amic oxides can be es ablished om sys ema ic s udies o hese dependencies. The compa ison o he Fe3+/2+ CTL in 5 % Fe-doped BCZT and in 2 % Fe-doped BaTiO3 e eals a di e ence o only 0.1 eV on an absolu e ene gy scale. The la e has been ob ained om ene gy band alignmen . The close ag eemen sugges s ha de ec ene gy le els can be ans- e ed om BaTiO3 o Ba0.82Ca0.18Ti0.92Z 0.08O3. The de ailed knowledge o de ec p ope ies o BaTiO3should hus cons i u e a di ec s a ing poin o se up quan i- a i e de ec models o BCZT o be e unde s and he ma e ial and enhance he piezoelec ic p ope ies. 7 SUPPLEMENTARY MATERIAL The Supplemen a y Ma e ial con ains addi ional in o - ma ion on sample p ocessing, addi ional pho oelec on spec a and da a on band alignmen om in e ace s ud- ies. ACKNOWLEDGEMENT The p esen ed wo k has been ca ied ou in he amewo k o he collabo a i e esea ch cen e FLAIR (Fe mi le el enginee ing applied o oxide elec oce am- ics), which is suppo ed by he Ge man Resea ch Foun- da ion (DFG), p ojec -ID 463184206 – SFB 1548, and by he Aus ian Fonds zu F¨o de ung de wissenscha lichen Fo schung (FWF, Aus ian Science Fund), p ojec [G an -DOI: 10.55776/I6450]. Measu emen s ca ied ou a he ISISS s a ion a he BESSY II ope a ed by he Helmhol z-Zen um Be lin ¨u Ma e ialien und Ene gie we e suppo ed wi hin p oposal-ID 251-13190-ST. Fo he pu pose o open access, he au ho has applied a CC BY public copy igh licence o any Au ho Accep ed Manusc ip e sion a ising om his submission. [1] A. Klein, K. Albe, N. Bein, O. Clemens, K. A. C eu z, P. E ha , M. F e icks, E. Gho bani, J. P. Ho mann, B. Huang, B. Kaise , U. Kolb, J. Ko uza, C. K¨ubel, K. N. S. Lohaus, J. R¨odel, J. Roh e , W. Rheinheime , R. A. De Souza, V. S eibel, A. Weidenka , M. Widen- meye , B.-X. Xu, and H. Zhang, The Fe mi ene gy as common pa ame e o desc ibe cha ge compensa ion mechanisms: A pa h o Fe mi le el enginee ing o oxide elec oce amics, J. Elec oce am. 51, 147–177 (2023). [2] B. A. Wechsle and M. B. Klein, The modynamic poin de ec model o ba ium i ana e and applica ion o he pho o e ac i e e ec , J. Op . Soc. Am. B 5, 1711 (1988). [3] C. A. Randall and P. Youse ian, Fundamen als and p ac- ical dielec ic implica ions o s oichiome y and chem- ical design in a high-pe o mance e oelec ic oxide: BaTiO3, J. Eu . Ce am. 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-en opy enginee ed BaTiO3-based ce amic capaci- o s wi h g ea ly enhanced high- empe a u e ene gy s o - age pe o mance, Na u e Commun. 16, 885 (2025). [5] M. Acos a, N. No ak, V. Rojas, S. Pa el, R. Vaish, J. Ko- uza, G. A. Rosse i, and J. R¨odel, BaTiO3-based piezo- elec ics: Fundamen als, cu en s a us, and pe spec- i es, Appl. Phys. Re . 4, 041305 (2017). [6] H.-J. Hagemann, A. He o, and U. Gonse , The alence change o Fe in BaTiO3s udied by M¨ossbaue e ec and g a ime y, phys. s a . sol. (a) 61, 63 (1980). [7] R. N. Schwa z and B. A. Wechsle , Elec on- pa amagne ic- esonance s udy o ansi ion-me al-doped BaTiO3: E ec o ma e ial p ocessing on Fe mi-le el po- si ion, Phys. Re . B 48, 7057 (1993). [8] H. K ¨ose, R. Scha schwe d , O. F. Schi me , and H. Hesse, Ligh -induced cha ge anspo in BaTiO3 ia h ee cha ge s a es o hodium, Appl. Phys. B 61, 1 (1995). [9] H. Kishi, Y. Mizuno, and H. Chazono, Base-Me al Elec ode-Mul ilaye Ce amic Capaci o s: Pas , P esen and Fu u e Pe spec i es, Jpn. J. Appl. Phys. 42, 1 (2003). [10] Y. A. Genenko, J. Glaum, M. J. Ho mann, and K. Albe, Mechanisms o aging and a igue in e oelec ics, Ma e . Sci. Engin. B 192, 52 (2015). [11] S.-H. Yoon, C. A. Randall, and K.-H. Hu , Di e ence be ween esis ance deg ada ion o ixed alence accep o (Mg) and a iable alence accep o (Mn)-doped BaTiO3 ce amics, J. Appl. Phys. 108, 064101 (2010). [12] T. Baia u, R. Wase , and K. H. H¨a d l, dc Elec ical Deg ada ion o Pe o ski e-Type Ti ana es: III, A Model o he Mechanism, J. Am. Ce am.c Soc. 73, 1663 (1990). [13] K. Hong, T. H. Lee, J. M. Suh, S.-H. Yoon, and H. W. Jang, Pe spec i es and challenges in mul ilaye ce amic capaci o s o nex gene a ion elec onics, J. Ma e . Chem. C 7, 9782 (2019). [14] H.-J. Hagemann, Accep o -ions in BaTiO3and S TiO3 and hei e ec on he p ope ies o i ana e ce amics, Ph.D. hesis, RWTH Aachen Uni e si y (1980). [15] H.-J. Hagemann and D. Hennings, Re e sible Weigh Change o Accep o -Doped BaTiO3, J. Am. Ce am. Soc. 64, 590 (1981). [16] I. Denk, W. M¨unch, and J. Maie , Pa ial conduc i i ies in S TiO3: Bulk pola iza ion expe imen s, oxygen con- cen a ion cell measu emen s, and de ec -chemical mod- eling, J. Am. Ce am. Soc. 78, 3265 (1995). [17] E. Possen iede, O. F. Schi me , H. J. Donne be g, G. Gode oy, and A. Mailla d, ESR iden i ica ion o Fe con aining de ec s in BaTiO3, Fe oelec ics 92, 245 (1989). [18] S. Zhang, B. Maliˇc, J.-F. Li, and J. R¨odel, Lead- ee e - oelec ic ma e ials: P ospec i e applica ions, J. Ma e . Res. 36, 985 (2021). [19] I. Suzuki, L. Gu a, and A. Klein, The ene gy le el o he Fe2+/3+ - ansi ion in BaTiO3and S TiO3single c ys- als, Phys. Chem. Chem. Phys. 21 (2019). [20] D. K. Sch ode , Semiconduc o ma e ial and de ice cha - ac e iza ion (John Wiley & Sons, Hoboken, 2006). [21] S. Chaoudha y, A. Aumen, E. C. Dickey, and A. Klein, Di ec de e mina ion o he cha ge ansi ion le el o dopan s in oxides by x- ay pho oelec on spec oscopy, Phys. Re . Ma e . 9, 075002 (2025). [22] J. F. Moulde , W. F. S ickle, P. E. Sobol, and K. D. Bomben, Handbook o X- ay Pho oelec on Spec oscopy (Physical Elec onics, Inc., Eden P ai ie, 1995). [23] M. P ades, N. Mas´o, H. Bel ´an, E. Co doncillo, and A. R. Wes , Polymo phism o BaTiO3Accep o Doped wi h Mn3+, Fe3+, and Ti3+, J. Am. Ce am. Soc. 91, 2364 (2008). [24] W. Liu and X. Ren, La ge Piezoelec ic E ec in Pb-F ee Ce amics, Phys. Re . Le . 103, 257602 (2009). [25] J. Ko uza, A. J. Bell, T. F ¨omling, K. G. Webbe , K. Wang, and J. R¨odel, Requi emen s o he ans e o lead- ee piezoce amics in o applica ion, J. Ma e iomics 4, 13 (2018). [26] R. Haya i, M. A. Bah e a , T. Ebadzadeh, V. Rojas, N. No ak, and J. Ko uza, E ec s o Bi2O3addi i e on 8 sin e ing p ocess and dielec ic, e oelec ic, and piezo- elec ic p ope ies o (Ba0.85Ca0.15)(Z 0.1Ti0.9)O3lead- ee piezoce amics, J. Eu . Ce am. Soc. 36, 3391 (2016). [27] M. Sayagu´es, A. O e o, L. San iago-And ades, R. Poya o, M. Monz´on, R. Paz, F. Go o , and R. Mo iche, Fine- g ained BCZT piezoelec ic ce amics by combining high- ene gy mechanochemical syn hesis and ho -p ess sin e - ing, J. Alloys Comp. (a) 1007, 176453 (2024). [28] A. M. Paulik, S. Kunz, M. Widenmeye , O. Clemens, A. Weidenka , and J. Ko uza, Solid-s a e o ma ion e- ac ions o Ba0.82Ca0.18Z 0.08Ti0.92O3piezoce amics, J. Eu . Ce am. Soc. 46, 117736 (2026). [29] B. Gong, F. Huang, Y. Shao, L. Lei, L. Liu, J. Wang, S. Yan, X. Lu, and J. Zhu, Mul i e oic p ope ies and magne oelec ic coupling o e-doped (ba0.7ca0.3) io3- ba(z 0.2 i0.8)o3 ce amics, physica s a us solidi (a) 217, 1900826 (2020). [30] K. N. S. Schuld , H. Ding, J.-C. Jaud, J. Ko uza, and A. Klein, In luence o de ec s on he Scho ky ba ie heigh a BaTiO3/RuO2in e aces, phys. s a . sol. (a) 218, 2100143 (2021). [31] A. Klein, In e ace P ope ies o Dielec ic Oxides, J. Am. Ce am. Soc. 99, 369 (2016). [32] J. J. Velasco-V´elez, V. P ei e , M. H¨a ecke , R. Wang, A. Cen eno, A. Zu u uza, G. Alga a-Sille , E. S o z, K. Sko upska, D. Teschne , P. Kube, P. B aeuninge - Weime , S. Ho mann, R. Schl¨ogl, and A. Knop-Ge icke, A mosphe ic p essu e X- ay pho oelec on spec oscopy appa a us: B idging he p essu e gap, Re . Sci. Ins . 87, 053121 (2016). [33] C. B undle, T. Chuang, and K. Wandel , Co e and a- lence le el pho oemission s udies o i on oxide su aces and he oxida ion o i on, Su ace Science 68, 459 (1977). [34] T.-C. Lin, G. Seshad i, and J. A. Kelbe , A consis en me hod o quan i a i e XPS peak analysis o hin oxide ilms on clean polyc ys alline i on su aces, Appl. Su . Sci. 119, 83 (1997). [35] K. N. S. Lohaus, De ec le el iden i ica ion in accep o - doped polyc ys alline ba ium i ana e, Ph.D. hesis, Tech- nical Uni e si y o Da ms ad (2023). [36] G. G eczynski and L. Hul man, Comp omising science by igno an ins umen calib a ion - need o e isi hal a cen u y o published XPS da a, Angew. Chem. In . Ed. 59, 5002 (2020). [37] G. G eczynski and L. Hul man, A s ep-by-s ep guide o pe o m x- ay pho oelec on spec oscopy, J. Appl. Phys. 132 (2022). [38] H.-I. Yoo, T.-S. Oh, H.-S. Kwon, D.-K. Shin, and J.- S. Lee, Elec ical conduc i i y–de ec s uc u e co ela- ion o a iable- alence and ixed- alence accep o -doped BaTiO3in quenched s a e, Phys. Chem. Chem. Phys. 11, 3115 (2009). [39] M. Salme on and R. Schl¨ogl, Ambien p essu e pho o- elec on spec oscopy: A new ool o su ace science and nano echnology, Su . Sci. Rep. 63, 169 (2008). [40] I. C. G. an den Bosch, J. U. Zaman, G. Sh e k, M. H. Hamed, M. Schneide , V. Ra o skii, Y. Bu, P. M. Di- e ich, G. Kos e , and C. Baeume , Labo a o y-based in si u and ope ando icolo x- ay pho oelec on spec- oscopy, Science Ad ances 11, eadw6673 (2025). [41] H. Lee, Y. S. Kang, S.-J. Cho, B. Xiao, H. Mo ko¸c, T. D. Kang, G. S. Lee, J. Li, S.-H. Wei, P. G. Snyde , and J. T. E ans, Dielec ic unc ions and elec onic band s uc u e o lead zi cona e i ana e hin ilms, J. Appl. Phys. 98, 094108 (2005). [42] R. D. King-Smi h and D. Vande bil , Fi s -p inciples in- es iga ion o e oelec ici y in pe o ski e compounds, Phys. Re . B 49, 5828 (1994). [43] L. Bjaalie, B. Himme oglu, L. Wes on, and A. J. C. G. V. de Walle, Oxide in e aces o no el elec onic applica- ions, New J. Phys. 16, 025005 (2014). [44] S. Balaz, Z. Zeng, and L. J. B illson, He e ojunc ion band o se s and dipole o ma ion a BaTiO3/S TiO3in e - aces, J. Appl. Phys. 114, 183701 (2013). [45] E. A. K au , R. W. G an , J. R. Wald op, and S. P. Kowalczyk, P ecise De e mina ion o he Valence-Band Edge in X-Ray Pho oemission Spec a: Applica ion o Measu emen o Semiconduc o In e ace Po en ials, Phys. Re . Le . 44, 1620 (1980). [46] S. Lee, R. D. Le i, W. Qu, S. C. Lee, and C. A. Ran- dall, Band-gap nonlinea i y in pe o ski e s uc u ed solid solu ions, J. Appl. Phys. 107, 023523 (2010). [47] A. Klein, Ene gy band alignmen a in e aces o semi- conduc ing oxides: A e iew o expe imen al de e mi- na ion using pho oelec on spec oscopy and compa i- son wi h heo e ical p edic ions by he elec on a ini y ule, cha ge neu ali y le els, and he common anion ule, Thin Solid Films 520, 3721 (2012). [48] A. Klein, Applica ion o pho oelec on spec oscopy o align he ene gy le els o pho oca alys s, in Applica ions o X- ay pho oelec on spec oscopy o ca aly ic s udies, edi ed by S. Za ei a os (Wo ld Scien i ic, 2023) pp. 193– 230. [49] A. Zunge , Composi ion dependence o deep impu i y le els in alloys, Phys. Re . Le . 54, 849 (1985). [50] A. Zunge , Theo y o 3d T ansi ion A om Impu i ies in Semiconduc o s, Ann. Re . Ma e . Sci. 15, 411 (1985). [51] J. M. Lange and H. Hein ich, Deep-Le el Impu i ies: A Possible Guide o P edic ion o Band-Edge Discon inu- i ies in Semiconduc o He e ojunc ions, Phys. Re . Le . 55, 1414 (1985). [52] J. M. Lange , C. Dele ue, M. Lannoo, and H. Hein ich, T ansi ion-me al impu i ies in semiconduc o s and he - e ojunc ion band lineups, Phys. Re . B 38, 7723 (1988). [53] C. G. Van de Walle and J. Neugebaue , Uni e sal align- men o hyd ogen le els in semiconduc o s, insula o s and solu ions, Na u e 423, 626 (2003). [54] C. Linde ¨al , A. Lindman, and P. E ha , A Uni ying Pe spec i e on Oxygen Vacancies in Wide Band Gap Ox- ides, J. Phys. Chem. Le . 9, 222 (2018). [55] Y. Liu, S. F ick, K. N. S. Lohaus, and A. Klein, Co2+/3+ and Fe2+/3+ cha ge ansi ion le els in (La,S )CoO3−δ and (La,S )FeO3−δ, Phys. Re . Ma e . 9, 075405 (2025). [56] N. S. Bein, P. Machado, M. Coll, F. Chen, M. Maka o ic, T. Rojac, and A. Klein, Elec ochemical Reduc ion o Undoped and Cobal -Doped BiFeO3Induced by Wa e Exposu e: Quan i a i e De e mina ion o Reduc ion Po- en ials and De ec Ene gy Le els Using Pho oelec on Spec oscopy, J. Phys. Chem. Le . 10, 7071 (2019). [57] S. M. Sze and K. K. Ng, Physics o Semiconduc o De- ices (John Wiley & Sons, Hoboken, 2007). [58] R. Moos and K. H. H¨a d l, De ec chemis y o dono - doped and undoped s on ium i ana e ce amics be ween 1000 deg ees and 1400 deg ees C, J. Am. Ce am. Soc. 80, 2549 (1997). [59] G. A l and H. Neumann, In e nal bias in e oelec ic ce amics: O igin and ime dependence, Fe oelec ics 87, 9 109 (1988). [60] P. Hansen, D. Hennings, and H. Sch einemache , Dielec- ic P ope ies o Accep o -Doped (Ba,Ca)(Ti,Z )O3 Ce- amics, Jou nal o Elec oce amics 2, 85 (1998). [61] T. Rojac and D. Damjano ic, Domain walls and de ec s in e oelec ic ma e ials, The Japan Socie y o Applied Physics 56, 10PA01 (2017).