# Fileset

[Darminto2023 - Darminto - On the Origin of the Non‐Arrhenius Na‐ion Conductivity in Na3OBr.pdf](https://mdr.nims.go.jp/filesets/dc5e7a7e-ebb6-448b-bb54-a02a439f5405/download)

## Creator

Brigita Darminto, Gregory J. Rees, John Cattermull, [Kenjiro Hashi](https://orcid.org/0000-0002-0320-4768), Maria Diaz‐Lopez, [Naoaki Kuwata](https://orcid.org/0000-0002-0736-6967), Stephen J. Turrell, Emily Milan, Yvonne Chart, Camilla Di Mino, Hyeon Jeong Lee, Andrew L. Goodwin, Mauro Pasta

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[On the Origin of the Non‐Arrhenius Na‐ion Conductivity in Na<sub>3</sub>OBr](https://mdr.nims.go.jp/datasets/041fbd1d-bd54-487b-9149-41587e066109)

## Fulltext

On the Origin of the Non‐Arrhenius Na‐ion Conductivity in Na3OBrSolid-State BatteriesOn the Origin of the Non-Arrhenius Na-ion Conductivity in Na3OBrBrigita Darminto, Gregory J. Rees, John Cattermull, Kenjiro Hashi, Maria Diaz-Lopez,Naoaki Kuwata, Stephen J. Turrell, Emily Milan, Yvonne Chart, Camilla Di Mino,Hyeon Jeong Lee, Andrew L. Goodwin, and Mauro Pasta*Abstract: The sodium-rich antiperovskites (NaRAPs) with composition Na3OB (B=Br, Cl, I, BH4, etc.) are a family ofmaterials that has recently attracted great interest for application as solid electrolytes in sodium metal batteries. Non-Arrhenius ionic conductivities have been reported for these materials, the origin of which is poorly understood. In thiswork, we combined temperature-resolved bulk and local characterisation methods to gain an insight into the origin ofthis unusual behaviour using Na3OBr as a model system. We first excluded crystallographic disorder on the anion sites asthe cause of the change in activation energy; then identified the presence of a poorly crystalline impurities, not detectableby XRD, and elucidated their effect on ionic conductivity. These findings improve understanding of the processing-structure-properties relationships pertaining to NaRAPs and highlight the need to determine these relationships in othermaterials systems, which will accelerate the development of high-performance solid electrolytes.IntroductionSodium metal batteries (SMBs) have attracted significantattention as a potential alternative to lithium-ion batteries(LiBs) due to their several key advantages: the abundanceof sodium (1200 times more common than lithium), hightheoretical specific capacity (1166 mAhg compared to372 mAhg of graphite anodes in LiBs), and the lowsolubility of sodium in aluminium, enabling its use as alightweight and inexpensive current collector.[1–3] Despitethese benefits, combining a Na metal anode with an organicliquid electrolyte presents very serious safety risks. The highchemical activity of Na metal anodes can lead to unavoid-able side reactions and uncontrolled dendrite growth,causing a continuous decline in electrochemical performanceand potentially leading to internal short circuits, thermalrunaway, and even fires.[4–9] Solid-state sodium metalbatteries (SSMBs) have been proposed as alternatives.There has been an increasing interest in Na-rich antiper-ovskite (NaRAP) materials as solid-state electrolytes (SSEs)for SSMBs over the last decade owing to their stabilityagainst Na metal, low synthesis temperatures (<500 °C), andrelatively soft mechanical properties (bulk modulus:�60 GPa) which facilitate processing conditions and im-prove contact with the active materials.[10–14] Their generalformula is Na3AB, where A is a divalent anion and B iseither a monovalent or a cluster anion.The room temperature Na-ion conductivity varies widelydepending on composition and processing conditions, withthe highest value of 4.4×10� 3 Scm� 1 reported by Sun et al.for Na3O(BH4); however, attempts to reproduce their resultshave been unsuccessful.[10,11,15] Interestingly, non-Arrheniusbehaviour resulting in a decrease in activation energy for ionconduction above 250 °C has been observed across differentantiperovskite compositions (Li, K and Na).[16–23] The originof this behaviour is not yet well understood. For antiper-ovskite compositions with cluster anions like Na3O(NO2)and Na2(NH2)(BH4), deviation from Arrhenius behaviour isoften assigned to activation of the cluster anion rotationsand correlated motion between these rotations and thetranslating cation.[17,18]For the compositions with single halide anions, suchbehaviour is usually attributed to structural instability, whichincludes phase transitions, order-disorder transitions of theion sublattice, and crystal melting.[16,24,25] Zheng et al. arguedthat the non-Arrhenius behaviour observed in K3OI could[*] B. Darminto, Dr. G. J. Rees, J. Cattermull, Dr. S. J. Turrell, E. Milan,Y. Chart, Dr. C. Di Mino, Prof. H. Jeong Lee, Prof. M. PastaDepartment of Materials, University of Oxford,Oxford, OX1 3PH, United KingdomE-mail: mauro.pasta@materials.ox.ac.ukDr. G. J. Rees, Dr. S. J. Turrell, Y. Chart, Dr. C. Di Mino,Prof. H. Jeong Lee, Prof. M. PastaThe Faraday Institution, Harwell Campus,Oxford, OX11 0RA, United KingdomJ. Cattermull, Prof. A. L. GoodwinDepartment of Chemistry, University of Oxford,Oxford, OX1 3TA, United KingdomDr. K. Hashi, Dr. N. KuwataNational Institute for Materials Science,Tsukuba, 305-0044, JapanDr. M. Diaz-LopezDiamond Light Source,Oxford, OX11 0DE, United KingdomProf. H. Jeong LeeDepartment of Materials Science and Engineering,Ulsan National Institute of Science and Technology,Ulsan, 44919, South Korea© 2023 The Authors. Angewandte Chemie International Editionpublished by Wiley-VCH GmbH. This is an open access article underthe terms of the Creative Commons Attribution License, whichpermits use, distribution and reproduction in any medium, providedthe original work is properly cited.AngewandteChemieResearch Articleswww.angewandte.orgHow to cite: Angew. Chem. Int. Ed. 2023, 62, e202314444doi.org/10.1002/anie.202314444Angew. Chem. Int. Ed. 2023, 62, e202314444 (1 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbHhttp://orcid.org/0000-0002-2613-4555https://doi.org/10.1002/anie.202314444http://crossmark.crossref.org/dialog/?doi=10.1002%2Fanie.202314444&domain=pdf&date_stamp=2023-11-14be related to the effect of K vacancies on the ion kinetics ofthe system.[16] K3OI stays crystalline up to 350 °C—wellabove the transition temperature of 240 °C. Through ab-initio molecular dynamics (AIMD) simulations, the authorsdemonstrated that at high temperatures local anion disor-dering is activated, and K-ions migrate not only via the Kvacancies, but also through the relatively large vacanciesaround the disordered I� O sites, thus enhancing K-iontransport.[16] Zhang et al. reported similar behaviour as themain driver for enhanced Li ionic mobility in Li3OCl, andreferred to the phenomenon as ‘sublattice melting’. Toensure the formation of energetically unstable Li3OBr, theyperformed in situ variable temperature neutron scatteringexperiments at high pressures (3.1 and 6.5 GPa). An abruptvolume increase was observed between 350 °C and 400 °C,which corresponds to �0.8 Tm, in agreement with AIMDsimulations reported by their group.[24,25] On the other hand,Dawson et al. performed molecular dynamics (MD) calcu-lations on Li3OB (B=Cl or Br) and Na3OB (B=Cl or Br)for fast-conducting solid electrolytes with an alkali-halidepartial Schottky defect concentration of δ=0.038 and didnot observe any deviations from Arrhenius behaviour intheir ionic conductivities at elevated temperatures.[26] Thesame behaviour is observed in Li2OHCl and suggested to bedifferent from that reported for K3OI and Li3OBr. Wanget al. argued that the defects on the Li site and themovement of H around the O atom give rise to a change inthe ionic conduction pathway as temperature increases.[27]Despite the initial similarities, these varying reports onseveral antiperovskite families indicate that the origin ofnon-Arrhenius behaviour is influenced by the type andconcentration of defects present in the structure and is notnecessarily the same across different families.Na3OBr was chosen for this experiment as it is reportedto show non-Arrhenius behaviour, the origin of which is notfully understood.[20] Furthermore, this composition is pre-dicted to be stable in the cubic antiperovskite structure onthe basis of its Goldschmidt tolerance factor, t, of 0.89. Thisis a dimensionless parameter calculated from the ratio of theionic radii of the constituent ions in the material.[28,29] Avalue of t=1 represents the perfect case for the formation ofa cubic structure.[28] Multiple cubic NaRAPs with t between0.84 and 0.97 have been synthesised and the value forNa3OBr is within this range.[15] We studied variable-temper-ature X-ray diffraction (XRD), variable-temperature nu-clear magnetic resonance (NMR), variable-temperatureelectrochemical impedance spectroscopy (EIS), differentialscanning calorimetry (DSC), and scanning electron micro-scopy (SEM). The combination of these techniques allowedus to gain new insights into this unusual behaviour.Results and DiscussionWe synthesised Na3OBr with commercially available Na2O,which contains �80% Na2O and �20% Na2O2. Thesignificant amount of Na2O2 was taken into account in theprecursor ratio calculation, as seen in the equation below.0:8Na2Oþ 0:2Na2O2 þNaBr! Na3OBrþ 0:1O2 (1)A combination of high-energy ball milling and subse-quent heating is expected to decompose Na2O2 into Na2Oand O2. This synthesis procedure has been reported multipletimes in the literature and led to successful formation ofNaRAPs.[12,17] Differential scanning calorimetry (DSC) wasperformed to study the thermal behaviour of the Na3OBr,shown in Figure 1a. Upon heating, an endothermic peak wasobserved with an onset at 247 °C and maximum at 258 °C.Previous studies on Li, Na and K antiperovskites alsoshowed similar endothermic peaks (K3OI: 245 °C, Li2OHBr:254 °C, Na3OBr: 255 °C, and Na3OCl: 191 °C). Some studiesattributed these peaks to melting of the antiperovskitephases and others to anion disordering.[14,16,17,19] Except forthis peak, no other peaks were observed in the DSC plotduring heating to 300 °C. Additionally, the bulk materialremains solid up to 300 °C, which is confirmed by variableFigure 1. In-house characterisations of as-synthesised Na3OBr. (a) DSCof Na3OBr showing an endothermic peak with an onset at 247 °C.(b) Arrhenius plot of total ionic conductivity of Na3OBr with a stepincrease observed at 250 °C.AngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (2 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensetemperature synchrotron XRD, Figure 2b, meaning that thepeak does not correspond to the melting of Na3OBr.Temperature-dependent potentiostatic electrochemicalimpedance spectroscopy (PEIS) was used to investigate theeffect of temperature on ionic conductivity. A pressure of60–70 MPa was applied to the cell during the measurementsto maintain good contact between gold blocking electrodesand the Na3OBr pellet. To determine the relative contribu-Figure 2. (a) Rietveld refinement of the room temperature synchrotron X-ray diffractograms of synthesised Na3OBr with an X-ray wavelength of0.825318 Å. The black line is the experimental data, the red line is the fit, and blue line is the difference between them. (b) Variable temperatureexperiment. Sequential Rietveld refinements were performed on the variable temperature data, and the changes in (c) lattice parameter, (d) thermalparameter of Na (beq Na), and (e) thermal parameter of Br (beq Br) of Na3OBr are plotted as a function of temperature. These show thermalexpansion with no presence of crystalline deviation. (f) Differential correlation function D(r) calculated from experimental X-ray total structurefactor F(Q). Measurement was done with an X-ray wavelength of 0.161669 Å.AngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (3 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensetions of the electrolyte bulk and boundaries to the totalimpedance, the setup was modelled as an equivalent circuitshown in Figure S1. R1 and CPE1 correspond to bulk Na+conduction while R2 and CPE2 correspond to conductionthrough the boundaries. The calculated total ionic conduc-tivities are presented in Figure 1b. There is a step change inthe total ionic conductivity at 250 °C, the temperature atwhich the endothermic peak was observed in the DSC,which corresponds to a reduction of activation energy from0.47–0.43 eV. These activation energies are slightly lowerthan the reported values (0.52–1.14 eV), possibly due to thehigher pressure applied to the cell during themeasurements.[10,12,20,30,31] Refinements were performed overvariable temperature synchrotron XRD data to elucidatethe origin of this phenomenon.In agreement with previous reports, the room temper-ature (RT) XRD pattern was fitted to the cubic space groupPm�3m with a unit cell of a=4.562(6) Å, using a Rietveldrefinement (Rwp=1.02%). No Bragg peaks other than thosewhich belong to the Na3OBr are present in the diffractionpattern, thus no other crystalline species are present atsignificant concentrations in our sample. To understand howthe crystal structure evolves with temperature, sequentialRietveld refinements were performed on the synchrotronXRD data between RT and 300 °C, with a step size of 2.5 °C(Figure 2b). Any sign of discontinuity in the lattice parame-ter and elemental thermal parameters (beq) could give anindication of an order to disorder transition. The intensitiesof the Na3OBr peaks do not change drastically above 247 °C,hence the Na3OBr powder did not melt. The Bragg peakpositions shift to lower angles due to thermal expansion ofthe Na3OBr lattice, but the rate of change in latticeparameter is constant with temperature, even above 267 °Cwhere the endothermic peak ends, as seen in Figure 2c.When fitted with a straight line, the deviation of each pointfrom the fit can be plotted (Figure S2). A change in thedeviation trend above 247 °C was observed; however, thedeviations are of the order of 10� 4 Å, making themnegligible compared to the size of any atom in the structure.Out of the three elements contained in our sample, wedecided to focus our attention on the thermal parameters ofNa and Br. This is because the thermal parameters of Ohave significant error bars (10–20%) due to O being theweakest X-ray scatterer of the three. The thermal parameterof Na (Figure 2d) was found to be higher than that of Br(Figure 2e), but in both cases the changes are only verysmall (�1%) around 250 °C. So, just as there is no evidencefor abrupt behaviour in the lattice parameter variation,neither is there any sign in atomic displacements ofdiscontinuous behaviour across the temperature rangestudied.In order to investigate the local structure of Na3OBr, welooked at the pair distribution function (PDF) in the rangeof 0 Å to 5 Å (Figure 2f, full range is available in Figure S3).The majority of the peak positions shift to higher distanceswhile the widths broadens as the temperature is increased.This behaviour can be attributed to thermal expansion andthe thermal population of phonons. We note that an order-disorder transition and a phase change (melting of Na3OBr)are absent in this temperature range. From this synchrotronXRD study, we can deduce that the endothermic peak in theDSC and the non-Arrhenius behaviour of the ionic con-ductivity arise from neither the melting of Na3OBr nor anorder-disorder transition.Having excluded Na3OBr melting and order-disordertransition as the potential origin of non-Arrhenius behav-iour, we probed the local Na environment by static 23NaNMR spectroscopy. The spectra in Figure 3a show twocomponents, a broad (136 kHz at 11.7 T, 80 kHz at 23.5 T)quadrupolar peak and a narrower component at �0 ppm(shown in purple), which arises from impurities. The wide-line Na3OBr (orange) is deconvoluted to give δiso=44 ppm,XQ=11.9 MHz and ηQ=0, which are similar to the reportedvalues for Na3OCl.[32] The large quadrupolar peak arisesfrom the non-symmetric octahedral coordination of Na ionswith two O ions and four coplanar Br ions. Solid-state 23Namagic angle spinning (MAS, νR=20 kHz) NMR was per-formed to improve the resolution and deconvolute thenarrower component shown in Figure 3a and 3b. The 23NaMAS spectrum shows two Na2O2 resonances (purple at 9.5and 4.5 ppm), a minor Na2O resonance (blue at 53 ppm),and a broader quadrupole resonance attributed to NaOH(green at 38.5 ppm), shown in Figure 3c. The isotropic shiftof the attributed NaOH resonance is at a higher frequencythan previous work published on NaOH (21 ppm).[33]However, the quadrupole coupling constant (CQ=3.8 MHz)is similar to that of NaOH. This discrepancy in shift could bedue to the lack of periodicity, as NaOH will be neighbouringNa3OBr, Na2O2 and Na2O in the mixture. The presence ofNaOH is also detected by Raman as seen in Figure S5. Thetotal percentage of impurities is 5%, with an atomic ratiobetween Na2O2, Na2O, and NaOH of 40 :3 :57. This amountis substantial and yet no additional Bragg peaks weredetected in the XRD patterns. Therefore, the impuritiescontained in the sample are likely to be poorly crystalline,which could be a result of the 10 hours of ball milling at350 rpm carried out during the synthesis.[34,35]To understand how the Na environment changes withtemperature, we performed temperature-controlled static23Na NMR. As shown in Figure 3d, the intensity of theNa3OBr peak remains constant over the sampled temper-ature range, although there is a reduction in the quadrupolecoupling constant at elevated temperatures. This shows thatthe antiperovskite structure is maintained even after thetransition, but the mobility of the sodium ions inside thestructure increases slightly at these elevated temperatures.The variable temperature NMR also shows a significantintensity increase and peak width decrease in the narrowercomponent on heating above 230 °C.[36] These changes aretypically observed for melting. Therefore, our temperature-dependent NMR results indicate that the step increase inionic conductivity around 250 °C is likely not an intrinsicproperty of Na3OBr.Na2O2 is an impurity contained in our Na2O precursor. Itis very hygroscopic, so it could readily react with watermolecules inside the glovebox to form NaOH. Each of theimpurities detected by the MAS NMR is proposed to bethermodynamically stable to above 300 °C.[37] However, aAngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (4 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licenseeutectic between Na2O2, Na2O, and NaOH has beenreported at �270 °C.[38] Therefore, the endothermic peakseen in our DSC could correspond to the eutectic melting ofa solid phase consisting of Na2O2, Na2O, and NaOH. Cross-sectional EDX maps of our pellet after the variable temper-ature EIS (Figure 4) show small areas at the grain andparticle boundaries that are rich in oxygen and deficient inbromine. Further, in some areas (Figure 4e), eutectic-likelamellae are visible, which is a strong indication that anNa2O2-Na2O-NaOH eutectic mixture formed on cooling.Based on these findings, we propose that the melting of aeutectic mixture of Na2O2, Na2O, and NaOH is the mainfactor responsible for enhanced Na+ ionic mobility above250 °C.The importance of non-bulk factors in determining theoverall ionic conductivity of solid electrolytes has beenpreviously investigated, with the majority of work focusingon the effect of grain boundaries.[30,39–43] However, theFigure 3. The room temperature 23Na static WURST-echo NMR spectra(black) of Na3OBr at (a) 23.5 T (ν0=264.6 MHz) and (b) 20 T(ν0=224.9 MHz).The orange dotted lines represent the sum of thesimulated Na3OBr (orange) and impurity resonances (purple). (c) Highresolution MAS NMR (MAS, νR=20 kHz) of the sample; impurityresonances were deconvoluted into three components: Na2O2 (purple),Na2O (blue), and NaOH (green). (d) The variable temperature 23Nastatic NMR spectra of Na3OBr (ν0=132.32 MHz).Figure 4. (a) Cross-sectional secondary electron image of the Na3OBrpellet taken after variable temperature EIS measurements. Correspond-ing EDX analysis of the different elements: (b) Na (purple), (c) O (red),and (d) Br (green). Na is homogeneously distributed throughout thepellet but there are areas around some of the grain and particleboundaries which are rich in O and deficient in Br. The oxygen richareas are not uniformly distributed throughout the pellet. (e) Cross-sectional secondary electron image of the same pellet showingeutectic-like lamellae between Na3OBr particles (circled).AngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (5 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licenseinfluence of impurities and their associated interphaseboundaries is rarely considered. This work demonstrateshow small volume fractions of impurities can have asignificant impact on the macroscopic properties of a solidelectrolyte. While it has been established that the presenceof grain boundaries is detrimental to ionic conduction inmany antiperovskite electrolytes,[39,40,43] in this study theliquid phase that forms at the particle boundaries on heatingabove 250 °C has been found to improve the high-temper-ature ionic conduction. This is supported by the Arrheniusplot of bulk and boundary ionic conductivities shown inFigure S1c and Figure S1d. The change in bulk ionicconductivity is linear with temperature, with a constantactivation energy of 0.43 eV. On the contrary, there is a stepchange in the boundary ionic conductivity at 250 °C, thetemperature at which an endothermic peak was observed inthe DSC. This corresponds to a reduction in activationenergy from 0.51 eV to 0.31 eV. Despite improving the high-temperature ionic conduction, other properties of the liquidboundary phase such as an increased electronic conductivitymay ultimately render this secondary phase undesirable.[44]ConclusionWe synthesised Na3OBr using commercially available Na2Oand observed a non-Arrhenius change in ionic conductivityat a temperature of about 250 °C. This coincided with anendothermic peak measured by DSC. Bulk characterisationof this material using variable temperature in situ synchro-tron XRD did not show a detectable change in its crystalstructure, which suggests that this behaviour does notoriginate from a melting or order-disorder transition inNa3OBr as previously speculated. Further local character-isation using variable temperature in situ NMR revealedthat the sample contained a minor fraction of poorlycrystalline impurities consisting of Na2O, Na2O2, and NaOH.These impurities form a eutectic mixture which melts ataround 250 °C, promoting Na+ mobility. We highlight theeffects of precursor purity and the synthesis method on thefinal quality of the product. While the use of the sameprecursors and similar synthesis methods is common in priorliterature, the likelihood that these lead to the formation ofpoorly crystalline impurity phases which affect the electro-chemical performance of the material has not been under-stood until now. It is therefore advisable not to usecommercial Na2O, which contains �20% Na2O2, as a solid-state synthesis precursor. Ball-milling may reduce thecrystallinity of impurity phases, which could explain whythey are often overlooked in solid-state electrolyte charac-terisation studies. Therefore, if a mechanochemical synthesismethod is employed, it is important also to check for poorlycrystalline impurities. In light of these findings, the non-Arrhenius ionic conductivity behaviour seen in a range ofother solid electrolyte materials—particularly the Li-and K-rich antiperovskites—should be re-investigated. These find-ings will be of significant practical value to researchersattempting to optimise the synthesis of NaRAPs. Further-more, they improve understanding of the processing-struc-ture-properties relationships pertaining to NaRAPs andhighlight the need to determine these relationships in othermaterials systems, which will accelerate the development ofhigh-performance solid electrolytes.AcknowledgementsThis work was supported by the Faraday InstitutionSOLBAT project (grant numbers FIRG056). The UK High-Field Solid-State NMR Facility used in this research wasfunded by EPSRC, BBSRC (EP/T015063/1), and the Uni-versity of Warwick, as well as part funding throughBirmingham Science City Advanced Materials Projects 1and 2 supported by Advantage West Midlands (AWM), theEuropean Regional Development Fund (ERDF) and, forthe 1 GHz instrument, EP/R029946/1. The Science &Technology Facilities Council (SFTC) is aknowledged forI15-1 beamtime allocation CM31137 and I11 beamtimeallocation CY29776 at the Diamond Light Source, U.K. Theauthors acknowledge use of characterisation facilities withinthe David Cockayne Centre for Electron Microscopy,Department of Materials, University of Oxford, alongsidefinancial support provided by the Henry Royce Institute(Grant ref EP/R010145/1). B.D. acknowledges support fromthe Indonesia Endowment Fund for Education (LPDP)doctoral studentship. A.L.G. thanks the European ResearchCouncil for funding (Advanced Grant 788144).Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are availablein the supplementary material of this article.Keywords: non-Arrhenius behaviour · poorly crystallineimpurities · sodium antiperovskites · solid sodium-ionconductors · solid-state batteries[1] S. Wang, B. Peng, J. Lu, Y. Jie, X. Li, Y. Pan, Y. Han, R. Cao,D. Xu, S. Jiao, Chem. Eur. J. 2023, 29, e2022023.[2] J. Ma, S. Zheng, P. Das, P. Lu, Y. Yu, Z.-S. Wu, SmallStructures 2020, 1, 2000053.[3] S. Lilley, Faraday Insights 2021.[4] Z. Li, P. Liu, K. Zhu, Z. Zhang, Y. Si, Y. Wang, L. Jiao,Energy Fuels 2021, 35, 9063–9079.[5] L. Ye, M. Liao, T. Zhao, H. Sun, Y. Zhao, X. Sun, B. Wang, H.Peng, Angew. Chem. Int. Ed. 2019, 58, 17054–17060.[6] C. Wang, H. Wang, E. Matios, X. Hu, W. Li, Adv. Funct.Mater. 2018, 28, 1802282.[7] C. Yang, S. Xin, L. Mai, Y. You, Adv. Energy Mater. 2021, 11,2000974.[8] B. Sun, P. Li, J. Zhang, D. Wang, P. Munroe, C. Wang,P. H. L. Notten, G. Wang, Adv. Mater. 2018, 30, 1801334.AngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (6 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1021/acs.energyfuels.1c00347https://doi.org/10.1002/anie.201910202[9] Q. Shi, Y. Zhong, M. Wu, H. Wang, H. Wang, Angew. Chem.Int. Ed. 2018, 57, 9069–9072.[10] E. Ahiavi, J. A. Dawson, U. Kudu, M. Courty, M. S. Islam, O.Clemens, C. Masquelier, T. Famprikis, J. Power Sources 2020,471, 228489.[11] Y. Sun, Y. Wang, X. Liang, Y. Xia, L. Peng, H. Jia, H. Li, L.Bai, J. Feng, H. Jiang, J. Xie, J. Am. Chem. Soc. 2019, 141,5640–5644.[12] H. Nguyen, S. Hy, E. Wu, Z. Deng, M. Samiee, T. Yersak, J.Luo, S. P. Ong, Y. S. Meng, J. Electrochem. Soc. 2016, 163,A2165–A2171.[13] H.-x. Li, X.-y. Zhou, Y.-c. Wang, H. Jiang, Inorg. Chem. Front.2021, 8, 425–432.[14] C. Liang, W. Yang, R. Zou, C. Jin, Y. Zhao, Inorg. Chem.2016, 55, 5993–5998.[15] J. Zheng, B. Perry, Y. Wu, ACS Mater. Au 2021, 1, 92–106.[16] J. Zheng, H. Fang, L. Fan, Y. Ren, P. Jena, Y. Wu, J. Phys.Chem. Lett. 2021, 12, 7120–7126.[17] P. C. Tsai, S. Mair, J. Smith, D. M. Halat, P. H. Chien, K. Kim,D. Zhang, Y. Li, L. Yin, J. Liu, S. H. Lapidus, J. A. Reimer,N. P. Balsara, D. J. Siegel, Y. M. Chiang, Adv. Energy Mater.2023, 13, 2203284.[18] L. Gao, H. Zhang, Y. Wang, S. Li, R. Zhao, Y. Wang, S. Gao,L. He, H. F. Song, R. Zou, Y. Zhao, J. Mater. Chem. A 2020, 8,21265–21272.[19] Y. Li, P.-C. Tsai, F. Wang, S. Mair, D. Zhang, L. Yin, S.Lapidus, Y.-M. Chiang, ECS Meeting Abstracts 2021, MA2021-02, 43.[20] M. Jansen, C. Feldmann, W. Müller, Z. Anorg. Allg. Chem.1992, 611, 7–10.[21] Z. Mehmedović, V. Wei, A. Grieder, P. Shea, B. C. Wood, N.Adelstein, Philos. Trans. R. Soc. London Ser. A 2021, 379,20190459.[22] W. Feng, L. Zhu, X. Dong, Y. Wang, Y. Xia, F. Wang, Adv.Mater. 2023, 35, 2210365.[23] F. Wang, P.-c. Tsai, Y. Li, S. L. Eiler, S.-B. Son, Y.-M. Chiang,ECS Meeting Abstracts 2019, MA2019-02, 70.[24] J. Zhang, L. Wang, J. Zhu, Y. Zhao, Appl. Phys. Lett. 2020,117, 081904.[25] Y. Zhang, Y. Zhao, C. Chen, Phys. Rev. B 2013, 87, 134303.[26] J. A. Dawson, H. Chen, M. Saiful Islam, J. Phys. Chem. C2018, 122, 23978–23984.[27] F. Wang, Y. Li, P.-C. Tsai, L. Yin, H. Evens, K. Kim, S. H.Lapidus, D. J. Siegel, Y.-M. Chiang, ECS Meeting Abstracts2020, MA2020-02, 967.[28] J. A. Dawson, T. Famprikis, K. E. Johnston, J. Mater. Chem. A2021, 9, 18746–18772.[29] H. J. Lee, B. Darminto, S. Narayanan, M. Diaz-Lopez, A. W.Xiao, Y. Chart, J. H. Lee, J. A. Dawson, M. Pasta, J. Mater.Chem. A 2022, 10, 11574–11586.[30] J. Zhu, S. Li, Y. Zhang, J. W. Howard, X. Lü, Y. Li, Y. Wang,R. S. Kumar, L. Wang, Y. Zhao, Appl. Phys. Lett. 2016, 109,101904.[31] Y. Wang, Q. Wang, Z. Liu, Z. Zhou, S. Li, J. Zhu, R. Zou, Y.Wang, J. Lin, Y. Zhao, J. Power Sources 2015, 293, 735–740.[32] G. Klösters, M. Jansen, Solid State Nucl. Magn. Reson. 2000,16, 279–283.[33] K. J. MacKenzie, M. E. Smith, Multinuclear Solid-State NMRof Inorganic Materials, in Pergamon Materials Series, vol. 6,Pergamon 2002.[34] B. R. Ortiz, P. M. Sarte, A. H. Avidor, S. D. Wilson, Phys. Rev.Mater. 2022, 6, 104413.[35] M. Enterría, M. Reynaud, J. I. Paredes, L. Medinilla, R.Younesi, N. Ortiz-Vitoriano, J. Energy Chem. 2022, 68, 709–720.[36] A. R. Jones, R. Winter, P. Florian, D. Massiot, J. Phys. Chem.B 2005, 109, 4324–4332.[37] C. H. Wang, D. Y. Gui, Q. Xia, M. Avdeev, C. D. Ling, B. J.Kennedy, Inorg. Chem. 2020, 59, 14439–14446.[38] J.-i. Saito, M. Tendo, K. Aoto, Pseudo-ternary Phase Diagramin the Na2O� Na2O2-NaOH System, Technical report 1997.[39] E. Milan, M. Pasta, Mater. Futures 2023, 2, 013501.[40] J. A. Dawson, P. Canepa, T. Famprikis, C. Masquelier, M. S.Islam, J. Am. Chem. Soc. 2018, 140, 362–368.[41] X. Lü, G. Wu, J. W. Howard, A. Chen, Y. Zhao, L. L. Daemen,Q. Jia, Chem. Commun. 2014, 50, 11520–11522.[42] X. Lü, J. W. Howard, A. Chen, J. Zhu, S. Li, G. Wu, P.Dowden, H. Xu, Y. Zhao, Q. Jia, Adv. Sci. 2016, 3, 1500359.[43] J. A. Quirk, J. A. Dawson, Adv. Energy Mater. 2023, 13,2301114.[44] O. Gerbig, Defect Chemistry in Alkali Peroxides and Super-oxides, Ph.D. thesis, University of Stuttgart 2014.Manuscript received: September 26, 2023Accepted manuscript online: October 30, 2023Version of record online: November 14, 2023AngewandteChemieResearch ArticlesAngew. Chem. Int. Ed. 2023, 62, e202314444 (7 of 7) © 2023 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2023, 51, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202314444 by National Institute For, Wiley Online Library on [02/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1002/anie.201803049https://doi.org/10.1002/anie.201803049https://doi.org/10.1016/j.jpowsour.2020.228489https://doi.org/10.1016/j.jpowsour.2020.228489https://doi.org/10.1021/jacs.9b01746https://doi.org/10.1021/jacs.9b01746https://doi.org/10.1149/2.0091610jeshttps://doi.org/10.1149/2.0091610jeshttps://doi.org/10.1039/D0QI00921Khttps://doi.org/10.1039/D0QI00921Khttps://doi.org/10.1021/acsmaterialsau.1c00026https://doi.org/10.1021/acs.jpclett.1c01807https://doi.org/10.1021/acs.jpclett.1c01807https://doi.org/10.1039/D0TA07110Bhttps://doi.org/10.1039/D0TA07110Bhttps://doi.org/10.1149/MA2021-02143mtgabshttps://doi.org/10.1149/MA2021-02143mtgabshttps://doi.org/10.1002/zaac.19926110502https://doi.org/10.1002/zaac.19926110502https://doi.org/10.1149/MA2019-02/2/70https://doi.org/10.1021/acs.jpcc.8b08208https://doi.org/10.1021/acs.jpcc.8b08208https://doi.org/10.1149/MA2020-025967mtgabshttps://doi.org/10.1149/MA2020-025967mtgabshttps://doi.org/10.1039/D1TA03680Ghttps://doi.org/10.1039/D1TA03680Ghttps://doi.org/10.1039/D2TA01462Ahttps://doi.org/10.1039/D2TA01462Ahttps://doi.org/10.1016/j.jpowsour.2015.06.002https://doi.org/10.1016/S0926-2040(00)00079-5https://doi.org/10.1016/S0926-2040(00)00079-5https://doi.org/10.1016/j.jechem.2021.12.014https://doi.org/10.1016/j.jechem.2021.12.014https://doi.org/10.1021/jp045705shttps://doi.org/10.1021/jp045705shttps://doi.org/10.1021/acs.inorgchem.0c02234https://doi.org/10.1088/2752-5724/aca703https://doi.org/10.1021/jacs.7b10593