# Fileset

[Advanced Science - 2026 - Setiawan - Mixed‐Coordination Electrolytes With Molecular Additives for Robust Interphases in.pdf](https://mdr.nims.go.jp/filesets/b10ef59c-84be-4b4c-baf5-d50080df3699/download)

## Creator

Dedy Setiawan, [Toshihiko Mandai](https://orcid.org/0000-0002-2403-7794)

## Rights

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

## Other metadata

[Mixed‐Coordination Electrolytes With Molecular Additives for Robust Interphases in High‐Voltage Rechargeable Magnesium Batteries](https://mdr.nims.go.jp/datasets/18ffa4cb-f04d-418b-80a0-4c57bf290832)

## Fulltext

Mixed‐Coordination Electrolytes With Molecular Additives for Robust Interphases in High‐Voltage Rechargeable Magnesium BatteriesAdvanced Science www.advancedscience.comRESEARCH ARTICLEMixed-Coordination Electrolytes With Molecular Additives for Robust Interphases in High-Voltage Rechargeable Magnesium Batteries Dedy Setiawan Toshihiko Mandai Functional Electrolyte Synthesis Team, Research Center for Energy and Environmental Materials (GREEN), National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan Correspondence: Dedy Setiawan ( SETIAWAN.Dedy@nims.go.jp) Toshihiko Mandai ( MANDAI.Toshihiko@nims.go.jp) Received: 23 March 2026 Revised: 1 May 2026 Accepted: 11 May 2026 Keywords: anion coordination | electrolyte | interphase | magnesium batteries ABSTRACT Rechargeable magnesium batteries (RMBs) offer a low-cost and high-capacity alternative to the current energy storage systems, yet achieving high-energy density and stable operation at high-voltage remains difficult, particularly due to oxidative decomposition in conventional ether-based electrolytes. In this work, we introduce a mixed-coordination electrolyte (MCE) combined with a molecular additive (MCE-MA) designed to promote anion-derived interphases and enhance cycling stability in high-voltage RMBs. The MCE formulation integrates dissociative and associative coordinating salts to optimize both Mg2 + transport and interphase formation, while the molecular additive facilitates the development of uniform, robust interphases. MCE-MA exhibits improved Mg plating/stripping efficiency in Cu|Mg asymmetric cells and sustains over 250 h of stable cycling in Mg|Mg symmetric cells, significantly outperforming conventional halide-free ether-based electrolytes. Furthermore, MCE-MA enables long-term cycling of a full cell with oxide-based cathode, achieving 200 cycles at 100 mA g− 1 . X-ray photoelectron spectroscopy (XPS) and time-of- flight secondary ion mass spectrometry (ToF-SIMS) were employed systematically to reveal the origin of enhanced performance. 1T  h  o  a  m  a  a  a  r  R  s            To©Ah Introduction he growing global reliance on lithium-ion batteries (LIBs)as raised concerns over the long-term availability and costf lithium resources [ 1 ]. This has stimulated the search forlternative energy-storage systems based on earth-abundant ele-ents. Among these, rechargeable magnesium batteries (RMBs)re promising candidates owing to the low cost and naturalbundance of magnesium, combined with its high volumetricnd gravimetric capacities (3833 mAh cm− 2 and 2205 mAh g− 1 ,espectively). When coupled with high-energy-density cathodes,MBs could, in principle, deliver energy densities comparable totate-of-the-art LIBs [ 2 ]. his is an open access article under the terms of the Creative Commons Attribution Licenriginal work is properly cited. 2026 The Author(s). Advanced Science published by Wiley-VCH GmbH dvanced Science , 2026; 13:e75749 ttps://doi.org/10.1002/advs.75749Despite significant progress, the practical realization of high-energy RMBs remains hindered by long-standing challengesin electrolyte design, particularly when operated at high-voltage [ 3–5 ]. Electrolytes based on weakly coordinatinganion salts (WCA) such as Mg[B(hfip)4 ]2 and Mg[Al(hfip)4 ]2 (hfip:hexafluoroisopropanol) in ether solvents exhibit excellentreductive stability and highly reversible Mg plating/stripping( > 99% efficiency for 0.3 m Mg[Al(hfip)4 ]2 in diglyme), owing toweak cation–anion interactions that suppress surface passivationon the Mg anode [ 6–10 ]. However, the cycle performance ofhigh-voltage RMBs with this electrolyte system remains poor,which is mainly attributed to the high fraction of uncoordinatedether molecules that are readily exposed at the cathode interfacese, which permits use, distribution and reproduction in any medium, provided the 1 of 13http://www.advancedscience.comhttps://doi.org/10.1002/advs.75749mailto:SETIAWAN.Dedy@nims.go.jpmailto:MANDAI.Toshihiko@nims.go.jphttp://creativecommons.org/licenses/by/4.0/https://doi.org/10.1002/advs.75749http://crossmark.crossref.org/dialog/?doi=10.1002%2Fadvs.75749&domain=pdf&date_stamp=2026-05-25FIGURE 1 (a) Schematic illustration of issues in WCA-based electrolyte and the introduction of MCE-MA to enable robust interphases formation. (b) Molecular structure of Mg[B(hfip)4 ]2 ∙3DME salt, Mg[B(tfe)4 ]2 ∙DME salt, and BisTFE molecule. Hydrogen and disordered atoms are not shown for figure clarity. (c) Raman spectra of the as-prepared electrolytes in the anion-sensitive spectral region, together with peak deconvolution analysis. The percentages indicate the relative contributions of solvent-separated ion pair (SSIP) and pseudo ion pair (pIP) in each electrolyte. (d) Raman spectra of as-prepared electrolytes in the selected region sensitive to solvent with assigned free and coordinated diglyme (G2). a  A  a  c  c  A  s  sI  m  i  v  u  M  t  i  A  a  i  e  FA  b  a  s                    2 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatind undergo oxidative decomposition at high voltage [ 3, 11–13 ].s a result, issues related to solvent-derived decomposition,ctive proton formation, interphase instability, and cathodeomponent dissolution were recently discovered, and are alsoommonly found in the current LIBs research [ 3, 4, 14–16 ].lthough employing high-concentration electrolytes has shownome promise in LIBs as well as RMBs, the relatively high cost ofome Mg salts makes this approach suboptimal [ 17, 18 ]. n this study, we introduce a WCA-based electrolyte incorporatingixed anionic species and a molecular additive to promote stablenterphase formation for extended cycle performance of high-oltage RMBs. Specifically, Mg[B(hfip)4 ]2 in diglyme (G2) wassed as the base electrolyte, while a certain concentration ofg[B(tfe)4 ]2 salt (tfe:trifluoroethanol) was added to modulatehe solvation structure and help the formation of anion-derivednterphase, forming a mixed coordination electrolyte (MCE).dditionally, Bis(2,2,2-trifluoroethyl) ether (BisTFE) was chosens a molecular additive to promote homogeneous and robustnterphase. The resulting mixture, termed mixed-coordinationlectrolyte with molecular additive (MCE-MA), is summarized inigure 1a . lthough both Mg[B(hfip)4 ]2 and Mg[B(tfe)4 ]2 salts containulky weakly coordinating anions, their cation–anion inter-ctions differ significantly. In Mg[B(hfip)4 ]2 , the coordinationtructure is fully dissociative, where Mg2 + –glyme complexes areof 13separated from [B(hfip)]− anions (Figure 1b ) [ 6 ]. The strongelectron-withdrawing hfip groups suppress anion coordination,while the strong electric field of Mg2 + enhances glyme solva-tion, resulting in a glyme-dominated solvation shell [ 12 ]. Incontrast, Mg[B(tfe)4 ]2 salt has an associative coordination struc-ture (Figure 1b ), where Mg2 + binds to six oxygen atoms fromtwo [B(tfe)4 ]− anions and one glyme molecule. This strongercation–anion association arises from the weaker charge disper-sion of [B(tfe)4 ]− , as the tfe group contains only one ─CF3 group compared to two in the hfip group [ 19 ]. Consequently,Mg[B(tfe)4 ]2 coordinates fewer glyme molecules (one vs. three inMg[B(hfip)4 ]2 ). The combination of these two salts is expected tobalance efficient Mg2 + transport with anion-derived interphaseformation, while the addition of BisTFE, which has a linearfluorinated structure (Figure 1b ), further promotes a uniforminterphase. 2 Results and Discussion 2.1 Electrolyte Solvation Structure The synthesis of Mg salts and the preparation of electrolytes aredetailed in the Experimental Section. Both Mg[B(hfip)4 ]2 and Mg[B(tfe)4 ]2 form adducts with coordinated DME(dimethoxyethane) used during the synthesis, as confirmedby 1 H NMR (Figure S1 ), consistent with previous reports [ 9, 19 ].Advanced Science, 2026ve Commons LicenseT  d  r  5  t  t  i  c  a  bT  c  a  f  M  a  r  a  i  d  (  (  i  (  a  e  M  (  i  p  t  a a  s  a  i  d  [  r  t  s  e  mI  c  M  s  a  w  t  w  B  i  i  a  c  p  a  t                                                       A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatihe base electrolyte was prepared using 0.3 m Mg[B(hfip)4 ]2 iniglyme (G2), a concentration selected based on its previouslyeported electrochemical performance [ 9 ]. To formulate the MCE, mM Mg[B(tfe)4 ]2 was included in the base electrolyte, whilehe MCE-MA was obtained by introducing 1 vol% BisTFE intohe MCE. The detailed electrolyte compositions are summarizedn Table S1 . The concentration of Mg[B(tfe)4 ]2 and BisTFE washosen after initial screening with a symmetric Mg|Mg cell and full-cell test. 5 mM Mg[B(tfe)4 ]2 and 1 vol% BisTFE providealanced performance in both electrochemical tests. o investigate the influence of mixing WCA salts with a distinctoordination behavior, as well as the effect of a moleculardditive, Raman spectra were collected at room temperatureor the base electrolyte, base + BisTFE (MA), MCE, and MCE-A. The regions sensitive to anion and solvent coordinationre shown in Figure 1c,d , respectively. In the 720–740 cm− 1egion (Figure 1c ), corresponding to anion coordination, MCEnd MCE-MA exhibit a blueshift relative to the base electrolyte,ndicating the formation of more “pseudo ion-pair” clusters,espite a minimum amount of Mg[B(tfe)4 ]2 being incorporatedonly 5 mM) [ 20–22 ]. According to the peak deconvolutionFigure 1c ), the base electrolyte is dominated by solvent-separatedon pairs (SSIP), with partial contribution from pseudo ion pairpIP). Here, the term “pseudo ion pair (pIP)” is used to describe weak cation–anion association characteristic of WCA-basedlectrolytes, in which the anion resides in close proximity tog2 + without forming a classical inner-sphere contact ion pairCIP). Although Mg[B(hfip)4 ]2 is widely regarded as prototyp-cal WCA-salt, recent studies have reported and discussed theresence of such cation-anion association [ 10, 23 ]. Notably, inhe MCE and MCE-MA, the contribution of this cation-anionssociation becomes more dominant (Figure 1c ). As Mg[B(tfe)4 ]2lone is known to favor pIP in ethers, introducing even amall fraction of this salt into the base electrolyte significantlylters the equilibrium between cation–solvent and cation–anionnteractions. The [B(tfe)4 ]− anion possesses a higher local chargeensity and weaker electron-withdrawing ability compared toB(hfip)4 ]− , making it more nucleophilic toward Mg2 + . As aesult, it can outcompete glyme for coordination sites, promotinghe formation of pIP and partially disrupting the dissociativeolvation structure characteristic of the base electrolyte. Thisnhanced anion participation reduces the availability of free G2olecules and weakens Mg2 + –G2 interactions. n the 820–900 cm− 1 region (Figure 1d ), associated with solventoordination, the coordinated G2 peak intensity is reduced inCE and MCE-MA compared to the base electrolyte. Thisuggests a weakening of Mg2 + –G2 interactions due to increasednion participation in the pIP formation, in good agreementith the anion-sensitive region peaks [ 24 ]. In the MCE-MA,he further decrease in free G2 intensity likely arises fromeak but non-negligible interactions between Mg2 + and theisTFE molecule. While BisTFE is generally non-coordinating,ts electron-withdrawing CF3 groups can polarize the surround-ng solvation shell, subtly stabilizing Mg2 + –anion complexesnd homogenizing the overall solvation environment [ 25–27 ]. Inontrast, addition of BisTFE alone to the base electrolyte does noterturb the solvation structure substantially because, without thessociative anion [B(tfe)4 ]− , there is no competitive driving forceo reorganize the coordination equilibrium. dvanced Science, 20262.2 Anion-Derived Solid Electrolyte Interphase Modulation of the Mg2 + solvation environment is ultimatelymanifested in the chemical composition of the solid electrolyteinterphase (SEI) formed on Mg metal. In particular, the formationof pIP-rich coordination structures in the MCE and MCE-MAelectrolytes is expected to shift the initial interfacial reac-tions from solvent-dominated decomposition toward preferentialanion participation. To elucidate how these solvation changestranslate into interphase chemistry, ex situ X-ray photoelectronspectroscopy (XPS) was performed on Mg metal electrodesrecovered after 50 h cycled in a Mg|Mg symmetric cell at 0.5 mAcm− 2 and 0.25 mAh cm− 2 (Figure S2a ) using the three electrolytes(Figure 2a–d ). The C 1s spectra (Figure 2a ) reveal the presence of organic surfacespecies and fluorinated carbonaceous components (C ─F) for allelectrolytes. Notably, MCE-MA exhibits the highest C ─F peakintensity, indicating enhanced formation of fluorocarbons andrelated species, which are attributed to the controlled decom-position of fluorinated anions and/or the BisTFE additive. Thisobservation already suggests a transition toward an anion-derivedSEI chemistry in the mixed-coordination systems. Consistentwith this trend, the Mg 2p spectra (Figure 2b ) show contributionsfrom Mg ─F species ( ≈ 52.0 eV) coexisting with Mg ─O–relatedcomponents arising from oxides, hydroxides, and/or carbonates( ≈ 50.9 eV) [ 28 ]. Both MCE and MCE-MA display a substan-tially higher Mg ─F fraction than the base electrolyte, furtherevidencing a stronger anion-derived interphase. Interestingly, asmall metallic Mg0 signal is detected in the MCE sample but isabsent in MCE-MA, implying that the SEI formed in MCE-MA ismore homogeneous and likely provides more complete surfacecoverage. The F 1s spectra (Figure 2c ) provide direct insightinto fluorine-containing interphase species. A dominant peak at∼ 689 eV corresponds to C ─F bonds and is most pronounced inMCE-MA, in agreement with the C 1s analysis [ 29 ]. In addition,a shoulder at ∼ 685 eV assigned to Mg ─F is observed in allelectrolytes but is significantly intensified in MCE, suggestingsubstantial anion-derived inorganic fluoride formation. Giventhe boron-centered anions employed, a B ─F contribution near∼ 688 eV was also included in the peak fitting. The enhancedB ─F intensity observed for both MCE and MCE-MA relative tothe base electrolyte further confirms the active involvement ofanion decomposition in shaping the SEI [ 30, 31 ]. This conclusionis reinforced by the B 1s spectra (Figure 2d ), which show markedlystronger B ─O ( ∼ 193 eV) and B ─F ( ∼ 195 eV) signals for MCE andMCE-MA than for the base electrolyte, consistent with increasedboron-containing decomposition products within the interphase[ 32 ]. To provide a more quantitative comparison, the fitted peak arearatios of the Mg 2p and F 1s spectra are summarized in Figure 2e,f .As shown in Figure 2e , Mg ─F species become the dominant Mg-containing component in both MCE (94%) and MCE-MA (68%),in sharp contrast to the base electrolyte (39%). Furthermore,the fluorinated interphase composition (Figure 2f ) is dominatedby C ─F species across all samples: base electrolyte (71%), MCE(51%), and MCE-MA (71%). On the other hand, B ─F contributionsare significantly more pronounced in the mixed-coordinationelectrolytes, MCE (22%) and MCE-MA (9%), reflecting enhancedand controlled anion participation during SEI formation. Overall,3 of 13ve Commons LicenseFIGURE 2 Ex situ XPS survey spectra of Mg metal cycled in Mg|Mg symmetric cell (0.5 mA cm− 2 , 0.25 mAh cm− 2 ) for 50 h using base electrolyte, MCE, and MCE-MA; (a) C 1s, (b) Mg 2p, (c) F 1s, and (d) B 1s. (e,f) The fitted peak area ratio of (e) Mg 2p and (f) F 1s provides quantitative results of XPS survey spectra. b  S  h  aB  o  r  g  (  m  t  1  d  f  t  e  h  s  u2PT  p  t                             4 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatioth MCE and MCE-MA promote the formation of anion-derivedEI layers; however, MCE-MA likely has a more robust andomogeneous SEI structure, due to the incorporation of BisTFEdditive. isTFE is expected to exhibit preferential reductive and/orxidative decomposition at the electrode interface due to itselatively labile C ─O bonds and electron-withdrawing ─CF3roups. Such decomposition can generate fluorinated speciese.g., C ─F ─containing fragments), which contribute to the for-ation of a fluorine-rich interphase. This is consistent withhe enhanced C ─F signals observed in the XPS (C 1s and Fs) spectra, where fluorinated fragments are more uniformlyistributed in the MCE-MA system. In addition to contributingluorinated species, the presence of BisTFE is also consideredo modulate interfacial reactions by altering the local solvationnvironment near the electrode surface. This can lead to moreomogeneous decomposition pathways, suppressing localizedolvent degradation and promoting the formation of a moreniform and stable interphase layer [ 33, 34 ]. .3 Impact of SEI on Mg Plating/Stripping erformance o evaluate the impact of SEI composition on the long-term Mglating/stripping stability, symmetric Mg|Mg cells were furtherested in the three electrolytes at 0.5 mA cm− 2 and 0.25 mAh cm− 2of 13for longer hours. For the base electrolyte, repeated polarization(activation process) is required to initiate Mg plating/stripping(Figure S2b ), reflecting the highly insulating nature of the Mgmetal surface, as also observed previously [ 35 ]. In contrast,such behavior is absent in the MCE and MCE-MA electrolytes,suggesting more favorable initial interfacial conditions for Mgelectrochemistry. Despite these differences, all electrolytes exhibitcomparable voltage profiles during the initial cycling period(Figure 3a ), indicating similar interfacial kinetics at the earlystage. However, pronounced differences emerge upon prolongedoperation. Notably, the MCE-MA electrolyte sustains stable Mgplating/stripping for more than 250 h, whereas both the baseelectrolyte and MCE suffer from soft short-circuiting withinapproximately 100 h. It is worth noting that the Mg plat-ing/stripping stability in the base electrolyte and the MCE isnearly identical. This observation indicates that the MCE alonedoes not significantly improve Mg plating/stripping homogene-ity, despite the formation of an anion-derived interphase, likelydue to the inhomogeneous nature of the resulting interphase. Incontrast, the incorporation of the molecular additive in MCE-MA promotes a more homogeneous interphase, thereby enablingmarkedly enhanced and prolonged Mg plating/stripping stability.The evolution of Mg plating/stripping overpotential providesfurther insight into the origin of these stability differences. Assummarized in Figure 3b , all three electrolytes display similarlylow overpotentials ( < 0.1 V) during the initial cycling stage ( < 50cycles). Upon extended cycling, the overpotentials of the baseAdvanced Science, 2026ve Commons LicenseFIGURE 3 (a) The discharge-charge profile of symmetric Mg|Mg cells with base electrolyte, MCE, and MCE-MA, cycled at 0.5 mA cm− 2 and 0.25 mAh cm− 2 . (b) Evolution of Mg plating/stripping overpotential as a function of cycle number in the symmetric cell. The overpotential is determined as the voltage plateaus of Mg plating/stripping at the end of the cycling test. (c,d) The magnified discharge/charge profile of the symmetric cell at (c) intermediate cycles, (d) later cycles. (e) Nyquist plot of EIS measurement of symmetric Mg|Mg cell with MCE-MA electrolyte taken at different stages: before cycle, 10th cycle, 30th cycle, and 100th cycle. (f–h) SEM image of Mg metal surface after cycling in (f) base electrolyte, (g) MCE, and (h) MCE-MA. e  a  c  v  n  o  c  m  c  I  t  d  t  i  w  S  cT  s  M  d                      A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatilectrolyte and MCE gradually increase from ∼ 0.1 to ∼ 0.15 Vfter ∼ 90 cycles, followed by a sudden drop in polarization asso-iated with soft short-circuit behavior (Figure 3c ). Such abruptoltage collapse is characteristic of internal shorting caused byon-uniform Mg deposition [ 36, 37 ]. In contrast, although theverpotential of MCE-MA also increases to ∼ 0.15 V after ∼ 100ycles, no sudden voltage drop is observed. Instead, MCE-MAaintains a stable overpotential of ∼ 0.2 V over more than 200ycles (Figure 3d ), demonstrating markedly enhanced durability.t is noteworthy that the Mg plating/stripping overpotential in allhree electrolytes exhibits only a weak dependence on currentensity over the range of 0.25–2.0 mA cm− 2 , indicating thathe use of MCE and MCE-MA does not substantially alter thentrinsic Mg plating/stripping kinetics. In contrast, all cells failhen subjected to a high current density of 4.0 mA cm− 2 (Figure3 ), suggesting a common limitation under aggressive operatingonditions rather than electrolyte-specific kinetic effects. he Nyquist plots (Figure 3e ) from electrochemical impedancepectroscopy (EIS) measurement of a symmetric Mg|Mg cell withCE-MA reveal a clear evolution of the interfacial resistanceuring cycling. The initial impedance before cycling is relativelydvanced Science, 2026high, which can be attributed to the absence of a well-formedinterphase. After 10 cycles, the interfacial resistance decreases sig-nificantly, indicating the formation of an electrochemically activeand ion-conductive interphase. Upon further cycling (30th and100th cycles), the impedance gradually increases, which is con-sistent with the growth and stabilization of the interphase layerduring prolonged operation. Importantly, despite this increaseat later stages, the overall impedance remains stable withoutan abrupt rise, in agreement with the sustained electrochemicalperformance observed for the MCE-MA electrolyte. This behaviorsupports the formation of a robust and stable interphase thateffectively regulates Mg plating/stripping over extended cycling. We also observed that these differences in electrochemical sta-bility are closely correlated with the morphology of Mg metalafter cycling. SEM images of Mg electrodes retrieved from thesymmetric cells are shown in Figure 3f–h . After cycling in thebase electrolyte, the Mg surface becomes highly mossy andcontains pronounced localized pits, indicative of uneven, 3D Mgdeposition (Figure 3f ). In the case of MCE, the surface is generallysmoother; however, distinct localized holes remain, suggestingspatially heterogeneous Mg plating/stripping arising from a non-5 of 13ve Commons Licenseu  e  p  O  t  M  a  l  g  eA  t  e  e  C  o  t  2  t  S  o  f  B  t  a  a  c  C  M  h  e  M  r  M  h  tB  s  e  r  C  o  m  s  s  b  e  u  r  a  d  c  rA  o  p  (  m                                                   6 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatiniform SEI (Figure 3g ). By contrast, Mg cycled in MCE-MAxhibits a relatively compact and homogeneous surface mor-hology, with significantly suppressed pit formation (Figure 3h ).verall, the electrochemical and morphological results indicatehat the superior long-term Mg plating/stripping stability inCE-MA originates from the formation of a more homogeneousnd robust SEI. This stabilized interphase effectively mitigatesocalized current concentration and suppresses non-uniform Mgrowth, thereby preventing premature short-circuiting duringxtended cycling. part from Mg plating/stripping stability, we also discoveredhat the three electrolytes exhibited different plating/strippingfficiency trends. The Mg plating/stripping efficiency in the baselectrolyte, MCE, and MCE-MA was examined in asymmetricu|Mg cells at a current density of 1 mA cm− 2 and areal capacityf 1 mAh cm− 2 . All three electrolytes exhibited comparable short-erm stability, with short-circuiting observed after approximately5 h (Figure S4a ). However, the initial Mg nucleation overpoten-ial revealed different behavior between the electrolytes (Figure4c ). The MCE and MCE-MA display lower Mg nucleationverpotential compared to the base electrolyte, indicating moreavorable Mg nucleation and improved interfacial reversibility.esides, their Coulombic efficiency (CE) profiles revealed distinctrends. In the initial cycle, MCE showed the highest CE (96.67%)nd the lowest overpotential, outperforming the base (89.55%)nd MCE-MA (93.13%). By the 14th cycle (before the short-ircuit), however, MCE-MA surpassed both systems, achieving aE of 94.84% with the lowest overpotential, while the base andCE stabilized at 92.28% and 91.30%, respectively. These resultsighlight the plausible differences in SEI evolution among thelectrolytes, driven by variations in their solvation structures. InCE-MA, SEI formation proceeds more gradually but yields aobust and stable interphase after repeated cycling. Conversely,CE promotes rapid SEI formation during the initial cycles;owever, the resulting interphase appears less resilient, leadingo performance degradation over time. eyond SEI chemistry, the electrolyte composition also exerts atrong influence on Mg deposition behavior, as reflected by differ-nces in the nucleation overpotential (Figure S5 ). To examine theesulting Mg plating morphology, Mg was electrodeposited ontou foil at a current density of 0.25 mA cm− 2 with an areal capacityf 1.0 mAh cm− 2 , followed by SEM analysis. The correspondingorphologies for the base electrolyte, MCE, and MCE-MA arehown in Figure S6 . In the base electrolyte, Mg deposits consist ofmall, sparsely distributed nuclei, indicative of a high nucleationarrier. In contrast, Mg deposited from the MCE and MCE-MAlectrolytes exhibits larger feature sizes and a more compact,niform morphology. This trend is consistent with previouslyeported correlations between reduced nucleation overpotentialnd increased metal deposit size [ 38, 39 ]. Such compact Mgeposition is advantageous for minimizing localized currentoncentration and suppressing undesired anode–electrolyte sideeactions during repeated charge–discharge cycling. dditionally, XRD analysis was also conducted for the Mg depositbtained from the base electrolyte and MCE-MA. The diffractionatterns reveal a noticeable change in preferred orientationFigure S14 ): the (002) and (101) reflections become slightlyore intense relative to the (100) peak when MCE–MA isof 13used. This suggests that MCE–MA promotes crystal growthalong the c -axis of the hexagonal close-packed Mg lattice. Suchbehavior can be attributed to additive-induced modifications ofthe electrode surface energy and the adsorption environment atthe metal–electrolyte interface, which can selectively stabilizecertain crystallographic planes during nucleation and growth[ 40 ]. However, the detailed mechanism of this preferred (002) and(101) crystal growth requires further investigation and is beyondthe scope of this study. 2.4 Insight Into High-Voltage RMBs Application To evaluate the applicability of the developed electrolyte systemsfor high-voltage RMBs, linear sweep voltammetry (LSV) wasperformed first to provide insight into their oxidative stabilities.Figure 4a shows the LSV profiles with platinum (Pt) as the work-ing electrode. Both MCE and MCE-MA exhibit slightly enhancedoxidation stability compared with the base electrolyte. Thisimprovement can be attributed to thermodynamic stabilizationarising from the modified solvation structure in the mixed-anion environment. In MCE and MCE-MA, the coexistenceof [B(hfip)4 ]− and [B(tfe)4 ] − anions—along with the weaklycoordinating BisTFE molecule in MCE-MA—modulates the localcoordination sphere around Mg2 + . This mixed coordinationweakens the strong Mg2 + –G2 interactions dominant in the baseelectrolyte, leading to a more delocalized charge distribution andimproved oxidative stability [ 41 ]. As shown in Figure 4b , small oxidation features appear below3.3 V vs. Mg/Mg2 + for MCE and MCE-MA, which are attributedto localized, self-limiting cathode electrolyte interphase (CEI)formation. The mixed-anion formulation slightly raises the bulkoxidative stability; however, local interfacial heterogeneities—such as regions enriched in [B(tfe)4 ]− or BisTFE and high-fieldsites on the electrode—can still host species with higher highestoccupied molecular orbital (HOMO) energies that oxidize pref-erentially. Oxidation of these localized species, likely involving[B(tfe)4 ]− or BisTFE, generates reactive radicals that decomposeinto boron- and fluorine-containing CEI components. The limitedcharge associated with these peaks and the rapid current decayindicate the formation of a thin, passivating interphase thatsuppresses further oxidation. While LSV provides a comparative assessment of the oxidativestability of the electrolytes, it primarily reflects dynamic elec-trochemical behavior. Therefore, the long-term stability of thecathode–electrolyte interphase (CEI) is more reliably evaluatedthrough galvanostatic full-cell cycling with a cathode material at areliable test condition. In this regard, the consistency between theLSV results and the extended cycling performance of the full cellshould support the enhanced oxidative stability and interphaserobustness of the MCE-MA electrolyte. Considering these oxidative stability results, two key conditionswere established for assembling RMBs full cells: (1) operationat room temperature, due to the relatively low boiling point ofBisTFE (62◦C–63◦C); and (2) a charge cut-off voltage around3.3 V to minimize parasitic electrolyte decomposition (Figure 4c ).Based on these criteria, FeV3 O9 ⋅1.1H2 O (FeVO) was selected asa model cathode material because of its stable operation at theAdvanced Science, 2026ve Commons LicenseFIGURE 4 (a) LSV profile of the base electrolyte, MCE, and MCE-MA measured using a Pt working electrode. (b) Magnified view of the LSV profile over the selected high voltage region. (c) Schematic illustration of high-voltage RMBs comprising base electrolyte, MCE, or MCE-MA with certain condition requirements. (d–f) Three representative cycles of galvanostatic discharge-charge profile of RMBs comprising FeVO cathode with (d) base electrolyte, (e) MCE, (f) MCE-MA at 10 mA g− 1 . (g) Cycling performance of RMBs with FeVO cathode in different electrolyte systems at 100 mA g− 1 after 1 cycle conditioning at 10 mA g− 1 . The Coulombic efficiency was defined as Qcharge /Qdischarge × 100%. (h) Selected discharge-charge profiles highlighting the failure of the RMBs with base electrolyte and MCE. d  o  v  w  t  dF  c  M  a  a  d  b  1  g  c  m  c                   A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatiesired voltage range [ 42, 43 ]. Additionally, the intrinsic stabilityf molecular H2 O molecules in FeVO has been systematicallyerified in previous studies, ensuring safe operation when pairedith Mg metal anodes [ 43, 44 ]. The FeVO cathode was syn-hesized following earlier protocol, and was confirmed by X-rayiffraction (XRD) and SEM–EDX analyses (Figure S7 ) [ 43 ]. igure 4d–f presents the charge–discharge profiles of RMBs fullells with FeVO cathode at 10 mA g− 1 using base electrolyte,CE, and MCE-MA. The applied current corresponds to anreal current density of approximately 0.01 mA cm− 2 for the Mgnode. Despite identical testing conditions, clear performanceifferences are observed among the electrolytes. The cell withase electrolyte delivers the highest first discharge capacity of97 mAh g− 1 but suffers from rapid capacity fading to 150 mAh− 1 (76% of the first discharge capacity) after three cycles. Theell with MCE delivers the lowest first discharge capacity of 142Ah g− 1 and degrades to 121 mAh g− 1 (85% of the first dischargeapacity) after three cycles. The cell with MCE-MA delivers advanced Science, 2026first discharge capacity of 166 mAh g− 1 and has relatively morestable cycling behavior compared to the base electrolyte andMCE, with the third discharge capacity of 149 mAh g− 1 (90%of the first discharge capacity). To elucidate the origin of thecapacity discrepancy between different electrolyte systems, ICP–MS analysis was performed to quantify Mg content in the FeVOcathode after discharge (Figure S8 ). The fraction of dischargecapacity attributed to Mg intercalation differs across electrolytes:41% for the base electrolyte, 65% for MCE, and 64% for MCE-MA.These results indicate that parasitic side reactions, likely protoninsertion, are more pronounced in the base electrolyte, despitethe low water content. As reported previously, such reactionsoccur during both discharge and charge, consistent with thehigher charge capacity relative to discharge capacity observedafter the second and third cycles in the base electrolyte [ 3 ].Solvent-derived decomposition not only reduces efficiency butalso leads to interphase crossover, disrupting the uniformity ofMg plating/stripping at the anode and thereby compromisinglong-term cycling stability [ 3 ] 7 of 13ve Commons LicenseS  d  f  R  a  s  e  d  c  f  (U  s  c  (  e  c  f  n  o  i  s  o  o  i  e  t  l  w  2  m  g  tA  e  n  t  l  c  a  e  s  I  t  c  i  i  hI  M  c  p  t  t  S  t  l  d                                                      8 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatiince MCE and MCE-MA demonstrated the formation of anion-erived interphases in the Mg anode, the impact on the RMBsull cell should be investigated. The cycling performance of theMB full cell with the FeVO cathode was further evaluated at higher current density of 100 mA g− 1 . Direct application ofuch high current densities led to poor cell performance for alllectrolytes (not shown here), likely due to the insufficientlyeveloped interphase discussed earlier. To address this, the fullells were preconditioned at a lower current density (10 mA g− 1 )or one formation cycle before subsequent cycling at 100 mA g− 1Figure 4g ). nder these optimized conditions, MCE-MA exhibited markedlyuperior long-term stability, sustaining over 200 cycles withapacity retention of 40% after 100 cycles and 35% after 200 cyclesFigure 4g ). In sharp contrast, cells employing the base or MCElectrolytes became inoperable within 30 cycles. Representativeharge–discharge profiles of the base and MCE cells prior toailure are shown in Figure 4h . Both displayed elongated andon-linear charge curves below the cut-off voltage, suggesting theccurrence of abnormal electrochemical processes. Such behav-or could stem from multiple degradation pathways, includingoft short-circuiting caused by electrode component migrationr possible corrosion of the positive electrode components underxidative conditions [ 45 ]. While these observations point towardnterfacial and mechanical instabilities in the less-optimizedlectrolytes, further systematic studies are required to elucidatehe precise failure mechanisms and identify the critical factorsimiting the high-rate durability of RMBs. Besides, full-cell testingith MCE-MA was also conducted at a higher current density of00 mA g− 1 (Figure S13 ). The cell delivers a limited capacity ( < 50Ah g− 1 ), which is primarily attributed to the intrinsically slug-ish Mg2 + diffusion kinetics in the FeVO cathode, indicating fur-her needs for cathode development with higher rate capability. lthough a comparison with literature employing WCA-basedlectrolyte and oxide cathodes could be important to further sig-ify the role of MCE-MA, direct quantitative comparison of elec-rochemical performance across different studies remains chal-enging, as reported values are highly sensitive to variations inell configuration, electrode preparation, electrolyte composition,nd testing protocols. Recent studies on the use of WCA-basedlectrolytes for high-voltage RMBs have shown that achievingtable interphase formation at high voltages remains challenging.n many cases, long-term cycling is limited by continuous elec-rolyte decomposition and interfacial instability [ 3, 5, 46 ]. In thisontext, the introduction of MCE-MA to regulate cation–anionnteractions and promote the formation of robust, fluorinatednterphases provides insight into improved cycling stability inigh-voltage RMBs systems employing oxide-based cathode. n addition to oxide-based cathodes, the applicability of MCE-A was further examined using a benchmark Chevrel-phaseathode (Mo6 S8 ) and an organic cathode material, 3,4,9,10-erylenetetracarboxylic dianhydride (PTCDA). When paired withhe Mo6 S8 cathode, the cycling performance of cells employinghe base electrolyte and MCE-MA is largely comparable (Figure9 ), with similar discharge–charge overpotentials observed overhe initial 20 cycles. This behavior is attributed to the relativelyow operating voltage of Mo6 S8 , under which oxidative electrolyteegradation is minimal, thereby limiting the advantages con-of 13ferred by MCE-MA. Nevertheless, considering the pronounceddifferences in Mg plating/stripping stability between the baseelectrolyte and MCE-MA observed in symmetric cells, the long-term cycling behavior of low-voltage cathode systems requiresfurther investigation. For the organic cathode system, cellsemploying MCE-MA exhibit improved cycling stability comparedto the base electrolyte at a current density of 20 mA g− 1 (FigureS10 ). Although the PTCDA-based full cell displays more rapidcapacity fading than the oxide-based cathode systems, theseresults demonstrate the broader compatibility of MCE-MA andhighlight organic cathodes as a promising platform for furtheroptimization and mechanistic study. 2.5 Interfacial Chemistry After Full-Cell Cycling To critically assess the influence of MCE and MCE-MA oncathode–electrolyte interphase (CEI) formation, systematic inter-facial characterizations were conducted. The interfacial charac-terizations were performed on electrodes harvested after the sameelectrochemical protocol, namely after 3 cycles of full-cell opera-tion at a current density of 10 mA g− 1 , rather than after prolongedcycling. This condition was intentionally selected to ensurecomparable interphase formation while avoiding significantdegradation or thickness accumulation associated with long-termcycling. Figure 5a–d shows the XPS spectra of FeVO electrodes inthe pristine state and after cycling in the base, MCE, and MCE-MA electrolytes. In the C 1s spectra, a carbonate-related peak( ∼ 290 eV) appears after cycling in the base electrolyte (Figure 5a )but is absent in MCE and MCE-MA, indicating that the mixed-anion systems effectively suppress solvent-derived decomposi-tion. All electrolytes exhibit C–F features with varying intensities,suggesting localized and nonuniform interphase formation. The O 1s spectra reveal broadly similar CEI compositions amongthe three systems, except for a more pronounced C ─O signal inthe base electrolyte (Figure 5b ), consistent with greater organicaccumulation observed in the C 1s spectra. The Mg 2p spec-tra confirm the presence of Mg(CO3 )2 , Mg(OH)2 , and/or MgOspecies across all samples (Figure 5c ), accompanied by minorMg ─F components. In the F 1s region (Figure 5d ), the baseelectrolyte exhibits relatively stronger C ─F, Mg ─F, and B ─Fpeaks than MCE and MCE-MA, reflecting a more extensiveanion-derived fluorinated interphase. These results contrast withthe SEI analysis, where MCE and MCE-MA displayed higherfluorinated species intensity, correlating with their enhanced Mgplating/stripping stability. However, XPS alone provides limitedinsight into CEI robustness, as fluorinated species buried beneathseveral nanometers of surface layers may not be fully detected dueto signal attenuation of F 1s photoelectrons. To verify this, detailed depth-profiling analyses using ToF-SIMS were performed on FeVO electrodes cycled in MCE-MA,alongside comparative measurements with the base electrolyte(Figure 5e–h ). The ToF-SIMS depth profiles (Figure 5e,f ) revealthat fluorine-containing species (MgF3 − , CF3 − , and BF3 − ) persistmuch deeper within the CEI of MCE-MA, whereas their inten-sity in the base electrolyte rapidly diminishes with sputteringtime. The abundance of organic fragments remains compara-ble between the two systems, suggesting that the fluorinatedinterphase in MCE-MA is not only chemically distinct but alsoAdvanced Science, 2026ve Commons LicenseFIGURE 5 (a – d) Ex situ XPS survey spectra of FeVO cathode after cycling in base electrolyte, MCE, and MCE-MA; (a) C 1s, (b) O 1s (c) Mg 2p (d) F 1s. (e–h) ToF-SIMS spectra of FeVO cathode cycled in MCE-MA compared to base electrolyte; (e) Depth profiling, and (f) 3D mapping of selected ion spectroscopy for base electrolyte. (g) Depth profiling and (h) 3D mapping of selected ion spectroscopy for MCE-MA. p  n  t  r  B  h  sI  o  a  o              A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creatihysically thicker and more uniform. In contrast, the fluori-ated components in the base electrolyte are concentrated nearhe surface, forming a thinner and patchier interphase. 3Deconstructions (Figure 5g,h ) further confirm that Mg–F and–F species form a more continuous, extended, and spatiallyomogeneous interphase layer in MCE-MA, underscoring itsuperior interfacial robustness. nterfacial processes at the cathode are expected to influence notnly CEI stability but also the evolution of the anode interphasend the homogeneity of Mg plating/stripping during full-cellperation [ 3 ]. To probe this coupling, ToF-SIMS was furtherdvanced Science, 2026employed to examine the Mg anode interphase after full-cellcycling in MCE-MA, with direct comparison to the base elec-trolyte. In the MCE-MA system, fluorinated species are detectednot only at the immediate Mg surface but also throughout deeperregions of the interphase (Figure 6c ), indicating the formationof a more integrated and chemically continuous interfacial layer.By contrast, the base electrolyte exhibits a more surface-confinedand spatially heterogeneous distribution of fluorinated species(Figure 6a ). 3D ToF-SIMS reconstructions further emphasizethis distinction: MCE-MA yields a continuous and uniformlydistributed interphase across the Mg surface (Figure 6d ), whereasthe base electrolyte produces a thin and uneven interphase9 of 13ve Commons LicenseFIGURE 6 (a–d) ToF- SIMS spectra and 3D mapping of selected ion spectroscopy of Mg metal anode cycled RMBs full cell with different electrolytes; (a,b) base electrolyte and (c,d) MCE-MA. (e–g) SEM images of Mg metal anode at (e) pristine stage, and after cycled in RMBs full cell with different electrolytes; (f) base electrolyte and (g) MCE-MA. (  m  p  F  m  M  s  c  s  d2T  s  e  m  t  e  2S  u  i  a  f  T  a  m                        1 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 CreatiFigure 6b ). These chemical differences are corroborated byorphological observations. SEM images of Mg metal in theristine state and after full-cell cycling are shown in Figure 6e–g .ollowing cycling in the base electrolyte, the Mg surface becomesarkedly rough, consistent with localized and inhomogeneousg plating/stripping, similar to what has been observed in theymmetric Mg|Mg cell test (Figure 6f ). In contrast, Mg metalycled in MCE-MA retains a comparatively smooth and compacturface morphology (Figure 6g ), indicative of more uniform Mgeposition and dissolution. .6 Toward Broader Electrolyte Component o provide additional insight into future electrolyte engineeringtrategies based on mixed-coordination chemistries, we brieflyxamined the effects of introducing a different ether-basedolecular additives such as 1,1,2,2-tetrafluoroethyl 2,2,3,3-etrafluoropropyl ether (TTE), ethyl 1,1,2,2-tetrafluoroethylther (ETFE), 1-fluoro-2-(2-fluoroethoxy)ethane (HFE), and,2-bis(trifluoromethyl)-1,3-dioxolane (BisTMD). ymmetric Mg|Mg cell tests revealed that replacing the molec-lar additive of BisTFE with TTE maintained stable Mg plat-ng/stripping for more than 500 cycles (Figure S11 ), while otherdditives performed differently. However, when evaluated inull cells using the FeVO cathode, the electrolyte containingTE experienced rapid degradation and undesirable reactionst moderate operating voltages (Figure S12 ). The underlyingechanisms responsible for this instability remain uncertain and0 of 13require further systematic investigation. Above all, these find-ings highlight the critical need to balance Mg plating/strippingperformance with the stability of high-voltage full cells whenincorporating broader electrolyte components. 3 Conclusion In this work, we establish a mixed-coordination electrolyte inte-grated with a molecular additive (MCE-MA) as a new electrolytedesign concept for stabilizing high-voltage rechargeable magne-sium batteries. Rather than relying on electrolytes with a singlesalt, this strategy deliberately combines weakly coordinating saltswith distinct cation–anion coordination characteristics and afunctional molecular additive to simultaneously regulate Mg2 + solvation and interphase formation. Specifically, the coexistenceof Mg[B(hfip)4 ]2 and a trace amount of Mg[B(tfe)4 ]2 enablescontrolled anion participation in the solvation structure, whilethe BisTFE additive promotes the formation of a homogeneousand robust interphase. As a result of this cooperative electrolyte design, MCE-MA deliv-ers markedly improved Mg plating/stripping stability, sustainingover 250 h of operation in Mg|Mg symmetric cells, whereasthe base electrolyte (0.3 m Mg[B(hfip)4 ]2 in G2) short-circuitswithin 100 h. Moreover, MCE-MA enables long-term cycling ofRMBs full cell with oxide-based cathodes for 200 cycles, in sharpcontrast to the rapid failure of the base electrolyte within 30cycles. Mechanistic analyses reveal that the MCE–MA providesa practical route to forming thicker and more robust fluorinatedAdvanced Science, 2026ve Commons LicenseC  i  fB  w  i  i  a  s  d  o44T  d  d  T  T  m  t  F  T  d  p  p  2  C  c  9  p  a  K4M  d  [M  S  p  ( i  o  a  p  t  f  f  9  w  f  4  A                                        A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 CreatiEI and SEI layers compared to the base electrolyte, therebymproving the homogeneity of Mg plating and stripping duringull-cell operation. eyond the specific electrolyte formulation studied here, thisork demonstrates that integrating mixed-coordination chem-stry with molecular additive engineering constitutes a general-zable strategy for electrolyte design. This approach establishes new framework for decoupling ion transport from interphasetability and offers broad guidance for the development ofurable, high-voltage rechargeable magnesium batteries andther multivalent energy-storage systems.  Experimental Section .1 Materials etrahydrofuran (THF), Dimethoxyethane (ethylene glycolimethyl ether, DME), and diglyme (diethylene glycolimethyl ether, G2) were purchased from Kanto Chemical.rifluoroethanol (tfe, ≥ 99.0%) was purchased from Nacalaiesque, Inc. All of the above solvents were treated witholecular sieves for a minimum of one day prior to use, andheir water content was less than 10 ppm as measured by Karl-ischer moisture titration (MKC710, KEM). 1 m Zn(BH4 )2 inHF was purchased from BetaPharma (Shanghai) Co., Ltd. andirectly used without further treatment. For the electrolytereparation, bis(2,2,2-trifluoroethyl) ether (BisTFE) wasurchased from Sigma–Aldrich (98%), and 1,1,2,2-tetrafluoroethyl,2,3,3-tetrafluoropropylether (TTE) was purchased from TCIhemicals ( > 95.0%). Both solvents were used as received. Forathode preparation, NH4 VO3 ( ≥ 99.0%) and Fe(NO3 )3 .9H2 O ( ≥8%) were purchased from Sigma–Aldrich. PTCDA ( > 98%) wasurchased from TCI Chemicals and used as received. Cu2 Mo6 S8 , precursor of Mo6 S8 cathode material, was purchased fromojundo Chemical Laboratory Co., Ltd. .2 Salt Synthesis g[B(hfip)4 ]2 was synthesized by transmetalation reaction usingi- n -butyl magnesium following the previously reported method 8 ]. g[B(tfe)4 ]2 was synthesized via dehydrogenation of Mg(BH4 )2 .pecifically, Mg(BH4 )2 in THF solution was prepared first byrecipitation reaction from 1 m Zn(BH4 )2 in THF (BetaPharmaShanghai) Co., Ltd.). In a reaction flask, 50 mL of 1 m Zn(BH4 )2n THF was mixed with an additional 50 mL of THF. Then, 1.2 gf Mg turnings were gradually added to the solution and stirredt room temperature for one week. During this period, Zn-basedrecipitates gradually formed. The resulting Mg(BH4 )2 /THF solu-ion was obtained by filtering the mixture. Next, 20 mL of theiltered Mg(BH4 )2 /THF solution was transferred into a reactionlask containing 20 mL of DME. 7.2 mL of trifluoroethanol (tfe,9%, Nacalai Tesque, Inc.) was added slowly to the mixture, inhich hydrogen evolution could be observed, followed by stirringor 20 h. Finally, the solvent was removed by vacuum drying at5◦C for at least 12 h. The salt was collected and stored in anr-filled glovebox, while the purity was confirmed by 1 H NMRdvanced Science, 2026(JNM-ECA 400, JEOL, 1 H resonance frequency = 400 MHz).The crystallographic structures of Mg[B(hfip)4 ]2 and Mg[B(tfe)4 ]2 were reproduced from CIF files deposited in the CambridgeCrystallographic Data Centre (CCDC 1537493 for Mg[B(hfip)4 ]2 and CCDC 2091598 for Mg[B(tfe)4 ]2 ). 4.3 Electrolyte Preparation The base electrolyte, 0.3 m Mg[B(hfip)4 ]2 in G2, was prepared bydissolving the Mg[B(hfip)4 ]2 salt in G2 solvent with 1 h stirringat room temperature in an Ar-filled glovebox. G2 ( > 99.5 %, KantoChemical CO., INC., Japan) solvent was treated with molecularsieves overnight prior to electrolyte preparation. MCE and MCE-MA were prepared following the same method according tothe composition in Table S1 . The water content of as-preparedelectrolytes was in the range of 30 – 150 ppm, as measured byKarl–Fischer moisture titration (MKC710, KEM). The Ramanspectra of as-prepared electrolytes were measured by a Ramanspectrometer (NRS-4500, Jasco). 4.4 Cathode Preparation FeV3 O9 ⋅1.1H2 O (FeVO) powder was prepared following the pre-vious method using NH4 VO3 and Fe(NO3 )3 .9H2 O as precursor[ 43 ]. Specifically, 24 mmol of NH4 VO3 was stirred in 800 mL ofdistilled water at 90◦C for 1 h until it dissolved completely, namelysolution A. Separately, 8 mmol of Fe(NO3 )3 ∙9H2 O was dissolvedin 80 mL of distilled water, namely solution B. Solution B wasadded slowly into solution A without stirring, producing yellow-colored insoluble colloidal products. Then, the mixture was agedat 90◦C for 24 h. The resulting brown precipitates were filtered,then thoroughly washed with distilled water multiple times anddried at 80◦C for 24 h under vacuum. Mo6 S8 powder was produced by Cu leaching from Cu2 Mo6 S8 following the reported procedure and stored under dry air [ 47 ]. Cathode material was prepared by mixing active material, Acety-lene Black, and PVDF binder with a mass ratio of 8: 1: 1, dispersedin N -methyl-2-pyrrolidone (NMP), and cast onto carbon-coatedaluminum foil as a current collector. For the XPS analysis,polyacrylonitrile (PAN) binder was used instead of PVDF to avoidF 1s overlapping. The loading mass of the cathode was 0.8 –1.5 mg/cm2 . 4.5 Electrochemical Characterization A two-electrode type cell (SB2A-EC-frontier) was used in sym-metric cell, asymmetric cell, and full cell tests [ 48–50 ]. For thesymmetric and asymmetric cell test, a glass fiber separator (GF/D,Whatman) (2.01 cm2 ) was used, while for the Mg deposition test, apolyethylene (PE, ◎wscope, SB-20D) separator was used. For thefull cell, cathode active material (2.01 cm2 ), a glass fiber separator(GF/D, Whatman) (2.01 cm2 ), and polished Mg metal (2.01 cm2 )were assembled in a two-electrode cell for full cell test. Theamount of electrolyte for each cell was 250 µL. The LSV test wasperformed using a home-made beaker-type cell. Ag/Ag+ was usedas a reference electrode and calibrated as 2.49 V vs. Mg/Mg2 + [ 9 ].11 of 13ve Commons Licensehttps://www.ccdc.cam.ac.uk/services/structures?id=doi:10.1002/advs.75749https://www.ccdc.cam.ac.uk/services/structures?id=doi:10.1002/advs.75749F  f  0  aT  a  E  m  T  c  g4T  t  t  f  A  H  t  a  t  i  r  (  t  t  cAT  N  a  M  e  tFT  NCTDRR O  (2  B  T  1                                           1 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 Creaor XRD sampling of the Mg deposit, graphite-coated aluminumoil is used. The deposition was done with a current density of.5 mA cm− 2 for 2 h. The resulting deposit was washed with THFnd dried inside an Ar-filled glovebox. he full cell discharge-charge measurement was conducted using battery cycler (HJ1001SD8 C, HD Meiden Hokuto, Japan). WhileIS was conducted using EC-Lab software on a Biologic VMP3ultichannel potentiostat (Biologic Science Instruments SAS).he cut-off voltage of the full cell was 0.8–3.3 V, with 3 h constant-urrent constant voltage mode (CCCV) during charge for 10 mA− 1 , and 1 h for 100 mA g− 1 . .6 Cathode and Anode Characterization he surfaces of the Mg anode and FeVO cathode after elec-rochemical cycling were characterized by X-ray photoelec-ron spectroscopy (XPS, VersaProbe II, ULVAC-PHI), time-of-light secondary ion mass spectrometry (ToF-SIMS, ToF-SIMS5-D-GCIB), and scanning electron microscopy (SEM, SU8200,itachi, Japan) equipped with energy-dispersive X-ray spec-roscopy (EDX). ToF-SIMS measurements were performed using 30 kV Bi3 + primary ion beam, with a sputtering rate calibratedo 4 nm min− 1 based on a SiO2 reference film. All electrochem-cal cells were disassembled inside an Ar-filled glovebox. Theetrieved electrodes were rinsed three times with tetrahydrofuranTHF) and dried inside the glovebox for several hours prioro analysis. All interfacial characterizations, including sampleransfer, were carried out under strictly air- and moisture-freeonditions. cknowledgements his work was financially supported by the GteX Program Japan (Grantumber JPMJGX23S1) of the Japan Science and Technology Agency. Theuthors acknowledge Dr. Falyouna, Ms. Sasajima, Ms. Watanabe, ands. Naya for their help in the electrochemical test validation and thelectrolyte preparation. The authors also thank BatteryPF at NIMS forheir support in the SEM, XPS, and ToF-SIMS characterization. unding his work was financially supported by the GteX Program Japan (Grantumber JPMJGX23S1) of the Japan Science and Technology Agency. onflicts of Interest he authors declare no conflicts of interest. ata Availability Statement esearch data are not shared. eferences 1 . J. T. Frith, M. J. Lacey, and U. Ulissi, “A Non-Academic Perspectiven The Future Of Lithium-Based Batteries,” Nature Communications 142023): 420, https://doi.org/10.1038/s41467- 023- 35933- 2 .  . D. Aurbach, Y. Gofer, Z. Lu, et al., “A Short Review On The Comparisonetween Li Battery Systems And Rechargeable Magnesium Batteryechnology,” Journal of Power Sources 97-98 (2001): 28–32, https://doi.org/0.1016/s0378-7753(01)00585-7 . 2 of 133 . D. Setiawan, O. Falyouna, and T. Mandai, “Beyond Half-Cell Success:Cathode-Electrolyte Reactivity Driving Magnesium Battery Full-CellDegradation at Elevated Temperature,” Advanced Science 12 (2025): 11416,https://doi.org/10.1002/advs.202511416 . 4 . Z. Yang, J. Cai, E. Wang, et al., “Electrochemical Reactivity And Passi-vation Of Organic Electrolytes At Spinel MgCrMnO4 Cathode InterfacesFor Rechargeable High Voltage Magnesium-Ion Batteries,” Journal ofMaterials Chemistry A 12 (2024): 22220–22232, https://doi.org/10.1039/d4ta03765k . 5 . X. Ye, H. Li, T. Hatakeyama, et al., “Examining Electrolyte Compatibil-ity on Polymorphic MnO2 Cathodes for Room-Temperature RechargeableMagnesium Batteries,” ACS Applied Materials & Interfaces 14 (2022):56685–56696, https://doi.org/10.1021/acsami.2c14193 . 6 . Z. Zhao-Karger, M. E. G. Bardaji, O. Fuhr, and M. Fichtner, “A NewClass Of Non-Corrosive, Highly Efficient Electrolytes For RechargeableMagnesium Batteries,” Journal of Materials Chemistry A 5 (2017): 10815–10820, https://doi.org/10.1039/c7ta02237a . 7 . J. Luo, Y. Bi, L. Zhang, X. Zhang, T. L. Liu, and A. Stable, “AStable, Non-Corrosive Perfluorinated Pinacolatoborate Mg Electrolyte forRechargeable Mg Batteries,”Angewandte Chemie International Edition 58(2019): 6967–6971, https://doi.org/10.1002/anie.201902009 . 8 . T. Mandai, “Critical Issues of Fluorinated Alkoxyborate-Based Elec-trolytes in Magnesium Battery Applications,” ACS Applied Materials& Interfaces 12 (2020): 39135–39144, https://doi.org/10.1021/acsami.0c09948 . 9 . T. Mandai, Y. Youn, and Y. Tateyama, “Remarkable ElectrochemicalAnd Ion-Transport Characteristics Of Magnesium-FluorinatedAlkoxyaluminate–Diglyme Electrolytes For Magnesium Batteries,”Materials Advances 2 (2021): 6283–6296, https://doi.org/10.1039/d1ma00448d . 10 . T. Mandai, “Enhanced Reversibility of Mg Plating/Stripping via Solva-tion Sheath Regulation by a Multidentate Linear Oligoether,” Batteries &Supercaps 8 (2025): e202500348, https://doi.org/10.1002/batt.202500348 . 11 . T. Kaneko, Y. Fujihara, T. Mandai, H. Kobayashi, and K. Sodeyama,“Ether Molecule Decomposition on Mg M2 O4 ( M = Mn, Fe, Co) SpinelSurface: A First-principles Study,” Electrochemistry 92 (2024): 027003–027003, https://doi.org/10.5796/electrochemistry.23-00087 . 12 . T. Mandai, H. Naya, and H. Masu, “Comparative Studies on[B(HFIP)4 ]-Based Electrolytes With Mono- and Divalent Cations,” TheJournal of Physical Chemistry C 127 (2023): 7987–7997, https://doi.org/10.1021/acs.jpcc.3c01160 . 13 . T. Mandai, M. Yao, K. Sodeyama, A. Kagatsume, Y. Tateyama, and H.Imai, “Toward Improved Anodic Stability of Ether-Based Electrolytes forRechargeable Magnesium Batteries,” The Journal of Physical Chemistry C127 (2023): 10419–10433, https://doi.org/10.1021/acs.jpcc.3c01452 . 14 . J. Holoubek, Q. Yan, H. Liu, et al., “Oxidative Stabilization of DiluteEther Electrolytes via Anion Modification,” ACS Energy Letters 7 (2022):675–682, https://doi.org/10.1021/acsenergylett.1c02723 . 15 . X. Ren, L. Zou, S. Jiao, et al., “High-Concentration Ether Electrolytesfor Stable High-Voltage Lithium Metal Batteries,” ACS Energy Letters 4(2019): 896–902, https://doi.org/10.1021/acsenergylett.9b00381 . 16 . S. Jiao, X. Ren, R. Cao, et al., “Stable Cycling Of High-Voltage LithiumMetal Batteries In Ether Electrolytes,” Nature Energy 3 (2018): 739–746,https://doi.org/10.1038/s41560- 018- 0199- 8 . 17 . S. Chen, J. Zheng, D. Mei, et al., “High-Voltage Lithium-Metal Bat-teries Enabled by Localized High-Concentration Electrolytes,” AdvancedMaterials 30 (2018): 1706102, https://doi.org/10.1002/adma.201706102 . 18 . A. Marojević, T. Pavčnik, O. Lužanin, et al., “Influence of SaltConcentration on the Electrochemical Performance of Magnesium Hex-afluoroisopropoxy Aluminate Electrolyte,” Batteries & Supercaps 9 (2025):202500497, https://doi.org/10.1002/batt.202500497 . 19 . W. Ren, D. Wu, Y. NuLi, et al., “An Efficient Bulky Mg[B(Otfe)4 ]2 Electrolyte and Its Derivatively General Design Strategy for RechargeableAdvanced Science, 2026tive Commons Licensehttps://doi.org/10.1038/s41467-023-35933-2https://doi.org/10.1016/s0378-7753(01)00585-7https://doi.org/10.1002/advs.202511416https://doi.org/10.1039/d4ta03765khttps://doi.org/10.1021/acsami.2c14193https://doi.org/10.1039/c7ta02237ahttps://doi.org/10.1002/anie.201902009https://doi.org/10.1021/acsami.0c09948https://doi.org/10.1039/d1ma00448dhttps://doi.org/10.1002/batt.202500348https://doi.org/10.5796/electrochemistry.23-00087https://doi.org/10.1021/acs.jpcc.3c01160https://doi.org/10.1021/acs.jpcc.3c01452https://doi.org/10.1021/acsenergylett.1c02723https://doi.org/10.1021/acsenergylett.9b00381https://doi.org/10.1038/s41560-018-0199-8https://doi.org/10.1002/adma.201706102https://doi.org/10.1002/batt.202500497M  o2  P  E  h2  “  S  h2  o  o  j2  AC  c2  S  n  (2  “  B  o2  I  S2  S  M  12  U  22  F  I3  “  o  13  S  L  j3  O  A3  C  P  (3  EJ  j3  FE  2                                        A 21983844, 2026, 45, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75749 by National Institute For, Wiley Online Library on [15/08/2026]. 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 applicaagnesium Batteries,”ACS Energy Letters 6 (2021): 3212–3220, https://doi.rg/10.1021/acsenergylett.1c01411 . 0 . W. A. Narvaez, S. J. Park, and B. J. Schwartz, “Competitive Ionairing and the Role of Anions in the Behavior of Hydrated Electrons inlectrolytes,” The Journal of Physical Chemistry B 126 (2022): 7701–7708,ttps://doi.org/10.1021/acs.jpcb.2c04463 . 1 . G. S. Perera, C. B. Nettles, Y. Zhou, S. Zou, T. K. Hollis, and D. Zhang,Direct Observation of Ion Pairing at the Liquid/Solid Interfaces byurface Enhanced Raman Spectroscopy,” Langmuir 31 (2015): 8998–9005,ttps://doi.org/10.1021/acs.langmuir.5b01903 . 2 . T. Kimura, K. Fujii, Y. Sato, M. Morita, and N. Yoshimoto, “Solvationf Magnesium Ion in Triglyme-Based Electrolyte Solutions,” The Journalf Physical Chemistry C 119 (2015): 18911–18917, https://doi.org/10.1021/acs.pcc.5b04626 . 3 . J. Bitenc, K. Pirnat, O. Luz ̌anin, and R. Dominko, “Organic Cathodes, Path Toward Future Sustainable Batteries: Mirage or Realistic Future?,”hemistry of Materials 36 (2024): 1025–1040, https://doi.org/10.1021/acs.hemmater.3c02408 . 4 . T. Mandai, K. Tatesaka, K. Soh, et al., “Modifications in Coordinationtructure of Mg[TFSA]2 -Based Supporting Salts For High-Voltage Mag-esium Rechargeable Batteries,” Physical Chemistry Chemical Physics 212019): 12100–12111, https://doi.org/10.1039/c9cp01400d . 5 . C. Li, R. D. Guha, A. Shyamsunder, K. A. Persson, and L. F. Nazar,A Weakly Ion Pairing Electrolyte Designed For High Voltage Magnesiumatteries,”Energy & Environmental Science 17 (2023): 190–201, https://doi.rg/10.1039/d3ee02861e . 6 . D. Ruan, Z. Cui, J. Fan, D. Wang, Y. Wu, and X. Ren, “Recent Advancesn Electrolyte Molecular Design For Alkali Metal Batteries,” Chemicalcience 15 (2024): 4238–4274, https://doi.org/10.1039/d3sc06650a . 7 . M. Karbak, K. Yim, Y. S. Meng, Y. Shao, and W. Xu, “Weaklyolvating Electrolytes: A Solvation-Centric Paradigm For Rechargeableetal Batteries,”Chemical Science 16 (2025): 20694–20717, https://doi.org/0.1039/d5sc06221g . 8 . J. S. Corneille, J.-W. He, and D. W. Goodman, “XPS Characterization Ofltra-Thin MgO Films on a Mo(100) Surface,” Surface Science 306 (1994):69–278, https://doi.org/10.1016/0039- 6028(94)90071- x. 9 . Y. Lei, K. Ng, L. Weng, C. Chan, and L. Li, “XPS C 1s Binding Energiesor Fluorocarbon–Hydrocarbon Microblock Copolymers,” Surface andnterface Analysis 35 (2003): 852–855, https://doi.org/10.1002/sia.1615 . 0 . Y. Kawamoto, K. Ogura, M. Shojiya, M. Takahashi, and K. Kadono,F1s XPS of Fluoride Glasses And Related Fluoride Crystals,” Journalf Fluorine Chemistry 96 (1999): 135–139, https://doi.org/10.1016/s0022-139(99)00068-8 . 1 . M. W. Herdiech, H. Mönig, and E. I. Altman, “An X-Ray Photoelectronpectroscopy Study of BF3 Adsorption On Positively And Negatively PolediNbO3 (0001),” Surface Science 626 (2014): 53–60, https://doi.org/10.1016/.susc.2014.04.004 . 2 . W. E. Moddeman, A. R. Burke, W. C. Bowling, and D. S. Foose, “Surfacexides of Boron and B12 O2 as Determined by XPS,” Surface and Interfacenalysis 14 (1989): 224–232, https://doi.org/10.1002/sia.740140503 . 3 . J. Xiao, X. Zhang, H. Fan, et al., “Cosolvent-Assisted Formation ofharged Ion-Solvent Clusters and Solid Electrolyte Interphase for High-erformance Magnesium Metal Batteries,” Advanced Energy Materials 122022): 2202602, https://doi.org/10.1002/aenm.202202602 . 4 . R. May, J. C. Hestenes, N. A. Munich, and L. E. Marbella, “Fluorinatedther Decomposition In Localized High Concentration Electrolytes,”ournal of Power Sources 553 (2023): 232299, https://doi.org/10.1016/j.powsour.2022.232299 . 5 . T. Mandai, U. Tanaka, and M. Watanabe, “Mg–Zn–Cl-Integratedunctional Interface For Enhancing The Cycle Life of Mg Electrodes,”nergy Storage Materials 67 (2024): 103302, https://doi.org/10.1016/j.ensm.024.103302 . dvanced Science, 2026ble C36 . X. Liu, A. Du, Z. Guo, et al., “Uneven Stripping Behavior, anUnheeded Killer of Mg Anodes,” Advanced Materials 34 (2022): 2201886,https://doi.org/10.1002/adma.202201886 . 37 . X. Liu, G. Wang, Z. Lv, A. Du, S. Dong, and G. Cui, “A Perspective onUniform Plating Behavior of Mg Metal Anode: Diffusion Limited Theoryversus Nucleation Theory,” Advanced Materials 36 (2024): 2306395,https://doi.org/10.1002/adma.202306395 . 38 . A. Pei, G. Zheng, F. Shi, Y. Li, and Y. Cui, “Nanoscale Nucleation andGrowth of Electrodeposited Lithium Metal,” Nano Letters 17 (2017): 1132–1139, https://doi.org/10.1021/acs.nanolett.6b04755 . 39 . K. Yan, J. Wang, S. Zhao, et al., “Temperature-Dependent Nucleationand Growth of Dendrite-Free Lithium Metal Anodes,” AngewandteChemie 131 (2019): 11486–11490, https://doi.org/10.1002/ange.201905251 . 40 . R. Davidson, A. Verma, D. Santos, et al., “Mapping MechanismsAnd Growth Regimes Of Magnesium Electrodeposition At High CurrentDensities,”Materials Horizons 7 (2019): 843–854, https://doi.org/10.1039/c9mh01367a . 41 . J. Chen, H. Zhang, M. Fang, C. Ke, S. Liu, and J. Wang, “Designof Localized High-Concentration Electrolytes via Donor Number,” ACSEnergy Letters 8 (2023): 1723–1734, https://doi.org/10.1021/acsenergylett.3c00004 . 42 . D. Setiawan, J. Hwang, M. S. Chae, and S.-T. Hong, “High VoltageCathode Materials For Rechargeable Magnesium Batteries: StructuralAspects And Electrochemical Perspectives,” Journal of Magnesium andAlloys 13 (2025): 4167–4188, https://doi.org/10.1016/j.jma.2025.07.018 . 43 . D. Setiawan, H. J. Kim, J. Lyoo, S.-T. Hong, and M. S. Chae, “NovelLayered Iron Vanadate As A Stable High-Voltage Cathode Material ForNonaqueous Magnesium-Ion Batteries,” Chemical Engineering Journal474 (2023): 145596, https://doi.org/10.1016/j.cej.2023.145596 . 44 . M. S. Chae, D. Setiawan, H. J. Kim, and S.-T. Hong, “LayeredIron Vanadate as a High-Capacity Cathode Material for NonaqueousCalcium-Ion Batteries,” Batteries 7 (2021): 54, https://doi.org/10.3390/batteries7030054 . 45 . C. Hogrefe, T. Waldmann, M. Hölzle, and M. Wohlfahrt-Mehrens,“Direct Observation Of Internal Short Circuits By Lithium Dendrites InCross-Sectional Lithium-Ion In Situ Full Cells,” Journal of Power Sources556 (2023): 232391, https://doi.org/10.1016/j.jpowsour.2022.232391 . 46 . H. Kobayashi, Y. Fukumi, H. Watanabe, et al., “Ultraporous, Ultra-small MgMn2 O4 Spinel Cathode for a Room-Temperature MagnesiumRechargeable Battery,” ACS Nano 17 (2023): 3135–3142, https://doi.org/10.1021/acsnano.2c12392 . 47 . M. D. Levi, E. Lancry, H. Gizbar, et al., “Kinetic and ThermodynamicStudies of Mg2 + and Li+ Ion Insertion Into the Mo6 S8 Chevrel Phase,”Journal of The Electrochemical Society 151 (2004): A1044–A1051, https://doi.org/10.1149/1.1758811 . 48 . Z. Song, Z. Zhang, A. Du, S. Dong, G. Li, and G. Cui, “InsightsInto Interfacial Speciation And Deposition Morphology Evolution AtMg-Electrolyte Interfaces Under Practical Conditions,” Journal of EnergyChemistry 48 (2020): 299–307, https://doi.org/10.1016/j.jechem.2020.02.019 . 49 . Z. Li, T. Diemant, Z. Meng, et al., “Establishing a Stable Anode–Electrolyte Interface in Mg Batteries by Electrolyte Additive,” ACSApplied Materials & Interfaces 13 (2021): 33123–33132, https://doi.org/10.1021/acsami.1c08476 . 50 . E. Winter, T. J. Schmidt, and S. Trabesinger, “Identifying Pitfalls inLithium Metal Battery Characterization,” Batteries & Supercaps 5 (2022):202100145, https://doi.org/10.1002/batt.202100145 . Supporting Information Additional supporting information can be found online in the SupportingInformation section. Supporting File: advs75749-sup-0001-SuppMat.docx. 13 of 13reative Commons Licensehttps://doi.org/10.1021/acsenergylett.1c01411https://doi.org/10.1021/acs.jpcb.2c04463https://doi.org/10.1021/acs.langmuir.5b01903https://doi.org/10.1021/acs.jpcc.5b04626https://doi.org/10.1021/acs.chemmater.3c02408https://doi.org/10.1039/c9cp01400dhttps://doi.org/10.1039/d3ee02861ehttps://doi.org/10.1039/d3sc06650ahttps://doi.org/10.1039/d5sc06221ghttps://doi.org/10.1016/0039-6028(94)90071-xhttps://doi.org/10.1002/sia.1615https://doi.org/10.1016/s0022-1139(99)00068-8https://doi.org/10.1016/j.susc.2014.04.004https://doi.org/10.1002/sia.740140503https://doi.org/10.1002/aenm.202202602https://doi.org/10.1016/j.jpowsour.2022.232299https://doi.org/10.1016/j.ensm.2024.103302https://doi.org/10.1002/adma.202201886https://doi.org/10.1002/adma.202306395https://doi.org/10.1021/acs.nanolett.6b04755https://doi.org/10.1002/ange.201905251https://doi.org/10.1039/c9mh01367ahttps://doi.org/10.1021/acsenergylett.3c00004https://doi.org/10.1016/j.jma.2025.07.018https://doi.org/10.1016/j.cej.2023.145596https://doi.org/10.3390/batteries7030054https://doi.org/10.1016/j.jpowsour.2022.232391https://doi.org/10.1021/acsnano.2c12392https://doi.org/10.1149/1.1758811https://doi.org/10.1016/j.jechem.2020.02.019https://doi.org/10.1021/acsami.1c08476https://doi.org/10.1002/batt.202100145 Mixed-Coordination Electrolytes With Molecular Additives for Robust Interphases in High-Voltage Rechargeable Magnesium Batteries 1 | Introduction 2 | Results and Discussion 2.1 | Electrolyte Solvation Structure 2.2 | Anion-Derived Solid Electrolyte Interphase 2.3 | Impact of SEI on Mg Plating/Stripping Performance 2.4 | Insight Into High-Voltage RMBs Application 2.5 | Interfacial Chemistry After Full-Cell Cycling 2.6 | Toward Broader Electrolyte Component 3 | Conclusion 4 | Experimental Section 4.1 | Materials 4.2 | Salt Synthesis 4.3 | Electrolyte Preparation 4.4 | Cathode Preparation 4.5 | Electrochemical Characterization 4.6 | Cathode and Anode Characterization Acknowledgements Funding Conflicts of Interest Data Availability Statement References Supporting Information