Quantifying electrokinetics of NaCa_0.6V₆O₁₆·3H₂O cathode in aqueous zinc-ion batteries with ZnSO₄ electrolyte

Abstract

Aqueous zinc-ion batteries (AZIBs) have been actively studied in recent years as a promising solution for next-generation stationary energy storage due to their inherent safety, low cost, and high energy density. However, their practical deployment remains hindered by the limited cycling stability of cathode materials. Overcoming this challenge requires a detailed understanding of cathodic electrokinetics and degradation mechanisms. In this study, we investigate the electrokinetic behavior of a NaCa_0.6V₆O₁₆·3H₂O (NaCaVO) cathode in ZnSO₄ electrolyte through a combined application of the galvanostatic intermittent titration technique (GITT) and electrochemical impedance spectroscopy (EIS). For the first time, we quantify the exchange current density (i₀) and interfacial charge-transfer resistance (R_CT) of NaCaVO as a function of states of charge (SOCs). The results reveal that the V⁴⁺ ⇌ V³⁺ redox reaction exhibits significantly slower kinetics than the V⁵⁺ ⇌ V⁴⁺ counterpart. Further GITT-EIS studies using D₂O–ZnSO₄ electrolyte, complemented by X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS), indicate that the sluggish V⁴⁺ ⇌ V³⁺ process is predominantly associated with proton insertion. Distribution of Relaxation Time (DRT) analysis correlates the increased interfacial resistance with the intermediate phase formation induced by this proton insertion. The electrokinetic insights obtained in this work fill critical knowledge gaps in AZIB research and provide a foundation for designing more durable and efficient cathode materials in the future.

---

1. Introduction

Rechargeable aqueous zinc-ion batteries (AZIBs) are an emerging class of electrochemical energy storage devices well-suited for stationary energy storage where cost-effectiveness, safety, and environmental sustainability are important considerations.^1–5 A typical AZIB consists of a high-energy-density zinc-metal anode, a layered or open-framework cathode, and a zinc-ion-conducting aqueous electrolyte – commonly a zinc salt solution.^6–9 With Zn²⁺ and/or H⁺ serving as the working ions, the electrochemical reactions at both cathode/electrolyte and anode/electrolyte interfaces are generally reversible, enabling rechargeable operation – a key advantage over conventional non-rechargeable alkaline Zn/MnO₂ batteries.^10–14

Despite their promise, the commercial development of AZIBs faces serious challenges, particularly due to degradations associated with the cathode and its interfacial interactions with the electrolyte.^15–17 Therefore, the advancement of AZIB technology critically hinges on the development of high-capacity, durable cathode materials, which in turn requires a thorough understanding of cathodic degradation mechanisms.^1,18–21 Additionally, there remains a substantial knowledge gap in the understanding of cathodic kinetics – especially those that govern the ion storage processes at various states of charge (SOCs).

To date, a major effort in aqueous zinc-ion battery (AZIB) research has been focused on engineering a stable zinc-anode/electrolyte interface to suppress the hydrogen evolution reaction (HER) at the zinc anode and mitigate zinc corrosion. Strategies such as surface coatings and electrolyte additives have been widely explored for this purpose. However, many of these additives inadvertently hinder cathodic kinetics, ultimately compromising overall battery performance despite effectively suppressing HER and reducing corrosion.^22,23

In comparison, cathode-focused research has primarily centered on compositional modifications of MnO₂⁻ and V₂O₅⁻ based materials to enhance capacity and cycling stability. A notable example is the pre-insertion of alkaline and alkaline-earth metal cations into V₂O₅-based structures.^8,24–29 In these studies, a variety of in situ and ex situ surface and bulk characterization techniques – such as X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), Electron Microscopy, Fourier Transform Infrared Spectroscopy (FTIR), and X-ray absorption spectroscopy (XAS) – are routinely employed to probe material properties and elucidate underlying reaction mechanisms. In parallel, conventional electrochemical techniques³⁰ such as Galvanostatic/Potentiostatic Intermittent Titration (GITT/PITT),^31,32 electrochemical impedance spectroscopy (EIS),³³ and cyclic voltammetry (CV)^34,35 are used to study electrochemical behaviors. Despite these efforts, rigorous investigation of cathodic electrokinetics during actual charge and discharge processes remains rare. Furthermore, many early studies only evaluated charge-transfer resistance (R_CT) under open-circuit conditions, which do not accurately represent real operating states of AZIBs, thereby limiting their value for mechanistic understanding and battery performance modeling.

In this study, we present a combined GITT and EIS approach to probe cathodic electrokinetics under actual operating conditions. The cathode material of choice is NaCa_0.6V₆O₁₆·3H₂O (NaCaVO), previously identified in our work as a stable and promising candidate for AZIB applications with outstanding cyclability (e.g., 94% capacity retention after 2000 cycles at 2 A g⁻¹ and 83% after 10 000 cycles at 5 A g⁻¹).³⁶ Electrokinetic measurements were conducted using a three-electrode configuration, with NaCaVO as the working electrode (WE), zinc metal as the counter electrode (CE), Ag/AgCl as the reference electrode (RE), and 2 M ZnSO₄ as the electrolyte – an industry-relevant formulation. To deconvolute overlapping electrochemical processes and better interpret cathodic behavior, we also apply Distribution of Relaxation Times (DRT) analysis to EIS spectra, correlating observed features with the formation of interfacial secondary phases. Additionally, to investigate the co-intercalation mechanisms of Zn²⁺ and H⁺ into the NaCaVO structure, we employ water isotope solvent D₂O in the ZnSO₄ electrolyte system to distinguish the contributions of zinc ions from those of protons.

2. Results and discussion

2.1 Electrokinetics quantification using combined GITT-EIS method

The cyclic voltammetry (CV) profiles of the NaCaVO cathode in 2 M H₂O–ZnSO₄ electrolyte at scan rates ranging from 0.1 to 1.0 mV s⁻¹ are shown in Fig. 1(a). Two distinct pairs of redox peaks, corresponding to the redox transitions of V-oxidation state, are consistently observed across all scan rates. The high-potential redox couple (D₁ at ∼0.81–0.82 V and C₁ at ∼0.91–0.93 V vs. Zn/Zn²⁺) is assigned to the V⁵⁺ ⇌ V⁴⁺ redox reaction, while the lower-potential couple (D₂ at ∼0.56 V and C₂ at ∼0.70 V) is related to the V⁴⁺ ⇌ V³⁺ reaction.^37–41 With increasing scan rate, both anodic and cathodic peak currents rise systematically, indicating charge-transfer-controlled behavior. However, the lower-potential peaks (D₂/C₂) are consistently broader and exhibit smaller peak currents relative to D₁/C₁, suggesting that the V⁴⁺ ⇌ V³⁺ reaction proceeds with slower kinetics and greater polarization than the V⁵⁺ ⇌ V⁴⁺ reaction. Moreover, the potential separation between cathodic and anodic peaks slightly widens with scan rate, particularly for the V³⁺/V⁴⁺ redox couple, further indicating that kinetic limitations and sluggish charge transfer dominate at low potentials. Overall, there is no significant change in redox peak position at different scan rates, justifying our use of potential to characterize electrokinetic behaviors of the V⁵⁺ ⇌ V⁴⁺ and V⁴⁺ ⇌ V³⁺ redox reactions during the charge and discharge cycles.

Fig. 1 H₂O–ZnSO₄ system: (a) CV and (b) I_p vs. v^1/2 of NaCaVO|2 M ZnSO₄|Zn cell scanned at different rates. RE: Ag/AgCl; CE: Zn; WE: NaCaVO.

Moreover, to investigate the ion diffusion characteristics in the NaCaVO cathode in H₂O–ZnSO₄ system, cyclic voltammetry (CV) conducted at varying scan rates (0.1–1.0 mV s⁻¹) was further analyzed by plotting the peak currents corresponding to distinct redox reactions vs. square root of scanning rate to determine diffusion coefficient of the electroactive species of a diffusion-controlled redox process using Randles–Ševčík equation:^42,43

I_p = (2.69 × 10⁵) × n^3/2 × A × D^1/2 × C × v^1/2 (1)

Here I_p is the peak current (A), n is the number of electrons transferred, A is the electrode area (cm²), D is the diffusion coefficient (cm² s⁻¹) of the electroactive species, C is the molar concentration of electroactive species (mol cm⁻³), and ν is the scan rate (V s⁻¹).

Fig. 1(b) shows that a plot of I_p versus v^1/2 of the four redox peaks follow a linear relationship, indicating semi-infinite diffusion as the rate-limiting process. The calculated diffusion coefficients related to the four redox peaks are 4.74 × 10⁻¹⁰ and 1.37 × 10⁻¹⁰ cm² s⁻¹ for C₁ and D₁, respectively, located at the high-potential region and 1.14 × 10⁻¹¹ and 3.11 × 10⁻¹¹ cm² s⁻¹ for C₂ and D₂, respectively, located at the low-potential peaks. The higher D values for the high-potential V⁴⁺/V⁵⁺ redox couple than the V³⁺/V⁴⁺ counterpart suggests a facile kinetics for the former, which will be further discussed in the following sections.

The potential (E) profiles collected by GITT during charge and discharge cycle are shown in Fig. 2(a). As potential E increases and decreases with time during charge and discharge cycle, respectively, pseudo-plateaus regions (quasi-flat regions) show good alignment with the two redox peaks identified in the CV (see Fig. 1(a)). During the charge cycle, the profile exhibits a mid-E pseudo-plateau around 0.85–0.95 V (corresponding to V⁴⁺ → V⁵⁺ oxidation peak) and a lower pseudo-plateau near 0.60–0.70 V (corresponding to V³⁺ → V⁴⁺ oxidation peak). Similar pseudo-plateaus are also observed during the discharge cycle, corresponding to V⁵⁺ → V⁴⁺ and V⁴⁺ → V³⁺ reduction peaks, respectively. The greater E changes in lower E region than in higher E region imply higher polarization and sluggish kinetics in the region.

Fig. 2 ZnSO₄–H₂O system: (a) E vs. time profile collected under ±50 mA g⁻¹; EIS spectra measured with ±50 mA g⁻¹ bias during charge (b) and discharge (c) cycles.

The EIS spectra collected at each GITT step during charge and discharge cycled at ±50 mA g⁻¹ (equivalent to ±0.024 mA cm⁻²) are shown in Fig. 2(b) and (c), respectively, as an example. Each spectrum is measured under different bias currents to obtain R_CT= R_CT(i) at a specific E (or SOC) created by the preceding galvanic polarization. From Fig. 2(b) collected during the charging cycle, it is evident that EIS spectra feature a charge transfer impedance at high frequencies, followed by low-frequency Warburg impedance (related to semi-infinitive surface diffusion of the active species) over low E range (step 1–10, 0.53–0.70 V). As E increases, the spectra gradually transit to a mixed Warburg (step 11–14, 0.72–0.78 V) and charge-transfer only feature (step 15–28, 0.80–0.93 V) over intermediate E range, and finally to charge-transfer only feature over high E range (step 29–45, 0.95–1.50 V). During the discharging cycle, the above trend remains, i.e., Warburg impedance appears at low E, whereas charge-transfer-only feature appears at high E. Note that the appearance of Warburg impedance at low E range implies diffusion limitation to the active species; we will further correlate it with the formation of an intermediate phase in the following section. The original E vs. time profiles and corresponding EIS spectra collected under OCV and other current densities such as 75–125 mA g⁻¹ can be found in Fig. S1.† By comparing Fig. 2 and .S1,† it is concluded that the shape of all profiles (two distinct pseudo-plateau regions) remains the same, suggesting that the electrochemical reaction mechanisms remain unchanged by the applied current.

With the equivalent circuit model shown in Fig. 2(b-1), R_CT under low E is extracted from the interception length on Z′-axis of the fitted curve (dotted), and for the high E cases, R_CT (sum of R_CT1 and R_CT2) is similarly taken using the equivalent circuit model in Fig. 2(c-1). The obtained R_CT is then plotted against E and shown in Fig. 3 for both charge and discharge cycle in a current range of 0–125 mA g⁻¹. A clear dependence of R_CT on E is observed. In both charge and discharge cycles, the R_CT–E profiles can be divided into three stages. Stage-1, the lowest E range (0.40–0.60 V for both charge and discharge), R_CT is generally the highest but decreases with E. Stage-2, the intermediate E range (0.68–0.91 V for charge and 0.60–0.83 V for discharge), a maximum is observed. Stage-3, the high E range (>0.91 V for charge and >0.83 V for discharge), a plateau is observed. The minimal R_CT appears to occur at a high E close to the V⁵⁺ ⇌ V⁴⁺ as indicated in CV, see Fig. 1(a). In contrast, high R_CT appears at lower E range close to the V⁴⁺ ⇌ V³⁺. Therefore, we can conclude from Fig. 3 that the V⁴⁺ ⇌ V³⁺ redox reaction is more sluggish than the V⁵⁺ ⇌ V⁴⁺ counterpart. This finding could be explained by a greater difficulty in inserting/extracting ions into/from NaCaVO at a deeper SOC. Another reason could be related to the formation of intermediate phase on the surface of NaCaVO cathode due to proton insertion; we will show evidence to support this assertion in a later section. On the other hand, in the higher E and shallow SOC regime, the charge transfer kinetics is facile with low R_CT, suggesting a lower presence of the intermediate phase. In addition, the ohmic resistance R_O of all cells studied was roughly ∼0.45 Ω cm² (see Fig. 2 and S1†), independent of DC bias applied.

Fig. 3 H₂O–ZnSO₄ system: R_CT vs. E during (a) charge and (b) discharge cycle under different current densities.

The trend of enhanced charge transfer kinetics by the applied current agrees with Butler–Volmer law that predicts higher overpotential drives a lower R_CT. The agreement also suggests that the charge transfer kinetics is a rate limiting step of the ion insertion process into NaCaVO. Although the shapes of two R_CT–E profiles between charge and discharge cycle are close to a mirror image, the discharge cycle does exhibit a slightly higher R_CT than the charge cycle at the same E, indirectly reflecting influence of the intermediate phase formation on charge transfer process.

2.2 Exchange current density (i₀)

With the obtained R_CT vs. i, we apply the “low-field” approximation of Butler–Volmer equation to extract exchange current density (i₀), a key parameter of electrokinetics, as the current density range applied in this study generally falls into “weak polarization” regime:⁴⁴

η = (RT/nF) × i/i₀) (2)

where η is the overpotential; i is the applied current density; n = 1 is the number of electrons transferred; i₀ is the exchange current density; R and T have their usual meanings. To obtain i-dependent η, we further integrate the experimental R_CT = R_CT(i) shown in Fig. 4(a) and obtain η–i relationship shown in Fig. 4(b). The selected potentials correspond to the regions where the charge transfer resistance (R_CT) reaches a minimum or forms a plateau in the R_CT vs. E plot (see Fig. 3), specifically at ∼0.68 V and ∼0.91 V during the charge, and ∼0.60 V and ∼0.83 V during the discharge, close to those of the redox reactions of V⁵⁺ ⇌ V⁴⁺ and V⁴⁺ ⇌ V³⁺. As is shown, a linear relationship is observed with a current density up to 125 mA g⁻¹, implying that the “low-field” approximation is valid.

Fig. 4 H₂O–ZnSO₄ system. (a) R_CT vs. i, (b) η vs. i and (c) i₀ comparison at 0.68 V and 0.91 V during charge and 0.83 V and 0.6 V during discharge.

A comparison of i₀ at different E of the charge and discharge cycle is given in Fig. 4(c); it informs that i₀ is higher at higher E than at lower E. Between the charge and discharge cycle for the same pair of redox reaction, i₀ is higher for the charge cycle than the discharge cycle. Again, these comparisons suggest that the V⁵⁺ ⇌ V⁴⁺ is a more facile redox reaction than the V⁴⁺ ⇌ V³⁺ counterpart. We hypothesize that the sluggish V⁴⁺ ⇌ V³⁺ redox reaction could be attributed to the intermediate phase formation induced by the H⁺ insertion at a deeper SOC. To verify this, we performed a further analysis on EIS spectra with DRT method and replacing of H₂O with D₂O as a solvent for the ZnSO₄ electrolyte.

2.3 Analysis of EIS spectra with DRT method

Even though our GITT-EIS method provides E-dependent R_CT data under different i, we still do not fully understand the reasons why the kinetics of V⁵⁺ ⇌ V⁴⁺ is faster than V⁴⁺ ⇌ V³⁺. We here apply DRT method to facilitate the understanding.⁴⁵ Fig. 5 shows an exemplary contour plot of γ-function of DRT analysis over a time domain (τ, relaxation time constant, unit in s) superposed with E (SOC) during a constant polarization at 50 mA g⁻¹, where τ is related to the frequency (f, in Hz) in EIS by:By correlating the E vs. time profiles with the corresponding DRT contour plots, specific phase formation and deformation processes can be revealed by the emergence or suppression of the γ(ln τ) response in the contour map within defined potential regions. At the peak redox potentials (see Fig. 1(a)), i.e. 1.50 to ∼0.80 V and ∼0.65 to 1.50 V corresponding to the V⁵⁺ ⇌ V⁴⁺ reaction during charging and discharging, respectively, while 0.40 to ∼0.80 V and ∼0.65 to 0.40 V corresponding to the V⁴⁺ ⇌ V³⁺ reaction during charging and discharging, respectively, the DRT plots unveil significant impedance response. At shorter relaxation time (<0.001 s, high frequency), the impedance is generally low in either charge or discharge regimes, implying that the charge transfer process of NaCaVO cathode with ZnSO₄ electrolyte is facile. At longer relaxation time (0.1–0.001 s, low frequency), the impedance is high in either charge or discharge regimes, implying possible diffusion limitation. The intermediate phase formed during the discharge process could be responsible for the high impedance at low-frequencies. In fact, the Warburg impedance tails become steeper at higher current densities, see Fig. 2 and S1,† inferring an increasing diffusion resistance. The intermediate phase formed during the discharge cycle could be a reason; we will show the evidence in the following section. For all current densities (OCV, 50, 75 and 100 mA g⁻¹), Fig. 5 suggest that this “intermediate phase effect” is reversible upon the charge and discharge cycle.

Fig. 5 E vs. time profiles (left) and corresponding DRT contour plots (right) derived from EIS of the NaCaVO cathode in H₂O–ZnSO₄ electrolyte during charge–discharge cycles under (a) OCV, (b) ±50 mA g⁻¹, (c) ±75 mA g⁻¹, and (d) ±100 mA g⁻¹. The DRT maps reveal phase formation and deformation processes via intensity variations in γ(ln τ) across specific potential ranges.

2.4 Evidence of intermediate phase formation

To identify the composition of the secondary phase, we performed ex situ XRD and XPS. Fig. 6 shows XRD patterns of NaCaVO cathode after being cycled under different SOCs. From the main peak of NaCaVO in Fig. 6(a), it is evident that the ion (de)insertion processes are generally reversible during the charge/discharge cycle, exhibiting lattice contraction during the discharge due to strong Zn²⁺-O^2- interactions after Zn²⁺ are inserted and lattice expansion during the charge after Zn²⁺ are extracted. More importantly, Fig. 6(b) indicates that an intermediate phase with a composition of Zn₄SO₄(OH)₆·xH₂O (herein denoted as Zn-LDH) is detected in NaCaVO only after being cycled at low E between 0.40 and 0.80 V (related to V³⁺ ⇌ V⁴⁺). The finding of Zn-LDH at low-E-cycled NaCaVO provide chemical reason for the slow electrokinetics of V³⁺ ⇌ V⁴⁺ and aggravated diffusion impedance. Since it is well known to the AZIB community that Zn-LDH is a product of H⁺ insertion into the cathode, it further suggests that H⁺ insertion might take place at lower E.^46,47 This contrasts with some theoretical calculations suggesting that H⁺ insertion takes place mainly at higher E.^48–50

Fig. 6 H₂O–ZnSO₄ system. Ex situ XRD patterns of NaCaVO cathode after (a) charge and discharge at different SOCs and (b) cycled at low and high E compared with the pristine.

The XPS spectra shown in Fig. S4† indicate a weaker V-2p signal in NaCaVO after cycling at low E than those of the pristine and cycled at high E, suggesting possible blocking effect by the Zn-LDH formed. Furthermore, the ratio of V³⁺/V⁴⁺ in NaCaVO becomes significantly higher after cycling at low E compared to the pristine (with V⁵⁺ and trace V⁴⁺) and the high-E-cycled one (with much lower V³⁺/V⁴⁺ ratio), which is consistent with our previous results.³⁶ These XPS results further suggest that the low-E redox reaction is associated with V³⁺ ⇌ V⁴⁺ and high-E redox reaction is related to V⁵⁺ ⇌ V⁴⁺.

2.5 Effect of D₂O on electrokinetics

To further verify if the H₂O-derived H⁺ insertion into NaCaVO is responsible for the redox reaction of V³⁺ ⇌ V⁴⁺, we performed the same GITT-EIS experiment using the same ZnSO₄ electrolyte but isotope D₂O as the solvent. Fig. S2† shows the original CV curves of NaCaVO in D₂O–ZnSO₄, which is notably different from the H₂O–ZnSO₄ case, especially in the low E range (0.5–0.7 V). In particular, the redox peaks associated with the V³⁺ ⇌ V⁴⁺ become less pronounced and shifted toward lower E, reflecting the impact of heavier D⁺ ions on this redox reaction. The more pronounced broadening and reduced peak current density suggests a retarded reaction kinetics. These observations support the hypothesis that H⁺/D⁺ impacts on the V³⁺ ⇌ V⁴⁺ redox reaction at lower E.

The E-profiles collected from the D₂O–ZnSO₄ system under 50 mA g⁻¹ by GITT during charge and discharge cycle are shown in Fig. 7 as an example; similar plots at other current densities can be found in Fig. S3.† The two-step pseudo-plateau feature remains, suggesting two active charge transfer processes: V³⁺ ⇌ V⁴⁺ and V⁴⁺ ⇌ V⁵⁺. However, a quick glance of EIS spectra indicate that R_CT at low E is appreciably higher than the H₂O–ZnSO₄ counterpart, inferring that the V³⁺ ⇌ V⁴⁺ kinetics has become more sluggish by the slower D⁺ (de)insertion process.

Fig. 7 D₂O–ZnSO₄ system: (a) E vs. time profile collected under ±50 mA g⁻¹; EIS spectra measured with ±50 mA g⁻¹ bias during (b) charge and (c) discharge cycles. Red circles mark the extra semicircle at high frequency range.

A noticeable difference of EIS spectra of the D₂O–ZnSO₄ system from H₂O–ZnSO₄ system, see Fig. 7(b), is the appearance of a high-frequency semicircle (highlighted by red dotted circles) during the charge cycle. The magnitude of this small semicircle does not seem to change appreciably with SOC. Since D⁺ extraction from NaCaVO occurs during the charge cycle, the appeared semicircle (an extra charge transfer process) implies a reduced D⁺ extraction kinetics, which leads to a more sluggish V³⁺ ⇌ V⁴⁺ redox reaction and increased relaxation time constant. Another notable observation is that the Warburg impedance becomes less pronounced as E increases, the same trend as the H₂O–ZnSO₄ system.

For the subsequent discharge cycle, see Fig. 7(c), the EIS collected under high E (Fig. 7(c-1), step 1–19, 1.5–0.83 V) show a much less pronounced, small high-frequency semicircle than that observed during the charge cycle. As E decreases, see Fig. 7(c-2) (steps 20–39, 0.82–0.66 V) and (c-3) (steps 40–56, 0.65–0.4 V), this small high-frequency semicircle appears to increase slightly, while the Warburg impedance becomes more pronounced, like the H₂O–ZnSO₄ system. A comparison of EIS spectra vs. SOC between the H₂O–ZnSO₄ and D₂O–ZnSO₄ systems suggests similar charge transfer mechanisms, but the heavier D⁺ is more difficult to extract from the cathode, thus increasing the relaxation time constant of the process and depicting itself on the EIS spectrum within the frequency range studied. Since D₂O has the most pronounced impact on R_CT at low E, it is reasonable to speculate that H⁺/D⁺ (de)insertion into the cathode leads to the V³⁺ ⇌ V⁴⁺ redox reaction.

To better illustrate how R_CT vary with SOC, we plot R_CT vs. E under different current densities in Fig. 8. In general, the overall trend of each curve resembles Fig. 3 of the H₂O–ZnSO₄ system, i.e., they can be divided into three stages, featuring monotonically decreasing, peaking and flattening of R_CT with SOC. However, the magnitude of R_CT for the V³⁺ ⇌ V⁴⁺ is significantly higher (by nearly 2×) in the D₂O–ZnSO₄ system than the H₂O–ZnSO₄ counterpart, whereas R_CT of the V⁴⁺ ⇌ V⁵⁺ is close for the two electrolyte systems. This observation further supports the assertion that it is H⁺/D⁺ that are mainly responsible for the V³⁺ ⇌ V⁴⁺ redox reaction.

Fig. 8 D₂O–ZnSO₄ system: R_CT vs. E during (a) charge and (b) discharge cycle under different current densities.

The dependence of R_CT and η on the applied current density for the D₂O–ZnSO₄ system is plotted in Fig. 9(a) and (b), where R_CT for the charge and discharge cycle at high E is seen to remain close but notably different at lower E, i.e. R_CT during the charge is lower than the discharge at a given SOC, which is like the H₂O–ZnSO₄ system.

Fig. 9 D₂O–ZnSO₄ system: (a) R_CT vs. i; (b) η vs. i and (c) i₀ at 0.66 V and 0.92 V during charge and 0.90 V and 0.62 V during discharge, with relative percentage of decrease compared to the H₂O–ZnSO₄ system.

Using the “low-field” approximation, the obtained i₀ of the D₂O–ZnSO₄ system is shown in Fig. 9(c). The results indicate a greater decrease in i₀ than the H₂O–ZnSO₄ system, i.e., it is −38% and −42% for the charge and discharge cycle, respectively, for the V³⁺ ⇌ V⁴⁺ occurring at low E and it is −25% and −23% for the charge and discharge cycle, respectively, for the V⁴⁺ ⇌ V⁵⁺ occurring at high E. Such an increased difference could be attributed to the slower D⁺ (de)insertion kinetics than H⁺.

3. Conclusions

By applying a combined GITT-EIS approach, we quantitatively characterized the electrokinetics of NaCaVO cathode in AZIBs for the first time. Analysis of the charge transfer resistance (R_CT) and exchange current density (i₀) reveals that the V³⁺ ⇌ V⁴⁺ redox reaction, which occurs at lower potentials during cycling, is intrinsically more sluggish than the V⁴⁺ ⇌ V⁵⁺ counterpart at higher potentials. This difference is likely attributed to two possible factors: (1) a higher energy barrier for charge transfer at a deeper SOC, and/or (2) the formation of an intermediate Zn-LDH phase induced by the H⁺ insertion. The possibility of H⁺ insertion into the NaCaVO structure is further supported by XRD and XPS analyses, as well as by the observed increase in R_CT upon replacing solvent H₂O with D₂O in the electrolyte. Since H⁺ insertion into oxide-based cathodes is inevitable, the resultant Zn-LDH formation can also be viewed as an alternative way of Zn-ion storage. Overall, this study provides important fundamental insights into ion storage mechanisms in V₂O₅-based cathodes and offers electrokinetic parameters for future computational modeling of AZIB systems.

4. Experimental methods

4.1 Cathode synthesis and ink preparation

The NaCaVO cathode is synthesized using a hydrothermal method we developed previously.³⁶ Briefly, VCl₃ and Ca(NO₃)₂·6H₂O are mixed in deionized water with the pH adjusted to 7. The solution is then thermally treated at 190 °C for 24 hours in an autoclave, followed by washing and drying the solid precipitates at 60 °C for 12 hours. The synthesized material is characterized using X-ray diffraction (XRD) to confirm phase purity.

NaCaVO cathode ink is prepared by thoroughly mixing 66 wt% NaCaVO with 20 wt% Super-P and 14 wt% polyvinylidene fluoride (PVDF) in N-methyl pyrrolidone (NMP) solvent. The resultant slurry is then coated uniformly onto ϕ10 mm stainless steel meshes with ∼0.6 mg cm⁻² active mass loading, followed by vacuum drying at 120 °C for about 12 h and compression under 10 MPa. The aqueous H₂O–ZnSO₄ and D₂O–ZnSO₄ electrolytes are prepared by dissolving 2 M ZnSO₄–7H₂O (Sigma-Aldrich, ACS reagent, 99%) in DI water and D₂O (Sigma-Aldrich, 99.9 atom% D), respectively.

4.2 Electrochemical measurements

In this study, a 3-electrode electrochemical cell is constructed to characterize electrochemical behaviors of the NaCaVO cathode,⁵¹ in which NaCaVO is used as the WE, a zinc metal as CE, an Ag/AgCl as RE, and 2 M ZnSO₄ as the electrolyte. During the measurement, Cyclic Voltammetry (CV) is first performed at different scanning rates (mV s⁻¹) to determine the number of redox reactions and their respective potentials. Once the redox potential is determined, we then cycle the NaCaVO–Zn 3-electrode cell 5 times from 1.5 V to 0.4 V vs. Zn/Zn²⁺ at 0.2A g⁻¹ to fully activate the cathode. We then perform stepwise GITT-EIS measurements within a voltage window of 1.5 V and 0.4 V vs. Zn/Zn²⁺ to simulate charge and discharge cycles. Each GITT-EIS run consists of three phases. Phase-1: a ten-minute constant current polarization at 50 mA g⁻¹, either positive (charge) or negative (discharge), is applied to mockup galvanic charge/discharge operation in a battery; this operation creates many potential intervals representing different SOCs (see later Fig. 2(a)). Phase-2: EIS measurements are carried out right after the galvanic interruption under OCV as well as DC bias current in the range of ±50 to ±125 mA g⁻¹ in step of ±25 mA g⁻¹ within a frequency range of 1 M Hz to 1 Hz with a base AC amplitude of 1 mA. The DC bias current is superimposed on the AC signal. Phase-3: there is a 1-minute resting period under OCV between each bias current. From the measured EIS spectra, R_CT under different bias are obtained using equivalent circuits. The obtained R_CT–i relationship is then used to attain overpotential (η)–i through integration, from which i₀ of NaCaVO at a specific potential can be obtained from the “low-field” approximation of Butler–Volmer equation given the low current density range used in this study. In addition, we also use DRT method to help analyze EIS data and facilitate the identification of elementary steps.

To illustrate the phase compositions and the variations of V-oxidation state with SOCs, we employe XRD and XPS, respectively, to examine NaCaVO cycled at high voltage range (1.5–0.75 V at 0.2 A g⁻¹ for 50 cycles) and low voltage range (0.75–0.4 V at 0.2 A g⁻¹ for 50 cycles). The cycled samples are thoroughly rinsed with DI water before examinations to avoid interference from the electrolyte.

4.3 Materials characterization

The XRD is performed on fresh and post-tested NaCaVO samples using a Rigaku D/MAX-2100 with Cu Kα radiation (λ = 1.5418 Å). The data are recorded from 5 to 80° with an interval of 0.02° and a scan speed of 2° min⁻¹.

The surface oxidation states of the sample are investigated with AXIS Ultra DLD XPS (Kratos Analytical) instrument. The XPS system is equipped with a non-monochromatic Al Kα source (1486.8 eV) operated at 150 W X-ray gun power, a hemispherical analyzer and a load lock chamber for rapid introduction of samples without breaking vacuum. The X-rays were incident at an angle of 45°, with respect to the surface normal. Analysis was performed at a pressure of ∼1 × 10⁻⁹ Torr and high-resolution core level spectra were acquired in the constant analyzer energy mode using a pass energy of 10 eV and a 0.07 eV step size (for survey scans 80 eV pass energy is used with 0.08 eV step size). The XPS experiments were performed by using a low energy electron beam, directed at the sample, for charge neutralization. The Binding energies (BE) of all peaks are corrected in reference to the C ls peak at 285 eV and given with an accuracy of ±0.2 eV. The curve fitting procedure was carried out using the Avantage software.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was sponsored by the U.S. Army Research Office (ARO) and was accomplished under Grant Number W911NF-21-1-0308 and the U.S. Air Force Office of Scientific Research under Grant Number FA9550-23-1-0505. The views and conclusions in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of ARO or AFOSR or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes, notwithstanding any copyright notation herein. The XPS facility receives financial support from the Office of the Vice President for Research, University of South Carolina.

References

1. G. Li, L. Sun, S. Zhang, C. Zhang, H. Jin, K. Davey, G. Liang, S. Liu, J. Mao and Z. Guo, Adv. Funct. Mater., 2024, 34, 2301291.
2. S. W. Gourley, R. Brown, B. D. Adams and D. Higgins, Joule, 2023, 7, 1415–1436.
3. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657.
4. B. Dunn, H. Kamath and J.-M. Tarascon, Science, 2011, 334, 928–935.
5. Z. Yang, J. Zhang, M. C. Kintner-Meyer, X. Lu, D. Choi, J. P. Lemmon and J. Liu, Chem. Rev., 2011, 111, 3577–3613.
6. C. Xia, J. Guo, P. Li, X. Zhang and H. N. Alshareef, Angew. Chem., 2018, 130, 4007–4012.
7. K. Zhu, T. Wu, S. Sun, Y. Wen and K. Huang, ChemElectroChem, 2020, 7, 2714–2734.
8. K. Zhu, T. Wu, W. van den Bergh, M. Stefik and K. Huang, ACS Nano, 2021, 15, 10678–10688.
9. Y. Lu, T. Zhu, W. van den Bergh, M. Stefik and K. Huang, Angew. Chem., Int. Ed., 2020, 59, 17004–17011.
10. K. Zhu, T. Wu, S. Sun, W. van den Bergh, M. Stefik and K. Huang, Energy Storage Mater., 2020, 29, 60–70.
11. J. J. Ye, P. H. Li, Z. Hou, W. Zhang, W. Zhu, S. Jin and H. Ji, Angew. Chem., 2024, 136, e202410900.
12. C. Geng, P. Zhang, J.-M. Wu, J. Qin and W. Wen, ACS Nano, 2024, 18, 33119–33130.
13. W. Sun, F. Wang, S. Hou, C. Yang, X. Fan, Z. Ma, T. Gao, F. Han, R. Hu and M. Zhu, J. Am. Chem. Soc., 2017, 139, 9775–9778.
14. X. Gao, H. Wu, W. Li, Y. Tian, Y. Zhang, H. Wu, L. Yang, G. Zou, H. Hou and X. Ji, Small, 2020, 16, 1905842.
15. B. Zhang, P. Dong, S. Yuan, Y. Zhang, Y. Zhang and Y. Wang, Chem Bio Eng., 2024, 1, 113–132.
16. D. S. Liu, Z. Zhang, Y. Zhang, M. Ye, S. Huang, S. You, Z. Du, J. He, Z. Wen and Y. Tang, Angew. Chem., Int. Ed., 2023, 62, e202215385.
17. B. Tang, L. Shan, S. Liang and J. Zhou, Energy Environ. Sci., 2019, 12, 3288–3304.
18. S. Zuo, X. Xu, S. Ji, Z. Wang, Z. Liu and J. Liu, Chem.–Eur. J., 2021, 27, 830–860.
19. B. Yong, D. Ma, Y. Wang, H. Mi, C. He and P. Zhang, Adv. Energy Mater., 2020, 10, 2002354.
20. Y. Lu, T. Zhu, W. van den Bergh, M. Stefik and K. Huang, Angew. Chem., 2020, 132, 17152–17159.
21. V. Mathew, B. Sambandam, S. Kim, S. Kim, S. Park, S. Lee, M. H. Alfaruqi, V. Soundharrajan, S. Islam and D. Y. Putro, ACS Energy Lett., 2020, 5, 2376–2400.
22. X. Gao, H. Dong, C. Su, Y. Dai, Y. Liu, I. P. Parkin, C. J. Carmalt and G. J. E. He, Energy Environ. Sci., 2025, 18, 13–18.
23. X. Li, J. Xiang, L. Qiu, X. Chen, Y. Zhao, Y. Wang, Q. Yue, T. Gao, W. Liu and D. Xiao, J. Energy Chem., 2025, 100, 770–778.
24. P. He, G. Zhang, X. Liao, M. Yan, X. Xu, Q. An, J. Liu and L. Mai, Adv. Energy Mater., 2018, 8, 1702463.
25. Y. Yang, Y. Tang, G. Fang, L. Shan, J. Guo, W. Zhang, C. Wang, L. Wang, J. Zhou and S. Liang, Energy Environ. Sci., 2018, 11, 3157–3162.
26. C. Liu, Z. Neale, J. Zheng, X. Jia, J. Huang, M. Yan, M. Tian, M. Wang, J. Yang and G. Cao, Energy Environ. Sci., 2019, 12, 2273–2285.
27. C. Xia, J. Guo, P. Li, X. Zhang and H. N. Alshareef, Angew. Chem., Int. Ed., 2018, 130, 4007–4012.
28. F. Ming, H. Liang, Y. Lei, S. Kandambeth, M. Eddaoudi and H. N. Alshareef, ACS Energy Lett., 2018, 3, 2602–2609.
29. T. Lv, G. Zhu, S. Dong, Q. Kong, Y. Peng, S. Jiang, G. Zhang, Z. Yang, S. Yang and X. Dong, Angew. Chem., Int. Ed., 2023, 135, e202216089.
30. Y. Tan, F. An, Y. Liu, S. Li, P. He, N. Zhang, P. Li and X. Qu, J. Power Sources, 2021, 492, 229655.
31. W. Weppner and R. A. J. Huggins, J. Electrochem. Soc., 1977, 124, 1569.
32. E. Deiss, Electrochim. Acta, 2002, 47, 4027–4034.
33. S.-I. Pyun and J.-S. Bae, Electrochim. Acta, 1996, 41, 919–925.
34. X. Chen, L. Wang, H. Li, F. Cheng and J. Chen, J. Energy Chem., 2019, 38, 20–25.
35. B. E. J. Conway, J. Electrochem. Soc., 1991, 138, 1539.
36. K. Zhu, T. Wu and K. Huang, Adv. Energy Mater., 2019, 9, 1901968.
37. N. Zhang, Y. Dong, M. Jia, X. Bian, Y. Wang, M. Qiu, J. Xu, Y. Liu, L. Jiao and F. Cheng, ACS Energy Lett., 2018, 3, 1366–1372.
38. M. Yan, P. He, Y. Chen, S. Wang, Q. Wei, K. Zhao, X. Xu, Q. An, Y. Shuang and Y. Shao, Adv. Mater., 2018, 30, 1703725.
39. S. Deng, Z. Yuan, Z. Tie, C. Wang, L. Song and Z. Niu, Angew. Chem., Int. Ed., 2020, 59, 22002–22006.
40. J. Kim, S. H. Lee, C. Park, H. S. Kim, J. H. Park, K. Y. Chung and H. Ahn, Adv. Funct. Mater., 2021, 31, 2100005.
41. J. Ding, Z. Du, B. Li, L. Wang, S. Wang, Y. Gong and S. Yang, Adv. Mater., 2019, 31, 1904369.
42. A. J. Bard, L. R. Faulkner and H. S. White, Electrochemical Methods: Fundamentals and Applications, John Wiley & Sons, 2022.
43. B. He, Y. Ling, Z. Wang, W. Gong, Z. Wang, Y. Liu, T. Zhou, T. Xiong, S. Wang and Y. Wang, eScience, 2024, 4, 100293.
44. T. F. Fuller and J. N. Harb, Electrochemical Engineering, John Wiley & Sons, 2018.
45. T. H. Wan, M. Saccoccio, C. Chen and F. Ciucci, Electrochim. Acta, 2015, 184, 483–499.
46. Z. Li, S. Ganapathy, Y. Xu, Z. Zhou, M. Sarilar and M. Wagemaker, Adv. Energy Mater., 2019, 9, 1900237.
47. A. Macrelli, M. Olivieri, A. Lamperti, V. Russo, B. Bozzini, M. Menegazzo, G. Bussetti, C. S. Casari and A. L. Bassi, Electrochim. Acta, 2023, 442, 141909.
48. M. Liu, Z. Li and Y. Zhang, J. Electroanal. Chem., 2023, 942, 117539.
49. F. Gong, Y. Feng, Y.-H. Fang, Y.-K. Hsu and Y.-C. Chen, ACS Appl. Mater. Interfaces, 2023, 15, 18808–18818.
50. D. Zhang, J. Cao, Y. Yue, T. Pakornchote, T. Bovornratanaraks, J. Han, X. Zhang, J. Qin and Y. Huang, ACS Appl. Mater. Interfaces, 2021, 13, 38416–38424.
51. S. Sun, A. Billings, B. Wang and K. Huang, ACS Electrochem., 2024, 1, 195–204.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta04992j

---