Phase-engineered cathode for super-stable potassium storage

The study addresses how crystal phase engineering influences the electrochemical performance and stability of potassium-ion battery (PIB) cathodes. While numerous strategies (doping, coating, morphology/size control) have been explored, atomic-scale dispersion and phase state critically determine nanomaterial properties. Amorphous phases can offer open architectures, abundant defects, isotropy, higher pseudocapacitive contributions, and more uniform volume accommodation, potentially improving ion transport and durability. Vanadium oxides (VOx) favor intercalation over conversion, mitigating volume change, and VO2 is known for rapid ion diffusion and high rate capability across Li-, Na-, and Zn-ion systems. This work investigates VO2 with controlled phases (amorphous, B, and M) supported on conductive carbon fiber cloth to evaluate K+ storage, hypothesizing that amorphization will enhance K insertion sites, diffusion, and stability compared to crystalline counterparts.

Prior research on VO2 as an electrode has largely focused on crystalline M and B phases due to their tunnel and layered structures that enable fast ion diffusion in Li-, Na-, and Zn-ion batteries. The M phase (monoclinic P2_1/c) features corner-sharing VO6 octahedra with alternating V4+–V4+ pair distances; the B phase (triclinic C2/m) comprises edge-sharing VO6 bilayers with 1D tunnels. VOx generally undergo intercalation rather than conversion, reducing volume expansion. Vanadium’s electronic structure (strong correlation, three d electrons) provides versatile properties and avoids Jahn–Teller effects typical of Mn3+. However, for PIBs, reports suggest challenges with crystalline VO2 phases for K storage. Amorphous phases in other materials have shown advantages for cation storage due to open networks and defect-rich structures, with theoretical indications of better insertion energetics and kinetics; nonetheless, more experimental validation combined with theory is needed to confirm these benefits in VO2 for K-ion storage.

Materials synthesis: VO2(a) on carbon fiber cloth (CF) was prepared hydrothermally by reacting 1 mmol V2O5 with 3.3 mmol citric acid in water at 80 °C for 30 min, adding 2.5 mL H2O2 (30 wt%), 25 mL alcohol, and a CF substrate, then sealing in a 50 mL Teflon autoclave at 180 °C for 3 h. Product was washed and vacuum-dried at 60 °C for 12 h. VO2(M) on CF used oxalic acid instead of citric acid followed by annealing at 500 °C for 4 h under Ar. VO2(B) on CF used oxalic acid with all precursors tripled (no anneal). Amorphous VO2 powder was synthesized separately by precipitating from VOSO4·xH2O with NaOH, drying at 80 °C, and calcining at 300 °C for 8 h; B and M phase powders were similarly prepared without CF. Characterization: XRD (Bruker D8, Cu Kα) identified phases; in situ XRD used a Be-disc-based cell under 1–4 V at 20–50 mA g−1. Morphology and microstructure were analyzed by FE-SEM (JEOL JSM-7610FPlus, EDS) and STEM/HRTEM (Thermo Talos F200X, 200 kV), including FFTs and simulated diffraction (CrysTBox). XPS (SHIMADZU AXIS SUPRA+) analyzed V valence states. True density of amorphous VO2 powders was measured by AccuPyc II 1340. Strain mapping used geometric phase analysis (GPA) on HRTEM images. Electrochemical testing: 2025 coin cells were assembled in Ar glovebox with as-prepared freestanding VO2/CF as cathode (loading ~0.8 mg cm−2, 12 mm diameter), K metal counter, glass microfiber separator, and 2.5 M KFSI in triethyl phosphate electrolyte. Voltage window: 1–4 V (half cells). Galvanostatic cycling/rate/GITT tests were run on Neware BTS-53 at 28 ± 0.5 °C. CV and EIS (Nyquist, 10−1 to 10^5 Hz) were measured on CHI660e. Control experiments included pure CF, and VO2 powders coated on Al foil. Full soft-pack batteries used VO2(a) cathode and pre-potassiated nano-graphite anode (80:10:10 nano-graphite:Super P:CMC on Al), pre-cycled 5 times vs K and discharged to 0.001 V to compensate irreversible capacity. N/P ~1.05; electrolyte same as half-cell; voltage window 0.7–3.8 V; bending tests performed with LED demonstration. Computations: An amorphous VO2 model was generated via ab initio molecular dynamics (AIMD). K insertion sites in VO2(a) were sampled by selecting 49 uniformly distributed positions (>1.45 Å from V/O), optimized to 25 distinct stable sites. For VO2(B) and VO2(M), three non-equivalent insertion sites were considered. Binding energies (relative to bulk K) were computed; migration barriers were obtained via climbing-image nudged elastic band (CI-NEB) along representative diffusion pathways: three long-range paths in VO2(a), main b-axis channel in VO2(B), and tunnel path in VO2(M). Electronic structures (total and projected DOS) were calculated before and after K insertion to assess conductivity changes. A diffusion-induced stress model compared equivalent stress distributions in particles for the three phases under identical K+ diffusion time (20 ms). Kinetics were further assessed by CV b-value analysis and GITT to extract chemical diffusion coefficients (Dk).

• Structural-phase dependence: TEM/EDS mapping shows K uniformly inserts into amorphous VO2, while VO2(M) exhibits K segregation in localized amorphous regions and largely remains crystalline elsewhere; VO2(B) remains crystalline with homogeneous K distribution but undergoes strain.
• Binding energetics: DFT shows all K insertion sites in VO2(M) have positive binding energies vs bulk K (thermodynamically unfavorable). VO2(B) has one eightfold-coordinated stable site but with higher energy than most amorphous sites. VO2(a) provides many favorable insertion sites with negative binding energies, confirming superior K accommodation.
• Diffusion barriers: Calculated K+ migration energies (eV): VO2(a) paths 2.15, 2.74, 1.85; VO2(B) 3.54; VO2(M) high (reported for reference due to no stable sites). Amorphous disorder locally widens channels, lowering barriers.
• Electronic structure: DOS indicates pristine VO2(a) is spin-polarized metallic (no gap at EF), VO2(B) is semi-metal with ~0.29 eV gap, VO2(M) is semiconducting with ~0.99 eV gap (reduced with localized states upon K insertion). VO2(a) thus offers better electronic conductivity.
• Electrochemical performance (half cells): • VO2(a): Capacity up to 111 mAh g−1 at 20 mA g−1 after activation (~200 cycles). Rate capability: 107 → 70 mAh g−1 from 20 to 500 mA g−1, returning to ~106 mAh g−1 when current is reduced. Long-term cycling: at 100 mA g−1 retains 56 mAh g−1 after 3700 cycles (>8 months); at 500 mA g−1 retains ~52 mAh g−1 after 8500 cycles (≈80% retention from ~60 mAh g−1 at 500 cycles). High coulombic efficiency throughout. • VO2(B) and VO2(M): Poor performance; at 500 mA g−1 discharge capacities ~39 and 5 mAh g−1, respectively. At 100 mA g−1, capacities ~50 and 25 mAh g−1 with rising CE but eventual cell failure. • VO2 powders (on Al): Amorphous VO2 powder outperforms crystalline powders; after 3500 cycles at 100 mA g−1, retains 46.3 mAh g−1 (slightly lower than VO2(a)/CF).
• Structural stability: In situ XRD shows VO2(a) remains amorphous without new crystalline peaks during cycling; VO2(B) peaks shift (lattice expansion on discharge) and recover on charge; VO2(M) peaks do not shift, indicating negligible K insertion.
• Kinetics: CV b-values: VO2(a) ~0.93 (dominant pseudocapacitive behavior); VO2(B) ~0.85; VO2(M) ~0.88 (likely influenced by residual amorphous regions). Pseudocapacitive contribution for VO2(a) increases from 83% to 94% when scan rate increases 0.1 → 1.0 mV s−1 (higher than VO2(B): 55%→79%).
• GITT diffusion coefficients: Average Dk ~10−11 cm2 s−1 for VO2(a); ~10−12 cm2 s−1 for VO2(B); ~10−10 cm2 s−1 reported for VO2(M) first cycle segments, but overall K insertion is not sustained; VO2(a) exhibits smallest polarization and best kinetics.
• EIS: Charge-transfer resistance (Rct) decreases with cycling for all, with VO2(a) consistently lower Rct than VO2(B)/VO2(M), indicating improved electronic/ionic transport.
• Stress analysis: Diffusion-induced stress modeling shows uniform, low stress in VO2(a) due to isotropy; VO2(B) develops anisotropic stress with concentrations near boundaries, hindering diffusion; VO2(M) shows large von Mises stress if K is forced in, consistent with instability. GPA strain mapping of VO2(B) after 100 cycles reveals significant εyy along K insertion direction.
• Full cell performance (VO2(a)/pre-potassiated nano-graphite): Reversible capacity 119 → 67 mAh g−1 from 20 to 500 mA g−1 (based on cathode mass). Cycling at 200 mA g−1 retains 63 mAh g−1 after 200 cycles with ~100% CE. Flexible soft-pack cells maintain LED illumination and unchanged voltage profiles after single and 1000 bends; electrode morphology remains intact.

The findings validate the hypothesis that amorphization enhances K-storage performance in VO2 by providing abundant, energetically favorable insertion sites, lower diffusion barriers, and metallic-like electronic conduction. Amorphous VO2 accommodates volume changes isotropically, minimizing diffusion-induced stress and preventing structural degradation, thereby enabling ultra-stable long-cycle operation. In contrast, crystalline VO2(B) suffers from anisotropic strain and higher migration barriers, while VO2(M) lacks stable K insertion sites and fails to function as a K-storage host. The strong pseudocapacitive contribution of VO2(a) further contributes to its high rate capability and stability, with in situ XRD confirming structural persistence during cycling. These advances demonstrate phase engineering as an effective route to optimize PIB cathodes and more broadly to tailor electrode properties for rechargeable batteries.

Phase engineering of VO2 enables a super-stable PIB cathode: amorphous VO2 delivers up to 111 mAh g−1 at 20 mA g−1, excellent rate capability, and exceptional longevity (over 8 months at 100 mA g−1; 80% retention after 8500 cycles at 500 mA g−1). Combined DFT and experiments show VO2(a) offers favorable K insertion energetics, lower migration barriers, superior electronic conductivity, and minimal diffusion-induced stress, maintaining an amorphous framework throughout cycling. Crystalline VO2(B) undergoes strain accumulation and degradation, while VO2(M) is unsuitable for K storage. This work underscores phase engineering as a powerful strategy for designing durable cathodes for PIBs and other alkali-ion batteries, providing a pathway to optimize ion transport and structural stability.