The pursuit of sustainable and cost-effective energy storage solutions has positioned sodium-ion battery technology as a promising alternative to lithium-ion systems, particularly for large-scale grid storage. The abundance and low cost of sodium resources are key advantages. However, the development of high-performance anode materials remains a central challenge. Graphite, the ubiquitous anode in lithium-ion batteries, exhibits poor sodium storage capability due to thermodynamic limitations. In this context, hard carbon materials have emerged as the most promising anode candidates for sodium-ion battery applications, offering considerable reversible capacity through a combination of adsorption on defect sites, intercalation into pseudo-graphitic domains, and pore filling.
Biomass precursors are ideal for hard carbon synthesis due to their natural abundance, renewability, and complex hierarchical structures that can translate into favorable carbon architectures. Among various biomass sources, wood, especially fast-growing species, offers significant advantages including short cultivation cycles, high yield, and consistent quality. Balsa wood (Ochroma pyramidale) is a quintessential fast-growing timber with low inherent inorganic impurity content, which simplifies the production process by potentially eliminating the need for acid-washing steps typically required for other biomass. Despite these advantages, hard carbon derived directly from carbonized balsa wood often suffers from a critical drawback: a low initial Coulombic efficiency (ICE). An ICE of approximately 70%, as commonly observed, implies that a significant portion of the sodium ions extracted from the cathode during the first charge is irreversibly consumed to form the solid electrolyte interphase (SEI) and trapped within the anode structure. This irreversible loss depletes the finite sodium inventory in the cell, drastically reducing the practical energy density of the full sodium-ion battery.
The low ICE is primarily attributed to the highly porous and disordered structure of biomass-derived hard carbons. The large specific surface area provides extensive sites for electrolyte decomposition, while the abundance of defects and micropores can lead to irreversible sodium trapping. Therefore, engineering the structure of balsa wood-derived hard carbon to reduce specific surface area, minimize unstable defect sites, and enhance structural ordering is paramount for improving ICE without compromising its high capacity.
This work presents a strategic approach to significantly enhance the ICE of balsa wood-based hard carbon anodes for sodium-ion battery. The methodology involves two key steps: first, the identification of an optimal diglyme-based electrolyte system that provides stable electrochemical operation; and second, a material modification process using graphite oxide (GO) hydroSol vacuum impregnation followed by co-carbonization. The GO acts as a structural guide and pore-filling agent, leading to a hard carbon with reduced surface area, smaller interlayer spacing, and improved stacking order. The synergistic effect of the optimized electrolyte and the modified carbon structure results in a dramatic increase in ICE alongside excellent rate capability and cycling stability, paving the way for the practical application of this sustainable anode material in high-performance sodium-ion battery systems.
Electrolyte System Screening and Optimization
The electrochemical performance of a carbonaceous anode is intrinsically linked to the stability and compatibility of the electrolyte system. For sodium-ion battery anodes, ether-based electrolytes, particularly those using diglyme (diethylene glycol dimethyl ether) as a solvent, are known for their superior stability with sodium metal and certain anode materials, forming a thinner and more stable SEI compared to conventional carbonate esters. We evaluated three common sodium salts dissolved in diglyme: sodium hexafluorophosphate (NaPF6), sodium trifluoromethanesulfonate (NaCF3SO3), and sodium perchlorate (NaClO4), all at a concentration of 1 mol L-1.
The initial electrochemical characterization of the pristine balsa wood-derived hard carbon (BHC) revealed distinct behaviors in these electrolytes. While the reversible specific capacities in all three systems were similarly high (~230 mAh g-1 at 1 A g-1), the ICE values differed markedly. The ICE was around 70% in both NaPF6/diglyme and NaCF3SO3/diglyme, but significantly lower at 65.6% in NaClO4/diglyme. Furthermore, cyclic voltammetry (CV) curves at 0.1 mV s-1 showed distorted shapes for the NaClO4-based system, and subsequent rate performance tests indicated severe capacity degradation and erratic cycling behavior at high current densities, suggesting ongoing electrolyte decomposition.
To understand the root cause of this instability, quantum chemical calculations were performed to determine the HOMO-LUMO energy gaps of the electrolyte species. The energy gap is a crude indicator of kinetic stability; a smaller gap generally correlates with higher reactivity. The calculated HOMO-LUMO gaps are summarized below:
$$ \text{HOMO-LUMO Gap (NaClO}_4\text{/diglyme)} = 6.670 \text{ eV} $$
$$ \text{HOMO-LUMO Gap (NaCF}_3\text{SO}_3\text{/diglyme)} = 6.691 \text{ eV} $$
$$ \text{HOMO-LUMO Gap (NaPF}_6\text{/diglyme)} = 7.768 \text{ eV} $$
The NaClO4-based system possesses the smallest energy gap, rationalizing its propensity for reductive decomposition on the carbon anode surface, which consumes sodium ions irreversibly and leads to poor ICE and cycling stability.
A deeper kinetic analysis of the charge storage mechanism was conducted by examining the CV response at varying scan rates. The current (i) obeys a power-law relationship with the scan rate (v):
$$ i = k v^b $$
where the b-value indicates the storage mechanism: b ≈ 0.5 suggests diffusion-controlled intercalation, while b ≈ 1 suggests surface-controlled capacitive behavior. Analysis of the redox peaks in the low-potential region yielded b-values of 0.56 for NaPF6/diglyme and 0.43 for NaCF3SO3/diglyme, confirming a diffusion-dominated process in both. However, the CV peaks for the NaPF6 system showed less potential shift and maintained higher peak currents at fast scan rates (e.g., 2, 5 mV s-1) compared to the NaCF3SO3 system. This indicates faster ion diffusion kinetics and better high-rate capability in the NaPF6/diglyme electrolyte, which was corroborated by superior capacity retention at 5 A g-1 during rate testing. Therefore, 1M NaPF6 in diglyme was selected as the optimal electrolyte for further studies due to its combination of higher kinetic stability (largest HOMO-LUMO gap) and favorable ion transport properties.
| Electrolyte (1M in Diglyme) | Initial Coulombic Efficiency (%) @ 1 A g-1 | HOMO-LUMO Gap (eV) | b-value (from CV) | Stability Observation |
|---|---|---|---|---|
| NaPF6 | ~70 | 7.768 | 0.56 | Stable CV, good rate performance |
| NaCF3SO3 | ~70 | 6.691 | 0.43 | Stable CV, poorer high-rate capacity |
| NaClO4 | 65.6 | 6.670 | N/A (unstable) | Distorted CV, severe decomposition |
Material Modification via Graphite Oxide Impregnation and Co-carbonization
With a stable electrolyte identified, the focus shifted to modifying the balsa wood precursor to address its intrinsically low ICE. The strategy involved impregnating the porous balsa wood powder with a dilute graphite oxide (GO) hydroSol under vacuum. This process allows the GO sheets, rich in oxygen-containing functional groups, to uniformly coat the internal surface of the wood’s vascular structure and pores. Subsequent high-temperature carbonization at 1100°C under an inert atmosphere serves a dual purpose: it converts the balsa wood lignin and cellulose into hard carbon while simultaneously reducing the GO coating to thermally reduced graphene oxide (rGO)-like structures. This co-carbonization process is critical for structural modification.
The modified hard carbon (denoted as M-BHC) exhibits a distinctly different microstructure compared to the pristine BHC. Scanning electron microscopy reveals that while BHC has relatively smooth surfaces, M-BHC surfaces are covered with fine wrinkles and folds characteristic of stacked graphene-like layers, indicative of the rGO coating derived from the impregnated GO.
Nitrogen physisorption analysis provides quantitative evidence of the structural changes. The specific surface area, calculated using the Brunauer-Emmett-Teller (BET) theory, plummets from 41.6 m2 g-1 for BHC to just 7.7 m2 g-1 for M-BHC. The BET equation in its linear form is:
$$ \frac{1}{v[(P_0/P)-1]} = \frac{1}{v_m C} + \frac{C-1}{v_m C} \left( \frac{P}{P_0} \right) $$
where \(v\) is the adsorbed volume, \(P/P_0\) is the relative pressure, \(v_m\) is the monolayer capacity, and \(C\) is a constant. The pore size distribution (PSD) derived using non-local density functional theory (NLDFT) models shows a complete disappearance of micropores (pores < 2 nm) in M-BHC. This dramatic reduction in surface area and micropore volume is directly attributed to the GO/rGO layer effectively covering and sealing the inherent micropores of the balsa-derived carbon, which are primary sites for irreversible sodium trapping and excessive SEI formation.
X-ray diffraction (XRD) patterns further elucidate the structural evolution. Both materials show the broad (002) peak typical of hard carbon, signifying turbostratic disorder. The interlayer spacing (d002) can be calculated using Bragg’s law:
$$ n\lambda = 2d\sin\theta $$
where \(n=1\), \(\lambda\) is the X-ray wavelength (Cu Kα, 1.54056 Å), and \(\theta\) is the diffraction angle. The most probable d-spacing decreases from 0.385 nm for BHC to 0.367 nm for M-BHC. This contraction suggests that the presence of GO during carbonization guides the graphitization process of the biomass carbon towards a more compact and ordered arrangement, likely by acting as a structural template.
Raman spectroscopy offers complementary insights into the graphitic ordering and defect density. The spectra for both materials show prominent D band (~1357 cm-1, disorder-induced) and G band (~1590 cm-1, graphitic lattice) peaks. The intensity ratio ID/IG remains similar, indicating a comparable level of structural defects like sp3 carbon. However, a key difference emerges: M-BHC exhibits a discernible 2D band in the range of 2460-3270 cm-1, which is very faint or absent in BHC. The 2D band is a second-order overtone related to the stacking order of graphene layers. Its appearance in M-BHC provides strong evidence for the formation of better-stacked, more graphitic domains due to the co-carbonization with GO, consistent with the XRD findings.
| Sample | BET SSA (m2 g-1) | Micropore Volume (cm3 g-1) | d002 (nm) | Raman ID/IG | 2D Band Presence |
|---|---|---|---|---|---|
| BHC (Pristine) | 41.6 | 0.015 | 0.385 | ~1.0 | No |
| M-BHC (GO-modified) | 7.7 | <0.005 | 0.367 | ~1.0 | Yes |
Electrochemical Performance Evaluation in Sodium-Ion Battery
The impact of the structural modifications on the electrochemical performance as a sodium-ion battery anode was evaluated in half-cells using the optimized NaPF6/diglyme electrolyte. The improvement is profound and multifaceted.
Most notably, the Initial Coulombic Efficiency is drastically enhanced. At a high current density of 1 A g-1, the ICE of M-BHC reaches 96.6%, a remarkable increase from the 69.6% observed for BHC. Even at a low current density of 20 mA g-1, where irreversible reactions have more time to proceed, M-BHC maintains a high ICE of 86.9%. This dramatic improvement is the direct consequence of the reduced specific surface area and sealed micropores, which minimize the exposed area for electrolyte decomposition and reduce sites for irreversible sodium ion trapping during the first discharge (sodiation) cycle.
The galvanostatic charge-discharge profiles and cyclic voltammetry confirm this. The CV curves of M-BHC show that the irreversible hump in the 0-0.9 V range, associated with SEI formation and irreversible reactions, is significantly suppressed after the first cycle compared to BHC. The reversible capacity is also improved. M-BHC delivers a specific capacity of approximately 314 mAh g-1 at 0.1 A g-1, outperforming BHC. The rate capability is excellent, with M-BHC retaining high capacities at increased current densities. Most importantly, the long-term cycling stability at 1 A g-1 is outstanding. After 500 charge-discharge cycles, M-BHC retains a capacity above 240 mAh g-1 with a Coulombic efficiency consistently near 100% throughout the test. The capacity curve shows a slight decrease followed by a gradual increase, a phenomenon sometimes observed in rGO-containing carbons attributed to the gradual activation of additional storage sites and improved wetting over cycles.
The kinetic behavior also evolves with modification. Re-evaluating the b-value from CV analysis for M-BHC yields a different value compared to BHC, indicating that the charge storage dynamics have been altered by the structural changes, likely due to the enhanced graphitic ordering and new interface properties introduced by the rGO component.
The overall electrochemical performance metrics are summarized below:
| Performance Metric | Value / Observation | Conditions |
|---|---|---|
| Initial Coulombic Efficiency (ICE) | 86.9% | 20 mA g-1 |
| Initial Coulombic Efficiency (ICE) | 96.6% | 1 A g-1 |
| Reversible Capacity | ~314 mAh g-1 | 0.1 A g-1 |
| Rate Capability | High capacity retention | 0.1 – 5 A g-1 |
| Cycling Stability (Capacity Retention) | >240 mAh g-1 after 500 cycles | 1 A g-1 |
| Average Coulombic Efficiency (Cycling) | ~99.9% | Over 500 cycles @ 1 A g-1 |
Conclusion
This study demonstrates a highly effective and scalable strategy to overcome the major limitation of low initial Coulombic efficiency in fast-growing wood-derived hard carbon anodes for sodium-ion battery. By first identifying NaPF6/diglyme as an electrolyte system offering superior stability and kinetics, and subsequently implementing a graphite oxide hydroSol vacuum impregnation and co-carbonization process, the structure of balsa wood-derived carbon was successfully engineered.
The modification strategy led to a hard carbon material (M-BHC) with a drastically reduced specific surface area (from 41.6 to 7.7 m2 g-1), elimination of micropores, smaller interlayer spacing (0.367 nm vs. 0.385 nm), and improved graphitic stacking order, as evidenced by the appearance of a Raman 2D band. These structural changes directly address the root causes of low ICE: reduced surface area limits electrolyte decomposition, while fewer micropores and a more ordered structure decrease irreversible sodium trapping.
The electrochemical results are compelling. The ICE was elevated to 96.6% at 1 A g-1, a critical advancement for practical energy density. Concurrently, the reversible capacity was improved to 314 mAh g-1 at 0.1 A g-1, and exceptional long-term cycling stability was achieved, retaining over 240 mAh g-1 after 500 cycles at a high current density. This work validates that sustainable biomass like fast-growing balsa wood can be transformed into high-performance hard carbon anodes for sodium-ion battery through rational structural design. The GO-assisted modification method presented here is not merely a coating process but a synergistic co-carbonization that fundamentally alters the carbonization pathway of the biomass, offering a generalizable route to enhance the performance of various porous carbon precursors for energy storage applications.