A Restacking Inhibited Carbonization Pathway for Graphitization‐Prone Precursors Toward Long‐Life Sodium‐Ion Batteries

ABSTRACT

Theory‐guided design of hard carbon anode from graphitization‐prone precursors remains challenging because oxidation and carbonization routes lack mechanisms to selectively disrupt ordered π–π stacking while preserving structural integrity during carbonation for performance. We propose an intrinsic heteroatom‐assisted site‐preferential oxidation mechanism that enables framework disruption and kinetically inhibits restacking during carbonization of petroleum asphaltenes rich in heteroatoms. Electronic inhomogeneity of nitric acid makes its acid‐derived radicals preferentially anchor on heteroatom‐modified sites, inducing steric hindrance and oxidation‐guided pore evolution that yields turbostratic hard carbon with expanded interlayer spacing and closed pores. Operando characterization and density functional theory (DFT) calculations identified this heteroatom‐mediated localized reactivity as the origin of suppressed graphitization and enhanced sodium‐storage kinetics. The resulting material delivers a high initial Coulombic efficiency (ICE) of 89.7% and a reversible capacity of 404.1 mAh g⁻¹, with a 93.2% capacity retention after 2200 cycles, outperforming most reported pitch‐derived hard carbons. Practical applicability is demonstrated in a 1.2 Ah pouch‐cell, while cradle‐to‐gate life cycle assessment (LCA) indicates substantially reduced environmental impacts as compared with representative commercial hard carbons. Beyond offering a generalizable strategy for converting low‐quality thermoplastic carbon sources into durable sodium‐ion battery anode materials, this study also offers mechanistic insights into selective carbonization pathways.

Heteroatom‐assisted nitric‐acid oxidation suppresses restacking of graphitization‐prone asphaltenes during carbonization, enabling hard carbon with expanded interlayer spacing, regulated defects, and closed pores. The resulting anode delivers high initial Coulombic efficiency, high reversible capacity, and durable cycling, while operando analysis and DFT uncover a selective carbonization pathway for sustainable sodium‐ion storage.

1. Introduction

The soaring demand for clean‐energy storage infrastructure has driven the widespread adoption of lithium‐ion batteries (LIBs) in renewable energy systems, electric vehicles and portable electronics, with their global installed capacity projected to surpass 3200 GWh by 2030 [1, 2]. However, concerns over energy security and lithium resource sustainability have stimulated growing interest in sodium‐ion batteries (SIBs), which have gradually emerged as a promising alternative to LIBs due to the abundant global reserves of sodium and lower environmental impact of raw material extraction [3, 4, 5]. Despite these advantages, two critical challenges for SIBs remain unresolved: namely, relatively poor cycle stability and the incompatibility of conventional graphite anodes [6, 7, 8, 9]. Conventional graphite anodes cannot accommodate the intercalation of Na⁺. This incompatibility arises because of its larger ionic radius than that of Li⁺ [10]. In this context, rational design of novel high‐performance and sustainable anode materials or the systematic optimization of existing candidates is considered a pivotal route to overcome these bottlenecks and accelerate the practical advancement of SIB technologies, which are poised to play a vital role in the next generation of sustainable energy‐storage systems.

Current design of hard carbons for sodium‐ion batteries is largely constrained by the lack of controllable strategies to transform graphitization‐prone carbon precursors into long range disordered structures without sacrificing structural integrity [11]. Conventional oxidation or pre‐treatment approaches typically induce the damage of random frameworks or insufficient disruption of π‐π stacking, limiting the precise regulation of interlayer configuration and formation of closed pores that govern effective sodium storage behavior [12, 13, 14]. Petroleum asphaltenes, despite their high carbon yield and industrial relevance, inherently favor ordered stacking during carbonization, making the selective disordering a persistent challenge [15, 16, 17, 18]. Developing feasible strategies guided by insightful mechanisms that enable localized framework disruption while preserving overall carbon connectivity therefore represents a critical yet unresolved problem for advancing sustainable hard carbon anode materials for SIBs.

Despite extensive efforts to tailor the structure of hard carbon materials through oxidative pre‐treatment or chemical functionalization, current approaches largely rely on non‐site‐preferential oxidation processes that either introduce excessive structural damage or fail to sufficiently disrupt the intrinsic π–π stacking of graphitization‐prone precursors [19, 20, 21, 22]. Strong oxidants typically generate random defects and unstable functional groups, leading to uncontrolled pore evolution that compromises structural stability during subsequent carbonization [23, 24]. Conversely, mild treatments often preserve the original aromatic stacking, allowing graphitic ordering to recover at elevated temperatures [25]. As a result, achieving controlled structural disorder that simultaneously suppresses restacking while maintaining carbon framework continuity remains elusive. The lack of mechanisms that enable localized and chemically guided framework disruption therefore represents a critical limitation in existing oxidation strategies for hard carbon synthesis.

Here, we demonstrate that intrinsic heteroatoms in graphitization‐prone asphaltenes can serve as chemically vulnerable sites that enable site‐preferential oxidation and localized framework disruption, providing a mechanism to kinetically suppress π–π restacking during carbonization for hard carbon formation. By coupling nitric‐acid‐mediated oxidation with heteroatom‐directed reactivity, ordered aromatic stacking can be selectively perturbed without compromising the overall integrity of the carbon frameworks, allowing controlled molecular cross‐linking and deterministic evolution of interlayer spacing and closed‐pore architecture. This heteroatom‐assisted pathway establishes a chemically guided route for transforming graphitization‐prone thermoplastic precursors into turbostratic hard carbons, revealing a previously underexplored mechanism for regulating the soft‐to‐hard carbon transition. The resulting structural configuration supports enhanced sodium storage of high reversible capacity and long‐term stability, demonstrating the potential of waste asphaltenes from petroleum refining as a sustainable precursor for the production of durable sodium‐ion battery anode materials.

2. Results and Discussion

The proposed mechanism, in which nitric‐acid oxidation regulates steric hindrance to disrupt the π–π conjugated stacking of asphaltenes and thereby optimize Na‐storage performance, was examined using three deliberately designed pathways of synthesis. The corresponding schemes are summarized in Figure 1a, and detailed synthesis parameters are provided in the Supporting Information. Each route yields a representative sample that isolates the effects of oxidation modality and intrinsic heteroatoms. In route I, raw asphaltenes were pre‐oxidized in air at 320°C for 4 h to obtain pitch‐derived hard carbon obtained by air pre‐oxidation only (AP‐PHC), serving as a non‐chemical oxidation control. Under these mild conditions, oxidation is expected to be largely surface‐confined and may be insufficient to introduce pronounced steric constraints or effectively perturb compact π–π stacking. In route II, heteroatom‐containing asphaltenes were treated with 0.5 mol L⁻¹ HNO₃ for 2 h at ambient temperature and subsequently carbonized at 1400°C to produce PHC‐0.5N. Here, the intrinsic heteroatoms are hypothesized to act as chemically vulnerable sites that guide nitric‐acid reactivity, enabling localized oxidation and steric regulation around aromatic segments, which can kinetically suppress restacking and facilitate the formation of a turbostratic hard carbon framework favorable for Na⁺ storage. In route III, a heteroatom‐free synthetic pitch was subjected to the same nitric‐acid treatment and carbonization protocol to prepare heteroatom‐free pitch‐derived hard carbon (HF‐PHC‐0.5N), which decouples the role of intrinsic heteroatoms from that of nitric acid. Comparing the properties of PHC‐0.5N with those of HF‐PHC‐0.5N allows the critical synergistic requirement of heteroatoms and nitric‐acid oxidation to be evaluated, which provides a direct basis for attributing the observed structural and electrochemical differences to heteroatom‐assisted site‐preferential oxidation rather than simple acid treatment.

FIGURE 1.

(a) Schematic illustration of pathways for the synthesis of hard carbon materials; (b) XRD patterns of PSC and PHC‐0.5N; (c) Raman spectra of PSC and PHC‐0.5N; (d) Comparison of graphite/disordered domain fractions, L _a/L _c and d ₀₀₂ obtained from XRD patterns of PSC and PHC‐0.5N; (e,f) HR‐TEM images of PSC; and (g,h) HR‐TEM images of PHC‐0.5N, with electron diffraction (SAED) patterns of selected areas shown in the inset.

The air pre‐oxidation route in Figure 1a was adopted as a control, and its processing window was first established by differential scanning calorimetry (DSC) and thermogravimetric (TG) analyses to ensure cross‐linking occurs without excessive mass loss. As shown in Figure S1, a substantial mass loss was observed between 100 and 200°C in air, which is attributed mainly to desorption of physically adsorbed water [26]. With further heating in air, raw asphaltenes exhibit a gradual decrease in mass, followed by an apparent plateau over 300 to 350°C and accompanied by a slight gain in mass toward the end of the plateau. Such a mass gain is attributed to oxygen uptake and the formation of cross‐linked oxygen‐containing functionalities [14]. In contrast, the TG trace in argon shows a negligible mass change over this temperature range, indicating a stable carbon framework without oxidative reactions. When the temperature exceeded 350°C in air, a pronounced mass loss together with multiple DSC events are observed, leading to a carbon yield below 10%. Such behavior is indicative of over‐oxidation and excessive framework degradation. Based on these results, 320°C was selected as the air pre‐oxidation temperature for raw asphaltenes.

XRD patterns and Raman spectra reveal clear structural distinctions between PSC and PHC‐0.5N (Figure 1b–d). Both samples show broad (002) diffraction features that can be deconvoluted into two contributions at about 22.0° and 25.4°, which are assigned to disordered and graphite‐like domains, respectively (Figure 1b; Figure S2) [24, 27]. Relative to PSC, PHC‐0.5N displays a shift of the (002) diffraction peak to lower angles, indicating an expanded d ₀₀₂ that results from selective disruption of π–π conjugated stacking. The fraction of disordered domains increases from 21.2% for PSC to 39.1% for PHC‐0.5N (Table S1). This increase is attributed to nitric acid‐regulated steric effects, accompanied by oxidation‐induced defect formation and oxygen incorporation [26, 28]. Furthermore, Raman spectra in Figures 1c and S3 and S4 reveal a more pronounced disordering of PHC‐0.5N compared with PSC. The I _D/I _G ratio increases from 0.13 for PSC to 1.73 for PHC‐0.5N, indicating a substantially higher density of defects and a more disrupted carbon network. In contrast, the I _2D/I _G ratio decreases, which is indicative of a weakened graphitic stacking. As the nitric acid concentration increases, the concurrent increase of I _D/I _G and decrease of I _2D/I _G (Figure S4a) further support that oxidation by nitric acid promotes the incorporation of oxygen and the dissociation of carbon layers, thereby suppressing π–π‐induced molecular aggregation. A defect‐related component at ∼1180 cm⁻¹, represented by I _D1/I _G, shows a similar trend, suggesting an enhanced localized disorder [8, 14]. These Raman spectral features corroborate a selective, heteroatom‐mediated oxidation pathway that disrupts the molecular stacking while enabling the formation of storage‐relevant microstructures (Figure S3). Beyond the (002) shift and I _D/I _G, the disordering was further quantified by L_a and L_c , as summarized in Figure 1d and Table S1. Consistently, PHC‐0.5N exhibits a reduced fraction of graphite‐like structures and lower L _a/L _c ratios, together with a decreased L_c but only moderate changes in L _a. These results collectively suggest a preferential disruption of structures along the c‐axis that suppresses vertical restacking while preserving largely the lateral framework of carbons.

The results of high‐resolution transmission electron microscopy (HR‐TEM), SAED and scanning electron microscopy (SEM) characterization in Figures 1e–h and S5 further reveal distinct microstructural differences between PHC‐0.5N and PSC. PHC‐0.5N exhibits a significantly higher density of nanoscale pores embedded in a heterogeneous carbon matrix, composed of short‐range ordered turbostratic domains and amorphous phases. The lattice fringes in PHC‐0.5N show an expanded interlayer spacing of ∼0.349 nm, compared with that of PSC, which is dominated by extended graphite‐like domains with a narrower spacing of ∼0.341 nm, characteristic of soft carbons [6, 14, 25]. The disruption of π–π stacking, expansion of interlayer spacing and increase in nanoscale porosity in PHC‐0.5N collectively define a more open, yet continuous carbon framework, in contrast to the more ordered structure of PSC [14, 25]. These findings further highlight the mechanistic role of heteroatoms in mediating the oxidation that governs the carbonization process leading to hard‐carbon formation and drives the evolution of sodium storage sites.

The interplay between nitric‐acid‐directed oxidation and the formation of hard‐carbon microstructures was analyzed using N₂ adsorption‐desorption isotherms, X‐ray photoelectron spectroscopy (XPS) and small‐angle X‐ray scattering (SAXS), with an emphasis on stacking disruption, pore development and graphitic ordering. The XPS O 1s spectra in Figure 2a highlights chemical drivers of steric hindrance. The content of heteroatoms changes only marginally at 6.7 at.% for pristine pitch (PP) and 4.9 at.% for PHC‐0.5N (Figure S4b), without the introduction of new nitrogen species [29, 30]. However, PHC‐0.5N shows a significantly higher oxygen content as listed in Table S2 and Figure S6. The oxygen‐containing groups such as C─O and ‐COOH introduced could create steric repulsion between carbon chains, prevent re‐formation of compact π‐π stacking and preserve disordered structures that ultimately improve sodium‐ion storage. N₂ adsorption‐desorption isotherms in Figures 2b and S7 and Table S3 reveal clear differences in pore structures among the samples investigated. PHC‐1N features the lowest specific surface area of 2.0 m² g⁻¹ with minimal amounts of open nano‐mesopores and macropores, while PHC‐2N and PSC exhibit larger pore volumes due to excessive etching or insufficient pre‐oxidation. PHC‐0.5N exhibits the highest differential pore volume of 0.129 cm³ g⁻¹ nm⁻¹ and a total pore volume of 0.028 cm³ g⁻¹, dominated by 3–7 nm nanopores. This nanopores‐rich feature results from the disruption of π‐π stacking, creating interlayer voids that provide extra sodium storage sites [3, 8, 12]. The results from SAXS analysis in Figures 2c and S8 further reveal nanoscale structural features. The absence of a mid‐q shoulder peak in PSC indicates the presence of only a negligible fraction of closed or semi‐closed micropores [3, 8]. In contrast, PHC‐0.5N exhibits a distinct shoulder feature that can be associated with closed pores of an average diameter of 3.33 nm and radius of gyration (R _g) of 1.29 nm (Table S3). The pronounced slope change in the low‐q region, consistent with the Porod law, suggests a complex internal structure with tortuous pore surfaces [8, 12]. These observations further support the oxidation‐induced disruption of π–π stacking, which introduces structural heterogeneity and favors the stabilization of nanoscale closed pores.

FIGURE 2.

Structural evolution of hard carbon during nitric acid oxidation. (a) High‐resolution XPS O 1s spectra of PSC and PHC‐0.5N; (b) N₂ adsorption‐desorption isotherms with corresponding pore size distributions shown in the inset; (c) SAXS profiles; (d) XRD patterns; (e) Raman spectra; (f) FT‐IR spectra; (g) High‐resolution XPS C 1s spectra with the inset showing the relative area of the fitted peaks; (h) N₂ adsorption‐desorption isotherms for uncarbonated precursor; and (i) in situ TG‐MS curves for m/z = 17.

The oxidation‐driven structural evolution was examined using XRD analysis and Raman spectroscopy. Prior to oxidation, the precursor exhibited a pronounced graphite‐like (002) diffraction peak at 26.4°, indicating strong stacking order and extensive graphitic domains (Figure 2d). Once established, such a graphitization‐prone structural state is generally difficult to reverse [18]. With increasing nitric acid treatment, the sharp (002) diffraction peak is markedly suppressed and evolves into a broad diffraction band over the 21°–28° range (Figure 2d; Figure S9a), which is indicative of disrupted stacking and increased turbostratic disorder. After subsequent carbonization, this oxidation‐induced disorder is retained, giving rise to a broadened (002) diffraction peak centered at 25.2° in PHC‐0.5N‐1400. Such a diffraction pattern indicates the presence of amorphous carbon featuring short‐range order but long‐range disorder [15, 31]. Raman spectroscopy further corroborates the disordering process (Figure 2e). Deconvolution of the Raman spectra (Figure S9b,c) shows that nitric acid treatments of 30 and 240 min increase the I _D/I _G ratio from 0.13 for the unoxidized precursor to 1.33 and 1.83, respectively, indicating a substantially increased density of defects. This evolution results from intrinsic heteroatom‐assisted oxidation that introduces localized defects and framework fragmentation, thereby kinetically suppressing restacking and graphitic ordering during subsequent carbonization [32].

FT‐IR and XPS analyses were used to track the chemical and structural evolution during nitric acid oxidation (Figure 2f,g). PHC‐0.5N‐30 and PHC‐0.5N‐240 show clear evidence of the incorporation of nitro functionalities, as indicated by the characteristic bands over the 1150–1620 cm⁻¹ spectral range (Figure 2f; Figure S9d). Oxidation also strengthened C─H‐related spectral features, which disappeared after carbonization. This feature suggests an oxidant‐induced bond cleavage and reformation process, followed by dehydrogenation and aromatization during pyrolysis [11, 33]. Notably, elevated C─H motifs are known to impede pyrolytic graphitization of petroleum‐derived carbons [23, 34]. Consistently, high‐resolution XPS C 1s spectra reveal increased C─O and C═O contributions after oxidation, indicating enhanced oxygen incorporation (Figures 2 g; Figure S9 and S10). Such oxygen incorporation can promote the formation of defects and structural reorganization of disordered carbon domains, thereby providing additional Na‐storage sites [29]. In the N 1s spectral region, distinct nitrogen species are observed for most oxidized PHC precursors but become negligible after carbonization, in agreement with the disappearance of nitrate‐related FT‐IR signatures (Figure 2f) and the evolution of XRD patterns (Figure 2d). Porosity evolution was further probed by N₂ adsorption–desorption on the uncarbonized precursors, denoted as PSC‐UC and PHC‐0.5N‐UC (Figure 2h; Figures S11 and S12). PHC‐0.5N‐UC exhibits a Type IV isotherm with a pronounced hysteresis loop at 0.4‐1.0 P/P₀, whereas PSC‐UC shows a Type II isotherm associated with large open pores which likely originate from volatile release during the drying of petroleum asphaltenes [35, 36]. The pore‐size distributions (Figures S11 and S12) indicate a hierarchical porosity in PHC‐0.5N‐UC samples with dominant features at about 4 nm. Importantly, the subsequent carbonization converts part of these accessible pores into closed or semi‐closed configurations that are not fully captured by conventional N₂ adsorption, as supported by the results in Figure S12 and Table S3 and corroborated by HR‐TEM and SAXS results. Such closed or semi‐closed micropores can provide additional Na‐storage sites and facilitate rapid ion and electron transport [3, 37].

In situ TG‐MS analysis from 25 to 1000°C (Figure 2i; Figure S13) reveals distinct thermal evolution between PHC‐0.5N‐UC and PSC‐UC. PHC‐0.5N‐UC exhibits two sharp NH₃ evolution peaks at 68°C and 660°C, corresponding to the decomposition of surface nitrate species and thermally stabilized nitrogen functionalities, respectively [38, 39]. In contrast, PSC‐UC displays a broader NH₃ release between 200°C–600°C, indicative of a more gradual decomposition of nitrogen‐containing species. Notably, despite the intensified and sharper gas evolution, PHC‐0.5N‐UC retains a significantly higher residual mass of 47.97% at 1000°C, as compared with a lower value of 34.62% for PSC‐UC. This distinct difference suggests that nitric acid treatment introduces thermally reactive functional groups into the PHC precursor, which decompose over defined temperature windows while contributing to stabilization of the carbon framework at elevated temperatures [38, 40]. The higher carbon yield implies an enhanced structural condensation or cross‐linking during carbonization, which counteracts excessive mass loss [38, 40]. The concurrent detection of NH₃, NO, NO₂ and CO₂ in the released gas phase indicates that nitrate‐derived species participate in redox reactions which are capable of restructuring the carbon matrix. More importantly, the results from CO₂ adsorption analysis (Figure S14) show a dramatic increase in ultramicroporosity for PHC‐0.5N‐UC. The total pore volume below 1.05 nm reaches 0.0144 cm³ g⁻¹, nearly two orders of magnitude higher than that of PSC‐UC (0.00020 cm³ g⁻¹). Correspondingly, the surface area of ultramicropores increases from 2.9 to 62.7 m² g⁻¹. The pore‐size distribution data further confirm the generation of sub‐nanometer pores in the range of 0.4‐1.0 nm. These results collectively demonstrate that nitric acid oxidation does not merely enhance volatilization but also induces a controlled process of gas‐assisted structural reconstruction. The staged decomposition of nitrate species generates a localized gas evolution that carves ultramicropores while simultaneously promoting the condensation of carbon frameworks, leading to both increased carbon yield and substantially enriched ultramicroporosity.

HR‐TEM analysis of the uncarbonized precursors reveal distinct structural states prior to carbonization. As shown in Figure S15, PSC‐UC preserves largely parallel carbon layers with an interlayer spacing of about 0.344 nm, whereas PHC‐0.5N‐UC displays clear distorted lamellae with nascent pore features and an expanded layer spacing of around 0.365 nm, indicating early‐stage disruption of graphitizable stacking during pre‐oxidation through the synergistic action of nitric acid and the intrinsic heteroatoms in asphaltenes [6, 8, 14, 25]. When the oxidation duration is further extended to 240 min, the precursor framework becomes excessively disrupted rather than progressively stabilized, as evidenced by the highly fragmented and disordered features in Figure S16. More importantly, this over‐disruption is retained after carbonization at 1400°C. XPS survey quantification in Table S4 further shows that the residual N content after 1400°C carbonization remains very low, whereas O remains clearly detectable, suggesting that oxygen functionalities are more likely to influence the irreversible sodium consumption and interfacial chemistry, while the contribution from N is expected to be minor. Raman spectra of the PHC‐0.5N‐1400 time‐series samples in Figure S17a–c show a clear increase in the degree of disorder for the samples treated for 240 min, and Table S1 quantitatively confirms a higher I _D/I _G of 1.94 together with a markedly enlarged d ₀₀₂ of 0.368 nm for PHC‐0.5N‐240min‐1400, as compared with the samples treated for a shorter period of time. The corresponding HR‐TEM images in Figure S17d–f corroborate a carbon framework with many more defects in PHC‐0.5N‐240min‐1400, implying that prolonged oxidation leads to persistently high disorder and defect density. These results collectively imply that nitric acid pre‐oxidation in the presence of intrinsic heteroatoms in asphaltene carbon precursors can guide pore evolution and suppress restacking within an appropriate window, while excessive oxidation drives structural over‐disruption that is unfavorable for forming a regulated carbon framework.

Control experiments were performed to decouple the roles of oxidation mode and intrinsic heteroatoms in selective structural transformation into hard carbon. AP‐PHC, obtained by air pre‐oxidation alone, shows only a low reversible capacity of 124.89 mAh g⁻¹ with an ICE of 46.6% (Figure 3a; Table S5), indicating that air oxidation is insufficient to induce the selective structural disruption required for hard‐carbon formation. In contrast, PHC‐0.5N exhibits the electrochemical signature of a typical turbostratic hard carbon, delivering an initial discharge/charge capacity of 450.70/404.11 mAh g⁻¹ and a high ICE of 89.7% (Figure 3b). This transformation cannot be reproduced in the heteroatom‐deficient control HF‐PHC‐0.5N, which remains electrochemically inactive and shows a reversible capacity below 30 mAh g⁻¹ after 200 cycles at 500 mA g⁻¹ (Figure 3c; Figure S18). The contrast among AP‐PHC, PHC‐0.5N, and HF‐PHC‐0.5N indicates that neither air oxidation alone nor nitric acid treatment in the absence of intrinsic heteroatoms is sufficient to generate the structural motifs required for hard‐carbon sodium storage. Together, these controls support the view that selective hard‐carbon formation is enabled not by nitric acid treatment alone, but by its coupling with intrinsic heteroatoms in the precursor.

FIGURE 3.

Electrochemical performance and strategy validation of hard carbon anodes for SIBs. (a–c) First two GCD cycles of AP‐PHC, PHC‐0.5N, and HF‐PHC‐0.5N. (d–f) Rate capability, cycling performance, and long‐term cycling stability of PHC‐0.5N. (g) Rate performance of hard carbons derived from asphaltenes carbonized at different temperatures. (h) GCD curves of PSC‐ and PHC‐based half‐cells prepared from n‐pentane‐extracted asphaltenes. (i,j) First two GCD cycles and cycling performance of full cells with PHC‐0.5N as the anode. (k) Cycling performance of the pouch cell. (l) Performance comparison of PHC‐0.5N with previously reported hard carbon anodes for SIBs.

Half‐cell tests further confirm the superiority of PHC‐0.5N anodes. The ICE and reversible capacity of PHC‐0.5N outperform most reported asphaltene‐based carbon anodes for sodium‐ion batteries (Table S6) [14, 24, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50]. Irreversible capacity loss is mainly attributed to the formation of the solid electrolyte interphase (SEI) and side reactions between Na⁺ and surface functional groups [16, 17]. For comparison, the PSC derived from pristine asphaltenes shows a much lower reversible capacity of 84.19 mAh g⁻¹ and an ICE of 45.5%, which can be attributed to its inferior carbon structure and lack of active Na⁺ storage sites. It is worth mentioning that PSC derived from pristine asphaltenes displays a sloping GCD profile without distinct plateaus, characteristic of sodium storage in soft carbon [6, 14, 21, 25]. This characteristic further confirms the little if any selective structural transformation in the absence of heteroatoms in asphaltenes [31]. The control experiments designed here provide direct experimental evidence that intrinsic heteroatoms govern the selectivity of nitric acid oxidation rather than merely enhancing electrochemical performance.

The rate performance of PSC and PHCs was evaluated under various current densities (Figure 3d; Figure S19a). PHC‐0.5N exhibited exceptional reversible capacities of 403.2, 354.2, 306.8, 267.8, and 223.9 mAh g⁻¹ at 50, 100, 200, 300, and 500 mA g⁻¹, respectively, and maintained at 211.37 mAh g⁻¹ even at 1 A g⁻¹. This high‐rate capability originates from the disruption of π–π conjugated stacking in asphaltenes induced by nitric acid oxidation. The introduced oxygen‐containing functional groups create steric hindrance and disrupt compact graphite‐like domains, while also expanding the interlayer spacing from 0.341 nm for PSC to 0.349 nm for PHC‐0.5N (Figure 1c) and providing channels for rapid Na⁺ migration. In contrast, PSC showed much lower capacities of 128.94 mAh g⁻¹ at 50 mA g⁻¹ and 8.56 mAh g⁻¹ at 0.5 A g⁻¹. In this case, unbroken π‐π stacking forms dense carbon layers that impede Na⁺ migration under high‐current conditions. When the current density was reset to 100 mA g⁻¹, PHC‐0.5N recovered a discharge capacity of 334.28 mAh g⁻¹, further confirming that the disordered structure derived from π–π breakage avoided irreversible structural collapse during high‐rate cycling, thus ensuring excellent stability.

The long‐term cycling performance of PSC and PHCs at 0.5 A g⁻¹ is shown in Figures 3e and S19b. After 200 cycles, PHC‐0.5N retained a reversible capacity of 227.04 mAh g⁻¹, with an ultra‐low decay rate of 0.05% per cycle and a Coulombic efficiency (CE) approaching 100%. Furthermore, PHC‐0.5N retained 93.2% of its initial capacity after 2200 cycles (Figure 3f) even at 1 A g⁻¹, underscoring an excellent long‐term cycling stability. This superior performance is rooted in its structural features derived from the breakage of π–π stackings in original asphaltenes. It is evident that nitric acid oxidation not only expands the interlayer spacing of the resulting hard carbons, but also induces the formation of closed/semi‐closed pores and abundant defects. These pores act as stable Na⁺ storage sites to avoid active site loss, while the generated defects promote reversible Na⁺ adsorption/desorption, together mitigating capacity decay. In contrast, the capacity of PSC dropped to only 5.67 mAh g⁻¹ after 200 cycles, which results from a dense and low‐defect structure with insufficient storage sites and poor volume buffering from its intact original π‐π stacking during the carbonization, leading to irreversible Na⁺ trapping and structural collapse during cycling.

To further clarify the role of carbonization temperature in regulating hard‐carbon structure and sodium storage, PHC‐0.5N samples were carbonized at 1100°C–1500°C. Half‐cell GCD tests (Figure 3g) revealed distinct variations in performance. PHC‐0.5N‐1500 achieved a maximum ICE of 71.07% but suffered a rapid capacity fading, retaining only 63.78% of its initial capacity after 200 cycles at 0.5 A g⁻¹ (Figure S19c,d). This rapid deterioration in performance stems from the ordering of carbon layers induced at this elevated temperature, which reduces the number of closed pores and tightens interlayer spacing to limit sodium storage. This phenomenon is a direct consequence of intensified π‐π stacking at high temperatures, which counteracts the effect of disordering by nitric acid oxidation.

To verify the generality of the nitric acid‐heteroatom synergistic oxidation strategy, asphaltenes extracted with different n‐alkane solvents solvents were converted into hard carbons using the same procedure. Electrochemical evaluations (Figure S19e,f) showed that the hard carbon derived from n‐nonane‐extracted asphaltenes exhibited much poorer performance, with an ICE of only 46.6% and a reversible capacity of 178.74 mAh g⁻¹ at 50 mA g⁻¹. This performance is comparable to that of PSC and inferior to that of most benchmark pitch‐derived carbons [48, 49]. In contrast, hard carbons derived from n‐pentane‐extracted asphaltenes after nitric‐acid pre‐oxidation delivered a reversible capacity of 363.42 mAh g⁻¹ with an ICE of 69.41% (Figure 3h), outperforming most reported asphaltene‐based hard carbons [13, 26]. Systematic investigation of the pre‐oxidation time within the PHC‐0.5N‐1400 series (Figure S19g–i; Table S5) shows that PHC‐0.5N‐240min‐1400 delivers the highest ICE of 69.2% together with a first‐cycle charge capacity of 363.4 mAh g⁻¹. However, its rate capability and long‐term stability are inferior to the optimally oxidized counterpart. At 1 A g⁻¹, the reversible capacity of PHC‐0.5N‐240min‐1400 is limited to 91.1 mAh g⁻¹, markedly lower than 136.4 mAh g⁻¹ for PHC‐0.5N‐60min‐1400, with its capacity retention after 200 cycles reduced slightly from 75.2% to 73.4%. XPS survey quantification further showed that the residual N remained comparable across the samples of PHC‐0.5N‐1400 time series at 0.52‐0.68 at.% after 1400°C carbonization, whereas O content increased from 8.97‐9.39 at.% to 12.41 at.% for the samples obtained by processing for 240 min. This finding suggests that oxygen‐containing surface functionalities are more likely to modulate SEI chemistry and long‐term interfacial stability than nitrogen‐containing surface functionalities (Table S4) [4, 5]. These results indicate that both insufficient and excessive oxidation undermine the structural robustness required for sustained deep sodiation [4, 5]. Insufficient oxidation leads to incomplete interlayer expansion, whereas excessive oxidation over‐disrupts the carbon framework. In both cases, the balance between the breakage of π–π stacking and the stability of the resulting frameworks is perturbed, ultimately compromising the rate performance and cycling durability.

Practical applicability of this current synergistic strategy was validated using full cells and pouch cells. The NVP//PHC‐0.5N full‐cell (NVP cathode, Na₃V₂(PO₄)₃) exhibited an ICE of 83.3% at 50 mA g⁻¹ (Figure 3i) and retained 92.69% of its initial capacity after 200 cycles (Figure 3j). This superior performance is attributed to the high ICE of the PHC‐0.5N anode, which reduces electrolyte consumption during the formation of the solid electrolyte interphases (SEI). A 1.2 Ah pouch‐cell (NFM cathode, NaFe_0.33Mn_0.33Ni_0.33O₂) delivered a high energy density of 139.83 Wh kg⁻¹, which is comparable to that of commercial lithium iron phosphate (LFP) batteries, and retained 86.5% of its initial capacity after 100 cycles at 0.5 C with a discharge capacity of 995 mAh (Figure 3k), demonstrating an excellent scalability. Collectively, PHC‐0.5N, synthesized via a scalable oxidation‐carbonization process, suppresses the tendency of graphitization of low‐softening‐point pitch‐derived carbons through disruption of conjugated π‐π stacking induced by nitric acid pre‐oxidation, achieving a high specific capacity, an excellent rate capability and a long cycling durability that outscores the existing hard carbon anodes of sodium‐ion batteries reported in open literature as shown in Figure 3l and Table S6 [14, 24, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50].

To further clarify the sodium‐storage mechanism and kinetic characteristics of PSC and PHC electrodes, cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and galvanostatic intermittent titration technique (GITT) were employed to probe the effects of nitric acid oxidation and intrinsic heteroatoms on charge‐storage behavior. In the case of CV tests at scan rates of 0.1‐1.0 mV s⁻¹ (Figure S20), both samples exhibited a mixed diffusion‐capacitive charge‐storage behavior, with curves maintaining similar profiles and showing only gradual broadening in peak shapes as the scan rate increased [51, 52, 53]. Notably, PHC‐0.5N benefits from a disordered carbon structure and abundant surface adsorption sites induced by disruption of π–π stacking, thus presenting a dominant capacitive contribution that increased from 59.05% to 94.96% as the scan rate increased from 0.1 to 1.0 mV s⁻¹ (Figure 4a). The b‐values of PHC‐0.5N were calculated from fitting the log(i)–log(v) relationship to be 0.71 (oxidation) and 0.72 (reduction) (Figure 4b), confirming a predominantly capacitive‐controlled sodium storage process [54, 55].

FIGURE 4.

Kinetic analysis of PSC and PHC electrodes. (a) CV curve of PHC‐0.5N at 1.0 mV s⁻¹; (b) b‐value fitting for the CV oxidation peaks; (c) Nyquist plots of PSC and PHC‐0.5N half‐cells, with the inset showing the fitted results based on the equivalent‐circuit model; (d,e) Na⁺ diffusion coefficients of PSC and PHC‐0.5N during discharge and charge, calculated by GITT; (f) GCD curves; (g) in situ EIS; (h) in situ DRT of the PHC‐0.5N electrode; and (i,j) 3D and 2D AFM images of PSC and PHC‐0.5N electrodes after 100 cycles.

In contrast, PSC exhibits a typical diffusion‐limited behavior of soft carbon due to unbroken conjugated π‐π stacking. All PHC electrodes showed higher b‐values than PSC (Figures S21g–i), further verifying that the breakage of π‐π stacking shifts the charge‐storage mechanism toward capacitive dominance. In terms of sodium diffusion kinetics, EIS tests on fresh electrodes revealed an optimized carbon network in PHC‐0.5N, featuring closed pores and defects generated by disruption of π–π stacking (Figure 4c), which favors controlled SEI growth. Moreover, PHC‐0.5N also exhibits the lowest solid electrolyte interphase resistance (R _SEI) and charge‐transfer resistance (R _ct), both of which are favorable for rapid Na⁺ transport. The results obtained from GITT analysis further confirmed such finding. It is worth mentioning that PHC‐0.5N exhibited extended titration duration in the high‐capacity potential region and Na⁺ diffusion coefficients (D _Na+) of 10⁻⁷‐10⁻⁹ cm² s⁻¹ during discharging and charging, which results from its low interfacial resistance of EIS (Figures 4d,e; Figure S22). All PHC electrodes had comparable Na⁺ diffusion coefficients in the regions of 0.005‐2.8 V. The results further prove that the enhanced ion kinetics induced by disruption of π‐π stacking underpin the excellent rate performance of PHCs.

In situ EIS and DRT analyses allow the investigation on sodium insertion/extraction mechanisms during cycling (Figure 4f–h). PSC showed significant fluctuations in R _SEI with incomplete recovery upon post‐desodiation, stemming most likely from repetitive peeling of SEI that caused a continuing consumption of electrolytes and structural degradation, which correlates well with its poor cycling stability. In contrast, PHC‐0.5N exhibited steadier in situ EIS spectra (Figure 4g) with minimal R _SEI/R _ct changes, indicating a stable structure and consistent ion conduction. DRT analysis (Figure S23a) revealed strong and SOC‐dependent low‐frequency peaks (τ > 1 s, Na⁺ diffusion in SEI) during cycling of PSC, reflecting an unstable SEI, in contrast to stable peaks of post‐desodiation for PHC‐0.5N [56]. Figure S23 presents the in situ EIS and DRT spectra of other PHC samples, including PHC‐0.05N, PHC‐0.1N, PHC‐1N and PHC‐2N, a one‐cycle GCD curve of PSC. The results complement well with the analysis described above. In mid‐frequency regions (10⁻³< τ < 1 s), PHC‐0.5N displayed smaller, less fluctuating charge‐transfer resistance (P _ct) peaks and reduced P _ct post‐desodiation due to enhanced carbon networks [20, 21]. The lower overall resistance (P _ec) and more negligible volume expansion of PHC‐0.5N, compared with PSC, underpinned its superior performance. To further validate the interfacial stability, the electrodes after 100 cycles were examined by atomic force microscopy (AFM). As shown in Figure 4i,j, the PHC‐0.5N electrode after 100 cycles at 0.5 A g⁻¹ exhibited a much smoother surface (R _a = 64.14 nm) as compared with the reference of PSC electrode (R _a = 80.06 nm). Meanwhile, its average Young's modulus (11.9 GPa) was nearly twice that of PSC (6.1 GPa), indicating a mechanically more robust and uniform SEI layer. The electrochemical and morphological analyses together confirm that the structural regulation of PHC‐0.5N by nitric acid not only enhances charge‐transfer kinetics but also yields a mechanically more stable SEI, ensuring a more durable sodium‐ion storage.

The evolution in SEI and interfacial polarization of PSC and PHC‐0.5N electrodes was further investigated after 200 cycles (Figure 5). The photographs of disassembled cells in Figures S24 and S25 show pronounced swelling and powder shedding on PSC with orange‐colored deposits on the GF/D separator, whereas the electrode of PHC‐0.5N and the corresponding separator remained nearly pristine, suggesting markedly suppressed undesired side reactions of PHC‐0.5N. XPS high‐resolution spectra reveal distinct SEI chemistries on the two electrodes (Figure 5a,b; Figure S26). In C 1s XPS spectra, PHC‐0.5N exhibits more pronounced carbonate (CO₃ ²⁻) and C‐F signals, while the corresponding contributions on PSC are much weaker, indicating a more inorganic/carbonate‐containing interphase on PHC‐0.5N than on PSC [4, 5, 13]. The F 1s spectra further differentiate the dominant fluorine environments, with PSC being mainly governed by C‐F_x species (64.18%), whereas PHC‐0.5N shows a higher P‐F contribution (59.03%). The results indicate a suppressed formation and accumulation of fluorinated organic species and a more stable salt‐derived interphase, consistent with the finding of thinner and more uniform morphology of the SEI on PHC‐0.5N than on PSC [4, 5].

FIGURE 5.

Evolution of SEI and interfacial properties of PSC and PHC‐0.5N electrodes. (a,b) high‐resolution XPS spectra of C 1s, O 1s, F 1s, and Na 1s for PSC and PHC‐0.5N electrodes after 200 cycles; (c,d) contour plots of discharge profiles over 200 cycles for PSC and PHC‐0.5N electrodes; (e,f) TEM images of PSC and PHC‐0.5N electrodes after 200 cycles; and (g,h) 3D TOF‐SIMS images of PSC and PHC‐0.5N electrodes after 200 cycles.

Such interfacial contrast correlates well with the evolution of polarization visualized by the contour plots of discharge profiles (Figure 5c,d). For PSC, the highly polarized regions below 0.1 V and in the 1.0–0.1 V region expanded progressively with cycling, suggesting a continuous buildup of its impedance and deteriorated interfacial kinetics [57]. In sharp contrast, PHC‐0.5N maintained nearly unchanged contour distributions over 200 cycles, indicating an effectively suppressed growth of polarization [57]. Post‐cycling TEM images in Figure 5e,f further reveal the pronounced morphological divergence, showing that PSC develops a thick and patchy surface layer with an average thickness of about 77 nm, whereas PHC‐0.5N exhibits a more continuous interphase with a reduced thickness of about 52 nm. Consistently, the results from 3D TOF‐SIMS mapping demonstrate deeper penetration depth and more homogeneous 3D distributions of such key fragments as NaF₂ ⁻, CHO₂ ⁻, C₂H₂O⁻, CO₃ ⁻ and NaO⁻ on PHC‐0.5N than on PSC (Figures 5 g,h; Figures S27–S32), indicating a more regulated interphase chemistry and spatially consistent ion‐conducting pathways in PHC‐0.5N than in PSC [2, 4, 13].

The sodium storage mechanism of PHC‐0.5N is elucidated by a combination of in situ and ex situ characterizations. In situ XRD analysis (Figure 6a–c) revealed a broad (002) peak. This peak reflects the structural disorder of hard carbon, which is consistent with prior results. The minimal peak attenuation during initial sodiation indicates surface adsorption of Na⁺ being the dominant mechanism [14, 24]. Subsequently, the (002) peak shifted from 25.4° to 24.9° on one edge while the other edge remained the same at 22.8°, showing a reduced width and a red‐shifted center. The peak intensity decreased sharply and reached a minimum at 0.002 V, corresponding to Na⁺ intercalation, which led to expanded interlayer spacing and enhanced disorder [27, 58]. Upon desodiation to 2.8 V, the peak recovered only partially, resulting from the formation of “dead Na” as reported [14, 24, 41]. Notably, no metallic Na peaks appeared even at 0.002 V, indicating Na⁺ pore‐filling rather than the formation of low‐activity clustered sodium [59, 60].

FIGURE 6.

In situ and ex situ characterization of sodium‐storage mechanisms in PHC electrodes. (a) GCD profiles; (b) in situ XRD patterns; (c) magnified XRD patterns of PHC‐0.5N; (d) in situ Raman spectra; (e) ex situ EPR spectra of PSC and PHC‐0.5N electrodes at selected sodiation/desodiation states; (f) representative deconvolution of the EPR spectrum of PSC at 0.002 V into narrow and broad components; (g) Synchrotron Na K‐edge XANES spectra collected at SSRF; (h) XPS Na 1s spectra; and (i) proposed sodium‐storage mechanism.

In situ Raman spectra (Figure 6d) further tracked the structural response of PHC‐0.5N during sodiation/desodiation. Notably, the G band progressively red‐shifts over the entire discharge process, indicating that Na uptake continuously perturbs the sp² carbon framework. This red‐shift becomes more pronounced in the low‐potential region, accompanied by a clear attenuation of the D‐band intensity from 0.1 to 0.002 V. Such coupled spectral evolution suggests substantially strengthened Na─C interactions and a pronounced reorganization of the electronic/vibrational environment at deep sodiation, rather than a negligible framework response above 0.1 V. The attenuation of the D band in the plateau region likely originates from charge‐transfer‐induced modification of the Raman resonance condition, leading to weakened defect‐activated scattering in the highly sodiated state. Combined with the concurrent low‐angle shift of the (002) reflection in the in situ XRD patterns (Figure 6c), these Raman results support progressive Na insertion into the turbostratic carbon framework, with the strongest structural response occurring in the plateau region.

Ex situ EPR spectra collected at selected sodiation/desodiation states were analyzed by two‐component fitting, resolving a broad line assigned to adsorbed/trapped Na in heterogeneous paramagnetic environments and a narrow line assigned to quasi‐metallic Na in a more electronically delocalized state (Figures 6e,f; Figures S33–S35). During discharge from 0.5 to 0.1 V, all electrodes exhibit superimposed broad and narrow contributions, indicating the coexistence of multiple Na‐storage environments [21, 28]. Upon further discharge to 0.002 V, however, PHC‐0.5N undergoes the most pronounced lineshape reconstruction: the resonance becomes markedly sharper and increasingly dominated by the narrow component, while the fitted g values of the two components become much closer. This evolution is consistent with the emergence of a low‐potential quasi‐metallic Na state, in line with Na accumulation in confined closed pores during the plateau region. Upon charging back to 0.1 and 0.5 V, the broad component reappears, indicating that this low‐potential Na state is largely reversible [21, 28]. By contrast, PSC retains a more evident residual narrow contribution after recharge, suggesting less reversible Na trapping and poorer structural/interfacial stability upon repeated deep sodiation. PHC‐2N also shows improved reversibility relative to PSC, but the clearest reversible low‐potential spectral reconstruction is observed for PHC‐0.5N.

Synchrotron X‐ray absorption near‐edge structure (XANES) analysis on Na K‐edge at 0.002 V showed a clear shift of the Na absorption edge by 0.38 eV toward lower energy (Figure 6g). This shift suggests the reduction of Na⁺ to quasi‐metallic Na, a trend that aligns well with previously documented sodium storage behavior in porous carbon structures [61]. When charged to 0.1 V, the absorption edge was similar to that at 0.002 V, with a slight upward shift in energy due to the polarization of sodium.

Na 1s and C 1s XPS spectra shown in Figure 6h revealed a hybrid sodium‐storage mechanism. Specifically, C─O─Na signals shifted toward the binding energy of metallic Na (∼1071.8 eV) during discharge when etching depths reached 240 nm, reflecting the filling of internal nanopores with quasi‐metallic Na species. Upon recharging, the C─O─Na peak returned to its initial position, which confirms the reversibility of this pore‐filling process as also reported in earlier studies [62, 63]. These results support the proposed sodium‐storage mechanism shown in Figure 6i, which encompasses surface adsorption, interlayer intercalation and pore filling. The results collectively verify that nitric acid oxidation regulates steric hindrance to break conjugated π‐π stacking in asphaltenes thereby leading to hard carbons with short‐range order and long‐range disorder. This structural optimization enhances the reversibility of quasi‐metallic Na storage, suppresses the formation of dead Na and ultimately underpins the superior sodium ion storage performance of PHC electrodes.

Previous electrochemical comparisons revealed that only PHC‐0.5N obtained from heteroatom‐containing asphaltenes with pre‐oxidation by nitric acid achieves a high reversible capacity and ICE, whereas hard carbons obtained by air‐oxidation of heteroatom‐containing asphaltenes or nitric acid oxidation of heteroatom‐free asphaltenes exhibit an inferior performance, suggesting that the intrinsic heteroatoms in asphaltenes play a decisive role in the oxidative transformation process. Consistently, the air‐pre‐oxidized sample of heteroatom‐containing asphaltenes showed a highly graphitized structure with severe layer restacking, as evidenced by its sharp XRD peak at 26.4°, long‐range ordered carbon‐layer arrangement in HR‐TEM and a low I _D/I _G ratio of 0.83 in Raman (Figure S36). The results further confirm that air oxidation fails to induce effective structural regulation in the final hard carbon product. To elucidate this mechanism, density functional theory (DFT) calculations were performed using heteroatom‐containing and pristine carbon frameworks (Figures 7a–d; Figure S37). Charge‐density analysis shows a pronounced local electronic inhomogeneity induced by intrinsic heteroatoms, generating chemically activated regions adjacent to defect sites [28, 29]. More importantly, quantitative adsorption calculations reveal a strong energetic preference for nitric acid‐derived oxidizing species at heteroatom‐modified sites, with adsorption energies of −4.611 eV (H⁺) and −4.57 eV (·OH) being substantially higher than those of −2.654 and −2.072 eV, respectively, on pristine carbon frameworks. This energetic differentiation provides a direct evidence that intrinsic heteroatoms govern site‐selective radical anchoring and initiate localized oxidation rather than uniform framework degradation [28, 29]. As a consequence, site‐preferential oxidation enhances local steric hindrance, which disrupts conjugated π–π stacking during pre‐oxidation and promotes the formation of defect‐rich turbostratic carbon domains. Consistent with this mechanism, the calculated diffusion barriers for Na⁺ decreased from 1.17 eV in ordered pristine carbon to 0.56 eV in heteroatom‐containing oxidized frameworks, confirming that heteroatom‐mediated site‐preferential oxidation governs both evolution of structures and transport kinetics of ions. Together, these results establish a mechanistic pathway in which the intrinsic heteroatoms in the asphaltene carbon source define the landscape of selectivity in oxidation by nitric acid, enabling controlled hard‐carbonization rather than random oxidative damage.

FIGURE 7.

DFT calculation systems and life‐cycle assessment. (a) carbon‐framework models of HF‐PHC and PHC; (b) differential charge‐density diagrams of HF‐PHC and PHC; (c) adsorption energies of nitric acid molecules and their derived radicals on the two carbon frameworks; (d) energy barriers for Na⁺ diffusion in the two carbon frameworks; (e,f) Sankey diagrams illustrating the GWP‐based carbon‐emission flows of the PHC‐0.5N and commercial CHC production processes; and (g‐l) contributions of TET, MEC, LU, IR, HNCT, and HCT to the production of 1 kg of PHC‐0.5N and CHC.

A cradle‐to‐gate life cycle assessment (LCA) was conducted to benchmark the production of 1 kg PHC‐0.5N via a simple oxidative route against the commercial bamboo‐derived hard carbon (CHC) process, revealing distinct environmental implications (Tables S7‐S8). As illustrated in Figures 7e,f, the global warming potential (GWP) decreases markedly from 18.28 to 11.20 kg CO₂‐eq when shifting from the commercial CHC route to the PHC‐0.5N process, representing an emission reduction of about 7.08 kg CO₂‐eq, or 38.7%. This improvement is mainly attributed to the carbonization stage, which contributes 67.74% of the total emissions in our route, whereas the commercial pathway exhibits multiple high‐impact contributors, including carbonization, pyrolysis, devolatilization and alkali leaching, each exceeding 15%. A comprehensive comparison across 18 impact categories further demonstrates substantial reductions by 40%–95% for PHC‐0.5N relative to CHC, except for fossil resource scarcity (FRS) and stratospheric ozone depletion (ODP), as shown in Figures 7g–l and S38–S40 and Tables S9–S11. The slight increases in ODP originate from the upstream nitric acid production where substantial N₂O emission occurs, while the higher FRS results mainly from the fossil‐based asphaltene feedstock. Nevertheless, using asphaltenes, a typically underutilized petroleum waste enables its high‐value valorization, which is environmentally beneficial from a resource‐utilization perspective. Figure 7i reveals a striking 95% reduction in land use (LU), dropping from 2.707 to 0.139, largely because the commercial CHC route depends on land‐intensive bamboo cultivation, whereas our process valorizes asphaltenes (a waste product of petroleum refining) without requiring additional land. Furthermore, the human carcinogenic toxicity (HCT) category shows a remarkable reduction, mainly due to the avoidance of NaOH‐based alkali leaching, whose upstream chlor‐alkali production and wastewater treatment introduce substantial carcinogenic burdens, independent of electricity consumption (Figure 7l). In summary, this work demonstrates that the oxidation of heteroatom‐containing polyaromatic asphaltenes by nitric acid to break π–π stacking and induce desired steric hindrance is an effective strategy to optimize asphaltene‐derived hard carbons. This strategy simultaneously enables substantial environmental advantages through waste valorization and markedly lower resource burdens. Our findings provide clear mechanistic and sustainability insights for designing high‐performance carbon‐based anodes for sodium‐ion batteries.

3. Conclusion

This work identifies a site‐preferential oxidation mechanism mediated by intrinsic heteroatoms in polyaromatic asphaltenes that enables controlled carbonization of graphitization‐prone asphaltene precursors into hard carbons of desired structure and mophologies. Unlike conventional oxidative treatments that induce non‐selective structural modification, reactive species from nitric acid anchor preferentially at heteroatom‐modified sites, generating localized electronic inhomogeneity that disrupts π–π stacking and kinetically suppresses graphitization. This selective framework disruption governs deterministic evolution of turbostratic domains and closed nanopores, thereby establishing a structural basis for efficient sodium storage. Density functional theory further reveals that intrinsic heteroatoms govern reaction selectivity during pre‐oxidation, thereby directing the formation of hard‐carbon structures that simultaneously enhance Na⁺ adsorption and diffusion kinetics. As a result, the optimized hard carbon exhibits a high ICE of 89.7%, a reversible capacity of 404.1 mAh g⁻¹, an excellent rate capability of 211.37 mAh g⁻¹ at 1 A g⁻¹, and a long‐term stability of 93.2% capacity retention over 2200 cycles. Practical relevance is demonstrated by stable operation in a 1.2 Ah pouch cell. Beyond performance metrics, cradle‐to‐gate LCA confirms substantial environmental advantages arising from waste‐derived feedstocks and simplified processing routes. More broadly, this study establishes a mechanistic framework for selective disruption of ordered carbon domains, offering a generalizable strategy for transforming graphitization‐prone precursors into high‐performance hard carbons and advancing the rational design of sustainable sodium‐ion battery anodes.

4. Experimental Section

Petroleum asphaltenes were obtained from Maoming Petrochemical Co., Ltd. Nitric acid and n‐alkanes (n‐pentane, n‐hexane, n‐octane, n‐nonane) were purchased from Dongjiang Chemical Reagents Co., Ltd. Heteroatom‐free pitch (No. 10 petroleum pitch block, Chinese national standard) was supplied by Zhongtie Weiye Co., Ltd.

The soft carbon precursor (petroleum pitch‐derived asphaltenes, PPA) was obtained by dissolving low‐softening‐point pitch (70°C) in n‐heptane. Typically, 2 g of pitch was mixed with 60 mL n‐heptane, ultrasonicated for 30 min, and then centrifuged at 8000 rpm three times with consecutive dispersion of precipitates in fresh n‐heptane until the supernatant became clear. The product (n‐heptane asphaltene precursor) was vacuum‐dried, ground and sieved to <100 µm. Then 0.5 g of asphaltene powder was treated with 60 mL nitric acid (various concentrations) for 120 min. After filtration through a microporous filter and drying, the product was labeled as PHC‐xN‐UC (where x denotes to the molarity of nitric acid). Finally, PHC‐xN‐UC was carbonized under argon at 1400°C for 2 h with a heating rate of 1.5°C/min to yield PHC‐xN.

As a comparison, asphaltene‐derived soft carbon without nitric acid oxidation (PSC) was prepared under the same carbonization conditions. As an air‐pre‐oxidation control, the asphaltene precursor was treated in air at 320°C for 4 h prior to carbonization under identical conditions, and the resulting sample was denoted as AP‐PHC. For comparison, a heteroatom‐free pitch precursor was extracted and processed using the same procedure as that for the n‐heptane asphaltene precursor, followed by treatment with 0.5 mol L⁻¹ nitric acid and subsequent carbonization under identical conditions to obtain HF‐PHC‐0.5N. Under the identical preparation conditions as n‐heptane asphaltenes, the effect of nitric‐acid treatment duration of other alkanes on the performance of asphaltene‐derived hard‐carbon anodes was also investigated. The samples treated for varying times were labeled as PHC‐0.5N‐y, where y denotes to the treatment time in minutes.

CR2032 coin cells were assembled in an Ar‐filled glovebox. The anode slurry (70 wt% active material, 20 wt% acetylene black, 5 wt% SBR and 5 wt% sodium carboxymethyl cellulose in deionized water) was doctor‐bladed onto an Al foil, dried at 90°C under vacuum for 12 h and punched into 10 mm discs (mass loading 1–1.5 mg cm⁻²). Sodium metal (14 mm) served as counter/reference electrode with glass fiber (Whatman GF/D) as separator and 1 mol L⁻¹ NaPF₆ in DEGDME as electrolyte. Electrochemical tests were performed on a Neware CT‐3008 W platform at 30°C. Galvanostatic charge–discharge was conducted at 50 mA g⁻¹ between 0.002 and 2.8 V. Cyclic voltammetry (CV) was measured at 0.2–1.0 mV s⁻¹. Electrochemical impedance spectroscopy (EIS) measurements covered a frequency range of 0.1 Hz to 100 kHz. Galvanostatic intermittent titration technique (GITT) used 50 mA g⁻¹ pulses for 30 s with 2 min rests.

For pouch‐cell fabrication, both cathode and anode electrodes were prepared by a double‐side coating process on 12 µm Al foil current collectors. The cathode slurry consisted of NFM333, Super P, carbon nanotubes and PVDF in a mass ratio of 95.3:1:1:2.7 using NMP as solvent. The areal mass loading of the cathode was 14.7 mg cm⁻², and the electrode was calendered to a compaction density of 2.0 g cm⁻³. The anode slurry was prepared using hard carbon, CMC‐Na, SBR and Super P in a mass ratio of 94.5:2:2:1.5 with deionized water as solvent. The anode areal mass loading was 6.25 mg cm⁻², and the electrode was calendered to a compaction density of 0.95 g cm⁻³. The cathode and anode sheets were cut into 60 mm × 80 mm and 63 mm × 84 mm, respectively, and stacked alternately into 8 cathode layers and 9 anode layers, with the N/P ratio controlled at 1.15.

A combination of structural, chemical, morphological and electrochemical‐interface characterizations was employed to elucidate the evolution of the asphaltene‐derived carbons. X‐ray diffraction (XRD), Raman spectroscopy and synchrotron X‐ray absorption near‐edge structure (XANES) were used to probe carbon ordering, defect structure and local electronic environments. Fourier‐transform infrared spectroscopy (FT‐IR) and X‐ray photoelectron spectroscopy (XPS) were performed to identify functional groups and elemental chemical states. The pore structure was evaluated by N₂/CO₂ adsorption and small‐angle X‐ray scattering (SAXS), while scanning electron microscopy (SEM), high‐resolution transmission electron microscopy (HR‐TEM) and atomic force microscopy (AFM) were used to examine morphology, lattice texture and surface topography. Thermal evolution and gaseous products were analyzed by thermogravimetric analysis/differential scanning calorimetry (TGA/DSC) and in situ TG‐MS. Time‐of‐flight secondary ion mass spectrometry (TOF‐SIMS) and electron paramagnetic resonance (EPR) were further applied to investigate interfacial chemistry and unpaired‐electron environments.

Conflicts of Interest

The authors declare no conflicts of interest.

Author Contributions

Y.L., H.Y. and Z.X. conceived the idea for the study and designed the experiments. Q.L., W.L. and X.K. performed nanostructure preparation, device fabrication, and characterization. R.B. and J.L. performed the theoretical simulation. F.J. assisted the device fabrication and characterization. Y.L. and J.P. carried out morphology characterization. Y.L., H.Y., Z.L. and Z.X. analyzed the data and wrote the manuscript. Y.L., Z.L. and Z.X. edited the manuscript. All authors contributed to the manuscript.

Supporting information

Supporting File: adma73451‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.