Lithium-ion batteries (LIBs) have exhibited a significant surge in demand across various fields owing to their efficient energy storage capabilities and convenient usability. Additionally, owing to their inherent advantages, including high specific capacity and voltage, low self-discharge, and the absence of memory effects, LIBs demonstrate superior performance over any preceding battery systems [1–3]. However, recent incidents such as fires and explosions have underscored the safety concerns with LIBs, making them a pressing challenge. Consequently, the constrained safety performance of LIBs impedes their market expansion. Lithium plating on the negative electrodes is a significant issue affecting the safety of LIBs. The conventional graphite anode undergoes lithium plating owing to the proximity of the lithium insertion potential (0.1 vs. Li/Li⁺) to the deposition potential of lithium metal [4–6]. Excessive Li deposition can lead to the formation of dendrites, which can result in internally short-circuit the cell. To effectively mitigate the risks associated with lithium plating, it is imperative to develop negative electrode materials that exhibit higher lithiation potentials.

Transition metal oxides (TMOs) have been extensively studied as alternative materials to replace commercially used graphite [7–11]. TMOs can be broadly categorized into two groups based on their reactivity with lithium: conversion and insertion types. Among them, insertion-type TMOs are characterized by stronger metal–oxygen bonding than conversion-type TMOs. This robust bonding ensures that during lithiation the metal–oxygen bonds remain intact, allowing Li⁺ ions to permeate into the lithium storage sites within the structure of metal oxides and engage in the reaction (MO + xLi⁺ + xe⁻ → Li_xMO) [7,8,11–13]. Insertion-type TMOs, although relatively lower in capacity than conversion-type TMOs, exhibit superior cycling performance owing to their smaller volume changes (<3%) during lithiation/delithiation cycles. Insertion-type TMOs include TiO₂, Li₄Ti₅O₁₂, V₂O₅, and MoO₂. Among these, titanium dioxide (TiO₂) exists in various phases such as anatase [13–17], rutile [18,19], TiO₂(B) [20,21], and brookite [22,23]. Research efforts have predominantly focused on investigating the anatase phase, which exhibits the best performance for lithium-ion storage owing to its inherently stable crystal structure. Furthermore, anatase TiO₂ has gained attention as a safe negative electrode material for LIBs because of its higher operating potential (>1.5 V vs. Li/Li⁺) than that of carbonaceous negative electrode materials, which minimizes lithium plating issues. This elevated reaction potential prevents the reduction of the electrolyte and the formation of a solid electrolyte interphase (SEI) on the electrode surface [24,25]. Anatase TiO₂ is characterized by a three-dimensional network formed by the stacking of one-dimensional zigzag chains of TiO₆ octahedra through distorted edge-sharing (space group I4₁/amd). This stacking process results in the formation of vacant zigzag channels within the anatase framework, facilitating the insertion of Li⁺ ions into these octahedral sites. The majority of anatase TiO₂ materials utilize only half of their capacity through the insertion reaction of Li⁺ ions (TiO₂ + xLi⁺ + xe⁻ → Li_xTiO₂), wherein half a Li⁺ ion is inserted per TiO₂ molecule [26]. Given the high working potential previously mentioned, nano-sized TiO₂ with a large surface area has a relatively minor decline in reversible capacity owing to less electrolyte decomposition. However, owing to its low electronic conductivity (~10⁻¹² S cm⁻¹), anatase TiO₂ tends to exhibit poor rate characteristics [27]. To enable fast charging, it is imperative to enhance the electronic conductivity of anatase TiO₂.

To enhance electron transport, various methods have been developed, including hybridization with highly conductive materials and the introduction of anion dopants [17,28–31]. Surface treatment of anatase TiO₂ is a prominent approach widely employed to improve its rate characteristics because of its inherently poor electronic conductivity [32]. Anion dopants including N [33–37], S [28,38], C [39–41], F [21,42], and B [31,43] have been introduced into TiO₂ lattice to mitigate the electron transport resistance. Among them, N dopants, which have been extensively researched in the field of photocatalysis, exhibit a predominant tendency to be located on the surface layer of TiO₂. This behavior can be attributed to the higher energy level exhibited by the surface relative to the bulk and the poor solubility of the dopants [44,45]. The surface of N-doped TiO₂ can be transformed into TiN or TiO_2–xN_x through heat treatment in NH₃ gas. This modification significantly enhances the electrochemical performance of the TiO₂ electrode by introducing an electrically conductive phase on the surface [46,47]. However, the use of corrosive and toxic NH₃ gas hinders their commercial application. Therefore, research on less-toxic alternatives capable of improving the poor electronic conductivity of TiO₂ is imperative [48,49].

In this study, we investigated the surface nitridation of TiO₂ via the thermal decomposition of urea (NH₂CONH₂) in an autogenic reactor (AGR) in a closed reaction environment. Additionally, for the synergistic effects of N and S, thiourea (NH₂CSNH₂) was introduced. The thermal decomposition of thiourea was conducted under an inert atmosphere. Urea decomposes below 200°C and generates NH₃ gas [50]. Similarly, thiourea decomposes at 180–220°C, releasing gases including NH₃ and H₂S [51]. Therefore, both substances can serve as coating materials for AGR-based treatments designed to facilitate nitridation at temperatures above 700°C [52]. This method is simple and applicable to mass production. Surface-nitridated TiO₂ demonstrated enhanced electrochemical performance compared to that of pristine TiO₂. In particular, the thiourea-treated sample exhibited excellent rate capability across all current densities. The thiourea-treated sample exhibited a nearly threefold increase in capacity compared to the pristine sample at 1.0 C and demonstrated nearly double the capacity retention.

The experiment was conducted using an autogenic reactor (AGR), as shown in Fig. 1. The AGR is designed to establish and sustain high-pressure and high-temperature conditions within a confined space, which is essential for the synthesis of various materials. These extreme conditions provide controlled environments for the formation and transformation of materials. The AGR treatment enables the generation of highly electrically conductive materials through pyrolysis reactions at elevated temperatures and pressures. In addition, the AGR process is relatively simple, facilitating scalability from laboratory-scale experiments to industrial production and ensuring the production of materials in quantities suitable for commercial applications [48,49].

Fig. 2 shows the alterations in powder colors and particle shapes in response to urea or thiourea through powder images and SEM images. In Fig. 2a–c, the pristine TiO₂ powder is white, whereas the surface-nitridated TiO₂ powder displays a different color in its synthesized form. Urea treatment results in a dark yellow color [53], whereas thiourea treatment yields an even darker hue [54]. Considering that the color of each powder was uniform, the synthesis was performed well overall. SEM analyses (Fig. 2e–g) were conducted to investigate the surface morphology after AGR treatment. The pristine TiO₂ powder exhibits a spherical morphology, as observed in the SEM image (Fig. 2e). The SEM images of the surface-nitridated samples (Fig. 2f,g) indicate no significant changes in both morphology and particle size after heat treatment. Additionally, the particle size distribution for the samples was determined using ImageJ software (Fig. 2h–j). For each sample, a measurement of 400 particles was conducted, wherein the diameter of each particle was measured. The pristine TiO₂ powder exhibits an average particle size of 158.4 ± 2.8 nm (Fig. 2h). For the urea-treated sample, the average particle size was measured to be 153.6 ± 2.9 nm, and for the thiourea-treated sample, it was 155.9 ± 1.4 nm (Fig. 2i,j). The constancy in particle size and shape suggests that these variables do not influence performance. Consequently, the performance is determined solely by the material used for surface modification. Meanwhile, surface-nitridated TiO₂ powder is expected to exhibit a uniform distribution of N and S elements. This is based on previous studies that demonstrated a homogeneous distribution of Ti, O, N, and S elements in nitridated TiO₂ [55,56].

The alterations in the powder color prompted an investigation into the crystal structures of the materials after surface treatment with urea or thiourea. Fig. 3 shows the XRD patterns of the pristine, urea-treated, and thiourea-treated TiO₂ samples. The XRD pattern for pristine TiO₂ (Fig. 3a) exhibits seven distinct peaks at 2θ = 25.13°, 37.63°, 47.88°, 53.73°, 54.92°, 62.71°, and 68.64°, corresponding to (1 0 1), (0 0 4), (2 0 0), (1 0 5), (2 1 1), (2 0 4), and (1 1 6) crystal planes, respectively. These peaks aligned with the tetragonal crystal planes of anatase TiO₂, which is consistent with the standard spectrum (JCPDS no.: 21-1272). Notably, there is no evidence of other crystalline phases, such as rutile and brookite. The XRD patterns of surface-nitridated TiO₂ (Fig. 3b,c) exhibit no discernible differences in the bulk structure, indicating that the AGR heat treatment method selectively modifies the surface without altering the crystal structure of the mother phase. That is, no deformation of TiO₂ occurs after the AGR heat-treatment process. The crystallite size of each particle is calculated by applying the full-width half-maximum (FWHM) values of the (1 0 1) main peaks in Fig. 3 to Scherrer’s equation (D=Kλ/βcosθ). The particle sizes of pristine and surface-nitridated TiO₂ were determined to be in the range of 46–63 nm based on the equation. The consistent crystallite sizes suggest similar crystallinities for all three samples. However, the diffraction peaks of compounds containing nitrogen or sulfur were not evident in the XRD pattern, likely because of their minimal presence in the total powder composition. Additionally, the substitution of O²⁻ by N³⁻ and S²⁻ during AGR treatment resulted in peaks appearing at the same positions, contributing to the lack of significant differences in the XRD pattern [56]. Furthermore, previous studies established that TiO₂ annealed in the presence of urea or thiourea exhibits the substitution of N for O sites, thereby resulting in the formation of a Ti–N layer on the TiO₂ surface [49,56,57]. The absence of heterogeneous XRD peaks indicates that the nitridation process predominantly affected the surface.

Fig. 4a presents the initial voltage profiles and the capacity retention for pristine, urea-treated, and thiourea-treated TiO₂. The galvanostatic charge/discharge test was conducted over a voltage range of 1.0–2.5 V at a rate of 0.1 C. The three types of TiO₂ exhibited typical voltage profiles of an anatase TiO₂ negative electrode [58]. All three samples displayed a lithiation (discharge) plateau observed at ~1.76 V and a region of voltage drop to cut-off 1.0 V. The delithiation (charge) plateau starts at ~1.9 V. These potentials align with those reported in previous studies (~1.75 and ~1.88 V, respectively) [59,60]. A magnification of the voltage profile provides additional insights into internal resistance. A detailed analysis of the 1.9–2.0 V range revealed that the overpotential decreased in the order of pristine, urea-treated, and thiourea-treated TiO₂. This suggests that internal resistance also decreased in the same sequence. These analytical results demonstrate that the surface-treatment process in this study effectively enhances the electronic conductivity of the samples by doping N or S onto the surface. Consequently, the reduced internal resistance of surface-nitridated TiO₂ is expected to improve both cycle performance and rate capability. The electrochemical performance of our samples was comparable to that reported in previous studies [14,58]. The pristine TiO₂ exhibited 158.03 mAh g⁻¹ (lithiation), 140.53 mAh g⁻¹ (delithiation), during the first cycle. The lithiation capacity of pristine TiO₂ was close to its theoretical capacity (168 mAh g⁻¹, 0.5 mol of Li). The first-cycle lithiation and delithiation capacities were 128.64, 120.60 mAh g⁻¹ for urea-treated TiO₂, and 154.68, 138.23 mAh g⁻¹ for thiourea-treated TiO₂, respectively. This demonstrates that for the surface-nitridated samples, the irreversible capacity was smaller during the first-cycle lithiation/delithiation than that of the pristine counterpart. Additionally, the initial Coulombic efficiency (ICE) of the surface-nitridated samples improved compared to that of the pristine sample. Fig. 4b shows the reversibility of the electrochemical reactions of the three samples over 30 lithiation/delithiation cycles at a 0.1 C-rate. The capacity retention rates for each sample after 30 cycles were 36.15%, 51.11%, and 70.01% for pristine, urea-treated, and thiourea-treated samples, respectively. The capacity of the cells with the pristine TiO₂ electrode experienced sharp degradation within the first 10 cycles. Conversely, the surface-nitridated samples exhibited superior capacity retention compared to their pristine counterparts. In particular, the cycling performance of the cell with the thiourea-treated TiO₂ electrode demonstrated the most significant improvement, nearly doubling the capacity retention of the pristine TiO₂ electrode.

Rate capability tests of the three samples were conducted to investigate the power densities, as shown in Fig. 5. Thiourea-treated TiO₂ exhibited higher reversible capacities than the other samples at all the current densities tested. At the low C-rate (0.1 C, 1st cycle), the sample heat treated in the AGR with urea (119.32 mAh g⁻¹) showed lower specific capacities than the pristine TiO₂ (133.82 mAh g⁻¹). However, from 0.1 C (2nd cycle) onward, all surface-nitridated samples demonstrated higher specific capacities than the pristine sample. This improved rate capability of the surface-nitridated TiO₂ can be confirmed by comparing the cycle performance at high C-rates (≥0.2 Crate). At current densities of 0.2 C, 0.5 C, and 1.0 C, the urea-treated sample exhibited approximately 1.2, 1.3, and 1.5 times higher specific capacities compared to the pristine sample, respectively. The thiourea-treated sample demonstrated approximately 1.5, 2.0, and 3.0 times higher specific capacities than the pristine sample at the respective current densities. Note that the capacity ratio between 0.1 C and 1.0 C of pristine, urea-treated, and thiourea-treated TiO₂ were 22.1%, 37.0%, and 65.1%, respectively. These results indicate that the thiourea-treated sample exhibited the most improved rate capabilities compared to other samples across all current densities, notably achieving an approximately threefold improvement in capacity at 1.0 C compared to pristine TiO₂.

Previous studies have yielded valuable insights into the outstanding electrochemical performance of the thiourea-treated TiO₂ sample compared with that of the urea-treated sample. This can be attributed to the following two factors. The first reason appears to be that the band gap of N/S–TiO₂ is narrower than that of N–TiO₂. Viswanath et al. [61] reported a considerable reduction in the band gap of N/S–TiO₂ (0.22 eV reduction) compared to that of N–TiO₂ (0.11 eV reduction). This reduction underscores the inevitable enhancement of the electronic conductivity in N/S–TiO₂ compared to that in N–TiO₂. The second reason is the reduction in resistance resulting from improved electronic conductivity. As reported by Jiao et al. [62], in the case of N/S–TiO₂ (by NH₃ and H₂S gases), a substantial decrease was noted in both the ohmic resistance (R_e) and charge-transfer resistance (R_ct) compared to N–TiO₂ (by NH₃ gas). Compared to pristine, N–TiO₂ exhibited a reduction of 18.2% in R_e and 35.6% in R_ct. By contrast, N/S–TiO₂ showed more substantial reductions, with a 31.4% reduction in R_e and a 68.2% reduction in R_ct relative to the pristine TiO₂. For these two reasons, surface-nitridated TiO₂ demonstrates enhanced electrochemical performance compared to pristine TiO₂. Among the surface-nitridated TiO₂ samples, the thiourea-treated sample exhibits the highest rate capability and capacity retention.

In this study, to enhance the transport of electrons, we applied the surface nitridation of TiO₂ through the thermal decomposition of urea or thiourea. The thermal decomposition was conducted in an AGR at 700°C for a closed reaction environment. Facilitated by the AGR, the reaction between TiO₂ and urea induces surface nitridation, consequently enhancing the electronic conductivity and thereby improving the electrochemical performance. Furthermore, substituting urea with thiourea promoted surface nitridation owing to the synergistic effect arising from the coexistence of both N and S, resulting in further enhanced electrochemical performance. In terms of capacity retention, thiourea-treated TiO₂ exhibited a nearly two-fold improvement over pristine TiO₂. Moreover, the thiourea-treated sample exhibited superior rate performance across all current densities, particularly at 1.0 C where the capacity demonstrated an approximately threefold increase compared to that of the pristine sample. These enhanced performance outcomes are attributed to the improvement in the electronic conductivity of the samples through the AGR treatment, with the use of thiourea notably enhancing the electronic conductivity more than when using urea. Compared to the other samples, the thiourea-treated TiO₂ demonstrated the narrowest band gap, thereby exhibiting the highest electronic conductivity. This leads to lower resistance, resulting in superior electrochemical performance, including rate capability and capacity retention. In the context of this study, a one-step surface-nitridation process was employed, which avoided the use of toxic NH₃ and H₂S gases, making it particularly suitable for largescale production. This observation strengthens the potential of TiO₂ as a versatile material for energy storage given its enhanced safety and excellent performance.