Online Monitoring of Faulty Gases (O₃, NO₂, CO) in Substation Secondary Equipment Based on Cr-Doped BN Sensor: Insights from Density Functional Theory

Abstract

The secondary equipment of a substation is pivotal for maintaining the safe and reliable operation of the power grid. However, over time, insulation defects can inevitably arise in this equipment. Gas detection in substation secondary equipment has proven to be an effective method for assessing its insulation status. In this paper, we employed a density functional theory (DFT) approach to simulate the adsorption process of three types of fault gases from substation secondary equipment onto Cr-modified BN nanosheets. From the doped and adsorption models, two stable structures were chosen, and by calculating their band structures, density of states, and differential charge density, we uncovered the relevant adsorption and sensing mechanisms. Our findings reveal that Cr-modified BN nanosheets possess robust gas-sensing capabilities, particularly in capturing O₃, which is primarily attributable to the contribution of Cr’s 4d orbital electron layer. Specifically, the adsorption capacity of Cr-modified BN nanosheets for fault gases in substation secondary equipment follows the order: O₃ > NO₂ > CO. The adsorption of Cr-BN on the three target gases mainly tends to be chemisorption accompanied by chemical bond breaking. Notably, there are significant changes in the electronic properties of the adsorbent substrate before and after the adsorption of the target gas molecules, resulting in alterations in its overall conductivity. This research lays the theoretical groundwork for the experimental development of high-performance gas-sensitive sensors designed to detect fault gases in substation secondary equipment.

1. Introduction

The power industry stands as a vital cornerstone supporting national economic and social development; thus, ensuring a stable and reliable power supply is of the utmost importance [1,2,3]. Substation secondary equipment plays a crucial role in the overall power system, significantly contributing to the safe operation of the power grid [4,5]. However, over time, insulation defects such as overheating inevitably develop in substation secondary equipment, leading to degraded operational states and potential equipment malfunctions. These issues pose a serious threat to the stable operation of associated primary substation equipment. Additionally, these faults can cause the decomposition of surrounding air into ozone (O₃), carbon monoxide (CO), and nitrogen dioxide (NO₂). Therefore, monitoring the gas composition around substation secondary equipment through surrounding gas analysis (SGA) has been recognized as a convenient and practical method [6,7,8]. This approach can be used to assess the operational status of substation secondary equipment, ensuring stable power system operation and minimizing potential damage to the national economy. Among the various viable methods, utilizing ideal resistive sensors for online detection of gases around substation secondary equipment is a promising solution due to its simplicity, high sensitivity, rapid response, low cost, and low power consumption [9,10].

Two-dimensional boron nitride (2D BN) is a material in which boron and nitrogen atoms are uniformly distributed in a heterohexagonal structure, resembling graphene [11]. This material boasts exceptional electrical insulation, thermal conductivity, corrosion resistance, and chemical stability, remaining stable and unreactive with water, acids, and bases at room temperature. Consequently, BN finds applications in various fields, such as refractory materials, semiconductor solid-phase doping sources, atomic stack structural materials, neutron radiation shielding materials, rocket engine components, high-temperature lubricants, and release agents [12,13,14,15]. However, the inherent gas adsorption performance of 2D BN is not optimal due to its chemically inert surface and strong interlayer van der Waals forces [16,17]. To address this, the research suggests modifying pristine BN to enhance its gas adsorption capacity. By doing so, the sensitivity, accuracy, and overall performance of BN-based gas sensors can be improved. This modification holds promise for the advancement of the use of 2D BN in gas-sensing applications.

The research mentioned highlights that pristine BN demonstrates good adsorption for SO₂ but shows limited sensitivity to CO₂ and NO₂ due to low adsorption amounts, energies, and minimal charge transfer [18]. This limitation necessitates exploring methods to enhance the gas adsorption behavior of BN. One effective approach involves doping metal atoms onto the perfect crystal surfaces of BN, which is relatively easier to achieve compared to creating vacancy defects, and it offers higher enhancement effects. However, this method does come with the disadvantage of increased raw material costs. Transition metal atom doping has proven to be particularly effective in improving gas adsorption behavior; this is primarily due to the significant electron hybridization between the adsorbent and gas molecules after doping. Common transition metal-doped particles include chromium (Cr), nickel (Ni), and silver (Ag). For instance, Hao Cui employed Rh monoatoms to optimize the band structure of MoSe₂ monolayers, resulting in improved adsorption capacity for CO, NO, and NO₂ [19]. Similarly, Y Chen investigated the adsorption characteristics of Pd cluster-doped GaNNTs on C₂H₂ and H₂ and found that Pd_n-GaNNTs (n = 1–2) exhibited higher reactivity and sensitivity to these gases compared to pristine GaNNTs [20]. Despite these advancements, few scholars have focused on the adsorption and sensing studies of conventional transition metal-modified BNs. This suggests a potential research gap that could be explored further to develop BN-based gas sensors with enhanced sensitivity and performance. By investigating the effects of different transition metal dopants on BN, researchers can identify optimal doping strategies to improve the gas adsorption behavior of BN and potentially pave the way for more sensitive and accurate BN-based gas sensors. Cr, a traditional transition metal, has excellent physicochemical properties, as well as sizable oxidation and heat resistance, and can substantially enhance the carrier capacity of two-dimensional material surfaces. Its strong d-orbital electronic activity has been demonstrated using a combination of experiments and simulations to significantly enhance the responsiveness of gas sensors and has therefore been selected as the dopant in this paper [9].

Drawing upon the exceptional electronic attributes and promising gas-sensing potential of BN nanomaterials, this research introduces, for pioneering exploration, the utilization of chromium (Cr)-functionalized BN nanosheets to examine the adsorption efficiency of fault gases emanating from substation secondary equipment. Leveraging density functional theory, a spectrum of plausible Cr-doped BN configurations was formulated. Based on considerations of doping energy and geometric arrangement, the most stable Cr-BN models across two doping categories were identified. Employing these stable models, the study delved into the adsorption and sensing characteristics of diagnostic fault gases (O₃, CO, NO₂) relevant to substation secondary equipment. An in-depth analysis was conducted, scrutinizing alterations in band structures, density of states, differential charge density, and molecular frontier orbital theory throughout the surface interaction process. This investigation unveiled the adsorption mechanisms of fault gases from substation secondary equipment onto Cr-BN surfaces. Consequently, this research offers theoretical insights to propel the advancement of Cr-BN nanosensors for insulation evaluations and real-time monitoring of substation secondary equipment.

2. Materials and Methods

In this paper, all calculations were conducted using Materials Studio 2023, specifically leveraging the density functional theory (DFT) implemented within Dmol3 [21]. The exchange-correlation function was formulated using the Perdew–Burke–Ernzerhof (PBE) correction within the generalized gradient approximation (GGA) framework [22]. To avoid interactions among neighboring supercells, periodic boundary conditions were applied, with a supercell configuration of 4 × 4 × 1 and a vacuum layer of 10 Å × 10 Å × 12 Å. Grimme’s dispersion adjustment was used to account for the long-range interaction force between the dopants and the basement. The spin was treated as unrestricted, starting with formal spin as the initial configuration. Sampling of the Brillouin zone was performed using the Monkhorst–Pack method, with a k-point grid of 6 × 6 × 1. For the calculations, the double numerical polarization (DNP) basis set was employed, and the density functional theory semi-core pseudopotential was selected for core electron treatment [23]. Convergence criteria were established at 1 × 10⁻⁵ Ha, 2 × 10⁻³ Ha/Å, and 5×10⁻³ Å for maximum displacement. Additionally, a self-consistent field convergence threshold of 1 × 10⁻⁶ Ha was imposed [24,25,26].

In this paper, various doping sites were taken into account, and the doping energy (E_b) was computed to achieve an optimized, modified structure. The formula utilized for calculating E_b in the case of Cr-doped BN is as follows:

E_b = E_Cr-BN − E_BN − E_Cr

(1)

where E_Cr-BN denotes the total energy of the doped system, E_BN represents the total energy of pure BN, and E_Cr stands for the energy of an isolated Cr atom. The doping energy, E_b, quantifies the energetic cost or benefit of the doping process. By examining Eb, one can assess the stability of the resulting material.

Once the optimized doping structure was secured, models were developed to simulate the adsorption of three target gases onto the surfaces of two doped materials. The adsorption energy E_ads, defined by Equation (2), serves as a metric to quantify the interaction strength between the target gas molecules and the adsorbent surface.

E_ads = E_gas/Cr-BN − E_Cr-BN − E_gas

(2)

where E_gas/Cr-BN signifies the total energy of the system subsequent to gas adsorption, E_Cr-BN denotes the total energy of the doped structure, and E_gas stands for the energy of the gas molecule prior to adsorption. The adsorption energy (E_ads), defined in this context, reflects the energetic change associated with the surface reaction process during gas adsorption. A negative Eads indicates that energy is released, pointing towards spontaneous adsorption.

Moreover, to investigate how doping modifications and adsorbed gases influence electron movement, we calculated the Mulliken population numbers for all doping and adsorption models. The charge transfer (ΔQ) is defined as follows:

ΔQ = Q_a − Q_b

(3)

where Q_a and Q_b represent the total charge quantities after and before the process of adsorption or doping, respectively. A positive ΔQ signifies that electrons are lost by the gas molecules, whereas a negative ΔQ indicates that electrons are gained by the gas molecules from the doped material.

3. Results and Discussion

3.1. Target Gases and Doping Models

3.1.1. Geometric Structures of Target Gases and Doping Models

Figure 1 displays the geometrically refined structures of pristine BN and three notable fault gases typically found in secondary equipment within substations. For this study, we focused on carbon monoxide (CO), nitrogen dioxide (NO₂), and ozone (O₃) as the primary gases indicative of faults in such equipment. The molecular configurations for these gases are presented in Figure 1a–d. Specifically, CO exhibits a linear molecular form, whereas both NO₂ and O₃ share similar characteristics, featuring planar symmetric structures. Prior to gas adsorption, the C-O bond length is the shortest at 1.142 Å, in contrast to the N-O and O-O bond lengths in the NO₂ and O₃ molecules, which are 1.209 Å and 1.279 Å, respectively. BN, on the other hand, possesses a two-dimensional hexagonal lattice reminiscent of pristine graphene, with evenly spaced B and N atoms, and a B-N bond length of 1.467 Å, as illustrated in Figure 1d.

To enhance the performance of BN, this study introduces Cr doping for modification. During the Cr modification process, multiple potential sites are available. Two doping sites in the defect modification process were selected for structure and energy optimization. The optimized results are presented in

Table 1. The energy associated with Cr doping in BN obtained through various doping techniques.

	Cr-BN(B-Vac)	Cr-BN(N-Vac)
Emod	−1.289	−1.889
QCr	0.292	−0.358

and Figure 2a,b, where Cr-BN_(B-Vac) and Cr-BN_(N-Vac) represent Cr-doped BN after modification with B and N defects, respectively. The columns show the modified energy (E_mod) and the charge on Cr (Q_Cr).

To improve the performance of BN, this study proposes the incorporation of Cr doping as a modification method. Throughout the Cr doping process, numerous potential sites are accessible for modification. For the purpose of structural and energy optimization, two specific doping sites were chosen during the defect modification process. The outcomes of these optimizations are displayed in Table 1 and Figure 2a,b, where Cr-BN_(B-Vac) and Cr-BN_(N-Vac) signify BN doped with Cr following modifications involving B and N defects, respectively. The tables and figures detail the modified energy (E_mod) and the charge on Cr (Q_Cr).

Purple spheres depict potential locations for Cr metal particles, divided into B-defect sites (where doping atoms substitute B atoms in pristine BN) and N-defect sites (where doping atoms replace N atoms in pristine BN). Figure 2a,b exhibit the most stable doping configurations achieved after optimization, with doping energies of −1.289 eV and −1.889 eV, respectively. Furthermore, Table 1 reveals that the charge transfer quantities for these two materials are 0.292 e and −0.358 e, implying that Cr atoms function as electron acceptors and donors, respectively, in these materials. This underscores the distinct impacts of doping sites on the material’s properties.

3.1.2. Electronic Performance Analysis of Doping Process

The electronic band structures provide insights into a material’s electronic properties. To gain an understanding of the electronic behavior of the material following Cr doping modification, we computed the band structures of both pristine BN and BN modified with two distinct types of Cr doping, as depicted in Figure 3. Our findings show considerable changes in the band structures before and after doping, suggesting low effective masses for electrons. Prior to the introduction of metal particles, BN exhibits a substantial band gap of 4.561 eV, which is characteristic of a typical semiconductor. However, upon substituting B and N atoms with Cr particles, the band gap values diminish to 1.924 eV and 0.971 eV, respectively. This demonstrates a profound effect of doping on the material’s band gap, primarily due to the influence of the 4d orbital electrons of Cr atoms.

In this study, we employed the density of states (DOS) theory and differential charge density surface analysis to delve deeper into the alterations in the electronic properties of BN before and after doping, as presented in Figure 4 and Figure 5, respectively. Figure 4 demonstrates that, although the overall distribution of the total DOS remains largely unchanged before and after doping, the introduction of Cr dopant atoms introduces certain modifications to the system’s total DOS. Specifically, the total DOS for both Cr-doped BN materials shifts slightly towards the left when compared to pristine BN, accompanied by a minor reduction in the electron count on both sides of the Fermi level. Based on the band structure after Cr doping, we hypothesize that under the influence of the dopant, the electrons gradually move from the lower end of the conduction band to the upper end of the valence band, which leads to a decrease in the overall resistivity of the system and an increase in the electronic conductivity of the crystal. Furthermore, the emergence of additional small peaks in the total DOS after doping hints at a significant impact of Cr atom doping on the electronic behavior of pristine BN, introducing more activated states within the material.

The differential charge density (DCD) analysis before and after doping was also conducted, as illustrated in Figure 5. In this figure, red and blue hues signify an increase and decrease in charge density, respectively, with a color scale spanning from −0.1 e/Å to 0.1 e/Å. Notably, there are distinct areas of electron depletion and electron accumulation surrounding Cr atoms and BN units, indicating that charge transfer predominantly takes place between Cr and BN. Additionally, the arrangement of these electron depletion and accumulation regions in proximity to Cr atoms varies between the two materials obtained through different doping techniques. This variation confirms that Cr assumes different roles in the electron donor–acceptor relationship under distinct doping methods, aligning with our previous analysis of charge transfer outcomes.

3.2. Adsorption and Electronic Performance Analysis of Adsorption Process

3.2.1. Adsorption Performance Analysis of Adsorption Process

In this section, we develop models for the adsorption of target gases onto the materials under investigation. The most stable configurations of these adsorption systems were determined based on various criteria, as depicted in Figure 6. The adsorption energy, adsorption distance, and charge transfer values are summarized in

Table 2. Performance metrics for the adsorption of fault gases from secondary equipment in substations using Cr-doped BN materials.

Material	Gas	Model	Eads/eV	Distance/Å	Charge Transfer/eV
Cr-BN(B-Vac)	CO	M1	−1.029	2.018	0.124
	O3	M2	−4.452	1.607	−0.512
	NO2	M3	−2.576	1.982	−0.386
Cr-BN(N-Vac)	CO	N1	−1.550	1.973	0.032
	O3	N2	−8.577	1.736	−1.017
	NO2	N3	−3.024	1.866	−0.339

. Specifically, Figure 6 M1–M3 and N1–N3 display the most stable adsorption structures for three target gases on Cr-BN_(B-Vac) and Cr-BN_(N-Vac), respectively. These figures reveal distinct adsorption behaviors of the two materials towards the three fault gases emanating from secondary equipment in substations. Our findings indicate that both Cr-BN_(B-Vac) and Cr-BN_(N-Vac) undergo notable surface reactions with the target gas molecules, accompanied by a subtle curvature change on the BN surface. Furthermore, the adsorption capacity of these materials for the fault gases follows the sequence: O₃ > NO₂ > CO, with corresponding adsorption energies of −4.452 eV, −2.576 eV, and −1.029 eV for Cr-BN_(B-Vac) and −8.577 eV, −3.024 eV, and −1.550 eV for Cr-BN_(N-Vac). The adsorption mechanisms are categorized as physicochemical in nature. Notably, during O₃ adsorption, chemical bond cleavage occurs, forming Cr-O bonds with high adsorption energies of −4.452 eV and −8.577 eV for the two materials, respectively, showcasing their robust adsorption capability for O₃. Additionally, from a geometric structure perspective, CO and NO₂ molecules undergo activation during adsorption, with elongated C-O and N-O bond lengths compared to their pre-adsorption states, primarily due to weak van der Waals interactions. In terms of charge transfer, all target gases except CO function as electron donors during the surface reaction, with the amount of charge transfer correlating with the adsorption strength. Consequently, Cr-modified BN exhibits a robust gas-sensing response. The adsorption of Cr-BN on the three target gases mainly tends to be chemisorption accompanied by chemical bond breaking.

3.2.2. Electronic Performance Analysis of Adsorption Process

Figure 7 displays three distinct band structures resulting from the adsorption of fault gases from substation secondary equipment onto two varieties of Cr-BN, providing a comprehensive understanding of the information on the electronic structure of the system.

Upon analyzing the trends in band gap changes, it is evident that the adsorption of CO, O₃, and NO₂ molecules has contrasting effects on the band gaps of Cr-BN_(B-Vac) and Cr-BN_(N-Vac). Specifically, for Cr-BN_(B-Vac), the band gap widens upon adsorption of these gases. In contrast, for Cr-BN_(N-Vac), the band gap narrows when O₃ is adsorbed but widens to varying degrees upon adsorption of CO and NO₂. A deeper examination reveals that the adsorption of CO, O₃, and NO₂ leads to a narrowing of both the valence and conduction bands in Cr-BN_(B-Vac), ultimately increasing the band gap value of the whole system after the adsorption of the target gas. A similar pattern is observed in Cr-BN_(N-Vac) upon gas adsorption. For Cr-BN_(B-Vac), the band gap values decrease from 1.924 eV to 1.103 eV, 0.271 eV, and 1.312 eV after CO, O₃, and NO₂ adsorption, respectively, with percentage reductions of 42.67%, 85.91%, and 31.81%. Notably, O₃ has the most significant impact on reducing the band gap, which may be attributable to the strong chemical interaction during adsorption, facilitating electron flow, enhancing excitation effects, and promoting electron transitions. Similar mechanisms explain the impact of gas adsorption on the band gap of Cr-BN_(N-Vac). In addition, the behaviors described above will contribute to the improvement of the quality of the electronic validity.

The differential charge density distribution illustrated in Figure 8 indicates that the primary charge transfer occurs in the vicinity of the adsorbed gas molecules and near the Cr dopants. During the reaction, CO, O₃, and NO₂ act as electron acceptors by gaining electrons, while the Cr dopants primarily function as electron donors. Notably, in comparison to the CO and NO₂ adsorption systems, the electron dissipation and accumulation regions in the O₃ adsorption system are larger and more pronounced, suggesting a more intense electron movement during the adsorption of O₃.

4. Conclusions

Utilizing density functional theory, this research explores the adsorption and sensing capabilities of two varieties of Cr-modified BN nanosheets towards fault gases present in the secondary equipment of substations. A comprehensive analysis of the adsorption characteristics and electronic attributes of three fault gases on two Cr-BN materials is carried out by assessing their band structures, density of states, and charge density variations. The key findings can be summarized as follows:

• The incorporation of metallic chromium (Cr) doping substantially boosts the native sensitivity of BN nanosheets to fault gases found in the secondary equipment of substations, offering an abundance of active adsorption sites for gases.
• The adsorption capabilities of the two Cr-modified BN materials for fault gases in the secondary equipment of substations are ranked as follows: O₃ > NO₂ > CO. The corresponding adsorption energies are −4.452 eV and −8.577 eV for O₃, −2.576 eV and −3.024 eV for NO₂, and −1.029 eV and −1.550 eV for CO, respectively. The nature of these adsorptions is characterized as physical–chemical in type.
• After adsorption, various fault gases exhibit different impacts on the electronic properties of the entire system. These variations in adsorption types and strengths lead to the generation of distinct electrical signals, which allows the differentiation of different fault gases.

These results underscore the substantial promise of Cr-BN nanosensors as gas-sensitive devices for real-time monitoring of fault gases in the secondary equipment of substations. The theoretical framework established in this study lays the groundwork for the development of Cr-BN nanosensors. Future endeavors will encompass the fabrication of these sensors and their gas-sensitive response testing, aiming to further propel their development and practical deployment.