Metal-modified C₃N₁ monolayer sensors for battery instability monitoring

Abstract

The pressing need for affordable gas sensors with enhanced sensitivity and selectivity in identifying hazardous gases released after the battery becomes unstable cannot be overstated. In this study, a C₃N₁ monolayer modified with Cu and Ag atoms (Cu/Ag–C₂N₁) was selected to achieve selective adsorption of NO₂ under the coexistence of multiple gases (PF₅, NH₃, H₂O, C₂H₄, and C₂H₆) based on density functional theory. The results demonstrate that securely anchoring metal atoms to the monolayer, as indicated by cohesion energy and ab initio molecular dynamics simulations, concurrently enhances the material's conductivity. Analyses of electrostatic potential and work function identified high activity sites and electron-releasing capabilities. Furthermore, the gas–solid interface structures of multiple gases on the Cu/Ag–C₂N₁ monolayers are revealed by the adsorption energy and distance. Importantly, NO₂ exhibits stronger adsorption energy on Cu/Ag–C₂N₁, reaching −3.54 and −3.27 eV, respectively. Crystal Orbital Hamilton Population and d-band center theory unveiled differences in adsorption energy resulting from the modification involving the two metals. Fascinatingly, density of states calculation demonstrates, for the first time, that the two doped metal monolayers generate a distinct response solely to NO₂ in a multi-gas coexistence setting, effectively excluding interference from water. In practice, based on Gibbs free energy and Einstein diffusion law calculations, Cu–C₂N₁ exhibits superior hydrophobicity, a broader temperature range and a lower diffusion activation energy barrier (2.5 kJ mol⁻¹). Our theoretical calculations demonstrate Cu's efficacy in substituting expensive Ag, yielding cost-effectiveness without compromising selectivity, response, stability, and versatility.

---

1 Introduction

Widespread in portable electronics and integral to the surge in electric vehicles, batteries play a pivotal role, significantly impacting the tapestry of modern society.^1–3 However, the pursuit of high-energy-density batteries raises safety concerns due to increased gas evolution, particularly with lithium metal.^4,5 Safety monitoring is vital as battery defects can trigger severe risks like electrolyte-induced heat leading to fire and explosion. This thermal runaway results in the release of harmful gases.⁶ Lithium metal poses challenges due to its reactivity, potentially causing dendritic formation and a hot mixture release, a precursor to fire or explosion. Early detection of gases like NO₂, PF₅, NH₃, C₂H₄, H₂O, and C₂H₆ before thermal runaway is imperative to prevent safety hazards in practical applications.^7,8

In recent years, an abundance of theoretical studies has delved into the design and fabrication of monolayers for detecting toxic gases. Theoretical calculations, employing density functional theory (DFT), have played a pivotal role in accurately predicting the sensing efficiency of two-dimensional (2D) monolayers and elucidating their micro-sensing mechanisms.^9–17 For instance, Sun et al.¹⁸ utilized DFT calculations to analyze the adsorption behavior of six gases on an indium nitride monolayer, demonstrating its potential in detecting SO₂ and NO₂ molecules. Additionally, first-principles calculations were employed to explore the structural and electronic properties of group III nitrides and phosphides in gas adsorption. It was concluded that significant potential exists for the detection of CO₂ by these sensors. Furthermore, the introduction of transition metal doping can enhance the sensing capabilities of 2D monolayers.^19–21

In gas-sensitive applications, 2D monolayers demonstrate exceptional performance owing to their high specific monolayer area, thermal conductivity, unique optoelectronic properties, and catalytic activity.^22–25 As chemical gas sensors, these monolayer offer abundance, cost-effectiveness, and easy manufacturability, and play a pivotal role in sensor technology. Among numerous 2D material-based gas sensors, their sensitivity to various toxic and explosive gases stems from gas-induced resistance changes,²⁶ finding wide application across diverse fields. Graphene and C₃N monolayer films, among others, have theoretically shown promising gas-sensing characteristics.^27–29 Despite this, inherent defects limit the full application potential of 2D monolayers in sensing technologies. Hence, exploring and discovering novel types of two-dimensional monolayers for gas detection become imperative to address these limitations and expand the sensor monolayer repertoire.

Creating new gas sensors to detect typical electrolyte decomposition products under harsh conditions is crucial. 2D monolayers stand out as prime candidates for such sensing applications, operating within a high temperature range and exhibiting the highest response levels.³⁰ Enhancing gas sensor characteristics involves monolayer material modification. Semiconductor-based metal oxide gas sensors also require temperature resistance during thermal runaway, making safety-based detectors particularly promising. These devices offer stability under high humidity conditions but demand high-temperature resilience during thermal runaway.^31,32 Metal-doped 2D material gas sensors exhibit notably high responses, showcasing how metals create a platform to merge diverse physical and chemical properties within a system. Past research has amalgamated experiments and intricate computational analyses to validate the applicability of metal-doped structures in sensor applications.^33–36 These studies have visualized the gas sensing traits, portraying the adsorption performance of gas sensors for practical implementation.

However, the electrical response of the sensing monolayer originates from the cumulative effect of each gas. The significant scientific challenge lies in the absence of an effective method to systematically allocate the contribution of each gas. Consequently, in the case of the simultaneous adsorption of multiple gases, the strategy to enhance selectivity involves eliminating the electrical response to other interfering gases, focusing solely on one specific gas.

In this study, DFT was employed to calculate the response of Cu/Ag–C₂N₁ in the presence of multiple gases (PF₅, NO₂, NH₃, H₂O, C₂H₄, and C₂H₆). Initially, 2D monolayers doped with these two metals were constructed, and their stability and plausibility were examined from various perspectives. Subsequently, the adsorption behavior of multiple gases on the doped monolayer was comprehensively calculated, elucidating differences in adsorption performance based on methods such as d-band center theory and Crystal Orbital Hamiltonian Population (COHP). Utilizing the density of states (DOS) method, the electrical signal response behavior of the monolayer during the adsorption of multiple gases for selective adsorption was evaluated. Finally, the practical applicability of the sensors was explored by computing the Gibbs free energy of gas adsorption and the recovery time of the electrical signal. Our theoretical exploration offers novel insights for designing highly selective gas sensors in battery instability monitoring.

2 Results and discussion

2.1 Structural characterization

The optimized structures of the Cu/Ag–C₂N₁ monolayer and six gas molecules (PF₅, NO₂, NH₃, H₂O, C₂H₄, and C₂H₆) are shown in Fig. 1a. It exhibits a top-down view of the metal-modified C₂N₁ monolayer, comprising a pattern of corner rings formed by C–N–C–C sequences, while also illustrating the interaction sites of Ag and Cu. In detail, the P–F bond length in PF₅ is 1.58 Å, the N–O bond length in NO₂ is 1.27 Å, the N–H bond length in NH₃ is 1.03 Å, and the H–O bond length in H₂O is 0.98 Å. For C₂H₄, the C–H bond length is 1.09 Å, and the C–C bond length is 1.33 Å. For C₂H₆, the C–H bond length is 1.06 Å, and the C–C bond length is 1.53 Å. These data calculated in this work are consistent with those previously reported,^37–44 demonstrating that the selected calculation parameters are reasonable and accurate. Fig. 1b and c depict the main view of the monolayer of Ag–C₂N₁ and Cu–C₂N₁, respectively. Clearly, both doped metals are situated atop the 2D material, creating the most stable structure, in accordance with earlier research observations.²⁸ So as to understand the monolayer stability of doped metal atoms from the point of view of charge density, Fig. 1b and c also illustrate distinct charge density distribution. The noticeable increase in red indicates a heightened charge density of electronic wavefunction, predominantly around the C and N atoms, confirming its high charge acceptance. Both Ag and Cu act as electron donors, transferring electrons to C₂N₁. Therefore, to maintain an optimal stable structure, the doped metal atoms ultimately position themselves above the monolayer, resulting in substrate deformation.

Fig. 1 (a) Structures of the Cu/Ag–C₂N₁ monolayer, PF₅, NO₂, NH₃, H₂O, C₂H₄, and C₂H₆. (b and c) Stable structure and charge density distribution of (b) Cu–C₂N₁ and (c) Ag–C₂N₁. (d and e) DOS and PDOS of (d) Ag–C₂N₁ and (e) Cu–C₂N₁. Densities are displayed with an isosurface and the interval of isovalues is between −0.1 and 0.1 e Å⁻³.

Additionally, cohesion energy (E_coh) analyses were conducted to quantitatively explore the anchoring of metal atoms onto the monolayer. First, the E_coh was analyzed as:⁴⁵

where E_C and E_N represent the energies (eV) of C and N atoms, respectively. n₁, n₂, and n₃ are the numbers of metal, C, and N atoms. The E_coh of the initial C₃N₁ monolayer was calculated to be 7.01 eV. After metal doping, it decreased to 6.67 eV for Cu–C₂N₁ and 6.70 eV for Ag–C₂N₁, showcasing a reduction within the expected range of 4%. This highlights the stable bonding of metals to the monolayer without undergoing aggregation. To evaluate the dynamic stability of this structure, ab initio molecular dynamics simulation was conducted. A total of 5000 steps were calculated over 50 ps, using a time step of 10 fs for output. The results, confirming the stability of the monolayer structures, are illustrated in Fig. S1.† In the realm of top-notch gas sensors, it's crucial to understand the electron transport capabilities and active sites of the material for optimal performance. The electrostatic potential and work function were calculated, with the work function representing the minimum energy required to remove electrons from the monolayer. As shown in Fig. S2,† the results reveal a comparable work function for Ag–C₂N₁ (3.67 eV) and Cu–C₂N₁ (3.72 eV), suggesting a similar electron migration capability. Furthermore, based on the high electrostatic potential, it can be confirmed that the doped metal sites exhibit heightened activity.

Based on previous research,^{12,19,46–49} the height of the DOS curve at the Fermi level can reflect the quality of material conductivity. In other words, higher peaks indicate better conductivity. Conductivity stands as a crucial criterion for high-performance gas sensors. Therefore, this study employs this approach to assess the impact of introducing doping metals on the material's conductivity. In Fig. 1d and e, the overall DOS demonstrates a slight increase in the curve around the Fermi level after the introduction of doping metals, indicating an enhancement in the material's conductivity. Additionally, partial density of states (PDOS) calculations were conducted to delve deeper into the behavior of the doped metal atoms. Fig. 1d and e display the d orbitals of the metal atoms and the p-orbitals of nonmetallic C/N in the monolayer. Typically, transition metal d-orbitals play a significant role in bonding. The outcomes reveal a noticeable alignment between the d orbital of Ag and the p orbital of C/N within the material. Furthermore, the hybridization of Cu's d orbital with C/N's p orbital is more prominent and situated at a higher energy level. The findings indicate that Cu and Ag can be securely affixed to the C₂N₁ monolayer. This indicates a robust interaction between the doped metal atoms and the monolayers, ensuring structural stability.

2.2 Gas sensing performance

2.2.1 Gas adsorption behavior. In this section, the optimal gas adsorption configurations were initially identified. Subsequently, the adsorption energy and distance were quantitatively analyzed. Finally, the differences in gas adsorption performance were elucidated using COHP and d-band center theory. As depicted in Fig. S3–S14,† various gas adsorption sites and configurations were explored on Ag–C₂N₁ and Cu–C₂N₁. Initially, all gas molecules positioned 2.5 Å above the monolayer underwent structural optimization calculations to determine their equilibrium positions. Based on the positive correlation between the adsorption distance and energy,^50,51 the initial configurations following this criterion were selected, as illustrated in Fig. S15 and S16.†Fig. 2 and 3 show the most stable adsorption configurations and charge densities of Ag–C₂N₁ and Cu–C₂N₁ for the six gases (C₂H₄, PF₅, C₂H₆, H₂O, NH₃, and NO₂).

Fig. 2 Stable configurations and charge density of gas adsorbed on the Ag–C₂N₁ monolayer: (a) Ag–C₂N₁–C₂H₄, (b) Ag–C₂N₁–C₂H₆, (c) Ag–C₂N₁–PF₅, (d) Ag–C₂N₁–H₂O, (e) Ag–C₂N₁–NH₃, and (f) Ag–C₂N₁–NO₂. Densities are displayed with an isosurface and the interval of isovalues is between −0.1 and 0.1 e Å⁻³.

Fig. 3 Stable configurations and charge density of gas adsorbed on the Cu–C₂N₁ monolayer: (a) Cu–C₂N₁–C₂H₄, (b) Cu–C₂N₁–C₂H₆, (c) Cu–C₂N₁–PF₅, (d) Cu–C₂N₁–H₂O, (e) Cu–C₂N₁–NH₃, and (f) Cu–C₂N₁–NO₂.

In Fig. 2, the shortest atomic distances observed are 2.3 Å for Ag–C₂N₁–C₂H₄, 2.5 Å for Ag–C₂N₁–PF₅, 2.2 Å for Ag–C₂N₁–C₂H₆, 2.5 Å for Ag–C₂N₁–H₂O, 2.3 Å for Ag–C₂N₁–NH₃, and 2.2 Å for Ag–C₂N₁–NO₂. In Fig. 3, the shortest atomic distances are 2.1 Å for Cu–C₂N₁–C₂H₄, 2.5 Å for Cu–C₂N₁–PF₅, 2.0 Å for Cu–C₂N₁–C₂H₆, 2.1 Å for Cu–C₂N₁–H₂O, 2.1 Å for Cu–C₂N₁–NH₃, and 1.9 Å for Cu–C₂N₁–NO₂. After optimization, the positions of all gas molecules have shifted. The shortest distance of each gas molecule in the Ag–C₂N₁ system is slightly longer than that of the Cu–C₂N₁ system. Additionally, for a more visual examination of the charge accumulation and depletion during gas interaction with Cu/Ag–C₂N₁, Fig. 2 and 3 present the results of charge density analysis. The charge densities of electronic wavefunction illustrate that both Ag–C₂N₁ and Cu–C₂N₁ transfer a significant amount of charge during gas adsorption. The distinct dark red color observed around the gas molecules (C₂H₄, C₂H₆, PF₅, H₂O, NH₃, and NO₂) clearly indicates the acquisition of electrons from Ag–C₂N₁ and Cu–C₂N₁ by the gas molecules, demonstrating a strong interaction between the gas and Cu/Ag–C₂N₁. For specific charge transfer values, refer to Table S1.† The results indicate that among the six gases considered, NO₂ exhibited the most significant charge transfer to Cu/Ag–C₂N₁, suggesting its superior adsorption effect on it, particularly on the Cu–C₂N₁ monolayer.

For a more comprehensive understanding of gas adsorption behavior and to quantitatively assess the interaction forces between gases and Cu/Ag–C₂N₁, it becomes imperative to compare adsorption distances and energies. Fig. 4a and b show respectively the adsorption distance and adsorption energy. The adsorption energy is defined as:^52–54

E_ads = E_total − (E_monolayer + E_gas) (2)

where E_ads is the adsorption energy (eV), E_total is the total energy of the monolayer with the gas molecule (eV), E_monolayer is the total energy of the monolayer (eV), and E_gas is the total energy of an isolated gas molecule (eV). For the adsorption distance, the darker the color, the shorter the adsorption distance. From Fig. 4a it can be intuitively seen that the color of the row where Cu is located is darker than the color of the row where Ag is located, that is, the adsorption distance of the gas under the Cu system is less than that of the Ag system, indicating that the Cu–C₂N₁ monolayer has better adsorption performance for the gas. The adsorption distance of Cu–C₂N₁–NO₂ is the smallest, showcasing the exceptional adsorption effectiveness of Cu–C₂N₁ for NO₂. Additionally, Cu–C₂N₁ exhibits relatively small adsorption distances for C₂H₄, C₂H₆, NH₃, and H₂O, indicating noteworthy adsorption performance. However, the analysis of adsorption energy suggests that Cu–C₂N₁ interacts with C₂H₄, C₂H₆, and NH₃ through physical adsorption based on our previous work.¹⁹ The column where NO₂ is situated appears significantly darker compared to the columns of the other gases (C₂H₄, C₂H₆, NH₃, H₂O, and PF₅) in Fig. 4b. This implies that among the six gases, only NO₂ exhibits a pronounced adsorption effect on the C₂N₁ monolayer after metal doping, and its adsorption energy is much stronger than that of the other five gases. This realization enables the selective adsorption of NO₂ gas on the monolayer in various gas environments, unaffected by the presence of other gases. Additionally, it's noteworthy that the adsorption energy of H₂O on Cu–C₂N₁ is weaker, indicating that Cu–C₂N₁ exhibits virtually no adsorption effect on H₂O. Consequently, humid environments do not affect the performance of Cu–C₂N₁.

Fig. 4 (a) The adsorption distance and (b) adsorption energy of gases (C₂H₄, C₂H₆, NH₃, NO₂, H₂O, and PF₅) on the Ag–C₂N₁ and Cu–C₂N₁ monolayers. (c) The d-band center of C₂N₁ doped with metals (Cu and Ag). (d) Metal–adsorbate interaction by altering the metal d-band center. (e) ICOHP of gases on the Ag–C₂N₁ and Cu–C₂N₁ monolayers.

Combining the different gas adsorption properties caused by the above two doped metal atoms, it can be found that the performance of Cu is better than that of Ag. For example, the adsorption energy of Cu–C₂N₁–NO₂ is −3.53 eV, which is stronger than that of Ag–C₂N₁–NO₂ (−3.26 eV). Given that d-band center theory is a useful tool for elucidating the interaction between transition metals and gas molecules, the calculation of the d-band center (ε_d) becomes imperative to unveil the underlying mechanism. The ε_d can be calculated as:^55–57

where ε is the energy and n_d(ε) is the density. Fig. 4c shows the ε_d of Cu/Ag–C₂N₁. The larger the ε_d (close to the Fermi level), the better the adsorption effect on gas. It is better proved that NO₂ has the greatest adsorption strength on Cu–C₂N₁. Fig. 4d shows that the ε_d of Cu–C₂N₁ is closest to the Fermi level, demonstrating that the proportion of unoccupied antibonding states increases, resulting in the strongest interaction between NO₂ and Cu–C₂N₁.

To expose disparities in gas adsorption energy on Cu–C₂N₁ and Ag–C₂N₁ from a chemical bonding perspective, the chemical bond is categorized into bonding and anti-bonding, with bonding playing the decisive role. COHP analysis is employed for a more accurate examination of interatomic forces during gas adsorption.^58–61 The detailed calculation results are displayed in Fig. S17 and S18.† COHP represents the respective contributions of bonding and antibonding and shows the strength of the chemical bond between atoms, and the larger the bonding state below the Fermi level, the smaller the antibonding state, the stronger the bonding ability between the two atoms. In the example Fig. S17† of Ag–C₂N₁–PF₅, there are a large number of anti-bonds below the Fermi level, resulting in a huge internal stress between Ag and F atoms, and the anti-bonds in the system cause repulsion between the atoms, and the interaction cannot be stable. In contrast, there is a large amount of bonding below the Fermi level in Cu–C₂N₁–NO₂ in Fig. S18,† indicating that there is a strong force between Cu and N, which confirms the strong interaction between Cu–C₂N₁ and NO₂. A more quantitative explanation can be provided by calculating the integral COHP (ICOHP) value obtained by calculating the energy integral from negative infinity to the Fermi level. The more negative the ICOHP value, the stronger the interaction between metal atoms and gas atoms, as shown in Fig. 4e. Compared to the other five gases, NO₂ exhibits the most prominent ICOHP values on Cu–C₂N₁ and Ag–C₂N₁. Therefore, this result substantiates the clear advantage of competitive adsorption for NO₂ on Cu–C₂N₁.

2.2.2 Electrical response behavior. The selectivity is the most crucial performance indicator for gas sensors. Based on our previous work,^19,47,48 the DOS at the Fermi level can effectively represent the material's conductivity. Therefore, the electrical response behavior induced by the adsorption of various gases on the Cu/Ag–C₂N₁ monolayer is comprehensively analyzed. Detailed calculation results are presented in Fig. S19 and S20.† To visually determine which material exhibits the best selective performance, all electrical response values are optimized to obtain the selectivity index shown in Fig. 5. This selectivity is derived from the comparison of the electrical conductivity before and after gas adsorption. If the conductivity remains unchanged, it indicates that the gas has no effect on the material. Conversely, any change suggests that the material demonstrates a sensitive response to the gas.

Fig. 5 Selectivity of gases on the (a) Ag–C₂N₁ and (b) Cu–C₂N₁ monolayers.

In Fig. 5, the conductivity of NO₂ adsorption was significantly higher than that of the background, leading to a higher electron count near this level, ultimately enhancing conductivity. In the context of NO₂ adsorption, the redistribution of charges is the primary driver behind this shift, causing a significant alteration in conductivity. The C₂N₁ monolayer modified with metal atoms (Ag and Cu) displays a selective response to NO₂ gas amidst various other gases (C₂H₄, C₂H₆, PF₅, H₂O, and NH₃). Notably, the DOS for all orbitals on the right side of the Cu–C₂N₁–NO₂ Fermi level surpasses that of Ag–C₂N₁–NO₂, indicating a more robust response of Cu–C₂N₁ to NO₂ compared to its Ag–C₂N₁ counterpart.

The discernible charge transfer path is directly reflected in the DOS alteration seen in Fig. S19 and S20.† The conductivity of C₂H₄, C₂H₆, H₂O, and NH₃ on the monolayer of Ag–C₂N₁ changed little after adsorption, and the conductivity decreased slightly after PF₅ adsorption. The conductivity of C₂H₆, PF₅, H₂O and NH₃ on the monolayer of Cu–C₂N₁ did not change significantly after adsorption, but the conductivity of C₂H₄ decreased slightly. There was no significant change in the conductivity of the gas after adsorption, which was consistent with the observed trend. Observing the changes of the DOS before and after adsorption of H₂O in Fig. S19 and S20d,† it can be found that the DOS curves before and after adsorption change very small, and there is no change at the Fermi level, which indicates that Ag–C₂N₁ and Cu–C₂N₁ have no electrical signal response to H₂O and the monolayer is hydrophobic. Interestingly, when NO₂ is adsorbed on the monolayer of Ag–C₂N₁ and Cu–C₂N₁, the Cu/Ag–C₂N₁ monolayer conductivity increases significantly, which is reflected in the obvious rise and left shift of the DOS curve at the Fermi level after NO₂ adsorption in the DOS diagram. This is in contrast to the DOS changes after adsorption of other gases. It is proved that the monolayers of Ag–C₂N₁ and Cu–C₂N₁ can generate a selective electrical signal response to NO₂ gas. It is worth noting that the DOS change of NO₂ on the monolayer of Cu–C₂N₁ is more obvious than that of Ag–C₂N₁, which proves that Cu–C₂N₁ has a stronger effect on NO₂ and has a better effect.

In order to explore the reason why Cu/Ag–C₂N₁ can respond selectively to NO₂ but not to the other five gases, the adsorption mechanism was explored by PDOS. The PDOS distribution analysis showed that the more overlapping peaks in different orbitals, the stronger the hybridization between the orbitals, as shown in Fig. S19 and S20.† The conspicuous overlap observed between the metal (Ag and Cu) d-orbitals and N p-orbitals substantiates strong chemical interactions between NO₂ and Ag–C₂N₁, as well as Cu–C₂N₁. Conversely, for the remaining five gases (C₂H₄, C₂H₆, PF₅, H₂O, and NH₃), which exhibit weak adsorption on Ag–C₂N₁ and Cu–C₂N₁, their PDOS plots indicate limited interactions. In addition, the interaction between NO₂ and Ag–C₂N₁ is weak, evident from the small overlapping region between Ag-d and N-p orbitals. However, during the adsorption of NO₂, the hybridization between the p orbital of the N atom and the d orbital of the Cu atom is significantly enhanced, resulting in a significant increase of the PDOS of the d orbital of the Cu atom at the Fermi level, which ultimately affects the overall electrical response of the material. In summary, compared with Ag–C₂N₁, Cu–C₂N₁ has the most obvious response to NO₂ selective electrical signals and has the best effect.

Notably, the conductivity of H₂O adsorbed on the Cu/Ag–C₂N₁ monolayer changes very little, as depicted in Fig. 5, and no discernible changes were observed. In Fig. S19 and S20,† the overlap between the metal (Ag and Cu) d orbital and the O p orbital is still small. This observation supports the conclusion that NO₂ can be selectively detected by Ag–C₂N₁ and Cu–C₂N₁, affirming that humidity does not impede the selective electrical signal response of the doped monolayer to NO₂ in multifarious gas environments.

2.3 Applications of gas sensors

In this section, firstly, based on the detailed discussion of high sensitivity and selectivity to NO₂, the gas sensor is shown in actual usage scenarios, while also showcasing the high humidity resistance of the Cu–C₂N₁ monolayer. Subsequently, as a response to the various hazardous gases generated after the battery becomes thermally unstable, it is imperative to conduct a thorough assessment of the operational environment to determine the temperature tolerance of high-performance sensors. Furthermore, the gas diffusion performance of gas-sensitive materials, which determines their responsiveness, is evaluated comprehensively through molecular dynamics simulations to assess the diffusion behavior of the six gases. Finally, the desorption time of the detected gases from the gas-sensitive material is analyzed.

In Fig. 6a, the practicality of Cu–C₂N₁ and Ag–C₂N₁ monolayer is illustrated when exposed to a range of gases emitted due to battery instability. This exposure highlights their distinctive capacity for selectively adsorbing NO₂, notably emphasized within Cu–C₂N₁. Both monolayers exhibit commendable hydrophobic properties. Additionally, considering the impact of temperature on the material's performance, it is essential to analyze the Gibbs free energy of the material to assess the practical temperature range for its use. The detailed calculation process can be referred to the ESI.†

Fig. 6 (a) Schematic diagram of the high selectivity and hydrophobicity of Cu/Ag–C₂N₁. The Gibbs free energy of the six gases on (b) Ag–C₂N₁ and (c) Cu–C₂N₁ as a function of temperature. Gas diffusion and mean-squared displacement on (d) Ag–C₂N₁ and (e) Cu–C₂N₁ at different temperatures.

The more negative the free energy, the more likely a spontaneous adsorption occurs. This study calculates Gibbs free energy at various temperatures to identify intervals where the free energy is below zero. Therefore, the actual applicable temperature range of Ag–C₂N₁ and Cu–C₂N₁ is determined. Fig. 6b and c show the linear relationship between the Gibbs free energy and temperature of the six gases adsorbed on Cu/Ag–C₂N₁ at different temperatures. In Fig. 6b, ΔG < 0 is observed at a minimum temperature of 500 K, indicating that Ag–C₂N₁ cannot adsorb certain gases at temperatures higher than 500 K. Meanwhile, Fig. 6c displays ΔG < 0 at a minimum temperature of 600 K. Consequently, the high-temperature resistance of Cu–C₂N₁ to the adsorption of the six gases surpasses that of Ag–C₂N₁. Therefore, the application scope of Cu–C₂N₁ is broader. Selective detection of NO₂ on Cu–C₂N₁ in high-temperature environments is superior and applicable over a wider temperature range compared to Ag–C₂N₁. The noteworthy aspect is the maximum applicable temperature of 500 K on Ag–C₂N₁ and 600 K on Cu–C₂N₁, which are the upper limits for H₂O adsorption. The results of DOS analysis confirmed that humidity has no impact on the selective adsorption performance. Consequently, the actual response temperature upper limit of Ag–C₂N₁ is 700 K (determined by Ag–C₂N₁–PF₅), while for Cu–C₂N₁, it's 800 K (determined by Cu–C₂N₁–C₂H₆). The Cu–C₂N₁ monolayer's applicable temperature surpasses the hot mixture's maximum temperature released after battery damage (470 to 720 K) and remains minimally affected by humidity, making it suitable for a wide range of environments.

For the gas diffusion performance of Cu/Ag–C₂N₁, molecular dynamics simulations were used to calculate the diffusion coefficient. The detailed calculation can be found in the ESI.† The diffusion coefficient plays a critical role as it quantifies the gas's diffusion capacity. A larger diffusion coefficient corresponds to a faster diffusion rate, indicating a lower energy barrier required for diffusion.^62,63 In order to explore the diffusion behavior of gases on the monolayer, the diffusion coefficients of the six gas molecules on the monolayer were calculated. The initial and the final stable configurations are shown in Fig. S21 and S22.† Cu–C₂N₁–NO₂ has the smallest diffusion coefficient as shown in Table S2,† which further proved that the monolayer of Cu–C₂N₁ can achieve selective adsorption of NO₂ gas. Interestingly, the diffusion coefficient of water is the largest, indicating that the monolayer does not attract much water.

NO₂ exhibits the lowest diffusion rate on the monolayer. To comprehend its underlying mechanism from a physicochemical perspective, NO₂ diffusion activation energy was obtained by fitting the Arrhenius equation according to the diffusion coefficient. The detailed calculation can be referred to the ESI.† The stable structures of Cu–C₂N₁–NO₂ at different temperatures are shown in Fig. S23 and S24.† The level of activation energy directly affects both the difficulty and the rate of adsorption in the process.^64,65 The smaller the activation energy, the less energy is required, making the adsorption more likely to occur. As shown in Fig. 6d and e, the mean-squared displacement of NO₂ on Ag/Cu–C₂N₁ in the effective applicable temperature range shows that the diffusion activation energy of Cu–C₂N₁ (2.5 kJ mol⁻¹) is less than that of Ag–C₂N₁ (2.7 kJ mol⁻¹). Thus, it is confirmed that Cu–C₂N₁ has more advantages than Ag–C₂N₁ in the field of NO₂ detection.

In the practical sensor usage, assessing gas molecule recovery time performance is crucial. The recovery time (τ) is defined as the time taken by a sensor to return to its original state, when the gas is removed. According to transition state theory, τ is calculated as:¹⁶

τ = ν⁻¹ e^−E_ads/kT (4)

where ν is the attempt frequency (determined to be 10¹² s⁻¹), k represents the Boltzmann constant, and T is the temperature. Tables S4 and S5† detail the τ of the six gases (C₂H₄, C₂H₆, NH₃, NO₂, H₂O, and PF₅) adsorbed on the monolayers at varied temperatures. Overall, as temperatures rise, the τ decreases, indicating accelerated gas desorption. The Cu–C₂N₁ monolayer exhibits longer desorption times compared to Ag–C₂N₁, with NO₂ displaying prolonged τ due to its robust adsorption energy, aligning with experimental observations.²⁶ Notably, in scenarios of battery damage and high-temperature gas release, the Cu–C₂N₁ monolayer continues to exhibit high selectivity to NO₂, demonstrating its advanced nature.

3 Conclusions

In summary, the gas-sensitive behavior on Cu/Ag–C₂N₁ was clarified by dispersion-corrected DFT calculations. The results show that Cu and Ag can be anchored to the monolayer of the C₂N₁ material. After doping metal atoms, the cohesion energy was compared to illustrate the monolayer stability. By DOS analysis, Cu decorated C₂N₁ improved conductivity better than the Ag atom. Moreover, NO₂ prefers to adsorb on Cu–C₂N₁ due to its obvious electronic overlap. The adsorption energy of NO₂ for Cu–C₂N₁ is −3.54 eV, and that of NO₂ for Ag–C₂N₁ is −3.27 eV. The adsorption of NO₂ by Cu–C₂N₁ is significant as also demonstrated by charge transfer, bond length and COHP. Thus, Cu–C₂N₁ has a stronger affinity for NO₂. Notably, we clearly show for the first time in the DOS analysis that Cu–C₂N₁ achieves a single electrical response to NO₂ while eliminating other gaseous disturbances. Furthermore, we also use d-band-center theory to understand the underlying mechanism of gas adsorption on the Cu/Ag–C₂N₁ monolayer. In terms of practicality, the temperature application range of the modified Cu–C₂N₁ (<800 K) is higher than that of Ag–C₂N₁ (<700 K). Interestingly, the adsorption and response of H₂O were extremely poor, demonstrating the minimal effect of humidity on the monolayer. Molecular dynamics simulations revealed that the fundamental reason for the difference in the diffusion performance of NO₂ on the monolayers is that Cu–C₂N₁ has a lower diffusion activation energy barrier (2.5 kJ mol⁻¹). When considering cost, Cu proves to be considerably more economical than Ag. Our research confirms that Cu–C₂N₁ surpasses the Ag–C₂N₁ in both adsorption and response to NO₂ gas. Thus, the Cu–C₂N₁ monolayer can be used as a gas sensor for gas generated by degassed products, greatly increasing the chances of successfully developing a new prototype of safety monitoring.

Data availability

Data will be made available on request.

Author contributions

Mingyang Gu: data curation, formal analysis, investigation, methodology, writing – original draft, and writing – review & editing. Lin Tao: validation, data curation, methodology, supervision, funding acquisition, conceptualization, validation, and writing – review & editing. Davoud Dastan: methodology, supervision, and resources. Jie Dang: methodology, formal analysis, and software. Timing Fang: methodology, formal analysis, and software. Baigang An: supervision, formal analysis, and writing – review & editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The funding from the National Natural Science Foundation of China (Grant No. 52304330), the University of Science and Technology Liaoning Talent Project Grants (6003000317), the Outstanding Youth Fund of University of Science and Technology Liaoning (2023YQ11), and the Youth Fund of the Education Department of Liaoning Province (LJKQZ20222324) are gratefully acknowledged.

References

1. J. B. Goodenough and K. S. Park, The Li-ion rechargeable battery: a perspective, J. Am. Chem. Soc., 2013, 135, 1167–1176.
2. A. Manthiram, An Outlook on Lithium Ion Battery Technology, ACS Cent. Sci., 2017, 3, 1063–1069.
3. M. Armand and J. M. Tarascon, Building better batteries, Nature, 2008, 451, 652–657.
4. A. M. Bates, Y. Preger, L. Torres-Castro, K. L. Harrison, S. J. Harris and J. Hewson, Are solid-state batteries safer than lithium-ion batteries?, Joule, 2022, 6, 742–755.
5. Y. Chen, Y. Kang, Y. Zhao, L. Wang, J. Liu, Y. Li, Z. Liang, X. He, X. Li, N. Tavajohi and B. Li, A review of lithium-ion battery safety concerns: The issues, strategies, and testing standards, J. Energy Chem., 2021, 59, 83–99.
6. D. Santos-Carballal, O. Lupan, N. Magariu, N. Ababii, H. Krüger, M. T. Bodduluri, N. H. de Leeuw, S. Hansen and R. Adelung, Al₂O₃/ZnO composite-based sensors for battery safety applications: An experimental and theoretical investigation, Nano Energy, 2023, 109, 108301.
7. O. Lupan, N. Magariu, D. Santos-Carballal, N. Ababii, J. Offermann, P. Pooker, S. Hansen, L. Siebert, N. H. de Leeuw and R. Adelung, Development of 2-in-1 Sensors for the Safety Assessment of Lithium-Ion Batteries via Early Detection of Vapors Produced by Electrolyte Solvents, ACS Appl. Mater. Interfaces, 2023, 15, 27340–27356.
8. P. Liu, L. Yang, B. Xiao, H. Wang, L. Li, S. Ye, Y. Li, X. Ren, X. Ouyang, J. Hu, F. Pan, Q. Zhang and J. Liu, Revealing Lithium Battery Gas Generation for Safer Practical Applications, Adv. Funct. Mater., 2022, 32, 2208586.
9. H. Cui, C. Yan, P. Jia and W. Cao, Adsorption and sensing behaviors of SF₆ decomposed species on Ni-doped C₃N monolayer: A first-principles study, Appl. Surf. Sci., 2020, 512, 145759.
10. D. Chen, X. Zhang, H. Xiong, Y. Li, J. Tang, S. Xiao and D. Zhang, A First-Principles Study of the SF₆ Decomposed Products Adsorbed Over Defective WS₂ Monolayer as Promising Gas Sensing Device, IEEE Trans. Device Mater. Reliab., 2019, 19, 473–483.
11. S. Zhai, X. Jiang, D. Wu, L. Chen, Y. Su, H. Cui and F. Wu, Single Rh atom decorated pristine and S-defected PdS₂ monolayer for sensing thermal runaway gases in a lithium-ion battery: A first-principles study, Surf. Interfaces, 2023, 37, 102735.
12. K. Boonpalit, J. Kinchagawat, C. Prommin, S. Nutanong and S. Namuangruk, Efficient exploration of transition-metal decorated MXene for carbon monoxide sensing using integrated active learning and density functional theory, Phys. Chem. Chem. Phys., 2023, 25, 28657–28668.
13. B. Mondal, X. Zhang, S. Kumar, F. Long, N. K. Katiyar, M. Kumar, S. Goel and K. Biswas, A resistance-driven H(2) gas sensor: high-entropy alloy nanoparticles decorated 2D MoS(2), Nanoscale, 2023, 15, 17097–17104.
14. P. Li, R. Zhou, B. Liu, Y. Yuan, H. Cui, Z.-H. Pu and T. Wu, Adsorption of a MoSe₂-Based Sensor for Fluorocarbon Gas Decomposition Products in Gas-Insulated Switchgear: A First Principles Analysis, J. Phys. Chem. C, 2023, 127, 11176–11185.
15. H. Cui, T. Yang, X. Peng and G. Zhang, First-principles screening upon Janus PtXY (X, Y = S, Se and Te) monolayer under applied biaxial strains and electric fields, J. Mater. Res. Technol., 2022, 18, 1218–1229.
16. D. Chen, X. Zhang, J. Tang, Z. Cui and H. Cui, Pristine and Cu decorated hexagonal InN monolayer, a promising candidate to detect and scavenge SF6 decompositions based on first-principle study, J. Hazard. Mater., 2019, 363, 346–357.
17. C. Anichini, W. Czepa, D. Pakulski, A. Aliprandi, A. Ciesielski and P. Samori, Chemical sensing with 2D materials, Chem. Soc. Rev., 2018, 47, 4860–4908.
18. X. Sun, Q. Yang, R. Meng, C. Tan, Q. Liang, J. Jiang, H. Ye and X. Chen, Adsorption of gas molecules on graphene-like InN monolayer: A first-principle study, Appl. Surf. Sci., 2017, 404, 291–299.
19. L. Tao, D. Dastan, W. Wang, P. Poldorn, X. Meng, M. Wu, H. Zhao, H. Zhang, L. Li and B. An, Metal-Decorated InN Monolayer Senses N₂ against CO₂, ACS Appl. Mater. Interfaces, 2023, 15, 12534–12544.
20. H. Cui, X. Zhang, Y. Li, D. Chen and Y. Zhang, First-principles insight into Ni-doped InN monolayer as a noxious gases scavenger, Appl. Surf. Sci., 2019, 494, 859–866.
21. H. Ye, S. Liu, K. Peng, Q. Zheng, J. Hu and H. Cui, Exploration of Pt-doped Janus WSSe monolayer as a typical gas sensor for condition assessment in XLPE cables, Chem. Phys. Lett., 2023, 832, 140880.
22. B. Pecz, G. Nicotra, F. Giannazzo, R. Yakimova, A. Koos and A. Kakanakova-Georgieva, Indium Nitride at the 2D Limit, Adv. Mater., 2021, 33, 2006660.
23. P. Hess, Bonding, structure, and mechanical stability of 2D materials: the predictive power of the periodic table, Nanoscale Horiz., 2021, 6, 856–892.
24. T. Zhang, L. Li, T. Huang, H. Wan, W.-Y. Chen, Z.-X. Yang, G.-F. Huang, W. Hu and W.-Q. Huang, Correlation between spin state and activity for hydrogen evolution of PtN₂ monolayer, Appl. Phys. Lett., 2024, 124, 063903.
25. W.-Y. Chen, L. Li, T. Huang, Z.-X. Yang, T. Zhang, G.-F. Huang, W. Hu and W.-Q. Huang, Extending Schottky–Mott rule to van der Waals heterostructures of 2D Janus materials: Influence of intrinsic dipoles, Appl. Phys. Lett., 2023, 123, 171601.
26. Y. Liu, L. Gao, S. Fu, S. Cheng, N. Gao and H. Li, Highly efficient VOC gas sensors based on Li-doped diamane, Appl. Surf. Sci., 2023, 611, 155694.
27. X. Tang, G. Guo, Y. Peng, M. You, S. Luo, C. He, C. Tang, Z. Huang and J. Zhong, Tunable properties of two-dimensional bilayer C₃N as anode material: Bandgap, binding energy, and diffusion barrier, J. Energy Storage, 2024, 77, 109906.
28. X. Wang, H. Niu, X. Wan, A. Wang, F. R. Wang and Y. Guo, Impact of Coordination Environment on Single-Atom-Embedded C₃N for Oxygen Electrocatalysis, ACS Sustain. Chem. Eng., 2022, 10, 7692–7701.
29. Z. Chen, J. Zhao, C. R. Cabrera and Z. Chen, Computational Screening of Efficient Single-Atom Catalysts Based on Graphitic Carbon Nitride (g-C₃N₄) for Nitrogen Electroreduction, Small Methods, 2018, 3, 1800368.
30. P. Kaur, S. Bagchi, D. Gribble, V. G. Pol and A. P. Bhondekar, Impedimetric Chemosensing of Volatile Organic Compounds Released from Li-Ion Batteries, ACS Sens., 2022, 7, 674–683.
31. P. Kaur, S. Bagchi, V. G. Pol and A. P. Bhondekar, Early Detection of Mixed Volatile Organic Compounds to Circumvent Calamitous Li-Ion Battery Thermal Runaway, J. Phys. Chem. C, 2023, 127, 8373–8382.
32. M. Wu, S. Han, S. Liu, J. Zhao and W. Xie, Fire-safe polymer electrolyte strategies for lithium batteries, Energy Storage Mater., 2024, 66, 103174.
33. H. Xiong, H. Zhang and L. Gan, A new bifunctional C₃N nanosheet of NO₂, SO₂ gas sensor and CO₂ separation: A first-principles study, Phys. E, 2021, 126, 114463.
34. H. Cui, G. Zhang, X. Zhang and J. Tang, Rh-doped MoSe₂ as a toxic gas scavenger: a first-principles study, Nanoscale Adv., 2019, 1, 772–780.
35. L. Li, H. Cao, Z. Liang, Y. Cheng, T. Yin, Z. Liu, S. Yan, S. Jia, L. Li, J. Wang and Y. Gao, First-Principles Study of Ti-Deficient Ti₃C₂ MXene Nanosheets as NH₃ Gas Sensors, ACS Appl. Nano Mater., 2022, 5, 2470–2475.
36. D. Dastan, K. shan, A. Jafari, T. Marszalek, M. K. A. Mohammed, L. Tao, Z. Shi, Y. Chen, X.-T. Yin, N. D. Alharbi, F. Gity, S. Asgary, M. Hatamvand and L. Ansari, Influence of heat treatment on H₂S gas sensing features of NiO thin films deposited via thermal evaporation technique, Mater. Sci. Semicond. Process., 2023, 154, 107232.
37. P. Lv, D. Wu, B. He, X. Li, R. Zhu, G. Tang, Z. Lu, D. Ma and Y. Jia, An efficient screening strategy towards multifunctional catalysts for the simultaneous electroreduction of NO₃−, NO₂− and NO to NH₃, J. Mater. Chem. A, 2022, 10, 9707–9716.
38. M. Tang, D. Zhang, D. Wang, J. Deng, D. Kong and H. Zhang, Performance prediction of 2D vertically stacked MoS₂-WS₂ heterostructures base on first-principles theory and Pearson correlation coefficient, Appl. Surf. Sci., 2022, 596, 153498.
39. D. Chen, Y. Li, S. Xiao, C. Yang, J. Zhou and B. Xiao, Single Ni atom doped WS₂ monolayer as sensing substrate for dissolved gases in transformer oil: A first-principles study, Appl. Surf. Sci., 2022, 579, 152141.
40. L. T. Ta, I. Hamada, Y. Morikawa and V. A. Dinh, Adsorption of toxic gases on borophene: surface deformation links to chemisorptions, RSC Adv., 2021, 11, 18279–18287.
41. W. Nong, H. Liang, S. Qin, Y. Li and C. Wang, Computational Design of Two-Dimensional Boron-Containing Compounds as Efficient Metal-free Electrocatalysts toward Nitrogen Reduction Independent of Heteroatom Doping, ACS Appl. Mater. Interfaces, 2020, 12, 50505–50515.
42. M. I. Ahmed, L. J. Arachchige, Z. Su, D. B. Hibbert, C. Sun and C. Zhao, Nitrogenase-Inspired Atomically Dispersed Fe–S–C Linkages for Improved Electrochemical Reduction of Dinitrogen to Ammonia, ACS Catal., 2022, 12, 1443–1451.
43. Q. Yue, Z. Shao, S. Chang and J. Li, Adsorption of gas molecules on monolayer MoS₂ and effect of applied electric field, Nanoscale Res. Lett., 2013, 8, 425.
44. H. Wang, X. Li, J. Wu and D. Zhang, An Experimental and Density Functional Theory Simulation Study of NO Reduction Mechanisms over Fe(0) Supported on Graphene with and without CO, Langmuir, 2023, 39, 15369–15379.
45. L. Tao, J. Huang, D. Dastan, T. Wang, J. Li, X. Yin and Q. Wang, CO₂ capture and separation on charge-modulated calcite, Appl. Surf. Sci., 2020, 530, 147265.
46. I. S. Amiinu, X. Liu, Z. Pu, W. Li, Q. Li, J. Zhang, H. Tang, H. Zhang and S. Mu, From 3D ZIF Nanocrystals to Co-N_x/C Nanorod Array Electrocatalysts for ORR, OER, and Zn-Air Batteries, Adv. Funct. Mater., 2018, 28, 1704638.
47. M. Wang, W. Ma, C. Tan, Z. Qiu, L. Hu, X. Lv, Q. Li and J. Dang, Designing Efficient Non-Precious Metal Electrocatalysts for High-Performance Hydrogen Production: A Comprehensive Evaluation Strategy, Small, 2023, e2306631.
48. Z. Lv, W. Ma, M. Wang, J. Dang, K. Jian, D. Liu and D. Huang, Co-Constructing Interfaces of Multiheterostructure on MXene (Ti₃C₂T_x)-Modified 3D Self-Supporting Electrode for Ultraefficient Electrocatalytic HER in Alkaline Media, Adv. Funct. Mater., 2021, 31, 2102576.
49. K. He, W. Li, L. Tang, L. Chen, G. Wang, Q. Liu, X. Xin, C. Yang, Z. Wang, S. Lv and D. Xing, Insight into the design of a Ti₃C₂ MXene/Ti₄O₇ composite ceramic membrane boosts the electrocatalytic activity for 1,4-dioxane electro-oxidation, Appl. Catal., B, 2023, 338, 123077.
50. S. Nie, L. Tao, J. Li, W. Wang, P. Poldorn, Y. He, X. Yin and M. Wu, A single response to reducing gases by NiO-TiO₂ heterojunction nanocrystals, Appl. Surf. Sci., 2024, 644, 158821.
51. S. Nie, J. Li, L. Tao, Y. He, D. Dastan, X. Meng, P. Poldorn and X. Yin, Insights into Selective Mechanism of NiO-TiO(2) Heterojunction to H(2) and CO, ACS Sens., 2023, 8, 4121–4131.
52. L. Tao, J. Huang, D. Dastan, T. Wang, J. Li, X. Yin and Q. Wang, New insight into absorption characteristics of CO₂ on the surface of calcite, dolomite, and magnesite, Appl. Surf. Sci., 2021, 540, 148320.
53. Y. He, L. Tao, J. Li, M. Wu, P. Poldorn, D. Dastan, S. Abbasi, S. Nie, X. Yin and Q. Wang, Atomic-level insights into selective adsorption of H₂ and CO on SnO₂/CoO heterojunctions, Mater. Today Nano, 2023, 22, 100334.
54. Y. He, J. Li, L. Tao, S. Nie, T. Fang, X. Yin and Q. Wang, First-principles calculations on the resistance and electronic properties of H₂ adsorption on a CoO-SnO₂ heterojunction surface, Phys. Chem. Chem. Phys., 2021, 24, 392–402.
55. M. Wang, L. Kong, X. Lu and C.-M. Lawrence Wu, First-row transition metal embedded pyrazine-based graphynes as high-performance single atom catalysts for the CO₂ reduction reaction, J. Mater. Chem. A, 2022, 10, 9048–9058.
56. J. Li, C. Hou, C. Chen, W. Ma, Q. Li, L. Hu, X. Lv and J. Dang, Collaborative Interface Optimization Strategy Guided Ultrafine RuCo and MXene Heterostructure Electrocatalysts for Efficient Overall Water Splitting, ACS Nano, 2023, 17, 10947–10957.
57. L. Tao, Z. Li, G.-C. Wang, B.-Y. Cui, X.-T. Yin and Q. Wang, Evolution of calcite surface reconstruction and interface adsorption of calcite-CO₂ with temperature, Mater. Res. Express, 2018, 6, 025035.
58. H. Zhang, R. Zhang, Y. Ni, M. Chen, C. Sun and F. Dong, SO₂ adsorption and conversion on pristine and defected calcite {1 0 4} surface: A density functional theory study, Appl. Surf. Sci., 2022, 596, 153575.
59. L. Lin, S. Yao, R. Gao, X. Liang, Q. Yu, Y. Deng, J. Liu, M. Peng, Z. Jiang, S. Li, Y. W. Li, X. D. Wen, W. Zhou and D. Ma, A highly CO-tolerant atomically dispersed Pt catalyst for chemoselective hydrogenation, Nat. Nanotechnol., 2019, 14, 354–361.
60. X. Liu, Y. Jiao, Y. Zheng, M. Jaroniec and S.-Z. Qiao, Building Up a Picture of the Electrocatalytic Nitrogen Reduction Activity of Transition Metal Single-Atom Catalysts, J. Am. Chem. Soc., 2019, 141, 9664–9672.
61. H. Hu, P. Zhang, B.-B. Xiao and J.-L. Mi, Theoretical study of p-block metal–nitrogen–carbon single-atom catalysts for the oxygen reduction reaction, Catal. Sci. Technol., 2022, 12, 6751–6760.
62. L. Tao, J. Huang, X. Yin, Q. Wang, Z. Li, G. Wang and B. Cui, Adsorption Kinetics of CO₂ on a Reconstructed Calcite Surface: An Experiment-Simulation Collaborative Method, Energy Fuels, 2019, 33, 8946–8953.
63. L. Tao, J. Huang, D. Dastan, J. Li, X. Yin and Q. Wang, Flue gas separation at organic-inorganic interface under geological conditions, Surf. Interfaces, 2021, 27, 101462.
64. T. Fang, C. Wei, X. Meng, G. Zhou and X. Liu, How homogeneous and biphasic membranes contribute to the gas transfer: A molecular dynamics simulation study, Int. J. Heat Mass Transfer, 2023, 201, 123644.
65. X. Meng, T. Fang, G. Zhou, S. Liu and X. Liu, Selectivity and permeability of gas separation in SILMs: Effect of collapsed structure, J. Mol. Liq., 2023, 388, 122834.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta00645c

---