English

• Volatile organic compounds (VOCs) are known as dominate precursors to urban haze and photochemical smog, mainly from processes such as coal chemical, petrochemical and fuel paint manufacturing and use [1,2]. Most VOCs are toxic, irritating, teratogenic, and carcinogenic, especially formaldehyde, vinyl chloride, and benzene, which seriously threaten the ecological environment and human health [3–5]. Studies have shown that regular exposure to formaldehyde increases the chance of developing cancers, such as blood and lymphoma [6]. Vinyl chloride enters the body and suppresses the central nervous system, damages the liver and spleen, and produces diseases such as hepatic angiosarcoma [7]. The intake of benzene paralyzes the central nervous system and affects the body's hematopoietic function, leading to aplastic anemia and leukemia in severe cases [8]. Therefore, it is urgent to eliminate VOC pollution. Adsorption technology has gained numerous attentions as an economical and environmentally friendly method in VOC reduction [1].

  Because of the interesting physical and chemical performance of 2D metal oxides, they have recently been widely explored in transistors, inverters, photodetectors, memristors, sensors, etc. [9,10]. The developments of advanced synthesis methods such as mechanical exfoliation, vapor deposition, and solid solution methods have provided new technical supports for preparing novel layered 2D metal oxides [11,12]. Huang et al. [13] proposed that ultrathin 2D Al₂O₃ nanosheets designed by replicating graphene oxide from the bottom up exhibited excellent adsorption capacity for fluoride ions. In addition, HfO₂ [14], In₂O₃ [15], ZnO [16], SnO₂ [17], γ-Bi₂O₃ [18], etc., are also attracting attention as new 2D metal oxides.

  Al₂O₃ is widely used in ultrafiltration membranes, catalysts, adsorbents, etc., due to its good electronic properties, mechanical properties and low cost [19–22]. Eunmi et al. [23] successfully synthesized porous Al₂O₃ with controlled mesoporous and macroporous characteristics using the solvent defect method to achieve efficient removal of the organic pollutant Congo red dye from water. Kumari et al. [24] used nitric-acid-activated Al₂O₃ to increase the efficiency of defluoridation in the industrial wastewater from 74.18% to 97.43%, and the adsorbent exhibited better regeneration and recyclability, which is beneficial for practical industrial applications. Recently, Liu et al. [25] found that CO₂ can be effectively adsorbed and decomposed on the anoxic γ-Al₂O₃ (100) surface based on first-principles studies. Bai et al. [26] found good adsorption of both CS₂ and COF₂ on the surface of α-Al₂O₃ (0001) based on DFT studies. γ-Al₂O₃ showed good adsorption capacity for potentially carcinogenic chlorinated volatile organic compounds (Cl-VOC) such as benzyl chloride, chloroform, and tetrachlorine after 8 weeks of adsorption [27]. We assume and expect that the new AlO₂ in this work can possess some potentials and possibilities for the adsorption of VOCs.

  Besides, studying other 2D aluminides has also gained plenty of interests. Wang et al. [28] reported 2D AlN synthesized by metal-organic chemical vapor deposition between Si substrate and graphene with ultra-wide bandgap semiconductor (9.63–9.20 eV) showing great promise in deep-UV optoelectronic applications. The 2D Al₂C is a medium bandgap semiconductor (1.02 eV) that maintains its structural integrity at 1500 K, indicating good thermal stability [29] and it was reported to have large adsorption energy (−1.972 eV ~ −2.244 eV) for four toxic VOCs (acetaldehyde, ethylene oxide, vinyl chloride and benzene) and can be used as an adsorbent and sensing medium for toxic VOCs [30]. Wang et al. [31] theoretically predicted that the novel monolayer AlP₅ is promising for detecting toxic heavy metals (As, Hg and Cd) in the aqueous environment.

  Recently, 2D CaCl and Na₂Cl crystals with abnormal anion and cation ratios generated on graphene films have abnormal physical properties such as metallicity, ferromagnetism and piezoelectric behavior [32,33]. Therefore, the discovery of 2D AlO₂ with abnormal stoichiometry compared with Al₂O₃ is expected to expand the development and application of novel electronic devices. Additionally, AlO₂ molecule and thin film with anomalous stoichiometric ratios have been found experimentally. A new AlO₂ thin film is formed at the interface of diffusion bonding between Pt and α-Al₂O₃ at 1200 ℃ in Ar atmosphere [34]. Lester et al. [35] reacted Al atoms evaporated by pulsed laser with O₂ from a condensing Ar stream to obtain cyclic AlO₂ molecule. The successful syntheses of AlO₂ molecule and thin film promote more explorations of other new kinds of AlO₂ including 2D materials.

  Herein, a novel 2D AlO₂ with anomalous stoichiometric ratios is designed based on DFT simulations, and its stability is determined by phonon band structure, calculation of elastic constants and ab initio molecular dynamics (AIMD) simulations. The mechanical and electronic performance of monolayer and multilayer AlO₂ are investigated. Interestingly, AlO₂ has a rare in-plane and out-of-plane NPR. In addition, the adsorption performance of AlO₂ on three VOCs (CH₂O, C₂H₃Cl and C₆H₆) is explored. The results show that monolayer AlO₂ has good adsorption and desorption behaviors and is expected to be a potential gas sensor and a good adsorber for detecting and adsorbing VOCs.

  The geometry optimization of pristine AlO₂, the calculation of electronic properties and the simulation of adsorbed VOCs on AlO₂ are implemented in the SIESTA code [36] based on DFT [37]. The exchange-correlation term of the generalized gradient approximation (GGA) [38] is performed under the Perdew-Burke-Ernzerhof (PBE) [39] scheme. In order to be closer to the theoretical limit, the double zeta polarization (DZP) [40] function is chosen for the basis set. A 9 × 12 × 1 sampling is used for Monkhorst-Pack grid point sampling, and the set plane wave cut-off energy is 300 Ry. All atoms during the geometry optimizations are completely relaxed until the maximum force is less than 0.02 eV/Å. Furthermore, the thermal and dynamic stability of AlO₂ are confirmed by ab initio molecular dynamics (AIMD) [41] simulations based on the canonical ensemble (NVT) and phonon spectrum calculation, respectively. Herein, the canonical ensemble (NVT) refers to the simulated system in which the number of particles (N), volume (V), and temperature (T) remain constant.

  To determine the stability of multilayer AlO₂, we define the formation energy (E_f) as:

  (1)

  where n is the number of layers of AlO₂; E(AlO₂) and E(n-AlO₂) denote the total energy of single-layer and n-layer AlO₂, respectively.

  In order to accurately describe the adsorption properties of VOCs on the AlO₂ surface, we define the adsorption energy (E_ads) as:

  (2)

  where E(AlO₂) and E(VOCs) are the energy of the original AlO₂ and independent VOCs gas molecules, respectively, and E(AlO₂-VOCs) is the total energy of the adsorption system of VOCs molecules.

  The differential charge density between AlO₂ and VOCs molecules in the adsorption system is defined as:

  (3)

  where ρ(AlO₂) and ρ(VOCs) denote the charge values of original AlO₂ and independent VOCs gas molecules, respectively, and ρ(AlO₂-VOCs) denotes the total charge value of the adsorption system of VOCs molecules.

  The optimized 2D AlO₂ structure is shown in Figs. 1a and b, which has a P-4m2 space group and belongs to the tetragonal crystal system (a = b = 3.27 Å, c = 15 Å, α = β = γ = 90°), and each cell includes one Al atom and two O atoms (see the light blue area in Fig. 1b). Section S7 (Supporting information) lists the specific position of each atom in the unit cell. The 2D AlO₂ is a sawtooth-like structure with a monolayer height of d = 1.45 Å, as shown in the side view of Fig. 1b. The outer and inner atoms are O and Al ones, respectively. The Al-O bond has a bond length of 1.79 Å, which is close to the theoretical data of the Al-O bond length of α-Al₂O₃ (1.68–1.86 Å) [42] and γ-Al₂O₃ (1.75–1.77 Å) [43]. The bond angle of Al-O-Al is 132.14°. Additionally, AlO₂ has anomalous stoichiometric ratios. The oxidation states may be abnormal. Lu et al. [34] reported that the possible form of AlO₂ is (Al³⁺)₂(O₂)²⁻(O²⁻)₂. Possible oxidation states of AlO₂ may be Al³⁺, O²⁻ and O⁻.

  Figure 1

  

  Figure 1.  (a) Geometrically optimized configuration of 2D AlO₂, (b) top and lateral views. The unit cell is marked with a light blue rectangular area. Four possible adsorption sites (O1, O2, Al1 and I1) are marked. (c) Potential energy fluctuations of AlO₂ in AIMD simulations at 300 and 1000 K, and the inset shows the structure after 5 ps of AIMD simulations at 1000 K. (d) Phonon band structure and phonon DOS of AlO₂. (e) Band structure of AlO₂ and the corresponding DOS.

  DownLoad: Full-Size Img PowerPoint

  The stability of AlO₂ is evaluated by molecular dynamics (MD) simulations, phonon spectra and strain energy calculations. The results of MD simulation for AlO₂ are plotted in Fig. 1c. It is obvious that no significant deformation and bond breaking occur in the structure of AlO₂ heated at 300 K and 1000 K for 5 ps, and its potential energies at 300 K and 1000 K have few changes, indicating that AlO₂ still has good thermal stability at 1000 K. The phonon spectrum of AlO₂ is shown in Fig. 1d, and no imaginary frequencies are observed throughout the Brillouin zone, confirming the dynamic stability of AlO₂. The maximum vibrational frequency of 1080.5 cm⁻¹ is contributed by both Al and O atoms, which facilitates future experimental characterization. To further determine the mechanical stability of AlO₂, the calculated elastic constants of C₁₁ = 48.23 N/m, C₁₂ = −12.32 N/m, C₂₂ = 48.23 N/m and C₄₄ = 1.25 N/m fully satisfy the Born-Huang criterion: C₄₄ > 0 and C₁₁C₂₂-C₁₂² > 0, indicating that AlO₂ is mechanically stable.

  In addition, the conductivity of AlO₂ material is verified by calculating electronic properties. The valence band top of the band structure crosses the Fermi energy level (Fig. 1e), indicating that AlO₂ is metallic. The contribution of O atoms near the Fermi level leads to a non-zero density of states (DOS), further proving the metallicity of AlO₂. The electronic properties of AlO₂ are similar to those of graphene oxide, which has a negative differential resistance (NDR) phenomenon [44], suggesting that AlO₂ is promising for NDR nanodevices.

  The flexibility or rigidity of the material is defined by the in-plane stiffness C, and flexible materials have less in-plane stiffness. The strain energy curves of AlO₂ are shown in Fig. 2a. As shown in Fig. 2b, the in-plane stiffness along the x or y orientation is 48.23 N/m, which is considerably smaller than that of graphene (341 N/m) [45] and MoS₂ (120.1 N/m) [46], but a little larger than those of g-C₃N₅ (37 N/m) [47] and g-ZnO (47.8 N/m) [48]. It is noted that when θ equals 45°, 135°, 225° and 315°, the corresponding in-plane stiffness is only 1.08 N/m. Such small in-plane stiffness along the diagonal direction indicates more flexibility and more softness than those along the x and y axes, which is attributed to its pleated structure. Poisson's ratio ν is defined as the ratio of lateral versus vertical strain of the material, and reflects the elastic constant of the lateral distortion of the material. As illustrated in Fig. 2c, AlO₂ has a large ν of 0.97 along the diagonal direction. However, a rare negative Poisson's ratio (NPR) of −0.26 appears in the x and y orientations. Such NPR phenomenon mainly originates from its characteristic sawtooth configuration in the x and y directions, similar phenomena have been reported in structures such as black phosphorus [49], BP₅ [50] and penta-graphene [51]. To further compare the NPR properties of AlO₂, the Poisson's ratios of different two-dimensional materials with NPR properties predicted by recent theories are listed in Fig. 2d. The results show that the NPR performance of AlO₂ materials is superior compared to the majority of 2D materials such as borophane (−0.053) [52], penta-graphene (−0.068) [51], phosphorene (−0.027) [53], SiS (−0.19) [54], Ag₂S (−0.12) [55] and SnSe (−0.17) [56]. Due to the special mechanical properties of NPR materials, they have broad application prospects in aerospace, medical equipment, sensors, protective equipment and other fields [57].

  Figure 2

  

  Figure 2.  (a) AlO₂ strain energy under various strains. (b) In-plane stiffness C and (c) Poisson's ratio ν of AlO₂. (d) Comparison of in-plane Poisson's ratios of some 2D materials. The schematic images of in-plane NPR phenomenon caused by the angle ∠OAlO of AlO₂ under transverse (e) stretching and (f) compression.

  DownLoad: Full-Size Img PowerPoint

  In order to explore the origination of the in-plane NPR of the single-layer AlO₂ structure, we investigate the changes of the angle ∠OAlO under x-axis strain. As shown in Fig. S1, ∠OAlO increases by 0.88% and decreases by 1.27% under tensile and compressive strains, respectively. Therefore, the in-plane NPR of AlO₂ mainly comes from the contribution of the angle ∠OAlO, as schematically shown in Figs. 2e and f.

  Since it is much easier to experimentally synthesize multilayer structures, the formation energy, DOS, C, ν and electronic properties of 1–5 layers of AlO₂ are considered. The structures of 1–5 layers of AlO₂ after geometry optimizations are shown in Fig. 3a, and the total layer thickness from 1.45 Å to 17.46 Å as the number of layers increases. In order to study the stability of multilayer AlO₂, the calculation of formation energy is performed according to Eq. 1. Fig. 3b shows the variation of formation energy of 2–5 layers of AlO₂ nanosheets, with the increase of the number of layers, the formation energy is more negative, which indicates the increased stability of the multilayer structure.

  Figure 3

  

  Figure 3.  (a) Side view of the geometry of 1–5 layers AlO₂ with the intra-layer heights marked. (b) Forming energy of 2–5 layers of AlO₂. (c) The DOS of 1–5 layers AlO₂. (d) In-plane stiffness C and (e) Poisson's ratio ν of 1–5 layers AlO₂. The inset shows an enlarged view of the NPR in (e).

  DownLoad: Full-Size Img PowerPoint

  The total DOS of 1–5 layers of AlO₂ is shown in Fig. 3c. The DOS of all multilayer AlO₂ crosses the Fermi level, confirming their conducting behavior, which is consistent with the band structures (Fig. S3 in Supporting information). As the number of layers increases, the DOS also increases, which is consistent with the increase in the number of energy bands. In addition, the DOS maps of 1–5 layers of AlO₂ are similar, with all DOS peaks locate at nearly the same energy, suggesting that the multilayer AlO₂ structures relies mainly on van der Waals interactions between neighboring layers.

  The C and ν as a function of angle for 1–5 layers of AlO₂ are shown in Figs. 3d and e. The in-plane stiffness and positive Poisson's ratio decrease as the layers increase. The NPR performance of the monolayer structure is still maintained for multilayer AlO₂, however the NPR values along the x and y axes increase from −0.26 to −0.08 as the number of layers increases (Table S1 in Supporting information). Therefore, the tuning of mechanical properties such as in-plane stiffness and Poisson's ratio of AlO₂ can be achieved by different numbers of layers.

  To explore the out-of-plane NPR of the bilayer structure, we compare the variation of interlayer height d₁ and intra-layer height d₂ in the bilayer AlO₂ structure under different biaxial strains. The typical structures of the bilayer AlO₂ after geometric optimization under the biaxial strains of −5%, 0% and 5% are shown in Figs. 4a–c. Obviously, with the applied compressive and tensile strains, the interlayer height d₁ decreases and increases in the side view of Figs. 4a–c, respectively, revealing that the bilayer AlO₂ has an out-of-plane NPR. Furthermore, the comprehensive variations of d₁ and d₂ with different biaxial strains are demonstrated in Fig. 4d. The results exhibit that by applying −5% compressive (or 5% tensile) strain in the transverse direction, d₁ decreases by 5.5% (or increases by 5.1%); on the other hand, d₂ increases by 8.1% (or reduces by 9.1%). Thus the out-of-plane NPR of bilayer AlO₂ comes mainly from the contribution of the interlayer height d₁. The schematic diagram of the out-of-plane NPR of bilayer AlO₂ is shown in Figs. 4e and f. AlO₂ becomes thicker or thinner vertically when it experiences transverse tensile or compressive strain respectively (Figs. 4e and f), similar to other 2D materials such as black phosphorene [58] and borophane [52] possessing out-of-plane NPR.

  Figure 4

  

  Figure 4.  Top and side views of the bilayer AlO₂ structure under biaxial strain of (a) −5%, (b) 0%, and (c) 5%. d₁ denotes the interlayer height of bilayer AlO₂. d₂ denotes the intra-layer height of single-layer AlO₂. (d) Percentage change of interlayer height d₁ and intra-layer height d₂ with biaxial strain. (e) Longitudinal expansion phenomenon of bilayer AlO₂ in the case of transverse stretching. (f) Longitudinal shrinkage phenomenon of bilayer AlO₂ under lateral compression.

  DownLoad: Full-Size Img PowerPoint

  The adsorption performance is investigated to explore the potential application of AlO₂ in VOC detection. As shown in Fig. 1b, there are three types of adsorption sites in the AlO₂ structure: atomic tops (A1, O1, and O2), the middle of the bonds (Al-O1 and Al-O2), and hollow (I1). VOC molecules are placed in the positions mentioned above to find a suitable adsorption conformation, and then the geometry is optimized without imposing any constraints. The most negative adsorption energy is used as a criterion to find the most stable adsorption conformation. Figs. 5a–c show the most stable adsorption configurations of three VOCs adsorbed on AlO₂. The adsorption energies of the AlO₂-CH₂O, AlO₂-C₂H₃Cl and AlO₂-C₆H₆ parallel (and vertical) adsorption systems are −0.84 eV (−0.51 eV), −1.19 eV (−0.72 eV) and −1.18 eV (−0.29 eV), respectively. The adsorption energies of the parallel systems are more negative than those of the vertical systems. CH₂O, C₂H₃Cl and C₆H₆ are parallel to the AlO₂ plane with adsorption distances of 1.97 Å, 1.90 Å and 2.19 Å, respectively. Mulliken charge analysis is performed for the most stable adsorption configurations, as shown in Figs. 5d–f, revealing that all VOC molecules act as charge donors and AlO₂ acts as charge acceptors. The charge transfer amount of the adsorption system is defined as: ΔQ = Q(^*VOCs) - Q(VOCs), where Q(^*VOCs) denotes the charge of VOCs gas in the adsorption system, and Q(VOCs) denotes the charge values of independent VOCs gas molecules. The charge loss of CH₂O, C₂H₃Cl and C₆H₆ molecules are −0.36 e, −0.64 e and −0.77 e, respectively.

  Figure 5

  

  Figure 5.  Stable configurations and charge density differences of (a, d) CH₂O, (b, e) C₂H₃Cl and (c, f) C₆H₆ adsorbed on the surface of AlO₂. The red, blue, gray, white and green balls are used to denote O, Al, C, H and Cl atoms. The value of 0.025 e/ų is determined as the isosurface. The amount and direction of charge transferred in the adsorption system are marked. The DOS before and after adsorption of (g) CH₂O, (h) C₂H₃Cl and (i) C₆H₆ by monolayer AlO₂. (j) Comparison of adsorption energies and charge transfers of some 2D materials for the adsorption of VOCs.

  DownLoad: Full-Size Img PowerPoint

  To further compare the adsorption stability of AlO₂, the adsorption energies and charge transfer amounts of VOCs on the surfaces of different 2D materials (MoBOH [59], phospholene [60], MnO₂ [61], Pd-C₃N [62], Ni-HfSSe [63], Al₂C [30] and F-diamane [64]) are presented in Fig. 5j. As a comparison, the adsorption energies of VOCs for the structures provided in previous works (MnO₂ [61] and phospholene [60]) are calculated. The adsorption energies of MnO₂-CH₂O, MnO₂-C₆H₆ and phospholene-CH₂O are −0.52 eV, −0.24 eV and −0.56 eV, respectively. These data are close to those reported works [60,61]. The results show that the two-dimensional AlO₂ material exhibits a better adsorption effect on VOCs and the adsorption energy is better than most 2D materials. The AlO₂ surface has satisfactory VOCS desorption compared to Pd-C₃N and Al₂C, which have higher adsorption energy. It must be noted that too high adsorption energy is not conducive to VOCs desorption and material recovery. The charge transfer is much larger than that of several other 2D materials, confirming the strong interaction of VOCs with AlO₂. The low adsorption distance, high adsorption energy and large charge transfer demonstrate the high efficiency of 2D AlO₂ for removing these VOCs.

  As shown in Figs. 5g–i, DOS analysis is also performed to examine the effect of the adsorption of VOCs on the electronic characteristics of AlO₂. The total DOS before and after the adsorption of VOCs appear to be significantly different. After adsorption of VOCs molecules on AlO₂, VOCs molecules not only have a significant contribution to the valence band but also have a major contribution to the conduction band, which is consistent with the band structures (Fig. S4 in Supporting information). Significant changes in the DOS before and after adsorption of VOCs molecules indicate the presence of strong interactions between AlO₂ and VOCs, which is consistent with the results of large charge transfer amounts.

  The effectiveness of VOCs desorption is mainly determined by the recovery time, which is the time the sensor needed to return to its original state after the gas is removed. Therefore, according to the transition state principle, the recovery time (τ) of VOCs molecules on the AlO₂ surface is evaluated:

  (4)

  where A (10¹² s⁻¹) is the apparent frequency factor, k_B (8.62 × 10⁻⁵ eV/K) is the Boltzmann constant, E_ads (eV) is the adsorption energy which is the energy barrier of the desorption process, and T (K) is the system temperature. As shown in Fig. 6, the desorption times of VOCs molecules on the AlO₂ surface at different temperatures are calculated. CH₂O has a short recovery time (1.69 × 10² s) at room temperature, indicating that the AlO₂ sensor is expected to detect CH₂O at room temperature reversibly. The long desorption time of C₂H₃Cl (1.36 × 10⁸ s) and C₆H₆ (9.93 × 10⁷ s) implies that AlO₂ is suitable as an adsorbent for C₂H₃Cl and C₆H₆ at room temperature. In addition, the desorption time decreases with increasing temperature. For C₂H₃Cl and C₆H₆ molecules, the desorption times at 498 K are 1.11 s and 0.92 s, respectively, indicating that the high temperature accelerates desorption. Therefore, at room temperature, AlO₂ can be used as a gas sensor for the detection of CH₂O and an adsorbent for C₂H₃Cl and C₆H₆; and at higher temperatures, it can also act as a gas sensor for the detection of C₂H₃Cl and C₆H₆.

  Figure 6

  

  Figure 6.  Desorption time of CH₂O, C₂H₃Cl and C₆H₆ molecules on AlO₂ surface at different temperatures.

  DownLoad: Full-Size Img PowerPoint

  For the future applications, the O₂, N₂ and water effects should be considered. Since AlO₂ has the maximum adsorption energy for C₂H₃Cl, we continue to consider the effect of O₂, N₂ and H₂O molecules on the adsorption of C₂H₃Cl. As shown in Figs. 7a–f, the adsorption energies (−0.18 eV, −0.45 eV and −0.37 eV) and the amounts of the charge transfer (−0.11 e, −0.26 e and −0.14 e) of N₂, O₂ and H₂O on the AlO₂ surface are much smaller than those of C₂H₃Cl on AlO₂. Figs. 7g–l display the geometry, charge difference density distribution and DOS of C₂H₃Cl adsorbed on the AlO₂ surface in O₂, N₂ and water environment, respectively. As shown in Fig. 7m, the adsorption energies (−1.06 eV, −1.01 eV and −1.04 eV) and the amounts of charge transfer (−0.56 e, −0.57 e and −0.57 e) of C₂H₃Cl on AlO₂ in O₂, N₂ and water environment are slightly lower than those in the vacuum environment. Therefore, the effect of the presence of O₂, N₂ and H₂O on the adsorption of C₂H₃Cl on the AlO₂ surface is insignificant.

  Figure 7

  

  Figure 7.  Stable configurations, charge density differences and DOS of (a, d) N₂, (b, e) O₂, (c, f) H₂O, (g, j) N₂-C₂H₃Cl, (h, k) O₂-C₂H₃Cl and (i, l) H₂O-C₂H₃Cl adsorbed on the surface of AlO₂. (m) Comparison of adsorption energies and charge transfers of monolayer AlO₂ for the adsorption of gas molecular and C₂H₃Cl (in O₂, N₂ and water environment).

  DownLoad: Full-Size Img PowerPoint

  In summary, we propose a novel 2D AlO₂ with anomalous stoichiometric ratios. The phonon spectra, AIMD simulations and elastic constants confirm its stability. Electronic calculations show that AlO₂ is a conductor. Due to its special structure of serrated configuration, AlO₂ offers low in-plane stiffness and a large negative Poisson's ratio (NPR). The Poisson's ratio of 1–5 layers of AlO₂ is negative, and the absolute value of NPR decreases as the number of layers increases. The bilayer AlO₂ exhibits a rare NPR in both in-plane and out-of-plane directions and reveals the underlying mechanism. The potential applications of AlO₂ in the adsorption of CH₂O, C₂H₃Cl and C₆H₆ detection are explored. The low adsorption distance (1.90 Å to 2.19 Å), high adsorption energy (−0.84 eV to −1.19 eV) and large charge transfer (−0.36 e to −0.77 e) indicate the high removal efficiency of 2D AlO₂ for these VOCs. The DOS of AlO₂ changed significantly before and after the adsorption of VOC molecules, indicating the presence of strong interactions between AlO₂ and VOCs. The evaluation of the desorption time revealed that AlO₂ could be used as an adsorbent or gas sensors for VOCs at room or higher temperatures. Besides adsorption of VOCs on the bare AlO₂ situations, we also investigate the adsorption of C₂H₃Cl on the AlO₂ surface in N₂, O₂, and water environment to support the possible applications of AlO₂. Compared with normal Al₂O₃, 2D AlO₂ with abnormal stoichiometric ratios can not only extend aluminum-oxide materials, but also provide some interesting properties including a high NPR and good adsorption of VOCs. Moreover, 2D AlO₂ has potentials for the novel mechanical and electronic devices. Therefore, this research is expected to provide new insights into the development and application of AlO₂ and other 2D materials.

  Declaration of competing 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.

  Acknowledgments

  This work was financially supported by National Natural Science Foundation of China (No. 22275149), Fundamental Research Funds for the Central Universities (No. SWU118105) and the Next-Generation Advanced Energy Materials Program of BatteroTech Co., Ltd.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109586.

  
  
• 1. [1]

     X. Li, L. Zhang, Z. Yang, et al., Sep. Purif. Technol. 235 (2020) 116213. doi: 10.1016/j.seppur.2019.116213

  2. [2]

     X. Ma, L. Yang, H. Wu, J. Clean. Prod. 302 (2021) 126925. doi: 10.1016/j.jclepro.2021.126925

  3. [3]

     H. Dai, S. Jing, H. Wang, et al., Sci. Total Environ. 577 (2017) 73–83. doi: 10.1016/j.scitotenv.2016.10.071

  4. [4]

     B. Liu, W. Zhao, Q. Jiang, et al., Sustain. Mater. Technol. 21 (2019) e00103.

  5. [5]

     K. Vikrant, H. Yang, M.W. Chung, et al., Mater. Today Nano 21 (2023) 100283. doi: 10.1016/j.mtnano.2022.100283

  6. [6]

     M. Hauptmann, J.H. Lubin, P.A. Stewart, et al., J. Natl. Cancer Inst. 95 (2003) 1615–1623. doi: 10.1093/jnci/djg083

  7. [7]

     J. Kielhorn, C. Melber, U. Wahnschaffe, et al., Environ. Health Perspect. 108 (2000) 579–588. doi: 10.1289/ehp.00108579

  8. [8]

     V. Bollati, A. Baccarelli, L. Hou, et al., Cancer Res. 67 (2007) 876–880. doi: 10.1158/0008-5472.CAN-06-2995

  9. [9]

     Z. Hu, X. Xiao, H. Jin, et al., Nat. Commun. 8 (2017) 15630. doi: 10.1038/ncomms15630

  10. [10]

      P. Kumbhakar, C.C. Gowda, P.L. Mahapatra, et al., Mater. Today 45 (2021) 142–168. doi: 10.1016/j.mattod.2020.11.023

  11. [11]

      J. Mei, T. Liao, L. Kou, et al., Adv. Mater. 29 (2017) 1700176. doi: 10.1002/adma.201700176

  12. [12]

      H. Zhao, Y. Zhu, F. Li, et al., Angew. Chem. Int. Ed. 56 (2017) 8766–8770. doi: 10.1002/anie.201703871

  13. [13]

      Z. Huang, A. Zhou, J. Wu, et al., Adv. Mater. 28 (2016) 1703–1708. doi: 10.1002/adma.201504484

  14. [14]

      A. Zavabeti, J.Z. Ou, B.J. Carey, et al., Science 358 (2017) 332–335. doi: 10.1126/science.aao4249

  15. [15]

      A. Jannat, N. Syed, K. Xu, et al., ACS Nano 15 (2021) 4045–4053. doi: 10.1021/acsnano.0c06791

  16. [16]

      N. Mahmood, H. Khan, K. Tran, et al., Mater. Today 44 (2021) 69–77. doi: 10.1016/j.mattod.2020.11.016

  17. [17]

      C.H. Huang, H. Chang, T.Y. Yang, et al., ACS Appl. Mater. Interfaces 13 (2021) 52822–52832. doi: 10.1021/acsami.1c18329

  18. [18]

      J. Wu, F. Wang, H. Li, et al., Small 18 (2022) 2104244. doi: 10.1002/smll.202104244

  19. [19]

      J. Ma, Z. Liu, J. Song, et al., Green Chem. 20 (2018) 5188–5195. doi: 10.1039/C8GC02618A

  20. [20]

      W. Shi, C. Yang, M. Qiu, et al., J. Membr. Sci. 642 (2022) 119992. doi: 10.1016/j.memsci.2021.119992

  21. [21]

      W. Fu, Z. Li, W. Xu, et al., Mater. Today Nano 11 (2020) 100088. doi: 10.1016/j.mtnano.2020.100088

  22. [22]

      H. Wang, Y. Liu, J. Song, et al., Mater. Today Nano 20 (2022) 100244. doi: 10.1016/j.mtnano.2022.100244

  23. [23]

      E. Im, H.J. Seo, D.I. Kim, et al., Chem. Eng. J. 416 (2021) 129147. doi: 10.1016/j.cej.2021.129147

  24. [24]

      U. Kumari, S.K. Behera, H. Siddiqi, et al., J. Hazard. Mater. 381 (2020) 120917. doi: 10.1016/j.jhazmat.2019.120917

  25. [25]

      S. Liu, Z. Zhou, J. Chen, et al., Appl. Surf. Sci. 611 (2023) 155645. doi: 10.1016/j.apsusc.2022.155645

  26. [26]

      Y. Bai, G. Wei, Z. Cao, et al., Appl. Surf. Sci. 577 (2022) 151933. doi: 10.1016/j.apsusc.2021.151933

  27. [27]

      S.A. Halawy, A.I. Osman, N. Mehta, et al., Chemosphere 295 (2022) 133795. doi: 10.1016/j.chemosphere.2022.133795

  28. [28]

      W. Wang, Y. Zheng, X. Li, et al., Adv. Mater. 31 (2019) 1803448. doi: 10.1002/adma.201803448

  29. [29]

      Y. Li, Y. Liao, P.V.R. Schleyer, et al., Nanoscale 6 (2014) 10784–10791. doi: 10.1039/C4NR01972E

  30. [30]

      R. Rahimi, M. Solimannejad, A. Chaudhari, J. Hazard. Mater. 403 (2021) 123600. doi: 10.1016/j.jhazmat.2020.123600

  31. [31]

      M. Wang, X. Cai, Q. Yang, et al., Appl. Surf. Sci. 623 (2023) 157025. doi: 10.1016/j.apsusc.2023.157025

  32. [32]

      Y. Yang, B. Peng, X. Lei, et al., Prog. Chem. 34 (2022) 1524–1536.

  33. [33]

      X. Xia, Y. Huang, B. Peng, et al., Adv. Mater. 35 (2023) 2303072. doi: 10.1002/adma.202303072

  34. [34]

      Y.C. Lu, R. Dieckmann, S.L. Sass, J. Phys. Chem. Solids 55 (1994) 1083–1095. doi: 10.1016/0022-3697(94)90126-0

  35. [35]

      L. Andrews, T.R. Burkholder, J.T. Yustein, J. Phys. Chem. 96 (1992) 10182–10189. doi: 10.1021/j100204a018

  36. [36]

      J.M. Soler, E. Artacho, J.D. Gale, et al., J. Phys. Condens. Mater. 14 (2002) 2745–2779. doi: 10.1088/0953-8984/14/11/302

  37. [37]

      P.J. Stephens, F.J. Devlin, C.F. Chabalowski, et al., J. Phys. Chem. 98 (1994) 11623–11627. doi: 10.1021/j100096a001

  38. [38]

      J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868. doi: 10.1103/PhysRevLett.77.3865

  39. [39]

      B. Hammer, L.B. Hansen, J.K. Norskov, Phys. Rev. B 59 (1999) 7413–7421. doi: 10.1103/PhysRevB.59.7413

  40. [40]

      A. Canal Neto, F.E. Jorge, Chem. Phys. Lett. 582 (2013) 158–162. doi: 10.1016/j.cplett.2013.07.045

  41. [41]

      M. Thomas, M. Brehm, R. Fligg, et al., Phys. Chem. Chem. Phys. 15 (2013) 6608–6622. doi: 10.1039/c3cp44302g

  42. [42]

      S.Q. Zhou, X.H. Ju, F.Q. Zhao, et al., Appl. Surf. Sci. 258 (2012) 7334–7342. doi: 10.1016/j.apsusc.2012.03.187

  43. [43]

      R. Kiruba, S. Vinod, A.W. Zaibudeen, et al., Colloids Surf. A 555 (2018) 63–71. doi: 10.1016/j.colsurfa.2018.06.044

  44. [44]

      M. Wang, C.M. Li, Phys. Chem. Chem. Phys. 13 (2011) 1413–1418. doi: 10.1039/C0CP00828A

  45. [45]

      Q. Peng, C. Liang, W. Ji, et al., Phys. Chem. Chem. Phys. 15 (2013) 2003–2011. doi: 10.1039/C2CP43360E

  46. [46]

      Q. Peng, S. De, Phys. Chem. Chem. Phys. 15 (2013) 19427–19437. doi: 10.1039/c3cp52879k

  47. [47]

      Q. Yang, X.H. Cai, Y. Pang, et al., Int. J. Quantum Chem. 120 (2020) e26378. doi: 10.1002/qua.26378

  48. [48]

      Q. Peng, C. Liang, W. Ji, et al., Comput. Mater. Sci. 68 (2013) 320–324. doi: 10.1016/j.commatsci.2012.10.019

  49. [49]

      J.W. Jiang, H.S. Park, Nat. Commun. 5 (2014) 4727. doi: 10.1038/ncomms5727

  50. [50]

      H. Wang, X. Li, J. Sun, et al., 2D Mater. 4 (2017) 045020. doi: 10.1088/2053-1583/aa8abd

  51. [51]

      S. Zhang, J. Zhou, Q. Wang, et al., Proc. Natl. Acad. Sci. U. S. A. 112 (2015) 2372–2377. doi: 10.1073/pnas.1416591112

  52. [52]

      L. Kou, Y. Ma, C. Tang, et al., Nano Lett. 16 (2016) 7910–7914. doi: 10.1021/acs.nanolett.6b04180

  53. [53]

      H. Wang, X. Li, P. Li, et al., Nanoscale 9 (2017) 850–855. doi: 10.1039/C6NR08550D

  54. [54]

      T. Jing, D. Liang, M. Deng, et al., J. Mater. Chem. C 8 (2020) 10382–10389. doi: 10.1039/D0TC02415E

  55. [55]

      R. Peng, Y. Ma, Z. He, et al., Nano Lett. 19 (2019) 1227–1233. doi: 10.1021/acs.nanolett.8b04761

  56. [56]

      L.C. Zhang, G. Qin, W.Z. Fang, et al., Sci. Rep. 6 (2016) 19830. doi: 10.1038/srep19830

  57. [57]

      C. Huang, L. Chen, Adv. Mater. 28 (2016) 8079–8096. doi: 10.1002/adma.201601363

  58. [58]

      Y. Du, J. Maassen, W. Wu, et al., Nano Lett. 16 (2016) 6701–6708. doi: 10.1021/acs.nanolett.6b03607

  59. [59]

      D. Li, X. Chen, C. Shu, et al., Appl. Surf. Sci. 563 (2021) 150233. doi: 10.1016/j.apsusc.2021.150233

  60. [60]

      A. Aasi, S.M. Aghaei, B. Panchapakesan, J. Mater. Chem. C 9 (2021) 9242–9250. doi: 10.1039/D1TC01510A

  61. [61]

      J. Cao, F. Wu, M. Wen, et al., Appl. Surf. Sci. 539 (2021) 148164. doi: 10.1016/j.apsusc.2020.148164

  62. [62]

      L. Gao, P. Feng, Y. Sun, et al., Comput. Theor. Chem. 1209 (2022) 113611. doi: 10.1016/j.comptc.2022.113611

  63. [63]

      Q. Zhao, Y. Man, S. Li, et al., Comput. Theor. Chem. 1214 (2022) 113744. doi: 10.1016/j.comptc.2022.113744

  64. [64]

      Y. Liu, L. Gao, S. Fu, et al., Appl. Surf. Sci. 611 (2023) 155694. doi: 10.1016/j.apsusc.2022.155694

• 

  Figure 1  (a) Geometrically optimized configuration of 2D AlO₂, (b) top and lateral views. The unit cell is marked with a light blue rectangular area. Four possible adsorption sites (O1, O2, Al1 and I1) are marked. (c) Potential energy fluctuations of AlO₂ in AIMD simulations at 300 and 1000 K, and the inset shows the structure after 5 ps of AIMD simulations at 1000 K. (d) Phonon band structure and phonon DOS of AlO₂. (e) Band structure of AlO₂ and the corresponding DOS.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) AlO₂ strain energy under various strains. (b) In-plane stiffness C and (c) Poisson's ratio ν of AlO₂. (d) Comparison of in-plane Poisson's ratios of some 2D materials. The schematic images of in-plane NPR phenomenon caused by the angle ∠OAlO of AlO₂ under transverse (e) stretching and (f) compression.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Side view of the geometry of 1–5 layers AlO₂ with the intra-layer heights marked. (b) Forming energy of 2–5 layers of AlO₂. (c) The DOS of 1–5 layers AlO₂. (d) In-plane stiffness C and (e) Poisson's ratio ν of 1–5 layers AlO₂. The inset shows an enlarged view of the NPR in (e).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Top and side views of the bilayer AlO₂ structure under biaxial strain of (a) −5%, (b) 0%, and (c) 5%. d₁ denotes the interlayer height of bilayer AlO₂. d₂ denotes the intra-layer height of single-layer AlO₂. (d) Percentage change of interlayer height d₁ and intra-layer height d₂ with biaxial strain. (e) Longitudinal expansion phenomenon of bilayer AlO₂ in the case of transverse stretching. (f) Longitudinal shrinkage phenomenon of bilayer AlO₂ under lateral compression.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Stable configurations and charge density differences of (a, d) CH₂O, (b, e) C₂H₃Cl and (c, f) C₆H₆ adsorbed on the surface of AlO₂. The red, blue, gray, white and green balls are used to denote O, Al, C, H and Cl atoms. The value of 0.025 e/ų is determined as the isosurface. The amount and direction of charge transferred in the adsorption system are marked. The DOS before and after adsorption of (g) CH₂O, (h) C₂H₃Cl and (i) C₆H₆ by monolayer AlO₂. (j) Comparison of adsorption energies and charge transfers of some 2D materials for the adsorption of VOCs.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Desorption time of CH₂O, C₂H₃Cl and C₆H₆ molecules on AlO₂ surface at different temperatures.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Stable configurations, charge density differences and DOS of (a, d) N₂, (b, e) O₂, (c, f) H₂O, (g, j) N₂-C₂H₃Cl, (h, k) O₂-C₂H₃Cl and (i, l) H₂O-C₂H₃Cl adsorbed on the surface of AlO₂. (m) Comparison of adsorption energies and charge transfers of monolayer AlO₂ for the adsorption of gas molecular and C₂H₃Cl (in O₂, N₂ and water environment).

  下载: 全尺寸图片 幻灯片
•