Introduction

The recent emergence of new two-dimensional (2D) materials opened up a diversity of new applications in the fields of environment, energy storage, electronics, and biomedical engineering [1,2,3,4,5,6,7,8,9,10,11]. The first two-dimensional material known as graphene was synthesized by Novoselov et al. in 2004 [12]. Since its discovery, graphene has been a focus of numerous theoretical and experimental studies [13,14,15,16,17,18,19,20].

Very recently, new post-graphene 2D materials such as stanene [21,22,23,24], germanene [25], bismuthene [26], and borophene [27,28,29,30,31] were discovered after being theoretically predicted. Particular attention is drawn to the borophene, an allotropic form of two-dimensional boron defined as boron sheet which has been synthesized by depositing vaporizing boron on silver Ag(111) substrate [27]; the miscellaneous bonding configurations adopted by boron nanosheets allow the borophene to exhibit a polymorphism and different crystal structures, consisting of combined hexagonal and triangular motifs, thus resulting in the following most common polymorphs denominated β12 and χ3 and striped–borophenes [27, 28]. Diverse borophene allotropes could be obtained experimentally by controlling the synthesis conditions such as the nature of the substrate and its crystallographic orientation dependence [32].

New architectures based on 2D boron nanosheets have also been developed to enrich the class of borophene clusters [33,34,35,36,37]. Among them, the smallest borophene (B36) cluster [38] having sixfold symmetry was experimentally evidenced and described as a highly stable flattened bowl-shaped boron cluster bearing a hexagonal vacancy at the center [39,40,41]. The latest reports show that borophene, with its both metallicity and lightweight properties [27], has an interesting potential for catalysis, hydrogen storage, and high capacity electrode [42,43,44,45,46]. Additionally, borophene is a promising chemical sensor material and could be used for making sensing devices for detecting toxic substances [47, 48].

Acrolein (AC) or 2-propenal is an organic compound and the simplest unsaturated three-carbon aldehyde; it is used mainly as a pesticide and as a precursor for producing acrylic acid. However, exposure to the acrolein vapor causes mucosal inflammation; it is highly toxic if inhaled and irritating to the skin, eyes, and upper respiratory tract even at low levels of concentrations; acrolein released from tobacco smoke is associated with respiratory toxicity [49]; it was also reported that acrolein could potentially be human carcinogenic [50].

Due to the importance of designing high sensitive devices capable of detecting acrolein, a number of computational investigations focused on the adsorption of acrolein on various adsorbent materials were carried out [51,52,53,54,55,56,57,58]. The ability of pristine and Al-doped graphenes to adsorb acrolein has been studied using DFT by Rastegar et al. [51]; their investigations confirmed the sensitivity of Al-doped graphene toward acrolein. Karami [52] reported a DFT study of acrolein adsorption on graphyne nanotube and revealed the potential of graphyne sheets for acrolein sensing applications. Pirillo et al. [55] explored theoretically the adsorption properties of acrolein on Pt(111); the authors indicated that the adsorption process is mainly due to the formation of C–Pt bonds.

Quantum chemical calculations may provide a tremendous insight regarding the adsorption process including the nature of involved interactions; it is worth noting that numerous 2D materials were predicted theoretically before being experimentally confirmed.

Herein, the plausibility of adsorption of acrolein on borophene B36 was assessed with dispersion-corrected density functional calculations; geometrical parameters and electronic structures, as well as the adsorption properties of the interacting systems, have been examined systematically.

Computational details

The geometries of all studied structures in this work were optimized using DFT-D3 methodology where dispersion correction with Becke–Johnson D3(BJ) and standard zero D3(0) damping functions are applied to account for the dispersion interactions. B3LYP-D3(BJ) [59,60,61,62], B97-D3(BJ) [60, 63], and M06-2X-D3(0) [59, 60, 64] levels of calculations were employed and combined with the def2-SVP [65,66,67] basis set in vacuum. These computations were carried out with ORCA program package [68] (version-4.2.0). For improving the computational efficiency procedure, the resolution of identity (RI) and chain of spheres exchange (COSX) approximations were applied [69,70,71]. A geometrical counterpoise correction (gCP) was incorporated to account for the basis set superposition error (BSSE) [72]. The frequency calculations were performed in order to confirm and verify the minima of all obtained stationary points. The visualization of the molecular orbitals was performed with the Avogadro software [73]. Analysis of weak intramolecular and intermolecular interactions between AC molecule and B36 nanosheet was carried out with the non-covalent interaction-reduced density gradient analysis approach (NCI-RDG) [74] using the Multiwfn software [75] and VMD visualization program [76]. The energy decomposition analysis (EDA) was determined with the GAMESS code [77] by the M06-2X [62] functional and 6-31G(d,p) basis set [78].

The adsorption energy is calculated according to the following expression [79,80,81,82,83,84,85,86,87,88,89,90]:

$$ {E}_{\mathrm{ads}}=\left(E\left[{\mathrm{B}}_{36}{\left(\mathrm{AC}\right)}_{\mathrm{n}}\right]-E\left({\mathrm{B}}_{36}\right)-\mathrm{n}E\left(\mathrm{AC}\right)\right)/\mathrm{n} $$
(1)

where E[B36(AC)n] corresponds to the total energy of the B36(AC)n complex, E(B36) is the energy of the isolated B36 nanosheet, and E(AC) is the energy of the isolated single AC molecule.

The change in the HOMO–LUMO gap |ΔEg|, as an indicator of the electronic sensitivity of the B36 toward the dissociative adsorption of AC molecule, is obtained by:

$$ \mid {\Delta E}_{\mathrm{g}}\mid =\kern0.5em \left[\left({E}_{\mathrm{g}2}-{E}_{\mathrm{g}1}\right)/{E}_{\mathrm{g}1}\right]\times 100 $$
(2)

where Eg1 and Eg2 are, respectively, the values of Eg in the initial and the final states.

Results and discussion

Properties of B36 borophene

The optimized geometry of the B36 borophene nanosheet was determined by the following theoretical levels of calculations: B3LYP-D3(BJ)/def2-SVP, B97-D3(BJ)/def2-SVP, and M06-2X-D3(0)/def2-SVP. Our results are corroborated by the experimental data [38] and previous theoretical [79,80,81,82,83,84,85,86] studies, indicating that the B36 borophene is consisting of three different sizes of rings: internal (B1–B6), median (B7–B18), and external rings (B19–B36), as shown in Fig. 1. The B36 borophene bears a quasi-planar structure with a central hexagonal hole and exhibits a perfect hexagonal C6v symmetry [48, 81]. The optimized structure at B97-D3(BJ)/def2-SVP predicts different B–B bond lengths ranging between 1.609 and 1.749 Å and close to the values calculated by Rostami and Soleymanabadi [80]. Frequency calculations were performed at the same level of theory for the reported structures without producing imaginary frequencies.

Fig. 1
figure 1

Optimized geometry structure of the B36 nanosheet calculated at the B97-D3(BJ)/def2-SVP level. (A) Side view; (B), (C) top views; (B): with bond lengths (Å) and (C): with bond angles (°)

Full size image

The calculated electronic parameters, such as the frontier molecular orbital energy (EHOMO, ELUMO) and HOMO–LUMO gap energy |Eg|, offer a valuable method for evaluating the reactivity and stability of such molecular systems [81,82,83,84,85,86], where smaller Eg (absolute values) indicates high reactivity, whereas larger values of |Eg| show a greater stability [91, 92].

Based on the calculated values of |Eg| (Table 1), it was found that the neutral B36 nanosheet is more reactive than cationic and anionic species, following the order: B36 >\( {B}_{36}^{-} \)>\( {B}_{36}^{+} \). The larger HOMO–LUMO gaps predicted by M06-2X functional are due to its higher amount of exact Hartree-Fock (HF) exchange (54%). These results are in agreement with previous similar studies [93,94,95] which concluded that increasing HF exchange percentage in DFT functionals leads to the increase of HOMO–LUMO gap.

Table 1 The frontier molecular orbital energy (EHOMO and ELUMO) and HOMO–LUMO gap|Eg|, dipole moment (μ), and total energy (TE) in kcal mol−1 of B36, \( {B}_{36}^{+} \), and \( {B}_{36}^{-} \) nanosheets
Full size table

The condensed Fukui function (Fig. 2) and molecular electrostatic (ESP) potential (Fig. 3) were used as reactivity descriptors in assessment and prediction of the most reactive centers, especially nucleophilic and electrophilic sites.

Fig. 2
figure 2

The calculated Fukui functions of the B36 nanosheet; (left) nucleophilic attack f+and (right) electrophilic attack f−

Full size image
Fig. 3
figure 3

The molecular electrostatic potential (ESP) surfaces of the B36 nanosheet; negative ESP regions are indicated in blue, and the positive regions in red

Full size image

In the literature [91,92,93,94, 96], it has been reported that the preferred center for electrophilic attack is the atom in the molecule where the value of f− is the highest, while the site for nucleophilic attack is the atom in the molecule where the value of f+is the highest.

The calculated values of the Fukui functions (f+ and f−) are illustrated in Fig. 2. The results predict that \( {f}_{\mathrm{k}}^{+} \) values are highest on the external rings of neutral B36 borophene, precisely on the corner sites (B21, B24, B27, B30, B33, and B36 atoms), suggesting that these sites would be most favorable for nucleophilic attack. The most susceptible sites for electrophilic attacks are B2, B5, B7, and B13 atoms, followed by atoms of the corner sites B21, B24, B27, B30, B33, and B36. This observation clearly establishes that the most reactive sites are those located at the corner sites of neutral B36, susceptible to both nucleophilic and electrophilic attacks. These findings are supported by several studies [48, 82, 84] indicating that the edge boron atoms of neutral B36 borophene are more reactive than the inner atoms.

Figure 3 shows the ESP maps of the van der Waals surface of the optimized neutral B36, in which the blue and red regions on the scale bar indicate respectively the lowest and highest ESP energy values. The surface local minima and maxima of ESP are represented respectively by cyan and orange spheres.

The inspection of the calculated ESP plots on the van der Waals surface revealed that the local minima of ESP surface are dispersed near the outer side of B36 (external ring “B19–B36”), with larger negative values (− 5.874 to − 5.907 kcal mol−1); thus, they could act as reactive sites and also may favor the electrophilic reaction.

The positive ESP regions on the van der Waals surface with a large value (i.e., ˃ 17 kcal mol−1) are distributed in both convex and concave sides of B36 surface, in which their local maxima of ESP are significant but relatively larger on the concave side (21.618 to 21.653 kcal mol−1) than convex side (17.451 to 17.546 kcal mol−1). This interesting observation suggests that the concave side is chemically more reactive than the convex side.

Adsorption of single acrolein molecule on the B36 borophene

In order to find and examine the most stable adsorption geometries, several initial configurations of AC adsorption on the B36 borophene were investigated, taking into consideration the convex and concave shapes of the B36 borophene surface and various possible orientations of AC molecule. With this aim, we have explored various sites by placing the oxygen atom of the acrolein molecule on top of the boron atoms, and also by orienting the AC molecule both in perpendicular and parallel directions to the B36 borophene sheet, as shown in supplementary data (see Fig. S1).

Different configurations for the adsorption of the acrolein on B36 borophene were considered in this study. However, only geometries with local minima obtained at the B97-D3(BJ)/def2-SVP level of theory were kept, in which the calculated adsorption energy was higher than − 0.99 kcal mol−1, indicating that the adsorption is spontaneous and the process is exothermic. Three local minima are selected with an adsorption energy higher or equal to − 9.8 kcal mol−1. Figure 4(B)–(D) show the optimized geometries of the three most stable configurations, and their associated frontier molecular orbital distributions (HOMO and LUMO).

Fig. 4
figure 4

The optimized geometries and frontier molecular orbital distributions HOMO and LUMO of (A) B36 nanosheet and the three most stable complexes: (B) AC@B36 edge, (C) AC@B36 concave, and (D) AC@B36 convex

Full size image

The most stable formed complex is (B), in which the AC molecule is attached via the oxygen atom of its carbonyl group to the boron atom of corner (edge) site of B36 structure. The calculated adsorption energy is − 25.20 kcal mol−1, and the newly formed bond length B–O was 1.466 Å. As shown in Table 2, the maximum change of |ΔEg| has been observed in the edge configuration (|ΔEg| = 67.10%), where both HOMO and LUMO levels were shifted respectively to higher and lower energies, thus resulting in the significant decrease of HOMO–LUMO gap.

Table 2 Adsorption energy (Eads, kcal mol−1), HOMO and LUMO energies (EHOMO and ELUMO, eV), HOMO–LUMO energy gap (|Eg|, eV), and the change of |Eg| after the adsorption process (|ΔEg|) for studied B, C, and D complexes
Full size table

From Table 2, it can be observed that the adsorption energies decrease in the following order: Eads edge > Eads concave > Eads convex. The calculated density of states (DOS) for B36 nanosheet and its three complexes AC@B36 edge, AC@B36 concave, and AC@B36 convex is illustrated in Fig. 5. The results clearly reveal that the AC@B36 edge complex affects strongly the electronic properties of B36 borophene if compared with both AC@B36 concave and AC@B36 convex.

Fig. 5
figure 5

The density of states (DOS) of (A) B36 nanosheet and the three most stable complexes: (B) AC@B36 edge, (C) AC@B36 concave, and (D) AC@B36 convex

Full size image

For C and D complexes (Fig. 4), the AC molecule is horizontally adsorbed on the concave and convex surfaces of B36 borophene with the adsorption energies of − 15.28 and − 9.60 kcal mol−1 respectively, indicating a weak interaction. This observation is obviously consistent with our previous ESP analysis (Fig. 3) which predicted that the concave side of B36 borophene is relatively more reactive than the convex side.

Furthermore, we introduced the non-covalent interaction (NCI) analysis based on the reduced density gradient (RDG) function, in order to analyze the intermolecular interactions and to distinguish the nature of the interactions between the AC molecule and B36 nanosheet. The RDG quantities are calculated by the following equation [93, 94, 96,97,98,99,100]:

$$ s=\frac{1}{2{\left(3{\pi}^2\right)}^{1/3}}\frac{\left|\nabla \rho \right|}{\rho^{4/3}} $$
(3)

where ρ and ∇ρ are density and its first derivative, respectively. The regions and types of interactions between AC molecule and B36 nanosheet can be identified and visualized by colored RDG isosurface, in which the blue regions represent strong attractive interactions like H-bond, green denotes weak van der Waals interactions, and the red regions represent steric repulsion.

The RDG isosurfaces for the three complexes are shown in Fig. 6, with the color ranging from blue to red corresponding to the value of − 0.05 < ρ*sign(λ2) < 0.05 a.u.

Fig. 6
figure 6

The NCI-RDG isosurfaces (bottom) and scatter plot (upper) of (A) AC@B36 edge, (B) AC@B36 concave, and (C) AC@B36 convex

Full size image

For AC@B36 concave and AC@B36 convex complexes, larger green isosurfaces (dark green and light green) were observed between AC molecule and B36 nanosheet, corresponding to the ranges between − 0.010 and − 0.015 a.u. (light green) and between − 0.01 a.u. and − 0.015 a.u. (dark green), which are characteristic of intermolecular van der Waals interactions, confirming that van der Waals (vdW) interactions are the key driving force for the adsorption of acrolein molecule on the B36 borophene in both AC@B36 concave and AC@B36 convex complexes.

In addition, for the AC@B36 edge complex, two green isosurfaces were determined. The first is located between hydrogen atom (H41) of AC molecule and boron atom (B20) of B36 borophene. The second isosurface is an intramolecular interaction occurring between oxygen (O42) of the carbonyl group and hydrogen atom (H44) of AC molecule. These weak van der Waals interactions contribute to global stability and explain the orientation of AC molecule in the AC@B36 edge complex.

Energy decomposition analysis (EDA) was performed with GAMESS code for the analysis of intermolecular interactions between the acrolein and B36 borophene using M06-2X functional which is recommended for predicting and describing non-covalent interactions and 6-31G(d,p) basis set. The total interaction energy obtained from a complex calculation (AC@B36) in both concave and convex configurations is decomposed into electrostatic, exchange, repulsion, and polarization terms; the computed decomposed energy terms are reported in Table 3.

Table 3 Decomposed energy components (in kcal mol−1) computed at the M06-2X/6-31G(d,p) level
Full size table

From Table 3, it is clear that dispersive energy dominates the attractive energy components and provides the largest contribution to the total interaction energy, the dispersive energy of AC@B36 concave is greater by 7.42 kcal mol−1 than that of AC@B36 convex; however, the repulsion energy due to the destabilization is larger for AC@B36 concave (+ 35.44 kcal mol−1) and higher than AC@B36 convex by 5.36 kcal mol−1. These results are consistent with those obtained from the RDG-NCI analysis confirming that vdW dispersion interactions play a key role in the stabilization of both complexes.

Adsorption of multiple acrolein molecules on the B36 borophene

In order to examine the concentration effect on the adsorption of several acrolein molecules, we have investigated the adsorption of up to six acrolein molecules on the B36 surface. For this purpose, we firstly selected the most stable complex with a singly edge-adsorbed acrolein, and then, the additional acrolein molecules were added increasingly and successively at the edge, concave, and convex sides, while focusing primarily on the adsorption at the edge sides. A reoptimization procedure was performed for each configuration at B3LYP-D3(BJ)/def2-SVP, B97-D3(BJ)/def2-SVP, and M06-2X-D3(0)/def2-SVP computational levels. For each newly optimized complex, various possible configurations are compared as shown in Fig. 7. The calculated adsorption energy, energy of HOMO and LUMO, HOMO–LUMO gaps, and the percentage of variation of HOMO–LUMO gap |ΔEg| for studied complexes with multiple AC molecules at the B97-D3(BJ)/def2-SVP theoretical level are reported in Table 4. Computational results obtained at B3LYP-D3(BJ)/def2-SVP and M06-2X-D3(0)/def2-SVP theoretical levels are listed in Tables S1 and S2 (see supplementary data).

Fig. 7
figure 7

Optimized geometry structures for studied complexes with multiple adsorbed AC molecules at B97-D3(BJ)/def2-SVP levels

Full size image
Table 4 Adsorption energy (Eads, kcal mol−1), energy of HOMO and LUMO (EHOMO and ELUMO, eV), HOMO–LUMO energy gap (|Eg|, eV), and the percentage of variation of |Eg| after the adsorption process (|ΔEg|) for studied complexes with multiple AC molecules at B97-D3(BJ)/def2-SVP computational levels
Full size table

From the calculated adsorption energies at the B97-D3(BJ)/def2-SVP level (Table 4), it is evident that the most stable complexes are A1, B1, C1, D1, and E1, in which AC molecules are adsorbed at the edge sites of B36 nanosheet.

The results from Table 4 clearly show that the HOMO levels are shifted to upper energies whereas the LUMO levels are shifted to lower energies after the adsorption process, particularly when increasing the number of adsorbed AC molecules, leading to a significant decrease of the HOMO–LUMO energy gap. In addition, the percentage of variation of HOMO–LUMO gap |ΔEg| exhibits greater values following the order: 2AC@36 < 3AC@36 < 4AC@36 < 5AC@b36 ≈ 6AC@B36.

Based on these results, it can be concluded that the concentration of acrolein affects the electronic properties of borophene by decreasing the HOMO–LUMO gap and, therefore, increasing the electrical conductance. Thus, the borophene can serve as a sensing material for the detection of acrolein molecule.

Conclusions

Using DFT calculations, we extensively studied the electronic and geometrical properties of acrolein adsorption on B36 borophene nanosheet. Our results reveal that the B36 borophene bears a quasi-planar structure with a central hexagonal hole, and the neutral B36 nanosheet is more reactive than cationic and anionic species. The calculated values of the Fukui functions show that the most reactive sites are those located at the corner sites of neutral B36, susceptible to both nucleophilic and electrophilic attacks; more specifically, the edge boron atoms are the most stable sites of adsorption. The acrolein molecule is attached to the borophene surface via the oxygen atom of its carbonyl group with an adsorption energy of − 25.20 kcal mol−1. The NCI-RDG analysis highlighted the role of van der Waals interactions in stabilizing the adsorption of acrolein on both convex and concave surfaces of borophene, in accordance with the results of the energy decomposition analysis. Up to six acrolein molecules can be adsorbed preferentially at the edge sides of the B36 borophene. According to the calculated density of states, acrolein molecule affects strongly the electronic properties of B36 borophene by decreasing the HOMO–LUMO gap and, thus, inducing changes in its electrical conductance. B36 borophene nanosheet is expected to be applied as a nanosensor for detecting acrolein.