The effects of nonmetal dopants on the electronic, optical, and catalytic performances of monolayer WSe₂ by a first-principles study

Abstract

Doping modifies the electronic, optical, and catalytic behavior of materials through the newly formed chemical bonds and the localized electrons. With the aid of first-principles calculations, the electronic, optical, and catalytic performances of the nonmetal (NM = H, B, C, N, O, F, Si, P, S, Cl, As, Br, Te, or I)-doped monolayer WSe₂ were investigated. The results showed that the NM dopants substitute preferentially for Se under a W-rich condition and H, F, Cl, Br, and I atoms are willing to locate at the interstitial site. The electron-clouds around the dopants and nearby W or Se atoms were altered by the newly formed W–NM or Se–NM bonds, with the differences determined by the bonding strength between them. The band gap, optical absorption edge, and intensities were altered or shifted by less than 0.08 eV, 32 nm, and 9.5%, respectively. The H, F, P, Cl, As, Br, and I dopants were conductive to separating the photogenerated e⁻/h⁺ pairs, whereas the B, C, Si, and Te dopants became recombination centers for the photogenerated e⁻/h⁺ pairs. Compared with pristine monolayer WSe₂, NM atoms with odd free electrons reduced the reduction potential by 0.39–0.71 eV and enhanced the oxidation potential by 0.45–0.75 eV. Thus, we can adjust the redox potentials of monolayer WSe₂ by introducing different kinds of NM impurities for various catalytic reactions, and the H-, F-, P-, Cl-, As-, Br-, and I-doped specimens have excellent photocatalysis capability.

---

Introduction

Transition-metal dichalcogenide compounds with a layer structure have attracted considerable attention due to their wide range of electronic, optical, and catalytic performances.^1–3 This category of compounds have a sandwiched structure with one transition-metal layer coupled to two chalcogen layers, with the transition-metal atoms bonded to chalcogen atoms by strong covalent interactions and the adjacent sandwiched layers connected to each other by weak van der Waals forces.^4–7 Hence, two-dimensional transition-metal dichalcogenide compounds can be easily synthesized with a high specific surface area and with excellent physical and chemical performances. In this regard, monolayer tungsten diselenide (WSe₂) has a narrow direct band gap (1.60 eV) and potential applications in electronic devices, optical devices, and as a catalyst.^8–11 For instance, high performance field-effect transistors,¹² valleytronics,¹³ photovoltaic devices,^14–16 and catalysts¹¹ have been successfully fabricated.

To date, the physical and chemical performances of two dimensional WSe₂ have been modulated by strain,¹⁷ adsorbed molecules,^18,19 interface composites,²⁰ vacancies,²¹ and impurities.^22–24 In recent years, several reports have indicated that metal and nonmetal (NM) dopants can effectively alter the electronic,^25–27 magnetic,²⁸ optical,²⁹ and catalytic³⁰ performances of monolayer WSe₂ by introducing new chemical bonds and changing the binding energies of the nearby bonds. For instance, Hu et al.³¹ have demonstrated that Re-doped WSe₂ can enhance the electronic and optical performances. Kriener et al.²⁴ studied the modification of the electronic structure and the hole-doping effect for the layered WSe₂ with a multi-valley band structure, where a Te dopant was found to change the electronic states. Huang et al.³² synthesized a large area S-doped monolayer WSe₂ with a tunable band gap, which exhibited promising stable electronic properties. Zou et al.³³ reported the excellent hydrogen evolution reaction performance of layered WSe₂ by optimizing the substitution of Se with S. Although the physical and chemical performances of several doped layered WSe₂ compounds have been studied theoretically and experimentally, the detail on the bond relaxations of the doped specimens is ambivalent. It was reported that chemical bonds not only decide the electronic and optical performances but also affect the redox potential of materials and can be altered by bond relaxation.³⁴ Thus, a detailed study of the chemical bonds in doped WSe₂ is necessary to understand the fundamental mechanism of bond relaxation.

In the present work, the structural and thermal performances of monolayer WSe₂ after doping with an NM atom (H, B, C, N, O, F, Si, P, S, Cl, As, Br, Te, or I) at the substitutive and interstitial sites were calculated. Also, the electronic and optical performances of all the specimens with a stable configuration were determined by first-principles calculations. Then, their catalytic performances and redox potentials were considered, and the effects of the different NM dopants on the electronic performances and chemical bonds of W–NM or Se–NM were compared.

Computational methods

After cutting the bulk WSe₂ along its (001) direction, monolayer WSe₂ was built. The spurious interactions effect in the nonperiodic directions was considered, whereby a vacuum region of 20 Å was shown to be sufficient. Then, 4 × 4 supercell monolayer W₁₆Se₃₂ after doping with a NM atom at the substitutive and interstitial positions was constructed (see Fig. 1). For the substitutive systems, one Se atom was replaced by one NM atom. Although several doped systems were not neutral due to the different number of valence electrons from the Se atom, the larger distance between the nearest neighbor dopants (about 13.50 Å) was enough to avoid unphysical charge–charge interactions. The density functional theory calculations, including geometry relaxation the and electronic performances [band structure, partial density of states (PDOS), electron density difference, the highest occupied molecular orbital (HOMO), and the lowest unoccupied molecular orbital (LUMO)], were performed utilizing the CASTEP package.³⁵ The generalized gradient approximation (GGA)³⁶ with the Perdew–Wang 91 (PW91) functional and norm-conserving pseudopotentials³⁷ was applied as the exchange–correlation functional. A 720 eV cutoff energy and 7 × 7 × 1 k-point sampling set were used for the convergence. Both the size and atomic positions of the supercells were optimized with a fixed c-axis, and the force and energy tolerances were respectively set as 0.01 eV Å⁻¹ and 1.0 × 10⁻⁵ eV per atom, while the maximum displacement was set as 1.0 × 10⁻³ Å. Moreover, the ab initio molecular dynamics (MD) simulation as implemented in VASP^38,39 was carried out for all the substitutive (W₆₄Se₁₂₄NM₄) and interstitial (W₆₄Se₁₂₈NM₄) specimens to explore the stable structure. The simulations were performed in the NVT ensemble with a target temperature of 300 K, where the time step and total time were set as 3 fs and 5 ps, respectively.

Fig. 1 Schematic for the crystal structure of 4 × 4 supercell monolayer WSe₂: (a) substituted site, (b) interstitial site, where the blue, yellow, and red spheres denote the W, Se, and NM atoms, respectively.

The substituted E_sub and interstitial E_int energies of all the doped systems were defined by the following equation²² to estimate the thermal stability:

where E(pristine) and E(specimens) are the total energies of monolayer WSe₂ without and with modification by a NM atom, and μ_Se and μ_NM are the chemical potentials of the substituted Se host atom and the substitutional NM atom, respectively. The μ_NM values were calculated with respect to the elemental bulk or gas in nature. The formation energy of pristine monolayer WSe₂ itself, E_f(WSe₂), could be calculated from the following equation:

E_f(WSe₂) = E_WSe2 − E_W − 2E_Se (2)

where E_WSe2 = E(pristine) per WSe₂ formula unit, and E_W and E_Se are the energy per atom of W and Se in their reference phase (the reference phase of W used is the bulk metallic body centered cubic structure, whereas it is the monoclinic bulk for Se). The value of E_f(WSe₂) was then calculated to be −1.93 eV.

Considering the dynamic growth process of WSe₂, the atomic chemical potential depends on the growth condition (i.e., when changing from a W-rich condition to a Se-rich condition). Hence, under W-rich conditions, μ_W was assumed to be the energy per atom in bulk W, and thus μ_Se could be obtained according to following equation:

2μ_Se + μ_W = μ_WSe2 (3)

where μ_WSe2 denotes the total energy per WSe₂ formula unit. However, under Se-rich conditions, μ_Se is the total energy of bulk Se in its reference phase.

The bonding strengths of the W–NM and W–Se₂ bonds of all the doped systems were, respectively, defined as:

where E^bonding_specimen and E^bonding_specimen* denote the bonding energies of the specimens, including and excluding the dopant (see Fig. 1S†), and E^bonding_NM and denote the bonding energies of the NM atom and the Se₂–NM group.

Result and discussion

Structural and thermal performances

After geometry relaxation, the lattice parameter (a = 3.298 Å) of WSe₂ was consistent with the experimental (3.282 Å (ref. 40) and 3.260 Å (ref. 41)) and other simulated results (3.282 Å (ref. 42)); this ensured the rationality of the calculated parameters. After doping with NM atoms, the crystal configurations of the doped specimens were changed. The lattice parameter a and the bond lengths of the W–NM, Se₁–NM, W–Se₂, W–Se₃, and W–Se₄ doped specimens are displayed in Fig. 1 and listed in Table 1. The results show that the bond lengths of the W–NM bonds (d_W–NM) of the substituted systems and of the Se–NM bonds (d_Se1–NM) of the interstitial systems were changed synchronously, caused by the atomic radius of the NM dopants. Therefore, the nearest neighbor chemical bonds (W–Se₂, W–Se₃, and W–Se₄ bonds) were relaxed by the newly formed W–NM or Se–NM bonds (see Table 1), and the specific values of the W–Se bonds after (d^doped_W–Se) and before (d^pristine_W–Se) doping with the NM atoms are shown in Fig. 2, where the positive and negative values, respectively, denote the bond extension and shrinkage. It was found that all the W–Se bonds had been slightly relaxed and their deformations were less than ±2%. Besides, the deformations of lattice parameters of all the specimens were less than ±1%. Briefly, this shows that the NM dopants affect the crystal configurations of monolayer WSe₂ to some extent.

Table 1 The equilibrium lattice parameter a with unit Å. The bond lengths of W–NM, Se–NM, W–Se₁, W–Se₂, and W–Se₃ for bonds (d_W–NM, d_Se–NM, d_W–Se1, d_W–Se2, and d_W–Se3) with unit Å for the substituted and interstitial sites

	asub	aint	dW–NM	dSe1–NM
Pristine	3.298		—	—	2.596
H	3.280	3.310	2.079	1.573	2.594	2.616	2.581	2.589	2.596	2.603
B	3.296	3.309	2.174	2.104	2.618	2.618	2.589	2.597	2.588	2.602
C	3.292	3.331	2.071	1.870	2.643	2.601	2.599	2.623	2.609	2.606
N	3.286	3.309	2.051	1.882	2.636	2.596	2.603	2.599	2.620	2.591
O	3.278	3.211	2.119	1.925	2.610	2.570	2.588	2.601	2.610	2.587
F	3.287	3.211	2.313	2.482	2.596	2.581	2.575	2.588	2.594	2.578
Si	3.311	3.211	2.443	2.259	2.630	2.615	2.587	2.593	2.587	2.592
P	3.296	3.211	2.475	2.183	2.601	2.578	2.604	2.591	2.603	2.583
S	3.293	3.112	2.464	2.321	2.600	2.578	2.596	2.579	2.598	2.552
Cl	3.297	3.217	2.569	3.079	2.601	2.578	2.585	2.569	2.597	2.590
As	3.297	3.211	2.620	2.269	2.587	2.581	2.601	2.582	2.598	2.584
Br	3.301	3.211	2.708	3.212	2.600	2.586	2.589	2.586	2.599	2.581
Te	3.265	3.211	2.783	2.442	2.585	2.581	2.592	2.575	2.590	2.589
I	3.307	3.211	2.892	3.454	2.596	2.572	2.590	2.613	2.598	2.610

---

Fig. 2 Rates of bond length change of the: (a) W–Se₁ bond, (c) W–Se₂ bond, and (e) W–Se₃ bond; rates of lattice constant change of the: (b) W–Se₁ bond, (d) W–Se₂ bond, and (f) W–Se₃ bond.

After the MD simulations, most of the doped specimens were preserved even after running at 5 ps, and the configurations of the substituted and interstitial WSe₂ at 0, 1, 2, 3, 4, and 5 ps are show in Fig. 2S and 3S.† For the substituted specimens (see Fig. 3), all the dopants still occupy the substituted Se site, except for the Br, Te, and I dopants, which slightly stand out from the surface of monolayer WSe₂ due to their large dopant radius. For the interstitial specimens (see Fig. 4), all the dopants also keep in the interstitial site. However, some Se atoms in the B-, C-, Si-, S-, As-, Br-, Te-, or I-interstitial specimens also partly stand out from the surface of monolayer WSe₂, which may be caused by their large dopant radius or unstable systems obtained based on the energy below.

Fig. 3 Structure of the MD simulation for the substituted system.

Fig. 4 Structure of the MD simulation for the interstitial system.

The substituted energy E_sub and interstitial energy E_int of all the doped monolayer WSe₂ specimens were calculated by eqn (1) and the results are listed in Table 2, and where both W-rich and Se-rich conditions were considered for the substituted specimens. Under W-rich conditions, the O-, F-, S-, and Te-substituted specimens with negative E_sub values were thermodynamically stable; otherwise, the others with positive E_sub values were less stable than pristine monolayer WSe₂. Under Se-rich conditions, the O- and S-substituted specimens with negative E_sub values were still more stable than the pristine specimen, while the others with positive E_sub values were less stable than pristine monolayer WSe₂. This suggests that the preparation conditions have a great influence on the thermal stability of the substituted specimens; for instance, there are opposite E_sub values for F- and Te-substituted specimens under W-rich and Se-rich conditions. Moreover, it was found that the NM dopants substitute preferentially for Se under W-rich conditions for all substituted specimens. The O-substituted monolayer WSe₂ was the most stable due to its lowest E_sub of −2.83 eV under W-rich conditions, while the O- and S-substituted monolayer WSe₂ were more stable than the pristine specimen under both W-rich and Se-rich conditions. For the interstitial specimens, the F-, Cl-, and Br-interstitial specimens with negative E_int were thermodynamically stable, while the H-, N-, and I-interstitial specimens with small and positive E_int (0.80–1.57 eV) were thermodynamically metastable, and the others with large and positive E_int (beyond 2.37 eV) were thermodynamically unstable. Meanwhile, the substituted energies of H-, F-, Cl-, Br-, and I-doped specimens were higher than their interstitial energies, and thus, the H, F, Cl, Br, and I dopants are willing to locate at the interstitial site. Hence, we only discuss the more stable systems below.

Table 2 The energies E_sub of the substituted site under W-rich and Se-rich conditions and the energies E_int of the interstitial site with the unit eV. The bonding energies of W–NM and W–Se₂ bonds (E^bonding_W–NM and ) with the unit eV. The band gap E_g with unit eV. The energies of the Fermi level, the VBM, and the CBM (E_F, E_VBM, and E_CBM) with the unit eV. The wavelengths of the optical absorption edges λ_edge with the unit nm, and the integral area A of the optical absorption curves in visible light region with uniform unit

	E^W-richsub	E^Se-richsub	Eint	E^bondingW–NM		Eg	EF	EVBM	ECBM	λedge	A
Pristine	0.00	0.00	0.00	−5.19	—	1.75	−3.20	−3.20	−1.45	705	1471
H	1.70	2.66	1.11	—	−5.34	1.69	−2.96	−3.81	−2.11	713	1461
B	2.36	3.32	5.64	−6.25	—	1.80	−3.56	−3.84	−2.04	718	1451
C	1.97	2.93	3.57	−5.14	—	1.68	−3.21	−3.21	−1.53	737	1554
N	0.35	1.31	1.57	−6.22	—	1.79	−3.40	−3.71	−1.92	728	1517
O	−2.83	−1.87	2.64	−4.89	—	1.72	−3.00	−3.00	−1.28	699	1576
F	−0.55	0.41	−1.80	—	−5.43	1.71	−3.73	−3.83	−2.12	722	1337
Si	1.80	2.76	5.04	−4.98	—	1.68	−3.18	−3.18	−1.50	737	1531
P	0.26	1.22	2.37	−5.29	—	1.69	−3.59	−3.59	−1.90	688	1411
S	−1.24	−0.28	8.53	−5.07	—	1.74	−3.14	−3.14	−1.40	690	1504
Cl	0.58	1.54	−0.56	—	−5.31	1.72	−3.62	−3.78	−2.07	712	1399
As	0.72	1.68	4.57	−5.47	—	1.67	−3.75	−3.75	−2.08	690	1344
Br	1.04	2.00	−0.46	—	−5.54	1.77	−3.75	−3.88	−2.11	722	1391
Te	−0.74	0.22	6.96	−4.59	—	1.72	−3.04	−3.04	−1.32	700	1544
I	1.53	2.49	0.80	—	−6.00	1.71	−3.80	−3.91	−2.20	715	1354

---

Electronic performances

The calculated band structures of all the specimens are shown in Fig. 5, where the red dashed line at the zero point denotes the Fermi level (E_F), while the positive and negative values indicate the conduction band (CB) and valance band (VB), respectively. For pristine monolayer WSe₂, the Fermi level is locate at the valence band maximum (VBM) and the direct band gap (1.75 eV) is in good agreement with the experimental (1.60 eV) and other simulated (1.71 eV) results,^43,44 albeit the experimental and simulated values are both affected by the preparation conditions and calculation methods.⁴² After introducing the NM dopant, the conduction band minimum (CBM), the VBM, and the band gap values of all the doped specimens were all changed. Compared with the Fermi level of each specimen itself, the VBM is still located at the Fermi level for the C-, O-, Si-, P-, S-, As-, and Te-doped specimens, but their CBM moves down to lower energy levels, and thus their band gap values are reduced slightly (less than 0.08 eV). On the other hand, although the CBM and VBM synchronously move down to the lower energy levels for the H-, B-, N-, F-, Cl-, Br-, and I-doped specimens, the variation trends of the band gap are different. For instance, the band gap values increase less than 0.05 eV for the B-, N-, and Br-doped specimens, but decrease less than 0.06 eV for the H-, F-, Cl-, and I-interstitial specimens. Although the band gap variation is quite small (less than 0.08 eV), the result is in agreement with others experimental value of S-doped monolayer WSe₂,⁴⁵ which will be explained below. Moreover, the impurity levels in the H-, C-, N-, F-, Si-, Cl-, Br-, and I-doped specimens lie above the E_F, which can help to reduce the electronic transition energies of systems. Thus, the NM dopants affect the band gap values and electronic transition energies of systems.

Fig. 5 Band structures of pristine and NM-doped monolayer WSe₂.

To gain a detailed insight into the influence of the dopants or chemical bond relaxations on the electronic performances of monolayer WSe₂, the average PDOS value per W, Se, and NM atoms, for an equal and reasonable comparison, were calculated and are shown in Fig. 6. For pristine monolayer WSe₂ (Fig. 6a), the W-5d states interact with the Se-4p states in both VB (−6.0–0.0 eV) and CB (1.2–2.5 eV), which are similar results to the other simulation results.⁴² After introducing NM dopants, the variation of the W-5d and Se-4p states, compared to pristine monolayer WSe₂, are indistinctive, and thus the NM dopants have slight effects on the electronic structures of the W and Se host atoms. Otherwise, the W-5d states, respectively, interact with the Se-4p states in the W–Se bonds and the NM-p states in the newly formed W–NM bonds, while the Se-4p states interact with the NM-p states (or H-1s states) in the Se–NM bonds. However, the NM-p states (or H 1s states) are entirely different from each other, which means that the interaction intensities in the W–NM and Se–NM bonds are different. Moreover, an impurity level appears in some doped specimens; for instance, it is mainly made up of the W-5d states for the H-interstitial specimen, the C-2p, N-2p, and Si-2p states, respectively, for the C-, N-, and Si-substituted specimens, and the F-2p, Cl-3p, Br-4p, and I-5p states for the F-, Cl-, Br-, and I-interstitial specimens. Briefly, the electronic structures of the W–NM and Se–NM bonds are decided by the NM dopants, and the W–Se bonds of monolayer WSe₂ are broadly the same before and after doping NM atoms. Thus, the electronic performances of the doped specimens are decided by the newly formed W–NM or Se–NM bonds.

Fig. 6 Average PDOS of W, Se, and NM atoms of (a) pristine and (b) H-, (c) B-, (d) C-, (e) N-, (f) O-, (g) F-, (h) Si-, (i) P-, (j) S-, (k) Cl-, (l) As-, (m) Br-, (n) Te-, and (o) I-doped monolayer WSe₂.

Further, the electron density differences of all specimens were calculated and are exhibited in Fig. 7, where the red and blue regions indicate the electron accumulation and loss, respectively. The red areas around the Se atom and the blue regions around the W atom represent the ionic bonds between W and Se atoms in pristine monolayer WSe₂ (Fig. 7a). Moreover, the red areas also appear in the middle of W and Se atoms, which indicate the covalent bonding between them. After introducing a NM atom into the monolayer WSe₂, the electronic distributions around Se₃ and Se₄ atoms adding W–Se₃ and W–Se₄ bonds were almost invariant, while the electronic distributions around Se₂ and W–Se₂ bonds were altered slightly; however, the electronic distributions around Se₁ or NM atoms adding W–Se₁ or W–NM bonds were altered obviously. This means that the NM dopants only affected the electronic structures of the neighboring W and Se₁ or Se₂ atoms, and thus the band gap values of the doped specimens were only affected by the newly formed W–NM or Se–NM bonds. Meanwhile, the electronic distributions around the NM dopants and W–NM (or Se–NM) bonds were totally different from pristine monolayer WSe₂, while the color depth of W, Se, and NM atoms adding the W–NM and Se–NM bonds were different for each doped specimen (Fig. 7b–o). This means that the electronic performances of the doped specimens are mostly altered by the NM dopants and W–NM (or Se–NM) bonds. Thus, the bonding strength of W–NM or Se–NM bonds should be determined by NM dopants.

Fig. 7 Electron density difference of (a) pristine and (b) H-, (c) B-, (d) C-, (e) N-, (f) O-, (g) F-, (h) Si-, (i) P-, (j) S-, (k) Cl-, (l) As-, (m) Br-, (n) Te-, and (o) I-doped monolayer WSe₂, where the red and blue regions mean the electron accumulation and loss, respectively.

Optical and catalytic parameters

The optical absorption curves of all the considered specimens were calculated and are shown in Fig. 8. It was found that the wavelength of the optical absorption edge (λ_edge) was near 705 nm for pristine monolayer WSe₂ (Fig. 8a). After introducing the NM dopants, the optical absorption curves of specimens were changed. For instance, compared to pristine WSe₂, the optical absorption edges are red-shifted less than 32 nm for the H-, B-, C-, N-, F-, Si-, Cl-, Br-, and I-doped specimens but are blue-shifted less than 17 nm for the others (see Table 2). Meanwhile, the integrated area (A) of the optical absorption curves in the visible light regions are also listed in Table 2, where the results indicate that the A values reduce less than 8% for the H-, B-, F-, P-, Cl-, As-, Br-, and I-doped specimens, but increase less than 9.5% for the others. In a word, the NM dopants can affect the optical performance of monolayer WSe₂.

Fig. 8 Optical absorption curves of W, Se, and NM atoms of (a) pristine and (b) H-, (c) B-, (d) C-, (e) N-, (f) O-, (g) F-, (h) Si-, (i) P-, (j) S-, (k) Cl-, (l) As-, (m) Br-, (n) Te-, and (o) I-doped monolayer WSe₂.

The active sites of all the specimens were defined by the HOMO and LUMO levels (see Fig. 9). For pristine monolayer WSe₂, the HOMO was made up of W-5d orbits, while the LUMO was mainly composed of W-5d and Se-4p orbits, and thus the excited electrons and holes occurring on the neighboring W and Se atoms can be recombined easily, which reduces the photocatalytic active. However, the HOMO and LUMO are changed by the NM dopants, whereby the HOMO mainly locates in the dopant and its surrounding atoms, while the LUMO is far away from the dopants for the F-, P-, Cl-, As-, Br-, and I-doped specimens, and the H-doped specimen show the opposite trends, and in which the visible separation of HOMO and LUMO contributes to the separation of photogenerated e⁻/h⁺ pairs. Moreover, the HOMO and LUMO were only partly separated after substituting one Se atom by one N, O, or S atom, which may slightly enhance the separation of photogenerated e⁻/h⁺ pairs. However, both the HOMO and LUMO mainly locate in the dopant and its surrounding atoms for the B-, C-, Si-, and Te-doped specimens, and thus the dopants will become the recombination centers for photogenerated e⁻/h⁺ pairs. In brief, the H-, F-, P-, Cl-, As-, Br-, and I-doped specimens could have a high photocatalytic efficiency due to the separation of HOMO and LUMO.

Fig. 9 HOMO (pink areas) and LUMO (yellow areas) of pristine and NM-doped monolayer WSe₂. The isosurface is taken at a value of 0.003 e Bohr⁻³.

Further, the adjustment of the redox potential by the dopants was studied by experiments and by using calculation methods,^46–48 and the energies of the Fermi level, the CBM, and the VBM (E_F, E_CBM, and E_VBM) of all the specimens, compared with the vacuum level, are listed in Table 2 and shown in Fig. 10. It was found the E_CBM and E_VBM values are affected by the number of free electrons of the dopants, which denote the redox potentials of the specimens. For instance, the NM atoms with odd free electrons (one free electron for H, F Cl, Br, or I atoms, three free electrons for N, P, or As atoms, and five free electrons for the B atom) obviously reduce the reduction potential by 0.39–0.71 eV and enhance the oxidation potential by 0.45–0.75 eV. The reduced reduction potential and enhanced oxidation potential may be caused by the lone electrons in the NM atoms with one, three, or five free electrons. The one lone electron in the NM atoms can also interact with the W or Se₁ atom to enhance the interaction intensity of the W–NM or W–Se₂ bonds. The bonding energy of the W–NM or W–Se₂ bonds (E^bonding_W–NM and ) for the H-, F-, Cl-, Br-, I-, N-, P-, As-, and B-doped specimens were lower than that of pristine specimen (see Table 2), which displays its stronger interaction intensity. On the other hand, the O, S, and Te atoms with two free electrons enhance the reduction potential and reduce the oxidation potential, while the Si and C atoms with four free electrons have only slight effects on the redox potentials of monolayer WSe₂. Thus, this demonstrates that we can adjust the redox potentials of monolayer WSe₂ by introducing different kinds of NM atoms for various catalytic reactions.

Fig. 10 E_CBM (a) and E_VBM (b) of pristine and NM-doped monolayer WSe₂, as distinguished by the number of free electrons of the dopants.

Conclusions

In summary, a series of NM-doped monolayer WSe₂ was studied by first-principles calculations and the effects of the dopants on the structural, thermal, electronic, optical, and catalytic performances were investigated in detail. In terms of the structural and thermal aspects, the O-doped monolayer WSe₂ was the most stable due to its lowest energy under W-rich conditions, while the NM dopants substituted preferentially for Se under W-rich conditions. In terms of the electronic performances, the dopants affect the electronic structures with the newly formed W–NM or Se–NM bonds, and the bonding strength of the W–NM or Se–NM bonds. In terms of the optical aspect, the optical absorption edges are red-shifted or blue-shifted less than 32 nm and the optical absorption intensities are enhanced or reduced less than 9.5% for all the doped specimens. In terms of the photocatalytic active aspect, the H, F, P, Cl, As, Br, and I dopants contributed to the separation of the HOMO and LUMO, but the B, C, Si, and Te dopants became recombination centers for the photogenerated e⁻/h⁺ pairs. In terms of the redox ability aspect, the NM atoms with odd free electrons obviously reduced the reduction potential by 0.39–0.71 eV and enhanced the oxidation potential by 0.45–0.75 eV. Thus, we demonstrated that we can adjust the redox potentials of monolayer WSe₂ by introducing different kinds of NM atoms for various catalytic reactions and that the H-, F-, P-, Cl-, As-, Br-, and I-doped specimens have excellent photocatalysis capability.

Acknowledgements

This work was supported in part by the National Natural Science Foundation of China (No. 51402274, 21401180 and 51402277), the Science and Technology of Zhejiang Province for Returned Researcher and the Special Program for Applied Research on Super Computation of the NSFC-Guangdong Joint Fund (the second phase). Computational resources were provided by the Jilin University.

References

1. Z. Yin, H. Li, H. Li, L. Jiang, Y. Shi, Y. Sun, G. Lu, Q. Zhang, X. Chen and H. Zhang, ACS Nano, 2012, 6, 74–80.
2. H. S. Liu, N. N. Han and J. J. Zhao, RSC Adv., 2015, 5, 17572–17581.
3. N. Latiff, W. Z. Teo, Z. Sofer, S. Huber, A. C. Fisher and M. Pumera, RSC Adv., 2015, 5, 67485–67492.
4. P. P. Hankare, A. H. Manikshete, D. J. Sathe, P. A. Chate, A. A. Patil and K. M. Garadkar, J. Alloys Compd., 2009, 479, 657.
5. Z. Ni, H. Zhong, X. Jiang, R. Quhe, G. Luo, Y. Wang, M. Ye, J. Yang, J. Shi and J. Lu, Nanoscale, 2014, 6, 7609–7618.
6. A. Kumar and P. K. Ahluwalia, J. Alloys Compd., 2014, 587, 459–467.
7. M. Kertesz and R. Hoffmann, J. Am. Chem. Soc., 1984, 106, 3453–3460.
8. G. L. Hao, L. Z. Kou, D. L. Lu, J. Peng, J. Li, C. Tang and J. X. Zhong, J. Appl. Phys., 2016, 119, 035301.
9. N. Flöry, A. Jain, P. Bharadwaj, M. Parzefall, T. Taniguchi, K. Watanabe and L. Novotny, Appl. Phys. Lett., 2015, 107, 2123106.
10. B. Wang, S. M. Eichfield, D. Wang, J. A. Robinson and M. A. Haque, Nanoscale, 2015, 7, 14489–14495.
11. J. H. Guo, Y. T. Shi, X. G. Bai, X. C. Wang and T. L. Ma, J. Mater. Chem. A, 2015, 3, 24397–24404.
12. W. Liu, J. Kang, D. Sarkar, Y. Khatami, D. Jena and K. Banerjee, Nano Lett., 2013, 13, 1983.
13. K. Gong, L. Zhang, D. P. Liu, L. Liu, Y. Zhu, Y. H. Zhao and H. Guo, Nanotechnology, 2014, 25, 43.
14. A. Y. Gao, E. F. Liu, M. S. Long, W. Zhou, Y. Y. Wang, T. L. Xia, W. D. Hu, B. G. Wang and F. Mao, Appl. Phys. Lett., 2016, 108, 223501.
15. J. R. Mckone, R. A. Potash, F. J. Disalvo and H. D. Abruna, Phys. Chem. Chem. Phys., 2015, 17, 13984–13991.
16. X. Y. Yu, M. S. Prévot, N. Guijarro and K. Sivula, Nat. Commun., 2015, 6, 7596.
17. B. Amin, T. P. Kaloni and U. Schwingenschlogl, RSC Adv., 2014, 4, 34561–34565.
18. S. F. Wang, W. J. Zhao, F. Giustiniano and G. Eda, Phys. Chem. Chem. Phys., 2016, 18, 4304–4309.
19. P. D. Zhao, D. Kiriya, A. Azcatl, C. X. Zhang, M. Tosun, Y.-S. Liu, M. Hettick, J. S. Kang, S. McDonnell, J. H. Guo, K. Cho, R. M. Wallace and A. Javey, ACS Nano, 2014, 8(10), 10808–10814.
20. Y. J. Zheng, Y. L. Huang, Y. F. Cheng, W. J. Zhao, G. Eda, C. D. Spataru, W. J. Zhang, Y.-H. Chang, L.-J. Li, D. Z. Chi, S. Y. Quek and A. T. S. Wee, ACS Nano, 2016, 10(2), 2476–2484.
21. Y. Guo, D. Liu and J. Robertson, Appl. Phys. Lett., 2015, 106, 173106.
22. S. J. Wi, M. K. Chen, D. Li, H. Nam, E. Meyhofer and X. G. Liang, Appl. Phys. Lett., 2015, 107, 062102.
23. C.-H. Chen, C.-L. Wu, J. Pu, M.-H. Chiu, P. Kumar, T. Takenobu and L.-J. Li, 2D materials., 2014, 1, 3.
24. M. Kriener, A. Kikkawa, T. Suzuki, R. Akashi, R. Arita, Y. Tokura and Y. Taguchi, Phys. Rev. B: Condens. Matter Mater. Phys., 2015, 91, 075205.
25. A. Allain and A. Kis, ACS Nano, 2014, 8(7), 7180–7185.
26. Y. Zhang, M. M. Ugeda, C. H. Jin, S. F. Shi, A. J. Bradley, H. Ryu, S. J. Tang, B. Zhou, F. Wang, M. F. Crommie, Z. Hussain, Z. X. Shen and S. K. Mo, Nano Lett., 2016, 16(4), 2485–2491.
27. J. J. Zhu and U. Schwingenschlögl, J. Mater. Chem. C, 2015, 3, 3946.
28. G. J. Gil, A. Pham, A. B. Yu and S. Li, J. Phys.: Condens. Matter, 2014, 26, 30.
29. A. A. Mitioglu, P. Plochocka, A. Granados del Aguila, P. C. M. Christianen, G. Deligeorgis, S. Anghel, L. Kulyuk and D. K. Maude, Nano Lett., 2015, 15(7), 4387–4392.
30. H. T. Wang, D. S. Kong, P. Johanes, J. J. Cha, G. Y. Zheng, K. Yan, N. Liu and Y. Cui, Nano Lett., 2013, 13(7), 3426–3433.
31. S. Y. Hu, M. C. Cheng, K. K. Tiong and Y. S. Huang, J. Phys.: Condens. Matter, 2005, 17, 23.
32. J. Huang, W. H. Wang, Q. Fu, L. Yang, K. Zhang, J. Y. Zhang and B. Xiang, Nanotechnology, 2016, 27, 23.
33. M. L. Zou, J. D. Chen, L. F. Xiao, H. Zhu, T. T. Yang, M. Zhang and M. L. Du, J. Mater. Chem. A, 2015, 3, 18090–18097.
34. Y. F. Zhao, C. Li, Y. Y. Gong, T. Wang and C. Q. Sun, Phys. Chem. Chem. Phys., 2014, 16, 21446.
35. M. D. Segall, P. J. D. Lindan, M. J. Probert, C. J. Pickard, P. J. Hasnip, S. J. Clark and M. C. Payne, J. Phys.: Condens. Matter, 2002, 14, 2717.
36. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865.
37. D. H. Hamalm, M. Schluter and C. Chiang, Phys. Rev. Lett., 1979, 43, 1494.
38. G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 47, 558.
39. C. Su, H. Jiang and J. Feng, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 075453.
40. A. H. Reshak and S. Auluck, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 68, 195107.
41. N. D. Boscher, C. J. Carmalt and I. P. Parkin, J. Mater. Chem., 2006, 16, 122.
42. L.-P. Feng, J. Su and Z.-T. Liu, J. Alloys Compd., 2015, 639, 210–214.
43. A. Kumara and P. K. Ahluwalia, Eur. Phys. J. B, 2012, 85, 186.
44. H. Zeng, G.-B. Liu, J. Dai, Y. Yan, B. Zhu, R. He, L. Xie, S. Xu, X. Chen, W. Yao and X. Cui, Sci. Rep., 2013, 3, 1608.
45. J. Huang, W. H. Wang, Q. Fu, L. Yang, K. Zhang, J. Y. Zhang and B. Xiang, Nanotechnology, 2016, 27, 13.
46. T. Shiraishi, G. Juhász, T. Shiraki, N. Akizuki, Y. Miyauchi, K. Matsyda and N. Nakashima, J. Phys. Chem. C, 2016, 120(29), 15632.
47. S. Lu, C. Li, Y. F. Zhao, Y. Y. Gong, L. Y. Niu and X. J. Liu, Appl. Surf. Sci., 2016, 384, 360–367.
48. S. Lu, Z. W. Chen, C. Li, H. H. Li, Y. F. Zhao, Y. Y. Gong, L. Y. Niu, X. J. Liu, T. Wang and C. Q. Sun, J. Mater. Chem. A, 2016, 4, 14827–14838.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra15697e

---