REVIEW 3 major objections 5 minor 31 references
Effect of arsenic doping on structural and electronic properties of MoSe$_2$ monolayer: an ab initio study
T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash
Pith's one-line read Arsenic substituting for molybdenum in a MoSe2 monolayer is energetically favorable, widens the band gap to 1.73 eV, and shifts the Fermi level toward the valence band, indicating p-type behavior.
desk verdict Routine but useful DFT data point on As-doped MoSe2; the reported numbers are unreliable because the odd-electron defect cells were apparently run spin-unpolarized. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The analysis rests on three density-functional-theory supercell models built from a 4×4×1 repeat of the MoSe2 unit cell with a 15 Å vacuum along the out-of-plane direction: a Mo vacancy, an As substitutional at that vacancy, and an As interstitial in the hollow center of a ring. The formation energy is evaluated with chemical potentials fixed by Se-rich conditions through the formula $E_f = E_{\rm def} - E_{\rm pure} - \sum_i n_i \mu_i$, and the electronic structure is interpreted through band structures plus total and per-atom projected densities of states, which assign each mid-gap level to Mo-4d, Se-4p, and As-4s/4p orbitals.
What would settle it
A converged supercell-size series (3×3×1, 4×4×1, and 5×5×1 with the same 15 Å vacuum) recomputing the As(Mo) formation energy and band gap would settle the central claim: if 1.839 eV or 1.73 eV shifts by more than about 0.1 eV between cell sizes, the favourability and the gap-widening are artifacts of the chosen supercell rather than properties of the isolated defect.
Extended reading notes
Core claim
The paper's central claim is that arsenic sitting on a molybdenum site in a MoSe2 monolayer is likely to form under Se-rich conditions and improves the material's electronic profile: the formation energy is 1.839 eV, the band gap widens to 1.73 eV from the pristine 1.41 eV, and the Fermi level moves toward the valence band, making the system a p-type semiconductor. The authors interpret the mid-gap states as a mixture of the vacancy effect and the arsenic atom itself, since the smaller As atom leaves some Mo-4d and Se-4p orbitals non-bonding while adding As-4s character. They further claim that the As interstitial is metastable ($E_f = 17.422$ eV) and n-type, with the Fermi level shifted toward the conduction band and a gap of 1.5 eV.
Load-bearing premise
The load-bearing assumption is that the adopted 4×4×1 supercell with a 15 Å vacuum is large enough to isolate each defect, but the paper reports no convergence test against cell size or vacuum thickness and contains inconsistent atom counts (stated 48 atoms versus 36 Se plus 16 Mo), so defect-defect interactions could change the formation energies and mid-gap levels.
Editorial extensions
If this is right
- Under Se-rich growth conditions, arsenic substituting for molybdenum in MoSe2 is energetically favorable, with a formation energy of 1.839 eV, making it a realistic doping route.
- The As(Mo) system has a band gap of 1.73 eV, larger than the pristine 1.41 eV, and its Fermi level shifts toward the valence band, indicating p-type semiconductor behavior.
- The mid-gap defect levels in the As(Mo) system are traced to Mo-4d, Se-4p, and As-4s orbitals, consistent with a combination of the vacancy effect and the arsenic atom.
- The As interstitial is metastable with a formation energy of 17.422 eV yet shifts the Fermi level toward the conduction band, suggesting n-type behavior with a 1.5 eV gap.
- If these electronic changes hold, the As(Mo) system should absorb in the visible range, making it a candidate for photocatalysis and photovoltaics, while the interstitial system could serve infrared electronics.
Reading between the lines
- The paper's stated atom count (48 atoms) does not match its breakdown (36 Se + 16 Mo = 52), suggesting the simulated cell may differ from the described one; if so, the nominal 6.25% As-doping concentration and possibly the defect-level positions would need revision.
- Because PBE-GGA is known to underestimate band gaps, the absolute values reported here are probably lower bounds; hybrid or GW calculations would test whether the ordering of gaps across the four systems survives.
- The p-type assignment for the As(Mo) system could be checked experimentally by Hall-effect or Seebeck measurements on arsenic-doped MoSe2: a positive Hall coefficient would support the Fermi-level placement, while n-type conduction would refute it.
- The mid-gap defect levels that widen the gap may also act as non-radiative recombination centers, which could lower carrier lifetimes and reduce the predicted photocatalytic efficiency; time-resolved photoluminescence on As-doped samples would be a direct test.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports PBE-GGA DFT calculations with Quantum Espresso of pristine MoSe2 monolayers and three defect systems: an Mo vacancy, As substitutional on the Mo site (As(Mo)), and an As interstitial. Formation energies are computed in Se-rich conditions via Eq. (3.1), and band structures, DOS, and PDOS are used to identify mid-gap defect levels and to assign p-type (Mo vacancy, As(Mo)) and n-type (As interstitial) behavior, with band gaps between 1.5 eV and 1.73 eV. The central claims are that As(Mo) is a favorable configuration (Ef = 1.839 eV) and that it could be useful for visible-light photocatalysis and photovoltaic applications, while the As interstitial is metastable (Ef = 17.422 eV).
Significance. If the results are correct, the paper would provide a straightforward DFT characterization of As-doped MoSe2, a material combination not widely studied, and would offer a concrete prediction that As(Mo) substitution is energetically accessible under Se-rich growth and induces p-type character with a widened gap. The work has no fitted parameters and the formation-energy trend is internally consistent with the stated method. However, the reliability of the quantitative claims is presently undermined by the absence of spin-polarized treatment of odd-electron defect cells, by an internal inconsistency in the reported supercell composition, and by the absence of convergence checks; these issues are load-bearing for the formation energies and the band-gap/Fermi-level assignments. The paper is reportable but needs substantial revision before its main conclusions can be accepted.
major comments (3)
- [Section 2] The neutral As(Mo) and As-interstitial supercells contain an odd number of valence electrons (a 4x4x1 stoichiometric MoSe2 cell has an even number; replacing Mo with As removes six and adds five electrons, and interstitial As adds five), yet no spin polarization, magnetization, or smearing/occupation scheme is reported. The band structure in Figure 4(a) shows a defect level exactly at 0 eV (the Fermi level), which is the textbook signature of a half-occupied Kohn-Sham state produced by a restricted, spin-unpolarized calculation. With spin polarization this level should split into occupied and unoccupied states, lowering the total energy and shifting both the apparent band gap and the Fermi level. Consequently, the reported formation energies (1.839 eV and 17.422 eV) and the p-/n-type assignments are not established as reported; the authors must either repeat the calculations spin-polarized or explicitly justify that spin polarization does not change the conclusions.
- [Section 2] The supercell composition is stated inconsistently: the computational-details text says a 4x4x1 cell has 48 atoms, 'of which 36 are Se atoms and 16 are Mo atoms' (totaling 52), while the PDOS discussion later says the As(Mo) cell contains 15 Mo, 32 Se, and 1 As (totaling 48). A 4x4x1 pristine MoSe2 supercell is 16 Mo + 32 Se = 48 atoms. This inconsistency affects the defect concentration, the odd-electron counting argument, and the interpretation of the PDOS, and it must be corrected.
- [Section 2] No convergence tests are reported for the supercell size, k-point mesh, plane-wave cutoff, or vacuum thickness. The 4x4x1 cell and 15 Å vacuum are simply stated. For defect formation energies and mid-gap levels, finite-size interactions between periodic images can be sizable, and for a 17.422 eV interstitial formation energy the result may be particularly sensitive to the cell size. The manuscript should include at least a demonstration that the formation energy and the defect-level positions are converged with respect to cell size and vacuum thickness.
minor comments (5)
- [Keywords] 'transient metal dichalcogenides' should be 'transition metal dichalcogenides'.
- [Section 2] The text states that chemical potentials are from 'the bulk bcc phase of Mo and Se'; elemental Se is not bcc under standard conditions, so the reference phase should be specified accurately.
- [Figure 4(b)] The legend labels an 'As-3d' contribution, but arsenic has no 3d electrons in its valence manifold; this is presumably a typo for As-4p or As-4s.
- [Figure 1 caption] The caption lists '(g) and (f) top and side views of MoSe2 with As interstitial'; the duplicate '(f)' should be corrected to '(h)'.
- [Section 4] The defect levels for the As(Mo) system are described as 'a pair of defect levels' and then as 'three different defect levels' (at 0 eV, +0.35 eV, and -0.35 eV); this wording should be made consistent.
Circularity Check
No significant circularity: DFT outputs are computed from independent total energies, not fitted to the conclusions.
full rationale
The paper's central outputs—formation energies, band structures, DOS/PDOS, band gaps, and Fermi-level positions—are direct outputs of converged PBE-GGA DFT total-energy and electronic-structure calculations in Quantum Espresso. The formation-energy formula (Eq. 3.1) is a standard definition that combines independently computed supercell total energies with bulk chemical potentials; it does not contain the claimed relative stability or the reported band gap as an input. The p-/n-type assignments are interpretations of the computed Fermi-level position in the resulting band structures, not quantities fitted to those conclusions. No load-bearing argument rests on a citation to the authors' own prior work; the references are to independent experimental and computational studies. The atom-count inconsistency and the absence of an explicitly reported spin-polarization treatment are potential correctness or reproducibility concerns, but they are not instances of a derivation reducing to its own inputs. The derivation is self-contained against first-principles inputs, so no circular step is found.
Assumptions & free parameters
assumptions (4)
- domain assumption DFT with PBE-GGA accurately describes the relative energetics of defects in MoSe2
- domain assumption The 4x4x1 supercell and 15 Å vacuum are sufficient to isolate the defect and avoid spurious interactions
- domain assumption The chemical potentials from bulk Mo, Se, and As phases, and the Se-rich condition, are appropriate references for formation energies
- domain assumption The Fermi level position in the supercell is a reliable indicator of p- or n-type doping character
Cite this review
Pith. "Pith review of Effect of arsenic doping on structural and electronic properties of MoSe$_2$ monolayer: an ab initio study." pith.science (2026). https://pith.science/paper/5F6DDOBG
@misc{pith2026241217664,
author = {Pith},
title = {Pith review of: Effect of arsenic doping on structural and electronic properties of MoSe$_2$ monolayer: an ab initio study},
year = {2026},
howpublished = {\url{https://pith.science/paper/5F6DDOBG}},
note = {Machine review of arXiv:2412.17664}
}
abstract
In this paper, we studied the structural and electronic properties of MoSe$_2$ monolayer in its pure and doped forms, using the density functional theory (DFT), and the calculations were performed using Quantum Espresso (QE) software package. The doped systems are a MoSe$_2$ monolayer with a vacancy in Mo site (Mo vacancy system), the MoSe$_2$ monolayer with an As atom as substitutional for Mo atom (As(Mo) doped system) and an MoSe$_2$ monolayer with As atom in an interstitial site in the hollow location of the center of one ring of the structure between the plane Mo atoms and the plane containing Se atoms (As interstitial system). We calculated the formation energy of various structures studied in Se-rich condition. We found that the As(Mo) doped system is a favorable configuration, whereas the As interstitial system is metastable. Different defects introduce midgap levels, which were interpreted according to the orbitals involved in their formation using the analysis of the band structure and DOS and PDOS of each system. The energy gap increases in all system structures and its value ranged between 1.5 eV and 1.73 eV, the Fermi level shifts toward the valence band for the Mo vacancy system, and As(Mo) doped system which suggests that it can be a $p$-type semiconductor, whereas Fermi level shifts to the conduction band for As interstitial system and suggests a $n$-type semiconductor behavior. The obtained results enable us to predict the possibility of using these systems in many applications, since it can be used in the As(Mo) doped system in photocatalysis or in photovoltaic applications in the visible light, and As interstitial system can be used in electronics applications in the infrared field.
Figures
Figures from the paper (4 more)
Reference graph
Works this paper leans on
-
[1]
Neto A. C., Guinea F., Peres N. M., Novoselov K. S., Geim A. K., Rev. Mod. Phys., 2009,81, No. 1, 109, doi:10.1103/RevModPhys.81.109
-
[2]
K., Eriksson O., Sanyal B., Phys
Haldar S., Vovusha H., Yadav M. K., Eriksson O., Sanyal B., Phys. Rev. B, 2015,92, No. 23, 235408, doi:10.1103/PhysRevB.92.235408
-
[3]
PaulJ.,SinghA.,DongZ.,ZhuangH.,RevardB.,RijalB.,AshtonM.,LinscheidA.,BlonskyM.,GluhovicD., J. Phys.: Condens. Matter, 2017,29, No. 47, 473001, doi:10.1088/1361-648X/aa9305
-
[4]
Bolotin K. I., Sikes K. J., Hone J., Stormer H., Kim P., Phys. Rev. Lett., 2008, 101, No. 9, 096802, doi:10.1103/PhysRevLett.101.096802
-
[5]
Nova, 2022,45, 654–658, doi:10.21577/0100-4042.20170864
Luo Y., Liu Q., Yang L., Yan Y., Quim. Nova, 2022,45, 654–658, doi:10.21577/0100-4042.20170864
-
[6]
MaJ.-J.,ZhengJ.-J.,LiW.-D.,WangD.-H.,WangB.-T.,Phys.Chem.Chem.Phys.,2020, 22,No.10,5832–5838, doi:10.1039/D0CP00047G
-
[7]
Menezes M. G., Ullah S., Phys. Rev. B, 2021,104, No. 12, 125438, doi:10.1103/PhysRevB.104.125438
-
[8]
Nair R. R., Blake P., Grigorenko A. N., Novoselov K. S., Booth T. J., Stauber T., Peres N. M., Geim A. K., Science, 2008,320, No. 5881, 1308–1308, doi:10.1126/science.1156965
Show all 31 references
-
[9]
A., Vancsó P., Szendro M., Dobrik G., Antognini Silva D., Popov Z
Koós A. A., Vancsó P., Szendro M., Dobrik G., Antognini Silva D., Popov Z. I., Sorokin P. B., Hen- rard L., Hwang C., Biró L. P., Tapasztó L., J. Phys. Chem. C, 2019, 123, No. 40, 24855–24864, doi:10.1021/acs.jpcc.9b05921
2019 doi
-
[10]
14, 1704150, doi:10.1002/smll.201704150
Gao D., Xia B., Wang Y., Xiao W., Xi P., Xue D., Ding J., Small, 2018, 14, No. 14, 1704150, doi:10.1002/smll.201704150
2018 doi
-
[11]
V., Tominaga J., From 3D to 2D: Fabrication Methods, Springer International Publishing, Cham, 2016, 79–107, doi:10.1007/978-3-319-31450-1_4
Kolobov A. V., Tominaga J., From 3D to 2D: Fabrication Methods, Springer International Publishing, Cham, 2016, 79–107, doi:10.1007/978-3-319-31450-1_4
2016 doi
-
[12]
51–52, 2743–2757, doi:10.1557/adv.2019.391
Zhang K., Robinson J., MRS Adv., 2019,4, No. 51–52, 2743–2757, doi:10.1557/adv.2019.391
2019 doi
-
[13]
I., Phys
Ayesh A. I., Phys. Lett. A, 2022,422, 127798, doi:10.1016/j.physleta.2021.127798
2022
-
[14]
H., Emerging trends in nanotechnology, Springer, 2021, doi:10.1007/978-981-15-9904-0
Khan Z. H., Emerging trends in nanotechnology, Springer, 2021, doi:10.1007/978-981-15-9904-0
2021 doi
-
[15]
Liang B., Li W., Ren Q., Zhu C., Li J., Results Phys., 2022,42, 105978, doi:10.1016/j.rinp.2022.105978
2022
-
[16]
4, 616–619, doi:10.1109/LED.2019.2901296
Zhang D., Li Q., Li P., Pang M., Luo Y., IEEE Electron Device Lett., 2019, 40, No. 4, 616–619, doi:10.1109/LED.2019.2901296
2019
-
[17]
Yang J., Zhu J., Xu J., Zhang C., Liu T., ACS Appl. Mater. Interfaces, 2017,9, No. 51, 44550–44559, doi:10.1021/acsami.7b15854
2017 doi
-
[18]
He J., Liu G., Zhang C., Wang Y., Zhang G., Micro Nanostruct., 2023, 180, 207612, doi:10.1016/j.micrna.2023.207612
2023
-
[19]
Rep., 2017,7, No
Zhao Y., Wang W., Li C., He L., Sci. Rep., 2017,7, No. 1, 17088, doi:10.1038/s41598-017-17423-w
2017 doi
-
[20]
L., Cococcioni M., Dabo I., et al., J
Giannozzi P., Baroni S., Bonini N., Calandra M., Car R., Cavazzoni C., Ceresoli D., Chiarotti G. L., Cococcioni M., Dabo I., et al., J. Phys.: Condens. Matter, 2009,21, No. 39, 395502, doi:10.1088/0953- 8984/21/39/395502
2009 doi
-
[21]
PerdewJ.P.,BurkeK.,ErnzerhofM.,Phys.Rev.Lett.,1996, 77,3865–3868,doi:10.1103/PhysRevLett.77.3865
1996 doi
-
[22]
J., Pack J
Monkhorst H. J., Pack J. D., Phys. Rev. B, 1976,13, 5188–5192, doi:10.1103/PhysRevB.13.5188
1976 doi
-
[23]
D., PNAS, 1944,30, No
Murnaghan F. D., PNAS, 1944,30, No. 9, 244–247, doi:10.1073/pnas.30.9.244
1944 doi
-
[24]
R., Ziegen- rücker R., Kentsch U., Krasheninnikov A
Prucnal S., Hashemi A., Ghorbani-Asl M., Hübner R., Duan J., Wei Y., Sharma D., Zahn D. R., Ziegen- rücker R., Kentsch U., Krasheninnikov A. V., Helm M., Zhou S., Nanoscale, 2021,13, No. 11, 5834–5846, doi:10.1039/D0NR08935D. 43602-9 B. Bradji, M. L. Benkhedir
2021 doi
-
[25]
K., Jasil T., Kiran G., Singh R., Singh A
Vinturaj V., Yadav A. K., Jasil T., Kiran G., Singh R., Singh A. K., Garg V., Pandey S. K., Bull. Mater. Sci., 2023, 46, No. 3, 121, doi:10.1007/s12034-023-02963-x
2023 doi
-
[26]
Kumar A., Ahluwalia P., Eur. Phys. J. B, 2012,85, 1–7, doi:10.1140/epjb/e2012-30070-x
2012 doi
-
[27]
Ramasubramaniam A., Phys. Rev. B, 2012,86, 115409, doi:10.1103/PhysRevB.86.115409
2012 doi
-
[28]
Nanotechnol., 2014,9, No
Zhang Y., Chang T.-R., Zhou B., Cui Y.-T., Yan H., Liu Z., Schmitt F., Lee J., Moore R., Chen Y., Lin H., Jeng H.-T., Mo S.-K., Hussain Z., Bansil A., Shen Z.-X., Nat. Nanotechnol., 2014,9, No. 2, 111–115, doi:10.1038/nnano.2013.277
2014 doi
-
[29]
O., Li J., Zhang D., Arora S
Zhao Y., Ren Y., Coileain C. O., Li J., Zhang D., Arora S. K., Jiang Z., Wu H.-C., Appl. Surf. Sci., 2021,564, 150399, doi:10.1016/j.apsusc.2021.150399
2021
-
[30]
Appl., 2019,3, No
Zhong M., Shen C., Huang L., Deng H.-X., Shen G., Zheng H., Wei Z., Li J., npj 2D Mater. Appl., 2019,3, No. 1, 1, doi:10.1038/s41699-018-0083-1
2019 doi
-
[31]
Balasubramaniam B., Singh N., Kar P., Tyagi A., Prakash J., Gupta R. K., Mol. Syst. Des. Eng., 2019,4, 804–827, doi:10.1039/C8ME00116B. Вплив легування миш’яком на структурнi та електроннi властивостi моношару MoSe2: першопринципнi дослiдження Б. Браджi, М. Л. Бенхедiр Лаборат...
2019 doi
Reviewed August 11, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.