{"id":"4eecd965-3015-404f-956f-b39c5eb53625","arxiv_id":"1910.05213","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A wavefunction projection diabatization method, using isolated monolayer states as references in a plane-wave basis, produces a diabatic Hamiltonian for MoS2/WS2 with a strong 0.330 eV coupling at Γ and weak 0.043 eV coupling at K.","lead":"This paper develops a way to build diabatic Hamiltonians for van der Waals heterostructures using plane-wave DFT, by projecting complex states onto isolated monolayer states. The method gives interlayer electronic couplings for MoS2/WS2, which are needed for quantum dynamics simulations of ultrafast charge transfer.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reported 0.330 eV Γ-point coupling is an unbenchmarked property of a Löwdin-rotated Kohn–Sham subspace; the paper never shows that this equals the physical hole-transfer coupling, and the K-point value may be sensitive to degenerate-state rotations.","rationale":"Read in good faith: the paper introduces a practical projection diabatization for plane-wave DFT and demonstrates it on MoS2/WS2, with convergence tests across 6×6 and 9×9 supercells and varying band numbers. Those checks support internal consistency but not external validity. The load-bearing step is the identification of the rotated Kohn–Sham Hamiltonian with the diabatic Hamiltonian for hole transfer. That identification is standard in the single-particle picture but is not benchmarked; CDFT is the natural reference because it variationally optimizes charge-localized states and yields total-energy couplings. The authors themselves call for such benchmarks (ref. 87) and acknowledge the missing excitonic effects. The reader's conditional verdict is appropriate; my concern strengthens the condition: the numerical coupling should be validated against CDFT before being used in dynamics. I therefore recommend keeping the reader's CONDITIONAL verdict, i.e., no change to the verdict.","tokens_in":15119,"tokens_out":9065,"duration_ms":103272,"concrete_test":"Perform CDFT (plane-wave PAW, same PBE+D2 and 6×6 supercell) constraining one hole on MoS2 and one on WS2, and compute the total-energy diabatic coupling using the Oberhofer–Blumberger CDFT formula. If the resulting Γ-point coupling differs from 0.330 eV by more than about 0.05 eV, the projection value is not the physical hole-transfer coupling and the central claim needs revision. Also report the singular values of the K-point 2×2 coupling block; if the largest and smallest singular values differ materially, the quoted 0.043 eV is basis-dependent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim (0.330 eV at Γ, 0.043 eV at K) is obtained by Eqs. (2)-(5): the complex's Kohn–Sham adiabatic states are projected onto isolated-monolayer Kohn–Sham reference orbitals, and the resulting overlap matrix is Löwdin-orthonormalized before rotating the adiabatic Hamiltonian. For this to be the diabatic coupling relevant to interlayer hole transfer, two conditions must hold: (i) the monolayer-reference manifold must span the hole-transfer active space of the complex, and (ii) the rotated single-particle Hamiltonian must represent the physical coupling between charge-localized states. Condition (i) is only tested by adding more bands (Figure S4, 9×9 supercell); condition (ii) is not tested at all. The references are neutral-monolayer ground-state orbitals, so they carry no orbital relaxation or density response to a localized hole; the diagonal energies obtained by rotation are Kohn–Sham eigenvalues, not total energies of charge-localized states. The paper itself flags the need for future benchmarks (ref. 87) and notes the chemically bonded case would require localized-orbital construction. For vdW systems it simply asserts that monolayer states suffice. The K-point value is likewise quoted as a single number even though the two degenerate K/K' VB states make the off-diagonal block ambiguous unless singular values are specified. The convergence checks therefore do not establish that 0.330 eV is the physical coupling. This is a correctness risk, not an internal inconsistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a wavefunction-projection diabatization scheme for van der Waals heterostructures computed with plane-wave DFT and the PAW method. The adiabatic Kohn-Sham states of the MoS2/WS2 complex are projected onto reference states built from the Kohn-Sham orbitals of the isolated monolayers; the resulting overlap matrix is Löwdin-orthonormalized and used to rotate the adiabatic Hamiltonian into a diabatic representation. The authors report diabatic state energies and couplings for the valence-band manifold, in particular a coupling of 0.330 eV between MoS2 and WS2 VB@Γ states and 0.043 eV between the VB@K states, and use the former to argue for ultrafast hole transfer with a Rabi-type oscillation period of 24.3 fs.","tokens_in":15512,"tokens_out":5224,"duration_ms":55803,"significance":"If the reported couplings are physically meaningful, the method would provide a practical, parameter-free route to diabatic Hamiltonians for photoinduced charge transfer in vdW heterostructures using only plane-wave electronic-structure output. The approach is transparent, is based on a standard projection-diabatization logic, and yields the expected physical trend (strong interlayer hybridization at Γ, weak at K), consistent with prior band-structure analyses. The paper also performs convergence checks on supercell size and the number of included bands. However, the central quantitative claim—that the rotated Kohn-Sham matrix elements equal the hole-transfer couplings—is not benchmarked against other diabatization schemes, and at least one derived dynamical consequence appears arithmetically inconsistent. These issues must be resolved before the numbers can be used as quantitative predictions.","major_comments":[{"comment":"The paper's central quantitative claim, that the diabatic coupling is 0.330 eV at Γ and 0.043 eV at K, is never validated as a physical charge-transfer coupling. These values are matrix elements of the Kohn-Sham Hamiltonian in a Löwdin-orthonormalized projection of neutral-monolayer Kohn-Sham orbitals; no comparison is made with other diabatization methods (e.g., CDFT, FODFT, or generalized Mulliken-Hush), nor with values extracted from experimental transfer rates. The convergence checks in Figure S4 address the stability of diabatic energies with respect to the number of bands, but not the correctness of the coupling as a descriptor of hole transfer. Because the abstract and the dynamical conclusions rely on these two numbers, the manuscript should either add such a benchmark or substantially soften the quantitative claims.","section":"Eqs. (2)-(5) and Fig. 3"},{"comment":"The reported K-point coupling of 0.043 eV is not well-defined as stated. The diabatic states D2/D3 (MoS2_VB@K) and A1/A2 (WS2_VB@K) are degenerate pairs, and the off-diagonal block of the diabatic Hamiltonian is not invariant under unitary rotations within each degenerate subspace. The paper reports a single number without specifying whether it is the largest singular value, an average over the degenerate manifold, or a particular matrix element after a chosen canonical rotation. This ambiguity is especially relevant because the K-point value is used as evidence of weak interlayer coupling. Please specify the convention and report the full 2×2 off-diagonal block or the singular values.","section":"Fig. 3 and the discussion of VB@K states"},{"comment":"The stated Rabi oscillation period of 24.3 fs appears inconsistent with the quoted coupling of 0.330 eV. For a two-state system with off-diagonal coupling V and diagonal energy difference Δ, the population oscillation period is h/sqrt(Δ² + 4V²), which for any Δ is at most h/(2V) ≈ 6.3 fs when V = 0.330 eV. A period of 24.3 fs corresponds to an energy scale of about 0.170 eV, not 0.660 eV. Either the formula used for the period, the value of the coupling, or the energy difference entering the two-state model is misreported, and the text should be corrected.","section":"Rabi period estimate (text near Fig. 4)"},{"comment":"The paper asserts that monolayer ground-state Kohn-Sham orbitals are adequate reference states for vdW complexes, but the only support is that adding more bands leaves the frontier diabatic energies approximately unchanged. This tests the completeness of the active space, not the appropriateness of the reference states for describing a localized hole. The references are neutral-monolayer orbitals and carry no orbital relaxation or density response to a hole, and the diagonal diabatic energies are Kohn-Sham eigenvalues rather than total energies of charge-localized states. The authors themselves note that chemically bonded cases require localized-orbital construction; the vdW case is asserted without an analogous test. A concrete check would be a CDFT calculation of the hole-localized states in the same plane-wave framework, or a comparison of the projected diabatic state densities with the self-consistent layer-projected densities of the complex.","section":"Discussion of limitations (p. 18-19)"}],"minor_comments":[{"comment":"The sentence beginning 'projection-operator approach and block diago nalization of Fock matrix, constrained density functional theory ( CDFT), molecular orbital based fragmentatio n approaches As pointed out...' appears to be a leftover fragment from the introduction and should be removed or rewritten.","section":"p. 18, first paragraph of the discussion section"},{"comment":"There are several typographical errors, including 'Femi level' (Figures 1 and 2 captions), 'Hatree' in 'multilayer multiconfiguratio n time-dependent Hatree (ML-MCTDH)', and inconsistent spacing in words like 'diago nalization' and 'fragmentatio n'. These should be corrected before publication.","section":"Throughout"},{"comment":"The statement 'All calculations were performed at the Γ point' is potentially confusing because the text discusses VB@K states. In the 6×6 and 9×9 supercells the primitive-cell K point folds to Γ, so this is consistent, but the manuscript should explain this explicitly so that readers do not misinterpret the k-point sampling.","section":"Computational Details"},{"comment":"The definition of the projection matrix t_ij would be clearer if the bra-ket notation and the PAW overlap correction in Eq. (9) were explicitly connected to the Hermitian conjugate of the overlap matrix, especially since the PAW transformation is not unitary in the pseudo-wavefunction space.","section":"Eq. (3) and surrounding text"},{"comment":"The caption lists donor and acceptor labels but does not explain how the reported 'absolute values of diabatic couplings' were extracted from the diabatic Hamiltonian matrix; a brief statement of the convention (e.g., the matrix element of the rotated Hamiltonian, or the singular value of the off-diagonal block) would help.","section":"Figure 3 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is a reasonable application of a standard projection-diabatization idea to a periodic plane-wave framework, and the method itself is sound in its derivation. My main concerns are (1) the lack of any benchmark of the coupling values against other diabatization approaches, which is a load-bearing issue for the paper's quantitative claims, and (2) the apparent arithmetic error in the Rabi period, which should be corrected. The K-point degeneracy issue also needs a clear convention. These are fixable but require additional work, hence major_revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Take this one seriously: it's a practical extension of a well-established diabatization scheme to plane-wave DFT for vdW heterostructures, and the demonstration on MoS2/WS2 gives couplings that match the known physics. The method is not conceptually new, but the implementation is genuinely useful for a class of systems where people currently get couplings from PDOS or band inspection.\n\nThe paper does well in several places. The PAW overlap correction in Eq. (9) is the right way to handle wavefunction projection in a plane-wave basis. The convergence tests with 6x6 vs 9x9 supercells and the number of bands included are appropriate. Most importantly, the authors don't oversell: they explicitly say the model is single-particle, that excitonic effects are missing, and that the chemically bonded case needs localized orbitals. The qualitative result, strong Gamma coupling and weak K coupling, is consistent with orbital character and previous dynamics studies. No parameters are fitted; the only inputs are the Kohn-Sham states and monolayer references, so the circularity burden is low.\n\nThe soft spots are real, though not fatal. The 0.330 eV coupling is a property of a Löwdin-rotated Kohn-Sham subspace, and the paper does not demonstrate that this equals the physical hole-transfer matrix element. The references are neutral monolayer orbitals, so there's no relaxation of the hole density. A benchmark against CDFT or a fragment-orbital approach for a simpler system would go a long way. The paper flags this as future work (ref 87), which is honest, but for a methods paper I'd want at least one comparison. The K/K' degeneracy is also glossed over: the off-diagonal between two degenerate states is not uniquely defined unless you specify how the rotation is resolved. The reader should ask for a statement about singular values or an averaged coupling. Minor issue, but easy to fix.\n\nThe other concern is reproducibility. The PIETAS code is on GitHub, but no raw input files or raw coupling values are deposited. For a method paper, that's a meaningful omission.\n\nBottom line: this is a solid, honest methods paper that deserves a serious referee. I'd accept it for review and ask for a bit more on degenerate-state handling and reproducibility. It's not a paradigm shift, but it fills a practical gap, and the citation pattern is appropriate.","headline":"Practical projection diabatization for vdW heterostructures; solid method paper with honest limitations, but coupling numbers need a benchmark and the K-point degeneracy should be clarified.","tokens_in":15934,"tokens_out":2349,"would_cite":true,"duration_ms":24910,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A wavefunction projection method builds a diabatic Hamiltonian for van der Waals heterostructures, reducing MoS2/WS2 hole transfer to layer-localized states and their couplings.","keywords":["diabatization","van der Waals heterostructures","wavefunction projection","hole transfer","MoS2/WS2","plane-wave basis","projector augmented wave","diabatic Hamiltonian"],"falsifier":"Recompute the same MoS2/WS2 diabatic Hamiltonian with a reference space expanded to include additional unoccupied bands, and compare with an independent diabatization such as constrained density functional theory; if the Γ-point coupling shifts by more than about 0.1 eV or the diabatic states gain substantial weight outside the monolayer manifold, the assumed reference space is not sufficient.","tokens_in":14917,"feed_emoji":"⚛️","tokens_out":9052,"duration_ms":83818,"temperature":0.7,"pith_summary":"This paper tries to establish a practical way to build diabatic Hamiltonians for weakly bound two-dimensional heterostructures directly from plane-wave density functional calculations. The key idea is to use the electronic states of each isolated monolayer as reference diabatic states and project the heterostructure's adiabatic states onto them. For MoS2/WS2, the resulting Hamiltonian gives valence-band states localized on each layer and interlayer couplings of 0.330 eV at the Γ point and 0.043 eV at the K point. This matters because a compact diabatic Hamiltonian is exactly what quantum-dynamics simulations need to describe photoinduced hole transfer in extended systems.","feed_headline":"Projection captures MoS2/WS2 hole transfer as coupled monolayer states","feed_subtitle":"Monolayer wavefunctions become the diabatic basis: 0.33 eV coupling at Γ, 0.04 eV at K, and a 24 fs transfer period.","key_machinery":"The central object is the adiabatic-to-diabatic transformation matrix $\\mathbf{T}$ built from wavefunction overlaps. The reference diabatic states are the single-particle orbitals of the isolated monolayers; the overlap between a heterostructure adiabatic state and a monolayer reference state is evaluated in the plane-wave/PAW representation through the all-electron reconstruction formula, then the raw projection matrix is orthogonalized as $\\mathbf{T} = (\\mathbf{T}^\\dagger \\mathbf{T})^{-1/2} \\mathbf{T}^\\dagger$. The diabatic Hamiltonian follows as $\\mathbf{V}^{(d)}_{\\rm el} = \\mathbf{T} \\mathbf{V}^{(a)}_{\\rm el} \\mathbf{T}^\\dagger$. This machinery turns expensive band-structure data for the full complex into a small Hamiltonian whose states are already localized on the donor and acceptor layers, with couplings ready for dynamics.","core_discovery":"The paper's central claim is that the wavefunction projection method, implemented with a plane-wave basis and projector augmented wave reconstruction of all-electron wavefunctions, provides a practical route to diabatic Hamiltonians for van der Waals heterostructures within the single-particle picture. Taking the isolated monolayer states of MoS2 and WS2 as reference diabatic states, the adiabatic states of the full heterostructure are transformed through a projection matrix with orthogonalization, producing a diabatic Hamiltonian whose diagonal elements are layer-localized valence band energies and whose off-diagonal elements are the interlayer couplings. For MoS2/WS2 the coupling between the two valence-band states at Γ is 0.330 eV and between those at K is 0.043 eV, explaining both the large Γ-point band splitting (0.67 eV) and the experimental ultrafast hole transfer.","pith_inferences":["A direct test would compare the projected diabatic couplings with nonadiabatic coupling matrix elements computed from ab initio molecular dynamics at Γ and K; agreement would support using the diabatic Hamiltonian as a dynamics generator, while disagreement would reveal missing reference states.","The method should generalize to electron transfer by taking isolated monolayer conduction band states as references, yielding the analogous donor-CB/acceptor-CB couplings without any change to the projection machinery.","Because the paper notes the reference states are ill-defined when chemical bonds connect the two parts, a natural extension is to generate fragment-localized orbitals for covalently linked donor-acceptor systems; until then the method's safe domain is genuinely van der Waals stacked complexes.","Including excitonic effects would require projecting many-body excited states rather than single-particle orbitals onto layer-localized exciton bases, a step the paper identifies as a future challenge."],"forward_implications":["The MoS2/WS2 diabatic Hamiltonian contains two strongly coupled valence-band states at Γ, with coupling 0.330 eV, so pure electronic two-state dynamics gives a coherent population oscillation with a period around 24.3 fs.","Because the coupling at K is only 0.043 eV, direct K-to-K hole transfer is inefficient; the likely route is an indirect pathway through the Γ-point states, consistent with previous phonon-assisted transfer mechanisms.","The strong Γ-point interlayer coupling produces a band splitting of 0.67 eV, shifting the heterostructure valence band maximum to Γ, about 0.2 eV above the K-point valence states.","The same projection scheme can be applied to other weakly stacked two-dimensional heterostructures to obtain diabatic Hamiltonians for either hole or electron transfer by choosing the appropriate valence or conduction band reference states.","Such diabatic Hamiltonians provide a foundation for nonadiabatic dynamics of extended systems using full quantum methods, since the electronic couplings and localized states are already in the required form."],"supporting_citations":[{"why":"Provides the wavefunction projection framework for constructing diabatic Hamiltonians that this work extends to periodic plane-wave systems.","marker":"19"},{"why":"Applies projection-based localized diabatization to molecular crystals, giving the localized reference-state idea used here.","marker":"20"},{"why":"Uses projection of adiabatic states onto pre-defined reference states at interfaces, the direct conceptual ancestor of the monolayer reference states.","marker":"23"},{"why":"Establishes projection diabatization within time-dependent density functional theory, supporting the transfer of the approach to periodic density functional calculations.","marker":"64"},{"why":"Defines the projector augmented wave method that lets the paper compute all-electron wavefunction overlaps from plane-wave pseudo-wavefunctions in the overlap formula.","marker":"71-72"}],"fun_headline_variants":["Projection builds diabatic model for MoS2/WS2 hole transfer","Wavefunction projection yields 0.33 eV MoS2/WS2 coupling","Plane-wave projection crafts diabatic Hamiltonians for vdW stacks","Diabatization via projection: 24 fs hole transfer in MoS2/WS2","Monolayer states as basis: projection constructs diabatic model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction assumes the heterostructure's relevant electronic states are faithfully described as combinations of the two isolated monolayers' valence and conduction states, so if interlayer interactions heavily rehybridize electrons beyond that space, the diabatic Hamiltonian is incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Projection builds diabatic model for MoS2/WS2 hole transfer","Wavefunction projection yields 0.33 eV MoS2/WS2 coupling","Plane-wave projection crafts diabatic Hamiltonians for vdW stacks","Diabatization via projection: 24 fs hole transfer in MoS2/WS2","Monolayer states as basis: projection constructs diabatic model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000172,"raw_usage":{"total_tokens":1237,"prompt_tokens":872,"completion_tokens":365,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":488,"completion_tokens_details":{"reasoning_tokens":264}},"tokens_in":488,"tokens_out":365,"duration_ms":3771,"temperature":1.0,"reasoning_tokens":264,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:46:44.666712+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the same MoS2/WS2 diabatic Hamiltonian with a reference space expanded to include additional unoccupied bands, and compare with an independent diabatization such as constrained density functional theory; if the Γ-point coupling shifts by more than about 0.1 eV or the diabatic states gain substantial weight outside the monolayer manifold, the assumed reference space is not sufficient.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the wavefunction projection framework for constructing diabatic Hamiltonians that this work extends to periodic plane-wave systems."},{"cited_title":"X.; Subotnik, J","cited_arxiv_id":null,"evidence_quote":"Applies projection-based localized diabatization to molecular crystals, giving the localized reference-state idea used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Uses projection of adiabatic states onto pre-defined reference states at interfaces, the direct conceptual ancestor of the monolayer reference states."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes projection diabatization within time-dependent density functional theory, supporting the transfer of the approach to periodic density functional calculations."}],"review_version":1}