{"id":"ead08d8d-51f5-4684-b841-a684485cc1e8","arxiv_id":"1908.02733","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Force-balance equations of the current density give exact determining equations for exchange-correlation potentials and produce local-exchange approximations that reduce to LDA and Slater Xα in the homogeneous limit.","lead":"This paper derives exchange-correlation potentials in density functional theory from Newton-like equations of motion for the electron current, instead of from energy formulas. If the approach works, it offers a cheaper and more unified way to build approximations for DFT, including systems with magnetic fields and time-dependent problems.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The local-exchange approximation n∇vHx = -FW[Φ] (Eq. 23) is ill-defined for generic inhomogeneous systems because FW[Φ]/n is not curl-free, so no local vHx exists; the claimed equivalence with the divergence form (Eq. 27) holds only for the exact curl-free FHxc.","rationale":"The reader correctly identified the unproven density-potential mapping for ground-state CDFT with the physical current as a weakness; that concern is real and explicitly acknowledged in Sec. III.B. However, the more directly fatal issue for the central claim is the local-exchange approximation in Eq. (23) itself: for a generic KS Slater determinant, FW[Φ]/n is not curl-free, so no local vHx exists. This is an internal mathematical inconsistency, not a missing external proof, and it affects the simplest standard-DFT setting. The exact force-balance relations are valuable, and the LDA reduction in Sec. IV is plausible, but the constructive approximation strategy as written requires either a longitudinal projection or a demonstration of curl-freeness that is absent. Because the defect is repairable by reformulating the approximation through the divergence equation, the appropriate verdict remains conditional rather than a full rejection, so I do not change the reader's CONDITIONAL verdict. My disagreement with the reader is only about which assumption is the single most load-bearing one: the curl-free condition is more central because it undermines the proposed approximation even when all density-potential mappings are assumed to exist.","tokens_in":17197,"tokens_out":22782,"duration_ms":252621,"concrete_test":"Implement a small three-dimensional KS Slater determinant in a non-spherical external potential, e.g., two spin-polarized electrons in the two lowest orbitals of an anisotropic harmonic trap. Compute the 1-RDM and density from the occupied orbitals, evaluate FW[Φ] via Eq. (50) on a real-space grid, and test whether ∇×(FW[Φ]/n) vanishes pointwise. If a grid point with nonzero curl is found, Eq. (23) has no solution. Then solve the divergence-based equation ∇·(n∇vHx) = -∇·FW[Φ] for the same system and compare the resulting vHx with an independent OEP or LDA calculation; a difference in the transverse channel would confirm that the two proposed routes are not equivalent for the approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The exact force-balance subtraction leading to Eq. (21) is algebraically correct. The central constructive step, however, is the replacement of FHxc by FW[Φ] in Eq. (23): n∇vHx = -FW[Φ]. For this equation to define a scalar potential vHx, the vector field FW[Φ]/n must be a pure gradient. For a KS Slater determinant, Eq. (50) gives FW[Φ] = -n∇v_H + (1/2)∇A, where A(r) = ∫ w(r-r2)|ρ(1)(r,r2)|^2 dr2. Hence FW[Φ]/n = -∇v_H + (1/(2n))∇A. The second term has curl -(1/(2n^2))∇n × ∇A, which is not identically zero for a generic Slater determinant in three dimensions. Therefore Eq. (23) has no solution in general, and the proposed local-exchange approximation is not well-defined as stated. The paper asserts that in the static case Eqs. (21) and (27) carry the same information and lead to the same approximations; this is true for the exact FHxc, whose curl-free nature is guaranteed by n∇vHxc = -FHxc, but it is not true for the approximate replacement FW[Φ]. If the intended construction is instead the divergence-based Poisson equation ∇·(n∇vHx) = -∇·FW[Φ], that is a different approximation and should be stated explicitly, since it discards the transverse part of FW[Φ]. This issue is independent of the density-potential mapping assumption and affects even standard ground-state DFT, the simplest setting of the paper.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a constructive route to exchange-correlation (xc) potentials in density-functional theories based on equations of motion for current quantities (force balance equations) rather than energy or action functionals. Starting from the equation of motion of the first-order reduced density matrix, the authors derive the current-density equation of motion (force balance equation) for both interacting and non-interacting Kohn-Sham systems. Subtracting these equations yields exact pointwise determining relations for the xc potentials: Eq. (21) for ground-state DFT, Eq. (27) for TDDFT, Eq. (31) for CDFT/TDCDFT with the physical current density, and Eq. (41)/(45) for paramagnetic CDFT. The paper then introduces local-exchange approximations by replacing the interacting internal forces with those of a Slater determinant, and shows in the homogeneous limit that these approximations reduce to a family of LDA-type exchange potentials, recovering exchange-only LDA and connecting to Slater's Xα method. The approach is advertised as avoiding the optimized-effective-potential (OEP) procedure and the differentiability and causality issues of energy- and action-based functionals.","tokens_in":17555,"tokens_out":9277,"duration_ms":89047,"significance":"If the proposed exact relations and approximations are valid, the paper offers a genuinely different route to orbital-dependent xc potentials that bypasses energy functionals and OEP, with potential consequences for time-dependent and current-carrying systems. The algebraic derivations are careful, the exact relations reproduce known zero-force and zero-torque constraints, and the paper gives useful clarification of the interrelations between the xc potentials of DFT, TDDFT, CDFT, and TDCDFT. However, the central constructive claim is weakened by two load-bearing issues: the pointwise local-exchange approximation is not well-defined for generic inhomogeneous systems because the approximate force is not curl-free, and the ground-state CDFT section with the physical current relies on an explicitly unproven density-potential mapping. These issues do not invalidate the exact formal relations but they do require substantial revision of the proposed approximation scheme and of the scope of the CDFT claims.","major_comments":[{"comment":"The local-exchange approximation n∇v_Hx = -F_W[Φ] in Eq. (23) is not well-defined for generic inhomogeneous systems. For a Kohn-Sham Slater determinant, Eq. (50) gives F_W[Φ] = -n∇v_H + (1/2)∇A(r) with A(r) = ∫ w(r-r2)|ρ^(1)(r,r2)|^2 dr2, so F_W[Φ]/n = -∇v_H + (1/(2n))∇A, whose curl is -(1/(2n^2))∇n × ∇A and is generically nonzero. Hence no scalar v_Hx satisfies Eq. (23) pointwise. The statement that in the static case Eqs. (21) and (27) share exactly the same information and lead to the same approximations is only true for the exact F_Hxc, whose curl-free nature follows from n∇v_Hxc = -F_Hxc; it does not carry over to the replacement F_W[Φ]. If the intended construction is the divergence-based Eq. (28), that is a distinct approximation that discards the transverse part of F_W[Φ] and should be stated explicitly as such. This issue affects standard ground-state DFT and is independent of the density-potential mapping assumptions.","section":"Sec. III.A, Eqs. (23), (27), (28)"},{"comment":"The exact determining relation (31) for ground-state physical-current CDFT relies on an unproven density-potential mapping. The paper states that it 'tacitly assumes that an appropriate density-potential mapping exists' for the physical current density in the static case. Unlike the time-dependent case (which rests on Vignale's theorem) and the paramagnetic-current case (Refs. 49, 50, 41), no ground-state Hohenberg-Kohn-type theorem for the physical current density is available to date; the cited positive examples (MDFT, QEDFT, kinetic-energy DFT on a lattice) concern different settings. If this density-potential map does not exist, the Kohn-Sham system matching both n and j in a static problem is not guaranteed to exist, and Eq. (31) with ∂tA_xc = 0 cannot serve as an exact determining relation for the xc potentials in that setting. The authors should either prove or otherwise substantiate the map for the relevant class of Hamiltonians, or explicitly restrict the ground-state CDFT claims to the paramagnetic current density treated in Sec. III.C.","section":"Sec. III.B, sentence after Eq. (30)"}],"minor_comments":[{"comment":"The homogeneous-limit result v^λLDA_x = 2λ vLDA_x contains a free parameter λ; the standard exchange-only LDA is recovered only for the midpoint choice λ = 1/2. The abstract and Sec. IV should state this explicitly, since otherwise the claim that the local-exchange approximation 'reduces to the exchange-only LDA' appears parameter-dependent.","section":"Sec. IV.B, Eq. (62)"},{"comment":"The Helmholtz decomposition F_Hxc/n = -∇φ + ∇×α is not unique without specifying boundary conditions; the authors should state the boundary conditions at infinity under which the scalar potential v_Hxc and the vector potential A_xc are uniquely defined in Eqs. (32)-(35).","section":"Sec. III.B, Helmholtz decomposition"},{"comment":"The subtraction leading to Eq. (31) is algebraically involved; providing a short verification in index notation or an appendix would improve readability and help readers check the signs of the various dyadic-product terms.","section":"Sec. III.A, Eq. (31)"},{"comment":"The statement that the approach avoids the OEP procedure should be nuanced: solving Eq. (27) or (28) for v_Hx still requires inverting a differential equation, and the numerical cost relative to OEP is not addressed in the manuscript.","section":"Sec. I, Introduction"}],"recommendation":"major_revision","confidential_remarks":"The exact force-balance relations are derived carefully and the paper is generally well written, but the advertised local-exchange approximation requires reformulation to be well-posed in inhomogeneous systems, and the ground-state physical-current CDFT section rests on an acknowledged unproven mapping. With these points addressed, I would support publication. The authors should also sharpen the novelty statement with respect to Tokatly's earlier work (Ref. 21), since Eq. (21) and related force-balance relations were already considered there; the new contribution appears to lie in the orbital-dependent approximations and the systematic treatment of the various DFT settings."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something genuinely useful: it takes the force-balance equation for the current density and turns it into exact, pointwise determining relations for the xc potentials in DFT, CDFT, and TDCDFT, avoiding the OEP bottleneck and the usual energy/action functional pathologies. The core scalar-DFT relation, Eq. (21), is not new—the paper credits Tokatly and Ruggenthaler–Bauer—but the systematic extension to physical- and paramagnetic-current CDFT, the gauge analysis, the Helmholtz decomposition, and the evolution equation for Axc are a real contribution. The derivations are clear and the authors are honest about the places where they rely on unproven density-potential maps, especially for ground-state physical-current CDFT. They also flag the loss of direct total-energy access. That is good scientific citizenship.\n\nThe exact part holds up. The proposed local-exchange approximation does not, at least not as stated. Eq. (23), n∇vHx = -FW[Φ], requires FW[Φ]/n to be a pure gradient. For a Slater determinant, FW[Φ] = -n∇vH + (1/2)∇A, so the exchange-force piece is (1/(2n))∇A, whose curl is -(1/(2n^2))∇n×∇A, generally nonzero in three dimensions. Thus no scalar local vHx solves Eq. (23) for a generic inhomogeneous system. The paper says Eqs. (21) and (27) carry the same information and lead to the same approximations, but that is only true for the exact FHxc, whose curl-free nature is guaranteed. For the approximate replacement FW[Φ], the divergence-based Poisson form is a different approximation—one that discards the transverse part of FW[Φ]. The stress-test note is right, and it affects the simplest setting of the paper, not just the CDFT extensions.\n\nThe LDA reduction in Sec. IV is also softer than advertised. The free parameter λ enters as where you expand the density; λ=1/2 recovers standard exchange LDA, but the derivation does not determine λ from first principles. The connection to Slater Xα via λ=3α/4 is nice but it reinterprets the empirical parameter rather than fixing it. Minor, but worth saying.\n\nThere is no numerical validation, so the practical value is unproven. Still, the exact relations and the framework deserve referee time. The paper is well-written, properly cites prior work, and the central exact derivation is solid. A serious referee should push on the well-definedness of Eq. (23), ask whether the divergence form is the intended replacement, and request at least one illustrative numerical example.\n\nRecommendation: send it to peer review, but expect revision on the local-exchange approximation.","headline":"A careful, honest derivation of force-balance relations for xc potentials in DFT/CDFT/TDCDFT, with a real flaw in the proposed local-exchange approximation that a referee should catch.","tokens_in":18139,"tokens_out":2126,"would_cite":true,"duration_ms":26936,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Force-balance equations exactly determine exchange-correlation potentials without energy functionals.","keywords":["density functional theory","exchange-correlation potential","force balance equation","current density functional theory","time-dependent density functional theory","optimized effective potential","local density approximation","Slater exchange"],"falsifier":"Compute both sides of Eq. (21) for a small inhomogeneous system using near-exact interacting and Kohn–Sham wave functions: if $F_T[\\Phi] - F_T[\\Psi] - F_W[\\Psi]$ divided by $n$ is not a gradient field, the exact pointwise relation fails.","tokens_in":17003,"feed_emoji":"⚖️","tokens_out":5797,"duration_ms":59677,"temperature":0.7,"pith_summary":"The paper proposes that exchange-correlation (xc) potentials in density-functional theories can be determined directly by subtracting the force-balance equations of the interacting and Kohn–Sham systems, instead of differentiating an energy or action functional. The central exact relation is $n\\nabla v_{Hxc} = F_T[\\Phi] - F_T[\\Psi] - F_W[\\Psi]$, with an analogous evolution equation for the xc vector potential in time-dependent current-density functional theory. These relations are pointwise, so they give orbital-dependent xc potentials without the numerically costly optimized-effective-potential step and without the differentiability and causality problems of functional-based theories. The same framework covers ground-state DFT, TDDFT, CDFT, and TDCDFT, and in the homogeneous limit its exchange-type approximation reduces to exchange LDA and Slater Xα. If correct, this gives a constructive route to advanced xc potentials that bypasses the usual functional machinery.","feed_headline":"No energy functional needed to fix exchange-correlation potentials","feed_subtitle":"Force-balance equations give exact xc potentials and recover LDA and Slater exchange.","key_machinery":"The machinery is the force-balance equation, the equation of motion for the physical current density, obtained from the equation of motion of the first-order reduced density matrix. Its two internal terms are the kinetic force $F_T[\\Psi]$ and the interaction force $F_W[\\Psi]$; subtracting them between the interacting and Kohn–Sham systems exposes the xc potential. A Helmholtz decomposition of the combined force term separates longitudinal (scalar potential) and transverse (vector potential) contributions. For the vector potential the determining relation is an evolution equation, so any imbalance in the forces is absorbed by $\\partial_t A_{xc}$.","core_discovery":"The paper's central claim is that equality of the densities and currents between interacting and non-interacting systems turns the current-density equations of motion into exact determining equations for the xc potentials. For static scalar potentials, subtracting the two zero-force balance equations gives Eq. (21), which states that $n\\nabla v_{Hxc}$ equals the difference of kinetic internal forces plus the interaction force. In the time-dependent current-density setting, subtracting the equations of motion gives Eq. (31), and with a Helmholtz decomposition this separates into a scalar potential determined by the longitudinal internal force and an evolution equation for $A_{xc}$ from the transverse part. These relations do not require the xc potentials to be functional derivatives; they exist whenever a matching Kohn–Sham system can be found. The exchange approximation obtained by replacing the interacting wave function with the Slater-determinant Kohn–Sham wave function reduces to exchange-only LDA in the uniform-gas limit, with the Slater Xα parameter reinterpreted as the position of the Taylor-expansion point.","pith_inferences":["One consequence the paper leaves implicit is that if the ground-state CDFT map with physical currents is later established, Eq. (31) would also supply a non-adiabatic xc vector potential with memory, so the approach could be benchmarked against existing memory approximations in linear response.","The reinterpretation of the Slater α as the expansion point λ suggests a concrete test: compare exchange potentials from Eq. (52) with exact-exchange optimized-effective-potential results on inhomogeneous atoms; the optimal λ need not be 1/2.","Because the xc potentials come from forces rather than an energy, total energies must be obtained a posteriori from the Kohn–Sham orbitals; whether a parent energy functional exists, and whether thermochemical consistency holds, remains an open question the paper acknowledges.","The same subtraction strategy could be carried over to coupled matter–photon systems, where the vector potential includes quantized modes; testing it would require building the corresponding photon-mode force-balance terms."],"forward_implications":["A direct, pointwise route to $v_{Hxc}$ exists that does not require evaluating any energy or action functional.","Orbital-dependent exchange potentials can be constructed without the optimized-effective-potential procedure, at least in the approximations considered.","The TDCDFT xc vector potential satisfies an evolution equation, which automatically carries memory effects and absorbs unbalanced forces.","The different xc potentials of DFT, TDDFT, CDFT, and TDCDFT become manifestations of one force-balance relation, with gauge choices distinguishing the cases.","In a homogeneous system the exchange approximation reduces to exchange-only LDA and exposes Slater Xα as a midpoint-parameter choice."],"supporting_citations":[{"why":"It supplies the EOM-based force-balance route and the integrated zero-force and zero-torque constraints that this paper extends.","marker":"[21]"},{"why":"It previously derived the local-exchange approximation from equations of motion, which this paper generalizes to CDFT and TDCDFT.","marker":"[28]"},{"why":"It provides the density-to-potential mapping that justifies matching densities in the time-dependent setting.","marker":"[19]"},{"why":"It provides the density- and current-to-potential mapping for TDCDFT that justifies matching physical currents.","marker":"[24]"},{"why":"It introduces the Slater Xα exchange potential that the homogeneous limit reproduces.","marker":"[36]"},{"why":"It supplies the rigorous paramagnetic CDFT setup and Kohn–Sham iteration that underpins the paramagnetic current case.","marker":"[41]"},{"why":"It provides the ground-state density-to-potential mapping underlying the static DFT relation.","marker":"[44]"}],"fun_headline_variants":["Force balance gives exact xc potentials with no energy functional","No energy functional: force balance yields xc potentials exactly","Force balance replaces OEP, recovers LDA and Slater Xα","Exact xc potentials from force balance equations alone"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a density-potential mapping exists for every setting used; in particular, for ground-state CDFT with the physical current density the paper only assumes such a map, and if that map fails the Kohn–Sham construction matching physical currents collapses.","fun_headline_variants_meta":{"raw":{"variants":["Force balance gives exact xc potentials with no energy functional","No energy functional: force balance yields xc potentials exactly","Force balance replaces OEP, recovers LDA and Slater Xα","Exact xc potentials from force balance equations alone"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00023,"raw_usage":{"total_tokens":1432,"prompt_tokens":843,"completion_tokens":589,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":459,"completion_tokens_details":{"reasoning_tokens":519}},"tokens_in":459,"tokens_out":589,"duration_ms":6006,"temperature":1.0,"reasoning_tokens":519,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:37:17.596750+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute both sides of Eq. (21) for a small inhomogeneous system using near-exact interacting and Kohn–Sham wave functions: if $F_T[\\Phi] - F_T[\\Psi] - F_W[\\Psi]$ divided by $n$ is not a gradient field, the exact pointwise relation fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the EOM-based force-balance route and the integrated zero-force and zero-torque constraints that this paper extends."},{"cited_title":"Ruggenthaler \\ and\\ author D","cited_arxiv_id":null,"evidence_quote":"It previously derived the local-exchange approximation from equations of motion, which this paper generalizes to CDFT and TDCDFT."},{"cited_title":"Vignale ,\\ 10.1103/PhysRevB.70.201102 journal journal Phys","cited_arxiv_id":null,"evidence_quote":"It provides the density- and current-to-potential mapping for TDCDFT that justifies matching physical currents."}],"review_version":1}