Abstract
This paper is concerned with the numerical analysis of linear and nonlinear Schrödinger equations with periodic analytic potentials. We prove that, for linear equations, when the potential is analytic in a strip of width A of the complex plane, the solution is analytic in the same strip, ensuring an exponential convergence of the planewave discretization of the equation with rate A. On the other hand, for nonlinear equations, we find that the solution may be analytic only in a strip of width smaller than A. This behavior is illustrated by two examples using a combination of numerical and analytical arguments.
Similar content being viewed by others
References
Babuška, I., Osborn, J.: Eigenvalue problems. In Handbook of Numerical Analysis, volume 2 of Finite Element Methods (Part 1), pages 641–787. Elsevier, (1991)
Bender, C. M., Orszag, S. A.: Advanced Mathematical Methods for Scientists and Engineers. 1: Asymptotic Methods and Perturbation Theory. Springer, New York Heidelberg, (1999)
Bernstein, S.: Sur la nature analytique des solutions des équations aux dérivées partielles du second ordre. Math. Ann. 59(1–2), 20–76 (1904)
Blatt, S.: On the analyticity of solutions to non-linear elliptic partial differential systems. arXiv:2009.08762 [math.AP], (2020)
Blöchl, P.E.: Projector augmented-wave method. Phys. Rev. B 50(24), 17953–17979 (1994)
Cancès, E., Chakir, R., Maday, Y.: Numerical analysis of nonlinear eigenvalue problems. J. Sci. Comput. 45(1), 90–117 (2010)
Cancès, E., Chakir, R., Maday, Y.: Numerical analysis of the planewave discretization of some orbital-free and Kohn-Sham models. ESAIM: Math. Modell. Numer. Anal., 46(2):341–388, (2012)
Friedman, A.: On the regularity of the solutions of non-linear elliptic and parabolic systems of partial differential equations. Indiana Univ. Math. J. 7(1), 43–59 (1958)
Giannozzi, P., Baroni, S., Bonini, N., Calandra, M., Car, R., Cavazzoni, C., Ceresoli, D., Chiarotti, G. L., Cococcioni, M., Dabo, I., Dal Corso, A., de Gironcoli, S., Fabris, S., Fratesi, G., Gebauer, R., Gerstmann, U., Gougoussis, C., Kokalj, A., Lazzeri, M., Martin-Samos, L., Marzari, N., Mauri, F., Mazzarello, R., Paolini, S., Pasquarello, A., Paulatto, L., Sbraccia, C., Scandolo, S., Sclauzero, G., Seitsonen, A. P., Smogunov, A., Umari, P., Wentzcovitch, R. M.: QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials. J. Phys. Condens. Matter: An Institute Phys. J., 21(39):395502, (2009)
Goedecker, S., Teter, M., Hutter, J.: Separable dual-space Gaussian pseudopotentials. Phys. Rev. B 54(3), 1703 (1996)
Gonze, X., Amadon, B., Antonius, G., Arnardi, F., Baguet, L., Beuken, J.-M., Bieder, J., Bottin, F., Bouchet, J., Bousquet, E., Brouwer, N., Bruneval, F., Brunin, G., Cavignac, T., Charraud, J.-B., Chen, W., Côté, M., Cottenier, S., Denier, J., Geneste, G., Ghosez, P., Giantomassi, M., Gillet, Y., Gingras, O., Hamann, D.R., Hautier, G., He, X., Helbig, N., Holzwarth, N., Jia, Y., Jollet, F., Lafargue-Dit-Hauret, W., Lejaeghere, K., Marques, M.A.L., Martin, A., Martins, C., Miranda, H.P.C., Naccarato, F., Persson, K., Petretto, G., Planes, V., Pouillon, Y., Prokhorenko, S., Ricci, F., Rignanese, G.-M., Romero, A.H., Schmitt, M.M., Torrent, M., van Setten, M.J., Van Troeye, B., Verstraete, M.J., Zérah, G., Zwanziger, J.W.: The Abinit project: Impact, environment and recent developments. Comput. Phys. Commun. 248, 107042 (2020)
Hartwigsen, C., Goedecker, S., Hutter, J.: Relativistic separable dual-space Gaussian pseudopotentials from H to Rn. Phys. Rev. B 58(7), 3641–3662 (1998)
Hashimoto, Y.: A Remark on the Analyticity of the Solutions for Non-Linear Elliptic Partial Differential Equations. Tokyo J. Math. 29(2), 271–281 (2006)
Herbst, M.F., Levitt, A., Cancès, E.: DFTK: A Julian approach for simulating electrons in solids. Proc. JuliaCon Conf. 3(26), 69 (2021)
Liu, D.C., Nocedal, J.: On the limited memory BFGS method for large scale optimization. Math. Progr. 45(1–3), 503–528 (1989)
Morrey, C. B.: On the Analyticity of the Solutions of Analytic Non-Linear Elliptic Systems of Partial Differential Equations: Part I. Analyticity in the Interior. Am. J. Math., 80(1):198, (1958)
Morrey, C. B.: On the Analyticity of the Solutions of Analytic Non-Linear Elliptic Systems of Partial Differential Equations: Part II. Analyticity at the Boundary. Am. J. Math., 80(1):219, (1958)
Nottoli, T., Giannì, I., Levitt, A., Lipparini, F.: A robust, open-source implementation of the locally optimal block preconditioned conjugate gradient for large eigenvalue problems in quantum chemistry. Theoret. Chem. Acc. 142(8), 69 (2023)
Perdew, J.P., Wang, Y.: Accurate and simple analytic representation of the electron-gas correlation energy. Phys. Rev. B 45(23), 13244–13249 (1992)
Petrovskii, I.G.: Sur l’analyticité des solutions des systèmes d’équations différentielles. Matematiceskij sbornik 47(1), 3–70 (1939)
Ratcliff, L.E., Dawson, W., Fisicaro, G., Caliste, D., Mohr, S., Degomme, A., Videau, B., Cristiglio, V., Stella, M., D’Alessandro, M., Goedecker, S., Nakajima, T., Deutsch, T., Genovese, L.: Flexibilities of wavelets as a computational basis set for large-scale electronic structure calculations. J. Chem. Phys. 152(19), 194110 (2020)
Reed, M., Simon, B.: Analysis of Operators. Number 4 in Methods of Modern Mathematical Physics. Academic Press, (1978)
Romero, A.H., Allan, D.C., Amadon, B., Antonius, G., Applencourt, T., Baguet, L., Bieder, J., Bottin, F., Bouchet, J., Bousquet, E., Bruneval, F., Brunin, G., Caliste, D., Côté, M., Denier, J., Dreyer, C., Ghosez, P., Giantomassi, M., Gillet, Y., Gingras, O., Hamann, D.R., Hautier, G., Jollet, F., Jomard, G., Martin, A., Miranda, H.P.C., Naccarato, F., Petretto, G., Pike, N.A., Planes, V., Prokhorenko, S., Rangel, T., Ricci, F., Rignanese, G.-M., Royo, M., Stengel, M., Torrent, M., van Setten, M.J., Van Troeye, B., Verstraete, M.J., Wiktor, J., Zwanziger, J.W., Gonze, X.: ABINIT: Overview and focus on selected capabilities. J. Chem. Phys. 152(12), 124102 (2020)
Slater, J.C.: A Simplification of the Hartree-Fock Method. Phys. Rev. 81(3), 385–390 (1951)
Troullier, N., Martins, J.L.: Efficient pseudopotentials for plane-wave calculations. Phys. Rev. B 43(3), 1993–2006 (1991)
Vrabel, R., Tanuska, P., Vazan, P., Schreiber, P., Liska, V.: Duffing-Type Oscillator with a Bounded from above Potential in the Presence of Saddle-Center Bifurcation and Singular Perturbation: Frequency Control. Abstr. Appl. Anal. 1–7, 2013 (2013)
Walter, W.: Ordinary Differential Equations. Number 182 in Graduate Texts in Mathematics ; Readings in Mathematics. Springer, New York, (1998)
Wazewski, T.: Systèmes des équations et des inégualités différentielles ordinaires aux deuxièmes membres monotones et leurs applications. Annales de la Société polonaise de mathématique 23, 112–166 (1950)
Acknowledgements
The authors would like to thank Geneviève Dusson and Michael F. Herbst for fruitful discussions. The authors would also like to thank the anonymous reviewers for their comments and suggestions.
Funding
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 810367).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
All the authors declare that they have no conflicts of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Extension to the Multidimensional Case with Application to Kohn–Sham Models
The goal of this section is to extend the previous results to the multidimensional case and apply them to the linear version of the Kohn–Sham equations (1). To this end, consider a Bravais lattice \(\mathbb {L}=\mathbb {Z}{\varvec{a}}_1+ \cdots + \mathbb {Z}{\varvec{a}}_d\) where \({\varvec{a}}_1,\dots , {\varvec{a}}_d\) are linearly independent vectors of \(\mathbb {R}^d\) (\(d=3\) for KS-DFT). Up to an affine change of variables (which preserves analyticity), we can take without loss of generality \(\mathbb {L}= 2\pi \mathbb {Z}^{d}\). We denote by \(\Omega =[0,2\pi )^{d}\) a unit cell, by \(\mathbb {L}^* = \mathbb {Z}^{d}\) the reciprocal lattice, by \(e_{{\varvec{G}}}({\varvec{x}})=|\Omega |^{-1/2}{\textrm{e}}^{i{\varvec{G}} \cdot {\varvec{x}}}\) the Fourier mode with wavevector \({\varvec{G}} \in \mathbb {L}^*\), and by
the \(\mathbb {L}\)-periodic Sobolev spaces endowed with their usual inner products. All the arguments in Sects. 3.1-3.3 can be extended to the multidimensional case by introducing the Hilbert spaces
where \(\displaystyle {{w}_{A,\mathbb {L}}({\varvec{G}}) = \sum _{n=1}^d w_A(G_{n})}\). Each \(u \in \mathcal {H}_{A,\mathbb {L}}\) can be extended to an analytic function \(u(z_1,\dots ,z_d)\) of d complex variables defined on a neighborhood on \(\mathbb {R}^d\), and it holds
with \(e_{1},\dots , e_d\) the canonical basis vectors. The approximation space \(X_{N,\mathbb {L}}\) is then defined as
and the inverse \(T_{22,\mathbb {L}}^{-1}\) of the restriction \(T_{22,\mathbb {L}}\) of the operator \(-\Delta \) on \(L^2_{{\textrm{per}},\mathbb {L}}\) to the invariant subspace \(X_{N,\mathbb {L}}^\perp = \text{ Span } (e_{{\varvec{G}}}, \, {\varvec{G}} \in \mathbb {L}^*, \, |{{\varvec{G}}}| > N )\) satisfies
The proofs of Theorems 1, 2 and 3 can thus be straightforwardly adapted to the multidimensional case.
Lastly, if \(V \in \mathcal {H}_{B,\mathbb {L}}\) for some \(B > 0\), the Schrödinger operator \(H = -\Delta + V\) considered this time as a Schrödinger operator on \(L^2(\mathbb {R}^d,\mathbb {C})\) with an \(\mathbb {L}\)-periodic potential, can be decomposed by the Bloch transform [22, Section XIII.16] and its Bloch fibers are the self-adjoint operators on \(L^2_{{\textrm{per}},\mathbb {L}}\) with domain \(H^2_{{\textrm{per}},\mathbb {L}}\) and form domain \(H^1_{{\textrm{per}},\mathbb {L}}\) defined as \(H_{{\varvec{k}}} = (-\i \nabla + {\varvec{k}})^2 + V\). The following result is concerned with the Bloch eigenmodes of the \(H_{{\varvec{k}}}\)’s.
Theorem 6
Let \(B > 0\) and \(V \in \mathcal {H}_{B,\mathbb {L}}\). For each \({\varvec{k}} \in \mathbb {R}^d\), the eigenfunctions of the Bloch fibers \(H_{{\varvec{k}}} = (- i\nabla +{\varvec{k}})^2 + V\) of the periodic Schrödinger operator \(H=-\Delta + V\) are in \(\mathcal {H}_{A,\mathbb {L}}\) for any \(0<A<B\). Let \(\lambda _{1,{\varvec{k}}} \leqslant \lambda _{2,{\varvec{k}}} \leqslant \cdots \) be the eigenvalues of \(H_{{\varvec{k}}}\) counted with multiplicities and ranked in non-decreasing order, and \(\lambda _{1,{\varvec{k}},N} \leqslant \lambda _{2,{\varvec{k}},N} \leqslant \cdots \leqslant \lambda _{d_{\mathbb {L},N},{\varvec{k}},N}\) the eigenvalues of the variational approximation of \(H_{{\varvec{k}}}\) in the \(d_{\mathbb {L},N}\)-dimensional space
Then, for each \(0< A < B\) and \(i \in \mathbb {N}^*\), there exists a constant \(C_{i,A} \in \mathbb {R}_+\) such that
where \(\Omega ^*\) is the first Brillouin zone (i.e. the Voronoi cell of the lattice \(\mathbb {L}\) of \(\mathbb {R}^d\) containing the origin).
Proof
It suffices to replace in the proofs of Theorems 2 and 3\(X_N\) with \(X_{\mathbb {L},{\varvec{k}}, N}\) and \(T_{22}\) with the restriction \(T_{22,\mathbb {L},{\varvec{k}}}\) of the operator \((- i\nabla + {\varvec{k}})^2\) to the invariant space \(X_{\mathbb {L},{\varvec{k}}, N}^\perp \). The latter is invertible and such that \(\Vert T_{22,\mathbb {L},{\varvec{k}}}^{-1} \Vert _{\mathcal {L}(X_{\mathbb {L},{\varvec{k}}, N}^\perp )} \leqslant N^{-2}\) and \(\Vert T_{22,\mathbb {L},{\varvec{k}}}^{-1} \Vert _{\mathcal {L}(\mathcal {H}_{A,\mathbb {L}} \cap X_{\mathbb {L},{\varvec{k}}, N}^\perp )} \leqslant N^{-2}\). The result then follows similarly. \(\square \)
Remark 3
As a conclusion, let us mention that, under appropriate assumptions, we can extend these results to the KS-DFT equations (1) with GTH pseudopotentials in the case where an analytic parametrization of the exchange-correlation functional is used. This is relevant for instance when studying condensed-phase systems, for which the periodic setting is well suited, and where the commonly used approximations of \(V_{\textrm{xc},\rho }\) are analytic on the positive density values that are of interest in such cases [19]. Then we can rewrite (1) as a system of elliptic PDEs:
We can then use known results for elliptic systems of PDEs [16, 17] to conclude, in a way similar to what we proved in Theorem 4, that the orbitals \(\varphi _i\) belong to \(\mathcal {H}_{A,\mathbb {L}}\) for some \(A>0\). In particular, this leads to the exponential convergence of planewave approximations, justifying the use of GTH pseudopotentials.
B Proof of Lemma 1
We start by rewriting (18) as a first-order ODE on \(\Psi _\varepsilon (y) {:}{=}\begin{bmatrix} \psi _\varepsilon (y) \\ \psi '_\varepsilon (y) \end{bmatrix} \in \mathbb {R}^2\), starting at \(y_\eta \geqslant B_0\):
We then use the following simplified version of more general comparison results on systems of differential inequalities [27, 28]. In the sequel, the inequality \(a \geqslant b\) for two vectors \(a,b \in \mathbb {R}^d\) means that \(a_i \geqslant b_i\) for all \(1 \leqslant \textrm{i} \leqslant d\).
Theorem 7
(See e.g. [27, p. 112]). Let \(d\geqslant 1\) and \(G: \mathbb {R}^d \rightarrow \mathbb {R}^d\) be locally Lipschitz and quasimonotone in the sense that for all \(X,Z \in {\mathbb {R}}^d\),
Let \(0 \leqslant y_0 < y_{\textrm{M}} \leqslant \infty \) and \(\Phi \in C^1 ([y_0,y_{\textrm{M}}),\mathbb {R}^d)\) and \(\Psi \in C^1 ([y_0,y_{\textrm{M}}),\mathbb {R}^d)\) satisfying respectively the ODE
and the differential inequality
Then we have
on \([y_0,y_\textrm{M})\).
To apply this result to (20), we introduce the function \(G_\varepsilon : \mathbb {R}^2 \rightarrow \mathbb {R}^2\) defined by
and the maximal solution \(\Phi _\varepsilon \) to
As \(\sinh (y) \geqslant 0\) for all \(y \geqslant 0\), we have the following differential inequality:
Note that, by convexity, \(\psi _\varepsilon (y) \geqslant \psi _\varepsilon (y_\eta ) > 1\) for \(y\geqslant y_\eta \): \(G_\varepsilon \) is indeed quasimonotone on the domain of interest. We are now ready to compute an upper bound of \(Y_\varepsilon \) with the use of Theorem 7, which yields
This leads us to the study of the ODE
We have
from which we deduce
with
where \(C(\eta )\) is computed from the initial conditions at \(y=y_\eta \) and \(B_0< y_0 < y_\eta \) is such that \(\psi _\varepsilon (y_0)=1\). Note that \(\eta \mapsto -\frac{1}{4}(1+\eta )^4 +\frac{1}{2}(1+\eta )^2 + \frac{\varepsilon }{2}(\psi '_\varepsilon (y_0))^2\) is decreasing on \(\mathbb {R}_*^+\) and takes the value \(\frac{1}{4} + \frac{\varepsilon }{2}(\psi '_\varepsilon (y_0))^2 > \frac{1}{4}\) at \(\eta = 0\). Thus, for \(\eta >0\) small enough, \(C(\eta ) \geqslant \frac{1}{4}\) and we have
hence
Finally, we consider the ODE
whose solution is
which is defined only up to \(Y_{\varepsilon ,\eta } {:}{=}\sqrt{\frac{\varepsilon }{2}}\log \left( 1 + \frac{2}{\eta } \right) + y_\eta \). Applying again Theorem 7, we have that \(\phi _\varepsilon (y) \geqslant \xi _{\varepsilon ,\eta }(y)\) for any \(y \geqslant y_\eta \) such that both functions are still finite. Putting everything together, we obtain that \(\psi _\varepsilon \) is only defined up to some \(Y_\varepsilon \) with \(B_0< Y_\varepsilon \leqslant Y_{\varepsilon ,\eta }<\infty \) and that
These results are illustrated on Fig. 7, where we plotted the lower bound \(\xi _{\varepsilon ,\eta }\) for \(\varepsilon =0.1\) and \(\eta =0.5\) as well as a numerical approximation of \(\psi _\varepsilon \).
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Cancès, E., Kemlin, G. & Levitt, A. A Priori Error Analysis of Linear and Nonlinear Periodic Schrödinger Equations with Analytic Potentials. J Sci Comput 98, 25 (2024). https://doi.org/10.1007/s10915-023-02421-0
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10915-023-02421-0