Many electronic structure methods rely on the minimization of the energy of the system with respect to the one-body reduced density matrix (1-RDM). To formulate a minimization algorithm, the 1-RDM is often expressed in terms of its eigenvectors via an orthonormal transformation and its eigenvalues. This transformation drastically alters the energy landscape. Especially in 1-RDM functional theory this means that the convexity of the energy functional is lost. We show that degeneracies in the occupation numbers can lead to additional critical points which are classified as saddle points. Using a Cayley or Householder parametrization for the orthonormal transformation, no extra critical points arise. In the case of Given's rotations or the exponential, additional critical points can arise, which are of no concern in practical minimization. These findings provide an explanation for the success of recent minimization procedures using second-order information.
Many electronic structure methods rely on the minimization of the energy of the system with respect to the one-body reduced density matrix (1-RDM). To formulate a minimization algorithm, the 1-RDM is often expressed in terms of its eigenvectors via an orthonormal transformation and its eigenvalues. This transformation drastically alters the energy landscape. Especially in 1-RDM functional theory this means that the convexity of the energy functional is lost. We show that degeneracies in the occupation numbers can lead to additional critical points which are classified as saddle points. Using a Cayley or Householder parametrization for the orthonormal transformation, no extra critical points arise. In the case of Given's rotations or the exponential, additional critical points can arise, which are of no concern in practical minimization. These findings provide an explanation for the success of recent minimization procedures using second-order information.
From drug design to material science, computational chemistry has become a valuable tool for many fields of applications in the past decades. Quantum chemistry in particular, can provide accurate descriptions of molecules and solids by relying on principles of quantum mechanics with the Schrödinger equation at its core. Unfortunately this equation becomes prohibitively expensive to solve for systems made of more than a few electrons and one has to rely on other approaches. Among the most popular is Density Functional Theory, relatively inexpensive and exact in principle. One has, however, to introduce approximations in practice, deteriorating its accuracy for challenging, so-called strongly correlated, systems. On the contrary, other methods like CASSCF remain reliable for these systems, but although cheaper than the solution of the Schrödinger equation, remain too expensive for extended systems. There is therefore room for methods in between, that could handle large correlated systems with reasonable accuracy. One such method is the One Body Reduced Density Matrix (1-RDM) Functional Theory (RDMFT), based on the fact that an electronic system can, in principle exactly, be described by its 1-RDM only. Although its theory dates back to the 1970s, computational hindrances have prevented this method from shining so far. Indeed, computations in RDMFT rely on a constrained optimization of the eigenvalues and eigenvectors of the 1-RDM, that can be reformulated as an unconstrained optimization by parameterizing the latter but in both cases suffers from convergence problems. In this thesis, the role of this parametrization is analyzed and motivates a now second order optimization algorithm, which improves the convergence speed by two orders of magnitude w.r.t. the state of the art algorithm. This understanding of the problem also inspires an alternative approach with the 1-RDM as the optimization variable. This however requires the development of new approximations and a first investigation in that direction, based on Machine Learning is performed. Lastly, a generalization of the method to excited states is attempted, highlighting some key computational challenges.
Functionals of the one-body reduced density matrix (1-RDM) are routinely minimized under Coleman’s ensemble N-representability conditions. Recently, the topic of pure-state N-representability conditions, also known as generalized Pauli constraints, received increased attention following the discovery of a systematic way to derive them for any number of electrons and any finite dimensionality of the Hilbert space. The target of this work is to assess the potential impact of the enforcement of the pure-state conditions on the results of reduced density-matrix functional theory calculations. In particular, we examine whether the standard minimization of typical 1-RDM functionals under the ensemble N-representability conditions violates the pure-state conditions for prototype 3-electron systems. We also enforce the pure-state conditions, in addition to the ensemble ones, for the same systems and functionals and compare the correlation energies and optimal occupation numbers with those obtained by the enforcement of the ensemble conditions alone.
Functionals of the one-body reduced density matrix (1-RDM) are routinely minimized under Coleman's ensemble N-representability conditions. Recently, the topic of pure-state N-representability conditions, also known as generalized Pauli constraints, received increased attention following the discovery of a systematic way to derive them for any number of electrons and any finite dimensionality of the Hilbert space. The target of this work is to assess the potential impact of the enforcement of the pure-state conditions on the results of reduced density-matrix functional theory calculations. In particular, we examine whether the standard minimization of typical 1-RDM functionals under the ensemble N-representability conditions violates the pure-state conditions for prototype 3-electron systems. We also enforce the pure-state conditions, in addition to the ensemble ones, for the same systems and functionals and compare the correlation energies and optimal occupation numbers with those obtained by the enforcement of the ensemble conditions alone.
In many of the approximate functionals in one-body reduced-density-matrix (1RDM) functional theory, the approximate two-body reduced density matrix (2RDM) in the natural orbital representation only depends on the natural occupation numbers. In Phys. Rev. A 92, 012520 (2015), Wang and Knowles initialized a discussion of to what extent this simplification is valid by introducing two different H4 geometries with identical natural occupation numbers but different 2RDMs. Gritsenko has argued that this feature is due symmetry [Phys. Rev. A 97, 026501 (2018)]. This work aims to contribute to the discussion on the following points: (1) one should rather speak of symmetry-restricted variants of the universal functional than saying that the universal functional is symmetry dependent; (2) the unitary invariance of degenerate NOs can lead to large deviations in the 2RDM elements, especially the phase of the NOs; (3) symmetry-restricted functionals are constructed for the H4 geometries considered by Wang and Knowles, whose structure could serve as guide in the construction of approximate 1RDM functionals.
The electronic Hamiltonian contains only pairwise interactions, allowing the energy of an electronic system to be expressed in terms of the two-electron reduced-density-matrix (2-RDM) in lieu of the many-electron wavefunction. The variable space for the exact N-electron wavefunction scales exponentially with the size of the system, while the 2-RDM is polynomial in scale. By using the 2-RDM as the primary variable in electronic structure calculations, it may be possible to obtain very accurate energies at a much more favorable scaling than wavefunction methods. In this thesis, we will use two existing 2-RDM methods to treat electronic systems. First, we will apply the active-space variational 2-RDM method, which directly minimizes the energy with respect to the 2-RDM, to a cadmium telluride polymer that was recently used to greatly enhance the conductivity of CdTe quantum dots. We find that this polymer is very highly correlated despite a deceptively simple structure. We will then turn to the parametric 2-RDM method (p2-RDM), which parameterizes the 2-RDM in terms of a truncated configuration interaction ansatz, but which includes additional flexibility in order to be size-extensive. We apply p2-RDM to the study of the olympicene molecule, which features both fully aromatic and diradical isomers. The parametric 2-RDM method predicts that all isomers are stable to dissociation, in contrast to coupled cluster methods which do not predict stable diradical states. We then present analytical nuclear gradients for p2-RDM, which greatly decrease the number of calculations required to perform geometry optimizations. We apply these gradients to the study of trans-polyacetylene, for which p2-RDM, unlike many wavefunction methods, is able to predict a bond length alternation (BLA) to within experimental values. Lastly, as a single-reference method, p2-RDM may encounter numerical difficulties when the reference wavefunction is of particularly poor quality. We propose a modification to the parameterization that may render the method more generally robust.
The variational quantum eigensolver (VQE) is a promising method for simulating molecular systems on near-term quantum computers. This approach employs energy estimation; however, other relevant molecular properties can be extracted from the one-particle reduced density matrix (1-RDM) generated by VQE. The accuracy of these properties strongly depends on the reliability and convergence of the 1-RDM, which is not guaranteed by energy-only optimization. Thus, we investigate the effect of optimizing the 1-RDM within VQE to improve the accuracy of both the energy and molecular properties. A two-step algorithm was implemented that optimizes the energy and 1-RDM by incorporating a penalty term in the cost function to enforce the convergence of the 1-RDM. The first step focuses on energy minimization, while in the second step, a weighted penalty is added to the cost function to promote simultaneous improvement of the energy and 1-RDM. This approach was tested and validated for the k-UpCCGSD and GateFabric ansätzes with active spaces (4,4) and (2,2), respectively. k-UpCCGSD produces energies close to CISD, so optimizing 1-RDM has little effect on the energy but significantly improves electronic properties such as electron density, dipole moments, and atomic charges. GateFabric initially shows higher energy deviations from CISD, but optimizing 1-RDM substantially improves both the energy accuracy and the quality of 1-RDM. These results demonstrate that simultaneous optimization of energy and 1-RDM is an effective strategy to improve the accuracy of energies and molecular properties in variational quantum algorithms.