In this thesis, a method to calculate eigenpairs is implemented for the Multipsi library. While the standard implemtentations use the Davidson method with Rayleigh-Ritz extraction to calculate the eigenpairs with the lowest eigenvalues, the new method uses the harmonic Davidson method with the harmonic Rayleigh-Ritz extraction to calculate eigenpairs with eigenvalues near a chosen target. This is done for Configuration Interaction calculations and for Multiconfigurational methods. From calculations, it seems the new addition to the Multipsi library is worth investigating further as convergence for difficult systems with a lot of near-degeneracy was improved. / I denna avhandling implementeras en metod för att beräkna egenpar för Multipsi-biblioteket. Medan standardimplementeringarna använder Davidson-metoden med Rayleigh-Ritz-extraktion för att beräkna egenparen med de lägsta egenvärdena, använder den nya metoden den harmoniska Davidson-metoden med den harmoniska Rayleigh-Ritz-extraktionen för att beräkna egenparen med egenvärden nära ett valt mål. Detta görs för konfigurationsinteraktionsberäkningar och för multikonfigurationsmetoder. Utifrån beräkningarna verkar det nya tillskottet till Multipsi-biblioteket vara värt att undersöka vidare eftersom konvergensen för svåra system med mycket nära degenerering förbättrades.

In the context of the search and tuning for novel magnetic materials, transition metal compounds exhibit remarkable features where the spin-orbit interaction is crucial. The collective interactions between various effects, like spins and charges, create different classes of unique magnetic systems. For heavy transition-metal compounds, the strength of spin-orbital coupling is enhanced. The jeff. = 1/2 Mott insulating state emerges from the combination of the spin-orbit interaction and the electronic correlations. The quantum-chemistry methods are employed in this thesis to investigate single- and two-site magnetic interactions of the selected transition-metal compounds. We also provide different estimations for the single- and two-site magnetic interactions based on the level of calculation accuracy. In this thesis, we apply ab initio quantum-chemistry methods to explore the electronic and magnetic properties of several d/f compounds. The thesis structure is as follows: In Chapter 1, the introduction of the thesis provides a short discussion of the electronic correlations and magnetism in transition metal compounds. In Chapter 2, the fundamentals of the quantum chemistry wavefunction-based approach are covered. This chapter gives an overview of the applied methods in this thesis. In Chapter 3, we discuss the quantum chemistry approach to investigate the material candidates to host Kitaev physics. The technique to obtain the strength of two-site magnetic couplings, including the Kitaev coupling, is discussed in-depth. In Chapter 4, we apply the technique, which is described in Chapter 3, to investigate the two-site magnetic interactions in the H3LiIr2O6, and Cu2IrO2 compounds as Kitaev candidates. The two-site magnetic couplings are reported in these compounds. In Chapter 5, we use quantum chemistry methods to investigate the on-site electronic and magnetic properties in the KCeO2 compound where 4f1 Ce3+ ions form a triangular two-dimensional lattice with sites of effective spin-1/2. Similar ytterbiumbased delafossites had been investigated as candidates for quantum spin liquid ground states. The absence of ordinary magnetic order is characteristic of quantum spinliquid states where quantum entanglements and fractionalized excitations are enriched. In Chapter 6, the magnetic properties of Co 3d8 ions doped in the Li3N crystalline solid are discussed. The results of the quantum chemistry investigation are been set side by side along with the experiment’s results. The Co ion in such a rare environment gives rise to single-site magnetism of an easy-plane anisotropy.:Table of Contents . . . . . . . . . . . . . . . . . . . . . . iv List of Figures . . . . . . . . . . . . . . . . . . . . . . . . .vi Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .i Acknowledgements . . . . . . . . . . . . . . . . . . . . . .iii 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.1 Electronic correlations and magnetism in transition metal compounds ...........1 1.2 Thesis outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Quantum chemistry methodology . . . . . . . . . . . . . . . . .6 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 Many-electron Hartree-Fock approximation . . . . . . . . . . . . . . . 9 2.3 Multi-configurational self-consistent field and multi-reference configuration methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4 Spin-orbit interaction and g-factors calculation . . . . . . . . . . . . . 15 2.5 Embedded cluster approach . . . . . . . . . . . . . . . . . . . . . . . 18 2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3 Quantum chemistry investigation of Kitaev material candidates . . . . . . . . . . .21 3.1 Introduction to the Kitaev model . . . . . . . . . . . . . . . . . . . . 23 3.2 Kitaev materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.3 Two-site quantum chemistry calculations . . . . . . . . . . . . . . . . 36 3.4 Effective Model of Two Spin-1/2 . . . . . . . . . . . . . . . . . . . . . 38 3.5 Non-canonical correspondence between two-site QC results and the effective Hamiltonian . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.6 Pseudospin coordinate system and canonical correspondence between two-site QC results and the effective Hamiltonian . . . . . . . . . . . 51 3.7 Signs of the g-tensor in the Kitaev limit . . . . . . . . . . . . . . . . 53 3.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 4 Kitaev material candidates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.2 Details of QC calculations . . . . . . . . . . . . . . . . . . . . . . . . 64 4.3 QC investigation of H3LiIr2O6 . . . . . . . . . . . . . . . . . . . . . . 66 4.4 QC investigation of Cu2IrO3 . . . . . . . . . . . . . . . . . . . . . . . 75 4.5 Impact of local symmetries on the obtained sets of magnetic couplings ......... 82 4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 5 Ce ions in two-dimensional triangular spin-1/2 lattices . . . . . . . . . . . . . . . . . . . . 89 5.1 Spin-1/2 frustrated triangular lattice . . . . . . . . . . . . . . . . . . 90 5.2 Correlated 4f -compounds as frustrated triangular lattices . . . . . . 94 5.3 Crystal structure of KCeO2 . . . . . . . . . . . . . . . . . . . . . . . 95 5.4 QC results for the electronic structure of Ce3+ ions in KCeO2 . . . . 100 5.5 The competition of SOC and crystal field splittings in KCeO2 . . . . 102 5.6 Chapter summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 6 Co-ion substitutes with linear coordination in Li3N . . . . . . . . . . . . . . . . . . . . . . 109 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 6.2 Crystal structure of Li2(Li(1−x)Cox)N and spectroscopic measurements .......112 6.3 QC computational details . . . . . . . . . . . . . . . . . . . . . . . . 115 6.4 Ab initio QC investigation of the Co+ 3d8 electronic structure doped into Li3N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 6.5 Chapter summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135

Diese Dissertation befasst sich mit der Modellierung von Spin- und Oxidationszuständen stark korrelierter Übergangsmetallkomplexe der ersten Reihe. Verschiedene ab initio-Methoden zur Modellierung der Elektronenkorrelation wurden verwendet, basierend auf einer oder mehreren Referenzwellenfunktionen, wobei ein besonderer Schwerpunkt auf den spektroskopischen Eigenschaften von offenschaligen Molekülenliegt. Die Arbeit untersucht die Stärken und Grenzen theoretischer Parameter und Effekte, wie der Elektronenkorrelation, Multiplett-Effekte sowie Größe und Natur des aktiven Raumes. Darüber hinaus setzt sich die Studie mit entscheidenden physikalischen und chemischen Konzepten wie Kovalenz, thermodynamische Stabilität, Ligandenfeldeffekte, magnetische Austauschwechselwirkungen und der Bestimmung von Spin- und Oxidationszuständen molekularer Systeme. / This dissertation delves into the challenges and successes encountered while modeling spin and oxidation states of strongly correlated first-row transition metal complexes using various single- and multireference quantum chemical methods with a focus on X-ray spectroscopy. It thoroughly examines the strengths and limitations of theoretical approaches, addressing electron correlation, multiplet effects, active space size and nature, dynamic electron correlation corrections, truncated configuration interaction (CI) techniques, and Density Functional Theory (DFT). Furthermore, the study explores pivotal physical and chemical factors such as covalency, stability, ligand-field effects, magnetic exchange interactions, and the determination of spin and oxidation states within molecular systems.

Doped ZnO nanoparticles were synthesized by three different methods – electrochemical deposition under oxidizing conditions (EDOC) , combustion method and wet chemical synthesis – for investigating the oxygen gas sensing response. Ga-doped ZnO was mostly synthesized but also In-doped ZnO was made. The samples were analyzed by XRD, SEM, EDX and TEM. Gas response curves are given alongside with Langmuir fitted curves and data for pure ZnO and Ga-doped ZnO. DFT quantum chemical investigation of cluster models ZnO nanoparticles were performed to evaluate defect effects and oxygen and nitrogen dioxide reactions with the ZnO surface. Defects were investigated by DOS and HOMO-LUMO plots , and are oxygen vacancy, zinc vacancy, zinc interstitial and gallium doping by replacing zinc with gallium. Oxygen and nitrogen dioxide reactions were investigated by computing Mulliken charges, bond lengths, DOS spectra and HOMO-LUMO plots.

<div><div><div><p>Pincer ligands are a cornerstone of modern transition metal catalysis. An increasing interest in bimetallic catalysis motivated us to develop binucleating variants of these mononucleating ligands. Expanded variants of the PDI and PyBOX ligands were targeted, leading to the development of the Naphthyridine Diimine (“NDI”) and Naphthyridine Bisoxazoline (“NapBOX”) ligands, respectively. Metalation of NDI with appropriate metal precursors yielded Fe2, Co2 and Ni2 complexes which exhibited unique stoichiometric and catalytic reactivity. Metalation of the NapBOX ligand with nickel carboxylate salts yielded Ni2 complexes which were capable of catalyzing an asymmetric intermolecular [4+1] cycloaddition reaction between 1,1-dichloroalkene-derived vinylidenes and 1,3-dienes. Each of these processes were studied experimentally and computationally in order to understand the fundamental reactivity of organic substrates across metal-metal bonds.</p></div></div></div>

The possibility of using methylamines instead of ammonia as a nitrogen precursor for the CVD of nitrides is studied using quantum chemical computations of reaction energies: reaction electronic energy (Δ𝑟𝐸𝑒𝑙𝑒𝑐) reaction enthalpy (Δ𝑟𝐻) and reaction free energy (Δ𝑟𝐺). The reaction energies were calculated for three types of reactions: Uni- and bimolecular decomposition to more reactive nitrogen species, adduct forming with trimethylgallium (TMG) and trimethylaluminum (TMA) followed by a release of methane or ethane and surface adsorption to gallium nitride for both the unreacted ammonia or methylamines or the decomposition products. The calculations for the reaction entropy and free energy were made at both STP and CVD conditions (300°C-1300°C and 50 mbar). The ab inito Gaussian 4 (G4) theory were used for the calculations of the decomposition and adduct reactions while the surface adsorptions were calculated using the Density Functional Theory method B3LYP. From the reactions energies it can be concluded that the decomposition was facilitated by the increasing number of methyl groups on the nitrogen. The adducts with mono- and dimethylamine were more favorable than ammonia and trimethylamine. 𝑁𝐻2 was found to be most readily to adsorb to 𝐺𝑎𝑁 while the undecomposed ammonia and methylamines was not willingly to adsorb.

Dans cette thèse, on étudie plusieurs modèles et problèmes issus de la mécanique quantique. Ces modèles interviennent naturellement en chimie quantique pour le calcul de la structure électronique de la matière. Ils présentent des difficultés théoriques liées aux problèmes d'existence de solutions et à leur calcul numérique. Cette thèse est une contribution à l'étude de ces problèmes. / This thesis is concerned with several mathematical problems in quantum mechanics. These problems arise naturally in quantum chemistry in connection with the electronic structure of matter. Of particular interest are the questions of existence of solutions and of ways to compute them effectively.

Quantum computers are appealing for their ability to solve some tasks much faster than their classical counterparts. In fact, they have a potential to perform the full configuration interaction (FCI) energy calculations with a polynomial scaling only. This is in contrast to con- ventional computers where FCI scales exponentially. We provide a detailed description of the quantum version of the FCI method and the results of numerical simulations of the ground and excited state energy calculations of the methylene molecule. We further generalize this method to the relativistic four component regime and show how to efficiently solve the eigenproblem of the Dirac-Coulomb(-Breit) Hamiltonian on a quantum computer. We demonstrate the func- tionality of the proposed procedure by numerical simulations of computations of the spin-orbit splitting in the SbH molecule. Finally, we propose quantum circuits with 3 qubits and 9 or 10 CNOTs, which implement a proof-of-principle relativistic quantum chemical calculation for this molecule and might be suitable for an experimental realization. 1

Cette thèse s'intéresse à la rationalisation du mécanisme de transfert d'excitation dans des polyphénylènes éthynylènes (PPE). Une étude statique approfondie a été réalisée en utilisant la TDDFT, permettant de confirmer la localisation des états excités de méta-PPE sur des fragments para, ainsi que la hiérarchie des interactions régissant les propriétés photochimiques des PPE. Des intersections coniques ont été identifiées, de même que les principales composantes de l'espace de branchement. Leur étude a soutenu l'hypothèse d'un transfert d'énergie par conversion interne entre états excités localisés sur des fragments para.D'autre part, nous avons proposé un modèle vibronique multiéchelles pour l'énergie des états électroniques. En particulier, nous avons exprimé les énergies des orbitales frontières de PPE en fonction des énergies des orbitales frontières du benzène et de l'acetylène via un Hamiltonien effectif de type Hückel. Un travail de mapping et d'optimisation nous a permis d'aboutir à une expression pour l'énergie de transition électronique en fonction d'un nombre réduit de coordonnées nucléaires locales. / The present work is focused on the rationalization of the excitation transfer mechanism in polyphenylene ethynylenes (PPEs). A static study was performed using TDDFT, allowing to confirm both the localization of the excited states of meta-PPEs on para building blocks and the hierarchy in the interactions governing the photochemical properties of PPEs. Conical intersections were identified, along with few components of their branching spaces. Studying those supported the assumption of an energy transfer proceeding through internal conversion between excited states localized on different building blocks.In addition, we proposed a multiscale vibronic model for the energy of the eletronic states. In particular, we expressed the energies of the frontier orbitals of PPEs in terms of the energies of the frontier orbitals of benzene and acetylene, using an effective Hückel-type Hamiltonian. Perfoming different optimizations, we achieved to propose an expression for the energy of the electronic transition in terms of a reduced number of local nuclear coordinates.

Ce travail de thèse est une recherche sur l'interaction entre des états atomiques excités électroniquement et un environnement non réactif. Nous avons étudié théoriquement et expérimentalement des situations où un atome métallique (K et Ba) est placé dans un environnement de taille finie (agrégat d’argon). La présence de l'environnement affecte les niveaux électroniques de l'atome. En retour, l'excitation de l'atome induit une dynamique de relaxation de l'énergie électronique via les déformations du système atome-agrégat. La partie expérimentale du travail porte sur les deux aspects : spectroscopie et dynamique. Dans les deux cas un premier laser porte l'atome métallique dans un état électronique excité et un second l'ionise. L'observable est le spectre de photoélectrons enregistré après photo ionisation éventuellement complétée par des informations sur les photo-ions qui sont également produits. Cette technique à deux lasers conduit à des mesures de dynamique selon la technique pompe-sonde quand les lasers utilisés sont à impulsion ultracourte (60 fs). L'utilisation de lasers nanosecondes, conduit à des mesures de résonance non résolues temporellement qui donnent des informations spectroscopiques sur la position des niveaux d'énergie du système étudié. D’un point de vue théorique, les états excités du système M-Ar_n ont été calculés ab-initio en utilisant des pseudo-potentiels à grand cœur pour limiter les électrons actifs aux seuls électrons de valence du métal. L’étude d’un métal alcalin (potassium) rend cette méthode particulièrement attractive car un seul électron est actif. Le calcul ab-initio et une simulation Monte-Carlo ont été couplés pour optimiser la géométrie d'agrégats KAr_n (n=1-10) quand K est dans l'état fondamental, excité dans les états 4p ou 5s ou ionisé vers l'état fondamental de l'ion. Des calculs ont également été conduits en collaboration avec B. Gervais (CIMAP, Caen) sur des agrégats KAr_n comportant plusieurs dizaines d'atomes Ar. Des spectres d'absorption ont également été calculés. D’un point de vue expérimental, nous avons pu caractériser les niveaux électroniques excités du potassium et du baryum perturbés par l’agrégat. Dans les deux cas une bande ∏, liante, et une bande ∑, anti-liante, ont été observées. Dans le cas du potassium, nous avons montré que l’excitation dans la bande ∑ conduisait à une éjection de l’agrégat en 1-2 ps alors que pour le baryum, l’état électronique relaxe majoritairement sur l’état ∏ en ≈ 6 ps et ne conduit pas à une éjection. L’interprétation fait appel aux structures et aux potentiels calculés. Une étude équivalente a été conduite sur la molécule de DABCO déposée sur agrégat. Au contraire de K et Ba, le premier état excité de cette molécule a un fort caractère isotrope et diffus, ce qui confère un caractère particulier à la dynamique photoinduite. / This thesis is a study on the interaction between electronically excited atomic states and a non-reactive environment. We have theoretically and experimentally studied situations where a metal atom (Ba or K) is placed in a finite size environment (argon cluster). The presence of the medium affects the electronic levels of the atom. On the other side, the excitation of the atom induces a relaxation dynamics of the electronic energy through the deformation of the cluster. The experimental part of this work focuses on two aspects : the spectroscopy and the dynamics. In both cases a first laser electronically excites the metal atom and the second ionizes the excited system. The observable is the photoelectron spectrum recorded after photoionization and possibly information on the photoion which are also produced. This pump/probe technique, with also two lasers, provide the ultrafast dynamic when the lasers pulses used are of ultrashort (60 fs ). The use of nanosecond lasers leads to resonance spectroscopic measurement, unresolved temporally, which give information on the position of the energy levels of the studied system. From a theoretical point-of-view, the excited states of M-Ar_n were calculated at the ab initio level, using large core pseudo-potential to limit the active electrons of the metal to valence electrons. The study of alkali metals (potassium) is especially well adapted to this method since only one electron is active. The ab-initio calculation and a Monte-Carlo simulation where coupled to optimize the geometry of the KAr_n (n = 1-10) cluster when K is in the ground state of the neutral and the ion, or excited in the 4p or 5s state. Calculations were also conducted in collaboration with B. Gervais (CIMAP, Caen) on KAr_n clusters having several tens of argon atoms. Absorption spectra were also calculated. From an experimental point-of-view, we were able to characterize the excited states of potassium and barium perturbed by the clusters. In both cases a binding ∏-state, and an anti-binding ∑ -state were observed. In the case of potassium, we observered that the excitation on the ∑ -state leads to the ejection of the metal within 1-2 ps while for barium, the excited state relaxes mainly on the ∏-state within 6 ps and does not followed by an ejection. The interpretation of these results uses the structures and potential calculated. A similar study was conducted on the DABCO molecule deposited argon cluster. Instead of K and Ba, the first excited state of this molecule is a diffuse isotropic state, which gives a specific relaxation to the photoinduced dynamics.