Search

Dalton is a powerful general-purpose program system for the study of molecular electronic structure at the Hartree–Fock, Kohn–Sham, multiconfigurational self-consistent-field, Møller–Plesset, configuration-interaction, and coupled-cluster levels of theory. Apart from the total energy, a wide variety of molecular properties may be calculated using these electronic-structure models. Molecular gradients and Hessians are available for geometry optimizations, molecular dynamics, and vibrational studies, whereas magnetic resonance and optical activity can be studied in a gauge-origin-invariant manner. Frequency-dependent molecular properties can be calculated using linear, quadratic, and cubic response theory. A large number of singlet and triplet perturbation operators are available for the study of one-, two-, and three-photon processes. Environmental effects may be included using various dielectric-medium and quantum-mechanics/molecular-mechanics models. Large molecules may be studied using linear-scaling and massively parallel algorithms. Dalton is distributed at no cost from http://www.daltonprogram.org for a number of UNIX platforms. © 2013 John Wiley & Sons, Ltd. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, (CC BY-NC-ND 3.0)

This thesis explores time-dependent coupled cluster (TDCC) and time-dependent equation-of-motion coupled cluster (TD-EOM-CC) methods in quantum chemistry. As part of the thesis work, these methods were integrated into the eT quantum chemistry program for simulating atomic and molecular interactions with electromagnetic fields. The primary objective was to apply the methods to study processes observed at state-of-the-art laser facilities and evaluate their mathematical properties. A key aspect of the thesis involves simulations of molecular photon absorption from short UV and x-ray laser pulses in sequence, creating time-resolved depictions of electron behavior. This approach can be viewed as making movies of electron dynamics, capturing the intricacies of their behavior. Methods were also developed to reduce the computational cost of these simulations. Additionally, the thesis encompasses simulations of other multiphoton phenomena such as collective Rabi oscillations in multiple atoms and impulsive stimulated x-ray Raman scattering in strong fields. During the research, specific limitations were identified for each method in certain settings: diverging time-dependent parameters in TDCC and substantial discrepancies from expected results due to system size in TD-EOM-CC. However, in other settings, the methods behaved without these limitations. These contributions enhance our understanding of how to simulate fundamental light-driven chemical processes, which may aid our comprehension of phenomena like photosynthesis, radiation-induced DNA damage, and eyesight mechanisms.

Theoretical studies of spectroscopic properties are of great value, both by themselves and in combination with experiments. The highly accurate coupled cluster methods are attractive options for such studies. Due to their high polynomial scaling, coupled cluster methods are limited by the system size which can be treated, and also run into issues when problematic degeneracies occur. The thesis is concerned with how coupled cluster can be used to obtain spectroscopic properties of systems exhibiting such problematic attributes, specifically photoswitches and molecular crystals. Most photoswitch studies are conducted using multireference methods due to the presence of degenerate states. However, we show that CC3 can generate potential energy surfaces in good agreement with high-level multireference results. Therefore, coupled cluster could constitute a valuable tool for investigating photoswitches. In the case of molecular crystals, which are “infinite” systems, we present a finite cluster approach for obtaining properties of a local region. This approach highlights a currently underexploited feature and makes it possible to use coupled cluster methods for the study of local spectroscopic properties. The approach is used to elucidate the potential of X-rays absorption spectroscopy for the study of methane and carbon-dioxide clathrate. Methane clathrate is a potential energy source, while carbon-dioxide clathrate could be used in carbon capture and storage.

Til tross for at eksplosiver og andre energetiske materialer har blitt benyttet og studert i over hundre år, er forståelsen av deres følsomhet fremdeles svært begrenset. Gode modeller som forutsier følsomhet, kan gjøre håndteringen av slike materialer sikrere og mer effektiv, både når det gjelder utvikling, produksjon, transport, lagring, bruk og eventuell avhending. Utviklingen av slike modeller er derfor et viktig forskningsmål. I de senere år har bruken av tetthetsfunksjonalteori vært økende innen forskning på energetiske materialer, og i denne masteroppgaven benyttes teorien til å studere termisk unimolekylær dekomponering av pikrinsyre (2,4,6-trinitrofenol, forkortet til PA), (mono)metylpikrinsyre (mPA) og dimetylpikrinsyre (dmPA). Til tross for at det er kun én eller to metylgrupper som skiller disse molekylenes strukturer fra hverandre, er deres slagfølsomhet drastisk forskjellig. Denne anomale oppførselen har vært kjent i mer enn 40 år, men årsaken fortsatt ukjent. I arbeidet blir tre unimolekylære reaksjonsveier, som er antatt å være viktige for initiering av orto-nitrofenoler og orto-nitrotoluener, studert for PA, mPA og dmPA. Energetiske forhold til disse reaksjonsveiene beregnes og sammenlignes for å rasjonalisere følsomhetsforskjellene. Resultatene viser få signifikante energidifferanser mellom PA, mPA, og dmPA, og resultatene reflekterer derfor ikke forskjellene i eksperimentelt målt følsomhet. Det kan være flere årsaker til dette. For det første kan bimolekylære reaksjonstrinn være viktigere enn unimolekylære når faste stoffer dekomponerer. For det andre kan energetiske forhold ved de unimolekylære reaksjonsveiene avhenge av om reaksjonene forløper i gassfase eller i fast fase. Dersom de funksjonelle gruppene til PA, mPA og/eller dmPA i stor grad bøyes ut av ringplanet på grunn av krystallpakkingen, kan energiberegningene påvirkes. For det tredje kan faststoffeffekter som krystalldefekter, molekylær stabling og partikkelstørrelse være avgjørende for slagfølsomheten til de studerte stoffene. Mer forskning er derfor nødvendig før klare konklusjoner kan trekkes. For videre arbeid foreslås karakterisering av krystallstrukturene til mPA og dmPA samt molekyldynamikksimuleringer av initieringsreaksjonene. Det kan gi informasjon som gjør det lettere å fastslå den relative vikigheten av uni- og bimolekylære reaksjoner i kondensert fase. I tillegg kan krystallstrukturene avdekke fysikalske egenskaper ved materialene, noe som er avgjørende for slag-følsomheten. Det kan også være nyttig å gjenta beregningene utført i denne masteroppgaven for de molekylære geometriene som finnes i krystallstrukturene. Although energetic materials such as explosives have been used and studied for more than a hundred years, fundamental understanding of the underlying processes that govern energetic material sensitivity is yet to be established. Theoretical sensitivity models with high predictive power could significantly improve the efficiency, safety, and sustainability of the production and use of energetic materials, and are thus highly desired. Recent years have seen an increase in the use of density functional theory in research on energetic materials. In this thesis, the theory is employed for the study of thermal unimolecular decomposition of picric acid (2,4,6-trinitrophenol, abbreviated to PA), (mono)methyl picric acid (mPA), and dimethyl picric acid (dmPA). While the molecular structures of these species only differ by one to two methyl substituents, their sensitivities to impact are drastically different. This anomalous sensitivity behaviour has been observed for more than 40 years, but its cause is nevertheless still unclear. In this thesis, three unimolecular reaction pathways believed important for the initiation of ortho-nitrophenols and ortho-nitrotoluenes are investigated for PA, mPA, and dmPA. The energetics of these reaction pathways are calculated and compared with the objective of rationalising the unexpected sensitivity behavio-ur of the three explosives. The results reveal few significant differences in the energetics between PA, mPA, and dmPA. In other words, the results do not reflect the large sensitivity differences observed in experiments. There are several possible explanations why this is the case. Firstly, bimolecular reactions may take place to a significant extent in the condensed phases of the studied explosives, altering the reaction energetics. Secondly, the gas phase calculations may not be representative for the studied unimolecular reaction pathways when they occur in the condensed phase, due to twisting and bending of the molecules as they are packed in their crystal structures. Thirdly, solid state effects such as crystal defects, molecular stacking, and particle size may be determining factors for the sensitivity. Hence, more work is needed before firm conclusions can be drawn. Recommendations for future work include determining the crystal structures of mPA and dmPA and performing molecular dynamics simulations of the initiation reactions. This will be helpful for assessing the feasibility and/or relative dominance of uni- and bimolecular reactions in the condensed phases. Redoing the calculations of this thesis using the molecular geometries from the crystal structures might also be useful in this context. Additionally, the crystal structures might reveal physical properties of the materials that can be determining for their impact sensitivities.

Nanoelektronikk er er fagfelt som blir mer og mer relevant nå som elektronikk blir mindre og raskere, og med rask utvikling av teknologier som superledere og kvantedatamaskiner. Dette er et forskningsfelt der det har vært mye fremgang i mange tiår. Denne fremgangen ser ikke ut til å stoppe. Et spesifikt område der det forsatt er stor utvikling er kvantemekanisk optoelektronikk, hvor vi ser på interaksjonen mellom lys og molekylære ledere, som det transporteres elektroner gjennom. Det finnes enda ikke et godt rammeverk for å teoretisk modellere denne interaksjonen. I denne avhandlingen introduserer vi en metode basert på lineær responsteori, et rammeverk hvor det å beskrive optiske egenskaper til molekyler er vanlig, for å beskrive elektronisk transport mellom forskjellige deler av et lukket molekylært system. Med dette gir vi noen enkle proof-of-concept beregninger for to enkle molekyler, ett hvor vi ser på transporten mellom et hydrogenatom og et litiumatom, og ett annet hvor vi ser på transport mellom et berylliumatom og et oksygenmolekyl. Disse beregningene er basert på modelperturbasjoner som ikke vil representere kvantitativt nøyaktige forutsigelser, men heller er gjort for å vise potensialet for videreutviklingen av en slik metode for å gi en god beskrivelse av like transportfenomener. For noen former av slike perturbasjoner har vi observert en forflytning på rundt 0.9 elektroner per perturbasjonsstyrke på 1 a.u. Dette har blitt gjort sammen med en beregning av elektrisk strøm gjennom et modellsystem av tre hydrogenmolekyl under en påført spenning, ved å bruke en ikke-likevekts Greens funksjonsmetode (Non-Equlibrium Green's Function, NEGF), som er en av de fundamentale metodene for å beskrive kvantemekanisk transport. Dette er gjort for å ha en lignende beregning for å kunne sammenligne responseteorimetoden med, både den interne matematikken av de to metodene, og de beregnede resultatene. Likevel, siden responseteorimetoden ikke enda inkluderer en beskrivelse av kontakter til elekstronreservoarer som er essensielt for å ha kontinuerlig strøm. Derfor kan vi ikke direkte sammenligne resultatene fra disse to metodene, som vi kunne ha gjort dersom metodene beregnet samme størrelse. The field of nanoelectronics is becoming increasingly relevant as electronics become smaller and faster and with the rapid development in technologies such as superconductors and quantum computers. This is a field that has seen a lot of progress for the last many decades and continues to do so. One particular area that is still seeing a lot of development is quantum optoelectronics, where we look at the interaction between light and molecular conducting channels undergoing electronic transport. There does not yet exist a rigorous framework for theoretically modeling this interaction properly. In this thesis we introduce a method based on linear response theory, a framework where describing optical properties of molecules is commonplace, for describing electronic transport between different parts of a closed molecular system. With this, we provide some simple proof-of-concept calculations for two simple molecules, one where we calculate the transport of electrons between a hydrogen and lithium atom, and the other calculating transport between a beryllium atom and an oxygen molecule. These calculations are based on model perturbations that will not represent quantitatively accurate predictions, but rather serves to show the potential of the method to be further developed for a good description of such transport phenomena. For certain types of perturbations, we observed a movement of the electronic density corresponding to about 0.9 electrons per perturbation strength of 1 a.u. This has been done alongside a calculation of the steady state current through a model system consisting of three hydrogen molecules under a potential bias, using the Non-Equilibrium Green's Function (NEGF) method, which is one of the fundamental pillars of describing quantum transport. This is done to have a similar calculation to compare the response theory method to, both in terms of the internal mathematics of the two methods, and the predicted results. However, the response theory method does not yet include a description of contacts to external electron reservoirs, which is required for steady state transport. As such, the results from the two methods is not as directly comparable as if they both calculated the same quantity.

We introduce a density-based multilevel Hartree-Fock (HF) method where the electronic density is optimized in a given region of the molecule (the active region). Active molecular orbitals (MOs) are generated by a decomposition of a starting guess atomic orbital (AO) density, whereas the inactive MOs (which constitute the remainder of the density) are never generated or referenced. The MO formulation allows for a significant dimension reduction by transforming from the AO basis to the active MO basis. All interactions between the inactive and active regions of the molecule are retained, and an exponential parametrization of orbital rotations ensures that the active and inactive density matrices separately, and in sum, satisfy the symmetry, trace, and idempotency requirements. Thus, the orbital spaces stay orthogonal, and furthermore, the total density matrix represents a single Slater determinant. In each iteration, the (level-shifted) Newton equations in the active MO basis are solved to obtain the orbital transformation matrix. The approach is equivalent to variationally optimizing only a subset of the MOs of the total system. In this orbital space partitioning, no bonds are broken and no a priori orbital assignments are carried out. In the limit of including all orbitals in the active space, we obtain an MO density-based formulation of full HF.

Netthinnen, som er et tynt lag med celler på baksiden av øyeeplet, inneholder to viktige typer proteiner, tapp- og stav-synspigmentene. Stav-synspigmenter eller rhodopsin er ansvarlig for synet i lav lysintensitet, og tapp-pigmenter, som består av røde, grønne og blå tapper, er ansvarlige for synet i dagslys. De inneholder alle retinal Schiff-base som kromofor omgitt av et protein som innstiller absorpsjonsspekteret fra 365-430 nm til 498 nm for rhodopsin og 425 nm for blå tapper. En rekke studier, fra spektroskopiske undersøkelser til teoretiske studier, er rettet mot å avdekke proteinmiljøets rolle i dette signifikante absorpsjonsskiftet. Elektrostatisk effekt av protein-residues, og enda vik tigere det antatte motionet av Schiff-basen, og fordreining av retinen påført av proteinet, er vidt aksepterte mekanismer som hittil er gitt. Ved å bruke coupled cluster med tilnærmet singles og doubles-metoden og resolution of identity-tilnærming (RICC2) som ab initio-metode sammen med tetthetsfunksjonalteori, har vi studert videre den mulige rollen residues spiller i absorpsjonsskiftet, til sammenligning med motionets rolle. Vår undersøkelse bekrefter den viktige rollen Glu-113 spiller i rhodopsin-absorpsjon. I blå tapper instilles derimot absorpsjonsspekteret ved deprotonering av Schiff-basen og en gruppe residues. Vi undersøker også effekten av forskjellige protoneringstilstander av Glu-181 og Glu-178 i henholdsvis rhodopsin og blå tappers absorpsjon. Vi konkluderer med at et negativt ladet glutamat kan forstyrre effektiviteten av fotoisomerisering i begge pigmenter. Til slutt undersøker vi protonaffiniteten til Schiff-basen i både rhodopsin og blå tapper. Retina, which is a thin layer of cells in the back of the eyeball, contains two important types of proteins, cone and rod visual pigments. Rod visual pigment or rhodopsin is responsible for vision in low light intensity and cone pigments, consisting of red, green, and blue cones, are responsible for day-light vision. All of them contain retinal Schiff base as the chromophore and a surrounding protein which tunes the absorption spectrum from 365-430 nm to 498 nm for rhodopsin and 425 nm for the blue cone. Numerous studies, from spectroscopic investigations to theoretical studies are devoted to unravel the protein environment role in this significant absorption shift. The electrostatic effect of residues, more importantly, the putative counterion of the Schiff base, and distortion of the retinal imposed by the protein are well-accepted mechanisms so far provided. Using coupled cluster with approximate singles and doubles method and the resolution of identity approximation (RICC2) as an ab initio method together with density functional theory allowed us to dig more into the possible role of protein residues in the shifting of the absorption spectrum, as compared to the role of the counterion. Our investigation confirms the important role of Glu-113 in rhodopsin absorption. Conversely, in the blue cone, deprotonation of the Schiff base and group of residues tune the absorption spectrum. We also investigate the effect of different protonation states of Glu-181 and Glu-178 in rhodopsin and blue cone absorption respectively. We conclude that a negatively charged glutamate might disrupt the efficiency of photoisomerization in both pigments. Finally, we investigate the proton affinity of the Schiff base in both rhodopsin and blue cone.

Optimering av likevektsgeometrier er et viktig trinn i studiet av diverse molekylære egenskaper. Mangle slike egenskaper er lokale, hovedsaklig knyttet til en liten region av det totale molekylære systemet. Dette gjør multilevel-metoder passende for beregning av lokale egenskaper, ettersom disse metodene modellerer ulike regioner på ulike nivå av teori, og derav tillater nøyaktige beskrivelser av regionen man er interessert i til redusert beregningskostnad. Denne avhandlingen ser på en nylig introdusert multilevel Hartree-Fock (MLHF) metode, og videreutvikler denne metoden for optimering av likevektsgeometrier. MLHF-metoden er en to-nivå-metode, som modellerer den aktive regionen med en HF-optimert tetthet og den inaktive regionen med et startgjett for tettheten. Den molekylære gradienten for MLHF utledes analytisk, og implementeres på en slik måte at den inaktive regionen er frosset i rommet mens geometrien optimeres. Implementering og testing av metoden utføres i det nylig utgitte elektronstrukturprogrammet eT. Likevektsgeometrier for fire små testsystemer beregnes og undersøkes i forhold til energi og et utvalg av bindingslengder og bindingsvinkler. Sammenlignet med HF-optimerte geometrier er MLHF-geometriene høyere i energi, med en energiforskjell på størrelsesorden 10^(-2) Hartree eller mindre. Videre sammenligninger gjøres med HF-geometrier der den samme regionen som var frosset for MLHF-beregningen også fryses for HF-beregningen. MLHF-geometriene er forsatt høyere i energi en disse ''frossen-HF''-geometriene, men forskjellene er kun på størrelsesorden 10^(-3) Hartree. Forskjellene i de utvalgte bindingslengdene og bindingsvinklene for disse to metodene er innenfor henholdsvis 2.8 pm og 2.6 grader. Dette er innenfor de vanlige intervallene for feil på HF-nivået av teori, og til gjengjeld gir MLHF-metoden en kraftig reduksjon i antall orbitaler som inngår i optimeringen. Effekten av å utvide det MLHF-aktive rommet utover den ikke-frosne regionen undersøkes, og de korresponderende geometriene nærmer seg ikke frossen-HF-geometrien. Dette motstrider hva som er forventet, og kan muligens skyldes en forglemmelse i implementeringen, men ingen definitiv forklaring er blitt funnet. Resultatene som er blitt oppnådd kommer bare fra små systemer, og sammenlignes kun innenfor HF-nivået av teori. Disse begrensningene for resultatene må riktignok huskes på, men MLHF geometrioptimering virker likevel lovende når det gjelder utsikter for bruk på store systemer, og for å fungere som et startpunkt for nøyaktige korrelerte metoder. Et av leddene i den molekylære MLHF-gradienten er beregningsmessig kostbart, og kan representere en utfordring for utviklingen av en effektiv implementering av metoden. Et par approksimasjoner til dette leddet utprøves, basert på enkle approksimative oppdateringsalgoritmer, men det viser seg at de introduserer en alvorlig forstyrrelse i gradienten, hvilket gjør dem ubrukelige. The optimization of equilibrium geometries is an important step in the study of various molecular properties. Many such properties are local, connected mostly to a small region of the entire molecular system. This makes multilevel methods appropriate for the calculation of localized properties, as these methods model different regions at different levels of theory, allowing accurate descriptions of the region of interest at reduced computational cost. This thesis concerns a recently introduced multilevel Hartree-Fock (MLHF) method, and develops this method for equilibrium geometry optimization. The MLHF method is a two-level method, modelling the active region with a HF optimized density and the inactive region with the starting guess density. The MLHF molecular gradient is analytically derived, and implemented in such a way that the inactive region is frozen in space during the geometry optimization. The implementation and testing of the method is done in the newly released electronic structure program eT. Equilibrium geometries of four small test systems are obtained and investigated in terms of energy and a selection of bond lengths and bond angles. When compared with HF optimized geometries, the MLHF geometries are higher in energy, with energy differences on the order of 10^(-2) Hartree or less. Further comparisons are made to HF optimized geometries where the same region that was frozen in space for the MLHF calculation is also frozen for the HF calculation. The MLHF geometries are still higher in energy than these ''frozen HF'' geometries, but the differences are only on the order of 10^(-3) Hartree. The differences in the selection of bond lengths and bond angles for these two methods are within 2.8 pm and within 2.6 degrees, respectively. This is within the usual error bounds of the HF level of theory, and in return the MLHF method gives a large reduction in the number of orbitals used in the optimization. The effect of extending the MLHF active space beyond the unfrozen region is explored, and the corresponding geometries fail to approach the frozen HF geometry. This is contrary to what is expected, and might be the result of an oversight in the implementation; however, no conclusive explaination has been determined. The results obtained are only from small systems, and are compared only within the HF level of theory. While these limitations of the results must be kept in mind, the MLHF geometry optimization nevertheless appears promising for the prospect of application on large systems and providing a starting point for accurate correlated methods. One of the MLHF molecular gradient terms is computationally expensive and could pose a challenge for the efficient implementation of the method. A couple of approximations to this term are attempted, based on simple approximate updating schemes, but they are found to introduce a severe disturbance in the gradient, which makes them unusable.

Modern quantum chemical calculations commonly use wave function based models or densitry functionaltheory (DFT). Using Green’s functions is an approach that has not been explored much, and is one thathas only been explored relatively recently, with older examples being useful, but limited in scope. Theuse of Green’s functions in literature, historically and contemporary, is outlined and explored, with afocus on the self-consistent second order Green’s function (GF2), as well as the GW-approximation.

Elektrolyttmodeller kan klassifiseres som primitiv eller ikke-primitiv. Primitive modeller representerer ikke løsningsmidlet som partikler eller molekyler, men som et dielektrisk kontinuum karakterisert av en permittivitet. Dette er i motsetning til ikke-primitive elektrolyttmodeller som representerer løsningsmidlet på et molekylært nivå. Denne oppgaven undersøker hvordan modellering av et løsningsmiddel ved bruk av en primitiv eller ikke-primitiv tilnærming forandrer oppførselen til fortynnede elektrolyttløsninger. Spesifikt så undersøker dette arbeidet løsningsmidlets rolle i elektrolyttløsninger ved å etablere en ikke-primitiv elektrolyttmodell og sammenligne dens dielektriske og termodynamiske egenskaper med den primitive Debye–Hückelteorien i grensen ved uendelig fortynning. Den ikke-primitive elektrolyttmodellen som brukes i dette arbeidet består av anioner, kationer og løsningsmiddelpartikler av samme størrelse. Ionene og løsningsmiddelpartiklene interager gjennom et mykt-kjernepotensial med en punktladning eller en punkt-dipol. Molekylær dynamiske simuleringer brukes for å estimere de termodynamiske og dielektriske egenskapene til den ikke-primitive elektrolyttmodellen. En etterbehandlingsprosedyre av partikkelkonfigurasjonene fra simuleringene viser at permittiviteten til den ikke-primitive elektrolyttmodellen avtar med økende ionkonsentrasjon. Dette fenomenet blir også observert i ekte elektrolyttløsninger og er kjent som dielektrisk dekrement. Debye–Hückelteorien tar ikke hensyn til dielektrisk dekrement da den modellerer løsningsmidlet som et kontinuerlig dielektrikum med konstant permittivitet. Dette arbeidet utleder et utrykk for den gjennomsnittlige ioniske aktivitetskoeffisienten ved bruk av en Debye–Hückel inspirert teori som ikke antar at løsningens permittivitet er uavhengig av ionkonsentrasjon. Det resulterende rekkeutviklingsutrykket viser at å introdusere en permittivitet som er avhengig av ionkonsentrasjon ikke forandrer grenselovene til Debye–Hückelteori. De uendrede grenselovene antyder at dielektrisk dekrement ikke endrer den ikke-primitive modellens termodynamikk i grensen ved uendelig fortynning. En regresjonsmetode brukes til å undersøke den ikke-primitive elektrolyttmodellens asymptotiske termodynamiske oppførsel. Siden Debye–Hückelteorien korrekt beskriver termodynamikken til fortynnede elektrolyttløsninger, bør termodynamikken til en riktig ikke-primitiv elektrolyttmodell konvergere til lignende fysisk oppførsel som Debye–Hückelteorien ved uendelig fortynning. Grensen for uendelig fortynning er uoppnåelig ved bruk av dette arbeidets molekylær dynamikkprosedyre. I stedet demonstrerer denne avhandlingen at det er mulig å gjenvinne Debye–Hückelteoriens asymptotiske oppførsel fra den ikke-primitive elektrolyttmodellens molekylær dynamikkresultater ved høye temperaturer. Det kreves ytterligere arbeid for å avgjøre om det er mulig å gjenvinne Debye–Hückelteoriens asymptotiske oppførsel ved å redusere ionenes ladning. Electrolyte models can be classified as primitive or non-primitive. Primitive models do not represent the solvent as particles or molecules, but as a dielectric continuum characterized by a permittivity. Oppositely, non-primitive electrolyte models explicitly consider the solvent on a molecular level. This thesis focuses on how modeling the solvent using a primitive or non-primitive approach changes the behavior of dilute electrolyte solutions. Specifically, the solvent's role in electrolyte solutions is investigated by establishing a non-primitive electrolyte model and comparing its dielectric and thermodynamic properties to the primitive Debye–Hückel theory in the limit of infinite dilution. The non-primitive electrolyte model used in this work consists of an anion, a cation, and a solvent species of the same size. The ions and solvent particles interact according to a soft-core potential with a point-charge or a point-dipole at their respective centers. Molecular dynamics simulations are used to estimate the thermodynamic and dielectric properties of the non-primitive electrolyte model. Post-processing the particle configurations from the molecular dynamics simulations demonstrate that the non-primitive electrolyte model's permittivity decreases with increasing ion concentration. This phenomenon is also observed in real electrolyte solutions and is known as dielectric decrement. Debye–Hückel theory does not consider dielectric decrement as it models the solvent as a continuous dielectricum of constant permittivity. This work derives an expression for the mean ionic activity coefficient using a Debye–Hückel inspired theory that does not assume that the solution's permittivity is independent of the concentration of ions. The resulting series expansion demonstrates that introducing a permittivity that depends on ion concentration does not alter the Debye–Hückel limiting law. The unaltered limiting law suggests that dielectric decrement does not change the non-primitive model's thermodynamics in the limit of infinite dilution. A regression method is used to investigate the non-primitive electrolyte model's asymptotic thermodynamic behavior. Since the Debye–Hückel theory correctly describes the thermodynamics of sufficiently dilute real electrolytes, the thermodynamics of a correct non-primitive electrolyte model should converge to that of the Debye–Hückel theory in the limit of infinite dilution. However, the limit of infinite dilution is unattainable using this work's molecular dynamics procedure. Instead, this thesis demonstrates that it is possible to recover the Debye–Hückel theory's asymptotic behavior in the limit of infinite dilution from the non-primitive electrolyte model's molecular dynamics results at high temperatures. Further work is required to determine if it is possible to recover the Debye–Hückel theory's asymptotic behavior by sufficiently reducing the magnitude of the ions' charge.

_ _

In this article, we present a black-box approach for the selection of orbital spaces when computing core excitation energies in the multilevel coupled cluster (MLCC) framework. Information available from the lower level of theory is used to generate correlated natural transition orbitals (CNTOs) for the high-level calculation by including both singles and doubles information in the construction of the transition orbitals. The inclusion of the doubles excitation information is essential to obtain a set of orbitals that all contain physical information, in contrast to the natural transition orbitals where only a small subset of the virtual orbitals contains physical information. The CNTOs may be included in an active space based on a cutoff threshold for the eigenvaluescorresponding to the orbitals. We present MLCC results for core excitation energies calculated using coupled cluster singles and doubles (CCSD) in the inactive space and CCSD with perturbative triples (CC3) in the active space. The use of CNTOs results in small errors compared to full CC3. Published by AIP Publishing. Locked until 31.3.2018 due to copyright restrictions. This article may be downloaded for personal use only. Any other use requires prior permission of the author and AIP Publishing. The following article appeared in Journal of Chemical Physics and may be found at https://aip.scitation.org/doi/full/10.1063/1.4979908

Embedding, or multilevel, schemes have become exceedingly popular in recent years. The idea is to divide the system into an active part, which is treated by an accurate quantum mechanical method, and then treat the rest of the system, referred to as the inactive part, with a less accurate, more computationally efficient method. In this work, recent developments in multilevel Hartree-Fock theory (MLHF) is connected with coupled cluster singles and doubles (CCSD) to enable computation of accurate size-intensive properties for large molecular systems. The method needs no a priori orbital assignment, no bonds are broken and one wavefunction describes the whole system. One part of the wave function is described at CCSD level, one part at HF level and the other kept constant after an initial density start guess. The objective is to explore whether a full HF calculation is needed for a reduced space CC model, or if only a region around the CCSD space needs to be optimized. Specifically, will the reduced space CCSD results be compromised when embedded in a MLHF wave function rather than a HF wave function. The MLHF method makes use of a molecular orbital (MO) density matrix driven optimization procedure to optimize only a subset of the MOs in the Slater determinant by using an orthogonal parametrization that ensures conserved symmetry, trace and idempotency properties for both the active and inactive density matrix. This is connected with CCSD by generating a set of canonical MOs from the active density, and using these MOs and the orbital energies as a starting point for the CCSD calculation. Since the reference wave function is not a HF wave function, this introduces some changes in the CCSD working equations. All changes enters through a modified Fock matrix. Some benchmarking calculations on core excitations have been carried out on smaller systems to establish the effect of only having a partly optimized HF state as a reference state, and the method is showing promising results. The errors are small, but still large enough to compromise the CCSD calculation. However, as this is for small systems, these preliminary results are promising for the application on larger molecular systems. From the results, it is expected that local CC calculations on large systems can be carried out in a MLHF framework without compromising accuracy.