Computer simulations of charged systems in partially periodic geometries

This work presents algorithms for the calculation of the electrostatic interaction in partially periodic systems. The framework for these algorithms is provided by the simulation package ESPResSo, of which the author was one of the main developers. The prominent features of the program are listed and the internal structure is described. In the following, algorithms for the calculation of the Coulomb sum in three dimensionally periodic systems are described. These methods are the foundations for the algorithms for partially periodic systems presented in this work. Starting from the MMM2D method for systems with one non-periodic coordinate, the ELC method for these systems is developed. This method consists of a correction term which allows to use methods for three dimensional periodicity also for the case of two periodic coordinates. The computation time of this correction term is neglible for large numbers of particles. The performance of MMM2D and ELC are demonstrated by results from the implementations contained in ESPResSo. It is also discussed, how different dielectric constants inside and outside of the simulation box can be realized. For systems with one periodic coordinate, the MMM1D method is derived from the MMM2D method. This method is applied to the problem of the attraction of like-charged rods in the presence of counterions, and results of the strong coupling theory for the equilibrium distance of the rods at infinite counterion-coupling are checked against results from computer simulations. The degree of agreement between the simulations at finite coupling and the theory can be characterized by a single parameter gamma_RB. In the special case of T=0, one finds under certain circumstances flat configurations, in which all charges are located in the rod-rod plane. The energetically optimal configuration and its stability are determined analytically, which depends on only one parameter gamma_z, similar to gamma_RB. These findings are in good agreement with results from computer simulations.

The massless two-loop two-point function and zeta functions in counterterms of Feynman diagrams

The main part of this thesis describes a method of calculating the massless two-loop two-point function which allows expanding the integral up to an arbitrary order in the dimensional regularization parameter epsilon by rewriting it as a double Mellin-Barnes integral. Closing the contour and collecting the residues then transforms this integral into a form that enables us to utilize S. Weinzierl's computer library nestedsums. We could show that multiple zeta values and rational numbers are sufficient for expanding the massless two-loop two-point function to all orders in epsilon. We then use the Hopf algebra of Feynman diagrams and its antipode, to investigate the appearance of Riemann's zeta function in counterterms of Feynman diagrams in massless Yukawa theory and massless QED. The class of Feynman diagrams we consider consists of graphs built from primitive one-loop diagrams and the non-planar vertex correction, where the vertex corrections only depend on one external momentum. We showed the absence of powers of pi in the counterterms of the non-planar vertex correction and diagrams built by shuffling it with the one-loop vertex correction. We also found the invariance of some coefficients of zeta functions under a change of momentum flow through these vertex corrections.

O (alpha 4 s) loop-by-loop contributions to heavy quark pair production in hadronic collisions

The present state of the theoretical predictions for the hadronic heavy hadron production is not quite satisfactory. The full next-to-leading order (NLO) ${cal O} (alpha_s^3)$ corrections to the hadroproduction of heavy quarks have raised the leading order (LO) ${cal O} (alpha_s^2)$ estimates but the NLO predictions are still slightly below the experimental numbers. Moreover, the theoretical NLO predictions suffer from the usual large uncertainty resulting from the freedom in the choice of renormalization and factorization scales of perturbative QCD.In this light there are hopes that a next-to-next-to-leading order (NNLO) ${cal O} (alpha_s^4)$ calculation will bring theoretical predictions even closer to the experimental data. Also, the dependence on the factorization and renormalization scales of the physical process is expected to be greatly reduced at NNLO. This would reduce the theoretical uncertainty and therefore make the comparison between theory and experiment much more significant. In this thesis I have concentrated on that part of NNLO corrections for hadronic heavy quark production where one-loop integrals contribute in the form of a loop-by-loop product. In the first part of the thesis I use dimensional regularization to calculate the ${cal O}(ep^2)$ expansion of scalar one-loop one-, two-, three- and four-point integrals. The Laurent series of the scalar integrals is needed as an input for the calculation of the one-loop matrix elements for the loop-by-loop contributions. Since each factor of the loop-by-loop product has negative powers of the dimensional regularization parameter $ep$ up to ${cal O}(ep^{-2})$, the Laurent series of the scalar integrals has to be calculated up to ${cal O}(ep^2)$. The negative powers of $ep$ are a consequence of ultraviolet and infrared/collinear (or mass ) divergences. Among the scalar integrals the four-point integrals are the most complicated. The ${cal O}(ep^2)$ expansion of the three- and four-point integrals contains in general classical polylogarithms up to ${rm Li}_4$ and $L$-functions related to multiple polylogarithms of maximal weight and depth four. All results for the scalar integrals are also available in electronic form. In the second part of the thesis I discuss the properties of the classical polylogarithms. I present the algorithms which allow one to reduce the number of the polylogarithms in an expression. I derive identities for the $L$-functions which have been intensively used in order to reduce the length of the final results for the scalar integrals. I also discuss the properties of multiple polylogarithms. I derive identities to express the $L$-functions in terms of multiple polylogarithms. In the third part I investigate the numerical efficiency of the results for the scalar integrals. The dependence of the evaluation time on the relative error is discussed. In the forth part of the thesis I present the larger part of the ${cal O}(ep^2)$ results on one-loop matrix elements in heavy flavor hadroproduction containing the full spin information. The ${cal O}(ep^2)$ terms arise as a combination of the ${cal O}(ep^2)$ results for the scalar integrals, the spin algebra and the Passarino-Veltman decomposition. The one-loop matrix elements will be needed as input in the determination of the loop-by-loop part of NNLO for the hadronic heavy flavor production.

Computer simulations of two-dimensional colloidal crystals in confinement

Monte Carlo simulations are used to study the effect of confinement on a crystal of point particles interacting with an inverse power law potential in d=2 dimensions. This system can describe colloidal particles at the air-water interface, a model system for experimental study of two-dimensional melting. It is shown that the state of the system (a strip of width D) depends very sensitively on the precise boundary conditions at the two ``walls'' providing the confinement. If one uses a corrugated boundary commensurate with the order of the bulk triangular crystalline structure, both orientational order and positional order is enhanced, and such surface-induced order persists near the boundaries also at temperatures where the system in the bulk is in its fluid state. However, using smooth repulsive boundaries as walls providing the confinement, only the orientational order is enhanced, but positional (quasi-) long range order is destroyed: The mean-square displacement of two particles n lattice parameters apart in the y-direction along the walls then crosses over from the logarithmic increase (characteristic for $d=2$) to a linear increase (characteristic for d=1). The strip then exhibits a vanishing shear modulus. These results are interpreted in terms of a phenomenological harmonic theory. Also the effect of incommensurability of the strip width D with the triangular lattice structure is discussed, and a comparison with surface effects on phase transitions in simple Ising- and XY-models is made

Hydrodynamical simulations of galaxy clusters in dark energy cosmologies

In this work we investigate the influence of dark energy on structure formation, within five different cosmological models, namely a concordance $\Lambda$CDM model, two models with dynamical dark energy, viewed as a quintessence scalar field (using a RP and a SUGRA potential form) and two extended quintessence models (EQp and EQn) where the quintessence scalar field interacts non-minimally with gravity (scalar-tensor theories). We adopted for all models the normalization of the matter power spectrum $\sigma_{8}$ to match the CMB data. For each model, we perform hydrodynamical simulations in a cosmological box of $(300 \ {\rm{Mpc}} \ h^{-1})^{3}$ including baryons and allowing for cooling and star formation. We find that, in models with dynamical dark energy, the evolving cosmological background leads to different star formation rates and different formation histories of galaxy clusters, but the baryon physics is not affected in a relevant way. We investigate several proxies for the cluster mass function based on X-ray observables like temperature, luminosity, $M_{gas}$, and $Y_{X}$. We confirm that the overall baryon fraction is almost independent of the dark energy models within few percentage points. The same is true for the gas fraction. This evidence reinforces the use of galaxy clusters as cosmological probe of the matter and energy content of the Universe. We also study the $c-M$ relation in the different cosmological scenarios, using both dark matter only and hydrodynamical simulations. We find that the normalization of the $c-M$ relation is directly linked to $\sigma_{8}$ and the evolution of the density perturbations for $\Lambda$CDM, RP and SUGRA, while for EQp and EQn it depends also on the evolution of the linear density contrast. These differences in the $c-M$ relation provide another way to use galaxy clusters to constrain the underlying cosmology.

Simulazioni Software-In-the-Loop per la verifica funzionale di un sistema di tracking satellitare (ALMATracker)

L'ALMATracker è un sistema di puntamento per la stazione di terra di ALMASat-1. La sua configurazione non segue la classica Azimuth-Elevazione, bensì utilizza gli assi α-β per evitare punti di singolarità nelle posizioni vicino allo zenit. Ancora in fase di progettazione, utilizzando in congiunta SolidWorks e LabVIEW si è creato un Software-in-the-loop per la sua verifica funzionale, grazie all'utilizzo del relativamente nuovo pacchetto NI Softmotion. Data la scarsa esperienza e documentazione che si hanno su questo recente tool, si è prima creato un Case Study che simulasse un sistema di coordinate cilindriche in modo da acquisire competenza. I risultati conseguiti sono poi stati sfruttati per la creazione di un SIL per la simulazione del movimento dell'ALMATracker. L'utilizzo di questa metodologia di progettazione non solo ha confermato la validità del design proposto, ma anche evidenziato i problemi e le potenzialità che caratterizzano questo pacchetto software dandone un analisi approfondita.

Numerical simulations of microphysical processes in pyro-convective clouds

Hochreichende Konvektion über Waldbränden ist eine der intensivsten Formen von atmosphärischer Konvektion. Die extreme Wolkendynamik mit hohen vertikalen Windgeschwindigkeiten (bis 20 m/s) bereits an der Wolkenbasis, hohen Wasserdampfübersättigungen (bis 1%) und die durch das Feuer hohen Anzahlkonzentration von Aerosolpartikeln (bis 100000 cm^-3) bilden einen besonderen Rahmen für Aerosol-Wolken Wechselwirkungen.Ein entscheidender Schritt in der mikrophysikalischen Entwicklung einer konvektiven Wolke ist die Aktivierung von Aerosolpartikeln zu Wolkentropfen. Dieser Aktivierungsprozess bestimmt die anfängliche Anzahl und Größe der Wolkentropfen und kann daher die Entwicklung einer konvektiven Wolke und deren Niederschlagsbildung beeinflussen. Die wichtigsten Faktoren, welche die anfängliche Anzahl und Größe der Wolkentropfen bestimmen, sind die Größe und Hygroskopizität der an der Wolkenbasis verfügbaren Aerosolpartikel sowie die vertikale Windgeschwindigkeit. Um den Einfluss dieser Faktoren unter pyro-konvektiven Bedingungen zu untersuchen, wurden numerische Simulationen mit Hilfe eines Wolkenpaketmodells mit detaillierter spektraler Beschreibung der Wolkenmikrophysik durchgeführt. Diese Ergebnisse können in drei unterschiedliche Bereiche abhängig vom Verhältnis zwischen vertikaler Windgeschwindigkeit und Aerosolanzahlkonzentration (w/NCN) eingeteilt werden: (1) ein durch die Aerosolkonzentration limitierter Bereich (hohes w/NCN), (2) ein durch die vertikale Windgeschwindigkeit limitierter Bereich (niedriges w/NCN) und (3) ein Übergangsbereich (mittleres w/NCN). Die Ergebnisse zeigen, dass die Variabilität der anfänglichen Anzahlkonzentration der Wolkentropfen in (pyro-) konvektiven Wolken hauptsächlich durch die Variabilität der vertikalen Windgeschwindigkeit und der Aerosolkonzentration bestimmt wird. rnUm die mikrophysikalischen Prozesse innerhalb der rauchigen Aufwindregion einer pyrokonvektiven Wolke mit einer detaillierten spektralen Mikrophysik zu untersuchen, wurde das Paketmodel entlang einer Trajektorie innerhalb der Aufwindregion initialisiert. Diese Trajektore wurde durch dreidimensionale Simulationen eines pyro-konvektiven Ereignisses durch das Model ATHAM berechnet. Es zeigt sich, dass die Anzahlkonzentration der Wolkentropfen mit steigender Aerosolkonzentration ansteigt. Auf der anderen Seite verringert sich die Größe der Wolkentropfen mit steigender Aerosolkonzentration. Die Reduzierung der Verbreiterung des Tropfenspektrums stimmt mit den Ergebnissen aus Messungen überein und unterstützt das Konzept der Unterdrückung von Niederschlag in stark verschmutzen Wolken.Mit Hilfe des Models ATHAM wurden die dynamischen und mikrophysikalischen Prozesse von pyro-konvektiven Wolken, aufbauend auf einer realistischen Parametrisierung der Aktivierung von Aerosolpartikeln durch die Ergebnisse der Aktivierungsstudie, mit zwei- und dreidimensionalen Simulationen untersucht. Ein modernes zweimomenten mikrophysikalisches Schema wurde in ATHAM implementiert, um den Einfluss der Anzahlkonzentration von Aerosolpartikeln auf die Entwicklung von idealisierten pyro-konvektiven Wolken in US Standardamtosphären für die mittleren Breiten und den Tropen zu untersuchen. Die Ergebnisse zeigen, dass die Anzahlkonzentration der Aerosolpartikel die Bildung von Regen beeinflusst. Für geringe Aerosolkonzentrationen findet die rasche Regenbildung hauptsächlich durch warme mikrophysikalische Prozesse statt. Für höhere Aerosolkonzentrationen ist die Eisphase wichtiger für die Bildung von Regen. Dies führt zu einem verspäteten Einsetzen von Niederschlag für verunreinigtere Atmosphären. Außerdem wird gezeigt, dass die Zusammensetzung der Eisnukleationspartikel (IN) einen starken Einfluss auf die dynamische und mikrophysikalische Struktur solcher Wolken hat. Bei sehr effizienten IN bildet sich Regen früher. Die Untersuchung zum Einfluss des atmosphärischen Hintergrundprofils zeigt eine geringe Auswirkung der Meteorologie auf die Sensitivität der pyro-konvektiven Wolken auf diernAerosolkonzentration. Zum Abschluss wird gezeigt, dass die durch das Feuer emittierte Hitze einen deutlichen Einfluss auf die Entwicklung und die Wolkenobergrenze von pyro-konvektive Wolken hat. Zusammenfassend kann gesagt werden, dass in dieser Dissertation die Mikrophysik von pyrokonvektiven Wolken mit Hilfe von idealisierten Simulation eines Wolkenpaketmodell mit detaillierte spektraler Mikrophysik und eines 3D Modells mit einem zweimomenten Schema im Detail untersucht wurde. Es wird gezeigt, dass die extremen Bedingungen im Bezug auf die vertikale Windgeschwindigkeiten und Aerosolkonzentrationen einen deutlichen Einfluss auf die Entwicklung von pyro-konvektiven Wolken haben.

Simulations of charge transport in organic compounds

To aid the design of organic semiconductors, we study the charge transport properties of organic liquid crystals, i.e. hexabenzocoronene and carbazole macrocycle, and single crystals, i.e. rubrene, indolocarbazole and benzothiophene derivatives (BTBT, BBBT). The aim is to find structure-property relationships linking the chemical structure as well as the morphology with the bulk charge carrier mobility of the compounds. To this end, molecular dynamics (MD) simulations are performed yielding realistic equilibrated morphologies. Partial charges and molecular orbitals are calculated based on single molecules in vacuum using quantum chemical methods. The molecular orbitals are then mapped onto the molecular positions and orientations, which allows calculation of the transfer integrals between nearest neighbors using the molecular orbital overlap method. Thus we obtain realistic transfer integral distributions and their autocorrelations. In case of organic crystals the differences between two descriptions of charge transport, namely semi-classical dynamics (SCD) in the small polaron limit and kinetic Monte Carlo (KMC) based on Marcus rates, are studied. The liquid crystals are investigated solely in the hopping limit. To simulate the charge dynamics using KMC, the centers of mass of the molecules are mapped onto lattice sites and the transfer integrals are used to compute the hopping rates. In the small polaron limit, where the electronic wave function is spread over a limited number of neighboring molecules, the Schroedinger equation is solved numerically using a semi-classical approach. The results are compared for the different compounds and methods and, where available, with experimental data. The carbazole macrocycles form columnar structures arranged on a hexagonal lattice with side chains facing inwards, so columns can closely approach each other allowing inter-columnar and thus three-dimensional transport. When taking only intra-columnar transport into account, the mobility is orders of magnitude lower than in the three-dimensional case. BTBT is a promising material for solution-processed organic field-effect transistors. We are able to show that, on the time-scales of charge transport, static disorder due to slow side chain motions is the main factor determining the mobility. The resulting broad transfer integral distributions modify the connectivity of the system but sufficiently many fast percolation paths remain for the charges. Rubrene, indolocarbazole and BBBT are examples of crystals without significant static disorder. The high mobility of rubrene is explained by two main features: first, the shifted cofacial alignment of its molecules, and second, the high center of mass vibrational frequency. In comparsion to SCD, only KMC based on Marcus rates is capable of describing neighbors with low coupling and of taking static disorder into account three-dimensionally. Thus it is the method of choice for crystalline systems dominated by static disorder. However, it is inappropriate for the case of strong coupling and underestimates the mobility of well-ordered crystals. SCD, despite its one-dimensionality, is valuable for crystals with strong coupling and little disorder. It also allows correct treatment of dynamical effects, such as intermolecular vibrations of the molecules. Rate equations are incapable of this, because simulations are performed on static snapshots. We have thus shown strengths and weaknesses of two state of the art models used to study charge transport in organic compounds, partially developed a program to compute and visualize transfer integral distributions and other charge transport properties, and found structure-mobility relations for several promising organic semiconductors.

Analysis of cone-beam ct dose in image-guided radiation therapy for brain and prostate cancer patients via gpu-based Monte Carlo simulations

La radioterapia guidata da immagini (IGRT), grazie alle ripetute verifiche della posizione del paziente e della localizzazione del volume bersaglio, si è recentemente affermata come nuovo paradigma nella radioterapia, avendo migliorato radicalmente l’accuratezza nella somministrazione di dose a scopo terapeutico. Una promettente tecnica nel campo dell’IGRT è rappresentata dalla tomografia computerizzata a fascio conico (CBCT). La CBCT a kilovoltaggio, consente di fornire un’accurata mappatura tridimensionale dell’anatomia del paziente, in fase di pianificazione del trattamento e a ogni frazione del medisimo. Tuttavia, la dose da imaging attribuibile alle ripetute scansioni è diventata, negli ultimi anni, oggetto di una crescente preoccupazione nel contesto clinico. Lo scopo di questo lavoro è di valutare quantitativamente la dose addizionale somministrata da CBCT a kilovoltaggio, con riferimento a tre tipici protocolli di scansione per Varian OnBoard Imaging Systems (OBI, Palo Alto, California). A questo scopo sono state condotte simulazioni con codici Monte Carlo per il calcolo della dose, utilizzando il pacchetto gCTD, sviluppato sull’architettura della scheda grafica. L’utilizzo della GPU per sistemi server di calcolo ha permesso di raggiungere alte efficienze computazionali, accelerando le simulazioni Monte Carlo fino a raggiungere tempi di calcolo di ~1 min per un caso tipico. Inizialmente sono state condotte misure sperimentali di dose su un fantoccio d’acqua. I parametri necessari per la modellazione della sorgente di raggi X nel codice gCTD sono stati ottenuti attraverso un processo di validazione del codice al fine di accordare i valori di dose simulati in acqua con le misure nel fantoccio. Lo studio si concentra su cinquanta pazienti sottoposti a cicli di radioterapia a intensità modulata (IMRT). Venticinque pazienti con tumore al cervello sono utilizzati per studiare la dose nel protocollo standard-dose head e venticinque pazienti con tumore alla prostata sono selezionati per studiare la dose nei protocolli pelvis e pelvis spotlight. La dose media a ogni organo è calcolata. La dose media al 2% dei voxels con i valori più alti di dose è inoltre computata per ogni organo, al fine di caratterizzare l’omogeneità spaziale della distribuzione.

Properties of cosmic voids in dark energy simulations

The last decade has witnessed the establishment of a Standard Cosmological Model, which is based on two fundamental assumptions: the first one is the existence of a new non relativistic kind of particles, i. e. the Dark Matter (DM) that provides the potential wells in which structures create, while the second one is presence of the Dark Energy (DE), the simplest form of which is represented by the Cosmological Constant Λ, that sources the acceleration in the expansion of our Universe. These two features are summarized by the acronym ΛCDM, which is an abbreviation used to refer to the present Standard Cosmological Model. Although the Standard Cosmological Model shows a remarkably successful agreement with most of the available observations, it presents some longstanding unsolved problems. A possible way to solve these problems is represented by the introduction of a dynamical Dark Energy, in the form of the scalar field ϕ. In the coupled DE models, the scalar field ϕ features a direct interaction with matter in different regimes. Cosmic voids are large under-dense regions in the Universe devoided of matter. Being nearby empty of matter their dynamics is supposed to be dominated by DE, to the nature of which the properties of cosmic voids should be very sensitive. This thesis work is devoted to the statistical and geometrical analysis of cosmic voids in large N-body simulations of structure formation in the context of alternative competing cosmological models. In particular we used the ZOBOV code (see ref. Neyrinck 2008), a publicly available void finder algorithm, to identify voids in the Halos catalogues extraxted from CoDECS simulations (see ref. Baldi 2012 ). The CoDECS are the largest N-body simulations to date of interacting Dark Energy (DE) models. We identify suitable criteria to produce voids catalogues with the aim of comparing the properties of these objects in interacting DE scenarios to the standard ΛCDM model, at different redshifts. This thesis work is organized as follows: in chapter 1, the Standard Cosmological Model as well as the main properties of cosmic voids are intro- duced. In chapter 2, we will present the scalar field scenario. In chapter 3 the tools, the methods and the criteria by which a voids catalogue is created are described while in chapter 4 we discuss the statistical properties of cosmic voids included in our catalogues. In chapter 5 the geometrical properties of the catalogued cosmic voids are presented by means of their stacked profiles. In chapter 6 we summarized our results and we propose further developments of this work.

Thermodynamic concepts in adaptive resolution simulations

This thesis work is devoted to the conceptual and technical development of the Adaptive Resolution Scheme (AdResS), a molecular dynamics method that allows the simulation of a system with different levels of resolution simultaneously. The simulation domain is divided into high and low resolution zones and a transition region that links them, through which molecules can freely diffuse.rnThe first issue of this work regards the thermodynamic consistency of the method, which is tested and verified in a model liquid of tetrahedral molecules. The results allow the introduction of the concept of the Thermodynamic Force, an external field able to correct spurious density fluctuations present in the transition region in usual AdResS simulations.rnThe AdResS is also applied to a system where two different representations with the same degree of resolution are confronted. This simple test extends the method from an Adaptive Resolution Scheme to an Adaptive Representation Scheme, providing a way of coupling different force fields based on thermodynamic consistency arguments. The Thermodynamic Force is successfully applied to the example described in this work as well.rnAn alternative approach of deducing the Thermodynamic Force from pressure consistency considerations allows the interpretation of AdResS as a first step towards a molecular dynamics simulation in the Grand Canonical ensemble. Additionally, such a definition leads to a practical way of determining the Thermodynamic Force, tested in the well studied tetrahedral liquid. The effects of AdResS and this correction on the atomistic domain are analyzed by inspecting the local distribution of velocities, radial distribution functions, pressure and particle number fluctuation. Their comparison with analogous results coming from purely atomistic simulations shows good agreement, which is greatly improved under the effect of the external field.rnA further step in the development of AdResS, necessary for several applications in biophysics and material science, consists of its application to multicomponent systems. To this aim, the high-resolution representation of a model binary mixture is confronted with its coarse-grained representation systematically parametrized. The Thermodynamic Force, whose development requires a more delicate treatment, also gives satisfactory results.rnFinally, AdResS is tested in systems including two-body bonded forces, through the simulation of a model polymer allowed to adaptively change its representation. It is shown that the distribution functions that characterize the polymer structure are in practice not affected by the change of resolution.rnThe technical details of the implementation of AdResS in the ESPResSo package conclude this thesis work.

Modeling and simulations of nanoparticles in liquid crystalline systems

The aim of this work is to investigate, using extensive Monte Carlo computer simulations, composite materials consisting of liquid crystals doped with nanoparticles. These systems are currently of great interest as they offer the possibility of tuning the properties of liquid crystals used in displays and other devices as well as providing a way of obtaining regularly organized systems of nanoparticles exploiting the molecular organization of the liquid crystal medium. Surprisingly enough, there is however a lack of fundamental knowledge on the properties and phase behavior of these hybrid materials, making the route to their application an essentially empirical one. Here we wish to contribute to the much needed rationalization of these systems studying some basic effects induced by different nanoparticles on a liquid crystal host. We investigate in particular the effects of nanoparticle shape, size and polarity as well as of their affinity to the liquid crystal solvent on the stability of the system, monitoring phase transitions, order and molecular organizations. To do this we have proposed a coarse grained approach where nanoparticles are modelled as a suitably shaped (spherical, rod and disk like) collection of spherical Lennard-Jones beads, while the mesogens are represented with Gay-Berne particles. We find that the addition of apolar nanoparticles of different shape typically lowers the nematic–isotropic transition of a non-polar nematic, with the destabilization being greater for spherical nanoparticles. For polar mesogens we have studied the effect of solvent affinity of the nanoparticles showing that aggregation takes places for low solvation values. Interestingly, if the nanoparticles are polar the aggregates contribute to stabilizing the system, compensating the shape effect. We thus find the overall effects on stability to be a delicate balance of often contrasting contributions pointing to the relevance of simulations studies for understanding these complex systems.