Thesis--Bochum. Ruhr-Universität. / In Periodical Room.

Thesis (Doctoral)--Technische Universität Carolo-Wilhelmina zu Braunschweig, 1983.

München, Techn. Univ., Diss., 2010.

Techn. University, Diss., 2006--Darmstadt.

No description available.

Die Thermodynamik von Adsorptionsvorgängen wurde mittels quantenchemischer Methoden und der statistischen Thermodynamik untersucht. Eine neue rechentechnische Methode wird vorgestellt, die den Mangel an Genauigkeit vorhandener Methoden zur Untersuchung periodischer Systeme, wie z. B. der Dichte-Funktional-Theorie, behebt, und es ermöglicht thermodynamische Funktionen mit chemischer Genauigkeit zu bestimmen. Das im Rahmen dieser Arbeit entwickelte Protokoll besteht aus verschiedenen rechentechnischen Schritten, als da wären, eine Strukturoptimierung in Normalkoordinaten anstatt in kartesischen Koordinaten, eine numerische Berechnung der harmonischen Frequenzen durch Abtasten der Potentialenergieoberfläche entlang der Normalkoordinaten und anschließender anharmonischer Korrektur. Die Normalkoordinatenoptimierung garantiert eine korrekte Relaxation der Struktur mit ausschließlich reellen harmonischen Frequenzen. Die anharmonischen Korrekturen ermöglichen eine gute Beschreibung der Schwingungsstrukturen von Systemen die durch eine besonders Flache PES gekennzeichnet sind. Gleichzeitig, wird durch die Verwendung eines QM:QM-Hybridverfahrens zur Bestimmung des elektronischen Anteils der Adsorptionsenergie eine höhere Genauigkeit in der Bestimmung der Korrelationsenergie im Vergleich zu DFT garantiert. Der elektronische Anteil der Adsorptionsenergie, sowie der Schwingungsthermische Anteil werden abschließend kombiniert um einen genauen Wert der thermodynamischen Funktionen zu erhalten. / The thermodynamic of adsorption is investigated from the vibrational point of view using quantum chemical methods via statistical mechanics. Due to the lack of accuracy of the present available methods for investigating periodic systems, such as plane-wave density functional theory (DFT), a novel computational strategy is presented to overcome these limitations and bring the estimate of the thermodynamic functions within chemical accuracy limits. The protocol presented in this work consists of different computational steps, namely a structure optimization using normal mode coordinates instead of Cartesians, a numerical harmonic frequency calculation via sampling of the potential energy surface along the normal mode coordinates and the inclusion of anharmonic correction to the latter. The normal mode coordinate optimization ensures a proper relaxation of the structure and a reliable set of real harmonic frequencies while the anharmonic corrections account for a proper description of the vibrational structure of a system characterized by a very flat potential energy surface. Parallel to these calculations the electronic part of the adsorption energy is corrected using a hybrid QM:QM scheme to account the electronic correlations effects more accurately than DFT. The hybrid electronic adsorption energy and the vibrational thermal contributions obtained using anharmonic corrections are finally combined to get accurate estimate of the adsorption thermodynamic functions.

Biological membranes are mainly composed of two layers of lipids, various kinds of proteins and organic macromolecules, forming the protective barriers that separate the inner milieu of living cells from the environment. The possibility of penetrating the membrane is of great importance for biomedical applications. Recently, a lot of attention has been given to the mechanisms and the details of the interactions between the membrane and nanoparticles, as well as to the development of effective delivery strategies. A manipulation of the hydrophobicity of nanoparticles can facilitate the translocation through the membrane. Modifying the physical/chemical properties of the membrane through oxidation can also influence the delivery of nanoparticles or macromolecules into the cell. In this work, using coarse-grained molecular dynamics simulations, the passive translocation of nanoparticles with a size of about 1 nm and with tunable degrees of hydrophobicity through lipid membranes is studied. It is shown that a window of nanoparticle translocation with a sharp maximum is located at a certain hydrophobicity in between fully hydrophilic and fully hydrophobic characters. By combining direct simulations with umbrella sampling simulations, the free energy landscapes for nanoparticles covering a wide range of hydrophobicities are obtained. The directly observed translocation rate of the nanoparticles can be mapped to the mean escape rate through the calculated free energy landscapes, and the maximum of translocation can be related with the maximally flat free energy landscape. For nanoparticles with the balanced hydrophobicity, the bilayer forms a remaining barrier of a few kBT and can be spontaneously surmounted. Further investigations are conducted to explore the cooperative effects of a larger number of nanoparticles and their impact on membrane properties such as membrane permeability for solvent, the area per lipid, and the orientation order of lipid tails. By calculating the partition of nanoparticles between water and oil phases, the microscopic parameter, i.e. the hydrophobicity of nanoparticles, can be mapped to an experimentally accessible partition coefficient. The studies reveal a generic mechanism for spherical nanoparticles to overcome biological membrane-barriers without the need of biologically activated processes. Two oxidatively modified lipids are studied on coarse-grained level using molecular dynamics simulations. The findings support the view that lipid oxidation leads to a change of the lipid conformation: lipid tails tend to bend toward the lipid head-tail interface due to the presence of hydrophilic oxidized beads. This change in conformation can further influence structural properties, elasticity and membrane permeability: an increase of the area per lipid, accompanied with decrease of the membrane thickness and order parameter of the lipid tails; a sharp drop of stretching modulus; a significant increase of the membrane permeability for water. Oxidized lipid bilayers interacting with NPs of various degrees of hydrophobicity are further studied. The critical hydrophobicity corresponding to the maximum translocation rate of NPs, shifts towards the hydrophilic region, which coincides with the same decrease in percentage of the average hydrophobicity in the core of the membrane upon oxidation. Around the critical point of NPs' hydrophobicity, a significant increase of the translocation rate of NPs through the oxidized bilayers is observed, when compared to non-oxidized bilayers. This is associated with a deterioration of the free energy barrier for NPs inside the oxidized bilayers, resulting from oxidation effects. These findings are consistent with the studies of the mean escape rate through the free energy landscapes using Kramers theory. Regarding the membrane perturbation induced by NPs of various hydrophobicity, the data obtained with oxidized lipid bilayers present the same general trend as in the case of the non-oxidized lipid bilayer. These findings provide a better understanding of the interaction between NPs and oxidized lipid bilayers, and open a possibility to facilitate drug delivery.:1 Introduction 1 1.1 Lipid Bilayers 1 1.2 Oxidized Lipid Bilayers 2 1.3 Experimental Methodology 4 1.4 Lipid Models 5 1.5 The Lipid Bilayer Interacting with NPs 6 1.6 Thesis Overview 7 2 State of the art 9 2.1 Molecular Dynamics Simulations of Lipid Bilayers 9 2.1.1 Equations of Motion and the Integrations of Equations of Motion 10 2.1.2 Interaction Potentials 12 2.1.3 Periodic Boundary Conditions 14 2.1.4 Barostats and Thermostats 15 2.2 Umbrella Sampling Simulation 19 2.2.1 The Basics of Umbrella Sampling Method 20 2.2.2 Analyzing Umbrella Sampling Results by WHAM 23 2.2.3 The Principle of Choosing Bias Potential 24 3 Lipid Membranes interacting with Nanoparticles of Various Degrees of Hydrophobicity 25 3.1 Introduction 25 3.2 Coarse-grained Model and Simulation Setups 27 3.3 Results and Discussions 31 3.3.1 NPs-membrane Interactions 31 3.3.2 NPs Translocation 33 3.3.3 Concentration Effect of NPs 35 3.3.4 The Effect of Hydrophobicity on Kinetic Pathways 38 3.3.5 Potential of Mean Force 39 3.3.6 Hydrophobicity Scale 41 3.3.7 Solvent Permeation and Membrane Perturbation Induced by NPs 45 3.4 Summary 47 4 Coarse-grained Model of Oxidized Lipids and their Interactions with NPs of Varying Hydrophobicities 51 4.1 Introduction 51 4.2 Coarse-grained Model and Simulation Details 52 4.3 Results and Discussions 54 4.3.1 Characterizing the Oxidized Lipid Membranes 54 4.3.2 Oxidized Lipid Membranes Interacting with NPs of Various Degrees of Hydrophobicity 59 4.4 Summary 65 5 Summary and Outlook 69

It is known that fission yeast Schizosaccharomyces pombe maintains its nuclear envelope during mitosis and it undergoes an interesting shape change during cell division - from a spherical via an ellipsoidal and a peanut-like to a dumb-bell shape. However, the biomechanical system behind this amazing transformation is still not understood. What we know is, that the shape must change due to forces acting on the membrane surrounding the nucleus and the microtubule based mitotic spindle is thought to play a key role. To estimate the locations and directions of the forces, the shape of the nucleus was recorded by confocal light microscopy. But such data is often inhomogeneously labeled with gaps in the boundary, making classical segmentation impractical. In order to accurately determine the shape we developed a global parametric shape description method, based on a Fourier coordinate expansion. The method implicitly assumes a closed and smooth surface. We will calculate the geometrical properties of the 2-dimensional shape and extend it to 3-dimensional properties, assuming rotational symmetry. Using a mechanical model for the lipid bilayer and the so called Helfrich-Canham free energy we want to calculate the minimum energy shape while respecting system-specific constraints to the surface and the enclosed volume. Comparing it with the observed shape leads to the forces. This provides the needed research tools to study forces based on images.

It is known that fission yeast Schizosaccharomyces pombe maintains its nuclear envelope during mitosis and it undergoes an interesting shape change during cell division - from a spherical via an ellipsoidal and a peanut-like to a dumb-bell shape. However, the biomechanical system behind this amazing transformation is still not understood. What we know is, that the shape must change due to forces acting on the membrane surrounding the nucleus and the microtubule based mitotic spindle is thought to play a key role. To estimate the locations and directions of the forces, the shape of the nucleus was recorded by confocal light microscopy. But such data is often inhomogeneously labeled with gaps in the boundary, making classical segmentation impractical. In order to accurately determine the shape we developed a global parametric shape description method, based on a Fourier coordinate expansion. The method implicitly assumes a closed and smooth surface. We will calculate the geometrical properties of the 2-dimensional shape and extend it to 3-dimensional properties, assuming rotational symmetry. Using a mechanical model for the lipid bilayer and the so called Helfrich-Canham free energy we want to calculate the minimum energy shape while respecting system-specific constraints to the surface and the enclosed volume. Comparing it with the observed shape leads to the forces. This provides the needed research tools to study forces based on images.