We report on a comprehensive computer simulation study of the liquid-crystal phase behaviour of purely repulsive, semi-flexible rod-like particles. For the four aspect ratios we consider, the particles form five distinct phases depending on their packing fraction and bending flexibility: the isotropic, nematic, smectic A, smectic B, and crystal phase. Upon increasing the particle bending flexibility, the various phase transitions shift to larger packing fractions. Increasing the aspect ratio achieves the opposite effect. We find two different ways in which the layer thickness of the particles in the smectic A phase may respond to an increase in concentration. The layer thickness may either decrease or increase depending on the aspect ratio and flexibility. For the smectic B and the crystalline phases, increasing the concentration always decreases the layer thickness. Finally, we find that the layer spacing jumps to a larger value on transitioning from the smectic A phase to the smectic B phase.

Compelling justification: Hard-core repulsion is the simplest interaction in Nature yet it drives the self-organization of many complex fluids. To investigate how enthalpy impacts upon entropy-dominated liquid crystalline states, we introduce a highly localized and tunable directional attractive interaction (or "patch") on one of the tips of rod-shaped colloids. Our experiments and computer simulations show that increasing the patch attraction dramatically stabilizes the lamellar phase, a structure desired in materials science due to its outstanding mechanical and optical properties. Our work demonstrates that introducing patches in anisotropic nanoparticles adds to the control of their self-assembly.Abstract: Dispersions of rod-like colloidal particles exhibit a plethora of liquid crystalline states, including nematic, smectic A, smectic B, and columnar phases. This phase behavior can be explained by presuming the predominance of hard-core volume exclusion between the particles.We show here how the self-organization of rod-like colloids can be controlled by introducing a weak and highly localized directional attractive interaction between one of the ends of the particles. This has been performed by functionalizing the tips of filamentous viruses by means of regioselectively grafting fluorescent dyes onto them, resulting in a hydrophobic patch whose attraction can be tuned by varying the number of bound dye molecules. We show, in agreement with our computer simulations, that increasing the single tip attraction stabilizes the smectic phase at the expense of the nematic phase, leaving all other liquid crystalline phases invariant. For sufficiently strong tip attraction the nematic state may be suppressed completely to get a direct isotropic liquid-tosmectic phase transition. Our findings provide insights into the rational design of building blocks for functional structures formed at low densities.

Weakly attractive interactions between the tips of rodlike colloidal particles affect their liquid-crystal phase behavior due to a subtle interplay between enthalpy and entropy. Here we employ molecular dynamics simulations on semiflexible, repulsive bead-spring chains where one of the two end beads attract each other. We calculate the phase diagram as a function of both the volume fraction of the chains and the strength of the attractive potential. We identify a large number of phases that include isotropic, nematic, smectic-A, smectic-B, and crystalline states. For tip attraction energies lower than the thermal energy, our results are qualitatively consistent with experimental findings: We find that an increase of the attraction strength shifts the nematic to smectic-A phase transition to lower volume fractions, with only minor effect on the stability of the other phases. For sufficiently strong tip attraction, the nematic phase disappears completely, in addition leading to the destabilization of the isotropic phase. In order to better understand the underlying physics of these phenomena, we also investigate the clustering of the particles at their attractive tips and the effective molecular field experienced by the particles in the smectic-A phase. Based on these results, we argue that the clustering of the tips only affects the phase stability if lamellar structures ("micelles") are formed. We find that an increase of the attraction strength increases the degree of order in the layered phases. Interestingly, we also find evidence for the existence of an antiferroelectric smectic-A phase transition induced by the interaction between the tips. A simple Maier-Saupe-McMillan model confirms our findings.

Agradecimentos O desenvolvimento deste trabalho contou com o apoio de diversas pessoas e instituições. São muitos os membros do grupo que me ajudaram, agradeço a todos pelo que aprendi e pelo convivío que tornou mais prazeiroso o meu período de mestrado. Em particular, agradeço à Mariana Ferraz que me acomapanhou nos experimentos, ensinando-me a utilizar os equipamentos do laboratório, discutindo procedimentos e análises, sendo companheira e amiga. Agradeço ao meu orientador Adriano Alencar por ter me orientado neste trabalho, por seu engajamento com o desenvolvimento científico e pessoal dos seus alunos e por inspirar-me com seu entusiasmo pela pesquisa e ciência. Agradeço ao professor Kees Storm, da Universidade Tecnológica de Eindhoven, que me recebeu e me orientou durante seis meses de estágio nesta universidade, mostrando-me novas abordagens e novos caminhos. Agradeço aos muitos membros do Instituto de Física da Universidade de São Paulo, por manterem a infraestrutura de que usufrui e pela convivência sempre descontraída. Por fim, agradeço às agências de fomento FAPESP e CNPq que financiaram este trabalho, concederam minha bolsa e me proporcionaram muitas outras oportunidades de aprendizado, como a participação em eventos e o estágio no exterior. Resumo Um problema proeminente em mecânica celular é relacionar como as modificações do citoesqueleto das células eucariontes resultam em alterações nas suas propriedades viscoelásticas. A proposta deste projeto é entender o comportamento mecânico das fibras de actina, principal componente do esqueleto celular, através de experimentos e do desenvolvimento de um modelo computacional que o descreva, na tentativa de entender as origens microscópicas da viscoelasticidade de redes de actina. Nesse contexto, existe uma extensa bibliografia voltada a modelos de comportamentos específicos das fibras do esqueleto celular, em particular, dos filamentos de actina. Contudo, a questão global relativa ao nível celular permanece em aberto. Experimentos com géis de actina in vitro foram o ponto de partida para conhecer os comportamentos mecânicos das fibras. Os dados experimentais obtidos foram a referência principal aos modelos matemáticos desenvolvidos para simular as propriedades viscoelásticas da rede de actina. Assim, foi montado um ambiente de trabalho, experimental e computacional, para estudar as propriedades mecânicas do citoesqueleto por meio de seu componente mais abundante.