An Experimental And Theoretical Research On Molecular Self-assembly

Self-assembly processes can assemble a limited number of fundamental building blocks into novel materials with ordered structures and superior functionalities via non-covalent interactions like van der Waals force and hydrogen bonding.In recent years,more and more self-assembled materials with excellent properties like biocompatibil-ity,environmental friendliness and high mechanical strength have been developed and widely applied.However,restricted by weak directionality of non-covalent interactions and complex collective effects in a large number of molecules,experimental methods and theoretical models for self-assembly systems are imperfect and hard to meet the de-mands of practical applications.Herein,this thesis chooses diphenylalanine and water molecules as main research targets and manages to further interpret or control their self-assembly processes through novel theoretical and experimental approaches.The main results are briefly described as follows:1.Diphenylalanine is one kind of self-assembling dipeptides.It can self-assemble into tubes,rods or fibers in liquid phases.However,hydrodynamic effects of so-lutions on self-assembly processes are usually not fully taken into consideration compared with intermolecular chemical reactions.Based on this understanding,we analyse possible hydrodynamic effects in liquids and their corresponding envi-ronmental conditions.Experiments based on evaporative dewetting processes on substrates are designed to build specific solution environments in which three dif-ferent hydrodynamic effects:Rayleigh convection,Marangoni flow and viscous fingering would play dominant roles in the self-assembly of diphenlyanine respec-tively.Organized crystalline structures with morphology related to their growing environments are produced.Significant roles of hydrodynamic effects and their act-ing mechanism on liquid-phase self-assembly processes are successfully verified.2.Self-assembly processes can be influenced by many environmental conditions and hard to form large-scale ordered structures.For the realization of facile and scalable self-assembly controlling techniques,we develop a novel method utilizing steady merging behaviour of double phase water/organic-solvent solutions to enable the synergy of self-assembly and hydrophobic aggregation.Organized long fiber bun-dles of diphenylalanine with centimeter-scale length are successfully prepared by this technique.Their superior mechanical strength and biochemical robustness are further proved by experimental measurements.Bending behaviors of these long fiber bundles can also be controlled by electric fields.These characteristics provide good basis for expanding their application fields.3.Hierarchy of self-assembly makes its incorporation with functional molecules at specific organizing level possible.It could lead to the formation of multi-functional composite materials.We introduce ultraviolet-responsive azobenzene molecules into the in vitro self-assembly of diphenylalanine molecules and acquire compos-ite materials with light-triggered bending ability.Due to the excellent mechani-cal strength of self-assembled multi-layer ordered structures,we have preliminarily explored the methods of preparing three-dimensional structures by template mold-ing and three-dimensional printing.The three-dimensional diphenylalanine struc-tures prepared by template molding can support heavier objects with lighter weight,which shows that it has good structural strength.After migrating the self-assembly process in solution to the three-dimensional printing platform,we have preliminar-ily realized the continuous assembly of long fiber structure.4.For a long time,the pattern emergence phenomenon and its physical implications have attracted much interests from researchers.However,with the existing com-puting capability,the usual molecular dynamics simulation methods based on first principles can only simulate the early stage of self-assembly process at nanoscale,but can hardly capture the characteristics of evolution at a broader time and space scale.Therefore,based on the cellular automata model and deep visual neural net-works,we construct a new theoretical framework for studying the morphology of self-assembly system.By identifying the features of the generated snowflake pat-terns,this framework connects similar snowflake patterns and constructs a large-scale similarity network of planar snowflakes.Using this network,we can match actual snowflake images with simulated snowflake patterns,and realize quantitative analysis and machine classification of snowflake patterns.