Study Of Iron Oxide Nanoparticles’ Crystal Formation And Translocation Across Membrane Based On Molecular Dynamics Simulation

Iron oxide nanomaterials play an import role in biomedical field,due to their excellent biosafety and unique magnetic response properties.Understanding the crystallization process of iron oxide nanomaterials and their interaction with cells in vivo can provide a theoretical basis for optimizing the material synthesis process,improving the performance of iron oxide nanomaterials,and realizing applications in clinical diagnosis and treatment.Limited by the low temporal/spatial resolution of current material characterization techniques,and the complex cellular physiological environment,it is difficult to observe the process of ions aggregation,structural phase transition,or nanomaterials entering cells in real time at a microscopic level.This thesis uses molecular dynamics（MD）simulation to describe the inter-particle interactions and motion states,and consequently investigate the microscopic mechanisms of iron oxide crystal nucleation and growth,and their interaction with cell membranes.Firstly,we fit and verify the iron oxide force field parameters based on the Coulombic-Buckingham potential with the partial charge model.Next,the process of ionic aggregation and subsequent iron oxide aggregates’structural transformation are further simulated.Besides,a local crystal structure identification algorithm for MD simulation systems based on the multi-population differential evolution algorithm is proposed,to implement the quantitative analysis of ion arrangement in the aggregates.Finally,adopting MARTINI coarse-grained force field,the effects of temperature and polyethylene glycol（PEG）grafting density on the translocation of nanoparticles（NPs）across asymmetric lipid membrane are discussed.The specific research content includes the following parts:（1）The force field parameters for iron oxide system based on the Coulombic-Buckingham potential are fitted.An objective function is constructed according to the tendency of expansion or compression,and the forces of each ion.Optimization algorithm for objective function（Zoutendijk algorithm）is improved in terms of calculating the gradient,achieving the optimal feasible direction,and searching for the optimal step length,to accelerate the convergence speed.Multiple initial value vectors are uniformly generated in the feasible domain space of the objective function.After preliminary screened and clustered,the initial value vectors are used as a starting point for iterative optimization method,to guarantee that the computational cost of fitting force field parameters is affordable.The obtained force field parameters can accurately reproduce the crystal structure and mechanical properties of iron oxide,and establish the basis for investigating the crystallization process of iron oxide nanomaterials.（2）Using the above-mentioned force field parameters,the process of iron oxide ions aggregation and rearrangement is simulated.Under the electrostatic attraction between the anions and cations,the ions sequentially added into simulation system formed iron oxide aggregates.When the number of Fe²⁺ions and O^2-ions in the aggregates reaches50,the force field parameters fitted in this study reproduce the transformation process from disordered structure to Fe O crystal.With the aggregation of more Fe³⁺ions and O^2-ions,the structure of Fe₂O₃aggregates become well-densified.The number of ions,of which the coordination number is consistent with that in iron oxide crystal(such as6-coordinated Fe³⁺ions and 4-coordinated O^2-ions),gradually increases.Some local long-range ordered structures are emerged in the interior of aggregates.During the temperature enhancement of the Fe₂O₃ aggregate（relaxing temperatures from 600 K to1500 K）,the amorphous or short-range ordered structures（likeε-Fe₂O₃）are transformed toβ-Fe₂O₃crystalline phase.In addition,the high temperature environment can also promote the rearrangement of ions adsorbed on the surface ofα-Fe₂O₃ and Fe₃O₄ crystals,which contributes to the epitaxial crystal growth.All the simulation results are in nice agreement with the published experimental observation.It provides theoretical guidance for the improvement of iron oxide nanomaterials synthesis process,and further exhibits the advantages of the force field parameters fitted in this thesis for the definition of ionic interactions in iron oxide systems.（3）In order to further analyze the structural transformation of the aggregates in simulations,a local crystal structure identification algorithm based on the multi-population differential evolution algorithm is proposed.The method defines an objective function through adopting the crystal structure as the matching template,and combining the geometric and topological structures of the simulation system.The multi-population differential evolution algorithm is utilized to obtain the optimal solution of the objective function,thus identifying all the atoms/ions of a local region with the arrangement similar to the crystal structure.The proposed algorithm cannot only achieve results generally consistent with common neighbor analysis（CNA）and polyhedral template matching（PTM）for the identification of simple crystal structures such as face-centered cubic（FCC）,hexagonal close packed（HCP）and body-centered cubic（BCC）structures,but also perform the identification of complex crystal structures such as corundum structure（α-Fe₂O₃）and anti-spinel structure（Fe₃O₄）.Besides,the aggregates can be directly segmented into several crystal regions according to the various orientations.The extent of influence of ions’random thermal motion on the identification results can be regulated via setting the threshold parameters in the algorithm.The proposed method in this paper is based on only the spatial coordinates of simulation system and target structures,and can theoretically solve the problem of identifying any structures,which provides a novel idea for the characterization of MD simulation results.（4）The translocation of NPs with various PEGylated density across the asymmetric membrane is simulated at different temperatures based on the MARTINI coarse-grained force field.The simulation results show that heating shortens the time of the translocation of NPs across the membrane,which is mainly attributed to the more disordered lipid structure and the more intense lateral diffusion at higher temperatures.On the other hand,steric barrier provided by PEG modification would inhibit the translocation of NPs.In the process of NPs’translocation across membranes,the conformation of PEG molecules changes dramatically,causing the so-called“snorkeling effect”to increase the contact between the surface of naked NPs and hydrophobic core of membranes.Higher grafting density weakens the conformational change of PEG molecules and facilitates the connection between hydrophilic particles in two leaflets of the membrane.Consequently,there are more lipid flip-flops and exchanges between two sides of the membrane,which is unfavorable for the maintenance of membrane asymmetry.The simulation results systematically illustrate the effects of temperatures and PEGylation on the interaction between NPs and membranes.It may offer guidance for the application of magnetic nanomaterials in the synergistic therapeutic strategy combining magnetic hyperthermia with chemotherapy.