Computational Reverse Engineering Analysis of the Scattering Experiment Method for Interpretation of 2D Small-Angle Scattering Profiles (CREASE-2D).

Small-angle scattering (SAS) is a widely used characterization technique that provides structural information in soft materials at varying length scales (nanometers to microns). The output of an SAS measurement is the scattered intensity I(q) as a function of q, the scattered wavevector with respect to the incident wave; the latter is represented by its magnitude |q| ≡ q (in inverse distance units) and azimuthal angle θ. While isotropic structural arrangement can be interpreted by analysis of the azimuthally averaged one-dimensional (1D) scattering profile, to understand anisotropic arrangements, one has to interpret the two-dimensional (2D) scattering profile, I(q, θ). Manual interpretation of such 2D profiles usually involves fitting of approximate analytical models to azimuthally averaged sections of the 2D profile. In this paper, we present a new method called CREASE-2D that interprets, without any azimuthal averaging, the entire 2D scattering profile, I(q, θ), and outputs the relevant structural features. CREASE-2D is an extension of the "computational reverse engineering analysis for scattering experiments" (CREASE) method that has been used successfully to analyze 1D SAS profiles for a variety of soft materials. CREASE-2D goes beyond CREASE by enabling analysis of 2D scattering profiles, which is far more challenging to interpret than the azimuthally averaged 1D profiles. The CREASE-2D workflow identifies the structural features whose computed I(q, θ) profiles, calculated using a surrogate XGBoost machine learning model, match the input experimental I(q, θ). We expect that this CREASE-2D method will be a valuable tool for materials' researchers who need direct interpretation of the 2D scattering profiles in contrast to analyzing azimuthally averaged 1D I(q) vs q profiles that can lose important information related to structural anisotropy.

Understanding Self-Assembly and Molecular Packing in Methylcellulose Aqueous Solutions Using Multiscale Modeling and Simulations.

We present a multiscale molecular dynamics (MD) simulation study on self-assembly in methylcellulose (MC) aqueous solutions. First, using MD simulations with a new coarse-grained (CG) model of MC chains in implicit water, we establish how the MC chains self-assemble to form fibrils and fibrillar networks and elucidate the MC chains' packing within the assembled fibrils. The CG model for MC is extended from a previously developed model for unsubstituted cellulose and captures the directionality of H-bonding interactions between the -OH groups. The choice and placement of the CG beads within each monomer facilitates explicit modeling of the exact degree and position of methoxy substitutions in the monomers along the MC chain. CG MD simulations show that with increasing hydrophobic effect and/or increasing H-bonding strength, the commercial MC chains (with degree of methoxy substitution, DS, ∼1.8) assemble from a random dispersed configuration into fibrils. The assembled fibrils exhibit consistent fibril diameters regardless of the molecular weight and concentration of MC chains, in agreement with past experiments. Most MC chains' axes are aligned with the fibril axis, and some MC chains exhibit twisted conformations in the fibril. To understand the molecular driving force for the twist, we conduct atomistic simulations of MC chains preassembled in fibrils (without any chain twists) in explicit water at 300 and 348 K. These atomistic simulations also show that at DS = 1.8, MC chains adopt twisted conformations, with these twists being more prominent at higher temperatures, likely as a result of shielding of hydrophobic methyl groups from water. For MC chains with varying DS, at 348 K, atomistic simulations show a nonmonotonic effect of DS on water-monomer contacts. For 0.0 < DS < 0.6, the MC monomers have more water contacts than at DS = 0.0 or DS > 0.6, suggesting that with few methoxy substitutions, the MC chains are effectively hydrophilic, letting the water molecules diffuse into the fibril to participate in H-bonds with the MC chains' remaining -OH groups. At DS > 0.6, the MC monomers become increasingly hydrophobic, as seen by decreasing water contacts around each monomer. We conclude based on the atomistic observations that MC chains with lower degrees of substitutions (DS ≤ 0.6) should exhibit solubility in water over broader temperature ranges than DS ∼ 1.8 chains.

Pair-Variational Autoencoders for Linking and Cross-Reconstruction of Characterization Data from Complementary Structural Characterization Techniques.

In materials research, structural characterization often requires multiple complementary techniques to obtain a holistic morphological view of a synthesized material. Depending on the availability and accessibility of the different characterization techniques (e.g., scattering, microscopy, spectroscopy), each research facility or academic research lab may have access to high-throughput capability in one technique but face limitations (sample preparation, resolution, access time) with other technique(s). Furthermore, one type of structural characterization data may be easier to interpret than another (e.g., microscopy images are easier to interpret than small-angle scattering profiles). Thus, it is useful to have machine learning models that can be trained on paired structural characterization data from multiple techniques (easy and difficult to interpret, fast and slow in data collection or sample preparation) so that the model can generate one set of characterization data from the other. In this paper we demonstrate one such machine learning workflow, Pair-Variational Autoencoders (PairVAE), that works with data from small-angle X-ray scattering (SAXS) that present information about bulk morphology and images from scanning electron microscopy (SEM) that present two-dimensional local structural information on the sample. Using paired SAXS and SEM data of newly observed block copolymer assembled morphologies [open access data from Doerk G. S.; et al. Sci. Adv.2023, 9 ( (2), ), eadd3687], we train our PairVAE. After successful training, we demonstrate that the PairVAE can generate SEM images of the block copolymer morphology when it takes as input that sample's corresponding SAXS 2D pattern and vice versa. This method can be extended to other soft material morphologies as well and serves as a valuable tool for easy interpretation of 2D SAXS patterns as well as an engine for generating ensembles of similar microscopy images to create a database for other downstream calculations of structure-property relationships.

Computational approach for structure generation of anisotropic particles (CASGAP) with targeted distributions of particle design and orientational order.

The macroscopic properties of materials are governed by their microscopic structure which depends on the materials' composition (i.e., building blocks) and processing conditions. In many classes of synthetic, bioinspired, or natural soft and/or nanomaterials, one can find structural anisotropy in the microscopic structure due to anisotropic building blocks and/or anisotropic domains formed through the processing conditions. Experimental characterization and complementary physics-based or data-driven modeling of materials' structural anisotropy are critical for understanding structure-property relationships and enabling targeted design of materials with desired macroscopic properties. In this pursuit, to interpret experimentally obtained characterization results (e.g., scattering profiles) of soft materials with structural anisotropy using data-driven computational approaches, there is a need for creating real space three-dimensional structures of the designer soft materials with realistic physical features (e.g., dispersity in building block sizes) and anisotropy (i.e., aspect ratios of the building blocks, their orientational and positional order). These real space structures can then be used to compute and complement experimentally obtained characterization results or be used as initial configurations for physics-based simulations/calculations that can then provide training data for machine learning models. To address this need, we present a new computational approach called CASGAP - Computational Approach for Structure Generation of Anisotropic Particles - for generating any desired three dimensional real-space structure of anisotropic building blocks (modeled as particles) adhering to target distributions of particle shape, size, and positional and orientational order.

Effects of solvent conditions on the self-assembly of heterotrimeric collagen-like peptide (CLP) triple helices: a coarse-grained simulation study.

We perform coarse-grained (CG) molecular dynamics (MD) simulations to investigate the self-assembly of collagen-like peptide (CLP) triple helices into fibrillar structures and percolated networks as a function of solvent quality. The focus of this study is on CLP triple helices whose strands are different lengths (i.e., heterotrimers), leading to dangling 'sticky ends'. These 'sticky ends' are segments of the CLP strands that have unbonded hydrogen-bonding donor/acceptor sites that drive heterotrimeric CLP triple helices to physically associate with one another, leading to assembly into higher-order structures. We use a validated CG model for CLP in implicit solvent and capture varying solvent quality through changing strength of attraction between CG beads representing the amino acids in the CLP strands. Our CG MD simulations show that, at lower CLP concentrations, CLP heterotrimers assemble into fibrils and, at higher CLP concentrations, into percolated networks. At higher concentrations, decreasing solvent quality causes (i) the formation of heterogeneous network structures with a lower degree of branching at network junctions and (ii) increases in the diameter of network strands and pore sizes. We also observe a nonmonotonic effect of solvent quality on distances between network junctions due to the balance between heterotrimer end-end associations driven by hydrogen bonding and side-side associations driven by worsening solvent quality. Below the percolation threshold, we observe that decreasing solvent quality leads to the formation of fibrils composed of multiple aligned CLP triple helices, while the number of 'sticky ends' governs the spatial extent (radius of gyration) of the assembled fibrils.

Mechanism of structural colors in binary mixtures of nanoparticle-based supraballs.

Inspired by structural colors in avian species, various synthetic strategies have been developed to produce noniridescent, saturated colors using nanoparticle assemblies. Nanoparticle mixtures varying in particle chemistry and size have additional emergent properties that affect the color produced. For complex multicomponent systems, understanding the assembled structure and a robust optical modeling tool can empower scientists to identify structure-color relationships and fabricate designer materials with tailored color. Here, we demonstrate how we can reconstruct the assembled structure from small-angle scattering measurements using the computational reverse-engineering analysis for scattering experiments method and use the reconstructed structure in finite-difference time-domain calculations to predict color. We successfully, quantitatively predict experimentally observed color in mixtures containing strongly absorbing nanoparticles and demonstrate the influence of a single layer of segregated nanoparticles on color produced. The versatile computational approach that we present is useful for engineering synthetic materials with desired colors without laborious trial-and-error experiments.

Computational Reverse-Engineering Analysis for Scattering Experiments for Form Factor and Structure Factor Determination ("P(q) and S(q) CREASE").

In this paper, we present an open-source machine learning (ML)-accelerated computational method to analyze small-angle scattering profiles [I(q) vs q] from concentrated macromolecular solutions to simultaneously obtain the form factor P(q) (e.g., dimensions of a micelle) and the structure factor S(q) (e.g., spatial arrangement of the micelles) without relying on analytical models. This method builds on our recent work on Computational Reverse-Engineering Analysis for Scattering Experiments (CREASE) that has either been applied to obtain P(q) from dilute macromolecular solutions (where S(q) ∼1) or to obtain S(q) from concentrated particle solutions when P(q) is known (e.g., sphere form factor). This paper's newly developed CREASE that calculates P(q) and S(q), termed as "P(q) and S(q) CREASE", is validated by taking as input I(q) vs q from in silico structures of known polydisperse core(A)-shell(B) micelles in solutions at varying concentrations and micelle-micelle aggregation. We demonstrate how "P(q) and S(q) CREASE" performs if given two or three of the relevant scattering profiles-I total(q), I A(q), and I B(q)-as inputs; this demonstration is meant to guide experimentalists who may choose to do small-angle X-ray scattering (for total scattering from the micelles) and/or small-angle neutron scattering with appropriate contrast matching to get scattering solely from one or the other component (A or B). After validation of "P(q) and S(q) CREASE" on in silico structures, we present our results analyzing small-angle neutron scattering profiles from a solution of core-shell type surfactant-coated nanoparticles with varying extents of aggregation.

Synergistic Role of Temperature and Salinity in Aggregation of Nonionic Surfactant-Coated Silica Nanoparticles.

The adsorption of nonionic surfactants onto hydrophilic nanoparticles (NPs) is anticipated to increase their stability in aqueous medium. While nonionic surfactants show salinity- and temperature-dependent bulk phase behavior in water, the effects of these two solvent parameters on surfactant adsorption and self-assembly onto NPs are poorly understood. In this study, we combine adsorption isotherms, dispersion transmittance, and small-angle neutron scattering (SANS) to investigate the effects of salinity and temperature on the adsorption of pentaethylene glycol monododecyl ether (C12E5) surfactant on silica NPs. We find an increase in the amount of surfactant adsorbed onto the NPs with increasing temperature and salinity. Based on SANS measurements and corresponding analysis using computational reverse-engineering analysis of scattering experiments (CREASE), we show that the increase in salinity and temperature results in the aggregation of silica NPs. We further demonstrate the non-monotonic changes in viscosity for the C12E5-silica NP mixture with increasing temperature and salinity and correlate the observations to the aggregated state of NPs. The study provides a fundamental understanding of the configuration and phase transition of the surfactant-coated NPs and presents a strategy to manipulate the viscosity of such dispersion using temperature as a stimulus.

Investigating the Hydrogen Bond-Induced Self-Assembly of Polysulfamides Using Molecular Simulations and Experiments.

In this paper, we present a synergistic, experimental, and computational study of the self-assembly of N,N'-disubstituted polysulfamides driven by hydrogen bonds (H-bonds) between the H-bonding donor and acceptor groups present in repeating sulfamides as a function of the structural design of the polysulfamide backbone. We developed a coarse-grained (CG) polysulfamide model that captures the directionality of H-bonds between the sulfamide groups and used this model in molecular dynamics (MD) simulations to study the self-assembly of these polymers in implicit solvent. The CGMD approach was validated by reproducing experimentally observed trends in the extent of crystallinity for three polysulfamides synthesized with aliphatic and/or aromatic repeating units. After validation of our CGMD approach, we computationally predicted the effect of repeat unit bulkiness, length, and uniformity of segment lengths in the polymers on the extent of orientational and positional order among the self-assembled polysulfamide chains, providing key design principles for tuning the extent of crystallinity in polysulfamides in experiments. Those computational predictions were then experimentally tested through the synthesis and characterization of polysulfamide architectures.

Polymer solution structure and dynamics within pores of hexagonally close-packed nanoparticles.

Using coarse-grained molecular dynamics simulations, we examine structure and dynamics of polymer solutions under confinement within the pores of a hexagonally close-packed (HCP) nanoparticle system with nanoparticle diameter fifty times that of the polymer Kuhn segment size. We model a condition where the polymer chain is in a good solvent (i.e., polymer-polymer interaction is purely repulsive and polymer-solvent and solvent-solvent interactions are attractive) and the polymer-nanoparticle and solvent-nanoparticle interactions are purely repulsive. We probe three polymer lengths (N = 10, 114, and 228 Kuhn segments) and three solution concentrations (1, 10, and 25%v) to understand how the polymer chain conformations and chain center-of-mass diffusion change under confinement within the pores of the HCP nanoparticle structure from those seen in bulk. The known trend of bulk polymer Rg2 decreasing with increasing concentration no longer holds when confined in the pores of HCP nanoparticle structure; for example, for the 114-mer, the HCP 〈Rg2〉 at 1%v concentration is lower than HCP 〈Rg2〉 at 10%v concentration. The 〈Rg2〉 of the 114-mer and 228-mer exhibit the largest percent decline going from bulk to HCP at the 1%v concentration and the smallest percent decline at the 25%v concentration. We also provide insight into how the confinement ratio (CR) of polymer chain size to pore size within tetrahedral and octahedral pores in the HCP arrangement of nanoparticles affects the chain conformation and diffusion at various concentrations. At the same concentration, the N = 114 has significantly more movement between pores than the N = 228 chains. For the N = 114 polymer, the diffusion between pores (i.e., inter-pore diffusion) accelerates the overall diffusion rate for the confined HCP system while for the N = 228 polymer, the polymer diffusion in the entire HCP is dominated by the diffusion within the tetrahedral or octahedral pores with minor contributions from inter-pore diffusion. These findings augment the fundamental understanding of macromolecular diffusion through large, densely packed nanoparticle assemblies and are relevant to research focused on fabrication of polymer composite materials for chemical separations, storage, optics, and photonics. We perform coarse-grained molecular dynamics simulations to understand structure and dynamics of polymer solutions under confinement within hexagonal close packed nanoparticles with radii much larger than the polymer chain's bulk radius of gyration.

Computational Reverse-Engineering Analysis for Scattering Experiments (CREASE) with Machine Learning Enhancement to Determine Structure of Nanoparticle Mixtures and Solutions.

We present a new open-source, machine learning (ML) enhanced computational method for experimentalists to quickly analyze high-throughput small-angle scattering results from multicomponent nanoparticle mixtures and solutions at varying compositions and concentrations to obtain reconstructed 3D structures of the sample. This new method is an improvement over our original computational reverse-engineering analysis for scattering experiments (CREASE) method (ACS Materials Au2021, 1 (2 (2), ), 140-156), which takes as input the experimental scattering profiles and outputs a 3D visualization and structural characterization (e.g., real space pair-correlation functions, domain sizes, and extent of mixing in binary nanoparticle mixtures) of the nanoparticle mixtures. The new gene-based CREASE method reduces the computational running time by >95% as compared to the original CREASE and performs better in scenarios where the original CREASE method performed poorly. Furthermore, the ML model linking features of nanoparticle solutions (e.g., concentration, nanoparticles' tendency to aggregate) to a computed scattering profile is generic enough to analyze scattering profiles for nanoparticle solutions at conditions (nanoparticle chemistry and size) beyond those that were used for the ML training. Finally, we demonstrate application of this new gene-based CREASE method for analysis of small-angle X-ray scattering results from a nanoparticle solution with unknown nanoparticle aggregation and small-angle neutron scattering results from a binary nanoparticle assembly with unknown mixing/segregation among the nanoparticles.

Impact of Polydopamine Nanoparticle Surface Pattern and Roughness on Interactions with Poly(ethylene glycol) in Aqueous Solution: A Multiscale Modeling and Simulation Study.

A significant research effort in the past few years has been devoted to engineering synthetic mimics of naturally occurring eumelanin. One such effort has involved the assembly of oligomers of 5,6-dihydroxyindole (DHI), a synthetic precursor of polydopamine (PDA), into melanin-mimicking nanoparticles for use in a variety of applications with desired optical, photonic, thermal, and electrical properties. In many of these applications, the PDA nanoparticles are mixed with other polymers or oligomers, thus motivating this specific study to understand how the surface characteristics of the assembled PDA-nanoparticles affect their interaction with poly(ethylene glycol) (PEG) chains in aqueous solution. We use molecular dynamics (MD) simulations to study the interaction of linear 20-mer PEG chains with different PDA-nanoparticles assembled using four types of oligomers of 5,6-DHI: two isomers of 5,6-DHI 2-mers with the monomers bonding either at the 2-2' position (A-type isomer) or 7-7' position (B-type isomer), denoted as A:2-mer and B:2-mer, respectively, and a 4-mer and an 8-mer of B-type chemistry denoted as B:4-mer and B:8-mer, respectively. Using explicit-solvent atomistic MD simulations, we find that PDA-nanoparticle surfaces assembled from B:8-mer exhibit smaller density fluctuations of water molecules and, as a result, are relatively more hydrophilic than the PDA-nanoparticle surfaces assembled from A:2-mer, B:2-mer, and B:4-mer. The surface composition of PDA-nanoparticles assembled from A:2-mer contains relatively fewer hydroxyl (-OH) groups compared to PDA-nanoparticles assembled from a B:2-mer, B:4-mer, or B:8-mer, yet the sample of PEG chains show more collapsed and adsorbed conformations on A:2-mer nanoparticles' surface. To explain the atomistically observed behavior of PEG chains on the nanoparticles' surfaces, we use coarse-grained (CG) MD simulations and explain the roles of the pattern formed by the attractive sites (e.g.,-OH groups) exposed on the surface and the roughness of the surface on interactions with a genric PEG-like copolymer chain. By comparing atomistic and CG MD simulation results, we confirm that the -OH groups' pattern on the surface of the PDA-nanoparticle assembled from A:2-mer is patchier than the random or string-like patterns on the PDA-nanoparticle assembled from B:2-mer, B:4-mer, or B:8-mer, and it is this -OH groups' surface pattern that dictates the PEG chain conformations and adsorption on the PDA-nanoparticle surface. Overall, these results guide the design of chemically and physically heterogeneous nanoparticle surfaces for the desired polymer interaction and conformations.

Impact of collagen-like peptide (CLP) heterotrimeric triple helix design on helical thermal stability and hierarchical assembly: a coarse-grained molecular dynamics simulation study.

Collagen-like peptides (CLP) are multifunctional materials garnering a lot of recent interest from the biomaterials community due to their hierarchical assembly and tunable physicochemical properties. In this work, we present a computational study that links the design of CLP heterotrimers to the thermal stability of the triple helix and their self-assembly into fibrillar aggregates and percolated networks. Unlike homotrimeric helices, the CLP heterotrimeric triple helices in this study are made of CLP strands of different chain lengths that result in 'sticky' ends with available hydrogen bonding groups. These 'sticky' ends at one end or both ends of the CLP heterotrimer then facilitate inter-helix hydrogen bonding leading to self-assembly into fibrils (clusters) and percolated networks. We consider the cases of three sticky end lengths - two, four, and six repeat units - present entirely on one end or split between two ends of the CLP heterotrimer. We observe in CLP heterotrimer melting curves generated using coarse grained Langevin dynamics simulations at low CLP concentration that increasing sticky end length results in lower melting temperatures for both one and two sticky ended CLP designs. At higher CLP concentrations, we observe non-monotonic trends in cluster sizes with increasing sticky end length with one sticky end but not for two sticky ends with the same number of available hydrogen bonding groups as the one sticky end; this nonmonotonicity stems from the formation of turn structures stabilized by hydrogen bonds at the single, sticky end for sticky end lengths greater than four repeat units. With increasing CLP concentration, heterotrimers also form percolated networks with increasing sticky end length with a minimum sticky end length of four repeat units required to observe percolation. Overall, this work informs the design of thermoresponsive, peptide-based biomaterials with desired morphologies using strand length and dispersity as a handle for tuning thermal stability and formation of supramolecular structures.

Computational Reverse Engineering Analysis for Scattering Experiments (CREASE) on Vesicles Assembled from Amphiphilic Macromolecular Solutions.

In this paper we present the development and validation of the "Computational Reverse-Engineering Analysis for Scattering Experiments" (CREASE) method for analyzing scattering results from vesicle structures that are commonly found upon assembly of synthetic, biomimetic, or bioderived amphiphilic copolymers in solution. The two-step CREASE method takes the amphiphilic polymer chemistry and small-angle scattering intensity profile, I exp(q), as input and determines the vesicles' structural features on multiple length scales ranging from assembled vesicle wall's individual layer thicknesses to the monomer-level packing and distribution of polymer conformations. In the first step of CREASE, a genetic algorithm (GA) is used to determine the relevant vesicle dimensions from the input macromolecular solution information and I exp(q) by identifying the structure whose computed scattering profile best matches the input I exp(q). Then in the second step, the GA-determined dimensions are used for molecular reconstruction of the vesicle structure. To validate CREASE for vesicles, we test CREASE on input scattering intensity profiles generated mathematically (termed as in silico I exp(q) vs q) from a variety of vesicle sizes with known dimensions. We also test CREASE on in silico I exp(q) vs q generated from vesicles with dispersity in all relevant dimensions, resembling real experiments. After successful validation of CREASE, we compare the CREASE-determined dimensions against those obtained from the traditional approach of fitting the scattering intensity profile to relevant analytical model in SASVIEW package. We show that CREASE performs better than or as well as the core-multishell analytical model's fitting in SASVIEW in determining vesicle dimensions with dispersity. We also show that CREASE provides structural information beyond those possible from traditional scattering analysis using the core-multishell model, such as the distribution of solvophilic monomers between the vesicle wall's inner and outer layers in the vesicle wall and the chain-level packing within each vesicle layer.

Anisotropic Synthetic Allomelanin Materials via Solid-State Polymerization of Self-Assembled 1,8-Dihydroxynaphthalene Dimers.

Melanosomes in nature have diverse morphologies, including spheres, rods, and platelets. By contrast, shapes of synthetic melanins have been almost entirely limited to spherical nanoparticles with few exceptions produced by complex templated synthetic methods. Here, we report a non-templated method to access synthetic melanins with a variety of architectures including spheres, sheets, and platelets. Three 1,8-dihydroxynaphthalene dimers (4-4', 2-4' and 2-2') were used as self-assembling synthons. These dimers pack to form well-defined structures of varying morphologies depending on the isomer. Specifically, distinctive ellipsoidal platelets can be obtained using 4-4' dimers. Solid-state polymerization of the preorganized dimers generates polymeric synthetic melanins while maintaining the initial particle morphologies. This work provides a new route to anisotropic synthetic melanins, where the building blocks are preorganized into specific shapes, followed by solid-state polymerization.