Statistical analysis, machine learning modeling, and text analytics of aggregation attachment efficiency: Mono and binary particle systems

Combining size distribution and shape of plastic and oxide particles to evaluate physicochemical interactions: Aggregation and attachment

Particles naturally have size distributions and shapes, but these are overlooked in the physicochemical theory used to estimate interaction energies for particle aggregation or attachment. Consequently, the objectives of this research were to implement size distribution and shape in physicochemical interactions, and to use machine learning (ML) to investigate physicochemical parameters to interpret aggregation or attachment. A deep neural network was trained on databases generated for the interactions of spheres, ellipsoids, and cylinders. The primary sizes of particles were measured and then used in a machine learning model to predict interaction profiles considering size distributions. Spherical polystyrene and polymethyl-methacrylate were used in stability and aggregation experiments. Bullet- and fragment-like silica particles were used in attachment experiments. Subsequently, ML predictions were used to interpret the results of the experiments. The size distribution provides an active zone for physicochemical interactions that is absent using the traditional mean particle diameter (one equivalent sphere or ellipsoid). This is relevant because the size distribution increases the estimates of favorable and unfavorable aggregation and attachment. For example, these zones increase as the particle size distribution increases (high polydispersity index). Finally, although the approach is appropriate for spherical, ellipsoidal, and bullet-like particles, it is inappropriate for fragment-like particles, such as microplastics.

Predicting a Wide Range of Fractal Dimensions of Salt-Induced Aggregates in Water Using a Random Forest Model

Salt-induced colloidal aggregates can significantly influence contaminant fate and transport in natural and engineered systems. These aggregates' fractal dimensions (df), ranging from 1.4 to 2.2, depend on various system variables. However, the quantitative relationship between these variables and df of aggregates has not been fully explored, especially in predicting a wide range of df. Here, we developed a random forest model capable of predicting the complete range of aggregate df using just four simple physical and chemical parameters of the aggregating system as inputs. The model accurately predicts the df of aggregates formed by colloids of different sizes, ranging from nano to micro sizes, after being trained and tested on appropriate data sets. Ionic strength (IS) has the most significant influence on the df of aggregates formed by microsized particles followed by the relative hydrodynamic radius of aggregates (Rh/Rp), particle concentration (Cp), and primary particle radius (Rp). For aggregates formed by both nano- and microsized particles, IS still has a strong influence on the df, with the significance of Rp increasing. All four inputs are negatively correlated with predicting the df of aggregates. The predictions align well with the physical interpretations.

The suspension stability of nanoplastics in aquatic environments revealed using meta-analysis and machine learning

Nanoplastics (NPs) aggregation determines their bioavailability and risks in natural aquatic environments, which is driven by multiple environmental and polymer factors. The back propagation artificial neural network (BP-ANN) model in machine learning (R2 = 0.814) can fit the complex NPs aggregation, and the feature importance was in the order of surface charge of NPs > dissolved organic matter (DOM) > functional group of NPs > ionic strength and pH > concentration of NPs. Meta-analysis results specified low surface charge (0 ≤ |ζ| < 10 mV) of NPs, low concentration (< 1 mg/L) and low molecular weight (< 10 kg/mol) of DOM, NPs with amino groups, high ionic strength (IS > 700 mM) and acidic solution, and high concentration (≥ 20 mg/L) of NPs with smaller size (< 100 nm) contribute to NPs aggregation, which is consistent with the prediction in machine learning. Feature interaction synergistically (e.g., DOM and pH) or antagonistically (e.g., DOM and cation potential) changed NPs aggregation. Therefore, NPs were predicted to aggregate in the dry period and estuary of Poyang Lake. Research on aggregation of NPs with different particle size,shapes, and functional groups, heteroaggregation of NPs with coexisting particles and aging effects should be strengthened in the future. This study supports better assessments of the NPs fate and risks in environments.

Attachment of various-shaped polystyrene microplastics to silica surfaces: Experimental validation of the equivalent Cassini oval extended DLVO model

Microplastics (MPs) vary in shape and surface characteristics in the environment. The attachment of MPs to surfaces can be studied using the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory. However, this theory does not account for the shape MPs. Therefore, we investigated the attachment of spherical, pear-shaped, and peanut-shaped polystyrene MPs to quartz sand in NaCl and CaCl2 solutions using batch tests. The attachment of MPs to quartz sand was quantified using the attachment efficiency (alpha). Subsequently, alpha behaviors were interpreted using energy barriers (EBs) and interaction minima obtained from extended DLVO calculations, which were performed using an equivalent sphere model (ESM) and a newly developed equivalent Cassini model (ECM) to account for the shape of the MPs. The ESM failed to interpret the alpha behavior of the three MP shapes because it predicted high EBs and shallow minima. The alpha values for spherical MPs (0.62-1.00 in NaCl and 0.48-0.96 in CaCl2) were higher than those for pear- and peanut-shaped MPs (0.01-0.63 in NaCl and 0.02-0.46 in CaCl2, and 0.01-0.59 in NaCl and 0.02-0.40 in CaCl2, respectively). Conversely, the ECM could interpret the alpha behavior of pear- and peanut-shaped MPs either by changes in EBs or interaction minima as a function of orientation angles and electrolyte ionic strength. Therefore, the particle shape must be considered to improve the attachment analyses.

Mixtures Recomposition by Neural Nets: A Multidisciplinary Overview

Artificial Neural Networks (ANNs) are transforming how we understand chemical mixtures, providing an expressive view of the chemical space and multiscale processes. Their hybridization with physical knowledge can bridge the gap between predictivity and understanding of the underlying processes. This overview explores recent progress in ANNs, particularly their potential in the 'recomposition' of chemical mixtures. Graph-based representations reveal patterns among mixture components, and deep learning models excel in capturing complexity and symmetries when compared to traditional Quantitative Structure-Property Relationship models. Key components, such as Hamiltonian networks and convolution operations, play a central role in representing multiscale mixtures. The integration of ANNs with Chemical Reaction Networks and Physics-Informed Neural Networks for inverse chemical kinetic problems is also examined. The combination of sensors with ANNs shows promise in optical and biomimetic applications. A common ground is identified in the context of statistical physics, where ANN-based methods iteratively adapt their models by blending their initial states with training data. The concept of mixture recomposition unveils a reciprocal inspiration between ANNs and reactive mixtures, highlighting learning behaviors influenced by the training environment.