Cluster characterization in atom probe tomography: Machine learning using multiple summary functions

In this work, we develop a machine learning-based method to characterize intracluster concentration (ρ_c), background concentration (ρ_b), clustering radius (r̄), and radius dispersity (δr) in simulated atom probe tomography data using multiple spatial statistics summary functions to train a Bayesian regularized neural network. We build upon previous work that utilized Ripley's K-function by incorporating additional features from nearest-neighbor spatial statistics summary functions to better characterize concentration-based metrics. The addition of nearest-neighbor based features allows for highly accurate estimates of ρ_c and ρ_b, both with 90% of the predictions within 4.0% of the real value; the root-mean-square errors are reduced by 81.5% and 92.8% from predictions using only K-function based features, respectively. Additionally, including these nearest-neighbor based features improves the ability to differentiate between r̄ and δr.

Understanding Fragmentation of Organic Small Molecules in Atom Probe Tomography

In atom probe tomography of molecular organic materials, field ionization of either entire molecules or molecular fragments can occur, but the mechanism governing this behavior was not previously understood. This work explains when a doubly ionized small molecule organic material is expected to undergo fragmentation. We find that multiple detection events arising from post-ionization fragmentation of a parent molecular dication into two daughter ions is well explained by the free energy and geometries of the molecules computed using density functional theory. Of the systems studied, exergonic free energies for formation of the daughter ions, smaller activation energies for dissociation, and increases in bond length are all found to be quantitative predictors for ion fragmentation. This work expands the applicability of atom probe tomography to organic materials by increasing the fundamental understanding of processes occurring during this analysis technique.

Three dimensional cluster analysis for atom probe tomography using Ripley's K-function and machine learning

The size and structure of spatial molecular and atomic clustering can significantly impact material properties and is therefore important to accurately quantify. Ripley's K-function (K(r)), a measure of spatial correlation, can be used to perform such quantification when the material system of interest can be represented as a marked point pattern. This work demonstrates how machine learning models based on K(r)-derived metrics can accurately estimate cluster size and intra-cluster density in simulated three dimensional (3D) point patterns containing spherical clusters of varying size; over 90% of model estimates for cluster size and intra-cluster density fall within 11% and 18% error of the true values, respectively. These K(r)-based size and density estimates are then applied to an experimental APT reconstruction to characterize MgZn clusters in a 7000 series aluminum alloy. We find that the estimates are more accurate, consistent, and robust to user interaction than estimates from the popular maximum separation algorithm. Using K(r) and machine learning to measure clustering is an accurate and repeatable way to quantify this important material attribute.

Trace organic contaminant (TOrC) mixtures in Minnesota littoral zones: Effects of on-site wastewater treatment system (OWTS) proximity and biological impact

On-site wastewater treatment systems (OWTSs) are an international wastewater management strategy for rural and semi-rural communities without access to centralized sewage treatment. These systems are a suspected source of trace organic contaminants (TOrCs) that may be responsible for endocrine disrupting effects to resident fish species in Minnesota Lakes. This study assessed localized porewater concentrations of TOrCs in near-shore environments across five Minnesota Lakes. Sampling sites were designated as either likely (HOME) or unlikely (REF) to receive OWTS discharges based on their proximity to shoreline households. Sampling sites also served as sunfish spawning habitats concurrently studied for biological impacts to resident adult males. Two-group hypothesis tests demonstrated significantly (p = .02) higher total TOrC concentrations in HOME (Mean = 841 ng/L) versus REF (Mean = 222 ng/L) sites. HOME sites also contained a wider suite of TOrC detections relative to REF sites. The distance to the nearest household (most proximal distance; MPD) negatively correlated (r = -0.62) with total TOrC concentrations. However, 2,4-D and DEET were major contributors to these total concentrations, suggesting that anthropogenic influence from households may not be exclusively attributed to OWTS discharges. Further, TOrC presence and elevated nitrogen concentrations in REF site porewater suggest additional, non-household TOrC discharges to these lakes. Significantly higher blood concentrations of vitellogenin (p = .03) and 11-ketotestosterone (p = .01) were observed in adult male sunfish captured from HOME versus REF sites. Comparisons between chemical and biological data indicate enhanced bioactive effects of co-contaminants. The findings from this study demonstrate multiple diffuse transport pathways contribute to the presence of biologically active TOrC mixtures in Minnesota Lakes, and mitigation efforts should consider minimizing residential inputs of chemicals associated with both outdoor and OWTS activity.

Effect of Diels-Alder Reaction in C60-Tetracene Photovoltaic Devices

Developing organic photovoltaic materials systems requires a detailed understanding of the heterojunction interface, as it is the foundation for photovoltaic device performance. The bilayer fullerene/acene system is one of the most studied models for testing our understanding of this interface. We demonstrate that the fullerene and acene molecules chemically react at the heterojunction interface, creating a partial monolayer of a Diels-Alder cycloadduct species. Furthermore, we show that the reaction occurs during standard deposition conditions and that thermal annealing increases the concentration of the cycloadduct. The cycloaddition reaction reduces the number of sites available at the interface for charge transfer exciton recombination and decreases the charge transfer state reorganization energy, increasing the open circuit voltage. The submonolayer quantity of the cycloadduct renders it difficult to identify with conventional characterization techniques; we use atom probe tomography to overcome this limitation while also measuring the spatial distribution of each chemical species.