A predicted-loss based active learning approach for robust cancer pathology image analysis in the workplace.

BACKGROUND: Convolutional neural network-based image processing research is actively being conducted for pathology image analysis. As a convolutional neural network model requires a large amount of image data for training, active learning (AL) has been developed to produce efficient learning with a small amount of training data. However, existing studies have not specifically considered the characteristics of pathological data collected from the workplace. For various reasons, noisy patches can be selected instead of clean patches during AL, thereby reducing its efficiency. This study proposes an effective AL method for cancer pathology that works robustly on noisy datasets. METHODS: Our proposed method to develop a robust AL approach for noisy histopathology datasets consists of the following three steps: 1) training a loss prediction module, 2) collecting predicted loss values, and 3) sampling data for labeling. This proposed method calculates the amount of information in unlabeled data as predicted loss values and removes noisy data based on predicted loss values to reduce the rate at which noisy data are selected from the unlabeled dataset. We identified a suitable threshold for optimizing the efficiency of AL through sensitivity analysis. RESULTS: We compared the results obtained with the identified threshold with those of existing representative AL methods. In the final iteration, the proposed method achieved a performance of 91.7% on the noisy dataset and 92.4% on the clean dataset, resulting in a performance reduction of less than 1%. Concomitantly, the noise selection ratio averaged only 2.93% on each iteration. CONCLUSIONS: The proposed AL method showed robust performance on datasets containing noisy data by avoiding data selection in predictive loss intervals where noisy data are likely to be distributed. The proposed method contributes to medical image analysis by screening data and producing a robust and effective classification model tailored for cancer pathology image processing in the workplace.

Access to essential cancer medicines for children: a comparative mixed-methods analysis of availability, price, and health-system determinants in east Africa.

BACKGROUND: Access to essential childhood cancer medicines is a core determinant of childhood cancer outcomes. Available evidence, although scarce, suggests that access to these medicines is highly variable across countries, particularly in low-income and middle-income countries, where the burden of childhood cancer is greatest. To support evidence-informed national and regional policies for improved childhood cancer outcomes, we aimed to analyse access to essential childhood cancer medicines in four east African countries-Kenya, Rwanda, Tanzania, and Uganda-by determining the availability and price of these medicines and the health system determinants of access. METHODS: In this comparative analysis, we used prospective mixed-method analyses to track and analyse the availability and price of essential childhood cancer medicines, investigate contextual determinants of access to childhood cancer medicines within and across included countries, and assess the potential effects of medicine stockouts on treatment. Eight tertiary care hospitals were included, seven were public sites (Kenyatta National Hospital [KNH; Nairobi, Kenya], Jaramogi Oginga Odinga Referral and Teaching Hospital [JOORTH; Kisumu, Kenya], Moi University Teaching and Referral Hospital [MTRH; Eldoret, Kenya], Bugando Medical Centre [BMC; Mwanza, Tanzania], Muhimbili National Hospital [MNH; Dar es Salaam, Tanzania], Butaro Cancer Centre of Excellence [BCCE; Butaro Sector, Rwanda], and Uganda Cancer Institute [UCI; Kampala, Uganda]) and one was a private site (Aga Khan University Hospital [AKU; Nairobi, Kenya]). We catalogued prices and stockouts for 37 essential drugs from each of the eight study siteson the basis of 52 weeks of prospective data that was collected across sites from May 1, 2020, to Jan 31, 2022. We analysed determinants of medicine access using thematic analysis of academic literature, policy documents, and semi-structured interviews from a purposive sample of health system stakeholders. FINDINGS: Recurrent stockouts of a wide range of cytotoxic and supportive care medicines were observed across sites, with highest mean unavailability in Kenya (JOORTH; 48·5%), Rwanda (BCCE; 39·0%), and Tanzania (BMC; 32·2%). Drugs that had frequent stockouts across at least four sites included methotrexate, bleomycin, etoposide, ifosfamide, oral morphine, and allopurinol. Average median price ratio of medicines at each site was within WHO's internationally accepted threshold for efficient procurement (median price ratio ≤1·5). The effect of stockouts on treatment was noted across most sites, with the greatest potential for treatment interruptions in patients with Hodgkin lymphoma, retinoblastoma, and acute lymphocytic leukaemia. Policy prioritisation of childhood cancers, health financing and coverage, medicine procurement and supply chain management, and health system infrastructure emerged as four prominent determinants of access when the stratified purposive sample of key informants (n=64) across all four countries (Kenya n=19, Rwanda n=15, Tanzania n=13, and Uganda n=17) was interviewed. INTERPRETATION: Access to childhood cancer medicines across east Africa is marked by gaps in availability that have implications for effective treatment delivery for a range of childhood cancers. Our findings provide detailed evidence of barriers to access to childhood cancer medicine at multiple points in the pharmaceutical value chain. These data could inform national and regional policy makers to optimise cancer medicine availability and affordability as part of efforts to improve childhood cancer outcomes specific regions and internationally. FUNDING: American Childhood Cancer Organization, Childhood Cancer International, and the Friends of Cancer Patients Ameera Fund.

Implementation of real-time TDDFT for periodic systems in the open-source PySCF software package.

We present a new implementation of real-time time-dependent density functional theory (RT-TDDFT) for calculating excited-state dynamics of periodic systems in the open-source Python-based PySCF software package. Our implementation uses Gaussian basis functions in a velocity gauge formalism and can be applied to periodic surfaces, condensed-phase, and molecular systems. As representative benchmark applications, we present optical absorption calculations of various molecular and bulk systems and a real-time simulation of field-induced dynamics of a (ZnO)4 molecular cluster on a periodic graphene sheet. We present representative calculations on optical response of solids to infinitesimal external fields as well as real-time charge-transfer dynamics induced by strong pulsed laser fields. Due to the widespread use of the Python language, our RT-TDDFT implementation can be easily modified and provides a new capability in the PySCF code for real-time excited-state calculations of chemical and material systems.

Degradation of Perfluorooctanoic Acid on Aluminum Oxide Surfaces: New Mechanisms from Ab Initio Molecular Dynamics Simulations.

Perfluorooctanoic acid (PFOA) is a part of a large group of anthropogenic, persistent, and bioaccumulative contaminants known as per- and polyfluoroalkyl substances (PFAS) that can be harmful to human health. In this work, we present the first ab initio molecular dynamics (AIMD) study of temperature-dependent degradation dynamics of PFOA on (100) and (110) surfaces of γ-Al2O3. Our results show that PFOA degradation does not occur on the pristine (100) surface, even when carried out at high temperatures. However, introducing an oxygen vacancy on the (100) surface facilitates an ultrafast (<100 fs) defluorination of C-F bonds in PFOA. We also examined degradation dynamics on the (110) surface and found that PFOA interacts strongly with Al(III) centers on the surface of γ-Al2O3, resulting in a stepwise breaking of C-F, C-C, and C-COO bonds. Most importantly, at the end of the degradation process, strong Al-F bonds are formed on the mineralized γ-Al2O3 surface, which prevents further dissociation of fluorine into the surrounding environment. Taken together, our AIMD simulations provide critical reaction mechanisms at a quantum level of detail and highlight the importance of temperature effects, defects, and surface facets for PFOA degradation on reactive surfaces, which have not been systematically explored or analyzed.

Ring currents modulate optoelectronic properties of aromatic chromophores at 25 T.

The properties of organic molecules can be influenced by magnetic fields, and these magnetic field effects are diverse. They range from inducing nuclear Zeeman splitting for structural determination in NMR spectroscopy to polaron Zeeman splitting organic spintronics and organic magnetoresistance. A pervasive magnetic field effect on an aromatic molecule is the aromatic ring current, which can be thought of as an induction of a circular current of π-electrons upon the application of a magnetic field perpendicular to the π-system of the molecule. While in NMR spectroscopy the effects of ring currents on the chemical shifts of nearby protons are relatively well understood, and even predictable, the consequences of these modified electronic states on the spectroscopy of molecules has remained unknown. In this work, we find that photophysical properties of model phthalocyanine compounds and their aggregates display clear magnetic field dependences up to 25 T, with the aggregates showing more drastic magnetic field sensitivities depending on the intermolecular interactions with the amplification of ring currents in stacked aggregates. These observations are consistent with ring currents measured in NMR spectroscopy and simulated in time-dependent density functional theory calculations of magnetic field-dependent phthalocyanine monomer and dimer absorption spectra. We propose that ring currents in organic semiconductors, which commonly comprise aromatic moieties, may present new opportunities for the understanding and exploitation of combined optical, electronic, and magnetic properties.

GPU-Enhanced DFTB Metadynamics for Efficiently Predicting Free Energies of Biochemical Systems.

Metadynamics calculations of large chemical systems with ab initio methods are computationally prohibitive due to the extensive sampling required to simulate the large degrees of freedom in these systems. To address this computational bottleneck, we utilized a GPU-enhanced density functional tight binding (DFTB) approach on a massively parallelized cloud computing platform to efficiently calculate the thermodynamics and metadynamics of biochemical systems. To first validate our approach, we calculated the free-energy surfaces of alanine dipeptide and showed that our GPU-enhanced DFTB calculations qualitatively agree with computationally-intensive hybrid DFT benchmarks, whereas classical force fields give significant errors. Most importantly, we show that our GPU-accelerated DFTB calculations are significantly faster than previous approaches by up to two orders of magnitude. To further extend our GPU-enhanced DFTB approach, we also carried out a 10 ns metadynamics simulation of remdesivir, which is prohibitively out of reach for routine DFT-based metadynamics calculations. We find that the free-energy surfaces of remdesivir obtained from DFTB and classical force fields differ significantly, where the latter overestimates the internal energy contribution of high free-energy states. Taken together, our benchmark tests, analyses, and extensions to large biochemical systems highlight the use of GPU-enhanced DFTB simulations for efficiently predicting the free-energy surfaces/thermodynamics of large biochemical systems.

Hexaazatriphenylene-Based Two-Dimensional Conductive Covalent Organic Framework with Anisotropic Charge Transfer.

The development of covalent organic frameworks (COFs) with efficient charge transport is of immense interest for applications in optoelectronic devices. To enhance COF charge transport properties, electroactive building blocks and dopants can be used to induce extended conduction channels. However, understanding their intricate interplay remains challenging. We designed and synthesized a tailor-made COF structure with electroactive hexaazatriphenylene (HAT) core units and planar dioxin (D) linkages, denoted as HD-COF. With the support of theoretical calculations, we found that the HAT units in the HD-COF induce strong, eclipsed π-π stacking. The unique stacking of HAT units and the weak in-plane conjugation of dioxin linkages leads to efficient anisotropic charge transport. We fabricated HD-COF films to minimize the grain boundary effect of bulk COFs, which resulted in enhanced conductivity. As a result, the HD-COF films showed an electrical conductivity as high as 1.25 S cm-1 after doping with tris(4-bromophenyl)ammoniumyl hexachloroantimonate.

Degradation of Per- and Polyfluoroalkyl Substances with Hydrated Electrons: A New Mechanism from First-Principles Calculations.

Per- and polyfluoroalkyl substances (PFASs) are synthetic contaminants found in drinking groundwater sources and a wide variety of consumer products. Because of their adverse environmental and human health effects, remediation of these persistent compounds has attracted significant recent attention. To gain mechanistic insight into their remediation, we present the first ab initio study of PFAS degradation via hydrated electrons─a configuration that has not been correctly considered in previous computational studies up to this point. To capture these complex dynamical effects, we harness ab initio molecular dynamics (AIMD) simulations to probe the reactivities of perfluorooctanoic (PFOA) and perfluorooctane sulfonic acid (PFOS) with hydrated electrons in explicit water. We complement our AIMD calculations with advanced metadynamics sampling techniques to compute free energy profiles and detailed statistical analyses of PFOA/PFOS dynamics. Although our calculations show that the activation barrier for C-F bond dissociation in PFOS is three times larger than that in PFOA, all the computed free energy barriers are still relatively low, resulting in a diffusion-limited process. We discuss our results in the context of recent studies on PFAS degradation with hydrated electrons to give insight into the most efficient remediation strategies for these contaminants. Most importantly, we show that the degradation of PFASs with hydrated electrons is markedly different from that with excess electrons/charges, a common (but largely incomplete) approach used in several earlier computational studies.

Harnessing deep reinforcement learning to construct time-dependent optimal fields for quantum control dynamics.

We present an efficient deep reinforcement learning (DRL) approach to automatically construct time-dependent optimal control fields that enable desired transitions in dynamical chemical systems. Our DRL approach gives impressive performance in constructing optimal control fields, even for cases that are difficult to converge with existing gradient-based approaches. We provide a detailed description of the algorithms and hyperparameters as well as performance metrics for our DRL-based approach. Our results demonstrate that DRL can be employed as an effective artificial intelligence approach to efficiently and autonomously design control fields in quantum dynamical chemical systems.

Plasmon-induced excitation energy transfer in silver nanoparticle dimers: A real-time TDDFTB investigation.

Using real-time quantum dynamics calculations, we perform theoretical investigations of light-induced interactions and electronic excitation transfer in a silver nanoparticle dimer. Real-time time-dependent density functional tight-binding (RT-TDDFTB) calculations provide details of the quantum dynamical processes at an electronic/atomistic level with attosecond resolution. The computational efficiency of RT-TDDFTB allows us to examine electronic dynamics up to picosecond time scales. With time scales varying over six orders of magnitude, we provide insight into interactions between the nanoparticle and laser and between nanoparticles. Our results show that the coupling between nanoparticle monomers is dependent on the separation distance between the nanoparticles in the dimer. As the interparticle distance is varied, the dipole-dipole interactions and electronic excitation transfer mechanisms are markedly different. At large distances (from 50 to 20 Å), the energy transfer from NP1 to NP2 becomes more efficient as the interparticle distance decreases. The total dipole moment of the Ag14 nanoparticle dimer increases linearly at an interparticle distance of 20 Å and reaches its maximum after 1.2 ps. The electronic excitation transfer is also the most efficient at 20 Å. At short distances, back-transfer effects reduce the ability of the dimer and NP1 to accept energy from the incident electric field. We attribute the distance-dependent features of the nanoparticle dimer to the beating between the laser acting on NP1 and the back transfer from NP2 to NP1.

Prosthetic Replacement of the Ocular Surface Ecosystem for Limbal Stem Cell Deficiency: A Case Series.

OBJECTIVES: To assess outcomes of limbal stem cell deficiency (LSCD) in patients treated with Prosthetic Replacement of the Ocular Surface Ecosystem (PROSE). METHODS: Retrospective case series. Patients with LSCD who received PROSE treatment were included. Data including best-corrected visual acuity (BCVA) and LSCD staging before and after PROSE dispensing were collected to characterize each case. RESULTS: Five eyes of four patients were included. All patients were female, with an age range of 21 to 80 years. Each patient received a PROSE device with diameters ranging from 16 to 18.5 mm. Follow-up ranged from 11 to 29 months. Tolerated wear times ranged from 3.5 to 10 hr daily. Four eyes showed improved BCVA and unchanged LSCD staging as per the global consensus after PROSE treatment. Three of these eyes had stage 3 and one had stage 1C LSCD at diagnosis. The fifth eye had worse BCVA and recurrence of stage 3 LSCD post-living-related conjunctival limbal allograft transplant despite PROSE treatment. CONCLUSIONS: Prosthetic Replacement of the Ocular Surface Ecosystem may be a viable treatment for LSCD, including severe cases, because it can provide symptom relief and improve vision. Its customizability, as demonstrated in this study, is beneficial for troubleshooting issues with fitting. Future studies are needed to further assess PROSE as treatment for LSCD.

Prosthetic Replacement of the Ocular Surface Ecosystem for Terrien Marginal Degeneration: A Case Series.

OBJECTIVES: To assess outcomes of the Prosthetic Replacement of the Ocular Surface Ecosystem (PROSE) treatment in patients with advanced Terrien marginal degeneration (TMD). METHODS: This is a retrospective case series of patients with advanced TMD who were assessed and fit with customized PROSE lenses. Data were collected on PROSE fitting details including visual acuity (VA) before and after PROSE, slit-lamp findings, and corneal tomography scans. RESULTS: Six eyes in four patients were included. All patients attempted at least one other contact lens (CL) modality before PROSE. Some patients had corneal comorbidities such as pseudopterygium and pseudobleb that contributed to intolerance to previous lenses and warranted extra considerations in the fitting process. With PROSE, VA improved in all six eyes. Patients with structural corneal comorbidities achieved improved vision, comfort, and lens tolerance with PROSE. Two eyes had noncorneal ocular comorbidities that limited PROSE efficacy. Another eye discontinued PROSE wear because of limbal stem-cell disease progression necessitating a limbal stem-cell transplant. CONCLUSIONS: PROSE treatment can be an effective option to improve vision and comfort for patients with advanced TMD who are intolerant to first-line therapeutic CL modalities, even in the presence of other corneal comorbidities.

Machine Learning: New Ideas and Tools in Environmental Science and Engineering.

The rapid increase in both the quantity and complexity of data that are being generated daily in the field of environmental science and engineering (ESE) demands accompanied advancement in data analytics. Advanced data analysis approaches, such as machine learning (ML), have become indispensable tools for revealing hidden patterns or deducing correlations for which conventional analytical methods face limitations or challenges. However, ML concepts and practices have not been widely utilized by researchers in ESE. This feature explores the potential of ML to revolutionize data analysis and modeling in the ESE field, and covers the essential knowledge needed for such applications. First, we use five examples to illustrate how ML addresses complex ESE problems. We then summarize four major types of applications of ML in ESE: making predictions; extracting feature importance; detecting anomalies; and discovering new materials or chemicals. Next, we introduce the essential knowledge required and current shortcomings in ML applications in ESE, with a focus on three important but often overlooked components when applying ML: correct model development, proper model interpretation, and sound applicability analysis. Finally, we discuss challenges and future opportunities in the application of ML tools in ESE to highlight the potential of ML in this field.

Predicting Hearing Aid Satisfaction in Adults: A Systematic Review of Speech-in-noise Tests and Other Behavioral Measures.

OBJECTIVES: Adults with hearing loss report a wide range of hearing aid satisfaction that does not significantly correlate to degree of hearing loss. It is not clear which auditory behavioral factors do contribute to hearing aid satisfaction. While poor speech understanding in noise is known to contribute to dissatisfaction, there are many categories of this type of assessment. The purpose of this systematic review is to answer the question, "Are behavioral pre-fitting measures using speech and nonspeech materials related to hearing aid satisfaction among adults?" DESIGN: Six electronic databases were searched to find peer-reviewed studies published before June 2020. The included studies reported on the relationship between auditory behavioral measures and hearing aid satisfaction alone or globally with other outcome domains among adults with hearing loss. Six types of behavioral prefitting measures were evaluated: speech recognition in quiet (% correct), speech recognition in noise (% correct), reception thresholds for speech-in-noise, speech-based subjective ratings, dichotic speech tests, and tests using nonspeech material. Each relevant study was independently reviewed by two reviewers. Methodological quality was evaluated in each included study using the American Speech-Language-Hearing Association's level of evidence ratings. RESULTS: There were 1342 articles identified in the systematic review process. After duplicates were removed and specific inclusion criteria were applied, 21 studies were included. All studies included had a 0 to 4 methodological quality rating indicating weak to moderate internal validity. The tests that showed potential for clinical application due to significant correlations with satisfaction were the QuickSIN, the synthetic sentence identification, the hearing in noise test, and the acceptable noise level test. Audibility, as measured by degree of hearing loss, was not significantly correlated to hearing aid satisfaction in the 13 studies that reported on this measure. CONCLUSIONS: Based on this review, results indicated that speech-in-noise tests had the highest associations to hearing aid satisfaction, suggesting a greater role for assessment of speech-in-noise perception in auditory rehabilitation. This is an important finding for clinical practice, given that audibility was not a significant factor in predicting satisfaction. Overall, the results from this review show a need for well-designed, high-quality, prospective studies assessing the predictive value of prefitting measures on hearing aid satisfaction with current hearing aid models.

Diagnoses and Outcomes of Prosthetic Replacement of the Ocular Surface Ecosystem Treatment-A Canadian Experience.

OBJECTIVES: To investigate underlying diagnoses and outcomes of patients undergoing Prosthetic Replacement of the Ocular Surface Ecosystem (PROSE) treatment at the first Canadian PROSE center. METHODS: A retrospective chart review was conducted on patients referred for PROSE treatment and fitted with PROSE devices from 2018 to 2020. Data were collected on diagnoses, presenting symptoms, previous lens modalities attempted, best-corrected visual acuities (BCVAs) pre-PROSE and post-PROSE, daily wear time, and failure rates. Best-corrected visual acuities pre-PROSE and post-PROSE were compared to evaluate visual improvement. RESULTS: In total, 78 patients (126 eyes) were analyzed. The most common diagnoses were keratoconus (n=39 eyes) and postcorneal graft (n=15) in the distorted cornea group, and limbal stem cell deficiency (n=17) and graft versus host disease (n=15) in the ocular surface disease (OSD) group. Most frequent symptoms included blur, photophobia, and pain. Most common lens modalities attempted pre-PROSE were conventional scleral lenses and glasses. The overall mean BCVA improvement was 0.40 logarithm of the minimal angle of resolution (logMAR) (4-lines Snellen) (P<0.0001). Best-corrected visual acuities improvement in the distorted cornea group (0.52 logMAR, 5-lines) was significantly greater than in the OSD group (0.29 logMAR, 3-lines) (P=0.004). CONCLUSIONS: Prosthetic replacement of the ocular surface ecosystem treatment can provide significant visual improvement for patients with distorted corneal surfaces and OSDs who failed other lens modalities.

Fractional occupation numbers and self-interaction correction-scaling methods with the Fermi-Löwdin orbital self-interaction correction approach.

We present a new assessment of the Fermi-Löwdin orbital self-interaction correction (FLO-SIC) approach with an emphasis on its performance for predicting energies as a function of fractional occupation numbers (FONs) for various multielectron systems. Our approach is implemented in the massively parallelized NWChem quantum chemistry software package and has been benchmarked on the prediction of total energies, atomization energies, and ionization potentials of small molecules and relatively large aromatic systems. Within our study, we also derive an alternate expression for the FLO-SIC energy gradient expressed in terms of gradients of the Fermi-orbital eigenvalues and revisit how the FLO-SIC methodology can be seen as a constrained unitary transformation of the canonical Kohn-Sham orbitals. Finally, we conclude with calculations of energies as a function of FONs using various SIC-scaling methods to test the limits of the FLO-SIC formalism on a variety of multielectron systems. We find that these relatively simple scaling methods do improve the prediction of total energies of atomic systems as well as enhance the accuracy of energies as a function of FONs for other multielectron chemical species.

Structural Dependence of Reductive Defluorination of Linear PFAS Compounds in a UV/Electrochemical System.

Per and polyfluoroalkyl substances (PFAS), legacy chemicals used in firefighting and the manufacturing of many industrial and consumer goods, are widely found in groundwater resources, along with other regulated compounds, such as chlorinated solvents. Due to their strong C-F bonds, these molecules are extremely recalcitrant, requiring advanced treatment methods for effective remediation, with hydrated electrons shown to be able to defluorinated these compounds. A combined photo/electrochemical method has been demonstrated to dramatically increase defluorination rates, where PFAS molecules sorbed onto appropriately functionalized cathodes charged to low cell potentials (-0.58 V vs Ag/AgCl) undergo a transient electron transfer event from the electrode, which "primes" the molecule by reducing the C-F bond strength and enables the bond's dissociation upon the absorption of a hydrated electron. In this work, we explore the impact of headgroup and chain length on the performance of this two-electron process and extend this technique to chlorinated solvents. We use isotopically labeled PFAS molecules to take advantage of the kinetic isotope effect and demonstrate that indeed PFAS defluorination is likely driven by a two-electron process. We also present density functional theory calculations to illustrate that the externally applied potential resulted in an increased rate of electron transfer, which ultimately increased the measured defluorination rate.

Degradation of Perfluoroalkyl Ether Carboxylic Acids with Hydrated Electrons: Structure-Reactivity Relationships and Environmental Implications.

This study explores structure-reactivity relationships for the degradation of emerging perfluoroalkyl ether carboxylic acid (PFECA) pollutants with ultraviolet-generated hydrated electrons (eaq-). The rate and extent of PFECA degradation depend on both the branching extent and the chain length of oxygen-segregated fluoroalkyl moieties. Kinetic measurements, theoretical calculations, and transformation product analyses provide a comprehensive understanding of the PFECA degradation mechanisms and pathways. In comparison to traditional full-carbon-chain perfluorocarboxylic acids, the distinct degradation behavior of PFECAs is attributed to their ether structures. The ether oxygen atoms increase the bond dissociation energy of the C-F bonds on the adjacent -CF2- moieties. This impact reduces the formation of H/F-exchanged polyfluorinated products that are recalcitrant to reductive defluorination. Instead, the cleavage of ether C-O bonds generates unstable perfluoroalcohols and thus promotes deep defluorination of short fluoroalkyl moieties. In comparison to linear PFECAs, branched PFECAs have a higher tendency of H/F exchange on the tertiary carbon and thus lower percentages of defluorination. These findings provide mechanistic insights for an improved design and efficient degradation of fluorochemicals.

Harnessing deep neural networks to solve inverse problems in quantum dynamics: machine-learned predictions of time-dependent optimal control fields.

Inverse problems continue to garner immense interest in the physical sciences, particularly in the context of controlling desired phenomena in non-equilibrium systems. In this work, we utilize a series of deep neural networks for predicting time-dependent optimal control fields, E(t), that enable desired electronic transitions in reduced-dimensional quantum dynamical systems. To solve this inverse problem, we investigated two independent machine learning approaches: (1) a feedforward neural network for predicting the frequency and amplitude content of the power spectrum in the frequency domain (i.e., the Fourier transform of E(t)), and (2) a cross-correlation neural network approach for directly predicting E(t) in the time domain. Both of these machine learning methods give complementary approaches for probing the underlying quantum dynamics and also exhibit impressive performance in accurately predicting both the frequency and strength of the optimal control field. We provide detailed architectures and hyperparameters for these deep neural networks as well as performance metrics for each of our machine-learned models. From these results, we show that machine learning, particularly deep neural networks, can be employed as cost-effective statistical approaches for designing electromagnetic fields to enable desired transitions in these quantum dynamical systems.

Real-time degradation dynamics of hydrated per- and polyfluoroalkyl substances (PFASs) in the presence of excess electrons.

Per- and polyfluoroalkyl substances (PFASs) are synthetic chemicals that are harmful to both the environment and human health. Using self-interaction-corrected Born-Oppenheimer molecular dynamics simulations, we provide the first real-time assessment of PFAS degradation in the presence of excess electrons. In particular, we show that the initial phase of the degradation involves the transformation of an alkane-type C-C bond into an alkene-type C[double bond, length as m-dash]C bond in the PFAS molecule, which is initiated by the trans elimination of fluorine atoms bonded to these adjacent carbon atoms.