The Effects and Mechanisms of pH and Dissolved Oxygen Conditions on the Release of Arsenic at the Sediment-Water Interface in Taihu Lake.

The pH and dissolved oxygen (DO) conditions are important environmental factors that control the migration of arsenic (As) at the sediment-water interface. This study investigates the distribution differences of reactive iron, manganese, and arsenic at the sediment-water interface under anaerobic and aerobic conditions at different pH levels. The strong buffering capacity of sediment to water pH results in a shift towards neutral pH values in the overlying water under different initial pH conditions. The level of DO becomes a key factor in the release of As from sediment, with lower DO environments exhibiting higher release quantities and rates of As compared to high DO environments. Under low DO conditions, the combined effects of ion exchange and anaerobic reduction lead to the most significant release of As, particularly under pH 9.5 conditions. The formation of amorphous ferrous sulfide compounds under low DO conditions is a significant factor contributing to increased arsenic concentration in the interstitial water. Therefore, the re-migration of endogenous arsenic in shallow lake sediments should consider the combined effects of multiple driving forces.

How Well Does Kohn-Sham Regularizer Work for Weakly Correlated Systems?

Kohn-Sham regularizer (KSR) is a differentiable machine learning approach to finding the exchange-correlation functional in Kohn-Sham density functional theory that works for strongly correlated systems. Here we test KSR for a weak correlation. We propose spin-adapted KSR (sKSR) with trainable local, semilocal, and nonlocal approximations found by minimizing density and total energy loss. We assess the atoms-to-molecules generalizability by training on one-dimensional (1D) H, He, Li, Be, and Be2+ and testing on 1D hydrogen chains, LiH, BeH2, and helium hydride complexes. The generalization error from our semilocal approximation is comparable to other differentiable approaches, but our nonlocal functional outperforms any existing machine learning functionals, predicting ground-state energies of test systems with a mean absolute error of 2.7 mH.