Search

The sorption properties of TiO nanoparticles (NPs) on sand were investigated in the presence of Suwannee River humic acid (SRHA) as a surrogate for natural organic matter. SRHA drastically reduced the sorption preference regardless of the initial pH and ionic strength of the solution. Despite the higher SRHA loading onto sand versus the TiO NPs, the consequent zeta ( ζ)-potential drop was more significant in the TiO NPs, indicating that the decreased sorption should be primarily ascribed to the fraction of SRHA sorbed onto the TiO NP surfaces. The hindering effect of SRHA on the sorption increased with SRHA concentration, but stabilized at the concentration equivalent to full surface coverage of the TiO NPs and sand in the reaction system. Interaction force measurements indicated that SRHA inhibits the approach of the TiO NPs to the sand surface, as expected by the Derjaguin-Landau-Verwey-Overbeek calculation. However, the adhesion force of TiO NPs to sand was not critically affected by the presence of SRHA. TiO NP sorption to natural sand was limited as to the acid-treated sand in the presence of SRHA, implying that the organics dissolved from natural sand might facilitate the transport of TiO NPs in aquifers by inhibiting TiO NP sorption to aquifer materials. [ABSTRACT FROM AUTHOR]

The ultimate goal of our study is to establish thin-layer chromatography (TLC) as a quick and simple method for identifying the type of refined petroleum products present in the environmental media. As a preliminary step, TLC chromatograms of different petroleum products, including gasoline, kerosene, and diesel, were characterized and compared. Methanol was determined as the optimum carrier solution in TLC analysis. The spherical-shaped TLC chromatogram of gasoline showed the longest migration distance, and thus the highest retardation factor ( R ) of 0.91. This was followed by that of kerosene (0.63) with an elliptical-shaped, and diesel (0.24) with an elongated trapezoid-shaped chromatogram. R of kerosene and diesel increased with the dilution factor, while gasoline showed a constant value. Additionally, it was observed that the TLC chromatograms of oils produced the same peak pattern with the corresponding petroleum products in gas chromatography (GC). A mixed sample of kerosene and diesel presented a triangular shaped chromatogram, underlining the need to consider the shape of chromatogram in addition to the R value, as an indicator of the petroleum type. The findings indicate that TLC has a huge potential to be used as a quick and reliable method for identifying the type of refined petroleum products in the environmental media. [ABSTRACT FROM AUTHOR]

In this study, indigenous microbes enabling Fe reduction under saline groundwater conditions were identified, and their potential contribution to Fe release from aquifer sediments during managed aquifer recharge (MAR) was evaluated. Sediment and groundwater samples were collected from a MAR feasibility test site in Korea, where adjacent river water will be injected into the confined aquifer. The residual groundwater had a high salinity over 26.0 psu, as well as strong reducing conditions (dissolved oxygen, DO < 2.0 mg/L; oxidation-reduction potential, ORP < − 100 mV) with high Fe 2 + concentrations. The indigenous microbes that mediate the reduction of Fe-minerals in this deep saline aquifer were found to be Citrobacter sp. However, column experiments to simulate field operation scenarios indicated that additional Fe release would be limited during MAR, as the dominant microbial community in the sediment would shift from Citrobacter sp. to Pseudomonas sp. and Limnohabitans sp. as river water injection alters the pore water chemistry. [ABSTRACT FROM AUTHOR]

We propose a parameter server system for distributed ML, which follows a Stale Synchronous Parallel (SSP) model of computation that maximizes the time computational workers spend doing useful work on ML algorithms, while still providing correctness guarantees. The parameter server provides an easy-to-use shared interface for read/write access to an ML model's values (parameters and variables), and the SSP model allows distributed workers to read older, stale versions of these values from a local cache, instead of waiting to get them from a central storage. This significantly increases the proportion of time workers spend computing, as opposed to waiting. Furthermore, the SSP model ensures ML algorithm correctness by limiting the maximum age of the stale values. We provide a proof of correctness under SSP, as well as empirical results demonstrating that the SSP model achieves faster algorithm convergence on several different ML problems, compared to fully-synchronous and asynchronous schemes.