Effects of soil colloids on adsorption and migration of benzo(a)pyrene on contaminated sites under runoff infiltration processes.

In the environment, soil colloids are widespread and possess a significant adsorption capacity. This makes them capable of transporting different pollutants, presenting a potential risk to human and ecological well-being. This study aimed to examine the adsorption and co-migration characteristics of benzo(a)pyrene (BaP) and soil colloids in areas contaminated with organic substances, utilizing both static and dynamic batch experiments. In the static adsorption experiments, it was observed that the adsorption of BaP onto soil colloids followed the pseudo-second-order kinetic model (R2 = 0.966), and the adsorption isotherm conformed to the Langmuir model (R2 = 0.995). The BaP and soil colloids primarily formed bonds through π-π interactions and hydrogen bonds. The dynamic experimental outcomes revealed that elevating colloids concentration contributed to increased BaP mobility. Specifically, when the concentration of soil colloids in influent was 500 mg L-1, the mobility of BaP was 23.2 % compared to that without colloids of 13.4 %. Meanwhile, the lowering influent pH value contributed to increased BaP mobility. Specifically, when the influent pH value was 4.0, the mobility of BaP was 30.1 %. The BaP's mobility gradually declined as the initial concentration of BaP in polluted soil increased. Specifically, when the initial concentration of BaP in polluted soil was 5.27 mg kg-1, the mobility of BaP was 39.1 %. This study provides a support for controlling BaP pollution in soil and groundwater.

AMSUnet: A neural network using atrous multi-scale convolution for medical image segmentation.

In recent years, Unet and its variants have gained astounding success in the realm of medical image processing. However, some Unet variant networks enhance their performance while increasing the number of parameters tremendously. For lightweight and performance enhancement jointly considerations, inspired by SegNeXt, we develop a medical image segmentation network model using atrous multi-scale (AMS) convolution, named AMSUnet. In particular, we construct a convolutional attention block AMS using atrous and multi-scale convolution, and redesign the downsampling encoder based on this block, called AMSE. To enhance feature fusion, we design a residual attention mechanism module (i.e., RSC) and apply it to the skip connection. Compared with existing models, our model only needs 2.62 M parameters to achieve the purpose of lightweight. According to experimental results on various datasets, the segmentation performance of the designed model is superior for small, medium, and large-scale targets. Code will be available at https://github.com/llluochen/AMSUnet.