3 March 2022 in Nature Machine Intelligence
14 October 2015 in Nature Communications
Efficient desalination of water continues to be a problem facing the society. Advances in nanotechnology have led to the development of a variety of nanoporous membranes for water purification. Here we show, by performing molecular dynamics simulations, that a nanopore in a single-layer molybdenum disulfide can effectively reject ions and allow transport of water at a high rate. More than 88% of ions are rejected by membranes having pore areas ranging from 20 to 60 Å2. Water flux is found to be two to five orders of magnitude greater than that of other known nanoporous membranes. Pore chemistry is shown to play a significant role in modulating the water flux. Pores with only molybdenum atoms on their edges lead to higher fluxes, which are ∼70% greater than that of graphene nanopores. These observations are explained by permeation coefficients, energy barriers, water density and velocity distributions in the pores.
10 June 2020 in ACS Catalysis
SVSantiago Jimenez Villegas
Santiago Jimenez Villegas
AMAmir Hassan Bagherzadeh Mostaghimi
Amir Hassan Bagherzadeh Mostaghimi
+2
2 February 2023 in Journal of Computational Physics
22 April 2023 in npj Computational Materials
Accurate and efficient prediction of polymer properties is of great significance in polymer design. Conventionally, expensive and time-consuming experiments or simulations are required to evaluate polymer functions. Recently, Transformer models, equipped with self-attention mechanisms, have exhibited superior performance in natural language processing. However, such methods have not been investigated in polymer sciences. Herein, we report TransPolymer, a Transformer-based language model for polymer property prediction. Our proposed polymer tokenizer with chemical awareness enables learning representations from polymer sequences. Rigorous experiments on ten polymer property prediction benchmarks demonstrate the superior performance of TransPolymer. Moreover, we show that TransPolymer benefits from pretraining on large unlabeled dataset via Masked Language Modeling. Experimental results further manifest the important role of self-attention in modeling polymer sequences. We highlight this model as a promising computational tool for promoting rational polymer design and understanding structure-property relationships from a data science view.