Machine learning-based approach for efficient prediction of toxicity of chemical gases using feature selection.

Toxic gases can be fatal as they damage many living tissues, especially the nervous and respiratory systems. They can cause permanent damage for many years by harming environmental tissue and living organisms. They can also cause mass deaths when used as chemical weapons. These chemical agents consist of organophosphates, namely ester, amide, or thiol derivatives of phosphorus, phosphonic or phosphinic acids, or can be synthesized independently. In this study, machine learning models were used to predict the toxicity of chemical gases. Toxic and non-toxic gases, consisting of 144 gases, were identified according to the United States Environmental Protection Agency, Occupational Safety and Health Administration, and the Centers for Disease Control and Prevention. Six machine-learning models were used to predict the toxicity of these chemical gases. The performance of the models was verified through internal and external validation. The results showed that the model's internal validation accuracy was 86.96% with the Relief-J48 algorithm. The accuracy value of the model was 89.65% with the Bayes Net algorithm for external validation. Our results reveal that identifying the toxicity of existing and potential chemicals is essential for the early detection of these chemicals in nature.

Strain engineered SiGe multiple-quantum-well nanomembranes for far-infrared intersubband device applications.

SiGe/Si quantum wells are of great interest for the development of Group-IV THz quantum cascade lasers. The main advantage of Group-IV over III-V materials such as GaAs is that, in the former, polar phonon scattering, which significantly diminishes the efficiency of intersubband light emission, is absent. However, for SiGe/Si multiple-quantum-well structures grown on bulk Si, the lattice mismatch between Si and Ge limits the critical thickness for dislocation formation and thus the number of periods that can be grown. Similarly, the use of composition-graded SiGe films as a lattice-matched substrate leads to the transfer of dislocations from the graded buffer substrate into the quantum wells, with a consequent decrease in light emission efficiency. Here we instead employ nanomembrane strain engineering to fabricate dislocation-free strain relaxed substrates, with lattice constants that match the average lattice constants of the quantum wells. This procedure allows for the growth of many periods with excellent structural properties. The samples in this work were grown by low-pressure chemical vapor deposition and characterized via high-resolution X-ray diffraction and far-infrared transmission spectroscopy, showing narrow intersubband absorption features indicative of high crystalline quality.