Nanotechnology for green applications: How far on the anvil of machine learning!

Almost 40 years since Eric Drexler’s 1991 MIT dissertation, “Molecular Machinery and Manufacturing with Applications to Computation” (Drexler, Nanosystems: molecular machinery, manufacturing, and computation, Wiley, 1992), a lot of exciting has happened in nanotechnology. By far, this nascent field of technology in combination with biotechnology has contributed heavily in medicine and biomaterials R&D sector, but nanomaterials for “green applications” are still quite away from realization. There are several issues underlying this delay, and these can be classified into two broad categories: (i) issues intrinsic to nanomaterials: associated with sustainable tailor-made production, optimization, scale-up, and stability and (ii) environmental issues of nanomaterials associated with their interaction and safe disposal. To have further clarity on first category above, the major intrinsic issues of nanomaterials are grouped into six attributes of studies. These are composition, synthesis, internal and external properties, stability, toxicity, and lifecycle assessment. Although these appear independent attributes but actually they interact with each other and a term “Ensemble Heterogeneity” (h_e) is defined here to understand their interdependence. Machine learning (ML) can provide deeper insights into both visible and hidden levels of these interdependencies. In the present chapter, firstly, these six attributes are reviewed and an understanding is developed that these attributes must be studied for hidden parameters and patterns using ML for producing optimized design solutions for nanomaterials. Secondly, an overview of different types of ML approaches is given that can contribute to the development of optimized design of nanomaterials. Finally, life cycle assessment (LCA) of nanomaterials is discussed. LCA is a tool to assess the environmental impact and sustainability of any technology to be “green.”