Environmental Phosphorus Recovery Based on Molecular Bioscavengers : From Quantum Mechanics to Continuum Physics

Abstract

Phosphorus is a ubiquitous element of all known life and as such it is found throughout numerous key molecules related to various cellular functions. The supply of phosphorus is tightly linked to global food security, since phosphorus is used to produce agricultural fertilizers, without which it would not be possible to feed the world population. Sadly, the current supply of phosphorus is based on the gradual depletion of limited fossil reserves, and some estimates predict that within 15-25 years we will consume more phosphorus than we can produce. There is therefore a strong international pressure to develop sustainable phosphorus practices as well as new technologies for phosphorus recovery. Nature has spent billions of years refining proteins that interact with phosphates. This has inspired the present work where the overall ambitions are: to facilitate the development of a recovery technology based on biological phosphorus scavengers, to examine fundamental molecular system aspects relevant for such a technology, and to motivate the use of computational techniques throughout an iterative design process of such a technology. A wide spectrum of computational methods, from atomic-scale quantum calculations to macro-scale fluid simulations, are employed to hint at the potential of a recovery technology based on molecular bioscavengers. As a first approach, data mining is used to obtain statistical information about how proteins in nature interact with phosphate groups, thereby revealing characteristic amino acid distributions of the binding sites. Quantum mechanical methods are used to investigate how phosphate moieties are described using electronic structure methods, and molecular dynamics in combination with quantum mechanics are used to show how the dynamical interaction between phosphates and proteins can be described – it is found that certain commonly used computational methods, including B3LYP, are ill-suited for characterizing interactions with phosphate groups, but nevertheless that phosphate-protein interactions can efficiently be quantified using other methods, e.g. wB97XD or PM6. Finally, it is shown how computational fluid dynamics can be used to optimize large-scale industrial processes using an open-source model, which we have made freely available online to the membrane community, and the advantages/disadvantages of different potential physical implementations of the proposed scavenger technology are discussed.