(359b) Systematic Chemicals-Based Product Development, Analysis and Chemical Substitution

Regardless of geographical location, society needs to use a variety of products that are directly or indirectly connected to chemicals [1]. Over 95 percent of all manufactured goods used by modern society, rely on some form of industrial chemical processes; nearly 900,000 chemical substances have been identified; an estimated 40,000 to 60,000 industrial chemicals are currently found in commerce globally; for less than 20,000 chemicals, measured quantitative property data are available. As the number of chemicals being used and their applications in processes and/or products grow, their implications on human health and environment as well as the economy need to be understood and analysed [2]. Also, where necessary, substitutes need to be found before the identified chemical(s) causes damage. What is urgently needed is a systematic computer-aided system that can quickly, reliably and efficiently analyse a chemicals-based product or a production process to identify chemicals that could potentially cause problems and find more acceptable substitutes before it is too late.

The goal of this presentation is to highlight the latest version of a framework [3] consisting of methods and associated software components for chemical product design, chemicals analysis, and based on it, chemical substitution, if necessary. Through the framework, a new prototype software, ChemSub, is able to access a suite of databases, property estimation models, product design-analysis methods and a chemical substitution method. The suite of databases, linked through a specially developed ontology for knowledge representation, consists of a products database, where the identified chemicals in the product are linked to other databases, such as several properties databases, a hazardous effects database, banned chemicals database, and a database of known chemical substitutes. The combined databases cover close to 900,000 chemicals. The associated software components are based on properties estimation methods, product design-analysis methods, developed and reported earlier by the co-authors [4, 5, 6]. The work-flow for a consists of 3 algorithms. In algorithm-1, the product type together with available information on the product in terms of basic information, specifications, applications-use and, user-defined chemicals present in the product are established. Here, using the identified chemicals as the link, data on properties (from a properties database), hazardous effects (from a hazards database), restrictions on use (from a banned-restricted chemical database) are retrieved. If any chemical is flagged for its properties of hazardous effects or restrictions on use, it is put in a list for substitution. In algorithm-2, if chemical substitution is desired, in-house model-based methods for molecular and/or mixture design and/or database search are applied. Note that substitution may also be desired for economical and operability reasons, for example, with respect solvents, refrigerants, and additives of various types. In algorithm-3, a final product analysis in terms sustainability, LCA indicators, safety, costs, etc., are established. Application of ChemSub and the main features of the framework will be highlighted through analysis and improvement of well-known chemicals-based products.

References:

1. L Zhang, H Mao, Q Liu, R Gani, Chemical product design–recent advances and perspectives, Current Opinion in Chemical Engineering, 27, 22-34, 2020.
2. S.R. Syeda, E.A. Khan, O. Padungwatanaroj, N. Kuprasertwong, A.K. Tula, A perspective on hazardous chemical substitution in consumer products, Current Opinion in Chemical Engineering, 36, 100748, 2022.
3. S. R. Syeda, E. A. Khan, N. Kuprasertwong, O. Padungwatanaroj, R. Gani,A Model-Data Driven Chemical Analysis System for Products and Associated Processes, Proceedings of 14^th Symposium of Process Systems Engineering (PSE2021), Elsevier, June, 2022
4. A. S. Alshehri, A. K. Tula, F. You, R. Gani, Next Generation Pure Component Property Estimation Models: With and Without Machine Learning Techniques, AIChE J, e17469, (https://doi.org/10.1002/aic.17469)
5. Q. Liu, L. Zhang, L. Liu, J. Du, A.K. Tula, M. Eden, R. Gani, “OptCAMD: An optimization-based framework and tool for molecular and mixture product design”. Computers & Chemical Engineering, 124, 285-301, 2019.
6. K. Udomwong, A. Robin, N. Kuprasertwong, O. Padungwatanaroj, A.K. Tula, L. Zhu, L. Zhou, B. Wang, S. Wang, R. Gani, “ProREFD: Tool for Automated Computer-Aided Refrigerant Design, Analysis, and Verification”, Computer-Aided Chemical Engineering, 50, 457-462, 2021.