NEIL BURNS

Managing Partner Neil A Burns LLC

Substantial investment and progress has already been made in the direct application of AI in businesses using surfactants. Unilever has been particularly active across the whole value chain, as the following accounts, published the company, show.

• Unilever is using AI to optimise its portfolio of brands and products, including surfactants, by using a data-driven tool that recommends which Stock Keeping Unit (SKUs) to delist, watch, protect or grow, based on the benefit to customers, consumers and Unilever (1). This helps to reduce waste, improve productivity and focus on core and innovative SKUs that meet consumer needs and preferences.
• Unilever is using AI to innovate by discovering new ingredients and new science, such as plant-based, sustainable palm oil alternatives for surfactants, by designing and simulating products and processes virtually, such as the world's first green carbon detergent, and by deploying products more efficiently and effectively, such as using digital twins and smart sensors to monitor and optimise production (2).
• Hindustan Lever in India uses AI to improve its supply chain resilience and agility, by using data analytics and machine learning to forecast demand, optimise inventory, reduce costs and enhance customer service (3). This helps to ensure the availability and quality of surfactants and other products, especially during times of disruption and uncertainty.
• Unilever in the UK is using AI to enhance its talent acquisition and retention, by using an artificial intelligence system to assess job interviews, reducing the time and cost of hiring and improving the diversity and fit of candidates (4).
• The company is using AI to improve its products and consumer experience, by using molecular models to improve its product formulas, faster and more effectively than in a physical lab (5), by using data to identify and predict consumer trends and preferences, such as using voice assistants and chatbots to interact with consumers. This helps to create value and impact for society and the environment, as well as for Unilever and its customers.

In the area of surfactant synthesis, there is surprisingly little fully documented AI work. However, there is plenty of work in the field of chemical synthesis in general. Professor Alan Aspuru-Guzik of the University of Toronto co-authored a paper, published at the end of October in the journal Autonomous Robots (7). In summary: The paper proposes an approach to automate chemistry experiments using robots by translating natural language instructions into robot-executable plans, using large language models together with task and motion planning. So you can tell your chemist robot - “make me a methyl ester of that acid” and off it goes, no further instruction needed. Adding natural language interfaces to autonomous chemistry experiment systems lowers the barrier to using complicated robotics systems and increases utility for non-expert users. However, translating natural language experiment descriptions from users into low-level robotics languages is not that easy. Furthermore, while recent advances have used large language models to generate task plans, reliably executing those plans in the real world ie. by a robot in a lab, remains challenging. To enable autonomous chemistry experiments and alleviate the workload of chemists, robots must interpret natural language commands, perceive the workspace, autonomously plan multi-step actions and motions, consider safety precautions, and interact with various laboratory equipment. The approach in the paper, CLAIRify, combines automatic iterative prompting with program verification to ensure syntactically valid programs in a data-scarce domain-specific language that incorporates environmental constraints. The generated plan is executed through solving a constrained task and motion planning problem using PDDLStream solvers to prevent spillages of liquids as well as collisions in chemistry labs (kind of a minimum requirement before even making any product). The effectiveness of the approach is demonstrated in planning chemistry experiments, with plans successfully executed on a real robot using a repertoire of robot skills and lab tools. This demo includes fundamental chemical experiments for materials synthesis: solubility and recrystallization. Further details about CLAIRify can be found at (8). It goes without saying that the methodologies and tools referenced here could have a big impact on surfactant synthesis work.

In the area of process technology for surfactants, there is similarly little fully documented work specifically concerning surfactants. However, applications of AI in the field of chemical engineering are widely reported and have obvious implications for surfactant process technology. A good example is a joint project between MIT and the University of Ghent which was written up in the journal, Frontiers in Chemical Engineering (9). The paper, entitled Artificial Intelligence for Computer Aided Synthesis in Flow, takes a machine learning approach to figuring out whether and how a particular reaction may benefit from being done in a flow system. The surfactant industry has a history of certain reactions being transitioned from batch to continuous flow, particularly sulfonation and sulfation used in the production of Linear Alkylbenzene Sulfonates, Alcohol Sulfates, Alcohol Ether Sulfates and Alpha Olefin Sulfonates. This MIT/Ghent project therefore could be of particular interest to companies considering flow chemistry for other surfactant-relevant reactions, such as alkoxylation.

As the paper explains, running a process in a continuous fashion introduces many additional considerations for process design. Reactions in which some reagents are solids or those in which the equilibrium is driven in a certain direction via precipitation of the products are difficult to develop in flow as the solids can clog the channels. Improper combinations of solvents and reactants, products, or catalysts can also result in precipitation of certain compounds, again increasing the risk of clogging. Reactions requiring long residence times may require impractically long reactors or inefficiently low flow rates. At very low flow rates, mixing becomes an important issue, implying that the use of a stirred spheroidal reactor is much more economical. As a result, it is very difficult to anticipate whether a synthesis can be performed continuously in an economic way. The MIT/Ghent project used a data-driven method, based on a statistical analysis of published reactions, that can identify such continuous syntheses and can potentially guide retrosynthetic software toward such syntheses.