![\"\"](\"http:\/\/ulprospector.ul.com\/wp-content\/uploads\/2024\/04\/Nylon66_300.jpg\")
<\/p>\n<\/p>\n
In Part 2 of this two-part series<\/a>, we look at the much more complex and advanced – and therefore challenging – potential of Artificial Intelligence (AI). These take general automation opportunities a stage further, or get into the fundamentals of designing new materials, fine-tuning existing grades and predicting their properties, often with an eye on their biodegradable properties and sustainability.<\/p>\n More complex aspects of the production process are also to the fore, with the development of deep learning models to produce consistent product grades from complex, non-linear polymerization reaction dynamics. This area is sometimes referred to as Polymer Infomatics.<\/p>\n What is Polymer Informatics?<\/strong><\/p>\n Traditional design strategies for synthetic polymers and organic molecules have been empirical, guided by experience and intuition, and driven by application requirements. However, with the growing demand for new materials and the vast number of existing organic molecules, these methods face significant challenges. Polymer synthesis is costly and labour-intensive, making it important to minimize the number of experiments needed.<\/p>\n Due to the vast macromolecular structural variety of polymers, new approaches are needed to identify and develop novel applications. The emerging field of polymer informatics addresses this challenge by replacing traditional trial-and-error with analysis of historical and real-time sensor data.<\/p>\n Artificial intelligence (AI) and machine learning (ML) identify patterns and relationships between structure, processing, and final properties. Digital twins (virtual models of a real situation) can be created to simulate material behaviour before any physical synthesis is commenced. Artificial Intelligence (AI) has shifted from a research curiosity to a core driver of efficiency and innovation in polymer manufacturing.<\/p>\n With polymer informatics, a large pool of chemically or synthetically feasible polymers can be screened for potential candidates by applying predictive models (algorithms) relevant to the desired material properties. The result is faster R&D cycles and quicker time-to-market for innovative and high-performance, custom and sustainable materials. It helps create new polymers, fine-tune formulations, reduce waste and customize products efficiently.<\/p>\n Interestingly, a related approach is being used in metallurgy. Here, there is a vast landscape of metallurgical literature \u2013 papers, patents, scientific reports, small databases – a treasure trove of (unstructured) data. Work at the University of Sheffield in the UK is using ChatGPT to extract and interpret relevant information from the literature and determine the handful of alloys which are worth pursuing commercially from hundreds of thousands of options…<\/p>\n Key Applications of AI in Polymer Manufacture<\/strong><\/p>\n Finding the polymers of the future<\/strong><\/p>\n Amongst the companies and organisations active in this field are Citrine Informatics, ResolveMass, Georgia Institute of Technology, Imubit, NRL and Matmerize.<\/p>\n Georgia Institute of Technology<\/em><\/p>\n Polymer Genome is a data-driven ML-based online tool that can rapidly predict various polymer properties using models trained on polymer databases, experimental data, or first-principles computations. A group in the School of Materials Science led by Prof Rampi Ramprasad is developing and adapting AI algorithms to accelerate materials discovery.<\/p>\n A paper featured in Nature Reviews Materials showcases recent breakthroughs in polymer design across critical and contemporary application domains: energy storage, filtration technologies, and recyclable plastics. A second paper, published in Nature Communications, focuses on the use of AI algorithms to discover a sub-class of polymers for electrostatic energy storage, with the designed materials undergoing successful laboratory synthesis and testing.<\/p>\n \u201cIn the early days of AI in materials science, propelled by the White House\u2019s Materials Genome Initiative over a decade ago, research in this field was largely curiosity-driven,\u201d says Ramprasad. \u201cOnly in recent years have we begun to see tangible, real-world success stories in AI-driven accelerated polymer discovery. These successes are now inspiring significant transformations in the industrial materials R&D landscape.\u201d<\/p>\n Ramprasad\u2019s team has developed groundbreaking algorithms that can instantly predict polymer properties and formulations before they are physically created. The process begins by defining application-specific target property or performance criteria. Machine learning (ML) models train on existing material-property data to predict these desired outcomes. Additionally, the team can generate new polymers, whose properties are forecasted with ML models. The top candidates that meet the target property criteria are then selected for real-world validation through laboratory synthesis and testing. The results from these new experiments are integrated with the original data, further refining the predictive models in a continuous, iterative process.<\/p>\n According to Ramprasad, while AI can accelerate the discovery of new polymers, it also presents unique challenges. The accuracy of AI predictions depends on the availability of rich, diverse, extensive initial data sets, making quality data paramount. Additionally, designing algorithms capable of generating chemically realistic and synthesizable polymers is a complex task.<\/p>\n Matmerize<\/em><\/p>\n Matmerize, also Georgia-based, partners with R&D teams to define polymer and formulation challenges, set goals, and execute transparent, results-focused roadmaps.<\/p>\n Seamless integration of the Matmerize PolymRize platform into existing R&D activities allows AI to integrate with an existing workflow, accelerating digital transformation, boosting efficiency, and unlocking greater value from R&D intelligence.<\/p>\n <\/p>\n Additionally, Matmerize has introduced AskPOLY, a natural language based polymer “expert”.<\/p>\n Matmerize has announced a partnership with the semiconductor equipment manufacturer Screen Holdings focused on discovering PFAS-free polymers with high chemical resistance for semiconductor equipment applications.<\/p>\n The collaboration aims to enhance the safety and sustainability of semiconductor equipment manufacturing through Matmerize’s PolymRize platform, which supports key milestones in polymer selection and testing. Matmerize has developed AI models to predict polymer chemical resistance based on a custom dataset curated by Screen Holdings. Utilizing its proprietary database of over 10,000 commercially available sustainable polymers, Matmerize has recommended promising candidates for lab testing, with the goal of identifying new polymer candidates with high chemical resistance and suitability as PFAS-free alternatives.<\/p>\n Citrine Informatics<\/em><\/p>\n Citrine Informatics is headquartered in the Silicon Valley and has a global support team in place. The company provides a cloud-based software platform that enables material scientists, researchers and chemists with to access cutting edge AI and ML with a minimum learning curve. Users do not need to be a data scientist or a software engineer to use it effectively.<\/p>\n The AI is tailored to learn from small and sparse data sets typical of the chemicals and materials world. There is a graphical user interface to teach these AI models with information which is not only captured in the training data but also in the brains of the scientists.<\/p>\n ResolveMass<\/em><\/p>\n Custom polymer synthesis has become a cornerstone for advancing materials in industries such as healthcare, electronics, and sustainable packaging. Artificial Intelligence (AI) is emerging as a game-changer, transforming the way researchers and manufacturers design, develop, and produce polymers. By leveraging AI, the field of custom polymer synthesis is experiencing enhanced precision, speed, and scalability.<\/p>\n Canadian company ResolveMass Laboratories is at the forefront of leveraging artificial intelligence (AI) to revolutionize custom polymer synthesis. By integrating advanced AI algorithms and machine learning models in polymer chemistry, ResolveMass accelerates the design and optimization of polymers with tailored properties for specific applications such as drug delivery. AI enables the prediction of polymer behaviour, optimization of synthesis pathways, and identification of novel monomers, significantly reducing development time and costs.<\/p>\n Imubit <\/em><\/p>\n Imubit is primarily based in Houston, Texas, USA, where its headquarters is located. It also has a significant R&D presence in Israel and a new R&D centre in Romania.<\/p>\n Starting with costly catalyst and raw materials, identifying and running at the optimum reactor temperature to maximize conversion for the current fouling level is a 7-figure task. AI is paving a flexible path to grow margins by shifting yields, reducing variability, and improving quality.<\/p>\n Imubit’s AI Optimization (AIO) models learn these complex, dynamic relationships and put them to work in closed loop, extracting untapped value to maximize yield by 1-3% across various product grades while respecting unit constraints and reducing natural gas usage by 15-30%.<\/p>\n NRL (formerly NREL)<\/em><\/p>\n The National Renewable Energy Laboratory (NREL) is a long-established national laboratory of the US Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy LLC.<\/p>\n Under the Trump administration, the Energy Department has renamed it the National Laboratory of the Rockies (NRL), reflecting\u00a0 wider energy brief.<\/p>\n Developed with funding from the Bioenergy Technologies Office, Office of Energy Efficiency and Renewable Energy, US Department of Energy. PolyID (standing for Polymer Inverse Design) helps screen millions of potential eco-friendly polymer designs to identify sustainable biodegradable alternatives to established polymer grades.<\/p>\n PolyID predicts thermal, transport, and mechanical properties for six different types of polymers based on monomer structure. To use, the user enters a SMILES string below (or uses the drawing tool) and presses “Polymerize”. To enter multiple monomers, they click the “New” button in the drawing tool or place a period between the monomer SMILES strings.<\/p>\n References<\/strong><\/p>\n Polymer Informatics: Current and Future Developments Matmerize PolymRize The New Standard For Accelerated and Cost-effective Development of Polymers & Formulations Citrine Informatics ResolveMass Prove the Value of AI Optimization (AIO) at Your Plant at No Cost PolyID – Polymer Inverse Design – Machine learning predictions of polymer properties
\nULTRUS Collection solve customer problems across product stewardship, ESG, renewable energy, learning and workplace safety. Learn more here<\/a>!<\/h2>\n
\n\n
\nhttps:\/\/www.azom.com\/article.aspx?ArticleID=20730<\/p>\n\n
\nJ. Appl. Phys., 128, 171104 (2020).
\nhttps:\/\/doi.org\/10.1063\/5.0023759<\/li>\n<\/ol>\n
\nhttps:\/\/www.matmerize.com\/<\/a><\/p>\n
\nApplying best-in-class AI to accelerate innovation in materials and chemistry
\nhttps:\/\/citrine.io\/\u00a0<\/a><\/p>\n
\nResolving Complexity in Pharma R&D with Custom Polymer Synthesis, and Mass Spectrometry Analytical Services
\nhttps:\/\/resolvemass.ca\/<\/a><\/p>\n
\nhttps:\/\/insight.imubit.com\/prove-the-value-of-ai-optimization<\/a><\/p>\n
\nhttps:\/\/polyid.nrel.gov\/#\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"