Machine Learning Identifies Best Polymers for Improving Performance of Film Capacitors

     Researchers at Lawrence Berkeley National Laboratory used machine learning to identify the best polymer materials to improve the performance of film capacitors in electronic devices. This could have implications for clean energy technologies. The method was used to screen 49,700 candidate chemical compounds, polysulfate polymers, to find the best material(s) that would improve performance. The qualities they were looking for were resistance to high temperatures and strong electric fields, high energy storage density, and the ability to synthesize easily. They found three polymers that were best suited. Resistance to heat is required in some applications. While ceramic capacitors, which are more heat resistant, can be used in some of those cases, film capacitors are much cheaper. The newly identified materials “exhibited a never-before-seen combination of heat resistance, insulating properties, energy density, and efficiency.”

 

Capacitor Types, Film Capacitors, and Comparisons

     Film capacitors are a type of non-polar fixed capacitors. The variety and types of capacitors are shown below. Film capacitors consist of an insulating material sandwiched between two conductive metal sheets. Batteries store and discharge energy over long time periods while capacitors utilize applied electric fields to charge and discharge energy much faster. According to Tech Xplore:

“Film capacitors are used for regulating power quality in diverse types of power systems. For example, they can prevent ripple currents and smooth voltage fluctuations, ensuring stable, safe, reliable operations.”

Source: Panasonic

 

     According to Panasonic:

“Nonpolar capacitors have no restrictions on the polarity of the voltage applied to their terminals. In other words, either terminal can be positive. Nonpolar capacitors can be directly used in AC circuits because a voltage that rises or falls from a zero potential can be applied.”

 

“Film capacitors, which use a plastic film as their dielectric, have the following features.

·        Non polar

·        Excellent high frequency characteristics (low ESR)

·        Excellent temperature characteristics (small rate of change in capacitance due to temperature)

·        Compatible with highly accurate capacitance

·        Long life

 

Source: Panasonic
    

     As explained below by EE Power there are many different types of film capacitors using different dielectric materials suited to different applications and price ranges.

“There are many types of film capacitors, including polyester film, metallized film, polypropylene film, PTFE film and polystyrene film. The core difference between these capacitor types is the material used as the dielectric, and the proper dielectric must be chosen according to the application.”

“PTFE film capacitors, for example, are heat-resistant and used in aerospace and military technology, while metallized polyester film capacitors are used in applications that require long term stability at a relatively low. Cheaper plastics are used if cost is a bigger concern than performance.”

“Power film capacitors are used in power electronics devices, phase shifters, X-ray flashes and pulsed lasers, while the low power variants are used as decoupling capacitors, filters and in A/D converters. Other notable applications are safety capacitors, electromagnetic interference suppression, fluorescent light ballasts and snubber capacitors.”

 

Source: Panasonic (part of table)

Source: Panasonic (part of flow chart)

Machine Learning Identifies Best Candidate Polymers: Assembly and Testing Verifies Qualities

     The research at Berkeley Labs used machine-learning models known as feedforward neural networks to find the polymers with the desired qualities. After the three best polymers were identified, researchers at Scripps Research Institute utilized the method of click chemistry to synthesize the polymers. They were then assembled into film capacitors and tested at Berkeley’s Molecular Foundry. Tech Xplore explains the results:

“Capacitors made from one of the polymers exhibited a record-high combination of heat resistance, insulating properties, energy density, and efficiency. (A high-efficiency capacitor wastes very little energy when it charges and discharges.) Additional tests on these capacitors revealed their superior material quality, operational stability, and durability.”

     The abstract of the paper in Nature Energy, shown below, compares the machine learning technique for identifying the best polymers to the ‘intuition-driven polymer design’ commonly used.

 Abstract

“The development of heat-resistant dielectric polymers that withstand intense electric fields at high temperatures is critical for electrification. Balancing thermal stability and electrical insulation, however, is exceptionally challenging as these properties are often inversely correlated. A traditional intuition-driven polymer design approach results in a slow discovery loop that limits breakthroughs. Here we present a machine learning-driven strategy to rapidly identify high-performance, heat-resistant polymers. A trustworthy feed-forward neural network is trained to predict key proxy parameters and down select polymer candidates from a library of nearly 50,000 polysulfates. The highly efficient and modular sulfur fluoride exchange click chemistry enables successful synthesis and validation of selected candidates. A polysulfate featuring a 9,9-di(naphthalene)-fluorene repeat unit exhibits excellent thermal resilience and achieves ultrahigh discharged energy density with over 90% efficiency at 200 °C. Its exceptional cycling stability underscores its promise for applications in demanding electrified environments.”

 

Future Implications of Machine Learning Applied to Polymers Research

     This successful application of machine learning to speed up materials selection could have future implications for optimizing performance via these improvements to heat resistance, strong electric field resistance, higher energy density, and ease of synthesis. Beyond capacitor research it would seem the techniques, machine learning followed by click chemistry, could be used to engineer improved materials with desired properties in other applications. Below from Tech Xplore, the team from Berkeley and their collaborators noted where follow-up polymer research should be focused:

"One idea is to design machine learning models that provide more insights into how the structure of polymers influences their performance," said Zongliang Xie, a postdoctoral researcher at Berkeley Lab.

"Another potential research area is to develop generative AI models that can be trained to design high-performance polymers without having to screen a library," added Tianle Yue, a graduate student at the University of Wisconsin–Madison.

"Our AI analysis quickly identified some key variables in the polymer design details that were predicted to add big improvements in the shielding properties of these polysulfate membranes. As reported in our new Nature Energy study, these earliest machine learning predictors for improving the capacitors are dramatically born-out by experiment," said Sharpless, W.M. Keck Professor of Chemistry at Scripps Research.

 

 

References:

 

Machine-learning models help discover a material for film capacitors with record-breaking performance. Michael Matz, Lawrence Berkeley National Laboratory. Tech Xplore. December 5, 2024. Machine-learning models help discover a material for film capacitors with record-breaking performance

Researchers inch closer to perfecting a futuristic energy tech — here's how they achieved this breakthrough method. Stephen Proctor. The Cool Down. January 21, 2025. Researchers inch closer to perfecting a futuristic energy tech — here's how they achieved this breakthrough method

Film Capacitor. Chapter 2 - Capacitor Types. EE Power. Film Capacitor | Capacitor Types | Capacitor Guide

Basic Knowledge of Capacitors(2)-Types, Characteristics, Applications. Panasonic Industry. June 26, 2018. Basic Knowledge of Capacitors(2) - Panasonic

Machine learning-accelerated discovery of heat-resistant polysulfates for electrostatic energy storage. He Li, Hongbo Zheng, Tianle Yue, Zongliang Xie, ShaoPeng Yu, Ji Zhou, Topprasad Kapri, Yunfei Wang, Zhiqiang Cao, Haoyu Zhao, Aidar Kemelbay, Jinlong He, Ge Zhang, Priscilla F. Pieters, Eric A. Dailing, John R. Cappiello, Miquel Salmeron, Xiaodan Gu, Ting Xu, Peng Wu, Ying Li, K. Barry Sharpless & Yi Liu. Nature Energy volume 10, pages90–100 (2025). Machine learning-accelerated discovery of heat-resistant polysulfates for electrostatic energy storage | Nature Energy