Developing and Validating a High-Throughput Robotic System for the Accelerated Development of Porous Membranes
Hongchen Wang, Sima Zeinali Danalou, Jiahao Zhu, Kenneth Sulimro, Chaewon Lim, Smita Basak, Aimée Tai, Usan Siriwardana, Jason Hattrick-Simpers, Jay Werber · 2025
Read Developing and Validating a High-Throughput Robotic System for the Accelerated Development of Porous Membranes by Hongchen Wang, Sima Zeinali Danalou, Jiahao Zhu, Kenneth Sulimro, Chaewon Lim, Smita Basak, Aimée Tai, Usan Siriwardana, Jason Hattrick-Simpers, Jay Werber (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-10973.txt.
Introduction
Hongchen Wang and co-authors built a robot line that makes and tests porous membranes in a 2025 paper. Membranes with fine pores serve in filters and other gear, but finding the best recipe takes slow hand work. The robot mixes the brew, spreads it thin, dips it in water to set the pores, and then tests each sheet. Settings such as mix strength and room dampness stay under tight control.
A squeeze test reads sheet stiffness, which hints at pore share and even build. In trial runs the robot copied known trends: stronger mixes gave stiffer, more even sheets. Damp air shifted pore shape and strength as past work would predict. The team pitch the line as a base for self-driving labs. Fast, repeat runs could feed data tools that hunt the best recipe on their own.
Catalog record
| CREATOR | Hongchen Wang, Sima Zeinali Danalou, Jiahao Zhu, Kenneth Sulimro, Chaewon Lim, Smita Basak, Aimée Tai, Usan Siriwardana, Jason Hattrick-Simpers, Jay Werber |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-10973
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Developing and Validating a High-Throughput Robotic System for the Accelerated Development of Porous Membranes Hongchen Wang, Sima Zeinali Danalou, Jiahao Zhu, Kenneth Sulimro, Chaewon Lim, Smita Basak, Aimée Tai, Usan Siriwardana, Jason Hattrick-Simpers, Jay Werber Abstract The development of porous polymeric membranes remains a labor-intensive process, often requiring extensive trial and error to identify optimal fabrication parameters. In this study, we present a fully automated platform for membrane fabrication and characterization via nonsolvent-induced phase separation (NIPS). The system integrates automated solution preparation, blade casting, controlled immersion, and compression testing, allowing precise control over fabrication parameters such as polymer concentration and ambient humidity. The modular design allows parallel processing and reproducible handling of samples, reducing experimental time and increasing consistency. Compression testing is introduced as a sensitive mechanical characterization method for estimating membrane stiffness and as a proxy to infer porosity and intra-sample uniformity through automated analysis of stress-strain curves. As a proof of concept to demonstrate the effectiveness of the system, NIPS was carried out with polysulfone, the green solvent PolarClean ® , and water as the polymer, solvent, and nonsolvent, respectively. Experiments conducted with the automated system reproduced expected effects of polymer concentration and ambient humidity on membrane properties, namely increased stiffness and uniformity with increasing polymer concentration and humidity variations in pore morphology and mechanical response. The developed automated platform supports high-throughput experimentation and is well-suited for integration into self-driving laboratory workflows, offering a scalable and reproducible foundation for data-driven optimization of porous polymeric membranes through NIPS. 1 Introduction Porous polymeric membranes are widely used in separation processes, including ultrafiltration (UF), gas separation, and as support layers for thin-film composite (TFC) membranes used in nanofiltration (NF) and reverse-osmosis (RO) [ 1 , 2 , 3 ] . Emerging high-pressure reverse osmosis (HPRO) processes extend the operation to pressures of 120+ bar, requiring support membranes that can withstand substantial compressive stresses [ 4 ] . As a result, developing porous membranes with robust and tunable mechanical properties is essential to ensure reliable performance in pressure-driven separation processes [ 4 ] . Currently, most commercial porous membranes are manufactured using nonsolvent-induced phase separation (NIPS) due to its processing simplicity, scalability, and cost-effectiveness [ 5 , 6 ] . In the NIPS process, polymer solution is cast onto a substrate and immersed in a nonsolvent bath to initiate the solvent-nonsolvent exchange and subsequent phase separation to form a solid membrane [ 6 , 7 ] . Despite its advantages, the process is highly sensitive to changes in composition, environmental conditions, and processing parameters, often resulting in substantial variability in membrane structure and performance [ 8 , 9 , 10 ] . The kinetics of this demixing process strongly influence the resulting membrane structure [ 8 ] . Specifically, rapid demixing typically forms finger-like macrovoids, which offer high flux but poor mechanical performance, while delayed demixing forms sponge-like structures, which are mechanically robust but less permeable [ 8 ] . Key fabrication parameters, including polymer concentration, solvent and nonsolvent types, temperature, humidity, and casting thickness, play critical roles in governing membrane morphology and performance [ 8 , 11 , 12 ] . While many studies have examined how these fabrication parameters influence membrane structure and performance [ 9 , 10 , 13 , 14 , 15 ] , optimizing these variables through traditional experimentation remains slow and labor-intensive. Manual workflows are often inconsistent and difficult to reproduce, particularly given the complex and sensitive dynamics of NIPS [ 8 , 16 ] . Furthermore, membrane characterization often involves multiple time-consuming steps, such as filtration testing and scanning electron microscopy (SEM) [ 17 , 18 ] . These limitations present a significant barrier to rapid iteration and systematic exploration of the NIPS fabrication space. To overcome challenges similar to those faced in NIPS, high-throughput experimentation (HTE) and self-driving laboratories (SDLs) have emerged as promising paradigms for accelerated materials development [ 19 , 20 ] . HTE enables rapid and systematic exploration of large combinatorial design spaces by automating the fabrication and/or characterization of samples under a wide range of conditions [ 19 ] . While we are not aware of a high-throughput system for NIPS, notable innovations in HTE for membrane science have been demonstrated in areas such as diafiltration, dead-end filtration, and membrane electrode assembly [ 21 , 22 , 23 ] . SDLs take automated experimentation one step further by integrating HTE with real-time data analysis and algorithmic decision-making to iteratively synthesize, test, and optimize material systems with minimal human intervention [ 20 , 24 , 25 ] . In other words, SDLs close the loop during automated materials research and development, which can dramatically accelerate materials development and optimization for design spaces with many dimensions where brute-force HTE would be onerous and inefficient. For example, Snapp et al. [ 26 ] demonstrated an SDL framework that autonomously performed over 25,000 compression tests on 3D-printed polymer structures, efficiently exploring an 11-dimensional design space via robotic experimentation and Bayesian optimization to identify structures with exceptional energy absorption properties. This approach highlights how integrating closed-loop automation with compression testing can yield rapid, data-rich insights into structure-property relationships, even in complex, high-dimensional materials systems. Interest in SDLs has increased markedly in recent years, with applications now developed in pharmaceuticals, catalysis, and materials discovery [ 27 , 28 , 29 ] . This manuscript presents our work towards a NIPS-based SDL for the accelerated development of porous membranes. In particular, we have developed an automated NIPS fabrication and testing platform (i.e., an HTE system for NIPS) that integrates liquid handling, robotic blade casting, controlled environmental conditions, and mechanical testing into a single automated system. By combining automated sample preparation with rapid mechanical compression testing, the system allows systematic exploration of the membrane design space under controlled and reproducible conditions. A key innovation of this platform is the use of compression testing as a high-throughput proxy for membrane performance. Drawing from previous research on porous mechanical systems [ 30 , 31 , 32 ] , we interpret the stress-strain response of membrane samples to extract two performance-relevant metrics: 1) elastic modulus as a measure of stiffness and 2) normalized plateau length as a proxy for porosity. As a proof of concept exploration using the automated NIPS fabrication platform, we investigate how polymer concentration and ambient humidity influence membrane morphology and mechanical performance, using polysulfone (PSf), the green solvent PolarClean ® , and water as the polymer, solvent, and nonsolvent, respectively. Multiple tests across each membrane demonstrate intra-sample consistency, with comparisons to manually fabricated membranes used to validate the membrane uniformity achieved through automation. The presented system is now capable of high-throughput synthesis and characterization of porous membranes through NIPS, achieving a throughput of less than an hour per sample from formulation to mechanical testing. Our future work will "close the loop" by integrating active learning algorithms and orchestrations to enable full SDL workflows. 2 Materials and Methods 2.1 Materials The chemicals used in this study are polysulfone (PSf) (Sigma-Aldrich, M_{w} - 35000) and PolarClean ® (Rhodiasolv PolarClean ® ). The commercially available tools used to build the automated platform are a liquid handling robot (Opentrons OT-2), a robotic arm (Ufactory xArm7), and a compression tester (TestResources 140).