Honglin Liu
Advisor: Dr. Youjiang Wang, Co- Advisor: Dr. Karl I. Jacob
will propose a doctoral thesis entitled,
Prediction and Optimization of Graphene Aerogel for Membrane Distillation Using Experimental, Molecular Dynamics, and Machine Learning Approaches
On
Wednesday, June 25 at 9:30 a.m. (EDT)
Virtually via MS Teams
[Meeting link]
Meeting ID: 215 050 562 617 6
Passcode: vH3QQ2Vy
Committee
Dr. Youjiang Wang – School of Materials Science and Engineering (advisor)
Dr. Karl I. Jacob – School of Materials Science and Engineering, George W. Woodruff School of Mechanical Engineering (co-advisor)
Dr. Donggang Yao – School of Materials Science and Engineering
Dr. Hamid Garmestani – School of Materials Science and Engineering
Dr. S. Mostafa Ghiaasiaan – George W. Woodruff School of Mechanical Engineering
Abstract
The growing global freshwater scarcity has been driving the development of advanced desalination technologies, with membrane distillation (MD) recognized as a promising next-generation approach due to its ability to utilize solar or low-grade thermal (LoT) energy and its insensitivity to high-salinity water. This dissertation work adds to that growing literature in this area by investigating the performance prediction and optimization of graphene aerogel (GA)-based membranes for cost-effective membrane distillation through a systematic approach, integrating experimental, molecular simulations, and machine learning methodologies. This study encompasses three key components: (1) design and construction of a fully automated direct contact membrane distillation (DCMD) testing platform using 3D printing and Python scripting to evaluate the influence of operational parameters on permeate flux. The results highlight the significant impacts of flow rate, feed temperature, membrane material, and long-term inorganic scaling phenomena. (2) use of fully-atomistic molecular dynamics simulations to quantitatively elucidate the relationships among structural parameters (average length of graphene sheets and dummmy inclusions' distance to zero potential), graphene aerogel’s morphological characteristics (pore channel diameter, density, thickness, porosity, specific surface area, and tortuosity), and its heat and mass transfer properties (water molecule diffusivity, permeate flux, thermal conductivity, and localized phonon transport) in a vacuum membrane distillation (VMD) process. A predictive equation set with high accuracy and strong physical interpretability for transfer performance was also developed. (3) Development of a novel Transformer-enhanced 3D convolutional neural network (TRM-CNN) model, leveraging a scalable dataset generated from high-throughput molecular dynamics simulations, to predict graphene aerogel’s morphological and heat and mass transfer properties, achieving superior computational efficiency and predictive accuracy compared to conventional convolutional neural network models. The findings from this work collectively form a scalable and reliable end-to-end framework for predicting and optimizing graphene aerogel’s structural, mass, and heat transfer properties, providing actionable guidance for designing energy-efficient membrane distillation processes.