實驗室核心理念與研究願景
Core Philosophy & Research Vision
本實驗室致力於將生成式 AI、高分子薄膜實驗、分子模擬結合機器學習力場三大技術主軸應用於化學工程。
Our laboratory is dedicated to applying three major technological pillars in chemical engineering:
1. Generative AI
2. Polymer Membrane Experimentation
3. Molecular Simulation with Machine Learning Force Fields
核心研究方向｜Research Directions
主軸一：生成式 AI 重塑化工新境界
Theme 1: Generative AI Redefining the Chemical Industry
我們將 AI 技術與化工系統工程深度結合，整合深度學習（Deep Learning）、生成式 AI（Generative AI）與智能代理人系統（Agentic AI），以加速化工製程的數位轉型與智慧製造。
We deeply integrate AI technologies with chemical systems engineering by combining Deep Learning, Generative AI, and Agentic AI systems to accelerate the digital transformation and smart manufacturing of chemical processes.
Real-Time Optimization and Dynamic Control of Chemical Processes
運用物理資訊神經網路（PINN）與數據驅動模型，實現反應器與蒸餾塔等關鍵單元的動態先進控制（APC）與節能減碳。
Utilizing Physics-Informed Neural Networks (PINN) and data-driven models to achieve advanced process control (APC) and energy-saving carbon reduction for key units such as reactors and distillation columns.
Domain-Specific Knowledge Base Construction for Chemical Engineering
結合檢索增強生成（RAG）與大語言模型（LLMs），建構專屬化工法規、操作 SOP 與文獻數據的智慧決策輔助平台。
Combining Retrieval-Augmented Generation (RAG) with Large Language Models (LLMs) to build an intelligent decision-support platform dedicated to chemical regulations, operational SOPs, and literature data.
Autonomous Flowsheet Design
打造能自主呼叫化工模擬軟體的 AI Agents，實現化學製程流程圖自動建立、參數優化與驗證。
Developing AI Agents capable of autonomously calling chemical process simulation software to enable automatic flowsheet generation, parameter optimization, and verification.
主軸二：薄膜創新淨化地球
Theme 2: Membrane Innovation for a Cleaner Planet
以高分子薄膜材料為核心，專注於能源與環保相關應用。
Focusing on polymer membrane materials as the core technology for energy and environmental applications.
Plasticized Solid Polymer Electrolyte for Lithium Batteries
開發兼具高安全性與優異離子導電率之增塑型固態高分子電解質。
Developing plasticized solid polymer electrolytes that combine high safety with excellent ionic conductivity.
Membrane Separation Processes
調控薄膜雙連續相結構與孔隙特性，應用於水資源回收、精密物質分離與環境工程。
Controlling the bicontinuous phase structure and pore characteristics of membranes for applications in water resource recovery, precision material separation, and environmental engineering.
Electromechanical Energy Conversion and Self-Powered Smart Sensing
開發壓電/摩擦奈米發電（TENG/PENG）功能薄膜，實現軟體機器人用自供電智慧電子皮膚。
Developing piezoelectric/triboelectric nanogenerator (TENG/PENG) functional membranes to enable self-powered smart electronic skin for soft robotics.
Radiative Cooling
設計光學微奈米光譜調控薄膜，達成全天候零耗能被動散熱。
Designing optical micro-nano spectral control membranes to achieve all-day, zero-energy passive heat dissipation.
Photocatalytic Hydrogen Generation Technology
開發高分子複合光催化材料系統，提升太陽能轉換效率與綠色潔淨氫能生產。
Developing polymer composite photocatalytic material systems to enhance solar energy conversion efficiency and the production of clean green hydrogen.
主軸三：分子模擬結合機器學習力場機制探討
Theme 3: Molecular Simulation with ML Force Fields for Mechanism Insights
我們結合分子模擬（Molecular Simulation）與機器學習力場（Machine Learning Force Fields），建立高精度且高效能的計算平台，深入探索高分子薄膜的結構、性質與傳輸機制，並應用於增塑型固態電解質、自供電機電轉換、被動式輻射冷卻、直接空氣捕捉二氧化碳分離（DAC）及光催化產氫等領域，加速新材料的理性設計與開發。
We integrate molecular simulation with machine learning force fields to establish a high-accuracy and high-efficiency computational platform. This enables in-depth investigation of the structure, properties, and transport mechanisms of polymer membranes, and is applied to plasticized solid electrolytes, self-powered electromechanical conversion, passive radiative cooling, direct air capture (DAC) of carbon dioxide, and photocatalytic hydrogen generation, thereby accelerating the rational design and development of new materials.
Development and Validation of Machine Learning Force Fields
建構適用於高分子系統的機器學習力場，提升分子動力學模擬的準確度與計算效率。
Constructing machine learning force fields tailored for polymer systems to enhance both the accuracy and computational efficiency of molecular dynamics simulations.
Structural and Transport Mechanism Analysis of Polymer Membranes
透過分子層級模擬，揭示薄膜微觀結構、自由體積分布與氣體／離子傳輸行為之間的關聯，並應用於固態電解質離子傳導、二氧化碳分離及光催化反應機制研究。
Revealing the relationships between membrane microstructure, free volume distribution, and gas/ion transport behavior through molecular-level simulations, with applications to ion conduction in solid electrolytes, carbon dioxide separation, and photocatalytic reaction mechanisms.
High-Throughput Virtual Screening and Multifunctional Materials Design
結合機器學習與分子模擬，進行高通量虛擬篩選，加速高性能增塑型固態高分子電解質、TENG/PENG 自供電薄膜、輻射冷卻薄膜、DAC 分離膜及光催化產氫材料的發現與優化。
Combining machine learning with molecular simulation for high-throughput virtual screening to accelerate the discovery and optimization of high-performance plasticized solid polymer electrolytes, TENG/PENG self-powered membranes, radiative cooling membranes, DAC separation membranes, and photocatalytic hydrogen generation materials.
Demo
1. 人工智慧 AI
- 大語言模型應用於化工程序設計與控制 Large Language Models in chemical process design and control
2. 高分子薄膜實驗 Polymer Membrane Experiments
- 鋰電池固態高分子電解質、輻射冷卻材料與分離程序 Lithium metal batteries solid-state polymer electrolytes, radiative cooling materials, and separation processes
3. 分子模擬 Molecular Simulation
- 高分子物性計算，並結合機器學習力場加速模擬 Computing polymer properties and accelerating simulations using machine-learning interatomic potentials