Embedded Files
- AI and machine learning in environmental research
While artificial intelligence and machine learning offer powerful tools for environmental and chemical research, their impact is often limited by small datasets, missing values, and poor generalizability of models trained on heterogeneous literature data. These challenges are especially pronounced in water treatment and catalysis, where experiments are costly and external validation data are scarce. Our research develops machine-learning frameworks tailored for these constraints, combining robust data handling, interpretability, and reliability assessment to enable trustworthy predictions. We apply these methods to predict disinfection byproduct concentrations and toxicity, as well as catalytic activity and selectivity, with the goal of guiding mechanistic understanding and practical decision-making.
Representative publications:
Sikder, R.; Hua, G.; Ye, T. Toward Generalizable Machine Learning Models for Dichloroacetonitrile Formation: Interpretable Insights and A Framework for Model Reliability. Water Res. 2026, 289, 124823. (DOI)
Hossain, M. M.; Sikder, R.; Hua, G.; Ye, T. From Model Development to Mitigation: Machine Learning for Predicting and Minimizing Iodinated Trihalomethanes in Water Treatment. Environ. Sci. Technol. 2025, 59, 11638-11652. (DOI)
Supported by:
Institutional support and ongoing proposal development; we welcome collaborations and funding opportunities in data-driven environmental research.
2. Catalysis
2. Catalysis
The growing concern of impaired water sources by the presence of oxyanion contaminants that are only amenable to reduction (i.e., nitrate, perchlorate, and bromate) necessitates the development of advanced reduction technologies. Pd-based catalytic hydrogenation appears as a promising water treatment technology that has been demonstrated to effectively catalyze reductive transformation of oxyanions to nontoxic products. The challenges of Pd-based catalytic hydrogenation in water treatment are to lower the high catalyst cost, optimize activity, maximize selectivity toward desirable end-products. Therefore, our research aims to provide sustainable, low-cost, and potentially scalable methods to prepare Pd-based catalysts with superior catalytic performance toward oxyanion reduction.
Representative publications:
Ye, T.; Banek, N. A.; Durkin, D. P.; Hu, M.; Wang, X.; Wagner, M. J.; Shuai, D. Pd Nanoparticle Catalysts Supported on Nitrogen-Functionalized Activated Carbon for Oxyanion Hydrogenation and Water Purification. ACS Appl. Nano Mater. 2018, 1, 6580-6586. (DOI)
Durkin, D. P.+; Ye, T.+; Choi, J.; Livi, K. J.; De Long, H. C.; Trulove, P. C.; Fairbrother, D. H.; Haverhals, L. M.; Shuai, D. Sustainable and Scalable Natural Fiber Welded Palladium-Indium Catalysts for Nitrate Reduction. Appl. Catal., B. 2018, 221, 290-301. (+ Equal Contribution) (DOI)
Ye, T.; Durkin, D. P.; Banek, N. A.; Wagner, M. J.; Shuai, D. Graphitic Carbon Nitride Supported Ultrafine Pd and Pd-Cu Catalysts: Enhanced Reactivity, Selectivity, and Longevity for Nitrite and Nitrate Hydrogenation. ACS Appl. Mater. Interfaces. 2017, 9, 27421-27426. (DOI)
Supported by:
USGS 104b (FY2022, link).
South Dakota Mines Nelson grant.
NSF (2327715).
3. Disinfection chemistry
3. Disinfection chemistry
Drinking water disinfection has dramatically reduced the outbreaks of waterborne diseases (e.g., gastroenteritis, cholera and typhoid). However, reactions between dissolved organic matter in water and the widely used low-cost disinfectant, chlorine, result in the production of undesired disinfection by-products (DBPs) with potential health concerns. Trihalomethanes and haloacetic acids are the two most abundant DBPs formed during chlorination and has been regulated by the U.S. Environmental Protection Agency. To comply with these regulations, many water utilities have opted for alternative disinfection methods, such as chloramines, chlorine dioxide, and potassium permanganate. However, each of these disinfection methods produces its own suite of DBPs, and sometimes offers more problems than solutions.
Representative publications:
Liu, Z.; Ye, T.; Xu, B.; Zhang, T.-Y.; Li, M.-Y.; Hu, C.-Y.; Tang, Y.-L.; Zhou, X.-R.; Xian, Q.-M.; Gao, N.-Y. Formation and Control of Organic Chloramines and Disinfection By-products during the Degradation of Pyrimidines and Purines by UV/Chlorine Process in Water. Chemosphere 2022, 286, 131747. (DOI)
Ye, T.; Zhang, T.-Y.; Tian, F.-X.; Xu, B. The Fate and Transformation of Iodine Species in UV Irradiation and UV-based Advanced Oxidation Processes. Water Res. 2021, 206, 117755. (DOI)
Ye, T.; Xu, B.; Wang, Z.; Zhang, T.-Y.; Hu, C.-Y.; Lin, L.; Xia, S.-J.; Gao, N.-Y. Comparison of Iodinated Trihalomethanes Formation during Aqueous Chlro(am)ination of Different Iodinated X-Ray Contrast Media Compounds in the Presence of Natural Organic Matter. Water Res. 2014, 66, 390–398. (DOI)
Ye, T.; Xu, B.; Lin, Y.-L.; Hu, C.-Y.; Lin, L.; Zhang, T.-Y.; Gao, N.-Y. Formation of Iodinated Disinfection By-Products during Oxidation of Iodide-Containing Waters with Chlorine Dioxide. Water Res. 2013, 47, 3006–3014. (DOI)
Supported by:
4. Electrospinning
4. Electrospinning
Electrospinning is a fiber production method which uses electric force to draw charged threads of polymer solutions or polymer melts up to fiber diameters in the order of some hundred nanometers. Electrospun nanofibers have a broad range of applications, including membrane and filtration, catalyst supports, medical device.
Representative publications:
Ye, T.; Durkin, D. P.; Hu, M.; Wang, X.; Banek, N. A.; Wagner, M. J.; Shuai, D. Enhancement of Nitrite Reduction Kinetics on Electrospun Pd-Carbon Nanomaterial Catalysts for Water Purification. ACS Appl. Mater. Interfaces. 2016, 8, 17739-17744. (DOI)
Zhu, W.+; Ye, T.+; Lee, S.-J.; Cui, H.; Miao, S.; Zhou, X.; Shuai, D.; Zhang, L. G. Enhanced Neural Stem Cell Functions in Conductive Annealed Carbon Nanofibrous Scaffolds with Electrical Stimulation. Nanomedicine: NBM 2018, 14, 2485-2494. (+ Equal Contribution) (DOI)
Supported by:
We are actively seeking external funding to advance this research direction.
5. Data-driven environmental chemistry
5. Data-driven environmental chemistry
Water chemistry is complex. Though analytical technology (e.g., gas chromatography [GC] and liquid chromatography [LC] combined with mass spectrometry [MS]) are currently available, significant efforts are required to characterize water parameters, such as DOC, SUVA, and pH, and correlate them with water quality, such as the formation of disinfection byproducts. Machine learning is a useful and powerful tool that can be used to predict water quality and help to improve the operation of water treatment.
Representative publications:
Sikder, R.; Zhang, T.; Ye, T. Predicting THM Formation and Revealing Its Contributors in Drinking Water Treatment Using Machine Learning. ACS EST Water 2024, 4, 899-912. (DOI)
Sikder, R.; Zhang, H.; Gao, P.; Ye, T. Machine Learning Framework for Predicting Cytotoxicity and Identifying Toxicity Drivers of Disinfection Byproducts. J. Hazard. Mater. 2024, 469, 133989. (DOI)
Supported by:
Page updated
Google Sites
Report abuse