“The result makes no sense according to any of the existing theories.”
Tan Zhang, assistant professor of Materials Science at Duke Kunshan University, and his student researchers had spent the summer testing several hypotheses about how to catalyze the synthesis of copper nanocrystals.
The goal was to find a more energy-efficient and safer way to produce the nanocrystals by making the chemical reaction easier to carry out.
The experiment was designed to establish a baseline for how readily different copper precursors could be reduced, or converted into metallic copper. The team had planned to use sodium nitrate as the electrolyte, but the lab had none in stock. So they used sodium chloride, the compound in table salt, instead.
Zhang said the team expected it to be “just a rehearsal to get familiar with the procedure.”
But the results were unexpected. In the sodium chloride solution, the measurements showed that all of the copper precursors were much easier to reduce.
At first, the team thought it might be “a glitch or an error.”
A week later, sodium nitrate arrived. They tested the same precursors again under the same conditions. This time, the results looked normal and matched their previous observations.
That made the earlier sodium chloride result harder to dismiss. Zhang asked Chung-Li Lin, then an undergraduate researcher in the Class of 2026 at DKU, to repeat the sodium chloride experiment. The unusual result appeared again.
That was when Zhang began to believe they had found what he called the “sweet spot.”
The unexpected result opened a new question: Why was chloride making copper easier to form?
The team designed a series of experiments and calculations to find out. They added chloride ions to systems using other copper salts and found that metallic copper could again be produced. Electrochemical measurements also showed that adding chloride made the copper ions significantly easier to reduce.
The evidence led the team to ask whether chloride could lower the energy required for the reaction itself.
They began lowering the reaction temperature. Copper nanocrystals still formed at 90 degrees Celsius. Then they formed at 80 degrees Celsius.
“For me, seeing copper nanocrystals form below the boiling point of water was the moment when the catalytic role of chloride became especially convincing,” Lin said. “It connected the mechanism we had been developing with a very tangible improvement in the synthesis process: making the reaction both more energy-efficient and safer.”
Together, the results supported the team’s explanation that chloride was catalyzing the synthesis of copper nanocrystals. Using sodium chloride, the team lowered the reaction temperature from 110 degrees Celsius to 80 degrees Celsius, reducing energy use and making the process safer. The grant report describes chloride as forming a stable intermediate complex and catalyzing the chemical reduction reaction.
Lin conducted the experiments, wrote the initial draft of the resulting paper. Published in Inorganic Chemistry, a leading peer reviewed journal covering fundamental research in inorganic chemistry, the paper was co-authored by Lin and Zhang and later highlighted on the journal’s cover.
The two had worked on the project for several years, with the summer grant supporting one phase of the research.
Zhang’s project was one of 14 supported through DKU’s Summer Research Program Grant in the 2025-26 academic year. The program provided nearly 1.25 million yuan in support of work across the arts and humanities, natural and applied sciences, and social sciences.
The projects ranged from fundamental questions to practical applications, and their methods often crossed disciplinary lines. Art history research drew on scientific imaging, economic history used artificial intelligence to process genealogical records, and applied mathematics intersected with biomedical research.
Student involvement was another common thread. Undergraduates ran experiments, built datasets, interpreted evidence and developed models. Some became coauthors or carried the work into Signature Work projects.
Traces left by craftspeople
In the storeroom of Duke University’s Nasher Museum of Art, Kent Cao, assistant professor of Art and Archaeology at DKU, identified a red-lacquer bowl dating to the late Ming to early Qing period.
Its surface preserves dozens of alternating lacquer layers, with thick red layers separated by thin black ones. Cao found that the black layers served as depth guides, helping the carver judge how far to cut without reaching the body of the bowl.
The black layers formed part of a deliberate carving strategy, preserving evidence of how the craftsperson worked. Cao’s broader project examines what material traces can reveal about craftspeople who are largely absent from written records. The report describes the black layers as a visual depth guide that helped prevent cuts into the bowl body.
At the center of that project are incidental marks: doodles, graffiti, informal inscriptions and casual figurative scratchings left by artisans in the course of their work. The craftspeople who produced them, Cao wrote in his project report, leave “virtually no trace” in the textual and official historical record.
Cao wrote that such marks offer “a rare and direct window” into the informal cognitive and creative lives of the people who built, decorated and maintained some of ancient China’s most celebrated monuments and objects.
To study them, he combines traditional art history research with scientific analysis. X-ray fluorescence provides information about pigments and material composition, while X-ray radiography can reveal marks and features hidden beneath a surface.
During the grant period, Cao conducted primary research during seven institutional visits across China, Japan and North America. DKU students joined parts of the research, helping with museum work, assisting with X-ray fluorescence instrument operation and contributing to the organization and preliminary interpretation of scientific data.
Cao is preparing a journal manuscript, and early findings are already incorporated into his HIST 301 Arts of China course at DKU.
Turning genealogies into data
Peiyuan Li faces a different problem of historical evidence: there is too much of it to process by hand.
Chinese genealogies, or jiapu, preserve information about families and relationships across generations. Li, assistant professor of political economy, is developing a system that combines optical character recognition with fine-tuned large language models to turn those historical records into structured data.
His team has collected 20,000 genealogy books and is acquiring another 25,000, bringing the projected corpus to approximately 45,000 volumes. An initial set of 100 books has been processed to create training and validation data for further development of the system.
But turning those records into reliable data still depends heavily on human judgment.
Four DKU undergraduates helped design annotation templates for the genealogies’ varied page layouts, manually corrected optical character recognition output and checked AI-extracted information including names, dates, kinship relationships and indicators of rank or status. They also joined weekly reviews to identify recurring errors and refine the extraction process.
That work is already generating new research. Yujia Zhou, Class of 2026, used the dataset as the empirical foundation for her Signature Work project. Kuang Sheng, also Class of 2026, and Li are developing a paper using the data to examine gender in long-run intergenerational transmission. Part of the extracted data is already being incorporated into a manuscript being revised and resubmitted to the Journal of Economic History, one of the top journals in the field.
Li wrote that DKU’s summer research grant had been “instrumental in launching a project that would otherwise have required years of fragmented academic-term work.” The project’s data and research infrastructure are expected to support further work on kinship, mobility and elite reproduction in late-imperial China.
Taking ownership of the question
Undergraduate Xiaoming Shi, Class of 2026, played a central role in a project led by Shixin Xu, associate professor of Mathematics, on noninvasive glucose monitoring. The research examined how glucose enters sweat and how the movement of ions and water shapes the resulting glucose measurements.
The project developed both a theoretical framework for transport across semi-permeable biological interfaces and a more detailed mathematical model of sweat secretion. It also developed a preliminary workflow for sweat collection and glucose monitoring.
Shi participated in literature review, model formulation, coding, numerical simulations and interpretation of results. He worked with Xu to develop a multicompartment model describing how ions, water and glucose move through different parts of the sweat gland. A manuscript based on the model is being prepared for submission.
Then Shi took the work in a new direction.
Drawing on that experience, Shi proposed SciDynBench-R, a benchmark for evaluating models that predict how scientific systems change over time. Rather than judging a model only by numerical error, the benchmark asks whether it captures scientifically meaningful behavior, remains robust when conditions change and produces interpretable predictions.
Undergraduates also took leading roles in Professor Zuchuan Li’s research on ocean biogeochemistry. DKU undergraduate Qianxun Xu, Class of 2026, was lead author of work presented at the 2025 Conference on Neural Information Processing Systems on a partial physics-informed diffusion model for reconstructing ocean chlorophyll concentrations. The model incorporates known physical principles into the reconstruction process. Another undergraduate, Yiwen Hu, was lead author of a study using statistical and machine-learning approaches to estimate marine nitrogen fixation, the process by which nitrogen gas is converted into forms organisms can use. That research, submitted to Journal of Geophysical Research: Oceans, also became Hu’s Signature Work project.
Qianhui Huang, Class of 2026, also took on a substantial research role in Professor Kim Hunter Gordon’s research on 19th-century Chinese martial theater. She examined actor Tan Xinpei, Empress Dowager Cixi’s theatrical patronage and the development of the wenwu laosheng, an older male role combining moral seriousness with controlled martial strength. Her analysis contributed directly to Gordon’s larger project by helping identify a later phase in the changing gendered meanings of martial authority.
Chung-Li Lin, who worked on Professor Zhang’s copper nanocrystals experiments, described a similar shift in how he saw his own role in research. Before this project, he said, he mainly saw himself as someone curious about science who enjoyed experiments. The experience taught him to stay with a question when the answer was unclear and keep refining his understanding through evidence.
“That transition from being interested in science to feeling that I could genuinely take ownership of a scientific problem was probably the most important thing I gained from the project,” he said.
All 14 projects and key outcomes
- Kent Cao – Craftsmen’s hand: Doodles, scribbles and the casual mind in ancient China
A journal article manuscript is in progress, a presentation proposal is planned for the 2027 Association for Asian Studies-in-Asia conference, and early findings have been incorporated into HIST 301 Arts of China.
- Titas Chakraborty – Concubinage in the making of the early colonial state in India
Chakraborty completed the project’s archival data collection and two article manuscript drafts as of July 2026. Related research has also been presented at academic conferences and is contributing to her second book manuscript.
- Kim Hunter Gordon – Ethnicity, gender and power in 19th century theatre: The politics of cross-gender martial roles in Qing China
The project resulted in an international conference presentation, a teaching module for GCHINA 201 and a completed student research project. A peer-reviewed article manuscript is in progress, and the research is also contributing to a longer-term book project.
- Zhenghui Huo – Interpolation methods and mapping properties for operators in complex analysis
Huo presented the research at the 2025 Several Complex Variables Annual Meeting and the 2025 DKU-SJTU Joint Workshop. One related paper has been submitted, and another is in progress.
- Joohyun Lee – Epigenetic regulation of cadmium tolerance: Role of HISTONE DEACETYLASE 2A in plant stress response mechanisms
The project established standardized cadmium stress conditions, generated sequencing-ready materials and identified IMA3 as a priority target for further testing. The work is being developed toward a mechanistic manuscript. Seven students participated in the core research activities, and six have since been admitted to major U.S. Ph.D. programs with full tuition and living stipend support.
- Peiyuan Li – AI-driven extraction of Chinese genealogical data using fine-tuned LLMs
The project produced an annotation framework and a growing genealogy corpus, with 20,000 books collected, another 25,000 being acquired and an initial 100 books processed for training and validation. Extracted data was incorporated into a manuscript now has been conditionally accepted at the Journal of Economic History. The project has also supported student Signature Work and a student-faculty research paper.
- Zuchuan Li – Learning to reconstruct the ocean’s respiration
Qianxun Xu was lead author of work published at NeurIPS 2025 on a partial physics-informed diffusion model for reconstructing ocean chlorophyll concentrations. Yiwen Hu led a study on marine nitrogen fixation submitted to JGR-Oceans, which also became her Signature Work. A separate study on ocean masked autoencoders was submitted to NeurIPS.
- Fan Liang – Privacy at the crossroads: Understanding concerns, cynicism, and management of online privacy in the digital era
The research was presented at the 76th International Communication Association Conference in Cape Town, and a paper has been submitted to Information, Communication & Society.
- Rui Liu – Photoacid/base enabled red or NIR light triggered ketyl and amidyl radical generation by PCET
The project produced a paper published in Scientific Reports, volume 16, article 26703 (2026), and a second paper accepted by Molecules. The project also produced a publicly released 1,437-molecule dataset and reproducible code, along with the BeyondPhoton software and web platform for molecular design and spectral prediction.
- Rasoul Namazi – The political thought of Ahmad Fardid: Archival investigation
Namazi located and examined archival materials concerning Fardid’s student years in Paris and identified evidence that, to his knowledge, had not previously been used in scholarship on Fardid. The research has contributed to an article invited for revision and resubmission, a book chapter in progress and two international conference presentations.
- Lie Philip Santoso – Shifting opinions: The role of partisan cooperation in shaping policy preferences
The project resulted in an article published in Comparative Political Studies. The research was also presented at the University of Macao, the University of Hong Kong, Soochow University and the DKU Social Science Brown Bag.
- Shan Wang – Computational modelling of the motor-cognitive processes underpinning age-related changes in rapid online control
The project contributed to an article published in Psychology and Aging on age-related decline in double-step reaching. A second manuscript on age-related differences in rapid movement adjustment is in preparation.
- Shixin Xu – Mathematical model of sweat gland function for noninvasive glucose monitoring
The project contributed to a published article in the Journal of Computational Physics. Two further manuscripts are in preparation or revision, including work on a multicompartment model of sweat gland dynamics and the student-led SciDynBench-R benchmark.
- Tan Zhang – Novel catalytic strategy for sustainable synthesis of nanocrystals
The project reported two completed journal publications, including the Inorganic Chemistry paper on chloride-catalyzed synthesis of copper nanocrystals, along with three additional articles or reviews in progress. The research was also presented at two chemistry conferences, and the Inorganic Chemistry paper was highlighted on the journal’s cover.
