From purification, spectroscopy, and microscopy of carbon dots to synthesis modeling and AI-assisted spectral data extraction

Bian, Zhengyi

Permalink

https://hdl.handle.net/2142/132802 Copy

Description

Title

From purification, spectroscopy, and microscopy of carbon dots to synthesis modeling and AI-assisted spectral data extraction

Author(s)

Bian, Zhengyi

Issue Date

2025-12-05

Director of Research (if dissertation) or Advisor (if thesis)

Gruebele, Martin

Doctoral Committee Chair(s)

Gruebele, Martin

Committee Member(s)

• Nie, Shuming
• Link, Stephan
• Landes, Christy F.

Department of Study

Chemistry

Discipline

Chemistry

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Carbon dots
• STM
• Data Collection

Abstract

Carbon dots (CDs) occupy a unique niche among nano-sized fluorescent materials. This dissertation integrates purification-first experiments, single-particle multimodal characterization, physical modeling, and AI-assisted data curation. Part A develops and applies rigorous purification and fractionation to disentangle bottom-up products of CD synthesis from confounding small molecules; it then combines ensemble spectroscopy with single-particle fluorescence (Eric Gomez) and scanning tunneling microscopy to quantify intrinsic absorption, bandgaps, blinking behavior, and structure–emission correlations. Building on this foundation, I engineer an impurity-free CD–dye hybrid that converts blue–green emissive CDs to red emission via near-ideal spectral overlap and short donor–acceptor separation, demonstrating a scalable path to color-tunable emitters. To enable electronic and optical probing, I fabricate ultrathin, atomically flat, and semi-transparent template-stripped Au films that simultaneously support scanning tunneling microscopy and single-particle photoluminescence, unlocking direct structure–property mapping at the single-dot level simultaneously. Part B advances a mechanistic view of bottom-up CD synthesis by formulating a Monte-Carlo–based dynamics framework for CD assembly. It then addresses the data bottleneck that limits ML for spectroscopy by creating an LLM-assisted, high-throughput pipeline that collect machine-readable structure–solvent–spectrum data at scale. Overall, the dissertation (i) establishes purification and single-particle standards that separate CD signals from artifacts, (ii) delivers practical routes to color-tunable CD emitters, STM/PL characterization of single CDs, and assembly process modelling, and (iii) bridges experiments and AI by converting the spectroscopy literature into large, usable datasets. These advances provide a reproducible foundation and scalable data infrastructure for CD photophysics and, more broadly, AI materials discovery.

Graduation Semester

2025-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/132802

Copyright and License Information

Copyright 2025 Zhengyi Bian

Owning Collections