Hyunjin Moon

Hyunjin Moon

Auto-generated profile, based on publicly available information

Hyunjin Moon

Department of Chemical Engineering, University of Arkansas, Fayetteville, USA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 March 2020 in Chemical Science
Verified
Journal Article
Open access
PDF file
HMHyunjin Moon
Hyunjin Moon
SHSongi Han
Songi Han
SSSusannah L. Scott
Susannah L. Scott
Surface polarity plays a key role in controlling molecular adsorption at solid–liquid interfaces, with major implications for reactions and separations. In this study, the chemical composition of periodic mesoporous organosilicas (PMOs) was varied by co-condensing Si(OEt)4 with organodisilanes, to create a homologous series of materials with similar surface areas, pore volumes, and hydroxyl contents. Their relative surface polarities, obtained by measuring the fluorescence of a solvatochromic dye, cover a wide range. In this series of PMO materials, EPR spectra of tethered nitroxide radicals show monotonically decreasing mobility as larger fractions of the radicals interact strongly with increasingly non-polar surfaces. The surface properties of the materials also correlate with their affinities for organic molecules dissolved in various solvents. The most polar PMO has negligible affinity for phenol, p-cresol, or furfural when these molecules are dissolved in water. However, stronger solute–surface interactions and favor adsorption as the surface polarity decreases. The trend is reversed for furfural in benzene, where weaker solvent–surface interactions result in higher adsorption on polar surfaces. In DMSO, furfural adsorption is suppressed due to the similar strengths of solute-surface and solvent–surface interactions. Thus, the polarity of the surface relative to the solvent is critical for molecular adsorption. These findings show how adsorption/desorption can be precisely and systematically tuned by appropriate choice of both solvent and surface, and contribute to a predictive strategy for the design of catalytic and separations processes.

29 October 2025 in Nature Communications
Verified
Journal Article
Open access
PDF file
HMHyunjin Moon
Hyunjin Moon
JDJason S. DesVeaux
Jason S. DesVeaux
ERElisabeth C. Van Roijen
Elisabeth C. Van Roijen
CSChristine A. Singer
Christine A. Singer
HAHannah M. Alt
Hannah M. Alt
GBGregg Beckham
Gregg Beckham
Oxidative catalytic depolymerization of polystyrene (PS) can produce benzoic acid, but the annual consumption of benzoic acid is ~40 times lower than PS. For this catalytic oxidation method to be a viable means to manage PS waste, benzoic acid should be converted to higher-volume chemicals. We demonstrate a hybrid chemical and biological process that uses PS as feedstock for production of adipic acid, a high-volume co-monomer for nylon 6,6 via benzoic acid. Mn/Br co-catalyzed autoxidation of PS to benzoic acid proceeds with a yield of up to 94% in a solvent mixture of benzoic acid and water. The PS-derived benzoic acid undergoes bioconversion at near-quantitative yield to muconic acid, which is readily converted to adipic acid through catalytic hydrogenation. Process modeling, techno-economic analysis, and life cycle assessment estimate an adipic acid minimum selling price of $3.18/kg, with a 61% decrease in greenhouse gas emissions relative to production from fossil fuels.

23 September 2026 in Scientific Reports
Verified
Journal Article
Open access
PDF file
YLYeonseo Lee
Yeonseo Lee
TKTae Hyung Kim
Tae Hyung Kim
HMHyunjin Moon
Hyunjin Moon
Machine learning-based yield prediction in chemical reaction systems is often limited by small experimental datasets and condition-dependent variability. Here, we develop an ensemble regression framework that combines Natural Gradient Boosting (NGBoost) with a heteroscedastic deep learning with Multi-Layer Perceptron (MLP) to predict reaction yields from limited data while estimating condition-specific uncertainty. The framework was evaluated using 137 experimental samples from polystyrene autoxidation reactions. The ensemble model achieved a mean absolute error (MAE) of 5.4 ± 0.2, up to a 62% reduction relative to linear baseline methods (Linear Regression, Ridge, Lasso), and an approximately 24% reduction relative to the strongest single-model baseline evaluated (NGBoost alone, MAE = 7.1), outperforming the individual NGBoost and MLP models. Shapley Additive Explanations (SHAP)-based interpretation showed that the model captured non-linear response patterns associated with NaBr/Mn acetate loading, and reaction time, which were consistent with experimentally observed trends. These results suggest that uncertainty-aware ensemble learning can support reaction condition screening and prioritization in data-limited chemical systems.