Dong-Yeun Koh

Microwave-driven carbothermal shock (CTS) synthesis has attracted considerable attention as an ultrafast route for nanomaterial production. However, controlling nanoparticle nucleation and growth during the extremely short microwave irradiation period remains challenging. Although defect-rich carbon materials have frequently been employed in microwave-driven CTS synthesis, the influence of a separately introduced defect-containing surface layer on nanoparticle formation remains insufficiently understood. In this study, carbon cloth (CC) was coated with a polyacrylonitrile (PAN) solution and thermally oxidized at 200–400 °C to modify the surface carbon structure through the evolution of PAN-derived polyaromatic clusters prior to microwave irradiation. Raman spectroscopy and X-ray photoelectron spectroscopy indicated that increasing the oxidation temperature promoted the formation of larger and more ordered polyaromatic domains, accompanied by changes in the surface defect characteristics. These structural variations were associated with distinct Cu nanoparticle sizes and distributions following CTS synthesis. Among the investigated oxidation temperatures, the PAN-coated CC treated at 300 °C produced the smallest and most uniformly dispersed Cu nanoparticles, with an average diameter of 58 ± 25 nm. These findings suggest that controlling the structural evolution of PAN-derived surface carbon layers can provide a practical approach for regulating nanoparticle size and dispersion during microwave-driven CTS synthesis.

Practical direct air capture (DAC) demands a sorbent contactor that simultaneously delivers high active-material loading, low flow resistance, efficient mass transfer, and rapid regeneration─requirements that conventional sorbent architectures rarely satisfy together. Here, we present a woven fiber contactor that addresses these requirements by interlacing dry-fabricated zeolite Y/PTFE fibers with copper wire supports into a structured contactor. Compared to a stacked laminate configuration, the woven sorbent exhibited markedly more efficient mass transfer, yielding a narrower mass transfer zone (LMTZ = 4.4 mm and 11.1 mm at 50 sccm) and a steeper breakthrough front. This structural advantage translated into robust sorption performance, with CO2 capture capacity varying by only 11.7% across a 10-fold increase in feed flow rate (50–500 sccm, corresponding to superficial velocities 5.79 × 10–3 to 5.79 × 10–2 m/s) compared to 19.2% for the stacked laminate. The woven architecture further reduced pressure drop by 7- to 18-fold over a wide range of superficial velocities (1.16 × 10–2–5.79 × 10–1 m/s). The integrated copper wires enabled electrified temperature swing adsorption (e-TSA) with near-instantaneous heating, driving CO2 desorption within 1 min. Over 10 consecutive adsorption–desorption cycles, the module retained an average capacity of 1.01 mmol/gsorbent with less than 6.93% variance. Together, these results establish the woven fiber contactor as a scalable platform that unites high sorbent loading, low flow resistance, efficient mass transfer, and rapid electrified regeneration for energy-efficient DAC.

Carbon molecular sieve (CMS) membranes represent a high-performance alternative to polymers, surpassing the trade-off limits between permeability and selectivity typically seen in conventional polymeric membranes. Their efficacy stems from a precise molecular sieving mechanism, enabling exceptional gas separation performance. However, a challenge persists in fine-tuning their microporous structure to achieve effective kinetic separation of gas molecule pairs with small differences in kinetic diameter. We demonstrate that simple plasma treatment can fundamentally alter the CMS structures, drive ultraselective molecular hydrogen (H ₂ ) separation over carbon dioxide (CO ₂ ) and methane (CH ₄ ), and stabilize the structures over a year of physical aging. Two types of plasma, nitrogen- and argon-derived, were used, both of which induced surface oxidation by creating defects or incorporating nitrogen atoms into the carbon layer and strands, leading to distinct development of ultramicroporosity compared to pristine CMS structures. These structural modifications resulted in marked improvements in separation performance. The plasma-treated membranes achieved over 40-fold increases in H ₂ /CO ₂ selectivity and over 137-fold increases in H ₂ /CH ₄ selectivity compared to the pristine membrane, while maintaining high H ₂ permeabilities of over 1000 Barrers, surpassing the current upper bound set by CMS membranes. When combined with high-temperature pyrolysis, the plasma-treated membrane exhibited a record-high combination of H ₂ permeability (exceeding 100 Barrers) and H ₂ /CO ₂ selectivity (exceeding 260). Plasma-treated CMS membranes demonstrated exceptional stability over 14 months of aging, with less than 10% variation in observed performance. This study establishes plasma treatment as a powerful strategy, delivering CMS membranes with both enhanced separation performance and remarkable long-term stability.

Direct air capture (DAC) is a promising negative-emissions technology, but its performance depends strongly on ambient air temperature and humidity, whose temporal variability is often neglected in DAC analysis and operation. This work develops a customizable gPROMS-based DAC process model integrated with a custom Bayesian optimization framework to determine operating conditions that adapt to changing ambient conditions. For each set of air conditions, the framework optimizes key operating variables, including desorption pressure, desorption temperature, and adsorption/desorption times, to balance productivity and energy consumption. The results show that neglecting ambient variability can significantly misrepresent DAC performance. Under certain conditions, variations in temperature and humidity increase energy consumption by 6.98% and reduce productivity by 14.8% relative to constant-condition assumptions. When applied in a continuous cycle-by-cycle manner across representative climates and seasons, the proposed optimization framework improves productivity by up to 7.04% and reduces energy consumption by 25.2% compared with fixed-parameter operation under variable ambient conditions. The analysis also identifies favorable climate-dependent operating regions, providing a mapping from ambient conditions to optimal operating parameters and revealing operational thresholds beyond which DAC operation becomes unfavorable. A secondary simplified profit-based case study for Finland, incorporating time-varying electricity and carbon prices, shows that an adaptive-weight, cycle-by-cycle optimization strategy that dynamically adjusts the trade-off between productivity and energy consumption outperforms a constant-weight strategy across all seasons, with a maximum profit improvement of 61.7% in summer. Overall, this study demonstrates that explicitly accounting for ambient variability and dynamically adapting operating conditions can materially improve both the technical performance and economic viability of DAC systems.

Atmospheric and vacuum distillation consume more than 1,100 TWh year⁻¹ and emit more than 160 million metric tonnes of CO₂ equivalent annually^1,2, making membrane-based pre-fractionation a compelling retrofit strategy for lowering the energy and carbon intensity of petroleum refining^{3,4,5,6,7,8,9,10}. Here we demonstrate that porous polyacrylonitrile (PAN) membranes, typically used as support layers, achieve effective molecular refining of crude oil at steady state. Under tangential flow, PAN membranes exhibited high crude oil permeances of up to 0.591 ± 0.040 l m⁻² h⁻¹ bar⁻¹, a more than 23-fold increase over the previous benchmark (−2 h⁻¹ bar⁻¹)^1,11, selectively yielding enriched lighter hydrocarbon fractions such as naphtha and kerosene. This unexpected selectivity arises from the dynamic deposition of heavy hydrocarbons within the initially approximately 15-nm surface mesopores, which narrows the pore diameter to sub-2-nm dimensions. Depth-resolved chemical identification reveals selective accumulation of n-alkanes, suggesting a self-limiting pore constriction mechanism that stabilizes selective transport pathways. Once the n-alkane deposition is stabilized, selective enrichment of raw crude oils occurs with sustained stability over 4 weeks. Process simulations show that PAN-membrane-based pre-fractionation could reduce energy by 31.6%, cooling water by 20.7% and CO₂ emissions by 37.6% compared with traditional atmospheric distillation.

While electrochemical direct ocean capture (e‐DOC) offers a plausible pathway for gigaton‐scale decarbonization, its practical deployment remains bottlenecked by complex reactor architectures, constrained fluid‐electrode interfacial areas, massive Ohmic penalties, and severe mineral fouling. Herein, a compact, membrane‐integrated hollow fiber electrode assembly architecture that systematically overcomes these limitations for continuous marine carbon mineralization is reported. Macroporous stainless‐steel hollow fibers, fabricated via scalable fiber spinning and sintering, deliver an expansive electrochemically active surface area alongside exceptional mechanical robustness and corrosion resistance. Crucially, by seamlessly integrating a membrane and a counter electrode within the fiber lumen, a sub‐millimeter inter‐electrode spacing (≈1 mm) is established. This coaxial configuration reduces ionic transport pathways and minimizes Ohmic losses, driving localized hydroxide generation with effective pH swing for rapid dissolved inorganic carbon (DIC) removal. Furthermore, the membrane‐integrated working electrode design effectively mitigates fouling, enabling stable continuous operation for over 100 h while delivering a DIC removal efficiency exceeding 85%. This advanced materials‐to‐device platform establishes a robust, high‐throughput paradigm for practical ocean decarbonization.

Efficient separation of hydrogen isotopes is essential for sustainable hydrogen management and environmental protection in nuclear and other industrial systems. Conventional hydrogen isotope separation technologies, including cryogenic distillation, Girdler sulfide process and ammonia–hydrogen exchange, consume massive energy and involve corrosive or toxic reagents, leading to significant environmental and operational challenges. In contrast, the liquid phase catalytic exchange (LPCE) process, based on water–hydrogen gas isotope exchange, provides an efficient and energy-conserving alternative consistent with sustainable catalytic processes. In this study, a platinum decorated mesoporous silica-polydimethylsiloxane composite (Pt@SiO₂-PDMS) was developed and applied as a catalyst for LPCE process. Physico-chemical characterizations verified the decoration of Pt into the mesoporous silica without destruction of mesoporous structure. The Pt@SiO₂ particles were homogeneously distributed in the PDMS matrix and the resulting composite maintained an intrinsic hydrophobic surface with a water contact angle above 108°. The catalyst composite exhibited effective hydrogen isotope exchange performance up to column efficiency of 67.7% at 80 °C when Pt@SiO₂-PDMS randomly packed with Dixon rings in a volume ratio of 2:1 in the LPCE process. Furthermore, stable operation was maintained continuously for 14 days at 80 °C, demonstrating the durability of catalyst composite and application on actual process conditions. The selective hydrogen isotope exchange between liquid water and hydrogen gas with Pt@SiO₂-PDMS catalyst is attributed to the effective porous structure of SiO₂, Pt loading of Pt@SiO₂ and hydrophobic nature of the PDMS matrix.

As the global carbon budget rapidly depletes, the transition to a sustainable, stable climate demands the gigatonne-scale deployment of negative emission technologies.