Peichun Amy Tsai

Using high-pressure microfluidics, we visualized oil displacement by supercritical CO ₂ (scCO ₂ ) foam and examined pore-scale mechanisms affecting foam transport and mobility control in enhanced oil recovery (EOR) under reservoir-relevant conditions.

Liquid-metal droplet impact plays a key role in thermal management, coating, and additive manufacturing, yet its dynamics on rough and heated surfaces remain to be elucidated. This study experimentally investigates the impact of oxidized Galinstan droplets on solid substrates, isolating the effects of surface roughness and substrate temperature in the absence of solidification. Experiments were conducted over wide ranges of Weber number (0.15 ≤ We ≤ 720), average surface roughness (0.2 nm ≤ Ra ≤ 19.5 μm), and surface temperature (25 °C ≤ Ts ≤ 400 °C). Our results show that surface roughness promotes splashing by reducing the critical impact velocity. By fitting the experimental data within the explored parameter range, an empirical splashing threshold, ReOh4/5 = 9.8r–0.28, is established by incorporating the dimensionless roughness (r), Reynolds number (Re), and Ohnesorge number (Oh). Despite its strong influence on splashing, surface roughness has a negligible effect on the maximum spreading factor (βmax) due to Galinstan’s high surface tension. Substrate temperature likewise exerts only a minor influence, as the droplet temperature variation during spreading causes only minor changes in the liquid properties and subsequent impact dynamics under our conditions. For all Ts and Ra conditions, the experimental data of βmax collapse onto an empirical correlation, βmax = 0.1(2154 + We/Oh)0.3 for Oh∼O(10−4) and 0.15 ≤ We ≤ 720.

This work presents a novel microfluidic method for dynamically determining the minimum miscibility pressure (MMP) of CO ₂ –oil systems with high precision and direct visualization.

The transition to sustainable energy is crucial for mitigating climate change impacts, with hydrogen and carbon storage and utilization technologies playing pivotal roles. This review highlights the integral and useful role of microfluidic technologies in advancing subsurface fluid dynamics for carbon capture, utilization, and storage (CCUS), enhanced oil recovery (EOR), and underground hydrogen storage (UHS). In particular, microfluidic platforms provide clear and insightful visualization of fluid–fluid and fluid–solid interactions at the pore scale, crucial for understanding and further optimizing processes for CO₂ sequestration, hydrogen storage, and oil displacement in various geological formations. We first discuss the development of lab-on-a-chip devices that accurately mimic subsurface conditions, allowing detailed studies of complex phenomena including viscous fingering, capillary trapping, phase behavior during CCUS and EOR processes, and the hysteresis effects unique to hydrogen storage cycles. We also discuss the dynamics of CO₂ gas and foam in enhancing oil recovery and the innovative use of hydrogen foam to mitigate issues associated with pure hydrogen gas storage. The integration of advanced imaging, spectroscopic techniques, and machine learning (ML) with microfluidic experiments has enriched our understanding and opened new pathways for predictive capabilities and operational optimization in CCUS, EOR, and UHS applications. We further emphasize the critical need for continued research into microfluidic applications, e.g., incorporating state-of-the-art ML to optimize microfluidic experiments and parameters, and UHS enhancement through favorable microbial activities and suppression of reactions in H₂ foam, aiming at refining storage strategies and exploiting the full potential of these technologies towards a sustainable energy future.

Direct visualization reveals how CO ₂ foam enhances fluid mobility control in heterogeneous porous media, outperforming surfactant solution and CO ₂ gas injection to improve sweep efficiency and oil recovery.

Determining accurate drug dissolution processes in the gastrointestinal tract is critical in drug discovery as dissolution profiles provide essential information for estimating the bioavailability of orally administered drugs. While various methods have been developed to predict drug solubility based on chemical structures, no reliable tools currently exist for predicting the dissolution rate constant. This study presents a novel two-stage machine learning approach, termed Machine Learning based Quantitative Structure-Dissolution Profile Relationship, which integrates physics-informed neural networks (PINNs) and deep neural networks (DNNs) to predict drug dissolution profiles in water, with varying concentrations of surfactant Sodium Lauryl Sulfate. In the first stage, PINNs extract key dissolution parameters─namely the dissolution rate constant (k) and the dissolved mass fraction at saturation (ϕ_s)─from existing dissolution data. By leveraging a physical law governing the dissolution process, PINNs aim to enhance prediction performance and reduce data requirements. Assuming first-order kinetics of the drug dissolution process as described by the Noyes-Whitney equation, PINNs, with 8 hidden layers and 40 neurons per layer, may outperform traditional nonlinear regression by effectively filtering noise and focusing on physically meaningful data. In the second stage, these extracted parameters (k and ϕ_s) are used to train a DNN to predict dissolution profiles based on the drug's chemical structure and dissolution medium. Using the FDA-recommended metrics: the difference and similarity factors (f₁ and f₂), the DNN─with 128 neurons in two hidden layers and a learning rate of 10^-2.8─achieved an average testing accuracy of 61.7% at an 80:20 train-to-test split. Although this current accuracy is below the generally acceptable range of 70-80%, this approach shows significant potential as a low-cost, time-efficient tool for early phase drug formulation. Future improvements are expected as data quality and diversity increase.