Enhanced Nanoemulsion Engineering through Continuous Ultrasonication and Integrated Machine Learning Algorithms

Abstract

Ultrasonic flow reactors of modest size have become increasingly favored by researchers due to their utility as a valuable medium for investigating and regulating the operating mechanism of ultrasound technology. As a result, these reactors are employed for both research and a variety of applications in chemical, biological, and pharmaceutical processes, predominantly on laboratory scales and occasionally on pilot scales. Herein, an ultrasonic continuous-flow cell assembly (UCF) is utilized for the formulation of oil-in-water nanoemulsions (o/w NEs). The setup provides efficient energy input, inducing strong cavitation for effective droplet fragmentation and ensuring continuous o/w NEs production. The emulsification process is optimized by adjusting parameters (amplitude, pulse control mode, processing time, oil-to-surfactant ratio (OSR), and stability) supported by machine learning-based data analysis. At unit OSR, D_z-avg measures 124 nm, compared to 832 nm at OSR 12. At the optimal OSR, NE remains stable for 50 days; at higher OSR, coalescence and Ostwald ripening instabilities are observed. Optimized OSR is found to be 4 at optimal ultrasonic conditions. Higher surfactant concentration reduced D_z-avg, while increased oil concentration raised D_z-avg. A significant decrease in D_z-avg is observed up to 459 J/mL energy density, thereafter D_z-avg declined slowly. The D_z-avg prediction, modeled with energy density, exhibited a strong power law fit and is applicable for both scaling up and adjusting D_z-avg. This study proposes a standard process for formulating o/w NEs through an efficient continuous emulsification method. It offers valuable insights into optimizing emulsion formulation via ultrasonic cavitation techniques, contributing to scalable production in diverse industrial sectors.