Droplet microfluidics can generate and manipulate discrete droplets through immiscible multiphase flows inside microcha.nnels, which effectively controls the size and generation rate of droplets. Due to its remarkable advantages, droplet microfluidics bears a significant value in an extremely wide range of areas, such as bio(chemical) analysis, particle preparation or petroleum engineering. Droplet splitting is an important form of obtaining monodisperse and small droplets in microfluidic devices, which can obtain smaller droplets to meet the requirements of biochemical reaction and material encapsulation. Nanoparticle surfactants, formed at liquid-liquid interfaces by the interactions between functional groups on nanoparticles and polymers having complementary end-functionality, are recently proposed as an excellent interface stabilizer to cover liquid droplets for applications of substance encapsulation and delivery. Apparently. understanding droplet splitting behavior with consideration of nanoparticle surfactants is of great significance for obtaining smaller and better dispersive functional droplets. In this paper, by using the approach of microfluidic experiment and theoretical analysis, the effects of nanoparticle surfactants on droplet splitting behavior within T-junction microchannels were investigated. Comparisons were conducted among clean droplets and those stabilized by nanoparticles alone, surface-active polymers alone and nanoparticle surfactants, respectively. Besides, the effects of concentrations of both nanoparticles and amino-terminal polymers, which determined nanoparticle surfactants concentration on the fluid-fluid interface, on the droplet splitting behavior have been systematically investigated. We have found that the droplet splitting process can be divided into three stages: Entering stage, squeezing stage and post-splitting stage. The generation of nanoparticle surfactants can effectively reduce the time required for droplet splitting. Besides, the nanoparticle surfactants induced reduction of the total split time is mainly attributed to the decrease of the entering time, while no significant change is presented in the squeezing and post-splitting time. The dynamical behavior of droplets splitting under different reagent conditions can be divided into four categories: Blocking splitting, transition state, non-blocking splitting and non-splitting. The blocking splitting state refers to that the droplets arc always in contact with the channel wall during the splitting process. Otherwise, there is always a gap between the droplet and the channel wall, which is called non-blocking splitting. Besides, the transition state is a special state of non-blocking splitting, which is in contact with the upper wall of the channel and has a gap with the bottom of the channel. More importantly, by analyzing the transient decreasing of neck width, we found that the thinning rate of the droplet was suddenly accelerated in the process of neck thinning, i.e., the rapid fracture stage. During neck thinning, the depression is fanned out until it reaches a critical width and becomes funnel-shaped. By geometric analysis of neck shape, a theoretical model of neck width thinning was obtained, which is applicable to different reagent conditions. Besides, the influence mechanism of nanoparticle surfactants on droplet splitting was obtained, i.e., affecting necking rate via decreasing the interfacial tension. Finally, via analyzing the droplet splitting state, a theoretical formulation was established for predicting the transition of blocking splitting and non-blocking splitting, which shows a good agreement with the experimental data. These findings provide a theoretical basis for the preparation of smaller functional droplets by droplet splitting.
