Decoding and controlling droplet dynamics in ultrasonic spray coating: from atomization to diamond thin film applications

Abstract

Ultrasonic Spray Coating is a promising deposition technique for the fabrication of thin and ultrathin functional layers. The technique combines several advantages, including low material losses, compatibility with sensitive materials, scalability toward industrial processes, and the ability to deposit homogeneous coatings over large areas. As a result, increasing interest has emerged in applications related to electronics, energy devices, photovoltaics, sensors, and diamond coatings. Despite these advantages, the relationship between ink properties, spray parameters, droplet formation, and final layer quality remains highly complex and only partially understood. The ultrasonic spray coating process consists of multiple strongly coupled stages: ultrasonic atomization of the liquid, droplet transport through the spray cone, droplet impact on the substrate, and finally solvent evaporation and thin-film formation. Small variations in parameters such as viscosity, surface tension, flow rate, atomization power, substrate temperature, or gas pressure can significantly influence the resulting coating morphology. The objective of this thesis was therefore to obtain a more fundamental understanding of the physical processes occurring during ultrasonic spray coating. Particular focus was placed on the relationship between liquid properties, droplet formation, droplet transport, and final layer formation. In addition, new process strategies and data-driven modelling approaches were explored to better understand and optimize the ultrasonic spray coating process. In the first part of this work, the mechanisms governing droplet formation during ultrasonic atomization were investigated. For this purpose, phase doppler anemometry was employed as an advanced optical characterization technique capable of simultaneously measuring droplet size and droplet velocity. Different model inks were studied, including water, acetone, polyvinylidene fluoride based polymer solutions, and SiO2 nanoparticle dispersions. The influence of parameters such as flow rate, atomization power, and ink composition on the droplet size distribution was systematically investigated. The results demonstrated that classical atomization models are often insufficient to accurately describe complex inks. In particular, polymers and nanoparticles were found to significantly affect droplet formation and, in some cases, resulted in bimodal or multimodal droplet size distributions. Based on an extensive dataset consisting of hundreds of spray conditions, machine learning models were therefore used to predict droplet size as a function of both process parameters and liquid properties. These models showed a clear improvement compared to traditional empirical correlations and represented an important step toward predictive process optimization for ultrasonic spray coating. Subsequently, droplet behaviour during transport through the spray cone was investigated. The influence of shroud gas pressure, nozzle-to-substrate distance, and spatial position inside the spray cone on droplet size and velocity was studied in detail. The experiments revealed that the spray is not homogeneous throughout the entire cone. Both droplet velocity and droplet size varied depending on the radial position inside the spray. In addition, droplet collisions and solvent evaporation during flight were found to play an important role in determining the final droplet distribution reaching the substrate. An important part of this thesis focused on thin-film formation and the suppression of undesired drying patterns such as the coffee-ring effect. During conventional deposition processes, dissolved or dispersed materials often migrate toward the droplet edge during evaporation, resulting in non-homogeneous coatings. To study this phenomenon, the influence of substrate temperature was investigated. In this work, a novel approach was introduced in which substrates were actively cooled to temperatures below ambient conditions. Experiments demonstrated that lower substrate temperatures strongly influence the internal flow behaviour inside evaporating droplets. By reversing temperature gradients and modifying evaporation dynamics, coffee-ring formation could be significantly reduced. Furthermore, cooled deposition conditions resulted in more homogeneous ultrathin coatings in several cases. These findings suggest that temperature control can serve as a powerful tool for steering thin-film formation during spray coating. Besides the fundamental studies, the ultrasonic spray coating process was also applied to the deposition of nanodiamond seed layers for diamond growth using Microwave Plasma Enhanced Chemical Vapor Deposition. Different deposition techniques, including ultrasonic spray coating, spin coating, and dip coating, were compared. The results demonstrated that ultrasonic spray coating is capable of depositing homogeneous seed layers with good control over seeding density and coating homogenity. The quality of the deposited seed layer was shown to directly influence the final morphology, optical properties, and mechanical properties of the grown diamond coatings. In addition, the influence of surface pretreatments was investigated. Plasma and UV-ozone treatments were found to strongly modify the surface energy and wettability of substrates, which in turn significantly affected droplet spreading and layer formation. By combining surface pretreatment, optimized spray parameters, and temperature control, the operational window for homogeneous coating formation could be mapped more effectively. The results presented in this thesis demonstrate that ultrasonic spray coating is a highly versatile and powerful deposition technique, while simultaneously highlighting that a deep understanding of the underlying physical mechanisms remains essential for further optimization and industrial implementation. The combination of advanced characterization techniques such as phase doppler anemometry with data-driven modelling and innovative process strategies provides new opportunities to better understand and control complex spray processes. Our research demonstrates that the further development of ultrasonic spray coating depends not only on improvements in hardware, but also on an integrated approach in which fluid dynamics, materials science, data analysis, and process control are combined. The insights obtained throughout this thesis therefore represent an important step toward more predictable, reproducible, and industrially scalable ultrasonic spray coating processes for advanced thin-film technologies.