Characterization of Large-Area Plasma Jet Discharges and Sintering Performance of Printed Thermoelectric Thin-Films

Thermoelectric (TE) generators are a promising energy source as they naturally interconvert heat flux and induced voltages, and their solid-state nature makes them quite versatile. In order to synthesize TE thin-films with user-tunable patterns on a variety of conformal substrates, there is a demand for alternatives to pressurized and thermal methods of thin-film printing such as chemical vapor deposition and magnetron sputtering, both of which require high energy expenditure and have material limitations. One such alternative is aerosol-jet printing (AJP), in which the sample requires additional sintering processing upon printing to evaporate the organic ink surfactants and induce agglomeration of the nanoparticles into a continuous body.

Sintering of particles occurs when the free energy of solubilized nanoparticles decreases and their boundaries merge in a process called necking, proceeding towards the monocrystal state. External energy is typically provided by thermal means such as furnaces or joule heating induced by a spark discharge in a process called spark plasma sintering (SPS). A proposed non- thermal alternative is low temperature, atmospheric pressure plasma jets, characterized as plasmas with electron temperatures far exceeding that of the ionized gas particles and neutral species. A common way to generate this type of plasma jet is a dielectric barrier discharge (DBD), or an electrical discharge between two electrodes separated by a dielectric surface. Since these jets are typically cylindrical in nature, and the AJP samples are rectangular in shape, two expanded nozzle plasma jets were studied in this work: a bell nozzle jet (BNJ) and elongated slot jet (ESJ). The performance of these novel jets, along with a well-studied cylindrical jet, were characterized by their flow rate, applied voltage, and substrate-nozzle gap distance. Imaging the three types of plasma discharges has shown that the BNJ is prone to visual fluctuations, air breakdown, and external ionization wave propagation. It has also shown that applied voltage significantly affects plasma intensity, volume, and power, while flow rate and gap do not. Thus, the ESJ was selected for use of sintering bismuth telluride (Bi₂Te₃) and silver nanoparticle (Ag NP) AJP thin-films, as well as vacuum-filtrated silver selenide (Ag₂Se) thin-films.

The ESJ was shown to be unsuccessful in sintering Bi₂Te₃ and Ag NP thin-films. Bi₂Te₃ sintering was unsuccessful as the electrical resistance increased after sintering and reverted to >100 MΩ in all samples over the next two days. Though the target measured resistance of <100 Ω was achieved for four total Ag NP samples, the wide variability within treatment groups across 15 total samples for no discernable reason indicates unsuccessful sintering. Ag₂Se sintering was successful in achieving fairly consistent resistances less than 100 Ω, but TE properties, namely Seebeck coefficient (S) and power factor (PF) had diminished. Scanning electron microscopy (SEM) and energy dispersive x-ray analysis (EDX) showed severe selenium (Se) loss of 89%, thus indicating that the silver was being sintered causing conductivity. Cold-pressing (CP) the samples at 25 MPa upon synthesis caused reduction of Se loss to 44% after ESJ sintering, a 50% improvement. Proposed future paths of fine-tuning an ESJ sintering protocol for Ag₂Se include using helium instead of argon plasmas, as helium ions deliver softer collisions to the sample surface, and intensively characterizing the sintering results from tuning flow rates and applied voltages.