Machine learning assisted in-situ non-thermal plasma aerosol jet printing

Non-thermal plasmas have recently emerged as a promising post-printing treatment method for aerosol jet and inkjet printing, owing to their capability to remove surfactants from the nanoparticle-containing inks and promote sintering and densification, ultimately forming a continuous film with desired electrical conductivity. In this work, we develop a novel aerosol jet printing method that incorporates a non-thermal, atmospheric pressure plasma jet to enable in-situ sintering of printed silver nanoparticles right at the point-of-printing. A machine learning algorithm is integrated into the system control to provide real-time, online defect detection and parameter optimization, improving the production of high-quality films via autonomous compensation for any detected anomalies. Investigation of various processing parameters reveals that the optimal conditions yield silver nanoparticle films with electrical conductivities comparable to those processed through other offline plasma sintering methods. Since the non-thermal plasma sintering process is hypothesized to be driven primarily by energetic, chemically reactive plasma species and/or photons rather than by thermal effects, the processing temperature of this work remains near room temperature. Additionally, because the printed films require minimal post-processing, the overall manufacturing time is shortened by more than tenfold. This method holds significant potential to advance the implementation of non-thermal plasma sintering in printed electronics, wearable devices, and biomanufacturing.