This is the authors' abstract. We don't add key points for this paper.
Charge trapping layers suppress parasitic surface conduction on high-resistivity silicon substrates for radiofrequency (RF) applications. This work investigates silicon nitride (Si x N y ) as a charge trapping layer, using variants deposited by physical vapor deposition (PVD) and plasma-enhanced chemical vapor deposition (PECVD). Coplanar waveguide (CPW) insertion loss is measured to assess RF suitability. Layers are characterized by spectroscopic ellipsometry, Time-of-Flight Secondary-Ion-Mass-Spectroscopy, and Fourier transform infrared spectroscopy. Insertion loss is low for wafers with PVD Si x N y and high for PECVD Si x N y . CPW measurements at different temperatures show insertion loss increases for PVD Si x N y but decreases for PECVD Si x N y . Chemical and physical analysis reveals negligible hydrogen in PVD Si x N y , while PECVD layers contain hydrogen at least an order of magnitude higher. Finite Element Method (FEM) simulations calculate insertion loss for high and low parasitic surface conductivity scenarios. Hydrogen presence appears to dominate insertion loss differences: traps in Si x N y saturate with hydrogen, preventing suppression of parasitic surface conduction. The hypothesized mechanism involves changes in sheet conductivity in the Si carrier's space charge region. Simulations using varying conductivity values show 450 S/m matches PECVD Si x N y measurements, while 2 S/m aligns with PVD Si x N y . Variations in Si x N y thickness do not significantly affect CPW insertion loss in simulations or measurements. Overall, hydrogen-free PVD Si x N y is well suited as a charge trapping layer to reduce parasitic surface conduction on high-resistivity Si substrates.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Naguib et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: