Comparing XPS on bare and capped ZrN films grown by plasma enhanced ALD: Effect of ambient oxidation

Abstract

In this article we compare x-ray photoelectron spectroscopy (XPS) measurements on bare- and capped- zirconium nitride (ZrN) films to investigate the effect of ambient sample oxidation on the detected bound O in the form of oxide ZrO2 and/or oxynitride ZrOxNy. ZrN films in both bare- and Al2O3/AlN capped- XPS samples were grown by plasma-enhanced atomic layer deposition (PEALD) technique using tetrakis dimethylamino zirconium (TDMAZr) precursor, forming gas (5% H2, rest N2) inductively coupled plasma (ICP), and as received research grade process gases under identical process conditions. Capped samples were prepared by depositing 1 nm thick PEALD AlN on ZrN, followed by additional deposition of 1 nm thick ALD Al2O3, without venting of ALD reactor. On bare ZrN sample at room temperature, spectroscopic ellipsometry (SE) measurements with increasing ambient exposure times (texp) showed a self-limiting surface oxidation with the oxide thickness (dox) approaching 3.7 ± 0.02 nm for texp > 120 min. In XPS data measured prior to sample sputtering (tsput = 0), ZrO2 and ZrOxNy were detected in bare- samples, whereas only ZrN and Al2O3/AlN from capping layer were detected in capped- samples. For bare-ZrN samples, appearance of ZrO2 and ZrOxNy up to sputter depth (dsput) of 15 nm in depth-profile XPS data is in contradiction with measured dox = 3.7 nm, but explained from sputtering induced atomic inter-diffusion within analyzed sample. Appearance of artifacts in the XPS spectra from moderately sputtered (dsput = 0.2 nm and 0.4 nm) capped-ZrN sample, provides an evidence to ion-bombardment induced modifications within analyzed sample.