Verification of layered structures in SnO2/metal-based gas sensors by X-ray microanalysis: comparison with X-ray photoelectron spectroscopy

The depth profile of thin film layers on bulk substrate, avoiding the cross-sectioning of samples, is commonly performed by techniques such as X-ray photoelectron spectroscopy (XPS), Auger electron Spectroscopy (AES), and secondary ion mass spectroscopy (SIMS). Techniques based on X-ray emission intensity measurements by energy dispersive spectroscopy (EDS), with conventional matrix or ZAF correction, are normally applied to cross-sectioned samples. This article compares XPS with surface Xray intensity measurements by EDS, carried out with a more realistic X-ray generation and absorption model, known as the phi(rho Z) model. The phi(rho Z) approach has been adopted together with Monte Carlo simulation for the proper selection of SEM accelerating voltages, in conjunction with the analysis of SEM morphological images for thin film density correction. The method discussed hereafter and compared with the XPS technique, has advantages of higher lateral resolution, non-destructive elemental analyses, morphological visualization, low cost, and faster performance. This methodology has been followed to verify the layered structure of SnO2/metal-based gas sensors. X-ray intensities were measured using an EDS ultra-thin window detector. Two different porous layers, 25-nm thick of SnO2 and 10-nm thick of Cu, were detected, showing better agreement with their nominal thickness compared to results obtained using XPS measurements where porosity affects XPS data. If confirmed to be reliable and as effective as XPS depth profiling, this technique may be adopted for process quality control purposes.