edfas.org 19 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 1 Thermally grown interfacial layers, dopant diffusion, interfacial characterization, and the separate processes leading to element segregation in the material can be analyzed with higher accuracy using APT than other methods in this context, meaning that important conclusions can be drawn from experiments such as these, which ultimately leads to improved device function. ELEMENTAL CLUSTERING AND PROXIMITY HISTOGRAMS APT can also offer insight into the size and composition of extremely small clusters hidden within layers of a material or device. Figure 6 shows an example of GeTe-based thermoelectric device analysis using APT, with the authors ultimately linking Ga-rich cluster formation to changes in the electron and phonon transport.[15] The authors produced a sandwiched Ga2Te3 sample (Fig. 6b) and after APT analysis, separate atom maps for the elements present in the GeTe section of the tip (Fig. 6a) reveal Ga-rich clusters. APT analysis also offers the ability to integrate concentrations across arbitrarily shaped 2D interfaces defined by a particular element or molecular concentration to create what is called a proximity histogram. Here, a single Ga cluster was selected in the APT reconstruction, visualized to show the cluster size (Fig. 6c), and a proximity histogram of the cluster was used to show an in- crease of Ga concentration over the surrounding area (Fig. 6d). Ultimately, APT can show complex 3D structures with accurate composition analysis that can tie back to the overall device performance. OUTLOOK Recent advances involving shorter wavelength light sources have resulted in improved analysis of a broad range of materials that have historically been difficult to measure with conventional near-UV APT methods, such as wide-bandgap insulators. Commercial systems are now utilizing wavelengths in the deep-UV region of ~257 nm, while Fig. 6 APT reconstruction data of a GeTe sample showing (a) separate ion distributions in the reconstructed APT tip, (b) scanning electron microscope image of the tip, (c) 3D distribution of a selected Ga cluster in the APT data, and (d) proximity histogram calculated by integrating the composition from a 1.5 at.% Ga isoconcentration surface (x = 0 nm, not shown) around the Ga-rich cluster, showing changes in composition upon crossing the cluster surface. Reproduced from Ref 15 under the terms of the Creative Commons Attribution License. Fig. 5 (a) An APT reconstruction of a poly-Si MOSFET device, (b) a 1D top-down composition profile through the device layers, and (c) a 1D top-down composition profile of an unpatterned test zone from a blanket wafer subjected to the same silicide process as b). Adapted with permission from Ref 14. (a) (b) (c) (a) (b) (c) (d)
RkJQdWJsaXNoZXIy MTYyMzk3NQ==