Microstructural Analysis of TaAlN Nitride MAX Phases by Atom Probe Tomography

This is the authors' abstract. We don't add key points for this paper.

Nitride MAX phases are a promising class of functional materials for extreme-temperature applications due to their unique combination of thermal stability, oxidation resistance, and mechanical strength at high temperatures. These materials belong to the broader MAX phase family (Mn+1AXn phases), where M is a transition metal, A is a group 13 or 14 element, and X is carbon or nitrogen [1]. In this study we investigate functional TaAlN nitride MAX phases, with an emphasis on their nanostructural analysis by atom probe tomography (APT). TaAlN structures were grown by molecular beam epitaxy (MBE) on a 76 mm (3 inch) diameter semi-insulating 4H-SiC substrate. The grown sample consists of an epitaxial hexagonal Ta2N layer, nominally 10 nm thick, followed by a thicker TaAlN layer, which was intentionally grown Al-rich to ensure formation of the desired hexagonal phase. The heater thermocouple temperature was 1100 °C during the Ta2N growth and 850 °C during the TaAlN growth, a change that was necessary because the actual substrate temperature increases substantially during metal film growth on the transparent SiC substrate [2], thereby potentially influencing the Al sticking and growth of the subsequent Al-containing layer. The grown structures were analyzed in situ by reflection high-energy electron diffraction (RHEED) and post-growth by transmission electron microscopy (TEM), and atom probe tomography (APT). In-situ RHEED (results not shown) indicate that the Ta2N and TaAlN layers are hexagonal and epitaxial, with the TaAlN layer having a substantially larger lattice parameter than the 4H-SiC substrate, indicating that the film is AlN-rich. Cross-sectional TEM analysis, Figure 1, reveals sharp interfaces between the 4H-SiC and Ta2N layer, and between the Ta2N layer and the TaAlN layer. The Ta2N layer is 6.6 nm thick, somewhat thinner than the intended 10 nm thickness, with a measured growth rate of 0.168 Å/s, a reflection of the difficulty in controlling the Ta flux from the in-situ e-beam evaporator operating in open-loop conditions. The TaAlN film is 143 nm thick, substantially thicker than initially intended which assumed substantial Al desorption, with little Al desorbing from the surface before incorporation into the film. The measured 0.681 Å/s growth rate suggests that nearly all of the incident Al is incorporated into the film, resulting in an estimated Al:Ta ratio close to 3:1 rather than the ultimately desired ratio of 1:2 for an ideal Ta2AlN layer. APT analysis of the structure is shown in Figure 1, accompanied by a proximity histogram [2] displaying a composition profile through all three layers of the structure. APT confirms that the top TaAlN layer is enriched in both Al and N, consistent with the in situ RHEED analyses, with a measured composition of 45.8Al-41.5N-12.5Ta (at.%) corresponding to a 3.7 Al:Ta ratio. Trace amounts of O, B, and C impurities (each less than 0.1 at.%) were also detected. The APT analysis extended through the underlying Ta2N film, which had a measured composition of 49.8Ta-48.6N-1.2Al (at.%), again with B, C, and O impurities. The nominally Ta2N middle layer is thus closer to TaN and depleted in Ta, like the top TaAlN film. The bottom 4H-SiC substrate had a measured composition of 60.4Si-38.3C (at.%) with the apparent depletion in C likely due to well-known difficulties in accurately measuring C in Si by APT due to “multiple ion events,” wherein two or more ionic species strike the APT detector nearly simultaneously [4]. Such artifacts are also prevalent in the TaAlN and Ta2N phases themselves, each displaying a high propensity (∼40%) for multiple events and for molecular evaporation, similar to those reported previously in Ti2AlC MAX phases analyzed by APT [5]. These phenomena can lead to biases in detecting certain atomic species and hence inaccurate compositional measurements. We will discuss the effects of APT analysis conditions (laser pulse energy and specimen temperature) to minimize these artifacts in the analysis of these and similar MAX phases [6]. Cross-sectional TEM analysis of the grown TaAlN/Ta2N/4H-SiC structure. APT reconstruction of the grown TaAlN/Ta2N/4H-SiC structure. Only Al, Ta, and C atoms are shown for clarity and the interfaces are indicated by 30 at.% Al (blue) and 30 at.% C (black) isoconcentration surfaces. The accompanying proximity histogram, computed from the 30 at.% Al isoconcentration surface, displays the compositions of the TaAlN and Ta2N layers and underlying 4H-SiC substrate.