Exploring Phase Control in Sc_xAl_x-1N Heterostructures Grown by Molecular Beam Epitaxy Free

The unique properties of Sc_xAl_x-1N thin films have attracted significant attention for applications in next-generation solid-state and acoustoelectric RF devices [1]. Sc_xAl_x-1N is formed by adding cubic ScN, to wurtzite AlN, which both softens the bonding and increases the polarization compared to AlN. High Sc compositions, x = 0.43, exhibit a fivefold increase in piezoresponse compared to pure AlN. This makes Sc_xAl_x-1N an extremely attractive candidate for future high frequency broadband filters [2]. Sputtered polycrystalline Sc_xAl_x-1N films with x > 0.27 have been shown to exhibit ferroelectricity, providing evidence that Sc_xAl_x-1N is the first experimentally demonstrated nitride ferroelectric and has led to renewed interest in ferroelectric non-volatile memory [3]. Controlling the phase of epitaxially grown Sc_xAl_x-1N is challenging due to the interplay of misfit strain and Sc incorporation during growth, both which affect the hexagonal/cubic phase boundary in the Sc_xAl_x-1N alloy phase diagram. Transmission electron microscopy analysis is ideally suited to examine the local structural morphologic, defect, and phase/growth relationships resulting from film epitaxial growth.

In this work, we examine the structure of Sc_xAl_x-1N films grown via molecular beam epitaxy with in situ reflection high-energy electron diffraction and post-growth cross-sectional aberration-corrected scanning transmission electron microscopy (STEM). Experimental HAADF and EELS measurements were acquired with an aberration-corrected Nion UltraSTEM 200X operated at 200 kV, with a convergence angle of 30 mrad, equipped with a MerlinEELS direct electron counting detector. Initial attempts to grow Sc_0.40Al_0.60N thin films resulted in the presence of rock-salt grains near the film nucleation layer. By introducing a compositionally graded layer starting with Sc_0.32Al_0.68N before the Sc_0.40Al_0.60N layer we were able to suppress the formation of rock-salt grains and reduce the density anomalously oriented grains [4]. Continuing work is combining 4D-STEM strain analysis with STEM-EELS in order to examine the nature of strain and chemical bonding in these heterostructures.

References