Stability and residual stresses of sputtered wurtzite AlScN thin films

Scandium-alloying of aluminum nitride (AlScN) enhances the piezoelectric properties of the material and increases the performance of piezoelectric microelectromechanical systems (MEMS). However, this enhancement is caused by the destabilization of the wurtzite phase and so far the stability of AlScN thin films has not been sufficiently studied. Stability is especially important for piezoelectric devices because changes to the film microstructure or residual stress can lead to drastic changes in the device behavior. The stability of AlScN is investigated by annealing sputtered films and characterizing the resulting changes. It is found that the wurtzite phase of ${\mathrm{Al}}_{0.7}{\mathrm{Sc}}_{0.3}\mathrm{N}$ is stable at least up to $1000{}^{\circ }\mathrm{C}$ and annealing increases the crystal quality, reaching a maximum at $800{}^{\circ }\mathrm{C}$. When annealed for more than 100 h at $1000{}^{\circ }\mathrm{C}$, argon used in sputtering segregates into the grain boundaries and causes compressive strains and formation of rock-salt phase. Additionally, annealing at $1000{}^{\circ }\mathrm{C}$ for 5 h reduces the average tensile stress by approximately 1 GPa.

Figure 1

Stress evaluation by combining out-of-plane and in-plane strains using the ${sin}^2\psi$ method.

Figure 2

Simulated lattice constants for AlScN as a function of Sc-fraction $x$. Solid symbols and error bars represent average values and standard deviation, respectively. Dashed lines give a polynomial fit to the data, quadratic for $a$ and cubic for $c$.

Figure 3

SEM micrographs of the surfaces of as-deposited and annealed films. (Contrast and brightness enhanced.) (a) As-deposited film and (b) Film annealed at $1000{}^{\circ }\mathrm{C}$ for 5 h.

Figure 4

The amplitude of the piezoelectric response of the as-deposited and annealed AlScN films. (a) As-deposited film and (b) Film annealed at $1000{}^{\circ }\mathrm{C}$ for 5 h.

Figure 5

Measured $\chi –2\theta$ x-ray diffraction maps of as-deposited and annealed films. The measured intensity is normalized to the Si113 peak in each map and a logarithmic scale is used to present the intensity. AlScN reflections are labeled in white and substrate reflections in black. (a) As-deposited film and (b) Film annealed at $1000{}^{\circ }\mathrm{C}$ for 5 h.

Figure 6

XRD results for as-deposited film and films annealed at $400{}^{\circ }\mathrm{C}–1000{}^{\circ }\mathrm{C}$ for 5 h. The square root of the intensity is used for visualization purposes. (a) Measured 2θ–ω diffractograms of the 0002 reflection and (b) X-ray rocking curves about the 0002 reflection.

Figure 7

TEM analysis of AlScN films before (left) and after annealing at $1000{}^{\circ }\mathrm{C}$ for 10 h (right). (a), (b) Cross-sectional STEM micrographs. (c), (d) SAED patterns. The AlScN spots are labeled in white and the Si spots in red. Contrast and brightness enhanced. (e) Cross-sectional STEM micrograph of a grain at the surface of the AlScN film after annealing.

Figure 8

XRD results of film annealed at $1000{}^{\circ }\mathrm{C}$ for 300 h. (a) A $\chi –2\theta$ diffraction map. Intensity is normalized to Si113 peak and presented using a logarithmic scale. AlScN reflections in white and Si in black. (b) Out-of-plane XRD of the AlScN 0002 reflection. The three components $f_{1-3}$ of the pseudo-Voigt sum are shown as orange dashed lines. (Only every fifth measurement plotted for clarity.) (c) In-plane XRD of the AlScN $10i0, 11i0$ and $20i0$ reflections. New reflections are highlighted with arrows. (Only every fortieth measurement plotted for clarity.) The vertical axis in the diffractograms presents the square root of the intensity for visualization purposes.

Figure 9

TEM results of AlScN film after annealing at $1000{}^{\circ }\mathrm{C}$ for 300 h. (a) SAED pattern, the selected area is near the film-substrate interface, shown in the inset. New diffraction spots next to 0002, $10i0$, and $10i2$ are highlighted with arrows, (b) EDS map of the Ar K-line intensity, showing the excess Ar concentration near the film-substrate interface, and (c) BF STEM micrograph for EDS mapping in panel (b).

Figure 10

FTIR absorbance spectra for the as-deposited film and films annealed at $1000{}^{\circ }\mathrm{C}$. Intensity presented as a square root for visualization purposes.

Figure 11

Schematic of Ar segregation-induced strain in AlScN crystallites in the small crystallite and columnar growth regions.