ICMCTF2019 Session C2-WeA: Novel Oxide Films for Active Devices

Beta-phase gallium oxide (β-Ga₂O₃) is an ultra-wide bandgap semiconductor that holds great promise for future power electronics due to its ease of bulk crystal growth and facile ability to extrinsically dope n-type over a wide range of dopant concentrations. Power devices often operate at elevated temperature with large surge currents and high blocking voltages. Reliable operation under these aggressive conditions requires that the metal and dielectric interfaces, as well as the doping profiles within the device, be highly stable under thermal and electrical stress, including thermal cycling. Obtaining stable materials interfaces is a particular challenge for oxide semiconductors, due to the ready supply of oxygen within the bulk, and to thermodynamic competition amongst adjacent metal atoms to oxidize or reduce. This property has been exploited in other oxide devices, for instance in oxygen-vacancy based resistive memory (ReRAM). However such behavior must be strictly avoided in power devices to ensure reliable and stable operation.

In this talk, I will present our experimental investigations on the stability of dielectric and contact electrode interfaces to beta-phase gallium oxide, using comprehensive materials characterization and electrical measurements. We study novel dielectrics that can be used as gate insulators in field-effect transistors, and for field plates and passivation layers. We also investigate metal contacts to bulk Ga₂O₃ with a variety of dopants and doping levels, to determine pathways to form low-resistance and stable ohmic contacts. The experimental results will be compared to thermodynamic predictions based on Gibbs’ free energies of reactions. Using these results, we will identify and discuss strategies for designing stable dielectric and contact interfaces in order to enable high-performance Ga₂O₃ power electronics.

Ga₂O₃ has emerged as a promising material for next generation power electronics and UV photodetectors applications due to its large bandgap (4.9 eV) and the availability of affordable native substrates from melt-grown bulk crystals. While β-Ga₂O₃ (monoclinic) is the most stable and studied of five Ga₂O₃ polymorphs, the slightly less energetically favorable α- and ε-Ga₂O₃ phases have unique characteristics that can be exploited. The α -Ga₂O₃ (rhombohedral corundum) has the largest bandgap of 5.3 eV and can be alloyed with α-Al₂O₃ and α-In₂O₃ for bandgap engineering. T he ε-Ga₂O₃ phase (hexagonal wurtzite) is a polar phase, with a calculated polarization strength that is 10 and 3 times larger than that of GaN and AlN, respectively. Like the III-N system, polarization induced charges can lead to higher charge densities and mobilities in two-dimensional electron gases formed at heterojunctions, which would improve the viability of Ga₂O₃ electronic devices. In this work, we use atomic layer epitaxy (ALEp) to produce high-quality heteroepitaxial Ga₂O₃ films and investigate phase selectivity as a function of substrate type and orientation, growth temperature (T_g), plasma gas phase chemistry and gas pressure.

β-Ga₂O₃ is a wide bandgap semiconductor that has the potential for high-power device applications. While heterostructures of thin film β-Ga₂O₃ grown on various materials such as GaN^1,2 and sapphire^3,4 have been reported, the understanding of direct wafer bonding β-Ga₂O₃ to form heterostructures is still limited. The benefits of wafer bonding β-Ga₂O₃ would be two-fold: (1) enables even more novel materials combinations to be realized and (2) other unprecedented orientations of β-Ga₂O₃ could be integrated with various materials. This last point is important for β-Ga₂O₃ due to its anisotropic properties especially since β-Ga₂O₃ exhibits the low-symmetry monoclinic crystal structure.

Currently there are reports of mechanically exfoliating β-Ga₂O₃ along cleavage planes parallel to the (100) and (001) planes.^5,6 However, this approach will be challenging to integrate in large-scale processing. Demonstrated with Si,^7,8 III-V’s,^9-12 and CdZnTe,¹³ the “SMART-Cut” method¹⁴ is more compatible with large-scale processing. In this approach, ions are implanted in a handle substrate and subsequently annealed. Under the appropriate processing conditions, the implanted ions diffuse and agglomerate into bubbles during annealing and surface blistering, accompanied by exfoliation, can occur. In this work, (010) β-Ga₂O₃ substrates are implanted with hydrogen ions and subsequently annealed. Both high-resolution X-ray diffraction (XRD) measurements and transmission electron microscopy (TEM) images were employed to monitor the structural evolution with annealing. Symmetric ω:2θ XRD scans showed that annealing at 150 °C and 300 °C did not change the implantation-induced strain appreciably; while annealing at 500 °C for 1 hour was sufficient in removing this strain. Additionally, TEM images showed subsurface hydrogen bubble formation. Despite implanting along a non-cleavage plane, ion implantation for exfoliation provides a promising pathway for obtaining thin layers of β-Ga₂O₃ with orientations other than (100) and (001).

References

1.A. Kalra, et al., Appl. Phys. Ex. 11, 2018

2.P. Li, et al., J. Mat. Chem. C. 5, 2017

3.S. Ghose, et al., J. Appl. Phys. 122, 2017

4.V.I. Nikolaev, et al., Mat. Sci. Semi. Proc. 47, 2016

5.Y. Kwon, et al., Appl. Phys. Lett. 110, 2017

6.M.J. Tadjer, et al., ECS 1233 2017

7.C.M. Varma, Appl. Phys. Lett. 71, 1997

8.C. Miclaus, et al., J. Phys. D: Appl. Phys. 36, 2003

9.S. Hayashi, et al., Appl. Phys. Lett. 85, 2004

10.S. Hayashi, et al., J. Electro. Soc. 153, 2006

11.S. Hayashi, et al., J. Electro. Soc. 154, 2007

12.E. Padilla, et al., ECS Trans. 33, 2010

13.C. Miclaus, et al., J. Elec. Mat. 34, 2005

14.M. Bruel, et al., Jpn. J. Appl. Phys. 36, 1997

In this study, we focused on the piezoelectric coefficients (d₃₃) depended on doping rare earth element, yttrium (Y), into ZnO thin films. Pure ZnO and Y doped ZnO (Y:ZnO) thin films with different yttrium contents were synthesized on p-type Si (111) substrates via RF magnetron sputtering, and the thickness of these thin films were fixed at 650 nm. In X-ray diffraction (XRD) patterns, Y:ZnO films with low yttrium contents (<3 at%) showed preferential orientation of (002) plane and still maintained columnar wurtzite structures. The positions of diffraction peaks shifted to lower angle as the concentration of yttrium increased due to the substitution of smaller host ions (Zn²⁺: 0.74Å) by larger dopant ions (Y³⁺: 1.04Å). When the concentration of yttrium is over 3 at%, the drastic transformation of morphology was observed in scanning electron microscopy (SEM) images, which consisted with the results of XRD analysis. This phenomenon caused by the lattice distortion, owing to the larger radii of Y³⁺ ions. The piezoelectric coefficients of pure ZnO and Y:ZnO thin films were measured by piezoresponse force microscopy (PFM) as well. The piezoelectric coefficients had achieved higher values (d₃₃ ~ 86.8 pm/V) when the contents of yttrium was 3 at%, which compared to the theoretical values of pure ZnO thin films (d₃₃ = 12.4 pm/V). Therefore, Y doped ZnO thin film was regarded as the promising candidate for piezoelectric nanogenerators (NGs).