# Fileset

[gamma_SL.docx](https://mdr.nims.go.jp/filesets/918a379a-3d8c-4401-9f12-082b3ee29670/download)

## Creator

Yuji Kato, [Masataka Imura](https://orcid.org/0000-0002-4236-9549), [Yoshiko Nakayama](https://orcid.org/0000-0002-7607-6779), [Masaki Takeguchi](https://orcid.org/0000-0002-0282-6020), [Takayoshi Oshima](https://orcid.org/0000-0001-8550-9735)

## Rights

© 2019 The Japan Society of Applied Physics 
<br>
This is an author-created, un-copyedited version of an article accepted for publication in Applied Physics Express, Volume 12, Number 6. IOP Publishing Ltd is not responsible for any errors or omissions in this version of the manuscript or
any version derived from it. The Version of Record is available online at https://doi.org/10.7567/1882-0786/ab2196.[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Fabrication of coherent γ-Al2O3/Ga2O3 superlattices on MgAl2O4 substrates](https://mdr.nims.go.jp/datasets/1ebce2e6-dc2d-4558-b4e0-edced3c6e801)

## Fulltext

Template for APEX (Jan. 2014)Fabrication of coherent γ-Al2O3/Ga2O3 superlattices on MgAl2O4 substratesYuji Kato1, Masataka Imura2, Yoshiko Nakayama3, Masaki Takeguchi3, Takayoshi Oshima1,*1Department of Electrical and Electronic Engineering, Saga University, 1 Honjo, Saga, Saga 840-8502, Japan2National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan3NIMS, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, JapanE-mail: oshima@cc.saga-u.ac.jpAbstractWe succeeded in fabricating 10-period coherent γ-Al2O3/Ga2O3 superlattices (SLs) on MgAl2O4 substrates by molecular beam epitaxy. By varying the each layer thickness, we tuned the average Al composition (xave) of the coherent SLs from 0.20 to 0.85, and obtained nearly-lattice-matched SLs to the substrate at xave ~ 0.5. The lattice-matched SLs maintained coherent interfaces up to a period length of 7.2 nm in spite of a large lattice mismatch between the end members. Our results suggest lots of flexibility in designing γ-(AlxGa1−x)2O3-based heterostructures for future functional heterojunction devices.Behind the rapid progress of research and development of ultra-wide-band-gap semiconductor of Ga2O3, (AlxGa1−x)2O3 alloy and heterostructures have also gained lots of attentions recently.1,2) Among the existed polymorphs of Ga2O3,3–6) α, β, γ, ε-phases are verified that their (AlxGa1−x)2O3 heteroepitaxial films can be obtained on the proper substrates under the controlled growth conditions,7–14) whose facts pave the way to further studies regarding their respective heterostructures. In particular, development of β-(AlxGa1−x)2O3/Ga2O3 heterojunctions have preceded those of other phases owing to the thermodynamic stability4) and the use of high quality β-Ga2O3 single crystal substrates15). Such β- (AlxGa1−x)2O3/Ga2O3 heterojunctions10,11,16,17) including superlattices (SLs)18) were fabricated to understand the band-alignments,11) which are in good agreement with those extracted from theoretical calculation,19) and to confine two dimensional electron gas,16,17) which brought the realization of modulation-doped transistors20,21). In contrast to the β-phase, α-(AlxGa1−x)2O3/Ga2O3 heterojunctions22,23) and superlattices24,25) were prepared on isostructural sapphire substrates to clarify band-alignments22,23,25) and are ready for device applications. In contrast, γ- and ε-(AlxGa1−x)2O3-based heterostructures have rarely been studied.In the case of γ-(AlxGa1−x)2O3 alloy and heterojunctions, the last author (TO) and coworkers have performed some preliminary studies using epitaxial growth. The end members of the alloy, or γ-Ga2O3 and γ-Al2O3, exist in metastable form having a defective-spinel-type structure with a-lattice constants of aGa = 8.237605) and aAl = 7.911 Å26), respectively. To obtain such metastable films, we performed epitaxial stabilization using spinel-type-structured (001) MgAl2O4 substrate (aS = 8.0806 Å27)) for γ-Ga2O328) and γ-(AlxGa1−x)2O3 alloy12) films. By analyzing the epitaxial films, we found that the band gap of γ-(AlxGa1−x)2O3 can be controlled in a wide range (4.96–6.97 and 4.80–6.86 eV for direct and indirect transitions, respectively).12) In addition, we recently have prepared very thin coherent γ-Ga2O3 and γ-Al2O3 single layer films on MgAl2O4 substrates to reveal Poisson’s ratios of 0.31 and 0.28 and to estimate critical thicknesses (hC) of 5.3 and 4.0 nm, respectively.29) We also confirmed that the coherent γ-Ga2O3/Al2O3 heterojunction interface has a type-I band alignment with conduction- and valence-band-offsets of 1.6 and 0.2 eV, respectively.29) Moreover, we verified that heavy doping on γ-Ga2O3 is effective to a carrier concentration level as high as ~1019 cm−3.30) These fundamental investigation results make us expect future realization of coherent γ-(AlxGa1−x)2O3/Ga2O3 heterojunction devices. In this study, we further attempted to fabricate γ-(AlxGa1−x)2O3-based heterostructures as a successive study. Among possible heterostructures, we demonstrated the fabrication of γ-Al2O3/Ga2O3 SLs to attract attentions, because the γ-Al2O3/Ga2O3 structure is considered to be the most difficult to maintain high crystallinity due to the largest lattice mismatches to the substrate (1.9 and −2.1 % for γ-Ga2O3 and γ-Al2O3, respectively).10-period γ-Al2O3/Ga2O3 multilayers were prepared on (001) MgAl2O4 substrates by plasma-assisted molecular beam epitaxy (MBE). Before the growth, the substrate was degreased through ultrasonic cleaning in acetone and methanol. Epitaxial growth was carried out by supplying evaporated Ga and Al and oxygen radicals to the heated rotating substrate from conventional effusion cells and an inductively coupled plasma gun, respectively. The Ga and Al beam equivalent pressures were set at 1.75±0.05×10-7 and 3.08±0.10×10-7 Pa, respectively. While the plasma gun was operated under an input RF power of 200 W and an O2 flow rate of 0.50 sccm. The temperature of a SiC thermal diffusion plate, on which the substrate was mounted, was maintained at 575±5°C, which was monitored through a pyrometer. Under these growth conditions, growth rates of γ-Ga2O3 and γ-Al2O3 layers were 0.56±0.05 and 0.21±0.03 nm/min, respectively. By controlling the growth times for the comprised γ-Ga2O3 and γ‑Al2O3 layers, whose layer thicknesses are labeled as dGa and dAl, respectively, various multilayers (samples A–G) having different average Al compositions (xave) and period lengths (L = dGa + dAl) were prepared.To evaluate epitaxial and layered structures of the samples, X-ray diffraction (XRD) measurement was conducted using monochromatic Cu Kα1 radiation. We found that all the samples except for samples F and G exhibited satellite peaks in the θ-2θ scan patterns, indicating SL structures. For these SL structures, the measured θ-2θ scan patterns were analyzed based on theoretical calculation using models with the parameters (a period number, dGa, and dAl), which were optimized by simulation and curve fitting [See blue curves in Figs. 1 and 3]. In the simulation model, SLs were assumed to be coherent to the substrate and have deformed lattices, which can be defined using bulk lattice constants (aGa, aAl, aS), and the ratios of elastic stiffness tensors of γ-Ga2O3 and γ-Al2O3 (C11/C12 = 0.44 and 0.39, respectively)29). The sample labels and corresponding optimized structural parameters by the curve fitting were summarized in Table I. In this table, values for samples F and G were estimated from the growth rates and times because they were not SL. Note that xave is also extracted using dGa and dAl values using the following equations:where, aGa and aAl represent the strained lattice constants of γ-Ga2O3 (8.376 Å) or γ-Al2O3 (7.778 Å) films along the growth direction29).We first varied xave by controlling the fraction of dAl with maintaining L. Figure 1 shows XRD θ-2θ scan patterns near 004 reflections for the SL samples A, B, C with different xave. [See Table I for the corresponding dGa, dAl, and L]. According to xave, 2θ angle of the 0th satellite peaks shift around the substrate one. We note that a nearly lattice-matched SL to the substrate can be realized at xave = 0.48 (sample B), whose composition is very close to x = 0.47 of our previously reported nearly-lattice matched γ-(AlxGa1−x)2O3 alloy film.12) As for the crystallinity of these SLs, even 8 small peaks between the satellite ones are clearly observed, indicating relatively high-crystalline 10-perid SLs. The full widths at half maximums (FWHMs) of 004 XRD rocking curve of the samples A, B, and C were 304, 256, and 266 (0th) arcsec, respectively, as shown in Fig. 2 (a). We then controlled L under the nearly-lattice matched conditions. Figure 3 shows XRD θ-2θ scan patterns for the samples D, B, E, F, and G having different L. We can see satellite reflections for the samples with L = 3.4–7.2 nm (samples D, B, and E), where intervals between the neighboring ones become narrower accordingly as L increases. The crystallinity of these SLs are relatively high as well taking observable small sub-peaks between the satellite ones into consideration. FWHMs of 004 rocking curve for the samples D, B, and E were 478, 256, and 220 arcsec, respectively, as shown in Fig. 2(b). Note that the larger FWHM value for sample D is considered to be attributed to small intensity of peak signal owing to the small sample volume. Whereas, at L ≥ 10.7 nm (samples F and G), weak and broad reflections appear instead of satellite ones.  In particular, at L = 17.4 (sample G), two weak peaks are observed at the 2θ angles corresponding to γ-Ga2O3 and γ-Al2O3 single epitaxial films12) [Compare peak positions of XRD patterns of sample G with those of γ-Ga2O3 and γ-Al2O3 films], indicating that long-range lattice order no longer exists in the layered structure by lattice relaxation due to increased strain energy induced by thicker dGa, and dAl. All the SLs were verified to be coherent to the substrate. Figure 4 shows XRD reciprocal space maps taken near asymmetric 115 spots for the SL samples A, B, C, and E. For these SL samples, satellite spots of the SLs and the substrate one are aligned vertically at the same QX, indicating coherent SLs to the substrate. As for the sample D, we could not get a map with significant 115 spots from the SL because XRD intensity was too weak to detect owing to the small total thickness of the SL. However, it is highly likely that the SL also has structural coherence to the substrate because a cumulative strain in the SL (L = 3.4 nm) is considered as smaller than that of sample B (L = 5.8 nm). These results indicate that we succeeded in controlling the γ-Al2O3/Ga2O3 SL structure in wide xave and dGa ranges of at least 0.20  xave  0.85 and dGa  4.0 nm and dAl  3.2 nm (L = 7.2 nm) with maintaining coherent interfaces to the substrate, which well agrees with our previously estimated hC (5.3 and 4.0 nm for γ-Ga2O3 and γ-Al2O3 films, respectively)29). Finally, the crystal lattices of the one of the coherently grown SL (sample B) was observed by cross-sectional scanning transmission electron microscopy (STEM). For this purpose, a TEM specimen was prepared by conventional focused ion beam lift out by the following procedure. A 40 kV beam was employed to roughly dig the both sides of the target position near the center of the sample, followed by two step thinning process with a 30 and a 10 kV beams to obtain 70-nm thick lamella specimen, which was finally polished gently with a 5 kV beam for 60 s for each side to eliminate the damaged layer. The measurement was carried out under an acceleration voltage and a viewing direction of 200 kV and [100], respectively. Figure 5(a) shows low-magnification bright-field (BF) STEM image of the sample B. 10-period γ-Al2O3 (lighter contrast) and γ-Ga2O3 (darker contrast) strips are seen without any observable dislocations, indicating a relatively high crystalline quality of the SL. To view more detailed lattice images, magnified BF- and HAADF-STEM images for the same area near the SL/substrate interface are shown in Figs 5(b) and (c). The crystal lattices of the SL in both figures are well ordered but are curved at the γ-Al2O3/Ga2O3 interfaces, reflecting strain attributed to the large lattice mismatch. In particular, in the HAADF image, we can clearly see that cross-stripes, which correspond to the network of heavier metal atoms of the spinel framework [Compare the cross-strips in Fig. 5(c) and Ga atom positions in (d)], maintain the coherency from MgAl2O4 to the upper SLs with some distortion at the interfaces. This result indicates that these heterojunctions (γ-Ga2O3/MgAl2O4 and γ-Al2O3/Ga2O3) had coherent interfaces without misfit dislocation. In summary, γ-Al2O3/Ga2O3 SLs were successfully fabricated on MgAl2O4 substrates by MBE despite the metastability and the large lattice mismatch. The average lattice constant as well as xave of the SL were able to be controlled with tuning dAl and dGa, and the nearly lattice-matched SL was obtained at xave ~ 0.5. XRD reciprocal space map and STEM characterization revealed coherent interfaces without misfit dislocation. Considering that γ-Al2O3 and γ-Ga2O3 are end members of γ-(AlxGa1−x)2O3 alloy, our results also suggest that high quality γ-(AlxGa1−x)2O3/Ga2O3 heterojunctions including SLs with entire x can be fabricated as well. Such outlook regarding the wide controllability of high quality γ-(AlxGa1−x)2O3-based heterostructures will trigger the successive studies on characterization and device application in the future.AcknowledgmentsWe would like to thank NIMS Microstructural Characterization Platform as a program of the “Nanotechnology Platform” of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan for STEM observation. This study was partially supported by the Murata Science Foundation.References1) M. Higashiwaki and G.H. Jessen, Appl. Phys. Lett. 112, 060401 (2018).2) S.J. Pearton, J. Yang, P.H. Cary, F. Ren, J. Kim, M.J. Tadjer, and M.A. Mastro, Appl. Phys. Rev. 5, 011301 (2018).3) M.A. Mastro, A. Kuramata, J. Calkins, J. Kim, F. Ren, and S.J. Pearton, ECS J. Solid State Sci. Technol. 6, P356 (2017).4) R. Roy, V.G. Hill, and E.F. Osborn, J. Amer. Chem. Soc. 74, 719 (1952).5) H.Y. Playford, A.C. Hannon, E.R. Barney, and R.I. Walton, Chem. - A Eur. J. 19, 2803 (2013).6) H. von Wenckstern, Adv. Electron. Mater. 3, 1600350 (2017).7) R. Kumaran, T. Tiedje, S.E. Webster, S. Penson, and W. Li, Opt. Lett. 35, 3793 (2010).8) H. Ito, K. Kaneko, and S. Fujita, Jpn. J. Appl. Phys. 51, 100207 (2012).9) G.T. Dang, T. Yasuoka, Y. Tagashira, T. Tadokoro, W. Theiss, and T. Kawaharamura, Appl. Phys. Lett. 113, 062102 (2018).10) T. Oshima, T. Okuno, N. Arai, Y. Kobayashi, and S. Fujita, Jpn. J. Appl. Phys. 48, 070202 (2009).11) R. Wakabayashi, M. Hattori, K. Yoshimatsu, K. Horiba, H. Kumigashira, and A. Ohtomo, Appl. Phys. Lett. 112, 232103 (2018).12) T. Oshima, Y. Kato, M. Oda, T. Hitora, and M. Kasu, Appl. Phys. Express 10, 051104 (2017).13) H. Nishinaka, D. Tahara, and M. Yoshimoto, Jpn. J. Appl. Phys. 55, 1202BC (2016).14) V. Gottschalch, S. Merker, S. Blaurock, M. Kneiß, U. Teschner, M. Grundmann, and H. Krautscheid, J. Cryst. Growth 510, 76 (2019).15) A. Kuramata, K. Koshi, S. Watanabe, Y. Yamaoka, T. Masui, and S. Yamakoshi, Jpn. J. Appl. Phys. 55, 1202A2 (2016).16) T. Oshima, Y. Kato, N. Kawano, A. Kuramata, S. Yamakoshi, S. Fujita, T. Oishi, and M. Kasu, Appl. Phys. Express 10, 035701 (2017).17) Y. Zhang, A. Neal, Z. Xia, C. Joishi, J.M. Johnson, Y. Zheng, S. Bajaj, M. Brenner, D. Dorsey, K. Chabak, G. Jessen, J. Hwang, S. Mou, J.P. Heremans, and S. Rajan, Appl. Phys. Lett. 112, 173502 (2018).18) S.W. Kaun, F. Wu, and J.S. Speck, J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 33, 041508 (2015).19) H. Peelaers, J.B. Varley, J.S. Speck, and C.G. Van de Walle, Appl. Phys. Lett. 112, 242101 (2018).20) S. Krishnamoorthy, Z. Xia, C. Joishi, Y. Zhang, J. McGlone, J. Johnson, M. Brenner, A.R. Arehart, J. Hwang, S. Lodha, and S. Rajan, Appl. Phys. Lett. 111, 023502 (2017).21) E. Ahmadi, O.S. Koksaldi, X. Zheng, T. Mates, Y. Oshima, U.K. Mishra, and J.S. Speck, Appl. Phys. Express 10, 071101 (2017).22) K. Kaneko, K. Suzuki, Y. Ito, and S. Fujita, J. Cryst. Growth 436, 150 (2016).23) T. Uchida, R. Jinno, S. Takemoto, K. Kaneko, and S. Fujita, Jpn. J. Appl. Phys. 57, 040314 (2018).24) T. Oshima, Y. Kato, M. Imura, Y. Nakayama, and M. Takeguchi, Appl. Phys. Express 11, 065501 (2018).25) T. Oshima, Y. Kato, E. Kobayashi, and K. Takahashi, Jpn. J. Appl. Phys. 57, 080308 (2018).26) R. ‐S Zhou and R.L. Snyder, Acta Crystallogr. Sect. B Struct. Sci. 47, 617 (1991).27) T. Yamanaka, Y. Takéuchi, and M. Tokonami, Acta Crystallogr. Sect. B Struct. Sci. 40, 96 (1984).28) T. Oshima, T. Nakazono, A. Mukai, and A. Ohtomo, J. Cryst. Growth 359, 60 (2012).29) T. Oshima, Y. Kato, E. Magome, E. Kobayashi, and K. Takahashi, Submitted (2019).30) T. Oshima, K. Matsuyama, K. Yoshimatsu, and A. Ohtomo, J. Cryst. Growth 421, 23 (2015).FiguresFigure 1. XRD θ-2θ scan patterns (red) of samples A, B, and C with various xave. Simulated XRD patterns (blue) are also plotted, which are calculated under the assumption that grown films are coherent to the substrates.Figure 2. XRD ω-rocking curves of 004 reflection for the samples showing (a) xave dependence (samples A, B, and C) and (b) L dependence (samples D, B, and E).  The curves correspond to 0th (sample A and C) and –1th (samples B, D, and E) reflections.Figure 3. Experimental (red) and simulated (blue) XRD θ-2θ scan patterns of samples D, B, E, F, and G with nearly matched lattices to the substrate and various L. The patterns of γ-Ga2O3 and γ-Al2O3 epitaxial single layers (green) are also plotted as references.12) Figure 4. XRD reciprocal space maps in vicinity of 115 spots of samples A, B, C, and E. Directions of QZ and QX are along [001] and [110], respectively.Figure 5. (a) and (b) BF- and (c) HAADF-STEM images of the SL sample B with incident electron beam direction along [100]. (b) and (c) are magnified images taken from the same area near the substrate. (d) Unit cell of γ‑Ga2O3 viewing from the same direction as the STEM.Table I. Summary of each layer thickness and xave for the samples. Sample dGa (nm) dAl (nm) L(nm) xave A 0.8 4.2 5.0 0.85 B 3.1 2.7 5.8 0.48 C 4.7 1.1 5.8 0.20 D 2.0 1.4 3.4 0.44 E 4.0 3.2 7.2 0.48 F 6.2 (est.) 4.5 (est.) 10.7 (est.) 0.44 (est.) G 9.3 (est.) 8.1 (est.) 17.4 (est.) 0.48 (est.)6image3.emf40 42 44 46 48 50 17.4G-3+3FEBSub.-2-1+2+10-2-1+2+10+1-1Intensity (arb. units)2 (deg)0004D 10.7   7.2   5.8L (nm)= 3.4γ-Al2O3γ-Ga2O3image4.pngimage5.pngimage1.emf40 42 44 46 48 50CBsub.-2-1+2+10-2-1+2+10+1-1Intensity (arb. units)2 (deg)0004A   xave= 0.85   0.48   0.20image2.emf-0.5 0.0 0.50.00.51.0Normalized Intensity(deg) A B C-0.5 0.0 0.5  D  B   E(a) (b)