Wide band gap oxide semiconductors are basically “green” materials constituted with abundant and safe elements.  Recent evolution has been brought by gallium oxide (Ga₂O₃) semiconductors, which possess wide band gap of about 5 eV and thus are promising for high-breakdown voltage power devices.  Being supported by ideal homoepitaxial growth by MBE or HVPE on thermally-stable orthorhombic β-Ga₂O₃ substrates, excellent properties of MOSFETs and Schottky barrier diodes (SBDs) with high breakdown voltage as well as high temperature operation have been reported by collaboration of NICT, Tamura Co., Koha Co., and Tokyo University of Agricultural and Technology[1-3].  On the other hand, our interest is to develop safe, inexpensive, and environmental-friendly growth technology.  This is the mist CVD technology, where we prepare water and/or alcohol solution of safe chemicals containing the target metal element (for Ga₂O₃, such as gallium chloride or gallium acetylacetonate) and the mist particles generated by ultrasonic atomization of the source solution are transferred to the growth chamber to cause oxidation reaction forming Ga₂O₃ on a substrate[4-6].  This technology allowed the growth of highly crystalline-quality corundum-structured Ga₂O₃ (α-Ga₂O₃) on sapphire substrates[5,6] which possess corundum structure, followed by the band gap tuning with α-(Al,Ga)₂O₃ and α-(In,Ga)₂O₃ alloys from 3.8 to 8.8 eV.  In this presentation, the author reports the efforts for the property and conductivity control of Ga₂O₃-based single crystalline semiconductors by the green synthesis technology of mist CVD as well as the promising potential of these materials for devices leading green innovation.

     Unintentionally-doped (UID) α-Ga₂O₃ showed unmeasurable high resistivity.  Sn doping is a promising technology for conductivity control.  With improvements of the crystal quality, n-type conductivity has been controlled in the range of 10¹⁷-10¹⁹ cm^-3 with the Hall mobility of upto 24 cm²/Vs.  Test MOSFETs were fabricated and showed modulation of drain current by the gate voltage, but the mutual conductance was too small to be actually utilized and the off-operation was not performed.  This was partly because the activation of donors is significantly affected by dislocation defects originating by heteroepitaxy on sapphire.  Our efforts are directing to superlattice buffer layers and epitaxial later overgrowth (ELOG), and will be introduced at the meeting.  Recently, SBDs showing very low on-resistance and high breakdown voltage of 0.1 mW×cm² and 531 V or 0.4 mW×cm² and 855 V, respectively, were reposted by FLOSFIA Inc. (Kyoto, Japan) using Sn-doped α-Ga₂O₃ grown by mist CVD[7], and this clearly indicates the potential of α-Ga₂O₃ for future green innovation.

     Besides α-Ga₂O₃, mist CVD has contributed the evolution of various wide band gap oxide semiconductors.  One of the examples is corundum-structured α-In₂O₃, whose band gap is 3.8 eV, showing the x-ray diffraction (0006) rocking curve FWHM of as small as 185 arcsec and the electron Hall mobility of as high as 130 cm²/Vs.  An α-In₂O₃ MOSFET demonstrated the modulation of drain current in terms of the gate voltage from -50 to +5 V with the on-off ratio of >10⁶, while the gate leakage current was as small as <100 pA. Off-state breakdown voltage tends to be 70 V at the present stage.  This was the clear demonstration that corundum-structured oxide semiconductors can work as electron devices.  Another example is tin oxide, which can be either n-type (SnO₂) or p-type (SnO) by controlling the growth conditions of mist CVD so that oxidation reaction is enhanced or reduced, respectively, by adding oxidizing or reducing agent.  This is also an attractive advantage of mist CVD, where reaction atmosphere can be chemically controlled.

     It is the responsibility of us to promote green innovation by green materials synthesized by green technology.  The combination of mist CVD and wide band gap oxide semiconductors can do, we hope.

[1] M. Higashiwaki et al., Appl. Phys. Lett. 103, 123511 (2013).

[2] K. Sasaki et al., IEEE Electron Device Lett. 34, 493 (2013).

[3] M. Higashiwaki et al., Conference Digest; 73rd Device Research Conference, 2015, p.29.

[4] S. Fujita et al., J. Cryst. Growth 401, 588 (2014).

[5] D. Shinohara et al., Jpn. J. Appl. Phys. 47, 7311 (2008).

[6] K. Kaneko et al., Jpn. J. Appl. Phys. 51, 020201 (2012).

[7] M. Oda et al., presented at Int. Workshop Gallium Oxide and Related Materials, Kyoto, Japan (2015).