This document is the Accepted Manuscript version of a Published Work that appeared in final form in Crystal Growth and Design, copyright © American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see: https://dx.doi.org/10.1021/acs.cgd.9b01669. Page 1 of 42 Crystal Growth & Design 1 2 3 4 5 6 7 P-type ultra-wide bandgap spinel ZnGa2O4: new perspectives for 8 9 10 energy electronics 11 12 13 14 15 Ekaterine Chikoidze*1, Corinne Sartel1, Ismail Madaci1, Hagar Mohamed1,2, Christele Vilar1, Belén 16 17 18 Ballesteros3, Francisco Belarre3, Elena del Corro3, Pablo Vales-Castro3, Guillaume Sauthier3, Lijie Li4, 19 20 Mike Jennings4, Vincent Sallet1, Yves Dumont1 and Amador Pérez-Tomás*3 21 22 23 24 1 Groupe d’Etude de la Matière Condensée (GEMaC), Université de Versailles Saint Quentin en Y. – 25 26 27 CNRS, Université Paris-Saclay, 45 Av. des Etats-Unis, 78035 Versailles Cedex, France. 28 29 2 Solid State Physics Department, National Research Center, El-Behooth St. 12311 Dokki, Giza, Egypt. 30 31 3 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of 32 33 34 Science and Technology, Barcelona, Spain. 35 4 36 College of Engineering, Swansea University, Bay Campus, Fabian Way, Crymlyn Burrows, Swansea, 37 38 SA1 8EN, UK. 39 40 41 42 43 e-mail: ekaterine.chikoidze@uvsq.fr; amador.perez@icn2.cat 44 45 46 47 Keywords: Ultra-wide Bandgap Semiconductor, Spinel Oxides, ZnGa2O4, p-type Conductivity 48 49 50 51 52 53 54 55 56 57 58 1 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 2 of 42 1 2 3 4 5 6 Abstract 7 8 9 10 The family of spinel compounds is a large and important class of multi-functional materials of general 11 12 13 formulation AB2X4 with many advanced applications in energy and optoelectronic areas such fuel cells, 14 15 batteries, catalysis, photonics, spintronics or thermoelectricity. In this work, it is demonstrated that the 16 17 ternary ultra-wide band gap (~5 eV) spinel zinc gallate (ZnGa2O4) arguably is the native p-type ternary 18 19 20 oxide semiconductor with the largest Eg, (in pair with recently discovered binary p-type monoclinic - 21 22 Ga2O3 oxide). For nominally undoped ZnGa2O4 high temperature Hall effect hole concentration was 23 24 determined to be as large as p = 2×1015 cm-3 while hole mobilities were found to be µh = 7-10 cm2/Vs (in 25 26 27 the 680-850 K temperature range). An acceptor-like small Fermi level was further corroborated by x-ray 28 29 spectroscopy and by density functional theory calculations. Our findings, as an important step towards p- 30 31 type doping, opens up further perspectives for ultra-wide bandgap bipolar spinel electronics and further 32 33 promotes ultra-wide bandgap ternary oxides such as ZnGa2O4 to the forefront of the quest of the next 34 35 36 generation of semiconductor materials for more efficient energy optoelectronics and power electronics. 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 2 59 60 ACS Paragon Plus Environment Page 3 of 42 Crystal Growth & Design 1 2 3 4 5 6 7 8 1. Introduction 9 10 11 12 13 Ultra-wide bandgap semiconductor oxides [1,2] are a class of materials that currently are attracting a lot of 14 [3–5] 15 attention as an emerging multi-functional platform owing to their unusual material properties , 16 17 endurance (high critical fields, radiation hardess,etc) [6–9] and unique optoelectronic properties [10–12]. The 18 19 emerging transparent semiconductor oxide (TSO) technology based on ultra-wide bandgap oxides holds 20 21 22 the promise of extending many energy and optoelectronic applications further into the deep ultraviolet 23 24 range, overpassing the conventional wide bandgap TSO’s near ultraviolet limit (~3.5 eV) (e.g. ZnO, SnO2 25 26 or In2O3) [13–15]. Nevertheless, to exploit the full potential of any emerging ultra-wide bandgap solid-state 27 28 29 semiconductor technology for electronics, a good control (over several orders of magnitude) of the n-type 30 31 and p-type conductivity should be attained. 32 33 34 35 36 Recently, single crystal n-type gallium oxide (Ga2O3) and zinc gallate (ZnGa2O4) are being revisited as 37 38 ultra-wide bandgap (~5 eV) oxide semiconductors. In particular, some very recent works such as the ones 39 40 from Hrong et al.[1] and Galazka et al. [16] have evidenced the great prospects of n-type ZnGa2O4 as 41 42 [17] potential semiconductor platform for future ultra-wide bandgap oxide optoelectronics . In this work, 43 44 45 we show that engineered ZnGa2O4 epitaxial single crystal layers may also be an ultra wide bandgap native 46 47 p-type semiconductor with great dopability prospects owing to the spinel‘s inherent diversity of choices 48 49 of cation coordination [18–19]. Spinels generally refer to compounds with formulation AB2X4 where A2+ is 50 51 52 a divalent cation such as Zn, Ni, Cu, Sn, Mg, Cr, Mn, Fe, Co or Cd, B 3+ is a trivalent cation such as Ga, 53 54 Al, In, Ti, V, Fe, Co, Ni, V, Cr, Mn and X2- is a divalent anion such as O, S or Se [20]. There are over 1,000 55 56 57 58 3 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 4 of 42 1 2 3 known compounds that crystalize in the spinel structure and the sub-family of spinel oxides is a large and 4 5 6 important class of multi-functional oxide semiconductors with many energy optoelectronics applications 7 [21–23] [24,25] [26] 8 in areas such batteries , fuel cells , catalysis (water spliting[27,28], photocatalysis [29,30] ), 9 10 photonics (phosphors [31–33] , bio-imaging [34,35] , transparent electrodes[36], photodetectors [37,38] ), 11 12 13 spintronics (magnets[39], bio-magnets[40]) or thermoelectricity [41]. Normal spinels have all A cations in the 14 15 tetrahedral site and all B cations in the octahedral site, e.g. Zn-tetrahedral site Zn2+(Td) and Ga-octahedral 16 17 site Ga3+(Oh) so that normal ZnGa2O4 is Zn(2+[Td])Ga2(3+[Oh])O4(2-). The spinel’s off-stoichiomerty, 18 19 which leads to a deviation from the ideal 1:2:4 stoichiometry, or the creation of cation antisite defects are 20 21 22 known routes for doping these compounds. Dominant defects in spinels are antisite acceptors (e.g. ZnGa) 23 24 with acceptor-like Zn2+(Td)-on-Oh antisite defects or donor-like Ga3+(Oh)-on-Td as . GaZn defects[42–46]. 25 26 Although spinel’s dopability was already predicted, experimentally demonstrated on very few visible 27 28 29 transparent (i.e. Eg ~ 3 eV) oxide spinels (e.g. ZnRh2O4 or ZnIr2O4) [47,48], and virtually no direct evidence 30 31 of p-type conductivity (only indirectly by Hrong et al.[49] to the best of our knowledge) has been reported 32 33 for UWBG spinel’s yet. It is worth mentioning that acceptor doping in other ultra-wide band gap (UWBG) 34 35 36 non-spinel oxides such as monoclinic -Ga2O3 is also still challenging [50],[51]. Therefore, the new prospects 37 38 opening up by p-type ternaries such as ZnGa2O4 may represent a good opportunity to mitigate the acceptor 39 40 issue in binary UWBGs. To achieve efficient p-type doping in wide band gap materials, we believe that 41 42 43 the first step is the realization of undoped semiconductor with hole conductivity originated from native 44 45 acceptor defects. It is clear that such semiconductor should be in off-stoichiometric state, but due to deep 46 47 acceptor centre levels it will be highly resistive. Folowing this approach, the goal of the work was 48 49 50 fabrication of undoped ZnGa2O4 with native hole conductivity. By demonstrating the intrinsic p-type 51 52 behaviour of the ultra-wide bandgap ZnGa2O4 we make feasible the predicted p-type doping, which paves 53 54 55 56 57 58 4 59 60 ACS Paragon Plus Environment Page 5 of 42 Crystal Growth & Design 1 2 3 the way to bipolar ultra-wide bandgap spinel electronics for a range of emerging power and energy 4 5 6 optoelectronic applications. 7 8 9 10 What does bipolar ultra-wide bandgap spinel electronics would mean? Oxide electronics is currently based 11 12 13 on classic and very well-known binary oxide semiconductors while emerging applications mostly rely on 14 15 complex oxides. Binary oxide examples include CuO2 for photovoltaics, In2O3, SnO2 and ZnO for 16 17 transparent electronics and, recently, Ga2O3 has disrupted for power and UV opto-electronics. Owing to 18 19 the flatness of the oxygen-related valence band and the deep energy of acceptor extrinsic dopant atoms, 20 21 22 the hole conductivity of wide- and ultra-wide binary oxides it is well-known to be very challenging. 23 24 Therefore, electronic devices made of these binary WBG semiconductors are mostly unipolar and n-type. 25 26 Among complex oxides, ternary perovskite oxides, (with general formulation ABO3), is perhaps the most 27 28 29 common platform of choice for advanced oxide electronics. Perovskite oxides, while exhibiting a range 30 31 of complementary properties such as ferroelectricity, colossal magnetoresistance or anomalous conducting 32 33 interfaces (e.g. SrTiO3/LaAlO3) still have not been demonstrated to deliver any competitive advantage for 34 35 36 energy or power electronics due to their relative low bandgap (~3 eV) and modest dielectric breakdown 37 38 field. Here, it is demonstrated that spinel ZnGa2O4 may be a native p-type ultra-wide bandgap 39 40 semiconductor (~5 eV). Hence, spinel oxides (a large family of AB2O4 compounds mostly unexplored as 41 42 wide bandgap power semiconductors) may represent a brand new opportunity for bipolar oxide energy 43 44 45 electronics (i.e., for managing large quantities of switching electrical energy) as they would join the 46 47 required qualities of sustaining large electrical fields in p-n junctions in the off-state together with low 48 49 losses in the on-state. The prospects of growing spinel ZnGa2O4 with high quality, either, as thin-film onto 50 51 52 cheap sapphire substrates or as a single-crystal (in a fashion similar to -Ga2O3) would make this approach 53 54 feasiable. Further, ZnGa2O4 may represent an unprecedented energy electronic platform for futuristic 55 56 57 58 5 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 6 of 42 1 2 3 energy generation, storage and power system integration since oxide spinels have been demonstated as 4 5 6 excellent phosphors, magnets, fuel cell and batery electrodes thus suggesting a promissing and yet fairly 7 8 unexplored smart integrated energy systems for the internet of things. 9 10 11 12 13 1. From Ga2O3/ZnGa2O4 mixed phases to monophase ZnGa2O4 14 15 16 17 In this experiment, the spinel’s ZnGa2O4 stoichiometry and its structural morphology have been optimized 18 19 within a metalorganic chemical vapor deposition (MOCVD) chamber during the growth of strongly 20 21 22 compensated p-type -Ga2O3 [51] in a progressively richer Zn-ambient. The ZnGa2O4 samples were grown 23 24 in a RF-heated horizontal MOCVD reactor on c-oriented sapphire (Al2O3) substrates, as described in the 25 26 27 experimental section. In short, during the growth, the flow rate of the gallium precursors and oxygen were 28 29 kept at 11 µmol/min and 1200 sccm respectively. The growth temperature, pressure and time were set at 30 31 775°C, 38 torr 150 minutes, respectively. Then, a number of Ga-Zn-O samples were grown varying the 32 33 34 Zn-flux (diethylzinc (DEZn)) in the range of 0-9 µmol/min. Pure -Ga2O3 was grown at a Zn-flux of 0 35 36 mol/min (termed Zn-0), while, for simplicity, in the following, the analysis is restricted to the most 37 38 representative Zn-fluxes of 3,9 µmol/min (Zn-4), 7 µmol/min (Zn-7) and 8,9 µmol/min (Zn-9). The Ga- 39 40 41 Zn-O layer thickness (ranging 900 to 1200 nm) was determined by cross-sectional scanning electron 42 43 microscope (SEM) and transmission electron microscopy (TEM), as described in Methods. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 6 59 60 ACS Paragon Plus Environment Page 7 of 42 Crystal Growth & Design 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Figure 1. (a) Full XRD pattern for -Ga2O3 and Ga-Zn-O thin-films with increasing Zn incorporation 48 49 (Zn-flux [mol/min] Zn-0, Zn-4, Zn-5, Zn-7, Zn-9). (b) An enlarged view of 4 peaks at 2θ=18.91°, 38.32°, 50 51 52 58.99° and 110.29° to show the phase transition from monoclinic -Ga2O3 to ZnGa2O4 spinel phase. For 53 54 low Zn-flux, the Ga-Zn-O layers show a double peak feature of mixed phases (GaO/GaZnO) of pure - 55 56 57 58 7 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 8 of 42 1 2 3 Ga2O3 (GaO) and pure ZnGa2O4. (c) Bragg distance as a function of Zn-flux from the (-804)/(444) 4 5 6 reflection in the GaO, mixed phase GaO/GaZnO and pure spinel ZnGa2O4 region. The inset in (c) are from 7 8 Zn-4 and Zn-9 2D composition maps from transmission microscopy (TEM) electron energy loss 9 10 spectroscopy (EELS) which further corroborates the mixed phase region and the pure ZnGa2O4 regions 11 12 13 (white scale bar is 100 nm). (d) Raman spectra for pure -Ga2O3, mixed GaO/GaZnO and pure spinel 14 15 ZnGa2O4. Although observed XRD family of Bragg reflections of monoclinic and spinel-cubic are 16 17 compatible, the different symmetry of the lattices resulted in very distinct Raman vibration modes with 18 19 20 the main being Ag and (T2g, A1g) for -Ga2O3 and ZnGa2O4, respectively. 21 22 23 24 The phase purity and crystallinity of the films was first analyzed by X-ray diffraction. Figure 1-a shows 25 26 27 the XRD spectrum for the pure monoclinic -Ga2O3 (Zn-0) and Ga-Zn-O thin films with various Zn-flux 28 29 ratios. The XRD patterns corroborate that -Ga2O3 was epitaxially grown following the monoclinic - 30 31 32 Ga2O3 {-2m0m} plane family. The pure -Ga2O3 (GaO) exhibited the presence of 5 peaks at 18.91°, 33 34 38.32°, 58.99°, 82.09° and 110.29° corresponding to the (-201), (-402), (-603), (-804) and (-10 0 5) 35 36 crystalline plans of the monoclinic β-Ga2O3 which are indexed in the card PDF-00-041-1103. The X-ray 37 38 39 diffractograms of the Ga-Zn-O samples show how they transitioned from a mixed solution of Ga2O3 and 40 41 ZnGa2O4 phases (termed GaO/GaZnO) to a pure spinel ZnGa2O4 phase. A double peak feature becomes 42 43 obvious for the smallest Zn-fluxes, while larger Zn-fluxes samples (Zn-flux > 7 mol/min) only exhibited 44 45 46 the shifted peak characteristic of the spinel {111} plane family reflections (Figure 1-b). As the Zn-flux 47 48 increases, the pure -Ga2O3 reflections gradually vanish while, at the same time, a new spinel reflection 49 50 (shifted to lower 2θ values) appears. This new reflection appears at 18.47°, 37.40°, 57.50°, 79.70° and 51 52 53 106.42° corresponding to the (111), (222), (333), (444) and (555) crystalline planes of the cubic spinel 54 55 ZnGa2O4, conformed to PDF-00-038-1240 The monoclinic-to-spinel 2-theta shift is more pronounced for 56 57 58 8 59 60 ACS Paragon Plus Environment Page 9 of 42 Crystal Growth & Design 1 2 3 the larger index planes (i.e., larger thetas angles) as expected from Bragg’s Llaw. Figure 1-c shows the 4 5 6 Bragg’s distance as a function of Zn flux for the spinel 𝐹𝑑3̅𝑚 (444) / monoclinic 𝐶2/𝑚 (-804) peak. The 7 8 inter-plane d-spacing is almost constant for -Ga2 O3 phases (either in pure or mixed phases), and only a 9 10 11 small increment of ~0.01 Å is observed with increasing Zn content for the GaZnO in the coexistence 12 13 region (GaO/GaZnO). In complement to these Bragg reflections separation, electron energy loss 14 15 spectroscopy (EELS) profiles at nano-scale supports a spinodal decomposition during the total growth 16 17 (and cooling) process. For pure ZnGa2O4 phases plane d-spacing becomes constant again. The 18 19 20 corresponding cubic a-parameter is equal to 8.33 Å, very close to reference values of PDF-00-038-1240 21 22 and values reported by Galazka et al [38] . Therefore, spinel 𝐹𝑑3̅𝑚 (111) / monoclinic 𝐶2/𝑚 (-201) 23 24 orientation are epitaxially compatible, as studied by Horng et al.[1] and very close interplane spacing (1.17- 25 26 27 1.20 Å) perpendicular to growth direction. Owing to their different lattice symmetries, however, they show 28 29 distinct vibrational modes under Raman excitation (Figure 1 -d). The primary ZnGa2O4 cubic spinel 30 31 vibrational modes detected are A1g (709.7 cm−1) and T2g modes (603.1 cm−1) while the main β-Ga2O3 32 33 34 vibrational modes detected are related to Ag (651.2 cm-1), in agreement with previous literature (e.g. [52]). 35 36 The oxide’s optical bandgaps were determined by optical transmittance and reflectance spectroscopy. The 37 38 optical transparency of the double polished sapphire substrate allows the determination of the intrinsic 39 40 optical characteristics of the Ga2O3 and ZnGa2O4 thin-films. The optical transmittance and reflectance of 41 42 43 -Ga2O3 and ZnGa2O4, pure and mixed phases (GaO/GaZnO), were determined by transmission 44 45 spectroscopy all the range from UV-C to short wavelength infrared. As seen in Figure 2-a, -Ga2O3, 46 47 48 ZnGa2O4 and mixed GaO/GaZnO exhibited a transparency above 80% in all wavelength range from ~300 49 50 nm to 2500 nm. All films have sharp absorption edges of ~250 nm. The optical bandgaps were determined 51 52 by means of a Tauc’s plot to be in the range of 4.6-4.8 eV (as shown in Figure 2-b). The optical band gap 53 54 55 56 57 58 9 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 10 of 42 1 2 3 for the pure -Ga2O3 and the GaO/GaZnO mixed phase is around ~4.6 eV, while it is slightly lager 4.8 eV 4 5 6 for the pure ZnGa2O4 (Figure 2-c-). 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Figure 2. (a) Optical transmittance. (b) Tauc’s plots (for the determination of the optical bandgap energy) 46 47 48 for GaO/GaZnO mixed phase and for the pure ZnGa2O4. (c) Optical bandgap energy versus Zn-flux 49 50 (µmol/min) used during the growth. (d) and (e) top view SEM images of the GaO/GaZnO mixed phase 51 52 and the pure ZnGa2O4. (f) and (g) cross sectional SEM images of the GaO/GaZnO mixed phase and the 53 54 55 pure ZnGa2O4. 56 57 58 10 59 60 ACS Paragon Plus Environment Page 11 of 42 Crystal Growth & Design 1 2 3 The structural characteristics of the different composition of ternary oxides were further 4 5 6 investigated by means of scanning electron microscope (SEM) and transmission electron microscope 7 8 (TEM) equipped by energy-dispersive x-ray spectroscopy (EDS) and electron energy loss spectroscopy 9 10 (EELS). Figure 2(d)-(g) presents a representative SEM top view and cross-sectional view of the ternary 11 12 13 Ga-Zn-O having mixed phase GaO/GaZnO and pure spinel zinc gallate ZnGa2O4. It is already evidenced 14 15 by SEM that a pure phase ZnGa2O4 (Zn-9) exhibited a much planar morphology together with the absence 16 17 of grains and grains boundaries as compared with a mixed GaO/GaZnO (Zn-4) sample. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Figure 3. (a) Transmission electron microscope (TEM) images for (a) mixed phase GaO/GaZnO and (b) 54 55 pure spinel ZnGa2O3. Higher resolution images are shown in (c) and in (d), for GaO/GaZnO and ZnGa2O3, 56 57 58 11 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 12 of 42 1 2 3 respectively. The inset shows the fast Fourier transform of the images, with multiple reflections 4 5 6 evidencing the phase mixture for the former and a perfectly oriented [011] view for ZnGa2O4. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Figure 4. Cross-section STEM EELS Ga (a,h.), Zn (b.,i.), Ga/Zn (e.,m.) and EDX Ga (c.,j.), Zn (d.,k.) 41 42 compositional maps and high-angle annular dark-field imaging (HAADF) STEM image (g.,o.) for (a) 43 44 mixed phase GaO/GaZnO and (b) pure spinel ZnGa2O4. A compositional EDX profile is shown for the 45 46 47 mixed and pure GaZnO showing its granular and homogeneous nature, respectively. 48 49 50 51 At higher resolution than the scanning electron microscopy, transmission electron microscopy (TEM) 52 53 54 imaging corroborates the granular nature of the mixed GaO/GaZnO. In contrast, ZnGa 2O4 exhibited an 55 56 epitaxial single crystal structure as shown in Figure 3-a,-c and in Figure 3-b,-d, for mixed GaO/GaZnO 57 58 12 59 60 ACS Paragon Plus Environment Page 13 of 42 Crystal Growth & Design 1 2 3 and spinel ZnGa2O4, respectively. For the samples with the lowest Zn-fluxes, there is an intermix of pure 4 5 6 GaO and GaZnO phases resulting in distorted grains of varying sizes of ~10-100 nm in average and the 7 8 corresponding large number of grain boundaries, coherent with a spinodal decomposition. In contrast, for 9 10 Zn-fluxes above 7 mol/min, the Zn atoms are accommodated epitaxially within a single crystal spinel 11 12 13 ZnGa2O4 structure as shown in Figure 3-c and 3-d. The Ga-Zn-O layers stoichiometry and atomic 14 15 composition were further investigated by performing EELS and spectroscopy as shown in Figure 4. 16 17 Corroborating the x-ray diffraction and Raman analysis, for the smallest Zn-fluxes (i.e., those that resulted 18 19 20 in mixed GaO/GaZnO phases), the Zn EELS and EDX maps are non-homogeneous but Zn atoms are 21 22 concentrated in ZnGa2O4 grains within the -Ga2O3 matrix (Figure 4-a). It is worth noting that the 23 24 EELS/EDX Zn concentration variations do not perfectly follow the high-angle annular dark-field imaging 25 26 27 (HAADF) contrast variation is the STEM mode. This may be due to the fact that Zn (30) and Ga (31) have 28 29 very similar atomic numbers. Pure spinel ZnGa2O4 was found to be, in contrast, much more homogeneous 30 31 in Zn and Ga compositions all over the layer. The stoichiometry of the Ga-Zn-O ternary oxide has been 32 33 34 determined for a Zn-flux of 8.9 mol/min (Zn-9) to be 32% Zn and 68% Ga, and therefore, within the 35 36 experimental error, it fits very well with a ZnGa2O4 formulation. 37 38 An atomic fine chemical analysis of the surface was performed by x-ray photoelectron spectroscopy (XPS). 39 40 41 XPS provides information about the Ga, Zn and O local bonding environment (not just element 42 43 identification) either on pure ZnGa2O4 or mixed GaO/GaZnO phases. The local atomic binding energies 44 45 are affected by the cation formal oxidation state, the identity of its nearest-neighbor (Zn or Ga) atoms and 46 47 48 its bonding hybridization. The two extreme compositions, monoclinic -Ga2O3 and wurtzite ZnO, grown 49 50 in the same MOCVD chamber (see methods), were also included for comparison. Figure 5 shows a 51 52 detailed view of the main XPS peaks for (a) gallium (Ga2p), (b) oxygen (O1s), (c) zinc (Zn2p) and (d) 53 54 55 56 57 58 13 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 14 of 42 1 2 3 carbon (C1s) at the surface of pure -Ga2O3 (GaO), pure wurtzite ZnO, mixed phase GaO/GaZnO (Zn-4) 4 5 6 and pure spinel ZnGa2O4. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Figure 5. Detail (intensity normalized) of the main XPS peaks for (a) gallium (Ga2p - Ga2p3/2), (b) oxygen 43 44 (O1s), (c) zinc (Zn2p – Zn2p3/2) and (d) adventitious carbon contamination (C1s, calibration for C-C bonds 45 46 47 at 284.8 eV) at the surface of pure -Ga2O3 (GaO) with Zn-flux = 0 mol/min), mixed phase GaO/GaZnO 48 49 (Zn-4), pure ZnGa2O4 and a pure ZnO (wurtzite) reference (grown in the same chamber). Deconvolution 50 51 of the main contributions of the (e) Ga2p3/2, (f) Zn2p3/2 and (g) O1s oxygen peak for spinel ZnGa2O4. 52 53 54 55 56 57 58 14 59 60 ACS Paragon Plus Environment Page 15 of 42 Crystal Growth & Design 1 2 3 Adventitious carbon contamination was used as a charge reference for XPS spectrum (by setting C1s 4 5 6 maxima to a binding energy of 284.8 eV), which is a very common practice (Figure 5-d). Regarding O1s, 7 8 a notable feature is the binding energy difference of the metal oxide oxygen peak in ZnO (~530 eV) 9 10 (e.g.[53]) and the metal oxide oxygen peak binding energy (~531 eV) in monoclinic -Ga2O3 (e.g.[54,55]), 11 12 13 GaO/GaZnO and spinel ZnGa2O4. According to Biesinger et al.[56], the ZnO(II) O1s lattice oxide peak 14 15 (OI) appears at 529.76 eV while the hydroxide, hydrated or defective oxygen (e.g. oxygen vacancies) 16 17 shoulder (OII) in the O1s peak of ZnO appears shifted 1.49 eV, at 531.25 eV. In our case, the ZnO’s lattice 18 19 20 peak appears at 529,99 eV and the shoulder maxima appears at 531.69 eV which also presents the broad 21 22 shape typical of multi-component contributions (i.e., -OH, H2O and VO). The -Ga2O3 (Zn-0) peak 23 24 appears at 530,94 eV exhibiting a much less pronounced shoulder (see Fig. 5(g)). There also is no 25 26 27 appreciable shoulder in GaO/GaZnO and spinel ZnGa2O4. The absence of shoulder features in O1s may 28 29 be correlated with a lower donor compensation (very low concentration of oxygen vacancies). This 30 31 correlates with the intrinsic p-type nature of the -Ga2O3, GaO/GaZnO and ZnGa2O4 layers, as it will be 32 33 34 shown below by Hall transport measurements. For the reference spinel ZnGa2O4 (Zn-9) the O1s lattice 35 36 oxide peak appears at 530.99 eV while the Zn2p3/2 and Ga2p3/2 peak appears at 1022.20 eV and 1118.20 37 38 eV, respectively. The XPS quantification for the reference ZnGa2O4 gives a stoichiometry of 30.4% Zn 39 40 41 and 69.6% Ga, which is in close agreement with the quantification from TEM/EDX and the spinel 42 43 stoichiometry. 44 45 As shown in Figures 5-a and 5-c, there is a significant shift of the binding energy of the Zn2+ and Ga3+ 46 47 48 cations towards larger (Ga-O and Zn-O) binding energies which usually is regarded as an indication of 49 50 cation oxidation (loss of electrons or an increase in the oxidation state). According to the first-principles 51 52 calculations based on the density functional theory of Chao et al. [57] , with the introduction of the Zn 53 54 impurity (Ga substitutional), the Zn atom loses almost all valence electrons (since the Zn electronegativity 55 56 57 58 15 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 16 of 42 1 2 3 is lower than Ga and the Zn2+ valence electrons are less than Ga3+) which may explain the Zn2p shift 4 5 6 towards larger binding energies with increasing Zn-flux. The Zn2p3/2 peak for pure ZnO appears at 7 8 1021.34 eV while the Zn2p shifts to 1021.68 eV for the mixed GaO/GaZnO and 1022.20 eV for pure 9 10 ZnGa2O4 (Zn-9). Normal AB2X4 spinels have all A cations in the tetrahedral site and all B cations in the 11 12 13 octahedral site. Therefore, in the case of zinc gallate ZnGa2O4, Zn atoms lie in Zn-tetrahedral sites 14 15 (Zn2+(Td)) and Ga atoms occupy the Ga-octahedral sites (Ga3+(Oh)) or Zn(2+[Td])Ga2(3+[Oh])O4(2-). In 16 17 contrast to simple binary oxides (such as pure -Ga2O3), where dominant defects are vacancies (Vo as 18 19 20 donor and VGa as acceptor) and interstitials, dominant defects in spinels are suggested to be antisites [18]. 21 22 These antisites are donor (GaZn) [donor-like Ga3+(Oh)-on-Td antisite defects] and acceptor (ZnGa) 23 24 [acceptor-like Zn2+(Td)-on-Oh antisite defects]. Intrinsic p-type self-doping of spinels implies therefore a 25 26 27 minimization of the electrical influence of donor-like Ga3+-on-Td antisite defects while maximizing the 28 29 electrical influence of acceptor-like Zn2+-on-Oh antisite defects. It has been suggested that in order to 30 31 achieve p-type dopability [18] , the concentration of Zn2+-on-Oh should be maximized either, (i) by 32 33 inducing Zn-rich off-stoichiometry using non-equilibrium growth or (ii) by choosing an Zn2+ element that 34 35 36 naturally prefers Oh-coordination. Regarding the Zn favorite coordination, it has been suggested that Zn 37 38 prefers a tetrahedral coordination when substituting Ga in -Ga2O3 . The -Ga2O3 belongs to space [58] 39 40 group C2/m with two-fold rotation axis b. There are two different Ga sites, denoted as Ga(1) and Ga(2), 41 42 43 and three different O sites, denoted as O(1), O(2), and O(3). Ga atoms are surrounded by O atoms in either 44 45 tetrahedral Ga(1) or octahedral Ga(2) coordination. Skachkov and Lambrecht recently published a 46 47 computational study of the electron paramagnetic resonance g-tensors of Zn-doped Ga2O3 [58] where they 48 49 50 determined that Zn would prefers the tetrahedral Ga(1) substitutional site, (Zn in ZnO has a tetrahedral 51 52 bonding indeed as it has in normal ZnGa2O4). 53 54 55 56 57 58 16 59 60 ACS Paragon Plus Environment Page 17 of 42 Crystal Growth & Design 1 2 3 The cation distribution can be characterized by the inversion parameter y, which is defined as 2 times the 4 5 6 fraction of Zn2+ ions in the octahedral sites. The degree of inversion in zinc gallate has been reported to 7 8 be potentially estimated by the deconvolution of the Ga2p and Zn2p XPS peaks in two components at 9 10 different binding energies (BEs) [59]. A Ga2p3/2 normal Ga3+(Oh) contribution peaks at high BEs (1118.3 11 12 13 eV) and an inversion Ga3+(Td) contribution is at significantly lower BEs (1116.8 eV). Analogously, a 14 15 Zn2p3/2 normal Zn2+(Td) is at lower BEs (~1022 eV) and an inversion Zn2+(Oh) is at higher BEs (~1023 16 17 eV). Therefore, the tail contribution at larger binding energies in Zn2p3/2 would directly be a measure of 18 19 ZnGa. In our case, the Ga2p3/2 peak appears precisely at 1118.2 eV and there is a small contribution shifted 20 21 22 ~2 eV at lower BE (Figure 5-e) of ~2.8% the total area, while there would be around ~5.3% of ZnGa sites 23 24 when the Zn2p3/2 peak is deconvoluted (Fig. 5(f)). An additional contribution may be observed for even 25 26 larger BE in the ZnGa2O4 Ga2p3/2 peak (1120.1 eV) which, as far as we know, has not been discussed in 27 28 29 previous literature yet. 30 31 32 33 34 35 36 2. ZnGa2O4 spinel electrical transport properties: An intrinsic p-type semiconductor 37 38 39 40 For studying the actual electrical properties of the spinel zinc gallate layers, the ZnGa2O4 thin-films 41 42 resistivities and their corresponding Hall Effect were measured at high temperatures (450-850 K) in a Van 43 44 45 Der Pauw configuration. Four electrical contacts by silver paint were made on each corner of a square 46 47 shape (1x1 cm²) samples. As shown in Figure 6-a, a linear, non-rectifying current-voltage (I-V) behavior 48 49 has been first corroborated at different temperatures, thus evidencing the Ohmic characteristic of the 50 51 52 contacts. The resistivity vs temperature for spinel ZnGa2O4 samples is shown in Figure 6-b. 53 54 55 56 57 58 17 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 18 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Figure 6. (a) Typical spinel ZnGa2O4 Ohmic I-V characteristic at 850K. (b) Resistivity versus 41 42 temperature for averaged (error bars) p-type spinel ZnGa2O4 and the reference p-type -Ga2O3. (c) Hall 43 44 45 voltage (VH) versus magnetic field at different temperatures for a typical spinel ZnGa2O4 sample. The Hall 46 47 voltage slope is positive thus an indication of p-type conductivity. (d) Temperature dependence for Hall 48 49 free hole concentration (p [cm-3]) for pure monoclinic reference -Ga2O3 and spinel ZnGa2O4 films. An 50 51 52 activation energy of the acceptor center (Ea) was determined from the log (p) vs 1/T plot (Arrhenius) to 53 54 be 1.2±0.5 eV and 1.0±0.5eV for reference -Ga2O3 and spinel ZnGa2O4 films, respectively. 55 56 57 58 18 59 60 ACS Paragon Plus Environment Page 19 of 42 Crystal Growth & Design 1 2 3 4 5 6 The determination of the majority carrier type is, in practice, challenging when the sample is of high 7 8 resistance ( > 107 Ω) due to the difficulty in correctly extracting the Hall voltage (VH) from the total 9 10 measured voltage. Elevating the temperature of measurement in our high temperature high impedance 11 12 13 home-built Hall set-up enables measuring up to 1GΩ (109) resistance sample. In a non-magnetic material, 14 15 VH is linearly proportional to the applied magnetic field and a positive sign slope for VH (H)indicates that 16 17 the majority charge carriers are p-type (holes). To further validate the sign of majority carriers, the Hall 18 19 20 voltage dependence on the applied magnetic field was measured at different elevated temperatures and 21 22 varying magnetic fields from 0 to 1.6 T. As shown in Figure 6-c, the positive VH linearly increases with 23 24 perpendicularly applied magnetic field. This confirms that the layer was p-type. The temperature 25 26 27 dependence for Hall hole concentration is shown in Figure 7-d. ZnGa2O4 p-type films and, at the highest 28 29 available temperature of 850 K, the free hole concentration was measured to be as large as p = 2×1015 cm- 30 31 3 , while for pure -Ga2O3 the hole concentration was 100 times smaller (i.e., p = 1.4×1013 cm-3). From the 32 33 34 log(p) vs 1/T plot, the activation energy (Ea) of the acceptor centers have been determined to be 1.2±0.5 35 36 eV and 1.0±0.5 eV. 37 38 39 40 [42–46] As mentioned before, according to previous literature , the most plausible origin of p-type 41 42 43 conductivity in ternary spinel ZnGa2O4 would be the anti-site ZnGa center, in contrast to normal binary 44 45 oxides where it is generally ascribed to a cation vacancy or interstitial. In our case, undoped -Ga2O3 and 46 47 48 spinel ZnGa2O4 exhibit similar value of acceptor center activation energies (Ea) but still 100 times more 49 50 free holes in ZnGa2O4 films. A canonical interpretation of our Hall Effect measurements would suggest 51 52 that the point defect responsible for p-type conductivity would be the same (or at least to lie in a similar 53 54 energetic position within the bandgap) in both materials, while the differences in p [cm-3] (free hole 55 56 57 58 19 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 20 of 42 1 2 3 concentration) would be related to a different compensation ratio. The compensation by donor defects, in 4 5 6 another word ND/NA ratio, would be smaller in ZnGa2O4 films than in pure -Ga2O3, grown with the same 7 8 parameters (i.e., oxygen partial pressure and temperature). The increase of acceptors, NA might be 9 10 explained with creation of anti-site ZnGa acceptor defects in ZnGa2O4 spinel. This defect being deep 11 12 13 cannot play an “effective” acceptor role with small ionization energy, but acts as a compensator-donor 14 15 killer, thus resulting enhancement of hole concentrations. Experimental Hall hole mobilities for spinel 16 17 ZnGa2O4 were found to be in the range of µ = 7 - 10 cm2/Vs (680-850 K). Such a remarkable high value 18 19 20 of acceptor mobilities for p-type TCO confirms that conductivity is by band activation mechanism and 21 22 not by small polaron as it often has been reported to be the case in other spinels [47]. That is a very important 23 24 feature for p-type ternary ZnGa2O4 oxide, since having delocalized holes in undoped state, gives strong 25 26 27 promises of efficient doping capability for increasing free carriers. 28 29 30 31 32 33 3. The ZnGa2O4 Valence Band: Experiment vs ab-initio simulations 34 35 36 Once the ZnGa2O4 layers p-type conductivity has been determined by high-temperature Hall Effect, the 37 38 nature of the oxide’s semiconductor valence band has been further investigated by means of high- 39 40 resolution X-ray photoelectron spectroscopy and ab-initio calculations. XPS also is an excellent chemical 41 42 43 method to investigate a material’s valence band characteristics and brings additional evidence of the 44 [60–62] 45 presence of states within the bandgap . When used in high resolution in the valence band vicinity 46 47 (i.e., for the lowest binding energies), it is possible to directly detect whether there are states in the lower 48 49 50 part (i.e., smaller than the intrinsic energy level or ~mid-gap) of the bandgap (those responsible for p-type 51 52 character) [63]. 53 54 55 56 57 58 20 59 60 ACS Paragon Plus Environment Page 21 of 42 Crystal Growth & Design 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Figure 7. (a) Experimental XPS high-resolution valence band for the p-type -Ga2O3 and a reference n- 38 39 type -Ga2O3. (b) Experimental XPS high-resolution of the first valence band group for pure p-type - 40 41 42 Ga2O3, p-type mixed phases of Ga2O3/ZnGa2O4 (GaZnO) and p-type pure spinel ZnGa2O4. An undoped 43 44 ZnO specimen is included for comparison. The valence band maxima (VBM) is from O2p states for all 45 46 the samples but those compounds containing zinc atoms exhibit an important additional contribution of 47 48 49 Zn3d states (c) Idem, for the first and the second group of valence bands. The second group of bands is 50 51 contributed primarily by Ga3d (in the compounds containing Gallium) and O1s. (d) A zoom of the valence 52 53 band region (VBM) showing the presence of tail states in the lower part of the bandgap for the p-type 54 55 56 compounds (the value of the bandgap is depicted for ZnO, -Ga2O3 and ZnGa2O4). 57 58 21 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 22 of 42 1 2 3 4 5 6 Our starting point was pure monoclinic p-type -Ga2O3 (Zn-0). As shown in Figure 7-a, when -Ga2O3 7 8 was grown in strong compensation conditions at high oxygen flux, the layers were found to be highly 9 10 11 resistive p type in nature. When compared with a reference (Si-doped) n-type -Ga2O3 single crystal, they 12 13 exhibited a shift towards smaller binding energies and a number of tail states arise at ~1.5-1.0 eV from 14 15 the conduction band edge (Figure 7-d). The n-type control sample is a commercial (Novel Crystal 16 17 18 Technology, Inc.) nominally n-type Si-doped -Ga2O3 (ND-NA = 1.3×1018 cm-3) epitaxy (500 nm) grown 19 20 on a single crystal -Ga2O3 (labelled ‘‘n-type’’). For our samples, the valence band edge is shifted towards 21 22 more p-type using the adventitious C-C 284.8 eV as the calibration peak (lower Fermi level implies more 23 24 25 p-type character). Figures 7b) and 7-c show the experimental valence region for p-type -Ga2O3, p-type 26 27 mixed phases of Ga2O3/ZnGa2O4 (GaZnO) and p-type pure spinel ZnGa2O4. These samples were defined 28 29 by increasing progressively the Zn-flux within the MOCVD chamber. The other stoichiometric end, an 30 31 32 undoped ZnO specimen, is included for comparison (i.e. Ga-flux 0 mol/min). Indeed, the spinel ternary 33 34 ZnGa2O4, in terms of composition, can be regarded as a metastable alloy of Ga2O3 and ZnO binary oxides 35 36 37 in a cubic 𝐹𝑑3̅𝑚 arrangement, where Zn and Ga atoms have tetrahedral (Td) and octahedral (Oh) 38 39 coordination bounds with oxygen, respectively. While -Ga2O3 have Ga-O (Td) and (Oh) bounds (1:1), 40 41 wurtzite ZnO presents only Td Zn-O bounds. Figure 7-d is a zoom of the valence band region (VBM) 42 43 44 showing the presence of tail states in the lower part of the bandgap for the p-type compounds (the value 45 46 of the bandgap is depicted for ZnO, -Ga2O3 and ZnGa2O4. For reference ZnO, the Fermi level (as 47 48 measured from the valence band maxima or VBM) lies in the mid-gap but in the upper-half of the bandgap 49 50 51 (~2.50 eV), an indication of their native highly resistive but n-type character. In contrast, the ZnGa2O4 52 53 spinel exhibit an smaller Fermi level energy (~0.5 eV) and thus, the acceptor states would lie within the 54 55 bandgap closer to the valence band maximum than in p-type -Ga2O3, an observation that is coherent with 56 57 58 22 59 60 ACS Paragon Plus Environment Page 23 of 42 Crystal Growth & Design 1 2 3 the experimental higher free hole concentrations by Hall Effect in spinel ZnGa2O4. In agreement with this 4 5 6 fact, it may be observed that, even for the mixed phase of -Ga2O3 and spinel ZnGa2O4, the acceptor tail 7 8 states are significantly shifted towards the valence band (label “GaZnO” in Figure 7). 9 10 11 12 13 Ab-initio simulation (see methods) are used to compare (vs experiment) the oxide’s theoretical bandgap 14 15 energies and to identify the main components of the experimental XPS valence bands in Figures 7(-c and 16 17 7-d. The calculated bandgap energies (Eg) and the semiconductor band structure in the vicinity of VBM 18 19 20 are shown in Figure 8. The theoretical direct band gaps are 4.73 eV, 3.40 eV and 5.06 eV, for monoclinic 21 22 -Ga2O3, wurtzite ZnO and spinel ZnGa2O4, respectively. These Eg values are in agreement with the 23 24 values extracted for the optical bandgap measured by optical spectroscopy (see section 2). Further 25 26 27 simulation details are given in the method section. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 Figure 8. Calculated bandgap energy for (a) monoclinic -Ga2O3, (b) wurtzite ZnO and (c) spinel 53 54 ZnGa2O4 by using DFT. 55 56 57 58 23 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 24 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Figure 9. Determination of the valence bands main atomic contribution and their density of states (DOS) 41 42 for (a) monoclinic -Ga2O3, (b) wurtzite ZnO and (c) spinel ZnGa2O4 by using DFT. 43 44 45 46 47 A closer look of the theoretical valence band electronic structure and its electronic configuration is shown 48 49 in Figure 9. The valence band maxima (VBM) is from O2p states for all the oxides under study but those 50 51 52 compounds containing zinc atoms and gallium atoms exhibit further contributions of Zn3d states (ZnO) 53 54 and Ga3d states (Ga2O3), respectively (Figures 9-a and 9-b). In the case of the spinel ZnGa2O4, there are 55 56 three main features observed in the experimental which are ascribed to O2p, Zn3d and Ga3d respectively 57 58 24 59 60 ACS Paragon Plus Environment Page 25 of 42 Crystal Growth & Design 1 2 3 as shown in Figure 7-c. When calibrated with respect adventitious C-C (284.8 eV), it is noticeable the 4 5 6 shift observed for the Zn3d and Ga3d towards smaller binding energies for the ternary spinel ZnGa2O4 7 8 oxide. The Ga3d peak appears at 20.73 eV, 19.75 eV and 17.60 eV, for -Ga2O3, mixed phases and spinel 9 10 ZnGa2O4, respectively. For the Zn3d contribution, the peak values are 10.79 eV, 10.40 eV and 8.05 eV, 11 12 13 for ZnO, mixed phases and spinel ZnGa2O4, respectively. This decrease in the first valence band binding 14 15 energies is coherent with the smaller Fermi level energy derived for spinel ZnGa2O4 (~0.5 eV). As shown 16 17 in Figure 9-b, the GGA+U approximation reproduces well the valence band experimental features for our 18 19 20 ZnO. Semi-empirical approach (LDA-1/2) and Meta_GGA (one of semi-local approximations functionals 21 22 as an alternative of the DFT-1/2) were used for spinel ZnGa2O4 and -Ga2O3, respectively. Although, the 23 24 binding energies for the Ga3d states are particularly understimated in gallium oxide related compounds 25 26 27 (we are using for the sake of simplicity local or semilocal exchange correlation), the simulation procedure 28 [64] 29 still reproduces the main valence band experimental peak contributions and ordering . The BE 30 31 divergences would be related to the fact that, in monoclinic -Ga2O3 and spinel ZnGa2O4, shallow Ga3d 32 33 34 core electrons have to be treated as valence electrons (for example, by using many-body perturbation 35 36 theory within Hedin’s GW approach) [65] . 37 38 39 40 41 As demonstrated for -Ga2O3 [49] , the Kroger’s thermodynamic analysis predicted that high oxygen 42 43 pressures guarantee the appearance of uncompensated hole conductivity effectively suppressing the 44 45 compensation by native donors, which are associated primarily to oxygen vacancies (VO). As opposed to 46 47 48 ZnO, thanks to the -Ga2O3 particular point defect chemistry and the large formation energy of oxygen 49 50 vacancies [66] (the native donor which can play the role of compensators for holes), it is possible to achieve 51 52 the realization of stable p-type conductivity even in undoped layers. Indeed, as mentioned before, the ZnO 53 54 55 layer presents a much larger Fermi energy level of ~2.5 eV which suggest a natural n-type character (in 56 57 58 25 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 26 of 42 1 2 3 line with n-type (Si-doped) reference -Ga2O3) and contrary to the rest of the p-type layers grown. As 4 5 6 shown in Figure 7-d-, the entire valence band is further shifted towards smaller Fermi levels for the spinel 7 8 ZnGa2O4 when compared to the starting p-type -Ga2O3 and mixed GaZnO phases. It is worth noticing 9 10 11 that the C1s calibration shift is believed to have no effect in the determination of the ZnGa 2O4 valence 12 13 band (i.e. Fermi level, Ga3d or Zn3d energies - see methods). 14 15 16 17 This enhanced p-type dopability may be ascribed to the natural spinel tendency to a certain degree of 18 19 20 inversion and the corresponding creation of acceptor antisites (ZnGa). In this sense, preliminary DFT 21 22 simulations for the ZGO antisite pair were conducted (see methods). A 112-atom supercell was built 23 24 (Zn16Ga32O64) and the simulated band structure and DOS shown the creation of two acceptor energy levels, 25 26 27 which is consistent with what was already reported by De Vos et al. [46]. These two energy states are 28 29 contributed by the electrons in Zn-3d and O-2p orbitals. The antisite induced acceptor energy states mainly 30 31 attributed to ZnGa is therefore a strong candidate for explaining the origin of measured p-type conductivity. 32 33 34 Here, the p-type spinel character has been achieved by tuning our growth methodology by defining highly 35 36 compensated -Ga2O3 in high oxygen pressure (to remove the remote donor) and then, adding the Zn- 37 38 atoms at high Zn-flux. As the gallium vacancy is the most plausible acceptor in our -Ga2O3 layers, the 39 40 41 Zn-atoms will incorporate naturally in some Td and Oh sites and play role of donor killer, promoting native 42 43 hole conductivity of our spinel ZnGa2O4. 44 45 46 47 48 4. Conclusions 49 50 To exploit the full potential of any emerging ultra-wide bandgap (UWBG) solid-state semiconductor 51 52 technology, a good control (over several orders of magnitude) of the n-type and p-type conductivity should 53 54 55 be attained. Although spinel’s dopability was already theoretically predicted, no p-type conductivity has 56 57 58 26 59 60 ACS Paragon Plus Environment Page 27 of 42 Crystal Growth & Design 1 2 3 been corroborated for UWBG ZnGa2O4 spinel’s yet. To achieve efficient p-type doping in wide band gap 4 5 6 materials, the first step is the realization of native hole conductivity in the undoped state. Following this 7 8 approach, it is further demonstrated in this work that the ternary ultra-wide band gap (~5 eV) spinel zinc 9 10 gallate (ZnGa2O4) is the native p-type ternary oxide semiconductor with the widest bandgap. This 11 12 13 achievement opens-up an important perspective for succesful UWBG p-type doping, thus paving the way 14 15 to bipolar extreme spinel electronics for a range of emerging energy applications (e.g. opto and power 16 17 electronics). Furthermore, as a new acceptor member of the emerging gallium oxide family technology, 18 19 this spinel ZnGa2O4 phase, (stabilized after Zn incorporation), is structurally compatible with the 20 21 22 monoclinic and more stable -Ga2O3 phase where suitable growth directions are [111] and [-201] for 23 24 ZnGa2O4 and -Ga2O3, respectively. 25 26 27 28 29 30 31 Acknowledgements 32 33 34 Hagar Mohammed would like to acknowledge Cultural Affairs and Massion Sector, Egyptian Ministry 35 36 for Higher Education for her fellowship giving possibility work in France. EdC acknowledges Spanish 37 38 MINECO Juan de la Cierva Fellowship JC-2015-25201. APT acknowledges Agencia Estatal de 39 40 Investigación (AEI) and Fondo Europeo de Desarrollo Regional (FEDER) under contract ENE2015- 41 42 43 74275-JIN. The ICN2 is funded by the CERCA programme / Generalitat de Catalunya and by the Severo 44 45 Ochoa programme of the Spanish Ministry of Economy, Industry and Competitiveness (MINECO, grant 46 47 no. SEV-2017-0706). We express our thanks to G.Bouchez for his assistance for optical measurements. 48 49 50 51 52 Experimental details 53 54 55 56 57 58 27 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 28 of 42 1 2 3 A. Thin-film Growth 4 5 6 Ga2O3, ZnO and ZnGa2O4 samples were grown in a RF-heated horizontal metalorganic chemical vapour 7 8 deposition MOCVD reactor with separate inlets to avoid premature reactions in the manifold between 9 10 oxygen and organometallics precursors. The reactor can operate at low pressure, between 30 and 760 torr, 11 12 13 and at high growth temperature, up to 1000 °C. Trimethylgallium (TMGa), diethylzinc (DEZn) and 5.5N 14 15 pure oxygen were used as gallium, zinc and oxygen sources, respectively. Argon was used as carrier gas. 16 17 18 19 Ga2O3 and ZnGa2O4. The TMGa and DEZn bubbler temperatures were fixed at -10°C and 0°C, 20 21 22 respectively. ZnGa2O4 layers were grown on C-oriented sapphire substrates. During the growth, the flow 23 24 rate of TMGa and oxygen were kept at 11µmol/min and 1200 sccm, respectively. The growth temperature 25 26 was set at 775°C, the reactor pressure was 38 torr and growth time was 150minutes. The DEZn flow was 27 28 29 varied in the range (2.8-8.9 µmol/min). It were defined several reference pure -Ga2O3 samples without 30 31 Zn (or Zn0) and several Ga-Zn-O specimens under the following conditions: 2,8µmol/min (Zn3), 32 33 3,9µmol/min (Zn4), 5µmol/min (Zn5), 7µmol/min (Zn7), 8,9µmol/min (Zn9). 34 35 36 37 38 ZnO. ZnO films were grown on c-oriented sapphire substrates. During the growth, the flow rate of DeZn 39 40 and nitrous oxide (N2O) were kept at 20µmol/min and 3700 sccm, respectively. The growth temperature 41 42 43 was fixed at 950°C, the reactor pressure was 50 torr. The growth rate was 280nm/h in this growth 44 45 conditions. 46 47 48 49 50 51 52 B. Characterization Methods 53 54 55 56 57 58 28 59 60 ACS Paragon Plus Environment Page 29 of 42 Crystal Growth & Design 1 2 3 B.1. XRD. X-ray diffraction profiles were recorded in θ/2θ configuration, using Rigaku SmartLab 4 5 6 equipped with Cu-Kα1 source (λ=0,1541nm). 7 8 9 10 B.2. Raman Spectroscopy. Raman spectra were acquired in a Witec spectrometer coupled to a 488.0 nm 11 12 13 laser excitation. The laser was focused on the sample by a 50× objective with a 1 mW power. We used a 14 15 600 g/mm grating, covering a spectral window from 100 to 4000 cm-1 with a spectral resolution of 3 cm- 16 17 1 /pixel. Single Raman spectra of 60 seconds were registered. 18 19 20 21 22 B.3. Transmittance/Reflectance. Optical transmission spectra were measured in 200-2000 nm spectral 23 24 range with a Perklin Elmer 9 spectrophotometer. 25 26 27 28 29 B.4. SEM/EDX. SEM images were done by FEG JEOL 7001-F electron microscopy, hich is equiped with 30 31 Everhart-Thornley detector of secondary electrons. EDX analyses has been carried out by EM FEG 32 33 OXFORD machine with SSD X-MAX detector. 34 35 36 37 38 B.5. TEM/EDX/EELS. Crystallographic characterization and imaging were conducted using a field 39 40 emission gun FEI Tecnai F20 microscope at 200 kV with a point-to-point resolution of 0.24 nm. Energy 41 42 dispersive X-ray (EDX) and electron energy loss spectroscopy (EELS) spectrum images and profiles were 43 44 45 obtained in high angle annular dark-field (HAADF) STEM mode with an EDAX super ultra-thin window 46 47 (SUTW) X-ray detector a Gatan Quantum SE 963 imaging filter respectively. TEM cross-sections were 48 49 prepared by conventional mechanical polishing and ion milling. 50 51 52 53 54 55 56 57 58 29 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 30 of 42 1 2 3 B.6. XPS. X-ray photoemission spectroscopy (XPS) measurements were performed with a Phoibos 150 4 5 6 analyzer (SPECS GmbH, Berlin, Germany) in ultra-high vacuum conditions (base pressure 3×10-10 mbar). 7 8 XPS measurements were performed with a monochromatic Al Ka X-ray source (1486.74 eV). The main 9 10 ZnGa2O4 Ga2p and Zn2p peak energies were consistent with previous literature after adventitious C1s 11 12 13 calibration (284.8 eV). The C1s calibration shift for the reference ZnGa 2O4 was only of 0,595 eV. The 14 15 C1s calibration shift energies were also relatively small for the other specimens with -0.3 eV and 2,8 for 16 17 reference (n-type) ZnO and (highly resistive) Ga2O3, respectively. 18 19 20 21 22 B.7. Transport. Hall Effect measurement set-up. Ohmic contacts were prepared by silver paint at the four 23 24 corners of the sample. Hall Effect measurements were performed in a Van der Pauw configuration in the 25 26 temperature range of 300 K to 850 K and for magnetic fields perpendicular to the film plane varying from 27 28 29 -1.6 T to 1.6 T, using a high impedance measurement set-up which was custom designed for measurement 30 31 of high resistance. 32 33 34 35 36 C. Ab-initio Methods 37 38 For the hexagonal ZnO cell containing 2 Zn atoms and 2 oxygen atoms. The lattice constants are: a = b = 39 40 3.249 Å, c = 5.207 Å. The generalized gradient approximation with the mean-field Hubbard correction 41 42 (GGA+U), together with the Perdew-Burke-Ernzerhof functional (PBE), cut-off energy of 140 Hartree 43 44 45 and 23×23×13 k-points grid were used for the simulation. The U parameters are set to 4 eV and 9.5 eV 46 47 for O2p and Zn3d respectively. For the face-centered cubic ZnGa2O4, local density approximation with 48 49 semi-empirical approach (LDA-1/2) was used in the simulation. k-points sampling are set to 20×20×20, 50 51 52 and the lattice constant is a = b = c =8.338 Å. For the monoclinic β-Ga2O3, Meta_GGA (one of semi- 53 54 local approximations functionals, an alternative of the DFT-1/2) was used in the simulation. The k-points 55 56 57 58 30 59 60 ACS Paragon Plus Environment Page 31 of 42 Crystal Growth & Design 1 2 3 sampling was set to 4×8×8, and the monoclinic cell contains 16 Ga atoms and 24 O atoms. The lattice 4 5 6 constants are: a = 12.23 Å, b = 6.08 Å, c = 5.8 Å, and β = 103.7°. 7 8 For the ZnGa2O4 antisite pair DFT simulations, a 112-atom supercell was built (Zn16Ga32O64). The 9 10 calculation was conducted with the Quantumwise software. The antisite model was firstly optimized 11 12 13 geometrically using the GGA PBE functional. The optimized model was then calculated with a 4×4×4 k- 14 15 point grid using the meta-GGA method. The c parameter was chosen as 1.45 to match with the 16 17 experimental bandgap of the ZnGa2O4 host. Density mesh cut-off was set to 125 Hartree, and the 18 19 broadening was set to 25 meV, which had the energy convergence up to 1×10-4 eV. 20 21 22 23 24 25 26 27 28 29 30 31 References 32 33 [1] Horng, Ray-Hua, Chiung-Yi Huang, Sin-Liang Ou, Tzu-Kuang Juang, and Po-Liang Liu, “Epitaxial Growth 34 35 of ZnGa2O4: A New, Deep Ultraviolet Semiconductor Candidate” Cryst. Growth Des. 2017, 17, 6071–78. 36 37 38 39 [2] J. Y. Tsao, S. Chowdhury, M. A. Hollis, D. Jena, N. M. Johnson, K. A. Jones, R. J. Kaplar, S. Rajan, C. G. Van 40 41 de Walle, E. Bellotti, C. L. Chua, R. Collazo, M. E. Coltrin, J. A. Cooper, K. R. Evans, S. Graham, T. A. Grotjohn, 42 43 E. R. Heller, M. Higashiwaki, M. S. Islam, P. W. Juodawlkis, M. A. Khan, A. D. Koehler, J. H. Leach, U. K. Mishra, 44 45 46 R. J. Nemanich, R. C. N. Pilawa-Podgurski, J. B. Shealy, Z. Sitar, M. J. Tadjer, A. F. Witulski, M. Wraback, J. A. 47 48 Simmons, "Ultrawide‐Bandgap Semiconductors: Research Opportunities and Challenges" Adv. Electron. 49 50 Mater. 2018, 4, 1600501. 51 52 53 54 55 56 57 58 31 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 32 of 42 1 2 3 [3] L. Nagarajan, R. A. De Souza, D. Samuelis, I. Valov, A. Börger, J. Janek, K.-D. Becker, P. C. Schmidt, M. 4 5 6 Martin, "A chemically driven insulator–metal transition in non-stoichiometric and amorphous gallium oxide" 7 8 Nat. Mater. 2008, 7, 391. 9 10 11 [4] E. Chikoidze, D. J. Rogers, F. H. Teherani, C. Rubio, G. Sauthier, H. J. Von Bardeleben, T. Tchelidze, C. Ton- 12 13 That, A. Fellous, P. Bove, E. V. Sandana, Y. Dumont, A. Perez-Tomas, "Puzzling robust 2D metallic 14 15 16 conductivity in undoped β-Ga2O3 thin films" Materials Today Physics 2019, 8, 10. 17 18 19 [5] J. Kim, T. Sekiya, N. Miyokawa, N. Watanabe, K. Kimoto, K. Ide, Y. Toda, S. Ueda, N. Ohashi, H. Hiramatsu, 20 21 H. Hosono, T. Kamiya, "Conversion of an ultra-wide bandgap amorphous oxide insulator to a semiconductor" 22 23 24 NPG Asia Materials 2017, 9, e359. 25 26 27 [6] P. Gorai, R. W. McKinney, N. M. Haegel, A. Zakutayev, V. Stevanovic, "A computational survey of 28 29 semiconductors for power electronics" Energy Environ. Sci. 2019, 12, 3338; S. B. Reese, T. Remo, J. Green, 30 31 32 A. Zakutayev, "How Much Will Gallium Oxide Power Electronics Cost?" Joule 2019, 3, 903. 33 34 35 [7] A. Pérez-Tomás, E. Chikoidze, M.R. Jennings, S.A.O. Russell, F.H. Teherani, P. Bove, E.V. Sandana, D.J. 36 37 Rogers, "Wide and ultra-wide bandgap oxides: where paradigm-shift photovoltaics meets transparent 38 39 power electronics" Proc. SPIE 10533, Oxide-based Materials and Devices IX, 105331Q (2018); 40 41 42 https://doi.org/10.1117/12.2302576 43 44 45 [8] S. J. Pearton, F. Ren, M. Tadjer, J. H. Kim, "Perspective: Ga2O3 for ultra-high power rectifiers and MOSFETS" 46 47 J. Appl. Phys. 2018, 124, 220901. 48 49 50 51 [9] M. Higashiwaki, G. H. Jessen, "The dawn of gallium oxide microelectronics" Appl. Phys. Lett. 2018, 112, 52 53 060401. 54 55 56 57 58 32 59 60 ACS Paragon Plus Environment Page 33 of 42 Crystal Growth & Design 1 2 3 [10] A. Pérez-Tomás, E. Chikoidze, Y. Dumont, M. R. Jennings, S. O. Russell, P. Vales-Castro, G. Catalan, M. 4 5 6 Lira-Cantú, C. Ton –That, F. H. Teherani, V. E. Sandana, P. Bove, D. J. Rogers, "Giant bulk photovoltaic effect 7 8 in solar cell architectures with ultra-wide bandgap Ga2O3 transparent conducting electrodes" Mater. Today 9 10 Energy. 2019, 14, 100350. 11 12 13 [11] Y. S. Lee, D. Chua, R. E. Brandt, S. C. Siah, J. V. Li, J. P. Mailoa, S. W. Lee, R. G. Gordon, T. Buonassisi, 14 15 16 "Atomic Layer Deposited Gallium Oxide Buffer Layer Enables 1.2 V Open‐Circuit Voltage in Cuprous Oxide 17 18 Solar Cells" Adv. Mater. 2014, 26, 4704. 19 20 21 [12] W.-Y. Kong, G.-A. Wu, K.-Y. Wang, T.-F. Zhang, Y.-F. Zou, D.-D. Wang, L.-B. Luo, "Graphene‐β‐Ga2O3 22 23 24 Heterojunction for Highly Sensitive Deep UV Photodetector Application" Adv. Mater. 2016, 28, 10725. 25 26 27 [13] K. Ellmer, "Past achievements and future challenges in the development of optically transparent 28 29 electrodes" Nat. Photonics. 2012, 6, 809. 30 31 32 33 [14] X. Yu, T. J. Marks, A. Facchetti, "Metal oxides for optoelectronic applications" Nat. Mater. 2016, 15, 383. 34 35 36 [15] A. Pérez‐Tomás, "Functional Oxides for Photoneuromorphic Engineering: Toward a Solar Brain" Adv. 37 38 Mater. Interfaces. 2019, 0, 1900471. 39 40 41 42 [16] Z. Galazka, S. Ganschow, R. Schewski, K. Irmscher, D. Klimm, A. Kwasniewski, M. Pietsch, A. Fiedler, I. 43 44 Schulze-Jonack, M. Albrecht, T. Schröder, M. Bickermann, "Ultra-wide bandgap, conductive, high mobility, 45 46 and high quality melt-grown bulk ZnGa2O4 single crystals" APL Mater. 2019, 7, 022512 [17] L.-C. Cheng, 47 48 49 C.-Y. Huang, R.-H. Horng, "Thickness Effect on Operational Modes of ZnGa 2O4 MOSFETs" IEEE J. Electron 50 51 Devices Soc. 2018, 6, 432. 52 53 54 55 56 57 58 33 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 34 of 42 1 2 3 [18] M. Stoica, C. S Lo, "P-type zinc oxide spinels: application to transparent conductors and spintronics" New 4 5 6 J. Phys. 2014, 16, 055011. 7 8 9 [19] J. D. Perkins, T. R. Paudel, A. Zakutayev, P. F. Ndione, P. A. Parilla, D. L. Young, S. Lany, D. S. Ginley, A. 10 11 Zunger, N. H. Perry, Y. Tang, M. Grayson, T. O. Mason, J. S. Bettinger, Y. Shi, M. F. Toney, " Inverse design 12 13 approach to hole doping in ternary oxides: Enhancing p-type conductivity in cobalt oxide spinels" Phys. Rev. 14 15 16 B 2011, 84, 205207; A. Zakutayev, T. R. Paudel, P. F. Ndione, J. D. Perkins, S. Lany, A. Zunger, D. S. Ginley, 17 18 "Cation off-stoichiometry leads to high p-type conductivity and enhanced transparency in Co 2ZnO4 and 19 20 Co2NiO4 thin films" Phys. Rev. B 2012, 85, 085204. 21 22 23 24 [20] Q. Zhao, Z. Yan, C. Chen, J. Chen, "Spinels: Controlled Preparation, Oxygen Reduction/Evolution Reaction 25 26 Application, and Beyond" Chem. Rev. 2017, 117, 10121. 27 28 29 [21] S. Chu, Y. Cui, N. Liu, "The path towards sustainable energy" Nat. Mater. 2016, 16, 16. 30 31 32 33 [22] A. Kwade, W. Haselrieder, R. Leithoff, A. Modlinger, F. Dietrich, K. Droeder, "Current status and 34 35 challenges for automotive battery production technologies" Nat. Energy 2018, 3, 290. 36 37 38 [23] L. A.-W. Ellingsen, C. R. Hung, G. Majeau-Bettez, B. Singh, Z. Chen, M. S. Whittingham, A. H. Strømman, 39 40 41 "Nanotechnology for environmentally sustainable electromobility" Nat. Nanotechnol. 2016, 11, 1039. 42 43 44 [24] P. Canepa, S.-H. Bo, G. Sai Gautam, B. Key, W. D. Richards, T. Shi, Y. Tian, Y. Wang, J. Li, G. Ceder, "High 45 46 magnesium mobility in ternary spinel chalcogenides" Nat. Commun. 2017, 8, 1759. 47 48 49 50 [25] C. Wu, S. Gu, Q. Zhang, Y. Bai, M. Li, Y. Yuan, H. Wang, X. Liu, Y. Yuan, N. Zhu, F. Wu, H. Li, L. Gu, J. Lu, 51 52 "Electrochemically activated spinel manganese oxide for rechargeable aqueous aluminum battery" Nat. 53 54 Commun. 2019, 10, 73. 55 56 57 58 34 59 60 ACS Paragon Plus Environment Page 35 of 42 Crystal Growth & Design 1 2 3 [26] Y. Wang, Y. Yang, S. Jia, X. Wang, K. Lyu, Y. Peng, H. Zheng, X. Wei, H. Ren, L. Xiao, J. Wang, D. A. Muller, 4 5 6 H. D. Abruña, B. J. Hwang, J. Lu, L. Zhuang, "Synergistic Mn-Co catalyst outperforms Pt on high-rate oxygen 7 8 reduction for alkaline polymer electrolyte fuel cells" Nat. Commun. 2019, 10, 1506. 9 10 11 27] C. Zeng, T. Hu, N. Hou, S. Liu, W. Gao, R. Cong, T. Yang, "Photocatalytic pure water splitting activities for 12 13 ZnGa2O4 synthesized by various methods" Mater. Res. Bull. 2015, 61, 481. 14 15 16 17 [29] M. Zhong, Y. Li, I. Yamada, J.-J. Delaunay, "ZnO–ZnGa2O4 core–shell nanowire array for stable 18 19 photoelectrochemical water splitting" Nanoscale 2012, 4, 1509. 20 21 22 23 [28] W. Zhang, J. Zhang, Z. Chen, T. Wang, "Photocatalytic degradation of methylene blue by ZnGa 2O4 thin 24 25 films" Catal. Commun. 2009, 10, 1781. 26 27 28 [30] S. C. Yan, S. X. Ouyang, J. Gao, M. Yang, J. Y. Feng, X. X. Fan, L. J. Wan, Z. S. Li, J. H. Ye, Y. Zhou, Z. G. Zou, 29 30 "A Room‐Temperature Reactive‐Template Route to Mesoporous ZnGa 2O4 with Improved Photocatalytic 31 32 33 Activity in Reduction of CO2" Angewandte Chemie Int.Ed. 2010, 49, 6400. 34 35 36 [31] W. Yang, J. Li, B. Liu, X. Zhang, C. Zhang, P. Niu, X. Jiang, "Multi-wavelength tailoring of a ZnGa2O4 37 38 nanosheet phosphor via defect engineering" Nanoscale 2018, 10, 19039. 39 40 41 42 [32] A. Bessière, S. Jacquart, K. Priolkar, A. Lecointre, B. Viana, D. Gourier, "ZnGa2O4:Cr3+: a new red long- 43 44 lasting phosphor with high brightness" Opt. Express 2011, 19, 10131. 45 46 47 [33] X. Hou, T. Xuan, H. Sun, X. Chen, H. Li, L. Pan, "High-performance perovskite solar cells by incorporating 48 49 50 a ZnGa2O4:Eu3+ nanophosphor in the mesoporous TiO2 layer" Sol. Energy Mater. Sol. Cells 2016, 149, 121. 51 52 53 54 55 56 57 58 35 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 36 of 42 1 2 3 [34] T. Maldiney, A. Bessière, J. Seguin, E. Teston, S. K. Sharma, B. Viana, A. J. J. Bos, P. Dorenbos, M. Bessodes, 4 5 6 D. Gourier, D. Scherman, C. Richard, "The in vivo activation of persistent nanophosphors for optical imaging 7 8 of vascularization, tumours and grafted cells" Nat. Mater. 2014, 13, 418. 9 10 11 [35] T. Maldiney, B. Ballet, M. Bessodes, D. Scherman, C. Richard, "Mesoporous persistent nanophosphors 12 13 for in vivo optical bioimaging and drug-delivery" Nanoscale 2014, 6, 13970. 14 15 16 17 [36] H. Kawazoe, K. Ueda, "Transparent Conducting Oxides Based on the Spinel Structure" J. Am. Ceram. Soc. 18 19 2004, 82, 3330. 20 21 22 23 [37] S.-H. Tsai, S. Basu, C.-Y. Huang, L.-C. Hsu, Y.-G. Lin, R.-H. Horng, "Deep-Ultraviolet Photodetectors Based 24 25 on Epitaxial ZnGa2O4 Thin Films" Sci. Rep. 2018, 8, 1. 26 27 28 [38] Z. Lou, L. Li, G. Shen, "High-performance rigid and flexible ultraviolet photodetectors with single- 29 30 crystalline ZnGa2O4 nanowires" Nano Res. 2015, 8, 2162. 31 32 33 34 [39] N. W. Grimes, "The spinels: versatile materials" Physics in Technology 1975, 6, 22–27; S. Mesoraca, S. 35 36 Knudde, D. C. Leitao, S. Cardoso, M. G. Blamire, "All-spinel oxide Josephson junctions for high-efficiency spin 37 38 filtering" J. Phys.: Condens. Matter 2018, 30, 015804. 39 40 41 42 [40] J. Muro-Cruces, A. G. Roca, A. López-Ortega, E. Fantechi, D. del-Pozo-Bueno, S. Estradé, F. Peiró, B. 43 44 Sepúlveda, F. Pineider, C. Sangregorio, J. Nogues, "Precise Size Control of the Growth of Fe3O4 Nanocubes 45 46 over a Wide Size Range Using a Rationally Designed One-Pot Synthesis" ACS Nano 2019, 13, 7716. 47 48 49 50 [41] Y. Okamoto, S. Niitaka, M. Uchida, T. Waki, M. Takigawa, Y. Nakatsu, A. Sekiyama, S. Suga, R. Arita, H. 51 52 Takagi, "Band Jahn-Teller Instability and Formation of Valence Bond Solid in a Mixed-Valent Spinel Oxide 53 54 LiRh2O4" Phys. Rev. Lett. 2008, 101, 086404. 55 56 57 58 36 59 60 ACS Paragon Plus Environment Page 37 of 42 Crystal Growth & Design 1 2 3 [42] H. Peng, A. Zakutayev, S. Lany, T. R. Paudel, M. d’Avezac, P. F. Ndione, J. D. Perkins, D. S. Ginley, A. R. 4 5 6 Nagaraja, N. H. Perry, T. O. Mason, A. Zunger, "Li-Doped Cr2MnO4: A New p-Type Transparent Conducting 7 8 Oxide by Computational Materials Design" Adv. Funct. Mater. 2013, 23, 5267. 9 10 11 [43] M. N. Amini, H. Dixit, R. Saniz, D. Lamoen, B. Partoens, "The origin of p-type conductivity in ZnM2O4 (M = 12 13 Co, Rh, Ir) spinels" Phys. Chem. Chem. Phys. 2014, 16, 2588. 14 15 16 17 [44] T. R. Paudel, A. Zakutayev, S. Lany, M. d’Avezac, A. Zunger, "Doping Rules and Doping Prototypes in A2BO4 18 19 Spinel Oxides" Adv. Funct. Mater. 2011, 21, 4493. 20 21 22 23 [45] Y. Xia, T. Wang, X. Zhao, X. Jiao, D. Chen, "Theoretical and Experimental Investigations on Effects of Native 24 25 Point Defects and Nitrogen Doping on the Optical Band Structure of Spinel ZnGa 2O4" J. Phys. Chem. C 2018, 26 27 122, 5509. 28 29 30 [46] A. De Vos, K. Lejaeghere, D. E. P. Vanpoucke, J. J. Joos, P. F. Smet, K. Hemelsoet, "First-Principles Study 31 32 33 of Antisite Defect Configurations in ZnGa2O4:Cr Persistent Phosphors" Inorg. Chem. 2016, 55, 2402. 34 35 36 [47] A. R. Nagaraja, N. H. Perry, T. O. Mason, Y. Tang, M. Grayson, T. R. Paudel, S. Lany, A. Zunger, "Band or 37 38 polaron: The hole conduction mechanism in the p-Type spinel Rh2ZnO4" J. Am. Ceram. Soc. 2012, 95, 269. 39 40 41 42 [48] M. Dekkers, G. Rijnders, D. H. A. Blank, "ZnIr2O4, a p-type transparent oxide semiconductor in the class of 43 44 spinel zinc-d6-transition metal oxide" Appl. Phys. Lett. 2007, 90, 021903. 45 46 47 [49] Ray- Hua Hrong, Yu-Yuan Zeng, Wei-Kai Wang, Chia-Lung Tsai, Yi-Keng Fu, and Wei-Hung Kuo, 48 49 50 "Transparent electrode design for AlGaN deep-ultraviolet light-emitting diodes" Optics Express 2017, 25, 51 52 32206–13. 53 54 55 56 57 58 37 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 38 of 42 1 2 3 [50] E. Chikoidze, A. Fellous, A. Perez-Tomas, G. Sauthier, T. Tchelidze, C. Ton-That, T. T. Huynh, M. Phillips, 4 5 6 S. Russell, M. Jennings, B. Berini, F. Jomard, Y. Dumont, "P-type β-gallium oxide : a new perspective for 7 8 power and optoelectronic devices" Mater. Today Phys. 2017, 3, 118. 9 10 11 [51] E. Chikoidze, C. Sartel, H. Mohamed, T. Tchelidze, M. Modreanu, C. Vales-Castro, C. Rubio, C. Arnold, V. 12 13 Sallet, Y. Dumont, A. Perez-Tomas, "Enhancing the intrinsic p-type conductivity of the ultra-wide bandgap 14 15 16 Ga2O3 semiconductor" J. Mater. Chem. C 2019, 10.1039.C9TC02910A. 17 18 19 [52] M. S. Relvas, M. R. N. Soares, S. O. Pereira, A. V. Girão, F. M. Costa, T. Monteiro, "Trends in Cr3+ red 20 21 emissions from ZnGa2O4 nanostructures produced by pulsed laser ablation in a liquid medium" J. Phys. Chem. 22 23 24 Solids 2019, 129, 413. 25 26 27 [53] R. Al-Gaashani, S. Radiman, A. R. Daud, N. Tabet, Y. Al-Douri, "XPS and optical studies of different 28 29 morphologies of ZnO nanostructures prepared by microwave methods" Ceram. Int. 2013, 39, 2283. 30 31 32 33 [54] N. Winkler, R. A. Wibowo, W. Kautek, G. Ligorio, E. J. W. List-Kratochvil, T. Dimopoulos, "Nanocrystalline 34 35 Ga2O3 films deposited by spray pyrolysis from water-based solutions on glass and TCO substrates" J. Mater. 36 37 Chem. C 2019, 7, 69. 38 39 40 41 [55] D.-Y. Cho, "Chemical and structural properties of ternary post-transition metal oxide thin films: InZnO, 42 43 InGaO and GaZnO" Curr. Appl. Phys. 2015, 15, 1337. 44 45 46 [56] M. C. Biesinger, L. W. M. Lau, A. R. Gerson, R. St. C. Smart, "Resolving surface chemical states in XPS 47 48 49 analysis of first row transition metals, oxides and hydroxides: Sc, Ti, V, Cu and Zn" Applied Surface Science 50 51 2010, 257, 887. 52 53 54 55 56 57 58 38 59 60 ACS Paragon Plus Environment Page 39 of 42 Crystal Growth & Design 1 2 3 [57] C. Li, J.-L. Yan, L.-Y. Zhang, G. Zhao, "Electronic structures and optical properties of Zn-doped β-Ga2O3 with 4 5 6 different doping sites" Chinese Phys. B 2012, 21, 127104. 7 8 9 [58] D. Skachkov, W. R. L. Lambrecht, "Computational study of electron paramagnetic resonance parameters 10 11 for Mg and Zn impurities in β-Ga2O3" Appl. Phys. Lett. 2019, 114, 202102. 12 13 14 15 [59] N. Li, X. Duan, F. Yu, H. Jiang, "Effects of preparation method and temperature on the cation distribution 16 17 of ZnGa2O4 spinel studied by X-ray photoelectron spectroscopy" Vacuum 2017, 142, 1. 18 19 20 [60] G. K. Wertheim, S. Hufner, "X-Ray Photoemission Band Structure of Some Transition-Metal Oxides" Phys. 21 22 23 Rev. Lett. 1972 28, 4. 24 25 26 [61] J. Xu, Y. Teng, F. Teng, "Effect of Surface Defect States on Valence Band and Charge Separation and 27 28 Transfer Efficiency" Sci. Rep. 2016, 6, 32457. 29 30 31 32 [62] B.J. Carey, J. Z. Ou, R.M. Clark, K.J. Berean, A. Zavabeti, A.S.R. Chesman, S.P. Russo, D.W.M. Lau, Z-Q. Xu, 33 34 Q. Bao, O. Kavehei, B. C. Gibson, M. D. Dickey, R. B. Kaner, T. Daeneke, K. K-Zadeh, "Wafer-scale two- 35 36 dimensional semiconductors from printed oxide skin of liquid metals" Nat Commun 2017 8, 14482. 37 38 39 40 [63] M. Michling, D. Schmeißer, "Resonant Photoemission at the O1s threshold to characterize β-Ga2O3 single 41 42 crystals" IOP Conf. Ser.: Mater. Sci. Eng. 2012, 34, 012002. 43 44 45 [64] F. Fuchs, J. Furthmüller, F. Bechstedt, M. Shishkin, G. Kresse, "Quasiparticle band structure based on a 46 47 generalized Kohn-Sham scheme" Phys. Rev. B 2007, 76, 115109. 48 49 50 51 [65] A. Navarro-Quezada, S. Alamé, N. Esser, J. Furthmüller, F. Bechstedt, Z. Galazka, D. Skuridina, P. Vogt, " 52 53 Near valence-band electronic properties of semiconducting β−Ga2O3 (100) single crystals" Phys. Rev. B 2015, 54 55 92, 195306. 56 57 58 39 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 40 of 42 1 2 3 [66] J. B. Varley, J. R. Weber, A. Janotti, C. G. Van de Walle, "Oxygen vacancies and donor impurities in β- 4 5 6 Ga2O3 " Appl. Phys. Lett. 2010, 97, 142106. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 40 59 60 ACS Paragon Plus Environment Page 41 of 42 Crystal Growth & Design 1 2 3 4 5 6 7 8 For Table of Contents Use Only 9 10 11 12 P-type ultra-wide bandgap spinel ZnGa2O4: new perspectives for 13 14 15 energy electronics 16 17 18 19 20 21 Ekaterine Chikoidze, Corinne Sarte, Ismail Madac, Hagar Mohamed, Christele Vilar, Belén Ballesteros, 22 23 Francisco Belarr, Elena del Corro, Pablo Vales-Castr, Guillaume Sauthier, Lijie Li, Mike Jennings, 24 25 Vincent Salle, Yves Dumont and Amador Pérez-Tomás 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 41 59 60 ACS Paragon Plus Environment Crystal Growth & Design Page 42 of 42 1 2 3 We have demonstrated that Ultra-wide band gap (~5 eV) spinel zinc gallate (ZnGa2O4) is the native p- 4 5 6 type ternary oxide semiconductor. This is important step towards p-type doping, opens up further 7 8 perspectives for ultra-wide bandgap bipolar spinel electronics to be integrated into the next generation 9 10 more efficient energy optoelectronics and power electronics. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 42 59 60 ACS Paragon Plus Environment