Open Access
Add to BibSonomyAbstract
Single crystallinity (AlGa)2O3 solar-blind photodetectors are epitaxially grown on sapphire. Oxygen pressure during the growth has a great impact on the Al composition in (AlGa)2O3, which is investigated utilizing X-ray photoelectron spectroscopy and X-ray diffraction measurements. Measured transmittance spectra and responsivity demonstrate that (AlGa)2O3 photodetectors achieve a wider bandgap compared to a Ga2O3 device. An (Al0.12Ga0.88)2O3 device obtains 10 times higher photocurrent Iphoto than a Ga2O3 photodetector. However, as Al composition increases, significant Iphoto degradation is observed in an (Al0.35Ga0.65)2O3 photodetector. Meanwhile, an (Al0.35Ga0.65)2O3 device exhibits the stronger persistent photoconductivity compared to the Ga2O3 control device. Analysis shows that defect states in a bandgap of (AlGa)2O3 might be associated with the performance change in (AlGa)2O3 photodetectors.
© 2017 Optical Society of America
1. Introduction
Ga2O3 has attracted great interests for high power and ultraviolet detection applications due to its wide bandgap EG, chemical and thermal stability. Ga2O3 photodetectors with the solar-blind sensitivity covering major deep ultraviolet (DUV) regions have been demonstrated experimentally [1–6]. Incorporating Al into Ga2O3 to form the ternary alloys, i.e. (AlGa)2O3, makes it feasible to adjust the cut-off wavelength λ of the devices by varying the Al compositions. Furthermore, wider EG engineering is critical to realize the higher breakdown voltage of the materials, benefitting the scaling of the photodetectors, which helps to improve the frequency response characteristics of the devices. (AlGa)2O3 compounds with single crystallinity have been obtained by pulsed laser deposition [7, 8], molecular beam epitaxy (MBE) [9, 10], and mist chemical vapor deposition [11], and their EG values were studied using transmittance and X-ray photoelectron spectroscopy (XPS) measurements [7,12]. EG dependence on Al composition in the materials can be directly obtained by characterizing the (AlGa)2O3 photodetectors. However, there is still a lack of study on fabrication of (AlGa)2O3 photodetectors.
In this paper, (AlGa)2O3 solar-blind photodetectors with wider EG compared to Ga2O3 control device are fabricated on sapphire. Crystal qualities and material compositions of (AlGa)2O3 are characterized. Electrical characteristics of the devices are investigated.
2. Experiments
β-(AlGa)2O3 films were grown on double-polished (0001)-oriented sapphire with a Ga2O3 buffer by laser MBE at 610 °C. The cylindrical stoichiometric (Al0.12Ga0.88)2O3 and Ga2O3 targets with a purity of 99.99% were irradiated by the focused KrF excimer laser beam with a frequency of 3 Hz and a pulse energy density 2.0 J/cm2. The oxygen pressure PO2 in the chamber could be tunable. (AlGa)2O3 layers for device samples were epitaxially grown with the oxygen pressure of 0.01 and 0.05 mbar. It was found that PO2 had a great impact on Al composition in (AlGa)2O3. Ga2O3 buffer layer was necessary to improve the (AlGa)2O3 crystalline quality and avoid the formation of polycrystalline film. Ga2O3 control sample was also grown using the similar conditions. The thickness of the epitaxial layer was determined by the total number of the laser pulses. Growth conditions for the samples used in this work are listed in Table 1. (AlGa)2O3 and Ga2O3 layers have an unintentional n-type doping, and the carrier concentration is ranging from 1015 to 1018 cm−3.
Table 1. Growth Conditions for (AlGa)2O3 and Ga2O3 Samples.
The fabricated (AlGa)2O3 and Ga2O3 photodetectors have the interdigital Schottky contacts, consisting of 10 nm Ti and 100 nm Au. The finger spacing and total length are 100 μm and 17.5 mm, respectively.
3. Results and discussion
3.1 Material characterization
Cross-sectional transmission electron microscope (XTEM) image in Fig. 1(a) shows that (AlGa)2O3 grown with 4000 pulses at 0.01 mbar PO2 has a thickness of 67 nm. Figure 1(b) depicts that β-Ga2O3 plane is parallel to the α-Al2O3 (0006) plane. Some contrast observed at sapphire/Ga2O3 interface indicates that there might be a small amount of α-phase in the transition region at interface [13]. Figure 1(c) illustrates high resolution TEM image of the (AlGa)2O3/Ga2O3 interface, showing that single crystal (AlGa)2O3 is epitaxially grown on Ga2O3 buffer.
Fig. 1 (a) XTEM image of (AlGa)2O3 epitaxially grown on sapphire with a Ga2O3 buffer. (b) and (c) High resolution TEM images of the selected regions in red dotted and blue dashed boxes, respectively, in (a), showing the single crystallinity of (AlGa)2O3 and Ga2O3 layers.
The stoichiometries of (AlGa)2O3 and Ga2O3 films were investigated by XPS measurement, and Al2p, Ga2p3/2, and O1s peaks of the samples are illustrated in Fig. 2(a)-(c). Before XPS measurement, a few nanometers of surface were removed using in situ plasma etching, and the peaks were calibrated using adventitious C1s of 284.6 eV. As compared to the Ga2O3, (AlGa)2O3 samples exhibit the obvious peaks shift towards the higher binding energy. The Al2p peaks are fitted using two Gaussian peaks, corresponding to (AlGa)2O3 and AlOx phases [Fig. 2(a)]. The observed AlOx is probably due to the fact that there is AlOx phase in the target. Based on the area of the peaks and the corresponding sensitivity factors, the relative ratios of Al to Ga in (AlGa)2O3 phases are calculated to be 35:65 and 12:88 for sample A and B, respectively. To confirm the XPS results, Energy-dispersive X-ray spectroscopy (EDS) measurement was carried out on (AlGa)2O3 Sample B using scanning electron microscope (SEM), as shown in Fig. 2(d). To ensure the X-ray excitation totally came from the (AlGa)2O3 film, during the EDS measurement, we used the 45 degree grazing incidence and reduced the accelerating voltage of electron beam less than 10 kV. EDS results demonstrate a ratio of Al and Ga atoms of 12.1:87.9 in (AlGa)2O3 Sample B, which is well consistent with the XPS measurement. Sample A, i.e. (Al0.35Ga0.65)2O3, has a higher Al composition than that in the (Al0.12Ga0.88)2O3 target, which is because the decomposition of GaOx on surface during growth leads to the loss of Ga atoms. As PO2 increases, decomposition rate of GaOx is greatly reduced, resulting in the increased growth rate and reduced Al composition in sample B. There is another practical reason for the Al composition dependence on PO2. Al atoms, much lighter than Ga, are influenced by the stronger scattering in the growth chamber, which leads to the reduction of the ratio of Al/Ga flux reaching the substrate, as PO2 increases. It is noted that the calculated oxygen atom percent in (AlGa)2O3 and Ga2O3 samples is in the range of 52%~54%, indicating that the formation of a large number of oxygen vacancies cannot be avoided by increasing PO2.
Fig. 2 XPS (a) Al2p, (b) Ga2p3/2, and (c) O1s spectra of (AlGa)2O3 and Ga2O3 samples. (d) EDS of (AlGa)2O3 Sample B, and inset shows the SEM image and the detection region.
Figure 3 presents the typical ,, andX-ray diffraction (XRD) ω-2θ scans for (AlGa)2O3 samples. The peaks from (AlGa)2O3 and Ga2O3 buffer can be clearly separated for (Al0.35Ga0.65)2O3 sample. (Al0.35Ga0.65)2O3 peaks are located right next to those of Ga2O3 buffer, demonstrating the smaller lattice constants in (Al0.35Ga0.65)2O3 compared to Ga2O3. For (Al0.12Ga0.88)2O3, single peak is observed for each face – that is, the XRD peaks corresponding to the same face of (AlGa)2O3 and Ga2O3 buffer might get too close to be distinguished [Fig. 3(b)]. This also proves that Al composition in (Al0.12Ga0.88)2O3 is much lower in comparison with (Al0.35Ga0.65)2O3. As pseudomorphically grown on Ga2O3 buffer, (Al0.35Ga0.65)2O3 contains a high in-plane tensile strain. The measured interplanar spacing of faces of (Al0.35Ga0.65)2O3 is 0.1554nm, which is 0.96% smaller than that calculated based on the data in [12]. This indicates that (Al0.35Ga0.65)2O3 is really under a very obvious out-of-plane compressive strain.
The values of EG of (AlGa)2O3 were characterized by measuring the transmittance spectra of the samples, as shown in Fig. 4(a). It is clearly observed that the (AlGa)2O3 samples have a blue shift in the absorption edges compared to the Ga2O3 sample, and the absorption edge of (AlGa)2O3 shifts towards shorter λ with the increasing of Al composition. The absorption coefficient α of the film was calculated using the relation α = (1/t)ln[(1-Rc)2/T] [14], where T and Rc are the transmittance and reflectance, respectively, and t is the film thickness. The relation between α and incident photon energy can be expressed by (αhν) = B(hν-EG)1/2, where hν is the energy of the incident photon and B is the absorption edge width parameter [14]. Figure 4(b) shows the plots of (αhν)2 as a function of hν for the samples. The EG of Ga2O3 control, (Al0.12Ga0.88)2O3, and (Al0.35Ga0.65)2O3 samples, evaluated by the extrapolation of the linear regions to the horizontal axis, are 4.89, 5.06, and 5.29 eV, respectively.
Fig. 4 (a) Transmittance spectra of (AlGa)2O3 and Ga2O3 samples. (b) (αhν)2 as a function of hν for the samples. The extrapolation of the linear regions to the horizontal axis determines the EG values.
3.2 Electrical characteristics of the devices
To investigate the responsivity R and photocurrent Iphoto characteristics of (AlGa)2O3 photodetectors, optical measurements using a low-pressure mercury lamp U3900 with the illumination λ from 254 to 365 nm and the various power densities Plight were carried out. The R is defined as R = (Iphoto-Idark)/(PlightS), where Iphoto and Idark are the photo current and the dark current, respectively, and S is the effective illuminated area [15]. Figure 5 shows that (Al0.35Ga0.65)2O3 and (Al0.12Ga0.88)2O3 detectors demonstrate 290 and 180 meV blue shift in maximum R, Rmax, compared with the Ga2O3 device. This proved that (AlGa)2O3 photodetectors have a larger EG over Ga2O3 device, and EG increases with the Al composition. These are consistent with the results obtained using the transmittance spectra in Fig. 4. EG value of (Al0.12Ga0.88)2O3 is also consistent with the reported results in [7] and [12]. Nevertheless, (Al0.35Ga0.65)2O3 has a smaller EG relative to the reported ones [7,12], and we infer that decreased EG is attributed to the in-plane tensile strain in the epitaxial (Al0.35Ga0.65)2O3 film. The Rmax of (Al0.35Ga0.65)2O3 photodetector is located at 238 nm, which is very close to that of AlGaN device with Al composition up to 70% [16].
Fig. 5 (a)-(c) R versus illumination optical λ for the (AlGa)2O3 and Ga2O3 photodetectors at Vbias of 20 and 40 V. (AlGa)2O3 devices exhibit the obvious blue shift in Rmax compared with the Ga2O3 device.
(Al0.12Ga0.88)2O3 photodetector achieves the much improved R in comparison with the other two devices. Figure 6 compares the Rmax of the devices, showing that a 15.7 times higher Rmax is obtained in (Al0.12Ga0.88)2O3 device in comparison with the Ga2O3 control. Rmax of (Al0.12Ga0.88)2O3 device is 1.5 A/W, and external quantum efficiency (EQE) is calculated to be 783% using EQE = hcRmax/(eλ), where h is Planck’s constant, c is the velocity of light, and e is the electronic charge. This high EQE might be due to the optical gain induced by avalanche-like multiplication [17]. From Fig. 5(a), it is observed that the R curves of (Al0.35Ga0.65)2O3 have an absorption tail extending to the longer λ, up to 300 nm. This indicates that there are a large number of defect states with various energy levels in the EG of (Al0.35Ga0.65)2O3, some of which could lead to the degradation of R. It seems that there are only shallow defects located in the EG of (Al0.12Ga0.88)2O3, hardly causing the broadened absorption edge [Fig. 5(b)]. R of devices decreases rapidly at λ shorter than that at Rmax, which is similar to that in [18]. This could be because energy loss occurs during the relaxation process of carriers in case of photon energy above EG of materials.
Fig. 6 Statistical plots showing that (Al0.12Ga0.88)2O3 photodetectors have the significantly improved Rmax compared to (Al0.35Ga0.65)2O3 and Ga2O3 devices.
Figure 7 represents Iphoto versus bias voltage Vbias characteristics for the devices at different Plight with λ corresponding to the Rmax for each photodetector. As shown in Fig. 7(b), Iphoto increases almost linearly with the Vbias, which is similar to Ref [19]. Iphoto of devices show obvious Plight dependence. About 10 times higher Iphoto is achieved in (Al0.12Ga0.88)2O3 photodetector compared to the Ga2O3 photodetector at Plight of 50 μW/cm2 and Vbias of 40 V. Moreover, (Al0.12Ga0.88)2O3 photodetector shows several times higher Idark than the Ga2O3 control. It is speculated that shallow impurity-like defects enhance the conductivity of (Al0.12Ga0.88)2O3, leading to the both increased Iphoto and Idark in the device. For example, oxygen vacancies in Ga2O3 were reported to act as the donor-like defects [20, 21], which improved the material conductivity. Oxygen vacancies might have the much higher activation rate in (Al0.12Ga0.88)2O3 than in Ga2O3 due to the presence of Al atoms. The impact of Al on the defects states induced by oxygen vacancies needs to be further investigated. It should be noted that, although Ti/Au Schottky contacts are used in the photodetectors, the obvious Schottky characteristics are not observed in the Iphoto-Vbias curves. This is because the undoped (AlGa)2O3 has the high resistivity, which results in the low on/off current ratio in the devices [18].
Fig. 7 (a) Logarithmic and (b) linear Iphoto versus Vbias with various Plight characteristics for (AlGa)2O3 and Ga2O3 photodetectors. (Al0.12Ga0.88)2O3 device achieved significantly enhanced Iphoto over the other two.
Compared to the Ga2O3 control, the (Al0.35Ga0.65)2O3 detector has the much lower Iphoto and the similar Idark. The possible reasons for the performance degradation are clarified. First, with the widening of EG, energy levels of defects get too deep to be activated, losing much of contribution to conductivity and photo absorption of (Al0.35Ga0.65)2O3. Second, deep defects formed with the increasing of Al composition capture the generated carriers by optical absorption, leading to the reduction of Iphoto. Finally, through experiment, we found that the wider EG in (Al0.35Ga0.65)2O3 resulted from the incorporation of Al leads to a larger Schottky barrier height at metal/semiconductor interface in comparison with the Ga2O3 device.
Time-dependent Iphoto characteristics of the photodetectors measured with a 4 s period at Plight of 50 μW/cm2 and Vbias of 20 V are shown in Fig. 8. After several illumination pulses, devices exhibit the stable Iphoto at the given Plight and Vbias. This is due to the influence of defect states on the optical transition behavior [22]. It is noted that, after the illumination is off, (AlGa)2O3 devices have the much slower decay process, compared to Ga2O3 photodetector. Decay processes of the devices demonstrate the two distinct relaxation durations: fast-response and slow-response. The former is directly related to the interband optical transition, while the latter might be resulted from the recombination of the carriers through the defects in EG. With the illumination λ corresponding to the Rmax, in the fast-response of decay, Iphoto values of (Al0.35Ga0.65)2O3, (Al0.12Ga0.88)2O3 and Ga2O3 devices drop to ~24%, ~37%, and ~58% of on-state photocurrent Ion, respectively. This indicates that the role of defects assisted optical transition in the photo response for (AlGa)2O3 photodetector is more pronounced compared with the Ga2O3 device. It is speculated that, compared to (Al0.12Ga0.88)2O3, (Al0.35Ga0.65)2O3 contains a larger number of deep traps, which release the more captured carriers at off-state, leading to the more significant persistent photoconductivity (PPC). Time-dependent Iphoto of devices measured at sub-EG λ exhibits a more serious PPC effect, which suggests that defect-assisted recombination of the carriers plays a more crucial role as illumination photon energy is less than EG of the material.
Fig. 8 Time-dependent Iphoto characteristics for (AlGa)2O3 and Ga2O3 photodetectors. As measured using a sub-EG λ, photodetectors exhibit the more serious PPC effect in time dependent Iphoto curve.
4. Conclusion
(Al0.35Ga0.65)2O3 and (Al0.12Ga0.88)2O3 photodetectors with the EG of 5.29 and 5.06 eV are fabricated on sapphire. (Al0.12Ga0.88)2O3 device demonstrates the significantly improved Iphoto and R in comparison with the Ga2O3 control detector, which might be due to the shallow impurity-like defects leading to the enhanced conductivity of the materials. R and time-dependent Iphoto measurements indicate that there are a large number of deep defects in (Al0.35Ga0.65)2O3, resulting in the degradation of Iphoto and pronounced PPC in the device.
Funding
National Natural Science Foundation of China (NSFC) (61534004 and 6133402).
Acknowledgments
Q. Feng wishes to thank State Key Laboratory of Crystal Materials (Shandong University)for the support and Prof. Dabing Li from Changchun Institute of Optics, Fine Mechanical and Physics of CAS for responsivity measurements.
References and links
1. T. Oshima, T. Okuno, and S. Fujita, “Ga2O3 thin film growth on c-plane sapphire substrates by molecular beam epitaxy for deep-ultraviolet photodetectors,” Jpn. J. Appl. Phys. 46(11), 7217–7220 (2007). [CrossRef]
2. F.-P. Yu, S.-L. Ou, and D.-S. Wuu, “Pulsed laser deposition of gallium oxide films for high performance solar-blind photodetectors,” Opt. Mater. Express 5(5), 1240–1249 (2015). [CrossRef]
3. D. Guo, Z. Wu, P. Li, Y. An, H. Liu, X. Guo, H. Yan, G. Wang, C. Sun, L. Li, and W. Tang, “Fabrication of β-Ga2O3 thin films and solar-blind photodetectors by laser MBE technology,” Opt. Mater. Express 4(5), 1067–1076 (2014). [CrossRef]
4. S. Oh, Y. Jung, M. A. Mastro, J. K. Hite, C. R. Eddy Jr, and J. Kim, “Development of solar-blind photodetectors based on Si-implanted β-Ga2O3.,” Opt. Express 23(22), 28300–28305 (2015). [CrossRef] [PubMed]
5. T.-C. Wei, D.-S. Tsai, P. Ravadgar, J.-J. Ke, M.-L. Tsai, D.-H. Lien, C.-Y. Huang, R.-H. Horng, and J.-H. He, “See-through Ga2O3 solar-blind photodetectors for use in harsh environments,” IEEE J. Sel. Top. Quantum Electron. 20(6), 3802006 (2014).
6. Q. Feng, L. Huang, G. Han, F. Li, X. Li, L. Fang, X. Xing, J. Zhang, W. Mu, Z. Jia, D. Guo, W. Tang, X. Tao, and Y. Hao, “Comparison study of β-Ga2O3 photodetectors on bulk substrate and sapphire,” IEEE Trans. Electron Dev. 63(9), 3578–3583 (2016). [CrossRef]
7. F. Zhang, K. Saito, T. Tanaka, M. Nishio, M. Arita, and Q. Guo, “Wide bandgap engineering of (AlGa)2O3 films,” Appl. Phys. Lett. 105(16), 162107 (2014). [CrossRef]
8. C. Kranert, M. Jenderka, J. Lenzner, M. Lorenz, H. von Wenckstern, R. Schmidt-Grund, and M. Grundmann, “Lattice parameters and Raman-active phonon modes of β-(AlxGa1-x)2O3,” J. Appl. Phys. 117(12), 125703 (2015). [CrossRef]
9. T. Oshima, T. Okuno, N. Arai, Y. Kobayashi, and S. Fujita, “β-Al2Ga2-2xO3 thin film growth by molecular beam epitaxy,” Jpn. J. Appl. Phys. 48(77R), 070202 (2009). [CrossRef]
10. Y. Oshima, E. Ahmadi, S. C. Badescu, F. Wu, and J. S. Speck, “Composition determination of β-(AlxGa1-x)2O3 layers coherently grown on (010) β-Ga2O3 substrates by high-resolution X-ray diffraction,” Appl. Phys. Express 9(6), 061102 (2016). [CrossRef]
11. H. Ito, K. Kaneko, and S. Fujita, “Growth and band gap control of corundum-structured α-(AlGa)2O3 thin films on sapphire by spray-assisted mist chemical vapor deposition,” Jpn. J. Appl. Phys. 51(10R), 100207 (2012).
12. B. W. Krueger, C. S. Dandeneau, E. M. Nelson, S. T. Dunham, F. S. Ohuchi, and M. A. Olmstead, “Variation of Band Gap and Lattice Parameters of β-(AlxGa1-x)2O3 Powder Produced by Solution Combustion Synthesis,” J. Am. Ceram. Soc. 99(7), 2467–2473 (2016). [CrossRef]
13. R. Schewski, G. Wagner, M. Baldini, D. Gogova, Z. Galazka, T. Schulz, T. Remmele, T. Markurt, H. von Wenckstern, M. Grundmann, O. Bierwagen, P. Vogt, and M. Albrecht, “Epitaxial stabilization of pseudomorphic α-Ga2O3 on sapphire (0001),” Appl. Phys. Express 8(1), 011101 (2015). [CrossRef]
14. S. K. Gullapalli, R. S. Vemuri, and C. V. Ramana, “Structural transformation induced changes in the optical properties of nanocrystalline tungsten oxide thin films,” Appl. Phys. Lett. 96(17), 171903 (2010). [CrossRef]
15. M. Zhong, Z. Wei, X. Meng, F. Wu, and J. Li, “High-performance single crystalline UV photodetectors of β-Ga2O3,” J. Alloys Compd. 619, 572–575 (2015). [CrossRef]
16. D. Walker, V. Kumar, K. Mi, P. Sandvik, P. Kung, X. H. Zhang, and M. Razeghi, “Solar-blind AlGaN photodiodes with very low cutoff wavelength,” Appl. Phys. Lett. 76(4), 403–405 (2000). [CrossRef]
17. T. Oshima, T. Okuno, N. Arai, N. Suzuki, S. Ohira, and S. Fujita, “Vertical solar-blind deep-ultraviolet Schottky photodetectors based on β-Ga2O3 substrate,” Appl. Phys. Express 1(1), 011202 (2008). [CrossRef]
18. G. C. Hu, C. X. Shan, N. Zhang, M. M. Jiang, S. P. Wang, and D. Z. Shen, “High gain Ga2O3 solar-blind photodetectors realized via a carrier multiplication process,” Opt. Express 23(10), 13554–13561 (2015). [CrossRef] [PubMed]
19. X. Z. Liu, P. Guo, T. Sheng, L. X. Qian, W. L. Zhang, and Y. R. Li, “β-Ga2O3 thin films on sapphire pre-seeded by homo-self-templated buffer layer for solar-blind UV photodetector,” Opt. Mater. 51, 203–207 (2016). [CrossRef]
20. J. B. Varley, J. R. Wever, A. Janotti, and C. G. V. D. Walle, “Oxygen vacancies and donor impurities in β-Ga2O3,” Appl. Phys. Lett. 97(14), 142106 (2010). [CrossRef]
21. K. Irmscher, Z. Galazka, M. Pietsch, R. Uecker, and R. Fornari, “Electrical properties of β-Ga2O3 single crystals grown by the Czochralskimethod,” J. Appl. Phys. 110(6), 063720 (2011). [CrossRef]
22. P. Ravadgar, R.-H. Horng, S.-D. Yao, H.-Y. Lee, B.-R. Wu, S.-L. Ou, and L.-W. Tu, “Effects of crystallinity and point defects on optoelectronic applications of β-Ga2O3 epilayers,” Opt. Express 21(21), 24599–24610 (2013). [CrossRef] [PubMed]
