Open Access
Add to BibSonomy
Abstract
Three AlGaN-based Schottky detector samples grown with varying pressure conditions are prepared and their responsivity is investigated. It is found that the responsivity of the three samples first increases and then decreases with the increase of pressure. In addition, the vacancy defect concentration increases and carbon impurities concentration decreases when the reactor pressure increases from 100 mbar to 200 mbar during the i-AlGaN layer growth. It is assumed that carbon impurities and vacancy defects play a negative role in detector’s performance, which increase the recombination of photogenerated carriers and reduce detector responsivity. The relationship between growth pressure and detector responsivity is not linear. It is necessary to select a suitable growth pressure to improve the performance of AlGaN detectors.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
Corrections
Yujie Huang, Jing Yang, Zongshun Liu, Feng Liang, Wei Jia, and Degang Zhao, "Non-monotonic effect of growth pressure on the responsivity of AlGaN ultraviolet Schottky detectors: publisher’s note," Opt. Mater. Express 13, 3365-3365 (2023)https://opg.optica.org/ome/abstract.cfm?uri=ome-13-11-3365
26 October 2023: A correction was made to an author name and affiliation.
1. Introduction
AlGaN-based ultraviolet (UV) photodetectors have important applications in many civil and military fields, such as flame detection, missile early warning, and space exploration [1–4]. In recently years, AlGaN-based photoresistors, p-(i)-n homo/heterojunction photodiodes, Schottky photodiodes, metal-semiconductor-metal (MSM) photodetectors, and phototransistors have been reported [5–9]. Among them, Schottky photodetectors have attracted attention due to their simple structure, easy fabrication and integration, short response time and high quantum efficiency [10]. Therefore, Schottky barrier detectors are favored for UV detection. However, AlGaN-based photodetectors still face many great challenges. The responsivity of AlGaN-based Schottky barrier detectors needs to be further improved. In addition, to realize weak signal detection, it is urgent to develop AlGaN-based photodetectors with high sensitivity and high quantum efficiency. Many important factors can also have influences on the responsivity of Schottky diode photodetectors, such as Schottky contact metal, doping concentration defect density in GaN epilayers, and interface states and so on [11]. To achieve a breakthrough in AlGaN-based photodetectors, high quality epitaxial growth is the key and also the focus of future research. Therefore, finding the right growth conditions is crucial to improving the performance of AlGaN-based photodetectors.
In this paper, we have investigated the effect of the growth pressure of i-AlGaN layer on 3 AlGaN-based Schottky detectors T1-T3. Firstly, the FWHM of the (002) and (102) of the samples T1-T3 were obtained by XRD, and the dislocation density was estimated, which showed that the obtained dislocation density of the samples T1-T3 was not significantly different. Secondly, the carbon impurity concentrations of the samples T1-T3 were obtained using Secondary Ion Mass Spectrometry (SIMS). With the increase of growth pressure, the concentration of carbon impurities decreases. The positron annihilation test showed that the vacancy defects concentration increases with the increase of growth pressure. The optical responsivity of three samples increases first and then decreases with the increase of growth pressure. By linking the value of responsivity to carbon impurity concentration and vacancy defect concentration, it is found that too high carbon impurity concentration or vacancy concentration will reduce the responsivity of the photodetectors. Therefore, by appropriately increasing the growth pressure of the i-AlGaN layer, the vacancy defect concentration and carbon impurity concentration of the i-AlGaN layer in the epitaxial material can be kept at a relatively low level, which will be very beneficial to improve the responsivity of the AlGaN-based photodetectors.
2. Experimental
Three metal-AlGaN Schottky barrier detectors, named as T1, T2 and T3, were grown by metal-organic chemical vapor deposition (MOCVD) on c-plane sapphire substrates. Their schematic diagram is shown in the Fig. 1. Trimethylaluminium (TMAl), Trimethylgallium (TMGa), and ammonia (NH3) were used as precursors in the epitaxial growth process, respectively. Samples were prepared as follows. First, in order to reduce the lattice mismatch, an unintentionally doped GaN buffer layer was grown on the sapphire substrate. Next, a 1.7 µm n-type GaN layer, a 500 nm n-type AlGaN layer and a 250 nm unintentionally doped i-AlGaN layer were grown sequentially. The i-AlGaN layer is a Schottky contact layer in the metal-i-n Schottky diode structure. The n-type GaN layer and the n-AlGaN layer were doped with Si. The three Schottky detectors T1-T3 have AlGaN active layers grown with different pressure of 100, 150, 200 mbar, respectively. The growth pressure of other layers is the same. The growth temperature of i-AlGaN layer of T1, T2 and T3 is 1070°C. Except for the different growth pressure of i-AlGaN layer, other growth conditions were the same in the three samples. After the epitaxial growth, the detectors were fabricated through processing such as photolithography, etching, and coating to make Schottky photodetector devices. To ensure the Schottky contact and reduce the loss of light on the metal, most of the surface was covered by a thin metal layer of 10/10 nm Ni/Au. Above that, a metal layer of Ti/Al/Ti/Au which had a thickness of 15/150/50/150 nm was used to make electrical contact.
Fig. 1. The cross-sectional scheme of the fabricated AlGaN Schottky photodetector and its working diagram.
The in-plane ω scan high resolution X-ray diffraction (HRXRD) measurements were performed by using Rigaku SmartLab X-Ray Diffractometer to determine the structural parameters of samples. The full width at half maximum (FWHM) at (002) and (102) reflection obtained by $\omega$-scan rocking curves can help to determine the edge and screw dislocation. In addition, the impurity concentration distribution in the samples was measured by secondary ion mass spectroscopy (SIMS), and positron annihilation was used to measure the vacancy defects of samples. Moreover, photoluminescence (PL) results of the samples were measured with a 325 nm laser as the excitation source. Furthermore, in the spectral response test system, a xenon lamp is used as the light source, followed by a monochromator. A calibrated Si detector is used to determine the accurate value of responsivity.
3. Result and discussion
Table 1 shows that the growth conditions and characterization parameters of the AlGaN layer in T1-T3 samples. Figure 2 shows that the spectral responsivity of three AlGaN Schottky junction detectors at zero bias voltage. It is noticed that the peak responsivity is 0.1445 A/W for T1, 0.1581 A/W for T2, and 0.1296 A/W for T3, which means that the peak responsivity increases first and then decreases with the increase of growth pressure of i-AlGaN layer.
Table 1. Growth conditions and characterization parameters of the AlGaN layer in T1-T3 samples
The growth pressure of i-AlGaN layer has an obvious influence on the response of Schottky detector. The essential reasons affecting responsivity need to be further studied and analyzed. A study shows that the dislocation cores are normally negatively charged in GaN [12]. It has been proved that the negatively charged acceptors introduced by the dislocation lines (mainly the edge dislocations) act as non-radiative recombination centers in GaN [13,14]. These non-radiative centers will increase the recombination probability of the photogenerated carriers and then affect the responsivity of detectors. The XRD was used to evaluate crystal quality. The FWHM of the XRD rocking curve and the calculated TD density of samples T1, T2, and T3 are listed in Table 1. The screw-component threading dislocation (STD) density (DS) and the edge-component threading dislocation (ETD) density (DE) are estimated by the formulas.
Where β and θ are the XRD FWHM of the (002) and (102) planes, $\varphi $ is the inclination angle of the (102) plane toward the (002) plane, and the b1 and b2 are the modules of the Burgs vector of the screw dislocation and the edge dislocation respectively. The dislocation densities of T1, T2 and T3 are listed in Table 1, and the dislocation densities of the three samples are relatively small and of the same order of magnitude. However, the spectral responsivity of samples T1-T3 has quite a distinct difference. This suggests that the different responsivity of these three samples is not strongly dependent on the dislocation density.In addition, some previous reports have discussed that the residual carbon concentrations in GaN is closely related to the growth pressure [15]. Actually, under different MOCVD growth conditions, i.e. the changes of growth pressure, temperature and NH3/TMGa flow rate ratio, can all lead to a change of the residual carbon doping concentration. In this study, the growth pressure of i-AlGaN layer was changed during MOCVD growth and it may induce a remarkable change of residual carbon concentration. Based on this consideration, we have measured the carbon concentration of the samples by secondary ion mass spectroscopy (SIMS), as shown in Fig. 3(a). It is clearly demonstrated that the carbon concentration of sample T1 at the i-AlGaN layer is as high as $5.42 \times {10^{16}}$. It is relatively low, being $4.26 \times {10^{16}}$ and $2.32 \times {10^{16}}$ for sample T2 and sample T3, respectively. This SIMS result means that the carbon impurity concentration decreases with increasing growth pressure. It has known that carbon impurities exist in different forms in GaN material, such as substitutional defects (CN,CGa), interstitial defects (Ci), or C-related complexes [16,17]. Whether they are donors or acceptors in GaN depends mainly on the position of the Fermi energy and the growth condition of the material. In general, the formation energy of C-related deep trap levels is much lower than the shallow energy levels.
Fig. 3. (a) Depth profiles of carbon concentration for samples T1, T2 and T3 measured by SIMS. (b)Normalized PL spectra at room temperature for samples T1, T2 and T3.
Furthermore, in order to study deep-level defects, the normalized room temperature photoluminescence (PL) data were obtained at 8-10 different locations on the samples T1, T2 and T3, a set of data is shown in Fig. 3(b) due to their good consistency. of three samples were measured and described, as shown in Fig. 3(b). Interestingly, a wide yellow luminescence (YL) band (centered on 2.25 eV) is present in the normalized photoluminescence spectral curves of all three samples, but its intensity is very different. With the increase of pressure, the intensity of YL decreases obviously. The presence of YL indicates that there are indeed some deep level defects inside the i-AlGaN layer, which may include carbon impurities and other point defects. Therefore, we speculate that there are other factors that also affect the responsivity of the detectors, such as vacancy defects in GaN.
Positron annihilation is a commonly used test method to detect defects in semiconductors [18]. The analysis method of Doppler spectrum generally adopts linear parameter method, and the commonly used linear parameters are mainly S parameter and W parameter. Parameter S was defined as the ratio of the count in the central region of the peak to the total number of peaks in the Doppler broadening spectrum of 511 keV, reflecting the annihilation of positron - electron pairs in the low momentum region; parameter W was defined as the ratio of the count on the two sides of the peak to the total number of peaks, reflecting the annihilation of positron - electron pairs in the high momentum region. When the S parameter increases, the W parameter decreases, indicating that the vacancy increases [19]. Figure 4 (a) and (b) show the S parameter and W parameter as a function of positron incident energy for the three samples T0, T1 and T2. It is observed that as the growth pressure increases, the S-parameter value increases and the W-parameter value decreases, suggesting that the vacancy defects in AlGaN increased. When positrons are annihilated at metal vacancies, the value of parameter S will increase and that of parameter W will decrease, since a larger fraction of annihilation happens to low momentum electrons. We used positron annihilation test to qualitatively analyze the vacancy defect concentration in three samples. Therefore, the results in Fig. 4 (a) and (b) suggest that the increase in the ascertained vacancy defect concentration with the increase of growth pressure. In combination with the responsivity results of the three samples shown in Fig. 2, it is found that the concentration of vacancy defects does not fully correspond to the value of responsivity. Based on the SIMS results, we speculate that both vacancy defects and carbon impurities jointly affect the responsivity of AlGaN Schottky detector. In order to further obtain the information about the point defects existing in the three samples, the relationship between the S and W parameters of these samples was investigated. As shown in Fig. 4(c), it is observed that the parameter S varies linearly with the parameter W in 3 samples. The slopes of the 3 linear curves are almost equal to each other. This result indicates that only one type of point defect exists in these samples. Since this defect is a negative center and the material of the i-layer is unintentionally doped AlGaN, the defect is most likely to be a metal vacancy, that is, a Ga vacancy or a Al vacancy. According to research by Warnick and Puzyrev, the formation energy of Ga vacancy is slightly lower than that of Al vacancy, which means that Ga vacancy is easier to form in AlGaN material. Therefore, we think this defect is more likely Ga vacancy [20,21].
Fig. 4. (a)The dependences of the high momentum parameter W on positron incident energy in the three samples. (b) The dependences of the low momentum parameter S on positron incident energy in the three samples T0, T1 and T2. (c)The relationship curves between the low-momentum parameter S and the high-momentum parameter W of the three samples.
Positron annihilation analysis of the samples demonstrates that a strong correlation exists between growth pressure and recombination processes within the AlGaN film. Wickenden et.al [21] had proposed that a white deposit was observed in the reactor for the growth under 130 torr pressure, which increases as a function of TMAl molar flow. In their research, growth at 65 torr pressure provides a reasonable Al content in AlGaN corresponding to the expected gas phase aluminum content, with no evidence of any additional deposit, which suggests that the aluminum is more efficiently incorporated into the growing film at a pressure of 65 torr compared to 130 torr. As the pressure increases, the growth rate will decrease and the pre-reaction will be enhanced, so more vacancy defects may be formed during the growth of AlGaN. Ga vacancy as a negative center will trap photo-generated holes. The concentration of photogenerated carriers in the depletion region has a great influence on the responsivity. Ga vacancies can trap photogenerated carriers or increase their recombination probability, resulting in serious decrease in responsivity. In this research, it seems that the responsivity of AlGaN Schottky photodetector is not entirely affected by the vacancy factor alone.
Figure 5 (a) shows the relationship between peak responsivity, vacancy defect concentration, carbon impurity concentration and growth pressure of the three samples T1, T2 and T3. It can be seen that the peak responsivity first increases and then decreases with the increase of growth pressure, the vacancy defect concentration increases with the increase of growth pressure, and the carbon impurity concentration decreases with the increase of growth pressure. The responsivity of the detector is the best when the pressure is 150 mbar where the carbon impurities and vacancy defects are in the middle level. It is speculated that both carbon impurity concentration and vacancy defects will affect the responsivity of the detectors, so at the pressure of 100 mbar, although vacancy concentration is very low, carbon impurity concentration is too high. While at 200 mbar, the carbon impurity concentration is low, but the vacancy defects concentration is too high. In conclusion, whether the vacancy defects concentration is too high or the carbon impurity concentration is too high, the responsivity of Schottky detector will be reduced. When both vacancy concentration and carbon impurity concentration are at moderate values, the responsivity of Schottky detector reaches the maximum. As shown in Fig. 5 (b), both vacancy defects and carbon impurity defects can act as deep-level defects. On the one hand, the metal vacancy is negatively charged and can trap holes, on the other hand, the carbon impurity will act as non-radiative recombination center to increase the probability of non-radiative recombination of photogenerated carriers. Actually, the vacancy and carbon impurities are considered to be the deep center in the band gap. When incident light irradiates AlGaN sample, the photo-generated holes in i-AlGaN are easily captured or more likely to non-radiatively recombine at deep-level defects, so that the number of electron hole pairs entering the circuit is reduced. The electrical signal generated by the light injection will be less. Both of these defects will reduce the responsivity of the detector.
Fig. 5. (a)Peak responsivity (black) and S parameter (blue) and carbon concentration (red) versus growth pressure. (The connected lines are used only for helping eyes). (b) Schematic diagram of deep level mechanism.
4. Conclusions
In summary, we find that the relationship between the responsivity of AlGaN Schottky detector and growth pressure is not a monotonous one, but increases first and then decreases with the increase of growth pressure. SIMS results show that the carbon impurity concentration decreased with the increase of growth pressure. The result of positron annihilation indicates that the vacancy defect concentration increases with the increase of growth pressure. This reveals that the responsivity is not a single function of carbon impurity concentration or vacancy. Both vacancy defects and carbon impurities play a negative role in AlGaN Schottky detector. Both high concentration vacancy defects and carbon impurities will reduce the responsivity of AlGaN Schottky detector. The presence of vacancy defects and carbon impurities as non-radiative recombination centers will increase the probability of photogenerated carrier recombination. Therefore, selecting the appropriate growth pressure can effectively control the vacancy and carbon impurity concentration, which is very meaningful to improve the response characteristics of AlGaN Schottky detector.
Funding
Shanxi-Zheda Institute of Advanced Materials and Chemical Engineering (2022SX-TD016); Strategic Priority Research Program of Chinese Academy of Sciences (XDB43030101); Beijing Nova Program (202093); Beijing Municipal Science and Technology Project (Z161100002116037, Z211100007921022); National Natural Science Foundation of China (61904172, 61974162, 62034008, 62074140, 62074142, 62250038); Youth Innovation Promotion Association of the Chinese Academy of Sciences (2019115).
Disclosures
The authors declare no conflicts of interest.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
1. M. A. Khan, M. Shatalov, H. P. Maruska, H. M. Wang, and E. Kuokstis, “III–Nitride UV Devices,” Jpn. J. Appl. Phys. 44(10R), 7191 (2005). [CrossRef]
2. D. Li, K. Jiang, X. Sun, and C. Guo, “AlGaN photonics: recent advances in materials and ultraviolet devices,” Adv. Opt. Photonics 10(1), 43 (2018). [CrossRef]
3. J. Li, M. Zhao, and X. F. Wang, “High performance Schottky UV photodetectors based on epitaxial AlGaN thin film,” Phys. B 405(3), 996–998 (2010). [CrossRef]
4. E. Munoz, E Monroy, J. L. Pau, F. Calle, F. Omnes, and P. Gibart, “III nitrides and UV detection,” J. Phys.: Condens. Matter 13(32), 7115–7137 (2001). [CrossRef]
5. R. Dahal, T. M. Al Tahtamouni, Z. Y. Fan, J Fan, Y. Lin, and H. X. Jiang, “Hybrid AlN-SiC deep ultraviolet Schottky barrier photodetectors,” Appl. Phys. Lett. 90(26), 263505 (2007). [CrossRef]
6. N. Biyikli, I. Kimukin, O. Aytur, and E. Ozbay, “Solar-blind AlGaN-Based p-i-n photodiodes with low dark current and high detectivity,” IEEE Photonics Technol. Lett. 16(7), 1718–1720 (2004). [CrossRef]
7. J. L. Pau, E. Monroy, M. A. Sanchez-Garcia, E calleja, and E. Munoz, “AlGaN ultraviolet photodetectors grown by molecular beam epitaxy on Si(111) substrates,” Materials Science and Engineering 93(1-3), 159–162 (2002). [CrossRef]
8. S. Rathkanthiwar, A. Kalra, S. V. Solanke, N. Mohta, R. Muralidharan, S. Raghavan, and D. N. Nath, “Gain mechanism and carrier transport in high responsivity AlGaN-based solar blind metal semiconductor metal photodetectors,” J. Appl. Phys. 121(16), 164502 (2017). [CrossRef]
9. C.-K. Wang, Y.-Z. Chiou, S.-J. Chang, W.-C. Lai, S.-P. Chang, C.-H. Yen, and C.-C. Hung, “GaN MSM UV Photodetector With Sputtered AlN Nucleation Layer,” IEEE Sens. J. 15(9), 4743–4748 (2015). [CrossRef]
10. F. Bouzid, L. Dehimi, and F. Pezzimenti, “Performance Analysis of a Pt/n-GaN Schottky Barrier UV Detector,” J. Electron. Mater. 46(11), 6563–6570 (2017). [CrossRef]
11. M. A. Kadaoui, W. B. Bouiadjra, A. Saidane, S. Belahsene, and A. Ramdane, “Electrical parameters of Au/n-GaN and Pt/n-GaN Schottky diodes,” Superlattices Microstruct. 82, 269–286 (2015). [CrossRef]
12. A. Krtschil, A. Dadgar, A. Krost, and A. K. A. D. A. Krost, “Decoration effects as origin of dislocation-related charges in gallium nitride layers investigated by scanning surface potential microscopy,” Appl. Phys. Lett. 82(14), 2263–2265 (2003). [CrossRef]
13. N. G. Weimann, L. F. Eastman, D. Doppalapudi, H. M. Ng, and T. D. Moustakas, “Scattering of electrons at threading dislocations in GaN,” J. Appl. Phys. 83(7), 3656–3659 (1998). [CrossRef]
14. J. Speck and S. Rosner, “The role of threading dislocations in the physical properties of GaN and its alloys,” Phys. B 273-274, 24–32 (1999). [CrossRef]
15. X. G. He, D. G. Zhao, D. S. Jiang, Z. S. Liu, P. Chen, L. C. Le, J. Yang, X. J. Li, S. M. Zhang, J. J. Zhu, H. Wang, and H. Yang, “Control of residual carbon concentration in GaN high electron mobility transistor and realization of high-resistance GaN grown by metal-organic chemical vapor deposition,” Thin Solid Films 564, 135–139 (2014). [CrossRef]
16. A. F. Wrigh, “Substitutional and interstitial carbon in wurtzite GaN,” J. Appl. Phys. 92(5), 2575–2585 (2002). [CrossRef]
17. U. Honda, Y. Yamada, Y. Tokuda, and K. Shiojima, “Deep levels in n-GaN doped with carbon studied by deep level and minority carrier transient spectroscopies,” Jpn. J. Appl. Phys. 51(4S), 04DF04 (2012). [CrossRef]
18. A. Dupasquier and A. Mills Jr, Positron Spectroscopy of Solids, (IOS press, 1995).
19. K. Saarinen, T. Laine, S. Kuisma, J. Nissilä, P. Hautojärvi, L. Dobrzynski, J. Baranowski, K. Pakula, R. Stepniewski, and M. Wojdak, “Observation of native Ga vacancies in GaN by positron annihilation,” MRS Online Proceedings Library (OPL) 482, 3030 (1997).
20. K. H. Warnick, Y. Puzyrev, T. Roy, D. M. Fleetwood, R. D. Schrimpf, and S. T. Pantelides, “Room-temperature diffusive phenomena in semiconductors: The case of AlGaN,” Phys. Rev. B 84(21), 214109 (2011). [CrossRef]
21. Y. S. Puzyrev, T. Roy, M. Beck, B. R. Tuttle, R. D. Schrimpf, D. M. Fleetwood, and S. T. Pantelides, “Dehydrogenation of defects and hot-electron degradation in GaN high-electron-mobility transistors,” J. Appl. Phys. 109(3), 9 (2011). [CrossRef]
