The n-ZnO/p-Si heterojunction with an ultrathin Al2O3 buffer layer was prepared by atomic layer deposition. X-ray diffraction revealed that the crystalline quality of (100)-oriented ZnO films was improved with an Al2O3 buffer layer. The n-ZnO/p-Si heterojunction with 5 nm inserted Al2O3 layer shows the best electrical characteristics, with a dark current of 0.5 μA at a reverse bias of −2 V and increasing the photo-to-dark current ratio effectively by 8 times. These results demonstrated that Al2O3 buffer layer with optimized thickness exhibits significant advantages in enhancing the crystal quality of ZnO film and improving the photoelectrical properties of n-ZnO/p-Si photodetectors.
© 2014 Optical Society of America
Zinc oxide (ZnO), as a promising semiconductor material, has attracted increasing attention for potential electronic and optoelectronic applications [1-3]. Over the past decade, various nanoscale ZnO structures have paved the way for the fabrication of ultraviolet (UV) optoelectronic devices due to its wide direct bandgap of 3.37 eV and high exciton binding energy of 60 meV [4,5]. However, as-grown ZnO usually exhibits n-type conductivity due to its intrinsic donor type defects such as oxygen vacancies (Vo) and zinc interstitials (Zni). The lack of obtaining high quality and stable p-ZnO makes it difficult to develop p-n ZnO homojunction devices. As an alternative approach, a lot of work have been conducted on the fabrication of ZnO-based heterojunctions with various p-type semiconductors, such as p-GaAs , p-SiC [7,8], and p-Si . Among all of these p-type semiconductors, the commercial silicon is highly focused for its low cost, large-area availability of Si wafers, and the compatibility with mature semiconductor integrated circuit processing technology.
Many studies on n-ZnO/p-Si heterojunction photodetectors have been published in recent years [9–11]. Mridha et al. reported the fabrication of a n-ZnO/p-Si thin-film heterojunction using a sol-gel technique . Jeong et al. fabricated a n-ZnO/p-Si UV/visible photodiode that shows induced photocurrent proportional to reverse bias for UV illumination and saturated at moderate bias for visible illumination . However, high performance n-ZnO/p-Si heterojunction photodetectors for practical applications are difficult to achieve due to the fact that direct growth of crystalline ZnO thin film on Si substrate is not easy. The crystallographic structure and lattice parameters of Si (diamond, a = 0.543 nm)  and ZnO (wurtzite, a = 0.324 nm and c = 0.520 nm)  exhibit that these two materials are not particularly compatible. The device performance usually degenerates due to the lattice mismatch at the n-ZnO/p-Si interface. As a result, inserting an ultrathin buffer layer between ZnO and Si substrate has been recently proposed to passivate defects and improve the interface quality. Intermediate layers including GaN , AlN , MgO [16,17], LaAlO3 , and Gd2O3(Ga2O3)  have been selected as buffer layers for the n-ZnO/p-Si heterojunction. For example, Zhang et al. fabricated a visible-blind UV photodetector based on a double heterojunction of n-ZnO/insulator-MgO/p-Si grown by molecular beam epitaxy . The key role of the middle MgO layer in visible-blind UV detecting was demonstrated. Besides, a thin LaAlO3 layer exhibited electrical insulating characteristics and served as a blocking layer for photo-excited electrons from p-Si to n-ZnO, leading to a high rectification ratio and visible-blind UV detectivity of n-ZnO/LaAlO3/p-Si photodiodes .
As a widely used insulator material, Al2O3 is also being a promising buffer layer to improve not only the crystalline quality of ZnO films but also the optical and electrical properties of n-ZnO/p-Si heterojunction [20–22]. Since the band offsets of n-ZnO/Al2O3/p-Si could create a much larger potential barrier for electrons in p-Si than for holes in n-ZnO, it is difficult for the photogenerated electrons under visible light to cross over the interface between p-Si and Al2O3. Under UV illumination, the holes generated in ZnO side can transmit through the Al2O3 layer easily. Therefore, the Al2O3 buffer layer could also serve as a block layer that leads to an efficient suppression of photo-response to visible light, which makes it a promising material for visible-blind UV detection. It is expected that the microstructural, optical, and electrical properties of the n-ZnO/p-Si heterojunction would be influenced by the thickness of the inserted thin Al2O3 layer. Up to now, however, there is no study on such heterojunction properties by changing the thickness of the Al2O3 layer. In this work, we present a UV photodetector based on the n-ZnO/Al2O3/p-Si heterojunction grown by atomic layer deposition (ALD). The in situ growth of ZnO thin film on the Al2O3 buffer layer during the ALD process benefits to the excellent control of interface quality. A UV photodetector with better structural and electrical characteristics is achieved by optimizing the thickness of Al2O3 buffer layer.
To fabricate n-ZnO/Al2O3/p-Si photodetectors, Al2O3 and ZnO layers were sequentially deposited on RCA standard cleaned p-Si (100) (ρ = 1-10 Ωcm) at 200°C in a BENEQ TFS-200 ALD reactor (Beneq Oy, Vantaa, Finland). Trimethlaluminum (TMA), diethylzinc (DEZn), and deionized water (H2O) were employed as the sources of Al, Zn, and O, respectively. To fully demonstrate the role of inserted Al2O3 layer in affecting the electrical properties of n-ZnO/p-Si photodetectors, 20, 50, and 100 ALD cycles of TMA/H2O were carried out to form nominal 2, 5, and 10 nm Al2O3 layers respectively. Next, the nominal 100 nm thick n-ZnO layer was deposited using 500 ALD cycles of DEZn/H2O for all the samples. The thicknesses of the ZnO/Al2O3 laminate were measured by spectroscopic ellipsometry (SE, Sopra GES5E, SOPRA, France) where the incident angle was fixed at 75°. High resolution transmission electron microscopy (HRTEM) was used to examine the microstructure and interface of the ZnO/Al2O3/Si heterojunction. The crystal structures of the ZnO films were determined by an x-ray diffractometer (XRD, D8 ADVANCE, Bruker AXS, Inc.) with Cu Kα radiation (40 kV, 40 mA, λ = 1.54056 Å). To characterize the electrical properties, Au (50 nm thick) electrodes in circular form with a diameter of 100 μm were deposited by a thermal evaporator through a shadow mask on the ZnO layer as well as on the backside of p-Si substrate. The dark current-voltage I-V measurements were done by keeping the sample in the dark for several hours and measuring the current between two contacts using an Agilent B1500A semiconductor device analyzer at room temperature in ambient condition. And a 5.0 W Hg lamp was employed to provide the 365 nm UV light for the photocurrent measurement.
3. Results and discussions
A simple four-phase model consisting of p-Si substrate/Al2O3 layer/thin ZnO film/surface rough layer has been used to fit the obtained ellipsometric spectra. The surface rough layer is modeled by the Bruggeman effective medium approximation of 50% ZnO and 50% voids. The experimental and fitted ellipsometric spectra of the selected sample with 50 ALD cycles of Al2O3 were shown in Fig. 1(a). It can be found that the experimental and fitting curves match very well, with the accuracy of the regression (R2) greater than 0.998. The thicknesses of Al2O3 and ZnO layer for this sample is then determined to be 4.9 and 96.6 nm, suggesting the growth rates of Al2O3 and ZnO thin film deposited by ALD are 0.10 and 0.19 nm/cycle, respectively. Table 1 lists the thicknesses of ZnO and Al2O3 layers for all the samples grown on Si substrate. As can be seen, film thicknesses measured by spectroscopic ellipsometry are very close to the expected values, suggesting the excellent thickness control of ALD films. Furthermore, the cross-sectional HRTEM image for the n-ZnO/p-Si sample with 50 ALD cycles of Al2O3 is shown in Fig. 1(b). The typical morphology of as-grown ZnO and Al2O3 layer is presented, with evident lattice observed in the ZnO film, indicating its polycrystalline nature. The ultrathin Al2O3 layer between the ZnO film and the Si substrate has a thickness of approximately 4.85 nm, which is in good agreement with the value obtained by spectroscopic ellipsometry. In addition, the amorphous Al2O3 layer with a sharp interface is suitable to be used as a carrier blocking layer for the heterojunction.
The XRD patterns of ZnO thin films grown on p-Si with Al2O3 buffer layers of different thicknesses are presented in Fig. 2. For the ZnO thin film grown directly on p-Si substrate, four diffraction peaks are observed at 2θ = 31.8°, 34.4°, 36.6°, and 57.1°, which correspond with the spacing in (100), (002), (101), and (110) directions of the ZnO layer respectively. No reflection peaks related to Al2O3 are detected in the XRD patterns detected from the ZnO/Al2O3/Si samples. The obtained diffraction patterns of as-grown samples are well-matched with the wurtzite hexagonal phase of pure bulk ZnO (JCPDS36-1451), which confirms that the synthesized products are well-crystalline. Moreover, the polycrystalline nature of ALD-ZnO films is consistent with the HRTEM image. It is also found that the intensity of (100) peak is the strongest, suggesting the α-axis growth is dominant in the growth of ZnO thin film. This kind of growth mode is one of typical characteristics of the ZnO thin film grown by ALD in the temperature around 200 °C [23,24]. Moreover, the (100) and (002) peaks tend to get stronger with the increasing thickness of inserted Al2O3 layer. It is noticed that the Al2O3 layer has no effect on the intensity of (110) peak. However, the intensity of the (101) peak decreases drastically with the increase of Al2O3 thickness. It is known that the amorphous Al2O3 layer has more hydroxyl than native SiO2, which benefits the growth of compact ZnO films. These findings demonstrated that the quality of ALD-ZnO films grown on Si (100) has been enhanced with the insert of Al2O3 layer.
Figure 3(a) shows the reverse dark I-V curves of n-ZnO/Al2O3/p-Si heterojunctions with various Al2O3 thicknesses on a logarithmic scale. For n-ZnO/p-Si heterojunction without Al2O3 buffer layer, the reverse dark current is over 1 mA at −2 V. According to the linear I-V characteristic of the n-ZnO/p-Si heterojunction shown in the inset of Fig. 3(a), the rectification ratio is calculated to be 15 at ± 5 V. This kind of heterojunction with a poor rectifying property and an unacceptable leakage current is not suitable for practical application. However, an evident decrease of reverse dark current is found for the heterojunctions with Al2O3 buffer layer. It is observed that the value of dark current decreases from 1.3 mA for the sample with 2 nm thick Al2O3 to 16 μA for the sample with 20 nm Al2O3 at −2 V. This variation trend indicates the Al2O3 buffer layer could be a good insulator and effectively prevent the electron injection from p-Si substrate to n-ZnO layer at a reverse bias. Nevertheless, the forward bias current will also decrease correspondingly when the insulator layer gets thicker. Evidently, the decreasing forward bias current would lead to a lower rectification ratio. Comparing the five samples, the largest rectification ratio of 429 can be obtained at ± 5 V for the heterojunction with 5 nm thick Al2O3 buffer layer. Therefore, the n-ZnO/p-Si heterojunction with 5 nm Al2O3 buffer layer exhibits better electrical properties than other samples, with both relatively low leakage current and better rectifying characteristic.
Figure 3(b) presents the log-log I-V plot of the heterojunction with 5 nm thick Al2O3 buffer layer under forward bias at room temperature. The curve can be divided into three distinct regions depending on the applied voltage. At a very low forward voltage for V<0.4 V (region I), a linear dependence of the current on the voltage (I~V) is observed, suggesting a transport mechanism obeying the Ohm’s law. At a moderately higher junction voltage, i.e., 0.4V<V<1.4 V (region II), the current increased exponentially as a function of I~exp(αV). It results from the recombination-tunneling mechanism, as is usually observed in wide band gap p-n heterojunctions [25,26]. The recombination tunneling path is provided by a highly dislocated and thin n-type region generated at the interface between ZnO and Si, which is generated to relieve the strain caused by the lattice mismatch . The constant α can be given as28]. For a thin film n-ZnO/p-Si heterojunction, a higher value of α indicates higher carrier injection. There are donor-like defects such as oxygen vacancies involved in the Al2O3 layer grown by ALD, which would lead to higher electron density and provide the recombination-tunneling path. Thus n-ZnO/p-Si heterojunction with an inserted Al2O3 buffer layer exhibits higher carrier injection. In region III (1.4 V<V<6 V), the I-V characteristic is deviated from the ideal thermionic emission and behaves as I~(V−V0)2 relation, which is generally attributed to space-charge-limited current (SCLC) conduction .
The performance of the UV photodetector is usually evaluated by photo-to-dark current ratio (PDCR) , which is a sensitive factor and can be measured using the formula as follows,Fig. 4. The PDCRs of n-ZnO/5 nm thick Al2O3/p-Si and n-ZnO/p-Si photodetectors under 2.0 V reverse bias are then calculated to be 0.48 and 0.06, respectively. The inserted Al2O3 layer increases the PDCR of the n-ZnO/p-Si photodetector effectively by 8 times, indicating the significant role of Al2O3 buffer layer in improving the photoelectrical properties of ZnO/p-Si photodetectors. The thin Al2O3 buffer layer effectively prevents the electron injection from p-Si side to n-ZnO side at a reverse bias in the dark due to the high potential barrier of Al2O3 layer for electrons. However, UV light with a wavelength centered at around 378 nm could be fully absorbed in the depletion region of n-ZnO and generate electron-hole pairs in the ZnO side. The photo-generated electrons drift toward Au electrode while holes do toward the p-Si side under the electric field. The holes can transmit through the Al2O3 layer in large quantities due to the relatively low barrier height, which leads to a large photocurrent. The influence of Al2O3 barrier layer on UV photocurrent is relatively less than dark current. On the other hand, the enhancement of ZnO quality due to Al2O3 buffer layer also brings better photoelectrical performance. As a result, a higher PDCR is obtained in n-ZnO/p-Si photodetector with an optimized Al2O3 buffer layer of 5 nm thick.
In summary, a UV photodetector based on n-ZnO/Al2O3/p-Si heterojunction has been fabricated and characterized by structural and photoelectrical measurements. The Al2O3 layer works as both a buffer layer for the ALD growth of ZnO and a barrier layer for the UV photodetector. This work demonstrates that the 5 nm Al2O3 layer is the optimized thickness to achieve the highest photo-to-dark current ratio of n-ZnO/Al2O3/p-Si photodetector.
This work is supported by the National Natural Science Foundation of China (no. 51102048, 61376008), the SRFDP (no. 20110071120017), Innovation Program of Shanghai Municipal Education Commission (14ZZ004), and Hui Chun Chin and Tsung Dao Lee Chinese Undergraduate Research Endowment (CURE). We are grateful to Prof. Xiao-Sheng Fang from Department of Materials Science, Fudan University for the experimental assistance.
References and links
1. U. Özgur, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Dogan, V. Avrutin, S. J. Cho, and H. Morkoc, “A comprehensive review of ZnO materials and devices,” J. Appl. Phys. 98(4), 041301 (2005). [CrossRef]
2. D. C. Look, “Recent advances in ZnO materials and devices,” Mater. Sci. Eng. B 80(1–3), 383–387 (2001). [CrossRef]
3. J. C. Fan, K. M. Sreekanth, Z. Xie, S. L. Chang, and K. V. Rao, “p-Type ZnO materials: Theory, growth, properties and devices,” Prog. Mater. Sci. 58(6), 874–985 (2013). [CrossRef]
5. H. Zhou, G. Fang, L. Yuan, C. Wang, X. Yang, H. Huang, C. Zhou, and X. Zhao, “Deep ultraviolet and near infrared photodiode based on n-ZnO/p-silicon nanowire heterojunction fabricated at low temperature,” Appl. Phys. Lett. 94(1), 013503 (2009). [CrossRef]
6. P. Köç, S. Tekmen, A. Baltakesmez, S. Tüzemen, K. Meral, and Y. Onganer, “Stimulated electroluminescence emission from n-ZnO/p-GaAs:Zn heterojunctions fabricated by electro-deposition,” AIP Adv. 3(12), 122107 (2013). [CrossRef]
7. Ya. I. Alivov, Ü. Özgür, S. Doğan, D. Johnstone, V. Avrutin, N. Onojima, C. Liu, J. Xie, Q. Fan, and H. Morkoç, “Photoresponse of n-ZnO/p-SiC heterojunction diodes grown by plasma assisted molecular-beam epitaxy,” Appl. Phys. Lett. 86(24), 241108 (2005). [CrossRef]
8. Z. F. Shi, X. C. Xia, W. Yin, S. K. Zhang, H. Wang, J. Wang, L. Zhao, X. Dong, B. L. Zhang, and G. T. Du, “Dominant ultraviolet electroluminescence from p-ZnO:As/n-SiC(6H) heterojunction light-emitting diodes,” Appl. Phys. Lett. 100(10), 101112 (2012). [CrossRef]
9. A. E. Rakhshani, “Optoelectronic properties of p-n and p-i-n heterojunction devices prepared by electrodeposition of n-ZnO on p-Si,” J. Appl. Phys. 108(9), 094502 (2010). [CrossRef]
10. S. Mridha and D. Basak, “Ultraviolet and visible photoresponse properties of n‐ZnO/p‐Si heterojunction,” J. Appl. Phys. 101(8), 083102 (2007). [CrossRef]
11. S. Jeong, J. H. Kim, and S. Im, “Ultraviolet-enhanced photodiode employing n-ZnO/p-Si structure,” Appl. Phys. Lett. 83(14), 2946–2948 (2003). [CrossRef]
12. S. M. Sze, Physics of Semiconductor Devices, 2nd ed. (Wiley, 1981).
13. D. R. Lide, ed., CRC Handbook of Chemistry and Physics, 82nd ed. (CRC, 2001).
14. A. Nahhas, H. K. Kim, and J. Blachere, “Epitaxial growth of ZnO films on Si substrates using an epitaxial GaN buffer,” Appl. Phys. Lett. 78(11), 1511–1513 (2001). [CrossRef]
15. L. Wang, Y. Pu, Y. F. Chen, C. L. Mo, W. Q. Fang, C. B. Xiong, J. N. Dai, and F. Y. Jiang, “MOCVD growth of ZnO films on Si(111) substrate using a thin AlN buffer layer,” J. Cryst. Growth 284(3–4), 459–463 (2005). [CrossRef]
16. T. C. Zhang, Y. Guo, Z. X. Mei, C. Z. Gu, and X. L. Du, “Visible-blind ultraviolet photodetector based on double heterojunction of n-ZnO/insulator-MgO/p-Si,” Appl. Phys. Lett. 94(11), 113508 (2009). [CrossRef]
17. D. C. Kim, B. O. Jung, J. H. Lee, H. K. Cho, J. Y. Lee, and J. H. Lee, “Dramatically enhanced ultraviolet photosensing mechanism in a n-ZnO nanowires/i-MgO/n-Si structure with highly dense nanowires and ultrathin MgO layers,” Nanotechnology 22(26), 265506 (2011). [CrossRef] [PubMed]
18. D. S. Tasi, C. F. Kang, H. H. Wang, C. A. Lin, J. J. Ke, Y. H. Chu, and J. H. He, “n-ZnO/LaAlO3/p-Si heterojunction for visible-blind UV detection,” Opt. Lett. 37(6), 1112–1114 (2012). [CrossRef] [PubMed]
19. B. H. Lin, W. R. Liu, S. Yang, C. C. Kuo, C.-H. Hsu, W. F. Hsieh, W. C. Lee, Y. J. Lee, M. Hong, and J. Kwo, “The growth of an epitaxial ZnO film on Si(111) with a Gd2O3(Ga2O3) buffer layer,” Cryst. Growth Des. 11(7), 2846–2851 (2011). [CrossRef]
20. S. Bang, S. Lee, Y. Ko, J. Park, S. Shin, H. Seo, and H. Jeon, “Photocurrent detection of chemically tuned hierarchical ZnO nanostructures grown on seed layers formed by atomic layer deposition,” Nanoscale Res. Lett. 7(1), 290 (2012). [CrossRef] [PubMed]
21. H. L. Lu, M. Yang, Z. Y. Xie, Y. Geng, Y. Zhang, P. F. Wang, Q. Q. Sun, S. J. Ding, and D. W. Zhang, “Band alignment and interfacial structure of ZnO/Si heterojunction with Al2O3 and HfO2 as interlayers,” Appl. Phys. Lett. 104(16), 161602 (2014). [CrossRef]
22. T. Wang, H. Wu, C. Chen, and C. Liu, “Growth, optical, and electrical properties of nonpolar m-plane ZnO on p-Si substrates with Al2O3 buffer layers,” Appl. Phys. Lett. 100(1), 011901 (2012). [CrossRef]
24. Z. Y. Ye, H. L. Lu, Y. Geng, Y. Z. Gu, Z. Y. Xie, Y. Zhang, Q. Q. Sun, S. J. Ding, and D. W. Zhang, “Structural, electrical, and optical properties of Ti-doped ZnO films fabricated by atomic layer deposition,” Nanoscale Res. Lett. 8(1), 108 (2013). [CrossRef] [PubMed]
25. S. Y. Liu, T. Chen, Y. L. Jiang, G. P. Ru, and X. P. Qu, “The effect of post-annealing on the electrical properties of well-aligned n-ZnO nanorods/p-Si heterojunction,” J. Appl. Phys. 105(11), 114504 (2009). [CrossRef]
26. J. B. Fedison, T. P. Chow, H. Lu, and I. B. Bhat, “Electrical characteristics of magnesium-doped gallium nitride junction diodes,” Appl. Phys. Lett. 72(22), 2841–2843 (1998). [CrossRef]
27. D. C. Look and R. J. Molnar, “Degenerate layer at GaN/sapphire interface: Influence on Hall-effect measurements,” Appl. Phys. Lett. 70(25), 3377–3379 (1997). [CrossRef]
28. J. D. Ye, S. L. Gu, S. M. Zhu, W. Liu, S. M. Liu, R. Zhang, Y. Shi, and Y. D. Zheng, “Electroluminescent and transport mechanisms of n-ZnO/p-Si heterojunctions,” Appl. Phys. Lett. 88(18), 182112 (2006). [CrossRef]
29. M. Dutta and D. Basak, “p-ZnO/n-Si heterojunction: Sol-gel fabrication, photoresponse properties, and transport mechanism,” Appl. Phys. Lett. 92(21), 212112 (2008). [CrossRef]
30. W. C. Lien, D. S. Tsai, S. H. Chiu, D. G. Senesky, R. Maboudian, A. P. Pisano, and J. H. He, “Low-temperature, ion beam-assisted SiC thin films with antireflective ZnO nanorod arrays for high-temperature photodetection,” IEEE Electron Device Lett. 32(11), 1564–1566 (2011). [CrossRef]