Fast and sensitive lateral photovoltaic effects in Fe₃O₄/Si Schottky junction

Abstract

In this paper, we report the fast and sensitive lateral photovoltaic (LPE) effects in a Fe₃O₄/Si junction. Good rectifying I–V characteristics and large LPE are observed in the Fe₃O₄/Si junction. The LPE exhibits a linear dependence on the position of the laser spot, and the position sensitivity can reach 32.5 mV mm⁻¹. The optical response time and relaxation time of LPE were ∼60 ns and 5 μs, respectively. The enhanced LPE properties and the fast relaxation time in the Fe₃O₄/Si junction were caused by the formation of the inversion layer in the interface of Fe₃O₄/Si.

---

1. Introduction

Intensive research on the photovoltaic properties of semiconductors has been pursued for a long time due to their practical and wide application in optoelectronic devices. Both vertical and lateral photovoltaic effects have been observed in Schottky junction based on silicon with low cost, large photo-response and compatibility with the technology. The lateral photovoltaic effect (LPE) originates from the lateral diffuse flow and recombination of the photo-generated electron–hole pairs away from the laser spot.^1,2 Since the LPE was first discovered by Wallmark in floating Ge p–n junction, this effect has been observed in many different systems.^1–6 The LPE output changes linearly with laser spot position and thus the LPE effect can be used in position-sensitive detectors (PSD) which can detect very small displacements. The response time and relaxation time of LPE are important characters for PSD applications. Because the LPE originates from the lateral diffuse transport of the excited carriers to the two electrodes from the laser spot, the photovoltage response time and relaxation time will be greatly influenced by the carrier mobility and resistance of the film. The Si-based devices of PSD had been used widely, and the response time of PSD was enhanced from milliseconds in amorphous silicon (a-Si) to microseconds in crystalline devices (c-Si), which due to the inherently slower carrier mobilities of a-Si compared with that of c-Si.⁷ But the high resistance of silicon results in the slowly relaxation time of photovoltaic, which limited the practical purpose of PSD in the extreme conditions.^7,8

Interestingly, the LPE was observed in some simple metal–semiconductor (MS) structures.^9–11 The LPE in these MS shows large position sensitivities and good linearity, which is necessary for developing and designing high-sensitivity PSD. But the LPE is very sensitive to the metal thickness due to the shorting effect, so the metal thickness must be in nano-scale which will result in the instability and high resistance. Recently, Wang et al. have shown that the LPE of ferromagnetic metal film on Si substrate (FM/Si) can be tuned by the magnetic field, which suggested the LPE of FM/Si junction is strongly coupled to the magnetic alignment of the FM layer.¹² Furthermore, a large LPE of 97 mV mm⁻¹ was observed in oxide–metal–semiconductor (TiO₂/Ti/Si) structure due to the increase of the surface resistivity.^13–15 These results suggest that large LPE can be achieved in Schottky junction with high resistivity films. It is well-known that the oxide-film is stable in air and its' resistance is bigger than that of metal, so the oxide-film is a potential candidate for LPE in the PSD applications. But the LPE results in oxide-film/Si structure is still lacking.

In this paper, we report the fast and sensitive LPE effects in Fe₃O₄/Si junction. Large LPE and fast response/relaxation time are achieved. The inversion layer in the interface of Fe₃O₄/Si plays an important role to the photovoltaic properties. The parallel resistance of inversion layer and Fe₃O₄ film is higher than that of metal films and is lower than that of semiconductor, so a highest position sensitivity of 32.5 mV mm⁻¹ and fast relaxation-time are achieved.

2. Experimental

The Fe₃O₄ thin films were prepared by pulsed laser deposition (PLD) on n-type Si (100) substrates with a 1.5 nm thick native SiO₂ layer on the surface. The target was α-Fe₂O₃. The films were prepared in a vacuum of 10⁻⁷ Torr at substrate temperature of 350 °C with different deposition time.¹⁶ The pulsed excimer laser uses KrF (λ = 248 nm) and produces a laser beam with an intensity of 1–2 J cm⁻² and a repetition rate of 2 Hz. After the deposition, the film and substrate were annealed for 20 min under the same condition.

The current–voltage (I–V) curve was measured with Keithley 2400 source and Keithley 2000 multimeter. The LPE was done using He–Ne laser (2 mW at 632 nm, made by Peking University) and Keithley 2000 multimeter. The time response of the LPE of Fe₃O₄/Si junction was measured using a Ti: sapphire regeneration amplified laser system with a wavelength of 800 nm and 130 fs pulse duration and an oscilloscope of 100 MHz (Owon POS7102T).

3. Results and discussion

Fig. 1 shows the X-ray diffraction pattern of 20 nm Fe₃O₄ film on Si with a native SiO₂ layer. Only diffraction lines of (111) family of Fe₃O₄ are observed, suggesting that a single-phase Fe₃O₄ film is formed and Fe₃O₄/Si junction is achieved.

Fig. 1 The XRD pattern of Fe₃O₄ film on Si.

Fig. 2(a) shows the dependence of resistivity of Fe₃O₄ film on temperature, which is measured on the surface of Fe₃O₄ film, as shown in the inset of Fig. 2(b). At low temperature, the resistance decreases with increasing temperature, which shows a typical behavior of thin Fe₃O₄ film, as the resistance data of Fe₃O₄ film on MgO substrate in Fig. 2(a). When the temperature is increased beyond 230 K, the resistance drops rapidly because the conduction path starts to switch from the Fe₃O₄ film into the inversion layer underneath the native SiO₂ via thermally assisted tunneling.^16,17 The inversion layer can provide a low resistive path for carrier transport along the surface of the Si substrate, so the temperature dependence of resistivity shows remarkable current-controlled channel switching between Fe₃O₄ film and inversion layer in Fe₃O₄/Si structure, as shown in Fig. 2(a). This result suggested that the measured resistivity at room temperature is the parallel resistance of inversion layer and Fe₃O₄ film in the Fe₃O₄/Si structure. Fig. 2(b) shows the I–V curve measured at 300 K in the plane of the Fe₃O₄ film. It is clear that the current increases nonlinearly with increasing voltage, which further confirms the existent of the inversion layer in the Fe₃O₄/Si junction.^16,17

Fig. 2 (a) The variation of the resistivity of Fe₃O₄ with temperature; (b) the I–V curve of Fe₃O₄/Si structure at 300 K. The inset exhibits the schematic setup for measurement.

Fig. 3 shows the I–V curve of Fe₃O₄/Si junction with 20 nm Fe₃O₄ film at 300 K. Two indium electrodes with a 0.5 mm diameter were placed on the surface of Fe₃O₄ film and Si substrate, as shown in the inset of Fig. 3. It is clear that the I–V curve exhibits a good backward diode-like rectifying character, which means the Fe₃O₄/Si structure is a Schottky junction. The junction current increases dramatically with increasing forward bias voltage. But the current remains very low for the reverse bias voltage even when the voltage reaches −4 V. Similar results had been observed by Qu et al. in Fe₃O₄/Si structure.²⁰ The work functions of Fe₃O₄ and Si are 5.3 and 4.4 eV, respectively.^18,19 Once the Fe₃O₄ thin film was deposited on the Si substrate, a large built-in field forms in the Fe₃O₄/Si and the Schottky junction is achieved,^20–23 which is good for the separation of electron–hole pairs in Si created by the radiation of the laser.

Fig. 3 The I–V curve of Fe₃O₄/Si junction at 300 K, the inset shows the schematic circuit of the junction measurement.

The LPE of the Fe₃O₄/Si junctions was investigated in detail. Two indium electrodes were placed on the surface of Fe₃O₄ film, as shown in the inset of Fig. 4. The sample was always kept in the dark during measurements. When the Fe₃O₄ surface was partially illuminated by a He–Ne laser spot, a large LPE was found. The transmittance spectra of Fe₃O₄ films on MgO or Al₂O₃ substrates were shown in the inset of Fig. 4. Obviously, there are slight absorbability of Fe₃O₄ from 400 to 800 nm, which is mainly due to its' thin thickness of Fe₃O₄ film. So a large absorption appears in Si substrates and the electron–hole pairs will be created in Si substrate under the radiation of the laser. Fig. 4 shows the dependence of the induced photovoltage on the laser position for the Fe₃O₄/Si junction. The LPE shows an approximately linear dependence as the laser spot moves in the region between the two electrodes on the Fe₃O₄ surface and decreases when the laser spot moves out of the region, and the largest position sensitivity of 32.5 mV mm⁻¹ in the linear region is achieved in 20 nm Fe₃O₄ film. Considering that the power of the laser illuminating the junction is only 2 mW and the contacts' distance is about 5 mm, the position sensitivity in Fe₃O₄/Si structure is very high. The position dependence of LPE can be well fitted using a traditional theory: LPV = K₀[exp(−|L − x|/d) − exp(−|L + x|/d)], where the K₀ is a proportional coefficient, L is the electrodes' position (as shown in Fig. 4), d is the electron/hole diffusion length and x is the spot position.^10,24 The fitted data was shown in red line in Fig. 4. This result clearly suggests that the LPE can be explained by the traditional theory. The LPE originates from the lateral diffuse flow and recombination of the created carriers away from the laser spot, and the large sensitivity value is due to the higher parallel resistance of inversion layer and Fe₃O₄ film. The photon energy of He–Ne laser is larger than the bandgap of Si, the electron–hole pairs will be created in Si substrate under the radiation of the laser. The laser spot is very small, and the electron–hole pairs can only be created at the illuminated region and separated by the built-in field. So the created holes will flow to the Fe₃O₄ film side, and then move away from the illuminated spot to the two electrodes on Fe₃O₄ film. The concentration of the created carriers is different at different distance from the illuminated spot, therefore, the lateral field is set up and the LPE is observed. It is well-known that the concentration of the created carriers will be greatly influenced by the resistivity—higher resistivity can suppress the diffusing of created carriers to the two electrodes and result in higher spatial variation of the LPE. The measured resistivity on the Fe₃O₄ surface in Fe₃O₄/Si structure is about 500 Ω at room temperature, which is higher than that of metal-based structure, resulting in the created carriers hardly diffusing to the two electrodes compared to the Ti(Al)/Si structure.¹³ This increases the difference in the concentration of the created carriers arriving at the two electrodes when the laser spot is not in the middle, so the enhanced LPE in Fe₃O₄/Si junction is achieved with ultrathin nano-crystalline Fe₃O₄ film.

Fig. 4 The dependence of the lateral photovoltage on the laser position for Fe₃O₄/Si junction, the inset exhibits the schematic setup for LPE measurement and the transmittance spectra of Fe₃O₄ films on MgO or Al₂O₃.

The time response of the LPE of Fe₃O₄/Si junction was further investigated using a laser with a wavelength of 800 nm and 130 fs pulse duration and an oscilloscope of 100 MHz. Fig. 5(a) shows the variation of the open-circuit LPE with time. The schematic circuit of the measurement is shown in the inset of Fig. 5(a). The rise time is about 3.8 μs and the full width at half maximum (FWHM) is about 72.5 μs when the LPE is directly measured. The FWHM stands for the relaxation time of LPE. The sharp rise time means that the LPE originates from the photo-induced voltage instead of thermoelectric effect.^25,26 The input impedance of the oscilloscope is 1 MΩ which results in longer relaxation time due to the gradual decrease of LPE. To reduce the influence of the measurement circuit and simulate the true circuit, some resistances with different value were connected in parallel to the Fe₃O₄ film, as shown in the inset of Fig. 5(b). Interestingly, ultrafast response and relaxation time of LPE was observed. The rise time dramatically reduces to about 60 ns and the FWHM is also reduced to about 5 μs, which means the optical response time and relaxation time of LPE were ∼60 ns and 5 μs, respectively.

Fig. 5 (a) The variation of the open-circuit LPE with time, the schematic circuit of the measurement is shown in the inset; (b) the variation of the LPE with time, a resistance was connected in parallel to the Fe₃O₄ film, the schematic circuit of the measurement is shown in the inset.

The vertical photovoltage effect (VPE) based on silicon has been widely used in solar cell and shows a picosecond response time which is much faster than the response time of LPE.^27,28 The basic mechanism of both VPE and LPE is the separation of photo-generated electron–hole pairs spatially by the built-in field of the junction when the laser illuminates on the junction surface, but the origination of VPE and LPE is different. Once the laser illuminates on the junction, the created electron–hole pairs will be spatially separated by the built-in field and the VPE will be observed immediately, so an ultrafast photoelectric response is achieved. However, the LPE originates from the lateral diffuse flow and recombination of the created holes away from the laser spot after the electron–hole pairs was spatially separated by the built-in field, which will result in slower time response of LPE than that of VPE. But the response of LPE in Fe₃O₄/Si structure is much faster than that used in LPE devices,¹³ and suggests good quality of Fe₃O₄/Si junction. The relaxation time of LPE is greatly influenced by the measured resistivity between the two electrodes. The measured resistivity is the parallel resistance of inversion layer and Fe₃O₄ film in the Fe₃O₄/Si structure, which is higher than that of metal-based structure and is lower than that of semiconductor, so a fast response with microseconds relaxation time are achieved.

4. Conclusions

In conclusion, we have fabricated a Fe₃O₄/Si junction with good quality using PLD and investigated the LPE properties at room temperature. The rectifying behavior and large LPE were observed in the Fe₃O₄/Si Schottky junction. The dependence of the LPE on the position shows very highly sensitivity of 32.5 mV mm⁻¹ as well as good linearity. The measured resistivity of Fe₃O₄/Si structure is higher than that of metal films, so a high position sensitivity and fast response with microseconds relaxation time are observed. The high position sensitivity and ultrafast response speed make the Fe₃O₄/Si junction promising for wide range of optoelectronic device applications.

Acknowledgements

This work is supported by National Natural Science Foundation of China (No. 51472064), NSF (DMR-0852862 and CBET-0754821), UW SER and the U. S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering under Award DE-FG02-10ER46728, Program for Innovation Research of Science in Harbin Institute of Technology (PIRS of HIT A201413).

Notes and references

1. J. Wallmark, Proc. Inst. Radio Eng., 1957, 45, 474.
2. B. F. Levine, R. H. Willens, C. G. Bethea and D. Brasen, Appl. Phys. Lett., 1986, 49, 1537.
3. R. H. Willens, Appl. Phys. Lett., 1986, 49, 663.
4. K. J. Jin, K. Zhao, H. B. Lu, L. Liao and G. Z. Yang, Appl. Phys. Lett., 2007, 91, 081906.
5. S. Q. Xiao, H. Wang, Z. C. Zhao, Y. Z. Gu, Y. X. Xia and Z. H. Wang, Opt. Express, 2008, 16, 3798.
6. K. Zhao, K. J. Jin, H. B. Lu, Y. H. Huang, Q. L. Zhou, M. He, Z. H. Chen, Y. L. Zhou and G. Z. Yang, Appl. Phys. Lett., 2006, 88, 141914.
7. J. Henry and J. Livingstone, IEEE Sens. J., 2006, 6, 1557.
8. E. Chen and S. Y. Chou, Appl. Phys. Lett., 1997, 70, 753.
9. J. Henry and J. Livingstone, Adv. Mater., 2001, 13, 1022.
10. S. Q. Xiao, H. Wang, Z. C. Zhao and Y. X. Xia, J. Phys. D: Appl. Phys., 2007, 40, 5580.
11. S. Liu, H. Wang, Y. J. Yao, L. Chen and Z. L. Wang, Appl. Phys. Lett., 2014, 104, 111110.
12. S. H. Wang, W. X. Wang, L. K. Zou, X. Zhang, J. W. Cai, Z. G. Sun, B. G. Shen and J. R. Sun, Adv. Mater., 2014, 26, 8059.
13. L. Du and H. Wang, Opt. Express, 2010, 18, 9113.
14. C. Q. Yu, H. Wang, S. Q. Xiao and Y. X. Xia, Opt. Express, 2009, 17, 21712.
15. C. Q. Yu, H. Wang and Y. X. Xia, Appl. Phys. Lett., 2009, 95, 263506.
16. X. J. Wang, Y. Sui, J. K. Tang, C. Wang, Z. Lu, X. Q. Zhang, Z. G. Liu, W. H. Su, X. K. Wei and R. C. Yu, Appl. Phys. Lett., 2008, 92, 012122.
17. J. K. Tang, J. B. Dai, K. Y. Wang, W. L. Zhou, N. Ruzycki and U. Diebold, J. Appl. Phys., 2002, 91, 8411.
18. M. Fonin, R. Pentcheva, Y. S. Dedkov, M. Sperlich, D. V. Vyalikh, M. Scheffler, U. Rüdiger and G. Güntherodt, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 104436.
19. F. G. Allen and G. W. Gobeli, Phys. Rev., 1962, 127, 150.
20. T. L. Qu, Y. G. Zhao, H. F. Tian, C. M. Xiong, S. M. Guo and J. Q. Li, Appl. Phys. Lett., 2007, 90, 223514.
21. W. B. Mi, E. Y. Jiang and H. L. Bai, J. Appl. Phys., 2010, 107, 103922.
22. Z. Viskadourakis, M. L. Parames, O. Conde, M. Zervos and J. Giapintzakis, Appl. Phys. Lett., 2012, 101, 033505.
23. J. Panda, P. Banerjee and T. K. Nath, J. Phys. D: Appl. Phys., 2014, 47, 415103.
24. L. M. Chi, P. F. Zhu, H. Wang, X. A. Huang and X. T. Li, J. Opt., 2011, 13, 015601.
25. P. X. Zhang, W. K. Lee and G. Y. Zhang, Appl. Phys. Lett., 2002, 81, 4026.
26. J. R. Sun, C. M. Xiong, B. G. Shen, P. Y. Wang and Y. X. Weng, Appl. Phys. Lett., 2004, 84, 2611.
27. H. B. Lu, K. J. Jin, Y. H. Huang, M. He, K. Zhao, B. L. Cheng, Z. H. Chen, Y. L. Zhou, S. Y. Dai and G. Z. Yang, Appl. Phys. Lett., 2005, 86, 241915.
28. J. Xing, K. J. Jin, M. He, H. B. Lu, G. Z. Liu and G. Z. Yang, J. Phys. D: Appl. Phys., 2008, 41, 195103.

---