- Email: [email protected]
Study on the sensitivity decrease of reflection-mode GaN NEA photocathode
Optik 125 (2014) 6040–6043
Contents lists available at ScienceDirect
Optik journal homepage: www.elsevier.de/ijleo
Study on the sensitivity decrease of reflection-mode GaN NEA photocathode Jiangtao Fu ∗ , Guoqiang Zheng, Songchun Zhang, Huahong Ma Institute of Electronic Information Engineering, Henan University of Science and Technology, Luoyang 471003, People’s Republic of China
a r t i c l e
i n f o
Article history: Received 28 October 2013 Accepted 2 June 2014 Keywords: GaN photocathode Spectral response curve Incident wavelength Potential barrier Energy distribution
a b s t r a c t The experiment result discloses that the sensitivity of the reflection-mode NEA GaN photocathodes decrease with time after activation. Through analyzing the experiment result, it is found that the changes of the element content in the surface of the cathodes play an important role in the decrease course of the sensitivity. In order to explain the above result, a surface model of the reflection-mode NEA GaN photocathodes is put forward. By using this model, the inner mechanism leading to this result is that the dipoles direction changes because the harmful gases are adsorbed in the surface of the photocathodes. © 2014 Published by Elsevier GmbH.
1. Introduction
2. Experiment
Negative electric affinity (NEA) GaN photocathodes made by the coadsorption of cesium and oxygen alternatively on the surface have many superior characteristics such as high quantum efficiency, low dark current, good spectrum response and low energy spread, so they have a vast application in the field of low light intensifier, spin polarization electron source and so on [1–4]. The NEA GaN photocathodes are produced under the “three-step” photoelectric emission theory put forward by Spicer, according to this theory, the quantum efficiency formula of the NEA GaN photocathodes can be deduced through solving the diffusion equation of the unequilibrium carriers. Because the GaN photocathodes are negative electric affinity photocathodes, the escape probability in the quantum efficiency formula is generally thought to have nothing to do with the incident photo wavelength [5,6]. But when applying the theoretical quantum efficiency formula to analyze the spectrum response of the GaN photocathode, there is always a deviance between the theoretical result and the experimental result. From the angle of fitting the experimental data with theoretical data, the relationship between them is concluded.
A kind of uniform doping structure reflection-mode GaN sample was grown on the high quality p-type GaN oriented substrate with Be doping by molecular beam epitaxy (MB7/30/2014E). The thickness of the emission layer is 1.4 m and the doping concentration is 1 × 10−19 cm−3 . Before activation, the chemical cleaning of the sample is finished at first, after chemical cleaning the sample is placed into the heating position of the UHV system quickly, when the vacuum pressure in the system is lower than 1 × 10−17 Pa, the heat cleaning is carried out to purify the sample, the high temperature and the low temperature are 650 ◦ C and 410 ◦ C, respectively, the duration of the high temperature and the low temperature are 55 min and 15 min respectively. After heat cleaning, when the sample is cooled down to 60 ◦ C and the vacuum pressure in the system recover back to 1 × 10−17 Pa, then the two-step “yo-yo” activation technique is applied to activate the sample. After activation, the spectral response is measured through the spectrum response measuring instrument and the result is shown as Fig. 1. In Fig. 1(b), curves 5 and 2 represent the spectral response one hour and two hours later after low temperature activation respectively, curve 6 represents the spectral response right after low temperature activation. Comparing with curve 6, curves 5 and 2 drop down very clearly. In the process of the low temperature activation, the O–Cs dipoles are damaged because of the desorption of Cs, the spectral response decreases continuously, especially the spectrum response decrease more quickly in long wavelength band. This phenomenon shows the escape probability
∗ Corresponding author. E-mail address: [email protected] (J. Fu). http://dx.doi.org/10.1016/j.ijleo.2014.06.096 0030-4026/© 2014 Published by Elsevier GmbH.
J. Fu et al. / Optik 125 (2014) 6040–6043
6041
Fig. 1. Spectral response curves and fitting curves: (a) high temperature activation and (b) low temperature activation.
Fig. 2. Band structure of GaN photocathode.
within different wavelength range is different; otherwise the dropping tendency of the spectral responses should be the same. 4. Relationship between escape probability and the incident wavelength 3. Relationship between escape probability and quantum efficiency The escape probability P is the probability that the electronic can escape from the surface of the photocathode into the vacuum through tunneling through the surface potential barrier, so when the photon excited electrons reach the surface of the photocathode, their energy distribution and transmission rate will decide the value of P. Generally speaking, when the hot electrons excited by the photons transit from the valance band to the conduction band, they will lose some of their energy because of the scattering effect of the phonons and fall into the range within several Kt of the conduction band bottom, the energy distribution of the electrons satisfy the Fermi–Dirac distribution when they reach the band bending region. When the electrons pass the band bending region and get to the surface of the photocathode, their energy distribution n(E) is calculated by Barterlink [7–9]. When the electrons arrive in the surface of the photocathode, they must tunnel through the surface potential barrier into the vacuum, all the electrons can only get into the vacuum with a transmission rate because of the surface potential barrier, the transmission rate P(E) of the electrons with different energy can be acquired through solving the electron wave function equation [10], then the escape probability P can be decided by the transmission rate P(E) and the energy distribution n(E). Actually the escape probability P is the average value of the transmission rate of the electrons with different energy [9]. From the analysis above, the excited electrons quickly fall into the range within several Kt of the conduction band bottom, so their energy distribution does not have too much difference which means the energy distribution n(E) is independent on the energy of the incident photons, the only factor that can affect the energy distribution n(E) is the surface potential barrier, according to such conjecture, the quantum efficiency formula of the reflection-mode GaN photocathode can be acquired through the diffusion equation on the basis of the “three-step” model by Spicer [5,6,11]: QE =
P˛ LD (1 − R) 1 + ˛LD
(1)
where P is the escape probability, ˛ is the light absorption parameter, LD is the electronic diffusion length, R is the reflection parameter of the photocathode.
By applying the theoretical quantum efficiency formula (1), the data fitting of the experimental curves are carried out. The electron diffusion length LD can first be decided through fitting the experimental curves 2, 5, 6 and 7 within the wavelength ranging from 800 nm to 880 nm,the fitting results of the four curves have good consistency and the value is 1.8 m, then the fitting curves of the four experimental curves are acquired by using 1.8 m in formula (1) as shown in Fig. 1. The fitting curves do not match the experimental curves very well. This mismatch discloses the inconsistency between the experiment result and the theoretical formula. The reason leading to the inconsistency is the assumption that the escape probability independent on the incident wavelength cannot be applied for the NEA reflection mode photocathodes. In the reflection mode photocathodes, the light incident surface and the electrons emission surface are the same, so the photon excited electrons are mainly produced near the emission surface, then some electrons may escape into the vacuum before they fall into the bottom of the conduction band, this procedure can be realized through two different ways: one hand, the electrons in the conduction band can be thermalized at the bottom of the conduction band through relaxation, but the thermalization length is generally dozens of nanometers, so some electrons with high energy can escape into the vacuum with the hot electrons form; on the other hand, the photon excited electron first transmit from the valence band to valley of the conduction band, before they escape into the vacuum, some electrons may have the energy high enough to be scattered into the L valley of the conduction band, some of these electrons in the L valley can directly escape into the vacuum, the photon excited electrons with energy lower than the L valley but higher than the minimum of valley will probably escape into the vacuum in the shape of hot electrons. With the increase of the incident wavelength, the energy distribution of the electrons will move to the high energy direction. Some experiments have proved this result [12–15]. The surface band structure of the photocathode is shown as Fig. 2. The surface potential barrier is composed of two straight lines with different slopes: barriers I and II which is based on the double-dipole model. The dot lines b and c in Fig. 2 represent different barrier shapes under different situations [6,8,9]. In the process of activation, after first exposure to Cs, the surface potential barrier is just a approximate straight line, with continuously exposure to Cs
6042
J. Fu et al. / Optik 125 (2014) 6040–6043
Fig. 4. Spectral response curves and fitting curves by applying formula (3): (a) high temperature activation and (b) low temperature activation. Fig. 3. Relationship of P versus .
and O alternatively, barrier I becomes narrower smaller, at last the surface potential barrier is formed as the solid line in Fig. 2. In the process of degradation (illuminate the photocathode with white light), the variation of the surface potential barrier is contrary to the situation in the process of activation. The electrons excited by the short wavelength photons have relatively higher energy, the barrier for them to tunnel through is narrow, so their probability to escape is high; the electrons excited by the long wavelength photons have the opposite situation. Because of the different energy distribution of the electrons under different incident wavelength and the surface potential barriers, the width that the electrons excited by photons with different wavelength need tunnel through is different. The electrons excited by the short wavelength photons have higher energy than those by the long wavelength photons, the distance they need tunnel through is shorter, and naturally they have a higher probability to escape into the vacuum. This result can suggest the escape probability P maybe a function of the incident wavelength. In order to find out the relationship between the escape probability and the incident wavelength, the escape probability P is calculated within a small wavelength range of ±10 nm by fitting the experimental curves 5, 6 and 7. The fitting result of the escape probability P versus the incident wavelength is shown in Fig. 3. Because the logarithmic coordinates is applied in Fig. 3, so it is speculated that escape probability P and the incident wavelength satisfy the following exponential equation: P = Ps exp[−l( − s )]
(2)
In formula (2), s is the starting wavelength of the spectral response, Ps is the escape probability under the starting wavelength, l is a factor decided by the surface potential barrier, thus formula (1) can be modified as: QE =
Ps exp[−l( − s )]˛LD (1 − R) 1 + ˛LD
(3)
By applying formula (3), the data fitting curves of the experimental curves 7, 6, 5 and 2 are shown as Fig. 4, the performance parameters acquired through fitting the curves are listed in Table 1.
Table 1 Fitting parameters. Curves
Ps
l
LD (m)
2 5 6 7
0.51 0.76 0.85 0.90
4.90 2.40 1.36 3.64
1.8 1.8 1.8 1.8
It can be seen from Fig. 4 that all the fitting curves match the experimental curves very well. Actually the parameter l reflects the shape variation of the surface potential barrier. When the shape of the surface potential barrier changes, the electrons with low energy will be affected more than those with high energy. After the low temperature activation, the performance of the photocathode is improved because of the structure optimization of the Cs–O layer, the surface potential barrier will become smaller and narrower, and the parameter l reflects this through the value changing from 1.36 to 3.64. On the contrary, after low temperature activation, when the performance of the photocathode become poor because of the desorption of Cs, the surface potential barrier will become higher and wider, the parameter l reflects this through the value changing from 1.36 to 4.90. 5. Conclusion When applying the quantum efficiency formula to analyze the experimental result of spectral response, the theoretical results have deviance with the experimental result. In order to find out the reason leading to the differences, some analysis about the spectral response are carried out, then it is found that the escape probability of the electrons is dependent on the incident wavelength, the reasons that the incident wavelength change the escape probability are analyzed, and the relationship between the escape probability and the incident photon wavelength approximately satisfy the exponential equation, then this conclusion is proved through fitting the experimental curves by applying the new quantum efficiency formula, the result shows that the fitting curves have a good consistency with the experimental curves. References [1] X.Q. Du, B.K. Chang, J.J. Zou, High quantum efficiency GaAs photocathode by gradient doping, Acta Opt. Sin. 25 (10) (2005) 1411–1414. [2] Chang Benkang, Xu Denggao, Qian Yunsheng, Peak value position shifting of spectral response of multialkali photocathodes, Acta Opt. Sin. 18 (2) (1998) 233–237. [3] Du Yujie, Chang Benkang, Wang Xiaohui, Theoretical study of Cs adsorption on GaN (0 0 0 1) surface, Appl. Surf. Sci. 258 (19) (2013) 7425–7429. [4] Jiao Gangcheng, Hu Canglu, Chen Liang, Comparative research on Cs/O activation source ratio for GaAs photocathodes, Opt. Commun. 293 (2013) 65–68. [5] W.E. Spicer, A. Gomez, Modern theory and application of photocathodes, Proc. SPIE 2022 (1993) 18–33. [6] Zong Zhiyuan, Chang Benkang, Calculating the quantum yield of NEA photocathodes by integral, Acta Opt. Sin. 19 (9) (1999) 1177–1182. [7] D.J. Barterlink, J.L. Moll, N.I. Meyer, Hot-electron emission from sha p–n junctions in a silicon, Phys. Rev. 130 (3) (1963) 972–985. [8] K.L. Jensen, Scattering and the relationship between quantum efficiency and emittance, J. Appl. Phys. 113 (5) (2013). [9] Xiaohua Yu, Yujie Du, Chang Bengkang, Study on the electron structure and optical properties of Ga0.5 Al0.5 As (1 0 0) beta(2)(2 × 4) reconstruction surface, Appl. Surf. Sci. 266 (2013) 380–385.
J. Fu et al. / Optik 125 (2014) 6040–6043 [10] J. Xie, X. Ni, M. Wu, J.H. Leach, M.A. Py, U. Ozgur, H. Morkoc, Electron mobility in InGaN channel heterostructure field effect transistor structures with different barriers, Appl. Phys. Lett. 91 (2007) 2664–2667. [11] Jijun Zou, Yijun Zhang, Xincun Peng, Energy distributions of electrons emitted from reflection-mode Cs-covered GaAs photocathodes, Appl. Opt. 51 (31) (2013) 7662–7667. [12] C.Y. Su, W.E. Spicer, I. Lindau, photoelectron spectroscopic determination of the structure (Cs,O) activated GaAs (1 1 0) surface, Appl. Phys. 54 (3) (1983) 1413–1422.
6043
[13] Y.J. Zhang, B.K. Chang, J. Niu, J. Zhao, J.J. Zou, F. Shi, H.C. Cheng, High-efficiency graded band-gap Alx Ga1−x As/GaAs photocathodes grown by metallorganic chemical vapor deposition, Appl. Phys. Lett. 99 (2011) 356–359. [14] X.Y. Guo, X.H. Wang, B.K. Chang, Y.J. Zhang, P. Gao, High quantum efficiency of depth grade doping negative-electron-affinity GaN photocathode, Appl. Phys. Lett. 98 (2011) 256–259. [15] Xinlong Chen, Yijun Zhang, Benkang Chang, Research on quantum efficiency of reflection-mode GaAs photocathode with thin emission layer, Opt. Commun. 287 (2013) 35–39.
Contents lists available at ScienceDirect
Optik journal homepage: www.elsevier.de/ijleo
Study on the sensitivity decrease of reflection-mode GaN NEA photocathode Jiangtao Fu ∗ , Guoqiang Zheng, Songchun Zhang, Huahong Ma Institute of Electronic Information Engineering, Henan University of Science and Technology, Luoyang 471003, People’s Republic of China
a r t i c l e
i n f o
Article history: Received 28 October 2013 Accepted 2 June 2014 Keywords: GaN photocathode Spectral response curve Incident wavelength Potential barrier Energy distribution
a b s t r a c t The experiment result discloses that the sensitivity of the reflection-mode NEA GaN photocathodes decrease with time after activation. Through analyzing the experiment result, it is found that the changes of the element content in the surface of the cathodes play an important role in the decrease course of the sensitivity. In order to explain the above result, a surface model of the reflection-mode NEA GaN photocathodes is put forward. By using this model, the inner mechanism leading to this result is that the dipoles direction changes because the harmful gases are adsorbed in the surface of the photocathodes. © 2014 Published by Elsevier GmbH.
1. Introduction
2. Experiment
Negative electric affinity (NEA) GaN photocathodes made by the coadsorption of cesium and oxygen alternatively on the surface have many superior characteristics such as high quantum efficiency, low dark current, good spectrum response and low energy spread, so they have a vast application in the field of low light intensifier, spin polarization electron source and so on [1–4]. The NEA GaN photocathodes are produced under the “three-step” photoelectric emission theory put forward by Spicer, according to this theory, the quantum efficiency formula of the NEA GaN photocathodes can be deduced through solving the diffusion equation of the unequilibrium carriers. Because the GaN photocathodes are negative electric affinity photocathodes, the escape probability in the quantum efficiency formula is generally thought to have nothing to do with the incident photo wavelength [5,6]. But when applying the theoretical quantum efficiency formula to analyze the spectrum response of the GaN photocathode, there is always a deviance between the theoretical result and the experimental result. From the angle of fitting the experimental data with theoretical data, the relationship between them is concluded.
A kind of uniform doping structure reflection-mode GaN sample was grown on the high quality p-type GaN oriented substrate with Be doping by molecular beam epitaxy (MB7/30/2014E). The thickness of the emission layer is 1.4 m and the doping concentration is 1 × 10−19 cm−3 . Before activation, the chemical cleaning of the sample is finished at first, after chemical cleaning the sample is placed into the heating position of the UHV system quickly, when the vacuum pressure in the system is lower than 1 × 10−17 Pa, the heat cleaning is carried out to purify the sample, the high temperature and the low temperature are 650 ◦ C and 410 ◦ C, respectively, the duration of the high temperature and the low temperature are 55 min and 15 min respectively. After heat cleaning, when the sample is cooled down to 60 ◦ C and the vacuum pressure in the system recover back to 1 × 10−17 Pa, then the two-step “yo-yo” activation technique is applied to activate the sample. After activation, the spectral response is measured through the spectrum response measuring instrument and the result is shown as Fig. 1. In Fig. 1(b), curves 5 and 2 represent the spectral response one hour and two hours later after low temperature activation respectively, curve 6 represents the spectral response right after low temperature activation. Comparing with curve 6, curves 5 and 2 drop down very clearly. In the process of the low temperature activation, the O–Cs dipoles are damaged because of the desorption of Cs, the spectral response decreases continuously, especially the spectrum response decrease more quickly in long wavelength band. This phenomenon shows the escape probability
∗ Corresponding author. E-mail address: [email protected] (J. Fu). http://dx.doi.org/10.1016/j.ijleo.2014.06.096 0030-4026/© 2014 Published by Elsevier GmbH.
J. Fu et al. / Optik 125 (2014) 6040–6043
6041
Fig. 1. Spectral response curves and fitting curves: (a) high temperature activation and (b) low temperature activation.
Fig. 2. Band structure of GaN photocathode.
within different wavelength range is different; otherwise the dropping tendency of the spectral responses should be the same. 4. Relationship between escape probability and the incident wavelength 3. Relationship between escape probability and quantum efficiency The escape probability P is the probability that the electronic can escape from the surface of the photocathode into the vacuum through tunneling through the surface potential barrier, so when the photon excited electrons reach the surface of the photocathode, their energy distribution and transmission rate will decide the value of P. Generally speaking, when the hot electrons excited by the photons transit from the valance band to the conduction band, they will lose some of their energy because of the scattering effect of the phonons and fall into the range within several Kt of the conduction band bottom, the energy distribution of the electrons satisfy the Fermi–Dirac distribution when they reach the band bending region. When the electrons pass the band bending region and get to the surface of the photocathode, their energy distribution n(E) is calculated by Barterlink [7–9]. When the electrons arrive in the surface of the photocathode, they must tunnel through the surface potential barrier into the vacuum, all the electrons can only get into the vacuum with a transmission rate because of the surface potential barrier, the transmission rate P(E) of the electrons with different energy can be acquired through solving the electron wave function equation [10], then the escape probability P can be decided by the transmission rate P(E) and the energy distribution n(E). Actually the escape probability P is the average value of the transmission rate of the electrons with different energy [9]. From the analysis above, the excited electrons quickly fall into the range within several Kt of the conduction band bottom, so their energy distribution does not have too much difference which means the energy distribution n(E) is independent on the energy of the incident photons, the only factor that can affect the energy distribution n(E) is the surface potential barrier, according to such conjecture, the quantum efficiency formula of the reflection-mode GaN photocathode can be acquired through the diffusion equation on the basis of the “three-step” model by Spicer [5,6,11]: QE =
P˛ LD (1 − R) 1 + ˛LD
(1)
where P is the escape probability, ˛ is the light absorption parameter, LD is the electronic diffusion length, R is the reflection parameter of the photocathode.
By applying the theoretical quantum efficiency formula (1), the data fitting of the experimental curves are carried out. The electron diffusion length LD can first be decided through fitting the experimental curves 2, 5, 6 and 7 within the wavelength ranging from 800 nm to 880 nm,the fitting results of the four curves have good consistency and the value is 1.8 m, then the fitting curves of the four experimental curves are acquired by using 1.8 m in formula (1) as shown in Fig. 1. The fitting curves do not match the experimental curves very well. This mismatch discloses the inconsistency between the experiment result and the theoretical formula. The reason leading to the inconsistency is the assumption that the escape probability independent on the incident wavelength cannot be applied for the NEA reflection mode photocathodes. In the reflection mode photocathodes, the light incident surface and the electrons emission surface are the same, so the photon excited electrons are mainly produced near the emission surface, then some electrons may escape into the vacuum before they fall into the bottom of the conduction band, this procedure can be realized through two different ways: one hand, the electrons in the conduction band can be thermalized at the bottom of the conduction band through relaxation, but the thermalization length is generally dozens of nanometers, so some electrons with high energy can escape into the vacuum with the hot electrons form; on the other hand, the photon excited electron first transmit from the valence band to valley of the conduction band, before they escape into the vacuum, some electrons may have the energy high enough to be scattered into the L valley of the conduction band, some of these electrons in the L valley can directly escape into the vacuum, the photon excited electrons with energy lower than the L valley but higher than the minimum of valley will probably escape into the vacuum in the shape of hot electrons. With the increase of the incident wavelength, the energy distribution of the electrons will move to the high energy direction. Some experiments have proved this result [12–15]. The surface band structure of the photocathode is shown as Fig. 2. The surface potential barrier is composed of two straight lines with different slopes: barriers I and II which is based on the double-dipole model. The dot lines b and c in Fig. 2 represent different barrier shapes under different situations [6,8,9]. In the process of activation, after first exposure to Cs, the surface potential barrier is just a approximate straight line, with continuously exposure to Cs
6042
J. Fu et al. / Optik 125 (2014) 6040–6043
Fig. 4. Spectral response curves and fitting curves by applying formula (3): (a) high temperature activation and (b) low temperature activation. Fig. 3. Relationship of P versus .
and O alternatively, barrier I becomes narrower smaller, at last the surface potential barrier is formed as the solid line in Fig. 2. In the process of degradation (illuminate the photocathode with white light), the variation of the surface potential barrier is contrary to the situation in the process of activation. The electrons excited by the short wavelength photons have relatively higher energy, the barrier for them to tunnel through is narrow, so their probability to escape is high; the electrons excited by the long wavelength photons have the opposite situation. Because of the different energy distribution of the electrons under different incident wavelength and the surface potential barriers, the width that the electrons excited by photons with different wavelength need tunnel through is different. The electrons excited by the short wavelength photons have higher energy than those by the long wavelength photons, the distance they need tunnel through is shorter, and naturally they have a higher probability to escape into the vacuum. This result can suggest the escape probability P maybe a function of the incident wavelength. In order to find out the relationship between the escape probability and the incident wavelength, the escape probability P is calculated within a small wavelength range of ±10 nm by fitting the experimental curves 5, 6 and 7. The fitting result of the escape probability P versus the incident wavelength is shown in Fig. 3. Because the logarithmic coordinates is applied in Fig. 3, so it is speculated that escape probability P and the incident wavelength satisfy the following exponential equation: P = Ps exp[−l( − s )]
(2)
In formula (2), s is the starting wavelength of the spectral response, Ps is the escape probability under the starting wavelength, l is a factor decided by the surface potential barrier, thus formula (1) can be modified as: QE =
Ps exp[−l( − s )]˛LD (1 − R) 1 + ˛LD
(3)
By applying formula (3), the data fitting curves of the experimental curves 7, 6, 5 and 2 are shown as Fig. 4, the performance parameters acquired through fitting the curves are listed in Table 1.
Table 1 Fitting parameters. Curves
Ps
l
LD (m)
2 5 6 7
0.51 0.76 0.85 0.90
4.90 2.40 1.36 3.64
1.8 1.8 1.8 1.8
It can be seen from Fig. 4 that all the fitting curves match the experimental curves very well. Actually the parameter l reflects the shape variation of the surface potential barrier. When the shape of the surface potential barrier changes, the electrons with low energy will be affected more than those with high energy. After the low temperature activation, the performance of the photocathode is improved because of the structure optimization of the Cs–O layer, the surface potential barrier will become smaller and narrower, and the parameter l reflects this through the value changing from 1.36 to 3.64. On the contrary, after low temperature activation, when the performance of the photocathode become poor because of the desorption of Cs, the surface potential barrier will become higher and wider, the parameter l reflects this through the value changing from 1.36 to 4.90. 5. Conclusion When applying the quantum efficiency formula to analyze the experimental result of spectral response, the theoretical results have deviance with the experimental result. In order to find out the reason leading to the differences, some analysis about the spectral response are carried out, then it is found that the escape probability of the electrons is dependent on the incident wavelength, the reasons that the incident wavelength change the escape probability are analyzed, and the relationship between the escape probability and the incident photon wavelength approximately satisfy the exponential equation, then this conclusion is proved through fitting the experimental curves by applying the new quantum efficiency formula, the result shows that the fitting curves have a good consistency with the experimental curves. References [1] X.Q. Du, B.K. Chang, J.J. Zou, High quantum efficiency GaAs photocathode by gradient doping, Acta Opt. Sin. 25 (10) (2005) 1411–1414. [2] Chang Benkang, Xu Denggao, Qian Yunsheng, Peak value position shifting of spectral response of multialkali photocathodes, Acta Opt. Sin. 18 (2) (1998) 233–237. [3] Du Yujie, Chang Benkang, Wang Xiaohui, Theoretical study of Cs adsorption on GaN (0 0 0 1) surface, Appl. Surf. Sci. 258 (19) (2013) 7425–7429. [4] Jiao Gangcheng, Hu Canglu, Chen Liang, Comparative research on Cs/O activation source ratio for GaAs photocathodes, Opt. Commun. 293 (2013) 65–68. [5] W.E. Spicer, A. Gomez, Modern theory and application of photocathodes, Proc. SPIE 2022 (1993) 18–33. [6] Zong Zhiyuan, Chang Benkang, Calculating the quantum yield of NEA photocathodes by integral, Acta Opt. Sin. 19 (9) (1999) 1177–1182. [7] D.J. Barterlink, J.L. Moll, N.I. Meyer, Hot-electron emission from sha p–n junctions in a silicon, Phys. Rev. 130 (3) (1963) 972–985. [8] K.L. Jensen, Scattering and the relationship between quantum efficiency and emittance, J. Appl. Phys. 113 (5) (2013). [9] Xiaohua Yu, Yujie Du, Chang Bengkang, Study on the electron structure and optical properties of Ga0.5 Al0.5 As (1 0 0) beta(2)(2 × 4) reconstruction surface, Appl. Surf. Sci. 266 (2013) 380–385.
J. Fu et al. / Optik 125 (2014) 6040–6043 [10] J. Xie, X. Ni, M. Wu, J.H. Leach, M.A. Py, U. Ozgur, H. Morkoc, Electron mobility in InGaN channel heterostructure field effect transistor structures with different barriers, Appl. Phys. Lett. 91 (2007) 2664–2667. [11] Jijun Zou, Yijun Zhang, Xincun Peng, Energy distributions of electrons emitted from reflection-mode Cs-covered GaAs photocathodes, Appl. Opt. 51 (31) (2013) 7662–7667. [12] C.Y. Su, W.E. Spicer, I. Lindau, photoelectron spectroscopic determination of the structure (Cs,O) activated GaAs (1 1 0) surface, Appl. Phys. 54 (3) (1983) 1413–1422.
6043
[13] Y.J. Zhang, B.K. Chang, J. Niu, J. Zhao, J.J. Zou, F. Shi, H.C. Cheng, High-efficiency graded band-gap Alx Ga1−x As/GaAs photocathodes grown by metallorganic chemical vapor deposition, Appl. Phys. Lett. 99 (2011) 356–359. [14] X.Y. Guo, X.H. Wang, B.K. Chang, Y.J. Zhang, P. Gao, High quantum efficiency of depth grade doping negative-electron-affinity GaN photocathode, Appl. Phys. Lett. 98 (2011) 256–259. [15] Xinlong Chen, Yijun Zhang, Benkang Chang, Research on quantum efficiency of reflection-mode GaAs photocathode with thin emission layer, Opt. Commun. 287 (2013) 35–39.