- Email: [email protected]
Improved quantum efficiency and stability of GaAs photocathode using favorable illumination during activation
Ultramicroscopy 202 (2019) 128–132
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Improved quantum efficiency and stability of GaAs photocathode using favorable illumination during activation
T
⁎
Cheng Fenga,c, Yijun Zhanga, , Yunsheng Qiana, Jian Liua, Jingzhi Zhanga, Feng Shib, Xiaofeng Baib, Jijun Zoud a
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China Science and Technology on Low-Light-Level Night Vision Laboratory, Xi'an 710065, China c School of Information and Communication Engineering, Nanjing Institute of Technology, Nanjing 211167, China d Engineering Research Center of New Energy Technology of Jiangxi Province, East China University of Technology, Nanchang 330013, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: GaAs photocathode Activation Illumination condition Photoemission performance
Considering that illumination using the light source is required to monitor the activation progress of negativeelectron-affinity GaAs-based photocathodes, it is important to understand if and how the illumination affects the activation. To improve the photoemission performance of GaAs photocathode, epitaxial GaAs samples are Cs/O2 activated and recesiated under illumination of monochromatic light of blue (460 nm), green (532 nm), and red (633 nm) with approximately the same incident photons, and halogen tungsten lamp as white light source, respectively, to induce photoemission. The performance characteristics including quantum efficiency and photocurrent degradation among the samples treated by different illumination conditions are compared to investigate their photoemission capability and stability. The results show that GaAs photocathodes activated and recesiated under illumination of monochromatic red light, can obtain higher quantum efficiency, and better stability. This work provides a favorable illumination condition during activation for obtaining satisfactory capability of GaAs-based photocathodes, which will be propitious for applications as spin-polarized electron sources.
1. Introduction Photoemission from III-V group GaAs-based negative-electron-affinity (NEA) photocathodes has shown numerous applications in important physic fields due to the advantages in high quantum efficiency, low mean transverse energy, high spin polarization, and good longwavelength response [1-3]. It has already been put to practical use in vacuum photodetectors, solar cells, polarized electron sources [4-7]. In recent years, such photocathodes have shown good application prospect as electron sources in low-energy electron microscopes and transmission electron microscopy, which requires high spin polarization and high brightness to obtain excellent image contrast and fast image acquisition [8-10]. The spin-polarized electron beam is generally produced by NEA GaAs-based photocathodes, while the NEA condition of the photocathode surface, a prerequisite for achieving high quantum efficiency, can be usually realized by Cs/O2 (or Cs/NF3) activation. The quantum efficiency and lifetime are important performance characteristics of GaAs-based photocathodes, which mainly depend on the material
⁎
structures and the technique during preparation [11-13]. An ideal photocathode should have a high quantum efficiency and a long operational lifetime, moreover, it should have the potential to withstand an adverse environment, such as the influence of residual gas and desorbed molecules. As everyone knows, NEA GaAs-based photocathodes are sensitive to contaminants in the vacuum environment, and the inferiority of short lifetime results in repetition of NEA preparation. Extensive efforts have been done to further promote the quantum efficiency and increase the operational lifetime, such as designing graded bandgap, strained, and superlattice structures to optimize cathode structures [14-17], finding more efficacious methods to remove the adsorbed contaminants in surface cleaning procedures [18-20], using more ideal solid O dispensers to release oxygen effectively during Cs/O activation process [13], and enhancing the immunity of photocathodes by introducing different electronegative substances [1,21,22]. As is well known, illumination using the light source is required during activation to monitor the activation process and adjust the operating current passing through the Cs and O dispensers [13,23]. White light and monochromatic laser light such as argon ion laser, He-Ne
Corresponding author. E-mail address: [email protected] (Y. Zhang).
https://doi.org/10.1016/j.ultramic.2019.04.010 Received 11 May 2018; Received in revised form 18 March 2019; Accepted 22 April 2019 Available online 23 April 2019 0304-3991/ © 2019 Elsevier B.V. All rights reserved.
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
laser, and 532 nm laser have been utilized for the photoelectron excitation from photocathodes [13,22-26]. These illumination conditions during the activation and recesiation process have not been comparatively analyzed, and how they affect the photoemission performance needs to be investigated. In this paper, the monochromatic light with blue (460 nm), green (532 nm), and red (633 nm) wavelength and the white light are respectively illuminated to GaAs photocathodes during the activation and recesiation process. We focus on the effect of different illumination conditions on quantum efficiency and emission stability of the GaAs photocathode samples. This work may further optimize the activation technique and enhance the photoemission performance of NEA GaAs-based photocathodes. 2. Experimental Fig. 2. Photocurrent change curves during Cs/O activation for the four GaAs cathode samples illuminated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
A kind of reflection-mode GaAs photocathode epitaxial wafer grown by the metalorganic chemical vapor deposition technique was chosen in our experiments, and four 11×11 mm2 square GaAs samples were cleaved from the same wafer to carry out the research on the activation and recesiation process under different illumination conditions. The GaAs epilayers consisting of a GaAs emission layer and an AlxGa1−xAs buffer layer are grown on the n-type (100)-oriented GaAs substrate, where the 1 µm-thick GaAs emission layers is quasi-exponentially doped with the zinc concentration ranging from 1 × 1019 cm−3 to 1 × 1018 cm−3, and the 0.5 µm-thick AlxGa11−xAs buffer layer with the doping concentration of 1 × 1019 cm−3 is of a graded Al proportion varying from 0.9 to 0. This graded bandgap structure was reported to have more advantages in quantum efficiency than the traditional structure [27]. Before activation, the GaAs cathode samples were chemically etched by the HF (40%) solution for 5 min right after being degreased. Then the samples were mounted on the stainless holder and transferred into the ultra-high vacuum (UHV) chamber as shown in Fig. 1. In the UHV chamber with the base pressure of about 1 × 10−7 Pa, the samples underwent a 650 ℃ heat-clean treatment for 20 min to obtain an atomically clean surface [28,29], and the Cs/O activation process was began to form NEA surface after the samples were cooled down to room temperature. During the activation process, for the four GaAs samples numbered by sample 1 to sample 4, four different illumination conditions corresponding to the samples were respectively adopted, i.e. (a) a halogen tungsten lamp with an intensity of 100lx as the white light source, (b) a blue (460 nm) monochromatic light with the power of 40.0 μW, (c) a green (532 nm) monochromatic light with the power of 34.6 μW, (d) a red (633 nm) monochromatic light with the power of 29.1 μW. The monochromatic light with the different wavelengths was generated by the monochromator and the light power was different to ensure their incident photons are approximately the same. As shown in Fig. 1, the incident light irradiated on the masked photocathode sample surface with a diameter of 10 mm was introduced by the optical fiber into the vacuum chamber. The activation process for the GaAs samples was carried out with the improved Cs/O co-deposition technique based on the solid Cs and O dispensers [13]. When there was no significant
increase between the present photocurrent peak and the previous one, the activation process was terminated with the Cs flux. The quantum efficiency was measured immediately followed by the measurement of photocurrent degradation, during which all samples were measured under illumination of the same halogen tungsten lamp with intensity of 100lx. After the first degradation, four recesiations for each sample were implemented by feeding fresh cesium flux to investigate the photocurrent recovery and degradation. During the recesiation process, the illumination condition was still different for each sample just as that in the initial Cs/O activation process, while in the multiple degradation processes, the samples were still illuminated with the same intensive white light of 100lx. 3. Results and discussion 3.1. Cs/O2 activation In the continuous activation process, the GaAs sample surface was exposed to alternate doses of Cs and O to increase the photocurrent peak, and the entire change of photocurrent with time was recorded online by the computer, as shown in Fig. 2. The photocurrent change curves show that sample 1 has much higher photocurrent peaks and shorter period compared with other samples, which is due to the high intensity of halogen tungsten lamp. For samples 2–4, they need more time to arrive at their first photocurrent maximum than sample 1 in the stage of only Cs adsorption. When only Cs flux is applied at the beginning of the activation, the photocurrent of the sample under illumination of 460 nm monochromatic light takes off earlier than those under illumination of 532 nm and 633 nm monochromatic light. More interestingly, the sample under illumination of 633 nm monochromatic light needs more time. This phenomenon could mean that for the activation using the monochromatic light, the Cs adsorption on the GaAs surface under illumination of blue photons during the initial activation stage will get easier. In other words, it turns out that electrons can be more easily excited to the conduction-band minimum by shortwave illumination light with higher photon energy, which is the main determinant to shorten the time of reaching the first Cs saturation. During the subsequent Cs/O co-deposition process, under illumination of 633 nm monochromatic light, the Cs/O alternation frequency is obviously higher, and more Cs/O alternation cycles can be achieved. Thus it is inferred that although the activation using the 633 nm monochromatic light takes a longer time, it is more conducive to the deposition of Cs and O on the GaAs surface, which can be favorable for the full formation of surface dipoles. The quantum efficiency curves of these GaAs samples were measured in situ right after the activation, as shown in Fig. 3. The quantum efficiency of the NEA GaAs cathode samples keeps at a high level when
Fig. 1. Schematic diagram of the activation equipment for preparing NEA GaAs photocathodes. 129
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
Fig. 3. Measured quantum efficiency curves after Cs/O activation for the four GaAs cathode samples illuminated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
Fig. 4. First photocurrent degradation curves under 100lx white light illumination after Cs/O activation for the four GaAs cathode samples activated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
the photon energy exceeds the GaAs bandgap. By comparison of the quantum efficiency curves of these samples, it is found that the GaAs photocathode activated under white light illumination using the halogen tungsten lamp has a passable quantum efficiency in the shortwave region, however the quantum efficiency drops a lot with the increase in wavelength. Unlike the sample activated under the white light illumination, the samples activated under illumination of monochromatic light possess a relatively flatter quantum efficiency curve over the whole wavelength region, except for the sample activated using the blue monochromatic light. The experimental results turn out that the GaAs photocathode activated under illumination of the 633 nm monochromatic light exhibits more excellent photoemission performance, especially in the long-wavelength region. The reason for the improved quantum efficiency ascribed to the Cs/O activation using the 633 nm monochromatic light can be complex and microscopic. It is considered that, on one hand, the height and width of the surface barrier related to the surface dipoles are changed under different illumination conditions, which affect the surface escape probability for the photoelectrons at different wavelengths. It might be more beneficial to reduce the barrier height and width by illumination with the red monochromatic light during activation, and the emission probability of the low energy electrons can be enhanced. On the other hand, the proper illumination during activation is an important factor to increase the probability of Cs and O adsorption, and more intense illumination will accelerate the activation process because the surface diffusion and migration and the bonding states transition are aggravated [23]. Intensive light will result in a decrease of photoemission performance because it may destroy the bonding of Cs-GaAs dipoles and Cs-O diploes. Meanwhile, the intensive illumination may promote the adsorption of unnecessary oxygen-containing residual gases. Accordingly, we believe that in the case of the monochromatic light accompanied by the same incident photons, the red monochromatic light illumination during activation can be conducive to the higher quantum efficiency in the entire wavelength region.
tungsten lamp irradiates the activated GaAs cathode samples, the photocurrent of sample 1 instantly degrades. By contrast, the photocurrent curves of samples 2, 3, and 4 rise at the beginning, which could be related to the rearrangement of surface dipoles under the intense illumination. During the initial Cs/O activation experiments, the activation processes for these samples were terminated by the excess Cs flux. For the samples activated using the monochromatic light, the higher light intensity after activation will lead to additional absorption of residual cesium in the vacuum, cause reconfiguration of surface dipoles, and promote the photoelectric conversion until Cs resaturation and new equilibrium under the new illumination condition. After that, the cathode performance begins to deteriorate due to the destruction of the optimum arrangement of surface dipoles. The residual gases in the chamber were mainly H2, H2O, CO and CO2 detected by the quadrupole mass spectrometer, wherein the partial pressures of H2, H2O, CO and CO2 were in the magnitude of 10-7 Pa, 10-8 Pa, 10-8 Pa and 10-9 Pa, respectively. The intense white light involving more high energy photons will contribute energy to more cracking of oxygen-containing residual gases and promote the adsorption of unnecessary residual gases, which is adverse to the formation of surface dipoles and degrade the photoemission. Thus the photocurrent of the GaAs sample activated with the white light exhibits the feature of faster degradation speed. In addition, the saturated photocurrent values show that the GaAs photocathode activated under illumination of the 633 nm monochromatic light can achieve higher photocurrent peak in contrast to those activated under other illumination conditions. In general, the capability of GaAs samples can be recovered by feeding the fresh Cs flux, and the illumination conditions may have effect on the recovery of cathode performance. Illumination of appropriate light is favorable, which can help delay the adsorption of residual gases, and at the same time, not hinder the adsorption of Cs. Accordingly, four recesiations were implemented after the first degradation of each sample, and the illumination conditions during the recesiations are the same as the corresponding ones during Cs/O activations. The lifetime after each regeneration was measured, as shown in Fig. 5. It is found from Fig. 5 that although the introduced the fresh cesium can help recover the capability, it can hardly eliminate the influence of the substrate oxides formed during degradation, and the complementary cesium can't form robust dipoles with oxygen atoms or Ga-As atoms. Therefore, there is an obvious and fast drop in photocurrent of the regenerated cathode. During the next several recesiations, the recovery of their initial photocurrent values become increasingly worse. It turns out that the lifetimes after the first recesiation decrease to less than half their original levels after the Cs/O activation
3.2. Degradations after activation and recesiations The illumination conditions during activation can influence the Cs/ O activation process and the final quantum efficiency, furthermore, the stability of the GaAs photocathodes activated under the different illumination conditions during activation can also be affected. The photocurrent degradation curves after activation were measured under the white light illumination with intensity of 100lx, which are presented in Fig. 4. It is obviously seen from Fig. 4 that the processes of photocurrent degradation among the four samples are different. When the halogen 130
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
Fig. 5. Photocurrent degradation curves under 100lx white light illumination after four recesiations for (a) sample 1 recesiated with white light, (b) sample 2 recesiated with 460 nm monochromatic light, (c) sample 3 recesiated with 532 nm monochromatic light, and (d) sample 4 recesiated with 633 nm monochromatic light.
whether the Cs/O activation under illumination of the broadband red light via band-pass filter or other monochrome red light has the same positive effect requires further exploration.
for these samples. The photocurrent degradation curve shows that the decline of initial photocurrent and operational lifetime of sample 1 activated using the white light are particularly obvious after each recesiation, while the samples activated using the monochromatic light are relative stable. Despite the photoemission capability can be recovered by recesiation under the illumination, the intensive illumination is not conducive to the result because it may destroy the Cs/O bonding with GaAs. The degradation shape of sample 1 shows that the photoemission performance of GaAs photocathode will degrade rapidly after recesiations under the illumination of white light, which is distinctly different from that of samples 2–4. Just like the degradation case after Cs/O activation, the intense white light illumination after recesiation can re-optimize the arrangement of surface dipoles. The photocurrent values of the samples recesiated using the monochromatic light decay gently at the beginning, which can help keep high photoemission capability for a relatively longer time. For the monochromatic light illumination with the same number of incident photons, the red monochromatic light with 633 nm can obtain a longer lifetime after each recesiation. Besides, in view of the higher recovered photocurrent after each recesiation, we believe that the recesiations under illumination of the red monochromatic light should bring the higher quantum efficiency in the longwave region, especially near the bandgap threshold, which is important for the practical application of polarized electron sources.
Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant nos. 61771245, 61301023, 61661002 and 61701220), and Science and Technology on Low-Light-Level Night Vision Laboratory Foundation of China (Grant no. J20150702). References [1] N. Chanlek, J.D. Herbert, R.M. Jones, L.B. Jones, K.J. Middleman, B.L. Militsyn, High stability of negative electron affinity gallium arsenide photocathodes activated with Cs and NF3, J. Phys. D 48 (2015) 375102. [2] S. Karkare, L. Boulet, L. Cultrera, B. Dunham, X. Liu, W. Schaff, I. Bazarov, Ultrabright and ultrafast III–V semiconductor photocathodes, Phys. Rev. Lett. 112 (2014) 097601. [3] J.J. Zou, X.W. Ge, Y.J. Zhang, W.J. Deng, Z.F. Zhu, W.L. Wang, X.C. Peng, Z.P. Chen, B.K. Chang, Negative electron affinity GaAs wire-array photocathodes, Opt. Express 24 (2016) 4632–4639. [4] K. Chrzanowski, Review of night vision technology, Opto-Electron. Rev. 21 (2013) 153–181. [5] J.W. Schwede, T. Sarmiento, V.K. Narasimhan, S.J. Rosenthal, D.C. Riley, F. Schmitt, I. Bargatin, K. Sahasrabuddhe, R.T. Howe, J.S. Harris, N.A. Melosh, Z.X. Shen, Photon-enhanced thermionic emission from heterostructures with low interface recombination, Nat. Commun. 4 (2013) 1576. [6] X. Jin, S. Ohki, T. Ishikawa, A. Tackeuchi, Y. Honda, Analysis of quantum efficiency improvement in spin-polarized photocathode, J. Appl. Phys. 120 (2016) 164501. [7] K. Mitsuno, T. Masuzawa, Y. Hatanaka, Y. Neo, H. Mimura, Activation process of GaAs NEA photocathode and its spectral sensitivity, 3rd International Conference on Nanotechnologies and Biomedical Engineering, Springer, 2016, pp. 163–166. [8] M. Kuwahara, S. Kusunoki, Y. Nambo, K. Saitoh, X. Jin, T. Ujihara, H. Asano, Y. Takeda, N. Tanaka, Coherence of a spin-polarized electron beam emitted from a semiconductor photocathode in a transmission electron microscope, Appl. Phys. Lett. 105 (2014) 193101. [9] X. Jin, S. Matsuba, Y. Honda, T. Miyajima, M. Yamamoto, T. Utiyama, Y. Takeda, Picosecond electron bunches from GaAs/GaAsP strained superlattice photocathode, Ultramicroscopy 130 (2013) 44–48. [10] M. Kuwahara, Y. Takeda, K. Saitoh, T. Ujihara, H. Asano, T. Nakanishi, N. Tanaka, Development of spin-polarized transmission electron microscope, J. Phys. 298 (2011) 012016. [11] J.K. Bae, L. Cultreta, P. Digiacomo, I. Bazarov, Rugged spin-polarized electron sources based on negative electron affinity GaAs photocathode with robust Cs2Te coating, Appl. Phys. Lett. 112 (2018) 154101. [12] T. Nishitani, M. Tabuchi, K. Motoki, T. Takashima, A. Era, Y. Takeda, Superlattice photocathode for high brightness and long NEA-surface lifetime, J. Phys. 298
4. Conclusions In summary, the GaAs photocathode samples were activated and recesiated under illumination of the white light, and the blue(460 nm), green(532 nm), and red(633 nm) monochromatic light, respectively. When the GaAs photocathode was activated using the monochromatic light, the relatively higher quantum efficiency, especially in the long wavelength region was obtained. Moreover, the slower degradation feature with the longer lifetime after activation and recesiations was presented, which is quite different from the case using the white light. It clearly shows that GaAs photocathode activated under illumination of the red monochromatic light can yield higher quantum efficiency and obtain better stability than those activated under other illumination conditions, which verifies an improved activation technique for preparing high performance GaAs-based photocathodes. In future work, 131
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
[21] Y. Sun, R.E. Kirby, T. Maruyama, G.A. Mulhollan, J.C. Bierman, P. Pianetta, The surface activation layer of GaAs negative electron affinity photocathode activated by Cs, Li, and NF3, Appl. Phys. Lett. 95 (2009) 174109. [22] Z. Liu, Y. Sun, S. Peterson, P. Pianetta, Photoemission study of Cs-NF3 activated GaAs(100) negative electron affinity photocathodes, Appl. Phys. Lett. 92 (2008) 241107. [23] T. Guo, The adsorption of Cs and O2 on a clean GaAs(110) surface under light illumination, J. Vac. Sci. Technol. A 7 (1989) 1563–1567. [24] I.V. Bazarov, B.M. Dunham, Y. Li, X. Liu, D.G. Ouzounov, C.K. Sinclair, F. Hannon, T. Miyajima, Thermal emittance and response time measurements of negative electron affinity photocathodes, J. Appl. Phys. 103 (2008) 054901. [25] S. Zhang, S.V. Benson, C. H-Garcia, Observation and measurement of temperature rise and distribution on GaAs photo-cathode wafer with a 532 nm drive laser and a thermal imaging camera, Nucl. Instrum. Methods A 631 (2011) 22–25. [26] N. Chanlek, J.D. Herbert, R.M. Jones, L.B. Jones, K.J. Middleman, B.L. Militsyn, High stability of negative electron affinity gallium arsenide photocathodes activated with Cs and NF3, J. Phys. D 47 (2014) 055110. [27] Y.J. Zhang, B.K. Chang, J. Niu, J. Zhao, J.J. Zou, F. Shi, H.C. Cheng, High-efficiency graded band-gap AlxGa1-xAs/GaAs photocathodes grown by metalorganic chemical vapor deposition, Appl. Phys. Lett. 99 (2011) 101104. [28] N. Chanlek, R.M. Jones, J.D. Herbert, L.B. Jones, K.J. Middleman, B.L. Militsyn, Gallium arsenide preparation and QE lifetime studies using the ALICE photocathode preparation facility, Proc. of IPAC’10, 2010 TUPEC018. [29] N.B. Clayburn, K.W. Trantham, M. Dunn, T.J. Gay, Method for monitoring GaAs photocathode heat cleaning temperature, Rev. Sci. Instrum. 87 (2016) 124903.
(2011) 012010. [13] Y.J. Zhang, Y.S. Qian, C. Feng, F. Shi, H.C. Cheng, J.J. Zou, J.Z. Zhang, X. Zhang, Improved activation technique for preparing high-efficiency GaAs photocathodes, Opt. Mater. Express 7 (2017) 3456–3465. [14] C. Feng, Y. Zhang, Y. Qian, B. Chang, F. Shi, G. Jiao, J. Zou, Photoemission from advanced heterostructured AlxGa1-xAs/GaAs photocathodes under multilevel builtin electric field, Opt. Express 23 (2015) 19478–19488. [15] W. Liu, Y. Chen, W. Lu, A. Moy, M. Poelker, M. Stutzman, S. Zhang, Record-level quantum efficiency from a high polarization strained GaAs GaAsP superlattice photocathode with distributed Bragg reflector, Appl. Phys. Lett. 109 (2016) 252104. [16] X. Jin, B. Ozdol, M. Yamamoto, A. Mano, N. Yamamoto, Y. Takeda, Effect of crystal quality on performance of spin-polarized photocathode, Appl. Phys. Lett. 105 (2014) 203509. [17] X. Jin, S. Ohki, T. Ishikawa, A. Tackeuchi, Y. Honda, Analysis of quantum efficiency improvement in spin-polarized photocathode, J. Appl. Phys. 120 (2016) 164501. [18] Z. Liu, Y. Sun, F. Machuca, P. Pianetta, W.E. Spicer, R.F.W. Pease, Optimization and characterization of III–V surface cleaning, J. Vac. Sci. Technol. B 21 (2003) 1953–1958. [19] D.A. Orlov, C. Krantz, A. Wolf, A.S. Jaroshevich, S.N. Kosolobov, H.E. Scheibler, A.S. Terekhov, Long term operation of high quantum efficiency GaAs (Cs,O) photocathodes using multiple recleaning by atomic hydrogen, J. Appl. Phys. 106 (2009) 054907. [20] X. Jin, M. Suzuki, T. Yasue, T. Koshikawa, Y. Takeda, Recovery of quantum efficiency in spin-polarized photocathodes by atomic hydrogen cleaning, Ultramicroscopy 183 (2017) 89–93.
132
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Improved quantum efficiency and stability of GaAs photocathode using favorable illumination during activation
T
⁎
Cheng Fenga,c, Yijun Zhanga, , Yunsheng Qiana, Jian Liua, Jingzhi Zhanga, Feng Shib, Xiaofeng Baib, Jijun Zoud a
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China Science and Technology on Low-Light-Level Night Vision Laboratory, Xi'an 710065, China c School of Information and Communication Engineering, Nanjing Institute of Technology, Nanjing 211167, China d Engineering Research Center of New Energy Technology of Jiangxi Province, East China University of Technology, Nanchang 330013, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: GaAs photocathode Activation Illumination condition Photoemission performance
Considering that illumination using the light source is required to monitor the activation progress of negativeelectron-affinity GaAs-based photocathodes, it is important to understand if and how the illumination affects the activation. To improve the photoemission performance of GaAs photocathode, epitaxial GaAs samples are Cs/O2 activated and recesiated under illumination of monochromatic light of blue (460 nm), green (532 nm), and red (633 nm) with approximately the same incident photons, and halogen tungsten lamp as white light source, respectively, to induce photoemission. The performance characteristics including quantum efficiency and photocurrent degradation among the samples treated by different illumination conditions are compared to investigate their photoemission capability and stability. The results show that GaAs photocathodes activated and recesiated under illumination of monochromatic red light, can obtain higher quantum efficiency, and better stability. This work provides a favorable illumination condition during activation for obtaining satisfactory capability of GaAs-based photocathodes, which will be propitious for applications as spin-polarized electron sources.
1. Introduction Photoemission from III-V group GaAs-based negative-electron-affinity (NEA) photocathodes has shown numerous applications in important physic fields due to the advantages in high quantum efficiency, low mean transverse energy, high spin polarization, and good longwavelength response [1-3]. It has already been put to practical use in vacuum photodetectors, solar cells, polarized electron sources [4-7]. In recent years, such photocathodes have shown good application prospect as electron sources in low-energy electron microscopes and transmission electron microscopy, which requires high spin polarization and high brightness to obtain excellent image contrast and fast image acquisition [8-10]. The spin-polarized electron beam is generally produced by NEA GaAs-based photocathodes, while the NEA condition of the photocathode surface, a prerequisite for achieving high quantum efficiency, can be usually realized by Cs/O2 (or Cs/NF3) activation. The quantum efficiency and lifetime are important performance characteristics of GaAs-based photocathodes, which mainly depend on the material
⁎
structures and the technique during preparation [11-13]. An ideal photocathode should have a high quantum efficiency and a long operational lifetime, moreover, it should have the potential to withstand an adverse environment, such as the influence of residual gas and desorbed molecules. As everyone knows, NEA GaAs-based photocathodes are sensitive to contaminants in the vacuum environment, and the inferiority of short lifetime results in repetition of NEA preparation. Extensive efforts have been done to further promote the quantum efficiency and increase the operational lifetime, such as designing graded bandgap, strained, and superlattice structures to optimize cathode structures [14-17], finding more efficacious methods to remove the adsorbed contaminants in surface cleaning procedures [18-20], using more ideal solid O dispensers to release oxygen effectively during Cs/O activation process [13], and enhancing the immunity of photocathodes by introducing different electronegative substances [1,21,22]. As is well known, illumination using the light source is required during activation to monitor the activation process and adjust the operating current passing through the Cs and O dispensers [13,23]. White light and monochromatic laser light such as argon ion laser, He-Ne
Corresponding author. E-mail address: [email protected] (Y. Zhang).
https://doi.org/10.1016/j.ultramic.2019.04.010 Received 11 May 2018; Received in revised form 18 March 2019; Accepted 22 April 2019 Available online 23 April 2019 0304-3991/ © 2019 Elsevier B.V. All rights reserved.
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
laser, and 532 nm laser have been utilized for the photoelectron excitation from photocathodes [13,22-26]. These illumination conditions during the activation and recesiation process have not been comparatively analyzed, and how they affect the photoemission performance needs to be investigated. In this paper, the monochromatic light with blue (460 nm), green (532 nm), and red (633 nm) wavelength and the white light are respectively illuminated to GaAs photocathodes during the activation and recesiation process. We focus on the effect of different illumination conditions on quantum efficiency and emission stability of the GaAs photocathode samples. This work may further optimize the activation technique and enhance the photoemission performance of NEA GaAs-based photocathodes. 2. Experimental Fig. 2. Photocurrent change curves during Cs/O activation for the four GaAs cathode samples illuminated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
A kind of reflection-mode GaAs photocathode epitaxial wafer grown by the metalorganic chemical vapor deposition technique was chosen in our experiments, and four 11×11 mm2 square GaAs samples were cleaved from the same wafer to carry out the research on the activation and recesiation process under different illumination conditions. The GaAs epilayers consisting of a GaAs emission layer and an AlxGa1−xAs buffer layer are grown on the n-type (100)-oriented GaAs substrate, where the 1 µm-thick GaAs emission layers is quasi-exponentially doped with the zinc concentration ranging from 1 × 1019 cm−3 to 1 × 1018 cm−3, and the 0.5 µm-thick AlxGa11−xAs buffer layer with the doping concentration of 1 × 1019 cm−3 is of a graded Al proportion varying from 0.9 to 0. This graded bandgap structure was reported to have more advantages in quantum efficiency than the traditional structure [27]. Before activation, the GaAs cathode samples were chemically etched by the HF (40%) solution for 5 min right after being degreased. Then the samples were mounted on the stainless holder and transferred into the ultra-high vacuum (UHV) chamber as shown in Fig. 1. In the UHV chamber with the base pressure of about 1 × 10−7 Pa, the samples underwent a 650 ℃ heat-clean treatment for 20 min to obtain an atomically clean surface [28,29], and the Cs/O activation process was began to form NEA surface after the samples were cooled down to room temperature. During the activation process, for the four GaAs samples numbered by sample 1 to sample 4, four different illumination conditions corresponding to the samples were respectively adopted, i.e. (a) a halogen tungsten lamp with an intensity of 100lx as the white light source, (b) a blue (460 nm) monochromatic light with the power of 40.0 μW, (c) a green (532 nm) monochromatic light with the power of 34.6 μW, (d) a red (633 nm) monochromatic light with the power of 29.1 μW. The monochromatic light with the different wavelengths was generated by the monochromator and the light power was different to ensure their incident photons are approximately the same. As shown in Fig. 1, the incident light irradiated on the masked photocathode sample surface with a diameter of 10 mm was introduced by the optical fiber into the vacuum chamber. The activation process for the GaAs samples was carried out with the improved Cs/O co-deposition technique based on the solid Cs and O dispensers [13]. When there was no significant
increase between the present photocurrent peak and the previous one, the activation process was terminated with the Cs flux. The quantum efficiency was measured immediately followed by the measurement of photocurrent degradation, during which all samples were measured under illumination of the same halogen tungsten lamp with intensity of 100lx. After the first degradation, four recesiations for each sample were implemented by feeding fresh cesium flux to investigate the photocurrent recovery and degradation. During the recesiation process, the illumination condition was still different for each sample just as that in the initial Cs/O activation process, while in the multiple degradation processes, the samples were still illuminated with the same intensive white light of 100lx. 3. Results and discussion 3.1. Cs/O2 activation In the continuous activation process, the GaAs sample surface was exposed to alternate doses of Cs and O to increase the photocurrent peak, and the entire change of photocurrent with time was recorded online by the computer, as shown in Fig. 2. The photocurrent change curves show that sample 1 has much higher photocurrent peaks and shorter period compared with other samples, which is due to the high intensity of halogen tungsten lamp. For samples 2–4, they need more time to arrive at their first photocurrent maximum than sample 1 in the stage of only Cs adsorption. When only Cs flux is applied at the beginning of the activation, the photocurrent of the sample under illumination of 460 nm monochromatic light takes off earlier than those under illumination of 532 nm and 633 nm monochromatic light. More interestingly, the sample under illumination of 633 nm monochromatic light needs more time. This phenomenon could mean that for the activation using the monochromatic light, the Cs adsorption on the GaAs surface under illumination of blue photons during the initial activation stage will get easier. In other words, it turns out that electrons can be more easily excited to the conduction-band minimum by shortwave illumination light with higher photon energy, which is the main determinant to shorten the time of reaching the first Cs saturation. During the subsequent Cs/O co-deposition process, under illumination of 633 nm monochromatic light, the Cs/O alternation frequency is obviously higher, and more Cs/O alternation cycles can be achieved. Thus it is inferred that although the activation using the 633 nm monochromatic light takes a longer time, it is more conducive to the deposition of Cs and O on the GaAs surface, which can be favorable for the full formation of surface dipoles. The quantum efficiency curves of these GaAs samples were measured in situ right after the activation, as shown in Fig. 3. The quantum efficiency of the NEA GaAs cathode samples keeps at a high level when
Fig. 1. Schematic diagram of the activation equipment for preparing NEA GaAs photocathodes. 129
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
Fig. 3. Measured quantum efficiency curves after Cs/O activation for the four GaAs cathode samples illuminated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
Fig. 4. First photocurrent degradation curves under 100lx white light illumination after Cs/O activation for the four GaAs cathode samples activated with white light, 460 nm monochromatic light, 532 nm monochromatic light, and 633 nm monochromatic light, respectively.
the photon energy exceeds the GaAs bandgap. By comparison of the quantum efficiency curves of these samples, it is found that the GaAs photocathode activated under white light illumination using the halogen tungsten lamp has a passable quantum efficiency in the shortwave region, however the quantum efficiency drops a lot with the increase in wavelength. Unlike the sample activated under the white light illumination, the samples activated under illumination of monochromatic light possess a relatively flatter quantum efficiency curve over the whole wavelength region, except for the sample activated using the blue monochromatic light. The experimental results turn out that the GaAs photocathode activated under illumination of the 633 nm monochromatic light exhibits more excellent photoemission performance, especially in the long-wavelength region. The reason for the improved quantum efficiency ascribed to the Cs/O activation using the 633 nm monochromatic light can be complex and microscopic. It is considered that, on one hand, the height and width of the surface barrier related to the surface dipoles are changed under different illumination conditions, which affect the surface escape probability for the photoelectrons at different wavelengths. It might be more beneficial to reduce the barrier height and width by illumination with the red monochromatic light during activation, and the emission probability of the low energy electrons can be enhanced. On the other hand, the proper illumination during activation is an important factor to increase the probability of Cs and O adsorption, and more intense illumination will accelerate the activation process because the surface diffusion and migration and the bonding states transition are aggravated [23]. Intensive light will result in a decrease of photoemission performance because it may destroy the bonding of Cs-GaAs dipoles and Cs-O diploes. Meanwhile, the intensive illumination may promote the adsorption of unnecessary oxygen-containing residual gases. Accordingly, we believe that in the case of the monochromatic light accompanied by the same incident photons, the red monochromatic light illumination during activation can be conducive to the higher quantum efficiency in the entire wavelength region.
tungsten lamp irradiates the activated GaAs cathode samples, the photocurrent of sample 1 instantly degrades. By contrast, the photocurrent curves of samples 2, 3, and 4 rise at the beginning, which could be related to the rearrangement of surface dipoles under the intense illumination. During the initial Cs/O activation experiments, the activation processes for these samples were terminated by the excess Cs flux. For the samples activated using the monochromatic light, the higher light intensity after activation will lead to additional absorption of residual cesium in the vacuum, cause reconfiguration of surface dipoles, and promote the photoelectric conversion until Cs resaturation and new equilibrium under the new illumination condition. After that, the cathode performance begins to deteriorate due to the destruction of the optimum arrangement of surface dipoles. The residual gases in the chamber were mainly H2, H2O, CO and CO2 detected by the quadrupole mass spectrometer, wherein the partial pressures of H2, H2O, CO and CO2 were in the magnitude of 10-7 Pa, 10-8 Pa, 10-8 Pa and 10-9 Pa, respectively. The intense white light involving more high energy photons will contribute energy to more cracking of oxygen-containing residual gases and promote the adsorption of unnecessary residual gases, which is adverse to the formation of surface dipoles and degrade the photoemission. Thus the photocurrent of the GaAs sample activated with the white light exhibits the feature of faster degradation speed. In addition, the saturated photocurrent values show that the GaAs photocathode activated under illumination of the 633 nm monochromatic light can achieve higher photocurrent peak in contrast to those activated under other illumination conditions. In general, the capability of GaAs samples can be recovered by feeding the fresh Cs flux, and the illumination conditions may have effect on the recovery of cathode performance. Illumination of appropriate light is favorable, which can help delay the adsorption of residual gases, and at the same time, not hinder the adsorption of Cs. Accordingly, four recesiations were implemented after the first degradation of each sample, and the illumination conditions during the recesiations are the same as the corresponding ones during Cs/O activations. The lifetime after each regeneration was measured, as shown in Fig. 5. It is found from Fig. 5 that although the introduced the fresh cesium can help recover the capability, it can hardly eliminate the influence of the substrate oxides formed during degradation, and the complementary cesium can't form robust dipoles with oxygen atoms or Ga-As atoms. Therefore, there is an obvious and fast drop in photocurrent of the regenerated cathode. During the next several recesiations, the recovery of their initial photocurrent values become increasingly worse. It turns out that the lifetimes after the first recesiation decrease to less than half their original levels after the Cs/O activation
3.2. Degradations after activation and recesiations The illumination conditions during activation can influence the Cs/ O activation process and the final quantum efficiency, furthermore, the stability of the GaAs photocathodes activated under the different illumination conditions during activation can also be affected. The photocurrent degradation curves after activation were measured under the white light illumination with intensity of 100lx, which are presented in Fig. 4. It is obviously seen from Fig. 4 that the processes of photocurrent degradation among the four samples are different. When the halogen 130
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
Fig. 5. Photocurrent degradation curves under 100lx white light illumination after four recesiations for (a) sample 1 recesiated with white light, (b) sample 2 recesiated with 460 nm monochromatic light, (c) sample 3 recesiated with 532 nm monochromatic light, and (d) sample 4 recesiated with 633 nm monochromatic light.
whether the Cs/O activation under illumination of the broadband red light via band-pass filter or other monochrome red light has the same positive effect requires further exploration.
for these samples. The photocurrent degradation curve shows that the decline of initial photocurrent and operational lifetime of sample 1 activated using the white light are particularly obvious after each recesiation, while the samples activated using the monochromatic light are relative stable. Despite the photoemission capability can be recovered by recesiation under the illumination, the intensive illumination is not conducive to the result because it may destroy the Cs/O bonding with GaAs. The degradation shape of sample 1 shows that the photoemission performance of GaAs photocathode will degrade rapidly after recesiations under the illumination of white light, which is distinctly different from that of samples 2–4. Just like the degradation case after Cs/O activation, the intense white light illumination after recesiation can re-optimize the arrangement of surface dipoles. The photocurrent values of the samples recesiated using the monochromatic light decay gently at the beginning, which can help keep high photoemission capability for a relatively longer time. For the monochromatic light illumination with the same number of incident photons, the red monochromatic light with 633 nm can obtain a longer lifetime after each recesiation. Besides, in view of the higher recovered photocurrent after each recesiation, we believe that the recesiations under illumination of the red monochromatic light should bring the higher quantum efficiency in the longwave region, especially near the bandgap threshold, which is important for the practical application of polarized electron sources.
Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant nos. 61771245, 61301023, 61661002 and 61701220), and Science and Technology on Low-Light-Level Night Vision Laboratory Foundation of China (Grant no. J20150702). References [1] N. Chanlek, J.D. Herbert, R.M. Jones, L.B. Jones, K.J. Middleman, B.L. Militsyn, High stability of negative electron affinity gallium arsenide photocathodes activated with Cs and NF3, J. Phys. D 48 (2015) 375102. [2] S. Karkare, L. Boulet, L. Cultrera, B. Dunham, X. Liu, W. Schaff, I. Bazarov, Ultrabright and ultrafast III–V semiconductor photocathodes, Phys. Rev. Lett. 112 (2014) 097601. [3] J.J. Zou, X.W. Ge, Y.J. Zhang, W.J. Deng, Z.F. Zhu, W.L. Wang, X.C. Peng, Z.P. Chen, B.K. Chang, Negative electron affinity GaAs wire-array photocathodes, Opt. Express 24 (2016) 4632–4639. [4] K. Chrzanowski, Review of night vision technology, Opto-Electron. Rev. 21 (2013) 153–181. [5] J.W. Schwede, T. Sarmiento, V.K. Narasimhan, S.J. Rosenthal, D.C. Riley, F. Schmitt, I. Bargatin, K. Sahasrabuddhe, R.T. Howe, J.S. Harris, N.A. Melosh, Z.X. Shen, Photon-enhanced thermionic emission from heterostructures with low interface recombination, Nat. Commun. 4 (2013) 1576. [6] X. Jin, S. Ohki, T. Ishikawa, A. Tackeuchi, Y. Honda, Analysis of quantum efficiency improvement in spin-polarized photocathode, J. Appl. Phys. 120 (2016) 164501. [7] K. Mitsuno, T. Masuzawa, Y. Hatanaka, Y. Neo, H. Mimura, Activation process of GaAs NEA photocathode and its spectral sensitivity, 3rd International Conference on Nanotechnologies and Biomedical Engineering, Springer, 2016, pp. 163–166. [8] M. Kuwahara, S. Kusunoki, Y. Nambo, K. Saitoh, X. Jin, T. Ujihara, H. Asano, Y. Takeda, N. Tanaka, Coherence of a spin-polarized electron beam emitted from a semiconductor photocathode in a transmission electron microscope, Appl. Phys. Lett. 105 (2014) 193101. [9] X. Jin, S. Matsuba, Y. Honda, T. Miyajima, M. Yamamoto, T. Utiyama, Y. Takeda, Picosecond electron bunches from GaAs/GaAsP strained superlattice photocathode, Ultramicroscopy 130 (2013) 44–48. [10] M. Kuwahara, Y. Takeda, K. Saitoh, T. Ujihara, H. Asano, T. Nakanishi, N. Tanaka, Development of spin-polarized transmission electron microscope, J. Phys. 298 (2011) 012016. [11] J.K. Bae, L. Cultreta, P. Digiacomo, I. Bazarov, Rugged spin-polarized electron sources based on negative electron affinity GaAs photocathode with robust Cs2Te coating, Appl. Phys. Lett. 112 (2018) 154101. [12] T. Nishitani, M. Tabuchi, K. Motoki, T. Takashima, A. Era, Y. Takeda, Superlattice photocathode for high brightness and long NEA-surface lifetime, J. Phys. 298
4. Conclusions In summary, the GaAs photocathode samples were activated and recesiated under illumination of the white light, and the blue(460 nm), green(532 nm), and red(633 nm) monochromatic light, respectively. When the GaAs photocathode was activated using the monochromatic light, the relatively higher quantum efficiency, especially in the long wavelength region was obtained. Moreover, the slower degradation feature with the longer lifetime after activation and recesiations was presented, which is quite different from the case using the white light. It clearly shows that GaAs photocathode activated under illumination of the red monochromatic light can yield higher quantum efficiency and obtain better stability than those activated under other illumination conditions, which verifies an improved activation technique for preparing high performance GaAs-based photocathodes. In future work, 131
Ultramicroscopy 202 (2019) 128–132
C. Feng, et al.
[21] Y. Sun, R.E. Kirby, T. Maruyama, G.A. Mulhollan, J.C. Bierman, P. Pianetta, The surface activation layer of GaAs negative electron affinity photocathode activated by Cs, Li, and NF3, Appl. Phys. Lett. 95 (2009) 174109. [22] Z. Liu, Y. Sun, S. Peterson, P. Pianetta, Photoemission study of Cs-NF3 activated GaAs(100) negative electron affinity photocathodes, Appl. Phys. Lett. 92 (2008) 241107. [23] T. Guo, The adsorption of Cs and O2 on a clean GaAs(110) surface under light illumination, J. Vac. Sci. Technol. A 7 (1989) 1563–1567. [24] I.V. Bazarov, B.M. Dunham, Y. Li, X. Liu, D.G. Ouzounov, C.K. Sinclair, F. Hannon, T. Miyajima, Thermal emittance and response time measurements of negative electron affinity photocathodes, J. Appl. Phys. 103 (2008) 054901. [25] S. Zhang, S.V. Benson, C. H-Garcia, Observation and measurement of temperature rise and distribution on GaAs photo-cathode wafer with a 532 nm drive laser and a thermal imaging camera, Nucl. Instrum. Methods A 631 (2011) 22–25. [26] N. Chanlek, J.D. Herbert, R.M. Jones, L.B. Jones, K.J. Middleman, B.L. Militsyn, High stability of negative electron affinity gallium arsenide photocathodes activated with Cs and NF3, J. Phys. D 47 (2014) 055110. [27] Y.J. Zhang, B.K. Chang, J. Niu, J. Zhao, J.J. Zou, F. Shi, H.C. Cheng, High-efficiency graded band-gap AlxGa1-xAs/GaAs photocathodes grown by metalorganic chemical vapor deposition, Appl. Phys. Lett. 99 (2011) 101104. [28] N. Chanlek, R.M. Jones, J.D. Herbert, L.B. Jones, K.J. Middleman, B.L. Militsyn, Gallium arsenide preparation and QE lifetime studies using the ALICE photocathode preparation facility, Proc. of IPAC’10, 2010 TUPEC018. [29] N.B. Clayburn, K.W. Trantham, M. Dunn, T.J. Gay, Method for monitoring GaAs photocathode heat cleaning temperature, Rev. Sci. Instrum. 87 (2016) 124903.
(2011) 012010. [13] Y.J. Zhang, Y.S. Qian, C. Feng, F. Shi, H.C. Cheng, J.J. Zou, J.Z. Zhang, X. Zhang, Improved activation technique for preparing high-efficiency GaAs photocathodes, Opt. Mater. Express 7 (2017) 3456–3465. [14] C. Feng, Y. Zhang, Y. Qian, B. Chang, F. Shi, G. Jiao, J. Zou, Photoemission from advanced heterostructured AlxGa1-xAs/GaAs photocathodes under multilevel builtin electric field, Opt. Express 23 (2015) 19478–19488. [15] W. Liu, Y. Chen, W. Lu, A. Moy, M. Poelker, M. Stutzman, S. Zhang, Record-level quantum efficiency from a high polarization strained GaAs GaAsP superlattice photocathode with distributed Bragg reflector, Appl. Phys. Lett. 109 (2016) 252104. [16] X. Jin, B. Ozdol, M. Yamamoto, A. Mano, N. Yamamoto, Y. Takeda, Effect of crystal quality on performance of spin-polarized photocathode, Appl. Phys. Lett. 105 (2014) 203509. [17] X. Jin, S. Ohki, T. Ishikawa, A. Tackeuchi, Y. Honda, Analysis of quantum efficiency improvement in spin-polarized photocathode, J. Appl. Phys. 120 (2016) 164501. [18] Z. Liu, Y. Sun, F. Machuca, P. Pianetta, W.E. Spicer, R.F.W. Pease, Optimization and characterization of III–V surface cleaning, J. Vac. Sci. Technol. B 21 (2003) 1953–1958. [19] D.A. Orlov, C. Krantz, A. Wolf, A.S. Jaroshevich, S.N. Kosolobov, H.E. Scheibler, A.S. Terekhov, Long term operation of high quantum efficiency GaAs (Cs,O) photocathodes using multiple recleaning by atomic hydrogen, J. Appl. Phys. 106 (2009) 054907. [20] X. Jin, M. Suzuki, T. Yasue, T. Koshikawa, Y. Takeda, Recovery of quantum efficiency in spin-polarized photocathodes by atomic hydrogen cleaning, Ultramicroscopy 183 (2017) 89–93.
132