Preliminary Systematic Study of the Temperature Effect on the K--Cs--Sb Photocathode Performance Based on the K and Cs Co-Evaporation

Xu-Dong Li(李旭东)$^{1\hyperlink{s}{**}}$, Zeng-Gong Jiang(姜增公)$^{1}$, Qiang Gu(顾强)$^{1}$,\\ Ming-Hua Zhao(赵明华)$^{2}$, Li Guo(郭力)$^{2}$

doi:10.1088/0256-307X/37/1/012901

⇦Chinese Physics Letters, 2020, Vol. 37, No. 1, Article code 012901⇨ Preliminary Systematic Study of the Temperature Effect on the K–Cs–Sb Photocathode Performance Based on the K and Cs Co-Evaporation * Xu-Dong Li (李旭东)1**, Zeng-Gong Jiang (姜增公)1, Qiang Gu (顾强)1, Ming-Hua Zhao (赵明华)2, Li Guo (郭力)2 Affiliations 1Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210 2Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201800 Received 30 September 2019, online 23 December 2019 ^*Supported by the National Natural Science Foundation of China under Grant No. 11405251, and the Shanghai Municipal Science and Technology Major Project.
^**Corresponding author. Email: lixudong@zjlab.org.cn Citation Text: Li X D, Jiang Z G, Gu Q, Zhao M H and Guo L et al 2020 Chin. Phys. Lett. 37 012901 Abstract It is very important to increase the quantum efficiency (QE) and prolong the lifetime of the photocathode in a variety of applications. We have succeeded in preparing a high QE cesium potassium antimonide (K–Cs–Sb) photocathode by K and Cs co-evaporation in the photocathode preparation facility. In order to better understand the effect of the substrate (photocathode) temperature on the photocathode performance, the photocathode preparation, photocathode performance degradation, photocathode performance recovery and photocathode removal are studied in detail. DOI:10.1088/0256-307X/37/1/012901 PACS:29.25.Bx, 81.05.Hd, 07.30.Kf © 2020 Chinese Physics Society Article Text Photocathodes, which convert incident photons into free electrons, exhibit a low work function and therefore easily emit electrons into a vacuum form the basis of electronic devices for a wide range of applications, including photon detection,[1] electron sources[2] and solar energy conversion,[3] etc. Nowadays, there are some limitations on the photocathode, such as the high quantum efficiency (QE) photocathode with limited lifetime and the long lifetime photocathode with low QE. The main goal of further photocathode development is to prolong the lifetime of high QE photocathodes or to increase the QE of robust photocathodes. The photocathode performance is greatly influenced by the preparation conditions and methods.[4] Here, we focus on a simpler method of modulating the photocathode performance just by adjusting the temperature.[5–8] Due to the high sensitivity of the K–Cs–Sb photocathode to any form of pollution, it must be prepared, transported and operated under ultra-high vacuum conditions (UHV). We have established a semiconductor photocathode preparation facility based on UHV transfer technology. Its functions include loading, cleaning, deposition and transportation to ensure the quality of the prepared K–Cs–Sb photocathode. The facility consists of a substrate loading chamber, a photocathode preparation chamber and the corresponding photocathode transmission system.[9] In the substrate loading chamber it can be subjected to load and clean the substrate by an argon ion gun. The preparation chamber can be used to heat clean the substrate and prepare the photocathode. Our method for preparing a K–Cs–Sb photocathode is as follows: (i) The polished and chemically cleaned Mo substrate is heated to 370$^{\circ}\!$C for six hours to remove surface oxide and absorbed gas such as H$_{2}$, O$_{2}$, N$_{2}$ and H$_{2}$O. (ii) The substrate temperature is stabilized at about 75$^{\circ}\!$C prior to the preparation, and 10-nm Sb is deposited at 0.1 Å/s deposition rate. (iii) The K and Cs co-evaporation is also kept at 75$^{\circ}\!$C, while evaporating K and Cs. The ratio of the evaporation rate of K and Cs is set at about $1\!:\!1$, in order to keep the constant deposition rate. The heating power of the sources should be controlled in time. The preparation is terminated when the photocurrent stops increasing at a typical QE of 4%–6%, the K and Cs sources are turned off, the entire preparation process usually takes 8–12 hours. (iv) After preparation, the photocathode is cooled down to room temperature.

Fig. 1. The change of QE, reflectivity and thickness over time in the process of photocathode preparation.

Figure 1 shows the change of QE, reflectivity ($R$) and thickness over time during photocathode preparation. The reflectivity decreases with the deposition of Sb. With the deposition of K and Cs, the reflectivity increases firstly and then decreases. When the reflectivity drops to the lowest, the QE increases simultaneously with the increase of reflectivity (beginning to form K–Cs–Sb photocathode). After the QE reaches its the maximum value, if K and Cs continue to deposit, the reflectivity will continue to increase, where the QE will decrease. It is shown that there is no new K–Cs–Sb photocathode formed on the surface for the excessive K and Cs at this time. The relationship between the reflectivity and Sb thickness at different temperatures is shown in Fig. 2. The substrate is at room temperature, 48$^{\circ}\!$C, 68$^{\circ}\!$C, 88$^{\circ}\!$C and 108$^{\circ}\!$C. The amorphous-to-crystalline transition for the Sb film is observed at the film thickness of 40 Å–80 Å. After the reflectivity decreases to the lowest level, the reflectivity increases gradually, indicating that the Sb film changes from amorphous to crystalline,[10] and finally becomes the fully crystalline Sb, the reflectivity remains unchanged during the process. The results show that too high or too low a temperature is detrimental to the formation of the crystal structure.

Fig. 2. The relationship between reflectivity and Sb thickness at different substrate temperatures.

Fig. 3. The relationship between QE and K/Cs thickness (amount) in the process of photocathode preparation at different substrate temperatures.

A 5-nm Sb layer is deposited on the Mo substrate, and then a K–Cs–Sb photocathode is prepared by K and Cs co-evaporation at room temperature, 48$^{\circ}\!$C, 68$^{\circ}\!$C, and 88$^{\circ}\!$C. The relationship between QE and K/Cs thickness (deposition amount) is shown in Fig. 3. Although the K–Cs–Sb photocathode has the highest QE at the Mo substrate temperature of 88$^{\circ}\!$C, the K/Cs amount required to achieve the maximum QE is about twice the amount at other substrate temperatures, and the maximum QE is slightly higher than that at 68$^{\circ}\!$C. In future, the QE and the K/Cs amount will be both considered. The suitable substrate temperature can not only improve the photocathode QE but also save the amount of K and Cs sources. In our photocathode preparation facility, the photocathode is maintained at a certain temperature using a halogen lamp. The temperature effect on the photocathode is studied, which has guiding significance for the future photocathode operation.[11] It is shown that the 1/$e$ lifetime of K–Cs–Sb photocathode at different temperatures is about 96 hours (room temperature), 90 hours (37$^{\circ}\!$C) and 68 hours (68$^{\circ}\!$C), respectively. The initial QE is normalized, as shown in Fig. 4. This indicates that the photocathode temperature has a great influence on the photocathode lifetime. The lower the photocathode temperature, the longer the photocathode lifetime.

Fig. 4. The 1/$e$ lifetime of K–Cs–Sb photocathode at different temperatures.

Fig. 5. The reflectivity changes during the 1/$e$ lifetime of photocathode at different temperatures.

It is demonstrated that the prepared photocathode QE increases at the beginning of temperature dropping, which has been attributed to the photocathode forming a new lattice structure with higher QE or the excess K and Cs loss, then the QE decreases for the surface useful Cs and K loss or the photocathode adsorbs more and more pollutants, which destroys the photocathode crystal structure. In photocathode operation, if the surface temperature is too high, the surface will have a chemical change, the reflectivity will decrease, as shown in Fig. 5. Generally, the initial photocathode preparation takes a long time. However, the photocathode performance recovery time is relatively short, and at the same time, the amount of alkali metal source is relatively small. The K–Cs–Sb photocathode QE could be partially recovered by brief thermal annealing at a certain temperature, which desorbs the contaminations and transfers the alkali from the inside to the photocathode surface. The in situ thermal annealing technology prolongs the photocathode lifetime, reducing the time for replacing and preparing a new photocathode, and saving the evaporation source. The QE changes of the degraded K–Cs–Sb photocathode before and after thermal annealing at 88$^{\circ}\!$C, 95$^{\circ}\!$C and 100$^{\circ}\!$C are shown in Fig. 6.

Fig. 6. The change of photocathode QE at different annealing temperatures.

Fig. 7. The QE change of photocathode at different times before and after thermal annealing.

In our experiment, when the degraded photocathode is heated to 88$^{\circ}\!$C for 2 hours to recover its performance, the QE after thermal annealing is 0.4%, which is about 50% of the original value, the photocathode QE via thermal annealing can be partially recovered, but the photocathode QE after thermal annealing decreases faster. During the measurement of the K–Cs–Sb photocathode lifetime, the QE at different times is measured. The results are shown in Fig. 7. Figure 8 shows the change of the photocathode QE map before and after thermal annealing. The degraded K–Cs–Sb photocathode can be removed after heating at 300$^{\circ}\!$C. The center color changes from blue to black. During the removal of the K–Cs–Sb photocathode by heating, the reflectivity of the degraded photocathode changes with the temperature increases, the reflectivity increases firstly, then decreases, and finally increases to the stable value. Even if the temperature drops, the reflectivity will not change, as shown in Fig. 9. The degraded K–Cs–Sb photocathode can be completely removed by heating the photocathode to about 250$^{\circ}\!$C.

Fig. 8. The QE map of the photocathode before and after thermal annealing.

Fig. 9. The relationship between substrate reflectivity and temperature.

To conclude, the experimental results show that the substrate temperature has a great influence on the growth of Sb film. During the photocathode preparation, there is an optimum substrate temperature for the high QE K–Cs–Sb photocathode. The low temperature can prolong the photocathode lifetime. The thermal annealing can partially recover the photocathode performance and prolong the photocathode lifetime. Finally, the K–Cs–Sb photocathode can be removed by heating to 250–300$^{\circ}\!$C. In the long run, the preparation method development is ongoing. More parameters having influence on the photocathode performance will be investigated in our facility. The specific effect of different factors on the lifetime and QE needs further study. The photocathode preparation facility could also be used for the growth/treatment of metallic and multi-alkali photocathodes. References The LHC project: The accelerator and the experimentsPhotocathode behavior during high current running in the Cornell energy recovery linac photoinjectorPhoton-enhanced thermionic emission from p -GaAs with nonequilibrium Cs overlayersDevelopment of Preparation Systems with K ₂ CsSb Photocathodes and Study on the Preparation ProcessTemperature-dependent quantum efficiency degradation of K-Cs-Sb bialkali antimonide photocathodes grown by a triple-element codeposition methodTemperature dependence of alkali-antimonide photocathodes: Evaluation at cryogenic temperaturesPossible hidden-charm molecular baryons composed of an anti-charmed meson and a charmed baryonDirect observation of bi-alkali antimonide photocathodes growth via in operando x-ray diffraction studiesCharge lifetime measurements at high average current using a

K_{2} CsSb

photocathode inside a dc high voltage photogun

[1]	Nakamura K, Hamana Y, Ishigami Y and Matsui T 2010 Nucl. Instrum. Methods Phys. Res. Sect. A 623 1
[2]	Cultrera L, Maxson J and Bazarov I et al 2011 Phys. Rev. Spec. Top. Accel. Beams 14 120101
[3]	Zhuravlev A G, Romanov A S and Alperovich V L 2014 Appl. Phys. Lett. 105 251602
[4]	Zhang F, Li X P and Li X S 2019 Chin. Phys. Lett. 36 022901
[5]	Ding Z H, Karkare S, Feng J, Filippetto D, Johnson M, Virostek S, Sannibale F, Nasiatka J, Gaowei M, Sinsheimer J, Muller E, Smedley J and Padmore H 2017 Phys. Rev. Accel. Beams 20 11
[6]	Kuriki M, Shonaka C, Iijima H, Kubo D, Okamoto H, Higaki H, Ito K, Yamamoto M, Konomi T, Okumi S, Kuwahara M and Nakanishi T 2011 Nucl. Instrum. Methods Phys. Res. Sect. A 637
[7]	Mamun M A, Hernandez-Flores M R, Morales E, Hernandez-Garcia C and Poelker M 2017 Phys. Rev. Accel. Beams 20 103403
[8]	Li X D, Gu Q, Zhang M and Zhao M H 2012 Chin. Phys. C 36 6
[9]	Li X D, Chen L P, Gu Q, Lin G Q and Zhao M H 2012 Nucl. Tech. 35 7 (in Chinese)
[10]	Ruiz-Oses M, Schubert S, Attenkofer K, Ben-Zvi I, Liang X, Muller E, Padmore H, Rao T, Vecchione T, Wong J, Xie J and Smedley J 2014 APL Mater. 2 121101
[11]	Mammei R R, Suleiman R, Feingold J, Adderley P A et al 2013 Phys. Rev. Spec. Top. Accel. Beams 16 033401