GaAs cathode activation with Cs-K-Sb thin film

GaAs cathodes are unique devices which generate a spin-polarized electron beam by the photoelectric effect when illuminated with a circularly polarized laser. Thin-film Negative Electron Affinity (NEA) surfaces have an essential role in spin polarized beam production, but they have limited lifetimes. In this study, we activate GaAs as an NEA cathode by evaporating Cs, K, and Sb metal on its cleaned surface. The experimental results of quantum efficiency measurements taken after evaporative deposition of the multi-alkali surface are presented here.


Introduction
Electron beam performance in linear accelerators is determined by the quality of the initial beam and the amount of current provided by the electron source.Therefore, optimizing the photocathode for low emittance, short pulse duration, high spin polarization and long lifetime is critical to the development of advanced linear accelerators such as the planned ILC [1] and FCC [2] colliders, X-ray Free Electron Lasers [3], Energy Recovery Linacs [4] and other future accelerator projects.
Quantum Efficiency (QE) and robustness are consistent practical issues with photocathodes.GaAs-based Negative Electron Affinity (NEA) cathodes can demonstrate QE greater than 10%, but with markedly short lifetimes caused by back-bombardment of residual gas molecules, causing performance degradation and necessitating frequent replacement; in contrast, metal cathodes, e.g.Cu, Mg, and Pb, have much longer lifetimes at the expense of QE and requiring UV light for activation, and thus are unsuitable for production of a high-current beam.
Multi-alkali photocathodes have recently shown promise in generating high-brightness beams with improved lifetimes [5] and high QE [6], making them enticing candidates for electron generation in future accelerators.Such cathodes are produced by evaporative deposition of a mixture of alkali metals on a GaAs substrate.More recent work has shown improvements with Cs-Sb-O [7,8], Cs-Te [9,10,11], Cs-O-Te [12], and other alkali metals [13].Furthermore, such cathodes can also produce highly spin-polarized electron beams by using a circularly polarized laser.As future colliders plan to make use of spin-polarized beams in their physics programs, this is a desirable characteristic.
Although the NEA mechanism is not well-understood, the current understanding is that the thin film facilitates electron acceleration out of the cathode by reducing the effective br of the conduction band of the cathode, with the difference being imparted to freed electrons as kinetic energy.A graphic representation is shown in Fig. 1.
At Hiroshima University, we are investigating techniques to improve high-performance multialkali cathodes and understand their properties.Here we report on the use of NEA activation of a GaAs cathode using a thin CsKSb film.

Experimental Setup
Cathode preparation is carried out at Hiroshima University, utilizing a vacuum chamber for production and a laser system using a tunable Xe lamp and optical system for illuminating the cathode.

Test Chamber
The test chamber is made of electrically polished stainless steel (SUS) kept at a vacuum of 10 −8 − 10 −9 Pa by means of a NEG pump and an ion pump.The cathode substrate is affixed to a SUS 304 base attached to a ceramic heater for cleaning and temperature control.The holder assembly is electrically isolated and given a DC voltage bias to measure the photoelectric current during testing.
Thin-film materials (Cs, K, Sb) are inserted on an evaporative head designed to release vapor symmetrically in the test chamber for even coating on both the substrate and a quartz thickness Layout of the vacuum test chamber.
monitor.Two viewports allow for illumination of the cathode.The chamber setup is shown in Fig. 2 2.2.Optical System The optical system for illuminating the cathode and inducing photoelectric current consists of a Xe lamp followed by a spectrometer for wavelength selection and a series of lenses to focus the light on the cathode face.For wavelengths greater than 490 nm a filter is applied to remove unwanted higher-energy light.Figure .3shows the optical setup used to select desired wavelengths and focus light on the cathode face.The spectrometer is controlled and monitored with a Labview-based setup, which is also used to calculate and record the cathode QE.

Procedure
After introducing the cathode and thin-film materials into the test chamber, the chamber was brought down to vacuum and baked at high temperature to speed up outgassing.The NEG pump and ion pump were activated during this process and left on for the remainder of the experiment to maintain vacuum.Current was applied to the evaporative head terminals in succession, causing each metal to be individually deposited on the cathode and thin-film sensor faces.The recipe for deposition was planned to be 50 Å of Sb, followed by alternating 5 Å layers of Cs and K to a total thickness of 200 Å.During the experiment, however, evaporation occurred significantly more rapidly than expected and the initial base layer of Sb was thicker than anticipated.
After deposition, the cathode face was illuminated with the Xe lamp while biased at 200 V and monitored for photoelectric current.Using the spectrometer, the illumination wavelength was varied from 350 nm to 910 nm and the QE then calculated according to the formula where I PE is the photoelectric current, and λ and P are the wavelength and power of the incident light, respectively.The spectrum of measured power from the Xe lamp setup at the viewport of the vacuum chamber is shown in Fig. 4.

Results
Following the initial deposition, cathode QE was recorded across the entire effective spectrum of the cathode.QE values recorded across the entire recorded spectrum were consistent with zero, indicating that no NEA activation was observed.These measurements are summarized in Table 1.
As noted earlier, the initial deposit of Sb was thicker than anticipated by approximately a factor of five as a result of faster than expected evaporation of the Sb pellet.In light of the null result following the first test after deposition, we decided to adjust the ratio of materials in the thin film by adding further layers of K and Cs.A second experimental run was carried out after depositing an additional 200 Å each of Cs and K, bringing the film to a total thickness of 850 Å.At lower wavelengths, this configuration showed photoelectric current at shorter wavelengths but was consistent with zero at longer wavelengths, as shown in Table 2.These results indicate that the thin-film itself had been activated but was not effectively facilitating NEA activity.A final test run was carried out, again adding further layers of Cs and K to bring the total combined thickness of the thin film to 1250 Å.Similar to the previous result, shorter wavelengths (here 350 nm) demonstrate a photoelectric signal, but at longer wavelength the signal becomes negligible.Results for selected wavelengths are shown in Table 3.

Conclusion
We attempted to activate a NEA thin-film surface on a GaAs cathode using evaporative deposition at Hiroshima University.After depositing layers of Sb, K and Cs to form a multi-alkali thin film, we measured the photoelectric current induced by a Xe light source with wavelength selected by a spectrometer.We observe a photoelectric current induced by short wavelength light incident on the cathode with a maximum QE of (3.44 ± 0.14) × 10 −2 % at 350 nm.The presence of photoelectric current indicates that the deposition was performed successfully; however, at longer wavelengths the signal becomes consistent with zero for all tested thicknesses of thin-film materials.Based on this result, we conclude that the film is not successfully facilitating the NEA mechanism as described in the introduction and will require further testing to refine the deposition technique.

Figure 1 .
Figure 1.A representation of the NEA mechanism.The thin film (here, Cs-O) has a lower conduction band br than the GaAs bulk, effectively lowering the band gap for the cathode.The energy difference indicated by the red arrow is imparted to freed electrons as kinetic energy.

Figure 2 .
Figure 2.Layout of the vacuum test chamber.

Figure 3 .
Figure 3. Optical setup used for illuminating the test cathode.

Figure 4 .
Figure 4. Power for the spectrum provided by the Xe lamp, measured at the viewport of the vacuum test chamber.

Table 1 .
Quantum Efficiency of the Cathode after Initial Deposition of Sb, Cs, and K to a

Table 2 .
Quantum Efficiency of the Cathode at Selected Wavelengths after Initial Deposition of Sb, Cs, and K to a Total Thickness of 850 Å

Table 3 .
Quantum Efficiency of the Cathode at Selected Wavelengths after Initial Deposition of Sb, Cs, and K to a Total Thickness of 1250 Å