Optical and photoemission measurements were performed on alkali antimonide Na2KSb and Na2KSb/CsxSb photocathodes in order to determine their energy-band diagrams, elucidate the photoemission pathways, and explore the options for interface engineering in order to reach high quantum efficiencies of the photocathodes. This study is motivated by the recent discovery of optical orientation in Na2KSb and emission of spin-polarized electrons from Na2KSb/C(x)sSb photocathodes [V.S. Rusetsky et al., Phys. Rev. Lett. 129, 166802 (2022)]. We have shown that the band gap E-g of Na2KSb at =295 K lies within the range of 1.40-1.44 eV. The Na2KSb surface activation by the deposition of Cs and Sb results in effective electron affinity decrease by approximately 0.37 eV, and in an increase of the quantum efficiency up to 0.2 electrons per incident photon. The analysis of longitudinal energy distribution curves (EDCs) proves that the surface of activated Na2KSb/CsxSb photocathodes have negative effective electron affinity of approximately -0.1 and -0.25 eV at =295 and 80 K, respectively. EDC measurements under increasing photon energy demonstrate the transition of photoemission pathway from the surface states' photoionization at (h) over barw < E-g to the emission from the conduction-band bottom at approximate to and from the states with high kinetic energy in the conduction band at <(h)over bar>w > E-g. EDCs measured at 80 K reveal a highly directional photoelectron emission from the Na2KSb/CsSb photocathode, as compared to the p-GaAs(Cs,O) photocathode. This fact, along with the observed significant, by an order of magnitude, increase in the photoluminescence intensity under the Na2KSb surface activation by Cs and Sb, indicates relatively weak diffuse scattering in the "quasiepitaxial" CsxSb activation layer of a Na2KSb/CsSb photocathode, compared to strong scattering in the amorphous (Cs,O) activation layer of a p-GaAs(Cs,O) photocathode.

An analog of the optical polarizer/analyzer for electrons, a spin filter based on freestanding ferromagnetic (FM) nanomembrane covering the entrance of the microchannel plate (MCP) was applied for efficient spin filtering and electron amplification in the 2D field of view. To study the spin dependent transmission, we constructed a spin-triode device (spintron), which consists of a compact proximity focused vacuum tube with the Na2KSb spin-polarized electron source, the FM-MCP and phosphor screen placed to run parallel to each other. Here, we demonstrate the fabrication of FM nanomembranes consisting of a Co/Pt superlattice deposited on a freestanding 3 nm SiO2 layer with a total thickness of 10 nm. The FM-MCP has 10e6 channels with a single-channel Sherman function S=0.6 and a transmission of 1.5x10e-3 in the low electron energy range. The FM-MCP-based device provides a compact optical method for measuring the spin polarization of free electron beams in the imaging mode and is well suited for photoemission spectroscopy and microscopy methods.

In the spectra of the photoemission quantum yield of p-GaAs(Cs,O) measured in the reflection geometry, a peak was found at a photon energy lower than the band gap of GaAs. It is shown that the appearance of the peak is due to the trapping of weakly absorbed radiation, which scatters diffusely on the rough back face of the epitaxial structure. Possible microscopic mechanisms of the appearance of the peak are discussed: the Franz-Keldysh effect in the surface electric field and adsorption-modified optical transitions in cesium adatoms. Keywords: photoemission, GaAs, negative electron affinity, quantum yield spectra.

The minimum achievable particle beam emittance in an electron accelerator depends strongly on the intrinsic emittance of the photocathode electron source. This is measurable as the mean longitudinal and transverse energy spreads in the photoemitted electron beam (MLE and MTE respectively); consequently, MLE and MTE are notable figures of merit for photocathodes used as electron sources in particle accelerators. The overall energy spread is defined by the sum of the MTE and the MLE, and the minimization of MTE is crucial to reduce emittance and thus generate a high-brightness electron beam. Reducing the electron beam emittance in an accelerator that drives a Free-Electron Laser (FEL) delivers a significant reduction in the saturation length for an x-ray FEL, thus reducing the machine's construction footprint and operating costs while increasing the x-ray beam brightness. The ability to measure the transverse energy distribution curve of photoelectrons emitted from a photocathode is a key enabler in photocathode research and development that has prompted the Accelerator Science and Technology Centre (ASTeC) at the STFC Daresbury Laboratory to develop the Transverse Energy Spread Spectrometer to make these crucial measurements. We present details of the design for the upgraded TESS instrument with measured data for copper (100), (110), and (111) single-crystal photocathodes illuminated at UV wavelengths around 266 nm.

Abstract The photoelectron processes in a p-GaN(Cs) photocathode with the effective negative electron affinity were studied experimentally within the 90–295 K temperature range. It was found that the photocathode illumination at the photon energies, which are below the energy gap of the p-GaN layer, increases the band bending at a semiconductor surface due to the photoemission from surface states.

Temperature and optical power density dependences of the photovoltage at a p-GaN(Cs) photocathode surface were measured in the temperature range 90–295 K. The study demonstrated that band bending at the p-GaN(Cs) photocathode surface can be reduced by ∼ 0.5 eV without modifying the surface atomic structure. The surface photovoltage impact on the p-GaN(Cs) photocathode quantum efficiency and photoelectron energy distributions was analyzed.

Spontaneous changes in photoemission properties of a р -GaN(Cs)–vacuum interface with effective negative electronic affinity induced by rearrangements of its atomic structure have been studied for the first time. The optimum Сs coating that ensures both the maximum photoelectron escape probability and its stability has been found. A thermodynamic model has been proposed to escape the relation of the photoemission properties of the р -GaN(Cs)–vacuum interface to its free energy and entropy.

Photoelectron scattering in a p-GaN(Cs,O) photocathode was studied by photoelectron emission spectroscopy at low temperatures. It has been experimentally established that the transfer of photoelectrons from the bulk of a p-GaN(Cs,O) photocathode to vacuum is accompanied by the emission of a cascade of optical phonons in the bulk and at the GaN-vacuum interface. Based on how the high-energy features of the photoelectron energy distribution shift with the photon energy, we find m(hh)/m(0) = 2.3 +/- 0.3.

We present measurements of the transverse and longitudinal energy spread of photoelectrons emitted from a GaAsP photocathode as a function of its degradation state. The cathode was initially activated to a state of negative electron affinity in our photocathode preparation facility, achieving a quantum efficiency of 3% at a wavelength of 532 nm. It was then transferred under XHV conditions to our transverse energy spread spectrometer, where energy spread measurements were made while the photocathode was progressively degraded through a controlled exposure to oxygen. Data have been collected under photocathode illumination at 532 nm, and the changing photoelectron energy distribution associated with the changes in the level of electron affinity due to quantum efficiency degradation through an exposure to 0.25 L of oxygen has been demonstrated. Our experiments have shown that GaAsP boasts a significantly higher resilience to degradation under exposure to oxygen than a GaAs photocathode, though it does exhibit a higher level of mean transverse energy. Coupled with the favourable published data on GaAsP photoemission response times, we conclude that GaAsP is a viable candidate material as a particle accelerator electron source.

It has been experimentally established that the transfer of photoelectrons from the bulk of a p-GaN (Cs,O) photocathode to vacuum is accompanied by the emission of a cascade of optical phonons. In the quantum efficiency spectrum of the p-GaN (Cs,O) photocathode, an exciton peak has been identified, indicating a significant contribution of the electron–hole interaction to the generation of free electrons in heavily doped p-GaN.

It has been established that the probability of Cs-induced chemisorption of molecular oxygen on a p-GaAs(Cs) surface is mainly determined by the probability of the dissociation of the molecule during its collision with the surface. With an increase in the amount of adsorbed oxygen on the p-GaAs(Cs, O) surface, the probability of its chemisorption decreases and, depending on the value of the cesium coating, can either be still determined by the probability of the dissociation of the molecule or be limited by the probability of the "capture" of oxygen atoms by the local chemisorption centers or depend on these two processes.

The (Cs,O)-activation procedure for p-GaAs(Cs,O)-photocathodes was studied with the aim of demarcating the domains of validity for the two practical models of the (Cs,O)-activation layer: The dipole layer (DL) model and the heterojunction (HJ) model. To do this, the photocathode was activated far beyond the normal maximum of quantum efficiency, and several photocathode parameters were measured periodically during this process. In doing so, the data obtained enabled us to determine the domains of validity for the DL- and HJ-models, to define more precisely the characteristic parameters of the photocathode within both of these domains and thus to reveal the peculiarities of the influence of the (Cs,O)-layer on the photoelectron escape probability.

The brightness of a photoelectron injector is fundamen- tally limited by the mean longitudinal and transverse energy distributions of the photoelectrons emitted from its photo- cathode, and the electron beam brightness is increased sig- nificantly if the mean values of these quantities are reduced. ASTeC have commissioned a Transverse Energy Spread Spectrometer (TESS - an experimental facility designed to measure these transverse and longitudinal energy distribu- tions) which can be used for III-V semiconductor, alkali antimonide/telluride and metal photocathode research (1). GaAs photocathodes were activated in our photocathode preparation facility (PPF) (2,3), then transferred to TESS under XHV conditions and progressively degraded through controlled exposure to oxygen. We present commission- ing data and initial measurements showing the evolution of the transverse energy distribution of electrons from GaAs photocathodes as a function of their degradation state.

The dosage dependences of surface band bending and effective electron affinity under cesium deposition on the Ga-rich GaAs(001) surface, along with the relaxation of these electronic properties after switching off the Cs source are experimentally studied by means of modified photoreflectance spectroscopy and photoemission quantum yield spectroscopy. At small Cs coverages, below half of a monolayer, additional features in the dosage dependence and subsequent downward relaxation of the photoemission current are determined by the variations of band bending. At coverages above half of a monolayer the upward relaxation of the photocurrent is caused supposedly by the decrease of the electron affinity due to restructuring in the nonequilibrium cesium overlayer.

Domains of validity for dipole layer and heterojunction models of the (Cs,O) – activation layer for GaAs – photocathode are determined. Two – step photoelectron escape model from NEA-photocathode is proved. Dominated elastic and inelastic scattering processes, which are accompanied the photoelectron escape, are revealed.

ASTeC has spent several years developing a GaAs Photocathode Preparation Facility (PPF) which routinely produces cathodes with quantum efficiencies (Q.E. )u p to 20 % at 635 nm [1]. The goal is to use these and other cathode materials in high–average–current high–brightness injectors for particle accelerators. Electron injector brightness is limited by source beam emittance, and brightness will be increased significantly by reducing the longitudinal and transverse energy spread in the emitted electrons, thereby creating a cold beam. We are constructing an experimental system which is compatible with the PPF for measurement of the energy distribution in electrons emitted from photocathodes at room and LN2-temperature. The photocathode will be illuminated by a small, variable–wavelength light spot. The electron beam image will be projected onto a detector comprised of a 3-grid energy filter, a microchannel plate and a phosphor screen. A low–noise CCD camera will capture screen images, and the spatial and energy distribution of the emitted electrons will be extracted through analysis of these images as a function of the grid potentials. The system will include a leak valve to progressively degrade the cathode, and thus allow its properties to be measured as a function of gas exposure.

Accelerator drivers for Energy Recovery Linac (ERL) and Free-Electron Laser (FEL) based light sources demand electron injectors which deliver high brightness bunches on the several hundreds of picocoulomb scale, at repetition rates between 1 MHz and 1 GHz (or higher), corresponding to an average current between 0.1 and 100 mA. Simultaneous satisfaction of these injector requirements is considerably beyond the current state-of-the-art. Daresbury Laboratory is concentrating efforts on the development of high average current III-V and XnY3−nSb photocathode-based DC and SRF photocathode guns for ERL applications. The ultimate goal of this research is their integration with the ALICE ERL.

A modification of photoreflectance spectroscopy, which improves the precision and efficiency of measuring time-evolving surface electric fields, is proposed. This modification is explored for studying the band bending evolution under cesium adsorption on the reconstructed GaAs(001) surface and relaxation processes in the non-equilibrium adsorbate overlayer. Observation of several distinct maxima and minima in the coverage dependence of the band bending can be explained by the formation of adatom-induced surface states with a quasi-discrete spectrum.

Accelerators and Lasers In Combined Experiments (ALICE) is an Energy Recovery Linac (ERL) built at STFC Daresbury Laboratory to investigate the process of energy recovery [1]. The project is an accelerator research facility intended to develop the technology and expertise required to build new light sources in the UK based on of Free-Electron Lasers. The original design of the ALICE photoinjector derives from the Jefferson Laboratory IRFEL photocathode gun. In its current configuration, the ALICE gun accommodates only a single photocathode, and the system must be vented to atmospheric pressure to facilitate photocathode replacement. To meet the stringent vacuum demands for good photocathode lifetime, the system then requires baking for up to three weeks. A new load-lock photocathode preparation facility (PPF) has been designed as part of an upgrade to the ALICE photoinjector. The PPF can accommodate up to six photocathodes, and permits rapid transfer of photocathodes between the PPF activation chamber and the ALICE gun, thus maintaining the integrity of the photoinjector vacuum system. The PPF was successfully commissioned with a GaAs photocathode in spring 2009, and has since permitted a quantum yield of 15 % to be achieved at a wavelength of 635 nm. Presently, a new photoinjector gun vessel and photocathode transport system is under design/manufacture, with a view to this being fully-installed on ALICE in spring 2011. ALICE PHOTOINJECTOR UPGRADE The whole ALICE photoinjector upgrade program includes: the development of photocathodes capable of delivering of 10-100 mA CW current; an integrated gun and PPF with a transfer system allowing for replacement of the photocathode in the gun within few hours; and a new gun vessel with improved pumping capabilities [2]. The current ALICE photocathode is a 25 mm O GaAs wafer activated in-situ and illuminated over a 4 mm O area by a 532 nm laser to drive electron emission [3]. The new photocathode assembly has been designed to accommodate a semiconductor heterostructure (1) with a 9.5 mm O active area, as shown in Fig. 1. As the laser only illuminates an area ~ 4 mm O, use of a smaller active area could help reduce the beam halo. The photocathode wafer is bound to 1.4 mm recess on the bottom of upturned 2 mm thick molybdenum “Petri dish” (2) with a diameter of 19 mm and a height of 6.5 mm. The polished surface of the cup (3) improves heat flow away from the photocathode surface when a high power laser is used to drive emission of high beam currents. The photocathode is retained in a transport holder (4) by a flat inconel spring (5). During preparation and operation, the photocathode is installed into a socket (6). Figure 1. GaAs photocathode in the mounting socket. The design of the three-chamber extreme-high vacuum PPF has been described in detail [4]. It comprises a loading chamber (LC), an atomic hydrogen cleaning chamber (HCC) and a preparation chamber (PC) (Fig. 2). Photocathodes are transferred between the chambers and also from the PPF into the photoinjector gun using magnetically-coupled linear manipulators. The proposed upgrade scheme permits usage of a variety of III-V photocathodes which may be activated with the PPF. The upgrade design will allow for the easy replacement of the III-V PPF with a facility for preparation or transport of a different type of photocathodes without venting the gun. One of the possible options under consideration is usage of ultra fast alkali antimonide-based photocathodes, which may operate at a wavelength of 532 nm with reasonably high levels of quantum efficiency. ___________________________________________ _____________ *boris.militsyn@stfc.ac.uk Proceedings of IPAC’10, Kyoto, Japan TUPE095 02 Synchrotron Light Sources and FELs A16 Energy Recovery Linacs 2347 Embodiment of the ALICE photoinjector upgrade is currently underway. The PPF has been constructed and commissioned in ‘stand-alone’ mode using GaAs heterostructures. A sophisticated side-load photocathode transport mechanism has been designed, and will be tested first in atmosphere, then in an extra-high vacuum facility which duplicates the vacuum conditions of the real gun. The test facility is now being commissioned. OPERATION OF THE III-V PPF IN STAND-ALONE COMMISIONING MODE In order to commission and to optimize our photocathode preparation procedure, the PPF has been operating in stand-alone mode. The vacuum port which will connect the PPF to the gun via an all-metal gate valve has been connected to a turbo vacuum pump which is used during the bake-out procedure. This solution will permit connection of the PPF to the operational gun without its ventilation. Originally, the III-V photocathode family (such as GaAs, GaAsP, InGaAsP, as grown) have a Positive Electron Affinity (PEA), which for GaAs is 4 eV. In order to make GaAs photocathodes able to emit electrons when illuminated by 532 nm light typical for ERL DC guns, the electron affinity should be lowered to less than 1 eV or even brought to a negative value. This process requires deposition on the atomically-clean photocathode surface of alternating thin layers of Cs and an oxidant, typically O2 or NF3, and is called activation. In the vacuum laboratory where the PPF is installed, O2 oxidant is used. Figure 2. III-V Photocathode Preparation Facility in the ASTeC vacuum laboratory. 1 – Loading Chamber (LC), 2 Atomic Hydrogen Cleaning Chamber (HCC), 3Preparation Chamber (PC), 4 magnetic manipulator, 5 hydrogen cylinder, 6 oxygen piezo-electric leak valve, 7 control rack An ideal photocathode preparation procedure includes chemical etching of the photocathode surface in hydrochloric acid to remove As and Ga oxides in a dedicated glove box under a pure nitrogen atmosphere. The photocathode is then transferred to the LC via a nitrogenfilled transport vessel to prevent the formation of new oxide species (see Fig. 2). The transport vessel can accommodate up to four samples simultaneously, though the full glove box procedure has not yet been implemented, so photocathodes are placed into the transport vessel as delivered rather than etched. The LC pumping system includes an oil-free preliminary pumping station, backed by an ion pump. During loading, the LC is vented to dry nitrogen gas. It is then pumped down to a typical pressure of 10 mbar, and then subjected to a short bake at a temperature of 120150 C in order to remove water from the samples. Once acceptable vacuum in the LC is established, the photocathode is transferred to the HCC for exposure to atomic hydrogen. Although it was originally implemented for rejuvenation of spent photocathodes, atomic hydrogen cleaning is now used for initial photocathode cleaning. TUPE095 Proceedings of IPAC’10, Kyoto, Japan 2348 02 Synchrotron Light Sources and FELs A16 Energy Recovery Linacs Typically, photocathodes are heated to 450 °C (measured with a pyrometer) and exposed to hydrogen at 2 × 10 mbar for 15 minutes. Cs-O activation of GaAs photocathode with Yo-Yo procedure