Multiphoton Emission from Cesium Telluride Photocathodes

The High-Brightness Electron Source Lab (HBESL) at Fermilab operates an L-band RF gun (1.3-GHz) that incorporates a Cesium Telluride (Cs2Te) photocathode illuminated by an ultra-short (sub-100 fs) laser pulses. In this contribution we report recent studies on exploring the electron beam emission using 800 nm laser pulses (instead of the 266 nm nominal wavelength used for linear photoemission). INTRODUCTION Photocathodes are excellent sources for production of short electron bunches [1]. Production of short bunches is possible due to the advancements in the drive laser pulse times combined with availability of fast-response photocathodes. Photoemission allows generating higher charges compared to fieldand thermionic emission, and also with small beam transverse emittance. One good parameter to characterize the performance of a photocathode is quantum efficiency (QE), which is defined as the number of electrons emitted for a unit photon. Cesium Telluride is considered to be an exceptionally high QE (up to 20 %) semiconductor photocathode [2]. The work function of Cs2Te 4.6 eV corresponds to the laser wavelength of 266 nm falling in the ultraviolet (UV) region. Since most of the commercially available lasers include lasing media with high gain in the infrared (IR), frequency up-conversion to the UV is generally done for linear photoemission from metallic and some semiconductor cathodes. For titanium sapphire based laser systems ( ~800 nm), UV pulses for photoemission are obtained from frequency tripling of IR pulses using a two-stage process consisting of a second harmonic generation (SHG) stage followed by a sum frequency generation (SFG). In order to preserve the short pulse duration during the up-conversion process, both stages generally use thin BBO crystals which results in low IR-to-UV conversion efficiency typically < 10%. In this paper we present preliminary results toward attempting to generate electron bunches with commensurate charge directly using the amplifier IR pulse. We specifically report on the observation of twophoton emission from Cs2Te. PHOTOCATHODE THEORY Spicer developed the first model in 1958 to explain photoemission in semiconductors, which is commonly referred to as the Three Step Model, which models the photoemission mechanism as a bulk effect [3]. In the first step, an electron absorbs a photon and gets excited from the valence band to the conduction band. In the second step, the excited electron transports to the surface. In the third step, the electron escapes to the vacuum level resulting in emission. The QE equation for a semi-infinite photocathode slab is given by [3] Where is the reflectivity of the material; is the intensity attenuation coefficient of the photocathode material; represents the coefficient of absorption for the vacuum level, or the number of electrons that are excited to the vacuum level that can possibly photoemit per unit laser intensity available; is called the escape length which represents the strength of the electron scattering; is the probability that an electron at the surface with sufficient energy to escape, escapes. Here all the parameters are functions of . Fowler-Dubridge model is derived for photoemission for metallic photocathodes based on similar approach. It is the first step (photo-absorption and excitation) that distinguishes between single-photon emission from multiphoton emission. Once the electrons get excited, the rest of the processes viz. electron transport to the surface and electron escape remain the same. In single photon emission the intensity dependence of QE goes away. But in multi-photon emission additionally depends on the intensity of the light (temporal) besides , as the nonlinear “simultaneous” absorption of multiple photons (of lower energy) has to occur to excite an electron to the vacuum level. Hence shorter laser pulses are required for higher order photoemission.