Metal-semiconductor (MS) ultrathin film devices provide experimental systems which facilitate the detection of hot charge carriers ballistically transporting through a thin metal layer excited by low energy particle bombardment. The investigation of hot electrons at surfaces requires devices with homogeneous and crystalline metal layers and minimal defects to be grown only a few monolayer thick. Furthermore, growth and characterization in ultrahigh vacuum (UHV) ensures minimum contamination.
I. Deposition system
We designed a compact deposition chamber for the growth and characterization of ultrathin noble metal layers on solid crystalline substrates. Our deposition chamber is attached the the loadlock chamber of an Omicron VT-STM  and includes the capability to cool the surface during growth. We chose to grow Ag/n-Si(100) Schottky diodes as a prototype MS device because Ag grows layer-by-layer and has domains of (111) orientation at deposition temperatures easily obtainable by liquid nitrogen cooling. Additionally, there is minimal chemical reactivity between the substrate and film at room temperature so no wetting layer is required and the metal-semiconductor interface is abrupt.
II. Film growth and characterization
Ag films were deposited with the McAllister Technical Services thermal evaporator gun equipped with a Ta heating filament inside a stainless steel water cooled jacket. Ag wire (99.99% pure) was evaporated onto low temperature (~120 K) n-Si(100) substrates to thicknesses between 8-15 nm and annealed at room temperature. Finished devices where characterized with AFM and J-V measurements and are reported in Refs. 2, 3.
A. Ar sputter/Auger analysis
To characterize the mechanical stability of Ag films during interaction with an ion beam, sputter depth analysis data were recorded using Auger electron spectroscopy (AES). Specifically, the total height of the AES peak corresponding to Ag was recorded as a function of sputter time for a 500 eV Ar beam near normal incidence on a 15 nm thick Ag top layer. The Ag AES peak was extinguished after 105 minutes of sputtering. The decrease in the Ag AES signal intensity was fit linearly to determine its erosion rate t. The observed negative slope of 0.94 ± 0.02 % per minute was used to determine an upper limit for the amount of sputtering which takes place on the device during beam interaction.
For one minute of beam exposure on a 20 nm thick film, less than 2 Å of thickness is lost which is on the order of half of one monolayer for the Ag surface. Since detection of hot electrons induced by beams of particles only requires exposing the metal surface for a few seconds at a time to record a current response this sputter measurement shows that the MS structures will remain stable during ion beam exposure.
B. Ion scattering spectroscopy
Hyperthermal energy ion scattering spectroscopy was done by impinging singly charged 1 keV K ions at an incident angle of 45°. Energy loss spectra were recorded at a specular scattering angle and are presented in Fig. 3. The trajectories of ions scattered from the surface were analyzed in terms of the sequential binary collision approximation. The most prominent feature in the spectrum was a quasisingle (QS) scattering event occurring between the K ion and single atom on the Ag surface explained by a classical kinematic factor.
We have demonstrated a design and fabrication method for MS ultrathin film devices suited to the investigation of internal emission of hot charge carriers induced by hyperthermal energy particle bombardment. Our custom built compact deposition chamber is attached to the Omicron VT-STM for in situ growth and analysis of as grown thin films. AFM, J-V and ion scattering spectroscopy characterization our candidate devices confirm that the ultrathin metal layers that we have grown are continuous and that energy deposition from particle bombardment can be predicted according to binary collision kinematics.
 Omicron NanoTechnology GmbH. Taunusstein, Germany.
 R. E. Lake, J. R. Puls, M. P. Ray and C. E. Sosolik. J. Vac. Sci. Technol. A. 27, 1024 (2009).
 M. P. Ray, R. E. Lake and C. E. Sosolik. Nucl. Instrum. Methods Phys. Res. B. 267, 615 (2009).