Spin- and angle-resolved inverse photoemission setup with spin orientation independent from electron incidence angle

Response to "Comment on 'Spin- and angle-resolved inverse photoemission setup with spin orientation independent from electron incidence angle'" [Rev. Sci. Instrum. 93, 093904 (2022)]

We reply to the Comment by Donath et al. on our setup, which allows a total 3D control of the polarization direction of the electron beam in an inverse photoemission spectroscopy (IPES) experiment, a significant advance with respect to previous setups with partial polarization control. Donath et al. claim an incorrect operation of our setup after comparing their results, treated to enhance the spin asymmetry, with our spectra without the same treatment. They also equal spectra backgrounds instead of equaling peak intensities above the background. Thus, we compare our Cu(001) and Au(111) results with the literature. We reproduce previous results, including spin-up/spin-down spectral differences observed for Au and not observed for Cu. Also, spin-up/spin-down spectral differences appear at the expected reciprocal space regions. In the Comment, it is also stated that our tuning of the spin polarization misses the target because the spectra background changes when tuning the spin. We argue that the background change is irrelevant to IPES since the information is contained in peaks produced by primary electrons, those having conserved their energy in the inverse photoemission process. Second, our experiments agree with previous results from Donath et al. [Wissing et al., New J. Phys. 15, 105001 (2013)] and with a zero-order quantum-mechanical model of spins in vacuum. Deviations are explained by more realistic descriptions including the spin transmission through an interface. Consequently, the operation of our original setup is fully demonstrated. Our development corresponds to "the promising and rewarding angle-resolved IPES setup with the three-dimensional spin resolution," as indicated in the Comment, after our work.

Comment on "Spin- and angle-resolved inverse photoemission setup with spin orientation independent from electron incidence angle" [Rev. Sci. Instrum. 93, 093904 (2022)]

The paper under discussion promises a spin- and angle-resolved inverse-photoemission (IPE) setup, where the spin-polarization direction of the electron beam used for excitation "can be tuned to any preferred direction" while "preserving the parallel beam condition." We support the idea to improve IPE setups by introducing a three-dimensional spin-polarization rotator, but we put the presented results to the test by comparing them with the literature results obtained by existing setups. Based on this comparison, we conclude that the presented proof-of-principle experiments miss the target in several aspects. Most importantly, the key experiment of tuning the spin-polarization direction under otherwise allegedly identical experimental conditions causes changes in the IPE spectra that are in conflict with existing experimental results and basic quantum-mechanical considerations. We propose experimental test measurements to identify and overcome the shortcomings.

On the energy resolution of a GaAs-based electron source for spin-resolved inverse photoemission

The spin resolution in inverse photoemission spectroscopy is achieved by injecting spin-polarized electrons, usually produced by GaAs-based cold cathodes that replace hot-filament electron guns of spin-integrated setups. The overall energy resolution of the system can be enhanced by adjusting either the optical bandpass of the photon detector or the energy distribution of the electron beam. Here we discuss the influence of the photocurrent and the photocathode temperature on the energy broadening of the electron beam through the inverse photoemission spectra of the spin-splitted Shockley surface state of Au(111). First, we find that cooling down the GaAs photocathode to 77 K increases the band gap and reduces the number of allowed vertical transitions, monochromatizing the electron beam with an enhancement of about 30 meV for the energy resolution. Second, we observe a correlation between the generated photocurrent at the electron source, and the space-charge effects at the sample as a reduction of lifetime and spin asymmetry of a polarized bulk state. These observations allow defining a threshold of current density for the optimum acquisition in the measurements of spin-resolved inverse photoemission in Au.