Search

The photocurrent from a semiconductor photocathode with a negative-electron affinity surface can be arbitrarily controlled by the excitation laser power. Applying this characteristic to a scanning electron microscope allows the probe current to be arbitrarily controlled at any location on the sample. A photocathode with a fast time response is required to control the probe current at high speed. This study used an InGaN photocathode for pulsed electron beam generation and investigated its time response. A pulsed electron beam with 3.8 ns pulse width and 8.1 × 103 A cm−2 current density was observed, and the rise and fall times of the photocurrent were found to be 1.7 and 2.0 ns, respectively. The results show that despite the bottleneck of the time response of the laser power, the InGaN photocathode generates an electron beam that can control the probe current on a pixel-by-pixel for a 270 MHz scan speed.

Pulsed electron beams from a photocathode using an InGaN semiconductor have brought selectively scanning technology to scanning electron microscopes, where the electron beam irradiation intensity and area can be arbitrarily selected within the field of view in SEM images. The p-type InGaN semiconductor crystals grown in the metalorganic chemical vapor deposition equipment were used as the photocathode material for the electron beam source after the surface was activated to a negative electron affinity state in the electron gun under ultrahigh vacuum. The InGaN semiconductor photocathode produced a pulsed electron beam with a rise and fall time of 3 ns, consistent with the time structure of the irradiated pulsed laser used for the optical excitation of electrons. The InGaN photocathode-based electron gun achieved a total beam operation time of 1300 h at 15 μA beam current with a downtime rate of 4% and a current stability of 0.033% after 23 cycles of surface activation and continuous beam operation. The InGaN photocathode-based electron gun has been installed in the conventional scanning electron microscope by replacing the original field emission gun. SEM imaging was performed by selective electron beaming, in which the scanning signal of the SEM system was synchronized with the laser for photocathode excitation to irradiate arbitrary regions in the SEM image at arbitrary intensity. The accuracy of the selection of regions in the SEM image by the selective electron beam was pixel by pixel at the TV scan speed (80 ns/pix, 25 frame/s) of the SEM.

The dependence of the electron emission current density on the excitation power density of a Cs/O-activated negative electron affinity (NEA) InGaN photocathode was investigated. The emission current density of the NEA-InGaN photocathode increased monotonically with the excitation power density in the measured range. The emission current density reached 5.6 × 103 A/cm2 at an excitation power density of 2.6 × 106 W/cm2. Using the electron thermal energy estimated by comparing simulation and experimental results [D. Sato, H. Shikano, A. Koizumi, T. Nishitani, Y. Honda, and H. Amano, J. Vac. Sci. Technol. B 39, 062209 (2021)], the reduced brightness of 4 × 108 A/m2 sr V was derived.

In this study, we generated 25 multielectron beam (MEB) using an InGaN photocathode with a negative electron affinity state irradiating with 25 multilaser beam. The uniformity of the MEB and the total electron beam current were evaluated. A laser beam was split into 25 laser beams using a spatial light modulator. The coefficient of variation (CV) of laser power was 20%. The CV of quantum efficiency was 1.1%. The CV of electron beam current was 12%, and the total current was about 1.2 μA. These results will enhance the development of the MEB‐defect inspection using the InGaN photocathode.