Brightness evaluation of pulsed electron gun using negative electron affinity photocathode developed for time-resolved measurement using scanning electron microscope

Highlights

• Scanning electron microscope with pulsed electron gun using photocathode.

• Short-pulsed electron beam emitted from negative electron affinity (NEA) photocathode.

• Waveform measurement of pulsed electron beam using prototype detector consisting of an avalanche photodiode (APD).

Abstract

Temporal changes in carrier relaxations, magnetic switching, and biological structures are known to be in the order of ns. These phenomena can be typically measured by means of an optical-pump & electron-probe method using an electron microscope combined with a pulsed electron source. A photoemission-type pulsed electron gun makes it possible to obtain a short-pulsed electron beam required for high temporal resolution. On the other hand, spatial resolution is restricted by the brightness of the pulsed electron gun used in electron microscopes when a low brightness electron source is used and an irradiation current larger than a certain value is required. Thus, we constructed a prototype pulsed electron gun using a negative electron affinity (NEA) photocathode for time-resolved measurement using a scanning electron microscope (SEM) with high spatiotemporal resolution. In this study, a high-speed detector containing an avalanche photodiode (APD) was used to directly measure waveforms of the pulsed electron beam excited by a rectangular-shape pulsed light with a variable pulse duration in the range of several ns to several μs. The measured waveforms were the same rectangular shape as incident pulsed excitation light. The maximum peak brightness of the pulsed electron beam was 4.2×10⁷ A/m²/sr/V with a pulse duration of 3 ns. This value was larger than that of the continuous electron beam (1.6 × 10⁷ A/m²/sr/V). Furthermore, an SEM image with image sharpness of 6.2 nm was obtained using an SEM equipped with a prototype pulsed electron gun at an acceleration voltage of 3 kV.

Introduction

The scanning electron microscope (SEM) is an essential tool widely used for observing and analyzing various specimens at a high spatial resolution in many fields including semiconductor devices, materials sciences, and biology [1]. Recently, improvements to pulsed electron sources have made it possible to obtain short-pulsed electron beams with a pulse duration shorter than 1 ns [2]. By using electron microscopes combined with short-pulsed electron sources, measurements with high spatiotemporal resolution can be realized [4]. Such a measurement technique makes it possible to visualize local and fast phenomena such as temporal changes in carrier relaxations, magnetic switching, and biological structures [[3], [5], [6], [20]]. A photoemission-type electron gun is preferable for obtaining high temporal resolutionbecause an electron beam with a short pulse duration can be easily generated by exciting with a pulsed light [9].

In general, the spatial resolution of a microscope depends on diffraction and aberrations. However, for SEM, it is necessary to account for the limitation in spatial resolution due to the brightness of the electron gun. The brightness of an electron gun is defined as an irradiation probe current per unit area and unit solid angle. The brightness normalized for acceleration voltage ($V_0$) of the electron beam at a source is defined as a reduced brightness (${\beta }_{\mathrm{V}}$), which is represented as

$${\beta }_{\mathrm{V}}=I_{\mathrm{p}}/\left({\pi }^2{r}_{\mathrm{o}}^2{\alpha }_{\mathrm{o}}^2{V}_0\right),$$

where $I_{\mathrm{p}}$ is the probe current and $r_{\mathrm{o}}$ and ${\alpha }_{\mathrm{o}}$ are the beam radius and beam aperture angle at a virtual source, respectively [1]. The reduced brightness is normalized by the acceleration voltage, so it does not depend on the energy of the electron beam. If an electron beam emitted from the virtual source is focused on the specimen using an electron optics system with a magnification of M, the projected source size and the beam aperture angle at the specimen are calculated by $r_s=M{r}_0$, and ${\alpha }_s={\alpha }_0/M$, respectively. Given Eq. (1), a source size projected onto the specimen is represented as,

$$2r_s=2/\left(\pi {\alpha }_s\right)\sqrt{(l_p/{\beta }_v{V}_0))}$$

This indicates that the spatial resolution is limited by brightness when using a low brightness electron source and more than a certain value of the probe current. Eq. (2) can be used to estimate the required brightness for the electron gun. When aiming for a spatial resolution of 10 nm, a source size ($2r_s$) projected onto the specimen should be smaller than 10 nm. The optimal irradiation angle (${\alpha }_s$) at the specimen under the condition for high spatial resolution with an acceleration voltage about 1 kV is several mrad, which is determined by diffraction, aberrations, and brightness. Because the irradiated probe current ($I_{\mathrm{p}}$) should be larger than 10 pA to obtain SEM images with sufficient signal-to-noise ratio, brightness (${\beta }_{\mathrm{v}}$) must be higher than 1×10⁷ A/m²/sr/V. In addition, high brightness is expected to be advantageous for high-throughput measurement.

In this study, we focus on an electron gun using a high-brightness negative electron affinity (NEA) photocathode with p-type GaAs and Cs-O adsorbates. The NEA condition is defined as when the bottom of the conduction band inside GaAs is higher than vacuum level and the electrons excited from the valence band to the conduction band by light are spontaneously emitted to vacuum. The advantages of electron sources using the NEA photocathode are high brightness (1.3 × 10⁷ A/m²/sr/V) [12] and small energy spread (0.24 eV) [13]. By utilizing an electron gun with the NEA photocathode, the spatial resolution of an SEM in the low acceleration voltage can be improved by reducing chromatic aberration which is proportional to energy spread [10].

The purpose of this study is to evaluate pulsed electron beams emitted from a prototype electron gun using the NEA photocathode excited with a pulse duration of several ns. To evaluate the both time-average brightness and peak brightness of the pulsed electron beam, waveforms of the pulsed electron beams were measured by means of a high-speed detector which consists of a silicon avalanche photodiode (Si-APD) and a high-speed amplifier.