Dynamics of Beam Interaction and Damage within a Thin TEM Specimen

Specimen heating is an unavoidable consequence of the inelastic scattering of primary electrons. For a thin-slab specimen, the temperature at the center of the beam increases with a time constant that depends on the beam radius R and to a lesser extent on the distance R0 to the heat sink [1]. The temperature rise ΔT is typically below 10 K for beam currents of 1 nA or less; see Fig. 1. Electrostatic charging occurs for poorly conducting samples (density ρ, thickness t), with a rise time ε0εrρ/σ that ranges from fs to ms, depending on the electrical conductivity σ; see Table 1. For small-radius probes, an equilibrium potential Vs may be reached, resulting in hole drilling due to ion emission [2] if the local charge density exceeds about 0.1 electrons/atom, or in ionic drift or dielectric breakdown caused by the field dV/dr near the edge of the probe [3]. Estimated charging properties [1] that arise from a STEM probe (2R = 1 nm, Ib = 0.4nA) and from TEM illumination (2R = 5 μm, Ib = 0.4nA; values in parenthesis), taking t = 100 nm, R0 = 30 μm and SE yield (at surface potential Vs) = 0.01. Estimated charging properties [1] that arise from a STEM probe (2R = 1 nm, Ib = 0.4nA) and from TEM illumination (2R = 5 μm, Ib = 0.4nA; values in parenthesis), taking t = 100 nm, R0 = 30 μm and SE yield (at surface potential Vs) = 0.01. All of the above data are based on macroscopic values of the thermal and electrical properties. Further insight can be obtained by considering the effects of the secondary electrons (SE) generated by each inelastic scattering event of a primary electron. Nielsen et al. [4] used Monte Carlo calculations to predict the behavior of a single 100eV electron, created in diamond or GaN. Inelastic scattering reduces this kinetic energy to below 10 eV within 1 fs, generating 8 electron-hole pairs but sub-nm radial spread; see Fig. 2a. Between 1 fs and 1 ps, the SE become thermalized by phonon scattering and three more e-h pairs are created, with significant radial spread; see Fig. 2b. In this respect, electron irradiation differs from that of incident ions, which deposit most of their energy directly to atomic nuclei, generating highly localized thermal spikes that expand at the speed of sound, dissipating their energy [5]. In a beam-sensitive organic material, radiation damage (radiolysis) arises mainly from SE that are produced with energies of 10 – 30 eV (from valence-electron excitation) or several hundred eV if generated by K-shell excitation. In such a material, hole production is replaced by the creation of ionized molecules, which repel and react with each other, while the generated secondary electrons are mobile and create further damage within the “spur” [5]. The subsequent (chemical) stage of radiolysis involves diffusion processes that occur on a longer time scale. So mass loss can be reduced (for the same amount of acquired data) by using a high dose rate [6], even though Coulomb repulsion between primary electrons precludes outrunning the primary (physical) damage processes. Conversely, low dose rates can be beneficial where a back-reaction (repair) process acts to reduce damage, or where beam heating exacerbates the damage. The use of pulsed electron beams could lead to a better understanding of radiation damage or even the possibility of reducing it [7]. For this purpose, precise timing of the incident electrons might be achievable with laser-driven electron sources but is more problematic for a chopped thermionic beam. Fig. 3 shows a typical electron-arrival pattern for a beam current of 22 pA and a chopping frequency of 5.2 GHz, as used in damage studies on paraffin [8]. The spatial delocalization and time dependence of damage is relevant to the possibility of cooperative effects (between inelastic-scattering events of different primary electrons) that could in principle be reduced with a pulsed beam. However, such effects should show up as a measurable dose-rate dependence of damage [9]. Thermal time constant τ and temperature rise ΔT at the center of the beam (current = 1 nA) for electrons transmitted through thin samples of silicon, PMMA and ice. (a) Mean number of electron-hole pairs generated by a single 100eV electron in GaN, versus time after creation. (b) Radial displacement of the hot electrons created in GaN, as a function of time. Thirty electrons in a gated beam with 22% duty cycle (top) and in a continuous beam from a thermionic source (middle) with the same current (22 pA). The horizontal axis represents distance along the beam time, or arrival time at the specimen. The gating effect becomes apparent only at higher beam current (e.g. 1 nA, bottom red display).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.