Plasma cleaning effect on the stability of theEpon resin sections

Abstract

Low voltage TEM and STEM (transmission and scanning transmission electron microscope) can be regarded as the method of choice for many structural studies of very thin biological samples like ultrathin sections, viruses etc. [1]. Unfortunately, the specimen contamination increases with electron flux and therefore a specimen cleanliness is an important factor in obtaining of valuable data especially in STEM [2]. An important parameter for imaging of those samples is a sensitivity of the sample to degradation by electron beam. The mass loss indicates a degree of the radiation damage. We investigated the mass loss of embedding medium (Epon resin of middle hardness) in combination of different thickness of the sections (60 nm and 150 nm) with using or not‐using of plasma cleaning which is often used to removing of contamination from the sample.The repeated imaging was performed by SEM Magellan 400L (FEI) at the acceleration voltage of 30 kV and the lowest possible probe current of 1.6 pA in the bright field (BF) using the commercial STEM3 detector (FEI). An electron dose for each scan was about 60 e‐/nm2. The measurement was performed at the eighth day after their preparation by the ultramicrotome Leica Ultracut UCT. Some slices were cleaned by Plasma Cleaner (FEI) installed at the Magellan SEM for 20 seconds just before imaging, so they did not leave high vacuum. For each experiment, three different places on each sample were imaged for downgrading space dependency of the measurement. The mass loss was obtained from the normalized BF signals using the Monte Carlo simulation of electron scattering MONCA [3] using the methodology of mass measurements by STEM [4]. All data processing was programmed in MATLAB (Mathworks).As seen on Fig. 1a, the normalized BF signal (showing the fraction of scattered electron to the BF detector) has similar shape in both thicknesses of the slices without plasma cleaning; the curves are shifted only by the different thickness. However, slices treated with the plasma cleaner show higher sensitivity to incident electron beam with rapidly increasing BF signal; Fig. 1b) shows the remaining mass. At the total irradiation dose of 3000 e‐/nm2there still remains 89% of initial mass for the 150 nm slice and 70% for the 60 nm slice, respectively. However, the slices cleaned by plasma cleaning are much more sensitive to the electron beam, with remaining 55% for the 150 nm slice and local destruction for the 60 nm slice. We observed higher sensitivity of cleaned Epon thin sections under incident electron beam. This limits the usable dose for imaging by the low voltage STEM (for 30 kV) because cleaned sections are more susceptible to burn‐out than non‐cleaned ones.