Phase contrast transmission electron microscopy with hole‐free phase plate for material analysis

Abstract

In an analysis of material structures, transmission electron microscopy (TEM) stands out as an essential tool. A wide variety of information can be obtained from a sample by appropriately using TEM. Recently, phase contrast transmission electron microscopy (PC‐TEM) with a hole‐free phase plate (HFPP) has been established [1‐4]. The HFPP is placed at the back focal plane of the objective lens. The incident beam passing through the thin film of HFPP generates secondary electrons, and they in turn lead to an electrostatic charging (potential). As the intensity distribution of electron beam on the HFPP has sharp peak at the center of optical axis, the charging occurs only at region of the beam crossover. This causes an additional phase shift for the direct beam, resulting in high contrast images of samples. The difference of phase shift between scattered electrons and direct electrons, made by the charging, enhance the image contrast as the common phase microscopy. Thus, the contrast of materials made of light elements, e.g., polymer samples and biological samples, can be significantly enhanced by HFPP. Furthermore, we expect that it may be possible to derive phase, i.e., the mean inner potential, from the images obtainable from the PC‐TEM with HFPP. In this study, we demonstrate a couple of experimental results of PC‐TEM with HFPP. First example is an observation of nano‐scale periodic structures of block copolymer (BCP). BCPs consist of light elements and thus often require “staining” of one of the phases for contrast enhancement under TEM observations. Because the staining might change nano‐structures, it would be better not to use this method for increasing contrast. We used a field emission TEM (JEM‐2200FS) operated at 200 kV with the PC‐TEM with HFPP in this experiments. Figure.1 shows TEM images of a unstained block copolymer (poly(styrene‐ b ‐isoprene)) obtained with and without HFPP. The lamellar structure was hardly seen in the TEM images without HFPP at in‐focus, i.e., Figures 1a. Under‐focusing slightly enhanced the contrast between the polystyrene (PS) and polyisoprene (PI) phases in Figure 1b. In contrast, it was obvious that the use of HFPP significantly increased contrast of the two phases in lamellar structure. Next, we explore the possibility of HFPP imaging for phase retrieval. As a first trial we selected the simple case of nanoparticles, for which we can assume a weak phase object approximation. As far as the phase values is small (typically <0.5 rad), we propose to retrieve the phase from two images with and without HFPP at the same defocus condition. The knowledge of the two images allows then for a simple inversion of the contrast transfer function that directly leads to a phase image of the area. Figure 2a and 2b show images of platinum nanoparticles with and without HFPP. Figure 2c is the phase image retrieved from the two images using the simple inversion. Phase images have been also retrieved for sets of images taken at different defocus (0, ‐40nm, ‐80 nm): the same phase images were obtained (though with a better signal/noise ratio at zero defocus). For comparison, an in line holography processing earlier validated on gold nanoparticles [5] has been applied: a good agreement on the phase value has been observed when comparing to the phase image of the same area obtained with and without HFPP. The measured phase values are close to 0.3 rd for 2‐nm particles as expected from simple calculation based on the mean inner potential (V Pt = 25 V). The advantage of the PC‐TEM with HFPP is easier usability than conventional phase retrieving methods, such as off axis electron holography. From these applications, we emphasize that PC‐TEM with HFPP would be one of prospective analysis tools for materials characterization.