Size-Uniform Spherical Particles and Their Application in Electron Microscopy

The possibility of obtaining aerosols and hydrosols containing spherical particles sufficiently uniform in size is of primary importance for a number of areas of physics. Until recently this problem had remained essentially unresolved. Numerous attempts did, it is true, lead in a number of cases to comparatively sharp particle-size distribution curves; however, it proved impossible to achieve genuine uniformity of the particles. Therefore the brief note by Backus and Williams ), who in their investigations in electron microscopy accidentally encountered spherical particles exceptionally uniform in size, is of undoubted interest. Unfortunately, no information is given either about the nature of these particles or about the method by which they were obtained. The authors confine themselves to the statement that the particles are contained in one of the commercial polystyrene latexes *).

Fig. 1. Electron micrograph of a replica of a diffraction grating (15,000 lines per inch) with a suspension applied to it by the shadow-contrast method.

The authors measured the mean diameter of the particles and their size distribution. For this purpose they initially attempted to use—

) R. C. Backus and R. C. Williams, J. Appl. Phys. 20, 224 (1949).
*) “Dow Latex 590-G, Lot 3584.”

use organic replicas of a diffraction grating with a known spacing between the rulings, subjected to preliminary shadow contrasting. The particles were applied to the replica in the form of a suspension of polystyrene latex in distilled water (the authors note that latex readily forms such suspensions). Fig. 1 presents an electron microphotograph obtained by them of a portion of a collodion replica of a glass diffraction grating with 15,000 rulings per inch, with the suspension deposited on it (concentration 1:1000).

However, comparative measurements of the particle diameters and the distances between the grating rulings showed that the latter are considerably less uniform than the particle sizes and are unsuitable for measuring them.

The authors then resorted to another method. A piece of a thin glass thread was immersed in a concentrated suspension of the particles under investigation and was then examined under an ordinary light microscope equipped with a previously calibrated ocular scale. With the aid of this scale, the distances between sharply distinguished clusters of particles that had settled on one or several threads and were 40–50 microns apart were measured. Then the same thread was sought in the field of view of the electron microscope at a magnification of 10,000; the selected clusters were identified, and, by obtaining a series of photographs partially overlapping one another, the distance between the clusters was again measured (along the thread). At the same time, the diameters of numerous particles situated in the measured interval were measured.

As a result of the measurements, the authors obtained, for the mean particle diameter, the value \(2590 \pm 25\) Å; it is indicated that the probable error should be attributed not to the dispersion of the particles in size, but to errors of the measurement method. The distribution of particles by size is characterized by the following data. About 500 particles were measured on 20 different photographs. Within the limits of the internal probable error of measurement, \(\pm 10\) Å, the measured values of the diameters coincide. Only a few particles (about 2% of the total number) had dimensions differing sharply (approximately by a factor of two) from the normal ones. Observations show that all particles have a strictly spherical shape.

Fig. 2. Electron microphotograph of a replica of a diffraction grating with a suspension deposited on it before shadow contrasting with uranium.

On the basis of these data, the authors propose using the described particles as a scale for measuring images obtained in electron microscopy.

In addition to direct measurement of the magnification in various places of the image, they indicate the following two applications:

1) measurement of the shape of the replica surface, and
2) measurement of the thickness of the layer used for shadow contrasting.

Fig. 2 presents a microphotograph of a replica of a diffraction grating subjected to shadow contrasting with uranium after deposition of the suspension. The ratio of the length of the shadow to the particle diameter gives the value

of the shadowing angle at the given point. Since this angle may vary over the extent of the object, determining it at different points in this way is very desirable. It makes possible not only a qualitative but also a quantitative measurement of the relief of the replica. In Fig. 2 one can clearly see, in the length of the shadows from the particles, the difference between raised and depressed parts of the surface.

If the particles were deposited on the replica both before and after shadow contrast enhancement, then, obviously, the thickness of the contrasting layer can be determined. In Fig. 2 it may be noted that the particles are slightly elongated in the direction of the shadow. The elongation is due to the build-up of the contrasting layer on the side facing the evaporator (in Fig. 2 this elongation is 35 Å). The thickness of the contrasting layer on the surface of the specimen can be determined by measuring the difference between the mean diameter of uncontrasted particles and the mean maximum size of contrasted particles and dividing this difference by the ratio of the shadow length to the particle diameter.

There is no doubt that, along with their use as scale spheres, these particles will find wide application in studies on the optics of turbid media.

G. R.