On a Method of Spectral Analysis by Means of X-rays.

Manne Siegbahn, Axel E. Lindh and Nils Stensson. Über ein Verfahren der Spektralanalyse mittels Röntgenstrahlen. Z. f. Phys. 4, p. 61. 1921.

X-ray studies at the present time can answer the question whether a given element is present in a given sample of a substance, and tenths of one percent of content are sufficient for it to be discovered.

To extend the question in the direction of the simultaneous determination of all the elements contained in a sample, the X-ray spectrographs used practically have not yet made possible.

Theoretically the problem appears quite feasible, since X-ray spectra possess a very simple and regular structure. The most intense lines of the two principal \(K\) and \(L\) groups are so characteristic that even an approximate determination of the wavelength makes it possible to conclude to which elements they belong.

Moreover, the X-ray spectrum of an element is an atomic property of it and in no way depends on molecular compounds or on the presence of foreign bodies.

The authors of the article under review propose a practical method of qualitative chemical analysis by means of X-rays—a method which, under certain conditions, also makes it possible to judge the approximate quantitative ratio of the elements entering into the composition of the body, whose spectra \((Na—Ur)\) are known.

The X-ray tube employed is, in its essential features, arranged as follows: the body of the tube is a metal rectangular box a little larger than a matchbox, with a drilled opening for the cathode, anticathode, and the tube leading to the molecular pump.

The cathode is an incandescent spiral, giving a rather extended focus on the anticathode, which makes it possible to fulfill the basic condition of the construction in this apparatus—obtaining considerable \((40^\circ)\) solid angles both for the ray coming from the anticathode and for that reflected from the crystal. Opposite the anticathode, placed at the closest possible distance from the opposite wall of the tube, a slit is cut \((4.0 \times 0.1\ \mathrm{mm})\), to which, before assembly with the spectrograph, a thin aluminum or gold plate is attached, pressed by a copper ring screwed to the tube; to this ring the crystal is fastened with cement (in the experiments described, Iceland spar or gypsum). The inclination of the crystal is such that the reflected rays go at an angle of \(10^\circ—50^\circ\).

The closeness of the crystal and anticathode to the slit contributes to fulfilling the above-mentioned condition—obtaining large solid angles. The vacuum spectrograph, screwed to the tube, has in cross-section the form of a sector, along the arc of which a photographic film is introduced through a side opening. The center of the arc is at the place of the image of the slit on the reflecting plane of the crystal.

This makes it possible to regard the wavelengths measured on the spectrograms as proportional to the angles of reflection of the rays. During operation the spectrograph is connected by a tube with the fore-vacuum of the molecular pump.

The cathode, anticathode, and box are cooled with water.

The tube operates by means of a single-arm induction coil with a transformer at approximately \(15000\ V\) and \(50\) periods, and with a rectifier on the motor axis. Heating of the cathode is produced by a small transformer at \(10\ V\). The current in the tube during exposures is \(20—30\) milliamperes.

For first-order spectra, with the lattice constant of Iceland spar

\[ 2d = 6.060\ \overset{\circ}{A} \]

and angles of reflection \(\varphi\) from \(10^\circ—50^\circ\), from the relation \(\lambda = 2d \sin \varphi\) we obtain for \(\lambda\) \(1.0—4.6\ \overset{\circ}{A}\); with gypsum \((2d = 15.15\ \overset{\circ}{A})\), \(\lambda = 2.6—11.6\ \overset{\circ}{A}\), i.e. two exposures are sufficient for recording wavelengths \(\lambda\) from \(1.0—11.6 \cdot 10^{-8}\ \mathrm{cm}\). (The duration of an exposure is about 2 hours.)

For rapid orientation, instead of directly measuring spectrograms, one may use a special “key”: the principal lines of the \(K\), \(L\), and \(M\) groups of elements are plotted on a scale, uniformly divided into degree divisions, at the corresponding divisions. The spectrogram under study is placed against such a scale, and by the coincidence of its lines with the lines of the scale it is determined which elements are present in the substance being analyzed. By regulating the voltage of the tube, one can eliminate the lines of shorter wavelengths, i.e., spectra of higher order, and thereby simplify the analysis of the spectrogram.

The principle of the method described is a valuable addition to existing methods of X-ray spectroscopy and puts into one’s hands a new and interesting method of chemical analysis, possessing the particular advantage that no combinations of elements can affect the result of determining the minimal quantities of a substance.

A. Trapeznikov.