Full Text
MASS SPECTROMETER WITH TIME SWEEP
The reviewed articles describe two mass spectrometers with time sweep. One of them is intended for the separation of heavy isotopes, for which the usual methods give a significantly lower relative resolving power (this method is also suitable for instantaneous scanning of the entire mass spectrum); the other method is for the analysis of mixtures whose composition changes rapidly.
The first separation method is based on the constancy of the period of rotation of a charge moving in a homogeneous magnetic field:
$$ T=\frac{2\pi}{(e/m)H}=655\frac{M}{H}\ \text{microseconds}, \tag{1} $$
where \(T\) is the period of rotation of the charge, which in the general case moves along a spiral whose pitch is \(v_{\parallel}T\), where \(v_{\parallel}\) is the component of the velocity along the field; \(\frac{e}{m}\) is the charge-to-mass ratio; \(H\) is the magnitude of the magnetic field in gauss; \(M\) is the mass in atomic units.
As is seen from equation (1), \(T\) does not depend on the velocity or on the angle at which the charged particle enters the magnetic field; therefore, for particles with the same ratio \(\frac{e}{m}\), \(T\) is constant.
In the spectrometer that has been constructed, the magnetic field is directed along the tube; the ions, admitted into this field as a diverging beam, move along spirals, being focused after every \(360^\circ\). By the collector, ions that have made the same number of revolutions are collected. Since the period of rotation increases linearly with mass, the distances between pulses on the screen of an oscilloscope connected through an amplifier to the collector will be proportional to the ion masses. This gives high resolving power for heavy ions, for which the methods used up to now gave reduced accuracies. The distances between two adjacent mass peaks in a field of 100 gauss are approximately equal to 7 microseconds. By applying a pulsed deflecting electric field, it is not difficult to measure this distance with an accuracy of up to \(\frac{1}{10}\) microsecond.
The proposed method of isotope separation imposes no rigid requirements on the monochromaticity of the ions, on the axiality of their beam, or on the size of the slit of the ion source. All this greatly increases the luminosity of the apparatus. At the same time, the presence of instantaneous scanning substantially simplifies the amplification of the ion currents. The slit of the source may have any shape, and its linear dimensions are limited only by the radius and homogeneity of the magnetic field. If \(R_0\) is the radius of the field and \(R\) is the maximum radius of the ion trajectory, then the half-width of the ion beam may be:
\[ \frac{R_0 - R}{R} \]
radians.
Reasonable dimensions of the apparatus and the attainable magnitude of the vacuum impose limitations on the radius of the spiral, and thereby also on the magnitude of the magnetic field and on the ion velocity. The magnitude of the magnetic field, bounded above by the time of rotation of the ions and below by the magnitude of the radius, must be of the order of 100 gauss. The ion velocity must be less than 100 eV.
The authors point out that applying this method to the determination of absolute mass values, as well as to the measurement of relative values of large mass differences, will be very difficult; the measurement of small mass differences of heavy isotopes is more promising. In doing so, they recommend collecting ions that have made several revolutions.
One cannot agree with the last remark, since it is not only unjustified by anything, but, on the one hand, leads to an increase in the already considerable length of the spectrometer tube, and, on the other hand, to the impossibility of the above-mentioned measurements. Measurement of the masses of the isotopes of the entire spectrum becomes possible if the ions are collected after the first revolution, since then ions with multiple mass ratios that have made different numbers of revolutions will not simultaneously fall on the collector. In order to obtain the entire spectrum simultaneously on the oscilloscope screen, the oscilloscope sweep should be transferred from one line to several, the number of which is determined by the ratio of the flight times of the iso-
larger mass to the length (in microseconds) of the sweep trace at the operating sweep frequency.
In the second spectrometer the time sweep is carried out on the basis of the relation
\[ \frac{m v^{2}}{2}=eV, \]
whence
\[ m \sim V t^{2}, \tag{2} \]
where \(V\) is the field potential difference, and \(t\) is the time.
The shortcomings of this method are obvious. But because of its simplicity the arrangement can find application where only qualitative measurements are needed.
The working part of the spectrometer tube was \(317\) cm. The vacuum, \(\sim 10^{-5}\) mm Hg, was maintained by a pair of two-stage mercury pumps with a trap filled with liquid nitrogen. The accelerating-field voltage (320–480 V) was applied from batteries with a divider. The pulse at the deflecting plates was 5 μsec, but the observed width of the pulse on the oscilloscope screen was 20–30 μsec. This pulse width included about 10 mass units in the region of light masses and \(>50\) in the region of heavy masses. The authors suppose that the reason for this was either insufficient monochromaticity of the ions, or the fact that the ions travel paths of different length. However, as they state, this device was applied with some success for determining impurities in volatile halides.
B. A. Shulyak
References Cited
- S. A. Goodsmit, Phys. Rev. 74, 622 (1948).
- A. E. Cameron, D. F. Eggers, Rev. of Sciv Instr. 19, No. 9 (1948).
- W. Stephens, Phys. Rev. 69, 619 (1946).