Full Text
Ionization Gauges for Ultra-Low Pressures
For measuring pressures in vacuum systems and installations, gauges are used whose operation is based on various physical principles.
In particular, for measuring ultra-low pressures, the use of an ionization gauge appears very promising.
An ionization gauge of conventional design[^1] consists of a concentrically arranged heated cathode, a positively charged grid, and a negatively charged collector surrounding the entire system. Electrons emitted by the cathode, before reaching the grid, ionize the gas molecules present in the gauge. The resulting positive ions are collected on the collector.
The theory of the ionization gauge[^2] gives a linear dependence between the ion current to the collector and the pressure in the gauge, provided that the emission current is constant:
\[ I_i = kp, \tag{1} \]
where \(I_i\) is the ion current, \(p\) is the pressure, and \(k\) is a constant depending on the design and electrical operating conditions of the gauge.
This dependence is confirmed by experimental data[^3], obtained in the pressure range from \(3 \cdot 10^{-6}\) to \(1 \cdot 10^{-3}\) mm Hg.
Under ordinary electrical operating conditions (grid voltage from 150 to 250 V, collector voltage from \(-15\) to \(-40\) V), the emission current is from 1 to 10 mA, and the constant \(k \simeq 0.1\ \dfrac{\mathrm{a}}{\mathrm{mm\ Hg}\cdot \mathrm{cm}}\). Thus, even such a low pressure as \(10^{-10}\) mm Hg should correspond to a current of the order of \(10^{-11}\) A, if one assumes that the linear dependence (1) is not violated. Such a current can be measured without particular difficulty with the aid of modern direct-current amplifiers.
However, experiments carried out in this direction[^1] have shown that with an ordinary ionization gauge it is practically impossible to measure pressures below \(10^{-8}\) mm Hg.
The reason for this is apparently the “residual” collector current, caused by electron emission from the collector of the gauge under the action—
by the soft X-ray radiation incident on it, arising upon the slowing down of the electrons emitted by the cathode, in the material of the grid. Since, with an ionization manometer, which measures the current in the collector circuit, it is impossible to distinguish the emission of electrons from the arrival of positive ions, the magnitude of this “residual” current determines the lower limit of measurement of the ionization manometer.
Recently, descriptions have appeared in the literature of three designs of ionization manometers in which the magnitude of the “residual” current is considerably reduced.
The first of them4 is shown in Fig. 1 ($F$—cathode, $G$—grid, $K_1$ and $K_2$—collectors). Whereas the collector of ordinary ionization manometers completely surrounds the grid and absorbs practically all of its X-ray radiation, on collector $K_2$ of this manometer there falls approximately
\[ \frac{1}{500} \]
of the grid radiation, and on collector $K_1$—one half. The configuration of the electric field in the manometer is such that collector $K_2$ absorbs approximately
\[ \frac{1}{5} \]
of the ions formed in the manometer and, consequently, the ratio of the ionic current of this collector to the photoemission current from it is 50 times greater than for collector $K_1$. Removal of the plate of collector $K_2$ does not affect the sensitivity of the manometer, but reduces the photoemission current from this collector still more strongly.
When the pressure is reduced below $10^{-6}$ mm Hg, the current to collector $K_2$ continues to decrease, whereas the current of collector $K_1$ remains constant. The minimum current to collector $K_1$ was equal to $2 \cdot 10^{-9}$ a, which corresponds to a pressure of $2 \cdot 10^{-8}$ mm Hg. The minimum measured current to collector $K_2$ was equal to $2 \cdot 10^{-12}$ a, which corresponds to a pressure of $1 \cdot 10^{-10}$ mm Hg. Thus it can be assumed that this manometer is suitable for measuring pressures of the order of $10^{-10}$ mm Hg.
The device of the second manometer5, capable of measuring pressures down to $10^{-10}$ mm Hg, is shown in Fig. 2. Inside its cylindrical grid $G$, along its axis, a thin metal wire $K$ is stretched, serving as the ion collector; cathodes $F_1$ and $F_2$ are located outside the grid. Thanks to such a device, only a small fraction of the X-ray radiation of the grid falls on the collector, which leads to a considerable reduction of the photoemission current.
The lower limit of the pressures measured with the aid of the manometer described may be taken as $1 \cdot 10^{-10}$ mm Hg.
The sensitivity of this manometer is approximately the same as that of an ordinary one. It is higher than that of the preceding one, since the positive grid $G$ creates a potential barrier and prevents the possibility of ions escaping from the volume enclosed within the grid. Therefore all ions formed inside the grid of the manometer go to the collector.
It should be noted that the form of the grid characteristics of the manometer described at various pressures testifies in favor of the assumption that the lower limit of pressures measured with the aid of ionization manometers is determined by the photoemission current from the collector.
In both manometers described above, the lowering of the limit of pressures accessible for measurement is achieved by reducing the surface of the collector and, consequently, by reducing the fraction of the X-ray radiation of the grid incident on the collector.
The device of the third manometer6 (Fig. 3) is based on an entirely different principle of suppressing the “residual” current of the collector.
The manometer is a four-electrode system with cathode $F$, anode $A$, cylindrical suppressor $S$, and a collector $K$ coaxial with it. The cathode, made in the form of a spiral of thoriated tungsten, is located in immediate proximity to the slot of the suppressor.
If the same negative (with respect to the cathode) potential is applied to the collector and to the manometer suppressor, then the collector current will consist of the ion current and the current of photoemission of electrons from it. If, however, the suppressor potential is made lower than the collector potential, then the electrons emitted from the collector will be returned back to the collector by the retarding field of the suppressor, and the measuring instrument in the collector circuit will register only the ion current.
It was precisely in such a regime \((U_F = 0, U_A = 200\ \text{V}, U_S = -40\ \text{V}, U_K = -2\ \text{V})\) that this manometer was used to measure ultra-low pressures.
The shape of the suppressor was such that radiation from the grid could not fall on its inner surface. Otherwise, photoelectrons emitted by the suppressor would go to the collector and would distort the measurement.
A study of the energy distribution of the photoelectrons emitted by the collector showed that among them there are practically no electrons
Fig. 1.
Fig. 2.
Fig. 3.
with energy greater than \(30\ \text{eV}\), and \(85\%\) of the electrons have an energy less than \(10\ \text{eV}\).
Thus, a retarding suppressor potential equal to \(-38\ \text{V}\) ensures practically complete absence of photoemission from the manometer collector.
A drawback of this manometer is its low sensitivity, due to the fact that most of the ions go to the suppressor, and only a smaller part of them (about \(20\%\)) reaches the collector.
However, if the lower limit of measurement of three-electrode ionization manometers cannot be made much lower than that of manometers of the first two described constructions, because the surface area of the collector, though reduced, is nevertheless finite, then the four-electrode manometer permits further improvement, since so far no causes are apparent that limit the degree of suppression of photoemission of electrons from the collector.
Л. Х.
References Cited
- S. Dushman, Scientific Foundations of Vacuum Technique, I.L., Moscow, 1950.
- N. D. Morgulis, ZhTF 3, 1276 (1933).
- S. Dushman, Phys. Rev. 17, 7 (1921).
- J. J. Lander, Rev. Sci. Instr. 21, 672 (1950).
- T. Bayard and D. Alpert, Rev. Sci. Instr. 21, 571 (1950).
- G. H. Metson, Brit. Journ. Appl. Phys. 2, 47 (1951).