ELECTROSTATIC EXTRACTION OF ELECTRONS FROM A 15 MeV BETATRON

B. A. KONONOV

(Presented by the scientific seminar of the Faculty of Physics and Technology)

A source of high-energy electrons (5–25 MeV) may be a betatron equipped with a device for extracting electrons from the accelerator chamber into the atmosphere. The problem of extraction consists in directing the electrons, which move in the betatron chamber along a circular orbit, at a definite point along the tangent to it. At the Tomsk Polytechnic Institute this problem has been solved, in particular, by deflecting the electrons in the electric field of a capacitor.

At the end of the accelerating cycle the electrons are displaced by a winding that intensifies the flux in the central core of the betatron pole, to the instability boundary, i.e., to the circumference where the electrons leave the region of action of the focusing forces of the magnetic field. The pitch of their spiral trajectory increases sharply and becomes sufficient to pass around the end surface of the inner (cutoff) plate of the deflecting capacitor. After passing through the deflecting field, the electrons are freed from the action of the controlling magnetic field and leave the accelerating chamber into the atmosphere.

Figure 1 shows a horizontal section through the median plane of the accelerator (schematically), where the relative arrangement of the principal components of the chamber for electron extraction can be seen.

Fig. 1. Horizontal section of the vacuum chamber for electron extraction.
1—vacuum chamber, 2—above-orbit injector, 3—deflecting capacitor, 4—exit window, 5—fluorescent screen, 6—ionization chamber or Faraday cylinder, 7—expansion winding, 8—pumping pipe, 9—bolt 1.

Vacuum chamber. The vacuum chamber is made of molybdenum glass 3 mm thick (Fig. 2) and has a branch pipe for the injector, a branch pipe for installing the deflecting capacitor, an exit window, and a branch pipe for pumping. The branch pipe for the deflecting capacitor has an elliptical

Fig. 2. Vacuum chamber

cross-section with a horizontal axis of 80 mm and a vertical one of 50 mm. In fabricating the vacuum chamber, special care is required when sealing on the branch pipes: there must be no runs or protrusions inside the chamber. The outer radius of the chamber is 170 mm, and the radius of the limiting circle is 163 mm; taking into account the thickness of the glass wall, 3 mm, we find that from the limiting circle to the chamber wall there is only 4 mm. If a run of glass into the chamber is allowed, the electron beam may fail to enter the deflecting capacitor, located at a radius of 165 mm, and may be stopped by the glass protruding inward. To eliminate this danger and to create a path for the deflected electron beam, the chamber wall at points 1 and 2 is blown outward by 5 and 10 mm, respectively. The inner surface of the chamber is coated with a thin layer of silver, which is grounded in order to prevent the accumulation of static charge on the glass surface.

Injector. A conventional three-electrode injector (Fig. 3) is installed at the radius where \(n = 1\), but 7 mm above the median plane,

Fig. 3. Superorbital injector.

so that it does not interfere with the emerging electron beam. Mounting the injector on a bellows simplifies tuning the accelerator for maximum radiat—

However, after the optimal position of the injector has been determined, the sylphon becomes unnecessary and should be dispensed with by mounting the injector on a shortened glass base.

Exit window. The exit pipe is sealed with a brass flange (Fig. 4) with seven cells covered by aluminum vacuum-tight foil, 0.03 mm thick. The size of the electron beam at the exit is such that all electrons can pass without loss through one cell. However, for different energies of the accelerated electrons, one or another cell may be required. The beam does not emerge through cells 1, 2, 3, and 4, but they are not superfluous, since they reduce the mass of the brass flange located in a strong magnetic field. Aluminum foil 0.030 mm thick scatters the electron beam only slightly and is sufficiently strong to withstand atmospheric pressure with the cell dimensions indicated in the drawing, provided that the edges of the cells are carefully rounded. The foil is attached to the brass flange with bitumen No. 4. The method of attaching the foil is as follows. A sheet of foil, somewhat larger in size than the brass flange, is laid on a smooth metal plate, which is heated from below. The brass flange is heated, and the surface on which the foil will lie is coated, as far as possible, with a thin layer of bitumen. The flange is then placed on the foil, and the plate is heated from below until the bitumen on the flange melts, after which the flange is pressed tightly against the foil and cooled. With this method of joining, the foil lies in an even layer, without wrinkles, and can operate for a long time without danger of leakage or rupture. When the flange has cooled, the excess foil is cut off, and one can proceed to solder the flange to the exit aperture of the chamber. The method of this joint is similar to the usual joining of two flanges with bitumen: both flanges are heated, coated with bitumen, joined to one another, and the bitumen is melted. The sole condition here is that the cells must be in the middle plane.

Fig. 4. Exit window.

Deflecting capacitor. The deflecting capacitor (Fig. 5) consists of two plates spaced 2.5–3.2 mm apart and mounted on a glass base. The thin intercepting plate is grounded; a voltage of up to 30 kV is applied to the other plate (we shall call it the outer plate, since it is located at a larger radius than the intercepting plate).

The glass base on which the plates are mounted is similar to the base of the injector, but with somewhat enlarged discharge surfaces to reduce the probability of corona discharge.

It is desirable for the intercepting plate to be as thin as possible in order to reduce the loss of electrons on the frontal surface of the electron beam entering the capacitor. However, the plate must meet requirements of mechanical strength, and therefore it cannot be made infinitely thin. In addition, the required electrical strength of the gap between the plates determines the materials from which the plates may be made. In our case the intercepting plate is made of tantalum...

foil 0.05 mm thick, which is welded to a frame of stainless steel. It should be noted that the foil was welded to the frame only after the latter was completely ready, i.e., welded, ground, and drawn out in the proper manner. With this manufacturing procedure the deformation of the foil will be minimal. The plate is mounted on a glass base on two rings insulated from one another, so that the fastening parts and the plate do not form a closed loop in the magnetic field.

Fig. 5. Deflecting capacitor.

The outer plate is made of sheet stainless steel 1 mm thick, whose surface and edges are carefully polished. The vertical size of the plate was chosen after determining the dimensions of the electron beam, on the assumption that the beam should move in a uniform electric field not weakened at the edges of the plates by significant scattering. The plate is mounted on a 3-mm molybdenum electrode fused into the glass base.

Adjustment of the position of the plates does not require much time, since the joints of the plates with the base are movable. The deflecting capacitor is mounted on bellows, and adjustment of the position of the capacitor in the chamber can be carried out either manually or by means of selsyn sensors, whose control can be brought out to the control panel.

Power-supply circuit of the deflecting capacitor. The deflecting capacitor cannot be supplied with a constant voltage, since at the moment of injection almost all electrons would be thrown onto the chamber wall by the strong electric field. Therefore the capacitor power supply (Fig. 6) is provided by a peaking transformer, which gives a voltage pulse of duration about 500 μsec and amplitude up to 50 kV. Such a pulse is applied to the capacitor at the moment of displacement and does not affect the process of electron capture—

Fig. 6. Peaking transformer for 50 kV.

... into acceleration. The pulse transformer has 250 turns in the primary winding and 60,000 in the secondary, and the cross sections of the iron core are, respectively, \(40\ \mathrm{cm}^2\) and \(4\ \mathrm{cm}^2\). The air gap of the magnetic shunt is \(\delta = 2.0\ \mathrm{cm}\).

The transformer is mounted on brackets screwed to the betatron stand, so as to reduce to a minimum the length of the conductors bringing the high voltage to the deflecting capacitor. The beginning of the secondary winding, located at the iron core, is grounded; the outer end of the winding is connected to the central electrode of the capacitor. The pulse transformer is fed from the AC mains through a phase-shifting circuit (Fig. 7), which serves to change the phase of the voltage peak on the deflecting capacitor relative to the displacement pulse.

Fig. 7. Power-supply circuit of the pulse transformer.

By varying the capacitance from 10 to \(90\ \mu\mathrm{F}\), the phase of the peak can be changed within 80 electrical degrees in steps of approximately 10 electrical degrees. Smooth displacement of the phase within one step is provided by a variable resistance of 70 ohms. The magnitude of the voltage on the deflecting capacitor is regulated through an autotransformer. Structurally, the power-supply circuit of the pulse transformer constitutes a separate unit.

The voltage on the deflecting plates is monitored by means of an ohmic voltage divider and an EO-7 oscilloscope, on which one can observe the shape of the voltage pulse applied to the plate, its phase relative to the displacement pulse, and the magnitude of the voltage in kilovolts (after calibration of the voltage divider).

As indicated above, the deflecting capacitor is installed at the radius \(R = 163\ \mathrm{mm}\) (the radius of the limiting circle for this accelerator). With a smooth increase in the voltage on the capacitor plates, extraction of electrons was always observed, recorded by a thin-walled ionization chamber, as shown in Fig. 1. Further adjustment of the deflecting device is required in order to obtain the maximum intensity of the extracted beam. Figure 8 presents the dependence of the electron output on the displacement of the capacitor in the radial direction. Along the ordinate axis the output is plotted in arbitrary units; along the abscissa axis, the displacement of the capacitor \(\Delta r\), in turns of the screw dial (2 turns correspond to \(1\ \mathrm{mm}\) of displacement). The dependence of the output on the angle of rotation of the capacitor is presented in Fig. 9 for the case of the optimal radial position. Along the abscissa axis are plotted the turns of the screw (Fig. 1), which provides rotation of the capacitor.

Fig. 8. Dependence of the yield on the radial position of the capacitor.

Fig. 9. Dependence of the yield on the angle of rotation of the capacitor.

Detection of the electron extraction is carried out both with the aid of a thin-walled ionization chamber, installed at the presumed path of the electrons in air, and with the aid of a fluorescent screen installed near the exit window (Fig. 1). The emerging electron beam produces a bright spot on the screen, which can be observed by means of a periscope directly from the control panel. By moving such a screen away from the exit window, the path of the electron beam can be rapidly determined.

By placing photographic paper (or photographic film) in the path of the electron beam, one can determine the dimensions of the beam and its angle of divergence. At a distance of 2 cm from the exit window, the beam cross section has the form of a rectangle with rounded corners; its dimensions in the horizontal and vertical directions are, respectively, 12 and 7 mm. At a distance of 17 cm from the window, the beam cross section increases to \(35 \times 20\) mm. Hence the divergence angle of the beam is \(8.5^\circ\) in the horizontal plane and \(5^\circ\) in the vertical plane.

The current of the extracted beam is measured by a shielded graphite Faraday cylinder with a galvanometer. A current value of \(4 \cdot 10^{-9}\) A was the largest for this betatron. The intensity of the x-ray radiation near the exit window, measured with a thimble chamber, is about 3500 roentgens per minute. The figures given refer to an electron energy of 5 MeV. At energies of 10 and 13 MeV the focusing of the extracted beam is better, but the current is somewhat smaller. The electron energy was monitored from their range in aluminum.