On the Manufacture of a Sealed-Off Vacuum Chamber for a 15 MeV Betatron

B. N. Gulko

(Presented by the scientific seminar of the Faculty of Radio Engineering)

In connection with the broad use of betatrons in various branches of science and technology, much attention is being devoted to simplifying their design, improving ease of operation, and ensuring stable performance.

One important improvement is the replacement of continuously pumped vacuum accelerating chambers by sealed-off chambers, which do not require bulky vacuum systems and fore-vacuum and high-vacuum pumps for their operation. When sealed-off vacuum chambers are used, the betatron can be prepared for operation within several minutes, whereas with an ordinary chamber about one to one and a half hours are spent preparing the betatron for operation, as required to attain the necessary degree of vacuum. Sealed-off vacuum chambers make it possible to use electron guns with oxide cathodes.

Although in the recent literature there are references to the operation of such chambers [1], the technology of their manufacture has not been published anywhere.

The aim of the present work is to develop a manufacturing technology for sealed-off chambers and to produce experimental samples.

For the manufacture of sealed-off chambers, glass toroidal chambers for a 15 MeV betatron were used; these had been made under factory conditions by blowing individual halves into a mold, which were then fused into a whole by means of glass-blowing operations.

In the course of manufacturing sealed-off chambers, a technology for their fabrication was developed which, like the technology for any large electrovacuum device, consists of the technology of preparatory operations, the technology of sealing, pumping, and conditioning of the sealed-off chambers.

Technology of Preparatory Operations

The preparatory operations include the creation of a reliable electrical contact between the inner conducting layer of the chamber and the molybdenum lead-in, which is usually sealed into the tub-

used for evacuating the chamber. For this purpose, two molybdenum rods made of wire 0.8–1.0 mm in diameter are sealed into a section of glass tube 15–20 cm long and 2.5–3.0 cm in diameter. These lead-in rods must protrude 3–4 mm into the tube, so that subsequently a sufficiently reliable contact with the conducting layer can be obtained.

To create such a contact, metallization with silver is used on the inner part of the molybdenum rod and on some part of the inner surface of the glass tube. The strongest adhesion to the base is obtained when a layer of silver is applied by the firing-in method. The essence of this method is that readily decomposing silver carbonate, applied in the form of a paste, is firmly joined on heating to the surface layer of the base, aided by a special flux added in small quantity [2].

However, molybdenum does not combine directly with silver. Therefore, before the silver is fired in, the molybdenum lead-ins are coated with nickel, which is used as an intermediate layer between them. The nickel is applied by an electrolytic method.

The glass tube prepared in this way is sealed by glassblowing operations to the glass branch pipe of the chamber intended for its evacuation. The next stage is the application to the inner surface of the chamber and the evacuation tube of a conducting coating, which in our experiments was made of aquadag. The technology for applying aquadag does not differ in principle from that adopted in the electrovacuum industry [3].

The initial fixing of the conducting layer is carried out by blowing hot air into the chamber immediately after the operation of applying the layer. Further drying and fixing of the layer on the surface are carried out in an electric furnace at a temperature of 400–450°C. In this process the organic compounds of the aquadag solution decompose and the graphite layer is firmly fixed on the glass surface.

The electron gun is mounted separately on a glass stem of the usual type used for 15 MeV chambers. Before being sealed into the chamber, the electron gun was subjected to special vacuum treatment, which consisted in preliminary degassing of its anode. Preliminary degassing of the gun anode is a mandatory operation, since the removal of a large mass of gas from the anode material cannot be carried out by heating it when the gun is sealed to the chamber. The design of the latter is such that the sealed-in gun cannot be heated to a high temperature during evacuation by high-frequency currents.

In this connection, the gun is first sealed into a glass tube of the same diameter as its stem. The free end of the tube is drawn out and provided with a stem. The gun was thus sealed into a cylindrical bulb, over which a cylindrical inductor made of copper busbar had previously been fitted. The sealed-in electron gun was connected to a vacuum system evacuated by high-vacuum and roughing pumps. Upon reaching a vacuum of the order of \(10^{-5}\)–\(10^{-6}\) mm Hg, the gun anode was heated by means of a high-frequency generator to a light-red heat; in this process a large quantity of gases sorbed by the anode metal was released. After the main mass of gas had been removed, the bulb with the gun was sealed off from the apparatus and stored in this form until the operation of sealing it into the chamber.

Naturally, the described method of preliminary degassing of the guns cannot be profitable under conditions of serial manufacture of sealed-off chambers. In this case it should be recommended to degas the anodes of the guns under a vacuum bell by means of high-frequency currents.

The preparatory operations also include preparation of the gas absorber (getter). As the chemical absorber of residual gases we used a very intensive barium–titanium absorber, known under the name “bati,” pressed in the form of tablets. Its advantage over other evaporated absorbers, including aluminum–barium ones, consists in the fact that the titanium additive, at the moment of evaporation, evaporates earlier than the barium and gives a dispersed layer, increasing the absorption surface. The barium is then deposited on top. The mirror obtained when such an absorber is evaporated is dispersed; its color is almost black.

The absorber shelf, on which the “bati” tablets are fastened, is made of nickel strip 0.1 mm thick in the form of a ring whose outside diameter is 20 mm and inside diameter is 8 mm. On the surface of the shelf, around the circumference above and below, 12–14 absorber tablets are attached. Each tablet is covered on top by a strip of perforated nickel ribbon. The strips are welded to the shelf by contact spot electric welding. To fasten the shelf with the absorbers in the chamber pumping tube, six tantalum spring strips are welded to it, three on each side (Fig. 1).

Fig. 1. Gas-absorber shelf

Technology of the Operation of Sealing and Tip-Off of the Chamber

After all preparatory operations have been completed, the operation of sealing the electron gun into the chamber follows. Sealing is carried out by “butt fusing” the wide tube of the chamber to the base of the electron-gun stem. Owing to the impossibility of ensuring uniform rotation of the device during this operation, one resorts to “fusing in sections,” when the resulting seam is fused successively, section by section. In this case, as a rule, the seal obtained is uneven, with local thickenings and gaps. Sealing in the electron gun is further complicated by the fact that often the diameter of the gun base is smaller than the diameter of the chamber branch pipe intended for the gun. Before sealing, therefore, it is necessary to draw down the end of this branch pipe and bring its diameter to that of the base.

When sealing the electron gun into the chamber, it is very important to position its electrodes correctly. The tungsten cathode of the gun must be in a strictly vertical position at a distance of 15.8–15.9 cm from the axis of the instrument. If this condition is not observed, the conditions for injecting electrons into the equilibrium orbit may be disturbed. Therefore, before soldering the electron gun into the chamber, the place for cutting off the gun base and the drawn-out end of the chamber tube is carefully marked each time. The cutting is carried out as carefully as possible with the aid of a heated piece of chromonickel wire. With correct cutting of the base and the end of the tube, the base and tube, when joined end to end, provide exactly the required insertion of the gun into the chamber. After cutting, the glassblower heats the butted ends of the tubes with a soft flame from a gas burner. By gradually increasing the air supply to the burner, the flame is made sharper and hotter, the glass softens, and the edges of the tubes are welded together. Then, using ordinary glassblowing techniques, the seam is fused through and smoothed out.

It is very important during this operation to protect the chamber from excessive overheating, since the residual internal stresses in it, as well as the temporary stresses that arise, can cause it to crack. In this connection it is useful to wrap the chamber near the sealing point with asbestos tape. To reduce the stress in the seal itself, it is absolutely necessary to cool the sealing point gradually. To monitor the vacuum in the sealed-off chamber we used an ionization manometer. For this purpose, a baseless manometer tube LM-2 was soldered to the base of the electron gun above or below the sealing point. Fig. 2 shows the stages of the sealing operations.

Manual sealing of the electron gun into the chamber has a significant drawback. Even when this operation is carried out carefully, one cannot be certain that the electron gun has been installed with complete accuracy. When the seam is being fused, the glassblower inflates and upsets the glass; the latter may lead to the gun being displaced either toward the axis of the instrument or outward by several millimeters. The gun may also rotate about its axis, and the base may become distorted. In series manufacture of instruments, a sealing machine must be used for this purpose. Only in this case can one be sure that the gun is seated correctly.

After the operation of sealing in the gun comes the operation of installing the getter shelf in the chamber pumping tube. The getter shelf is inserted inside the tube and moved back from the edge of the tube by 6–7 cm, so that the subsequent constriction of the tube on the vacuum line will not noticeably heat the pellets with the flame of the gas burner. The end of the pumping tube is drawn out, and a thick-walled tubulation with a constriction, having an internal diameter of 4–5 mm, is soldered to it.

After the getter shelf has been sealed into the tube, it is pulled slightly outward with the aid of a wire hook so that it is located at a distance of 6–7 cm from the sealed-in molybdenum leads. This must be done in order to prevent cracking of the lead-in seals during degassing of the getter shelf and evaporation of the getter, when this assembly is heated almost to 1000°C. The tantalum springs, which bear against the glass from inside, firmly hold the getter shelf in the tube. This completes the sealing operation, and the chamber is immediately put on the pump. For this purpose it is sealed to the vacuum system of the pumping station.

Pumping of the chamber should be begun immediately after the sealing-off operations, since the presence of a large amount of moisture after glass-blowing work may lead to peeling of the conducting coating from the surface of the glass.

Fig. 2. Stages of the sealing-off operations
a—the chamber prepared for sealing off; b—the chamber tube drawn down to the diameter of the gun base; c—the chamber with the gun soldered on; d—the chamber with the gun and an LM-2 lamp soldered on.

The pumping station used for pumping the chambers consisted of a metal frame, in the lower part of which was located the vacuum system with high-vacuum and fore-vacuum pumps. The vacuum system was made of grade 3S-5 molybdenum glass. For ease of manipulation, a large glass stopcock was provided in the high-vacuum section. The fore-vacuum pump was an RVN-20 pump, and the high-vacuum pump was a TsVL-100.

In the upper part of the apparatus there was an electric furnace in the form of a flat horizontal cabinet, made of iron 0.5 mm thick. One chamber to be pumped could be placed in the furnace in a horizontal position. On the bottom of the furnace, on an asbestos-cement backing, three chromium-nickel spirals were fastened, connected in parallel to the electric mains. The power of the furnace heater was about three kilowatts. Inside the furnace, to improve thermal insulation, an asbestos lining 1 cm thick was made. On the outside the furnace was painted with aluminum paint in order to reduce radiation losses.

When the heater was switched on, temperatures of 480–500°C could be reached in the furnace after one hour. Temperature control was carried out by means of a chromel–copel thermocouple and a compensating direct-current potentiometer.

The pumping operation begins with removal of the main mass of air from the chamber. Upon reaching a vacuum of \(10^{-5}\)—\(10^{-6}\) mm Hg, vacuum treatment of the device is carried out, consisting in degassing, under pumping, the glass envelope of the chamber and the internal metal parts.

To remove water vapor and certain gases sorbed by the surface of the glass and by the conducting coating, the chamber was heated in a furnace to a temperature of 450—480°C. In this process, at a temperature of 300—320°C, maximum gas evolution was observed; after that it decreased, and at a temperature of 400—450°C it again increased owing to the release of gases from the inner layers of the glass during reactions between the glass components. For 0.5—1 hour the chamber was held at a temperature of 450—480°C, and then was slowly cooled over 1.5—2.0 hours. If, at a temperature of 300—350°C, the pressure inside the chamber and the vacuum system was \(1—2 \cdot 10^{-6}\) mm Hg, it was considered that degassing had been carried out quite satisfactorily. Sometimes, when the coating layer was too thick (application of aquadag twice), it was necessary to carry out a second heating, but with a shorter hold at high temperature. After the chamber had cooled to 100°C, the furnace was opened and the internal metal parts were degassed.

First of all, the metal parts of the manometric tube soldered to the chamber were subjected to heating. Heating of the grid and the tungsten cathode was achieved by passing current directly through these electrodes. For this purpose a laboratory autotransformer and a step-down transformer were used, making it possible to increase the heating of the parts slowly. Heating was continued until the gases were completely removed, which was monitored by means of an ionization gauge on the vacuum installation.

The filament of the gun was degassed in the same way; a current of up to 3 amperes was passed through it; due to heat radiation from the filament, the control electrode of the gun was also degassed.

Upon reaching a pressure of \(1 \cdot 10^{-6}\) mm Hg, degassing was stopped.

Degassing of the grid and cathode of the manometric tube takes 5—6 minutes. Degassing of the gun filament and control electrode requires more time—about 30 minutes.

Degassing of the anode (ion collector) of the tube and of the getter shelf was carried out by heating with high-frequency currents. For this purpose a high-frequency generator with a power of 1 kW was used. The inductor, 5 cm in diameter, was made of a copper strip and consisted of 6 turns.

Heating of the nickel anode of the tube to a light-red heat was carried out for 8—10 minutes. After the pressure had fallen to \(1 \cdot 10^{-6}\) mm Hg with the anode heated, degassing was stopped. The most labor-intensive operation was degassing the getter shelf and the “bati” tablets, because they contained a large amount of gases. Heating was carried out gradually, with a slow increase in temperature. The main mass of gas is released at a temperature of 300—400°C, and an increase in pressure by almost two orders of magnitude was often observed. Heating of the getter shelf was carried out with interruptions, since it was necessary to give the pump time to cope with the gases being evolved. Gradually, after 6—7 heatings, the getter shelf was sufficiently degassed and was heated to a light-red heat, i.e., almost to 900—1000°C. At the indicated temperature a prolonged ...

heating was carried out for 5–6 minutes; during this time the getter began to evaporate and was deposited as a dense layer on the glass of the pumping tube. When a thick, opaque layer of metal had been obtained, heating was stopped. The pressure in the vacuum system often reached the 7th order.

After evaporation of the getter, the chamber was immediately sealed off from the vacuum system.

For this purpose the constriction of the tubulation was gradually heated with a hand gas torch. After the main mass of gas had been evolved and removed by the pump, the tubulation was sealed off with the sharp flame of the burner, and the chamber was separated from the vacuum system.

Conditioning of the Chamber

To absorb the residual gases in the chamber, the latter was subjected to conditioning. In this process the vacuum in the chamber improved by one to two orders of magnitude. The conditioning, carried out under laboratory conditions, consisted in switching on the manometric tube soldered to the chamber. The manometric tube made it possible not only to monitor the pressure in the sealed-off chamber, but also to reduce the pressure in it, operating analogously to an electron pump. Ionized gas particles were sorbed by the collector of the tube, and were also partly absorbed by the mirror of evaporated getter.

The first conditioning of the chamber was carried out for 0.5–1 hour. It could then be observed that the pressure in the chamber gradually decreased and by the end of this period became equal to \(4\text{–}5 \cdot 10^{-7}\) mm Hg. Subsequent conditionings were carried out over a shorter interval of time; the pressure decreased considerably faster. As a result of gas absorption by the getter, the upper and lower limits of the pressure during the conditioning processes shifted toward lower pressures.

Fig. 3. Absorption of residual gases as a result of chamber conditioning

Figure 3 shows the change in pressures during the conditioning of one of the chambers.

Subsequently, before operating the chamber in the betatron, it is sufficient to switch on the manometer tube for 3–5 minutes in order to reduce the pressure to \(1\text{–}2 \cdot 10^{-7}\) mm Hg.

Conclusion

The work has shown that the manufacture of chambers for 15 MeV betatrons presents no fundamental difficulties under the conditions of a small laboratory equipped with modern apparatus.

In serial manufacture of such chambers under factory conditions, their cost should not exceed the cost of receiving television tubes.

REFERENCES

1. J. P. Girard and G. D. Adams, Application of the betatron to practical radiography, Electrical Engineering, v. 65, 1946.
2. F. E. Evteev and V. A. Zhukov, Technology of Radio Apparatus, Gosenergoizdat, 1952.
3. V. Espe and M. Knoll, Technology of Electrovacuum Materials, Oborongiz, 1939.