US2967012A

US2967012A - Getter-ion pump

Info

Publication number: US2967012A
Application number: US652952A
Authority: US
Inventors: Richard J Connor
Original assignee: High Voltage Engineering Corp
Current assignee: High Voltage Engineering Corp
Priority date: 1957-04-15
Filing date: 1957-04-15
Publication date: 1961-01-03
Anticipated expiration: 1978-01-03

Description

GETTER-ION PUMP Richard J. Connor, Medway, Mass, assignor to High Voltage Engineering Corporation, Burlington, Mass, a corporation of Massachusetts Filed Apr. 15, 1957, Ser. No. 652,952

3 Claims. (Cl. 230-69) This application is a continuation-in-part of my copending application, Serial Number 440,137, filed June 29, 1954, now US. Patent No. 2,796,555.

This invention relates to high-vacuum pumps and in particular to a getter-ion pump which comprehends a vacuum container, in which there is provided a source for the production of gas ions and excited gas molecules, and spaced from said source but connected thereto by a wide aperture means for the evaporation of getter-metal. Said vacuum container may itself be the volume to be pumped or it may be in communication therewith.

'Many proposals for producing a high vacuum have been made in which a getter metal is evaporated, while at the same time gas atoms are ionized or excited in the same space. All these pumps require a very large expenditure of mechanical or electric power, either for the getter-source, which must be furnished with a mechanical feed, or for the ion source, if this is in an external magnetic field.

The present invention concerns a surprisingly simple arrangement and stems from the surprising discovery that it is not at all necessary to excite the gas, which is to be pumped, in the same place as that in whiclrthe getter metal is evaporated; and makes use of the additional experience that it is not even necessary to produce metal vapor continuously, but that it suffices to keep the getter metal at an elevated working temperature corresponding to the gas which is to be pumped, provided that a gas excitation or ionization takes place at some place in the same container. However, it is important that there be no substantial narrowing of the cross-section of the tube or other aperture between the ion source and the getter region, since the cross-section and also the length of the connection between them has quite a substantial effect on the pumping speed.

It is surprising, that with the simple arrangement according to the invention, even noble gases are taken up.

A major use of high-vacuum pumps is in connection with the evacuation of those regions of charged-particle accelerators wherein charged particles are accelerated to high energy. Such regions must be highly evacuated in order to minimize collisions of the charged particles being accelerated with gas molecules or ions, since such collisions tend to prevent the individual charged particles from attaining the high energy desired.

At the present time, acceleration tubes are commonly evacuated and a high degree of vacuum maintained by means of diffusion pumps in conjunction with cold traps. A new type of high-vacuum pump to replace such diffusion pumps and cold traps has been developed at the University of Wisconsin. In the Wisconsin pump, the system to be evacuated is first partially evacuated by a fore pump, and the system is then sealed off. Titanium in wire form is evaporated continuously inside a chamber which forms a part of the sealed-off system. The gettering action of the titanium, which is thus continuously deposited on the inner wall of the chamber, results in a pumping action, and a high vacuum may be 57,012 P atented Jan. 3, 1961 obtained in this manner. A brief description of the Wisconsin pump appears in volume 89 (second series) of The Physical Review at page 897 (issue of February 15, 1953).

Major advantages of the continuous-gettering-type pump are its simplicity and the fact that it makes possible the construction of an acceleration tube as a sealedoff system.

Generally pumps of this type must include means for producing ionization within the system to be evacuated, in order to remove substances, such as argon and helium, which do not otherwise readily respond to the gettering action.

In the drawing:

Fig. l is a diagram illustrating the essential components of my improved high-vacuum pump;

Fig. 2 is a view illustrating in detail novel means for evaporating the getter material; and

Fig. 3 is a somewhat diagrammatic view in vertical cross-section of a preferred embodiment of my invention, and illustrates the incorporation of a vacuum gauge in my improved high-vacuum pump; and

Fig. 4 is a somewhat diagrammatic view in vertical cross-section of a closed-off multiple-electrode acceleratio-n tube incorporating my invention.

Referring to the drawings, and first to Fig. 1 thereof, the essential components of my improved high-vacuum pump are enclosed within a chamber 1 which is connected to the system 2 to be evacuated by a suitable length of tubing 3. A second length of tubing 4 connects the volume, enclosed by the chamber 1 and the system 2 to be evacuated, to a fore-pump 5. After the system 2 has been partially evacuated by the fore-pump 5, it is closed off from the atmosphere by a suitable valve 6.

Further evacuation is then accomplished by the gettering action of a suitable material, preferably titanium, assisted by suitable means for ionizing at least some of the residual gases. Any suitable ionization means may be employed, such as a conventional ionization gauge, and the ionization means is therefore indicated merely diagrammatically at 7.

In accordance with my invention, a tungsten filament 8 is supported within a chamber 1 by conductive members 9 which are connected to a suitable voltage source 10 by leads 11. The tungsten filament 8 is covered by titanium, which may be deposited thereon by an electrolytic plating process or in any other convenient manner. A simple manner of obtaining the required titanium covering is to overwind the tungsten filament 8 with small-diameter titanium wire, as indicated at 12 in Fig. 1. The tungsten filament 8, overwound with titanium wire 12, is illustrated in detail in Fig. 2.

The high-vacuum pumplng action is commenced merely by setting the ionization means '7 in operation, and by turning on the voltage source it), so that sufficient electric current flows through the tungsten filament 8 to heat the same and cause evaporation of titanium from the titanium wire 12. The titanium vapor is deposited on the walls of the chamber 1 and, by a gettering action, pumps the residual gases from the system to be evacuated.

Tests conducted by me have indicated that a continuousgettering pump constructed as hereinbefore described gives excellent performance, particularly in the evacuation of acceleration tubes of the type shown in Fig. 4 and including a multiplicity of electrodes 21, separated by insulators 22 and cemented thereto by an organic plastic cement 25. In one such test, an acceleration tube having 31 electrodes separated by insulators %-lHCh thick and sealed thereto by vinylseal cement, and including a metal tube extension 18 inches long, could be evacuated at will to a pressure of about l 10- mm. Hg solely by means of the high-vacuum pump herein described and claimed. After 1000 hours the apparatus had to be dismantled for other reasons, and the pump was still operating satisfactorily at that time. During the 1000 hours, the acceleration tube was periodically operated at 1 megavolt with an electron beam current of 250 microamperes for a total of 60 hours.

The acceleration tube just described evolves gases at a rate of about 10* mm.-liters per second when the tube is not in operation, and at a rate of about 10 mm.- liters per second when the tube is in operation. Hence, during the 1000 hours of operation of the acceleration tube, the pump absorbed a total of more than 2 atmosphere-cc. of gases.

A suitable forepump can evacuate such an acceleration tube to about mm. Hg in about minutes. After this preliminary evacuation, the system may be closed off and the forepump disconnected. Referring to Fig. l, the voltage source 10 may be turned on for a period of about 10 minutes to 1 hour, and the resultant layer of titanium deposited on the inner walls of the chamber 1 is sufiicient to provide the proper pumping action for about 8 hours operation of the acceleration tube. The ionization means 7, however, must be in operation continuously while the acceleration tube is being operated, and should be turned on at least 3 minutes before the acceleration tube is to be utillzed.

In the aforementioned test, a .045-inch-diameter tungsten filament 3 inches in length was overwound with .OZO-inch-diameter titanium wire (as illustrated in Fig. 2), and the voltage source 10 (Fig. 1) was a transformer with a Varlac, operating at 4 volts and 65 amperes. The ionization means 7 (Fig. 1) was a conventional Penningtype ionization gauge. As is well known, a conventional Penning-type ioniztion gauge includes a source of exciting electrons and a system of electrodes which are arranged in a constant magnetic guide field so that the existing electrons travel in spiral paths.

It will thus be noted that While the gettering action of the titanium acts continuously after the operation of the accelerator tube, it is not necessary to evaporate the titanium continuously. Consequently, very little power is required to produce the desired pumping action, and the necessity for cooling parts of the acceleration tube structure, which would result from prolonged heating of the titanium, is avoided. Moreover, the undesirable increase in the rate of gas evolution from organic plastcs used to cement the components of the acceleration tube, which results from the deccmposftion of complex gas molecules in the presence of heated surfaces within the tube, is also minimized by evaporating the titanium only intermittently.

The rate at which the getter-ion pump absorbs gas from the system to be evacuated may be increased by warming the source of getter material even though the temperature to which the getter material is raised is far short of the evaporization point. The reason for this is that as the getter material is warmed, the rate of diffusion of the gas into the volume of the getter material is increased with the result that the surface of the getter becomes more active in its ability to take up the gas being absorbed. The material which is thus warmed may be either the original source of the getter material or the layer of getter material which has been deposited on the walls of the container as a result of evaporation.

Although the embodiment of the invention in which the getter material is evaporated by coating a filament with the getter material and heating the filament is the preferred embodiment and is claimed in the application of which this is a continuation-in-part, the present invention is not limited to this method of evaporating the getter material but includes other methods of raising the temperature of the getter to the evaporation point, such as heating by induction, heating by electron bombardment within the electron container, or heating by means of a heat source located outside the vacuum container.

It is possible to use a carbon filament in place of the tungsten filament 8 (Fig. 1), but a carbon filament will require about twice as much power from the voltage source 10 as the tungsten filament 8. In further tests conducted by applicant, an acceleration tube was closed off from the atmosphere and, after the initial evacuation thereof, the vacuum of 10- mm. Hg was maintained for a period of 841 hours solely by means of the pump described and claimed herein. The acceleration tube was operated with a l-megavolt 250 microampere electron beam for a total period of 217 hours during the aforementioned 841 hours. During said 841-hour interval, the pump was turned on 6 times, each time for a 20- minute interval. Thus a 20-minute operation of the pump sufiices for a 30-hour operation of the acceleration tube.

Applicant discovered, as a result of his experiments,

that the continuous evaporation of titanium using a graphite crucible requires much greater power to operate than the intermittent evaporation of titanium in accordance with the invention described and claimed herein. It required about 1 kilowatt to heat the graphite crucible, while only Watts is required to heat the tungsten fi.ament. Moreover, the graphite crucible is operated continuously, While the filament is operated for only about 6 hours for every 1000 hours of tube operation. Hence the crucible requires 1000 kilowatt-hours per 1000 hours of operation, while the filament requires only 1.4 kilowatt-hours per 1000 hours of operation.

' Owing to the large amount of heat generated by the crucible, extensive water cooling is required in order to prevent overheating of nearby portions of the acceleration tube being evacucted. Such water cooling is unnecessary with rpplicznts invention.

Moreover, the crucible is wasteful of power, since only a small fraction of the heat generated thereby is delivered to the titanium.

As indicated hereinbefore, a major use of applicants invention is in the evacuation of acceleration tubes. Such acceleration tubes commonly comprise a multiplicity of electrodes separated by insu'ators and sealed there o by a suitable cement. Organic plastics which commonly are contained in such cement give off complex gas molecules. When these complex gas molecules approach a heated surface, they decompose, thereby increasing the number of gas particles in the acceleration tube. Since the pressure in the tube depends on the number of gas particles therein, this decomposition results in increased pressure. In general, a heated surface such as that provided by a hot filament increases the rate of gas evolution by a factor of about 10 to I00.

If a heated crucible is used to evaporate titanium, a relatively large hot surface area is present within the tube. In addition, such a crucible is generally heated by bombardment with electrons from a hot filament. Using such a crucible to evaporate titanium continuously, a large hot surface is present in the tube throughout operation of the acceleration tube.

By using applicants invention, not only is the hot surface area reduced by the omission of the crucible, but the decomposition of the complex gas molecules is reduced by the fact that the filament is heated for only a small portion of the operating time ofthe acceleration tube.

Titanium is by far the best material to provide the necessary gettering action, although it would be possible to use a suitable filament 9 (Fig. 1) overwound with zirconium or magnesium wire 12.

As hereinbefore stated, any suitable device may be employed as the ionization means 7 (Fig. 1). However, a very simple arrangement is illustrated in Fig. 3, wherein a hot cathode, comprising a tungsten filament i3 energized by a suitable voltage source 14, emits electrons which are attracted to a tantalum plate 15, which is maintained at a positive potential with respect tothe cathode by a plate-voltage source 16. The electron flow from the cathode 13 to the plate 15 provides the necessary ionization. The tungsten filament 8 overwound with titanium wire 12 may then be positioned as shown in Fig. 3. The entire high-vacuum pump may thus be enclosed within the chamber 1, which may comprise, for example, a 25-mrn. diameter Pyrex glass bulb.

Moreover, the arrangement of Fig. 3 may also serve as an ionization gauge, in addition to its function as a high-vacuum pump. To that end, the plate 15 may be perforated, as indicated by the multiplicity of apertures 17, so that many of the electrons attracted from the cathode 13 to the plate 15 continue to travel beyond the plate 15 and towards the tungsten filament 8 overwound with titanium 12. If a small negative voltage is applied to the tungsten-titanium filament 8, as by a battery 18, then any positive ions produced by the electrons after traveling through the apertures 17 will be collected by the tungsten-titanium filament 8. By means of two microammeters 19, 20, one 19 of which measures the current between the cathode 13 and the plate 15, and the other 20 of which measures the ion current collected by the tungsten-titanium filament 8, the pressure within the chamber 1 may be measured.

The voltages supplied by the plate voltage source 16 and the battery 18 are not critical. Merely by way of example, the plate 15 may be maintained at +200 volts and the tungsten-titanium filament 8 at -22 volts with respect to the cathode 13.

While some suitable ionization means is necessary, it is not necessary that this ionization means be operated during operation of the acceleration tube. While it is necessary that ionization occur during operation of the acceleration tube, at least some of this ionization is always provided by the charged particles accelerated by the acceleration tube during the operation thereof. However, the separate ionization means must be opoperation of the acceleration tube.

erated before operation of the acceleration tube if the acceleration tube has not been in operation for an appreciable time. Moreover, unless the beam of charged particles itself provides sufficient ionization, the additional ionization means must also be operated during Hence, the ionization means must always be operated during at least some of the time that the acceleration tube is closed off from the surrounding atmosphere.

Having thus described the preferred construction of a high-vacuum pump in accordance with my invention, together with several modifications thereof, including the incorporation of a vacuum gauge in such a highvacuum pump, it is to be understood that although specific terms are employed, they are used in a generic and descriptive sense and not for purposes of limitation, the scope of the invention being set forth in the following claims.

I claim:

1. A getter-ion pump for producing a very high vacuum below 10' mm. Hg, comprising a pump casing having a first chamber and a second chamber, an ion source supported within said first chamber, and a directly heated vaporizer source for getter metals supported within said second chamber, a gas conduit interconnecting said chambers, said conduit having sufiicient length so that the electric field associated with said ion source is excluded from said second chamber.

2. A getter-ion pump in accordance with claim 1, wherein said ion source includes a source of exciting electrons and a system of electrodes which are arranged in a constant magnetic guide field so that the existing electrons travel in spiral paths.

3. A getter-ion pump in accordance with claim 1, wherein said ion source includes a heated cathode.

Hertzler Apr. 28, 1953 Alpert Dec. 13, 1955