US2897036A - Method of evacuation - Google Patents

Description

' July 28, 1959 Filed April 18, 195'.

A. J. GALE El AL METHOD OF EVACUATION 2 Sheets-Sheet l Jul 28, 1959 Filed April 18, 1957 A. J. GALE ET AL 2,897,036

METHOD OF EVACUATION 2 Shee ts-Shet 2 /2 5 g 3 I: 22 T v #5 l i 7 v HM I 035mm l Po v 04 m 9 1 3 8 E 4 E 1 5 7 l I m T 6. 2 ,0: 1 m E I 5 95 -6 ABSOPPT/OA/ CHARAC'EP/ST/C OF GETTEE/lO/V PUM F01? DPYA/Q 54/471550 TIME M/A l/TES METHOD OF'EYACUATION Application April 18, D57, Serial No. 653,606

6 Claims. c1. -316--25) This invention relates to methods of processing and pumping systems which evolve limited amounts of gas and to the in situ' production of low pressures at temperatures at which the usual diffusion pump-fluids no longer operate. The invention makes useof getterion pumps and is particularly adapted to the process pumping of multiple electrode acceleration tubes. Howeventhe invention is not limited to the evacuationof such tubes but in'cludesthe evacuation of other closed systems such as the larger partially demountable power tubes. Throughout the specification and claims the term getter ion pump means a device in which the. pumping action is provided by a getter material which is vaporized in conjunction with means for ionizingor otherwiseex-citing the residual gas in the system being evacuated.

X-ray tubes for supervoltage radiography and radiation therapy at l-million volts pose design problems not present in lower voltage tubes. This is particularly true when adaptation to relatively small transportable equipment demands compactness and, therefore, operation of the tube at high electrical gradients. It has been a matter of some concern to operators of supervoltage X-ray equipment that they are faced with the daily round of icing traps and the common task of warming these same traps periodically. Component replacement exposes operating personnel to unfamiliar vacuum techniques usually without adequate leak detecting tools. The existence of the diffusion vacuum system requires continuous monitoring of the residual pressure together with automatic interlocks designed to secure or prevent operation of the machine if this pressure exceeds a certain value. These electrics are an added complexity. Mercury diffusion pumping systems require continuous cooling which is generally obtained by connection to water supplies. Transportable units can only be made with bulky and probably relatively unreliable cooling systems. Only a moderate degree of flexibility has been achieved in the conventional designs. One result of the use of diffusion pumped systems is that pressures inside the acceleration tube are in' order of magnitude or more higher than obtained in the fused sealed tubes common in the radio and allied industries. This higher pressure apparently has been having an adverse effect on cathodelife-the most unreliable component in most supervoltage X-ray generators at this date.

A sealed tube should therefore give at least'the following advantages: (1) greater overall convenience to the operator; (2) simpler designs and possibly lower overall costs; (3) increased flexibility in equipments and a larger range of users; and, most important of them all, (4) improved reliability, particularly'of the cathode.

The experience gained with the cold sealed tube is valuable no't o-nly for sealed X-ray tubes but also for the design of scaled ion systems. It will in due course have a major impact on electron processing equipments and an influence on the pumping systems of non-sealed ion accelerators.

"ice

The invention, comprehends not only a completely sealed system but also one in which one or more gaskets are used. .However, a completely sealed system has the advantage that it does not present field personnel with the difficulties associated with any form of gaskets or the procedures which would be necessary in getting a green system onthe air.

The invention may best be understood from the following detailed description thereof, having reference to the accompanying drawings in which:

Fig. 1 is a diagrammatic view of mobile one-million volt X-ray equipment having an acceleration tube which has been fabricated in accordance with theinvention;

Fig. 2 is a side elevation of the tube of the apparatus shown in Fig. 11; V

Fig. 3 is a somewhat diagrammatic view of the acceleration tube of Fig. 2 in combination with apparatus for carrying out the method of invention;

Fig. 4 is a diagram showing apparatus for the investigation of getter ion pumping; and

Fig. '5 is a graph showing getterion pumping characteristics for dry air.

Referring to the drawings, a typicaLequipment used for radiography of steel specimens up to 5 inches thick is shown in Fig. 1. The radiographic unit 1 is mounted on a hydraulicly operated jackstacker 2 powered by storage batteries and is, therefore, fully mobile. The tubef3 of this unit is shown in Fig. 2. It consists of a multiple sandwich of insulators 4 and electrodes 5 together with a cathode 6 ofthe conventional filamentary type and a hollow anode 7 which is some 4 feet long and at thebottom of which is a high atomic number target 8. Currentpractice on pumped electron tubes is the use of a relatively heavy tantalum filament. Although it is felt that the lower residual pressure existent in the cold sealed tube will permit smaller diameter filaments or perhaps other cathode materials, the heavier tantalum filaments may be retained if desired. Deep hollow anodes are required for radiography of circumferential welds of pressure vessels and similar equipments. Except for mechanical modifications to accord with the one piece construction, the targets are functionally the same as those in use on current pumped designs. .X- radiation outputs will therefore be identical with pumped models.

The insulating section of the one-million volt tube v3 shown inFig; 2 is 21 inches long, and about 30 insulators 4 each take an approximately equal increment of electrical stress. Unfortunately, insulator and electrode materials having compatible thermoco-efficients. o-fexpansion do not always have the best electrical characteristics at high gradients. Accordingly, a relatively low temperature fabrication technique has been evolved which permits selection of insulator and electrode materials for optimum electrical properties. These tubes are constructed by cementing the electrode/insulator multiple sandwich together with a thermosetting resina process which we have termed cold sealing because at no timedo the seal temperatures reach a value high enough to give significant outgassing of the material. Current designs of pumped tubes use aluminum electrodes .ce

' mented to glass insulators with vinyl acetate polymer.

3 temperature cured (375 400 F.) epoxy resin marketed originally by the Ciba Company as Araldite AN-l and now as Hysol 2501 by Houghton Laboratories. The gas evolution rate appears to be at least three orders of magnitude less than vinyl acetate, and the cured material has a high mechanical strength.

An alternative material has been selected for the insulatora porcelain AI-200 manufactured by the Coors Porcelain Company. In addition to its ability to Withstand electron bombardment, this material will withstand 200 kilovolts per inchsome individual samples going to as high as 400 kilovolts per inch. The tensile strength of AI-200 is in excess of 25,000 p.s.i. The yield strength of aluminum 52 SO electrodes is 14,000 p.s.i. During cooling after curing the seals at about 200 C., the differential contraction causes the aluminum to yield at a point 11,000 p.s.i. or more lower than the yield point of the insulator. The gas which is more or less continuously evolved from the tube structure during life must be absorbed by some means.

Some work started prior to 1950 had shown the abnormal ability of barium to absorb gas when an ionizing discharge was present, but later work has tended to use titanium, which is stable to the atmosphere at room temperature and whose compounds have extremely low vapor pressures.

Referring again to Fig. 2, the appendage at the side is a titanium/ion gas absorber or pump 9 which consists of a commercial Penning ion gauge 10 and one or more tungsten-filaments 11 overwound with titanium wire 12 and enclosed in a vaporization chamber 13 which is shown in detail in Fig. 4. Results have shown that this appendage is somewhat overdesigned and later models of high voltage tubes have the pump more aesthetically designed into the tube diameter. This component, in conjunction with the use of low gas evolution materials, has been the main factor permitting the development of the cold sealed tube. Basically the pump comprises a chemically active getter surface and a region in which residual gas molecules are ionized. Under these circumstances, all gases including the inactive rare group are immobilized in the getter surface. The getter surface is obtained on the inner surface of the wall of the chamber 13 by evaporating titanium from the overwind ing 12 on the tungsten heater 11, in the manner described and claimed in the co-pending application of Richard J. Connor, Serial No. 440,137, filed June 29, 1954, now Patent No. 2,796,555, issued June 18, 1957. Titanium is chosen because, unlike the conventional getter materials (e.g. barium) used in the radio industry, it will stand exposure to atmosphere at room temperature and has high chemical activity. Also its compounds have a uniformly low vapor pressure. The ionization region is obtained inside the Penning type gauge 10.

Two designs of getter-ion pump for X-ray tubes will now be described.

, In the case of one tube, the getter-ion pump is an external occasional water-cooled appendage and uses a con ventional Penning gauge modified to the extent of removing the gasket and rescaling the head with epoxy resin cement. the anode extension of the tube through a Kovar bobbin (to relieve stress during evaporation of the getter) and a porcelain insulator.

Other tubes use a getter-ion pump integral with the anode drift tube. A Penning gauge of much smaller dimension has been designed for operation at greater than 400 p.s.i. external pressure. It is expected that in due time the latter design will replace the former. Tubes may be made ready for duty by exhausting with a fore vacuum pump to a pressure about 10- mms. Hg and The pump body is connected to a T on v about 1% of the atmosphere, is being pumped. In the absence of an ionizing discharge the pressure drops slightly and then gradually rises after the first getter evaporation, but subsequent flashes give little further change. The pumping speed of the getter/ion system for the rare gases is considerably less than that for the chemically active gases and a tube of the type we are describing can be successfully checked for leaks by surrounding with helium and checking for a pressure rise followed by a return to base pressure when the helium is removed. The method appears comparable in sensitivity with the mass-spectrometer leak detector in a practical set-up.

The apparently anomalous pumping of the rare gases can hardly be accounted for by the usual explanation that the atoms are ionized and then accelerated and embedded in the getter. In pumps constructed according to the invention ionization regions are used which do not see the getter surface, and we have confirmed that pumping of the rare gases exist even when an atom must make at least four contacts with the walls before reaching the getter surface. It would appear that it is the excited state of the atom that is important. We know that these states are relatively long lived. It seems possible that the'excited rare gas atom can have physical properties enabling it to diffuse into a metal rather like hydrogen but once it returns to the ground state, diffusion is no longer possible and the atom becomes trapped.

Based on the experience of date the following fabrication procedure is used. Components are cleaned in the usual manner, and the epoxy resin coating is then mechanically sprinkled on the insulators only, preheated to C. Current designs require the cementing of the cathode sub assembly which is afterward cemented into the complete tube. This is not good design practice and the cathode will receive soldered seals as soon as they are developed. Meanwhile, a partial cure is carried out by baking the sub assembly for /2 hour at 375 F.

The main assembly is probably the most critical opera- I tion in the fabrication process. Of utmost importance is faultless jigging. Off center spots can to some extent be tolerated in gasketed systems and allowed for by slight elbowing at the gaskets. Cold sealed tubes must be aligned and remain in alignment throughout the curing process; This is particularly important for the electrodes and insulators at the cathode end of the tube in some cases and for the whole tube in other cases. In curing the tube, recommended treatment is 2 hours at 380390 F. Uniformity of temperature is important.

In accordance with the invention, the initial outgassing of the tube components is absorbed in an auxiliary high vacuum pump. The titanium/ion combination is ideal for this since the limited capacity both in total gas to be pumped and pump speed is not important. Fig. 3 shows the tube 3 of Fig. 2 being processed in this manner. No attention to the tube is required other than the occasional evaporation of getter in the auxiliary pump 14.

The following procedure is carried out according to the invention:

(1) Connect the cured tube assembly 3 to a processing getter/ion pump 14.

(2) Connect a fore pump 15 to the process pump 14 and reduce pressure to normal fore-vacuum pressure. Liquid nitrogen trapping is desirable, as shown at 16.

(3) Elevate the temperature of the process pump filaments to just below evaporation temperature by applying an appropriate voltage across the filament terminals 17. This will drive off absorbed gas which, in the case of getter assemblies not previously used, will be considerable.

4. Repeat step 3 for the filaments of the main pump 9. I

(5) Close off the process pump 14 from the fore pump 15 and evaporate titanium getter from the process pump filament. For example, in one embodiment 4 volts A.C.

asezcse was applied across the filament terminals 17 for 'ten minutes to evaporate getter. The process pump gauge '18 should be on continuously during the pump down procedure. The tube gauge need be usedonly-intermittently to check the pump down rate of the tube. During this pump down-period'the process pump pressureshould be kept at one-tenth that of the tubeor less. The purposes of this pump down period are to conserve the tube getter filaments, minimize possible contamination of the tube Penning gauge, to prepare the tube for storage or for use without frequent getter evaporation during the early part of its life, and by periodic checks on both gauges during the pump down interval to determine whether the rate of pump down is followingthat of known good tubes. The pump down interval can be reduced by putting a larger-load onthe tube getter/ion pump, but in the light of experience with the earliest two tubes which were not separately processed this seems not to be serious.

(6) Seal 01f tube. The technique of cold sealing the copper pumping line 19 requires that the point of seal be allowed to float during the sealing'process. No pressure should be permitted-other than the direct-thrust of the sealing heads. All designs use a pinched-off gas free copper tube.

(7) Evaporate tube getter by applying an appropriate voltage across the main filament terminals 20. The tube pressure should reduce to less than '10 -mm. Hg.

Tubes may be stored for limited periods-without the tube Penning gauge operating and-without evaporation of getter. A well-processed tube can evidently be stored for as long as 40 days without'operation of the Penning gauge. It is recommended that until :further information is collected that the tubes be stored with the Penning gauge operating and that getterbeevaporated when the pressure reading reaches 10- mm. Hg.

The fabrication techniques associated with the cold sealed tube may be modified without departing from the .spirit and scope of the invention. For example, process pumping may be performedat elevated temperature, which is not possible with difiusion pumps'Similarly, process pumping may be performed during the resin cure cycle. 'Outgasssing by ionization may also be performed, in conjunctionn with a suitable pump, such as a titanium getter/ion pump. These items are aimed toward longer storage life without Penning gauge operation and quicker processing.

The method of processing and pumping described and claimed herein appears to have wide application to 'sys tems which evolve limited amounts of gas and to the in situ production of low pressures at temperatures at which the usual diffusion pump fluids no longer operate. 'The usefulness of pumping the larger, partially demountable, power tubes by this means has been discussed with a number of organizations, and in some cases getter/ion pumps of the type described have been used for experimental work. As an example, one of these units has been used to maintain a pressure about -1()' mm. Hg in a multiplate current monitor in a microwave linear accelerator.

In order to determine the pumping characteristics of such systems, the arrangement of Fig. 4 has beenconstructed. The pump 9 and reservoir 21have equal volumes-about 700 cm. After-roughly outgassing'the pump 9 and reservoir 21 by heating the filaments 11 to just below the evaporation point with the whole system connected to a high vacuum pump (not shown), the right hand valve 22 is then closed and getter material evaporated according to an arbitrarily chosen schedule for the time and filament current. With the middle valve 23 closed, gas is fed into the reservoir21 to a predetermined pressure and this volume of gasis then shared with the pump 9 and the middle valver23 again closed. Time/- pressure curves are taken and the procedure thenvrepeated until the getter is exhausted.

Fig. 5 shows the results/for 9 runs using dry' air and for an initial pressure of 70 microns in the reservoir 21, or 3'5 microns when shared :with the pump 9. Readings are only taken over the full portions of the curves and the 'direction'backto the initial pressure .035 mm. Hg has I been arbitrarily indicated.

The curves are obviously divided into three parts. The final leveling of the-curve corresponds to equilibrium between the getter take-up rate and the outgassing of'the pump structure since,'if left alone long enough, the pressure will eventually rise. The slopes of the central portions of the curves are all equal and constant, and in this region, the pumping rate (about '1"6 cm. seekis'controlled by the Penning gauge 10-in other words, the absorptive capacity of the titanium is greater than the rate-at whichthe Penningngaugecan supplysuitable particles to the surface. The initial portions of the curves have slopes which depend on the previous history of the pump. The generally lower pumping speed in this region may in part be attributed to the Penning gauge currents levelingofi (due-to the associated circuit) for pressures in excess of IO- mms. Hg but in addition to this, the progressive delays in reaching the central portions of the curves must be attributed to gradual saturation of the getter, since a fresh evaporation will return the pumping characteristic to curve #1.

Pumping speed varies for different gases and Table I shows results for a few gases.

TABLE -I Pumping speed of getter/ion system for various gases Air 16cm. sec? Oxygen 28 cmfflsecr Nitrogen 18 cm. sec." Carbon dioxide 24 cm. sec? Helium 9 cm? sec? The curves'from whichthese pumping speeds were deduced all showed similar progressively increasing delays as those for dry air. Helium was no exception. Hydrogenispumped so rapidly that results cannot be plotted, and in this case, chemical pumping presumably predominates.

The apparatus of Fig. '4 has beenvused to show pumping by a gas discharge alone. The getter was evaporated with'the middle valve 23 open and the equipment allowed to pump down over a period of two or three days. A base pressure close to the extinction of Penning gauges 10, 24 was obtained. The middle valve '23 was closed, and dry air admitted'to the reservoir 21 until a pressure of 10- mms. Hg was read on the gauge 24. After about 2 minutes, the pressure'had been reduced to 10' mm.

Hg. The procedure of admitting dry air until the reservoir'pressure was 10 mms. Hg and then allowing a re duction to 10 mms. Hg was repeated 46 times without opening-the middle'valve 23 and apparently would have continued since the pumping speed for the 46th run was the same as that forthe first run-25 cms. sec. This-value was somewhat higher than that shown for dry air in Table I and is attributable presumably to differences inthe Penning gauges 10, 24 and their asso ciated circuits. It is concluded from this result that not only does the ion discharge act onitinerant gas molecules and'bring them to a level of ionization or excitation suitable for absorption in the getter, but that it also outgasses its own structure when connected to a system having a low basepressure andthis outgassed structure can then behave as a pump at somewhat higher pressures. The toal amount of gas absorbed in this series of runs was about ,30 cmr mm. Hg.

Epeedpumps can,,of course, be TnZdfiy incorporating .theionization region within the surface of the evaporated material, or by increasing the Penning 7 gauge gas-handling capabilities as by increasing the Penning gauge current. p

Tests which have been carried out to date include the following:

, Tube #1.--Tube comprising a separate getter/ ion pump metal gasketed to the tube assembly. After some preliminary experiences which included focusing adjustments, cathode changes and renewals of getter filaments, a total of 13 getter evaporations were made. The final one sufiiced for months including 300 hours of radiographic operation at which time the unit was stripped for examination since further prolongation of the test would not give significant information.

Tube #2.Tube which was unsatisfactory jigged during assembly thus giving an off center spot. The getter evaporation schedule appeared to be following that of tube #1 at the time of its removal to test tube #3.

Tube #3.-This tube was processed with an external titanium pump and the tubes own titanium/tungsten filaments were outgassed prior to seal off. After seal off and first getter evaporation, this tube was left without the Penning gauge operating for 40 days at which time it was installed in a generator and getter again evaporated. After a further 30 days including 90 hours operation, getter was again evaporated. This tube has now been sealed off 130 days and operated 150 hours. Indicated pressure in this tube is less than 10- mm. Hg.

Tube #4.Thirty-three inch tube. Imperfectly cured seals prevented tests beyond 1.7 mev.

T ube #5.-Forty-five inch tube. This tube was operated for short intervals in excess of 2.25 mev. and appeared to condition in a manner similar to the l mev. tubes. An excessive leak particularly under pressure caused the tests to be discontinued.

In addition to the tests on tubes, the following additional tests have been made:

Getter evaporation-Single getter filaments have been repeatedly operated 120 times before failure. This represents several years shelf life and several thousand hours operation of a l-mev. tube even without allowing for the decrease in gas evolution taking place during tube life.

Strength test.-An epoxy resin cemented assembly comprising simulated anode and cathode terminations, three porcelain insulators and two 52 SO aluminum electrodes was subjected to a moment in excess of 1350 lbs. feet without failure.

Shock test.The same assembly withstood immersion in alcohol cooled to Dry Ice temperature. Immersion in liquid nitrogen caused fracture of the porcelain cemented to the stainless steel and pieces. The aluminum electrodes were further cold drawn by the contractlon.

Penning gauge.The pressure at which these cold sealed tubes operate indicate that the commercial Penning gauge should have several years of useful life. Design of the small Penning gauge used with 2-mev. cold sealed tubes is aimed at preventing contamination and the expected life of the gauges under operating condition is currently being evaluated.

Having thus described the method of the invention together with several illustrative embodiments of apparatus for carrying out the method, 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.

We claim:

l. A method of process pumping a closed system which evolves limited amounts of gas, which method comprises the following steps: (1) connecting the closed system to a first getter-ion pump, (2) connecting the closed system to a second getter-ion pump, each of said getter-ion pumps including getter-evaporation means and ionization means, (3) connectinga fore pump to said second getterion pump and reducing pressure to normal fore-vacuum pressure, (4) heating the getter elements of said second getter-ion pump, but not to evaporation temperature, (5) heating the getter elements of said first getter-ion pump, but not to evaporation temperature, (6) closing ofi said second getter-ion pump from said fore pump, (7) evaporating at least part of the getter in said second getter-ion pump, (8) closing oif the connection between said second getter-ion pump and said closed system, and (9) evaporating at least part of the getter in said first getter-ion pump, at least one of said ionization means being operative throughout the process.

2. A method of process pumping a cold-sealed multiple-electrode acceleration tube wherein a multiplicity of alternating insulators and electrodes are cemented together by a resin, which method comprises the following steps: (1) connecting the assembled acceleration tube to a first getter-ion pump, (2) connecting the assembled acceleration tube to a second getter-ion pump by a connection which can be sealed-ofi by pinching, each of said getter-ion pumps including getter-evaporation means and ionization means, (3) connecting a fore pump to said second getter-ion pump and reducing pressure to normal fore-vacuum pressure, (4) heating the getter elements of said second getter-ion pump, but not to evaporation temperature, (5) heating the getter elements of said first getter-ion pumps, but not to evaporation temperature, (6) closing ofi said second getter-ion pump from said fore pump, (7) evaporating at least part of the getter in said second getter-ion pump, (8) pinching off the connection between said second getter-ion pump and said assembled acceleration tube, and (9) evaporating at least part of the getter in said first getter-ion pump, at least one of said ionization means being operative throughout the process.

3. A method of process pumping a cold-sealed multipleelectrode acceleration tube wherein a multiplicity of alternating insulators and electrodes are cemented together by a resin, which method comprises the following steps: (1) connecting the assembled acceleration tube to a first getter-ion pump, (2) connecting the assembled acceleration tube to a second getter-ion pump by a connection which can be sealed-off by pinching, each of said getterion pumps including getter-evaporation means, comprising a heatable filament about which getter material is supported, and ionization means, (3) connecting a fore pump to said second getter-ion pump and reducing pressure to normal fore-vacuum pressure, (4) heating the filament of said second getter-ion pump, but not to getterevaporation temperature, (5) heating the filament of said first getter-ion pump, but not to getter-evaporation temperature, (6) closing off said second getter-ion pump from said fore pump, (7) evaporating at least part of the getter from the filament of said second getter-ion pump, (8) pinching off the connection between said second getter-ion pump and said assembled acceleration tube, and (9) evaporating at least part of the getter from the filament of said first getter-ion pump, at least one of said ionization means being operative throughout the process.

4. A method of producing low pressures in situ in a closed system which evolves limited amounts of gas, which method comprises the following steps: (1) connecting the closed system to a first and a second getterion pump, each of said getter-ion pumps including getterevaporation means and ionization means, (2) connecting a fore pump to said second getter-ion pump and reducing pressure to normal fore-vacuum pressure, (3) heating the getter elements of said first and second getter-ion pumps, but not to evaporation temperature, (4) closing off said second getter-ion pump from said fore pump, (5) evaporating at least part of the getter in said second getterion pump, (6) closing oif the connection between said second getter-ion pump and said closed system, and (7) evaporating at least part of the getter in said first getterion pump, at least one of said ionization means being operative throughout the process, the closed system being at elevated temperatures at which the usual diffusion pump fluids no longer operate during at least part of the process.

5. A method of process pumping a. cold-sealed multipleelectrode acceleration tube wherein a multiplicity of alternating insulators and electrodes are cemented together by a resin, which method comprises baking the assembled acceleration tube so as to cure the resin and, during the curing process, performing the following steps: (1) connecting the assembled acceleration tube to a first getter-ion pump, (2) connecting the assembled acceleration tube to a second getter-ion pump by a connection which can be sealed-01f by pinching, each of said getterion pumps including getter-evaporation means and ionization means, (3) connecting a fore pump to said second getter-ion pump and reducing pressure to normal fore-vacuum pressure, (4) heating the getter elements of said second getter-ion pump, but not to evaporation temperature, (5) heating the getter elements of said first getter-ion pumps, but not to evaporation temperature, (6) closing ofii said second getter-ion pump from said fore pump, (7) evaporating at least part of the getter in said second getter-ion pump, (8) pinching off the connection between said second getter-ion pump and said assembled acceleration tube, and (9) evaporating at least part of the getter in said first getter-ion pump, at least one of said ionization means :being operative throughout the process.

6. A method 0? outgassing a closed system which evolves limited amounts of gas, which method comprises heating the closed system to produce evolution of gas therein and, while the temperature of the closed system is thus elevated, the following steps: (1) connecting the closed system to a first getter-ion pump, (2) connecting the closed system to a second getter-ion pump, each of said getter-ion pumps including getter-evaporation means and ionization means, (3) connecting a fore pump to said second getter-ion pump and reducing pressure to normal fore-vacuum pressure, (4) heating the getter elements of said second getter-ion pump, but not to evaporation temperature, (5) heating the getter elements of said first getter-ion pump, but not to evaporation temperature, (6) closing 01f said second getter-ion pump from said fore pump, 7 evaporating at least part of the :getter in said second getter-ion pump, (8) closing ofi the connection between said second getter-ion pump and said closed system, and (9) evaporating at least part of the getter in said first getter-ion pump, at least one of said ionization means being operative throughout the process.

References Cited in the file of this patent UNITED STATES PATENTS 2,636,664 Hertzler Apr. 28, 1953 2,715,993 Batina Aug. 23, 1955 2,755,014 Westendorp et a1. July 17, 1956

Images