AVS1997 Session VT-TuP: Assorted Topics in Vacuum Technology

Tuesday, October 21, 1997 5:30 PM in Room Exhibit Hall 1
Tuesday Afternoon

Time Period TuP Sessions | Topic VT Sessions | Time Periods | Topics | AVS1997 Schedule

VT-TuP-1 Measurement and Control of Sticking Probability of H2O on Stainless Steel Surfaces
Y. Shiokawa, M. Ichikawa (Joint Research Center for Atom Technology (JRCAT), Japan)
Sticking probability and residence time of H2O on metal surfaces for a chamber wall are very important in vacuum technology, because the time constant of pumping-down is proportional to their values. The residence time is known to be 104 s1 but the sticking probability is little known.2 We obtained sticking probability of H2O on stainless steel surfaces using molecular beam method3 and tried to control it. A mass filter with differential pumping can only detect molecules ejected directly from a sample surface without detecting molecules scattered from a chamber wall. First, the amount of H2O dose is measured from reflected H2O during its irradiation to a sample surface. Next, the amount of H2O adsorption is measured from released H2O during flash heating the sample. The sticking probability is obtained by the ratio of these amounts, because its residence time is much longer than the measuring time. Samples were polished stainless steels (SUS304) with no treatment and with treatment of vacuum baking(400°C, 20h, 10-5torr of H2O). Moreover, the samples were annealed at different temperatures in UHV condition. After cooling them down to room temperature, the sticking probability was measured. Without annealing, the sticking probability of the no treatment sample was about 1x10-2 and that of the vacuum baking sample was about 8x10-2. Over 600°C of annealing temperature, the sticking probabilities of the both samples drastically decreased and down to 1x10-4 for the no treatment sample after 850°C annealing. These values didn't change after exposing the samples to air in days, but further vacuum baking increased the small probability realized by high temperature annealing. However, the rise of the values by vacuum baking was restrained with H2 supply of 10-3torr during vacuum baking. We also discuss the mechanism of the sticking probability changes. This work partly supported by NEDO, was performed in JRCAT under the joint research agreement between NAIR and ATP.


1M.Morrw: Vacuum,36(1986)523
2Y.Tuzi et al.:Vacuum 47(6-8) (1996)705
3M.Kobayashi et al.:JVST,16(1979)685

VT-TuP-3 Dynamic Time-Of-Flight Mass Spectrometer for the Chemical and Isotope Analysis of Gas Mixtures
A.V. Kozlovsky, S.N. Markovsky (Mettek Ltd., Russia); G.P. Saksagansky (The D.V. Efremov Scientific Research Institute of Electro-Physical Apparatus, Russia); I.L. Fedichkin (Mettek Ltd., Russia)
The proposed development of the automated time-of-flight gas analyzes with the control system, the diagnostic and processing of the information is designed for working both as a separate analytic device and in the composition of various physical plants of the user, in which it is required to carry out the analysis of the composition of gasses, vapors, molecular beams and the simultaneous obtaining, accumulation, storage, processing and display of big scopes of information. The plant's equipment control various units of the user's installations. The Gas analyzes can be used for the study of various physical and chemical processes in the sciences and technics carrying out with gas emission. The gas analyzes consists of the mass analyzes, the vacuum joint and the recording system. The mass analyzes to be implemented according to the ion-optical scheme with the reflection of the ionic beams, ensuring additional focusing of ions on primary energies. The frequency of analyzises in the device is 10,000-20,000 analyzises per second. The spectrum on the monitor can be considered both panoramically (the scan period is 20 and 50 mks), and on separate groups of masses (the scan period is 2 and 10 mks) with the usage of the scan's delay. Two modifications of the device will be performed constructively: The first one consists of the mass analyzer with the recording system and allows the installation in vacuum units having pressure no more then 10 -3 Pa, with the further work in its composition; The second one is the gas analyzer with the vacuum system on the basis of the magnetic - discharge pump patented as "mass-spectrometer for gas analysis".
VT-TuP-4 Reliability of a Spinning Rotor Gauge in Medium Vacuum
M. Hirata, S. Suginuma (Electrotechnical Laboratory, Japan)
The performance of a spinning rotor gauge was measured and analyzed in medium vacuum by using a static expansion system. The gauge is a noble viscosity gauge which is essential to a transfer gauge and a reference gauge in high vacuum. A rotor ball made of steel is magnetically suspended and freely rotated in vacuum. Pressure is calculated from the relative decrement of the rotational speed of the ball due to the drag force of ambient gas. The drag force on the ball shows a tendency to saturate in the medium vacuum, while it is proportional to ambient gas pressure in high vacuum less than 0.1Pa. If the gauge is stable and repeatable, it is also effective as a transfer gauge in the medium vacuum. Argon was poured instantaneously into the chamber, to which a spinning rotor gauge was attached, and then the pressure in the chamber was kept constant for 1.5 hours. The indication of the gauge was not stable unfortunately. It increased gradually over a period longer than one hour. Judging from the instability of the indication, the gauge is not useful in the pressure higher than 15Pa for argon. At the pressure of 100Pa, increase in the indication from the initial value was 35% at the equilibrium and it was 40% higher than the value of the real ambient gas pressure. While the re-acceleration of the rotor was done frequently, change in the offset on the rotational speed was not dominant. Change in the viscosity of ambient gas was dominant, since the correction factor linearizing the indication against the real pressure was sensitive to the value of the viscosity. The temperature of the tube, in which the ball was installed, raised gradually 15K with time, too. Change in the temperature of 15K corresponds to change in the viscosity of 4%. The correction factor was 7 at the pressure. Correcting the value of viscosity for the temperature of the tube, most of the deviation can be compensated.
VT-TuP-5 Design of a Low Cost, Low Magnetic Susceptibility, UHV Compatible, Flanged Electrical Break
M. Mapes, H.C. Hseuh, P. Cameron (Brookhaven National Laboratory)
At Brookhaven National Laboratory, the g-2 experiment, a muon storage ring,1 requires a totally non-magnetic vacuum system. All the vacuum components must have a low magnetic susceptibility. The design of the beam vacuum system requires flanged electrical breaks on pumping manifolds in order to prevent ground loops. The electrical breaks developed at Brookhaven consist of a ceramic disc clamped between two Conflat flanges. A non-conductive retaining ring is used to support and center the ceramic disc and seals with respect to the flanges. Two Helicoflex Delta seals are used to seal the flanges to the ceramic disc which has metallized surfaces in the seal area. Threaded rods which are electrically insulated from the flanges with bushings are used to provide the clamping force necessary to compress the vacuum seals. The assembly remained leak tight after repeated bakeouts up to 200 C. This design offers several advantages over commercially available Conflat flanged ceramic breaks. These breaks do not use Kovar and are totally non-magnetic. In addition to the cost being lower the overall length of the break is much shorter than commercial breaks. The same design concepts can be used for different sized Conflat flange and other types of flanges. The detailed design and the performance of this break are presented in this paper. ext H.C. Hseuh, L.Snydstrup, M. Mapes, and C. Pai, J. Vac. Sci. Technol., A14, 1237(1996).
VT-TuP-6 Edwards Drystar Dry Vacuum Pumps.
M. Firth, A. Patil, S. Hoath, A. Rothwell, M.J. O'Brien, C. Tunna (Edwards High Vacuum International, United Kingdom)
Edwards Drystar dry vacuum pumps have performed repeatedly and with low maintenance needs on harsh metallurgical processes over the last decade. A major part of this success is that the patented dry pump designs have been manufactured directly by Edwards and tested in large numbers across a very wide application and customer base in several key industries. The technological basis for the low maintenance requirements of the arises from the gas packed seal purge for the dry pump, which tends to keep debris and process gases from damaging the shaft seals and entering the gearbox. The Edwards patented reverse claw mechanism ensures short gas paths and good particulate handling, while the pump inlet (high vacuum) bearings are readily accessible and are high vacuum grease lubricated. Variants on the pump design, cooling control and purging details have enabled Drystar to handle process chemical and condensing applications, while a reduced cost and features, the GV Drystar series, is for industrial use. The economic case for GV dry pumping is presented for vacuum metallurgy via a cost of ownership model comparison against competing piston pumping.
VT-TuP-7 Experimental Features of Cryosorptional Panels for Pumping of Fusion Reactors Fuel Mixtures
N.T. Kazakovsky, E.L. Koyra (All-Russian Experimental Physics Institute, Russia); A.P. Kryukov, S.B. Nesterov (Moscow Power Engineering Institute, Russia); G.L. Saksagansky, D.V. Serebrennikov (The D.V.Efremov Scientific Research Institute of Electrophysical Apparatus, Russia)
Features of research facilities for modelling and tritium tests of primary vacuum pumps of the ITER fusion reactor are presented. An ultra-high vacuum facility on base of helium cryosorption pumps is used for study of vacuum and physical characteristics. A cryogenic blok, containing three baths with independent cryoliquids supplement (LHe, LN2) to each bath is placed into the pump. The application of the pump with independent cooling of baths makes it possible to carry out the experiments in different modes. The facility is made for long-term operation with tritium content up to 30 g. Initial pumping speed of the pump is 4 m3/s (H2, T2); 5 m3/s (D2); 2 m3/s (He). Tritium impact on physical and mechanical characteristics and resouces of cryopanels are studied on the other test facility. The facility includes a vacuum system, a tritium generator, a container for samples exposure in tritium atmosphere, a thermostating system of container at cryogenic (LN2) and increased temperatures. Small fragments of cryosorption panels (80x15x1 mm3) put into the container are tested. After pumping a container is filled by tritium and is exposed to 80-300 K. Critical features of sorptional layer are analysed before and after exposition in tritium. The isotherms of protium, deuterium and tritium adsorption were determined for activated carbon chemviron SCII at 78 K. Static and dynamic characteristics of helium isotopes (3He, 4He) cryosorption at the temperature range 2-4 K are presented.
VT-TuP-8 Design and Pumping Characteristics of a New Compact Titanium-Vanadium Non-Evaporable Getter (NEG) Pump
Y. Li, D. Hess, R. Kersevan, N.B. Mistry (Cornell University)
A compact non-evaporable getter (NEG) pump is made by stacking sintered st185 (Ti 70-V 30, by wt%) NEG blades recently available from SAES Getters1. The 50 blades are stacked between copper spacers on a stainless steel tube, and is activated by a removable heater cartridge inserted into the tube. The NEG disks are 0.8mm thick squares, 24mm by 24mm with a 9mm diameter center hole. The room temperature pumping speeds for CO, N2 and H2 and for gas mixtures are measured as a function of sorbed gas quantities at various activation temperatures and with different NEG disk spacing, in order to optimize the pump design and operation. The study shows the pumping speed is almost linearly dependent on the activation temperature from 250 degC to 600 degC. Linear pumping speeds of about 450 liter/s*m and 2200 liter/s*m can be achieved for CO at activation temperatures of 300 degC and 550 degC for 1-hr, respectively. The low activation temperature of the st185 NEG and the compact design make the NEG pump cartridge particularly attractive for situations with tight space and with components sensitive to high temperature. We acknowledge gratefully the loan by SAES Getters of the 50 blades for evaluation.
VT-TuP-9 Multi-Technique Characterization of BATMAN and F82H Martensitic Steels for Fusion Reactors
A.D. Laine, C. Cepek, E. Magnano (Istituto Nazionale per la Fisica della Materia, Italy); G. Gervasini, G. Bonizzoni, M. Sancrotti (Associazione EURATOM/ENEA/CNR, Italy)
Structural materials for future controlled thermonuclear fusion reactors are the subject of exhaustive characterisation studies worldwide. At the European level, the Long Term Programme on Structural Materials for Fusion Reactors has been initiated by EURATOM. Several types of steel which are possible candidates for fusion reactors have been identified within this programme. Martensitic steels are under consideration as possible fusion reactor construction materials because they have very low radiation swelling and very low contents of nickel and molybdenum. Here we report on the results of degassing studies of BATMAN C and F82H along with depth-profiling x-ray photoemission measurements of new and aged samples of F82H and the results of model calculations for hydrogen isotope transport in MANET II (MArtensitic for NET). The temperature and time dependent degassing measurements of the steels BATMAN C and F82H indicate that BATMAN C has the superior degassing properties. In F82H, a large amount of the hydrogen remains trapped. This finding is consistent with the results of the model calculation for MANET II. This high hydrogen concentration is close to the critical concentration for the hydrogen induced cracking of the material which counts against its use in possible fusion reactors. The photoelectron spectroscopy results show that the surface elemental composition of F82H is much richer in chrominum than the bulk material and independent of the ageing procedure. The thickness of the enriched zone, is not large enough to effect hydrogen surface recombination rates and thus need not be considered in calculating the overall hydrogen transport rates.
VT-TuP-10 Clean Machining for Aluminum Alloy Materials using Corona Discharge Generating Ozone
H. Ishimaru, M. Nishiwaki (KEK, Japan)
To make clean surface for ultrahigh vacuum systems, clean machining with corona discharge generating ozone is used. Any aluminum alloy surface as received is contaminated with oil and water. Many years ago, EX process was developed to make clean surfaces for aluminum alloy. EX process means machining inside the environment of dry argon and oxygen. Work must be installed inside the vacuum chamber. After that contamination was removed by machining from the aluminun alloy surfaces. EX process was successfully against extremely high vacuum and fast pump-down process systems. But, EX process is complicated, cost high and not practical use. To simplify for clean machining with low-cost using corona discharge assisted in oxygen and argon mixture are developed. Corona discharge can generate ozone. Ozone is very active to react with carbon contamination, and to make clean oxide layer quickly with thin, high density & non-porous, and no-contamination. Clean surface was analyzed by Auger electron spectroscopy (AES). Carbon depth profile in AES was reduced. Next step, we will be measure thermal outgassing rate, dynamic outgassing from electron stimulated desorption, and residual gas analysis for 99.99% high purity aluminum material. This clean machining process is useful to manufacture a beam chamber of synchrotron radiation ring with weld assembly and a semiconductor process chamber. H.Ishimaru: Clean machning for aluminum alloy materials using corona discharge generating ozone, Contamination-Its measurementa and control in vacuum systems, A workshop for semiconductor fabrication and accelerator technology, Jefferson Laboratory, March 26-27, 1997
VT-TuP-11 Bakeable All Metal Burst Diaphragm for UHV Applications.
R.J. Todd, G.T. McIntyre, L.A. Smart (Brookhaven National Laboratory)
The Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory is comprised of superconducting magnets which encompass a cold-bore UHV beam line. Numerous bellows associated with the beam line must be protected from excessive internal pressures in the event of a cryogenic helium leak. A bakeable, all metal burst diaphragm for low pressure venting was developed for this purpose. The device incorporates a bellows which, under pressure, drives a piece of copper foil into a stationary blade. The spring constant of the bellows allows for a puncturing pressure of 8 psi, while complete rupture occurs at 15 psi. The copper foil can be replaced at minimal cost, and the device can be reliably baked at temperatures up to 300 C.
VT-TuP-12 Substrate Delivery System among Vacuum Instruments in an Extreme High Vacuum with Magnetic Levitation Transports
M. Tosa, A. Kasahara, K. Yoshihara (NRIM, Japan)
Extreme high vacuum less than 10-10 Pa (XHV) can result in no surface contamination by adsorption of gasses and offer an ideal ultraclean environment which lasts a long time to artificially synthesize advanced materials with atomic manipulation. An XHV integrated process with magnetic levitation transports has been developed in order to transfer substrates long distances from one vacuum chamber with instrument to another without any contamination on the ultraclean substrate surface. The process consists of vacuum chambers, pumps, gauges, gate valves, transports for main line and sidetrack line and so on. The integrated process has five main line chambers and six sidetrack line chambers with connected six chambers with instruments. A main line has a connection chamber to join another main line chamber and a sidetrack line chamber which can attach chambers with such vacuum instruments as surface analyses, film preparation and so on. Magnetic levitation transports are installed into the line chambers because they have no sliding part to generate dust particles and outgassing which may much damage ultraclean substrate surfaces and environment. One transport for a main line electromagnetically levitates a carrier and transfers it by a linear motor drive. The other transport for a sidetrack line levitates a carrier by YBa2Cu3O7-x superconducting magnet discs and mechanically transfers it by means of pinning effect. The levitation transports can deliver a substrate from one conncted chamber to another in the pressure change of less than 10-10Pa.
VT-TuP-13 Computer Modeling of the Effects of Grids in Ion Optics Using SIMION 3D
C.W. Baker, S.M. Colby (Scientific Instrument Services, Inc.)
Overview: It has been know for some time that the non-homogeneous fields produced at the small openings within grids can be an important factor in ion optics. We report a new technique for the simulation of ion scattering at grids. This tool is a macro for use with the ion optics program SIMION 3D v.6.0 [1]. Specific examples involving TOFMS are presented. Simulation: The simulation of grid scattering is accomplished through the use of SIMION's "user programming interface". When a ion passes through a simulated grid the ion is "jumped" to a region in which a small section of grid is modeled. Monte-Carlo methods are used to position the ion in relation to the grid opening and the program considers the appropriate electric fields. In this manner one piece of simulated grid can be used for all grid transitions found within the instrument and cumulative effects can be observed. Results: Our initial results showed that grids could have a considerable effect on the ion trajectories [2]. Effects were most pronounced when an ion was decelerated after passing through a grid. In these cases accelerations resulting from field non-homogeneity at grids could result in a significant fraction of the total velocity. We examined the cumulative effects of grid scattering in a Reflectron TOFMS. We assumed 70 and 333 line per inch (lpi) grids at 2000 and 10,000 eV. (100 AMU ion). Full details are available on the web site listed below. With 70 lpi grids, a surprisingly large fraction (93 %) never reached the detector. The results for 333 lpi grids were better, only 80% missed the detector. This is despite reported transmissions of 90 and 70% for the 70 and 333 lpi grids respectively. Results were independent of drift energy. 1. David A. Dahl 43ed ASMS 1995, 717. 2. S. M. Colby; C. W. Baker; J. J. Manura , 44th ASMS 1996. (Available at www.sisweb.com appl. note #47) Text of the full poster, an example simulation, and all related macros for this simulation are available at http://www.sisweb.com .
Time Period TuP Sessions | Topic VT Sessions | Time Periods | Topics | AVS1997 Schedule