vacuum technology

Introduction

      all processes and physical measurements carried out under conditions of below-normal atmospheric pressure. A process or physical measurement is generally performed in a vacuum for one of the following reasons: (1) to remove the constituents of the atmosphere that could cause a physical or chemical reaction during the process (materials processing) (e.g., vacuum melting of reactive metals such as titanium); (2) to disturb an equilibrium condition that exists at normal room conditions, such as the removal of occluded or dissolved gas or volatile liquid from the bulk of material (e.g., degassing of oils, freeze-drying) or desorption of gas from surfaces (e.g., the cleanup of microwave tubes and linear accelerators during manufacture); (3) to extend the distance that a particle must travel before it collides with another, thereby helping the particles in a process to move without collision between source and target (examples of uses are in vacuum coating, particle accelerators, television picture tubes); (4) to reduce the number of molecular impacts per second, thus reducing chances of contamination of surfaces prepared in vacuum (useful in clean-surface studies).

      For any vacuum process a limiting parameter for the maximum permissible pressure can be defined. It can be the number of molecules per unit volume (reasons 1 and 2), the mean free path (reason 3), or the time required to form a monolayer (reason 4).

      At room temperature and normal atmospheric pressure, 1 cubic foot (0.03 cubic m) of air contains approximately 7 × 10²³ molecules moving in random directions and at speeds of around 1,000 miles per hour (1,600 kilometres per hour). The momentum exchange imparted to the walls is equal to a force of 14.7 pounds for every square inch of wall area. This atmospheric pressure can be expressed in a number of units, but until relatively recently it was commonly expressed in terms of the weight of a column of mercury of unit cross section and 760 mm high. Thus, one standard atmosphere equals 760 mm Hg, but to avoid the anomaly of equating apparently different units, a term, torr, has been postulated; one standard atmosphere = 760 torr (1 torr = 1 mm Hg). This term was replaced in 1971 by an SI unit defined as the newton per square metre (N/m²) and called the pascal (one pascal = 7.5 × 10^-3 torr).

      The first major use of vacuum technology in industry occurred about 1900 in the manufacture of electric light bulbs. Other devices requiring a vacuum for their operation followed, such as the various types of electron tube. Furthermore, it was discovered that certain processes carried out in a vacuum achieved either superior results or ends actually unattainable under normal atmospheric conditions. Such developments included the “blooming” of lens surfaces to increase the light transmission, the preparation of blood plasma for blood banks, and the production of reactive metals such as titanium. The advent of nuclear energy in the 1950s provided impetus for the development of vacuum equipment on a large scale. Increasing applications for vacuum processes were steadily discovered, as in space simulation and microelectronics.

      Various kinds of devices have been developed for producing, maintaining, and measuring a vacuum. Several of the more significant types are described below.

Oil-sealed rotary pump.

      Capacities >are available from ¹/₂ to 1,000 cubic feet per minute, operating from atmospheric pressure down to as low as 2 × 10^-2 torr for single-stage pumps and less than 5 × 10^-3 torr for two-stage pumps. The pumps develop their full speed from atmosphere to about one torr, the speed then decreasing to zero at their ultimate pressures. One device of this type, useful for pumping both liquids and gases, is a two-bladed pump in which the rotor is eccentric to the stator, so forming a crescent-shaped volume that is swept by the blades through the outlet valve. Another variety, a rotary piston pump, is similar to a single-bladed pump, but the single blade is part of the sleeve fitting around the rotor. The blade is hollow and acts as an inlet valve, closing off the pump from the system when the rotor is at top centre.

      Ultimate pressures attainable are limited by leakage between the high- and low-pressure sides of the pump (due mainly to carryover of gases and vapours dissolved in the sealing oil that flash off when exposed to the low inlet pressure) and decomposition of the oil exposed to hot spots generated by friction.

      Typical applications of this pump are in food packaging, high-speed centrifuges, and ultraviolet spectrometers. It is also widely used as a forepump or a roughing pump, or both, for most of the other pumps described.

Mechanical booster.

      Capacities are available from 100 to 70,000 cu ft per minute, operating usually in the pressure range of 10 to 10^-3 torr. The peak speed of the pump is developed in the pressure range of 1 to 10^-2 torr, the speed at the lower end of the pressure range depending on the type of forepump used. A typical mechanical booster utilizes two figure-eight-shaped impellers, synchronized by external gears, rotating in opposite directions inside a stator. The gas is trapped between the impellers and the stator wall and transferred from the high vacuum to the fore vacuum side of the pump. The mechanical booster must be backed by another pump in series when working in its normal pressure range. The most frequently used type of forepump is the oil-sealed rotary pump. Typically, the mechanical booster is employed for pumping vacuum-melting furnaces, in an impregnation plant for electrical equipment, and in low-density wind tunnels.

Vapour diffusion pump.

      This pump is mainly used on equipment for the study of clean surfaces and in radio-frequency sputtering. Capacities are available up to 190,000 cu ft per minute with an operating pressure range of 10^-2 to less than 10^-9 torr when water-cooled baffles are used and less than 10^-11 torr when refrigerated baffles are employed. The pumping speed for a vapour pump remains constant from about 10^-3 torr to well below the ultimate pressure limitations of the pump fluid—that is, with the best fluids to pressures of better than 10^-9 torr. The diffusion pump is initially evacuated by an oil-sealed rotary pump to a pressure of about 0.1 torr or less. When the pump fluid in the boiler is heated, it generates a boiler pressure of a few torr within the jet assembly. High-velocity vapour streams emerge from the jet assembly, impinge and condense on the water- or air-cooled pump walls, and return to the boiler. In normal operation a portion of any gas arriving at the inlet jet is entrained, compressed, and transferred to the next stage. This process is repeated until the gas is removed by the mechanical forepump.

Sputter ion pump.

      Capacities are available up to 14,000 cu ft per minute, with an operating pressure range of 10^-2 torr to below 10^-11 torr. The full speed of the pump is developed in the pressure range from about 10^-6 to 10^-8 torr, although the characteristic at the lower pressure is dependent on pump design. This pump makes use of the sputtering principle, in which a cathode material such as titanium is vaporized—or sputtered by bombardment with high-velocity ions. The active gases are pumped by chemical combination with the sputtered titanium, the inert gases by ionization and burial in the cathode, and the light gases by diffusion into the cathode.

      A typical pump consists of two flat rectangular cathodes with a stainless steel anode between them made up of a large number of open-ended boxes. This assembly, mounted inside a narrow box attached to the vacuum system, is surrounded by a permanent magnet. The anode is operated at a potential of about seven kilovolts (kV), whereas the cathodes are at ground potential.

      Sputter ion pumps have a long life and can provide ultrahigh vacuum, free of organic contamination and vibration. They are employed mainly for clean-surface studies and in those applications where any organic contamination will give unsatisfactory results.

Titanium sublimation pump.

      Capacities are available up to many thousands of cu ft per minute, operating in the pressure range of 10^-3 to below 10^-11 torr. The full speed of the pump, which only pumps chemically reactive gases, is developed at pressures below 10^-5 torr. In this type of pump, titanium is sublimed onto the pump walls from either a resistance or an electron-beam heated source. Active gases are pumped by chemical combination, but inert gases are not pumped. As a consequence, it must always be used in conjunction with a diffusion or sputter ion pump. At pressures below 10^-5 torr the film will be deposited faster than it is being consumed, allowing deposition to be carried out at intervals rather than on a continuous basis. sublimation pumps are generally used in conjunction with a sputter ion pump in applications where a high speed is required and freedom from organic contamination is essential, as in the evaporation of materials onto a clean surface.

Sorption pump.

      Typically, the size of these pumps is about 1,000 grams of sorbent material, which retains gas molecules on its surface. They are capable of pumping from atmosphere to 10^-2 torr or can be used in series down to 10^-5 torr. In most cases the sorbent material is a molecular sieve—that is, a material that has been processed so that it is porous, with pore sizes comparable to the size of molecules, although activated charcoal can also be employed. The sorbent is positioned inside a cylindrical container that is connected to the vacuum system and that can be immersed in liquid nitrogen for supercooling to aid the sorption process. The gas is released when the sorbent returns to room temperature. This pump is used mainly for roughing systems in which the sputter ion and titanium sublimation pumps serve to ensure freedom from organic contamination.

Cryopump.

      This type of pump utilizes extremely low temperatures to condense gases and thus remove them from the system. Pumping speeds of millions of cu ft per minute are possible with the cryopump over the pressure range 10^-3 torr to well below 10^-10 torr. This type of pump can develop its full speed curve over the entire pumping range. Most cryopumps employ helium to cool the low-temperature surface; the helium can be in the form of gas at about 15 K or liquid helium at 4.2 K. A cryopump, which depends on the condensation of the gas for its pumping speed, will not effectively pump gases, such as helium and hydrogen, that have high vapour pressures at the low-temperature surface. Consequently, complementary diffusion or sputter ion pump capacity is a necessary adjunct to a cryopump vacuum-producing system. Most such pumps are used in high-altitude or space simulation.

McLeod gauge.

      The McLeod gauge takes advantage of Boyle's law (the product of pressure and volume for a given quantity of gas remains constant if a constant temperature is maintained) to determine gas pressure within a range of 10 to 10^-6 torr. Raising the mercury level in the McLeod gauge seals off the gas from the system to which the gauge is connected. When the level of mercury is raised further, the gas is compressed. The difference in levels of mercury between this trapped volume and the system being evacuated corresponds directly to the pressure in torr in the trapped volume. As the gauge depends only on the known initial volume trapped, the final compressed volume, and the pressure in this final volume—all of which can be directly measured—it is called an absolute gauge and is mainly a standard for calibrating other gauges.

Thermal conductivity gauges.

      Two types of thermal conductivity gauges, the Pirani and the thermocouple, determine pressure by the rate at which heat is dissipated from a hot filament. The Pirani gauge basically is a Wheatstone bridge with one arm in the form of a heated filament placed in the vacuum system. The resistance of the filament depends on its temperature, which, in turn, depends on the rate of dissipation of thermal energy through the residual gas. Thermal energy dissipation is affected by the pressure and thermal conductivity characteristics of the residual gas. The bridge is powered from a constant voltage source, and out-of-balance current due to temperature changes is indicated directly in torr. In the thermocouple gauge, the hot junction of the thermocouple is attached to a filament in the vacuum system and powered from a constant voltage source. The mode of operation is the same as that of the Pirani except that the temperature of the filament indicates the pressure. These gauges are rugged and simple to operate and cover a range of from 100 to 10^-4 torr.

Cold-cathode ionization gauge.

      This gauge makes use of the fact that the rate of ion production by a stream of electrons (electron tube) in a vacuum system is dependent on pressure and the ionization probability of the residual gas. Also called the Penning gauge, it consists of two cathodes opposite one another with an anode centrally spaced between them inside a metal or glass envelope. Outside the envelope a permanent magnet provides a magnetic field to lengthen the path travelled by the electron in going from cathode to anode, thus increasing the amount of ionization occurring within the gauge. Normally the anode is operated at about 2 kV, giving rise to a direct current caused by the positive ions arriving at the cathode. The pressure is indicated directly by the magnitude of the direct current produced. The pressure range covered by this gauge is from as low as 10^-7 torr. It is widely used in industrial systems because it is rugged and simple to use.

Hot-filament ionization gauge.

      The operating principles of this gauge are similar to the Penning gauge except that the electrons are produced by a hot filament and accelerated to a grid. The pressure range covered is either 1 to 10^-5 torr or 10^-2 to 10^-7 torr, depending on the electrode structure. Electrons emitted from the filament ionize residual gas molecules in the container being evacuated; the ion current arriving at the collector plates is directly proportional to the pressure and the ionization probability of the residual gas. This is a clean, accurate gauge that can be used down to about 10^-6 torr; below this pressure its accuracy is reduced due to the soft X-rays produced by electrons striking the grid. These X-rays generate a current in the collector circuit independent of pressure.

      Bayard-Alpert hot-filament ionization gauge. In this ionization gauge, the cross section of the collector is reduced to minimum to reduce the X-ray effect. This is achieved by inverting the gauge—that is, the collector (a fine wire) is surrounded by the grid. The pressure range covered is 10^-3 to 10^-9 torr or down to 10^-11 torr if a modulated instrument is used. Operating principles are the same as for the other ionization gauges described above.

* * *