By Alvin B. Kaufman and Edwin N. KaufmanCold traps are used in instrumentation and elsewhere to prevent the introduction of vapors or liquids into a measuring instrument from a system, or from a measuring instrument (such as a McLeod gauge) into the system. A cold trap provides a very low temperature surface on which such molecules can condense, and improves pump-down by one or two magnitudes.
However, cold traps improperly employed can impair accuracy, destroy instruments or systems, and be a physical hazard. For example, many of the slush mixtures used in cold traps are toxic or explosive hazards, and this is not indicated in the literature. .
The authors became aware of the deficiencies in tunnel instrumentation, where it was necessary to measure pressures in the micron to 760-torr region (The torr is equal to a pressure of 1 mm of mercury at standard condition). The instrumentation system used Statham gauges for ambient pressure down to 100 to 150 torr or about 2-3 psia and NRC Alphatron gauges for pressures to 5 x 10-2 torr to prevent calibration shifts and contamination of the NRC transducers by oil fumes from the vacuum pump and possible wind-tunnel contaminants, a cold trap was placed in the line.
The cold trap was filled with liquid nitrogen, and the valve to the tunnel line shut off. When the valve was opened, cold gas shot out, shown by condensation; the over-pressure developed in the system destroyed the Statham strain-gage bridge, although it was not sufficient to rupture the transducer diaphragm. As no satisfactory explanation was forthcoming, a glass cold trap was procured and set up in a dummy system. The cause of the phenomenon soon became apparent: air in the trap and system lines was becoming liquefied in the trap. When the valve was opened, this liquid air was being blown into the warmer lines by atmospheric pressure; the resultant volatilization liquid into gas was practically an explosion.
Nevertheless, cold traps are often the only satisfactory means of removing contaminants, although in ordinary experimental work the charcoal trap is occasionally acceptable. A charcoal trap will remove oil and condensable vapors so that pressures to 10-8 torr or better may be secured, but it presents a serious restriction on pumping speed and requires bakeout when it has become charged with oil and vapors. Molecular sieve traps place similar restrictions on pumping speed.
The errors introduced by water vapor, when measuring low pressures, depend on the vacuum gauge used. The presence of water vapor also affects the magnitude of vacuum that can be achieved. The equilibrium point of a dry-ice-acetone slush is -78°C (-108.4°F), which, although sufficient to trap mercury vapor effectively, does not remove water vapor; a temperature of at least -100°C (-148°F) is required to eliminate water vapor, or, alternatively, exposure to anhydrous phosphorus pentoxide (P205). This material is usually rejected for field use because of possible biological, fire and explosive hazards: in absorbing water it produces heat, and reacts vigorously with reducing materials. Slush mixtures using liquid air and liquid oxygen were considered and dropped, either because of the explosive hazard or toxicity of the vapor or because they were not cold enough. Table I lists many common thermal transfer and coolant fluids with their hazards and limitations.
Carbon dioxide is adequately caught by traps cooled by liquid air or nitrogen, its vapor pressure at the liquid-air temperature being 10-6 to 10-7 torr. Methane, ethylene and carbon monoxide have considerably higher vapor pressures and are not effectively trapped by even a liquid-nitrogen trap.
Vapor pressure of most standard roughing pump oils is 10-3 to 10-4 torr at 25°C (77°F), 1/5 of this value at 0°C, and negligibly small below the temperature of dry ice. Fractionating oils currently used in vacuum pumps have very low vapor pressures, ranging from 20 x 10-6 to 10-7 torr, and pose no problem for most work. Nevertheless, gases produced by thermal decomposition of the oil may contaminate the vacuum unless trapped.
TABLE I
Thermal Transfer Fluids (1) Used with Instrumentation Cold Traps
Compound (2) Temperature °C (3) Inhalation Toxicity Skin Toxicity Explosive or Fire Hazard (4) Vapor Pressure 70% Glycerine
in Water-39 None Slight Slight 2.5X10-3 T@50°C Ethyl Alcohol,
Dry Ice-78 Moderate Slight Dangerous 100 T @22°C
Chloroform, Dry Ice -64 Extreme Slight Slight Liquid SO2 -76 Extreme Extreme None Methanol, Dry Ice -78 Slight Slight Dangerous 100 T @22°C Acetone, Dry Ice -78 Moderate Slight Dangerous 400 T @39°C Methyl Bromide,
Dry Ice-78 Extreme Extreme Slight Freon 11, Dry Ice -78 Slight Slight None Methylene Chloride, Dry Ice -78 Moderate Moderate Slight 380 T @22°C Calcium Chloride -42 Slight Slight None Ethyl Methyl
Ketone-78 Moderate Moderate Dangerous (1) Transfer fluids will freeze solid and become colder if subject to temperatures lower than their freezing point. A slush mixture is secured by lowering temperature, such as by the introduction of limited quantities of dry ice until the mixture is quasi-frozen.
(2) These materials are often sold under trade names (listed in the Handbooks of Chemistry and Physics). In general, any combination of elements shown was selected for the coldest slush mixture obtainable.
(3) If the refrigerant is dry ice, the transfer fluid will not go below -78°C (-108.4°F), the temperature of solid CO2.
(4) The consensus is that many of these liquids while hazardous at room temperature, are not hazardous when cooled, since their evaporation at low temperatures is fairly low. For utmost safety those noted dangerous should not be employed unless venting or other special precautions are taken. For greater detail see Reference 2.
VIRTUAL LEAKS
If the cold trap is chilled too soon after the evacuation of the system begins, gases trapped will later re-evaporate, when the pressure reaches a sufficiently low value. The evaporation of the refrigerated and trapped gases is not rapid enough to be evacuated by the system, but is enough to degrade the vacuum, producing symptoms very similar to those of a leak.
To avoid these virtual leaks, keep the traps warm until a vacuum of about 10-2 torr is obtained. The tip of the trap is then cooled until ultimate vacuum is reached, at which time the trap may he immersed in the coolant to full depth.
SAFETY PRECAUTIONS
If liquid nitrogen is the coolant, liquid air can condense in the trap, inviting explosion. Liquid air, comprising a combination primarily of oxygen and nitrogen, is warmer than liquid nitrogen. Depending on the nitrogen content, air liquefies anywhere from -190°C (-310°F) (5°C warmer than liquid nitrogen) to -183°C (-297.4°F) (liquid oxygen). If liquid nitrogen is used, the trap should be charged only after the system is pumped down lest a considerable amount of liquid oxygen condenses, creating a major hazard.
Handle any liquid gas carefully; at its extremely low temperature, it can produce an effect on the skin similar to a burn. Moreover, liquefied gases spilled on a surface tend to cover it completely and intimately, and therefore cool a large area.
The evaporation products of these liquids are also extremely cold and can produce burns. Delicate tissues, such as those of the eyes, can be damaged by an exposure to these cold gases which is too brief to affect the skin of the hands or face.
Eyes should be protected with a face shield or safety goggles (safety spectacles without side shields do not give adequate protection). Gloves should be worn when handling anything that is or may have been in contact with the liquid; insulated gloves are recommended, but leather gloves may be used. The gloves must fit loosely so that they can be thrown off quickly if liquid should spill or splash into them. When handling liquids in open containers, high-top shoes should be worn with trousers (cuffless if possible) worn outside them.
Stand clear of boiling and splashing liquid and its issuing gas. Boiling and splashing always occur when charging a warm container or when inserting objects into the liquid. Always perform these operations SLOWLY to minimize boiling and splashing.
Should any liquefied gas used in a cold trap contact the skin or eyes, immediately flood that area of the body with large quantities of unheated water and then apply cold compresses. Whenever handling liquefied gases, be sure there is a hose or a large open container of water nearby, reserved for this purpose. If the skin is blistered, or if there is any chance that the eyes have been affected, take the patient immediately to a physician for treatment.
Oxygen is removed from the air by liquid nitrogen exposed to the atmosphere in an open Dewar. Store and use liquid nitrogen only in a well ventilated place; owing to evaporation of nitrogen gas and condensation of oxygen gas, the percentage of oxygen in a confined space can become dangerously low. When the oxygen concentration in the air becomes sufficiently low, a person loses consciousness without warning symptoms and will die if not rescued. The oxygen content of the air must never be allowed to fall below 19%.
The appearance of a blue tint in liquid nitrogen is a direct indication of its contamination by oxygen, and it should be disposed of, using all the precautions generally used with liquid oxygen. Liquid nitrogen heavily contaminated with oxygen has severe explosive capabilities. In addition, an uninsulated line used to charge Dewars will condense liquid air; liquid air dripping off the line and revaporizing causes an explosive hazard during the charging operation.
If the cold trap mixture is allowed to freeze, and the cold trap becomes rigid, slight movement in other parts of the apparatus could result in breakage of the trap or other glassware.
If a gas trap has to be lifted out of the Dewar cold bath for inspection, it will be difficult to reinsert into the slush. Therefore, it is preferable to use a liquid that will not freeze at -78.5 °C.
ACKNOWLEDGEMENTReprinted from Rimbach Publications, Pittsburgh, Pennsylvania, "Instrument and Control Systems." Vol. 36, pp. 109-111, July, 1963.REFERENCES1. Strong. J., Neher, H. V., Whitford, A. E., Cartwright, C. H., and Hayward, R., "Procedures in Experimental Physics," Prentice-Hall, Inc.. New York, (1938).2. Sax, N. Irving, "Dangerous Properties of Industrial Materials." Reinhold Pub. Co., New York, (1961).3. Dushman, Lafferty, "Scientific Foundation of Vacuum Techniques," John Wiley & Sons. New York. (1962).