Answer to Question #10502 Submitted to "Ask the Experts"
The following question was answered by an expert in the appropriate field:
There are a number of methods that are suitable for sampling of tritium (3H) in air. The method of choice depends on a number of factors, including whether you require real-time monitoring of the airborne tritium levels, the chemical and physical form that the tritium is in, the kinds of collection devices you have access to and/or prefer, and the types of measuring systems to which you have access.
If you desire real-time measurements of the airborne tritium levels, a flow through ionization chamber type of system is among the most common and useful choices. In the simplest version, a single flow-through ionization chamber is connected to an appropriate power supply and to a current measuring (electrometer) system. If particulate tritium is not a concern, the air flow passes through a high-efficiency particulate filter before entering the chamber. Such a system naturally must be calibrated and corrections for gamma radiation-induced background may have to be made. There are commercial versions available that use multiple chambers, including flow-through chambers for the tritium measurements and static chambers to assess the gamma background. See, for example, the systems at these websites (these examples are not intended as recommendations):
If you do not require real-time monitoring, the remaining sampling options involve collection of samples for a fixed interval of time (e.g., daily, weekly, monthly) and subsequent analysis of the collected tritium. It is more often the case at most facilities that airborne tritium is in gaseous form rather than particulate form. I shall assume that this is true for your case. The next concern is whether the tritium is in the common form of tritiated water vapor or the diatomic gas, likely HT. In a university research facility it is possible that some procedures might generate the diatomic gas alone or in combination with tritiated water, HTO vapor. The tritiated water may be collected fairly easily by any of several techniques. One of the simplest uses a flow-through gas bubbler containing either water or sometimes ethylene glycol. The air is drawn through the bubbler, and after collection an aliquot of the liquid is transferred to a liquid scintillation cocktail mix and counted in a liquid scintillation counter. If the diatomic gas is present it is generally necessary to pass the effluent gas from the bubbler (after removal of the tritiated water) through an appropriate heated catalyst (often palladium) to oxidize the gas to tritiated water, which is then passed again through one or more collection bubblers. If you put your own such system together, it is a simple matter to determine the tritiated water collection efficiency by using three or more bubblers in series and determining the fraction of activity collected in the first bubbler. There are also commercial systems available—e.g., see http://www.fjspecialty.com/mrb500.htm.
A couple of other fairly common collection methods are available for tritiated water vapor. These include collection on dessicant agents, such as silica gel or CaSO4, with subsequent removal by distillation from the solid collection agent. Collection times of these devices may be restricted by the amount of humidity in the air and the volume of the collection material. Collection by freezing out the trtitiated water by drawing the air through a cold trap is another alternative, although it is probably better suited to short-term sampling. In either of these methods, the most common measurement technique is liquid scintillation counting of all or a portion of the collected tritiated water.
You can find a considerable amount of information on the subject by searching the internet. The U.S. Nuclear Regulatory Guide NUREG-1400 (Section 1.3) summarizes some of the more common techniques for tritium air sampling. National Council on Radiation Protection and Measurements Report No. 47, Tritium Measurement Techniques, remains a good reference for the topic.
I hope you are able to implement your sampling program with minimal problems.
George Chabot, PhD