Not everyone who works with ultra-high purity gases is aware of the difficulties involved in establishing a primary gas standard. The problems associated with the necessary correlation procedures, the instability of certain gases, reaction of gas mixtures, etc., are complicated further by the fact that many firms maintain their own primary standards, with little or no effort given to correlate values with others. It is obvious, therefore, that the only remedy for this haphazard situation is to establish a single repository for primary standards. Because of its history and experience in such matters, the U.S. National Bureau of Standards (NBS) is the only likely choice to serve as this repository. As is commonly known, this is no panacea. All firms involved in the specialty gas business must work continuously to maintain the effectiveness of the primary standards established. An absolute, by definition, is impossible to attain. However, by diligent efforts in correlating gases, we, as an industry, can approach that absolute as closely as possible.

The use of by-weight mixtures is not always as accurate as one may think. This paper intends to point up, by way of specific examples, several correlation techniques we have found necessary in certifying a primary standard, and to point up our present and future needs for standard reference materials supplied by NBS.