; CF4-(C2F6); PF3-PF5-(NF3); SFe-(S2F10); Xe-PF3-PF5-(POF3). Because of the proximity of appearance temperatures, the detection of SF4 in the presence of SFs was limited to approximately 1 part in 100. The component gases of a mixture can be identified by determination of the appearance temperature of the individual gases-i.e., by calibration of the column-or by trapping out the components after they pass through the detector and analyzing them by infrared scan or other analytical means.The empirical nature of the method makes it difficult to correlate exactly the appearance temperature with other variables such as temperature gradient, rate of warming, flow rate of helium, and length and geometry of the separation column. Also, the separations become increasingly difficult as the gap between appearance temperature narrows.Hence, no attempt was made to fix appearance temperatures more precisely than to the nearest five degrees. The vapor pressure of the substances at their respective appearance temperatures, except those appearing at -195°and -175°C , are below 1 mm and no meaningful volatility correlation can be made.However, the order of the appearance temperatures generally conforms with that of the respective boiling points. Although the appearance temperatures have no absolute meaning, it was found that their relative order, as listed above, remained unchanged even when other variables were changed by as much as a factor or two. Further evidence of this unchanging relative order is the close agreement of the appearance temperature differentials for the two gases, CF4 and C02, common to this work and that of Cady and Siegwarth (7); despite the marked dissimilarities (geometry of equipment, warm-up rate, etc.) of the two studies, these differentials were 35°C and 39°C , respectively.
ACKNOWLEDGMENTThe author thanks Murray Barsky and Milton Ader for their cooperation and helpful discussions.