EVALUATION OF THERMAL DESORPTION FOR THE MEASUREMENT OF ARTIFICIAL SWINE ODORANTS IN THE VAPOR PHASE

Abstract

Quantification of odorants from animal production facilities is difficult. The current technique is to collect air samples in Tedlar bags and quantify odor using a trained olfactory panel. In this approach, relative differences between samples can be determined, but further quantification of odorants is limited. An alternative approach is to quantify odorants in air emissions using sorbent tubes. A sorbent tube is a glass tube packed with a specific adsorbent material (Tenax TA, Carboxen 1000, Carbosieve SIII, etc.) and has been used to collect volatiles and quantify emissions from various industrial sources. Each adsorbent has a limited range of chemical selectivity. Limited applications of sorbent tubes with single or dual adsorbents have been used to measure odorant emissions from animal production facilities. In this study, tri–packed sorbent tubes and Tedlar bags were compared in characterizing 19 major odorants found in artificial swine odor. The sorbent tubes were packed with Tenax TA, Carboxen 1000, and Carbosieve SIII. The artificial swine odor was directly desorbed onto the tri–packed sorbent tube. For comparison, a 10–L Tedlar bag was filled with nitrogen gas and artificial swine odor. The Tedlar bag was then desorbed onto the tri–packed sorbent tube. The sorbent tube was then thermally desorbed into a gas chromatography (GC) system with a flame ionization detector (FID) for quantification. The tri–packed sorbent tube demonstrated recoveries greater than 74% and detection limits less than 0.4 ng for all 19 odorants. Thus, a tri–packed sorbent tube may provide an analytical method to measure low concentrations of major odorants found in air emissions from swine production facilities. Tedlar bags showed limited recoveries of some odorants, less than 12% for indole and skatole. In addition, Tedlar bags immediately sampled after three flushings with nitrogen emitted 3.50 ng L–1 hr–1 of acetic acid (~35% above background levels) and 2.13 ng L–1 hr–1 phenol (~27% above background levels). These results suggest that air samples collected in Tedlar bags may bias olfactory analysis.