As previously described, the same GC sample introduction hardware was deployed as a CDS 7000C concentrator configured with a DHS module. On the automation side, the setup was connected to a CTC RTC rail. This automated system is controlled by Pal Sample Control (PSC) software with two plug-ins for the 7000C concentrator and the DHS module individually.
The DHS module was mounted on the RTC rack. A perfume sample was sealed in a 10 ml VOC vial and placed on the sample tray. During testing, the VOC vial is transported by the CTC Purge and Trap Tool, which is the robotic arm designed specifically to handle 10 ml, 20 ml and 40 ml VOC vials for purge and trap applications, into the DHS module. Once the sample was loaded into the DHS module, the dual jacketed needle inside the DHS module was lowered to pierce the top septum of the vial. Inlet purge gas flow followed to purge the sample in the head- space through a heated transfer line to be enriched in the analytical trap installed in the 7000C concentrator. The setup is shown in Figure 1.
In this experimental setup, the Full Evaporation Technique (FET) by Markelov  was followed. A commercially purchased perfume oil was diluted with methanol to a final 5% (v/v) concentration. A micro syringe was used to obtain 0.5 µL, 1.0 µL, 2.0 µL, 5.0 µL and 10.0 µL volume of sample from the diluted solution individually. Each aliquot was injected individually to the bottom of a clean 10 mL VOC vial and each sample vial was capped immediately after injection for future analysis.
Figure 2 presented the chromatogram data from a 2.0 µL run. All fragrance peaks were numbered based on the compounds list in Table1. For each of the 26 compounds identified in the Figure 2, a calibration curve was drawn by fitting the peak areas from 5 runs (0.5 µL, 1.0 µL, 2.0 µL, 5.0 µL and 10.0 µL sample volume) with a linear polynomial. The assumption is that the response factor obtained from FET method is independent of the sample volume. To better depict the data, 26 compounds were separated into two groups based on the peak area from the 10.0 µL sample volume run. If the peak area of a specific com- pound from the 10.0 µL run is below 4,500,000, this com- pounded is considered as a high response compound. If not, it is grouped into low response compounds. Figure 3 and 4 summarized the calibration curve for high response and low response compounds.