Lower GC Detection Limits by Large Volume Injection

Large volume injection (10µL) coupled with thermal desorption achieved lower GC detection limits by reducing solvent background.

Achieving Lower Detection Limits with Large Volume Injections and Thermal Desorption, Alltech Application Note DY001, September 21, 1998.

In a typical GC analysis, split injections of around 1-2µL result in most of the sample being vented, limiting the sensitivity. Injection of larger than normal volumes of sample onto a GC column can result in lower detection limits, but the background contribution of contaminants in the solvent usually interferes with the detection of the target analytes.

To get around this problem, large volume injection (LVI) of samples of around 10µL can be used in which the sample is first placed on the frit or glass wool of a sorbent tube. Carrier gas is then flowed through the tube to distribute the target analytes from the sample onto the adsorbent material. The solvent is not adsorbed and is mostly flushed out. The sorbent tube containing the target analytes is then placed in a thermal desorption unit where it is heated, backflushed with carrier gas and the analytes carried onto a conventional GC column for separation and analysis.

Several examples of the use of LVI with thermal desorption were presented to demonstrate the utility of this approach to sample concentration and analysis. The LVI’s (10µL) were performed on a Dynatherm Model 60, 6-Tube Conditioner (Part No. 50255) with the Injection Port Option (Part No. 5147). Standards prepared in methylene chloride were injected onto a TD-VOC Thermal Desorption Tube (Part 106613) for the ACEM 900 Thermal Desorber (Part No. 50241). The TD-VOC tubes were packed with Carbograph 2, Carbograph 1 and AT-347 sorbents. Tenax Thermal Desorption Tubes for the ACEM 900 (Part No. 106601) were used for the standards prepared in methanol. Before use, the blank tubes were purged.

Samples (blank solvent and solvent containing the target analytes) were run by injecting the sample onto the sorbent tube via the injection port device, installing the tube into the thermal desorber unit and then thermally desorbing the analytes (tube temperature 280°C) onto a narrow bore focusing trap where they were reconcentrated onto a smaller amount of adsorbent. This trap was then heated (300°C) and the analytes backflushed onto an AT-5 (30m x 0.32mm x 1.0µm) capillary column (transfer line temperature 125°C) and analyzed by FID. The column temperature was held at 40°C for 2 min and raised to 250°C at 8°C/min with the detector at 275C.

Results of analysis of a Tenax blank, a methanol blank and a 20ppm gasoline standard in methanol showed that most of the gasoline components could be resolved from the solvent background. The Tenax contributed little to the background.

Analysis of a 10ppm gasoline standard showed the methanol contribution was significant. Analysis of the same 10ppm standard in methylene chloride using the TD-VOC adsorbent tube resulted in acceptable resolution of the gasoline from the solvent and adsorbent background.

Analysis of a 1ppm gasoline standard in methylene chloride using the TD-VOC tube showed that the methylene chloride background was more significant, indicating that the LVI technique while useful, is ultimately limited by the solvent purity.