NJDEP-SRP Low Level TO-15 Series: Part 4 – How to calibrate your system Joisey style… with no digression
16 Dec 2015In the last installment of this series*, we talked about optimizing the internal standard (ISTD) responses on the Markes CIA Advantage. This is a critical step on the road to meeting the NJ LL TO-15 calibration criteria of a 5-point curve from 0.2 to 40 ppbv, for most of the target analytes. I say “most,” because NJ provides that acetone, chloroethane, chloromethane, carbon disulfide, dichlorodifluoromethane, 1,4-dioxane , ethanol, isopropanol, methylene chloride, methyl ethyl ketone, methyl methacrylate, tertiary butyl alcohol, tetrahydrofuran and 1,2,4-trimethylbenzene may have a lowest calibration point that is higher than 0.2 ppbv; however, the upper limit of 40 ppbv remains the same.
The last topic we need to cover before I show you the numbers is the injection volume (yes, no peeking at the table below). This I found to be almost as equally important as the ISTD details we discussed previously*. As mentioned last time, the whole calibration process is a fine balancing act between injection volume, breakthrough volume, detection limits, saturation limits, etc… for all volatile organic compounds (VOCs). I found that my system had a sweet spot around 250 mL. AND… this volume was largely dictated by chloromethane. That slippery bugger would start to breakthrough at larger injection volumes. Had it not been for chloromethane, I may have been able to use larger injection volumes. Regardless, 250 mL provides sufficient mass to the 5977 to have legitimate peaks at 0.2 ppbv. When I say legitimate, I mean signal to noise ratios of 10:1 or better. For the record, this on the same dirty source Colton and I gunked up for the e-cig work (post titled "Electronic Cigarettes Part X: Vapor Analysis – Application Note") over the last year.
Okay, I may have digressed a little back there, so back to the program now. Just like the U.S. Environmental Protection Agency (EPA) Method TO-15, NJ says that calculated % relative standard deviation (RSD) for the relative response factor (RRF) for each compound in the calibration must be less than 30% with at most two exceptions up to a limit of 40%. So here is what happens when we inject 250 mL of 0.2, 0.8, 2.0, 10, 20, and 40 ppbv of the target analytes… drum roll please:
As shown in the table above, the only 2 compounds that do not meet the NJ LL TO-15 criteria are ethanol and acetone. BUT no need to panic here, as it just so happens that those two compounds fall in the list of exceptions previously mentioned. It appears that NJ knows these compounds can present a challenge at the lower calibration levels due to background contamination. So if we drop the two lowest calibration points, which were definitely due to blank cleanliness (i.e., not a reflection of the CIA’s performance), the % RSDs become 7.41 and 6.31 for ethanol and acetone, respectively.
So there you have it, all target VOCs from 0.2 to 40 ppbv with a 6 point calibration curve, a dirty source, and no manual integrations (i.e., these are realistic results folks); and we meet the NJ LL TO-15 criteria with flying colors! Stay tuned for future discussions on precision and method detection limits (MDLs).