TO-15 + PAMS + TO-11A = China’s HJ759 + PAMS + HJ683 Part 3: Formaldehyde Sampling in Air Canisters

In my previous blog (TO-15 + PAMS + TO-11A = China’s HJ759 + PAMS + HJ683 part 2: Deans switching and TO-15/PAMS) I covered the combination of the TO-15 and PAMS (or HJ759 and PAMS) methods into a single run using a Deans Switch and FID/MS detector set up. I said that I’d revisit integrating TO-11A (i.e., HJ683 in China) methods, but first let’s talk about the methods a bit.

TO-11A differs from TO-15 since it involves the capture and derivitization of formaldehyde and other carbonyls on a coated adsorbent tube rather than air canister sampling. The carbonyls react with 2,4-dinitrophenylhydrazine (DNPH) to form a hydrazone derivative in the adsorbent tube. The DNPH derivative is then eluted off of the tube using a solvent and the resulting solution is analyzed by  HPLC/UV. Having two completely different sampling systems and instruments can be a rather large investment in both time and money, so it’s easy to see why there is a push to use canister sampling and GC/MS analysis instead.

So why the separate setup? Formaldehyde is capable of being analyzed directly by GC with no dervitization, although there are some issues with it being low mass and sharing ions with carbon dioxide. See Fig. 1 below illustrating our combined HJ759/PAMS method, using both air and helium as a fill gas for the canister.

Fig. 1 - Formaldehyde in helium (top) and air (bottom), showing interferences from air/CO2 in the bottom chromatogram.

Even if the chromatography can be made acceptable the issue is that formaldehyde is unstable in canisters, creating the risk of false negatives and low response. Canister coating has advanced since the TO methods were originally written though, and one maker of sampling canisters claims they can collect formaldehyde with no losses. If true that would be a game changer for air analysis, so we decided to put it to the test using their canisters. Six canisters were spiked with 100 ppbv of formaldehyde. Three were filled with dry air and three with air at 80% relative humidity to see if water content had any effect. The canisters were then tested every 12 hours for several days and we were not able to replicate their results, with the data showing a quick and drastic drop in formaldehyde response.

 

Fig. 2 – Average formaldehyde loss in dry and humid competitor canisters. N=3 for each data point and error bars show the standard deviation of the replicate canisters.

How bad was the formaldehyde stability? As seen in Fig. 2 in half a day the formaldehyde had dropped to less than 80% of the original amount in the dry canisters, and less than 60% in the humidified ones.  The humidified canisters held steady at near 50% afterwards, but the dry canisters continued to drop until they were less than 40% after 4 days. The water in the humidified canisters likely traps the formaldehyde initially (Day 0 to Day 3) then slowly releases it later (Day 3 on out), causing the initial lower and higher ending results. Given the quick initial drop in what is the simplest scenario of just formaldehyde in air it’s hard to see this as a viable sampling method, even if the samples are rushed to the lab as quick as possible.

While the push for a universal air method is understandable, at this point the sampling techniques don’t support it and it appears that TO-11A will live on as an independent method. And for those of you interested in aldehyde and ketone in air analysis, Restek has you covered with standards and HPLC columns.