Tetrachloroethylene recovery in a PE unknown

[LALman]

I just completed running a sample and found tetrachloroethylene (PERC) that was not in the unknown.

There was supposed to be no PERC. I found 14.2 ppb. Furthermore I ran it in duplicate and spiked in duplicate and got M = 14.21, MD = 12.98, MS = 53.29, and MSD = 53.89 ppb for a 40 ppb spike. My initial CC had 40.3 ppb (AvgRF 2.82 CCRF 2.85, for 40 ppb spike) and the immediately following blank was 0.40 ppb (less than my MDL of 1 ppb). My cal curve is a response factor curve with 20, 40, 80, 120, 160, 240, 320, 400, and 480 ppb. The RSD for the calibration is 4.52%.

Now my MS was a little high in recovery, so I let the MSD finish, ran a blank and then another unspiked matrix and another MS from the second VOA. The notes said the sample had to be run immediately so using the second VOA (which had been cooled to 4C in the sample fridge) was perhaps not best. But I thought it was better than reprepping the sample a second time. The M2 was slightly low on some analytes but the MS was right on the money.

I do not normally have a problem with PERC showing up in samples, nor with carryover of PERC from the CC or MS/MSD.

So, I am puzzled how I can find this in the sample, have good instrument and batch QC, and then be told it is not in the sample.

The only thing I can think of is that PERC might be a degradation product of something else in the sample?

[HPLC chemist]

What was the sample matrix?

[LALman]

Matrix is a 5mL aliquot of water. Method is EPA 8260C from 2006. Autosampler is an Archon 4551 with a Tekmar 3000 P&T. Trap is Vocarb 3000 (IIRC). Instrument is 6890/5973 Inert GC-MS with standard turbopump for vacuum and DB-VRX 20m x 180um x 1um column, helium carrier. Last calibration was May 2017. Last lens cleaning was Nov 2016. I have the 9mm opening in the drawout lense that makes it more linear when using hydrogen carrier.

I was in the accepted range for all but 3 of 67 or so target compounds. But had a carbon tetrachloride that was ~11 instead of ~22 and 1,3-****propylene that was also about half its value.

[Peter Apps]

Who says that there is no TCE in the sample, and why do you trust their opinion more than your own result ?

Peter

[LALman]

Its a purchased performance evaluation study sample. The study is closed now, and the results are back. So I thought it would be OK to ask outside advice for tracking down how such a puzzling thing could occur.

[Steve Reimer]

PERC is a degradation product of pentachloroethane. I have had issues with that in the past where one lot of standards will have a problem (20% degradation to PERC in 6 hours) and the next lot will be fine, all from the same supplier. Under neutral to basic pH pentachloroethane and to a lesser extent, some other chlorinated compounds, will dehydrohalogenate - lose HCl.
The solution is to acidify to get the pH down. I found that ~4 was low enough that the process didn't interfere with my results. Since I don't know what the variables are and can't reliably reproduce the problem, your results may differ.

[Peter Apps]

Its a purchased performance evaluation study sample. The study is closed now, and the results are back. So I thought it would be OK to ask outside advice for tracking down how such a puzzling thing could occur.

That would have been good to know from the outset. Assuming that you were not the only participant, did anyone else also find PERC ? and if they did do their methods have anything in common with yours ?

Peter

[James_Ball]

PERC is a degradation product of pentachloroethane. I have had issues with that in the past where one lot of standards will have a problem (20% degradation to PERC in 6 hours) and the next lot will be fine, all from the same supplier. Under neutral to basic pH pentachloroethane and to a lesser extent, some other chlorinated compounds, will dehydrohalogenate - lose HCl.
The solution is to acidify to get the pH down. I found that ~4 was low enough that the process didn't interfere with my results. Since I don't know what the variables are and can't reliably reproduce the problem, your results may differ.

This is the most common reason to find Tetrachloroethylene when it shouldn't be there, if Pentachloroethane is present.

I used to have problems with it all the time. A CCV run right at the start would be ok, but one that sat on the instrument until the next 12 hour CCV check would fail PERC high and Pent low. We solved the problem when we installed chillers on the autosamplers. We now run the autosamplers at roughly 6C and CCVs will last for two days sitting on the autosampler tray.

[LALman]

Who says that there is no TCE in the sample, and why do you trust their opinion more than your own result ?

Peter

The study states that there is no PERC (Tetrachloroethene) in the sample and gives no statistics from other labs. For compounds for which a result was expected, the study does give Zscore, Study Mean, and Study Standard Deviations.

There is TCE (Trichloroethene) present. I found 36.5 ppb. Accepted value was 42.2 ppb Z Score was -0.472 Study mean was 38.9 and Study Standard Deviation was 5.04.

There is also 1,1,1,2 Tetrachloroethane present. I found 39.0 ppb Accepted value was 53.1 ppb Z Score was -1.12 Study mean was 45.5 and Study Standard Deviation was 5.81.

1,2 Dichloroethane was present. I found 53.7 ppb, accepted value was 54.3 ppb. 1,1-dichloroethene was present. I found 44.9 ppb, accepted value was 51.0 ppb.

1,1,2,2-Tetrachloroethane was below my 1 ppb MDL. Hexachlorobutadiene was below my 2 ppb MDL. Neither 1,1,1- nor 1,1,2- Trichloroethane were above my 1 ppb MDL.

[LALman]

PERC is a degradation product of pentachloroethane. I have had issues with that in the past where one lot of standards will have a problem (20% degradation to PERC in 6 hours) and the next lot will be fine, all from the same supplier. Under neutral to basic pH pentachloroethane and to a lesser extent, some other chlorinated compounds, will dehydrohalogenate - lose HCl.
The solution is to acidify to get the pH down. I found that ~4 was low enough that the process didn't interfere with my results. Since I don't know what the variables are and can't reliably reproduce the problem, your results may differ.

This is the most common reason to find Tetrachloroethylene when it shouldn't be there, if Pentachloroethane is present.

I used to have problems with it all the time. A CCV run right at the start would be ok, but one that sat on the instrument until the next 12 hour CCV check would fail PERC high and Pent low. We solved the problem when we installed chillers on the autosamplers. We now run the autosamplers at roughly 6C and CCVs will last for two days sitting on the autosampler tray.

This interests me. I do not have pentachloroethane in my standard mix. I use the AccuStandard mixes M-502A-R (54 compounds), M465B-10X (7 gasses), and M4665D-ADD-R (8 additional compounds). It could be that the PE mix contained pentachloroethane to see if people in the study had running conditions that properly controlled for breakdown products. I've been running 8260 volatiles on this instrument since we bought it in 2004. But I consider myself more of an expert on elemental analysis.

Let me see if I can find where pentachloroethane elutes on a DB-VRX column and I will look for it in the unknown.

[LALman]

I just reexamined the spectra after checking a copy of J&W's DB-VRX chromatogram for 8260. There is no pentachloroethane but there is 14 ppb of hexachloroethane! Thanks so much for pointing this out to me.

[James_Ball]

If you are running out to Naphthalene it will elute a little before that. It will probably elute near Hexachlorobutadiene.

[LALman]

I run out to napthalene, hexachlorobutadiene, and the 1,2,4- 1,2,3- trichlorobenzenes. I see the hexachloroethane immediately after 1,2-Dibromo-3-chloropropane and that is right where the J&W chromatogram puts it.

So, one could add sodium bisulfate to every vial or refrigerate the autosampler tray? This sample must degrade very rapidly. After I prepared 100mL of the sample I immediately filled two VOA's and after using one to do M/MD/MS/MSD, I put the second voa in the sample fridge and then prepped an additional M and MS. They all have nearly identical areas of hexachloroethane. My Archon autosampler does not have the tray cooling mod.

Is this mostly or exclusively a degradation that occurs in the sample? Or, is this also a common problem within the purge and trap to inlet pathway?

I ask this because I was running hydrogen as a carrier for a couple years until a power failure pulled roughing pump vapor through my EPC vent. I typically purge with Argon because I have it plumbed for my ICP-MS and back when I was using hydrogen for carrier gas, I had my Tekmar 3000 plumbed with hydrogen. Thus a 1 minute dry purge with hydrogen after purgeing with argon was enough to prevent argon from causing any quenching problems at the filiments. Even though I am now using helium for my carrier, I still have that UHP hydrogen dry purge step. So, I wonder if that might be reactive enough on desorb to pull a couple chlorines off of the hexachloroethane.

[Steve Reimer]

From the paper I could find on the subject hexachloroethane is an order of magnitude slower in dechlorinating due to lack of hydrogen trans to a chlorine.

I was never able to reliably reproduce the problem except with the one lot of standard mix. Our fix is to acidify with HCl (1:1 dilution), one drop is plenty.

[James_Ball]

It will usually be Penatchloroethane that decomposes into Tetrachloroethene, and Pentachloroethan does break down quite easily at room temperature.

If you only have a few samples to run at a time you can store them in the cooler at 4c and that will retard the breakdown for a few hours while it warms to room temp on the autosampler.