Chromatography Forum

We are seeing loss of metribuzin of >40% during SPE (EPA 525). All other compounds, including BaP and Perylene-d12 are OK. Diluted standard recovers fine for all compounds. Could it be the extraction solvents, in this case ethyl acetate or dichloromethane? [The DCM we buy is preserved with amylene to prevent formation of radicals that could destroy PAHs]. Has anyone experienced this?
Thank you for your input!

We are seeing loss of metribuzin of >40% during SPE (EPA 525). All other compounds, including BaP and Perylene-d12 are OK. Diluted standard recovers fine for all compounds. Could it be the extraction solvents, in this case ethyl acetate or dichloromethane? [The DCM we buy is preserved with amylene to prevent formation of radicals that could destroy PAHs]. Has anyone experienced this?
Thank you for your input!

I run these often and have been averaging about 80-85% recovery when spiked at a mid level and 70-80% recovery spiked at the MRL using the same solvent, including preserved DCM.

Do you use the Waters HLB cartridges listed in the method?

How fast do you draw the eluting solvent through the cartridges?
We pull it through until it just starts to drip, then shut off flow for a minute or two, then pull it until all has passed through the sample delivery line to the cartridge then allow to gravity feed until the solvent gets to the top of the packing and stops, then pull through with vacuum slowly until all has passed through. We do this with both ethylacetate and DCM. Pulling through faster seems to give lower recoveries.

When doing solvent evaporation, we also bring it down to about 2ml then wash the tubes with another 2ml of ethyl acetate before continuing down to less than 1ml which also helps with recovery.

Metribuzin is prone to photolysis (DT50 0.2 days).

We are seeing loss of metribuzin of >40% during SPE (EPA 525). All other compounds, including BaP and Perylene-d12 are OK. Diluted standard recovers fine for all compounds. Could it be the extraction solvents, in this case ethyl acetate or dichloromethane? [The DCM we buy is preserved with amylene to prevent formation of radicals that could destroy PAHs]. Has anyone experienced this?
Thank you for your input!

I run these often and have been averaging about 80-85% recovery when spiked at a mid level and 70-80% recovery spiked at the MRL using the same solvent, including preserved DCM.

Do you use the Waters HLB cartridges listed in the method?

How fast do you draw the eluting solvent through the cartridges?
We pull it through until it just starts to drip, then shut off flow for a minute or two, then pull it until all has passed through the sample delivery line to the cartridge then allow to gravity feed until the solvent gets to the top of the packing and stops, then pull through with vacuum slowly until all has passed through. We do this with both ethylacetate and DCM. Pulling through faster seems to give lower recoveries.

When doing solvent evaporation, we also bring it down to about 2ml then wash the tubes with another 2ml of ethyl acetate before continuing down to less than 1ml which also helps with recovery.

Hi James, and thank you for your input. I should have mentioned we run EPA 525.2 method version, and as such we are not allowed to use any sorbent other than C18. That said, everything seems to be under control now. We had prepared the diluted spike in acetone which, from what I gathered, is not an optimal solvent for metribuzin. I switched to methanol and metribuzin is at around 90% recovery, like all others, even our PQL spike at 0.05 ug/L. I also started preparing the spike fresh with every batch.

The elution solvents are drawn at about 2 mL/min (on Fisher AutoTrace 280, over 10 years old), we cook it down to 1 mL (Turbovap), add 3 mL of ETAC and cook it down to around 1 mL once more. Bring it to volume of 1 mL with ETAC and analyze. If I do not put the extracts on the GCMS the same day, they go to freeze. When ready to analyze, they get sonicated for about 5 minutes before they go on the auto sampler. We run Shimadzu single quad QP2020NX and 2030 GC that is about 1 year old and performs quite well. Since I am in AZ, we must follow the AZ DHS QC requirements rather than the EPA’s and as such we must see 0.01 ug/L of endrin which may be challenging sometimes.

Agin, thank you for your reply!

Cheers!
Ivo

The low detection level for Endrin and for Benzo(a)pyrene is the main reason we switched to 525.3 so we could use the SIM/Scan option to get lower detection limits over pure Scan. Also not needing to extract the internal standard then add the monitoring compound simplifies the recovery and quant calculations so that our LIMS can handle it all, since with 525.2 you need an internal standard for the targets and another internal standard for the internal standard.

The low detection level for Endrin and for Benzo(a)pyrene is the main reason we switched to 525.3 so we could use the SIM/Scan option to get lower detection limits over pure Scan. Also not needing to extract the internal standard then add the monitoring compound simplifies the recovery and quant calculations so that our LIMS can handle it all, since with 525.2 you need an internal standard for the targets and another internal standard for the internal standard.

We looked into getting the cert for 525.3 but our QC person decided against it with no good explanation. Can you get pre-preserved bottles for 525.3? That may have been the reason since the people collecting samples would have to do it themselves if pre-preserved sample bottles weren’t available.

Anyway, I also investigated endcapped vs. unendcapped C18 and sure enough there’s a difference (endcapped showed better recovery specifically for metribuzin, 95+ vs 75-80, but I didn’t do any statistics on the small group of data, so it may be just a coincidence). All is working now.

Thanks again!

The low detection level for Endrin and for Benzo(a)pyrene is the main reason we switched to 525.3 so we could use the SIM/Scan option to get lower detection limits over pure Scan. Also not needing to extract the internal standard then add the monitoring compound simplifies the recovery and quant calculations so that our LIMS can handle it all, since with 525.2 you need an internal standard for the targets and another internal standard for the internal standard.

We looked into getting the cert for 525.3 but our QC person decided against it with no good explanation. Can you get pre-preserved bottles for 525.3? That may have been the reason since the people collecting samples would have to do it themselves if pre-preserved sample bottles weren’t available.

Anyway, I also investigated endcapped vs. unendcapped C18 and sure enough there’s a difference (endcapped showed better recovery specifically for metribuzin, 95+ vs 75-80, but I didn’t do any statistics on the small group of data, so it may be just a coincidence). All is working now.

Thanks again!

Yes, we get pre-preserved bottles from QEC and some other companies have them too.

As for C-18, we had one company substitute part numbers recently and our Formaldehyde recoveries went to nothing. We did not notice they had substituted the parts, and we found out later it was a slightly different C-18 but it made a big difference.

The only other requirement that most don't like with 525.3 is that you begin with a CCV ran at the MRL level that you report. Pass/fail is 50%+/- and normally we don't have any problems. We make that our Calibration level 2 with 1 being one half that level.

What columns are you guys running? Any issues with co-elutions? The lab we sub to is using the restek RXI columns. Last time anyone was working on bringing it back in house they were playing with a 30x0.25x0.25 HP-5MSUI

What columns are you guys running? Any issues with co-elutions? The lab we sub to is using the restek RXI columns. Last time anyone was working on bringing it back in house they were playing with a 30x0.25x0.25 HP-5MSUI

We use Restek Rxi-5SilMS 20m x 0.18mm x 0.36um and it separates well. The one to watch is the Simazine/Atrazine pair as they run very close to each other. If the column becomes dirty they can co-elute. I prefer the 0.18mm due to the higher head pressure compressing the injection solvent better and getting on column in a tighter band. You loose a little linear range, but these are drinking waters and not expecting any large target hits anyway.