Chromatography Forum

I am using this extraction method:

20g baked sand, 50ul of 200ppm spike, Sodium sulphate as drying agent
100ml SS
tumble 1hour
Concentrate using KD to 1ml
spike internal standard

analysis by GCMS

2,4 dimethylphenol consistently is recovered at less than twenty percent target while recovery of other phenols is good (around 60%). I don't have this problem with liquid-liquid extractions (similar method, 1000ml sample conc to 1ml).

I suspect it may be that my solid extracts get left out on the bench longer (overnight sometimes) than the LLE samples.

Any ideas?

I've tried using a brand new spike, clean system got a great reult for every compound except 2,4-dimethylphenol!

I have to ask a general question first. Why do you not add your internal standard much earlier, like in the first step? What are you using as internal standard?

I have to agree with CPG, the whole point of adding an Internal STD is to reduce variations caused by your extraction method.

Adding IS as a final step is more likely to cause you problems.

Correct me if I am wrong, but if you add it at the beginning of the procedure it is usually called a surrogate standard. I do agree that adding it early would be better.

As for what is happening to your 2,4-dimethylphenol, maybe the solid surface is catalyzing a reaction? Have you tried looking for unknown peaks that could be reaction products?

Correct me if I am wrong, but if you add it at the beginning of the procedure it is usually called a surrogate standard. I do agree that adding it early would be better.

As for what is happening to your 2,4-dimethylphenol, maybe the solid surface is catalyzing a reaction? Have you tried looking for unknown peaks that could be reaction products?

A surrogate standard is a totally different thing.

If you measure compound X concentration by calibrating with concentrations of compound Y, compound Y is a surrogate STD

However, in the current scenario IS is added to samples containing compound X and Standardised against STDs of compound X/IS. Calibration is peak ratio X/IS vs concentration X.

In every IS assay I have worked on the IS is added in the initial stage.

All,

tangaloomaflyer is doing classic EPA 8270 type analysis. "Surrogate" is added at the very begining along with LCS compounds when appropriate to follow efficiency of entire extraction process. Internal standards are added at the very end to give you a reference peak for calculations (and by default a snapshot of injection efficiency.)

My suggestion is to remove the sand matrix. If it is an LCS, then you don't really have to have a matrix per se, just spike the mix into a beaker with sulfate and go. I would also insert a KD blank (solvent with LCS and surrogate mix) that goes through process starting at the KD step.

Best regards,

My understanding is this:

Internal standard is used to quant (rather than using spiked compounds to quant) to allow for drift in responses. It is not to be used to compensate for extraction recoveries as different compounds may be affected differently by the extraction process (case in point: 2,4-dimethylphenol!)

To assess extraction recoveries surrogate compound spikes give an indication of recovery per sample

A matrix spike is similar but the compounds spiked in the matrix spike are identical to the target analytes. Surrogate spikes use compounds that aresimilar but not identical to target analytes (e.g. deuterated analogues) so they can be added to every sample.

The LCS gives an indication of recovery from an 'ideal' matrix. The point of using baked sand is to show that compounds can be extracted from a dry contamination-free solid matrix.

I would also insert a KD blank (solvent with LCS and surrogate mix) that goes through process starting at the KD step.

Best regards,

I'll give this a go though - good idea

If you haven't done it already; make up a new solution of dimethyl phenol.

Peter

If you haven't done it already; make up a new solution of dimethyl phenol.

Peter

It's part of a mix of phenols. i made up a fresh dilution and it did'nt make any difference.

tangaloomaflyer,

I would argue, and have seen it argued many times long ago when I worked in environmental labs, that LCS/LCSD can never mimic a real matrix and that is why MS/MSD is run as well. Thus, I have seen it argued, and won, that the lab would simply spike sulfate and extract. Really neither a here nor there argument for me at this point in my life.....

Simply stated, your objective is to determine if your blank sand, your sulfate or your extraction is causing the 24 DMP to crater. Thus, try two with sulfate and no sand, two with sand and no sulfate and two without either one and compare the results. My Snicker's bar is on the sand. I am assuming no cleanup, right?

Good luck.

I have seen erratic chromatographic behavior from 2,4-dimethylphenol when the GCMS is super clean - it's almost as if it prefers a slightly dirty inlet and gently used column. Of course, your 2,4-dinitrophenol will decrease as it gets dirty, so it ends up being a balancing act.

If eliminating the sand does not work I would suggest running the LCS after a matrix spike and see if the results improve.

tangaloomaflyer,

Simply stated, your objective is to determine if your blank sand, your sulfate or your extraction is causing the 24 DMP to crater. Thus, try two with sulfate and no sand, two with sand and no sulfate and two without either one and compare the results. My Snicker's bar is on the sand. I am assuming no cleanup, right?

Good luck.

I'm trying this right now, I also have my snickers on the sand, but we do bake it out, however not sure how long it sits on the bench...

OK,

ran LCS without sand and without NaSO4 both came back same as control.

Next up, going to try glassware washed by different methods.

results to follow....

tangaloomaflyer,

thanks for the response to my post regarding aniline, 4-chloroaniline. we seem to be having similar problems, but with different compounds.

the idea of a super-clean gc-ms causing problems for 2,4-dmp is interesting. i've never heard of anything like that, but i would think it would be easy to test by simply running a standard with no extraction involved.

in a similar way i have ruled out (i think) the gc inlet or column as the source of my problems with the anilines. in my experience 4-nitroaniline is more of a problem on the gc side of things than aniline or 4-chloroaniline, just as 2,4-dinitrophenol, 4-nitrophenol, and pentachlorophenol are more difficult than 2,4-dimethylphenol.

so it sure seems like our analytes are being lost in the extraction process. i've tried skipping the sodium sulfate altogether, and i've tried concentrating at different water bath temperatures. to no avail.

like you, i don't have a problem when concentrating in straight dcm (water samples). so acetone seems to be a key factor. i am curious to hear the result of your experiment with the glassware, so keep posting and good luck!