PAHs in Fatty Food Matrix

[Schmitty]

I need to determine PAHs in a fatty food. Typically, I would do an organic solvent shakeout of the matrix, then filter and analyze directly. For sensitivity I will be analyzing by GC/MS.

I tried 5mL DCM + 1g sample, followed by shaking then centrifugation and filtering. A significant portion of the oils from the matrix are present.

I also tried 5mL Methanol + 1g sample, then sonicated. I added ~3mL water to the sample and shook briefly (the oils precipitate/coagulate), then centrifugation. This provided for a much lower level of interference in the injection solution, but I do not want to inject this much junk onto my GC/MS.

I need to extract the PAHs while not also bringing the fats through the run. Or, what can I do after extraction to get rid of the oils?

One idea I had was extracting with BF3:Methanol (instead of just methanol). At least the fats would be GC-able instead of sticking to my injection port. yay/nay?

Thanks

[GOM]

Hi Schmitty,

Nice idea but BF3/methanol won't transesterify triglycerides - it will only methylate fatty acids. If you want to try that approach look at Christie's excellent site on this subject http://www.lipidlibrary.co.uk/GC_lipid/ ... /index.htm

The acid catalysed transesterification is simple and works nicely.

Best of luck

Ralph

[chromatographer1]

Ralph,

While I agree with you I also disagree only because BF3 in methanol WILL transesterify a triglyeride if a suitable solvent such as methylene chloride, toluene, THF, or ether is added so that the fat is dissolved. It is the low concentration of fat in the methanol that is the rate limiting step.

From the site you mentioned (a wonderful resource by the way), please note the specific comment "... in a reasonable time...."

"Non-polar lipids, such as cholesterol esters or triacylglycerols, are not soluble in reagents composed predominantly of methanol, and will not react in a reasonable time, unless a further solvent is added to effect solution."

Now to get the separation of fat from PAHs one might consider trying to base hydrolize the fat into the free acids and glycerol. Then a simple S P Extraction might concentrate the PAHs.

Surely there are EPA procedures in place for this analysis, but I don't have time to look them up.

best wishes,

Rod

[GOM]

Hi Rod,

Without wishing to go off topic, you are quite right. My statement was based on experience in which, as you say, methanol alone is a poor solvent.

Base hydrolysis is a nice idea - possibly refluxing with 0.5M KOH - it may even be possible to then precipitate the acids out as their calcium soaps.

Cheers,

Ralph

[chromatographer1]

Schmitty,

Back to the topic, Ralph has a good idea.

I would take your sample after DCM extraction and add TMAH tetraammoniumhydroxide in ether with some water to the solution and heat to hydrolyze the fats. Then I would extract again with aqueous calcium hydroxide. The glycerol and much of the FA salts should separate from the PAHs in the DCM. To facilitate the removal of the PAHs you might wish to add enough hexane to put the non-polar phase above the aqueous phase instead of below it.

You might need to dry the DCM/Hexane solution with magnesium sulfate or some other drying agent.

best wishes,

Rod

[Schmitty]

As of now, this is my method, as planned:

-5g sample into 20mL 1M KOH (in ethanol), sits overnight.
-Add 20mL H2O.
-LLE with 50:50 Hexane:Ethyl Ether (3x20mL).
-Backwash organic phase with H2O.
-Dry organic over Na2SO4.
-Take to near-dryness on rotory evaporator.
-Dilute with DCM/Methanol.

I will omit the acidification and re-extraction steps, as I believe the PAH I want is going to cross over to the organic in these first steps.

I have also set up a HPLC/FL for detection as well. So, I can inject this GC/MS or LC/FL. I have also been toying with the idea of suspending the sample in H2O or DMSO and doing Headspace GC/FID or Headspace Trap GC/MS. I'm sorry for all of the possible directions that this could go in, but the only methods we have found are for meats. An EPA method would be nice, but I think those might be limited to water/waste and soils. I have run PAHs before, but in toluene extracts of carbon black.

Thanks for the great link, and the suggestions!

[Schmitty]

Schmitty,

Back to the topic, Ralph has a good idea.

I would take your sample after DCM extraction and add TMAH tetraammoniumhydroxide in ether with some water to the solution and heat to hydrolyze the fats. Then I would extract again with aqueous calcium hydroxide. The glycerol and much of the FA salts should separate from the PAHs in the DCM. To facilitate the removal of the PAHs you might wish to add enough hexane to put the non-polar phase above the aqueous phase instead of below it.

You might need to dry the DCM/Hexane solution with magnesium sulfate or some other drying agent.

best wishes,

Rod

Thanks, Rod.

This is another workup I can try today/tomorrow.

edit: Rod: Is there another reagent that is more common than Tetramethylammonium hydroxide that will accomplish the same task?

edit2: I think I have TBAH on hand. Wouldn't the CaOH or KOH hydroylze the sample itself too?

[chromatographer1]

Schmitty

This looks like a good procedure to me.

I hope you do some std additions to verify your recovery.

wishing you great success,

Rod

[Steve Reimer]

Having done way too many fish, including lamprey, I'll throw in my 2 cents worth. Your plan looks workable. A base wash is usually effective in removing fats.
The EPA SW-846 3620 silica gel method works fairly well for fats and oils. PAHs elute with a weaker solvent than the triglycerides and free fatty acids so in our lab we load the column with the extract in hexane and then elute with 40% dcm in pentane. If done right the fats will stay on the column and the PAHs are recovered. If the column is overloaded or a more polar solvent is used the results aren't as good.
We have used 10g grams of silica gel from bulk (for 1 - 2 grams of fish oil)as well as commercially available cartridges.
One trick that can help for those really nasty samples: store the extracts in the freezer for a week or more, then remove the fat layer before it melts. We haven't had a problem with loss of analytes however the time factor precludes general use.

[Schmitty]

Thanks for the input, Steve.

I haven't been having much luck with all of the methods and method modifications that I've tried.

I am trying the freezer trick right now, with a methanol extract of the matrix.

It always seems like there is too much resideue left in my solvent when it comes to injecting the sample.

It appears that method 3620 is Fluorisil, not silica gel. Can you confirm which you use? I tried a fluorisil column, but the fats went right through with the hexane loading and hexane/DCM wash. Perhaps 3630/40 would work better, using the silica gel?

[chromatographer1]

Schmitty

The florisil in method 3620 must be very carefully activated for reproducible results. Even then there are batch to batch variations which cannot always be predicted.

One suggestion for a path in your research, If nothing seems to work for you, you might try the dioxin system sold by my company Supelco.

It is designed to remove fats from PAHs and PCBs and Dioxins. Fish, mother's milk, soil, and ground water have been processed by this system among other sundry matrices. It is used in some Japanese Industrial methodologies.

It is a flash chromatography column based on silica with 7 neutral, acidic, and basic silica layers to remove different interferences followed by a dual layer carboxen column to separate dioxins from PCBs. I believe the PAHs will elute ahead of the PCBs and Dioxins which are retained on the dual layer Carboxen column.

Techservice@sial.com would be glad to assist you.

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
rgeorge@sial.com

[Schmitty]

Hi Rodney,

Thanks to you and your company (I think my employer may be a good portion of your employer's business ).

The fluorisil we use is activated overnight at over 100°C I believe. It is cooled before dispensing into columns, followed by a small amount of Na2SO4. It is rinsed with 10-20mL hexane before the sample is loaded on. I just copied what our pesticides group does with their adsorption column prep.

I do have access to a GPC system, which would probably do the same job as the Dioxin System, but with either I would be looking at significant R&D to figure out the elution of my analytes vs the "junk." Currently I do not have the luxury of time.

[Steve Reimer]

Sorry for the wrong method number. ( 3630c is the current #)We use Florisil for pesticides and PCBs, silica gel for PAHs. Silica gel seems to work better for PAHs but that may be because of our limited experience using Florisil with PAHs.

[Schmitty]

Thanks for the follow-up Steve.

Well, I believe I have found the most inelegant sample work up.

2g sample + 5mL hexanes in 10mL centrifuge tube
shake
transfer to 10mL syringe with 0.45µ Teflon filter
Filter into new centrifuge tube
add 0.5 mL H2O
shake

Inject hexane onto DB-1.

My inlet + column were already contaminated by previous attempts, so I will perform standard addition on each sample...until the baseline goes to heck. I will report back if this method stops workin out for me. Currently I am seeing 2.9ng/g of my desired PAH in a ~25% fat/oil matrix. Linearity is 0.99+ in a renge of 2.9 to 29ng/g sample.

Thanks all!