HPLC determination of trace acrylamide in food

[qazwsx52]

Our lab is interested in getting a HPLC for trace acrylamide detection in starchy food (bread, potato chips, etc). But we have very tight budget to get a HPLC-MS or HPLC-MS/MS. So we are wondering if there is any other lower-cost detector equivalent to MS in terms of sensitivity or other methods (type of columns, sample clean-up, derivatization, etc) exist. The dynamic linear range of acrylamide we are looking for is 10-1000 ppb.

So far Waters contact me about their latest QDa Mass Detector with a affordable price. But I have no idea how this new instrument will work. Anyone recommend it? What is the disadvantage or any improvement required? What is the LOD, LOQ and linear range of QDa? Is QDa easier to operate and maintain compared to conventional MS? I used to work with Agilent 1200, never work with Waters before, so I am kind of worried about different operation system.

[MSCHemist]

My predecessor did it via GC/MS. I'm not sure what protocol she used but If I were developing the method I'd use ethyl or propyl chloroformate and ethanol/n-propanol. Those derivatize amide and amines very readily as they work well on amino acids, are cheap, work in aqueous solutions, and give high sensitivity. You can measure asparagine (the precursur) as well with it. The reaction takes ~20 minutes mostly of pipetting reagents and vortexing then extracting with chloroform.

[James_Ball]

We have been doing a study for Acrylamide in waste waters here using HPLC/UV. Direct injection of the waste water gives a detection limit of 10ppb ( the lowest calibration standard). If you are concentrating the extract from the food you may be able to go lower than that.

[MSCHemist]

I just got a request.

HPLC the sample prep is pretty simple but of course and HPLC is usually more expensive than GC.

I've reviewed the GC methods. My initial chloroformate impulse seems like it won't work. With chlroformate derivatization amides are usually untouched or converted to nitrile and 2-propene nitrile would be too polar to extract into solvent.

The GC/MS methods are either sample cleanup usually with SPE and injection of underivatized acrlamide onto a polar wax column or cyanopropyl column.

Bromination. Bromination followed by debromination to a monobrominated derivative that is more stable and also allows it do be done via ECD detector.

Finally cleanup, silylation, and in most cases SPME injection.

The protocol I like the best involves extraction with water [defatting with hexane if necessary], application to a diatomaceous earth SPE column, elution with ethyl acetate, then blow down to ~ 0.5-1ml and injection. This is similar to the HFBI derivatized 3-mcpd protocol. The cleanup removes the carbohydrates and asparagine and other precursors that can cuase acrylamide formation in the hot GC. Quote LOD is 3ppb you can probably go lower with the work on large volume splitless injection or of course with a GC/MS/MS.

http://www.teknoscienze.com/Articles/Ag ... d-gas.aspx

The one my predacessor used was easier but I worry does not cleanup the aforementioned precursors.

Journal of Food Composition and Analysis 20 (2007) 232–235
dehydrate, defat with hexanes, and dilute with isopropanol.

[MSCHemist]

I did some work on acrylamide Friday. Using the diamotaceous earth column as mentioned in the article and EI ionization on the GC/MS It looks like My LOD is at best 50-100ppb.

If the OP is looking for the cheapest way to do acrlamide in food I'd recommend brominating and using a GC-ECD. Most methods brominate with HBr and bromate then treat with triethylamine to eliminate the primary bromine creating 2-bromo propylamide. The dibromo deriviative is less stable and can lose a bromine during GC.

[James_Ball]

I just got a request.

HPLC the sample prep is pretty simple but of course and HPLC is usually more expensive than GC.

If you need to purchase instruments then it is about equal, especially if you are looking at GC/MS. Mobile phase for HPLC will cost more than Helium for a GC or GC/MS but you probably make up the difference in cost when you consider the derivatizing agents and time involved in prep for the GC methods.

I used to prefer GC or GC/MS for analysis until I began working with HPLC also. When you can eliminate derivatization or even completely eliminate extraction in favor of direct injection of a sample HPLC will always give better recoveries. When you still have to do the complex sample prep for HPLC then it comes down to which will give the best sensitivity due to larger volumes being able to be injected or if analytes suffer from thermal breakdown.

[MSCHemist]

True but A GC-ECD can run on N2 or H2 or acrylamide could also be done with an NPD not sure how the sensitivity compares to EI MSD SIM though.

DId some more work today. My ethyl acetate seems dirty as heck and is creating too much noise. I tried a standard in acetone and I could see it fairly clearly at 10ppb so my LOD is ~25-30ppb with a 5ul injection pulsed splitless at 150 deg C (low inlet temp seems to help). I need ethyl acetate to elute from diatomaceous earth columns (need a polar solvent immiscable with water) but hopefully if I evap to dryness and replace with acetone most of the garbage will go with it.

I am beginning to think though bromination then elimination with triethylamine to 2-bromopopenamide is the way to go. Bigger m/z less interference and cleanup and ECD is a good option.

[MSCHemist]

For anyone interested I found a great article on derivatizing acrylamide with xanthydrol and doing GC/MS which is not water sensitive. The sample is extracted with water, scavenged/defatted with hexanes, centrifuged and the aqueous layer is applied to a AC2 SPE cartridge (The paper also uses a C18 but with the hexane wash I find that redundant) and eluted with methanol derivatized, and extracted with ethyl acetate.

Alternately you can apply the aqueous extract to a diatomaceous earth column like Agilent Chemelut and wash the column with hexanes and elute with 25ml of ethyl acetate and blow it down and derivatize in methanol. The AC-2 is aparently much better as it hold the acrylamide so tightly you can apply 50ml of sample and then elute it with a few ml of methanol bypassing the need to blow it down so much.
http://www.ncbi.nlm.nih.gov/pubmed/22257340

I'll be testing the method next week. The xanthydrol is much less expensive than the bromination reagents. ~$50 for 50g which is enough for 1000 reactions.

[MSCHemist]

Hi all I have been working on a GC/MS method for acrylamide for a few months now on and off and if anyone else is working on it and has some suggestions please comment.

I started out trying to do underivatized acrylamide. I finally gave up on that because m/z 71 and 55 are too nonspecific resulting in too much noise on the chromatogram

I then found a nice article on derivatizing with xanthydrol. It is a nice simple derivatization in acidic water or methanol followed by extraction into ethyl acetate. I was very excited with the standards which gave nice clean peaks. When I tried it on some samples it fell apart. It seems everything that coextracts kills the reaction (I get as little as 10% yeild). I've tried several different liquid liquid extractions as well as diatomaceous earth SPE's. Nothing has worked. The original paper uses a piggy backed C18 and activated carbon SPE system and that makes the procedure too expensive for me as activated carbon cartidges alone are $10 each.

I am now working on bromination. My basic procedure is 3 grams of sample + 18g water, Defat with a hexane extraction, then centrifuge and collect 10ml of aqueous supernatant. Add H2SO4 KBrO3, and KBr and derivatize overnight at 4 deg in the fridge. Then I neutralize the bromine with 1ml 0.1N sodium thiosulfate and salt it into 1ml of ethyacetate, add 10% triethylamine, and inject 2ul splitless. Not quite sensitive enough my LOD is arround 20 ppb. 1701 column works well and detection is with sum of 105, 107, 149 and 151 m/z.

It looks like I will have to extract it out with more ethyl acetate and blow it down to a smaller volume say 100-200ul.


Any suggestions?