Problem with base peak seperation

[kouroshh1]

Hi Kenn,

I think your comment is missing in your reply message.

Kourosh

[Kenn]

Hi Mattmullaney and Kenn,

Actually, I tried infusing Indole 3 carbinol (I3C) into the MS directly by using syringe and I did not have a problem. The intensity was high and no condensation product could have been seen.

The problem is when the I3C reference standard is running through the LC the condensation will occur. Unfortunately, when the condensation product from I3C is reaching into the ion source it will be fragmented into I3C mass transitions before reaching into the first quadruple. So, in my multiple reaction monitoring (MRM) method as a precursor ion (Q1) I have 130 m/z and as a product ion Q3 I have 103.1 m/z in a positive polarity. Due to the in-source fragmentation, I only get the transition for I3C that is 130 m/z----> 103 m/z. So it is very hard to identify the mass of the second peak because it has a same mass transition as I3C unless as you said we purchase the standard which have not done it yet and compare it against the second peak.

I guess it depends on the objective of your study and what type of study it is. Although what you say here is almost assuredly true my QA would have a field day with what you are saying here. They would want proof in the data and that's a research project. If the larger and "real" I3C peak were quantitative I would use it for quant and if asked would say the second one is believed to be an artifact of source fragmentation. Again, it depends on what type of study you are doing.

Unfortunately, the ratio of those 2 peaks does not remains constant over the course of your sequence.

Does your high standard return the same area cts after say 20 injections? Is there a transformation actively taking place causing your peak of interest to change over time? If what you say above is true it should not. If your r value is 0.999 your QC's are all 100% it should not matter if there is another artifact in samples that generates a peak with your mass transition at another retention time. To tell the truth I would probably not bother to mention it but the report will include that c-gram and if a client or agency says what's that other peak out at x min in your I3C chromatogram, I would say I presume it's an artifact stemming from source fragmentation and it was not used for quantitating your analyte.

Regarding the second part, in my MRM method if I have several compounds and some of them are co-eluting together, it is acceptable to count the area peak of each analyte and do the quantification. Am I understood correctly? I can select each peak based on its MRM transition. So, would it be acceptable to calculate the area peak of each peak independently because it has its own MRM transition?

This I do virtually all the time. Are you running analyst? Set it up in quant method to quantitate each mass transition individually, there is no need for separation. Sure if easy I may separate with the gradient but try doing it with 2 isomers with the same mass transition. You can't always get separation and you don't need it. The number of documented methods for particular classes of compound that have a forest of peaks between say 6 and 8 minutes is endless but the methods still work for each one of them once your mass transition is extracted out.

Thank you for your time and help

No problem I hope I have helped. I just wish I knew what sort of study or guideline you are trying to satisfy because without that half of what I have said may not apply.

[Kenn]

Hi Kenn,

I think your comment is missing in your reply message.

Kourosh

Sorry about that, I walked away and came back later, not sure what hap.

[kouroshh1]

Hi Kenn,

Thank you so much for spending time to explain to me in a very clear way. I will try to answer your question below.


* I guess it depends on the objective of your study and what type of study it is.

I am planning to extract I3C from the plant shoot. In general, when the brassica plants are stressed with for instance pathogen ,they expresss the enzyme which break down glucobrassicin (plant secondary metabolite) into I3C. By quantifying I3C, I will ensure that plant are defending themselve against the pathogen. So, quantification of this conpound is important for me.

* Does your high standard return the same area cts after say 20 injections?

Actually, I have tried only 3 injections and the intensity of I3C across 3 injections was not that much different. But i will try it with more injections to enure the result is consistent.

In addition, I haven't made I3C calibration curve yet. I will make it soon and as you sai. If the R square is very high then I will not care as you said about the artifact from source fragmentation.

* Are you running analyst?

Yes, we have Linear iontrap triple quadruple LC-MS/MS 3200 from Sciex in our lab and I am working with an analyst software.

* Set it up in quant method to quantitate each mass transition individually, there is no need for separation.

I didn’t know, it would be acceptable for a publication. So, I will not care that much in the future to get a super perfect resolution of the peaks.

* Sure if easy I may separate with the gradient but try doing it with 2 isomers with the same mass transition. You can't always get separation and you don't need it. The number of documented methods for particular classes of compound that have a forest of peaks between say 6 and 8 minutes is endless but the methods still work for each one of them once your mass transition is extracted out.

Very very good point.

And may I finally ask how can I improve the sensitivity of I3C while i am using basic mobile phase such as methanol with 10 mM ammonium formate at pH of 8.5 with NH4OH? In basic mobile phase I3C will stay mainly as a single peak but the sensitivity is very poor. So, if I can improve the sensitivity my problem will be solved.

Kourosh

[mattmullaney]

Hi Kourosh, Kenn,

Kenn, I'm grateful for your sharing of your experience!

Kourosh, if you have an auxiliary LC pump and a tee connector, you can add acid post-column to the effluent from the LC before the Ms...that should boost the intensity of the signal out.

Apologies if I've asked if you have a spare pump already and the answer was "no," Kourosh.

[Kenn]

Hi Kenn,

Thank you so much for spending time to explain to me in a very clear way. I will try to answer your question below.


* I guess it depends on the objective of your study and what type of study it is.

I am planning to extract I3C from the plant shoot. In general, when the brassica plants are stressed with for instance pathogen ,they expresss the enzyme which break down glucobrassicin (plant secondary metabolite) into I3C. By quantifying I3C, I will ensure that plant are defending themselve against the pathogen. So, quantification of this conpound is important for me.

It sounds like you have a method that will accurately quantify your compound now. As long as across the board your source contamination happens with your stds your 100 ppb std is going to represent 100 ppb. If your study sample returns the same area cts how much is in your sample? Your source contamination may be greater with higher level samples because that condensation will carry more but it will basically all come out in the wash so to speak. Once again, if your curve is acceptable and you can fortify a blank sample at 100 ppb and your curve tells you it has 100 ppb in it, it works and an artifact that elutes a couple minutes later is meaningless.

* Does your high standard return the same area cts after say 20 injections?

Actually, I have tried only 3 injections and the intensity of I3C across 3 injections was not that much different. But i will try it with more injections to enure the result is consistent.

In addition, I haven't made I3C calibration curve yet. I will make it soon and as you sai. If the R square is very high then I will not care as you said about the artifact from source fragmentation.

That would be the first thing I would do. If you don't have acceptable linearity you haven't got a method. If you have matrix to fortify and can throw one of them in the sequence you may kill 2 birds with one stone. If the curve looks good what's the first thing you will do anyway.

* Are you running analyst?

Yes, we have Linear iontrap triple quadruple LC-MS/MS 3200 from Sciex in our lab and I am working with an analyst software.

* Set it up in quant method to quantitate each mass transition individually, there is no need for separation.

I didn’t know, it would be acceptable for a publication. So, I will not care that much in the future to get a super perfect resolution of the peaks.

Everything I do is proprietary so I don't know what is considered "acceptable" for publication but if your method works, it works.

And may I finally ask how can I improve the sensitivity of I3C while i am using basic mobile phase such as methanol with 10 mM ammonium formate at pH of 8.5 with NH4OH? In basic mobile phase I3C will stay mainly as a single peak but the sensitivity is very poor. So, if I can improve the sensitivity my problem will be solved.

The first thing I would try is something I have done on countless occasions is reduce or eliminate the mobile phase modifiers. Can you chromatograph your analyte without them? Formate typically robs you of signal unless it is absolutely essential. Others here may disagree but I have been handed methods on countless occasions that include a buffered mobile phase and the first thing I do is attempt to make the method without them. I would say in the majority of cases I do. I have about 6 sequences to run today, all with different analytes and one of those analyses has a mobile phase modifier and it is HFBA, you won't chromatograph that compound without it.

The next thing I would do is look at your products. I believe you said your parent mass was only 130 Da so you may not have many and some will be too noisy to use but have you ever summed multiple mass transitions in Analyst? When I initially scan a compound I typically tell it to optimize on three products and none below 80 Da. In your case you may have to go lower due to your parent mass. Now when you chromatograph your compound do you see those other products you are using for confirmation or do you only have one? If you have more than one you can check that little box in the upper right corner of the box when you are building your quant method and tell it to sum the mass transitions and then select which ones. I use 2 criteria when doing this. 1st, you must have a decent signal in your low std for the smaller product and 2, if one of them has an extremely noisy baseline it usually brings with it that noise so it better be significant or I won't use it. I modified a method the other day for more sensitivity with an antibiotic and practically doubled my sensitivity. Curves were always acceptable and now every one of them is 3 9's. (That's what I call 0.999).

[kouroshh1]

Dear Kenn,

I am so sorry that I forgot to appreciate you for your very useful comments. I will up to date you as soon as I do a new experiment regarding this topic.

Kourosh

[mattmullaney]

@ Kourosh and Kenn,

My thanks again to you both--and hey, I'm interested in the outcome, too!!