GC MS:Ion fragmentation pattern changed

[Sonia Oliva]

Hi all, I have an issue with GC MS, anabolic quantification: keeping standard conditions as usual, ion fragmentation pattern has changed!!!!! I mean, 4 diagnostic ions plus deutered-ion selected from internal standard appear in mass spectrum, but relative abundances have changed without aparently reason. Autotune is ok, other technics for steroids has been tried in the same instrument also. Do you have any suggestion for me? I have used HFBA as derivative, but it never showed problem before.Thanks a lot-

[Don_Hilton]

A few thoughts:

Is the peak saturating the detector? If so the ion ratios are probably not changing - the instrument is just not able to count them all. Look at the ion traces and see if the stronger ions flatten out at the top of the peak.

Is this a standard or an analytical sample? And if it is an analytical sample, how does a standard look?

There are a few things that can change the fragmentation in the ion source, but unless the changes are fairly significant, changes will not be particularly noticible. From the autotune being OK, I assume no changes in the fragmentation of PFTBA. YOu would expect changes in the filament to affect the tune. Without more informatin, I would not look at the mass spectrometer as the first suspect.

My first place to look would be at the samples and chromatography. Ion ratios will change when you have coelutions with shared ions. Look at peak shapes to see if they are changing. Also, look at the ions of interest and see if they all rise and fall together through the peak. If the shapes for the ion traces differ, you have a clue that you have coeluting compunds with shared ions. And if this is the case, you may be able to use deconvolution software (like AMDIS, which is free) to get a better idea of what is going on.

[Sonia Oliva]

Hi, thanks for your opinion. Let me add more information about this issue. Peak is not saturating detector. This effect appeared in both standard samples for quantification curve and post clean up urine samples. Your assumption was right, ion fragmentation parameters wiht PFTBA were successful. About posible coelution shared ions, it would apply if standard samples only were affected , but in this case all samples injected shows the same mass spectrum profile. This is very curiuos cause it happened twice, in a random procedure. what do you think?

[Peter Apps]

all samples injected shows the same mass spectrum profile. This is very curiuos cause it happened twice, in a random procedure. what do you think?

so is it all samples or only two of them that were affected ? Also it is rather confusing when you refer to standards as samples.

I'm with Don - this is very likely to be a co-eluting contaminant, and its occurrence in both standards and samples does not rule that out. Do you have large enough peaks to get a full scan mass spectrum ? What do you see if you run a blank of everything except the sustnace that gives the problem peak ?

Peter

[JI2002]

ion fragmentation parameters wiht PFTBA were successful.

What do you mean by this? In order to have consistent ion patterns for target compounds, the ion ratios of PFTBA tune need to be kept constant, i.e. the ratios of m/z 131, 219, 414, 502 over m/z 69 should be about the same. Was the tune file used in the run method changed?

[Sonia Oliva]

Sorry If I wasn´t clear enough (English limitation). When I was talking about samples, I meant both urine samples and standars showed same effect, relative abundance changed. I have run blank, no peaks appeared, only internal standar ion as it was expected. Also, other substances were detected without problems using different clean up technics in the same instrument and this effect didn´t appear. I didn´t mention before, I´m using SIM method, and only specific extracted ions are being monitored.
ALso when I mentioned ion fragmentation pattern normal, I meant Autotune using PFTBA was ok. But looking again 219with paramenter has changed from -0.008 to 0.018. Thank you all for your thoughts. SO-

[JI2002]

What are ions that has changed? Did the ion pattern of internal standard change?

[Sonia Oliva]

Ions changed are 666,453,318,306 for nortestost and 464,678,343,369 for trembolone. Deutered ion monitored is 669 and it changed as well, It is used as internal standard ion only.

[Don_Hilton]

Are we talking about derivatized nortestosterone and perhaps trenboloe? If so these two steroids have masses of only about 270. Each has ketone and a hydroxyl group that can be derivatized - but we would be having to add quite a bit of mass to get something up around 670.

I have to ask which derivatives are you making - let's see if there is something that makes sense chemically.

[Sonia Oliva]

yes, it was derivatized with HFBA.

[Don_Hilton]

I would expect nortestosterone (a monohydroy compund) to react with one mole of HFBA. Nortestosterone has a formula of C18H26O2. If one makes the heptafluorobutyrate ester, one replaces the OH Hydrogen with C4F7O Giving a molecue with the formula C22H25F7O2 - with a mass of 454. If you could get the enol of the nortestosterone to form and esterify, you would have something with a mass of 652. I don't see a way to get 666 out of the molecule. And trembolone is a ligher molecue, so 678 does not makes sense.

So the next question, is are you sure you are looking at the right peaks and ions? Or, what have I missed?

[Sonia Oliva]

Boldenone instead of Trembolone is the correct,my mistake when I answered before. However, for Nortest chosen ion are correct ones for this method. Unfortunately I am doing this method as routine, it was validated from a paper time ago and where this fragmentation pattern was best choice. Would you have and email where I can send scanned chromatograms as example?

[Don_Hilton]

Do you have a copy of the paper from which the method came?

e-mail: dchiton*mylink!net -- replace * with @ and ! with .

It would be good, however to post information on the forum for two reasons:

1) there may be someone with more relevant experience to jump in and
2) There are times I may be tied up with things for a few days and someone else may answer more quickly.

But either way, I'll be glad to look and see if I can offer any useful ideas for you.

[Sonia Oliva]

Hi, thanks. I was out of lab. Searching source papers for this method I found"
Validation of multi-residue methods for the detection of anabolic steroids by GC-MS in muscle tissues and urine samples from cattle†
E. Daeseleire, R. Vandeputte and C. Van Peteghem
Analyst, 1998, 123, 2595-2598". however, I was told this method was tuned and adapted to our CG-MS (HP5890-5972). I will try to post some chromatograms to this forum.

[Don_Hilton]

If you can e-mail me a copy of the paper I will look at it. I look at things from this forum at home and do not have access to this journal from home.