I have been working for a method development for perfluorocarboxylic acids (10 compounds) using LC/MS/MS (Agilent 6410). I started with MS parameter optimization by injecting individual compound (in methanol) directly into the mass spec and monitoring transition ions which I obtained from literature. My understanding is that, I would maximize the precursor ion in MS1 by changing fragmentor voltage (often called cone voltage), and then maximize the product ion in MS2 by changing collision energy.
However, for the compounds of interest, I am observing both precursor and product ions in the MS1 scan in comparable abundance. That also means, the product ion is maximized in MS2 for zero or very small collision energy. If I apply higher collision (> 3ev) energy, the product ions are decreasing.
I understand the collision and fragmentor voltage are instrument specific. The literature I am looking into has used a range 8-20 ev collision energy.
I would really appreciate any input in addressing the following concerns:
1. Would my method be valid if I observe both the ions in MS1 and report zero collision energy for product ion?
If not,
2. What can I do to remedy this situation?
3. I have seen a paper using pseudo MRM for similar compounds where they monitored same ion in MS1 and MS2. Can I use similar approach?
Thanks in advance.
Hello Hanna
First of all, what software version are you using on your Agilent MS? The newer versions come with a set of programs that basically do the optimization for you such as source parameters, fragmentor, collision energies etc. and also picks your most abundant fragment ions for setting up MRMs. I assume you first ran a scan experiment, that's when you observed both precursor and fragment ions? Basically you're right with your approach, you can maximise transmission of the precursor ions into the MS with the fragmentor, but when you see the fragment ions at relatively low fragmentor settings, that means your compounds may be fragmenting even earlier, i.e. in your source. That can be due to e.g. high drying gas temperatures, which is why optimising your source parameters matters.
Generally I would not be worried about detecting fragment ions in scan mode as long as you have a strong enough signal for the precursor ion. Once you set the instrument to MRM, all that matters is the precursor ions (everything else gets thrown out in quad one) which will then be fragmented in your quad 2 (collision energy here needs to be >0) and monitored in quad 3. So the manual approach (i.e. not using the helper programs) would be:
And don't bother running "pseudo MRM" on a QQQ when you can do the real thing ;-)
good luck and let us know how you went!
Ps have a look at the attached link for an agilent presentation on QQQ method development
https://www.chem.agilent.com/Library/posters/Public/QQQ_Method_Development_Triple_and_Optimization.pdf
Hi Hanna,
I've attached a couple of references for you. Firstly I will mention that this is a problematic analysis due to background concentrations of these compounds in the environment, in the laboratory, and in the instrumentation. US EPA has method for this EPA 537. This would be very helpful to read, if only to understand issues regarding this analysis. They spend years researching the development of these drinking water methods so it would be in your best interest to understand how they do it. It is amazing how much info you can get just reading an official method particularly a drinking water method.
I have not run this as my system has meters of Teflon tubing, and Teflon seals, gaskets, O-rings. However I can tell you that I run the same instrumentation (Waters Triple Quad) and that the collision energy they are using in Method 537 is not unusual. In other words, there is typical collision energy used for optimized fragmentation within the method.
Here is an excerpt from the method, as an example of information that you will find useful to you.
EPA 537 Excerpt:
PFOS has linear and branched isomers. There have been reports that not all the products ions in the linear PFOS are produced in all the branched PFOS isomers. (This phenomenon probably exists for PFHxS and PFBS also, although it has not been studied to date.) Thus, in an attempt to reduce PFOS bias, it is required that the m/z 499 ® m/z 80 transition be used as the quantitation transition. Some MS/MS instruments, such as conventional ion traps, may not be able to scan a product ion with such a wide mass difference from the precursor ion; therefore, they may not be used for this method if PFOS, PFBS, or PFHxS analysis is to be conducted. Literature reports indicate for the most abundant PFOS isomer, which is the linear isomer, that all the products ions obtained on an ion trap have less than 10% relative abundance. In addition, there is not a single ion trap MS/MS transition that encompasses the linear isomer and the majority of the branch isomers; thus, the bias would be unacceptably high.
http://www.epa.gov/oppt/pfoa/pubs/Analytical%20Methods-Reagen_finalrev1.pdf
Hello Christian,
Thanks a lot for your reply.
The software I am using can do the optimization for me. This is my first time working with LC/MS/MS method development and I want to get a better understanding by optimizing the system manually.
I really appreciate your recommendations. I will try them out and let you know how it goes.
Thanks again.
Hello Peter,
I am in fact using a draft EPA procedure. I have taken into consideration the background contamination and run some initial blanks. The instruments and tubing does not seem to have high PFC contamination. In future I'll definitely use QA/QC to to check for background contamination.
I'm aware about the issues you have mentioned for PFOS and other sunlfonic acids. Right now I am just working with perfluorocarboxilyc acid group.
Thank you for your reply.
There are also ASTM methods available. D7979-15 for water, and D7968-14 for soils. If you are getting fragmentation, make sure collision cell gas is off. Not sure how the Agilent works but on the Waters instrument there is a MS mode and a MS/MS mode. When optimizing on precursor you have to use MS mode, I always have to make sure I'm in the right mode (not mood), and that the collision cell gas is off. Good luck.
Hi Peter,
Looks like I am having the dissociation induced by the collision gas, similar to what you mentioned. I am not sure about the settings you mentioned. I will look into it.
Is there any documentation for what you are doing? I googled it, but could not find anything helpful.
Thank a lot for your help.
Check the Agilent web site for manuals. A quick look at the manual seems like that parameter is set up in the tune file - under context -> tune -> manual tune. Looks like that is where all your tune controls are. Maybe you set up a tune file first? You would be looking at doing either a full scan (narrow mass window) or sim scan to optimize on the precursor. Just a guess.... I use Waters Triple Quad, software is totally different.
Here is user manual for your product.
http://www.chem.agilent.com/Library/usermanuals/Public/g3335-90086_qqq_quickstart.pdf
I find that the software optimisation will take me to the beginning of optimisation for a compound that I am looking at. I usually find that I still have to do manual optimisation further. When you are scanning, you need not worry about a presence of product (daughter) ions. However, when you are optimising, then you have to adjust your cone voltage, N2 gas flows, probe temperature, etc. to get the max. precursor (parent) ion going in to the MS1. I have been told that new MS and MS software are made such that you really do not need an MS operator any longer, but it is the experience that you get from working on different compounds and instruments that will make you more conversant with MS's.