My work is wanting me to make a method to look for a very specific nonpolar analyte at a PPB level. I've never made a method before and I am concerned that I won't have time to do everything I need to do by the deadline. I have looked through the internet to see if I can find a method detecting this analyte and I can't find it anywhere. Does anyone have any advice on how to move forward?
We use a Thermoscientific triplus 500 headspace, a Trace 1600 GC with a nonpolar column, and we have a triple quadripole TSQ 9610 Mass Spec.
Are you able to let us know what the compound and matrix is please?
From what I gather, you're attempting targeted analysis of a single compound which may be present down to PPB level and you aren't interested in anything else that might be in your sample. Based on this information and given that you have a TQ instrument, I would consider SIM or MRM based methods which would allow you to cut some corners when it comes to method development saving you some time.
Because you're not creating a method that needs to resolve 20 or 30 key analytes of interest, you can focus on creating a very simple method that allows you to get your analyte onto the column and to the MSD nice and quickly, reliably, with a decent peak shape and with the minimum of peak area variability without having to compromise conditions to suit another 20 or so other analytes. Creating your chromatographic method is a matter of trial and error, but I suspect you could probably find a basic method for something that is reasonably similar to your analyte to help you to make a solid start and from there adapt conditions such that your analyte is well resolved, (ideally) doesn't co-elute with anything else in there and gives good precision.
To get PPB levels, you need to be using SIM or MRM. Personally I would begin method development with a simple scan run to determine the MS fragments produced by your analyte, or alternatively I would look them up on NIST. From either or both of these sources I would then select 3 ions based upon their specificity and abundance to create my SIM run parameters, using the most abundant fragment as my quant/target ion and the other 2 as my qualifier ions. For something like dichlorobenzene (
https://webbook.nist.gov/cgi/cbook.cgi?ID=C106467&Units=SI&Mask=200#Mass-Spec), I'd probably use 146 as my quant/target ion with 2 of 148, 111 or 75 as my qualifier ions.
With 3 well-selected ions coupled with a RT, you can be near certain that your analyte (if detected) is there and it is your actual analyte rather than something else. At this stage I would probably consider running a series of dilutions to determine approximate LOD of this method to see whether or not this gets you where you need to be or not.
It may well be the case that you could improve sensitivity (not always) by using MRM, particularly if your matrix is complex or difficult. Unlike SIM, MRM provides absolutely flat zero-noise baselines which improves SNR massively and impacts very favourably on LOD and LOQ in many instances. As a thermo user you should have a piece of software called AutoSRM which will help you to navigate through the process of MRM method development. It can be done manually, but it's a lot slower as it involves determining the daughter ions produced from each of your SIM ions (setting MS1 in SIM mode and MS2 in Scan mode) and then working out the optimum collision energy for each transition - this takes a while. At the end of the process you should have a number of transitions that are more or less 'unique' to your analyte and that you can select based upon preference and selectivity. For something like Dichlorobenzene, I suspect that you'll end up with 146>75, 146>111, 111>75 etc. from which you can select the most appropriate and abundant transitions.
Hope that helps a little.
R
ChemBuddy | 
Chem Reddit | 
Topic: HS GC-MS method xreation (Read 3581 times)